Fisiologia de Katz

Authors: Katz, Arnold M.
Title: Physiology of the Heart, 5th Edition

Copyright ©2011 Lippincott Williams & Wilkins
2011
Lippincott Williams & Wilkins

Philadelphia
Two Commerce Square, 2001 Market Street, Philadelphia, PA 19103 USA, LWW.com
978-1-60831-171-2
1-60831-171-6
© 2011 by LIPPINCOTT WILLIAMS & WILKINS, a WOLTERS KLUWER business
Two Commerce Square, 2001 Market Street, Philadelphia, PA 19103 USA, LWW.com
All rights reserved. This book is protected by copyright. No part of this book may be reproduced in any form or by
any means, including photocopying, or utilized by any information storage and retrieval system without written
permission from the copyright owner, except for brief quotations embodied in critical articles and reviews.
Materials appearing in this book prepared by individuals as part of their official duties as U.S. government
employees are not covered by the above-mentioned copyright.
Printed in China
Acquisitions Editor: Frances DeStefano
Product Manager: Leanne McMillan
Production Manager: Alicia Jackson
Senior Manufacturing Manager: Benjamin Rivera
Marketing Manager: Kimberly Schonberger
Design Coordinator: Stephen Druding
Production Service: Aptara, Inc.
Library of Congress Cataloging-in-Publication Data
Katz, Arnold M.
Physiology of the heart / Arnold M. Katz. — 5th ed.
p. ; cm.
Includes bibliographical references and index.
Summary: “Dr. Arnold Katz's internationally acclaimed classic is now in its thoroughly revised Fifth Edition,
incorporating the latest molecular biology research and extensively exploring the clinical applications of these
findings. In the single authored, expert voice that is this book's unique strength, Dr. Katz provides a
comprehensive overview of the physiological and biophysical basis of cardiac function, beginning with structure
and proceeding to biochemistry, biophysics, and pathophysiology in arrhythmias, ischemia, and heart failure.
Emphasis is on the interrelationships of basic processes among the cell, cardiac muscle function, and the
biophysics of contractile and electrical behavior. This edition includes new material on cell signaling and
molecular biology”—Provided by publisher.
ISBN-13: 978-1-60831-171-2 (hardback : alk. paper)
ISBN-10: 1-60831-171-6
1. Heart—Physiology. 2. Heart—Pathophysiology. I. Title.
[DNLM: 1. Heart—physiology. 2. Heart Diseases—physiopathology. WG 202 K19p 2011]
QP111.4.K38 2011
612.1′7—dc22
2010024639
Care has been taken to confirm the accuracy of the information presented and to describe generally accepted
practices. However, the authors, editors, and publisher are not responsible for errors or omissions or for any
consequences from application of the information in this book and make no warranty, expressed or implied, with
respect to the currency, completeness, or accuracy of the contents of the publication. Application of the
information in a particular situation remains the professional responsibility of the practitioner.
The author, editors, and publisher have exerted every effort to ensure that drug selection and dosage set forth in
this text are in accordance with current recommendations and practice at the time of publication. However, in
view of ongoing research, changes in government regulations, and the constant flow of information relating to
drug therapy and drug reactions, the reader is urged to check the package insert for each drug for any change in
indications and dosage and for added warnings and precautions. This is particularly important when the
recommended agent is a new or infrequently employed drug.
Some drugs and medical devices presented in the publication have Food and Drug Administration (FDA) clearance
for limited use in restricted research settings. It is the responsibility of the health care providers to ascertain the
FDA status of each drug or device planned for use in their clinical practice.
To purchase additional copies of this book, call our customer service department at (800) 638-3030 or fax orders
to (301) 223-2320. International customers should call (301) 223-2300.
Visit Lippincott Williams & Wilkins on the Internet: at LWW.com. Lippincott Williams & Wilkins customer service
representatives are available from 8:30 am to 6 pm, EST.
10 9 8 7 6 5 4 3 2 1
> Front of Book > Authors
Author
Arnold M. Katz MD, D.Med (Hon), FACP, FACC
Professor of Medicine Emeritus
University of Connecticut School of Medicine, Farmington, Connecticut, Visiting Professor
of Medicine and Physiology, Dartmouth Medical School, Lebanon, New Hampshire, Visiting
Professor of Medicine, Harvard Medical School, Boston, Massachusetts
> Front of Book > Foreword
Foreword
The cardiovascular pandemic is now advancing at an alarming pace in many parts of the
world. Epidemiologists inform us that by 2020, cardiovascular disease will be responsible
for 25 million deaths annually, 36% of all deaths, and for the first time in the history of
our species, it will be the most common cause of death. Thus, cardiovascular disease may
now be considered to be humankind's most serious health threat. The cardiovascular
epidemic is advancing steadily in developing nations in Asia, Africa and South America
with enormous and rapidly growing populations.
On a more positive note, age-adjusted cardiovascular mortality and morbidity have been
declining steadily for more than two decades in North America and Western Europe. These
improvements, translating into the extension of useful life for millions of persons, result
from advances in cardiovascular science leading to both the prevention and improved
treatment of patients with cardiovascular diseases. A few examples of the latter include
the impairment of conduction and of automaticity of specialized cardiac tissue, which
leads to heart block, and other serious bradyarrhythmias can be readily corrected with
implantation of a cardiac pacemaker; fatal ventricular fibrillation can be averted with an
implanted cardioverter defibrillator; asynchronous ventricular contraction in heart failure
can be corrected by biventricular pacing; hypertension secondary to increased activity of
the renin-angiotensin-aldosterone axis and of the adrenergic nervous system can be
relieved by pharmacologic blockers; and the imbalance between myocardial oxygen
supply and demand that can lead to debilitating angina pectoris or fatal myocardial
infarction can be relieved by increasing oxygen supply and/or reducing demand. The age-
adjusted incidence of most forms of ischemic heart disease has been declining steadily
with increasing attention to lifestyle—especially the reduction of smoking—and to the
widespread use of statins.
These landmark improvements in cardiac care have resulted directly from the advances in
cardiovascular physiology and pathophysiology that occurred during the first half of the
twentieth century, an era when physiology was devoted largely to the study of the
function of the intact heart. It then became clear that further understanding of
cardiovascular function required a focus on progressively smaller components of the
organ. Accordingly, there has been a steady march from the examination of the whole
heart to strips of cardiac muscle, to individual myocytes, to organelles within the
myocyte, to the proteins of which these organelles are composed, and to the genes that
encode these proteins. In other words, a reductionist approach has been dominant in
cardiovascular (and other biomedical) sciences for more than 50 years. An important next
step will be to obtain a clearer understanding of how the individual components affect
the function of the whole heart in the intact human.
This magnificent fifth edition of the classic text, Katz's Physiology of the Heart, a book
that improves with every edition, considers the normal and diseased heart at all of these
levels. After an incisive exposition of cellular, subcellular, molecular, and genetic
processes in the first half of the book, it then goes on to explain how these processes
affect the function of the entire organ, both in health and disease.
What is, of course, so remarkable is that Physiology of the Heart remains a single-
authored comprehensive text, probably among the last of its kind. It is a tour de force
that reflects Dr. Katz's rich experience as a creative scientist, a gifted educator, and an
experienced clinician. It flows smoothly without the repetition, inconsistencies, gaps, and
abrupt changes in style that are characteristic of so many multiauthored texts. The
explanatory diagrams are superb. Katz has the rare gift of explaining complex concepts so
that they can be readily understood by students and physicians without advanced training
in cardiovascular science. This book will also be especially useful to fundamental
cardiovascular investigators who today, more than ever before, need to understand how
the brick on which they are laboring fits into and is an integral part of the total structure.
Increasingly, the human is being recognized as a valid model for detailed investigation by
basic scientists. Physiology of the Heart will excite scientists, practitioners, and trainees
about the heart, and it will thereby help to move the field forward.
Eugene Braunwald, MD
Boston, MA
> Front of Book > Preface to the First Edition
Preface to the First Edition
Why write a textbook about the biophysical basis of cardiac function? Of what importance
are the energetics and chemistry of myocardial contraction to anyone but a physical
chemist or a biochemist? Why should electrical potentials at the surface of the myocardial
cell concern those who are not basic electrophysiologists? The answers to all of these
questions lie in the fact that virtually every important physiological, pharmacological, or
pathological change in cardiac function arises from alterations in the physical and
chemical processes that are responsible for the heartbeat.
Although it remains fashionable to consider the heart as a muscular pump, this organ is
much more than a hollow viscus that provides mechanical energy to propel blood through
the vasculature. It is an intricate biological machine that contains, within each cell, a
complex of control and effector mechanisms. Both the strength of cardiac contraction and
its electrical control are modulated by alterations in one or more of these cellular
mechanisms, which are involved in the fundamental processes of excitability, excitation-
contraction coupling, and contraction.
This text is written for medical students and graduate students in the biological sciences,
and for the physician who would like to find a simplified exposition of our current
understanding of the physiological and biophysical basis of cardiac function. Therefore,
this book is intended to provide a synoptic view of our present knowledge in this rapidly
expanding area. The major emphasis is on the relationships between the biochemical
properties of individual constituents of the myocardial cell, the biophysics of cardiac
muscle function, and the performance of the intact heart.
The task of relating these different aspects of cardiac function to each other has required
much selectivity, and undoubtedly, an excess of simplification and speculation. There can
be no doubt that much of this conceptual material will become invalid as our knowledge
of cardiac function advances. This is, after all, the lesson taught to us by the history of
science. The early neurophysiologists who tried to understand nerve conduction as the
passage of fluid down hollow tubes were trying to explain physiological phenomena in
terms of the limited biophysical knowledge of their time. With the development of an
understanding of animal electricity, the focus in neurophysiology shifted to studies of the
electrical properties of the nervous system, and attempts were made to explain
phenomena such as neuron-to-neuron communication and memory in terms of electrical
circuitry. More recently, the enormous advances in our knowledge of chemical
transmitters and the potential for information storage as newly synthesized
macromolecules has cast doubt on many of the theories of the great neurophysiologists of
the last century. Yet these were not unintelligent scientists. They were, however, required
to interpret their observations within the framework of knowledge that existed during
their lifetime. It would be presumptuous indeed for us now to assume that the evolution
of new principles of science has ended. For this reason, no apology is made for the
misconceptions and faulty interpretation that will inevitably accompany the present
attempt to organize our knowledge of cardiac function in terms of the broad principles
that are understood today.
The only true “facts” in biology are the results of individual experiments carried out
under controlled conditions by a carefully defined methodology. Yet, it is not the purpose
of this book to catalogue and discuss the biological “facts;” for this, the reader is referred
to the large number of reviews, symposia, multi-authored texts, and, most important,
individual scientific papers. Instead, the present text attempts to identify and describe
the unifying themes that connect different lines of investigation of the function of the
heart and, in so doing, to set out interpretations of these biological “facts.” The
bibliographies to each chapter are intentionally brief and generally include one or more
recent reviews to which the interested student may refer for more complete lists of
references. In some cases, “classic” articles are also cited.
Every effort has been made to keep this book simple—suitable for use as a text for
graduate and undergraduate teaching. Achievement of this goal, however, requires the
resolution, more or less arbitrarily as the case may require, of many serious conflicts, as
well as the addition of speculative material to connect important biochemical,
biophysical, physiological, and pathophysiological observations. It is the author's intention
that these departures into the realm of speculation be clearly identified in the text. Yet
the expert in these fields will undoubtedly be troubled by this attempt to provide a
coherent and unified text. While the author is not laboring under the illusion that all of
his interpretations will prove correct, it seems especially important to provide the
student with an indication of the significance of the many biological “facts” describing
the heart and its function rather than just to catalogue specific experimental findings. It
is, after all, the pattern on the fabric that holds the interest of most of us, rather than
the threads. For this reason, though with apologies to the protagonists of opposing
viewpoints, the author has chosen the present format for this text.
Arnold M. Katz
Heidelberg, Germany
1976
> Front of Book > Dedication
Dedication
To my father
Louis N. Katz
1897–1973
> Front of Book > Preface
Preface
The material covered in this text has undergone an unprecedented expansion since I
began the first edition of Physiology of the Heart in 1975. Thirty-five years ago I was able
to write from memory virtually all that I felt was essential to understand the physiology of
the heart. Reading a few papers, conversations with colleagues, and attendance at
meetings were all the background that I had needed. What a difference today! The
breadth of knowledge needed to understand cardiovascular physiology now includes topics
that did not exist in 1975, and material that could be summarized in a few sentences even
a decade ago is now the subject of reviews that are dozens of pages long and cite
hundreds of references.
This expansion poses a serious challenge to revising Physiology of the Heart as it increases
the difficulty in providing discussions that, while thorough and accurate, are not so
detailed as to defeat the purpose of this text which, as for the first edition, is to be
“simple—suitable for use as a text for graduate and undergraduate teaching.” In fact,
today's complexity raises the question as to why anyone should try to summarize all this
material in a single-authored book, especially because virtually all this information is
readily available from authoritative sources that can be located quickly using the
Internet. The answer to this question, and the reason that I have spent much of the past
year preparing this new edition, is that there is far more to understanding cardiovascular
physiology than knowing facts—or being able to look them up. It is necessary also to
understand how facts fit together to form patterns. This is because these patterns help
physicians and other health care providers to know what is happening to their patients,
and allow basic scientists to understand the relationships between specific areas of
biology and human disease. The importance of physiology in understanding disease is
obvious, but it is also true that efforts to understand heart disease have contributed to
our knowledge of normal cardiac physiology. My father often quoted his teacher, Carl
Wiggers, who observed that “every disease is an experiment that nature performs, and its
signs and symptoms are the manifestations of abnormal function.”
Comparison of the five editions of this text illustrates the extent to which cardiovascular
physiology has expanded during the past 35 years. Progress, however, has been uneven.
Fields like hemodynamics and electrocardiography have advanced within an established
framework of knowledge and so can be viewed as mature sciences. Advances in
biochemistry, molecular biology, and biophysics have been more significant, notably in
understanding energetics and metabolism, excitation–contraction coupling, and cardiac
electrophysiology. Most dramatic have been advances in signal transduction, which is
mentioned only briefly in the first edition, published in 1977. The second edition,
published in 1992, describes two types of regulation that I called “phasic” and “tonic”
because they mediate short-term and long-term responses, respectively; 10 years later, in
the third edition, these are called functional and proliferative. The former, which
activates short-term physiological responses, alters interactions between preexisting
structures to modify such physiological variables as heart rate, contractility, and
relaxation. Proliferative responses, on the other hand, bring about long-lasting changes in
the size, shape, and composition of the heart by changing myocyte structure, protein
synthesis, gene expression, and other molecular features of the heart. The fourth edition,
published in 2006, when signaling abnormalities were emerging as a major cause of
cardiovascular disease, contains separate chapters on each of these two types of signal
transduction. The importance of cytoskeletal signaling in mediating both adaptive and
maladaptive proliferative responses to cell deformation led me to expand a brief
discussion of the cytoskeleton in the third edition to a full chapter in the fourth edition.
This chapter became so detailed in the present edition that I decided to introduce it with
a brief passage from “The Catalogue of the Ships” in Homer's Iliad to explain why I
describe so many different proteins.
I am aware that my attempt to discuss a broad range of topics, which range from
molecular biology, through biochemistry and physiology, to clinical cardiology, might be
viewed as presumptuous. Because I am not an expert in all of the fields covered in this
text, it is inevitable that this book contains errors. In spite of this limitation, I have
prepared this new edition because I believe it important that readers have access to an
integrated discussion of cardiovascular physiology that is written in a single voice. I find
some comfort in a statement attributed to Dr. C. Sidney Burwell, who was Dean of
Harvard Medical School in the early 1950s, to the effect that “half of what the faculty
teaches medical students is wrong, but the faculty does not know which half.”
This revision of Physiology of the Heart challenges the view that there is a widening gap
between bench and bedside, between understanding the mechanisms of disease and how
to treat patients. History shows that this is not correct, and that new knowledge has been
filling, rather than widening the gap. The ancient Greeks and Romans, who viewed health
as a balance between opposing principles (the four humors), believed that the heart
generated heat that it distributed throughout the body in the blood; it is largely for this
reason that bleeding was viewed as a logical way to treat fever. It was not until 1628,
when William Harvey showed that the heart is a pump and not a furnace, that it became
possible to recognize the hemodynamic basis for the signs and symptoms of heart failure,
a syndrome that had been described many times during the preceding 2000 years.
However, the gap between science and medicine was so wide that this discovery was to
have little impact on patient care for the next 300 years. Throughout the 19th century,
when efforts to understand the causes of heart failure centered on cardiac hypertrophy,
correlations between clinical syndromes and autopsy findings led to the view, elegantly
stated in 1892 by William Osler, that hypertrophy begins as an adaptive response to
overload but eventually causes the heart to deteriorate.
Ernest Starling's description of the “Law of the Heart” and Carl Wiggers' work in the first
half of the 20th century made it possible to understand the hemodynamics of heart
disease. This narrowed the gap between bench and bedside and contributed to a
revolution in patient care when, in the 1940s, cardiac catheterization—pioneered by
Werner Forssmann, Andr© Cournand, and Dickinson Richards—made possible the precise
diagnoses of structural heart disease needed to allow surgeons to repair structurally
damaged hearts. In the 1950s, Stanley Sarnoff's description of “families of Starling curves”
clarified the concept of myocardial contractility, which a decade later led Eugene
Braunwald to demonstrate that contractility is depressed in failing hearts. The gap
between bench and bedside continued to narrow in the 1960s, when recognition of the
role of calcium in regulating cardiac contraction and relaxation led to the development of
new inotropic drugs.
The widely held view that heart failure is largely a hemodynamic disorder began to
unravel in the early 1990s, when clinical trials showed that although inotropic drugs cause
an immediate improvement in symptoms, they shorten long-term survival. At the same
time, direct-acting vasodilators, which because of their energy-sparing effects are of
short-term benefit in patients with heart failure, were found to have serious adverse
effects on long-term prognosis. Explanations for these and other unexpected findings
began to emerge when maladaptive consequences of overload-induced cardiac
hypertrophy were recognized to contribute to the poor prognosis in heart failure. The
practical importance of these discoveries became apparent in the 1990s when β-
adrenergic receptor blockers, whose negative inotropic effects transiently worsen
symptoms, were found to improve long-term outcome in part by inhibiting maladaptive
proliferative signaling. Today, as we enter the 21st century, insights from the emerging
fields of signal transduction and molecular biology, supplemented by discoveries in new
fields such as epigenetics, are stimulating an interplay between clinical cardiology and
molecular biology that encompasses the pathophysiology, treatment, and prevention not
only of heart failure, but also arrhythmias, sudden cardiac death, and vascular disease.
Even though the gap between bench and bedside is narrowing, the flood of new
information is making it difficult to find individuals who can teach basic science to
students of the clinical sciences, and who can teach students of biology the relevance of
basic science to patient care. The problem is especially serious in medical schools, where
an already overcrowded curriculum is generating pressures to shorten the time allocated
for teaching the mechanisms of disease. Because this threatens to lead to the graduation
of practitioners who lack the foundation needed to deliver optimal care to their patients,
a major goal of this revision is to help medical educators explain the interplay between
basic science and patient care in managing cardiac patients.
The importance of a scientific foundation in medical practice was noted by William Osler,
who in 1902 wrote: “A physician without physiology practices a sort of pop-gun pharmacy,
hitting now the disease and again the patient, he himself not knowing which.” Osler's
observation is of even greater relevance today because physicians have access to powerful
physiologically based therapy that, when used properly, is of immeasurable value to the
patient, whereas treatment lacking a solid basis in physiology often does more harm than
good.
The division of the fourth edition into four parts seemed to work well and so has not been
changed. Part I, Structure, Biochemistry, and Biophysics, reviews the structure,
biochemistry, and biophysics of the normal heart, the functions of the cytoskeleton, and
the chemistry of cardiac contraction, relaxation, and excitation–contraction coupling. Part
II, Signal Transduction and Regulation, contains an expanded discussion of the functional
signal transduction systems that bring about short-term responses, and the proliferative
signaling systems that control long-term changes in cardiac size, shape, and composition.
The physiology of the cardiac pump and the electrical signals that maintain the
homogeneity necessary for efficient contraction are described in Part III, Normal
Physiology. Part IV, Pathophysiology, begins with a description of the physiological basis
for the normal electrocardiogram, a marvelous diagnostic tool that uses physiological
principles to define pathophysiology, and concludes with chapters on arrhythmias,
ischemic heart disease, and heart failure, the major types of heart disease encountered in
developed societies.
Although there have been no changes in overall organization, virtually the entire text of
this edition has been rewritten, new figures have been added, and many of the older
figures revised. Much of the “mature” science has been condensed and material that no
longer seems important has been removed. The discussions of arrhythmias, ischemic heart
disease, and heart failure that end this text have been updated to reflect the growing
impact of molecular biology on our understanding of the pathophysiology and
management of these syndromes. As this book is designed primarily for the nonexpert, no
attempt has been made to document the many details; instead, bibliographies are
included to identify sources for further reading. I have kept a few references to classical
papers because these contain clear descriptions of important concepts and, perhaps more
important, they provide students with an understanding as to how we got to where we are
today.
My major goal in preparing this fifth edition of Physiology of the Heart, as for previous
editions, is to provide a readable and comprehensive text that explains normal cardiac
function and how altered function causes disease. This text is not a reference book to be
consulted to verify facts but instead is intended to be read from cover to cover. Stated
simply, my goal is to help physicians and other health care professionals understand the
basic sciences, and basic scientists to appreciate how specific areas of research relate to
the broad sweep of cardiac physiology.
Arnold M. Katz, MD, D.Med (Hon), FACP, FACC
Norwich, Vermont
> Front of Book > Acknowledgments
Acknowledgments
More than 35 years have passed since I had planned to coauthor a textbook on cardiac
physiology with my father, to whom this book is dedicated. Dad's death in 1973 made this
impossible, but those who remember him will, I hope, recognize his forthright and lucid
approach in this text.
The 1st edition of this book was published when I was Philip J. and Harriet L. Goodhart
Professor of Medicine (Cardiology) at the Mount Sinai School of Medicine of the City of
New York; most of the text was written at the Max-Planck-Institut-f©r Medizinische
Forschung in Heidelberg, Germany, and was supported in part by the Alexander von
Humboldt Stiftung. The 2nd edition was written in the Dana Medical Library at Dartmouth
Medical School during a sabbatical year when I was Professor of Medicine (Cardiology) at
the University of Connecticut. The 3rd, 4th, and present editions were written on our
hilltop in Norwich, Vermont, after I had retired from the University of Connecticut. I
thank Dartmouth Medical School for providing me with library privileges and an Internet
link that allowed me to work from my home. I thank Indu Jawwad and her team at Aptara
for dealing patiently with my many fussy corrections. I am especially grateful to Frances
DeStefano and Lippincott Williams & Wilkins for their confidence in asking me to write
this 5th edition.
I warmly acknowledge the probing questions posed by the students I have taught over the
past 45 years at Columbia University, the University of Chicago, the Mount Sinai School of
Medicine, the University of Connecticut, Dartmouth Medical School, and most recently
Harvard Medical School where I teach in a 2nd-year course that I took as a student in
1953. These students continue to serve as gentle but firm critics of my efforts to explain
things.
The understanding of my children and their families when I disappeared during their visits
to work on this edition is gratefully acknowledged, as are Eurykleia (“Kleia”), our Springer
Spaniel, who took me from my desk for walks in the woods that restored my circulation
and recharged my intellectual batteries, and Pyrrhus, our long-haired ginger cat who
occasionally sat in my inbox while I wrote this text. Above all, I thank Phyllis, my wife, for
her steadfast and loving support during the past 51 years and for providing an
intellectually stimulating and tranquil environment without which this text could not have
been written.
Arnold M. Katz, MD, D.Med (Hon), FACP, FACC
Professor of Medicine Emeritus, University of Connecticut School of Medicine, Visiting
Professor of Medicine and Physiology, Dartmouth Medical School, Visiting Professor of
Medicine, Harvard Medical School
> Table of Contents > Part One - Structure, Biochemistry, and Biophysics > Chapter 1 - Structure of the Heart and Cardiac
Muscle
Chapter 1
Structure of the Heart and Cardiac Muscle
It has been shown by reason and experiment that blood by the beat of the ventricles flows through the
lungs and heart and is pumped to the whole body … the blood in the animal body moves around in a circle
continuously, and … the action or function of the heart is to accomplish this by pumping. This is the only
reason for the motion and beat of the heart.
—William Harvey (1628).
Exercitatio Anatomica de Moto Cordis et Sanguinis in Animalibus.
Harvey's proof that the heart is a muscular pump, which overthrew the ancient view that the heart is the
source of the body's heat, along with increasing use of human autopsies, made it possible to understand
the pathogenesis of heart disease in terms of abnormal organ structure (Katz, 2008). Virchow's founding of
pathology in the 19th century supplemented this anatomical knowledge with histological pathology. In the
20th century, development of electron microscopy and new insights into the biochemistry and biophysics
of cardiac function further extended knowledge of the causes of heart disease. Today, rapid advances in
molecular biology are providing additional insights into the ultrastructural and molecular basis of
cardiovascular disease.
Organ Structure
Mammalian hearts can be viewed as two pumps that operate in series: the right atrium and the right
ventricle, which pump blood from the systemic veins into the pulmonary circulation, and the left atrium
and the left ventricle, which pump blood from the systemic veins into the pulmonary circulation (Fig. 1-
1). Within the heart, atrioventricular (AV) valves prevent blood from flowing backward from the
ventricles into the atria: on the right the tricuspid valve and on the left the mitral valve (Fig. 1-2).
Semilunar valves, named for their crescent-shaped cusps, separate each ventricle from its great artery:
the pulmonic valve between the right ventricle and the pulmonary artery, the aortic valve between the
left ventricle and the aorta. All four of these valves lie in a plane within a connective tissue “skeleton”
that separates the atria and ventricles in which the mitral, tricuspid, and aortic valves surround a fibrous
triangle, called the central fibrous body (Fig. 1-3). This connective tissue skeleton can be viewed as an
insulator that prevents electrical impulses from being conducted between the atria and the ventricles.
The AV bundle (also called the common bundle or bundle of His), a strand of specialized cardiac muscle
that penetrates this insulator, normally provides the only conducting pathway between the atria and the
ventricles. Damage to this critical conducting structure is an important cause of AV block (Chapter 16).
P.4
Fig. 1-1: Circulation of the blood. Light shading: deoxygenated blood; dark shading: oxygenated
blood. RA, right atrium; LA, left atrium; RV, right ventricle; LV, left ventricle. (Adapted and
modified from Starling, 1926).
The free margins of the semilunar aortic and pulmonary valve cusps are supported by thick tendinous
edges. Sinus of Valsalva lie behind each of the three aortic valve cusps; the anterior and left posterior
sinuses contain the orifices of coronary arteries (see below), whereas the right posterior sinus does not
give rise to a coronary artery and so is often called the “noncoronary” sinus (see Fig. 1-3). The larger
cusps of the mitral and tricuspid valves are tethered at their free margins by fibrous chordae tendinae
that attach to “fingers” of myocardium called papillary muscles that project into the right and left
ventricular cavities (see Fig. 1-2). Much as the strands of a parachute arise from a skydiver's harness,
several chordae tendinae fan out from each papillary muscle to support the valve margins (Becker and
deWit, 1979). Laxity of the connective tissue supporting the mitral valve can allow the leaflets to move
backward (prolapse) into the left atrium when left ventricular pressure rises during systole. In some
patients, abrupt opening of these abnormal valves causes an audible “click”; if blood subsequently leaks
from the left ventricle into the left atrium (mitral regurgitation), a late systolic murmur can be heard.
This syndrome, called mitral valve prolapse, is often of no hemodynamic significance (although leaky
valves are susceptible to bacterial infection); when severe, however, laxity of the chordae tendinae can
cause significant mitral regurgitation. Rupture of a papillary muscle, which can occur when coronary
artery occlusion interrupts the blood supply to these vital muscular structures (see Chapter 17), generally
causes severe mitral regurgitation and can be fatal.
P.5
Fig. 1-2: Major structures in a human heart opened after transection slightly anterior to the midline.
(Adapted and modified from Berne and Levy, 1967.)
Fig. 1-3: Schematic diagram of the connective tissue skeleton of the heart, viewed from above,
showing the four valves and the atrioventricular (AV) bundle that crosses this insulating structure
through the central fibrous body. Sinuses of Valsalva lie behind the aortic valve cusps, two of which
give rise to coronary arteries. The ostium of the left main (LM) lies in the left posterior (LP) sinus,
while the right coronary artery (RCA) originates in the anterior sinus (A); the third sinus of Valsalva,
the right posterior (RP), is called the “noncoronary” sinus because it does not give rise to a
coronary artery. The sharper right border of the heart forms the “acute margin,” the more rounded
left border is the “obtuse margin.”
P.6
Architecture of the Walls of the Heart

When the heart is viewed from above, the rounded margin of the left ventricle forms an above obtuse
angle, whereas the margin of the right ventricle is sharper, like an acute angle (see Fig. 1-3); this explains
the terms “obtuse marginal” and “acute marginal” used in naming branches of the coronary arteries (see
below). The thin-walled atria, which develop much lower pressures than do the ventricles, contain ridges
of myocardium called pectinate muscles that may provide preferential conducting pathways, often
referred to as internodal tracts or sinoatrial (SA) ring
P.7
bundles, that link the SA and AV nodes (Hayashi et al., 1982). The ventricles, which develop much higher
pressures than do the atria, have thicker muscular walls. The left ventricle, which has approximately
three times the mass and twice the thickness of the right ventricle, can be viewed as a “pressure pump”
whose cavity resembles an elongated cone in which the mitral valve, through which blood flows into the
ventricle, and the aortic valve, through which blood leaves the ventricle, lie side-to-side in the wider end
(Fig. 1-4). Peak systolic pressure in the left ventricle is normally about three times higher than that in the
right ventricle; the latter, which represents a “volume pump,” is shaped like a crescent with inflow
through the tricuspid valve at one end and outflow through the pulmonic valve at the other (see Fig. 1-4).
During systole, the interventricular septum normally moves toward the left ventricular free wall and so
participates in left ventricular ejection. In chronic right ventricular overload, for example, in patients
with pulmonary hypertension, the septum can move paradoxically away from the left ventricular cavity
during systole to aid right ventricular ejection.
Fig. 1-4: A: Schematic anterior views of the right and left ventricular chambers. The inflow
(tricuspid) and outflow (pulmonic) valves in the U-shaped right ventricle are widely separated,
whereas in the conical left ventricle, the mitral and aortic valves lie side-by-side, where they are
separated by the anterior leaflet of the mitral valve. B: Casts of canine right and left ventricles
showing approximate locations of the pulmonic (PV) and tricuspid (TV) valves in the right ventricle,
and the aortic (AV) and mitral (MV) valves in the left ventricle. Left: anterior view. Right: superior
view.
The left ventricular cavity, which is conical in shape during diastole, assumes a more spherical shape as
intraventricular pressure rises at the end of isovolumic contraction (Hawthorne, 1961, 1969) (Fig. 1-5).
During left ventricular ejection, its cavity again assumes its conical shape. Because ejection propels blood
superiorly (toward the head), according to Newton's third Law—which states that for every action there is
an equal and opposite reaction—the base of the heart moves inferiorly (toward the feet). This movement,
called “decent of the base,” explains the prominent “x descent” seen in the normal venous pulse during
ventricular ejection.
The heart, along with a small amount of fluid, is contained within a noncompliant fibrous sac called the
pericardium whose inner surface, the parietal pericardium, is continuous with the epicardium (see
below). The cavities of the atria and the ventricles, along with the valves, are lined with another
connective tissue layer called the endocardium (Brutsaert, 1989). Because the heart is contained within
the rigid pericardium (see below), the ventricles interact with one another. Ventricular interactions are
especially important in diastole, when dilatation of one ventricle can impair filling of the other (Yacoub,
1995; Santamore and Dell'Italia, 1998; Williams and Frenneaux, 2006).
The muscular walls of the ventricles are made up of overlapping fiber bundles, sometimes called
bulbospiral and sinuspiral muscles, that follow spiral paths as they sweep from the fibrous skeleton at the
base of the heart to its apex (Grant, 1965; Lower, 1669; Streeter et al., 1969;
P.8
Fenton et al., 1978; Sengupta et al., 2006). The bundles at the epicardial surface of the left ventricle are
oriented as a right-handed helix that tends to parallel the base-apex axis of the heart, whereas those at
the endocardial surface form a left-handed helix and are oriented more circumferentially (Cheng et al.,
2008) (Fig. 1-6).
Fig. 1-5: Schematic diagrams of the canine left ventricle at the ends of diastole, isovolumic
contraction, and ejection. The broad lightly shaded arrow in the right-hand side of the diagram
shows the “decent of the base” in an inferior direction as blood is ejected in a superior direction in
to the aorta (darkly shaded arrow). Ao, aorta; LA, left atrium; LV, left ventricle. (Adapted and
modified from Hawthorne, 1961.)
Fig. 1-6: Spiral musculature of the ventricular walls. A: Spiral bundles in the left ventricle. Left:
anterior view, Right: inferior view. (Modified from Lower, 1669.). B: Schematic drawing of the
spiral bundles that sweep from the fibrous skeleton at the base of the heart (above) to the apex
(below). C: Schematic diagram showing the different helical orientations of the fiber bundles in the
subendocardium (left) and the subepicardium (right).
Electrical Activation
The heartbeat is initiated and controlled by electrical impulses that are generated and conducted by
specialized myocardial cells in different regions of the heart. Activation normally begins in the SA node
(Fig. 1-7), a band of spontaneously depolarizing cells derived from the embryonic right sinus venosus that
lies between the superior vena cava and the right atrium (Oosthoek et al., 1993a; Anderson and Ho, 1998;
Verheijck et al., 1998). Because the firing rate of the SA node is more rapid than that of the other regions
of the heart, this structure normally serves as the cardiac pacemaker (see Chapter 15).
The wave of depolarization initiated by the SA node is propagated through atrial myocardial cells to the
right atrium, and then to the left atrium. After encountering a delay in the slowly
P.9
conducting cells of the AV node, which is derived from the embryonic left sinus venosus, the wave of
depolarization enters the AV bundle (Oosthoek et al., 1993b). The latter, a rapidly conducting structure
made up of Purkinje cells (see below), bifurcates into right and left bundle branches at the top of the
interventricular septum. The right bundle branch crosses the right ventricular cavity within the moderator
band, a muscular bundle that extends from the interventricular septum to the base of the papillary
muscle that supports the anterior leaflet of the tricuspid valve (Fig. 1-7). The left bundle branch is often
pictured as bifurcating into anterior and posterior fascicles, but this branching is highly variable (see
Chapter 15). The impulses conducted through the bundle branches reach the ventricular myocardium via
the His-Purkinje system, a subendocardial network of rapidly conducting cells that synchronizes
ventricular activation.
Fig. 1-7: Conducting system of the human heart (capitalized labels at right) and major anatomical features
(lower case labels at left). AV, atrioventricular; SA, sinoatrial. (Modified from Benninghoff, 1944.)
The Coronary Circulation
Major Epicardial Coronary Arteries

Large epicardial coronary arteries carry virtually all of the blood that supplies the heart. Although a few
layers of endocardial myocytes are perfused from the ventricular cavities via arteriosinusoidal and
arterioluminal vessels, this auxiliary blood supply is of no clinical importance when a large coronary
artery becomes occluded (Chapter 17). The major coronary arteries are the left main coronary artery
(LEFT MAIN), right coronary artery (RCA), left anterior descending (LAD), circumflex (CIRC), and posterior
descending artery (PDA) (Fig. 1-8). All lie
P.10
in grooves between the heart's chambers: the RCA and the CIRC between the atria and the ventricles, the
LAD and the PDA between the left and right ventricles.
Fig. 1-8: Major coronary arteries and their branches (labels at right and left) and key elements of the
cardiac conduction system (labels above and below). AV, atrioventricular; SA, sinoatrial.
The anatomy of these vessels can be summarized by the statement three out of two makes four. Two
coronary arteries arise from the aorta (RCA and LEFT MAIN) and, after the LEFT MAIN divides into the LAD
and CIRC, continue as three vessels (RCA, LAD, and CIRC). The PDA is a continuation of either the RCA or
the CIRC, so that the myocardium is supplied by four major arteries (RCA, LAD, CIRC, and PDA).
The LEFT MAIN, which originates in the left posterior sinus of Valsalva (see Fig. 1-3), continues as a single
vessel of variable length before dividing into its two major branches: the LAD and the CIRC (see Fig. 1-8).
The LAD, which courses down the anterior interventricular groove, gives rise to septal perforating
arteries that supply the anterior two-thirds of the interventricular septum, diagonal branches that supply
the anterior wall of the left ventricle, and right ventricular branches that provide blood to the anterior
wall of the right ventricle. After crossing the apex of the heart, the LAD usually turns upward, toward the
base, to run a short distance in the posterior interventricular groove (see Fig. 1-8; Fig. 1-9). The CIRC,
which courses to the left in the anterior AV groove, gives rise to obtuse marginal branches that supply the
lateral wall of the left ventricle. In most human hearts, the CIRC, after reaching the back of the left
ventricle, runs only a short distance down the posterior interventricular groove to end near the crux of
the heart, where the plane of the interventricular septum crosses the plane of the AV groove (Fig. 1-9B).
P.11
Fig. 1-9: Posterior view of the human heart showing left dominant (A) and right dominant (B) coronary artery
distribution. In the left dominant distribution, the posterior descending artery (PDA) is a continuation of the
circumflex branch of the left coronary artery (CIRC) that runs from the crux of the heart down the posterior
interventricular groove; more commonly, in the right dominant distribution, the posterior descending artery
is a continuation of the right coronary artery (RCA). The left anterior descending (LAD) coronary artery, after
wrapping around the inferior surface of the heart, usually courses upward for a short distance in the posterior
interventricular groove. LA, left atrium; RA, right atrium; LV, left ventricle; RV, right ventricle; SVC, superior
vena cava, crux; point at which the plane of the interventricular septum crosses that of the atrioventricular
groove.
The PDA, which can arise from either the CIRC or the RCA, runs inferiorly in the posterior interventricular
groove where it supplies septal perforating branches that perfuse the posterior third of the
interventricular septum. In approximately 90% of human hearts the PDA is supplied by the RCA (called
“right dominant”); in the remaining approximately 10% the CIRC turns downward at the crux to supply the
PDA (“left dominant”) (Fig. 1-9A).
The RCA, which arises from the anterior sinus of Valsalva, courses toward the right in the anterior AV
groove where it gives rise to right ventricular (acute marginal) branches that supply the free wall of the
right ventricle. The RCA then crosses the acute margin of the heart, turns to the left in the posterior AV
groove, and, after reaching the crux of the heart, usually continues in the posterior interventricular
groove as the PDA (“right dominant” coronary circulation, Fig. 1-9B).
Coronary occlusive disease is often described as “one-vessel,” “two-vessel,” and “three-vessel” disease,
terms that describe how many of the three major arteries (RCA, LAD, and CIRC) are narrowed
significantly. Obviously, the more vessels that are narrowed, the worse is the clinical prognosis. LEFT MAIN
disease is especially dangerous because this vessel supplies both of the arteries that supply blood to the
left ventricle (LAD and CIRC).
Collateral Vessels
Occlusion of a large epicardial artery generally causes an infarct (defined as an area in which cells have
died because of inadequate blood supply) whose borders are sharply demarcated from the adjacent
normally perfused myocardium supplied by other, nonoccluded arteries. Intracoronary collateral vessels
can connect the vascular beds supplied by different large epicardial arteries, such as the RCA, LAD, and
CIRC. Although collateral vessels are usually poorly developed in younger individuals, in older patients,
especially those with long-standing coronary atherosclerosis, intracoronary collaterals can enlarge and
provide blood flow to regions of the heart downstream from
P.12
an occluded coronary artery. However, collaterals are not found at the level of the microcirculation, so
that there is usually little or no “border zone,” defined as a region where a limited blood supply is
preserved at the edge of a myocardial infarct (Factor et al., 1981).
Blood Supply to the Ventricular Myocardium

Blood from the epicardial arteries reaches the myocardium via muscular branches that traverse the walls
of the ventricles (Fig. 1-10); normal compression of these muscular branches during systole explains why
virtually all nutrient coronary flow occurs during diastole, and why the subendocardial regions of the
thick-walled left ventricle are especially vulnerable to coronary artery narrowing.
The left ventricular papillary muscles receive their blood supply from large penetrating vessels called
perforators. The anterolateral papillary muscle, which supports the anterior leaflet of the mitral valve,
has a dual blood supply derived from branches of the CIRC and the LAD. The posteromedial papillary
muscle, which supports the posterior leaflet, receives its blood supply from the RCA or the CIRC via the
PDA.
Blood Supply to the Conduction System

The SA node is perfused by the SA node artery (see Fig. 1-8), which in slightly more than half of human
hearts is a branch of the RCA; in the remainder, this artery arises from the CIRC. The AV node is usually
supplied by an AV node artery that is a branch of the PDA, so that the blood supply to the AV node is
derived from the RCA in approximately 90% of human hearts and the CIRC in approximately 10%.
Fig. 1-10: X-ray microphotograph of a cross section of a left ventricle following injection of the coronary
arteries with radiopaque dye. The large coronary arteries that course over the epicardial surface of the
ventricle give rise to muscular branches that penetrate the myocardium and reach the endocardium after
traversing the thick ventricular wall (retouched).
P.13
The AV bundle, along with proximal portions of both right and left bundle branches, is perfused by septal
perforators that arise from both the LAD and the PDA. Because these critical conducting structures have a
dual blood supply, the appearance of a conduction block in the AV bundle or bundle branches in a patient
with an acute myocardial infarction implies that more than one major coronary artery is occluded. The
anterior division of the left bundle branch and midportion of the right bundle branch are supplied by
septal perforators arising from the LAD, whereas the posterior division of the left bundle branch is
perfused by septal perforators supplied by the PDA.
Coronary Venous Drainage

The venous effluent of the heart is collected in large veins that parallel the epicardial coronary arteries.
Most venous drainage of the left ventricle enters the coronary sinus, which parallels the CIRC in the left
posterior AV groove before emptying into the floor of the right atrium. A portion of the venous drainage of
the left ventricle, along with much of that derived from the right ventricle, enters the right atrium
through anterior cardiac veins. A small fraction of the venous drainage of the ventricular myocardium
flows directly into the cavities of the right and left ventricles by way of Thebesian veins.
Because the coronary sinus passes to the left behind the heart in the left AV groove, an electrode
catheter inserted into this venous channel can be used to record the electrical activity of the left atrium
and ventricle, stimulate these structures, and perform ablation therapy on the left side of the heart.
Fractal Anatomy of the Heart

Many structures in the heart, including the coronary blood vessels, chordae tendinae, and interventricular
conduction system, form networks whose seemingly disorganized branching actually follows complex
rules. The latter can be described mathematically as fractals, which define the order often found in
seemingly random biological structures (Goldberger et al., 1990). Goldberger et al. (2002) note that
disease and aging are often associated with a “breakdown of fractal physiological complexity [that] may
be associated with excessive order (pathological periodicity) on the one hand, or uncorrelated
randomness on the other.”
Lymphatics
Fluid that is transudated across the capillary endothelium and enters the cardiac interstitium is returned
to the circulation via the lymphatic system. In the heart, the larger lymphatic vessels run alongside the
coronary arteries and veins in the AV and interventricular grooves. Most cardiac lymphatic channels cross
the anterior surface of the pulmonary artery to reach pretracheal lymph nodes and a cardiac lymph node
situated between the superior vena cava and the right innominate artery. The lymph ultimately drains
into the thoracic duct (Miller, 1982).
Innervation
The heart is innervated by both sympathetic and parasympathetic nerves. Most postganglionic
sympathetic fibers reach the heart from the fourth and fifth thoracic segments of the spinal cord after
forming synaptic connections in the cervical and thoracic cervical ganglia (often
P.14
called stellate ganglia) and in the cardiac plexus, a network of sympathetic fibers located at the base of
the heart. Once they arrive at the heart, postsynaptic sympathetic nerves do not form specialized
junctions but instead lie in plasma membrane depressions on the surface of cardiac myocytes where they
release norepinephrine, the sympathetic neurotransmitter. The heart's parasympathetic innervation
originates in the dorsal efferent nuclei of the medulla oblongata and reaches the heart via the cardiac
branches of the vagus nerve. Preganglionic parasympathetic fibers impinge on postganglionic cells in the
SA and AV nodes, the atria, and the heart's blood vessels; parasympathetic innervation of the ventricular
myocardium is more limited.
Sensory fibers that originate in the heart reach the brain stem via the cardiac plexus. Activation of these
fibers in patients with coronary artery occlusive disease causes a chest discomfort, called angina pectoris.
Like other visceral pain, the discomfort caused by cardiac ischemia is poorly localized and perceived
differently by different individuals.
Stretch receptors located in the inferior and posterior walls of the left ventricle can evoke a powerful
vagal response, called the von Bezold-Jarisch reflex, that slows the SA node pacemaker, inhibits
conduction through the AV node, and causes peripheral vasodilatation (Dawes and Comroe, 1954). This
parasympathetic reflex is commonly activated in inferior and posterior wall myocardial infarction (see
Chapter 17).
Histology
The outer surfaces of the atria and the ventricles are lined by the epicardium, a layer of squamous cells
that overlies a network of fibroelastic connective tissue, that is continuous with the inner layer of the
pericardium. The endocardium, which lines the heart's chambers, is made up of squamous cells, a mesh of
collagen and elastic fibers, and a rudimentary layer of smooth muscle.
The myocardium, which makes up the vast majority of the heart's thickness, contains both myocytes and
connective tissue. Although cardiac myocytes represent most of the myocardial mass, approximately 70%
of the cells are smaller nonmyocytes, which include vascular smooth muscle, endothelial cells, and
fibroblasts. The latter secrete and maintain the connective tissue fibers that contribute to the heart's
tensile strength and stiffness. This connective tissue framework is organized into the endomysium, which
surrounds individual cardiac myocytes, the perimysium, which supports groups of myocytes, and the
epimysium, which encases the entire muscle (Fig. 1-11).
Several types of cardiac myocytes are found in the adult human heart (Fig. 1-12). Working myocytes,
which are specialized for contraction, are found in the atria and the ventricles; atrial myocytes are
smaller than those of the ventricles. Purkinje fibers, found in the AV bundle, bundle branches, and
ventricular endocardium, are large, pale, glycogen-rich cells that are specialized for rapid conduction and
have few myofilaments. Nodal cells in the SA and AV nodes, which are responsible for pacemaker activity
and an AV conduction delay, respectively, are small pale cells that also contain few myofilaments.
Additional heterogeneity is seen at the molecular level (Katz and Katz, 1989); for example, in human
atria, different cardiac myocyte molecular phenotypes are distributed in a mosaic pattern (Fig. 1-13)
(Sartore et el., 1981; Bouvagnet et al., 1984).
The many different types of cardiac myocytes form a branched network that was once believed to
represent an anatomical syncytium. However, the intercalated discs, which are densely staining
transverse bands that characteristically appear at right angles to the long axis of the cardiac myofibers,
are now known to represent specialized cell–cell junctions that form strong mechanical linkages between
cells. The intercalated discs also contain pores that reduce internal
P.15
electrical resistance (see below). Although the heart is not a true anatomical syncytium, all of its
myocytes are in free electrical communication.
Fig. 1-11: Connective tissue framework of the human heart showing groups of myocytes surrounded
by the perimysium (P). A weave of endomysium that surrounds the individual myocytes (W) forms
lateral struts (S) that connect adjacent cells. Collagen struts also connect myocytes to microvessels
(thin arrow) and to the perimysium (thick arrow). (From Rossi et al., 1998, by permission of the
American Heart Association.)
Working cardiac myocytes are filled with cross-striated myofibers and mitochondria and usually contain a
single centrally located nucleus (Fig. 1-12A). The rapidly conducting Purkinje fibers are large pale cells
that contain more glycogen but fewer contractile filaments and mitochondria (Fig. 1-12B). Cells
intermediate in appearance between the Purkinje fibers and the working cardiac myocytes are called
transition cells (Fig. 1-12E). The myocytes in the SA node (Fig. 1-12C) and the AV node (Fig. 1-12D), like
Purkinje fibers, are rich in glycogen and contain few contractile filaments; however, nodal cells conduct
slowly in part because of their small size.
P.16
Unlike working cardiac myocytes, which rely on oxidative metabolic pathways to generate adenosine
triphosphate (ATP), the myocytes that make up the heart's conduction system are capable of significant
anaerobic energy production (Henry and Lowry, 1983).
Fig. 1-12: Human cardiac myocytes. A: Working ventricular myocytes contain cross striations, central nuclei,
and intercalated discs. B: Purkinje fibers are large, poorly staining cells with sparse cross striations. The
sinoatrial node (C) and atrio-ventricular node (D) are networks of small, sparsely cross-striated cells. E:
Transition cells are seen where Purkinje fibers (left) impinge on the working myocardium (right). (Modified
from Benninghoff, 1944).
Atrial cardiac myocytes contain granules that represent stores of biologically active natriuretic peptides,
which are natriuretic and diuretic and relax vascular smooth muscle. These peptides are released when
the walls of the heart are stretched, which helps the body defend against expanded
P.17
blood volume (see Chapter 8). This means that the heart is not only a pump but also an endocrine organ!
Fig. 1-13: Microphotographs of serial sections from human right atrium stained with two different
immunofluorescent antiventricular human myosin antibodies (A, B) and a histochemical marker for myosin
ATPase activity (C). One antibody binds to all atrial myosin isoforms (A), the second binds selectively to some
cells (B). The arrowheads in (B) and (C) show a fiber that binds the second antibody (B) but exhibits weak
ATPase activity (C) while the arrows show a cell that binds weakly to the second antibody but exhibits high
ATPase activity. Bar = 20 µm. (Reprinted from Bouvagnet et al., 1984, by permission of the American Heart
Association, Inc.)
Ultrastructure
The contractile proteins, which make up almost half the volume of working cardiac myocytes (Table 1-1),
are organized in a regular array of cross-striated myofibrils (Figs. 1-14 through 1-16). Most of the
remaining cell volume is occupied by mitochondria, which generate the large
P.18
amounts of high-energy phosphate required for contraction. Key membrane systems that regulate the
performance of cardiac myocytes include the plasma membrane, which separates the cytosol from the
surrounding extracellular space, extensions of the plasma membrane called t-tubules that penetrate the
interior of these cells, and the intracellular membranes of the sarcoplasmic reticulum (Table 1-2).
Fig. 1-14: Electron microphotograph of two normal human left ventricular myocytes (above and
below) that are separated by a narrow extracellular space (oriented from right to left in the
center of the figure). Sarcomeres are aligned both within and between cells. Arrowheads:
endomysium between cells. M, mitochondria; Z, Z-lines; D, intercalated disc; L, lipid droplet; *, t-
tubule. Scale bar = 2 µM. (Reproduced with permission from Gerdes et al., 1995.)
Table 1-1 Components of a Working Myocardial Cell (Rat Left Ventricle)
Component Percentage of Cell Volume
Myofibrils 47
Mitochondria 36
Sarcoplasmic reticulum 3.5
Subsarcolemmal cisternae 0.35
Sarcotubular network 3.15
Nuclei 2
Other (mainly cytosol) 11.5
Modified from Page (1978).
Myofibrils
The cross-striated pattern in working atrial and ventricular cardiac myocytes reflects the highly ordered
distribution of two types of filaments: thick filaments, which extend the length of the A-band, and thin
filaments, which extend from the Z-lines toward the center of the sarcomere (see Chapters 4 and 6). The
fundamental unit of striated muscle, the sarcomere, is defined as the region between two Z-lines; each
sarcomere therefore includes a central A-band and the two adjacent half I-bands. The darkly staining
striations contain a parallel array of thick and thin filaments that strongly rotate polarized light and so
are highly birefringent (anisotropic), hence their designation A-bands. The lightly staining striations,
which contain only thin filaments, are less birefringent (more isotropic) and so are called I-bands. Both
thick and
P.19
P.20
thin filaments, along with the darkly staining Z-line that bisects each I-band, contain several cytoskeletal
proteins (see Chapter 5).
Fig. 1-15: Ultrastructure of a working cardiac myocyte. Contractile proteins are arranged in a regular array
of thick and thin filaments (seen in cross section at the left). The A-band represents the region of the
sarcomere occupied by the thick filaments into which thin filaments extend from either side. The I-band
contains only thin filaments that extend toward the center of the sarcomere from Z-lines that bisect each I-
band. The sarcomere, the functional unit of the contractile apparatus, lies between two Z-lines and contains
one A-band and two half I-bands. The sarcoplasmic reticulum, an intracellular membrane system that
surrounds the contractile proteins, consists of the sarcotubular network at the center of the sarcomere and
the subsarcolemmal cisternae. The latter form specialized composite structures with the transverse tubular
system (t-tubules) called dyads. The t-tubular membrane is continuous with the sarcolemma, so that the
lumen of the t-tubules contains extracellular fluid. Mitochondria are shown in the central sarcomere and in
cross section at the left. (Modified from Katz, 1975).
Fig. 1-16: Electron microphotograph of a sarcomere in normal human left ventricle. A grazing section on the
left side of the sarcomere shows the sarcotubular network (S) overlying the I-band. Three mitochondria are
seen above the sarcomere. M, M line; A, A-band; I, I-band; Z, Z-line; T, t-tubule. Scale bar = 2 µM.
(Reproduced with permission from Gerdes et al., 1995).
Table 1-2 Membrane Surface Areas in a Working Myocardial Cell (Rat Left Ventricle)
Membrane µm2 Membrane Area Per µm3 Cell Volume
Plasma membrane 0.465
Sarcolemma 0.31
t-Tubules 0.15
Nexus 0.005
Total sarcoplasmic reticulum 1.22

Subsarcolemmal cisternae 0.19
Sarcotubular network 1.03
Mitochondria 20
Modified from Page (1978).
In cross section, the A-band is a hexagonal array of thick filaments, each of which is surrounded by six
thin filaments that lie at the trigonal points between adjacent thick filaments (Figs. 1-17 and 1-18). In the
I-band, which lacks thick filaments, the thin filaments are less ordered. Radial cross-links, formed by
myosin-binding protein C, link the thick filaments in a hexagonal array at the center of the A-band.
The thick filaments are composed largely of myosin polymers (see Chapter 4) and a huge cytoskeletal
protein called titin. The central regions of the thick filaments contain several additional cytoskeletal
proteins, including myosin-binding protein C, M-protein, myomesin, and the MM isoform of creatine
phosphokinase (see Chapter 5). Cross-bridges that project from the thick filaments and interact with the
thin filaments represent the heads of myosin molecules. The thin filaments are double-stranded actin
polymers that include tropomyosin and the three proteins of the troponin complex. The Z lines, in which
the thin filaments are interwoven with a number of cytoskeletal proteins, link adjacent sarcomeres to one
another and to the extracellular matrix (see Chapter 5).
Fig. 1-17: Schematic cross sections at three regions of the sarcomere. A: In the A-band thin filaments lie at
the trigonal points in a hexagonal array of thick filaments. I: In the I-band, where thick filaments are absent,
the thin filaments are less ordered. M: In the M-band at the center of the A-band thin radial filaments made
up of myosin-binding protein C connect adjacent thick filaments.
P.21
Fig. 1-18: Cross section of a cat right ventricular papillary muscle, showing mitochondria (Mito)
and myofilaments cut at the level of the A-band (A), I-band (I), and M-band (M); in the latter, radial
filaments link adjacent thick filaments (Compare with Fig. 1-17). The Z-line (Z) appears as a dense
network. (From McNutt and Fawcett, 1974.)
One of most important discoveries in muscle physiology was that the lengths of the thick and thin
filaments remain constant during contraction and relaxation (Hanson and Huxley, 1953; Huxley, 1953).
These findings demonstrated that muscle contraction is not brought about by folding of elongated
contractile protein filaments but indicated instead that muscle shortening is caused by changes in the
extent of overlap between thick and thin filaments (Fig. 1-19) (see Chapter 6).
The Plasma Membrane and Transverse Tubules

Cardiac myocytes are surrounded by a plasma membrane (sarcolemma) that separates the intracellular
and extracellular spaces (see Fig. 1-15). This membrane contains channels, carriers, and pumps that
regulate cell composition and function; receptors and enzymes that participate in cell signaling; and
cytoskeletal molecules that link cells to each other and to the extracellular matrix. Extensions of the
plasma membrane, called transverse tubules (t-system), penetrate the cell where they play a key role in
excitation–contraction coupling by transmitting action potentials deep into the cell interior (see Chapter
7). The t-tubules, which are open to and communicate freely with the extracellular space, contain
extracellular fluid.
Intracellular Membrane Structures

Cardiac myocytes, like all eukaryotic cells, contain intracellular membrane-delimited organelles (see Fig.
1-15; Figs. 1-20 through 1-22). These include the nucleus, which contains the genetic material that
determines cell structure (Chapter 9), mitochondria, which catalyze
P.22
P.23
P.24
the oxidative reactions that generate most of the ATP used by the heart (Chapter 2), and the sarcoplasmic
reticulum, which plays a central role in excitation–contraction coupling and relaxation (Chapter 7).
Fig. 1-19: Schematic diagram of a sarcomere showing length-dependent changes in the overlap
between thick and thin filaments. A: At long sarcomere lengths in resting muscle, the myosin cross-
bridges are at right angles to the thick filament and the thin filaments are pulled away from the
center of the A-band. B: During contraction, the thin filaments, which are attached to thin
filaments, are drawn toward the center of the sarcomere by a shift in the orientation of the myosin
cross-bridges. C: As the sarcomere shortens further, the thin filaments of adjacent I-bands pass in
the center of the A-band (“double overlap”).
Fig. 1-20: Electron micrograph of rat ventricular muscle showing the sarcotubular network (SR) in a
“grazing” section overlying a sarcomere (center). The dark granules are glycogen. A faint linear
structure, composed of two parallel lines, that crosses the sarcotubular network over a Z line at
the lower right is probably a microtubule. Mito, mitochondria; A, A-band; I, I-band; Z, Z-line. Scale
= l µm. (Courtesy of Mrs. Judy Upshaw-Earley and Dr. Ernest Page.)
Fig. 1-21: Cross section of dyad in rat ventricular muscle. The transverse tubular system (t), seen
in cross section, is adjacent to the subsarcolemmal cisternae (sc). Electron-dense “feet” (arrows)
can be seen in the cytosol between the membranes of the t-tubule and subsarcolemmal cisterna.
Mito, mitochondria; A, A-band; I, I-band; Z, Z-line. Scale = 0.1 µm. (Courtesy of Mrs. Judy Upshaw-
Earley and Dr. Ernest Page.)
Fig. 1-22: Schematic diagram of a dyad showing sarcoplasmic reticulum calcium release channels (“ryanodine
receptors”) adjacent to plasma membrane calcium channels (“dihydropyridine receptors”) in the t-tubule.
The former, which form the “feet,” have a single opening into the cytosol and four openings into the lumen
of the subsarcolemmal cisterna.
Mitochondria originated as microorganisms that, hundreds of millions of years ago, crept into the cells of
our progenitors where, in return for a nutrient-filled environment, these symbiotic invaders provide
eukaryotic cells with a generous supply of ATP (Margulis, 1970). Mammalian mitochondria, which contain
circular DNA characteristic of prokaryotes, are surrounded by outer and inner membranes; infoldings of
the latter, called cristae, contain enzymes that participate in oxidative phosphorylation. Phase contrast
studies show that mitochondrial shape changes rapidly in living cardiac myocytes, enlarging and
contracting, branching, and fusing with one another. In hearts fixed under conditions that do not permit
oxidative phosphorylation (e.g., low oxygen tension), the cristae appear as stacks of flat membrane
sheets, whereas in hearts fixed when the mitochondria are carrying out oxidative phosphorylation, the
cristae are angulated in an “energized” configuration.
The sarcoplasmic reticulum, which takes up, stores, and releases the calcium that regulates contraction
and relaxation (Chapter 7), is a specialized form of the endoplasmic reticulum found in virtually every
mammalian cell. The endoplasmic reticulum generally includes a rough endoplasmic reticulum whose
outer surface is studded with ribosomes that carry out protein synthesis and a smooth endoplasmic
reticulum that participates in such processes as lipid metabolism and drug detoxification. In muscle, the
major function of these internal membranes, sometimes referred to as the sarcoendoplasmic reticulum
(SERCA), is to regulate cytosolic calcium concentration.
The cardiac sarcoplasmic reticulum consists of two regions (see Figs. 1-15 and 1-21). The sarcotubular
network, a network of tubules that surrounds the myofilaments, contains ATP-dependent calcium pump
molecules that relax the heart by pumping calcium out of the cytosol. Subsarcolemmal cisternae, which
contain channels that activate contraction by releasing calcium from the sarcoplasmic reticulum into the
cytosol in response to plasma membrane depolarization, are flattened structures that form composite
structures, called dyads, in which the sarcoplasmic reticulum and plasma membranes approach one
another but do not fuse. The narrow cytosolic space between these membranes contains huge electron-
dense proteins, often called feet because they resemble the feet of a caterpillar (Franzini-Armstrong and
Nunzi, 1983) (see Figs. 1-21 and 1-22). These proteins (called “ryanodine receptors” because they bind
with high affinity to this chemical) contain the calcium release channels whose opening initiates cardiac
contraction by allowing calcium to flow out of the sarcoplasmic reticulum into the cytosol (Chapter 7).
The sarcoplasmic reticulum calcium release channels differ from the L-type calcium channels (called
“dihydropyridine receptors”) found in the plasma membrane.
Cytoskeletal Proteins
Cells contain a network of proteins, called the cytoskeleton, that maintains cellular architecture, forms
mechanical linkages between cells and with the extracellular matrix, organizes enzymes that participate
in integrated catalytic cycles, maintains functionally important spatial relationships between membrane
pumps and channels that regulate key ion fluxes, and plays an important role in cell signaling (see
Chapter 5). The heart's cytoskeleton contains three types of filaments: microfilaments, microtubules, and
intermediate filaments. Microfilaments, which have an
P.25
actin backbone, include sarcomeric actin filaments, which are the thin filaments of the sarcomere (see
above), and cortical actin filaments. The latter form a network beneath the plasma membrane, link
various cell structures to one another and to the extracellular matrix, and, along with members of an
extended family of myosin molecules, transport membrane vesicles to and from the cell surface.
Microtubules, which have a tubulin backbone, transport cell organelles and participate in cell division and
the movements of cilia and flagella. Microtubular transport runs along polymers of tubulin, rather than
actin, and uses kinesins and dyneins, rather than myosin, as “motor proteins.” The third type of
cytoskeletal filament, the intermediate filaments, are desmin polymers that form strong rivet-like
structures, called desmosomes, which link cells to one another and attach cells to the extracellular
matrix. Unlike microfilaments and microtubules, intermediate filaments are not motile.
Intercalated Discs
Specialized cell-to-cell junctions, called intercalated discs (Fig. 1-23), form mechanical and electrical
connections between cardiac myocytes (Table 1-3) (Gallicano et al., 1998; Perriard et al., 2003). The
mechanical linkages are provided by the fascia adherens, in which sarcomeric actin filaments are
connected to networks of cytoskeletal actin filaments, and by desmosomes that connect intermediate
filaments in adjacent cardiac myocytes. A third structure, the gap junction, contains large nonselective
connexin channels that allow ions and other small molecules to diffuse freely between the cytosol of
adjacent cells. By providing low-resistance connections between cells, gap junction channels allow
electrical impulses to be conducted rapidly throughout the heart (see Chapter 13).
P.26
Fig. 1-23: Electron microphotographs of the intercalated disc. Top: Transverse section of cat
ventricular myocardium, showing insertions of sarcomeric actin microfilaments into the fascia
adherens of the intercalated disc (FA), which is made up of cortical actin microfilaments (AM). At
the right the intercalated disc continues as a nexus, or gap junction (N). Bottom: Oblique section
of intercalated disc in the mouse ventricle showing cortical actin myofilaments (AM), fascia
adherens (FA), a nexus (N), and two desmosomes or maculae adherens (MA). (Modified from McNutt
and Fawcett, 1974.)
Table 1-3 Cell-to-Cell Communication Across the Intercalated Disc
Type of Transmembrane
Structure Connection Proteins Cytoplasmic Proteins Cellular Structure
Fascia Mechanical N-cadherin β-Catenin Microfilament

adherens (actin, α-actinin)
β-1D Plakoglobin
integrin Vinculin
Desmosome Mechanical Desmoglein- Desmoplakin Intermediate
2 filament (desmin)
Desmocollin- Plakophilins
2 Plakogloblin
Gap Electrical Connexin 43 Ion channel

junction
Membrane Structure and Function

Biological membranes are made up of a phospholipid bilayer in which a lipid core is lined by two
hydrophilic surfaces (Fig. 1-24). The hydrophobic core provides a barrier to the passage of charged
molecules, while various charged “head groups” that line the hydrophilic surfaces interact with the
aqueous media on the two sides of the membrane. Proteins that are imbedded in the bilayer serve various
transport, signaling, and other functions.
Fig. 1-24: The membrane bilayer showing phospholipid molecules and cholesterol. The hydrophobic core,
which is made up of uncharged (apolar or hydrophobic) fatty acyl chains and cholesterol, is lined by charged
(polar or hydrophilic) “head groups.”
P.27
Fig. 1-25: Structure of a phospholipid, oriented with the surface of the bilayer at the top, showing the
glycerol “backbone” that is esterified to a head group and two fatty acyl chains. Left: Atomic structure, in
which the glycerol carbons are numbered 1, 2, and 3. The fatty acids in most phospholipids are esterified to
carbons 1 and 2 and the head group, which can be linked to various compounds (R), is esterified to carbon 3.
Right: Molecular model as shown in Figure 1-24.
Membrane Lipids
Membrane bilayers contain a mixture of lipids, most of which are amphipathic in that they are made up of
both hydrophilic (polar) and hydrophobic (apolar) moieties. Most membrane lipids are built upon glycerol,
a 3-carbon sugar that is generally esterified to a hydrophilic “head group” and one or two hydrophobic
fatty acyl chain “tails” (Fig. 1-25). Other membrane lipids include sphingolipids, which are built upon the
3-carbon amino acid serine, instead of glycerol. Most head groups contain charged anionic phosphate
compounds, and so are called phospholipids. Cholesterol, which is found in the plasma membrane (see
Fig. 1-24), reduces fluidity and “stiffens” the bilayer.
Virtually all of the fatty acids in membrane lipids contain an even number of carbon atoms; in mammalian
membranes these are mainly palmitic and stearic acids (saturated C16 and C18), and oleic, linoleic, and
linolenic acids (unsaturated C18 fatty acids that contain one, two, and three double bonds, respectively).
Saturated fatty acids form relatively ordered regions in membrane bilayers, whereas regions made up of
unsaturated fatty acids are more fluid (Klausner et al., 1980). Natural unsaturated membrane fatty acids
are cis-isomers; trans-fatty acids, which occur in artificially hydrogenated fats, are unnatural molecules
that can have adverse effects on membrane function (Mozaffarian and Willett, 2007).
Hydrolysis of membrane lipids by enzymes called phospholipases can contribute to membrane damage in a
number of diseases. Phospholipases A1 and A2 hydrolyze the ester bonds linking fatty acids to glycerol
carbons 1 and 2, respectively (Fig. 1-26). (Phospholipase B is a mixture of phospholipases A1 and A2 that
hydrolyzes both of these ester bonds.) Phospholipase C cleaves the phosphate head group from the
glycerol “backbone,” whereas phospholipase D removes organic structures from the head group leaving
the phosphatidic acid moiety attached to carbon 3 of glycerol.
Small amounts of membrane lipids released by phospholipases often serve as signaling molecules (see
Chapters 8 and 9). For example, two intracellular messengers, diacyl glycerol (DAG) and inositol
trisphosphate (InsP3), are released when phospholipase C hydrolyzes the membrane
P.28
phospholipid phosphatidylinositol 4,5-bisphosphate. Arachidonic acid, a polyunsaturated 20-carbon fatty
acid released from membrane phospholipids by phospholipase A2, is the precursor of several extracellular
messengers including prostaglandins, thromboxanes, and leukotrienes.
Fig. 1-26: Phospholipases A1 and A2 release the fatty acids esterified to glycerol at carbons 1 and
2, respectively, whereas phospholipases C and D release all or part of the head groups from glycerol
carbon 3. Diacylglycerol also serves as a second messenger.
Surface Charge and Transmembrane Potential
Anionic moieties in the head groups of membrane lipids give rise to a negative surface charge that
attracts cations in the aqueous media toward the membrane surface. The result is a gradual change in
surface potential as one moves away from the membrane (Fig. 1-27). The potential at
P.29
the plane of shear when the membrane moves through the surrounding aqueous medium is called the zeta
potential.
Fig. 1-27: Distribution of electrical potential at the surface of a membrane composed of phospholipids with
negatively charged head groups. Left: Surface charge falls sharply with increasing distance from the
membrane when ions are absent in the surrounding medium. Right: When salts are included in the medium
adjacent to the membrane, attraction of the cations to the anionic surface causes a more gradual fall in
surface charge. Some of these cations remain associated with the membrane when it is moved through to the
surrounding medium, giving rise to a “plane of shear” outside of which ions move freely. The potential at the
plane of shear is the zeta potential.
Biological membranes often separate regions of different electrical potential; cardiac Purkinje fibers, for
example, have a potential difference of about 90 mV across the resting plasma membrane. Changes in the
magnitude, and often the polarity, of this potential difference exert forces that modify the conformations
of intrinsic membrane proteins such as voltage-gated ion channels (Chapter 13). Although the absolute
potential differences across the plasma membrane are small, they create enormous electrical potential
gradients because they occur across a very thin surface. A resting potential of -90 mV (-90 × 103 V viewed
from within the cell) across the sarcolemma, which is approximately 30 Å (30 × 108 cm) thick, represents
a potential gradient of -300,000 V/cm (90 × 103 V ÷ 30 × 108 cm). During depolarization, this potential
difference reverses to +30 mV, so that the gradient becomes +100,000 V/cm. This means that the change
in transmembrane potential gradient is approximately 400,000 V/cm! These large changes in potential
gradient explain how what seem to be small changes in transmembrane potential generate forces that can
open and close ion channels.
Membrane Proteins
Most of the important activities of biological membranes are mediated by intrinsic membrane proteins
that are imbedded in one or both leaflets of the bilayer (Fig. 1-28). In the plasma membrane these
include receptors, ion channels, carriers, pumps, exchangers, and cell adhesion molecules that span the
phospholipid bilayer. Some cytosolic proteins mediate signal transduction after they are incorporated into
aggregates along the inner surface of the plasma membrane. Membrane proteins can make up more than
half of the weight of a membrane. Their extracellular portions often contain covalently bound lipid
(lipoproteins) or carbohydrate (glycoproteins).
The fluid nature of the lipid bilayer allows membrane proteins to move in the plane of the bilayer, much
as icebergs float in the sea. The lipids that surround the hydrophobic surfaces of membrane proteins,
sometimes called the boundary layer lipids or annulus, play an important role in regulating the activity of
these proteins (Katz and Messineo, 1981). Many cardioactive drugs are amphipathic molecules that can
reach binding sites on the hydrophobic surfaces of membrane proteins after they enter the hydrophobic
core and diffuse through the plane of the bilayer (Herbette and Mason, 1991).
Fig. 1-28: Membrane proteins (shaded) can span the bilayer (A), be incorporated into one leaflet of the
bilayer (B), or adsorbed to the membrane surface (C and D). A and B represent intrinsic membrane proteins.
C and D illustrate an aggregate, sometimes called a “scaffold,” formed when regulatory proteins become
organized along the inner surface of a membrane to form a signaling complex.
P.30
Membrane Transport
Transport of materials across membranes can be effected by two fundamentally different mechanisms.
The first utilize the pumps and exchangers described in Chapters 7 and 8 and the channels discussed in
Chapter 13 to move ions and other substances across membrane bilayers. A second, entirely different,
mechanism uses membrane-lined vesicles for the bulk movement of various substances. These transport
mechanisms begin when a membrane invaginates and then pinches off to form a vesicle that carries
materials through the cytosol. In exocytosis, intracellular membrane vesicles transport substances
manufactured within cells to the surface where the vesicles fuse with the plasma membrane, thereby
releasing the substances into the extracellular fluid by a process called trafficking. Bulk transport in the
opposite direction occurs by endocytosis in which molecules, often bound to a specific receptor, enter
cells within vesicles formed by invagination of the plasma membrane. These bulk transport processes
utilize cytoskeletal “molecular motors” that are powered by interactions between cortical actin filaments
and nonmuscle myosin, and tubulin with kinesins and dyneins (see above).
Endocytosis can be effected by several mechanisms, including pinocytosis, where vesicles formed from
plasma membrane invaginations enclose small amounts of extracellular fluid that is then transported into
the cell. Receptor-mediated endocytosis occurs when selected molecules in the extracellular fluid
(ligands) bind to specific receptors on the outer surface of the plasma membrane; the ligand-bound
receptors then stimulate the adjacent plasma membrane to invaginate. These invaginations, which are
called coated pits because their cytosolic surfaces are lined by proteins such as clathrin and caveolin,
form sealed coated vesicles that contain the receptor-bound ligands. These vesicles then fuse with other
intracellular vesicles, called endosomes, that can be transported within cells.
Bibliography
Anderson RH, Becker AE. The heart. Structure in health and disease. London: Gower Medical
Publishing, 1992.
Colicci WS, ed. Atlas of heart failure, 5th ed. New York: Springer, 2007.
Finean JB, Coleman R, Michell RH. Membranes and their cellular functions. Oxford: Blackwell, 1978.
Goldstein MA, Schroeter JP. Ultrastructure of the heart. In: Page E, Fozzard HA, Solara RJ, eds. The
cardiovascular system, Vol. I, The heart. Oxford: Oxford University Press, 2002:3–74.
Lodish H, Berk A, Zipursky SL, et al. Molecular cell biology, 4th ed. Basingstoke: Freeman, 1999.
Quinn PJ. The molecular biology of cellular membranes. Baltimore: University Park Press, 1976.
Robertson RN. The lively membranes. Cambridge: Cambridge University Press, 1983.
Sommer JR, Dolber PC. Cardiac muscle: ultrastructure of its cells and bundles. In: de Carvalho AP,
Hoffman BF, Lieberman M, eds. Normal and abnormal conduction in the heart. Mt Kisco, NY: Futura,
1982:1–28.
References
Anderson RH, Ho SY. The architecture of the sinus node, the atrioventricular conduction axis, and
the internodal atrial myocardium. J Cardiovasc Electrophysiol 1998;9:1233–1248.
Becker AE, deWit APM. Mitral valve apparatus. A spectrum of normality relevant to mitral valve
prolapse. Br Heart J 1979;42:680–689.
Benninghoff A. Lehrbuch der Anatomie des Menschen. Munich: JF Lehmanns, 1944.
P.31
Berne RE, Levy MN. Cardiovascular physiology. St. Louis: Mosby, 1967.
Bouvagnet P, Leger J, Dechesne C, et al. Fiber types and myosin types in human atrial and
ventricular myosin. An anatomical description. Circ Res 1984;55:794–804.
Brutsaert DL. The endocardium. Ann Rev Physiol 1989;51:263–273.
Cheng A, Nguyen TC, Malinowski M, et al. Heterogeneity of left ventricular wall thickening
mechanisms. Circulation 2008;118:713–721.
Dawes GS, Comroe JH Jr. Chemoreflexes from the heart and lungs. Physiol Rev 1954;34:167–201.
Factor SM, Okun EM, Kirk ES. The histological lateral border of acute canine myocardial infarction. A
function of microcirculation. Circ Res 1981;48:640–649.
Fenton TR, Cherry JM, Klassen GA. Transmural myocardial deformation in the canine left ventricular
wall. Am J Physiol 1978;235:H523–H530.
Franzini-Armstrong C, Nunzi G. Junctional feet and particles in the triads of a fast-twitch muscle
fibre. J Muscle Res Cell Motil 1983;4:233–252.
Gallicano GI, Kouklis P, Christoph C, et al. Desmoplakin is required early in development for
assembly of desmosomes and cytoskeletal linkage. J Cell Biol 1998;143:2009–2022.
Gerdes AM, Graves JH, Settles HE, et al. Improved preservation of myocardial ultrastructure in
perfusion-fixed human heart explants. In: Singal PK, Dixon IMC, Beamish RE, et al., eds Mechanisms
of heart failure. Boston: Kluwer, 1995:129–141.
Goldberger AL, Amaral LA, Hausdorff JM, et al. Fractal dynamics in physiology: alterations with
disease and aging. Proc Nat Acad Sci USA 2002;9[Suppl 1]:2466–2472.
Goldberger AL, Rigney DR, West BJ. Chaos and fractals in human physiology. Sci Am 1990;262:43–49.
Grant RP. Notes on the muscular architecture of the heart. Circulation 1965;32:301–308.
Hanson J, Huxley HE. Structural basis of the cross-striations in muscle. Nature 1953;172:530–532.
Harvey W. Movement of the heart and blood in animals. Franklin KJ, Trans. Oxford: Blackwell, 1957.
Hawthorne EW. Instantaneous dimensional changes of the left ventricle in dogs. Circ Res 1961;9:
110–119.
Hawthorne EW. Dynamic geometry of the left ventricle. Introduction. Fed Proc 1969;4:1323–1367.
Hayashi H, Lux RL, Wyatt RF, et al. Relation of canine atrial activation sequence to anatomical
landmarks. Am J Physiol 1982;242:H421–H428.
Henry CG, Lowry OH. Quantitative histochemistry of canine Purkinje fibers. Am J Physiol 1983;245:
H824–H829.
Herbette LG, Mason RP. Techniques for determining membrane and drug-membrane structures: a
reevaluation of the molecular and kinetic basis for the binding of lipid-soluble drugs to their
receptors in heart and brain. In: Fozzard H, Haber E, Katz A, et al. The heart and cardiovascular
system, 2nd ed. New York: Raven Press, 1991:417–462.
Huxley HE. X-ray analysis and the problem of muscle. Proc R Soc Lond B Biol Sci 1953;B141:59–62.
Katz AM. Congestive heart failure: role of altered myocardial cellular control. N Engl J Med
1975;293: 1184–1191.
Katz AM. The “modern” view of heart failure: how did we get here? Circ Heart Fail 2008;1:63–71.
Katz AM, Katz PB. Homogeneity out of heterogeneity. Circulation 1989;79:712–717.
Katz AM, Messineo FC. Lipid-membrane interactions and the pathogenesis of ischemic damage in the
myocardium. Circ Res 1981;48:1–16.
Klausner RD, Kleinfeld AM, Hoover RL, et al. Lipid domains in membranes. Evidence derived from
structural perturbations induced by free fatty acids and lifetime heterogeneity analysis. J Biol Chem
1980;255:1286–1295.
Lower R. Tractus de Corde. London: Allestry, 1669.
Margulis L. Origin of eukaryotic cells. New Haven: Yale University Press, 1970.
McNutt NS, Fawcett DW. Myocardial ultrastructure. In: Langer GA, Brady AJ, eds. The mammalian
myocardium. New York: Wiley, 1974:1–49.
Miller AJ. Lymphatics of the heart. New York: Raven, 1982.
P.32
Mozaffarian D, Willett WC. Trans fatty acids and cardiovascular risk: a unique cardiometabolic
imprint? Curr Atheroscler Rep 2007;9:486–493.
Oosthoek PW, Virágh S, Lamers WH, et al. Immunohistochemical delineation of the conduction
system. II: the atrioventricular node and Purkinje fibers. Circ Res 1993a;73:482–491.
Oosthoek PW, Virágh S, Mayen AEM, et al. Immunohistochemical delineation of the conduction
system. I: the sinoatrial node. Circ Res 1993b;73:473–481.
Page E. Quantitative ultrastructural analysis in cardiac membrane physiology. Am J Physiol 1978;

63:C147–C158.
Perriard JC, Hirschy A, Ehler E. Dilated cardiomyopathy: a disease of the intercalated disc? Trends
Cardiovasc Med 2003;13:30–38.
Rossi MA, Abreu MA, Santoro LB. Connective tissue skeleton of the human heart. A demonstration by
cell-maceration scanning electron microscope method. Circulation 1998;97:934–935.
Santamore WP, Dell'Italia LJ. Ventricular interdependence: significant left ventricular contributions
to right ventricular systolic function. Prog Cardiovasc Dis 1998;40:289–308.
Sartore S, Gorza L, Pierobon Bormioli S, et al. Myosin types and fiber types in cardiac muscle. I:
ventricular myocardium. J Cell Biol 1981;88:226–233.
Sengupta PP, Krishnamoorthy VK, Korinek J, et al. Left ventricular form and function revisited:
applied translational science to cardiovascular ultrasound imaging. J Am Soc Echocardiogr
2007;20:539–551.
Starling EH. Principles of human physiology. Philadelphia: Lea & Febiger, 1926.
Streeter DD, Spotnitz HM, Patel DP, et al. Fiber orientation in the canine left ventricle during systole
and diastole. Circ Res 1969;24:339–347.
Verheijck EE, Wessels A, van Ginneken ACG, et al. Distribution of atrial and nodal cells within the
rabbit sinoatrial node. Models of sinoatrial transmission. Circulation 1998;97:1623–1631.
Williams L, Frenneaux M. Diastolic ventricular interaction: from physiology to clinical practice. Nat
Clin Pract Cardiovasc Med 2006;3:368–376.
Yacoub MH. Two hearts that beat as one. Circulation 1995;92:160–161.

> Table of Contents > Part One - Structure, Biochemistry, and Biophysics > Chapter 2 - Energetics and Energy Production
Chapter 2
Energetics and Energy Production
L arge amounts of chemical energy are expended when the heart pumps blood under pressure into the aorta and pulmonary artery. Most of this
energy is derived from the oxidation of fats and carbohydrates so that the heart—which beats without pause—requires an uninterrupted delivery of
oxygen via the coronary circulation. Because of this high rate of energy expenditure, the heart's energy supply is “at the edge,” even under
normal conditions; this is why even a brief interruption of coronary flow severely impairs cardiac pump function.
The first clue as to how energy is produced by the heart emerged in the late 18th century when Lavoisier, Priestly, and others observed that
oxygen is essential for animal life. Berzelius, an early 19th-century Swedish chemist, identified a role for chemical reactions in muscular
contraction when he found that an acid present in sour milk (lactic acid) appeared in the muscles of a stag that had been exhausted during a long
hunt. In 1907, Fletcher and Hopkins discovered that this lactate disappears when a fatigued muscle is allowed to recover in the presence of
oxygen. Meyerhof, who in the 1920s detailed the enzymatic reactions in glucose metabolism (called glycolysis), showed that lactic acid production
is proportional to the amount of work done by a muscle contracting under anaerobic conditions, where lactate cannot be oxidized. This suggested
that energy released by glycolysis is directly coupled to muscle contraction:
This attractive hypothesis, although able to explain a large body of experimental data, collapsed in 1930 when Lundsgaard discovered that
muscles can contract when glycolysis is blocked with iodoacetic acid. Lundsgaard's subsequent observation that working muscles hydrolyze
phosphocreatine, a labile compound composed of creatine and phosphoric acid, and the demonstration by Eggleton and Eggleton that the
decrease in phosphocreatine is proportional to the amount of work performed during muscle contraction, led to a revised hypothesis, that energy
is provided for muscle contraction when phosphocreatine is broken down to creatine and inorganic phosphate (Pi):
These observations indicated that glycolysis, instead of delivering energy directly to the contractile machinery, supplies the energy needed to
form phosphocreatine from creatine and Pi.
After only a few years, however, the hypothesis that phosphocreatine hydrolysis is directly coupled to muscle contraction had to be abandoned
when adenosine triphosphate (ATP) was found to be essential for phosphocreatine breakdown. This led to the discovery that phosphorylation of
adenosine diphosphate (ADP), which forms ATP, is essential for the transfer of Pi and its chemical energy from phosphocreatine to energy-
consuming reactions. Recognition of the high-energy phosphate bond (∼P) demonstrated that the role of phosphocreatine in cellular energetics is
indirect and
P.34
depends on transfer of ∼P from phosphocreatine to ADP to form ATP and hydrolysis of ATP to yield ADP, Pi, and chemical energy:
and
For 30 years, ATP hydrolysis was believed to be coupled directly to muscle contraction, but because muscles contain creatine phosphokinase, an
enzyme that regenerates ATP by transferring ∼P from phosphocreatine to ADP (see below), ATP concentration could not be shown to decrease
during contraction. It was not until 1962 that discovery of a specific inhibitor of creatine phosphokinase made it possible to demonstrate that ATP
hydrolysis is coupled to energy release by working muscle:
The final chapter in this story was written when it was learned how myosin cross-bridges utilize energy released by ATP hydrolysis to effect muscle
contraction (see Chapters 4 and 6).
The role of adenine nucleotides in cellular energetics was likened by Albert Szent-Gyárgyi to that of money: ADP accepts chemical energy by
incorporating ∼P to form ATP, and ATP supplies chemical energy when its high-energy phosphate bond is hydrolyzed, Like money, ATP can be
obtained (regenerated) using energy derived from metabolism of a number of substrates, and ATP can be expended (hydrolyzed) to energize a
variety of energy-consuming processes including muscle contraction and relaxation, active transport, and biosynthesis. According to this monetary
analogy, the ∼P in phosphocreatine represents a cash reserve.
“Patterns” of Energy Production and Utilization by Different Muscles
Adaptation of the pathways of energy production and energy utilization to the functional needs of different muscle types can be understood by
examining the behavior of two long-eared European mammals, the rabbit and the hare (Fig. 2-1). The rabbits live in burrows from which they
venture only a short distance in search of food and adventure; to escape predators they rely on a rapid sprint to safety so that survival depends on
their ability to accelerate quickly and run rapidly for short distances. Although excellent sprinters, rabbits are poor distance runners, tiring
quickly if they cannot reach their burrow. Many species of hare, on the other hand, have no burrow but range widely in their habitat. These hares
are excellent distance runners, relying on their staying power to escape pursuers—indeed the coursing of hares has been known since antiquity.
Early in the 20th century, it became clear to scientists—although this information had long been known to hunters and cooks—that the back and
leg muscles of rabbits and hares differ in color; those in the rabbit are pale pink, almost white, whereas the back and leg muscles of the hare are
deep red. Similar relationships between muscle color and function are found in other vertebrate species: “white” muscle is generally found where
the need is for short bursts of intense activity, and “red” muscle where activity is sustained. In common culinary experience the chicken breast,
which powers wings that are used only briefly, is white meat, whereas the dark meat of the chicken leg is obtained from muscles that are used for
more continuous activity, such as walking
P.35
around a barnyard. In birds capable of sustained flight, unlike the chicken, the breast muscles are red (e.g., the dark meat of duck or goose
breast), whereas the muscles of a snake, which strikes, and a frog, which hops, are white.
Fig. 2-1: Functional specializations of the European rabbit and hare. When confronted with danger, such as a fox, the rabbit tries to escape
by sprinting to its burrow, whereas the hare tries to outrun, and outlast, the pursuing fox.
The following discussion focuses on red and white muscles, which represent the “extremes” of biochemical specialization (Table 2-1);
intermediate muscle types exist and there are exceptions
P.36
to the generalizations described below. For clarity in describing how biochemical specializations meet functional needs these exceptions are not
considered further.
Table 2-1 Biochemical Differences between Red and White Muscle
Biochemical Characteristic Red Muscle White Muscle

Pathways of energy production Aerobic Anaerobic
Substrates Lipid, carbohydrate Carbohydrate
Metabolites CO2, H2O Lactic acid
Glycolytic enzymes Sparse Abundant
Mitochondria Abundant Sparse
Phosphocreatine stores Minor Significant
Dependence on oxygen Marked Little
Intrinsic ATPase of contractile proteins Low High
Energy Production and Utilization By Red and White Muscle

Red muscle is specialized for sustained activity, during which the rate of energy utilization cannot exceed that of energy production. This requires
an uninterrupted supply of large amounts of ATP, which can be produced only by oxidative (aerobic) metabolism. The most important substrates
for aerobic energy production are lipids, and to a lesser extent carbohydrates. Although red muscles can carry out anaerobic glycolysis and have a
limited reserve of high-energy phosphate in the form of phosphocreatine, these are of little functional significance because neither can
regenerate the large supply of ATP needed to sustain intense muscular activity. The red color of these muscles is due mainly to myoglobin, a
heme-containing protein that facilitates oxygen diffusion through the muscle.
Red muscles pay a threefold price for their dependence on a high rate of ATP production. First, these muscles depend on an uninterrupted supply
of oxygen to generate the ATP consumed during sustained activity. If blood flow is interrupted, lack of oxygen quickly brings oxidative metabolism
to a stop and exhausts the phosphocreatine stores. Although anaerobic glycolysis can provide a limited supply of ATP, it is not sufficient for
sustained activity. Second, to carry out oxidative metabolism, red muscle myocytes must contain a large volume of mitochondria, which, by
occupying space that could otherwise contain contractile proteins, weakens the muscle. The third price, which helps match energy utilization and
energy production, is that the contractile proteins have a low intrinsic rate of ATP hydrolysis (ATPase activity); this slows intrinsic shortening
velocity and weakens these muscles (see Chapter 6).
In white muscles, a limited supply of ATP can be regenerated from phosphocreatine and by anaerobic glycolysis, but these muscles do not depend
on oxidative metabolism during their brief periods of intense activity. Because the rate of energy expenditure exceeds that of energy production,
white muscles require periods of rest to allow the lactate produced by glycolysis to be oxidized to CO2 and water. This represents an “oxygen
debt” that is repaid by the muscle when lactate is oxidized during periods of inactivity, and by the liver, after the lactate has entered the
circulation. White muscles are not required to balance the rates of energy production and energy utilization during their brief bursts of activity so
that their contractile proteins can have a high ATPase activity, which allows these muscles to achieve a high shortening velocity (see Chapter 6).
Muscle strength is also increased because cell volume otherwise occupied by mitochondria contains contractile proteins.
The physiological consequences of these biochemical specializations allow the rabbit to escape pursuit by a rapid dash to its burrow (the
“jackrabbit start”); should the rabbit fail to reach safety, it soon tires because phosphocreatine stores are quickly depleted, anaerobic glycolysis
ceases, and lactic acid accumulation causes its muscles to become acidotic. Once in its burrow, the rabbit requires a period of rest to replenish its
phosphocreatine reserves and to repay the “oxygen debt”—if the rabbit shown in Figure 2-1 is caught, repaying this debt becomes the fox's
problem. The hare, which relies on its staying power to elude pursuers, can run long distances because, even during intense activity, its red
muscles regenerate ATP at the same rate at which it is being consumed. These muscles, however, require the continuous delivery of oxygen and
substrate via the bloodstream. Because of the lower intrinsic ATPase of their contractile proteins and the larger volume of muscle occupied by
mitochondria, the running muscles of the hare are slower and weaker than those of the rabbit. These differences were summarized by W.F.H.M.
P.37
Mommaerts, who said that white muscle operates on a “twitch now, pay later” basis, whereas the modus operandi of red muscle is “pay as you
go.”
In humans, most skeletal muscles contain several fiber types. These include slow oxidative fibers, which are similar to the red muscles described
above, fast glycolytic fibers, which are like white muscle (although the muscles are pink in color), and fast oxidative-glycolytic fibers, which have
high ATPase contractile proteins but contain numerous mitochondria and so can produce ATP by oxidative reactions (Pette and Staron, 1990). The
heart, of course, functions like a red muscle.
Muscular Efficiency
The efficiency of muscular contraction can be viewed in several ways (Backx, 1993). Thermodynamic efficiency, the ratio between the mechanical
work performed during contraction and the free energy made available by substrate metabolism, is precise but impractical because the free
energy changes in most of the chemical reactions in muscle are not precisely known. It is therefore more useful to estimate mechanical
efficiency, the ratio between useful work and the enthalpy changes during substrate metabolism.
An interplay between the load and the intrinsic ATPase activity of the contractile proteins influences the mechanical efficiency of a muscle (see
Chapter 3). When contracting against a light load, fast muscles with high ATPase myosins are more efficient than slow muscles, whose contractile
proteins hydrolyze ATP at a slower intrinsic rate. However, when lifting a heavy load, slow muscles develop tension at less energy cost than do fast
muscles (Awan and Goldspink, 1972). For this reason, muscles with high ATPase contractile proteins are more efficient when shortening rapidly at
low tension, whereas muscles with low ATPase contractile proteins are more efficient when developing high levels of tension. Human ventricles,
which contain mostly a low ATPase myosin, contract relatively efficiently when wall stress is high.
Enthalpy changes in the working heart are readily calculated by measuring cardiac oxygen consumption because the myocardium regenerates ATP
almost entirely by oxidative metabolism. Even though the energy liberated by the oxidation of fat (∼9 kcal/g) is more than twice that of either
carbohydrate or protein (∼4 kcal/g) (also see Table 2-5), more oxygen is needed to metabolize fat. As a result, the enthalpies of the oxidation of
all three substrates, when calculated per liter of oxygen consumed, are similar: fat 4.69 kcal, carbohydrate 5.05 kcal, and protein 4.60 kcal.
Cardiac efficiency can therefore be estimated as the ratio between the work performed and the energy equivalent of the oxygen consumed (see
Chapter 12).
Cardiac efficiencies calculated from measurements of external work and oxygen consumption are usually less than 20% to 25%. The efficiency of
the contractile process itself is higher because some of the energy consumed by the heart is used for processes other than contraction (e.g.,
active ion fluxes). A major cause of the heart's inefficiency is heat liberation during relaxation, when energy that had been expended to stretch
elasticities in the walls of the contracting ventricle is dissipated as heat (see Chapter 12). Efficiencies as high as 40% can be calculated during
ejection (Suga et al., 1993); these compare favorably with efficiencies of approximately 30% for man-made machines, such as gasoline engines.
Overview of Energy Production by the Heart

The heart can be viewed as an omnivore because it is able to regenerate ATP by metabolizing fats, carbohydrates, and proteins. Most of the
energy utilized by the heart is derived from the oxidation of fats; carbohydrate metabolism makes an important contribution, whereas amino
P.38
acid metabolism normally contributes little to cardiac energy production. Unlike fat, which can be metabolized only in the well-oxygenated heart,
carbohydrates are metabolized by glycolysis under both aerobic and anaerobic conditions. Anaerobic glycolysis, however, has only a limited ability
to regenerate ATP and cannot meet the energy needs of the beating heart, which explains why interruption of oxygen supply to mammalian hearts
brings effective contraction to a halt within less than a minute (see Chapter 17). Aerobic glycolysis also generates only a fraction of the energy
used by normal hearts but plays a key role in supplying substrate for oxidative metabolism.
Glycogen Formation and Breakdown

Carbohydrate is stored in the heart as glycogen, a polysaccharide made up of glucose 1-phosphate. Glycogen formation and breakdown do not
occur by a single reversible reaction, but instead utilize two separate pathways that are catalyzed by different enzymes (Fig. 2-2). The key
enzymes in glycogen formation (glycogen synthase) and glycogen breakdown (phosphorylase) exist in active and inactive forms whose
interconversions are regulated by sympathetic stimulation, high-energy phosphate levels, and metabolic intermediates (Greenberg et al., 2006).
Glycogen Formation: Glycogen Synthase

Glycogen synthesis begins when phosphoglucomutase catalyzes the isomerization of glucose 6-phosphate, which forms glucose 1-phosphate. The
latter is polymerized to form glycogen in a two-step reaction (Fig. 2-2, left). The first, which is catalyzed by glucose 1-phosphate
uridylyltransferase, uses energy in uridine triphosphate to form uridine diphosphoglucose. The
P.39
latter is added to the glycogen polymer in the second step, which releases uridine diphosphate. The second step, which is rate-limiting, is
catalyzed by glycogen synthase whose activity is regulated by covalent modification (phosphorylation and dephosphorylation) and allosteric
responses to high-energy phosphate and glucose 6-phosphate levels. A third type of control, translocation to structures linked to the actin
cytoskeleton, also regulates glycogen synthase (Prats et al., 2005, 2009).
Fig. 2-2: Pathways of glycogen formation (left, reading upward) differ from those of glycogen breakdown (right, reading downward).
Glycogen synthesis involves two steps. The first transfers uridine from uridine triphosphate (UTP) to glucose 1-phosphate. The second step
adds uridine diphosphoglucose (UDPG) to glycogen and releases uridine diphosphate (UDP). Glycogen breakdown releases glucose 1-
phosphate that, after isomerization to glucose 6-phosphate, can enter the glycolytic pathway. Glycogen synthesis is regulated by glycogen
synthase, whereas phosphorylase regulates glycogenolysis.
Fig. 2-3: Phosphorylation and dephosphorylation reactions that control glycogen synthase activity. A: Phosphorylation by cyclic adenosine
monophosphate (AMP)-dependent protein kinase converts the active, dephosphorylated glycogen synthase a to the less active,
phosphorylated glycogen synthase b. B: Dephosphorylation by synthase phosphatase converts phosphorylated glycogen synthase b to the
more active, glycogen synthase a. ADP, adenosine diphosphate; ATP, adenosine triphosphate.
Glycogen synthase exists in two forms whose interconversions are controlled by phosphorylation and dephosphorylation reactions (Fig. 2-3).
Glycogen synthesis is slowed when sympathetic stimulation activates a cyclic adenosine monophosphate (AMP)-dependent protein kinase (protein
kinase A or PK-A) that converts the more active a (dephospho-) form of glycogen synthase to the less active b (phospho-) form (Fig. 2-3A).
Phosphorylation of glycogen synthase a does not directly inhibit the enzyme, but instead increases the stimulatory effects of several substrates
and metabolites. Most important are those of ATP and glucose 6-phosphate which, by increasing the catalytic activity of glycogen synthase b,
promote glycogen storage when an abundant supply of energy maintains high levels of these activators. Glycogen synthesis returns to its high
basal rate when glycogen synthase b is dephosphorylated by synthase phosphatase to form the more active glycogen synthase a (Fig. 2-3B).
Glycogen Breakdown: Phosphorylase
Glycogen breakdown, which releases glucose 1-phosphate (Fig. 2-2, right), is catalyzed by phosphorylase, which, like glycogen synthase, is
regulated by phosphorylation and dephosphorylation (Fig. 2-4). In contrast to glycogen synthase, in which the phosphorylated enzyme is less
active, phosphorylated phosphorylase a is the active enzyme. The less active phosphorylase b is inhibited by ATP and glucose 6-phosphate, and
activated by inorganic phosphate (Pi). Because these inhibitory effects are incomplete, phosphorylase b remains able to hydrolyze glycogen in
energy-starved hearts, where ATP and glucose 6-phosphate levels fall and Pi levels rise.
Phosphorylation of phosphorylase b is catalyzed by phosphorylase kinase, which increases glycogen breakdown when cyclic AMP levels are
increased in response to sympathetic stimulation.
P.40
However, cyclic AMP does not interact directly with phosphorylase kinase, but instead activates a cyclic AMP-dependent protein kinase, called
phosphorylase kinase kinase, that phosphorylates phosphorylase b kinase (Fig. 2-5). Increased levels of cytosolic calcium also mobilize glycogen, in
this case by activating a calcium/calmodulin-dependent protein kinase that, like phosphorylase kinase kinase, phosphorylates and activates
phosphorylase kinase. Glycogen breakdown slows to its basal level when phosphorylase kinase is dephosphorylated and inhibited by phosphorylase
kinase phosphatase, which allows phosphorylase a to be dephosphorylated and converted to the less active phosphorylase b by phosphorylase
phosphatase.
Fig. 2-4: Phosphorylation and dephosphorylation reactions that control phosphorylase activity. A: Phosphorylation by phosphorylase b kinase
converts the less active, dephosphorylated phosphorylase b to the more active phosphorylase a. B: Dephosphorylation by phosphorylase
phosphatase converts phosphorylase a to the less active, phosphorylase b. ADP, adenosine diphosphate; ATP, adenosine triphosphate.
The reactions described above provide an integrated control mechanism that allows cyclic AMP to regulate the flux of glucose 1-phosphate into
and out of glycogen stores (Fig. 2-6). Cyclic AMP-stimulated phosphorylations increase glucose supply by inhibiting glycogen synthase and
P.41
stimulating phosphorylase, while dephosphorylation of these enzymes shifts the balance toward glycogen synthesis. In this way, sympathetic
stimulation, which increases cardiac energy utilization (see Chapter 8), increases energy production by favoring glycogen breakdown.
Fig. 2-5: Phosphorylation and dephosphorylation reactions that control the activity of phosphorylase b kinase. A: Phosphorylation is
catalyzed by phosphorylase kinase kinase, a cyclic AMP-dependent protein kinase that converts the inactive, dephosphorylated form of
phosphorylase kinase to the active form. B: Dephosphorylation of phosphorylated phosphorylase kinase by phosphorylase kinase
phosphatase forms the inactive enzyme. ADP, adenosine diphosphate; AMP, adenosine monophosphate; ATP, adenosine triphosphate.
Fig. 2-6: Integrated control of glycogen formation and breakdown by phosphorylation and dephosphorylation of glycogen synthase and
phosphorylase. A: Glycogen synthesis is inhibited and glycogen breakdown is accelerated after phosphorylation has converted glycogen
synthase a to the less active b form and phosphorylase b to the more active a form. B: Glycogen synthesis is accelerated and glycogen
breakdown is inhibited when both enzymes are dephosphorylated, which converts glycogen synthase to the more active a form and
phosphorylase to the less active b form. UDP, uridine diphosphate; UTP, uridine triphosphate.
Debranching Enzymes
Glycogen is a highly branched polysaccharide that requires debranching enzymes to release glucose residues at the branch points. Although these
reactions are not normally rate-limiting, molecular abnormalities in the debranching enzymes can cause glycogen storage diseases by preventing
the complete breakdown of glycogen.
Glycolysis
Glycolysis describes the process by which glucose, a 6-carbon sugar, is broken down to form 2 moles of pyruvate, a 3-carbon product (Fig. 2-7).
Utilization of a series of steps, rather than a single reaction, allows the energy released during glycolysis to be captured as high-energy
P.42
phosphate bonds of ATP and the reactive electrons that energize oxidative phosphorylation (see below). In the initial steps of glycolysis, 2 moles
of ATP/mole of glucose are utilized to energize the carbohydrate, whereas generation of 4 moles of ATP/mole of glucose toward the end of the
glycolytic pathway yields a net of 2 moles of ATP/mole of glucose. Glycolysis also reduces coenzymes that, when oxidized in the mitochondria,
regenerate additional ATP (see below). In aerobic (well-oxygenated) hearts, the end-product of glycolysis is pyruvate that is oxidized and
decarboxylated to form acetyl-CoA. Under anaerobic conditions, on the other hand, pyruvate is reduced to form lactate.
Fig. 2-7: Overall reaction of glycolysis in which glucose is metabolized to form pyruvate. ADP, adenosine diphosphate; ATP, adenosine
triphosphate; NAD+, oxidized nicotinamide adenine dinucleotide; NADH, reduced nicotinamide adenine dinucleotide.
Enzymes, Coenzymes, and the Cytoskeleton

Glycolytic enzymes were once viewed as proteins and protein complexes that diffuse freely through the cytosol. It is now clear, however, that
organization of glycolytic enzymes by the cytoskeleton allows substrates to be delivered to appropriate enzymes and ATP to be released near
energy-utilizing structures. This structural organization explains why ATP regenerated by glycolytic pathways (“glycolytic ATP”) is used
preferentially in some energy-consuming reactions.
Coenzymes, which are much smaller than enzymes, play an essential role in glycolysis and other metabolic processes. Because coenzymes often
contain moieties that cannot be synthesized by mammalian cells, they must be included in the diet as vitamins. Several oxidized coenzymes
participate in redox reactions by accepting electrons released in a variety of reactions; this generates reduced coenzymes that can be oxidized in
the mitochondria to provide energy for ATP regeneration (see below). Coenzymes that participate in redox reactions include nicotinamide adenine
dinucleotide (NAD), which contains niacin; coenzyme Q (ubiquinone), an electron carrier with a quinone group; and flavine mononucleotide (FMN)
and flavine adenine dinucleotide (FAD), which contain riboflavin. Nicotinamide adenine dinucleotide phosphate (NADP), which is similar to NAD
except that it contains an additional phosphate, generally participates in biosynthetic rather than energy-producing reactions.
Additional coenzymes include thiamine pyrophosphate, which serves as a cofactor for decarboxylations, and lipoic acid, which participates in
transacetylations. Thiamine deficiency causes beriberi, which can be accompanied by high output heart failure. Coenzyme A (CoA), which
contains a vitamin called pantothenic acid, has a reactive sulfhydryl group (hence the abbreviation CoA-SH) that activates acetate, fatty acids,
and other substrates by forming a high-energy thioester bond analogous to the high-energy phosphate bond of ATP.
Overview of Glucose Metabolism

Glycolysis is regulated by four different signaling mechanisms (Table 2-2). Three are functional signals that modify the catalytic activity of
existing structures, whereas the fourth, proliferative or transcriptional signaling, alters gene expression. Many steps in glucose metabolism are
regulated by more than one type of signal.
Functional signals regulate glucose metabolism at the six numbered steps marked by asterisks in Figure 2-8. Humoral control by sympathetic
stimulation increases glycolytic rate by accelerating fructose 1,6-bisphosphate formation from fructose 6-phosphate, which is the rate-limiting
step for glycolysis in the normal heart. A second type of functional signal also operates at this step to match the rates of energy production and
energy utilization by allowing high ATP concentrations to inhibit glycolysis, and ∼P depletion, which increases ADP and AMP and
P.43
Pi levels, to stimulate glycolysis. The third type of functional signal responds to changes in redox potential, which defines the tendency of
electrons to flow between members of redox pairs like NADH and NAD+. This mechanism, which responds to the balance between oxidized and
reduced coenzymes, regulates glyceraldehyde 3-phosphate oxidation and conversion of pyruvate to lactate, and slows glycolysis when NADH
accumulates and NAD+ becomes depleted in energy-starved hearts.
Table 2-2 Signaling Mechanisms That Regulate Energy Production in the Heart
Functional
Humoral: hormones and neurotransmitters
Energy requirements: levels of ATP, ADP, etc.
Coenzymes: levels of oxidized and reduced NAD, FAD, etc.
Proliferative
Gene expression: transcription factors etc.
ATP, adenosine triphosphate; ADP, adenosine diphosphate; NAD, nicotinamide adenine dinucleotide; FAD, flavine adenine
dinucleotide.
The fourth type of control, long-term regulation by proliferative signaling, differs fundamentally from the three functional mechanisms because,
instead of modifying the activity of preexisting enzymes, transporters, and so forth, it alters their synthesis. Proliferative signaling, which
increases glycolytic capacity in hypertrophied hearts, plays an important role in heart failure (Chapter 18).
An analogy to a factory is useful in understanding how these four types of mechanism regulate glycolysis. Production (glycolytic rate) is
determined by the incentive to produce (neurotransmitters and hormones), delivery of raw materials (supply of substrates), stores (ATP level),
debt (ADP and AMP levels), supply of parts (oxidized coenzymes), and the availability and types of tools used in production (content and isoform
characteristics of glycolytic enzymes).
Glucose Transport
Glucose enters myocardial cells (*1, Fig. 2-8) by moving down a concentration gradient from the extracellular space into the cytosol and so does
not require the expenditure of energy. However, glucose uptake is mediated by GLUT4, a transporter that moves between intracellular vesicles,
called recycling endosomes, and the plasma membrane. Under basal conditions, most of the GLUT4 is sequestered in the endosomes where it is
inactive. Insulin, ischemia, and hypoxia accelerate glucose uptake by recruiting the GLUT4-containing endosomes to the plasma membrane where
the transporter is able to bind extracellular glucose for transport into the cell.
Hexokinase
Before glucose can be metabolized, it is phosphorylated by hexokinase in a reaction that “invests” the first of 2 moles of high-energy phosphate
per mole of glucose (*2, Fig. 2-8). Hexokinase is regulated by allosteric effects of glucose 6-phosphate, ATP, ADP, AMP, and Pi. Most
P.44
P.45
important is the inhibitory effect of the product, glucose 6-phosphate, which slows glycolysis when the heart has an abundant supply of energy.
ATP potentiates this inhibitory response, whereas ADP, AMP, and Pi—whose concentrations increase when the heart is energy-starved—accelerate
glycolysis by reducing the inhibitory effect of glucose 6-phosphate. Together, these responses help match the rates of energy production and
energy utilization by accelerating the hexokinase reaction when increased cardiac work lowers glucose 6-phosphate and ATP concentrations and
increases ADP, AMP, and Pi levels.
Fig. 2-8: Major control points in glucose metabolism: translocation of GLUT-4 to the plasma membrane, which facilitates glucose transport
into the cell (*1); hexokinase, which catalyzes glucose phosphorylation (*2); phosphofructokinase, which catalyzes fructose 6-phosphate
phosphorylation (*3); glyceraldehyde 3-phosphate dehydrogenase, which catalyzes glyceraldehyde 3-phosphate oxidation (*4); pyruvate
dehydrogenase, which catalyzes acetyl-CoA formation (*5); and lactate dehydrogenase, which catalyzes pyruvate reduction (*6). ADP,
adenosine diphosphate; ATP, adenosine triphosphate; NAD+, oxidized nicotinamide adenine dinucleotide; NADH, reduced nicotinamide
adenine dinucleotide.
6-Phosphofructo-1-Kinase
Following an isomerization reaction catalyzed by phosphoglucomutase, which converts glucose 6-phosphate to fructose 6-phosphate, a second
mole of ∼P is invested to phosphorylate fructose 6-phosphate, which forms fructose 1,6-bisphosphate (*3, Fig. 2-8). The latter reaction, which is
the rate-limiting step of aerobic glycolysis, is catalyzed by 6-phosphofructo-1-kinase (phosphofructokinase-1 or PFK). This highly regulated
enzyme complex, like hexokinase, is inhibited by ATP and stimulated by the products of ATP hydrolysis. PFK is especially sensitive to stimulation
by AMP, which helps increase energy production in response to increased energy utilization by allowing a small increase in AMP to accelerate
glycolysis. PFK activity is also increased by fructose 2,6-bisphosphate, which is produced from fructose 6-phosphate in a “side-reaction” that is
catalyzed by 6-phosphofructo-2-kinase (phosphofructokinase-2); the latter, a cyclic AMP-dependent protein kinase, accelerates glycolysis when
cardiac energy utilization is increased by sympathetic stimulation. Calcium, which like sympathetic stimulation accelerates cardiac energy
utilization, also stimulates PFK. Acidosis, a powerful inhibitor of PFK, slows glycolysis when conversion of pyruvate to lactate releases protons;
this is one reason why anaerobic glycolysis, although initially accelerated in the anaerobic heart, quickly slows after coronary artery occlusion (see
Chapter 17).
In 1861, Louis Pasteur found that ethanol production by fermenting grapes is inhibited by exposure to air. This inhibition of glycolysis by aerobic
metabolism, called the Pasteur effect, is due largely to a decrease in PFK activity that occurs when glucose oxidation increases ATP concentration
and reduces ADP, AMP, and Pi levels.
Allosteric control of PFK utilizes a mechanism, called amplification, that allows small changes in ATP, ADP, and especially AMP concentrations to
exert large effects on glycolytic rate (see below). Because the free energy available from ATP hydrolysis is proportional to the ATP/ADP ratio,
minimizing a fall in this ratio is important for the heart, where as little as a 15% to 25% reduction in the ATP/ADP ratio can reduce the free energy
available from ATP hydrolysis to levels that impair vital energy-dependent reactions (Kammermeier et al., 1982; Tian and Ingwall, 1996).
Glyceraldehyde 3-Phosphate Dehydrogenase

The next control point in glycolysis occurs after fructose 1,6-bisphosphate is hydrolyzed to form 2 moles of triose. This step, where glyceraldehyde
3-phosphate is oxidized and phosphorylated to form 1,3-bisphosphoglycerate, is catalyzed by glyceraldehyde 3-phosphate dehydrogenase (GAPDH)
(*4, Fig. 2-8). The GAPDH reaction does not determine glycolytic rate in the normal heart, where PFK activity is rate-limiting, but 1,3-
bisphosphoglycerate formation can become rate-limiting during hypoxia or ischemia, or when the heart is performing high levels
P.46
of work. Under conditions in which PFK is activated by lowered ATP concentration and increased ADP, AMP, and Pi levels (see above), control of
glycolysis shifts “downstream” to the GAPDH reaction. In energy-starved hearts, this reaction is eventually slowed by depletion of NAD+,
accumulation of NADH and 1,3-bisphosphoglycerate, and acidosis.
Phosphoglycerate Kinase and Pyruvate Kinase

There are two steps in glycolysis where ATP is regenerated by substrate-level phosphorylation. In the first, the phosphate used to form 1,3-
bisphosphoglycerate is energized and transferred to ADP by phosphoglycerate kinase. The second substrate-level phosphorylation is catalyzed by
pyruvate kinase, which forms pyruvate and ATP from phosphoenolpyruvate and ADP (Fig. 2-8). Although pyruvate kinase regulates glycolysis in
some tissues, it does not play an important regulatory role in the heart.
Pyruvate Metabolism
Pyruvate stands at a metabolic “crossroads” between aerobic and anaerobic energy production because it can be oxidized to form acetyl-CoA or
reduced to form lactate. The road taken is determined in part by the activities of two enzymes: pyruvate dehydrogenase (PDH) and lactate
dehydrogenase (LDH). In well-oxygenated hearts, PDH catalyzes the conversion of pyruvate to acetyl coenzyme A (acetyl-CoA) (*5, Fig. 2-8), which
can be oxidized in the citric acid cycle (see below). In anaerobic hearts, on the other hand, pyruvate is reduced to lactate by LDH, which
generates a limited amount of NAD+ (*6, Fig. 2-8).
Lactate Dehydrogenase
LDH functions as a tetramer made up of two isoforms called M and H; all combinations (H4, H3M, H2M2, H1M3, and M4) are found in muscle, but the
M isoform occurs mainly in skeletal muscle, whereas H is most prevalent in the heart. The metabolic fate of pyruvate is determined in part by the
affinity of LDH for pyruvate, which is higher for the M than the H subunits. In skeletal muscles where energy production depends mainly on
anaerobic glycolysis, the M subunits of LDH bind to pyruvate which is then reduced to lactate (*6, Fig. 2-8), whereas in well-oxygenated hearts
that have an abundant supply of NAD+, the lower pyruvate affinity of the H subunits of LDH allows pyruvate to be oxidized and decarboxylated by
PDH (see below).
Pyruvate Dehydrogenase
PDH, a huge enzyme complex, catalyzes an irreversible series of reactions that oxidize and decarboxylate pyruvate to form acetyl-CoA, the major
substrate for oxidative energy production (*5, Fig. 2-8). The reactions catalyzed by PDH, which require NAD+, FAD, CoA, thiamine, and lipoic acid
(Fig. 2-9), are regulated by the concentrations of substrate (pyruvate), product (acetyl-CoA), and cofactors (CoA-SH, lipoic acid, phosphorylated
thiamine, NADH, and NAD+).
PDH is regulated by phosphorylation and dephosphorylation of the enzyme complex. Activity is decreased when PDH is phosphorylated by pyruvate
dehydrogenase kinase and increased after it is dephosphorylated by pyruvate dehydrogenase phosphatase (Fig. 2-10). Lactate production is
increased in energy-starved hearts when the activity of PDH kinase, which catalyzes the
P.47
inhibitory phosphorylation, is increased by the products of the PDH reaction (NADH and acetyl-CoA) and reduced by decreases in the levels of
pyruvate, CoA, and NAD+; these effects reduce PDH activity when its substrates are exhausted and the reaction products accumulate.
Phosphorylation of PDH is also stimulated by NADH, which, as noted above, favors lactate production in energy-starved hearts by inhibiting
pyruvate oxidation.
Fig. 2-9: Overall reaction by which pyruvate is oxidized and decarboxylated to form acetyl-CoA. NAD+, oxidized nicotinamide adenine
dinucleotide; NADH, reduced nicotinamide adenine dinucleotide.
The inhibitory phosphorylation of PDH is reversed by pyruvate dehydrogenase phosphatase. The latter is inhibited by NADH, which, by slowing
pyruvate production, favors lactate production in energy-starved hearts. Pyruvate dehydrogenase phosphatase is also inhibited by citrate, which
slows acetyl-CoA production from glucose; this is one reason why the high citrate levels in the well-oxygenated heart favor fatty acid oxidation
rather than glucose oxidation. This effect of citrate contributes to the normal “preference” of well-oxygenated hearts for lipids rather than
carbohydrates. Sympathetic stimulation and calcium, both of which increase myo-cardial contractility, also accelerate acetyl-CoA formation by
activating pyruvate dehydrogen-ase phosphatase.
Other Roles of Pyruvate

Pyruvate can be carboxylated to form oxaloacetate and malate, both which can be oxidized in the citric acid cycle (see below). The heart can
also generate a limited amount of energy from protein metabolism after transamination of the amino acid alanine forms pyruvate.
Fig. 2-10: Phosphorylation and dephosphorylation reactions that control pyruvate dehydrogenase (PDH) activity. A: Phosphorylation by PDH
kinase converts active, dephosphorylated PDH to the inactive phosphorylated enzyme. B: Dephosphorylation by pyruvate dehydrogenase
phosphatase converts the inactive PDH to the active dephosphorylated form. ADP, adenosine diphosphate; ATP, adenosine triphosphate.
P.48
Fatty Acids
Fats can be viewed as “concentrated energy” whose oxidation yields 9 calories/g, compared with only 5 calories/g for carbohydrates and proteins.
Dietary fatty acids, after being absorbed into the bloodstream, are transported to the heart as glycerol esters (triacylglycerols or triglycerides)
and as free fatty acids (FFAs) (Fig. 2-11). The latter is a misnomer because FFAs bind to plasma proteins, mainly albumin; FFAs, therefore, are
“free” only because they are not bound as esters.
Hydrolysis of Triacylglycerol
Triacylglycerols cannot enter cells until the ester bonds linking the fatty acid to glycerol are hydrolyzed by lipoprotein lipase, an extracellular
enzyme that is located on the luminal surface of the capillary endothelium. Lipoprotein lipase is activated by β-adrenergic agonists so that, like
glycolysis and glucose uptake, fatty acid release from triacylglycerols is under humoral control.
Fatty Acid Uptake

Transfer of fatty acids across the capillary endothelium and cardiac plasma membrane is facilitated by fatty acid transport proteins and, after
entering the cytosol, fatty acids are transported in combination with fatty acid translocases. Fatty acid uptake by the myocardium is passive, and
so is determined by the law of mass action; this allows fatty acid uptake to be accelerated by high plasma FFA and low fatty acid levels in the
cytoplasm. The latter allows increased fatty acid utilization to accelerate fatty acid uptake by reducing the intracellular concentration of this key
substrate.
Fatty Acid Activation

Fatty acid metabolism, like glucose metabolism (see above), requires the “investment” of energy to form fatty acyl-CoA. This occurs by a two-
step reaction that begins when acyl-CoA synthase, located on the mitochondrial outer membrane, transfers AMP derived from ATP to the fatty
P.49
acid, which forms fatty acid∼AMP, a high-energy complex (Fig. 2-12). The pyrophosphate (PPi) released in this reaction is hydrolyzed by a
pyrophosphatase to form inorganic phosphate. In the second step, the adenylate moiety of fatty acid∼AMP is replaced by CoA to form fatty acyl-
CoA, which contains a high-energy thioester bond.
Fig. 2-11: Initial steps in fatty acid metabolism by myocardial cells showing key reactions (capital letters) and enzymes or carriers (italics).
Free fatty acids and fatty acids released by hydrolysis of plasma triacylglycerols (1) enter cardiac myocytes in combination with fatty acid
transport proteins and are transported through the cytosol by fatty acid translocases (2). The fatty acids are then activated by acyl-
coenzyme A (CoA) synthase to form fatty acyl-CoA (3). The latter is transferred from the cytosol to the mitochondrial matrix by carnitine
acyl transferases (4), after which hydrolysis (β-oxidation) within the mitochondria yields acetyl-CoA (5).
Fig. 2-12: Fatty acid activation occurs in a two-step reaction that is catalyzed by acyl-CoA synthase. In the first step, the fatty acid binds
to adenosine monophosphate (AMP) to form an activated fatty acid∼AMP complex and pyrophosphate (PPi). In the second step, AMP is
replaced by CoA to form fatty acyl-CoA, which contains a high-energy thioester bond. ATP, adenosine triphosphate.
Fatty Acid Transfer

In order for activated fatty acids to enter the mitochondrial matrix, where they are oxidized, they must cross the mitochondrial outer and inner
membranes. However, the inner membrane is impermeable to fatty acyl-CoA, so that the bound CoA is replaced with carnitine to form fatty acyl-
carnitine which is transferred to the matrix by an exchange-diffusion (Fig. 2-13). Because carnitine cannot be synthesized by the heart, it must be
present in the diet. Carnitine deficiency is rare but can be treatable cause of heart failure (Paulson, 1998).
Fig. 2-13: Carnitine-mediated exchange of fatty acids across the mitochondrial inner membrane. Carnitine acyl transferases I and II
exchange CoA and carnitine in the mitochondrial intermembrane space and matrix, respectively. Fatty acid derivatives in the former, which
lies between the inner and outer membranes, are freely available to the cytosol. Fatty acyl-carnitine in the intermembrane space is then
exchanged for carnitine in the matrix by a translocase on the mitochondrial inner membrane. Fatty acyl-CoAc and acyl-carnitinec, activated
fatty acids in the cytosol and intermembrane space; fatty acyl-CoAm and fatty acyl-carnitinem, activated fatty acids within the
mitochondrial matrix; carnitinec and carnitinem, carnitine in the intermembrane space and matrix, respectively.
P.50
Fatty acid transfer begins when carnitine acyl-transferase I, located on the mitochondrial outer membrane, replaces CoA bound to cytosolic fatty
acyl-CoA with carnitine to form fatty acyl-carnitine. A translocase on the inner membrane then exchanges the cytosolic fatty acyl-carnitine for
carnitine in the matrix. After the former enters the matrix, carnitine acyl-transferase II located on the inner membrane then replaces the bound
carnitine in fatty acyl-carnitine with CoA, which forms fatty acyl-CoA in the matrix.
Fatty acid uptake into the mitochondria is regulated by malonyl-CoA, a powerful inhibitor of carnitine acyl-transferase I. Synthesis of this 3-carbon
fatty acid is catalyzed by AMP-activated protein kinase, a key regulator of cardiac energy production (see below). Fatty acid transfer is also
determined by the concentrations of the reactants on the two sides of the inner membrane, which allows high rates of fatty acid oxidation, which
reduce mitochondrial fatty acyl-CoA, to accelerate release of fatty acyl-CoA and carnitine from mitochondrial acyl-carnitine. The carnitine
released in the mitochondrial matrix is then exchanged for cytosolic fatty acyl-carnitine, which increases the amount of carnitine available to
convert cytosolic fatty acyl-CoA to fatty acyl-carnitine. The latter releases CoA that can bind to cytosolic fatty acids, which increases fatty acid
uptake by reducing cytosolic FFA concentration.
β-Oxidation
Long-chain fatty acyl CoA is oxidized in the mitochondria by a process called β-oxidation (Fig. 2-14). This stepwise breakdown of long-chain fatty
acids, most of which contain 16 or 18 carbon atoms, can be likened to a spiral in which each “turn” releases a 2-carbon acetyl-CoA through four
reactions; two of oxidation, one of hydration, and one of thiolysis (Fig. 2-15). The oxidations reduce NAD+ and FAD to form NADH and FADH2,
respectively. Thiolysis then adds CoA to the oxidized β-ketoacyl CoA, which releases acetyl-CoA, thereby shortening the fatty acid chain by 2-
carbons. Although β-oxidation does not regenerate ATP directly, the electrons transferred to NADH and FADH2 are used to energize oxidative
phosphorylation (see below).
β-Oxidation is very rapid in the well-oxygenated heart, where the rate-limiting step in fat metabolism is fatty acid transfer into the mitochondria.
In energy-starved hearts, where levels of oxidized NAD+ and FAD are low, β-oxidation can become rate-limiting. In patients with diabetes, in whom
high levels of FFAs accelerate fat metabolism, cytotoxic effects of reactive oxygen species generated during β-oxidation may play a role in the
pathogenesis of diabetic cardiomyopathy (Boudina and Abel, 2007).
Fat Deposition and Fatty Acid Synthesis

Fat deposits are commonly found beneath the epicardium and can envelop the heart in obese individuals; these extracellular deposits, called
fatty infiltrates, are largely of exogenous origin and usually have no clinical significance. In fatty degeneration, where fatty acids are synthesized
P.51
by cardiac myocytes (Hillgartner et al., 1995), fat droplets that appear within myocardial cells may contribute to membrane damage (Katz and
Messineo, 1981).
Fig. 2-14: Overall reaction of β-oxidation which releases acetyl-CoA from activated fatty acids (acyl-CoA). NAD+, oxidized nicotinamide
adenine dinucleotide; NADH, reduced nicotinamide adenine dinucleotide; FAD, flavine adenine dinucleotide; FADH, reduced flavine adenine
dinucleotide.
Fig. 2-15: β-Oxidation. Each cycle in the stepwise breakdown of long-chain fatty acids involves four steps: two where oxidation generates a
reduced coenzyme for oxidative phosphorylation, one of hydration, and one where thiolysis shortens the fatty acyl chain by two carbons.
Fatty Acids as Regulators of Gene Expression: PPARs

FFAs can regulate gene transcription when they bind to peroxisome proliferator-activated receptors (PPARs), which are analogous to steroid
hormone receptors. Binding of fatty acids to PPARα, the cardiac isoform of this receptor, increases energy production by activating the synthesis of
several proteins that participate in fatty acid oxidation (Brown and Plutzky, 2007).
Acetyl-Coa Oxidation: the Citric Acid Cycle

Acetyl-CoA formed from pyruvate and fatty acids is oxidized within the mitochondrial matrix by the citric acid cycle (tricarboxylic acid cycle,
Krebs cycle). The overall reaction, which breaks down each mole of acetyl-CoA to release 2 moles of carbon dioxide, generates 1 mole of ∼P by
substrate-level phosphorylation, 3 moles of NADH, and 1 mole of FADH2 (Fig. 2-16). Much larger amounts of ATP are regenerated when the NADH
and FADH2 are oxidized by the respiratory chain (see below).
P.52
Fig. 2-16: Overall reaction of the citric acid cycle. Oxidation and decarboxylation of each mole of acetyl CoA form 2 moles of carbon
dioxide, 1 mole of ∼P that is transferred from GTP to ATP, 3 moles of NADH, and 1 mole of FADH2. GDP, guanosine diphosphate; GTP,
guanosine triphosphate; NAD+, oxidized nicotinamide adenine dinucleotide; NADH, reduced nicotinamide adenine dinycleotide; FAD,
oxidized flavine adenine dinucleotide, flavine adenine dinucleotide; FADH, reduced flavine adenine dinucleotide.
The citric acid cycle begins when acetyl-CoA condenses with oxaloacetate, a 4-carbon organic acid, to form citrate and release CoA (Fig. 2-17). A
configurational rearrangement converts citrate, a 6-carbon organic acid, to cis-aconitate which is converted to isocitrate. The latter is then
oxidized and decarboxylated to form α-ketoglutaric acid, a 5-carbon organic acid which, following decarboxylation and oxidation, yields succinyl-
CoA, a 4-carbon organic acid linked to CoA by a high-energy thiol bond. The energy in the thiol bond is transferred to succinyl CoA synthase
(labeled E in Fig. 2-17B) to form E∼CoA, after which the enzyme-bound CoA is replaced by phosphate to form the high-energy phosphorylated
intermediate E∼P (Fig. 2-17B). The ∼P of E∼P is transferred first to guanosie diphosphate (GDP), which forms guanosine triphosphate (GTP), and
then to ADP, which regenerates ATP. This is the only reaction in the citric acid cycle where substrate-level phosphorylation forms a high-energy
phosphate bond.
The succinate released from succinyl-CoA is oxidized to fumarate by a FAD-containing respiratory chain enzyme complex called succinate-
coenzyme Q reductase (labeled E-FAD in Fig. 2-17A). The FADH2 produced in this reaction is oxidized in the respiratory chain, while the fumarate
is hydrated to form malate. The latter is oxidized in a reaction that reduces NAD+ to NADH and regenerates oxaloacetate, which completes the
citric acid cycle.
Like all important cellular processes, the citric acid cycle is highly regulated. Citrate synthase, which catalyzes the condensation of acetyl-CoA
with oxaloacetate, is influenced by the supply of acetyl-CoA and oxaloacetate; inhibition of this reaction by ATP slows the cycle in well-
oxygenated hearts. Isocitrate dehydrogenase, which catalyzes the production α-ketoglutarate from isocitrate, is inhibited by ATP and stimulated
by ADP, which allows increased energy consumption to accelerate energy production. The increased cytosolic calcium responsible for most
increases in myocardial contractility also helps match energy consumption and energy production by activating isocitrate dehydrogenase and α-
ketoglutarate dehydrogenase, which catalyzes succinyl-CoA formation. In anaerobic hearts, inhibition of citrate synthase and α-ketoglutarate
dehydrogenase by NADH slows the citric acid cycle. These and other regulatory mechanisms allow the citric acid cycle to respond to changes in
substrate supply, energy requirements, and the availability of oxidized coenzymes.
Transport of Reduced Nadh from Cytosol to Mitochondria: the Malate–Aspartate Cycle

The large amounts of NADH formed in the mitochondria by oxidative phosphorylation are readily accessible for oxidation by the respiratory chain,
but NADH produced in the cytosol by glycolysis must cross the mitochondrial inner membrane before it can be oxidized. Virtually all of the NAD+
required for aerobic glycolysis is formed in the mitochondria, and so must cross the mitochondrial inner membrane to reach the cytosol. However,
reducing equivalents cannot
P.53
P.54
be transferred directly between the cytosol and mitochondrial matrix. Instead, NADH and NAD+ transfer use a complex exchange mechanism,
called the malate–aspartate cycle, that includes the oxidations, reductions, and transaminations shown in Figure 2-18.
Fig. 2-17: Citric acid cycle (A) and succinyl-CoA metabolism (B). A: Condensation of acetyl-CoA with oxaloacetate yields citrate, a 6-carbon
acid, which, after isomerization to isocitrate, is oxidized and decarboxylated to form α-ketoglutarate. The latter is oxidized and
decarboxylated to form succinyl-CoA which, after the CoA is transferred to succinyl CoA synthase, undergoes two steps of oxidation and one
of hydration to regenerate oxaloacetate. B: Succinyl CoA contains a high-energy thiol bond that is used to generate adenosine triphosphate
(ATP) by substrate-level phosphorylation in reactions that are catalyzed by succinyl CoA synthase (E). This enzyme transfers the high-energy
bond linking CoA to succinyl-CoA first to GTP, which forms GTP, and then to ADP to regenerate ATP. ADP, adenosine diphosphate; FAD, flavine
adenine dinucleotide; FADH, reduced flavine adenine dinucleotide; GDP, guanosine diphosphate; GTP, guanosine triphosphate; NAD,
oxidized nicotinamide adenine dinucleotide; NADH, reduced nicotinamide adenine dinucleotide.
Fig. 2-18: The malate–aspartate cycle. Four simultaneous reactions “transfer” NADH from the cytosol to the mitochondria. (1) Reduction of
oxaloacetate in the cytosol oxidizes NAD+ and forms malate (1a), after which the malate is transferred across the mitochondrial inner
membrane (labeled MITOCHONDRIAL MEMBRANE) into the mitochondrial matrix by a membrane carrier (I) where oxidation of the malate
regenerates oxaloacetate and releases NADH (1b). (2) The oxaloacetate formed in the matrix is transaminated with glutamate (Tm) to form
α-ketoglutarate and aspartate, after which aspartate is returned to the cytosol by another membrane carrier (II). (3) After entering the
cytosol, the aspartate is trans-aminated with α-ketoglutarate to form oxaloacetate and glutamate (Tc), after which the latter is returned to
the matrix by membrane carrier II in exchange for the aspartate produced in reaction 2. (4) The glutamate returned to the matrix is
transaminated with oxaloacetate (Tm, described in reaction 2) to yield α-ketoglutarate, after which the cycle is completed by exchange of
mitochondrial α-ketoglutarate for malate by membrane carrier I (reaction 1). NAD, oxidized nicotinamide adenine dinucleotide; NADH,
reduced nicotinamide adenine dinucleotide.
The complex reactions of the malate-aspartate cycle might, at first glance, seem to represent a needlessly elaborate way to move NADH from the
cytosol for oxidation in the mitochondrial matrix and to provide NAD+ for cytosolic reactions. The complexity of this transfer (Why don't
mitochondria simply have a carrier to exchange NAD+ for NADH?) illustrates nature's failure to adhere to Ockham's razor: Pluralitas non est
ponenda sine necessitate (Plurality should not be posited without necessity); this 14th-century precept can be paraphrased as meaning that when
choosing among competing explanations, it is best to start with the simplest. However, the overlapping layers of control that characterize
biological regulation, although violating Ockham's razor, are advantageous for homeostasis. In the malate–aspartate cycle, this complexity helps
match energy production and energy utilization by adjusting the rates of anaerobic and aerobic energy production to changes in the supply of key
substrates and cofactors.
Transaminations
Transfer of amino groups between organic acids and amino acids (e.g., α-ketoglutarate and glutamate, oxaloacetate and aspartate, pyruvate and
glutamate) is catalyzed by pyridoxal-dependent enzymes called transaminases or aminotransferases. In addition to their role in the malate–
aspartate cycle, transaminations enable the heart to metabolize a small amount of protein. In hearts operating under anaerobic conditions,
lactate formation and the onset of acidosis are delayed when pyruvate is transaminated to form alanine (Taegtmeyer et al., 1977).
P.55
Adenine Nucleotide Transfer

The mitochondrial inner membrane is impermeable to adenine nucleotides as well as to NADH and NAD+. For this reason, transfer of ATP and ADP
between the cytosol and mitochondrial matrix depends on an exchanger, called the ATP–ADP translocase (transferase), that couples ATP flux in
one direction to ADP flux in the opposite direction (Fig. 2-19). Because ATP has three negative charges and ADP has only two, ATP transfer from
the matrix into the cytosol adds to the negative potential across the mitochondrial inner membrane that is generated by the proton pumps (see
below).
The Phosphocreatine Shuttle

The high ATP concentration in myocardial cells, which is in the millimolar range, allows this high-energy compound to diffuse rapidly from the
mitochondria, where ATP is regenerated, to cytosolic structures where it is hydrolyzed. Diffusion of ADP through the cytosol for rephosphorylation
in the mitochondria, however, is slower because of the low cytosolic ADP concentration (see below). The slow diffusion of ADP poses a serious
problem for the heart, where energy utilization is very rapid, that is solved by the transfer of creatine and phosphocreatine, rather than ADP and
ATP, within the cytosol (McClellan et al., 1983; Jacobus, 1985; Kammermeier, 1987). This transfer, called the phosphocreatine shuttle (Fig. 2-19),
takes advantage of the high cytosolic concentrations of creatine and phosphocreatine to return creatine rather than ADP to the mitochondria for
rephosphorylation and to use phosphocreatine instead of ATP to carry ∼P from the mitochondria to energy-consuming structures in the cytosol.
Fig. 2-19: ATP–ADP translocase and the phosphocreatine shuttle. ATP hydrolysis by energy-utilizing systems in the cytosol (ATPase, left)
releases ADP that is immediately rephosphorylated by cytosolic creatine phosphokinase using ∼P derived from phosphocreatine. The
creatine produced by this reaction then diffuses to the mitochondria (right) where mitochondrial creatine phosphokinase uses ∼P from ATP
to regenerate phosphocreatine and release ADP outside the inner membrane. This ADP is exchanged for ATP in the mitochondrial matrix by
the ATP–ADP translocase. After entering the matrix, ADP is rephosphorylated by oxidative phosphorylation to form ATP which is exchanged
for cytosolic ADP by the ATP–ADP translocase. The ∼P in the ATP outside the inner membrane is transferred to creatine by mitochondrial
creatine phosphokinase to regenerate phosphocreatine that diffuses through the cytosol to supply energy to cytosolic ATPases (left). ADP,
adenosine diphosphate; ATP, adenosine triphosphate.
P.56
Transfer of ∼P between ADP and creatine by the phosphocreatine shuttle depends on the ATP–ADP transferase described above and two enzymes.
The first, mitochondrial creatine phosphokinase, is located on the mitochondrial inner membrane where it transfers ∼P from ATP generated in the
mitochondrial matrix to creatine, forming phosphocreatine in the cytosol and releasing ADP that is returned to the matrix in exchange for
additional ATP (Fig. 2-19, right). After entering the mitochondrial matrix, this ADP can be rephosphorylated by oxidative phosphorylation.
The phosphocreatine released in the cytosol by mitochondrial creatine phosphokinase diffuses to energy-utilizing sites where the second enzyme,
cytosolic creatine phosphokinase, regenerates ATP by transferring ∼P from phosphocreatine to ADP (Fig. 2-19, left). This ATP supplies ∼P to
cytosolic ATPases such as the contractile proteins, sarcoplasmic reticulum, and sodium pump. The cytosolic creatine formed by these and other
energy-consuming structures diffuses rapidly to the mitochondria where it can accept additional ∼P (Fig. 2-19, right).
The Respiratory Chain

The ATP generated by substrate-level phosphorylations in glycolysis and the citric acid cycle is not sufficient to meet the energy demands of the
beating heart; instead, most of this high-energy phosphate is regenerated by oxidative metabolism. The latter involves four “steps” (Table 2-3)
that begin when electrons produced by carbohydrate and fatty acid oxidation are transferred to NAD+ and FAD to form NADH and FADH2. In the
second step, NADH and FADH2 transfer the electrons to the respiratory chain (electron transport chain), after which the third step uses the energy
carried by these electrons to establish a proton electrochemical gradient across the mitochondrial inner membrane. The final step occurs when
proton flux down this gradient releases energy that regenerates ATP by respiratory chain-linked phosphorylations. Transfer of electrons through
these tightly coupled steps, rather than by a single explosive reaction, minimizes energy wastage and the “escape” of reactive oxygen species
within cells.
I came to appreciate the large amount of energy released when oxygen and hydrogen form water when, as a teenager, I combined atmospheric
oxygen with a small amount of hydrogen (generated by pouring hydrochloric acid over powdered zinc) that I had trapped in a rubber glove. After I
ignited the reactants using a match tied to a 6-foot pole, the resulting explosion rattled windows more that a hundred meters away. Oxidative
phosphorylation can be viewed as a
P.57
mechanism that retains this energy in the high-energy phosphate bonds of ATP. The “magic” of the respiratory chain is its ability to prevent the
highly reactive electrons and free radicals formed during electron transfer from interacting with structures within the cell, but instead to form
oxygen free radicals that combine with protons to form water.
Table 2-3 Four “Steps” in Oxidative Metabolism
Step 1: Transfer of electrons from substrates to coenzymes
Step 2: Passage of electrons from coenzymes through the respiratory chain
Step 3: Establishment of proton electrochemical gradient across mitochondrial inner membrane
Step 4: Downhill proton flux across the mitochondrial inner membrane and regeneration of ATP
ATP, adenosine triphosphate.
Electrons or Reducing Equivalents?

The respiratory chain uses the energy carried by the electrons in reduced NADH and FADH2 in a more controlled manner than occurred in my
rubber glove. One way to characterize these reactions is to view oxidation of the reduced coenzymes as the transfer of reducing equivalents and
protons to oxygen, thereby forming water:
and
Because each hydrogen atom (H) is a proton (H+) plus an electron (e), these reactions can also be characterized as electron transfer to molecular
oxygen; viewed in this manner, NADH is equivalent to (NAD+ - H+ - 2e) and FADH2 to (FAD - 2H+ - 2e), so that the overall reactions can be written:
and
The second pair of equations describes oxidative phosphorylation as the transfer of electrons carried by the reduced coenzymes to molecular
oxygen to form O2, a free radical that combines with the protons to form water.
The oxidations in carbohydrate and fat metabolism can also be viewed as the transfer of electrons (along with protons) to reduced coenzymes in
which the electrons become associated with nitrogen or oxygen atoms in organic ring structures:
and
P.58
Electrons carried by respiratory chain intermediates can also be associated with nitrogen, oxygen, iron sulfates, the heme iron of cytochromes,
and copper. The metals accept electrons according to the general reactions:
and
Electron Transfer Through the Respiratory Chain

The metabolic reactions described in this chapter generate both protons and highly reactive electrons. The electrons, after oxidizing coenzymes
like NADH and FADH2, are transferred to molecular oxygen to form O2 that combines with protons to form water according to the overall reaction:
As electrons are transferred from substrates to coenzymes and then move along the respiratory chain (Fig. 2-20), they lose energy. This energy is
captured in a series of tightly linked oxidations and reductions to activate a proton pump that establishes a proton gradient, called a proton
motive force, across the mitochondrial inner membrane. Like an old-fashioned “bucket brigade,” which relies on organization, proximity, and
careful handling of water-filled buckets to minimize spillage and maximize efficiency, the electron carriers of the respiratory chain are organized
to minimize the energy loss and cell damage that would occur if the highly reactive electrons were released into the cytosol.
The respiratory chain includes four multiprotein complexes, all of which are bound to the mitochondrial inner membrane. Three (NADH-CoQ
reductase, CoH2Q-cytochrome c reductase, and cytochrome c oxidase) are coupled to proton pumps that use energy released when electrons pass
through the complexes to move protons uphill, out of the matrix, which generates an electrochemical gradient that energizes ATP regeneration
(see below). The fourth complex, succinate-CoQ reductase, passes electrons down the chain but does not energize proton transport.
Electrons can enter the respiratory chain from NADH, whose oxidation provides a pair of electrons that release 52.6 kcal/mol, and from FADH2,
which releases 43.4 kcal/mol (Fig. 2-20). The former provides sufficient energy to regenerate approximately 3 moles of ATP, whereas the latter
regenerate only approximately 2 moles of ATP. The electrons carried by NADH are transferred to NADH-CoQ reductase (Complex I) through which
they are passed to FMN and several iron sulfate clusters (FeS) before being transferred to coenzyme Q (CoQH2, ubiquinone). After accepting an
electron, the latter is converted to a free radical called a semiquinone (labeled CoQH2•, in Fig. 2-20). The dot (•) denotes a free radical. The
electrons carried by the FADH2 formed during β-oxidation (Fig. 2-15) and succinate oxidation (Fig. 2-17A) enter the respiratory chain in reactions
that are catalyzed by succinate-CoQ reductase (Complex II), which transfers the electrons from FADH2 first to Fe3+S and then to CoQH2•. Because
electrons carried by FADH2 bypass
P.59
Complex I before entering the respiratory chain (Fig. 2-20) they activate only two of the three proton pumps, which is why less ATP is generated
by oxidation of FADH2 than of NADH.
Fig. 2-20: The respiratory chain. Reduced NADH (left) delivers electrons (shaded circles labeled “e”) to NADH-CoQ reductase (Complex I),
which transfers the electrons to CoQH2-cytochrome c reductase (Complex III) and cytochrome c oxidase (Complex IV). As the electrons pass
along the respiratory chain they transfer energy to three proton pumps that transport H+ uphill, out of the mitochondrial matrix. At the end
of the chain (right) the electrons form an oxygen free radical with molecular oxygen to form O2 (not shown) which then combines
immediately with protons to produce water. Unlike the electrons in NADH, which energize three proton pumps (below), the electrons in
FADH2 enter the respiratory chain via succinate-CoQ reductase (Complex II); because this bypasses the first proton pump, FADH2
regenerates less adenosine triphosphate than NADH. FMN, flavine mononucleotide; FeS, iron sulfate; CoQH2, oxidized coenzyme Q
(ubiquinone); CoQH2•·, semiquinone, a free radical; cyt, cytochrome; Cu, copper. NAD+, oxidized nicotinamide adenine dinucleotide;
NADH, reduced nicotinamide adenine dinucleotide; FAD, oxidized flavine adenine dinucleotide; FADH2, reduced flavine adenine
dinucleotide.
The third enzyme complex, CoQH2Q-cytochrome c reductase (Complex III), accepts electrons that enter the respiratory chain from both
Complexes I and II. After transfer from CoQH2•, the electrons move through Complex III via iron atoms in cytochrome b, a ferrous sulfate-
containing protein, and cytochrome c1 before reaching cytochrome c. Electron transfer through Complex III energizes the second proton pump
(Fig. 2-20). The electrons are then transferred from cytochrome c to the fourth enzyme complex, cytochrome c oxidase (Complex IV). The latter,
which contains copper and cytochromes a and a3, energizes the third proton pump (Fig. 2-20). Electron passage through the respiratory chain ends
when the electrons, having lost most of their energy, are transferred to molecular oxygen in a reaction that forms O2—a highly reactive oxygen
free radical that combines immediately with protons to form H2O.
P.60
Fig. 2-21: The proton electromotive force. Transport of protons (H+) out of the mitochondrial matrix generates an electrochemical gradient
across the inner membrane in which the matrix becomes electronegative and depleted of protons (alkaline). The downhill flux of protons
into the matrix establishes the proton electromotive force that provides energy for oxidative phosphorylation.
Proton Pumps and the Proton Electromotive Force

Much of the energy released by the passage of electrons through the respiratory chain is used to pump protons (H+) “uphill” out of the
mitochondrial matrix (Fig. 2-21). The resulting proton gradient across the mitochondrial inner membrane energizes ATP regeneration (see below).
Unlike substrate-linked phosphorylations, which transfer ∼P from phosphorylated substrates directly to ADP, there is no strict stoichiometry
between electron transport, proton pumping, and ATP regeneration.
Calcium Fluxes into and out of the Mitochondria

The negative electrical potential within the mitochondrial matrix favors entry of calcium ions into these cell organelles. In the normal heart,
where cytosolic calcium concentration is very low, uptake of activator calcium by the mitochondria does not contribute to relaxation, but when
cytosolic calcium is high, as occurs in energy-starved hearts, mitochondria accumulate calcium. When severe, calcium overload can cause
calcium-phosphate precipitates to form within the mitochondria.
Mitochondrial calcium uptake uncouples oxidative phosphorylation when these cations dissipate the proton electromotive force across the
mitochondrial inner membrane. For this reason, energy starvation can initiate a vicious cycle where low ATP levels impair active calcium transport
out of cardiac myocytes, which increases cytosolic calcium, which promotes mitochondrial calcium uptake, which uncouples oxidative
phosphorylation, which worsens energy starvation.
P.61
Increased cytosolic calcium also activates the energy-consuming reactions of the contractile proteins, which initiates another vicious cycle that
can lead to cardiac myocyte necrosis in ischemic (Chapter 17) and failing hearts (Chapter 18).
Oxidative Phosphorylation
Efforts to define the molecular mechanism of oxidative phosphorylation illustrate the danger of extrapolating from one process to another.
Because a number of phosphorylated intermediates had already been described in substrate-level phosphorylations, early workers followed what
seemed to be reasonable strategy by searching for analogous intermediates in oxidative phosphorylation. However, years of futile efforts failed to
identify a phosphorylated intermediate; instead, the solution came from a totally unexpected direction. In accord with George Bernard Shaw's
maxim: “The reasonable man adapts himself to the world; the unreasonable one persists in trying to adapt the world to himself. Therefore all
progress depends on the unreasonable man.” (Shaw, 1903), Peter Mitchell demonstrated that mitochondria do not regenerate ATP using chemical
energy in a phosphorylated intermediate but instead use osmotic energy derived from the proton motive force to rephosphorylate ADP. Subsequent
work demonstrated that this occurs when a torque developed by proteins related to bacterial flagella captures the energy generated by downhill
proton flux to transfer ∼P to ADP.
Atp Synthase
A multiprotein enzyme called ATP synthase regenerates ATP in the mitochondria. This membrane-spanning protein complex is made up of two
connected structures: F0, a protein complex within the membrane bilayer that contains a proton channel, and F1, which projects from the inner
membrane into the matrix (Fig. 2-22). ATP synthesis is energized when the downhill flux of protons through the proton channel rotates the
transmembrane F0 complex. This rotational energy is then captured by ADP- and ATP-binding sites on the F1 complex to form the high-energy
bonds that link Pi to ADP. When isolated, the F0 complex runs in “reverse” and hydrolyzes ATP; for this reason, it was originally called the
F1ATPase.
Some of the proteins in ATP synthase are homologous to those in bacterial flagella, which use chemical energy to move these prokaryotes by
spinning their flagella. The fact that mitochondria generate ATP when this process runs in the opposite direction provides further evidence that
mitochondria descended from prokaryotes that live within our cells (see Chapter 1).
Control of Oxidative Phosphorylation

The major determinant of the rate of oxidative phosphorylation in the normal heart is the availability of ADP. This reflects the tight coupling
between ATP regeneration from ADP and Pi, proton flux across the inner membrane, electron flux through the respiratory chain, oxidation of
NADH and FADH2, and substrate oxidation (Fig. 2-23). Because protons cannot flow through the F0 channel in ATP synthase unless ADP is available
to couple the proton flux to ATP regeneration by F1, all of these reactions come to a halt in the absence of ADP. In well-oxygenated hearts, control
of oxidative energy production by ADP helps match ATP production
P.62
P.63
and utilization. However, in ischemic hearts, in which ADP rephosphorylation is inhibited by oxygen lack, the ability of the respiratory chain to
generate the proton motive force becomes rate-limiting. As a result, control of these tightly coupled reactions shifts to the availability of oxygen,
rather than of ADP.
Fig. 2-22: Adenosine triphosphate (ATP) synthase and a hydraulic model showing its mechanism of action. A: ATP synthase, a multiprotein
complex that projects from the mitochondrial inner membrane into the matrix, is made up of two connected structures: F0, a
transmembrane segment that contains a channel through which protons flow down their electrochemical gradient into the mitochondrial
matrix and F1 which converts energy generated by this downhill proton flux to a torque that rotates the F1 ATPase. The latter regenerates
ATP by energizing the formation of high-energy bonds linking Pi to adenosine diphosphate (ADP). B: Utilization of rotational energy by F1 to
regenerate ATP is analogous to the generation of electricity by a water-powered turbine in a dam.
Fig. 2-23: The tight coupling between substrate oxidation, oxidation of NADH and FADH2, electron flux through the respiratory chain,
proton flux across the mitochondrial inner membrane, and adenosine triphosphate (ATP) regeneration from adenosine diphosphate (ADP)
and Pi by oxidative phosphorylation resemble a group of interlocking gears that must all turn together, NADH, reduced nicotinamide adenine
dinucleotide; FADH, reduced flavine adenine dinucleotide.
Integration of Glycolysis and Respiration

Although the well-oxygenated heart can metabolize both carbohydrates and lipids, fatty acids are the preferred substrates, providing more than
70% of the energy expenditure. Fatty acids inhibit carbohydrate metabolism when the heart is presented with both substrates, in part because the
high concentrations of ATP regenerated by oxidative phosphorylation, along with reduced concentrations of AMP and Pi, slow glycogen breakdown
by inhibiting phosphorylase b, and inhibit glycolysis by reducing phosphofructokinase-1 activity. The latter also increases glucose 6-phosphate
levels, which slow the hexokinase reaction. Increased levels of citrate, which is produced in oxygenated hearts when the large amounts of acetyl
CoA generated by β-oxidation condense with oxaloacetate, also slow glycolysis by inhibiting phosphofructokinase-1 and pyruvate dehydrogenase
phosphatase. These relationships change in chronically overloaded and failing hearts, where the preferred substrates shift from fats to
carbohydrates (see Chapter 18).
ATP Compartments and “Glycolytic ATP” in Cardiac Myocytes

The total ATP content (moles/weight) in the human heart is 6 to 8 mmoles/kg wet weight; if all of this ATP were distributed uniformly in the cell
water, ATP concentration (moles/volume) would be approximately 5 mM. However, differences in the rates of ATP production and consumption in
various regions of myocardial cells, coupled with the relatively slow transport of nucleotides across the mitochondrial inner membrane, give rise
to gradients for ATP, ADP, and especially AMP between the cytosol and mitochondrial matrix (Illingworth et al., 1975; Wallimann et al., 1992;
Rauch et al., 1994). Gradients also result from the structural organization of other energy-transferring systems (Dzeja and Terzic, 2003; Seppet et
al., 2006).
One consequence of this compartmentation is that some energy-dependent systems preferentially use ATP produced by glycolysis, often called
glycolytic ATP (Apstein, 2000). This preference is due partly to the ability of cytoskeletal elements to locate glycolytic enzymes that regenerate
ATP in close proximity to specific ATP-consuming structures.
Adenylate Kinase, AMP, and Amplification

Adenylate kinase (also called myokinase) plays an important role in cellular regulation by catalyzing the reaction:
P.64
This reaction has several important functions. By regenerating ATP from ADP, adenylate kinase increases the free energy available from ATP
hydrolysis, which is proportional to the ratio ATP/ADP. For example, in a heart where the total adenine nucleotide pool (ATP, ADP, and AMP) is 5
mM, the equilibrium constant for this enzyme would maintain ATP concentration at approximately 4.48 mM, ADP at approximately 0.5 mM, and
AMP at approximately 0.02 mM. The adenylate kinase reaction also allows a small decrease in ATP concentration to cause a proportionately large
increase in AMP concentration. This is one example of a mechanism called amplification, which in the heart allows a small change in ATP
concentration to stimulate energy production when cardiac work increases or coronary flow is interrupted. For example, according to the kinetics
of the adenylate kinase reaction, a 13% fall in ATP concentration to 3.89 mM would double ADP concentration to 1.0 mM and increase AMP
concentration more than fivefold to 0.11 mM. The changes would result in a marked increase in glycolytic activity when the rise in AMP causes
allosteric stimulation of hexokinase and phosphofructokinase-1.
The low AMP concentrations maintained by adenylate kinase also minimize the amount of AMP available for dephosphorylation, which is important
in preserving the total adenine nucleotide pool. This is because adenosine, the product of AMP dephosphorylation, is rapidly and irreversibly
deaminated to form hypoxanthine, whereas the de novo synthesis of adenine required to replenish the adenine nucleotide pool requires several
days.
AMP-Activated Protein Kinase

AMP plays an important role in regulating cell function by activating adenosine monophosphate-activated protein kinases (AMPK), which have
been likened to a metabolic fuel gauge or an orchestra conductor (Dyck and Lopaschuk, 2002; Arad et al., 2007; Young, 2008). These enzymes,
which regulate the activity of a number of metabolic enzymes and other proteins, increase energy production and inhibit energy utilization when
AMP levels rise in energy-starved hearts. Stimulation of AMPK by AMP can be direct, when the nucleotide induces a conformational change in
AMPK, and indirect, when AMP exposes a regulatory threonine on AMPK that, when phosphorylated by an AMPK kinase, increases AMPK activity.
AMPK plays an important physiological role by matching energy production and energy utilization, and by integrating carbohydrate and fatty acid
metabolism. The physiological effects of AMPK include increased glucose uptake caused when AMPK increases the amount of GLUT 4 at the cell
surface, and increased synthesis of fructose 2,6-bisphosphate which, by activating phosphofructokinase-2, increases phosphofructokinase-1
activity. AMPK increases fatty acid oxidation by accelerating fatty acid transport into the mitochondria; this occurs when AMPK inhibits acetyl-CoA
carboxylase, an enzyme that carboxylates acetyl-CoA to form malonyl-CoA, a powerful inhibitor of the carnitine acyl-transferase I that transports
fatty acid into the mitochondria (see above). In addition to inhibiting malonyl CoA production, AMPK increases fatty acid oxidation by stimulating
the release of fatty acids from lipoproteins by endothelial cell lipoprotein lipase, and the uptake of FFAs into cardiac myocytes. The latter occurs
when AMPK recruits CD36, a fatty acid transport protein, to the plasma membrane.
When a heart becomes energy-starved—as occurs during ischemia—AMPK increases energy production, slows energy-consumption, and exerts
protective effects that preserve myocardial viability (Table 2-4). AMPK also reduces energy utilization by inhibiting the synthesis of glycogen,
triglycerides, and proteins, and protects against cardiac myocyte death by inhibiting
P.65
programmed cell death (apoptosis, see Chapter 9) and promoting ischemic preconditioning (see Chapter 17).
Table 2-4 Some Effects of Adenosine Monophosphate-Activated Protein Kinases on the Heart
Increased plasma membrane GLUT 4 activity, which increases glucose uptake
Increased phosphofructokinase-1 activity, which accelerates glycolysis
Reduced synthesis of malonyl-CoA, an inhibitor of carnitine acyl-transferase I, which increases fatty acid transfer into mitochondria
Increased endothelial lipoprotein activity, which increases fatty acid release from lipoproteins
Increased plasma membrane activity of CD36, a fatty acid transport protein, which increases fatty acid uptake into myocytes
Decreased synthesis of glycogen, triglycerides, and proteins, which conserves energy
Induction of ischemic preconditioning, which increases survival of energy-starved myocytes
Inhibition of apoptosis, which inhibits cell death

GLUT-4; glucose transporter.
Energy Balances
The higher yield of ATP from the metabolism of fats than carbohydrates is seen in Tables 2-5A and 2-5B, which show that when calculated on the
basis of the weight, oxidation of a saturated fatty acid regenerates more than twice as much ATP as does oxidation of glucose.
Carbohydrate Metabolism
Anaerobic glycolysis yields a total of 2 moles of ATP per mole of glucose by substrate-level phosphorylations. In contrast, approximately 38 moles
of ATP are regenerated per mole of glucose by aerobic glycolysis (Table 2-5A). Of the additional approximately 36 moles of ATP regenerated in the
presence of oxygen, 2 are produced by substrate-level phosphorylation; this occurs in the citric acid cycle when the high-energy thiol bond linking
CoA to succinate in succinyl CoA is used to phosphorylate first GDP and then ADP (Fig. 2-17).
Most of the ATP regenerated by aerobic glucose metabolism uses energy released when reduced coenzymes are oxidized in the respiratory chain.
Each mole of triose formed in the initial reactions of glycolysis (Fig. 2-8) generates 1 mole of NADH during glyceraldehyde 3-phosphate oxidation,
so that because glycolysis generates 2 moles of triose per mole of glucose, a total of 2 moles of NADH are produced per mole of glucose. When
oxidized in the respiratory chain, each mole of NADH regenerates approximately 3 moles of ATP so that aerobic glycolysis, in addition to
regenerating 2 moles of ATP by substrate-level phosphorylation, provides approximately 6 moles of ATP per mole of glucose by respiratory chain-
linked phosphorylation. Aerobic glycolysis therefore yields a total of approximately 8 moles of ATP per mole of glucose (Table 2-5A). Oxidation of
pyruvate to form acetyl-CoA yields an additional mole of NADH that, when oxidized in the respiratory chain, provides approximately 3 moles of
ATP; because each mole of glucose yields 2 moles of pyruvate, pyruvate oxidation adds approximately 6 moles of ATP per mole of glucose (Table 2-
5A).
P.66
Table 2-5A Energy Balances: ATP Regeneration from Carbohydrate Metabolism
Carbohydrate Metabolism (moles ATP/mole Glucose)
Reaction Substrate-Level Phosphorylation Respiratory Chain-Linked Phosphorylation Total
Anaerobic
Anaerobic glycolysis
Glucose → lactate 2 0 2
Aerobic (Glucose oxidation)
Aerobic glycolysis
Glucose → pyruvate 2 0
2 NADH → 2 NAD+ + 2H 0 6a
Total 2 6a 8a
Pyruvate oxidation
2 pyruvate S 2 acetyl-CoA + 2 CO2 0 0
2 NADH → 2 NAD+ + 2H 0 6a
Total 0 6a 6a
Acetyl-CoA oxidation
2 Acetyl-CoA → 4 CO2 2 0
6 NADH → 6 NAD+ + 6H 0 18a
2 FADH2 → 2 FAD + 4H 0 4a
Total 2 22a 24a
Glucose oxidation (total)
Per mole 4 34a 38a
Per 100 g (MW = 192) 20a
ATP, adenosine triphosphate; NADH; NAD+; FADH; FAD, flavine adenine dinucleotide.
aThese values are approximations because the energy released during electron transport through the respiratory chain is not
stoichiometrically linked to ATP regeneration. NAD+, oxidized nicotinamide adenine dinucleotide; NADH, reduced nicotinamide
adenine dinycleotide; FADH, reduced flavine adenine dinucleotide.
Each mole of acetyl-CoA oxidized in the citric acid cycle generates 3 moles of NADH (one each during oxidation of isocitrate, α-ketoglutarate, and
malate) and 1 mole of FADH2 (generated during succinyl-CoA oxidation). Oxidation of each mole of NADH by the respiratory chain yields
approximately 3 moles of ATP, so that the acetyl-CoA oxidation in the citric acid cycle adds approximately 9 moles of ATP per mole of acetyl-CoA,
or approximately 18 per mole of glucose. FADH2 oxidation yields an additional approximately 2 moles of ATP per mole of acetyl-CoA, or
approximately 4 per mole of glucose. Respiratory chain-linked phosphorylation therefore regenerates a total of approximately 11 moles of ATP per
mole of acetyl-CoA, or approximately 22 per mole of glucose.
The total of approximately 38 moles of ATP regenerated by the aerobic metabolism of each mole of glucose compares with only 2 moles of ATP
produced by anaerobic glycolysis (Table 2-5A).
P.67
Table 2-5B Energy Balances: ATP Regeneration from Fat Metabolism
Fat Metabolism (moles ATP/mole Palmitic Acid)
Reaction Substrate-Level Phosphorylation Respiratory Chain-Linked Phosphorylation Total
Aerobic
Fatty acid activation -1 0 -1
β-Oxidation
8 acyl-CoA(n) → 8 acyl-CoA(n-2) + 8 acetyl-CoA 0 0
8 NADH → 8 NAD+ + 8H 0 24a
8 FADH2 → 8 FAD + 16H 0 16a
Total 0 40a 40a
Acetyl-CoA oxidation
8 Acetyl-CoA → 16 CO2 8 0
24 NADH → 24 NAD+ + 24 H 0 72a
8 FADH2 → 8 FAD + 16 H 0 16a
Total 8 88a 96a
Palmitate oxidation (total)
Per mole 7 128a 135a
Per 100 g (MW = 272) 50a
ATP, adenosine triphosphate; NADH; NAD+; FADH; FAD, flavine adenine dinucleotide.
aThese values are approximations because the energy released during electron transport through the respiratory chain is not
stoichiometrically linked to ATP regeneration. NAD+, oxidized nicotinamide adenine dinucleotide; NADH, reduced nicotinamide
adenine dinycleotide; FADH, reduced flavine adenine dinucleotide.
Fat Metabolism
Almost all of the ATP regenerated by fat metabolism depends on oxidative phosphorylation. The balances at the top of Table 2-5B are calculated
for the metabolism of palmitic acid, a saturated 16-carbon fatty acid. During β-oxidation, each mole of palmitate yields 8 moles of acetyl-CoA and
approximately 8 moles each of reduced NADH and FADH2 (Table 2-5B). Oxidation of the reduced coenzymes regenerates a total of approximately
40 moles of ATP. Subsequent oxidation of the 8 moles of acetyl-CoA regenerates an additional approximately 96 moles of ATP: 8 by substrate-level
phosphorylation in the citric acid cycle and approximately 88 by oxidation of the reduced coenzymes (Table 2-5B). After subtraction of the single
mole of ATP used to activate the fatty acid, palmitate oxidation yields approximately 135 moles of ATP. When these balances are normalized to
the molecular weights of each substrate (Tables 2-5A and 2-5B) palmitate oxidation can be seen to regenerate more than twice as much ATP per
gram as glucose oxidation. These calculations explain why more energy is released by oxidation of fats (9 cal/g) than carbohydrates (5 cal/g).
The heart must pay a “price” to use these efficient oxidative pathways; this is an absolute dependence on a supply of oxygen that must be
delivered via the coronary arteries. The
P.68
devastating effects of coronary artery occlusion on the heart (see Chapter 17) are readily understood because oxygen lack halts production of
almost 95% of the ATP potentially available from glucose metabolism and makes it impossible for the heart to regenerate ATP from fat
metabolism.
Overview of the Control of Energy Production by the Heart

Energy production by the heart is regulated by at least four different types of mechanism (Table 2-2). Humoral control allows circulating
hormones, neurotransmitters, and other extracellular messengers to modify the transporters that mediate the entry of carbohydrate and fat into
the myocardium, and several enzymes that control the metabolism of these substrates. Most important is sympathetic stimulation which, by
increasing energy utilization (see Chapter 8), accelerates ATP production. The second mechanism, which responds to changing high-energy
phosphate levels, helps match the rates of ATP production and ATP utilization. In the normal heart, ATP levels remain virtually constant as cardiac
work changes; this reflects the role of ADP and AMP in determining the rate of oxidative phosphorylation, the ability of high-energy phosphate
levels to modify key glycolytic enzymes such as phosphofructokinase-1 and PDH, and the many effects caused when AMPK is stimulated by AMP.
The third type of regulation responds to changes in redox state, which allows energy production to be regulated by levels of oxidized and reduced
coenzymes, for example, by increasing anaerobic glycolysis when depletion of oxidized coenzymes slows oxidative metabolism in energy-starved
hearts. The fourth mechanism, which is effected by changes in proliferative signaling, plays a central role in responses to long-term challenges
such as heart failure, where isoform shifts and changes in the contents of several enzymes increase the proportion of energy derived from glucose
oxidation, relative to fatty acid oxidation (see Chapter 18). The ability of these regulatory mechanisms to adjust energy production to meet the
needs of the heart allow ATP be regenerated at the same rate that it is consumed, and maximize cardiac energy production in disease illustrates
the observation, made by Stephen Hales who first measured blood pressure in 1733: “So curiously are we wrought, so fearfully and wonderfully
are we made.”
Bibliography
Alberts B, Johnson A, Lewis J, et al. Molecular biology of the cell. 5th ed. New York: Garland, 2008.
Alpert NR, Mulieri LL, Hasenfuss G. Myocardial chemo-mechanical energy transduction. In: Fozzard H, Haber E, Katz A, et al., eds. The heart
and cardiovascular system. 2nd ed. New York: Raven, 1991:111–128.
Chock PB, Stadtman ER. Superiority of interconvertible enzyme cascades in metabolic regulation: analysis of multicyclic systems. Proc Natl
Acad Sci USA 1977;74:2766–2770.
Depre C, Vanoverschelde J-LJ, Taegtmeyer H. Glucose for the heart. Circulation 1999;99:578–588.
Devlin TM, ed. Textbook of biochemistry, 4th ed. New York: Wiley, 1997.
Finck BN, Kelly DP. Peroxisome proliferator-activated receptor α (PPARα) signaling in the gene regulatory control of energy metabolism in the
diseased heart. J Mol Cell Cardiol 2002;42:124–1257.
Goodwin GW, Ahmad F, Taegtmeyer H. Preferential oxidation of glycogen in isolated working rat heart. J Clin Invest 1996;97:1409–1416.
Lodish H, Berk A, Kaiser Cam Kreiger M, et al. Molecular cell biology. 6th ed. New York: Freeman, 2008.
P.69
Needham DM. Machina carna. The biochemistry of muscular contraction and its historical development. Cambridge: Cambridge University
Press, 1971.
Nelson DL, Cox MM. Lehninger. Principles of biochemistry. New York: Freeman, 2008.
Neubauer S. The failing heart—an engine out of fuel. N Engl J Med 2007;356:1140–1151.
Saltiel AR, Pessin JE. Insulin signaling pathways in time and space. Trends Cell Biol 2002;12:65–71.
Steinberg D. Interconvertible enzymes in adipose tissue regulated by cyclic AMP-dependent protein kinase. Adv Cyclic Nucl Res 1977;7:157–
198.
Taegtmeyer H. Metabolism—the lost child of cardiology. J Am Coll Cardiol 2000;36:1386–1388.
Taegtmeyer H. Glycogen in the heart. An expanded view. J Mol Cell Cardiol 2004;37:7–10.
Van Bilsen M, Nieuwenhoven FA, van der Vusse GJ. Metabolic remodelling of the heart: beneficial or detrimental? Cardiovasc Res
2009;81:420–428.
van der Vusse GJ, Glatz JFC, Stam HCG, Reneman RS. Fatty acid homeostasis in the normoxic and ischemic heart. Physiol Rev 1992;72:881–
940.
References
Apstein CS. Increased glycolytic substrate protection improves ischemic cardiac dysfunction and reduces injury. Am Heart J 2000;139[2 Pt
3]:S107–S114.
Arad M, Seidman CE, Seidman JG. AMP-activated protein kinase in the heart. Role during health and disease. Circ Res 2007;100:474–488.
Awan MZ, Goldspink G. Energetics of the development and maintenance of isometric tension by mammalian fast and slow muscles. J
Mechanochem Cell Motil 1972;1:97–108.
Backx P. Efficiency of cardiac muscle: thermodynamic and statistical mechanical considerations. Basic Res Cardiol 1993;88 [Suppl 2]:21–28.
Boudina S, Abel ED. Diabetic cardiomyopathy revisited. Circulation 2007;115:3213–3223.
Brown JD, Plutzky J. Peroxisome proliferator–activated receptors as transcriptional nodal points and therapeutic targets. Circulation
2007;115:518–533.
Dyck JRB, Lopaschuk GD. Malonyl CoA control of fatty acid oxidation in the ischemic heart. J Mol Cell Cardiol 2002;34:1099–1109.
Dzeja PP, Terzic A. Phosphotransfer networks and cellular energetics. J Exp Biol 2003;206:2039–2047.
Greenberg CC, Jurczak MJ, Danos AM, et al. Glycogen branches out: new perspectives on the role of glycogen metabolism in the integration
of metabolic pathways. Am J Physiol Endocrinol Metab 2006;291:1–8.
Hillgartner FB, Salati LM, Goodridge AG. Physiological and molecular mechanisms involved in nutritional regulation of fatty acid synthesis.
Physiol Rev 1995;75:47–76.
Illingworth JA, Christopher W, Ford L, et al. Regulation of myocardial energy metabolism. In: Roy PE, Harris P, eds. Recent advances in
studies on cardiac structure and metabolism, Vol. 8. Baltimore: University Park Press, 1975:271–290.
Jacobus WE. Respiratory control and the integration of heart high-energy phosphate metabolism by mitochondrial creatine kinase. Ann Rev
Physiol 1985;47:707–725.
Kammermeier H. Why do cells need phosphocreatine and a phosphocreatine shuttle? J Mol Cell Cardiol 1987;19:115–118.
Kammermeier H, Schmidt P, Jángling E. Free energy change of ATP-hydrolysis: a causal factor of early hypoxic failure of the myocardium? J
Mol Cell Cardiol 1982;14:267–277.
Katz AM, Messineo FC. Lipid-membrane interactions and the pathogenesis of ischemic damage in the myocardium. Circ Res 1981;48:1–16.
McClellan G, Weisberg A, Winegrad S. Energy transport from mitochondria to myofibril by a creatine phosphate shuttle in cardiac cells. Am J
Physiol 1983;245:C423–C427.
Paulson DJ. Carnitine deficiency-induced cardiomyopathy. Mol Cell Biochem 1998;180:33–41.

Pette D, Staron RS. Cellular and molecular diversities of mammalian skeletal muscle fibers. Rev Physiol Biochem Pharmacol 1990;116:1–76.
P.70
Prats C, Cadefau JA, Cussá R, et al. Phosphorylation-dependent translocation of glycogen synthase to a novel structure during glycogen
resynthesis. J Biol Chem 2005;280:23165–23172.
Prats C, Helge JW, Nordby P, et al. Dual regulation of muscle glycogen synthase during exercise by activation and compartmentalization. J
Biol Chem 2009;284:15692–15700.
Rauch B, Schultze B, Schultheiss HP. Alteration of the cytosolic-mitochondrial distribution of high-energy phosphates during global
myocardial ischemia may contribute to early contractile failure. Circ Res 1994;75:760–769.
Seppet EK, Eimre M, Anmann T, et al. Structure-function relationships in the regulation of energy transfer between mitochondria and ATPases
in cardiac cells. Exp Clin Cardiol 2006;11:189–194.
Shaw GB. Man and superman. A comedy and a philosophy. Cambridge: The University Press, 1903.
Suga H, Goto Y, Kawaguchi O, et al. Ventricular perspective on efficiency. Basic Res Cardiol 1993;88 [Suppl 2]:43–65.
Taegtmeyer H, Peterson MB, Ragever VV, et al. De novo alanine synthesis in isolated oxygen-deprived rabbit myocardium. J Biol Chem
1977;252:5010–5018.
Tian R, Ingwall JS. Energetic basis for reduced contractile reserve in isolated rat hearts. Am J Physiol 1996;270:H1207–H1216.
Wallimann T, Wyss M, Brdiczka D, et al. Intracellular compartmentation, structure and function of creatine kinase isoenzymes in tissues with
high and fluctuation energy demands: the “phosphocreatine circuit” for cellular energy homeostasis. Biochem J 1992;281:21–40.
Young LH. AMP-activated protein kinase conducts the ischemic stress response orchestra. Circulation 2008;117:832–840.
> Table of Contents > Part One - Structure, Biochemistry, and Biophysics > Chapter 3 - Energy Utilization (Work and Heat)
Chapter 3
Energy Utilization (Work and Heat)
Muscle research in the early 19th century centered on the then new field of thermodynamics, notably the first law, which
states that the sum of the energies in an isolated system is constant. This means that when a muscle contracts, the chemical
energy consumed by the contractile machinery is liberated as work and heat. Muscle work had been quantified since the 17th
century, and the first effort to measure heat production by isolated muscle was made by the German physicist and physiologist
Hermann von Helmholtz in 1848. However, it was not until the 1920s that A. V. Hill was able to measure muscle heat with
sufficient accuracy to obtain insights into the chemistry of muscle contraction. An early effort to measure heat production by
cardiac muscle was made in 1925 by L. N. Katz, to whom this text is dedicated, but his effort proved fruitless because the
heart generated too little heat to be recorded by the instruments of that time. Subsequent research using better
instrumentation has demonstrated important differences between the energetics of cardiac and skeletal muscle, but Hill's
studies continue to provide valuable insights regarding the contractile machinery of the heart.
A Few Terms
Isometric and Isotonic Contractions

Physiologists generally study muscles when they contract at constant length (isometric contraction) or at constant load
(isotonic contraction). In an isometric contraction, the ends of the muscle are fixed so that the muscle cannot shorten; even
though developed tension is maximal under these conditions, work (the product of force × distance) is zero because muscle
length does not change. In an isotonic contraction, the muscle is allowed to shorten while bearing a constant load. If load is
zero, the muscle shortens to its maximal extent, but no work is done because no force is developed. Work is maximal when the
muscle shortens against an intermediate load. Cardiac myocyte contraction in the beating heart is neither isometric nor
isotonic because wall stress first increases and then decreases as the cavities empty (see Chapter 11).
Preload and Afterload

The difference between a preload and an afterload depends on the time that the muscle first interacts with the load. A
preload stretches a relaxed skeletal muscle before contraction begins, whereas the muscle does not encounter an afterload
until after contraction has begun. In a linear muscle (Fig. 3-1), a weight supported by the resting muscle is a preload, while a
weight that rests on a support until after contraction has begun is an afterload.
The beating heart operates with both a preload and an afterload (Chapter 11). In the left ventricle, preload is determined by
the pressures and volumes during diastole, while afterload is
P.72
determined by the pressures and volumes after left ventricular pressure exceeds aortic pressure. Preload and afterload are
important clinically because they are major determinants of the work of the heart and influence the energetics of ventricular
performance.
Fig. 3-1: Preload and afterload. A preload is supported by a resting muscle, before it begins to contract (left). An
afterload, such as a weight resting on a support, is not encountered by the muscle until developed tension exceeds its
weight (right).
Influence of Afterload on Muscle Work

The relationship between load (P) and work can be understood by comparing a contracted muscle to a spring that, when
loaded with a 10-g weight, has increased in length by 10 cm (Fig. 3-2). If the increase in length is linearly proportional to
increasing load (Hooke's law), length will
P.73
have increased 1 cm for each gram of added load; conversely, if an initial 10-g load is decreased in 1-g steps, the spring will
shorten 1 cm for each gram removed from the load. These relationships are shown in Table 3-1, where column a describes the
stepwise decrease in load when a series of lighter weights is hung on the spring after it had been stretched by the 10-g weight.
Column b describes the extent of shortening at each lighter weight, while column c lists the work that would have been done if
the stretched spring had lifted each of the lighter weights, calculated by multiplying each new load (column a) by the distance
that the spring would have shortened at each new load (10-column b). Plotting the work in column c as a function of load
demonstrates that no work is done when the load is the same as the maximal load (P = 10 g), because the spring cannot
shorten at this load, nor when the spring is completely unloaded (P = 0) because in spite of the large extent of shortening, no
force is generated. Instead, work is maximal at the intermediate loads (Fig. 3-3).
Fig. 3-2: Effect of load on the work done by a spring that obeys Hooke's law. When the unloaded spring (A) is stretched
by a 10-g weight, its length will increase by 10 cm (B). If the 10-g load is then replaced with smaller weights, the spring
will shorten by 1 cm for each gram of load that is removed. For example, when the final load is 5 g, the spring will
shorten 5 cm (C), and if the weight is removed completely, the spring will shorten 10 cm (A). The work done when the
stretched spring in (B) is presented with a series of lighter loads is shown in Table 3-1 and Figure 3-3.
Table 3-1 Relationship between Shortening and Work Performed When a Spring Stretched by a 10-g
Load Is Allowed to Shorten at Lighter Loads
(a) Load (g) (b) Shortening (cm) (c) Work (g × cm)
10 0 0
9 1 9
8 2 16
7 3 21
6 4 24
5 5 25
4 6 24
3 7 21
2 8 16
1 9 9
0 10 0
Fig. 3-3: Work-load curve of the spring shown in Figure 3-2. Starting with the spring in its stretched state, the 10-g load
is replaced by a series of lighter loads, after which the spring is allowed to life the new load. The amount of work
performed in lifting each load is plotted on the ordinate.
P.74
Work-load relationships like that for a spring also describe the influence of load on the work performed by an activated
muscle. This similarity led early physiologists to postulate that the transition from rest to activity occurred when a resting
muscle formed new spring-like bonds. Although now known to be incorrect, this new elastic body theory is described at this
point because recognition of the fundamental error provided the foundation for our modern understanding of muscle
energetics.
New Elastic Body Theory of Muscle Contraction

Work-load curves, such as that shown in Figure 3-3, were initially explained by postulating that the transition from rest to
activity in a muscle causes new elastic bonds to be formed within the contractile machinery (Fig. 3-4). Contraction was
therefore thought to be initiated when chemical energy is used to form spring-like bonds that increase the ability of the
muscle to shorten and generate tension. This theory explained why stretching a resting muscle (Fig. 3-4, left) generates only a
small amount of tension (resting tension) and the greater stiffness of active muscle (Fig. 3-4, right). According to this theory,
activation adds a fixed amount of energy that causes the muscle to become a stiffer spring; that is, a “new elastic body.”
A key assumption of the new elastic body theory was that excitation adds a fixed amount of energy to the muscle to establish
the new bonds; for this reason, the total energy available for release by the active muscle (work + heat) should be
independent of load. This theory therefore predicted that curves relating load to total energy release would resemble Figure
3-5. At maximum load, all of the energy added during the transition from rest to activity would appear as heat because no
work is done. In an unloaded contraction, all of the energy added to establish the new cross-links in the muscle would also
appear as heat, again because no work is done. At intermediate loads,
P.75
where the release of energy as work is maximal (see Table 3-1 and Fig. 3-3), heat liberation should be minimal because the
new elastic body theory predicts that total energy release is independent of load (Fig. 3-5).
Fig. 3-4: The new elastic body theory of muscle contraction. According to this theory, the transition from the resting
state (left) to the active state (right) is caused when activation adds a fixed amount of energy to form new elastic
bonds.
Fig. 3-5: Predicted relationship between loading and total energy released as work plus heat in a muscle that contracts
according to the new elastic body theory. Because a fixed amount of energy is added to the muscle during the transition
from rest to activity, the total energy released as work and heat should be independent of load. Because the shape of
the work-load curve is similar to that of a spring (Fig. 3-3), this theory predicted that heat production would decrease
at intermediate loads to maintain a constant amount of energy release.
The new elastic body theory was widely accepted until Fenn (1923) showed that the total energy liberated by a contracting
muscle is not constant, but instead depends on load.
Fenn Effect
Fenn's decisive contribution to our understanding of muscle contraction was that the total energy released as work and heat
increases when more work is performed (Fig. 3-6). This finding, called the “Fenn effect,” proved that the energy available for
release by a contracting muscle is not
P.76
determined at the time of activation, but instead depends on load because when muscle does more work, more energy is
liberated. This finding indicated that a muscle resembles a gasoline or electric motor, where fuel consumption increases when
the motor is more heavily loaded. The Fenn effect was confirmed directly in the early 1960s when high-energy phosphate
utilization, like energy release as work plus heat, was shown to be maximal at intermediate loads, where the highest levels of
work are performed (Fig. 3-7).
Fig. 3-6: Relationship between load (P) and total energy released during contraction of frog sartorius muscle. Total
energy release is not constant, but parallels the total work performed. This increase in total energy release when a
muscle does more work is the Fenn effect.
Fig. 3-7: Influence of load (P) on high-energy phosphate breakdown (∼P) during contraction of frog sartorius muscle.
More chemical energy is used at intermediate loads, where the muscle performs more work, than at heavy or light
loads.
It is a historical curiosity that the Fenn effect had been documented in cardiac muscle almost a decade before Fenn's report.
In a paper that had been overlooked by most muscle physiologists of the time, but which Fenn cited in his 1923 article, Evans
and Matsuoka (1914–1915) reported that cardiac oxygen consumption increases when the heart does more work. This
observation led Starling, in his Linacre Lecture describing the “Law of the Heart” (1918), to equate the extra oxygen
consumption caused by an increase in cardiac work to the added fuel consumption by a motorcycle when it is ridden up a hill.
Force-Velocity Relationship
The force-velocity relationship, which plots the influence of load on the velocity of muscle shortening, provides additional
evidence that contracting muscle does not behave like a stretched spring. The new elastic body theory, which viewed muscle
as made up of parallel elastic and viscous elements (Fig. 3-8), predicted that shortening velocity would increase in a linear
manner when load is decreased (curve A, Fig. 3-9). In 1935, however, Fenn and Marsh found this relationship to be hyperbolic
(curve B, Fig. 3-9). Although this hyperbolic relationship could be explained in the context of the new elastic body theory by
assuming special characteristics for muscle elasticity and viscosity, A. V. Hill demonstrated in 1938 that the hyperbolic shape is
also obtained from measurements of work and heat, and so reflects the fundamental energetic properties of the contractile
machinery.
P.77
Fig. 3-8: Representation of an active muscle as an elastic body containing the elastic element (depicted at left as a
spring) and a viscous element (depicted at right as a “dashpot”).
Heat Liberation By Muscle

Muscle liberates three types of heat (Fig. 3-10, Table 3-2). Maintenance (resting) heat, the slow liberation of heat by resting
muscle, is not related to contraction and so is not considered further. The other two types of heat, which together represent
the activity-related heat, are generated when the muscle contracts; these are initial heat and recovery heat. Initial heat
appears during contraction, while recovery heat is liberated after the contraction has reached its peak.
Initial Heat
The first component of the heat liberated after stimulation, which for historical reasons is called initial heat, is the extra heat
(extra in that it exceeds maintenance heat) released during activation, shortening, and the generation of tension (Fig. 3-10,
Table 3-2). The following description relates key biochemical and biophysical processes to different components of initial heat.
However, this is imprecise because many processes contribute to more than one phase of heat liberation.
Fig. 3-9: Force-velocity relationships. (A, dashed line): The linear relationship predicted for a muscle in which the
elastic element obeys Hooke's law and the viscous element has Newtonian characteristics. (B, solid line): The hyperbolic
force-velocity relationship measured for frog sartorius muscle by Fenn and Marsh.
P.78
Fig. 3-10: Heat liberation (solid line) and tension development (dashed line) for a frog sartorius muscle. Maintenance
heat is liberated by the resting muscle. Initial heat is liberated during contraction, and recovery heat is liberated during
and immediately after relaxation. Recovery heat is reduced when the load is removed from the muscle prior to
relaxation. s, stimulus.
Activation Heat, Tension-Independent Heat, and Tension-Dependent Heat

The initial heat that is liberated immediately after a muscle is stimulated is often called activation heat; most of this heat is
associated with the plasma membrane ion fluxes at the time of depolarization, calcium release from the sarcoplasmic
reticulum, and conformational changes among the proteins of the thin filaments that are initiated when calcium binds to
troponin C.
In a muscle in which the interactions between the thick and thin filaments are inhibited, so that no external work can be
done, stimulation is still followed by the appearance of a small amount of initial heat. The latter, which represents tension-
independent heat, includes that portion of the activation heat not related to contractile protein interactions, along with heat
liberated during processes that end activation, including plasma membrane repolarization, reuptake of activator calcium by
the sarcoplasmic reticulum, ion transport by the sodium pump that restores the resting gradients for sodium and potassium
across the plasma membrane, and rephosphorylation of ADP by oxidative metabolism.
The tension-dependent heat, which is the difference between total initial heat and tension-independent heat, is due mainly to
interactions between myosin cross-bridges in the thick filaments and actin in the thin filaments.
Recovery Heat
Recovery heat, most of which is generated by processes that restore the state of contracted muscle to that which existed
before excitation, plays an important role in determining muscle efficiency. In skeletal muscle, much of this heat is due to the
oxidation of lactate produced during activity, whereas some of the recovery heat produced by cardiac muscle is related to ADP
rephosphorylation by the mitochondria.
P.79
Table 3-2 Energy Liberated by Muscle

Heat Some Heat-Generating Processes
I. Maintenance Resting metabolism, maintenance of ion composition, protein synthesis

(resting) heat
II. Activity- Excitation, contraction, relaxation, recovery

related heat
A. Initial Excitation, contraction, relaxation

heat
1. Plasma membrane depolarization and repolarization, Ca release and reuptake by the

Tension- sarcoplasmic reticulum, Ca binding to troponin, conformational changes in the thin filaments,
independent ion transport by the sodium pump, oxidative reactions that rephosphorylate ADP
heat
Activation Plasma membrane depolarization, Ca release by the sarcoplasmic reticulum, Ca binding to

heat troponin, conformational changes in the thin filaments
2. Contractile protein interactions

Tension-
dependent
heat
Shortening Muscle shortening, myosin–actin interactions

heat
Tension- Cross-bridge turnover

time heat
B. Recovery Oxidative reactions that rephosphorylate ADP, potential energy degraded to heat as tension
heat falls during relaxation
Work
External work Load (force) times distance (shortening)
Internal work Cross-bridge turnover [f(P,t)] stretching of series elasticity and internal viscosity
A large additional quantity of recovery heat appears when a muscle is allowed to relax while bearing a weight (Fig. 3-10). This
extra heat, which is proportional to load, is generated by the dissipation of potential energy stored by elasticities within the
contracted muscle; for this reason, recovery heat is reduced if the load is removed from the muscle before it begins to relax
(Fig. 3-10). In the heart, dissipation of potential energy stored by elasticities in the walls of the contracted ventricle makes a
major contribution to recovery heat and so reduces cardiac efficiency. However, recovery heat is decreased by closure of the
aortic and pulmonic valves, which reduces ventricular wall stress as the heart begins to relax by isolating the relaxing
ventricles from the blood under pressure in the aorta and pulmonary artery. Recovery heat in the heart is also reduced by the
decrease in wall stress that normally occurs during ejection (see Chapter 12).
Liberation of Initial Heat during Isometric Contraction

Two components of initial heat are liberated in an isometric contraction (Fig. 3-11). The first, which appears immediately
after the muscle is stimulated, is the activation heat described above (A, Fig. 3-11). Additional heat is released when
isometric tension is maintained [f(P,t)
P.80
in Fig. 3-11]; this is called tension-time heat, and is associated with internal work (Wi). The total energy released (δE) during
an isometric contraction is therefore described by the following equation:
Fig. 3-11: Liberation of initial heat (solid line) and tension development (dashed line) by a tetanized muscle
contracting under isometric conditions. The heat liberated as tension develops (A) immediately after stimulation (s) is
the activation heat, while the slower release of heat after tension reaches its peak (B) is the tension-time heat [f(P,t)].
s, stimulus.
Internal work is performed during an isometric contraction because, even though the ends of the muscle are fixed and overall
muscle length cannot decrease, the contractile proteins stretch the series elasticity and cause shape changes in the muscle
(Chapter 12). In addition to the energy associated with internal work that appears as tension-time heat, potential energy
generated by stretch of the series elasticity during the performance of internal work is converted to recovery heat at the end
of the isometric contraction, when tension is dissipated and the elasticity shortens.
Liberation of Initial Heat during Isotonic Contraction

The energetics of an isotonic contraction, where the muscle is allowed to shorten, are more complex than those of an
isometric contraction because the change in length releases two additional forms of energy. The first is the external work
performed when a loaded muscle is allowed to shorten. The second additional component of initial heat is determined by the
extent of muscle shortening, and so is called shortening heat.
In 1938, Hill found that shortening of an activated muscle is accompanied by the release of a small quantity of extra heat
(labeled ax in Fig. 3-12). This is the shortening heat, which is independent of load but proportional to the distance that the
muscle shortens (x). If the muscle is presented with different loads and allowed to shorten a fixed distance, the total amount
of shortening heat stays the same, but the rate at which this heat appears decreases with increasing load (Fig. 3-12). When
the muscle is allowed to shorten different distances with the same load, the amount of shortening heat increases
proportionately with the extent of shortening (Fig. 3-13).
To analyze these findings, Hill introduced the term a to quantify the amount of heat liberated per centimeter of shortening.
This quantity, which is constant for a given muscle, has the dimensions of a force, so that a times x (the distance the muscle
shortens) is the total amount of heat liberated during shortening:
P.81
Fig. 3-12: Liberation of initial heat by a muscle, contracting under isotonic conditions, that is allowed to shorten a
constant distance while lifting a light load (dashed line, 1) or a heavy load (dotted line, 2). The time course of initial
heat liberation by the muscle contracting under isometric conditions is also shown (solid line). The additional heat
liberated when the muscle shortens is the shortening heat (ax), which is released more slowly when the muscle lifts the
heavier load. However, the total amount of shortening heat is independent of load. s, stimulus.
Unlike the amount of shortening heat liberated in an isotonic contraction, which depends on the distance shortened but is
independent of load, the rate at which shortening heat is generated is inversely proportional to the load (see above). Because
the velocity at which the muscle shortens is also inversely proportional to load (Fig. 3-9), the rate of total energy liberation,
as work and heat, decreases at heavier loads.
The total energy released during an isotonic contraction (δE) includes external work (We), internal work (Wi), activation heat
(A), and shortening heat (ax), so that:
Unlike an isometric contraction, where only internal work is performed, most of the work in an isotonic contraction is the
external work expended to lift the load (We).
Tension-Time Heat
Improved measurements of heat liberation during the 1960s showed that both activation heat (A) and shortening heat (ax)
increase at higher loads. A simple interpretation of these findings was provided by Mommaerts (1969), who separated
activation heat into two components: A
P.82
and f(P,t). The first component, A, is the activation heat described above (Table 3-2). The second term, f(P,t), represents a
“tension-time heat” whose magnitude is proportional to the tension on the muscle (P) and the length of time (t) that tension
is maintained.
Fig. 3-13: Liberation of initial heat when a muscle contracting under isotonic conditions with a constant load is allowed
to shorten to various lengths. The shortening heat (ax) increases in direct proportion to the distance shortened, so that
when the muscle shortens 2 cm it liberates twice as much shortening heat (2ax) as when it shortens 1 cm (ax); when the
muscle shortens 3 cm the shortening heat is 3ax. s, stimulus.
The major source of tension-time heat is the slow cycling of actin-bound myosin cross-bridges in activated muscle (see Chapter
4). Because much of the heat liberated as f(P,t) is related to internal work, this term can replace Wi in Equation [3-1]. The
energy liberated during an isometric contraction is therefore:
And the equation for an isotonic contraction, (Equation [3-3]), becomes:
Tension-time heat [f(P,t)] plays an important role in the energetic consequences of changes in myosin ATPase activity caused
by alterations in this molecule such as myosin heavy chain isoform shifts (see Chapter 18). For example, chronic hemodynamic
overloading, as occurs in hypertension and aortic stenosis, favors the synthesis of a low ATPase myosin that depresses
myocardial contractility by slowing the cycling of actin-bound myosin cross-bridges. At the same time, however, this change
increases efficiency by reducing the tension-time heat [f(P,t)]. The latter represents an important adaptive mechanism
because chronically overloaded hearts are generally energy-starved.
The tension-time heat, f(P,t), is not included in the following discussion of the Hill equation. Although this simplification
introduces a minor error, it clarifies the salient relationships between force, velocity, energy liberation, and the chemistry of
the contractile proteins.
The Hill Equation

The Hill equation, which describes the energetics of muscle contraction, relates the liberation of extra energy as work and
heat to the chemistry of the contractile process. This equation is based on the energetics of isotonic contraction, where a
muscle lifts a load (P) over a distance (x). The amount of extra energy liberated during shortening appears as external work
(Px which is the same as We) and shortening heat (ax), so that:
Activation heat, which appears before the muscle begins to shorten (Fig. 3-11), is not included in Equation [3-6] because,
when f(P,t) is ignored, it is not influenced by the amount of work performed.
The rate at which extra energy is liberated during shortening can be calculated by differentiating the quantity (P + a) x with
respect to time:
Because dx/dt = shortening velocity (v), the rate of extra energy liberation is (P + a) v. This allows Equation [3-7] to be
rewritten as:
Hill found experimentally that the rate of extra energy liberation described in Equation 3-8 decreases with increasing load,
and is a direct linear function of P0 - P, the difference between
P.83
maximal isometric tension (P0) and the actual load (P) (Fig. 3-14). In an isometric contraction, where P = P0 and the distance
shortened (x) is zero, the rate of extra energy liberation is zero because there is neither shortening heat (ax) nor work (Px).
The rate of extra energy release is maximal in an unloaded isotonic contraction, where P = 0, because even though work (Px) is
zero, both the distance shortened (x) and the release of shortening heat (ax) are large. This is why, as load (P) becomes
smaller, the rate of extra energy release increases (Fig. 3-14).
Fig. 3-14: Relationship between load (force) and the rate of extra energy liberation (work plus heat) measured during
isotonic contractions.
The direct linear proportionality between the rate of extra energy liberation [(P + a) v] and the difference between maximal
isometric tension and the actual load (P0 - P) shown in Figure 3-14 can be described by the following equation:
where b is a constant of proportionality. Rearranging Equation [3-9] to put all of the constants (a, b, P0) on the right generates
the Hill equation:
Because P and v, the only two variables, are expressed as a product on the left, and all of the terms on the right are
constants, Equation [3-10] describes a hyperbola (x times y = constant). For this reason, a graph of the Hill equation that plots
the rate of extra energy liberation [(P + a) v] as an inverse function of load (P0 - P), yields a hyperbolic curve (Fig. 3-15).
Force-velocity curves based on Hill's measurements of heat and work (Fig. 3-15) have the same hyperbolic shape as the force-
velocity relationships obtained by Fenn and Marsh a decade earlier (Fig. 3-9). The similarity between these direct
measurements and those based on the Hill equation demonstrates that the hyperbolic relationship between force and velocity
is an expression of the fundamental properties of muscle chemistry!
Significance of the Hill Equation

The reader who has toiled through the concepts (and algebra) described above is entitled to ask how these features of muscle
physics—load, energetics, heat production, velocity, etc.—help in understanding the physiology of the heart. At this point, this
question is answered in general terms; the relationship between muscle physics and muscle chemistry is discussed in more
detail in Chapter 12, after the biochemistry of cardiac contraction is described.
P.84
Fig. 3-15: Hyperbolic force-velocity relationship calculated from the Hill equation (Equation [3-10]). This curve is
similar to that observed directly during contractions of frog sartorius muscle (Fig. 3-9).
The significance of Equation [3-10] was eloquently stated by Hill (1938), who provided a remarkable prediction regarding the
chemistry of the working muscle:
“The control exercised by the tension P existing in the muscle at any moment, on the rate of its energy
expenditure at that moment, may be due to some such mechanisms as the following. Imagine that the
chemical transformations associated with the state of activity in muscle occur by combination at, or by
the catalytic effect of, or perhaps by passage through, certain active points in the molecular machinery,
the number of which is determined by the tension existing in the muscle at the moment. We can imagine
that when the force in the muscle is high the affinities of more of these points are being satisfied by the
attractions they exert on one another, and that fewer of them are available to take part in chemical
transformation. When the tension is low the affinities of less of these points are being satisfied by
mutual attraction, and more of them are exposed to chemical reaction. The rate at which chemical
transformation would occur, and therefore, at which energy would be liberated, would be directly
proportional to the number of exposed affinities or catalytic groups, and so would be a linear function of
the force exerted by the muscle, increasing as the force diminished.”
Hill's statement, written when virtually nothing was known of the biochemistry of the contractile proteins and a year before
myosin was discovered to be an ATPase enzyme, explains the energetics of muscle contraction in terms of interactions
between hypothetical “active points”
P.85
within the muscle. To account for the inverse relationship between load and the rate of extra energy liberation (Fig. 3-14), Hill
postulated that the active points can exist in either of two states. In one, the active points are attached and so maintain
tension, much as when a man pulls against a rope that is firmly anchored to a post (Fig. 3-16); in the other, all of the active
points are free to liberate chemical energy and thus cycle at their maximal velocity, as when the post is pulled from the
ground, the man is able to run at his top speed (Fig. 3-17).
Fig. 3-16: Anthropomorphic depiction of an active point in a muscle during an isometric contraction, after the active
point has become attached and has developed tension, but does not liberate energy.
Fig. 3-17: Anthropomorphic depiction of an active point in a muscle allowed to shorten at zero load, where the active
point liberates energy at its maximal rate but does not develop tension.
The inverse relationship between load and the rate of extra energy release (Fig. 3-14) reflects the effects of load on the
distribution of the active points between the attached and free states. In an isometric contraction, where tension is maximal
(P = P0), all of the active points are attached in the state where they develop tension (Fig. 3-18); under these conditions, force
is maximal and the rate of energy liberation is zero. Although energy must be expended to reach this state, which accounts for
the activation heat (Fig. 3-11), additional energy need not be expended to maintain tension. In the unloaded contraction,
where P = 0, the muscle shortens at its maximal velocity because all of the active points are free to cycle at their maximal
rate (Fig. 3-19).
Efficiency and Tension-Time Heat
The preceding discussion is simplified because it assumes that once tension has developed during an isometric contraction, all
active points remain attached and do not undergo further movement. However this is not correct; instead, a slow turnover of
active points during an isometric contraction generates the tension-time heat f(P,t) described in Equations [3-4] and [3-5]; in
anthropomorphic terms this is as if the “little men” in Figure 3-18 occasionally shifted their feet.
We now know, of course, that the active points in muscle are not little men, but instead are interactions between myosin
cross-bridges and actin (Chapters 1 and 4). This explains the
P.86
tension-time heat f(P,t), which is caused when the interactions between cross-bridges on the thick filaments and actin in the
thin filaments cycle between attached and detached states. Cycling occurs at a rate proportional to the intrinsic rate of
energy release by the contractile proteins, which can be measured in vitro as myosin ATPase activity. The rate at which the
tension-time heat wastes energy during an isometric contraction is greater in muscles with high myosin ATPase activity because
the intrinsic turnover rate of the myosin cross-bridges in vivo reflects the same properties that determine myosin ATPase
activity in vitro (see Chapter 4).
Fig. 3-18: Anthropomorphic depiction of three active points in a muscle contracting under isometric conditions (P = P0,
V = 0). After the active points become attached and develop tension, they no longer liberate energy.
Fig. 3-19: Anthropomorphic depiction of three active points in a muscle contracting against zero load (P = 0, V = Vmax).
All of the active points are liberating energy at their maximal rate, but do not develop tension.
Fast muscles, which contain a high ATPase myosin, contract more efficiently when lightly loaded than slow muscles, but are
less efficient in sustaining tension (Chapter 2). Slow muscles, although they shorten less rapidly, maintain tension more
efficiently because the slower cross-bridge cycling rate wastes less energy as tension-time heat. Returning to the analogy of
the little men, a faster runner finds it easier to lift his feet from the ground, like the rapid detachment of myosin cross-bridge–
actin interactions (see Chapter 4), and so is more efficient when running freely. In contrast, a slower athlete, whose feet
remain on the ground for a longer time, is more efficient when pulling on a tethered rope. Simply stated, an athlete wearing
lead shoes runs more slowly and less efficiently, but uses less energy when called upon to pull on a tethered rope, compared to
a faster runner wearing Hermes' winged shoes.
P0 and Vmax: The Intercepts of the Force-Velocity Curve

The Hill equation predicts that the shortening velocity of an unloaded muscle (Vmax) is independent of the number of active
points; this is apparent from the analogy of the little men because one runner capable of a top speed of 10 mph will pull an
unloaded rope at the same speed as three (or any number) of such runners linked together (Figs. 3-17 and 3-19). This is why
Vmax reflects the intrinsic velocity of myosin cross-bridge turnover (which is proportional to myosin ATPase activity), and is
independent of the number of interactions between the thick and thin filaments. According to the Hill equation, the other
intercept of the force-velocity curve, the maximal force generated during an isometric contraction (P0), is determined by the
number of active points in the muscle. According to the analogy of the little men (Figs. 3-16 and 3-18), P0 depends on the
number of men pulling on the rope and not how fast each can run when load is zero. P0 is therefore independent of the
maximal rate of energy expenditure, reflecting instead the number of active interactions between the myosin cross-bridges
and actin, which in the heart is determined largely by the amount of activator calcium bound to troponin C (see Chapter 4).
P.87
The elegant analyses of the energetics of tetanized frog sartorius muscle described above cannot be carried out in the heart,
where the active state is slow in onset and changes throughout the cardiac cycle (see Chapter 6). The once prevalent view
that Vmax is a valid index of contractility has had to be abandoned because changes in calcium delivery to the contractile
proteins, the major mechanism that regulates myocardial contractility, determines the number of interactions between the
myosin cross-bridges and actin, rather than their turnover rate, and so would be expected to modify mainly P0 and have little
or no effect on Vmax. On the other hand, identification of a growing number of molecular changes that change tension
development and shortening velocity in failing hearts has increased the clinical significance of these concepts in human
disease.
Bibliography
Alpert NA, Mulieri LA, Hasenfus G. Myocardial chemo-mechanical energy transduction. In: Fozzard H, Haber E, Katz A, et
al., eds. The heart and circulation. 2nd ed. New York, NY: Raven Press, 1991: 111–128.
Bárány M. ATPase activity of myosin correlated with speed of muscle shortening. J Gen Physiol 1967;50 (6, Pt. 2):197–
206.
Curtin NA, Woledge RC. Energy changes and molecular contraction. Physiol Rev 1978;58:690–761.
Gibbs CL. Muscle mechanics and energetics: a comparative view. Cardiac energetics: sense and nonsense. Clin Exper
Pharmacol Physiol 2003;30:598–603.
Rall JA. Sense and nonsense about the Fenn effect. Am J Physiol 1982;11:H1–H6.
References
Evans CL, Matsuoka Y. The effect of the various mechanical conditions on the gaseous metabolism and efficiency of the
mammalian heart. J Physiol (Lond) 1914–1915;49:378–405.
Fenn WO. The relation between the work performed and the energy liberated in muscular contraction. J Physiol (Lond)
1923;58:373–395.
Fenn WO, Marsh BS. Muscular force at different speeds of shortening. J Physiol (Lond) 1935;85:277–297.
Hill AV. The heat of shortening and the dynamic constants of muscle. Proc R Soc (Lond) [Biol] 1938; 126:136–195.
Mommaerts WFHM. Energetics of muscular contraction. Physiol Rev 1969;49:427–508.
Starling EH. The linacre lecture on the law of the heart. London: Longmans, Green & Co., 1918.
> Table of Contents > Part One - Structure, Biochemistry, and Biophysics > Chapter 4 - The Contractile Proteins
Chapter 4
The Contractile Proteins
Cardiac contraction and relaxation depend on interactions among the six proteins listed in Table 4-1. This
chapter describes each of these proteins, how they interact as actomyosins in vitro to cause the walls of
the heart to shorten and develop tension, and how these interactions are regulated.
Myosin
Myosins, along with kinesins and dyneins, are “motor proteins” that participate in a variety of functions,
including muscular contraction, cytokinesis, endocytosis, transport of cell organelles, signal transduction,
and sensory functions such as hearing and vision. Like railway engines, the motor proteins move along
tracks; in the case of myosin the interactions are with actin filaments, whereas kinesins and dyneins move
along microtubules. All utilize chemical energy released during ATP hydrolysis to perform mechanical
work.
Cardiac myosin is a tadpole-shaped molecule made up of two heavy chains and four light chains (Fig. 4-1).
The “tail,” which is a “coiled coil” made up of two α-helical heavy chains wound around each other,
provides rigidity to the thick filaments. The heavy chains contain paired “heads” which, along with the
light chains, make up the cross-bridges that project from the thick filaments (see Chapter 1). Interactions
between the cross-bridges, which contain the ATPase site of myosin, and actin filaments release the
chemical energy that powers contraction.
Two fragments, called meromyosins, are released from myosin by the proteolytic enzymes trypsin and
chymotrypsin. The smaller fragment, light meromyosin, is derived from the tail of the molecule, while
the larger heavy meromyosin includes the head, a small portion of the tail, and the light chains (Fig. 4-2).
Further digestion of heavy meromyosin with papain, another proteolytic enzyme, removes the rest of the
tail, leaving a globular protein called heavy meromyosin subfragment 1 that includes the paired heads of
the myosin heavy chains and the four light chains. Myosin ATPase activity and the ability to interact with
actin are present in heavy meromyosin and heavy meromyosin subfragment 1.
Myosin forms filaments in vitro in which the cross-bridges project away from the center of these
aggregates (Fig. 4-3). These structures, which resemble the thick filaments of striated muscle, allow
polarized interactions between the cross-bridges and the thin filaments that enter the A-band from each
of the two adjoining I bands. In resting hearts, where the thick and thin filaments are detached, the
cross-bridges are nearly perpendicular to the long axis of the muscle (Fig. 4-4). Following activation, the
cross-bridges attach and detach from actin in a series of steps that, like the oars of a racing shell, draw
the thin filaments toward the center of the sarcomere. Because muscle volume remains virtually constant
during contraction, the lateral distance between the thick and thin filaments is modified by changes in
sarcomere length. The effects of
P.89
P.90
P.91
these changes in lattice spacing on tension development and shortening velocity are minimized by
“hinges” in the myosin molecule (Fig. 4-1) which allow the cross-bridges to maintain contact with the thin
filaments at short sarcomere lengths by varying the extent to which the myosin heads extend from the
thick filaments (Fig. 4-5).
Table 4-1 Contractile Proteins of the Heart
Approximate
Molecular
Protein Location Weight Number of Components Salient Biochemical Properties
Myosin Thick 500,000 Two heavy chains, ATP hydrolysis,

filament four light chains interacts with actin
Actin Thin 42,000 One Activates myosin

filament ATPase, interacts with
myosin
Tropomyosin Thin 70,000 Two Modulates actin–

filament myosin interaction
Troponin C Thin 17,000 One (contains Calcium binding

filament four “EF-hand”
domains)
Troponin I Thin 30,000 One Inhibits actin–myosin

filament interactions
Troponin T Thin 38,000 One Binds troponin

filament complex to the thin
filament
Fig. 4-1: Each myosin molecule contains two heavy chains and four light chains. The “tail” of the
elongated molecule is a coiled coil (two α-helical chains wound around each other) made up of the
two heavy chains; the latter continue into the paired “heads” that, along with the light chains,
form the cross-bridge. Myosin has two points of flexibility, (“hinges”) at which proteolytic cleavage
releases the meromyosins (Fig. 4-2). One lies below the heads, and the other divides the tail into
two unequal lengths.
Fig. 4-2: Myosin fragments (left) and subunits (right). Light and heavy meromyosins are fragments produced
by proteolytic digestion, whereas the heavy and light chains are subunits released by denaturing agents. Mild
denaturation of myosin releases the light chains, while stronger denaturing agents are needed to dissociate
the heavy chains.
Fig. 4-3: Myosin aggregates make up the thick filament whose “backbone”, delineated by dashed lines,
contains the tails of the individual myosin molecules. The heads of the individual myosin molecules, which
project from the long axis of the thick filament, are the cross-bridges whose polarities are opposite in the
two halves of the filament (left and right). The bare area in the center of the thick filament is devoid of
cross-bridges because of the “tail-to-tail” organization of the myosin molecules.
Fig. 4-4: In the resting heart (left), the cross-bridges project almost at right angles to the
longitudinal axis of the thick filament. In the active heart (right) the cross-bridges draw the thin
filaments toward the center of the sarcomere.
Fig. 4-5: Relationship between the thick and thin filaments at long and short sarcomere lengths.
The myosin hinges (circles) allow the tips of the cross-bridges to interact with the thin filaments
when changes in sarcomere alter the lattice spacing between the thick and thin filaments.
Heavy Chains
Adult human atria and ventricles contain different myosin heavy chain isoforms (Table 4-2); additional
isoforms are found in fetal and neonatal hearts. The human ventricle contains mainly a low ATPase β-
myosin heavy chain along with a small amount (<10%) of a higher ATPase α-myosin
P.92
heavy chain. Human atria contain two atrial myosin heavy chain isoforms, both of which differ from those
in the ventricles. Myocardial cells containing the α- and β-heavy chain isoforms are distributed in a
“mosaic” pattern where different isoforms are found in adjacent cells (see Chapter 1). Chronic overload
induces a hypertrophic response that is accompanied by an isoform shift in which the low ATPase heavy
chain isoform replaces the high ATPase isoform (see Chapter 18).
Table 4-2 Cardiac Myosin Heavy and Light Chains
I. Heavy Chains
Structure Isoform Enzymatic Activity
Atria α (atrial) High ATPase
β (atrial) Low ATPase
Ventricles α (ventricular) High ATPase

β (ventricular) Low ATPase
II. Light Chains
Structure Isoform Characteristics
Atria LC1A Essential light chain
LC2A Regulatory light chain
Ventricles LC1V Essential light chain
LC2V Regulatory light chain
LC2V* Regulatory light chain
Light Chains
Cardiac myosins contain two pairs of light chains, often referred to as essential and regulatory light
chains (Table 4-2). These names refer to the fact that extraction of the essential light chains (also called
LC1, ELC, MLC-1, or alkali-light chains) inactivates myosin, whereas the regulatory light chains (also
called LC2, RLC, MLC-2, or DTNB- or EDTA-light chains) can be removed without abolishing myosin ATPase
activity. All of these myosin light chains are members of the family of EF-hand calcium-binding proteins
that includes troponin C and calmodulin (see below).
Five different light chain isoforms are found in human atrial and ventricular myosin. Two, LC1A and LC1V,
are the essential light chains of atrial and ventricular myosin, respectively. The other three are regulatory
light chains; two are found in the ventricles (LC2V and LC2V*) and one (LC2A) in the atria. In chronically
overloaded ventricles, the atrial essential light chain LC1A replaces some of the LC1V. LC1A is also found in
developing ventricles. The human cardiac myosin light chains do not bind calcium, but provide substrates
for phosphorylations that regulate muscle shortening velocity.
Actin
Actin, a highly conserved protein found in all eukaryotic cells, received its name because of its ability to
activate myosin ATPase activity. Actin can be stabilized in vitro as a monomer, called G-actin (G =
globular), or as the filamentous F-actin polymer (F = fibrous). Both G- and F-actin contain nucleotide- and
cation-binding sites. The bound nucleotide in G-actin is ATP while F-actin ordinarily contains bound ADP;
in both, the bound cation can be either calcium or magnesium. Actin monomers, which are much smaller
than myosin (Table 4-1), are ovoid globular proteins that are ∼55 á in diameter. The polymer, a
macromolecular helix in which two chains of actin monomers are wound around one another like two
strings of beads, makes up the backbone of the thin filaments (Fig. 4-6) and represents one of the three
types of cytoskeletal filament described in Chapters 1 and 5.
Fig. 4-6: The F-actin polymer, which forms the backbone of the thin filaments, is composed of two strands of
G-actin monomers (shaded and unshaded ovals) wound around each other. The internodal distance is
approximately 385 á.
P.93
Fig. 4-7: Tropomyosin is located in the groove between the two strands of F-actin in the thin filament.
Two actin isoforms are found in the human heart, α-cardiac actin and α-skeletal actin. Adult human
ventricles contain mainly α-cardiac actin; α-skeletal actin, which is present in smaller amounts, is the
fetal isoform.
Actomyosins reconstituted in vitro from purified actin and myosin can hydrolyze ATP and undergo
physicochemical changes similar to those that occur during muscle contraction. However, these two-
protein actomyosins are not regulated by calcium because they lack the regulatory proteins (see below).
Tropomyosin
Tropomyosin is a homodimer or heterodimer made up of two α-helical peptide chains, called α- and β-
tropomyosin. Although this protein had no known function for the first 15 years after it was first purified,
its structure was of interest to physical chemists because, like the tail of the myosin molecule, it is a rigid
coiled-coil. Tropomyosin itself has no biological activity, but when incorporated into reconstituted
actomyosins, this seemingly inert protein regulates the interactions between myosin and actin. These
regulatory effects arise from shifts in the position of tropomyosin in the grooves between the two F-actin
chains in the thin filament (Fig. 4-7). The activity of the cardiac contractile proteins is inhibited when
p38-MAPK, a stress-activated mitogen-activated protein kinase (see Chapter 9), causes a regulatory serine
in tropomyosin to be dephosphorylated.
The Troponin Complex

The troponin complex includes three proteins (Table 4-1) that, along with tropomyosin, are found in the
thin filaments (Figs. 4-8 and 4-9). Troponin I received its name because its most important effect is to
inhibit actin–myosin interactions, troponin T binds the troponin complex to tropomyosin, and troponin C
contains the calcium-binding sites that participate in excitation-contraction coupling and relaxation.
Fig. 4-8: Troponin complexes are distributed at ∼400 á intervals in the thin filament.
P.94
Fig. 4-9: Cross section of the thin filament in resting muscle at the level of a troponin complex,
showing relationships between actin, tropomyosin, and the three components of the troponin
complex.
Troponin I
Changes in the strength of a labile bond linking troponin I and actin allow calcium to regulate cardiac
contraction. In the relaxed heart, where troponin C is not bound to calcium (see below), troponin I binds
tightly to actin in a conformation that causes tropomyosin to “block” the myosin-binding sites of actin.
Calcium binding to troponin C induces an allosteric change in the thin filament that loosens the bond
linking troponin I to actin, which allows active sites on actin to interact with the myosin cross-bridges. In
this way, changes in the actin-binding affinity of troponin I provide a molecular switch that recognizes a
rise in cytosolic calcium as a signal to initiate contraction (see below).
Different troponin I isoforms are found in cardiac, fast skeletal, and slow skeletal muscles; an additional
troponin I isoform has been identified in developing muscle. Cardiac troponin I contains a regulatory
serine that, when phosphorylated by protein kinase A [PKA, cyclic adenosine monophosphate (cAMP)-
dependent protein kinase], reduces the calcium sensitivity of troponin C; this posttranslational response
accelerates relaxation during sympathetic stimulation by favoring dissociation of activator calcium from
the contractile proteins (see Chapter 10). Protein kinase C (PKC) also catalyzes the phosphorylation of
cardiac troponin I, but the consequences of this posttranslational change are not well understood.
Troponin T
Troponin T, the largest of the three troponin components, mediates allosteric effects within the thin
filament that influence the calcium sensitivity of tension development. Troponin T is not a substrate for
PKA, but phosphorylation of cardiac troponin T by PKC reduces calcium sensitivity, slows cross-bridge
cycling, and decreases force generation by the cardiac contractile proteins. Isoform shifts involving
cardiac troponin T, which result from alternate splicing of the gene that encodes this protein, modify
myocardial contractility and the calcium sensitivity of the contractile process.
Ef-Hand Proteins
Troponin C is one of a family of intracellular calcium-binding EF-hand proteins that includes the myosin
light chains and calmodulin. All contain four peptide chains of ∼30 amino acids in which two α-helical
regions, designated E and F, are separated by a short non-helical sequence. The term EF hand reflects the
fact that these α-helices, which evolved by duplication and
P.95
modification of the gene that encoded the ancestral EF-hand protein, resemble the extended index finger
and thumb of a right hand (Fig. 4-10). This structure includes several oxygen-containing amino acids that
form an anionic “pocket” that binds calcium ions with very high affinity. Amino acid substitutions in these
calcium-binding regions have caused some members of this family to lose their ability to bind calcium; in
others the specificity for calcium, relative to magnesium, was lost.
Fig. 4-10: Structure of an EF-hand protein (left) showing two α-helical regions (E and F) separated
by a non-helical loop. This structure, which orients six oxygen atoms (O) to form a high-affinity
calcium-binding site, is called an “EF hand” because it resembles a right hand (right).
Several mechanisms allow different EF-hand proteins to activate muscle contraction in response to a rise
in cytosolic calcium (Fig. 4-11). In the heart and mammalian skeletal muscle,
P.96
troponin C, an EF-hand protein located on the thin filament, causes a rearrangement of the regulatory
proteins when it binds activator calcium (see below). In a few other muscles, such as the scallop
adductor, contraction is activated when calcium binds to an EF-hand myosin light chain instead of
troponin C. In vascular smooth muscle, calcium binds to calmodulin, a soluble EF-hand protein, rather
than an EF-hand protein in the myofilaments; the resulting calcium-calmodulin complex activates a
myosin light chain kinase (MLCK) that initiates contraction by phosphorylating a smooth muscle myosin
light chain. The ability of the latter, which is an EF-hand protein that has lost the ability to bind calcium,
to stimulate contraction when it is phosphorylated in response to a calcium-mediated signal illustrates
how various members of a single protein family can perform similar functions but in different ways.
Fig. 4-11: Three mechanisms by which binding of calcium (small dark circles) to EF-hand proteins (shaded
ovals) can mediate excitation-contraction coupling. In troponin-linked regulation (left) the calcium-binding
protein is incorporated into the thin filament, whereas in myosin light chain-linked regulation (center)
calcium binds to an EF-hand myosin light chain in the myosin cross-bridge (right). In vascular smooth muscle,
the calcium receptor is calmodulin, a soluble EF-hand calcium-binding protein that forms a calcium-
calmodulin complex which activates a protein kinase called myosin light chain kinase (MLCK). The latter
phosphorylates an EF-hand myosin light chain (open circles) that, although it has lost its ability to bind
calcium, still participates in calcium-mediated signaling. ADP, adenosine diphosphate; ATP, adenosine
triphosphate.
Troponin C
Troponin C is a dumbbell-shaped molecule that contains four EF-hand amino acid sequences (Fig. 4-12).
Some of these sequences bind only calcium and so are designated calcium-specific sites, while others,
called calcium-magnesium-sites, also bind magnesium. Because the ionized magnesium concentration in
most cells is much higher than that of calcium, the calcium-magnesium-sites cannot mediate signals that
are initiated by small changes in cytosolic calcium because they remain occupied by magnesium. For this
reason, only the calcium-specific sites are able to recognize a rise in cytosolic calcium concentration as a
signal to activate contraction.
Three troponin C isoforms are found in mammalian striated muscle: cardiac troponin C, and fast and slow
skeletal troponin C. All contain two calcium-specific sites (sites I and II) and two calcium-magnesium-sites
(sites III and IV). In cardiac troponin C, two negatively charged aspartic acid residues in site I are replaced
by leucine and alanine, which causes this site to lose its ability to bind calcium with high affinity. For this
reason, only site II, the other calcium-specific site, is able to serve as the physiological calcium receptor
of the cardiac contractile proteins.
Calmodulin
Calmodulin, which contains four EF-hand regions, responds to a rise in cytosolic calcium by forming a
calcium-calmodulin complex in which a hydrophobic region becomes exposed on
P.97
the surface of the calmodulin molecule. Once exposed, this hydrophobic surface can interact with
hydrophobic domains on other proteins, including members of a family of protein kinases called calcium-
calmodulin kinases (CaM kinases) that participate in cell signaling by catalyzing a variety of regulatory
phosphorylations. As noted above, MLCK, one of the CaM kinases, regulates contraction in vascular smooth
muscle.
Fig. 4-12: Schematic diagram of troponin C showing the two globular regions, each of which
contains two EF-hand sites (cross-hatched ovals), that are separated by a nine-turn α-helix. In
cardiac troponin C, sites III and IV are calcium-magnesium sites, while site I has lost its cation-
binding properties; for this reason, only site II is a physiological calcium receptor.
Actomyosins
Because the insolubility of the contractile proteins made it difficult to study the kinetics of their ATPase
enzyme, early muscle biochemists were viewed with disdain by those whose elegant studies of the
kinetics of soluble enzymes represented the “state of the art” in biochemistry. Little heed was paid to the
fact that a soluble muscle would be of little use in developing tension! Combinations of actin and myosin,
called actomyosin did, however, provide striking in vitro models of muscular contraction. In the 1940s, A.
Szent-Gyárgyi found that addition of ATP causes actomyosin threads reconstituted from myosin and F-
actin to shorten and, if a load is attached, to perform work and even generate hyperbolic force-velocity
curves. Another fascinating observation was that addition of large amounts of ATP to milky suspensions of
insoluble actomyosin particles was followed by a biphasic response: the suspension initially became clear,
after which a cloud of smaller and denser particles appeared. Low-speed centrifugation of the initial and
final suspensions revealed that ATP had decreased the volume of the pellet because the actomyosin
particles had shrunk, much as a contracting sponge might squeeze water out of itself. This experiment
demonstrated that ATP has two effects on the contractile proteins; at high concentrations ATP has a
“plasticizing” effect that dissociates myosin and actin, which is why the actomyosin suspensions initially
became clear. Actomyosin particles reappeared because the ATPase activity of the dissociated myosin had
lowered ATP concentration, which allowed myosin to recombine with actin, after which the particles
shrank when ATP energized interactions between actin and myosin analogous to those that occur during
muscular contraction.
Both of the effects of ATP described above—dissociating actin and myosin which facilitates relaxation, and
energizing the interactions between actin and myosin that cause contraction—occur in the living heart.
During diastole, the plasticizing effect of ATP inhibits the interactions between thick and thin filaments
by dissociating myosin and actin, whereas during systole, ATP that is hydrolyzed when the catalytic site of
myosin is activated by actin provides energy for contraction.
In the 1950s, when calcium was recognized to be the physiological activator of contraction, actomyosins
reconstituted from highly purified actin and myosin were found to remain active after powerful calcium
chelators lowered ionized calcium concentrations almost to zero. This surprising observation stimulated a
search for the calcium receptor of the contractile proteins that ended with the discovery of the roles of
tropomyosin and the troponin complex. Inclusion of these regulatory proteins in reconstituted
actomyosins was found to allow removal of calcium to inhibit the in vitro manifestations of contraction
described above. The ability of calcium to reverse this inhibition documented two essential features of
the physiological control of actin–myosin interactions by calcium: that the regulatory proteins are
inhibitory at low calcium concentrations, and that calcium stimulates contraction by reversing this
inhibitory effect.
P.98
Calcium as an Intracellular Messenger

Calcium mediates a number of intracellular signals, generally serving as an activator when it binds to one
of the EF-hand proteins described above. In muscle, calcium signaling represents the final step in
excitation-contraction coupling, the process by which plasma membrane depolarization activates
contraction (see Chapter 7). Calcium serves a similar role in controlling excitation-secretion coupling,
cytoplasmic streaming, and the motion of cilia and flagella.
A series of hypotheses set forth by Kretsinger (1979) explains how calcium might have become an
intracellular messenger. These center on the very low ionized calcium concentration inside eukaryotic
cells, which is about 10,000-fold lower than that in surrounding fluids. Cytosolic calcium concentration in
resting cardiac myocytes is ∼0.2 µM, whereas the calcium concentration in the blood and extracellular
space, like sea water, is in the millimolar range. Kretsinger, noting the low solubility of calcium
phosphate, suggested that calcium was initially excluded from the cytosol to allow cells to retain the
phosphate needed to synthesize ATP, nucleic acids, phosphosugars, and other key molecules. With the
evolution of energy-dependent ion pumps and exchangers that could transport calcium uphill, out of the
cytosol, the passive entry of small amounts of this cation through plasma membrane channels provided a
useful mechanism for signal transduction. In addition to serving as a chemical signal recognized by the EF-
hand proteins, the influx of positively charged calcium ions generates an electrical signal by depolarizing
the plasma membrane (see Chapter 13). Although speculative, this scenario is useful in understanding the
signaling role of calcium influx into the cytosol.
Response of the Contractile Proteins to Calcium

Calcium binding to troponin C regulates the interactions between actin and myosin by initiating a series of
cooperative interactions among the proteins of the thin filament that shift the position of tropomyosin in
the thin filament (Fig. 4-13). In the relaxed heart, where troponin C is not bound to calcium, tropomyosin
lies toward the outside of the grooves of the double-stranded F-actin polymer where it prevents active
sites on actin from interacting with the myosin cross-bridges. Calcium binding to troponin C weakens the
bond connecting troponin I to actin (see above), which shifts tropomyosin away from its “blocking”
position (Fig. 4-14) and allows actin to interact with the myosin cross-bridges (Fig. 4-15). The muscle
relaxes when removal of calcium from troponin C returns tropomyosin to its inhibitory position in the thin
filament.
Fig. 4-13: Cross section of a thin filament at a region that does not contain the troponin complex.
Tropomyosin, which lies in the groove between the two strands of actin monomers, occupies
different position at rest (left) and during activity (right).
P.99
Fig. 4-14: Cross section of a thin filament at a region containing the troponin complex in resting
(left) and active (right) muscle. At rest, the troponin complex holds the tropomyosin molecules
toward the periphery of the groove between adjacent actin strands, which prevents actin from
interacting with the myosin cross-bridges. In active muscle, calcium binding to troponin C weakens
the bond linking troponin I to actin, which rearranges the regulatory proteins so as to shift
tropomyosin deeper into the groove between the strands of actin, thereby exposing active sites on
actin for interaction with the myosin cross-bridges.
In addition to providing an “on-off” switch that initiates contraction, tropomyosin and the troponin
complex regulate the intensity of the contractile response. The length–tension relationship and Starling's
law of the heart are due in part to the ability of changes in sarcomere length to modify the calcium
sensitivity of the contractile proteins (see Chapter 6). A number of other posttranslational modifications
of the proteins of the thin filament vary the interactions between the contractile proteins; these include
troponin I phosphorylation, which facilitates relaxation by reducing the calcium sensitivity of the cardiac
contractile proteins (see above), troponin T phosphorylation which modifies force generation, and an
allosteric effect that reduces contractility when protons bind to troponin I in acidotic hearts. Isoform
shifts and mutations involving the proteins of the thin filament also have important effects on cardiac
performance (see Chapter 18).
Fig. 4-15: Cross section of a muscle at a point of potential interaction between actin and an adjacent myosin
cross-bridge at rest (left) and during activity (right). The shift in the position of the tropomyosin molecules
toward the center of the groove between adjacent actin strands in the thin filament (see Fig. 4-14) exposes
active sites on actin for interaction with the cross-bridge.
P.100
Biochemistry of Contraction
Is ATP Essential for Contraction or for Relaxation?

The seemingly contradictory effects of ATP both to dissociate actomyosin suspensions and cause them to
contract are seen in living muscle, where hydrolysis of the terminal high-energy phosphate bond of ATP
provides the energy for contraction, and severe ATP depletion causes rigor (see above). These and other
observations led early investigators to ask whether the major function of ATP is to cause muscle to
contract or to relax. One answer focused on the role of ATP in contraction:
Actin + myosin (relaxed) + ATP → actomyosin (active) + ADP + Pi
The second, which emphasized the role of ATP in relaxation, highlighted the ability of ATP to dissociate
actin and myosin:
Actomyosin (active) + ATP → actin + myosin-ATP (relaxed)
It is now clear that both mechanisms operate when the heart contracts and relaxes, and that this dual
role reflects the interaction of ATP and its hydrolytic products, ADP and Pi, with the contractile proteins.
These can be understood by examining the roles of ATP, ADP, and Pi during the sequence of reactions
between actin and myosin that occur during each cardiac cycle. The following discussion begins with the
simpler ATPase reaction of myosin alone, which is then contrasted with that of actomyosin to clarify the
interactions among the contractile proteins that cause the heart to contract and relax.
Myosin ATPase Reaction

Myosin binds ATP with high affinity to form a myosin-ATP complex in which the chemical energy of the
nucleotide remains in its terminal phosphate (step 1, Fig. 4-16). The next step, ATP hydrolysis by the
enzymatic site of myosin (step 2, Fig. 4-17), does not immediately dissociate the products; instead, an
energized complex is formed in which ADP and Pi remain attached to myosin. The slow dissociation of ADP
and Pi from myosin, which returns the latter to its basal state (step 3, Fig. 4-16) is rate-limiting for the
low ATPase activity of myosin alone.
Fig. 4-16: Simplified three-step reaction mechanism for myosin ATPase. The sequence begins at the
bottom, where ATP binds with high affinity to myosin (step 1); in this step the energy in the
terminal phosphate of ATP remains in the bond linking ATP to myosin. After hydrolysis of myosin-
bound ATP (step 2), the phosphate-bond energy remains in the myosin-bound ADP and Pi.
Dissociation of these reaction products (step 3), which is rate-limiting, releases this energy and
returns myosin to its resting state, where it can again bind ATP. ADP, adenosine diphosphate; ATP,
adenosine triphosphate.
P.101
Fig. 4-17: Simplified, four-step, reaction mechanism for actomyosin ATPase. The sequence begins
in diastole (upper left), where ATP binding to myosin has caused the muscle to relax by dissociating
myosin and actin. Hydrolysis of myosin-bound ATP (step 1) transfers the energy of the nucleotide to
the cross-bridge, but because the latter is not attached to actin, the muscle remains in a relaxed,
energized state (upper right). Interaction of the cross-bridge with actin and release of Pi (step 2)
forms an active complex in which the energy derived from ATP remains associated with the cross-
bridge. Release of ADP (step 3) leads to the formation of a low-energy rigor bond between the
cross-bridge and thin filament; expenditure of chemical energy in this step allows the muscle to
perform mechanical work. The cycle ends, and the muscle returns to its resting state, when ATP
binding to the rigor complex (step 4) dissociates the myosin cross-bridge from actin. ADP, adenosine
diphosphate; ATP, adenosine triphosphate.
Actomyosin ATPase Reaction—the Cross-Bridge Cycle

Actin increases myosin ATPase activity when it interacts with the energized myosin that is bound to ADP
and Pi (step 3 in Fig. 4-16, which is rate-limiting). Actin converts the low myosin ATPase activity to the
higher ATPase activity of actomyosin by accelerating dissociation of these reaction products from myosin.
The key reactions of the cross-bridge cycle can be summarized in four steps: two in which actin and
myosin are detached (relaxed and relaxed, energized), and two in which the myosin cross-bridges are
bound to actin (active and rigor complexes) (Figs. 4-17 and 4-18). In both of the former, where myosin is
bound either to ATP or to ADP and Pi, the muscle is relaxed because the cross-bridges are not attached to
actin. The two states in which the cross-bridges are attached to actin are quite different from one
another. In the active complex, the energy of ATP remains in the cross-bridge, whereas formation of the
rigor complex, which is in a low-energy state, expends the chemical energy of the active complex to shift
the position of the actin-attached cross-bridge.
In choosing where to enter the cross-bridge cycle, we follow the convention of physiology, which begins
the cardiac cycle in diastole. This description, therefore, starts with the heart in the relaxed state, where
binding of ATP to myosin has dissociated the cross-bridge from the thin filament. The four steps shown in
Figure 4-17 are: (1) hydrolysis of myosin-bound ATP, (2) formation of the active complex with actin and
release of Pi, (3) formation of the rigor complex and release of ADP, and (4) dissociation of actin and
myosin. Actomyosins that lack the regulatory proteins remain active even at very low calcium
concentrations, so that the cross-bridge cycle continues as long as there is an adequate supply of ATP.
P.102
Fig. 4-18: Cartoon showing the four-step reaction mechanism for actomyosin ATPase depicted in Figure 4-17.
The sequence begins at the upper left, where ATP binding to myosin has dissociated the cross-bridge from
actin. Hydrolysis of myosin-bound ATP (step 1) transfers the energy of the nucleotide to the cross-bridge,
which still is bound to ADP and Pi. Because the energized cross-bridge is not attached to actin, the muscle
remains in a relaxed state (upper right). Interaction of the cross-bridge with actin and release of Pi (step 2)
lead to the formation of the active complex in which the energy derived from ATP remains associated with
the cross-bridge. Release of ADP (step 3) shifts the position of the myosin cross-bridge and leads to formation
of a rigor bond; the chemical energy released at this step can be used to perform mechanical work. Rebinding
of ATP to the rigor complex (step 4) dissociates the myosin cross-bridge from the thin filament, which ends
the cycle and returns the muscle to its relaxed state. ADP, adenosine diphosphate; ATP, adenosine
triphosphate.
Step 1. Hydrolysis of Myosin-Bound ATP

The high ATP concentration in the resting heart forms a relaxed myosin-ATP complex that dissociates the
thick and thin filaments by (Figs. 4-17 and 4-18), after which the cross-bridge cycle proceeds when the
catalytic site of myosin hydrolyzes the myosin-bound ATP. This step transfers the energy of the terminal
phosphate bond of ATP to myosin, where the energy in the relaxed-energized complex remains in the
cross-bridge and so has not been used to perform work. As occurs during the myosin ATPase reaction (see
above), the hydrolytic products, ADP, and Pi initially remain attached to myosin.
Step 2. Formation of the Active Complex with Actin and Release of Pi

Interaction of actin with the relaxed-energized cross-bridge forms an active complex between actin and
myosin that releases Pi. Although the myosin cross-bridge is bound to the actin in this active complex, the
energy released by ADP hydrolysis has not been expended to perform
P.103
work, but remains in the actin-myosin-ADP complex. This energy is released in the next step, which
energizes the cross-bridge movement that pulls the thin filament toward the center of the sarcomere.
Step 3. Formation of the Rigor Complex with Actin and Release of ADP
The heart performs mechanical work when release of the myosin-bound ADP allows the energy in the
active complex to shift the position of the cross-bridge. This forms the rigor complex in which the myosin
cross-bridge, still firmly attached to actin, is in a low-energy state. The rigor complex is therefore very
different from the active complex. Release of ADP from actomyosin, like ADP release from myosin (see
above), is rate-limiting in the cross-bridge cycle.
Rigor complexes persist when a muscle becomes severely energy-starved. In skeletal muscle, this is the
cause of rigor mortis, while in the heart, formation of rigor bonds causes ischemic contracture. Prior to
the development of effective cardioplegic agents, which arrest the heart in a relaxed state during open
heart surgery by inhibiting actin–myosin interactions, irreversible ischemic contracture sometimes
appeared during prolonged cardiopulmonary bypass (the “stone heart syndrome”). Less dramatic, but
more common clinically, is decreased diastolic compliance caused by the formation of rigor bonds in
energy-starved ischemic and failing hearts.
Step 4. Dissociation of Actin and Myosin

Completion of the cross-bridge cycle requires that ATP again binds to myosin. This occurs in a rapid
reaction where rebinding of ATP to the rigor complex detaches the bonds linking the cross-bridge to the
thin filament. Dissociation of the rigor complex explains the “plasticizing” effect of ATP described above.
Role of ATP in Contraction and Relaxation

The reactions of the cross-bridge cycle shown in Figures 4-17 and 4-18 explain two ways that ATP
participates in contraction and relaxation: ATP hydrolysis by actin-activated myosin energizes contraction
(steps 1–3), whereas ATP binding to myosin is essential for relaxation (step 4). High ATP concentrations, in
the millimolar range, are needed to dissociate the cross-bridges from actin, whereas much lower,
micromolar, ATP levels are needed to energize myosin ATPase activity.
The ability of ATP to relax the heart does not require expenditure of phosphate bond energy, as evidenced
by the ability of non-hydrolyzable ATP analogues to dissociate actin and myosin in vitro. In contrast,
hydrolysis of the high-energy phosphate bond of ATP is required for cross-bridge cycling.
Regulation of the Cross-Bridge Cycle by Calcium

Two steps must be added to the cross-bridge cycle shown in Figures 4-17 and 4-18 to explain the role of
calcium in activating cardiac contraction. In the first, which maintains the heart in a relaxed state,
calcium is removed from troponin C, which allows tropomyosin and the troponin complex to prevent actin
from interacting with myosin (Fig. 4-19, right). This halts the cycle at the step where the myosin cross-
bridges are energized but not bound to the thin filaments. In the second, binding of calcium to troponin C
allows the cycle to resume because the regulatory proteins no longer prevent the cross-bridges from
interacting with actin.
P.104
Fig. 4-19: Role of tropomyosin and the troponin complex in regulating the cross-bridge cycle. As long as
calcium is bound to troponin C, the muscle remains in its activated state (left four diagrams), which allows
the cross-bridge cycle to continue. Calcium removal from troponin C (right), which shifts the regulatory
proteins on the thin filament to their inhibitory state, relaxes the heart by preventing actin from interacting
with the energized myosin cross-bridges. Calcium binding to troponin allows the cycle to resume. ADP,
adenosine diphosphate; ATP, adenosine triphosphate.
Cross-bridge cycling can also be modified by isoform shifts involving the myosin heavy chains that alter
the cooperative interactions among the proteins of the thin filament, and by posttranslational changes
such as myosin light chain phosphorylation.
Regulation of the Cross-Bridge Cycle by Load

Hill interpreted his observation that the rate of energy release by contracting muscle is inversely
proportional to load as evidence suggests that load determines the distribution of “active points”
between two different states: one in which they cycle freely but do not develop tension, the other where
they maintain tension but do not cycle (see Chapter 3). In the following discussion, these different states
are explained in terms of the cross-bridge cycle.
Active Points Liberating Chemical Energy

The key to understanding the biochemical basis for Hill's observation that the rate of energy liberation is
maximal when muscle contracts at zero load was provided by the finding that the maximal shortening
velocity (Vmax) of different muscles correlates closely with the ATPase activity of their myosins (Bárány,
1967). This indicated that the maximal speed at which the “little men” described in Chapter 3 can run at
zero load (Fig. 3-17) corresponds to step 3 in the cross-bridge cycle (Figs. 4-18 and 4-19), which is the
rate-limiting step where ADP is
P.105
released from myosin. As long as the cross-bridges are allowed to cycle freely, Vmax in a given muscle is
proportional to the rate of ATP hydrolysis by its purified myosin.
Active Points Maintaining Tension

The ability of increasing load to slow the rate of energy liberation (Chapter 3) can be explained because
of increasing load inhibits the breaking of the rigor bonds linking myosin and actin, which slows cross-
bridge movement along the thin filaments. Stabilization of the rigor bonds inhibits the dissociation of
actin and myosin (step 4 in the cross-bridge cycle shown in Figs. 4-18 and 4-19); this is analogous to tying
a group of the “little men” to a post so as to prevent their feet from moving (Fig. 3-16). This analogy
helps to explain why the determinants of the rate of cross-bridge cycling are not the same as those that
determine the strength of the rigor bonds, and why maximal shortening velocity at zero load (Vmax) is not
correlated with maximum isometric tension (P0). The number of active cross-bridges does not influence
maximum shortening velocity at zero load for the same reason that changing the number of “little men”
does not modify the velocity at which the group can run when the rope breaks (Fig. 3-19).
The Force-Velocity Curve

The preceding discussion explains why the intercepts of the force-velocity curve (Fig. 3-15) are
determined by different properties of the contractile proteins. The maximal isometric tension (P0), which
is determined by the number of rigor bonds between the cross-bridges and actin, reflects the number of
troponin C molecules bound to calcium. For this reason, the amount of calcium released during
excitation-contraction coupling and the calcium affinity of troponin are the major determinants of P0.
Maximal shortening velocity (Vmax), on the other hand, is determined by the turnover rate of cross-bridge
cycling and so is independent of the number of active cross-bridges as long as load is zero. Recruiting
additional actin–myosin interactions increases P0, but cannot increase shortening velocity in an unloaded
muscle. It is clear that two independent mechanisms, the number of active cross-bridges and their
turnover rate, determine myocardial contractility, which can be defined as the ability of the heart to do
work at any given rest length (see Chapter 10). One needs to consider only the different body types of
world-class weight lifters and sprinters to realize that these need not be closely related to each other.
Bibliography
Alberts B, Johnson A, Lewis J, et al. Molecular biology of the cell. 5th ed. New York, NY: Garland,
2008.
Bers DM. Excitation-contraction coupling and cardiac contractile force. 2nd ed. Dordrecht: Kluwer,
2001.
de Tombe PP, Solaro RJ. Integration of cardiac myofilament activity and regulation with pathways
signaling hypertrophy and failure. Ann Biomed Eng 2000;28:991–1001.
Gunning P, O'Neill G, Hardeman E. Tropomyosin-based regulation of the actin cytoskeleton in time
and space. Physiol Rev 2008;88:1–35.
Huxley HE, Hanson J. The molecular basis of contraction in cross-striated muscles. In: Bourne GH.
Structure and function of muscle. Vol 1. Structure. New York, NY: Acad Press, 1960:183–227.
P.106
Kawai M, Saeki Y, Zhao Y. Crossbridge scheme and the kinetic constants of elementary steps deduced
from chemically skinned papillary and trabecular muscles of the ferret. Circ Res 1993;73:35–50.
Kobayashi T, Jin L, de Tombe PP. Cardiac thin filament regulation. Pflugers Arch Eur J Physiol 2008;
457:37–46.
Kostyukova AS. Tropomodulins and tropomodulin/tropomyosin interactions. Cell Mol Life Sci 2008;
65(4):563–569.
Kretsinger RH. The informational role of calcium in the cytosol. Adv Cyclic Nucl Res 1979;11:1–26.
Langer GA, ed. The Myocardium. San Diego, CA: Academic Press, 1979.
Lodish H, Berk A, Kaiser Cam Kreiger M, et al. Molecular Cell Biology. 6th ed. New York, NY:
Freeman, 2008.
Page E, Fozzard HA, Solaro RJ, eds. Handbook of Physiology Section 2: The cardiovascular system.
Vol I. The heart. New York, NY: Oxford, 2002.
Schiaffino S, Reggiani C. Molecular diversity of myofibrillar proteins: gene regulation and molecular
significance. Physiol Rev 1996;76:371–423.
Taylor EW. Mechanism and energetics of actomyosin ATPase. In: Fozzard H, Haber E, Katz AM, et al.,
eds. The heart and circulation. 2nd ed. New York, NY: Raven Press, 1991:1281–1294.
References
Bárány M. ATPase activity of myosin correlated with speed of muscle shortening. J Gen Physiol
1967;50 (6, Pt. 2):197–206.
Kretsinger R. Evolution of the informational role of calcium in eukaryotes. In: Wasserman RH,
Corradino RA, Carafoli E, et al., eds. Calcium-binding proteins and calcium function. New York, NY:
North Holland, 1977:63–72.
> Table of Contents > Part One - Structure, Biochemistry, and Biophysics > Chapter 5 - The Cytoskeleton
Chapter 5
The Cytoskeleton
Come to me now, you muses who live on Olympus for you are goddesses, are
everywhere and know all things, while we only hear of glory but do not truly know
it. Who were the leaders of the Greeks, who the rulers? I could not tell the names
or number of all the men, even if I had ten tongues, ten mouths, an unwearied
voice, and a heart of bronze, unless you, Olympian muses, daughters of Zeus who
bear the aegis, recall for me the men who came to Troy, so now I will recount the
leaders of the ships, and all their numbers.
—Homer (Iliad. Book 2, 484–493. tr. P.B. Katz)
L ike Book 2 of Homer's Iliad, this chapter includes a list of the many cytoskeletal proteins found in
cardiac myocytes. As is true for any list, the role and significance of each entry can be difficult to
appreciate. However, discoveries made over the past decade have made it clear that cytoskeletal proteins
are important in the pathogenesis of human disease; they generate and transmit maladaptive signals that
contribute to the progressive deterioration of failing hearts, and mutations in many of these proteins have
been identified as causes of cardiomyopathies and sudden cardiac death. The following pages are
intended to provide a brief introduction to many of these proteins, along with a framework by which
these important molecules can be organized. Because the overall goals of this introductory text require
that the descriptions of each protein be brief, the reader who seeks more detail is referred to the
authoritative reviews cited in the bibliography.
Cardiac myocytes are not simply fluid-filled bags in which enzymes and cell organelles like myofilaments,
the sarcoplasmic reticulum, nuclei, and mitochondria float freely in the cytosol; instead, a complex
cellular architecture is maintained by a protein framework called the cytoskeleton. Like the girders in a
modern building, these structural proteins hold cell constituents in place, organize functionally related
proteins, and facilitate mechanical interactions with adjacent cells and the extracellular matrix. In
addition to these structural functions, the cytoskeleton participates in cell signaling, and makes a major
contribution to the signal transduction pathways that modify the size and shape of the heart. The
cytoskeleton, therefore, is not simply a system of girders that maintains cell shape; because of its role in
cell signaling, the cytoskeleton is more accurately viewed as a structural framework that also serves as a
phone system!
In cardiac myocytes, the cytoskeleton organizes spatial relationships among intracellular structures,
transmits tension developed by the contractile proteins from one sarcomere to another and to the cell
surface, and links cells to other cells and to the extracellular matrix. The cytoskeleton also helps the
heart adapt its architecture in response to cell deformation and changes in mechanical stress. By
modifying the size, shape, and composition of the heart, cytoskeletal signaling serves a key role in
adapting form to function. This is seen in the activation of specific signaling pathways by various types of
cytoskeletal deformation, which contributes to the appearance of different
P.108
phenotypes of cardiac hypertrophy and helps explain why pressure overload causes concentric
hypertrophy (seen in aortic stenosis where systolic stress is increased), and volume overload causes
eccentric hypertrophy (as occurs in aortic insufficiency where diastolic stress is increased). These
proliferative responses are especially important in failing hearts, where cytoskeletal signaling contributes
to the different types of overload-induced hypertrophy. Although initially beneficial, these proliferative
responses generally become maladaptive when the overload is sustained. Cytoskeletal signaling plays a
role in determining prognosis in systolic heart failure, where deleterious transcriptional signaling
contributes to the progressive ventricular dilatation, called remodeling, that shortens survival (see
Chapter 18). For these reasons, cytoskeletal proteins can serve as targets for therapy directed to
modifying maladaptive growth responses in diseased hearts. Drugs that can modify the behavior of these
proteins are also likely to help patients with familial cardiomyopathies, many of which are caused by
mutations in cytoskeletal proteins.
The mechanical and signaling functions of the cytoskeleton are made possible by interactions among
dozens of proteins. Some of these proteins are largely structural, while other proteins associated with the
cytoskeleton, including protein kinases, phosphatases, and transcription factors, function mainly in signal
transduction. However, many cytoskeletal proteins serve both mechanical and signaling functions, so that
the designations “structural” and “signaling” in the following discussion are arbitrary. In many ways, these
proteins are like the front wheels of an automobile, which both hold the vehicle to the road and
participate in steering.
Cytoskeletal Filaments
The cytoskeleton contains three types of intracellular fiber (Table 5-1). Actin microfilaments are found in
two structures: the thin filaments of the sarcomeres (see Chapter 4) and microfilaments. The latter
connect the contractile proteins, plasma membrane, and other cytoskeletal elements to one another,
connect the cytoskeleton to the extracellular matrix, and anchor cells to each
P.109
other. Microtubules, which are made up mainly of tubulin, transport organelles and vesicular structures
within cardiac myocytes and participate in the mitosis occasionally seen in these terminally differentiated
cells (see Chapter 9). Specialized arrays of microtubules participate in the movements of cilia and
flagella, but these are not found in adult cardiac myocytes. The heart contains two types of intermediate
filament. Class III intermediate filaments contain desmin, and form strong rivet-like structures, called
desmosomes, which help maintain cell architecture, link cells to one another, and attach cells to the
extracellular matrix. Class V intermediate filaments, which contain lamin, support the nuclear
membrane. Unlike microfilaments and microtubules, intermediate filaments do not participate in cell
motion or intracellular transport. However, all of these cytoskeletal fibers participate in cell signaling.
Table 5-1 Cytoskeletal Filaments
Major
Filament Proteins Some Functions
Microfilaments Actin
Thin Sarcomeric contraction (with myosin)

filaments
Cortical Maintain cell shape, provide linkages to the cell surface,

filaments organize cell contents, intracellular transport (with
myosin)
Microtubules Tubulin Intracellular transport (with kinesin and dynein), mitosis
Intermediate Filaments
Class III Desmin Maintains cell shape, provides linkages to the cell surface
Class V Lamin, Strengthen nuclear membrane, support nuclear structure

emerin
Cytoskeletal Proteins Associated with the Myofilaments

Our view of the sarcomere has changed dramatically since the 1950s, when the sliding filament hypothesis
for muscular contraction was first proposed. At that time, sarcomeres were thought to be made up of two
proteins, actin and myosin; in the 1970s, after the roles of tropomyosin and the troponin complex were
recognized, this number increased to six (see Chapter 4). However, the classical studies of Huxley and
Hanson (1960) showed that although the A-bands disappeared when myosin was extracted from muscle
fibers, and that subsequent extraction of actin removed the I-bands, removal of these proteins did not
cause the fibers to fall apart. Instead, the sarcomeres remained intact and Z-lines retained their normal
alignment. For almost 30 years, these observations, while not forgotten, were generally overlooked. It
was not until the 1990s that the role of the cytoskeletal proteins, which maintained sarcomere structure
after the contractile proteins had been removed, came to be appreciated. Initially, the cytoskeleton was
believed simply to provide mechanical support that helped maintain sarcomere structure and transmit the
tension developed by the contractile proteins to adjacent sarcomeres, other myocytes, and the
extracellular matrix. It is now clear, however, that cytoskeletal proteins also regulate the lengths of the
thick and thin filaments and serve a number of vital signaling functions. Even more important clinically
has been the recognition that disturbances of these signaling functions are important in human disease.
Cytoskeletal Proteins of the Thick Filaments

The thick filaments, once thought to be composed entirely of myosin, are now known to contain a number
of cytoskeletal proteins. Some are incorporated into these filaments where they bind directly to myosin.
An even larger number of cytoskeletal proteins form aggregates that link the thick filaments to nearby
structures and participate in cell signaling. The resulting complexity makes it difficult to create visual
images that describe the interactions among these proteins; for this reason, the following descriptions of
these interactions use both diagrams and lists.
Titin, a huge protein with a molecular weight of ∼3,000,000 that extends from the Z-lines to the center
of the thick filament (Fig. 5-1), supports sarcomeric structure and contributes to the high resting stiffness
of the myocardium (see Chapter 6). The regions of titin that are incorporated into the thick filaments add
rigidity to these structures, while a more compliant region in the I-bands contributes to the heart's
elasticity. Phosphorylation of titin by PKA reduces sarcomere stiffness, and so facilitates ventricular filling
when β-adrenergic activity is increased, as during exercise.
P.110
Fig. 5-1: Cytoskeletal proteins of the thick filaments. Center: Thick filaments, made up of a myosin
backbone, include the giant protein titin, which extends from the Z-line to the center of the thick filament,
and telethonin (t-cap), which connects the thick filaments to the Z-line. The portions of the titin molecule
that lie within the A-band are rigid, while the regions in the I-band are more elastic. Above: Enlarged
depiction of some of the cytoskeletal proteins found in the M-bands. Below: List of some of the many
cytoskeletal proteins found in the M-band.
Two titin isoforms, N2B and N2BA that differ in the flexible region of the molecule, are found in adult
human ventricles. N2BA, which is normally the dominant isoform, is more compliant than N2B. Heart
failure patients with dilated left ventricles (“systolic heart failure”) overexpress the more compliant N2BA
isoform, whereas those whose left ventricles are stiff and have smaller cavities (“diastolic heart failure”)
have an increased content of the stiffer N2B isoform. Mutations in titin are associated with both dilated
and hypertrophic cardiomyopathies (see Chapter 18).
P.111
Titin forms transverse fibers with a number of other cytoskeletal proteins in the M-bands at the center of
thick filaments (Fig. 5-1). These include myosin binding protein C (C protein), M-protein, and myomesin,
which provide lateral mechanical stability to the sarcomere, and muscle creatine kinase, adenylate
kinase, and phosphofructokinase, which may participate in length-dependent modifications of energy
production. Obscurin, another titin-linked M-band protein, connects the sarcomeres to the sarcoplasmic
reticulum through small ankyrin-1 (see below). Ubiquitin and the ubiquitin-like modifier SUMO3 regulate
protein degradation. Additional M-band proteins participate in cell signaling; these include calmodulin,
myofibrillogenesis regulator 1, four-and-a-half LIM protein (FHL), the transcription factors p53, muscle-
specific ring finger protein (MURF-1), and the two zinc-finger proteins (nbr1 and p62) (see Chapter 9).
The M-band proteins provide an example of a “hot spot” where interactions between structural and
regulatory proteins generate proliferative responses.
The interactions among the cytoskeletal proteins shown in Figure 5-1 suggest several ways that M-band
proteins could mediate specific responses to cell deformation. For example, MURF, FHL, and the two zinc-
finger proteins (nbr1 and p62) regulate gene transcription and protein synthesis, which could allow these
cytoskeletal proteins to modify the heart's size, shape, and composition in response to abnormal
mechanical stresses. Interactions of the M-band proteins with obscurin, small ankyrin-1, and the
sarcoplasmic reticulum might contribute to length-dependent changes in calcium release by these
internal membranes (see Chapter 6).
Different responses of cytoskeletal proteins oriented parallel and perpendicular to the long axis of the
sarcomere probably explain why transverse and longitudinal stretch of cardiac myocytes evokes different
focal adhesion kinase responses (Senyo et al., 2007), and activation of different mitogen-activated
protein kinases when cardiac myocytes are stretched in systole and in diastole (Yamamoto et al., 2001).
Activation of different cytoskeletal signal transduction pathways by increased systolic and diastolic stress
can also explain how chronic pressure overload causes concentric hypertrophy, as occurs when aortic
stenosis increases systolic stress, and how volume overload causes eccentric hypertrophy, as occurs when
aortic insufficiency increases stress during diastole (Katz and Konstam, 2009).
Cytoskeletal Proteins of the Thin Filaments

Actin can be viewed both as part of the cytoskeleton and as a contractile protein. As the former,
sometimes called cortical actin, actin is found in the microfilaments (see above), while as sarcomeric
actin this protein makes up the backbone of the thin filaments (see Chapter 4). Two cytoskeletal proteins
cap the ends of the thin filaments; tropomodulin, which is found at the ends that interdigitate with the
A-bands, and cap Z (β-actinin), which anchors the other ends of the thin filaments in the Z-lines (Fig. 5-
2). In addition to linking the thin filaments to other proteins, tropomodulin and cap-Z regulate thin
filament length and participate in proliferative signaling. Nebulette, a cytoskeletal protein that runs
alongside the thin filaments from the Z-line into the I-bands, interacts with a variety of signaling proteins,
most of which are found at the Z-lines.
Cytoskeletal Proteins of the Z-Lines

The Z-lines (or Z-discs) (see Chapter 1) are rigid structures in which the thin filaments of adjacent
sarcomeres are interwoven with a number of cytoskeletal proteins that transmit tension developed by the
contractile proteins to the fascia adherens of the intercalated discs that link
P.112
adjacent cells (Fig. 5-3). The Z-lines also anchor a number of proteins that are linked to cytoskeletal
filaments that connect to the plasma membrane, sarcoplasmic reticulum, and other cellular structures
(see below). Rigid bands made up of desmin and α-actinin surround the Z-lines in structures called
costameres that resemble the staves of a barrel. Desmin, the major protein of the costameres, is one of a
family of intermediate filament proteins that includes vimentin, which supports connective tissue cells,
keratins found in epithelial cells, neurofilaments that help organize axon structure, glial fibrillary acidic
protein in astrocytes and other glial cells, and lamins that support the inner surface of nuclear
membranes. Desmin links the Z-lines to intermediate filaments that connect to desmosomes in the
intercalated disc (see Chapter 1), and so participates in transmitting tension between adjacent cardiac
myocytes. As desmin lacks signaling domains, this protein probably does not directly modify cell signaling;
however, by transmitting tension to other cytoskeletal proteins, desmin can play an indirect role in signal
transduction.
Fig. 5-2: Cytoskeletal proteins of the thin filaments. Below: Thin filaments have been added to the diagram
in Figure 5-1. Above: Enlarged view of a thin filament showing the two “capping” proteins tropomodulin and
cap Z. The latter, along with α-actinin and nebulette, connects the thin filaments to the Z-lines. Titin
extends from the thick filaments into the ends of the thin filaments that are bound into the Z-lines.
The Z-lines are connected to the plasma membrane by dystrophin, g-filamin, myotilin, and actin
microfilaments, while obscurin, in concert with small ankyrin-1, connects the Z-lines to the sarcoplasmic
reticulum (see below). Other Z-line cytoskeletal proteins whose functions are largely structural include
titin, telethonin, cap Z, tropomodulin, and nebulette (see above). Z-line proteins
P.113
that serve signaling as well as structural functions include several LIM proteins (see below) and the
paladin-myopalladin complex, which forms a signaling scaffold.
Cytoskeletal proteins at the Z-line that participate in proliferative signaling include protein kinases, a
tumor suppressor called myopodin, calcineurin, a calcineurin-binding protein called calsarcin
(calcineurin-associated sarcomeric protein), and S100, a calcium binding protein that regulates
myogenesis (Fig. 5-3). Protein degradation is regulated by arginine-binding protein 2 (ArgBP2), which is
found in intracellular signaling scaffolds, and atrogin, a ubiquitin ligand.
P.114
Muscle ankyrin repeat proteins (MARPs), which include CARP (cardiac ankyrin repeat protein) and DARP
(diabetes-related ankyrin repeat protein), link cytoskeletal elements to one another and connect
myocardial cells at the intercalated discs; these proteins also regulate gene transcription. MinK, a
regulatory subunit of a plasma membrane voltage-gated potassium channel (see Chapter 13), links the Z-
lines to the t-tubules and may help integrate signals generated by mechanical stretch and changing
electrical potential across the plasma and t-tubular membranes.
Fig. 5-3: Cytoskeletal proteins of the Z-lines. Below: Depiction of major cytoskeletal proteins whose
functions are primarily structural. Above: List and major functions of some of the cytoskeletal proteins found
in the Z-line.
The LIM/PDZ proteins, a family of cytoskeletal proteins that serve both structural and signaling functions,
contain one or more amino acid sequences called LIM (named for lin-11 and mec-3 found in the
roundworm C. elegans and Isl-1, an insulin binding protein found in rats). LIM proteins contain a double
zinc finger and a cysteine-rich sequence, and a sequence called PDZ (named for PSD-95/SAP90, a post-
synaptic protein, Discs-large, a Drosophila junction protein, and ZO-1, a tight junction protein). In
addition to their structural roles, some of the LIM/PDZ proteins incorporated into membrane-associated
protein modules provide “scaffolds” for proliferative signaling (see Chapter 9). LIM/PDZ proteins found at
the Z-line include actinin-associated LIM protein (ALP) which links to the α-actinin and participates in
stretch-activated growth responses, muscle-specific LIM (mlp), and zyxin which provide linkages to the
plasma membrane and respond to sarcomere stretch; zyxin can also inhibit apoptosis. Other Z-line
LIM/PDZ proteins are cypher/ZASP, an enigma protein that regulates protein phosphorylations, and four-
and-a-half LIM protein (FHL) that regulates responses to cyclic AMP.
Membrane-Associated Cytoskeletal Proteins

Several cytoskeletal proteins link intracellular structures to intrinsic plasma membrane proteins (Fig. 5-4).
Like those associated with the sarcomeres, most of these cytoskeletal proteins serve both structural and
signaling functions.
Ankyrin
Ankyrins, which contain 24 copies of a 33 amino acid sequence called an “ankyrin repeat,” link a variety
of plasma membrane proteins, including receptors, cell adhesion molecules, ion channels, ion pumps, and
ion exchangers, to the actin/spectrin cytoskeleton and to each other. Small ankyrin-1, one member of
this family, links the sarcoplasmic reticulum to obscurin in the M-bands and Z-lines. Ankyrins contain a
domain that binds to membrane proteins, a spectrin-binding domain that binds to the actin-spectrin
cytoskeleton, a death domain (see Chapter 9), and a C-terminal domain that binds to obscurin and serves
a variety of regulatory functions. The membrane-binding domain forms linkages with a variety of
membrane proteins, often in multiprotein aggregates that participate in proliferative signaling. The death
domain may play a role in apoptosis.
Ankyrins organize and integrate membrane function by creating microdomains that bring functionally
related proteins in proximity to one another. For example, by co-localizing voltage-gated sodium
channels, the sodium pump ATPase, and the sodium/calcium exchanger, ankyrins help coordinate sodium
fluxes across the plasma membrane. Ankyrins also organize interactions between plasma membrane
calcium channels (“dihydropyridine receptors”) and the intracellular calcium release channels in the
sarcoplasmic reticulum (“ryanodine receptors”) (see Chapter 7).
P.115
Fig. 5-4: Cytoskeletal proteins of the Z-lines and their relationship to membrane-associated cytoskeletal
proteins. Several Z-line proteins connect the costameres to plasma membrane proteins; these include γ-
filamin, which binds to integrins and the dystrophin complex in the plasma membrane, and obscurin, which
binds the costameres to the sarcoplasmic reticulum. γ-Filamin in a complex with myotilin binds titin to
integrins, while actin microfilaments link α-actinin in the costameres to spectrin and ankyrin. SERCA,
sarcoplasmic reticulum calcium pump ATPase.
Spectrin
Spectrins, which are tetrameric proteins made up of 2 α-subunits and 2 β-subunits, form a network on the
cytosolic side of the plasma membrane that is linked to actin microfilaments and ankyrins (Fig. 5-4).
Spectrins also participate in cell signaling; some contain an EF-hand calcium-binding domain, others an
amino acid sequence similar to that found in regulatory tyrosine kinases. Compartmentalization within the
spectrin cytoskeleton is partly responsible for the ankyrin-mediated organization of membrane proteins
described above.
Obscurin
Obscurin connects titin and nebulette at the Z-line, and titin in the M-line, to small ankyrin-1 that is
bound to the sarcoplasmic reticulum (Fig. 5-4). In addition to this structural function, obscurin regulates
sarcomere formation and participates in signal transduction, for example,
P.116
through interactions with calmodulin and Rho, a monomeric G protein. The name obscurin was chosen
because of the large size of this protein, its complex structure, and the small amounts found in muscle.
Filamin
Filamin links actin microfilaments to the sarcoglycan complex in the plasma membrane (see below); its
participation in cell signaling is evidenced by interactions with focal adhesion molecules and a LIM protein
called migfilin. γ-Filamin, the cardiac isoform, links the Z-lines to integrins and sarcoglycans in the
plasma membrane.
Myotilin
Myotilin, which binds sarcomeric actin, α-actinin, and γ-filamin in the Z-line, is one of the proteins that
link the cytoskeleton to integrins in the plasma membrane. In addition to its structural role, myotilin
interacts with cytoskeletal signaling proteins including calsarcin, a calcineurin regulator.
Cytoskeletal Proteins in the Intercalated Discs

Cardiac myocytes are connected to one another by intercalated discs (see Chapter 1). These composite
structures contain gap junctions that allow electrical current to flow across the intercalated disc (see
Chapter 13), and two structures that provide mechanical linkages between adjacent cardiac myocytes:
the fascia adherens (also called adherens junctions or focal adhesions) and desmosomes.
Fascia Adherens
The fascia adherens links actin microfilaments of adjacent cells through connections between similar
members of a protein family, called cadherins, which are held together by calcium-dependent linkages
(Fig. 5-5A). Complexes made up of α-, β-, and γ-catenins (the latter is also called plakoglobin) function as
anchoring proteins that link the intracellular domains of the cadherins in the fascia adherens to α-actinin
and vinculin, which connect the fascia adherens to the actin microfilaments. Catenins, which are
members of a family of signaling proteins called Wnt, participate in the G protein-mediated signaling
pathways that play an important role in overload-induced cardiac hypertrophy.
Desmosomes
Desmosomes, which are linked to intermediate filaments, provide strong, rivet-like, connections between
cells. Unlike the fascia adherens, in which the cadherins on adjoining cells are similar, two different
cadherins, called desmoglein and desmocollin, form a heterotypic connection in desmosomes (Fig. 5-5B).
The cadherins in the desmosomes are bound to plakoglobin (g-catenin) and plakophilin, which are
members of the Wnt family of proteins. A large protein called desmoplakin connects this multiprotein
complex to desmin in the intermediate filaments.
P.117
P.118
A clinical syndrome called arrhythmogenic right ventricular cardiomyopathy (see Chapter 16), which can
be caused by mutations in several of the genes that encode proteins in the outflow tract of the right
ventricle, can be viewed as a disease of the desmosomes (Dokuparti et al., 2005).
Fig. 5-5: Cytoskeletal proteins of the intercalated disc. A: In the fascia adherens, cadherins link adjacent
cells to actin microfilaments through α-, β-, and γ-catenins, α-actinin, and vinculin. B: Desmoglein and
desmocollin, the cadherins in the desmosomes, are linked to intermediate filaments through a complex that
includes plakoglobin (γ-catenin), plakophilin, and desmoplakin.
Cytoskeletal Proteins that Link Cells to the Extracellular Matrix

The cytoskeletal proteins described above differ from those that link cells to the extracellular matrix.
The latter include the dystrophin glycoprotein complex and integrins, both of which connect actin
microfilaments to extracellular matrix proteins. γ-Filamin, which also links actin microfilaments to
plasma membrane proteins, interacts with titin and myotilin in the Z-lines.
The Dystrophin Glycoprotein Complex

The plasma membrane is anchored to extracellular matrix proteins by the dystrophin glycoprotein
complex, which is connected to actin microfilaments (Fig. 5-6). Dystroglycans, which span the plasma
membrane and bind extracellular matrix proteins like laminin and fibronectin, are connected within the
cytosol to dystrobrevin and syntrophins in a multiprotein aggregate that links dystrophin to the plasma
membrane. Within the latter, dystroglycan forms a complex with a family of glycoproteins called
sarcoglycans and to sarcospan.
Dystroglycan, along with dystrophin, forms a complex with dystrobrevin and syntrophins that participates
in several signaling systems. Dystrobrevin, a substrate for regulatory phosphorylations, contains a
calcium-binding domain and, along with caveolin-3, regulates nitric oxide production by nitric oxide
synthetase (NOS). Caveolin-3, which, like ankyrin, can integrate the activities of functionally related
signaling molecules, interacts with dystroglycan in the dystrophin glycoprotein complex to mediate signal
transduction across the plasma membrane. Mutations in dystrophin, dystroglycan, and other proteins in
this complex can cause dilated cardiomyopathies by destabilizing cell-matrix linkages, and dystrobrevin
mutations have been implicated in noncompaction of the left ventricular wall. Disruption of dystrophin
linkages can exacerbate progressive dilatation in failing hearts.
Integrins
Integrins, a family of cell adhesion molecules that link cells to the extracellular matrix, form
heterodimers made up of different α- and β-subunits whose properties determine the binding specificity
and signaling effects of this cell adhesion molecule. Both the α- and β-subunits bind to extracellular
matrix proteins, but only the β-subunits are linked to the cytoskeleton. The intracellular domains of the
integrins bind to actin microfilaments through adaptor proteins that include α-actinin, vinculin, tensin,
and talin; paxillin, another intracellular protein that can bind to this complex, is a substrate for
regulatory phosphorylations that influence cell adhesion. Integrins bind to extracellular matrix proteins,
including fibronectin, laminin, and vitronectin, which along with proteoglycans, such as heparan, link the
integrins to collagen (Fig. 5-7A).
P.119
Fig. 5-6: Cytoskeletal proteins of the dystrophin glycoprotein complex. Dystrophin links actin microfilaments
to the plasma membrane through a protein complex that includes syntrophins α and β and dystrobrevin. The
latter are bound to the plasma membrane by dystroglycans α and β, six isoforms of sarcoglycan, caveolin-3,
and sarcospan in a complex that binds to laminin and collagen in the extracellular matrix. This complex also
participates in cell signaling through interactions with caveolin-3 and nitric oxide synthetase (NOS).
In addition to their structural role, integrins participate in a variety of proliferative signaling pathways
(Fig. 5-7B). Although integrins lack enzymatic activity, these cell adhesion molecules allow mechanical
stresses to regulate mitogen-activated protein (MAP) kinase pathways (see Chapter 9). Many of these
proliferative effects are mediated by small G-proteins like Rho, Ras, and Ras, and by non-receptor protein
kinases that are linked to the cytoskeleton. The latter include the tyrosine kinases focal adhesion-
associated kinases (FAKs), proline-rich tyrosine kinase 2 (PYK2), and c-Src-kinase (CSK), and
serine/threonine kinases such as integrin-linked kinase (ILK), p21-activated kinase (PAK), and protein
kinase Cá (PKCá). Many of these signaling pathways mediate maladaptive growth responses that
contribute to the poor prognosis in patients with heart failure (see Chapter 18). However, some integrin-
linked signaling proteins play a protective role in overloaded hearts; these include a pathway linked to
melusin which, through interactions with muscle LIM domain protein (mlp), can blunt the maladaptive
growth response and reduce apoptosis in pressure-overloaded hearts.
Fig. 5-7: Interactions of cytoskeletal proteins with integrins in the plasma membrane. A: Integrin
heterodimers, made up of α- and β-subunits, bind to fibronectin, laminin, proteoglycans, and other
proteins that are linked to the extracellular collagen matrix. Within the cell, the integrins bind to
actin microfilaments through a protein complex that includes talin, tensin, vinculin, paxillin, and
α-actinin. B: Integrin signaling occurs when these proteins interact with intracellular molecules
that include talin, tensin, and vinculin, which activate tyrosine kinases such as FAK, PYK2, and CSK
(unshaded) and serine/threonine kinases like ILK, PAK, and PKCá (shaded). These kinases can
modify proliferative signaling by stimulating the mitogen-activated protein kinase (MAP kinase
pathways) Raf, MEK, and ERK. Integrin signals are also mediated when melusin, along with muscle
LIM protein (mlp), activates the serine/threonine kinase Akt.
P.120
P.121
Cytoskeletal Proteins of the Nucleus

Nuclei are surrounded by an envelope that interacts with the cytoskeleton to position the nucleus within
eukaryotic cells, control chromosome movements, and participate in cell division. The nuclear envelope is
made up of two membranes: an outer membrane that is continuous with the endoplasmic reticulum
membrane, and an inner membrane that is continuous with the outer membrane at nuclear pores.
Linkages of tubulin, intermediate filaments, and actin microfilaments provide mechanical support for the
nucleus and initiate a variety of cell signal in response to mechanical stresses on the nuclear envelope
(Fig. 5-8).
Several nuclear envelope proteins interact with the cytoskeleton. These include lamin, a Class V
intermediate filament protein (see above), that forms a mesh beneath the inner membrane within the
nuclear matrix. Lamin-linked proteins connected to the cytoskeleton include emerin, which binds to both
microtubules and actin microfilaments in the cytosol. SUN proteins (named for the yeast gene Sad1 and
the C. elegans gene UNC-84) located within the inner membrane connect to nesprins that span the
perinuclear space; the latter, which are found in both nuclear membranes, are members of a family of
KASH proteins (homologous to Klarsicht, ANC-1, Syne in Drosophila, C. elegans, and humans,
respectively). Like emerin, SUN proteins and emerin link the cytoskeleton to lamin. Large proteins that
contain an α-helical coiled coil, called plectins, serve as connectors that link all three types of
cytoskeletal filament to nesprins and other cytoskeletal proteins.
Fig. 5-8: Cytoskeletal proteins of the nuclear membrane and nucleus. Emerins and SUN-nesprin complexes
link lamin within the nuclear matrix to actin microfilaments. Plectins connect nesprins to microtubules, actin
microfilaments, and intermediate filaments in the cytosol.
P.122
Summary
The ability of cytoskeletal deformation to “sense” mechanical stresses on cardiac myocytes plays an
important role in determining the size, shape, and composition of the heart. This local control of cell
size, shape, and composition by cytoskeletal signaling is largely responsible for the remarkable adaptation
of form to function in the normal heart (Katz and Katz, 1989), whose structure allows its violent
contractions to propel blood through the ventricles without causing turbulence; this explains why
murmurs are not heard in normal individuals.
Activation of proliferative responses by cytoskeletal deformation contributes to the hypertrophic response

in overloaded hearts, and plays an important role in determining the different phenotypes of hypertrophy
seen in heart disease (see Chapter 18). Among the best known of these growth responses are ability of
chronic pressure overload in patients with aortic stenosis and systemic hypertension to initiate
proliferative responses that cause concentric hypertrophy, which normalizes wall stress, and chronic
volume overload to increase chamber volume by initiating eccentric hypertrophy in patients with mitral
and aortic regurgitation. Specific cytoskeletal responses to abnormal stresses caused by myofibrillar and
cytoskeletal protein abnormalities may also explain how different mutations involving several proteins can
cause familial hypertrophic and dilated cardiomyopathies.
Bibliography
General
Assoian RK, Zhu X, Giancotti FG. Cell anchorage and the cytoskeleton as partners in growth factor
dependent cell cycle progression. Curr Opin Cell Biol 1997;9:93–98.
Capetanaki Y. Desmin cytoskeleton in healthy and failing heart. Heart Failure Rev 2000;5:203–220.
Hoshijima M. Mechanical stress-strain sensors embedded in cardiac cytoskeleton: Z disk, titin, and
associated structures. Am J Physiol Heart Circ Physiol 2006;290:H1313–H1325.
Kudoh S, Akazawa H, Takano H, et al. Stretch-modulation of second messengers: effects on

cardiomyocyte ion transport. Prog Biophys Mol Biol 2003;82:57–66.
Lange S, Ehler E, Gautel M. From A to Z and back? Multicompartment proteins in the sarcomere.
Trends Cell Biol 2006;16:11–18.
Sussman MA, McCulloch A, Borg TK. Dance band on the Titanic. Biomechanical signaling in cardiac
hypertrophy. Circ Res 2002;91:888–898.
Tarone G, Lembo G. Molecular interplay between mechanical and humoral signalling in cardiac
hypertrophy. Trend Mol Med 2003;9:376–382.
Thick and Thin Filaments, Z-Lines
Agarkova I, Ehler E, Lange S, et al. M-band: a safeguard for sarcomere stability. J Muscle Res Cell
Motil 2003;24:191–203.
Flashman E, Redwood C, Moolman-Smook J, et al. Cardiac myosin binding protein C. Its role in
physiology and disease. Circ Res 2004;94:1279–1289.
Frank D, Kuhn C, Katus HA, et al. The sarcomeric Z-disc: a nodal point in signaling and disease. J
Mol Med 2006;84:446–468.
Granzier HL, Labeit S. The giant protein titin. A major player in myocardial mechanics, signaling and
disease. Circ Res 2004;94:284–295.
P.123
Kontrogianni-Konstantopoulos A, Bloch RJ. Obscurin: a multitasking muscle giant. J Muscle Res Cell
Motil 2005;26:419–426.
Kontrogianni-Konstantopoulos A, Catino DH, Strong JC, et al. Obscurin modulates the assembly and
organization of sarcomeres and the sarcoplasmic reticulum. FASEB J 2006;20:2102–2111.
Linke WA. Sense and stretchability: the role of titin and titin-associated proteins in myocardial
stress-sensing and mechanical dysfunction. Cardiovasc Res 2008;77:637–648.
McElhinny AS, Perry CN, Witt CC, et al. Muscle-specific RING finger-2 (MURF-2) is important for
microtubule, intermediate filament and sarcomeric M-line maintenance in striated muscle
development. J Cell Sci 2004;117:3175–3188.
Moncman CL, Wang K. Architecture of the thin filament-Z-line junction: lessons from nebulette and
nebulin homologies. J Muscle Res Cell Motil 2000;21:153–169.
Pyle WG, Solaro RJ. At the crossroads of myocardial signaling. The role of Z-discs in intracellular
signaling and cardiac function. Circ Res 2004;94:296–305.
Samarel AM. Costameres, focal adhesions, and cardiomyocyte mechanotransduction. Am J Physiol

Heart Circ Physiol 2005;289:2291–2301.
Sheikh F, Bang ML, Lange S, et al. “Z” eroing in on the role of cypher in striated muscle function,
signaling, and human disease. Trends Cardiovasc Med 2007;17:258–262.
LIM/PDZ Proteins
Hung AY, Sheng M. PDZ domains: structural modules for protein complex assembly. J Biol Chem
2004;277:5699–5702.
Kadrmas JL, Beckerle MC. The LIM domain: from the cytoskeleton to the nucleus. Nat Rev Mol Cell
Biol 2004;5:920–931.
Intercalated Disc
Aberle H, Schwartz H, Kemler R. Cadherin-catenin complex: protein interactions and their

implications for cadherin function. J Cell Biochem 1996;61:514–523.
Garrod D, Chidgey M. Desmosome structure, composition and function. Biochim Biophys Acta 2008;
1778:572–587.
Saffitz JE, Kláber AG. Effects of mechanical forces and mediators of hypertrophy and remolding of
gap junctions in the heart. Circ Res 2004;94:585–591.
Shevtsov SP, Haq S, Force T. Activation of β-catenin signaling pathways by classical G-protein-
coupled receptors. Mechanisms and consequences in cycling and non-cycling cells. Cell Cycle
2006;5:2295–2300.
Stokes DL. Desmosomes from a structural perspective. Curr Opin Cell Biol 2007;19:565–571.
Yap AS, Brieher WM, Gumbiner BM. Molecular and functional analysis of cadherin-based adherens
junction. Annu Rev Cell Dev Biol 1997;13:119–146.
Membranes
Bennett V, Baines AJ. Spectrin and ankyrin-based pathways: metazoan inventions for integrating
cells into tissues. Physiol Rev 2001;81:1353–1392.
Bennett V, Chen L. Ankyrins and cellular targeting of diverse membrane proteins to physiological
sites. Curr Opin Cell Biol 2001;13:61–67.
Brancaccio B, Hirsch E, Notte A, et al. Integrin signalling: the tug-of-war in heart hypertrophy.
Cardiovasc Res 2006;70:422–433.
Dubreuil RR. Functional links between membrane transport and the spectrin cytoskeleton. J Membr
Biol 2006;211:151–161.
P.124
Goldsmith EC, Borg TK. The dynamic interaction of the extracellular matrix in cardiac remodeling. J
Card Fail 2002;8(Suppl 6):S314–S318.
Gratton JP, Bernatchez P, Sessa WC. Caveolae and caveolins in the cardiovascular system. Circ Res
2004;94:1408–1417.
Hillis GS, MacLeod AM. Integrins and disease. Clin Sci 1996;91:639–650.
Hsueh WA, Law RE, Do YS. Integrins, adhesion, and cardiac remodeling. Hypertension 1997;31:176–
180.
Lapidos KA, Kakka R, McNally EM. The dystrophin glycoprotein complex signaling strength and
integrity for the sarcolemma. Circ Res 2004;94:1023–1031.
Ozawa E, Mizuno Y, Hagiwara Y, et al. Molecular and cell biology of the sarcoglycan complex. Muscle
Nerve 2005;32:563–576.
Ross RS, Borg TK. Integrins and the myocardium. Circ Res 2001;88:1112–1119.
Ross RS. The extracellular connections: the role of integrins in cardiac remodeling. J Card Fail
2002;8 (Suppl 6):S326–S331.
Steinberg SF. β2-Adrenergic receptor signaling complexes in cardiomyocyte caveoli/lipid rafts. J Mol
Cell Cardiol 2004;37:407–415.
Nucleus
Hale CM, Shrestha AL, Khatau SB, et al. Dysfunctional connections between the nucleus and the
actin and microtubule networks in laminopathic models. Biophys J 2008;95:5462–5475.
Holaska JM. Emerin and the nuclear lamina in muscle and cardiac disease. Circ Res 2008;103:16–23.
Starr DA. A nuclear-envelope bridge positions nuclei and moves chromosomes. J Cell Sci
2009;122:577–586.
Wiche G. Role of plectin in cytoskeleton organization and dynamics. J Cell Sci 1998;111:2477–2486.
References
Dokuparti MVN, Pamuru PR, Thakkar B, et al. Etiopathogenesis of arrhythmogenic right ventricular
cardiomyopathy. J Hum Genet 2005;50:373–381.
Huxley HE, Hanson J. The molecular basis of contraction in cross-striated muscles. In: Bourne GH,
ed. Structure and function of muscle. Vol 1: Structure. New York, NY: Acad Press, 1960:183–227.
Katz AM, Konstam MA. Definition, historical aspects. In: Heart failure: pathophysiology, molecular
biology, clinical management. 2nd ed. Philadelphia, PA: Lippincott/Williams & Wilkins, 2009:1–49.
Senyo SE, Koshman YE, Russell B. Stimulus interval, rate and direction differentially regulate
phosphorylation for mechanotransduction in neonatal cardiac myocytes. FEBS Lett 2007;581:4241–
4247.
Yamamoto K, Dang QN, Maeda Y, et al. Regulation of cardiomyocyte mechanotransduction by the

cardiac cycle. Circulation 2001;103:1459–1464.
> Table of Contents > Part One - Structure, Biochemistry, and Biophysics > Chapter 6 - Active State, Length-Tension
Relationship, and Cardiac Mechanics
Chapter 6
Active State, Length-Tension Relationship, and Cardiac
Mechanics
T ension development and shortening by the heart depend on interactions between the contractile
proteins (Chapter 4) and their interplay with the cytoskeleton and matrix proteins (Chapter 5). The
classical studies of these mechanical properties described in Chapter 3 were carried out in tetanic
contractions of frog sartorius muscle, whose parallel fibers are ideal for these analyses. Unlike the
relatively simple mechanics of skeletal muscle, cardiac mechanics are complicated by the complex
architecture of the heart, in which branched myocytes are organized in spiral bundles (Chapter 1).
Furthermore, the heart cannot normally be tetanized, which makes it impossible to maintain the
contractile element in a steady state. This poses a formidable obstacle because it requires that
mechanical measurements be made under changing conditions, during the rise and fall of tension in each
cardiac cycle. In spite of these difficulties in analyzing time-dependent and length-dependent changes in
the contractile machinery, studies of the interplay between muscle chemistry and muscle performance
provide valuable insights into the mechanical behavior of cardiac muscle.
Series Elasticity
Analyses of the interactions among the contractile proteins of living muscle are complicated by
elasticities within the muscle. Efforts to understand the rise and fall of tension in skeletal muscle
contractions led early investigators to postulate that a spring-like element, called the series elasticity,
lies between the contractile element and the ends of the muscle (Fig. 6-1). Elongation of this elasticity at
the onset of a contraction absorbs some of the energy generated when the contractile elements shorten.
This contributes to a delay, called the latent period, between the time that a skeletal muscle is
stimulated and the first appearance of tension.
When a skeletal muscle is stretched during the latent period, stiffness is found to have increased, which
indicates that the contractile element had begun to stretch the series elasticity before tension appeared
at the ends of the muscle. Absorption of energy by the series elasticity as tension develops at the onset of
systole has important effects on the heart's performance; conversely, dissipation of some of the tension in
the series elasticity can be used to eject blood later during systole (see Chapter 12).
Twitch, Summation, and Tetanization

The brief response of a skeletal muscle to a single stimulus (S1 in Fig. 6-2) is a twitch. Because the
skeletal muscle action potential is much briefer than the mechanical response, these muscles can respond
to a second stimulus before they begin to relax (S2 in Fig. 6-2). The tension developed in the second
contraction is greater than that developed during the first, which is why this phenomenon is called
summation. If a skeletal muscle is stimulated so rapidly that
P.126
each successive stimulus arrives before the muscle has begun to relax, tension continues to rise until it
reaches a new steady state (Fig. 6-3). The strong sustained contraction is called a tetanus, and the new
level of tension is called tetanic tension. In a skeletal muscle contracting under isometric conditions,
tetanic tension exceeds twitch tension approximately threefold and corresponds to P0 of the force-
velocity curve described in Chapter 3. Twitch tension is less than tetanic tension because energy absorbed
by stretching of the series elasticity reduces the amount of the force developed by the contractile
element that is transmitted to the ends of the muscle (see below).
Fig. 6-1: Model of muscle showing the contractile element in series with a spring-like series
elasticity.
Tetanic contraction, tetanus, and tetany are not the same. Although the term “tetanus” is used to
describe both a tetanic contraction and the disease caused by the endotoxin of the bacteria Clostridium
tetani, the distinction is obvious. Tetany, which differs from both, is a hyperexcitable state of skeletal
muscle caused when lowering of the threshold of the motor endplate to physical and chemical stimuli
allows contractions to appear in response to what are normally subthreshold stimuli. This pathological
condition can be caused by systemic alkalosis or hypocalcemia.
The Series Elastic Element

Figure 6-1 follows a tradition that shows the series elastic element as separate from the contractile
element. In skeletal muscle, some of this elasticity is in the tendinous ends of the muscle. Stretching of
damaged regions adjacent to the clamps used to hold isolated muscles add to this
P.127
elasticity, as does asynchrony of contraction caused by asynchronous stimulation. Significant elasticity is
also found in the sarcomeres, much of which is attributable to elasticities in the cross-bridges, titin, and
other components of the cytoskeleton.
Fig. 6-2: Twitch and summation in a skeletal muscle contracting under isometric conditions. Left: The
contraction (solid line) that follows a single action potential (dashed line) produced by a single stimulus (S1)
is a twitch. Right: Application of a second stimulus (S2) during a twitch produces a second action potential
that causes the renewed development of tension. The dotted line is the tension that would have developed
had the second stimulus not been preceded by the first twitch. Addition of the second tension response to
the first is called summation.
Fig. 6-3: A twitch, as shown in Figure 6-2 (left), and a tetanic contraction (right) in a skeletal
muscle contracting under isometric conditions: Application of a train of stimuli (S) generates a
series of responses that cause tension to rise to a much higher level in a tetanus than is developed
during a twitch. Tetanic tension remains at this high level until stimulation ends or the muscle
fatigues.
Active State in Skeletal Muscle

The tension recorded at the ends of a muscle during a twitch is less than that generated by the
contractile element because the series elasticity absorbs some of the mechanical energy released by the
contractile proteins. The influence of this elasticity can be eliminated by applying a “quick stretch” or a
“quick release” to the ends of the muscle. In skeletal muscle, these experiments reveal the remarkably
rapid development of an active state, which is the tension developed by the interactions between the
myosin cross-bridges and actin. In cardiac muscle, however, similar studies yield results that are quite
different from those in skeletal muscle (see below).
Quick-Stretch Experiments
One way to compensate for the absorption of mechanical energy by the series elasticity during the onset
of a twitch is to stretch the muscle very rapidly. The key to understanding quick-stretch experiments is to
realize that the rapid increase in muscle length can, if the stretch is “just right,” equalize tension at
three places: between the ends of the muscle, within the contractile element, and across the series
elasticity. When this occurs, the tension recorded at the ends of the muscle will be the same as the active
state tension developed by the contractile element.
P.128
Fig. 6-4: Measurement of the active state by application of quick stretches to a skeletal muscle
contracting under isometric conditions. Solid line: Twitch, as shown in Figures 6-2 and 6-3. Dashed
lines: 1: When the muscle is quickly stretched a relatively short distance immediately after
stimulation, the tension at the end of the stretch is less than that produced by the contractile
element; as a result, tension continues to increase. 2: When the muscle is quickly stretched a
relatively long distance immediately after stimulation, the tension at the end of the stretch
exceeds that developed by the contractile element; as a result, tension declines after the stretch.
3: When the muscle is quickly stretched to a length at which tension is the same as the active state
of the contractile element, tension remains at a plateau that equals the active state.
In quick-stretch experiments, the tension developed by the muscle depends on the extent to which its
length is increased. This is depicted in Figure 6-4, which shows the tension recorded after a skeletal
muscle is stretched to three different lengths immediately after stimulation. If the stretch is small,
tension will be less than that which can be generated by the interactions between the myosin cross-
bridges and actin; as a result, the contractile element will develop more tension (curve 1 in Fig. 6-4). If,
on the other hand, the muscle is stretched to a length so long that the tension across the muscle exceeds
that developed by the contractile element, the latter will lengthen, causing tension to decrease until it
equals the tension developed by the contractile proteins (curve 2 in Fig. 6-4). If the quick stretch is “just
right,” and brings the tension between the ends of the muscle to the same level as that developed by the
contractile element, the muscle will neither lengthen nor shorten; instead, tension will remain at a level
equal to that of the contractile element (curve 3 in Fig. 6-4). This tension is the active state.
The time course of active state development can be estimated by applying quick stretches at different
times after stimulation (Fig. 6-5). When such experiments are done with skeletal muscle, active state is
found to develop very rapidly, reaching a brief plateau before twitch tension reaches its peak, after which
the active state begins to decline.
Because muscle tension exceeds active state tension at the end of a twitch (Fig. 6-5, inset), energy stored
in the stretched series elasticity is returned to the ends of the muscle; this allows some of the energy
that had been expended to stretch the series elasticity to be used to perform external work. Use of
energy stored in the stretched series elasticity is especially important in the heart, where the balance
between energy supply and energy demand is precarious even under normal conditions (see Chapter 2).
One reason that allowing a ventricle to eject increases its efficiency (see Chapter 12) is that the decrease
in wall stress which normally occurs during ejection allows some of this elastic energy to be used to pump
blood, rather than being degraded to heat.
The damping effect of the series elasticity explains why more tension is developed during a tetanus than
during a twitch. In the latter, the tension between the ends of the muscle decreases when the series
elasticity is stretched, whereas during a tetanus, repeated stimulation prevents the active state from
declining, which allows active state tension to become the same as the tension
P.129
at the ends of the muscle (Fig. 6-6). Tetanic tension therefore measures the tension developed by the
contractile element.
Fig. 6-5: Time course of the active state in an isometric skeletal muscle twitch. The tension
developed by the contractile element (active state), measured by a series of quick stretches,
greatly exceeds the tension that appears at the ends of the muscle during the twitch because
energy is absorbed by the series elasticity. Storage of this energy in the series elasticity explains
why, at the end of the twitch, muscle tension exceeds the active state; this is shown in the inset,
where the more darkly shaded area indicates the period when the active state exceeds the tension
on the series elasticity, and the lightly shaded region shows when series elasticity tension is greater
than that of the contractile element.
Quick-Release Experiments
The ability of a muscle to develop tension and shorten reflects two different properties of the contractile
process (see Chapter 3). One is the number of rigor bonds, which determines both active state tension
and P0, the maximum tension that a tetanized muscle is able to generate. The other property of
contracting muscle, the velocity of contractile element shortening (Vmax),
P.130
is an index of the rate of cross-bridge turnover, which is independent of the number of rigor bonds. The
latter can be evaluated by subjecting the contracting muscle to “quick releases.”
Fig. 6-6: Relation of tetanic tension (dashed line) to active-state intensity (dotted line). Repetitive
stimulation in the tetanic contraction, by sustaining the tension developed by the contractile
element, allows the full intensity of the active state to be manifest at the ends of the muscle.
Fig. 6-7: Quick releases to different loads in a tetanized skeletal muscle contracting under
isometric conditions. Top: The tension developed during the tetanic contraction falls abruptly when
the muscle is quickly released to a slightly reduced load (1), to a moderately reduced load (2), and
to a markedly reduced load (3). Bottom: The quick releases allow the muscle to shorten to lengths
that are inversely proportional to the new load. The initial, rapid length changes (dashed line) are
due to shortening of the series elasticity, while the slower rates of shortening (dotted lines)
represent the shortening velocities of the contractile element at each new load. The velocity of
the second phase of shortening increases when the muscle is released to progressively lighter
loads.
When a skeletal muscle that has been tetanized under isometric conditions is suddenly presented with a
series of new, reduced loads, the shortening that follows the quick releases occurs in two phases (Fig. 6-
7). The first is a very rapid decrease in length that is caused by shortening of the series elasticity (dashed
line in Fig. 6-7). The velocity of the second slower phase of shortening, which is the result of shortening
of the contractile element (dotted lines in Fig. 6-7), increases when the quick releases are made to
progressively lighter loads (1, 2, and 3 in Fig. 6-7, going from heaviest to lightest). Plots of the load
dependence of shortening velocity in this second phase yield hyperbolic curves (Fig. 6-8) that are similar
to the force-velocity relationship described in Chapter 3 (Fig. 3-15). Quick-release experiments in skeletal
muscle can therefore be used to measure Vmax, an index of the rate of cross-bridge cycling, as well as P0,
an index of the number of active force-generating sites. In cardiac muscle, however, the results of quick-
release experiments are very different from those shown in Figure 6-7.
The Length-Tension Relationship

A major determinant of the ability of a skeletal or cardiac muscle to develop tension is its resting length.
Curves describing this length-tension relationship (Fig. 6-9) are customarily scanned from left to right, so
that the increase in developed tension that occurs when a muscle
P.131
is stretched at shorter muscle lengths is called the ascending limb, while the decline of tension when the
muscle is stretched at longer lengths is the descending limb. The tension developed during tetanic
contractions in skeletal muscle is maximal at intermediate muscle lengths (lmax or l0).
Fig. 6-8: Force-velocity curve constructed from the quick-release experiment shown in Figure 6-7.
Points 1, 2, and 3 are the tensions achieved after the quick releases to the new lengths in the
upper part of Figure 6-7, and the corresponding shortening velocities calculated from the length
changes shown by the dotted lines in the lower part of Figure 6-7.
Shortly after the sliding filament hypothesis of muscular contraction became widely accepted, attempts
were made to explain the length-tension relationship in terms of the number of myosin cross-bridges on
the thick filaments that are opposite to, and so able to interact with, actin on the thin filaments. This
ultrastructural mechanism explains the descending limb of the length-tension relationship, where tension
decreases with increasing sarcomere length, but not the ascending limb, where tension increases with
increasing sarcomere length.
The key to understanding the length-tension relationship was provided by Gordon et al. (1966) who
measured length-tension curves generated by a single sarcomere in a frog semitendinosus muscle (Fig. 6-
10); these curves are narrower than those in the whole muscle because
P.132
these experiments eliminate inhomogeneities within the muscle and artifacts introduced when the
damaged ends of the muscle are attached to the recording device. In addition to being narrower,
sarcomere length-tension relationships demonstrate sharp changes in tension not seen in studies of whole
muscles (Fig. 6-10B and C).
Fig. 6-9: Length-tension curve during isometric tetanic contractions of a frog semitendinosus
muscle. Resting length is expressed as percent of lmax, the length at which developed tension is
maximal. As these curves are conventionally scanned from left to right, the ascending limb is to
the left, where tension rises with increasing muscle length, and the descending limb is to the right,
where tension decreases with increasing muscle length.
Fig. 6-10: Length-tension curve for a single sarcomere of a frog semitendinosus muscle (solid line)
compared with that for the whole muscle shown in Figure 6-9 (dashed line). Sarcomere length is
shown as the upper abscissa, muscle length is below. At a sarcomere length of 3.65 µm (A), no
tension is developed. Developed tension increases to a maximum as the sarcomere shortens to 2.2
µm (A → B), then remains constant when sarcomere length decreases from 2.2 to 2.0 µm (B → C).
After sarcomere length decreases below 2.0 µm, further shortening reduces the development of
tension (C → D). At sarcomere lengths below 1.65 µm (D → E), contraction bands appear and
tension declines very rapidly.
The Descending Limb: An Ultrastructural Phenomenon

To understand the sarcomere length-tension curve, it is useful to begin at the right, at the end of the
descending limb, where developed tension is zero (A, Fig. 6-10). Sarcomere length is 3.65 µm at this
point, and there is no overlap between the thick filaments, whose length in the skeletal muscle where
these measurements were obtained is 1.65 µm, and the two thin filaments, whose combined length is 2.0
µm (Fig. 6-11). Because there is no overlap, myosin cross-bridges cannot interact with the thin filaments
and no tension can be developed.
When sarcomere length is reduced (A → B, Fig. 6-10), tension increases, reaching a maximum at 2.2 µm
(B, Fig. 6-10). At this sarcomere length, which corresponds to lmax (see above), all of the myosin cross-
bridges in the two-halves of the sarcomere are able to interact with actin because they are opposite to
one of the thin filaments (Fig. 6-12).
Further reduction in sarcomere length, from 2.2 to 2.0 µm (B → C, Fig. 6-10), does not change active
tension because decreasing sarcomere length neither increases nor decreases the number of potential
interactions between the thick and thin filaments. This is due to the small, 0.2 µm, “bare” area in the
center of the thick filament that is devoid of cross-bridges (see Chapter 4, Fig. 4-3). Although the thin
filaments lose potential interactions with cross-bridges at the center of the sarcomere when they enter
this bare area, new interactions with cross-bridges are gained at the ends of the thick filament (Fig. 6-
13).
P.133
Fig. 6-11: At a sarcomere length of 3.65 µm (A, Fig. 6-10), there is no overlap between thick and the thin
filaments, so that myosin cross-bridges cannot interact with actin. The lengths of the thick and thin filaments
(dotted arrows) are shown at the bottom of the figure.
The Ascending Limb: Length-Dependent Variations in Cooperative

Interactions between the Contractile Proteins, the Cytoskeleton,
and Calcium
The fall in tension as sarcomeres shorten at lengths below 2.0 µm (C → D, Fig. 6-11) cannot be explained
by a change in the number of potential interactions between the thick and thin filaments because
interactions lost in the center of the thick filament are matched by gains at its
P.134
ends. The fall in tension with decreasing sarcomere length was initially attributed to mechanical
interference between the thin filaments in the region of “double overlap,” where the thin filaments from
opposite sides of the sarcomere pass each other (Fig. 6-14). This explanation seemed logical because the
thin filaments from the opposite halves of the sarcomere had crossed into the domains of the “wrong”
halves of the thick filament, where the polarities of the cross-bridges and actin do not allow interactions
between adjacent thick and thin filaments (see Chapter 4). However, neither mismatched polarities nor
mechanical interferences caused by double overlap explain the decrease in tension when sarcomeres
shorten on the ascending limb.
P.135
Instead, length-dependent changes in the lattice spacing of the myofilaments (Pearson et al., 2007)
modify the cooperative interactions among the contractile proteins of the thick and thin filaments, the
calcium sensitivity of these interactions, and interactions with cytoskeletal proteins such as titin and
myosin-binding protein C. The complexity of these mechanisms is seen in the ability of posttranslational
changes that change the calcium sensitivity of the contractile proteins, such as protein kinase A-catalyzed
phosphorylations, to modify the length-tension relationship (Konhilas et al., 2003).
Fig. 6-12: At a sarcomere length of 2.2 µm (B, Fig. 6-10), all of the myosin cross-bridges can interact with
the thin filaments, so that the number of potential interactions between the contractile proteins is maximal.
Fig. 6-13: Between sarcomere lengths of 2.2 and 2.0 µm (C, Fig. 6-10), the loss of potential cross-
bridge interactions at the ends of the thin filaments in the center of the sarcomere is matched by a
gain at the ends of the thick filaments, so that all myosin cross-bridges remain able to interact
with the thin filaments.
Fig. 6-14: At a sarcomere length of 1.7 µm (C → D, Fig. 6-10), all of the myosin cross-bridges can
interact with the thin filaments; however, the central ends of the thin filaments have crossed in
the middle of the sarcomere (“double overlap”).
Fig. 6-15: At sarcomere lengths less than 1.65 µm (D → E, Fig. 6-10), the Z-lines “collide” with the
ends of thick filaments; this causes the latter to “crumple” (thick, wavy lines) and so gives rise to
contraction bands at the periphery of the A-band. The collisions between the thick filaments and
the Z-lines cause the precipitous fall of tension as the sarcomere shortens on this nonphysiological
portion of the length-tension curve.
Contraction Bands: A Mechanical Phenomenon

The steep decline in tension as muscle length decreases at very short sarcomere lengths, below 1.65 µm
(D → E, Fig. 6-10), is accompanied by the appearance of contraction bands (Fig. 6-15). The latter appear
when the Z-lines collide with the ends of the thick filaments, whose length is 1.65 µm, which causes the
latter to crumple. In the heart, contraction bands are seen under conditions of severe calcium overload,
where active tension can become so high as to tear the myocytes apart. This phenomenon, which causes
cardiac myocyte death when plasma membrane damage allows large amounts of calcium to enter the
cytosol, is often referred to as “contraction band necrosis” (see Chapter 17).
Properties of the Resting Myocardium

Cardiac muscle has a low compliance, or high stiffness (Table 6-1), so that the resting length-tension
curve is very steep (Fig. 6-16). [These and other terms used to describe the passive properties of the
heart are reviewed by Mirsky and Parmley (1973)]. This low diastolic compliance distinguishes the heart
from skeletal muscle, in which resting tension is close to zero at lmax. However, the relationships between
sarcomere length and active tension are similar in skeletal and cardiac muscle. The high resting tension of
cardiac muscle helps prevent the ventricles from moving onto the descending limb of their length-tension
curves, which can establish a dangerous vicious cycle (see below).
P.136
Table 6-1 Terms Used to Describe the Passive Properties of the Myocardium
Tension: Force along a line, e.g., dyn/cm.
Stress: Force across an area, e.g., dyn/cm2.
Strain: Deformation of a material, the change in dimension caused by application

of stress.
Compliance Change in volume of a chamber caused by a change in pressure: dV/dP.

or
distensibility:
Stiffness: Change in pressure within a chamber caused by a change in volume:
dP/dV.
Elasticity: Ability of a material to return to its original conformation when a stress is

removed following a stress-induced change in conformation.
Elastic Slope of a stress-strain curve, i.e., amount of stress needed to cause a

stiffness: given strain. In the myocardium, elastic stiffness is the ability of the walls
of a chamber to resist stretch (a strain) following an increase in tension (a
stress).
The high diastolic stiffness of cardiac muscle is due in part to the extracellular matrix, notably collagen
(Weber, 1989). Stiffness of the cytoskeletal proteins, notably titin, also contributes to the low resting
compliance of the heart, and both titin isoform shifts and changes in titin phosphorylation can modify
diastolic compliance (Borbály et al., 2009; see also Chapter 5). Residual interactions between the thick
and thin filaments during diastole, which are intensified by calcium overload, represent an important
cause of increased diastolic stiffness in energy-starved hearts.
The high resting stiffness of the myocardium minimizes chamber enlargement, which has several
deleterious consequences. Dilatation, by increasing wall stress (The Law of Laplace, see Chapter 11),
decreases cardiac efficiency (see Chapter 12). On the other hand, the high diastolic
P.137
stiffness prevents the ventricles from entering the descending limb of the length-tension relationship,
where the heart cannot respond to an increase in filling by increasing its ability to eject (Chapter 10).
Instead, increased venous return to a ventricle operating on the descending limb will reduce ejection,
thereby causing a further increase in end-systolic volume (the volume of the ventricle at the end of
systole) that would reduce ejection further; the result would be a vicious cycle that can lead to acute
pulmonary edema (see Chapter 18).
Fig. 6-16: Isometric length-tension curves in skeletal and cardiac muscle. The tension developed during
isometric contractions at various rest lengths (dotted lines labeled “active tension”) is equal to the total
tension developed in each contraction (dashed line) minus the resting tension prior to stimulation (solid
line). The active length-tension curves for the two muscle types are similar, but the resting tension is much
higher in cardiac than skeletal muscle.
Active State in the Heart: Cardiac Mechanics

In the 1960s, efforts to quantify myocardial function were stimulated by advances in the surgical repair of
damaged cardiac valves and observations that, even when the surgery went well, patients sometimes
failed to recover. Recognition that prolonged hemodynamic overloading irreversibly damages the
myocardium (see Chapter 18) made it essential to distinguish the extent to which the hemodynamic
abnormalities in a given patient were determined by the valve abnormality, or whether the major
problem was heart muscle damage. It was quickly noted that the severity of hemodynamic abnormalities,
such as low cardiac output and high venous pressures, could not accurately define the contractile state of
the myocardium. These considerations led to the emergence of a new field, cardiac mechanics, which was
based on the work in isolated skeletal muscle described in Chapter 3. It was initially believed that
estimates of such variables as P0 and Vmax would allow cardiologists to quantify the intrinsic contractile
properties of the heart, called myocardial contractility (see Chapter 10) in order to identify the optimal
time for palliative procedures like valve replacement—before myocardial deterioration passed a “point of
no return,” but not so early as to expose patients prematurely to the complications of prosthetic valves.
Initial studies of cardiac mechanics were based on studies of frog sartorius muscle, where fibers are
parallel to each other, and the active state is rapid in onset and can be stabilized in tetanic contractions.
However, it is now clear that the relatively straightforward time-dependent and length-dependent
features of skeletal muscle contraction described in Chapter 3 are not seen in the heart. One reason is
that, because cardiac muscle cannot be tetanized, its mechanical properties do not achieve a steady
state. Another problem reflects the spiral arrangement of the heart's muscle bundles (see Chapter 1),
which adds to the elasticities in the walls of the heart that complicate the relationships between
sarcomere shortening, ventricular volume, and wall stress. The inability of classical skeletal muscle
mechanics to describe the contractile performance of the heart became obvious when the results of
quick-stretch and quick-release studies in cardiac muscle revealed an entirely unexpected type of
regulation in which time- and length-dependent changes in active state have profound effects on the
development of tension in the walls of the heart.
Quick Stretch in Cardiac Muscle

Unlike quick-stretch experiments in skeletal muscle, which demonstrate that the active state appears
rapidly and maintains a steady state following stimulation (Fig. 6-4), similar experiments in cardiac
muscle show that active-state tension does not achieve a plateau (Fig. 6-17). Quick stretches that initially
exceed the ability of cardiac muscle to hold tension, as evidenced by a fall in tension immediately after
the stretch, are followed by a slow rise in tension whose initial time course is similar to that of tension
development by the unstretched muscle (A in Fig. 6-17).
P.138
Increasing the quick stretch to longer lengths causes an even greater initial drop in tension, but again,
tension rises along a similar time course (B in Fig. 6-17).
Fig. 6-17: Results of a quick-stretch experiment in cardiac muscle. Solid line, tension developed
during isometric contraction. Dashed lines, tension developed after a smaller (A) and a larger (B)
quick stretch. Both quick stretches bring the muscle to a length where total tension exceeds that
developed by the contractile element—which is evidenced by a transient fall in tension—but in
both, total tension resumes its rise. Unlike the analogous experiment in skeletal muscle (Fig. 6-5),
a plateau of tension is not seen regardless of the extent to which the muscle is stretched.
Even more striking differences between cardiac and skeletal muscle are seen when quick stretches and
quick releases are applied at different times during tension development. Unlike skeletal muscle, where
the active state rises very rapidly (see Fig. 6-4), the active state in cardiac muscle rises slowly at the
onset of systole. This is apparent because the tension developed after a cardiac muscle is stretched at
different times after stimulation follows a time course that is virtually independent of the time when
stretch is applied (A and B in Fig. 6-18). Furthermore, the increased tension developed after cardiac
muscle is stretched when tension is developing is the same as that seen when the muscle is stretched to
the same length before stimulation (X in Fig. 6-18). This demonstrates that the tension increase which
appears after a quick stretch in cardiac muscle (Figs. 6-17 and 6-18) is simply a manifestation of the
length-tension relationship. These quick-stretch experiments therefore show that unlike skeletal muscle,
where active state develops rapidly and is sustained, active state in cardiac muscle is slow in onset and
does not achieve a plateau.
Effects of Changes in Length on Active State in Cardiac Muscle

Analysis of quick stretches and quick releases applied at different times after the onset of contractions in
heart muscle demonstrates that changes in muscle length have important effects on the time course of
subsequent tension development. The initial effect of stretching the heart early during systole, as noted
above (Figs. 6-17 and 6-18), can be explained by the ability of the higher tension to increase sarcomere
length along the ascending limb of the length-tension curve. If, however, the time course of tension is
followed for a longer time, stretch is seen also to prolong contraction (A in Fig. 6-19). Conversely, when
the contracting heart is allowed to shorten, contraction is abbreviated (B and C in Fig. 6-19) and the
earlier the muscle is allowed to shorten, the greater is the abbreviation of systole (compare B and C in
Fig. 6-19).
P.139
Fig. 6-18: Results of a quick-stretch experiment in cardiac muscle where the same stretch is
applied at two times after stimulation (A and B). Although the tension developed after each stretch
initially exceeds that developed by the contractile element—as evidenced by a transient fall in
tension—total tension resumes its rise and the two new curves become superimposed as tension
approaches its peak. When the muscle is stretched to the same new length prior to stimulation (X),
the increased tension is the same as that which develops after the quick stretches.
Equalization of Tension Developed in the Walls of the Heart

The effects of length changes on the active state in cardiac muscle play a vital role in maintaining
homogenous contractions in the walls of the heart. This is because the heart's architecture, in which
branched myocytes of different dimensions are connected in series (see Chapter 1), poses a serious
obstacle to achieving the homogeneous tension within the walls of the heart needed to maximize
efficiency (Katz and Katz, 1989).
Fig. 6-19: Effects of length changes on the subsequent development of tension in a cardiac muscle.
When an isometrically contracting muscle (solid line) is stretched early during systole, tension is
increased and the contraction is prolonged (A). If the muscle is released to a lower tension (B and
C), the muscle shortens and the duration of the contraction is abbreviated. The abbreviation of
contraction is increased when the muscle is allowed to shorten at a lower (C) than a higher (B)
tension.
P.140
The importance of length-dependent changes in active state can be understood by examining the
interactions between two myocytes of unequal strength that are linked in series, where the stronger
myocyte would stretch the weaker myocyte (Fig. 6-20B), or the interactions between two identical
myocytes are not activated synchronously, where the myocyte that was activated first would stretch the
myocyte in which stimulation was delayed (Fig. 6-21B). In both cases, these inequalities are reduced by
the
P.141
length-tension relationship and the length dependence of the active state. This is because when a
cardiac myocyte that contracts more weakly is stretched by the stronger myocyte, the active state of the
weaker myocyte is both strengthened (by the length-tension relationship) and prolonged (by the length
dependence of active state duration); at the same time, both the length-tension relationship and length
dependence of active state weaken the stronger myocyte. These mechanisms tend to equalize tension
because shortening of the myocyte with the stronger contraction (B and C in Figs. 6-20 and 6-21) reduces
both developed tension (by the length-tension relationship) and the duration of its contraction (by the
length dependence of active state), while both developed tension (by the length-tension relationship) and
the duration of its contraction (by the length dependence of active state) (B and C in Figs. 6-20 and 6-21)
are increased by lengthening of the myocyte with the weaker contraction.
Fig. 6-20: Influence of myocyte size on the extent of isotonic shortening by two cardiac myocytes
connected in series. A: When the two myocytes are identical, both shorten to the same extent. B:
If the left-hand myocyte is larger (and so stronger) than the right-hand myocyte, the greater extent
of shortening by the stronger myocyte stretches the weaker myocyte. C: In cardiac muscle,
because shortening weakens contraction and elongation strengthens contraction, the force
generated by the stronger myocyte will decrease as it begins to stretch the weaker myocyte;
conversely, the force developed by weaker myocyte will increase as it begins to stretch the
stronger myocyte. Together, these responses equalize the shortening of the mismatched myocytes
by reducing the extent of shortening by the stronger myocyte and increasing shortening of the
weaker myocyte.
The ability of these mechanisms to equalize the tension developed by the millions of myocytes in the
walls of the heart is blunted in disease by asynchrony of activation and
P.142
mechanical heterogeneities (see Chapters 17 and 18). The possibility of restoring normal homogeneity
within the walls of these hearts led to the development of “cardiac resynchronization therapy,” where
electrical stimulation of regions where activation is delayed can overcome the loss of heterogeneity seen
in many failing hearts.
Fig. 6-21: Influence of the synchrony of myocyte activation on the extent of isotonic shortening by
two cardiac myocytes connected in series. A: When the two myocytes are activated at the same
time, both shorten to the same extent. B: If the left-hand myocyte is activated before the right-
hand myocyte, initial shortening by the left-hand myocyte will stretch the right-hand myocyte. C:
In cardiac muscle, because shortening weakens contraction and elongation strengthens contraction,
the force generated by the left-hand myocyte will decrease as it begins to stretch the right-hand
myocyte; conversely, the force developed by the right-hand myocyte will increase as it begins to
stretch the left-hand myocyte. Together, these responses equalize the shortening of the
asynchronously activated myocytes by reducing the extent of shortening by the myocyte that is
activated first and increasing shortening of the myocyte that is activated later.
Bibliography
Brady AJ. Mechanical properties of isolated cardiac myocytes. Physiol Rev 1991;71:413–428.
Brady AJ. Time and displacement dependence of cardiac contractility: problems in defining the
active state and force velocity relations. Fed Proc 1965;24:1410–1420.
Brutsaert DL, Sys SU. Relaxation and diastole of the heart. Physiol Rev 1989;69:1228–1315.
Dos Remedios CG. The regulation of muscle contraction: as in life, it keeps getting more complex.
Biophys J 2007;93:4097–4098.
Moss RL, Fitzsimons DP. Frank-Starling Relationship. Long on importance, short on mechanism. Circ
Res 2002;90:11–13.
Pearson JT, Shirai M, Tsuchimochi H, et al. Effects of sustained length-dependent activation on in

situ cross-bridge dynamics in rat hearts. Biohys J 2007;93:4319–4339.
Solaro RJ. Mechanisms of the Frank-Starling Law of the heart: the beat goes on. Biophys J 2007;93:
4095–4096.
Solovyova O, Katsnelson L, Guriev S, et al. Mechanical inhomogeneity of myocardium studied in

parallel and serial cardiac muscle duplexes: experiments and models. Chaos Solitons Fractals
2002;13: 1685–1711.
References
Borbály A, Falcao-Pires I, van Heerebeek L, et al. Hypophosphorylation of the Stiff N2B titin isoform
raises cardiomyocyte resting tension in failing human myocardium. Circ Res 2009;104:780–786.
Gordon AM, Huxley AF, Julian FG. The variation in isometric tension with sarcomere length in
vertebrate muscle fibers. J Physiol (Lond) 1966;184:170–192.
Konhilas JP, Irving TC, Wolska BM, et al. Troponin I in the murine myocardium: influence on length-
dependent activation and interfilament spacing. J Physiol 2003;547:951–961.
Mirsky I, Parmley WW. Assessment of passive elastic stiffness for isolated heart muscle and the
intact heart. Circ Res 1973;33:233–243.
Weber KT. Cardiac interstitium in health and disease: the fibrillar collagen network. J Am Coll
Cardiol 1989;13:1637–1652.
> Table of Contents > Part One - Structure, Biochemistry, and Biophysics > Chapter 7 - Excitation-Contraction Coupling:
Extracellular and Intracellular Calcium Cycles
Chapter 7
Excitation-Contraction Coupling: Extracellular and
Intracellular Calcium Cycles
It is quite … impossible to explain the rapid development of full activity in a

[skeletal muscle] twitch by assuming that it is set up by the arrival at any point of
some substance diffusing from the surface: diffusion is far too slow. Either we must
suppose that [the muscle is stimulated by] excitation (natural or artificial)
throughout the interior, not merely at the surface: or we must look for some
physical or physico-chemical process which is released by excitation at the surface
and then propagated inwards.
—A. V. Hill, 1949
Hill (1949), writing before calcium was discovered to activate the contractile proteins, noted that
diffusion of an activator from the surface of a frog sartorius muscle is too slow to account for the rapid
onset of the active state in these large, rapidly contracting muscles. This led him to postulate that
excitation-contraction coupling, the mechanism by which plasma membrane depolarization initiates
contraction, depends on a process more rapid than diffusion across the plasma membrane, or on the
release of a diffusible activator from structures within the myocytes. Identification of the role of calcium
in activating the contractile proteins (see Chapter 4) led to the discovery that both of Hill's postulated
mechanisms overcome the limitations caused by the slowness of diffusion.
Excitation-contraction coupling in the small, slowly contracting myocytes of primitive and embryonic
hearts is effected by an extracellular calcium cycle in which activator calcium diffuses into the cytosol
from the extracellular fluid (Fig. 7-1). However, excitation-contraction coupling in the larger myocytes of
adult mammalian hearts, which are filled with myofilaments, contract more rapidly, and develop higher
levels of tension, requires an additional intracellular calcium cycle. In the latter, depolarization of plasma
membrane extensions called transverse tubules (t-tubules) propagates a signal into the cell interior that
triggers calcium release from intracellular stores in the sarcoplasmic reticulum (Fig. 7-2) (see Chapter 1).
The Sarcoplasmic Reticulum and Transverse Tubules

The first clue that calcium stored within the sarcoplasmic reticulum activates contraction was obtained in
the 1950s, when supernatants obtained after low-speed centrifugation of muscle minces were found to
relax actomyosin preparations in vitro (see Chapter 4). This effect, which required the presence of ATP
and could be abolished by calcium, was initially believed to be
P.144
caused by a “soluble relaxing factor” (Gergely, 1959). However, Hasselbach and Makinose (1961) and
Ebashi and Lipmann (1962) discovered independently that the relaxing effect of these supernatant
fractions depended on tiny membrane vesicles derived from the sarcoplasmic reticulum, called
microsomes, that use energy from ATP hydrolysis to pump calcium into their interior. The concurrent
discovery that actin–myosin interactions are activated under physiological conditions by micromolar-
ionized calcium concentrations (Weber and Winicur, 1961) made it clear that muscle relaxation is not
caused by a soluble factor, but instead occurs when calcium is sequestered by the sarcoplasmic reticulum.
Within a few years it was possible to show that the cardiac sarcoplasmic reticulum contains a calcium
pump with both sufficient capacity and calcium affinity to relax the heart (Katz and Repke, 1967;
Harigaya and Schwartz, 1969).
Fig. 7-1: Primitive and embryonic cardiac myocytes (A) are smaller and contain fewer
myofilaments than adult myocytes (B). Calcium influx from the extracellular space can deliver
enough of this activator to explain the more slowly developing, weaker contractions of embryonic
myocytes, but diffusion of calcium across the plasma membrane from the extracellular space is too
slow to activate the more rapidly developing, stronger contractions of the adult heart.
The answer to the second part of Hill's question, how a signal generated by activation at the cell surface
reaches the interior of the myocytes, was provided when the t-tubules were discovered to be plasma
membrane extensions whose lumens open into the extracellular space, and that the t-tubular membranes
can propagate action potentials into the interior of the muscle cells. In one of the classic experiments in
skeletal muscle physiology, Huxley and Taylor (1958) demonstrated that very small electrical stimuli
applied through microelectrodes placed near the mouth of a t-tubule can induce contractions that are
limited to the sarcomeres adjacent to the point of stimulation. Further evidence that t-tubules play a
critical role in excitation-contraction coupling was obtained when disruption of the connections between
the t-tubules and the plasma membrane was found to make it impossible to activate contraction
(Eisenberg and Eisenberg, 1968). This and other evidence that transmission of a wave of depolarization
down the t-tubules
P.145
into the cell interior, which is much more rapid than diffusion of an activator substance, completed the
answer to Hill's question.
Fig. 7-2: The sarcoplasmic reticulum. A: Intracellular calcium stores within the sarcoplasmic
reticulum overcome the limitations caused by slow diffusion of activator from the extracellular
space into the cytosol. B: Major membranes in a working myocardial cell. The sarcoplasmic
reticulum includes the subsarcolemmal cisternae, which form composite membrane structures with
the plasma membrane called dyads, and the sarcotubular network, which surrounds the contractile
proteins. The lumens of the t-tubules open to the extracellular space, which allows these
structures to propagate action potentials into the cell. Mitochondria are shown in the central
sarcomere and in cross section at the left side of the figure.
Extracellular and intracellular Calcium Cycles

Muscles can use two possible sources of calcium to activate contraction. As noted above, most of the
calcium that activates the small, slowly contracting myocytes of embryonic hearts enters the cytosol from
the extracellular fluid (the extracellular calcium cycle), whereas the larger, more rapidly contracting
myocytes of the adult mammalian heart depend mainly on calcium derived from intracellular stores (the
intracellular calcium cycle). In both cases, the fluxes of activator calcium into the cytosol are passive
(downhill) because ionized calcium concentrations in the extracellular fluid and within the sarcoplasmic
reticulum are much higher than cytosolic calcium concentration in resting muscle (see Chapter 4).
There are important differences between these calcium cycles. One is that both calcium influx and
calcium efflux across the plasma membrane in the extracellular calcium cycle are accompanied by
depolarizing currents. In contrast, the calcium fluxes into and out of the sarcoplasmic reticulum in the
intracellular calcium cycle do not influence the electrical potential across the plasma membrane. In
addition, the electrochemical gradient that drives calcium into the cytosol
P.146
across the plasma membrane in the extracellular calcium cycle is increased by the electronegativity
within resting cardiac myocytes.
The relative amounts of calcium delivered to the contractile proteins by the extracellular and
intracellular calcium cycles vary among different muscles. In skeletal muscle, virtually all of the calcium
that activates contraction is derived from the sarcoplasmic reticulum, whereas most of the activator
calcium in smooth muscle and embryonic hearts enters the cytosol from the extracellular space. There
are also differences in the sources of activator calcium in the hearts of different mammalian species; in
human ventricles, about two-thirds is derived from the sarcoplasmic reticulum and one-third from the
extracellular space.
Table 7-1 Structures That Participate in Cardiac Excitation-Contraction Coupling

and Relaxation
Structure Excitation-Contraction Coupling Relaxation
Plasma membrane
Sarcolemma
Na channel Depolarization
Open plasma membrane Ca channels
Ca channel Action potential plateau
Open intracellular Ca release channels
Ca pump (PMCA) Ca removal from

cytosol
Na/Ca exchanger Ca entry into cytosol Ca removal from

(NCX) cytosol
Na pump Establish Na
gradient
Transverse tubule
Na channel Propagate action potentials into cell

interior
Ca channel Open intracellular Ca release channels
Sarcoplasmic reticulum
Subsarcolemmal cisternae
Ca release channel Release of Ca for binding to troponin C
Calsequestrin Ca storage, regulation
Sarcotubular network
Ca pump (SERCA) Ca removal from

cytosol
Myofilaments
Actin and myosin Contraction
Troponin C Ca receptor
Tropomyosin, Regulate actin interactions with

troponins I and T myosin cross-bridges
P.147
In most mammalian skeletal muscle twitches, enough calcium is released from the sarcoplasmic reticulum
to bind to virtually all of the troponin C; as a result, the active state normally reaches its maximum.
However, in adult mammalian cardiac myocytes contracting under basal conditions, the sarcoplasmic
reticulum provides enough calcium to bind to only ∼40% of the troponin C. This allows the intensity of the
contractile response to be modified by variations in calcium release from the sarcoplasmic reticulum,
calcium influx across the plasma membrane, and the calcium affinity of troponin C. These variations allow
cardiac performance to be regulated by the many structures that regulate calcium fluxes into and out of
the cytosol (Table 7-1).
The Extracellular Calcium Cycle
Calcium Influx across the Plasma Membrane

Calcium influx thorough voltage-gated calcium channels participates in both electrical and chemical
signaling (Table 7-2). Most important are the L-type calcium channels, named because of their relatively
long openings (L = long-lasting); these channels are also called dihydropyridine receptors because they
bind with high affinity to this class of calcium channel-blocking drugs. In the atria, ventricles, and His-
Purkinje system, where the initial depolarizing current that causes the action potential upstroke is carried
by sodium, the subsequent opening of these channels contributes to the action potential plateau by
allowing positively charged calcium ions to cross the plasma membrane (see Chapter 14). Analogous
depolarizing calcium currents in the SA and AV nodes participate in pacemaker activity and impulse
propagation. Calcium entry via L-type calcium channels triggers the release of a larger amount of calcium
from intracellular stores in the sarcoplasmic reticulum (see below) and, because some of the calcium that
enters the cytosol from the extracellular space is taken up and stored by the sarcoplasmic reticulum
where it can be added to the calcium released in subsequent contractions, this calcium influx helps
determine the strength of the heartbeat.
Table 7-2 Major Functions of Calcium Entry Through L-Type Calcium Channels in
The Heart
Role of Calcium Functional Consequence

Electrical
Carries positive charge into the cell Depolarization
Working cells (atria and ventricles) Action potential plateau
His-Purkinje system Action potential plateau
Sinoatrial node Pacemaker activity
Atrioventricular node Atrioventricular conduction
Chemical
Triggers calcium release from sarcoplasmic reticulum Ca-triggered Ca release
Provides calcium for binding to troponin Activates contraction
Fills calcium stores in sarcoplasmic reticulum Maintains contractility
Activates potassium channels Repolarization
P.148
Fig. 7-3: Three P-type ion pump proteins. The plasma membrane calcium pump (A), sarcoplasmic
reticulum calcium pump (B), and the α-subunit of the sodium pump (C) all contain 10 membrane-
spanning α-helices and a large cytosolic domain that includes an ATPase site that provides the
energy for active transport. The ions that are transported bind to groups of membrane-spanning α-
helices. In the plasma membrane calcium pump, a portion of the C-terminal peptide chain provides
a regulatory site that binds the calcium-calmodulin complex. Phospholamban, which regulates
calcium transport into the sarcoplasmic reticulum, is homologous to the C-terminal portion of the
plasma membrane calcium pump. The sodium pump contains three subunits: the larger α-subunit
binds sodium, potassium, ATP, and cardiac glycosides. A glycosylated β-subunit and a small γ-
subunit regulate pump activity.
P.149
Calcium Efflux across the Plasma Membrane

At any steady state, the calcium that enters the cytosol through L-type calcium channels during each
action potential must be pumped out of the cell during diastole. In the heart, two mechanisms effect this
uphill transport: a plasma membrane calcium pump and a sodium/calcium exchanger. The latter has a
greater capacity than the plasma membrane calcium pump and is responsible for more than 80% of this
calcium efflux in human ventricles.
P-Type Ion Pumps: the Plasma Membrane Calcium Pump ATPase

(PMCA)
The plasma membrane calcium pump ATPase (PMCA) is one of a family of P-type ion pumps that couple
energy derived from ATP hydrolysis to active cation transport; other members of this family include the
sarcoplasmic reticulum calcium pump, the sodium pump, and proton pumps (Fig. 7-3). The uphill
transport of ions by P-type ion pumps resembles a boat moving upstream through a series of locks (Fig. 7-
4). As an ion moves across the membrane bilayer it
P.150
becomes “occluded,” which means that it becomes unable to exchange with ions in the aqueous solutions
on either side of the membrane—this is analogous to the boat in a closed lock. Transfer of chemical
energy from ATP to the occluded ions increases their activity, much as pumping water into the lock raises
the boat.
Fig. 7-4: Ion transport by a P-type ion pump. The uphill flux of a cation, such as calcium, across
the plasma membrane (left) resembles the passage of a boat up a river through a series of locks
(right). A cation in its low-energy state approaches the channel (A) and binds to negatively charged
sites in the channel where, like the boat in a closed lock, it becomes “occluded” (B). Energy
supplied to the occluded ion within the channel “raises” its activity in a process analogous to filling
the lock, which lifts the boat to the higher level on the upstream side (C). After its activity is
increased, the ion ceases to be occluded and, like the boat after the upstream gate is opened, can
move freely into the region of high activity (D).
The turnover of the plasma membrane calcium pump is activated when calcium binds to a transport site
on its cytosolic side. Increased cytosolic calcium also stimulates the pump indirectly because in its basal
state, when cytosolic calcium is low, the pump is inhibited by a portion of its C-terminal region that lies
within the cytosol (Fig. 7-5). Reversal of this inhibition when a calcium-calmodulin complex binds to this
C-terminal region of the pump helps cardiac myocytes avoid calcium overload by recognizing increased
cytosolic calcium concentration as a signal to accelerate calcium efflux from these cells.
The plasma membrane calcium pump also plays an important role in cell signaling. This occurs when the
pump protein is incorporated into caveoli, which allows it to interact with signaling
P.151
proteins that include protein kinases, calcineurin, RASSF1 (Ras-associated factor 1), nNOS (neuronal nitric
oxide synthase), PDZ domain-containing proteins, and syntrophins that mediate proliferative responses
(see Chapters 5 and 9).
Fig. 7-5: Regulation of the plasma membrane calcium pump by the calcium-calmodulin complex. A:
In the basal state, where low cytosolic calcium concentration prevents formation of the calcium-
calmodulin complex, a portion of the C-terminal peptide chain interacts with a regulatory site to
inhibit calcium transport. B: Binding of the calcium-calmodulin complex activates the pump by
abolishing the inhibitory effect of the C-terminal peptide.
The Sodium/Calcium Exchanger

The sodium/calcium (Na/Ca) exchanger (NCX), the most important mechanism that transports calcium
from the cytosol of cardiac myocytes into the extracellular fluid, is an antiport that uses the sodium
gradient across the plasma membrane, rather than ATP, to energize the uphill calcium flux. The discovery
of this exchanger has a convoluted history. A direct relationship between extracellular calcium and the
strength of cardiac contraction was discovered by Ringer in the late 19th century. However, in the 1940s,
changes in extracellular calcium were found not to modify the strength of cardiac contraction when
sodium concentration was also varied so as to maintain a constant ratio between extracellular calcium
and sodium concentrations (Wilbrandt and Koller, 1948). This unexpected observation suggested that
calcium and sodium compete for binding to an exchanger that can transport either ion in both directions
across the plasma membrane (Láttgau and Niedergerke, 1958). According to this mechanism, binding of
calcium to the extracellular side of the exchanger increases calcium influx, which causes the heart to
contract more strongly, whereas increasing extracellular sodium weakens contraction by causing the
exchanger to bind sodium, rather than calcium, for transport into the cell. Evidence was also obtained
that calcium efflux is determined by the relative concentrations of sodium and calcium at the
intracellular side of the exchanger. The importance of Na/Ca exchange was established by Reuter and
Seitz (1968), who found that the exchanger accounts for ∼80% of the calcium efflux from the mammalian
myocardium.
A simple way to understand Na/Ca exchange is to view the exchanger as a carrier that, after binding
either sodium or calcium both within and outside the cell, shuttles these ions in opposite directions across
the lipid bilayer (Fig. 7-6). The exchange is reversible, its direction depending on the electrochemical
activities of sodium and calcium on either side of the membrane. Although the driving force for uphill
calcium transport out of the cytosol is the sodium gradient across the plasma membrane, the ultimate
source of this energy is the ATP used by the sodium pump to establish the sodium gradient.
Fig. 7-6: Overview of Na/Ca exchange. A: The exchanger (NCX) transports 3 Na+ in either direction across the
plasma membrane in exchange for a single Ca2+; the directions of these fluxes are determined largely by
cytosolic calcium concentration. Calcium efflux by the exchanger is accompanied by an inward depolarizing
current, while calcium influx generates an outward repolarizing current (dotted lines). B: The exchanger can
be represented as a well in which two buckets move in opposite directions; one bucket contains a single
divalent calcium ion, the other three sodium ions.
P.152
Fig. 7-7: The NCX contains nine membrane-spanning α-helices organized in two groups linked by an
intracellular peptide chain; an additional membrane-spanning α-helix serves as a signaling peptide for
trafficking of the protein to the plasma membrane. Two groups of membrane-spanning α-helices form cation-
binding sites that allow sodium and calcium to move across the bilayer, while the intracellular peptide chain
contains regulatory sites that allow the activity of the exchanger to be modified by posttranslational
phosphorylations, changes in cytosolic calcium concentration, and an allosteric effect of ATP.
NCXs are intrinsic membrane proteins that include 10 membrane-spanning α-helices and a large cytosolic
domain (Fig. 7-7). One α-helix is a signal peptide that is removed from the functional exchanger in some
members of this family. The remaining nine α-helices, along with the intervening amino acid sequences,
are organized in two hydrophobic clusters of five and four α-helices, which together participate in cation
exchange across the plasma membrane. The intracellular domain contains sites that modify the activity of
the exchanger in response to protein kinase-catalyzed phosphorylations, changes in cytosolic calcium, and
an allosteric effect of ATP.
Although Na/Ca exchange does not require ATP hydrolysis, its turnover is stimulated by high
concentrations of ATP. This allosteric effect is one example of the general ability of ATP to accelerate ion
fluxes that are mediated by exchangers, channels, and pumps. A corollary to this effect is that Na/Ca
exchange is inhibited when ATP concentration falls which, by reducing calcium efflux, can exacerbate
calcium overload in energy-starved hearts. The exchanger is activated when it is phosphorylated by
protein kinases A and C, and by calcium-calmodulin-dependent protein kinase (CAM kinase). In addition,
elevated intracellular calcium increases the turnover of the exchanger by a regulatory effect that, like
phosphorylation by CAM kinase, helps alleviate calcium overload.
The NCX is electrogenic because it exchanges three sodium ions in exchange for one calcium ion; a ratio
of 4:1 has also been described (Dong et al., 2002). This generates an ionic current, defined as the flux of
positive charge, that flows in the same direction as the sodium flux and is opposite to the movement of
calcium (Fig. 7-6) (this can be remembered as “current follows sodium”). The currents generated by the
exchanger are normally small, and contribute only a few millivolts to membrane potential. However, they
can cause dangerous arrhythmias in patients with calcium-overloaded hearts, where calcium efflux across
the plasma membrane is increased (see Chapters 14, 16, 17, and 18).
P.153
The electrogenicity of Na/Ca exchange also allows membrane potential to influence the direction of ion
transport by the exchanger. In the resting heart, the negative intracellular potential favors the entry of
positively charged sodium ions, and so promotes calcium efflux. Reversal of membrane potential during
systole, when the inside of the cell becomes positively charged, has the opposite effect, to increase
calcium influx by favoring sodium efflux. These effects of membrane potential help the NCX operate in a
manner that favors calcium flux in whatever direction maintains the existing mechanical state: calcium
efflux—which relaxes the heart—is favored by the intracellular negativity during diastole, and calcium
influx—which increases contractility—is favored during systole.
Discovery of Na/Ca exchange provided the key to understanding the positive inotropic effect of cardiac
glycosides, which in the 1950s had been discovered to inhibit sodium transport out of cells (Schátzmann,
1953). These drugs increase myocardial contractility because the increased intracellular sodium caused by
sodium pump inhibition favors sodium efflux by the exchanger, which reduces calcium efflux. By retaining
calcium within the myocytes, this response increases contractility by adding to the amount of calcium
available for release during excitation-contraction coupling (Repke, 1964).
Other Ion Fluxes across the Plasma Membrane
The Sodium Pump

The sodium pump (also called the sodium-potassium ATPase or Na-K-ATPase because its ability to
hydrolyze ATP is stimulated when sodium and potassium are present together) transports sodium out of
the cytosol in exchange for potassium that enters the cell. This exchange of sodium and potassium cations
reduces the electrochemical work of the sodium pump. However, both sodium efflux and potassium influx
are uphill processes, so that the sodium pump requires energy that is derived from ATP hydrolysis. The
impact of the sodium pump on cellular energy expenditure is high, as evidenced by the fact that the
sodium pump accounts for 20% to 30% of the ATP consumed by non-motile tissues like the kidneys.
The major function of the sodium pump in working cardiac myocytes is to exchange the small amount of
sodium that enters the cytosol during each action potential with some of the potassium that leaves the
cytosol during repolarization. The electrochemical gradients established by the sodium pump also
contribute to the driving force for the sodium currents that depolarize the working myocytes of the atria
and ventricles, and the Purkinje fibers, and for the potassium currents that repolarize all regions of the
heart (Chapter 13). The sodium gradient across the plasma membrane energizes calcium efflux by the
NCX (see above), proton efflux by the NHE (see below), and the active transport of several substrates and
metabolites across the plasma membrane.
The sodium pump includes three subunits (Fig. 7-8). The largest is the α-subunit, a P-type ion pump
ATPase (see Fig. 7-3) that contains 10 membrane-spanning α-helices and a large intracellular domain. The
sites that bind potassium for transport into the cytosol are on the extracellular side of the α-subunit,
while sodium binds on the cytosolic side. The smaller β-subunit, which is a glycoprotein that includes a
single membrane-spanning domain, participates in trafficking of the sodium pump to the plasma
membrane and regulates cation-binding affinity. The γ-subunit (phospholemman) is a substrate for
phosphorylations by protein kinases A and C that stimulate the sodium pump. Pump activity is also
regulated by acylation of the α-subunit, glycosylation of the β-subunit, and an allosteric effect of high
P.154
ATP concentration that stimulates the pump. Long-term changes in pump activity, seen in diseased hearts,
result from isoform shifts and changes in the concentration of the pump in the plasma membrane (see
chapter 18).
Fig. 7-8: The sodium pump. The large cytoplasmic domain of the α-subunit contains a site that, when
phosphorylated by ATP, provides the energy needed for active ion transport. The extracellular portions of
several membrane-spanning α-helices bind potassium, while sodium binds to membrane-spanning helices from
the intracellular side of the membrane. Cardiac glycosides bind to a site formed by membrane-spanning α-
helices near the potassium-binding site. The intracellular domain, along with the glycosylated β-subunit and
small γ-subunit regulate sodium pump activity.
The sodium pump is electrogenic because it transports three sodium ions out of the cell in exchange for
two potassium ions that enter the cytosol (Fig. 7-9). This generates a repolarizing current which, like that
generated by Na/Ca exchange (Fig. 7-8), is in the same direction as the sodium flux. The contribution of
this current to membrane potential is small, usually <10 mV. Under some circumstances this repolarizing
current can be of functional importance; for example, the outward current generated by the sodium
pump helps maintain intracellular electronegativity in ischemic myocytes where resting potential is
reduced and sodium “leak” into the cell is increased. Conversely, sodium pump inhibition decreases
intracellular potassium concentration, which causes arrhythmias by reducing the electrochemical gradient
responsible for resting potential (see Chapter 16).
Fig. 7-9: Overview of ion transport by the sodium pump. Because three sodium ions are exchanged
for two potassium ions, the pump generates an outward repolarizing ionic current (dotted line); as
is also true of NCX, current follows sodium.
P.155
Cardiac glycosides, which were found to be useful in treating heart failure more than 200 years ago, have
four major actions, all due to sodium pump inhibition. Their ability to increase myocardial contractility,
which is traditionally viewed as the desired effect, occurs when sodium pump inhibition increases
intracellular sodium concentration near a region of the plasma membrane that is rich in NCX molecules.
Because the higher sodium concentration at the cytosolic side of the membrane inhibits calcium efflux via
the NCX (see above), less calcium leaves the cell which, by increasing intracellular calcium stores,
explains the positive inotropic effect (Chapter 10). The second action, resting membrane depolarization,
caused when reduced potassium influx decreases the electrochemical gradient for potassium across the
plasma membrane, is responsible for many of the arrhythmogenic effects of these drugs (see Chapter 16).
This depolarizing effect is increased by the third action, reduction of the small repolarizing current
generated by the sodium pump. However, many of the beneficial effects of the cardiac glycosides in heart
failure are the result of sodium pump inhibition in the brainstem, which increases vagal activity and
reduces sympathetic tone. These central responses decrease heart rate, contractility, and peripheral
resistance, which have energy-sparing effects that benefit failing hearts; the reduced sympathetic
outflow and increased vagal tone also have beneficial antiproliferative effects (see Chapter 18).
Cardiac glycosides inhibit the sodium pump when they bind to a site made up of several membrane-
spanning α-helices in close proximity to the potassium-binding site (Fig. 7-8) (Ogawa et al., 2009). The
ability of potassium to displace cardiac glycosides from this binding site explains why increased
extracellular potassium reduces the sensitivity of the sodium pump to inhibition by these drugs and,
conversely, the dangerous potentiation of the effects of the cardiac glycosides by hypokalemia.
The sodium pump forms several intermediates between the protein (designated E in the following
discussion) and its substrates (Fig. 7-10). Two non-phosphorylated intermediates, E1 and E2, have
different reactivities to ATP and bind differently to cations: sodium binds to the intracellular side of E1,
while potassium binds to the extracellular side of E2. There are also
P.156
2 phosphorylated intermediates, E1∼P and E2-P. Both contain acyl phosphate bonds, but these bonds have
different energetics: E1∼P is a high-energy intermediate, while a low-energy intermediate, E2-P, is formed
when energy is expended during the active transport of sodium and potassium. These phosphorylated
intermediates resemble those formed during the cross-bridge cycle (Chapter 4), but unlike the contractile
proteins, where the chemical energy liberated from ATP shifts the angle of the cross-bridges to cause
tension development and shortening, the sodium pump uses this energy to perform osmotic work.
Fig. 7-10: Reaction scheme for the sodium pump ATPase. The pump, labeled E, can be in a high-energy E1
state (shaded circles) or a low-energy E2 state (open circles). Enzyme-bound ions can be “occluded” (shaded
rectangles) or exchangeable (open rectangles). High- and low-energy bonds are indicated by tildes (∼) and
dashes (-), respectively, and the energy states of the ions are shown by the subscripts i (intracellular) and o
(extracellular). The reaction, described in the text, begins when sodium binds to the energized pump (upper
left).
The reaction shown in Figure 7-10, which requires magnesium, begins when three sodium ions bind to the
intracellular (cytosolic) side of the sodium pump (E), which is in a high-energy state:
Release of ADP then transfers energy to the bound sodium, which becomes occluded (indicated by
brackets). This raises the activity of the occluded sodium to the higher activity in the extracellular fluid
(Na+o).
Sodium is released into the extracellular fluid and the pump returns to the low energy state, E2-P.
(Sodium release occurs in two steps that, for simplicity, are treated as one in Figure 7-10 and Equation 7-
3.)
The next steps in this reaction, which transport potassium into the cytosol, begin when potassium at its
low activity in the extracellular space (K+o) binds to E2-P.
This is the step where cardiac glycosides inhibit the sodium pump by binding to the extracellular side of
the enzyme. Release of inorganic phosphate then causes the bound potassium, still at its low activity in
the extracellular fluid, to become occluded.
Binding of ATP to the E2•2{K+o} complex begins the transfer of energy that raises the activity of the
occluded potassium to that in the cytosol. However, most of the chemical energy of ATP initially remains
in the nucleotide, so that the activity of the enzyme-bound potassium remains low.
Transfer of energy from ATP to the E2•ATP•2K+o complex brings the pump to a high-energy state (E2) and
increases the activity of the occluded potassium to the higher level within the cytosol (K+i).
Further transfer of energy to the pump completes the reaction by releasing potassium into the cytosol.
P.157
The overall reaction of the sodium pump (see also Fig. 7-10) can be written as:
This reaction can run in reverse in a sequence of steps that couple downhill sodium and potassium fluxes
to the synthesis of ATP from ADP and Pi; however, pump reversal does not occur under physiological
conditions because it is strongly inhibited by ATP.
The Sodium/Hydrogen Exchanger

Because energy production is accompanied by the generation of protons, mechanisms are needed to
prevent intracellular acidosis, which has several deleterious effects on the heart. These include inhibition
of enzymes involved in energy production, a negative inotropic effect caused when protons compete with
calcium for binding to the troponin complex, and inhibition of calcium fluxes that relax the heart. The
normal intracellular pH in the heart is about 7.2, which is more alkaline than would be expected if
protons were distributed simply according to their electrochemical gradient. Transport of protons out of
the cell, which maintains this intracellular alkalinity, depends on a symport, the sodium/bicarbonate
transporter, and three antiports. The latter are a chloride/bicarbonate exchanger, a chloride/hydroxyl
exchanger and, most important, a sodium/hydrogen (Na/H) exchanger (NHE) that uses energy derived
from the sodium gradient across the plasma membrane to transport protons uphill, out of the cytosol.
Because the stoichiometry between proton efflux and sodium influx by NHE1 is 1:1, Na/H exchange is
electrically neutral.
NHEs are members of a large and ancient superfamily of transporters found in both prokaryotes and
eukaryotes that includes uniports (one ion moving in one direction), symports (two ions moving in one
direction), and antiports (two ions moving in opposite directions). NHE1, the major NHE isoform in the
heart, is a large glycoprotein that contains 12 membrane-spanning α-helices, an intracellular N-terminal
domain that participates in the ion exchange reaction, and a large intracellular C-terminal domain that
contains several regulatory sites (Fig. 7-11). The C-terminal domain also links NHE1 to the actin
cytoskeleton through a cytoskeletal “organizer” called ezrin, which is one of a family of ezrin-radixin-
moesin (ERM) proteins that connect membrane proteins and the cytoskeleton and allow the exchanger to
participate in proliferative signaling (Bretscher et al., 2002).
Intracellular acidosis stimulates Na/H exchange in two ways: a direct response to increased proton
concentration and an indirect effect that occurs when the protons bind to an intracellular pH “sensor”
that increases the turnover of the antiport. Both increase contractility because the increased proton
efflux is coupled to sodium influx that, by increasing intracellular sodium, reduces calcium efflux by the
NCX. NHE1 is also activated by an allosteric effect of ATP and phosphorylation by a calcium-calmodulin-
activated protein kinase; additional protein kinases phosphorylate the exchanger in response to agonist-
binding to G-protein-coupled receptors and receptor tyrosine kinases, and by cytoskeleton-mediated
signals. Impaired oxidative phosphorylation accelerates anaerobic glycolysis in energy-starved hearts (see
Chapter 2); this increases lactate production and releases protons. Exchange of these protons for sodium
by the NHE1 increases cytosolic sodium, which, by decreasing calcium efflux by the NCX, can initiate a
vicious cycle where increasing calcium overload worsens energy starvation by increasing energy
utilization.
P.158
Fig. 7-11: The sodium/hydrogen exchanger contains 12 membrane-spanning α-helices and a glycosylated
extracellular peptide between the first and second helices. The N-terminal region and adjacent membrane-
spanning helices participate in ion transport, while the large intracellular C-terminal peptide chain contains
phosphorylation, proton-binding, ATP-binding, and calcium-calmodulin-binding sites that regulate the
exchanger. Ezrin, a cytoskeletal organizer, links the exchanger to actin microfilaments, which allow the
exchanger to participate in signal transduction.
The Intracellular Calcium Cycle
Because ionized calcium concentration within the sarcoplasmic reticulum is several orders of magnitude
higher than that in the cytosol, calcium release is a passive, downhill, flux, and calcium uptake into this
internal membrane system requires the expenditure of energy. Unlike calcium fluxes across the
sarcolemma, which generate ionic currents (see above), movements of positively charged calcium ions
into and out of the sarcoplasmic reticulum do not generate electrical currents. This is because these
internal membranes contain both nonspecific anion channels that allow negatively charged chloride and
phosphate anions to accompany the calcium ions (Beil et al., 1977), and trimeric intracellular cation
(TRIC) channels that allow other cations to neutralize charge movements (see below).
The sarcoplasmic reticulum includes two regions: the subsarcolemmal cisternae, which contain the
calcium release channels that control the downhill calcium efflux that initiates systole, and an extensive
sarcotubular network that contains a densely packed array of calcium pump ATPase proteins that relax
the heart (see Chapter 1). The subsarcolemmal cisternae form composite structures with the plasma
membrane and t-tubules, called dyads, which mediate the ability of plasma membrane depolarization to
initiate calcium release from the sarcoplasmic reticulum. This occurs when calcium entry from the
extracellular fluid into the cytosol through L-type calcium channels in the plasma membrane and t-
tubules opens calcium release channels in the subsarcolemmal cisternae that activate contraction by
allowing calcium to flow out of the sarcoplasmic reticulum into the cytosol. During diastole, this calcium
is removed from the cytosol by a calcium pump that transports this cation back into the sarcotubular
network.
P.159
Calcium Release from the Sarcoplasmic Reticulum

Two different mechanisms were initially proposed to explain how plasma membrane depolarization
initiates calcium release from the sarcoplasmic reticulum (Fig. 7-12). The first is a mechanical coupling in
which depolarization of the t-tubule opens a calcium channel in the adjacent subsarcolemmal cisternae
by removing a “plug” (Fig. 7-12A). (This mechanism was initially called the “plumber's helper model”
because the plug was pictured as being similar to devices used to open blocked drains.) In the second
mechanism, a small amount of calcium that crosses the plasma membrane from the extracellular fluid
induces the release of a much larger amount of activator calcium from within the sarcoplasmic reticulum
(Fig. 7-12B). Called “calcium-triggered calcium release” (Fabiato, 1983), this mechanism is analogous to a
flintlock musket, where the small primer charge explodes when the flint strikes the primer pan—like
calcium entry across the plasma membrane—ignites the larger amount of powder within the barrel of the
musket—which is analogous to the larger calcium release from within the sarcoplasmic reticulum.
In both of the mechanisms depicted in Figure 7-12, the signal initiated by plasma membrane
depolarization causes a voltage-dependent conformational change in a protein related to an L-type
calcium channel (see Chapter 13). In skeletal muscles, the conformational change in the plasma
membrane protein shifts the position of an intracellular peptide that unplugs the intracellular channel
(Fig. 7-12A). In the heart, a similar voltage-dependent conformational change
P.160
in a different L-type channel protein allows a small influx of “trigger” calcium that binds to and opens the
intracellular calcium channel (Fig. 7-12B). This ability of different L-type calcium channels to activate
contraction by removing a plug or by supplying a small amount of trigger calcium is one of many examples
of the way that homologous structures can use different mechanisms to bring about similar responses.
Fig. 7-12: Two mechanisms by which action potentials propagated across the plasma or t-tubular membrane
can release calcium from the sarcoplasmic reticulum. A: Mechanical coupling. Plasma membrane
depolarization changes the conformation of a voltage-regulated peptide so as to remove a “plug” that
occludes the pore in the sarcoplasmic reticulum calcium release channel. B: Calcium-triggered calcium
release. A small amount of calcium that enters the cytosol through an L-type plasma membrane calcium
channel binds to and opens the intracellular channel that releases a much larger amount of calcium from
within the sarcoplasmic reticulum.
Couplons and Calcium Sparks

What appears to be a simple increase in the average level of ionized calcium within activated myocytes
during excitation-contraction coupling is, in fact, the average of a large number of localized increases in
cytosolic calcium. The latter, called calcium sparks, appear when calcium is released in small bursts from
macromolecular clusters that include L-type calcium channels, intracellular calcium release channels, and
several regulatory proteins. The bursts of calcium release from these clusters, often called couplons,
cause the appearance of the calcium sparks.
The closure of calcium release channels after a calcium spark can be viewed as paradoxical because the
increased cytosolic calcium might be expected to reopen these channels. However, the calcium release
channels are much less sensitive to an increase in “global” intracellular calcium than to the highly
localized rise in ionized calcium within the couplon, and the calcium signal generated by the couplon is
very brief (this is evidenced by the short duration of the calcium sparks). Together, these features
constrain calcium release to a series of brief openings of individual couplons.
Calcium Release Channels
The electron dense “feet” that lie between the plasma membrane and the membrane of the
subsarcolemmal cisternae in the dyad (see Chapter 1) are the intracellular calcium release channels that
are often called ryanodine receptors (RyR) because they bind to this plant alkaloid. These large
membrane proteins are made up of four subunits, each of which includes several membrane-spanning α-
helices segments and a large cytosolic domain (Fig. 7-13). The membrane-spanning α-helices of the four
subunits surround a single central pore, while each of the four cytosolic domains that make up the foot
contains a laterally located pore; the latter are not connected to the central pore when the channel is in
its closed state. Binding to the “trigger” calcium that enters the cytosol via L-type calcium channels
causes a conformational change that aligns the pores in the four subunits with the pore in the membrane-
spanning domain (Fig. 7-14C). This opens the calcium release channel and allows calcium to flow out of
the sarcoplasmic reticulum.
Cardiac calcium release channels are highly regulated. They can be phosphorylated by several protein
kinases, including a CAM kinase that inhibits channel opening at high cytosolic calcium concentrations;
this exerts a protective effect that reduces calcium release from the sarcoplasmic reticulum in calcium-
overloaded myocytes. Phosphorylation of these channels by protein kinase A participates in the inotropic
response to sympathetic stimulation by increasing calcium release. Additional proteins regulate calcium
release from the sarcoplasmic reticulum. These include triadin and junctin, which increase calcium flux
through the channel; calsequestrin and histidine-rich protein, which regulate calcium release channel
opening and bind calcium within the sarcoplasmic reticulum (see below); junctophilin, which mediates
interactions between the calcium release channel and the plasma membrane, and TRIC (trimeric
intracellular cation channel) which mediates a cation flux that is opposite to, and so neutralizes, the
current generated by calcium efflux (Fig. 7-13). FKBP12 (also called calstabin2), which binds cyclosporine
P.161
P.162
and other immunosuppressive drugs, coordinates the gating of the four channel subunits, which both
stabilizes and facilitates opening of the cardiac calcium release channels.
Fig. 7-13: Schematic diagram of the proteins that release calcium from the cardiac sarcoplasmic reticulum in
response to plasma membrane depolarization. This cluster is often referred to as a “couplon.” A plasma
membrane L-type calcium channel (above) is shown in close proximity to one of the four subunits of the
calcium release channel (shaded, below). The number of α-helical membrane-spanning segments in the latter
has not been established; six are shown in this figure. Calcium binding to sites on the cytoplasmic domain of
the calcium release channel (the foot) opens the channel. Other proteins that participate in calcium release
include junctin and triadin, which regulate calcium flux through the channel. Calsequestrin and the histidine-
rich calcium-binding protein bind calcium within the sarcoplasmic reticulum and regulate opening of the
calcium release channel, junctophilin serves a regulatory function, and the TRIC channel carries a counter-
ion that neutralizes the current generated when calcium ions cross the sarcoplasmic reticulum membrane.
Fig. 7-14: Schematic representation of a dyad showing the relationship between an L-type calcium
channel and a calcium release channel. A: View through the plane of the bilayer showing the
plasma membrane (above) and subsarcolemmal cisternal membrane (below). The latter contains
the sarcoplasmic reticulum calcium release channel, which is made up of a membrane-spanning
domain and a large “foot” that projects into the cytosolic space. Dark areas represent the pores
through which calcium crosses the membrane when the channels are in the open state. B: The
membrane-spanning domain of the intracellular calcium release channel, as seen from
subsarcolemmal cisterna space (left), and the foot, as seen from the dyad space (right). The
membrane-spanning domain contains a central pore while each of the four foot subunits contains a
radial pore. C: Depiction of channel opening. In the closed state, the four radial pores do not
connect with the central pore. Alignment of the central pore with the radial pores opens the
channel and releases calcium from the sarcoplasmic reticulum.
Inositol trisphosphate-gated calcium channels (InsP3 receptors), which are smaller and open and close
more slowly than the calcium release channels described above, play a major role in delivering the
calcium that activates smooth muscle contraction. In the heart, calcium released through a small number
of InsP3-gated calcium channels regulates protein synthesis, cell cycling, and apoptosis, and may modify
resting tension.
Calcium Uptake by the Sarcoplasmic Reticulum

Because the calcium that activates contraction in the adult mammalian heart is derived largely from the
sarcoplasmic reticulum, most of the calcium released into the cytosol during systole must be pumped
back into this internal membrane system during diastole. This is effected by the sarcoplasmic reticulum
calcium pump [also called the sarco(endo)plasmic reticulum calcium
P.163
pump ATPase or SERCA], one of the P-type ion pumps depicted in Fig. 7-3, which is found in densely
packed arrays within the membranes of the sarcotubular network (Fig. 7-15). The cardiac isoform (SERCA
2a), like the plasma membrane calcium pump ATPase described above, contains a calcium-transport site
made up of several α-helical membrane-spanning segments, and a large intracellular domain with an ATP
phosphorylation site and several regulatory sites.
Fig. 7-15: Three-dimensional depiction of the membrane of the sarcotubular network; the
cytosolic surface is above and the lumen below. The calcium pump ATPase molecules are packed
into the bilayer with most of their mass projecting into the cytosol.
Coupling of ATP hydrolysis to calcium transport by the calcium pump (Fig. 7-16) is similar to the coupling
of ATP hydrolysis to alkali metal ion transport by the sodium pump (Fig. 7-10), except that the calcium
pump does not exchange calcium for a counter-ion. In the following discussion, E1 and E2 refer to non-
phosphorylated states of the calcium pump that have different reactivities to ATP and calcium, while the
two phosphorylated states, E1∼P and E2-P, are high- and low-energy intermediates, respectively. The
subscripts refer to calcium activity in the cytosol (Ca2+c) and within the sarcoplasmic reticulum (Ca2+s).
Fig. 7-16: Reaction scheme for the sarcoplasmic reticulum ATPase. The pump, labeled E, can be in
a high-energy E1 state (shaded circles) or a low-energy E2 state (open circles). Enzyme-bound
calcium can be “occluded” (shaded rectangles) or exchangeable (open rectangle). High- and low-
energy bonds are indicated by tildes (∼) and dashes (-), respectively, and the energy states of
calcium ions are shown by the subscripts c (within the cytosol) and s (within the sarcoplasmic
reticulum). The reaction, described in the text, begins at the upper left, when ATP binds to the
enzyme.
P.164
The reaction begins when two calcium ions at low activity within the cytosol bind to the pump that,
because it is bound to ATP, is in a high-energy state.
In the next step, release of ADP transfers energy to the complex to form E1∼P•2Ca2+c in which the bound
calcium, although still at its low cytosolic activity, becomes occluded and so is unable to exchange with
cytosolic calcium (indicated by brackets).
The activity of the occluded calcium is then raised to the high level within the sarcoplasmic reticulum
(Ca2+s) in a reaction that uses energy in E1∼P, which returns to the low-energy state in E2-P.
Calcium is then released into the region of higher calcium activity within the sarcoplasmic reticulum
after which the low-energy phosphoenzyme E2-P is dephosphorylated to liberate Pi and form E2.
Binding of ATP to E2, by transferring energy to the pump, converts the latter to the high-energy E1.
The energy added when ATP binds to E2 increases the affinity of the calcium-binding site, which allows
the E1•ATP complex to bind two calcium ions at their low activity within the cytosol (Equation 7-10).
The overall reaction of the calcium pump ATPase can be written as:
All of the steps in the calcium pump ATPase reaction, like those of the Na-K-ATPase, are reversible, so
that under special conditions in vitro, the pump can be made to run backward in a reaction that couples
the energy of the calcium gradient across the sarcoplasmic reticulum to the synthesis of ATP from ADP
and Pi. However, to run the calcium pump in reverse, cytosolic ADP and Pi levels must be high, cytosolic
calcium and ATP levels must be low, and the calcium concentration within the sarcoplasmic reticulum
must be high. Because these conditions are not normally found in cardiac myocytes, the pump cannot
mediate a net flux of calcium out of the sarcoplasmic reticulum under physiological conditions.
The most important regulator of calcium uptake into the sarcoplasmic reticulum is the direct effect of
cytosolic calcium that stimulates pump turnover (Equation 7-10). Cytosolic calcium also stimulates
calcium uptake indirectly by forming a calcium-calmodulin complex that, by activating a CAM kinase,
phosphorylates a regulatory site on the pump. Energy starvation inhibits calcium uptake for several
reasons. Loss of the allosteric effect of the normally high concentration
P.165
of cytosolic ATP in the heart slows the calcium pump, which is one reason why energy starvation, by
attenuating this regulatory effect, impairs relaxation in energy-starved hearts. Also important is reduction
of the free energy available from ATP hydrolysis (-ΔG) caused by the decreased ATP concentration that,
along with a larger increase in ADP concentration, slows calcium uptake into the sarcoplasmic reticulum
(Tian and Ingwall, 1996). Together, these effects slow relaxation in ischemic and failing hearts (see
Chapters 17 and 18).
Phospholamban and Sarcolipin
Two closely related membrane proteins, phospholamban and sarcolipin, regulate the sarcoplasmic
reticulum calcium pump; both contain a single α-helical membrane-spanning segment and both are
substrates for phosphorylation by protein kinase A. Phospholamban is present in the atria and ventricles of
mammalian hearts, while sarcolipin is found mainly in the atria. The dephospho forms of both proteins
decrease the calcium sensitivity of the calcium pump, which slows relaxation. Reduced calcium uptake
into the sarcoplasmic reticulum also decreases contractility by allowing more of this activator to be
transported out of the cytosol into the extracellular space, which reduces intracellular calcium stores.
The inhibitory effects of phospholamban are reversed by protein kinase A-catalyzed phosphorylation (Fig.
7-17), which accelerates relaxation and increases contractility; these responses contribute to the
increased cardiac output following sympathetic stimulation. The dephosopho form of sarcolipin also
inhibits relaxation and reduces contractility by slowing calcium uptake into the sarcoplasmic reticulum.
Phosphorylation of phospholamban by calcium-calmodulin kinase stimulates calcium uptake into the
sarcoplasmic reticulum, which helps protect the heart in situations of calcium overload.
Comparison of the stimulatory effect of the calcium-calmodulin-binding domain of the plasma membrane
calcium pump (Fig. 7-5) with that of phospholamban (Fig. 7-17) provides a fascinating example of the way
that homologous structures in different systems can generate similar responses. Although the inhibitory
action of the calcium-calmodulin-binding domain of the plasma membrane calcium pump resembles that
of phospholamban, in the former this regulatory peptide is part of the pump protein whereas
phospholamban is a separate protein. Homologies between these two peptide chains suggest that a
regulatory peptide in the ancestral calcium pump evolved in two ways: in the plasma membrane calcium
pump, it remained attached to the pump molecule where it responds to a calcium-activated signal,
whereas the homologous region SERCA2a became a separate protein that mediates a signal initiated by
cyclic AMP.
Calcium-Binding Proteins within the Sarcoplasmic Reticulum

Much of the calcium taken up by the sarcoplasmic reticulum is associated with calcium-binding proteins in
the subsarcolemmal cisternae. These proteins include calsequestrin, each molecule of which contains
between 18 and 50 calcium-binding sites, histidine-rich calcium-binding protein, sarcalumenin, and
calreticulin. All help in maintaining a low ionized calcium concentration within the sarcoplasmic
reticulum, which is important because high levels of intraluminal calcium inhibit the calcium pump.
These calcium-binding proteins also serve a number of regulatory functions. Calsequestrin and the
histidine-rich calcium-binding protein regulate calcium release, sarcalumenin regulates calcium pump
activity, and calreticulin regulates cardiac development. Many of these regulatory effects are modified by
protein kinase-catalyzed phosphorylation reactions.
P.166
Fig. 7-17: Regulation of the cardiac sarcoplasmic reticulum calcium pump by phospholamban. A: In
the basal state phospholamban inhibits calcium transport when it interacts with the intracellular
regulatory domain. B: Phosphorylation of phospholamban reverses this inhibitory effect, which
accelerates calcium transport and increases calcium stores in the sarcoplasmic reticulum.
P.167
Mitochondria
The possibility that mitochondrial calcium uptake and release participate in cardiac excitation-
contraction coupling and relaxation has been considered for decades. A large electrochemical gradient for
calcium across the mitochondrial membrane allows calcium to enter these organelles by a passive
downhill flux, and both a NCX and a channel related to the calcium release channels of the sarcoplasmic
reticulum (“ryanodine receptors”) are found in mitochondria. These homologues, which are likely to have
been the result of gene transfer between the mitochondrial genome and the nuclear DNA of eukaryotes
during evolution, do not mean that the related proteins must carry out the same functions in different
cellular organelles.
Both the calcium release channels and a mitochondrial calcium uniporter have been suggested to help
relax the heart by mediating calcium uptake into mitochondria. However, it is not clear whether the
calcium affinity and turnover of the uniporter are sufficiently rapid to relax the heart, and the role of the
mitochondrial calcium release channels is controversial. The NCX, which like its plasma membrane
homologue exchanges three sodium ions for each calcium ion, can mediate uphill transport of calcium out
of the mitochondria, but the increases in sodium concentration within mitochondria needed to accelerate
this exchange have not been observed. For these and other reasons, a role for mitochondrial calcium
fluxes in cardiac excitation-contraction coupling and relaxation remains controversial.
In calcium-overloaded myocytes, mitochondrial calcium uptake can buffer increases in cytosolic calcium.
However, excessive mitochondrial calcium accumulation impairs oxidative phosphorylation, and in
ischemic hearts, where energy starvation is severe, mitochondrial calcium uptake causes calcium
phosphate to precipitate within mitochondria (Shen and Jennings, 1972).
Overview of the Extracellular and Intracellular Calcium Cycles

The extracellular and intracellular calcium cycles in the adult mammalian heart (Fig. 7-18) involve five
pools, or compartments; the extracellular space, sarcoplasmic reticulum, cytosol, contractile proteins,
and mitochondria (Fig. 7-18A). The major calcium fluxes are shown in Figure 7-18B, where upward arrows
represent active fluxes and downward arrows passive fluxes; the thickness of each arrow is roughly
proportional to the amount of the calcium flux.
In the extracellular calcium cycle, calcium influx across the plasma membrane is mediated by voltage-
gated L-type calcium channels (arrow A), while active calcium efflux is effected by the plasma membrane
calcium pump (PMCA, arrow B1) and Na/Ca exchanger (NCX, arrow B2). The calcium that enters the
cytosol from the extracellular space binds directly to the contractile proteins (arrow A1) and opens
calcium release channels in the sarcoplasmic reticulum (arrow A2, calcium-triggered calcium release).
The calcium transported into the extracellular space by the PMCA and NCX after it has dissociated from
troponin is indicated by arrow D1.
In the intracellular calcium cycle, the calcium fluxes out of (arrow C), into (arrow D), and within (arrow
G) the sarcoplasmic reticulum are greater than those of the extracellular calcium cycle (arrows A, B1, and
B2). The contractile proteins are activated when calcium binds to troponin C (arrow E), while the heart
relaxes when lowering of cytosolic calcium concentration causes the activator to dissociate from the high
affinity EF-hand calcium binding sites on this protein (arrow F). The limited ability of mitochondria to
buffer high cytosolic calcium levels is shown by the double arrow H.
P.168
P.169
Fig. 7-18: Major structures (A) and calcium fluxes (B) that control cardiac excitation-contraction
coupling and relaxation. In A, Calcium “pools” are labeled by bold capital letters. In B, the
thickness of the arrows indicates the magnitude of the calcium fluxes, while their vertical
orientations describe their “energetics”: downward arrows represent passive calcium fluxes;
upward arrows represent energy-dependent active calcium transport. The calcium that enters the
cell from the extracellular fluid via L-type calcium channels (arrow A) directly activates the
contractile proteins (arrow A1) and triggers calcium release from the sarcoplasmic reticulum
(arrow A2). Calcium is actively transported out of the cytosol into the extracellular fluid by the
plasma membrane calcium pump ATPase (PMCA; arrow B1), and the Na/Ca exchanger (NCX) (arrow
B2). The sodium that enters the cell in exchange for calcium (dashed line) is pumped out of the
cytosol by the sodium pump. Calcium fluxes mediated by the sarcoplasmic reticulum are calcium
efflux from the subsarcolemmal cisternae via calcium release channels (arrow C) and calcium
uptake into the sarcotubular network by the calcium pump ATPase (arrow D). (Arrow D1 identifies
the small amount of calcium that, after dissociating from troponin, is transported into the
extracellular space by the PMCA and NCX.) Calcium diffuses within the sarcoplasmic reticulum from
the sarcotubular network to the subsarcolemmal cisternae (arrow G), where it forms a complex
with calsequestrin and other calcium-binding proteins. Calcium binding to (arrow E) and
dissociation from (arrow F) high-affinity calcium-binding sites of troponin C activate and inhibit the
interactions of the contractile proteins. Calcium movements into and out of mitochondria (arrow H)
buffer cytosolic calcium concentration. The extracellular calcium cycle is shown as arrows A, B1,
and B2, while the intracellular cycle involves arrows C, D, and G.
P.170
Implications of the Energetics of Calcium Channels and Calcium

Pumps on Excitation-Contraction Coupling and Relaxation in the
Heart
Downhill ion flux through a single ion channel is much more rapid than active transport by an ion pump;
this is apparent when the velocity at which calcium ions pass through an intracellular calcium release
channel is compared with that of calcium uptake by the calcium pump of the sarcoplasmic reticulum
(Table 7-3). The downhill flux rate, calculated from the calcium conductance of the intracellular calcium
release channels (∼75 pS; Bers, 1991), is ∼10 times greater than that through an L-type calcium channel
(5 to 9 pS; Tsien and Tsien, 1990). As the latter can carry ∼75,000 calcium ions/sec when extracellular
calcium is 1 mM (calculated from data provided by Tsien, 1983), calcium flux into the cytosol through a
single intracellular calcium release channel is ∼750,000 ions/sec, assuming that calcium concentration
within the sarcoplasmic reticulum is also 1 mM.
The velocity of calcium uptake by each sarcoplasmic reticulum calcium pump, which has been measured
directly, is only ∼30 ions/sec at 37° (Shigekawa et al., 1976). Although the number of densely packed
calcium pump proteins in the sarcotubular network (Figure 7-15) is ∼170 fold greater than that of the
intracellular calcium release channels in the heart (Bers, 1991), the maximum rate of calcium delivery
into the cytosol via the calcium release channels during systole is ∼150 times faster than the maximum
rate of calcium uptake into the sarcoplasmic reticulum during diastole (Table 7-3). The actual difference
in the intact heart is
P.171
much greater because calcium uptake into the sarcoplasmic reticulum slows significantly as cytosolic
calcium concentration decreases during diastole.
Table 7-3 Comparison of Maximum Calcium Fluxes Rates Into and Out of the Cardiac
Sarcoplasmic Reticulum
Maximum calcium flux rates (ions/sec)
Intracellular calcium release channela 750,000
Sarcoplasmic reticulum calcium pumpb 30
Maximum release rate per channel/maximum uptake rate per pump 25,000
Content of pump and channel molecules (nmol/kg wet weight)c
Sarcoplasmic reticulum calcium pump ATPase 6,000
Intracellular calcium release channels 36
Calcium pump ATPase molecules per calcium release channel 170
Maximum calcium flux ratio
Maximal rate of calcium entry during systole/maximal rate of calcium ∼150

removal during diastole
aCalculated from the calcium conductance of 0.025 pA for an L-type plasma membrane
channel in 1 mM extracellular calcium, where a single channel current of 1 pA is equivalent
to a flux of 3,000,000 divalent cations/sec (Tsien, 1983), and the relative conductances of
plasma membrane and intracellular calcium channels (Tsien and Tsien, 1990; Bers, 1991).
bShigekawa et al., 1976.
c
Bers, 1991.
The differences between the rates of passive calcium diffusion into the cytosol and active calcium
transport out of the cytosol illustrate the well-established precept that it is easier to get into trouble
than out of trouble. The major reason that it is “easier” for heart to contract by allowing calcium to
diffuse into the cytosol than to relax by pumping calcium out of the cytosol is that the passive flux down
an electrochemical gradient is much faster than uphill transport. Furthermore, cardiac myocytes cannot
compensate for an increase in the rate at which calcium diffuses into the cytosol without an increase in
calcium concentration during diastole. The regulatory mechanisms that allow increased cytosolic calcium
to stimulate calcium removal from the cytosol of normal hearts' efflux described in this chapter can
compensate for increased cytosolic calcium in normal hearts. However, a rise in cytosolic calcium
concentration in ischemic or failing hearts, which are generally energy-starved, can impair relaxation by
increasing calcium binding to troponin.
By activating contraction and inhibiting relaxation in diseased hearts, increased cytosolic calcium sets the
stage for a vicious cycle in which energy starvation increases cytosolic calcium concentration, which
increases energy expenditure, worsens energy starvation, etc. This vicious cycle can lead to cell death, so
that calcium overload can be an important cause of the myocardial necrosis and reactive fibrosis
commonly seen in patients with heart disease (Katz and Reuter, 1979).
Bibliography
General
Berridge MJ, Bootman MD, Rederick HL. Calcium signalling: dynamics, homeostasis and remodelling.
Mol Cell Biol 2003;4:517–529.
2001.
Langer GA, ed. The Myocardium. 2nd ed. San Diego: Academic Press, 1997.
Maier LS, Bers DM. Role of Ca(2+)/calmodulin-dependent protein kinase (CaMK) in excitation-
contraction coupling in the heart. Cardiovasc Res 2007;73:631–640.
Sachs G, ed. Symposium on ion motive ATPases. Acta Physiol Scand 1998;163. Suppl. 643.
L-Type Calcium Channels (Dihydropyridine Receptor)
Bodi I, Mikala G, Koch SE, et al. The L-type calcium channel in the heart: the beat goes on. J Clin
Invest 2005;115:3306–3317.
Calcium Release Channels (Ryanodine Receptor)
Cheng H, Lederer WJ. Calcium sparks. Physiol Rev 2008;88:1491–1545.
Dulhunty A, Wei L, Beard N. Junctin—the quiet achiever. J Physiol (London) 2009;587:3135–3137.
Fill M, Copello JA. Ryanodine calcium receptor calcium release channels. Physiol Rev 2002;82:893–
922.
Franzini-Armstrong C, Protasi F, Ramesh V. Shape, size, and distribution of Ca2+ release units and
couplons in cardiac muscles. Biophys J 1999;77:1528–1539.
Garbino A, van Oort RJ, Dixit SS, et al. Molecular evolution of the junctophilin gene family. Physiol
Genomics 2009;37:175–186.
P.172
Rossi D, Sorrentino V. Molecular genetics of ryanodine receptors Ca2+-release channels. Cell Calcium
2002;32:307–319.
Yamazaki D, Komazaki S, Nakanishi H, et al. Essential role of the TRIC-B channel in Ca2+ handling of
alveolar epithelial cells and in perinatal lung maturation. Development 2009;136:2355–2361.
The Plasma Membrane Calcium Pump ATPase (PMCA)
Carafoli E. Biogenesis: plasma membrane calcium ATPase: 15 years of work on the purified enzyme.
FASEB J 1994;8:993–1002.
Oceandy D, Stanley PJ, Cartwright EJ, et al. The regulatory function of plasma membrane Ca2+
ATPase (PMCA) in the heart. Biochem Soc Trans 2007;35:927–930.
DiPolo R, Beaugá L. Metabolic pathways in the regulation of invertebrate and vertebrate Na+/Ca2+
exchange. Biochim Biophys Acta 1999;1422:57–71.
Egger M, Niggli E. Regulatory function of Na–Ca exchange in the heart: milestones and outlook. J
Memb Biol 1999;168:107–130.
Schillinger W, Fiolet JW, Schlotthauer K, et al. Relevance of Na+–Ca2+ exchange in heart failure.
Cingolani HE, Ennis IL. Sodium-hydrogen exchanger, cardiac overload, and myocardial hypertrophy.
Circulation 2007;115:1090–1100.
Karmayzn M, Kili A, Javadov S. The role of NHE-1 in myocardial hypertrophy and remodeling. J Mol
Cell Cardiol 2008;44:64–653.
Lytton J. Na+/Ca2+ exchangers: three mammalian gene families control Ca2+ transport. Biochem J
2007; 406:365–382.
Marger MD, Saier MH Jr. A major superfamily of transmembrane facilitators that catalyse uniport,
symport and antiport. Trends Biochem Sci 1993;18:13–20.
Orlowski J, Grinstein S. Diversity of the mammalian sodium/proton exchanger SLC9 gene family.
Pflugers Arch Eur J Physiol 2004;447:549–565.
Putney LK, Denker SP, Barber DL. The changing face of the Na+/H+ exchanger, NHE1: structure,
regulation, and cellular actions. Annu Rev Pharmacol Toxicol 2002;42:527–552.
Slepkov ER, Rainey JK, Sykes BD, et al. Structural and functional analysis of the Na+/H+ exchanger.
Biochem J 2007;401:623–633.
Wakabayashi S, Shigekawa M, Pouyssegur J. Molecular physiology of vertebrate Na+/H+ exchangers.

Physiol Rev 1997;77:51–74.
The Sodium Pump
Blanco G, Mercer RW. Isozymes of the Na-K-ATPase: heterogeneity in structure, diversity in function.
Am J Physiol 1998;275:F633–F650.
Dempski RE, Lustug J, Friedrich T, et al. Structural arrangement and conformational dynamics of the
g subunit of the Na+/K+ ATPase. Biochemistry 2008;47:257–266.
Han F, Bossuyt J, Despa S, et al. Phospholemman phosphorylation mediates protein kinase CV-
dependent effects on Na+/K+ pump function in cardiac myocytes. Circ Res 2006;99:1376–1383.
P.173
Han F, Tucker AL, Lingrel JB, et al. Extracellular potassium dependence of the Na+-K+-ATPase in
cardiac myocytes: isoform specificity and effect of phospholemman. Am J Physiol Cell Physiol
2009;297: C699–C705.
Kaplan JH. Biochemistry of Na,K-ATPase. Annu Rev Biochem 2002;71:511–535.
Levi AJ, Boyett MR, Lee CO. The cellular actions of digitalis glycosides on the heart. Prog Biophy
Molec Biol 1994;62:1–54.
Lingrel JB, Croyle ML, Woo AL, et al. Ligand binding sites of Na K-ATPase. Acta Physiol Scand 1998;
163(Suppl 643):69–77.
Calcium Uptake by the Sarcoplasmic Reticulum
MacLennan DH, Kranias EG. Phospholamban: a crucial regulator of cardiac contractility. Nat Rev Mol
Cell Biol 2003;4:566–577.
Periasamy M, Bhupathy P, Babu GJ. Regulation of sarcoplasmic reticulum Ca2+ ATPase pump
expression and its relevance to cardiac muscle physiology and pathology. Cardiovasc Res
2008;77;265:265–273.
Calcium-Binding Proteins within the Sarcoplasmic Reticulum
Beard NA, Laver DR, Dulhunty AF. Calsequestrin and the calcium release channel of skeletal and
cardiac muscle. Prog Biophys Mol Biol 2004;85:3369.
Bers DM. Macromolecular complexes regulating cardiac ryanodine receptor function. J Mol Cell
Cardiol 2004;37:417–429.
Jiao Q, Bai Y, Akaike T, et al. Sarcalumenin is essential for maintaining cardiac function during
endurance exercise training. Am J Physiol Heart Circ Physiol 2009;297:H576–H582.
Lynch JM, Chilibeck K, Qui Y, et al. Assembling pieces of the cardiac puzzle; calreticulin and
calcium-dependent pathways in cardiac development, health, and disease. Trends cardiovasc med
2006;16:65–69.
Pritchard TJ, Kranias EG. Junctin and the histidine-rich Ca2+ binding protein; potential roles in heart
failure and arrhythmogenesis. J Physiol (London) 2009;587:3125–3133.
Royer L, Rios E. Deconstructing calsequestrin. Complex buffering in the calcium store of skeletal
muscle. J Physiol (London) 2009;587:3103–3111.
Mitochondria
Dedkova EN, Blatter LA. Mitochondrial Ca2+ and the heart. Cell Calcium 2008;44:77–91.
Griffiths EJ. Mitochondrial calcium transport in the heart: physiological and pathological roles. J Mol
Cell Cardiol 2009;46:P789–P803.
Liu T, O'Rourke B. Regulation of mitochondrial Ca2+ and its effects on energetics and redox balance
in normal and failing heart. J Bioenerg Biomembr 2009;41:127–132.
Kurland CG, Collins LJ, Penny D. Genomics and the irreducible nature of eukaryote cells. Science
2006;312:1011–1014.
Murphy E, Eisner DA. Regulation of intracellular and mitochondrial sodium in health and disease.
Circ Res 2009;104:292–303.
Tekle YI, Parfrey LW, Katz LA. Molecular data are transforming hypotheses on the origin and
diversification of eukaryotes. Bioscience 2009;59:471–481.
References
Beil FU, von Chak D, Hasselbach W, et al. Competition between oxalate and phosphate during active
calcium accumulation by sarcoplasmic vesicles. Z Naturforsch 1977;32:281–287.
P.174
Bers DM. Excitation-contraction coupling and cardiac contractile force. Dordrecht: Kluwer, 1991.
Bretscher A, Edwards K, Fehon RG. ERM proteins and merlin: integrators at the cell cortex. Nat Rev
Mol Cell Biol 2002;3:586–599.
Dong, H, Dunn J, Lytton J. Stoichiometry of the cardiac Na+/Ca2+ exchanger NCX1.1 measured in
transfected HEK cells. Biophys J 2002;82:1943–1952.
Ebashi S, Lipmann F. Adenosine triphosphate-linked concentration of calcium ions in a particulate

fraction of rabbit muscle. J Cell Biol 1962;14:389–400.
Eisenberg B, Eisenberg RS. Transverse tubular system in glycerol-treated skeletal muscle. Science
1968;160: 1243–1244.
Fabiato A. Calcium-induced release of calcium from the cardiac sarcoplasmic reticulum. Am J

Physiol 1983;245:C1–C14.
Gergely J. The relaxing factor of muscle. Ann NY Acad Sci 1959;81:490–504.
Harigaya S, Schwartz A. Rate of calcium binding and uptake in normal and failing human cardiac
muscle. Circ Res 1969;25:781–794.
Hasselbach W, Makinose M. Die calciumpumpe der “ershlaffungsgrana” des muskels und ihre
abhangigkeit von der ATP-spaltung. Biochem Z 1961;333:518–528.
Hill AV. The abrupt transition from rest to activity in muscle. Proc Roy Soc B 1949;136:399–420.
Huxley AF, Taylor RE. Local activation of skeletal muscle fibres. J Physiol (Lond) 1958;44:426–441.
Katz AM, Repke DI. Quantitative aspects of dog cardiac microsomal calcium binding and calcium
uptake. Circulation Res 1967;21:153–162.
Katz AM, Reuter H. Cellular calcium and cardiac cell death. Am J Cardiol 1979;44:188–190.
Láttgau HC, Niedergerke R. The antagonism between Ca and Na ions on the frog's heart. J Physiol
(Lond) 1958;143:486–505.
Ogawa H, Shinoda T, Cornelius F, et al. Crystal structure of the sodium-potassium pump (Na+, K+-
ATPase) with bound potassium and ouabain. Proc Nat Acad Sci (US) 2009;106:13742–13747.
Repke K. áber den biochemischen Wirkungsmodus von Digitalis. Klin Wochschr 1964;41:156–165.
Reuter H, Seitz N. The dependence of calcium efflux from cardiac muscle on temperature and
external ion composition. J Physiol (London) 1968;195:451–470.
Schátzmann HJ. Herzglykoside als hemmstoffe fár den aktiven Kalium-und Natriumtransport durch
die Erythrocytenmembran. Helv Physiol Acta 1953;11:346–354.
Shen AC, Jennings, RB. Myocardial calcium and magnesium in acute ischemic injury. Am J Pathol
1972; 67:417–440.
Shigekawa M, Finegan J-AM, Katz AM. Calcium transport ATPase of canine cardiac sarcoplasmic
reticulum. A comparison with that of rabbit fast skeletal muscle sarcoplasmic reticulum. J Biol Chem
1976; 251:6894–6900.
Tian R, Ingwall JS. Energetic basis for reduced contractile reserve in isolated rat hearts. Am J
Physiol 1996; 270 (Heart Circ Physiol 39):H1207–H1216.
Tsien RW. Calcium channels in excitable cell membranes. Ann Rev Physiol 1983;45:341–358.
Tsien RW, Tsien RY. Calcium channels, stores, and oscillations. Ann Rev Cell Biol 1990;6:715–760.
Weber A, Winicur S. The role of calcium in the superprecipitation of actomyosin. J Biol Chem
1961;236:3198–3202.
Wilbrandt W, Koller H. Die Calcium-Wirkung am Froschherzen als Funktion des Ionengleichgewichts

zwischen Zellmembran und Umgebung. Helv Physiol Acta 1948;6:208–221.
> Table of Contents > Part Two - Signal Transduction and Regulation > Chapter 8 - Signal Transduction: Functional Signaling
Chapter 8
Signal Transduction: Functional Signaling
The vie constante, where life manifests itself independently of the external environment [is] characterized by freedom and
independence … Here life is never suspended, but flows steadily on apparently indifferent to alterations in its cosmic environment or
changes in its material surroundings. Organs, structural mechanisms and tissues all function uniformly … [Because] the milieu
intárieur surrounding the organs, the tissues, and their elements never varies; atmospheric changes cannot penetrate beyond … we
have an organism which has enclosed itself in a kind of a hot-house. The perpetual changes of external conditions cannot reach it; it
is not subject to them, but is free and independent.
—Claude Bernard, 1878
Bernard (1878) noted that constancy of the milieu intárieur is essential for independent life. Maintenance of this internal environment, which Cannon
(1932) called homeostasis, is made possible by mechanisms that compensate for changes in the labile, often hazardous, external environment. In disease,
the same mechanisms help alleviate problems created by malfunction of one or more of the body's components. The remarkable complexity of these
regulatory systems allow the body to meet a wide range of challenges, while numerous redundancies help avoid the consequences of Murphy's law, which
states: “If anything can go wrong, it will; and at the worst possible time.” Furthermore, damage to one or more of the mechanisms that maintain
homeostasis often produces few ill effects, and sometimes none, because the body addresses Murphy's law with another law: “If something is worth doing,
many different mechanisms will see to it that this is done correctly.” The latter is well known to modern engineers, who design elaborate systems with
redundancies that prevent failure of any component from causing the structure to crash.
Utilization of physiological regulatory mechanisms requires signals to inform a structure, such as the heart or a blood vessel, that its function must be
modified, along with the means to initiate appropriate responses. To ensure that these responses are turned on when needed, operate at an appropriate
intensity, and end when no longer useful, cardiovascular signaling systems collect and integrate information from many sources. Appropriate safeguards
are provided by interlocking control mechanisms that recognize challenges, amplify signals, and adjust and integrate the responses that compensate for
changing hemodynamics.
Two different types of response adjust cardiovascular function to the changing needs of the body (Fig. 8-1). The first are short-term functional responses
that adjust the circulation to stresses—such as exercise and hemorrhage—that develop rapidly, over a few seconds or minutes, and generally last no longer
than a few minutes or hours. The second are prolonged proliferative (transcriptional) responses that are evoked by chronic abnormalities, such as
hypertension, a
P.178
leaky or stenotic heart valve, or after a myocardial infarction has irreversibly damaged the left ventricle. The latter, which evolve over periods of weeks,
months, and often years, are mediated by changes in gene expression and protein synthesis that alter the architecture of the heart, and the size, shape,
and composition of its myocardial cells.
Fig. 8-1: Functional and proliferative responses. Functional responses, which alter the properties of preexisting structures by posttranslational
mechanisms, aid survival by evoking responses like fight and flight. Proliferative responses, which take longer to evolve, allow organisms to grow
out of trouble. (Images from Huber, 1981.)
Functional Signal Transduction

Functional responses are often referred to as neurohumoral responses because most are evoked by the autonomic nervous system and chemical mediators
that are secreted into the blood or extracellular fluid. The familiar cardiovascular response to exercise, which helps maintain arterial blood pressure and
increase blood flow to the exercising muscles, is initiated largely by increased sympathetic activity and epinephrine released by the adrenal medulla.
Along with decreased parasympathetic activity, these neurohumoral signals act on the heart and blood vessels to increase ventricular ejection and filling,
accelerate heart rate, and constrict peripheral arterioles and veins. Norepinephrine, the most important sympathetic mediator, activates multistep
signaling cascades that can be viewed as an old-fashioned bucket brigade (Fig. 8-2). However, unlike a bucket brigade, where a single substance (water) is
passed from fireman to fireman, biological signaling cascades utilize many different signaling molecules and chemical reactions.
The first step in the response of the heart to sympathetic stimulation is norepinephrine release from sympathetic nerve endings into the extracellular
space (Fig. 8-2). The second step, norepinephrine binding to β-receptors on the surface of cardiac myocytes, also takes place in the extracellular fluid,
after which all of the remaining steps occur within these cells. These include G protein activation (step 3), which increases the catalytic activity of
adenylyl cyclase (step 4), which generates cyclic AMP (step 5), which activates a cyclic adenosine monophosphate (cAMP)-dependent protein kinase (PKA,
step 6), which catalyzes the phosphorylation of L-type calcium channels (step 7), which increases the entry of “trigger” calcium into the cytosol (step 8),
which increases the opening of intracellular calcium release channels (step 9), which releases more calcium for binding to troponin C (step 10), which
increases the number of exposed actin molecules in the thin filaments (step 11), which increases the number of actin–myosin interactions
P.179
P.180
(step 12), which increases contractility (step 13), which increases the volume of blood ejected by the heart (step 14). Cascades similar to that in Figure 8-
2, although different in detail, mediate virtually every signal that modifies cardiovascular function.
Fig. 8-2: Depiction of the signal cascade by which sympathetic stimulation increases ejection. This analogy shows 14 steps as a series of buckets
where, in each step, an “upstream” signal causes the bucket to pour its contents into the next bucket, thereby transmitting the signal down the
cascade. The response begins when sympathetic stimulation releases norepinephrine from nerve endings at the surface of cardiac myocytes (step 1),
after which this neurotransmitter binds to and activates β-adrenergic receptors (step 2). The remaining steps, which all take place within
myocardial cells, include activation of a G protein (step 3), increased adenylyl cyclase activity (step 4), accelerated production of cAMP (step 5),
activation of a cAMP-dependent protein kinase (step 6), phosphorylation of plasma membrane L-type calcium channels (step 7), more calcium entry
into the cell (step 8), increased opening of calcium release channels (step 9), increased calcium binding to troponin C (step 10), greater availability
of actin in the thin filaments for interaction with myosin (step 11), participation of more cross-bridges in contraction (step 12), increased
contractility (step 13), and a greater extent of ejection (step 14).
Why are there so many Steps in a Signal Transduction Cascade?

One might ask why there are so many steps between a challenge, like exercise, and a physiological response like that depicted in Figure 8-2. The answer
lies in the ability of biological signal transduction cascades, whose complexity might, at first glance, seem to be almost perverse, to enhance regulatory
control. These cascades can amplify, inhibit, and fine-tune responses, integrate signal transduction cascades with one another, and prevent responses from
going out of control (“runaway signaling”) by allowing signals to turn themselves off automatically. In fact, the depiction of the signal cascade in Figure 8-2
is oversimplified because most biological signaling pathways are not linear, where each step is coupled to a single downstream reaction, but instead form
branches, loop forward and backward, interconnect with other pathways, and generate multiple signals at many steps. These intricacies organize
responses that provide options for both amplification and negative feedback, and for the integration of each step in the response with other steps in the
cascade and with additional signaling systems. This fine-tuning would not be possible if cell signaling generated only an “all-or-none” response.
An example of a mechanism that regulates signal transduction is the allosteric effect of ATP described in Chapter 7, which accelerates ion fluxes through
channels, pumps, and exchangers. This effect is important in energy-starved hearts, where a fall in ATP concentration inhibits several steps in the signaling
cascade shown in Figure 8-2; these include the opening of L-type calcium channels in the plasma membrane (step 8) and calcium release channels in the
sarcoplasmic reticulum (step 9), and the interactions between actin and myosin (step 12). These and other allosteric effects of decreased ATP
concentration help match the rates of ATP production and ATP consumption by inhibiting contraction, the most expensive energy-consuming reaction in the
heart. Acidosis, which occurs when glycolysis is accelerated in energy-starved hearts (Chapter 2), also reduces energy consumption by inhibiting many
steps in this cascade. A third example, which illustrates how a reaction near the “top” of a signaling pathway can respond to changes further down the
cascade, occurs when increased calcium release into the cytosol (step 9) inhibits adenylyl cyclase (step 4); this negative feedback, by slowing the cascade,
reduces contractility. Additional mechanisms that avoid calcium overload include the ability of increased cytosolic calcium to accelerate calcium transport
out of the cytosol by the plasma membrane and sarcoplasmic reticulum calcium pumps, and by the sodium/calcium exchanger (Chapter 7).
Biological signal transduction cascades often allow a single stimulus to evoke an integrated, multifaceted response. For example, the elevated level of
cAMP (step 5 in Fig. 8-2) increases the rate and extent of relaxation when phosphorylation of phospholamban accelerates calcium uptake into the
sarcoplasmic reticulum (Chapter 7). Another example of signal diversification occurs at step 3 in Figure 8-2, where norepinephrine binding to a single β1-
receptor activates both Gαs and Gβγ, each of which can modify a different downstream target (see below).
Control of cell function is enhanced further by a large number of isoforms of many signaling molecules. For example, binding of norepinephrine to β1- and
β2-receptors in the heart evokes different responses, while in vascular smooth muscle, activation of α1-receptors by norepinephrine causes
vasoconstriction, and β-receptor activation has a relaxing effect. These intricacies also
P.181
characterize intracellular signal transduction, as is evident by the ability of protein kinases to activate several pathways by phosphorylating a number of
different signaling proteins.
The Hemodynamic Defense Reaction

The hemodynamic defense reaction, which compensates for underfilling of the arterial system, is an integrated functional response that involves the
heart, blood vessels, and kidneys (Table 8-1). As detailed in a brilliant essay by Poter Harris (1983), exercise, hemorrhage, and heart failure all initiate this
response by stimulating cardiac performance, constricting blood vessels, and inhibiting salt and water excretion by the kidneys (Table 8-2). In exercise, a
short-term challenge in which arterial blood pressure falls because of vasodilatation in the active muscles, the hemodynamic defense reaction restores
blood pressure by stimulating the heart and constricting the arterioles supplying nonexercising tissues such as the gut and kidneys. Cardiac stimulation and
selective vasoconstriction appear rapidly, over a few seconds or minutes, and last until the exercise ends, usually for a few minutes or hours. Urine output
also decreases during exercise, but because the challenge is generally brief—marathon runners are a notable exception—fluid retention does not become
apparent. More prolonged responses, supplemented by thirst and fluid retention by the kidneys, are initiated when cardiac output is severely decreased in
a syndrome called shock. The latter can last for hours, at most a few days, after which the patient either dies or recovers. In heart failure, which is
almost always progressive, the hemodynamic defense reaction persists and fluid retention commonly becomes a major clinical problem (see Chapter 18).
Exercise
The neurohumoral response evoked by exercise, the briefest of the challenges listed in Table 8-1, is initiated by several mechanisms. Most important is the
baroreceptor reflex, which increases sympathetic activity and reduces parasympathetic tone in response to the fall in arterial blood pressure caused by
vasodilatation in the exercising muscles. Sympathetic activity is also increased when CO2 and lactic acid released by the exercising muscles stimulate the
chemoreceptors. These
P.182
responses are supplemented by nerve impulses that arise in the exercising muscles and the cerebral cortex; the latter explains the onset of sympathetic
stimulation before the start of exercise, for example, when a sprinter hears the starter's gun.
Table 8-1 Three Conditions That Evoke the Hemodynamic Defense Reaction
Condition Duration Challenge Response
Exercise Minutes/hours Increased blood flow to exercising muscles Cardiac stimulation, selective
vasoconstriction
Shock Hours Decreased arterial filling (e.g., hemorrhage, fluid loss, Cardiac stimulation, vasoconstriction,
impaired cardiac pumping) fluid retention
Heart Usually for life Chronically impaired pumping by a damaged or overloaded Cardiac stimulation, vasoconstriction,
failure (progressive) heart fluid retention
Table 8-2 The Hemodynamic Defense Reaction

The short-term neurohumoral response evoked by exercise is mediated by an inotropic effect that increases the amount of blood ejected during systole, a
lusitropic effect that augments filling during diastole, and a chronotropic effect that accelerates heart rate (Tables 8-1 and 8-2). Together, these responses
increase the flow of blood out of the heart (cardiac output). Sympathetic stimulation also causes contraction of vascular smooth muscle. Constriction of
small arterioles helps maintain blood pressure by increasing the resistance in vascular beds not dilated by exercise, while venoconstriction increases the
return of blood to the heart. Fluid retention, as noted above, becomes important only during prolonged exercise.
Shock
Shock, a clinical syndrome whose most obvious abnormalities are low blood pressure and inadequate tissue perfusion, also activates the hemodynamic
defense reaction. There are many causes of this syndrome, which usually lasts no more than several hours (Table 8-3). Hypovolemic shock occurs when
hemorrhage, diarrhea, or leaky capillaries severely decrease circulating blood volume. In distributive shock, which is caused by excessive vasodilatation,
there is enough blood in the vasculature but it is in the wrong place, that is, the veins rather than the arteries. Distributive shock can be caused by
vasodilator actions of endotoxins (e.g., toxic shock, gram-negative septicemia), reflex vasodilatation (vasovagal syncope, which is the old fashioned
“swoon”), and the von Bezolt–Jarisch reflex which can be activated in patients with inferior or posterior myocardial infarction (Chapter 17). A third type of
shock, cardiogenic shock, is seen after myocardial infarction, but in this case the cause is reduced ejection by a severely damaged left ventricle. Other
causes of cardiogenic shock include valve rupture, pulmonary embolus, pericardial tamponade, acute myocarditis, and arrhythmias. If the underlying cause
is reversible and appropriate therapy started promptly, patients usually
P.183
recover. However, if treatment is delayed this syndrome becomes irreversible because, even if blood pressure can be increased, the tissue damage
initiated by low blood flow causes the patient to die.
Table 8-3 Causes of Shock
Hypovolemic shock: Too little blood in the vascular system
Blood loss: Hemorrhage
Fluid loss: Endothelial damage (burns, vasculitis, endotoxin), excessive fluid loss (diarrhea)
Distributive shock: Enough blood in the vascular system but in wrong place; veins not arteries
Sepsis: Vasodilator actions of endotoxins (toxic shock, gram-negative septicemia)
Reflex: Vasovagal syncope (swoon), von Bezolt–Jarisch reflex (inferior myocardial infarction)
Cardiogenic shock: Acute failure of the left ventricle
Acute myocardial infarction: Severe left ventricular damage
Other: Arrhythmias, valve rupture, pulmonary embolus, pericardial tamponade, acute myocarditis
The neurohumoral response in shock differs from that evoked by exercise because the challenge is more intense and, more importantly, lasts longer. The
hemodynamic defense reaction, therefore, has time to cause the kidneys to retain salt and water, which facilitates restoration of blood volume. However,
when vasoconstriction is prolonged and severe, organ damage by low flow is increased.
Patients can recover from shock if the underlying cause is reversible and appropriate treatment is started promptly. Otherwise, severe tissue damage
causes this syndrome to become irreversible. If this occurs these patients are doomed to die even if blood pressure can be briefly restored.
Heart Failure
In heart failure, where underfilling of the arterial system usually lasts for a lifetime, most of the functional responses listed in Tables 8-1 and 8-2 become
maladaptive. Tachycardia and increased contractility, which initially help maintain cardiac output, also increase cardiac energy utilization which, because
most failing hearts are energy-starved, contributes to myocardial cell death. Arteriolar vasoconstriction helps maintain blood pressure when the heart
fails, but the increased afterload reduces cardiac output and increases myocardial energy demand. Fluid retention by the kidneys, which is beneficial in
patients with shock, represents a serious problem in heart failure because the elevated blood volume increases the already elevated pulmonary and
systemic venous pressures to levels that cause fluid to be transuded into the lungs, peripheral tissues, and body cavities. Furthermore, the increased
preload caused by venoconstriction and fluid retention can do little to increase cardiac output because Starling curves are often flattened in heart failure.
For all of these reasons, the neurohumoral response usually becomes deleterious in most patients with heart failure (see Chapter 18).
Cardiac stimulation, vasoconstriction, and salt and water retention, which provide short-term support for the circulation during the brief challenges posed
by exercise and shock,
P.184
become maladaptive in chronic heart failure. This paradox is vividly summarized by Harris (1983):
Success and survival in the animal kingdom have overwhelmingly depended on physical mobility and strength. To ensure this the
body makes use of the neuro-endocrine defense reaction which is also life-saving in injury … When the output of the [failing] heart
decreases, the body reacts in the way nature has programmed it. It cannot distinguish. But now the neuro-endocrine response
persists. Over weeks or months or years the retention of saline threatens the cardiac patient with drowning in his own juice. And
every hour of every day he is running for his life.
Regulatory and Counterregulatory Responses

The hemodynamic defense reaction activates two opposing types of functional responses (Table 8-4). Regulatory responses, which include cardiac
stimulation, vasoconstriction, and fluid retention help maintain blood pressure and cardiac output (see above). At the same time, however,
counterregulatory responses that oppose the dominant regulatory responses are also evoked; these include a negative inotropic effect on the heart,
vasodilatation, and diuresis. The ability of a single stimulus to activate opposing responses illustrates one way that biological signaling systems minimize
the risk of “runaway signaling” (see below).
Table 8-4 Regulatory and Counterregulatory Neurohumoral Responses
A. Regulatory Responses
Functional responses
Increased myocardial contractility, accelerated relaxation, faster heart rate
Vasoconstriction
Fluid retention by the kidneys

Proliferative responses
Stimulate cell growth and proliferation
Pro-apoptotic
B. Counterregulatory Responses
Decreased myocardial contractility, slowed relaxation, slower heart rate
Vasodilatation
Diuresis
Inhibit cell growth and proliferation
Anti-apoptotic
P.185
Table 8-5 Regulatory and Counterregulatory Neurohumoral Responses
A. Some Signaling Molecules Whose Major Role is Regulatory
Mediators
Norepinephrine—peripheral effects
Angiotensin II
Arginine vasopressin
Endothelin
Responses
Increased cardiac contractility, relaxation, heart rate
Vasoconstriction
Fluid retention by the kidneys
Stimulation of cell growth and proliferation

B. Some Signaling Molecules Whose Major Role is Counterregulatory
Mediators
Norepinephrine—central effects
Dopamine
Natriuretic peptides
Nitric oxide (NO)
Bradykinin
Agmatine
Responses
Decreased cardiac contractility, relaxation, heart rate
Vasodilatation
Reduced fluid retention by the kidneys, diuresis
Inhibition of cell growth and proliferation
The functional responses evoked by some neurohumoral mediators are predominantly regulatory, whereas others are mainly counterregulatory. As a rule,
regulatory mediators also have proliferative effects that promote cardiac hypertrophy, while counterregulatory mediators generally inhibit cell growth and
proliferation (Table 8-5). However, there are exceptions to this generalization, and many neurohumoral mediators evoke both regulatory and
counterregulatory responses.
Extracellular Signaling Molecules

The ability of chemical mediators to regulate cardiovascular function was discovered by Oliver and Scháfer (1895), who found that injection of adrenal
extracts increases heart rate in anesthetized cats. A decade later, Elliott (1905) suggested that the response to sympathetic nerve stimulation, which
resembles that evoked by adrenal extracts, might also be mediated by a chemical messenger. However, it was not until 1921 that Loewi (1921) carried out
a simple but inspired experiment
P.186
which proved conclusively that a chemical can mediate a neurohumoral response, in this case, to vagal stimulation. Early efforts to isolate a chemical
mediator had proven fruitless because the neurotransmitter is inactivated very rapidly, but Loewi overcame this problem by placing two frog hearts a short
distance apart in a slowly moving stream of Ringer's solution. When he stimulated the vagus nerve supplying the upstream heart, not only did this heart
slow, but the rate of beating also decreased in the unstimulated downstream heart; in contrast, stimulation of the vagus nerve supplying the downstream
heart had no effect on the upstream heart. This elegant experiment proved that vagal stimulation releases a chemical that slows the heart; a few years
later Loewi, who initially called this chemical vagusstoff, identified the mediator as acetylcholine.
In the 1930s, Cannon and Rosenblueth (1937) postulated that sympathetic stimulation releases a vasoconstrictor that they called sympathin E (for
excitatory) which is related to epinephrine (adrenaline), a vasodilator that had previously been isolated from the adrenal medulla (which they called
sympathin I, for inhibitory). Shortly after World War II, von Euler (1946) showed that the sympathetic neurotransmitter is norepinephrine, which differs
from epinephrine only in the absence of a methyl group.
A variety of signaling molecules are known to regulate cardiovascular function (Table 8-6). These include tyrosine metabolites called catecholamines (e.g.,
epinephrine, norepinephrine, dopamine), quaternary amines (e.g., acetylcholine), other small organic molecules (e.g., thyroxin,
P.187
purines), peptides (e.g., angiotensin II, arginine vasopressin, natriuretic peptides, endothelin, cytokines, growth factors), steroid hormones (e.g.,
aldosterone), fatty acid derivatives (e.g., prostaglandins), and even a free radical gas (nitric oxide). Because these chemical mediators act on the outsides
of cells, they are sometimes referred to as extracellular messengers.
Table 8-6 Some Receptors and Extracellular Messengers That Modify Cardiovascular Function
G Protein-Coupled Receptors Direct Binding to an Intracellular Target
Catecholamine Nitric oxide (NO)
Norepinephrine and epinephrine Agmatine
Dopamine Enzyme-linked receptors
Peptide Tyrosine kinase receptors
Angiotensin II Fibroblast growth factor (FGF)a
Bradykinin Platelet-derived growth factor (PDGF)a
Endothelin Insulin-like growth factor (IGF)a
Arginine vasopressin (ADH) Vascular endothelial growth factor (VEGF)a
Calcitonin gene-related peptides: Serine/threonine kinase receptors
adrenomedullin, intermedin Transforming growth factor-β (TGF-β)a
Apelin Receptor guanylyl cyclases
Neuropeptide Y Natriuretic peptides
Ghrelin Cytokine receptors
Other Tumor necrosis factor α (TNF-α)a
Acetylcholine (muscarinic) Interleukinsa
Adenosine (purinergic) Growth hormonea
Prostaglandins Leptin
Ion channel-linked receptors Nuclear receptors
Acetylcholine (nicotinic) Aldosterone
Thyroxina
aPrimary discussion in Chapter 9.

Interactions between Extracellular Messengers (Ligands) and their Receptors
Extracellular messengers are frequently called ligands because they bind with high affinity and specificity to receptors that recognize their presence as a
signal to modify cell function. Most physiological ligands, as well as a majority of clinically useful drugs, are amphipathic molecules whose hydrophilic
moieties prevent them from crossing the lipid barrier in biological membranes (Chapter 1). For this reason, their cellular actions depend on interactions
with plasma membrane receptors. Exceptions include steroid and thyroid hormones, which are hydrophobic molecules that enter the cytosol where they
interact with intracellular receptors.
The concept of specific receptors originated when Ahlquist (1948) found that low concentrations of epinephrine dilate blood vessels whereas higher
concentrations cause vasoconstriction. This led him to postulate that epinephrine can bind to two types of receptors: α-receptors, which have a low
affinity for this ligand and are responsible for the constrictor action seen at the higher epinephrine concentrations, and β-receptors which, because they
have a higher affinity for epinephrine, mediate the vasodilator response seen at low concentrations of this extracellular messenger. The subsequent
elucidation of the structures of these and other receptor molecules, along with the demonstration that different signal transduction pathways are
activated when a given ligand interacts with different receptors, have proven Ahlquist's hypothesis to be correct.
Signal Transmission
Endocrine (hormonal) signaling (Table 8-7) was discovered by Bayliss and Starling (1902), who observed that instilling acid into the duodenum stimulates
pancreatic secretion, even after both tissues are denervated. After finding that intravenous injection of a jejunal mucosa extract also
P.188
stimulates pancreatic secretion, they proposed that a substance, which they called secretin, travels through the bloodstream from the jejunum, where it
is produced, to the pancreas, where it evokes the response. Less than 20 years after this discovery, Loewi described neurotransmitter signaling, where the
extracellular messenger is released by the nervous system (see above). More recently, signaling molecules were found to reach their receptors by shorter
routes; in paracrine signaling an extracellular messenger released by one cell diffuses to a receptor on a nearby cell, while in autocrine signaling a cell
modifies its own function by releasing a ligand that binds to a receptor on its surface. Cytoskeletal signaling allows mechanical interactions within and
between neighboring cells and the surrounding extracellular matrix to alter cell structure and function. This type of signaling is initiated when cell
deformation modifies signaling proteins—many of which are homologous to the ligands, receptors, enzymes, and other proteins that participate in cell
signaling—that interact with the cytoskeleton (see Chapter 5).
Table 8-7 Routes by Which Extracellular Messengers Reach Cells
Endocrine (hormonal) signaling: An extracellular messenger generated by a distant cell is delivered via the bloodstream to the cells whose
structure or function is altered.
Neurotransmitter signaling: An extracellular messenger generated by a nerve ending is released at the surface of the cells whose structure or
function is altered.
Paracrine signaling: An extracellular messenger generated by a cell diffuses through the extracellular fluid to nearby cell whose structure or
function is altered.
Autocrine signaling: An extracellular messenger is released into the extracellular fluid by the cell whose structure or function is altered.
Cytoskeletal (mechanical) signaling: A mechanical stress recognized by a cytoskeletal protein generates a signal that modifies cell structure
or function.
Most extracellular messengers contain hydrophobic moieties, so that these molecules have high partition coefficients, a variable that quantifies the
distribution of a molecule between the lipid bilayer and the surrounding aqueous medium. Molecules with high partition coefficients tend to accumulate in
membranes; for example, when a cell in an aqueous medium encounters 10,001 molecules with a partition coefficient of 10,000, 10,000 of the molecules
will enter the bilayer, leaving only 1 in the aqueous medium. Because of their hydrophobicity, extracellular messengers and drugs often have partition
coefficients in the thousands. The anti-arrhythmic drug amiodarone has a partition coefficient >1,000,000 (Herbette et al., 1988) that explains its
remarkably long biological half-life, which is many months. Their high hydrophobicity allows these molecules to gain access to plasma membrane receptors
by first dissolving in the lipid bilayer and then diffusing to their receptors (Herbette et al., 1991). Utilization of this lipid pathway, which allows ligands to
gain rapid access to their membrane receptors, also explains why the ligand- and drug-binding sites of many intrinsic membrane proteins include
hydrophobic regions of the membrane-spanning α-helices within the bilayer.
LigandBinding to Receptors: Receptor Number and Binding Affinity

The interactions between ligands and their receptors can be characterized in terms of the number of receptors that can bind to the ligand and the affinity
with which the ligand attaches to the receptor. Receptor number is often quantified as Bmax, the maximal amount of ligand that binds specifically to the
receptor. A commonly used index of the affinity of a receptor for its ligand is the dissociation constant (kd), which is the ligand concentration at which 50%
of receptors are bound to the ligand; this means that the greater the affinity of the receptor for the ligand, the lower will be the kd. The constant kd also
describes the ratio between the rate of ligand dissociation from its receptor (the “off-rate”) and the rate of binding of the ligand (the “on-rate”). Affinity
can also be expressed as an association (binding) constant (kb), which is the reciprocal of kd, the dissociation constant. Ligands have a wide range of
receptor-binding affinities; some peptides bind to their receptors at concentrations below 1012 M, while most neurohumoral transmitters and drugs occupy
their receptors at concentrations between 108 and 106 M. A few compounds, like ethanol, interact with membrane proteins at much higher concentrations
(Table 8-8).
Binding affinity is an important determinant of the specificity of the biological response to a ligand. A ligand that binds to its receptor with high affinity
usually evokes a specific response
P.189
that is accompanied by few side effects; this is because low concentrations of the ligand can be recognized by the receptor, which minimizes interactions
of the ligand with other cell components. Ligands that bind to their receptors with low affinity, and so evoke responses only at high concentrations, often
have additional “nonspecific” actions that are usually undesirable. The toxic effects of most clinically useful drugs are seen at concentrations higher than
those that produce the desired therapeutic effects; exceptions include allergic or sensitivity reactions, which are generally much less dependent on the
concentration of the ligand. Specificity can be described as the ratio of the ligand concentrations that cause toxic and therapeutic effects; a higher
“toxic/therapeutic ratio” means that a drug is less likely to cause unwanted side effects when administered at doses that yield desirable therapeutic
effects.
Table 8-8 Ligand-Binding Affinities of Some Cardiac Plasma Membrane Receptors
a
Ligand Receptor Approximate kd
Tetrodotoxinb Sodium channel 1012 M
Nitrendipinec Calcium channel 1010 M
Epinephrine β-adrenergic receptor 108 M
Ouabaind Sodium pump 106 M
Ethanol Non-specific 103 M
ak , the dissociation constant, is the ligand concentration at which half of the receptors are occupied, so that a lower k means that the
d d
ligand binds more tightly to its receptor.
bA peptide toxin from puffer fish.
cA dihydropyridine calcium channel blocker.

dA cardiac glycoside.
Receptor Blockade
Characterization of specific receptors made it possible to identify drugs that could inhibit their ability to interact with their ligands. Such inhibitors, often
called blockers or antagonists, usually have structures similar to those of the physiological extracellular messengers. The inhibitory effects of many of
these drugs exhibit competitive kinetics and can be reversed by high concentrations of the physiological ligand. A few drugs inactivate a signal
transduction system completely and permanently; aspirin, for example, acetylates and irreversibly inactivates cyclooxygenase, an enzyme that releases a
thrombogenic prostaglandin (see below).
The difference between most receptor blockers (antagonists) and the physiological ligands (agonists) is not where these molecules bind, but what happens
after binding has occurred. Unlike an agonist, which activates the subsequent steps in a signal transduction cascade (e.g., after step 1 in Fig. 8-2),
antagonists occupy the receptor but do not generate an intracellular signal. The clinical value of an antagonist (e.g., a β-adrenergic blocker) therefore
reflects the fact that the receptor-bound antagonist inhibits the ability of the receptor to bind to, and thus become activated by, the physiological agonist
(e.g., norepinephrine).
P.190
Not all molecules that interact with receptors can be classified simply as agonists and antagonists. Some drugs, called partial agonists, bind to a receptor
where they cause a weak activation of its signal transduction cascade, while at the same time blocking the ability of the receptor to interact with the
more potent physiological agonists. In the case of the β-adrenergic blockers, the weak stimulatory effect of a partial agonist is referred to as intrinsic
sympathomimetic activity (ISA). By blocking the more potent effects of norepinephrine, partial β-adrenergic agonists inhibit the response to surges of
sympathetic activity while providing a low level of adrenergic stimulation in the basal state. Unfortunately, many partial agonists have turned out to be
less advantageous clinically than this description might suggest.
Types of Receptors
Cardiovascular function is regulated when a variety of ligands bind to their receptors (Table 8-6). Most functional signals are mediated by G protein-
coupled receptors (GPCRs), which are named for the heterotrimeric guanosine triphosphate (GTP)-binding proteins that mediate
P.191
their cellular actions (see below). Enzyme-linked receptors, most of which contain a kinase or other enzyme that is activated when the receptor binds to
its ligand, were initially thought to mediate only proliferative responses (see Chapter 9); however, these receptors are now known to participate in
functional signaling as well. Cytokine receptors resemble enzyme-linked receptors except that instead of having intrinsic enzyme activity, ligand-bound
cytokine receptors modify the catalytic activity of other enzymes. Ion channel-linked receptors contain channels that are opened when the receptor binds
its ligand. Nuclear receptors bind to hormones, like aldosterone and thyroxin, whose hydrophobic structure allows them to cross the plasma and nuclear
membranes; the resulting ligand-receptor complexes regulate gene expression by interactions with DNA in the nucleus. Cellular responses to nitric oxide
are not mediated by a receptor; instead, this signaling molecule binds directly to guanylyl cyclase, its target within cells.
Table 8-9 Some Important G Protein-Coupled Receptors in the Cardiovascular System
Ligand Receptor Gα Isoform “Target” Second Messenger/Effector
Norepinephrine α-Adrenergic Gαq Phospholipase C Diacylglycerol, InsP3 (↑)
Norepinephrine β-Adrenergic Gαs Adenylyl cyclase (↑) cAMP (↑)
Acetylcholine Muscarinic Gαo K channel Outward K current (↑)
Acetylcholine Muscarinic Gαi Adenylyl cyclase (↓) cAMP (↓)
Adenosine Purinergic (P1) Gαo K channel Outward K current (↑)
Adenosine Purinergic (P1) Gαi Adenylyl cyclase (↓) cAMP (↓)
Angiotensin II Angiotensin (AT1) Gαq Phospholipase C Diacylglycerol, InsP3 (↑)
Bradykinin Bradykinin (B2) Gαq eNOS (↑) Nitric oxide (↑)
Endothelin Endothelin (ETA) Gαq Phospholipase C Diacylglycerol, InsP3 (↑)
Vasopressin Vasopressin (V1a) Gαq Phospholipase C Diacylglycerol, InsP3 (↑)
Vasopressin Vasopressin (V2) Gαs Adenylyl cyclase (↑) cAMP (↑)
PGE2, PGI2 Prostanoid (EP, IP) Gαs Adenylyl cyclase (↑) cAMP (↑)
TxA2 Prostanoid (TP) Gαq Phospholipase C Diacylglycerol, InsP3 (↑)
Abbreviation: InsP3: inositol 1,4,5-trisphosphate; eNOS: endothelial nitric oxide synthase; PGE2: prostaglandin E2; PGI2: prostaglandin I2; TxA2:
thromboxane A; ↑: increased; ↓: decreased.
G Protein-Coupled Receptors
The most important regulators of cardiovascular function are G protein-coupled receptors (GPCRs) (Table 8-9), whose name describes their interactions
with guanyl nucleotide-binding proteins (G proteins). This family, which is among the largest in biology, includes more than 800 different proteins; in
humans, approximately one in 80 genes encodes members of this class of receptors (Clapham and Neer, 1997, Luttrell, 2006). GPCRs contain seven
membrane-spanning α-helices (Fig. 8-3) and so are sometimes called heptahelical or seven-membrane-spanning receptors. The ligand-binding sites include
hydrophobic regions of the membrane-spanning α-helices, while the sites that bind the G proteins include both the membrane-spanning α-helices and a
large intracellular C-terminal loop within the cytosol.
Fig. 8-3: A G protein-coupled membrane receptor showing the seven-membrane-spanning α-helices, portions of which contribute to the ligand-
binding site. Sites that bind the heterotrimeric G proteins are located on the intracellular peptide chain that links the fifth and sixth membrane
spanning helices. Phosphorylation of the C-terminal intracellular peptide chain participates in receptor desensitization.
P.192
Enzyme-Linked and Cytokine Receptors

Ligand-binding to enzyme-linked receptors activates an intracellular enzyme, usually a protein kinase, which is part of the receptor molecule. This family
includes tyrosine kinase receptors whose catalytic sites phosphorylate tyrosine, serine/threonine kinase receptors that phosphorylate serine or threonine,
receptor guanylyl cyclases that synthesize cGMP, and phosphatases that catalyze dephosphorylations. Ligand-bound cytokine receptors, which lack
enzymatic activity, form aggregates that activate latent tyrosine kinases in other membrane proteins.
Ion Channel-Linked Receptors

Ion channel-linked (ionotropic) receptors contain channels that open when the receptor binds to its ligand. These include nicotinic receptors in skeletal
neuromuscular junctions, which generate a depolarizing sodium current that activates the motor end-plate when the receptors are bound to acetylcholine.
(Nicotinic receptors differ from the G protein-coupled muscarinic receptors that mediate parasympathetic responses.) Ion channel-linked receptors
mediate the central responses to a variety of small molecules, including agmatine, glutamine, serotonin, and γ-aminobutyric acid.
Nuclear Receptors
Lipophilic ligands, like thyroxin and aldosterone, modify cardiovascular function after they cross the plasma membrane and bind to hormone receptors
within cells. When bound to their ligands, these receptors are transported to the nucleus where they regulate proliferative signaling.
The following discussion highlights the GPCRs, which are the most important mediators of functional signals in the heart and blood vessels; the enzyme-
linked, cytokine, and nuclear receptors, whose major roles are in proliferative signaling, are described in Chapter 9.
G Protein-Coupled Receptors and Heterotrimeric GTP-Binding Proteins (G Proteins)

GPCRs activate pathways in the heart and blood vessels whose cellular “targets” include enzymes like adenylyl cyclase and phospholipase C, which
synthesize intracellular second messengers, and voltage-gated potassium channels (Table 8-9). The responses initiated by these receptors are mediated by
heterotrimeric GTP-binding proteins (G proteins) that include one member of each of three protein families, Gα, Gβ, and Gγ. G proteins can generate two
signals; one carried by the GTP-binding protein Gα and the other by the Gβγ dimer. The rich signaling diversity made possible by this coupling system
reflects the fact that the large superfamily of GPCRs interacts with at least 20 Gα subunits, 5 Gβ subunits, and 12 Gγ subunits (Luttrell, 2006). Monomeric
G proteins, a family of GTP-binding proteins related to Gα that includes ras, rac, and rho, play a major role in proliferative signaling (see Chapter 9).
Gα subunits, which are associated with the intracellular surface of the plasma membrane, bind GTP and contain a GTPase site. Four types of Gα subunit
are important in cardiovascular regulation; Gαs, which participates in the activation of adenylyl cyclase by β-adrenergic agonists; Gαi, which mediates the
inhibition of cAMP production by muscarinic and purinergic agonists; Gαo, which opens potassium channels; and Gαq, which activates phospholipase C-
mediated signaling
P.193
cascades (Table 8-9). Gαs and Gαq are generally regulatory and Gαi and Gαo are generally counterregulatory. The heterodimers formed by Gβ and Gγ can be
both regulatory and counterregulatory. Some Gβγ dimers activate phospholipase C and phospholipase A2; another activates IK.Ach, an inward rectifying
potassium channel that mediates vagal slowing of heart rate (see Chapter 13).
Ligand-bound GPCRs generally interact with a single Gα to activate a single signaling pathway. However, some can interact with more than one Gα subtype
and/or Gβγ dimer so as to activate several signal transduction cascades, and ligand binding to a single receptor can activate up to 10 different Gα subunits,
including members of all four families (Laugwitz et al., 1996).
Interactions between G Protein-Coupled Receptors and Heterotrimeric G Proteins

The interactions between a GPCR, its ligand, and the heterotrimeric G protein can be described by the five-step sequence depicted schematically in Figure
8-4.
STEP 1, Binding of the ligand to its receptor and activation of Gα: Inactive receptors, whose ligand-binding sites are unoccupied (Fig. 8-4A), are bound to
the G protein trimer (Gabg), which increases the affinity of the unoccupied receptor (R) for its ligand (L). Binding of the ligand to the receptor initiates the
first step in the activation sequence (Fig. 8-4B):
STEP 2, Formation of Gα-GTP and dissociation of Gβγ: Ligand binding to the receptor G protein complex causes the guanosine diphosphate (GDP) bound to
Gα to be exchanged for GTP, which forms an activated Gα-GTP complex and releases activated Gβγ that can interact with its targets (Tβγ) (Fig. 8-4C).
STEP 3, Dissociation of Gα-GTP from the receptor: Gα-GTP dissociates from the R-L-Gα-GTP complex, which allows free Gα-GTP to activate its own targets
(Tα) (Fig. 8-4D). Dissociation of Gα-GTP reduces the ligand-binding affinity of the receptor, which releases the ligand.
STEP 4, Dephosphorylation of Gα-bound GTP: Dissociation of the Gα-GTP complex from the receptor stimulates the intrinsic GTPase activity of Gα which
dephosphorylates the Gα-bound GTP to form Gα-GDP (Fig. 8-4E). This returns Gα to its basal state (Gα-GDP), which can no longer participate in signal
transduction. Gα-GDP then rebinds and inactivates Gβγ released in step 2, which ends signal transduction by Gβγ.
GTP hydrolysis, which turns off both Gα- and Gβγ-mediated signals, is the major determinant of the duration of the response to the ligand.
STEP 5, Rebinding of Gα, Gβγ, and the receptor to form the receptor-bound Gabg complex: The system returns to the basal state when the Gabg complex
formed by Gα-GDP and Gβγ rebinds the free receptor, which increases its affinity for the ligand (Fig. 8-4A).
P.194
Fig. 8-4: Simplified scheme showing five steps in the interactions between a G protein-coupled receptor, its ligand, and the heterotrimeric G
proteins. Active states of these proteins are shaded. A: In the basal state, where the receptor (R) is not bound to its ligand (L), Gα is bound to GDP,
the Gβγ dimer, and the receptor in a R-L-Gα-GDP-Gβγ complex. The Gαβγ trimer in this complex increases the ligand-binding affinity of the receptor.
B: Binding of the ligand to the receptor causes Gα to exchange its bound GDP for GTP, which begins the dissociation of Gα from Gβγ. C: Dissociation
activates Gβγ, which interacts with its targets (Tβγ). D: Dissociation of the Gα-GTP complex from the receptor further activates Gα (increased
shading), which interacts with its targets (Tα). Dissociation of Gα also reduces the ligand-binding affinity of the receptor, which releases the ligand.
E: Dissociation of Gα activates its intrinsic GTPase activity, which dephosphorylates the bound GTP to form the inactive Gα-GDP complex. The latter
then rebinds both the receptor and Gβγ, which increases the ligand-binding affinity of the former and inactivates the latter. This returns these
signaling proteins to the basal state depicted in A.
P.195
Overview of the G Protein Cycle

Interactions between the heterotrimeric G proteins and their receptors allow a single ligand to generate two intracellular signals, one carried by Gα-GTP
and the other by Gβγ, both of which can activate their own downstream targets. At the same time, these reactions help avoid runaway signaling by turning
off the cycle; this is due to the instability of the active Gα-GTP complex, which spontaneously hydrolyzes the bound nucleotide to form the inactive Gα-
GDP that rebinds and inactivates Gβγ. G protein-mediated signals are also attenuated when dissociation of activated Gα-GTP from the ligand-bound
receptor reduces the affinity of the receptor for its ligand.
Regulation of the G Protein Cycle

Several proteins regulate the G protein cycle by modifying the Gα-bound nucleotide. GTPase-activating proteins (GAPs) inhibit the cycle by stimulating
GTP dephosphorylation, which is the rate-limiting step that turns off the cycle. Guanine nucleotide dissociation inhibitors (GDIFs), which inhibit the
release of bound GDP, also slow the cycle, while guanine nucleotide exchange factors (GEFs), which increase the release of Gα-bound GDP, accelerate the
cycle. Signaling by Gα can also be stimulated when Gβγ activates a nucleoside diphosphate kinase (NDPK) that transfers a high-energy phosphate first to Gβ
and then to GDP; this forms GTP that activates Gα.
The risk of runaway signaling is reduced by enzymes that phosphorylate ligand-bound receptors; these include G protein-coupled receptor kinases (GRKs)
and second messenger-dependent protein kinases like protein kinases A and C (see below). Glycosylation and palmitoylation of the receptors,
myristoylation and palmitoylation of the Gα subunits, and interactions of Gα with GAPs and regulators of G protein signaling (RGSs) also regulate the G
protein cycle.
Desensitization of G Protein-Coupled Receptors

Patients who have used nasal sprays containing a β-adrenergic agonist often find that repeated use causes the spray to lose its efficacy. This phenomenon,
once referred to as “tachyphylaxis” and now generally called desensitization, occurs because prolonged activation of these receptors decreases the
number that is available to bind to the ligand. An important clinical example of desensitization is the decreased number of β-receptors in patients with
heart failure that occurs when high levels of circulating norepinephrine cause sustained β-adrenergic stimulation (see Chapter 18).
Desensitization of GPCRs, which can begin a few seconds after the receptor binds its ligand, occurs by a three-step process: uncoupling, internalization,
and digestion (Fig. 8-5).
Uncoupling occurs when activated GPCRs are phosphorylated by protein kinases. In heterologous desensitization, the receptors are phosphorylated in a
single step, whereas homologous desensitization occurs in two steps. In the latter, the activated Gβγ recruits a specific GRK to the ligand-bound receptor
which, after the receptor is phosphorylated by the GRK, binds to β-arrestin, a cofactor that uncouples the receptor from its G protein (Fig. 8-5B). The GRK
that desensitizes the β-adrenergic rector, called βARK (β-adrenergic receptor kinase), prevents the ligand-bound receptor from activating its G protein by
phosphorylating the intracellular C-terminal peptide chain of
P.196
the receptor. Uncoupling is readily reversed when the receptor is dephosphorylated by a G protein-coupled receptor phosphatase, a process called
resensitization.
Fig. 8-5: β-Adrenergic receptor desensitization. A: Activated, ligand-bound receptor. B: Prolonged binding of the β-receptor to its agonist stimulates
a G protein receptor kinase (GRK) called β-adrenergic receptor kinase (βARK), which phosphorylates the C-terminal intracellular peptide chain of
the receptor. The latter then binds a cofactor called β-arrestin that inactivates the receptor. C: Transfer of the phosphorylated receptor from the
plasma membrane to clathrin-coated pits within the cell internalizes the receptor which, although structurally intact, can no longer interact with
either its agonists or G proteins. Dephosphorylation (not shown) allows the internalized receptors to return to the plasma membrane, which
resensitizes the receptor. D: Receptors that remain internalized for long periods are digested by intracellular proteolytic enzymes; this step, unlike
uncoupling and internalization, is irreversible.
Internalization, the second step in desensitization, removes the phosphorylated β-arrestin-bound receptor from the plasma membrane, after which the
receptor (which has already been uncoupled from its G protein) is transferred to a clathrin-coated pit within the cell (Fig. 8-5C). Internalized receptors
can no longer interact with their ligands, but initially remain structurally intact so that, like phosphorylation, internalization is reversible if the receptor
can return to the plasma membrane. A remarkable nuance in signal transduction is the ability of some internalized receptors, which are no longer able to
bind to extracellular messengers that activate functional responses, to participate in proliferative signaling (see Chapter 9); in the case of
P.197
internalized β2-receptors, this occurs when the β2-receptor-β-arrestin complex forms a “scaffold” that activates mitogen-activated protein kinases (MAP
kinases) (Luttrell et al., 1999).
Digestion by proteolytic enzymes, the final step in desensitization (Fig. 8-5D), occurs after receptors remain internalized. Unlike uncoupling and
internalization, digestion is irreversible, so that restoration of receptor function requires the synthesis of new receptors.
Denervation Sensitivity
Prolonged administration of β-adrenergic blockers increases the number of available β-receptors. When this occurs, sudden withdrawal of the β-blocker
finds the heart sensitized to β-adrenergic agonists. This results in a phenomenon called denervation sensitivity, which can have fatal consequences when
prolonged β-blocker therapy is suddenly discontinued in patients with high levels of circulating norepinephrine, as occurs in heart failure. Discontinuation
of β-blockers in these patients can allow the increased number of activated β-receptors to generate enough cAMP to cause sudden cardiac death
(Eichhorn, 1999).
Intracellular Second Messengers

Relatively few neurohumoral signals are mediated by direct coupling of an activated receptor or G protein to an effector (Fig. 8-6A); instead, most
intracellular signal transduction pathways stimulate the generation of small molecules within cells called second messengers, that include nucleotides,
lipids, and phosphosugars (Table 8-10). The prototype of these second messengers, cAMP, is generated when norepinephrine-bound β-receptors activate
Gαs, which stimulates adenylyl cyclase to synthesize cAMP from ATP (Fig. 8-6B). Norepinephrine binding
P.198
P.199
to α1-receptors generates a more complex signal; in this case the activated G protein, Gαq, stimulates phospholipase C (PLC), a lipolytic enzyme that
generates two intracellular messengers, inositol trisphosphate (InsP3) and diacylglycerol (DAG), when it hydrolyzes a membrane phospholipid called
phosphatidylinositol 4,5-bisphosphate (PIP2) (Fig. 8-6C). Other second messengers that participate in cardiovascular regulation include cGMP, which is
generated when natriuretic peptides bind to guanylyl cyclases (see below) and nitric oxide, a reactive gas released from arginine by nitric oxide syntheses.
Calcium which, although obviously not synthesized within cells, enters the cytosol by highly regulated fluxes across membranes (see Chapter 7).
Table 8-10 Major Intracellular Messengers
Second Messenger Initiation of Signal Termination of Signal
Cyclic AMP Synthesized from ATP by adenylyl cyclase Degraded to AMP by phosphodiesterases
Cyclic GMP Synthesized from GTP by guanylyl cyclase Degraded to GMP by phosphodiesterases
InsP3 Synthesized from PIP2 by phospholipase C Dephosphorylated by phosphatases
Diacylglycerol Synthesized from PIP2 by phospholipase C Phosphorylated to form a phosphatide or hydrolyzed to form a
monogylceride
Nitric oxide Synthesized from arginine by nitric oxide Degrades rapidly and spontaneously with a half-life of only a few
synthases seconds
Calcium Diffuses into the cytosol from regions of high Pumped out of cytosol
concentration
Abbreviations: PIP2: phosphatidylinositol 4,5-bisphosphate; InsP3: inositol 1,4,5-trisphosphate.

Fig. 8-6: Three mechanisms by which ligand-binding to a G protein-coupled receptor can modify cell function. A: Direct coupling, where an
activated G protein interacts directly with a target that alters cell function. This is seen when acetylcholine binding to muscarinic receptors
activates Gβγ, which then activates a plasma membrane potassium channel. B and C: Second messenger-mediated coupling, where the activated G
protein-coupled receptor modifies the production of one or more intracellular messengers. B: Activation of adenylyl cyclase by Gαs increases cAMP
production, which increases calcium entry by activating a cAMP-activated protein kinase that phosphorylates plasma membrane L-type calcium
channels. C: Activation of phospholipase C by Gαq stimulates the hydrolysis of phosphatidylinositol, a membrane phospholipid, which releases two
second messengers: diacylglycerol (DAG) and inositol trisphosphate (InsP3).
Signaling by most intracellular second messengers can be regulated by changes in the rates at which they are produced and broken down (Table 8-10);
cAMP, for example, is synthesized by adenylyl cyclase and degraded by phosphodiesterases, both of which are highly regulated. This allows cAMP levels to
be increased when adenylyl cyclase is activated by β-adrenergic agonists, when phosphodiesterase inhibitors inhibit cAMP breakdown, or both. Signals
mediated by calcium are generated when this cation enters the cytosol from a region of high concentration, and end when calcium is actively transported
out of the cytosol by membrane pumps and exchangers (see Chapter 7).
Cyclic AMP
Cyclic AMP, which mediates most of the regulatory cardiovascular responses to sympathetic stimulation, is generated from ATP by adenylyl cyclase (Fig. 8-
7). In the heart this enzyme is stimulated when Gαs is activated by norepinephrine binding to β1-adrenergic receptors. Most responses to cAMP are
mediated by cAMP-dependent protein kinases (PKA). In addition to increasing contractility, accelerating relaxation, and increasing heart rate, cAMP helps
provide substrates needed for the increased energy expenditure by accelerating glycogen breakdown and fatty acid metabolism (see Chapter 2). The major
responses to parasympathetic stimulation, whose cardiovascular effects generally oppose those of the sympathetic nervous system
P.200
and so are counterregulatory, occur when cAMP production is inhibited by Gαi (Fig. 8-7). These signals are turned off when cAMP is degraded by
phosphodiesterases.
Fig. 8-7: Opposing effects of sympathetic (left) and parasympathetic (right) activation on cAMP production. Norepinephrine binding to β1 receptors
on the heart activates Gαs which stimulates adenylyl cyclase to increase cAMP production, whereas acetylcholine binding to muscarinic receptors
activates protein Gαi, which inhibits adenylyl cyclase.
Cyclic GMP
The counterregulatory responses to natriuretic peptides and nitric oxide are mediated in part by cGMP whose effects, which generally oppose those of
cAMP, are mediated by cGMP-dependent protein kinases. The mechanisms that stimulate production of this intracellular messenger differ from those which
activate adenylyl cyclase. Natriuretic peptides increase cGMP production when they bind to receptor guanylyl cyclases, which are plasma membrane
enzymes that contain a single membrane-spanning α-helix. Parasympathetic stimulation causes vasodilatation when acetylcholine binds to muscarinic
receptors on endothelial cells that release nitric oxide that diffuses to adjacent vascular smooth muscle cells, where it increases cGMP levels by activating
soluble guanylyl cyclases in the cytosol (see below). These signals end when phosphodiesterases degrade the cGMP.
Inositol 1,4,5-trisphosphate and Diacylglycerol

Inositol 1,4,5-trisphosphate (InsP3) and diacylglycerol (DAG), the second messengers generated by phospholipase C (Fig. 8-6C), activate different
intracellular signal transduction pathways. InsP3 releases calcium from internal stores by opening intracellular InsP3-gated calcium release channels; the
resulting increase in cytosolic calcium constricts vascular smooth muscle. However, these channels are of little importance in regulating myocardial
contractility because systole is initiated by a much larger calcium flux through the sarcoplasmic reticulum calcium release channels that open during
excitation-contraction coupling (see Chapter 7). DAG has a weak effect to increase myocardial contractility and, in concert with InsP3-gated calcium
release, plays a major role in proliferative signaling (see Chapter 9).
Nitric Oxide
Nitric oxide, a free radical gas whose structural formula is N = O, is released from L-arginine by a family of enzymes called nitric oxide synthase (NOS).
Signaling by nitric oxide occurs when it binds to guanylyl cyclases that synthesize cyclic GMP (see above), and when NO is transferred to cellular proteins
by a process called S-nitrosylation (see below).
Calcium
Cellular responses to calcium, which generally mediates excitatory signals, are influenced by the location of the channels that allow this messenger to
enter the cytosol. In cardiac myocytes, rapid calcium entry through L-type calcium channels in the dyads activates contraction by opening adjacent
intracellular calcium release channels in the sarcoplasmic reticulum (see Chapter 7). The slower increases in cytosolic calcium caused by calcium entry
through plasma membrane T-type calcium channels and intracellular InsP3-gated calcium release channels, which are not located in proximity to the
sarcoplasmic reticulum calcium release channels, participate in proliferative signaling. Functional responses to calcium depend on the specific cell
P.201
type; calcium entry via L-type calcium channels is important for activation of contraction in working cardiac myocytes and conduction in the AV node,
while L- and T-type calcium channels both contribute to pacemaker activity by the SA node.
Signaling Enzymes
Many of the extracellular and intracellular second messengers listed in Tables 8-6 to 8-10 participate in signal transduction by activating enzymes that
modify other proteins. The most important of these signaling enzymes are protein kinases, which transfer phosphate from ATP to serine, threonine, or
tyrosine. Signaling enzymes also transfer methyl groups (methylation), fatty acids (acetylation, myristoylation, and palmitoylation), sugar residues
(glycosylation), and other moieties to proteins. The signals generated by these transfers are turned off when the added groups are removed by other
enzymes.
Protein Kinases
Protein kinases catalyze the formation of phosphoester bonds in a variety of proteins. Phosphoester bonds, which are of low energy, differ from the high-
energy acyl bonds in ATP that energize a variety of energy-consuming reactions. Most protein kinases that participate in functional signaling are
serine/threonine kinases, a large and diverse family of enzymes that include protein kinase A (cAMP-activated), protein kinase G (cGMP-activated), protein
kinase C (phospholipid-activated), and CAM kinase (calcium-activated). Tyrosine kinases and protein kinase B (also called akt), whose major roles are in
proliferative signaling, are discussed in Chapter 9. Protein kinase-generated signals are turned off by phosphoprotein phosphatases, which hydrolyze the
bonds that link the phosphate moieties to effector proteins.
Cyclic AMP-dependent protein kinases are tetramers made up of two catalytic subunits that transfer the terminal phosphate of ATP to the effector protein,
and two regulatory subunits that contain cAMP-binding sites that inhibit protein kinase activity. Binding of cAMP to the regulatory subunits stimulates
protein kinase activity by dissociating the catalytic subunits. Activation of this protein kinase ends when a fall in cytosolic cAMP levels dissociates the
nucleotide from the regulatory subunits, which then bind to and inactivate the catalytic subunits.
Major Extracellular Mediators of the Neurohumoral Response
Norepinephrine and Epinephrine

The most powerful components of the neurohumoral response to underfilling of the arterial system are initiated when norepinephrine released by the
sympathetic nervous system binds to both α- and β-adrenergic receptors (Table 8-11). The responses to epinephrine, which is released from the adrenal
medulla, occur when this catecholamine binds to β-receptors.
α-Adrenergic Receptors
Norepinephrine binding to α1- and α2-receptors generates different cardiovascular responses; activation of α1-receptors on blood vessels (Table 8-11) and,
to a lesser extent on the heart, generates
P.202
a regulatory response, whereas α2-receptors in the central nervous system inhibit sympathetic outflow, and so are counterregulatory.
Table 8-11 Major Adrenergic Receptor Subtypes
α1-Adrenergic Receptors
Increased myocardial contractility (minor)
Smooth muscle contraction—vasoconstriction
Sodium retention by the kidneys
Stimulation of protein synthesis, cell growth, and proliferation
α2-Adrenergic Receptors
Central inhibition of sympathetic activity
Vasodilatation
Cardiac inhibition
β1-Adrenergic Receptors
Cardiac stimulation—positive inotropy, lusitropy, and chronotropy
Stimulation of protein synthesis, cell growth, and proliferation
β2-Adrenergic Receptors
Increased myocardial contractility
Smooth muscle relaxation—vasodilatation
Anti-apoptotic
Regulation of protein synthesis, cell growth, and proliferation (internalized receptor)
Binding of norepinephrine to peripheral α1-receptors activates Gαq/11, which stimulates signaling pathways that generate InsP3 and DAG. The predominant
response is a powerful vasoconstrictor effect in vascular smooth muscle and, in human hearts, a weak positive inotropic effect; both are caused by InsP3-
mediated calcium release. Calcium release through InsP3-gated channels and activation of protein kinase C by DAG in response to α1-receptor activation
are important regulators of proliferative signaling (see Chapter 9).
Because α1-receptors in blood vessels mediate a vasoconstrictor response and α2-receptors in central nervous system and postsynaptic adrenergic neurons
cause vasodilatation, blood pressure can be reduced by both peripheral α1-adrenergic blockers and central α2-adrenergic agonists.
P.203
β-Adrenergic Receptors
Activation of β1- and β2-receptors plays a key role in cardiovascular regulation. Norepinephrine binding to β1-receptors, the major subtype in the human
heart, activates Gαs which stimulates adenylyl cyclase; the resulting increase in cAMP production initiates powerful inotropic, lusitropic, and chronotropic
responses. In addition to these functional responses, β1-receptor activation mediates hypertrophic responses and causes apoptosis (see Chapter 9).
Cardiac β2-receptors, which are coupled to both Gαs and Gαi, have several effects. These include Gαs-mediated regulatory responses that are weaker than
those initiated by β1-receptor activation. Cardiac β2-receptors also activate Gαi which mediates counterregulatory responses and has anti-apoptotic
effects. β2-Receptors in vascular smooth muscle mediate a vasodilator effect that is mediated by cAMP (Table 8-11), but this counterregulatory response is
normally overwhelmed by the vasoconstrictor effects of α1-receptor activation; for this reason, the dominant response of blood vessels to sympathetic
stimulation is vasoconstriction.
Activation of β3-receptors, which mainly regulate gastrointestinal motility and lipolysis, causes weak negative inotropic and vasodilator responses.
Dopamine
Dopamine, an intermediate in the biosynthesis of norepinephrine from tyrosine, acts as an extracellular messenger in both the central nervous system and
peripheral tissues. At low concentrations, dopamine interacts with peripheral DA1 receptors to exert a physiological counterregulatory effect that relaxes
vascular smooth muscle. Higher concentrations stimulate norepinephrine release from sympathetic nerve endings and activate β1-receptors in the heart,
both of which have regulatory effects. At still higher concentrations, dopamine cross-reacts with peripheral α1-receptors to cause vasoconstriction.
Imidazoline Agonists
A central imidazoline-mediated signaling system that has many similarities to the central α2-adrenergic system is activated by a derivative of arginine
called agmatine, the drug moxonidine, and other organic molecules. Binding of imidazoline agonists to I1 and I2 receptors, like central α2-adrenergic
activation, reduces sympathetic outflow and causes counterregulatory responses. Parallels between these counterregulatory systems are also seen in the
ability of α2-receptor agonists to activate central imidazoline receptors, and agmatine and moxonidine to bind to α2-adrenergic receptors. Ligand binding
to I2 receptors located on the mitochondrial outer membrane also inhibits monoamine oxidase, an enzyme that inactivates catecholamines; for this reason,
I2 receptor activation causes regulatory responses that include vasoconstriction and increases in heart rate and myocardial contractility. However, the
dominant effects, which include inhibition of proliferative signaling, are counterregulatory.
Muscarinic Agonists
Acetylcholine, the parasympathetic mediator, binds to at least five muscarinic receptor subtypes, called M1 to M5; all but the M4 receptors have been
found in the heart but, except for the M2 receptors, their functional importance remains unclear. Acetylcholine binding to M2 receptors
P.204
activates Gαi, which slows sinus node depolarization, inhibits conduction through the atrioventricular node, and causes a slight decrease in ventricular
contractility by inhibiting cAMP production. M2-receptor activation also reduces atrial contractility by shortening the atrial action potential; as noted
above, this response is effected by direct coupling of activated Gβγ to potassium channels.
Binding of acetylcholine to M3 receptors in vascular smooth muscle activates phospholipase C to produce InsP3, which causes vasoconstriction; however,
this response is normally outweighed when acetylcholine binds to M3 receptors on endothelial cells, which stimulate the release of NO, which is a powerful
vasodilator. Endothelial damage in patients with atherosclerosis and other vascular diseases can impair the counterregulatory response, which allows the
direct effects of acetylcholine (and other neurotransmitters like serotonin) to evoke an abnormal vasoconstrictor response. In the heart, this abnormal
response is an important cause of coronary vasospasm.
Purinergic Agonists
Adenosine can bind to at least eight GPCR subtypes that are sometimes called P2Y receptors to distinguish them from purine-gated channels, called P2X
receptors, that regulate salt and water transport by the kidneys.
Activation of A1, A2A, A2B, and A3 receptors, which are GPCR (P2Y) subtypes, evoke a variety of cardiovascular responses. Ligand binding to A1 and A3
receptors activates Gαi, which causes the dominant counterregulatory effects on the heart by inhibiting cAMP production. These receptors also regulate
apoptosis and protect the heart against reperfusion injury. Ligand binding to A1 receptors in the SA node slows pacemaker activity by activating inward
rectifying potassium channels that carry iK.Ach (see Chapter 14). In contrast, A2A and A2B receptors are coupled to Gαs so that, when bound to adenosine,
these receptors evoke regulatory responses in the heart by stimulating cAMP production.
In vascular smooth muscle adenosine activates A2B receptors, which cause vasodilatation by a Gαs-mediated increase in cAMP production. Adenosine also
activates Gαq which stimulates phospholipase to release InsP3; the latter opens intracellular calcium release channels and so has a vasoconstrictor effect.
However, the counterregulatory response predominates and adenosine relaxes vascular smooth muscle.
In cells where energy consumption exceeds energy production, ATP breakdown releases adenosine into the extracellular fluid. This exerts a paracrine
vasodilator effect that contributes to autoregulation, a response that increases local blood flow in metabolically active tissues such as exercising muscles.
The Renin-Angiotensin System

Activation of the renin-angiotensin system evokes powerful regulatory responses that include a weak inotropic effect on the heart, vasoconstriction, and
decreased fluid excretion by the kidneys. The renin-angiotensin system also activates counterregulatory responses, but the latter are normally
overwhelmed by the regulatory effects. This system generates proliferative signals that, although less apparent than those responsible for the functional
responses, are important clinically (see Chapters 9 and 18).
The substance initially thought to mediate the vasoconstrictor response was renin, a protein released by ischemic kidneys; however, renin turned out to be
a protease that catalyzes the
P.205
hydrolysis of angiotensinogen, an inactive 14-carbon precursor, to form a decapeptide called angiotensin I. This story became more complex when
angiotensin I was found to be the relatively inactive precursor of a more potent vasoconstrictor whose release from angiotensin I is catalyzed by another
protease called angiotensin converting enzyme (ACE). The active vasoconstrictor was discovered independently by Page and Helmer (1940) in Cleveland,
Ohio, who named it “angiotonin,” and by Braun-Menendez et al. (1940) in Buenos Aires, Argentina, who chose the name “hypertensin”—the term
angiotensin II was coined 18 years later as a compromise (Braun-Menendez and Page, 1958).
The extracellular messenger responsible for most responses of the renin-angiotensin system is the octapeptide angiotensin II, also called angiotensin (1–8),
which is released in both the circulation and tissues (Fig. 8-8). Circulating angiotensin II, which plays an important role in regulating vasomotor tone, is
synthesized when renin released into the bloodstream by the juxtaglomerular apparatus of the kidneys hydrolyzes circulating angiotensinogen made in the
liver to form angiotensin I that is digested further in the lung to form angiotensin II that circulates in the blood (Fig. 8-8A). Angiotensin II is also released
locally in tissues: from angiotensinogen by kallikrein and cathepsin G, and from angiotensin I by tissue ACE and chymase (Fig. 8-8B). The angiotensin II
formed by the circulating system acts on distant targets (endocrine signaling), whereas the locally produced peptide binds to adjacent cells (paracrine
signaling) and to receptors on the same cells that produce this extracellular messenger (autocrine signaling). The angiotensin II produced in the tissues has
important effects in proinflammatory and proliferative signaling (see Chapter 9).
A number of additional peptide mediators are formed by the renin-angiotensin system. Proteolytic cleavage of angiotensin I by ACE2, an ACE isoform,
generates the nonapeptide angiotensin (1–9). Hydrolysis of the latter by ACE forms the heptapeptide angiotensin (1–7). Further proteolysis of angiotensin II
generates two biologically active peptides: the heptapeptide angiotensin III [angiotensin (2–8)] and the hexapeptide angiotensin IV [angiotensin (3–8)].
The effects of angiotensin II are mediated by at least three receptor subtypes, called AT1, AT2, and AT4. The cardiovascular responses to activation of AT1
and AT2, which are GPCRs, generally oppose one another (Fig. 8-9). AT1 receptor activation evokes regulatory responses that include a weak effect to
increase myocardial contractility, vasoconstriction, inhibition of sodium excretion by the kidneys, and stimulation of cardiac myocyte hypertrophy. In
contrast, the AT2 receptor generally causes counterregulatory responses that include vasodilatation and growth inhibition.
Angiotensins II and III bind to both AT1 and AT2 receptors, but the predominant responses are regulatory. Angiotensin IV is both a weak AT1 receptor agonist
that evokes vasoconstrictor, proliferative, and inflammatory responses, and an AT2 receptor agonist. Angiotensin IV also activates AT4 receptors, which are
insulin-regulated aminopeptidases rather than GPCRs that have vasodilator and proinflammatory effects. The nonapeptide angiotensin (1–9) evokes a
vasoconstrictor response, but appears not to have important effects on human coronary arteries (Campbell et al., 2004). The heptapeptide angiotensin (1–
7), which is formed when angiotensin (1–9) is hydrolyzed by ACE, has counterregulatory actions that cause vasodilatation and inhibit proliferative signaling.
AT1, AT2, and AT4 receptors are found in the central nervous system, where activation of AT1 receptors amplifies regulatory responses by increasing
sympathetic outflow. Central AT2 receptor activation reduces sympathetic outflow, which has an opposing counterregulatory effect. Binding of angiotensin
IV to central AT4 receptors evokes a variety of neurological responses. The
P.206
ACE enzyme itself also participates in proliferative and proinflammatory signaling by activating MAP kinase pathways.
Fig. 8-8: Generation of active signaling peptides by the renin-angiotensin system. A: The “classical” description of the circulating system includes
two proteolytic cleavages that are catalyzed by renin and angiotensin converting enzyme (ACE). Renin catalyzes the formation of angiotensin I from
the inactive precursor angiotensinogen and ACE releases angiotensin II from angiotensin I. B: Additional proteolytic reactions in the tissue system
are catalyzed by chymase, cathepsin G, kallikrein, and ACE2, which form additional biologically active peptides.
The renin-angiotensin system, along with other signaling systems, has important effects on the release of other mediators of the neurohumoral response
(see below). The ability of angiotensin II to initiate regulatory amplifications by stimulating the secretion of aldosterone, catecholamines, vasopressin, and
endothelin can worsen many of the vicious cycles seen in patients with heart failure.
P.207
Fig. 8-9: Angiotensin II receptor subtypes. The responses to the AT1 and AT2 subtypes differ, and in many cases oppose one another. The AT1
receptors, which include AT1a and AT1b subtypes, exert regulatory effects, whereas AT2 receptor stimulation generally evokes counterregulatory
responses.
Bradykinin and Related Peptides

Kinins are peptides released from inactive precursors called kininogens by proteolytic enzymes called kallikreins (Fig. 8-10). Levels of bradykinin, the
octapeptide Arg-Pro-Gly-Phe-Ser-Pro-Phe-Arg, and kallidin (lysyl-bradykinin) are regulated by the rates at which the peptides are released from their
precursors and broken down into inactive fragments. Bradykinin hydrolysis is catalyzed by converting enzymes, including the ACE that releases angiotensin
II (see above); this means that ACE has synergistic effects on vascular tone by generating a vasoconstrictor (angiotensin II) and hydrolyzing a vasodilator
(bradykinin).
Low concentrations of most kinins cause vasodilatation, whereas high concentrations are proinflammatory. These biological actions are mediated by two
types of GPCR called B1 and B2, both of which activate Gαi and Gαq (Fig. 8-11). The responses to activation of B2 receptors, which are constitutive in most
cells, are mainly counterregulatory. These include vascular smooth muscle relaxation that is brought about by physiological amounts of NO released when
the activated B2 receptors stimulate endothelial cell nitric oxide synthase (eNOS), and when Gαq activates phospholipase A2; the latter increases the
production of prostacyclin (PGI2) and prostaglandin E2 (PGE2), which stimulate adenylyl cyclase to generate cAMP.
Fig. 8-10: Kinin production and breakdown. Kinins are released when kininogens are hydrolyzed by enzymes called kallikreins. Active kinins include
bradykinin and kallidin, which are inactivated by the same converting enzymes that release angiotensin II.
P.208
Fig. 8-11: Kinin receptor subtypes. Binding of bradykinin and kallidin to B2 receptors mediate physiological responses that participate in circulatory
responses, notably vasodilatation, while activation of B1 receptors mediates inflammatory responses. Both receptor subtypes modify proliferative
signaling, but they evoke different responses.
Proinflammatory cytokines increase expression of inducible B1 receptors, which are present at low levels in normal cells. The greater number of activated
B1 receptors increases the synthesis of inducible nitric oxide synthase (iNOS) which releases high concentrations of NO that participate in inflammation.
Like the constitutive B2 receptors, B1 receptor activation causes vasodilatation. Bradykinin can cause a cough and, rarely, a dangerous angioneurotic
edema.
Both receptor types also activate proliferative signaling pathways, but the responses are different. B1 receptors activate stress-activated MAP kinase
pathways that mediate maladaptive hypertrophy, while B2 receptors mediate an adaptive hypertrophic response that is beneficial in failing hearts (see
Chapters 9 and 18).
Endothelin
Endothelins (ET), a family of peptide vasoconstrictors first isolated from endothelial cells, are released by many cell types; active peptides include ET-1,
which is expressed in cardiac myocytes; ET-3, which mediates the release of nitric oxide by endothelial cells; and ET-2, which is expressed in ovaries and
the gastrointestinal cells but has no known role in cardiovascular regulation. Endothelin production is regulated by the rate of synthesis of large
precursors, called preproendothelins, which are processed by endopeptidases to release proendothelins (big endothelins) (Fig. 8-12). The latter are
hydrolyzed by endothelin converting enzymes to form endothelins. Other proteolytic enzymes form endothelins by hydrolyzing proendothelins, and
degrade the active peptides.
Preproendothelin synthesis is stimulated by several mediators of the neurohumoral response, including angiotensin II, norepinephrine, vasopressin, peptide
growth factors, and the cytokine interleukin-1; synthesis is inhibited by atrial natriuretic peptide (ANP), nitric oxide, and prostaglandins. Endothelins are
released from blood vessels in response to epinephrine, angiotensin II, cytokines, growth factors, and high shear stress along the endothelium. Although
P.209
endothelins were initially found to circulate in the plasma, most of their major actions are mediated by paracrine signaling.
Fig. 8-12: Endothelin production. Expression of endothelin genes produces preproendothelin, a large peptide that is processed by proteolytic
reactions that generate proendothelin (big endothelin) and then endothelin.
ET-1 binds to two types of GPCR, ETA and ETB (Fig. 8-13). ETA receptors, whose effects in vascular smooth muscle and the heart are mediated largely by
Gαq, activate phospholipase C to form InsP3 and DAG. These pathways allow ETA receptors to mediate regulatory functional responses, notably
vasoconstriction and increased myocardial contractility. The responses to endothelin-bound ETA receptors predominate over those evoked by the ETB
receptors, which are generally counterregulatory. ETB receptors, which are found in blood vessels, are coupled mainly to Gαi, which causes vasodilatation
by stimulating the release of NO and prostacyclin. However, some ETB receptor-mediated responses are mediated by Gαq, and so can cause
vasoconstriction. Activation of Gαq when ET-1 binds to the ETA receptor stimulates proliferative signaling by activating several MAP kinase pathways (see
Chapter 9).
Fig. 8-13: Endothelin receptors. ET-A and ET-B receptor subtypes, which have different affinities for endothelins, activate G protein-coupling
receptors that mediate a variety of responses. The responses to ET-A, which activates Gαs, are usually regulatory, whereas the responses to ET-B,
which activates Gαi, are usually counterregulatory. However, both cause vasoconstriction by activating Gαq.
P.210
Arginine Vasopressin (Antidiuretic Hormone, ADH)

Vasopressin, the major regulator of the body's water balance, is an octapeptide that is synthesized in the supraoptic and paraventricular nuclei of the
hypothalamus, after which it is transported to the posterior pituitary where it is stored and released. The human isoform is often called arginine
vasopressin because it contains an arginine in position 8. By increasing the water permeability of the renal collecting ducts, vasopressin increases water
reabsorbtion; as this inhibits diuresis, this peptide is often called antidiuretic hormone (ADH). Vasopressin also increases thirst, a physiological response
that drives water-deprived individuals to seek water.
Interactions of vasopressin with different receptors allow this peptide to initiate several responses (Fig. 8-14). Binding to V1a receptors can cause both
vasoconstriction and vasodilatation. Vasoconstriction, the predominant effect, occurs when vasopressin activates Gαq in vascular smooth muscle, which
activates phospholipase C that increases cytosolic calcium. V1a receptor activation also mediates a hypertrophic response in the heart. Vasopressin-binding
to V2 receptors in the kidneys is responsible for water retention; this response is initiated when vasopressin-bound V2 receptors activate Gαs, which
stimulates adenylyl cyclase activity; the resulting increase in cAMP levels causes a cAMP-dependent protein kinase to phosphorylate aquaporin in the
collecting ducts. Water retention occurs when the phosphorylated aquaporin is translocated to the plasma membrane where it forms water-permeable
channels.
The most important physiological stimulus for vasopressin secretion is increased plasma osmolarity, which stimulates hypothalamic osmoreceptors that
release this peptide from the posterior pituitary (Fig. 8-15). This response helps travelers survive a desert crossing because when plasma osmolarity
increases in a water-deprived individual, vasopressin increases water reabsorbtion by the kidneys. The latter also stimulates hypothalamic osmoreceptors
that drive the dehydrated traveler to seek water by increasing thirst. The antidiuretic response and thirst are turned off when plasma osmolarity returns
to physiological levels. Atrial stretch in volume-overloaded patients inhibits vasopressin release, which reduces blood volume by promoting water
excretion and decreasing thirst; conversely, fluid loss decreases atrial volume, which by causing vasopressin release from the posterior pituitary, helps
restore blood volume by promoting water retention and increasing thirst. Because vasopressin is a vasoconstrictor, this peptide also helps maintain blood
pressure in volume-depleted individuals.
Vasopressin release can become maladaptive in patients with heart failure because these physiological control mechanisms are blunted when chronic
reduction of arterial filling and a central effect of increased angiotensin II levels increase vasopressin levels. Excessive vasopressin
P.211
release, which occurs in severe heart failure—especially after administration of diuretics—causes a dilutional hyponatremia that may be fatal.
Fig. 8-14: Vasopressin receptor subtypes. V1 receptors mediate regulatory vasoconstriction, while counterregulatory vasodilatation is mediated by
V2 receptors.
Fig. 8-15: Regulation of vasopressin release. Under physiological conditions, vasopressin secreted in response to increased plasma osmolarity and
decreased blood volume increases water retention by the kidneys, increases thirst, and causes vasoconstriction. These adaptive responses restore
plasma osmolarity and blood volume, and increase blood pressure. In heart failure, where vasopressin release comes under neurohumoral control,
inappropriate secretion of this peptide becomes maladaptive when increased thirst and water cause dilutional hyponatremia.
Natriuretic Peptides
The finding that the density of granules in atrial myocardial cells changes in response to altered water and electrolyte balance led to the discovery that
the heart is an endocrine organ as well as a pump (DeBold, 1979). Analysis of these granules showed that they contain peptides which, when released in
response to atrial stretch, stimulate a counterregulatory response that defends against volume overload by inducing a diuresis and a natriuresis, along with
systemic arteriolar vasodilatation. The first of these peptides to be discovered was secreted by the atria, and so was named atrial natriuretic peptide
(ANP); the second, which was first isolated from brain, was named brain natriuretic peptide (BNP); this is the most important of these peptides in human
hearts. At the time this chapter was written, six natriuretic peptides that influence the mammalian cardiovascular system had been identified (ANP, BNP,
CNP, DNP, VNP, and urodilatin).
P.212
Fig. 8-16: Production of BNP. Pre-proBNP, a large peptide containing 134 amino acids, is processed by removal of a signal peptide to generate
proBNP, which contains 108 amino acids. Proteolysis of the latter releases BNP, which contains 32 amino acids.
Formation of BNP begins when a signal peptide is removed from pre-proBNP, a 134 amino acid prohormone, to form proBNP, which contains 108 amino acids
(Fig. 8-16). Proteolysis of the latter forms active BNP, which is composed of 32 amino acids; additional biologically active products of proBNP and BNP
proteolysis have been found.
Natriuretic peptides bind to natriuretic peptide receptors-A and -B (NPR-A and NPR-B) which are members of a family of receptor guanylyl cyclases that
synthesize cyclic GMP. ANP and BNP evoke their counterregulatory effects when they bind to NPR-A, which initiates diuretic, natriuretic, and vasodilatory
responses. ANP also has a weak negative inotropic effect and BNP inhibits fibrosis, due possibly to downregulation of the renin-angiotensin-aldosterone
system. Binding of CNP to NPR-B in vascular endothelium also has a vasodilator effect. Many of the counterregulatory responses to CNP are mediated by
natriuretic peptide receptor-C (NPR-C), an atypical inhibitory GPCR that is a homodimer of two peptides with only a single membrane-spanning α-helix.
NPR-C lacks a guanylyl cyclase domain, but instead exerts counterregulatory and anti-proliferative effects by activating both GαI, which inhibits cAMP
production by adenylyl cyclase, and a Gβγ component that activates a phospholipase C.
The proliferative responses to hemodynamic overload activate natriuretic peptide synthesis. For example, the gene encoding ANP, which is normally
expressed in fetal but not adult ventricles, is re-expressed in failing adult human ventricles (see Chapter 18). Stimulation of proBNP and BNP production in
hypertrophied human ventricles causes an increase in the blood levels of these peptides that can be used to help identify patients with heart failure and
characterize the severity of the hemodynamic abnormality.
Calcitonin Gene-Related Peptide, Adrenomedullin, Intermedin

Members of the calcitonin family of signaling peptides, which include calcitonin gene-related peptide (CGRP), adrenomedullin, and intermedin,
participate in cardiovascular regulation. All are powerful vasodilators, but their effects on the heart and kidneys are not well understood.
Calcitonin-related signaling peptides bind to two closely related GPCRs whose activity is regulated by receptor activity modifying proteins (RAMPs), which
are single membrane-spanning proteins that regulate trafficking of the receptors to the cell surface. When bound to their ligands, these receptors activate
Gαs, which increases adenylyl cyclase activity; activation of Gαq has also been described. The vasodilator effects of these peptides, which are mediated by
endothelium-dependent and endothelium-independent mechanisms that increase levels of cAMP, cGMP, and NO can be initiated by ischemia and hypoxia,
mechanical stress, glucocorticoids, and cytokines. Reports of positive inotropic and chronotropic effects can be explained by the ability of these peptides
to activate Gαs, while effects on cell growth and proliferation
P.213
have been attributed to stimulation of Gαq, phospholipase-β, and phosphoinositide 3-kinase (see Chapter 9).
Adrenomedullin evokes natriuretic and diuretic responses, and has both proinflammatory and anti-proliferative effects. CGRP also increases urine flow,
whereas intermedin has been reported to have an opposite effect, due possibly to its ability to lower blood pressure. Increased expression of both the
GPCRs that bind the calcitonin family of signaling peptides and RAMPs has been observed in heart failure.
Other Peptides
Apelin (APJ endogenous ligand) activates GPCRs called APJ in several cell types, including vascular epithelium and cardiac myocytes. Several apelin
isoforms, ranging in size from 36 to 12 amino acids, are generated by proteolysis of a 77 amino acid preproprotein. Apelin-mediated responses often
oppose those initiated by angiotensin II. In blood vessels, apelin mediates a counterregulatory vasodilatation, but the cardiac response, increased
contractility, is regulatory. Unlike most of the regulatory extracellular messengers that participate in the hemodynamic defense reaction, apelin does not
appear to stimulate myocardial hypertrophy. In contrast to angiotensin II levels, which rise in heart failure, apelin levels decrease in this syndrome.
Neuropeptide Y evokes a variety of functional and proliferative responses when it interacts with at least six different receptor subtypes, called Y1 to Y6, in
the central nervous system and peripheral tissues. The peripheral effects of neuropeptide Y, which are both regulatory and counterregulatory, include a
powerful centrally mediated regulatory response. In the heart, which has Y1, Y2, Y3, and Y5 receptors, neuropeptide Y slows pacemaker activity by
reducing the pacemaker current ih, increases contractility by opening iCaL channels, and modifies excitability and contractility by reducing the transient
outward current ito; some of these responses are due to activation of Gαi, others to activation of Gαq. Activation of the latter contributes to the ability of
neuropeptide Y to stimulate overload-induced cardiac hypertrophy. In blood vessels, neuropeptide Y causes vasoconstriction by activating α1-adrenergic
and angiotensin receptors.
Ghrelin stimulates the secretion of growth hormone when it binds to a G protein-coupled growth hormone secretagogue receptor. Cardiovascular
responses include vasodilatation and increased contractility; this peptide also has anti-inflammatory and anti-apoptotic effects that can alleviate cardiac
cachexia.
Leptin, a catabolic peptide released by adipocytes, binds to members of the family of gp130 cytokine receptors whose effects are mediated by the Janus
kinase/signal transducers and activators of transcription pathway (see Chapter 9). Cardiovascular responses to leptin include vasoconstriction caused by
endothelial cell-mediated vasodilatation, a negative inotropic effect that is mediated in part by NO, and proinflammatory and proliferative effects.
Responses similar to those of leptin can be initiated by other peptides, including corticotrophin-releasing factor and α-melanocyte-releasing factor.
Nitric Oxide (NO)

More than 30 years ago, Furchgott and Zawadzki (1980) discovered that the vasodilator responses to several extracellular messengers are not caused by a
direct effect on vascular smooth muscle, but instead depend on a vasodilator substance derived from endothelial cells. This physiological vasodilator,
initially called endothelial-derived relaxing factor, is now known to be nitric oxide (NO). In the cardiovascular system, three NOS isoforms generate NO.
NOS1
P.214
(neuronal NOS) and NOS3 (endothelial NOS) are constitutive enzymes that participate in physiological signaling, while iNOS (inducible NOS or NOS2)
generates the large amounts of NO that act as toxic-free radicals during inflammation. Both NOS1 and NOS3 act as autocrine and paracrine regulators that
generally mediate counterregulatory functions. In mammalian hearts, NOS1 is located in the sarcoplasmic reticulum and mitochondria, while NOS3
interacts with caveoli in the plasma membrane to modify the production and actions of cAMP.
The vasodilator effect occurs when NO activates soluble guanylyl cyclases. NO generated by NOS1 and NOS3 also slows the heart and evokes a bimodal
inotropic response in which contractility increases at lower NO concentrations and decreases at higher concentrations. These responses are mediated both
by stimulation of cyclic GMP formation and by S-nitrosylation, in which N = O groups are transferred to cysteine residues in a variety of receptors, ion
channels, structural proteins, and transcription factors; NOS-mediated S-nitrosylation is reversed when the NO groups are removed by S-nitrosoglutathione
reductases. The specificity of S-nitrosylation is determined in part by interactions between NOS and the PDZ-binding domains of cytoskeletal adaptor
proteins (see Chapter 5). NOS1 and NOS3 also alter the production of reactive oxygen and nitrogen species (ROS and RNS) that can modify signaling by S-
nitrosylation. The ability of these NOS to increase ROS production and decrease RNS production attenuates abnormalities in nitroso-redox balance in
ischemic and failing hearts.
Aldosterone
Aldosterone, a steroid produced by the zona glomerulosa of the adrenal cortex, acts on the renal tubules to increase sodium reabsorbtion and the
excretion of potassium and hydrogen ions. In addition to regulating blood volume and the mineral composition of the blood, aldosterone stimulates fibrosis
and promotes maladaptive proliferative responses in the heart.
Aldosterone secretion is regulated differently in normal individuals and patients with heart failure (Fig. 8-17). Low blood volume, an important
physiological stimulus for aldosterone secretion, causes the pituitary to release adrenocorticotrophic hormone (ACTH) which stimulates
P.215
aldosterone synthesis when it binds to receptors in the adrenal cortex; ACTH also modifies the normal diurnal changes in aldosterone secretion.
Hyperkalemia, another physiological stimulus for aldosterone release, increases potassium excretion by the renal tubules. In heart failure, different
mechanisms stimulate aldosterone release. When heart failure is severe, the most important stimulus is the high angiotensin II level (see Chapter 18);
other mediators of the neurohumoral response, notably sympathetic stimulation, vasopressin, and endothelin, also play an important role in stimulating
aldosterone secretion in milder heart failure.
Fig. 8-17: Regulation of aldosterone synthesis and release. Different mechanisms operate in normal individuals and patients with heart failure. In
the former, this steroid serves mainly to maintain normal serum sodium and potassium levels, whereas in heart failure, aldosterone secretion is
increased by mediators of the neurohumoral response and promotes fluid retention by the kidneys.
Prostaglandins and Related Compounds

Two classes of signaling molecules are formed by the insertion of double bonds into arachidonic acid, a 20 carbon polyunsaturated fatty acid; these are
prostaglandins, whose formation is catalyzed by cyclooxygenases (COX), and leukotrienes, which are formed by lipoxygenases. Both are short-lived lipid
messengers that reach their target cells by paracrine and autocrine pathways, where they stimulate functional, proinflammatory, and proliferative
responses. The prostaglandins that are most important in cardiovascular regulation are prostacyclin (PGI2) and prostaglandin E2 (PGE2), which are
counterregulatory mediators that relax vascular smooth muscle and inhibit platelet aggregation and adhesion, and thromboxane A (TxA2), a regulatory
prostaglandin that causes vasoconstriction, platelet aggregation, and proliferative responses. The responses to these physiologically active prostaglandins
are mediated by GPCRs; PGI2 by DP receptors, PGE2 by EP receptors, and TxA2 by TP receptors.
Two major COX isoforms synthesize prostaglandins: COX-1, a “housekeeping” protein, helps maintain homeostasis by synthesizing thromboxane, while COX-
2 catalyzes production of PGI2 and PGE2. Aspirin, which irreversibly inhibits TxA2 production by acetylating COX-1, decreases the risk of complications of
vascular disease, whereas selective inhibition of COX-2 has the opposite effect by increasing the risk of thrombotic damage and clotting by reducing PGI2
synthesis. These responses are important in the pathophysiology of atherosclerosis, plaque rupture, and clotting, the major causes of heart failure in
developed countries. Ventricular hypertrophy has been found to be stimulated by TxA2 and inhibited by reduced PGI2.
Interactions Among Mediators of the Neurohumoral Response
The functional components of the neurohumoral response enable the body to make the rapid adjustments needed to compensate for what Bernard (1878)
called “perpetual changes of external conditions” (see above). Minor short-term stresses, such as altered blood flow caused by changes in body position,
initiate responses that are analogous to the small adjustments in steering made under normal conditions by an automobile driver. In an emergency, such as
flight from a predator or after hemorrhage, these mechanisms are amplified, as occurs when the driver swerves to avoid an unexpected obstacle.
Interactions between different signaling systems help integrate regulatory responses, as seen, for example, in the ability of sympathetic stimulation to
promote the release of angiotensin II, aldosterone, endothelin, and vasopressin. Conversely, counterregulatory effects help prevent “runaway signaling,”
which could happen in the automobile if, when trying to avoid the obstacle, the driver turned the steering wheel too far and went off the road.
The long-term activation of the neurohumoral response in patients with heart failure has been described as a “grand design [that maintains] intravascular
volume and sufficient perfusion
P.216
pressure to vital organs” (Francis and McDonald, 1995). This description, which echoes the evolutionary interpretation provided by Harris (1983), helps in
understanding the interplay between activation of sympathetic outflow, the renin-angiotensin system, and vasopressin secretion, the three major
regulatory systems that are activate in heart failure, and how cardiac stimulation, fluid retention, and vasoconstriction interact in these patients. There
are important differences in the timing of these responses. Increased sympathetic outflow acts very rapidly on the heart and vasculature to maintain blood
pressure and cardiac output, while slower activation of the renin-angiotensin system and vasopressin secretion has prominent, but delayed, effects that
promote fluid retention by the kidneys. Equally important are proliferative responses to prolonged neurohumoral activation in heart failure (see Chapters 9
and 18). The latter, which from an evolutionary standpoint are more ancient and so more complex than the functional components of the hemodynamic
defense reaction described in this chapter, are now known to be major determinants of outcome in these patients. Understanding of this grand design and
its individual components is therefore essential for the formulation of an optimal therapeutic plan, and the ability to adjust this plan to meet the changing
pathophysiology in these patients.
Bibliography
General
Alberts B, Johnson A, Lewis J, et al. Molecular biology of the cell. 5th ed. New York, NY: Garland, 2008.
Devlin T. Textbook of biochemistry. New York, NY: Wiley-Liss, 1997.
Francis GS. Neuroendocrine activity in congestive heart failure. Am J Cardiol 1990;66:33D–39D.
Landry DW, Oliver JA. The pathogenesis of vasodilatory shock. New Engl J Med 2001;345:588–595.
Olson EN. A decade of discoveries in cardiac biology. Nature Med 2004;10:467–474.
G Protein-Coupled Receptors
DerMardirossian C, Bokoc GM. GDIs: central regulatory molecules in Rho GTPase activation. Trends Cell Biol 2005;15:356–363.
Hermans E. Biochemical and pharmacological control of the multiplicity of coupling at G-protein-coupled receptors. Pharmacol Ther 2003;99:25–44.
Clapham DE, Neer EJ. G protein bg subunits. Ann Rev Pharmacol 1997;37:167–203.
Smrcka AV. G protein bg subunits: central mediators of G protein coupled receptor signaling. Cell Mol Life Sci 2008;65:2192–2214.
Tang C-M, Insell PA. GPCR expression in the heart. “New” receptors in myocytes and fibroblasts. Trends Cardiovasc Med 2004;14:94–99.
Wieland T. Interaction of nucleoside diphosphate kinase B with heterotrimeric G protein betagamma dimers: consequences on G protein activation
and stability. NaunynSchmiedebergs Arch Pharmacol 2007;374:373–383.
Epinephrine, Norepinephrine, and Dopamine
Brodde OE, Bruck H, Leineweber K. Cardiac adrenoceptors: physiological and pathophysiological relevance. J Pharmacol Sci 2006;100:323–337.
Lefkowitz RJ, Pitcher J, Krueger K, et al. Mechanisms of β-adrenergic receptor desensitization and resensitization. Adv Pharmacol 1997;42:416–420.
P.217
Xiao RP, Zhu W, Zheng M, et al. Subtype-specific β-adrenoceptor signaling pathways in the heart and their potential clinical implications. Trends
Pharmacol Sci 2004;25:358–365.
Xiao RP, Zhuc W, Zheng M, et al. Subtype-specific α1- and β-adrenoceptor signaling in the heart. Trends Pharmacol Sci 2006;27:330–337.
Ursino MG, Vasina V, Raschi E, et al. The beta3-adrenoceptor as a therapeutic target: current perspectives. Pharmacol Res 2009;59:221–234.
Imidazoline Agonists
Head GA, Mayorov DN. Imidazoline receptors, novel agents and therapeutic potential. Cardiovasc Hematol Agents Med Chem 2006;4:17–32.
Raasch W, Scháfer, Chun J, et al. Biological significance of agmatine, an endogenous ligand at imidazoline binding sites. Br J Pharmacol
2001;133:755–780.
Muscarinic Agonists
Abrams P, Andersson KE, Buccafusco JJ, et al. Muscarinic receptors: their distribution and function in body systems, and the implications for treating
overactive bladder. Br J Pharmacol 2006;148:565–578.
Kuhn M. Structure, regulation, and function of mammalian membrane guanylyl cyclase receptors, with a focus on guanylyl cyclase-A. Circ Res
2003;93:700–709.
Olshansky B, Sabbah HN, Hauptman PJ, et al. Parasympathetic nervous system and heart failure: pathophysiology and potential implications for
therapy. Circulation 2008;118:863–871.
Fitzpatrick DA, O'Halloran DM, Burnell AM. Multiple lineage specific expansions within the guanylyl cyclase gene family. BMC Evol Biol 2006;6:26.
Purinergic Agonists
Banfi C, Ferrario S, De Vincenti O, et al. P2 receptors in human heart: upregulation of P2×6 in patients undergoing heart transplantation, interaction
with TNF alpha and potential role in myocardial cell death. J Mol Cell Cardiol 2005;39:929–939.
Cohen MV, Downey JM. Adenosine: trigger and mediator of cardioprotection. Basic Res Cardiol 2008;103: 203–215.
Villarreal F, Zimmermann S, Makhsudova L, et al. Modulation of cardiac remodeling by adenosine: in vitro and in vivo effects. Mol Cell Biochem
2003;251:17–26.
Yaar R, Jones MR, Chen JF, et al. Animal models for the study of adenosine receptor function. J Cell Physiol 2005;202:9–29.
The Renin-Angiotensin System
Ardaillou R, Chansel D. Synthesis and effects of active fragments of angiotensin II. Kidney Int 1997;52: 1458–1468.
Burns KD. The emerging role of angiotensin-converting enzyme-2 in the kidney. Curr Opin Nephrol Hypertens 2007;16:116–121.
Danilczyk U, Penninger JM. Angiotensin-converting enzyme II in the heart and the kidney. Circ Res 2006; 98:463–471.
Fleming I. Signaling by the angiotensin-converting enzyme. Circ Res 2006;98:887–896.
Gao L, Wang WZ, Wang W, et al. Imbalance of angiotensin type 1 receptor and angiotensin II type 2 receptor in the rostralventrolateral medulla:
potential mechanism for sympathetic overactivity in heart failure. Hypertension 2008;52:708–714.
P.218
Lemariá CA, Schiffrin EL. The angiotensin II type 2 receptor in cardiovascular disease. J Renin Angiotensin Aldosterone Syst 2010;11:19–31.
Martin P, Mehr AP, Kreutz R. Physiology of local renin-angiotensin systems. Physiol Rev 2006;86:747–803.
Stragier B, De Bundel D, Sarre S, et al. Involvement of insulin-regulated aminopeptidase in the effects of the renin-angiotensin fragment angiotensin
IV: a review. Heart Fail Rev 2008;13:321–337.
van Esch JH, Oosterveer CR, Batenburg WW, et al. Effects of angiotensin II and its metabolites in the rat coronary vascular bed: is angiotensin III the
preferred ligand of the angiotensin AT2 receptor? Eur J Pharmacol 2008;588:286–293.
Watanabe T, Barker TA, Berk BC. Angiotensin II and the endothelium: diverse signals and effects. Hypertension 2005;45:163–169.
Bradykinin and Related Peptides
Carretero OA, Scicli AG. Kinins as regulators of blood flow and blood pressure. In: Laragh JH, Brenner BM, eds. Hypertension. Pathophysiology,
Diagnosis, and Management. New York, NY: Raven Press, 1990:805–817.
Leeb-Lundberg LM. Bradykinin specificity and signaling at GPR100 and B2 kinin receptors. Br J Pharmacol 2004;143:931–932.
Marcondes S, Antunes E. The plasma and tissue kininogen-kallikrein-kinin system: role in the cardiovascular system. Curr Med Chem Cardiovasc
Hematol Agents 2005;3:33–44.
Mombouli JV, Vanhoutte PM. Heterogeneity of endothelium-dependent vasodilator effects of angiotensin-converting enzyme inhibitors: role of
bradykinin generation during ACE inhibition. J Cardiovasc Pharmacol 1992;20(Suppl 9):S974–S982.
Moreau ME, Garbacki N, Molinaro G, et al. The kallikrein-kinin system: current and future pharmacological targets. J Pharmacol Sci 2005;99:6–38.
Rose RA, Giles WR. Natriuretic peptide C receptor signalling in the heart and vasculature. J Physiol 2008;586:353–366.
Urata H, Arakawa K. Angiotensin II-forming systems in cardiovascular diseases. Heart Failure Rev 1998;3: 119–124.
Endothelin
Attina T, Camidge R, Newby DE, et al. Endothelin antagonism in pulmonary hypertension, heart failure, and beyond. Heart 2005;91:825–831.
Barton M, Yanagisawa M. Endothelin: 20 years from discovery to therapy. Can J Physiol Pharmacol 2008;86:485–498.
Kedzierski RM, Yanagisawa M. Endothelin system: the double-edged sword in health and disease. Ann Rev Pharmacol Toxicol 2001;41:851–876.
Levin ER. Endothelins. New Eng J Med 1996;333:356–362.
Marasciulo FL, Montagnani M, Potenza MA. Endothelin-1: the yin and yang on vascular function. Curr Med Chem 2006;13:1655–1665.
Schorlemmer A, Matter ML, Shohet RV. Cardioprotective signaling by endothelin. Trends CV Med 2008; 18:233–239.
Vasopressin and Aquaporin
Boone M, Deen PM. Physiology and pathophysiology of the vasopressin-regulated renal water reabsorbtion. Pflugers Arch 2008;456:1005–1024.
Goldsmith SR, Gheorghiade M. Vasopressin antagonism in heart failure. J Am Coll Cardiol 2005;46: 1785–1791.
P.219
Kozniewska E, Romaniuk K. Vasopressin in vascular regulation and water homeostasis in the brain. J Physiol Pharmacol 2008;59(Suppl 8):109–116.
Rosner MH. Hyponatremia in heart failure: the role of arginine vasopressin and diuretics. Cardiovasc Drugs Ther 2009;23:307–315.
Natriuretic Peptides
Abassi Z, Karram T, Ellaham S, et al. Implications of the natriuretic peptide system in the pathogenesis of heart failure: diagnostic and therapeutic
importance. Pharmacol Ther 2004;102:223–241.
Boomama F, Van derMeiracker AH. Plasma A- and B-type natriuretic peptides: physiology, methodology and clinical use. Cardiovasc Res 2001;51:442–
449.
Garbers DL, Chrisman TD, Wiegn P, et al. Membrane guanylyl cyclase receptors: an update. Trends Endocrinol Metab 2006;17:251–258.
Levin ER, Gardner DG, Samson WK. Natriuretic peptides. New Eng J Med 1998;339:321–328.
Martinez-Rumayor A, Richards AM, Burnett JC, et al. Biology of the natriuretic peptides. Am J Cardiol 2008;101(Suppl 3A):3–8.
Adrenomedullin, Intermedin, Calcitonin Gene-Related Peptide
Bell D, McDermott BJ. Intermedin (adrenomedullin-1): a novel counter-regulatory peptide in the cardiovascular and renal systems. Br J Pharmacol
2008;153:S247–S262.
Brain SD, Grant AD. Vascular actions of calcitonin gene-related peptide and adrenomedullin. Physiol Rev 2004;84:903–934.
Eto T, Kato J, Kitamura K. Regulation of production and secretion of adrenomedullin in the cardiovascular system. Regul Pept 2003;112:61–69.
Kuwasako K, Cao YN, Nagoshi Y, et al. Adrenomedullin receptors: pharmacological features and possible pathophysiological roles. Peptides
2004;11:2003–2012.
Tsuruda T, Burnett JC Jr. Adrenomedullin. An autocrine/paracrine factor for cardiorenal protection. Circ Res 2002;90:625–627.
Yanagawa B, Nagaya N. Zhang Z, et al. Adrenomedullin: molecular mechanisms and its role in cardiac disease. Amino Acids 2007;32:157–164.
Apelin, Neuropeptide Y, Ghrelin, Leptin
Chandrasekaran B, Kalra PR, Donovan J, et al. Myocardial apelin production is reduced in humans with left ventricular systolic dysfunction. J Card
Fail 2010;7:556–561.
Charo DN, Ho M, Fajardo G, et al. Endogenous regulation of cardiovascular function by apelin-APJ. Am J Physiol Heart Circ Physiol 2009;297:H1904–
H1913.
Diaz-Cabiale Z, Parrado C, Rivera A, et al. Galanin-neuropeptide Y (NPY) interactions in central cardiovascular control: involvement of the NPY Y
receptor subtype. Eur J Neurosci 2006;24:499–508.
Isgaard J, Barlind A, Johansson I. Cardiovascular effects of ghrelin and growth hormone secretagogues. Cardiovasc Hematol Disord Drug Targets
2008;8:133–137.
Jacques D, Abdel-Samad D. Neuropeptide Y (NPY) and NPY receptors in the cardiovascular system: implication in the regulation of intracellular
calcium. Can J Physiol Pharmacol 2007;85:43–53.
Koh KK, Park SM, Quon MJ. Leptin and cardiovascular disease: response to therapeutic interventions. Circulation 2008;117;3238–3249.
Korbonits M, Goldstone AP, Gueorguiev M, et al. Ghrelin—a hormone with multiple functions. Frontiers Neuroendocrinol 2001;25:27–68.
Luo JD, Zhang GS, Chen MS. Leptin and cardiovascular diseases. Drug News Perspect 2005;18:427–431.
P.220
Matsumura K, Tsuchihashi T, Fujii K, et al. Neural regulation of blood pressure by leptins and the related peptides. Reg Peptides 2003;114:79–86.
Nagaya N, Kanagawa K. Therapeutic potential of ghrelin in the treatment of heart failure. Drugs 2006;66: 439–448.
Protas L, Qu J, Robinson RB. Neuropeptide Y: neurotransmitter or trophic factor in the heart. News Physiol Sci 2003;18:181–185.
Pusztai P, Sarman B, Ruzicska E, et al. Ghrelin: a new peptide regulating the neurohormonal system, energy homeostasis and glucose metabolism.
Diabetes Metab Res Rev 2008;24:343–352.
Quazi R, Palaniswamy C, Frishman WH. The emerging role of apelin in cardiovascular disease and health. Cardiol Rev 2009;17:283–286.
Nitric Oxide (NO)
Massion PB, Feron O, Dessy C, et al. Nitric oxide and cardiac function ten years after, and continuing. Circ Res 2003;93:388–398.
Saraiva R, Hare LM. Nitric oxide signaling in the cardiovascular system: implications for heart failure. Cur Opinion Cardiol 2006;21:221–228.
Zimmet JM, Hare JM. Nitroso-redox interactions in the cardiovascular system. Circulation 2006;114: 1531–1544.
Prostaglandins
Coleman RA, Smith WL, Naruyima S. International Union of Pharmacology Classification of prostanoid receptors: properties, distribution, and
structure of the receptors and their subtypes. Pharmacol Rev 1994;46:205–229.
Colombo PC, Banchs JE, Celaj S, et al. Endothelial cell activation in patients with decompensated heart failure. Circulation 2005;111:58–62.
Dogná JM, Hanson J, Pratico D. Thromboxane, prostacyclin and isoprostanes: therapeutic targets in atherosclerosis. Trends Pharm Sci 2005;26:639–
644.
Warner TD, Mitchell JA. Cyclooxygenases: new forms, new inhibitors, and lessons from the clinic. FASEB J 2004;18:790–804.
Wright DH, Abran D, Bhattacharya M, et al. Prostanoid receptors: ontogeny and implications in vascular physiology. Am J Physiol Regulatory
Integrative Comp Physiol 2001;281:R1343–R1360.
References
Ahlquist PR. A study of the adrenotropic receptors. Am J Physiol 1948;153:586–600.
Bayliss W, Starling E. The mechanism of pancreatic secretion. J Physiol (London) 1902;28:325–352.
Bernard C. Leáonssur les phánománes de la vie communs aux animaux et aux vegetaux. Paris: Balliáe, 1878.
Braun-Menendez E, Page IH. Suggested revision of nomenclature—angiotensin. Science 1958;127:242.
Braun-Menendez E, Fasciolo E, Leloir JC, et al. The substance causing renal hypertension. J Physiol (Lond) 1940;98:283–298.
Campbell DJ, Zeitz CJ, Esler MD, et al. Evidence against a major role for angiotensin converting enzyme-related carboxypeptidase (ACE2) in
angiotensin peptide metabolism in the human coronary circulation. J Hypertens 2004;22:1971–1976.
Cannon WB. The Wisdom of the Body. New York, NY: WW Norton, 1932.
Cannon WB, Rosenblueth A. Autonomic neuro-effector systems. New York, NY: MacMillan, 1937.
Clapham DE, Neer EJ. G protein bg subunits. Ann Rev Pharmacol 1997;37:167–203.
P.221
DeBold AJ. Heart atria granularity. Effects of changes in water-electrolyte balance. Proc Soc Exp Biol Med 1979;161:508–511.
Eichhorn EJ. Beta-blocker withdrawal: the song of Orpheus. Am Heart J 1999;138:387–389.
Elliott TR. The action of adrenalin. J Physiol (London) 1905;32:401–467.
Francis GS, McDonald KM. Neurohumoral mechanisms in heart failure. In: McCall D, Rahimtoola SH, eds. Heart failure. New York, NY: Chapman &
Hall, 1995:91–116.
Furchgott RF, Zawadzki JV. The obligatory role of endothelial cells in the relaxation of arterial smooth muscle by acetylcholine. Nature 1980;288:373–
376.
Harris P. Evolution and the cardiac patient. Cardiovasc Res 1983;17:313–319, 373–378, 437–445.
Herbette LG, Mason RP. Techniques for determining membrane and drug-membrane structures: a reevaluation of the molecular and kinetic basis for
the binding of lipid-soluble drugs to their receptors in the heart and brain. In: Fozzard H, Haber E, Katz A, et al., eds. The heart and circulation. 2nd
ed. New York, NY: Raven Press, 1991:417–462.
Herbette LG, Trumbore M, Chester DW, et al. Possible molecular basis for the pharmacodynamics of three membrane active drugs: propranolol,
nimodipine, amiodarone. J Mol Cell Cardiol 1988;20:373–378.
Huber R. Treasury of fantastic and mythological creatures. New York, NY: Dover, 1981.
Laugwitz KL, Allgeier A, Offermanns S, et al. The human thyrotropin receptor: a heptahelical receptor capable of stimulating members of all four G-
protein families. Proc Acad Sci (USA) 1996;93:116–120.
Loewi O. áberhumoraleábertragbarkeitderHerznervenwirkung. I. Mitteilung. Pflágers Arch Ges Physiol 1921;189:239–242.
Luttrell LM. Transmembrane signaling by G protein-coupled receptors. Methods Mol Biol 2006;332:3–49.
Luttrell LM, Ferguson SSG, Daaka Y, et al. β-arrestin-dependent formation of β2 adrenergic receptor-src protein kinase complexes. Science
1999;283:655–661.
Oliver G, Scháfer EA. The physiological effects of extracts of the suprarenal capsules. J Physiol (London) 1895;18:230–276.
Page IH, Helmer OM. A crystalline pressor substance (angiotonin) resulting from the reaction between renin and renin activator. J Exp Med
1940;71:29–42.
von Euler US. The presence of a sympathomimetic substancein extracts of mammalian heart. J Physiol 1946;105:38–44.
> Table of Contents > Part Two - Signal Transduction and Regulation > Chapter 9 - Signal Transduction: Proliferative Signaling
Chapter 9
Signal Transduction: Proliferative Signaling
The signaling systems that regulate cell size, shape, and composition play an important role in the
pathophysiology of cardiac disease. These systems, along with those that control cell division, are
described as proliferative in this text to distinguish them from the functional systems that modify
interactions among preexisting structures. Functional and proliferative signaling was once thought to be
controlled by different, independently regulated mechanisms. However, as noted by Bourne (1995),
instead of “a few discrete clans” of signaling molecules organized in distinct functional and proliferative
pathways, control of these two types of response resembles a series of “wheels within wheels! … bustling
communication networks within and between clans of signaling proteins ….” For example, the G protein-
coupled receptors initially identified as functional regulators (see Chapter 8) are now known to modify
proliferative responses once thought to be under the exclusive control of receptor tyrosine kinases.
Conversely, the receptor tyrosine kinases initially viewed as mediating only proliferative responses have
been found to influence such functional variables as myocardial contractility and vascular tone. The
clinical importance of this crosstalk is enormous. It helps explain why β-adrenergic blockers, whose
negative inotropic effects have obvious short-term deleterious effects in systolic heart failure, improve
long-term outcome (see Chapter 18), and why direct-acting vasodilators, even though they reduce
afterload, often worsen prognosis in these patients. In accord with the principle that what is obvious is
not always important, and what is important is not always obvious, formerly overlooked proliferative
effects of the neurohumoral response (see Chapter 8) can be of greater clinical significance than the more
obvious functional effects.
Can Cardiac Myocytes Divide?

Unlike the connective tissue cells of adult mammalian hearts, which undergo rapid proliferation, cardiac
myocytes rarely divide, and synthesis of DNA and protein is slow—the half-lives of myofibrillar proteins,
for example, can be several days. However, these myocytes have not entirely lost their ability to
proliferate. In a classical review, Rumyantsev (1977) lists histological studies going back to the 1880s
which show that when severely stressed, cardiac myocytes have a limited capacity for mitosis (Fig. 9-1).
These findings have been confirmed in recent studies using the tools of modern molecular biology (Nadal-
Ginard et al., 2003). It remains clear, however, that this limited capacity for cell division cannot
regenerate significant amounts of functioning myocardium.
Although adult mammalian hearts cannot adapt to sustained overload or myocyte death by increasing
myocyte number (hyperplasia), cardiac myocytes readily enlarge (hypertrophy). This hypertrophic
response is well suited for the heart, which cannot suspend its pumping to
P.223
incorporate newly generated myocytes into the intricate architecture of its walls (see Chapter 1). As
noted by Goss (1966): “By giving up the potential for hyperplasia in favor of the necessity for constant
function, [the heart … has] adapted a strategy that enables [it] to become hypertrophic to a limited
extent while doing [its jobs] efficiently.” However, a “price” must be paid; because cardiac myocytes
have virtually no ability to divide, patients with damaged or overloaded hearts must depend on myocyte
hypertrophy to develop an adequate wall stress. Unfortunately, the proliferative responses that allow the
adult heart to hypertrophy both damage these cells and shorten cardiac myocyte survival (see Chapter
18).
Fig. 9-1: Histology at the edge of experimental myocardial infarctions produced by coronary artery
ligation in rabbit hearts. A: 5 days after ligation showing the infarcted area (right) and surviving
myocytes whose ends are slightly expanded (left). B: 16 days after ligation showing fibroblast
nuclei surrounding the ends of the surviving myocytes, one of whose nuclei is vesicular (arrow). C:
16 days after ligation showing a surviving myocyte undergoing mitosis (arrow). (Modified from Ring,
1950).
P.224
The Cell Cycle

Cells lose their ability to divide when they withdraw from the cell cycle. The latter is a highly regulated
process in which cell enlargement, DNA replication, and nuclear division (karyokinesis) are followed by
cell division (cytokinesis) that produces two daughter cells, each of which contains a set of genes copied
from the mother cell. The cell cycle plays a critical role in the survival of simple life forms, called
prokaryotes (the name comes from the absence of a formed nucleus) that lack the internal membrane
structures needed to control their milieu intárieur. Instead, prokaryotes can survive environmental
change because their ability to divide rapidly allows them grow their way out of trouble. Because each
prokaryote represents an independent entity with its own complement of genes, when a population of
these organisms encounters a major environmental change, selection of cells whose phenotype is best
able to withstand the new stress helps some of the population to survive. Even if a vast majority dies,
rapid cell cycling allows a few survivors to multiply and fill the new environment. This is evident in
bacteria, which are modern prokaryotes that can divide as rapidly as every 20 minutes. This allows a
single individual to generate more than 4 billion descendants—approximately the human population of this
planet—within 10 hours.
Cell cycling and DNA replication in eukaryotes (eukaryote = discrete nucleus), unlike that in prokaryotes,
occur in spurts. The pauses between mitotic events allow the newly formed daughter cells to carry out
their biological functions, replicate their DNA, and rebuild and reorganize their internal structure. The
rate of cell cycling in eukaryotic cells varies greatly, pauses average 10 to 20 hours, but can be much
longer. In adult human cardiac myocytes, most of which have lost their ability to divide, the pause can
last until the host dies.
Nuclear division and cell division are usually tightly coordinated in cells that undergo mitosis, so that the
cell cycle generates daughter cells that contain a single nucleus. In some cells, karyokinesis can occur
without cytokinesis, so that the cell cycle generates a multinucleate cell. In other cells, DNA replication
without either cytokinesis or karyokinesis generates polyploid cells whose nuclei contain more than two
sets of chromosomes. In the normal adult human heart, ∼25% of cardiac myocytes are diploid, and more
than half are tetraploid; the degree of ploidy increases with aging and when the heart hypertrophies
(Rumyantsev, 1977).
Phases of the Cell Cycle

The cell cycle can be divided into four phases; in two, the cycle is active, in the other two it is relatively
quiescent. The two active phases are the S phase, characterized by DNA synthesis and chromosome
duplication, and the M phase, in which the duplicated chromosomes become separated and the cell
divides (Fig. 9-2). These active phases are separated by two phases in which most processes related to
cell division are temporarily suspended to allow cells to enlarge and carry out their physiological
functions; these are the G1 phase, between M and S, and the G2 phase, between S and M. The G1
restriction point, which occurs late in G1, represents a major pause in cell cycling (Pardee, 1989). Once
this restriction point is passed the cell is committed to dividing, which means that the cell cycle, while
still governed by internal regulatory mechanisms, has become independent of external stimuli.
P.225
Fig. 9-2: Schematic depiction of the cell cycle and the transition to a terminally differentiated cell. Left:
The cell cycle, such as occurs in proliferating fetal cardiac myocytes, proceeds in a clockwise direction.
Mitosis (M) is followed by a “gap” phase (G1) during which cells enlarge and carry out their physiological
functions; from the standpoint of the cell cycle this phase is quiescent. Cells become “committed” to the cell
cycle when they pass a restriction point in G1, after which the cell cycle proceeds though a phase of DNA
synthesis (S), a second gap phase (G2), and cell division (M). Cells are most susceptible to programmed cell
death (apoptosis) late in the G1 phase and during the S phase. Right: Myogenic determinants cause cardiac
myocytes to withdraw from the cell cycle and enter a prolonged quiescent phase of terminal differentiation,
called G0.
In terminally differentiated cardiac myocytes, the cells are in a G1 phase that resists stimuli which would
otherwise activate cell cycling. This quiescent state, called G0 in this text, can be viewed as a “detour”
out of the cell cycle. The term G0 is not strictly correct in describing this state in cardiac myocytes
because G0 was initially coined to describe cell cycle arrest associated with low metabolic activity in
serum-depleted cultured cells. A more accurate characterization of adult cardiac myocytes might be
“permanent arrest in the G1 phase,” but with apologies to experts, this text uses the simpler term G0.
Adult cardiac myocytes have little or no ability to reenter the cell cycle, even when exposed to stimuli
that would activate mitosis in less differentiated cells. As noted above, a few severely stressed cardiac
myocytes can reverse the transition from G1 to G0, and so reenter the cell cycle, but these proliferative
responses have virtually no ability to regenerate functioning cardiac myocytes.
Although proliferative signaling in mammalian hearts cannot increase cell number (hyperplasia), it does
cause the myocytes to enlarge (hypertrophy). However, these efforts turn out badly because, in spite of
increasing sarcomere number, hypertrophy of diseased or overloaded hearts generally has deleterious
consequences that hasten progression of hemodynamic abnormalities, accelerate myocardial cell death,
and shorten life expectancy (see Chapter 18).
P.226
Fig. 9-3: Cyclin-dependent protein kinases (CDKs) regulate the cell cycle when they interact specifically with
regulatory proteins called cyclins to form cyclin/CDK pairs. Most cyclins activate CDK activity only partially;
full activation occurs when the cyclin-bound CDKs are phosphorylated by other protein kinases called CDK-
activating kinases (CAK). Phosphorylated cyclin/CDK pairs are inactivated by cyclin-dependent kinase
inhibitors (CDKIs). Additional regulators (not shown) include phosphatases that dephosphorylate the CDKs.
Cyclins, CDK, and Related Proteins

The cell cycle is controlled by heterodimeric serine/threonine kinases that include catalytic subunits
called cyclin-dependent kinases (CDKs) and regulatory subunits called cyclins that determine the
substrate specificity of the CDK (Fig. 9-3). Cyclin/CDK pairs regulate transitions within and between the
phases of the cell cycle shown in Figure 9-2; some stimulate cell growth in G1, others promote DNA
replication in S, induce cell division in M, and stimulate transitions from G2 to M and from G1 to S. The
cyclin D/CDK4 pair allows the cell cycle to pass the G1 restriction point by releasing the transcription
factor E2F from an inhibitory complex with a hypophosphorylated retinoblastoma protein (see below).
Cyclin/CDK pairs also regulate the activity of a variety of transcription factors.
CDKs are enzymatically inactive unless bound to an appropriate cyclin, but cyclins generally cause only
partial activation of their corresponding CDKs. Complete activation often requires the participation of
additional protein kinases, called CDK-activating kinases (CAK), that phosphorylate the cyclin/CDK pairs,
while cyclin-dependent kinase inhibitors (CDKIs) slow the cell cycle by inhibiting active cyclin/CDK pairs
(Fig. 9-3). Additional regulatory proteins, not shown in Figure 9-3, include phosphatases that inhibit cell
cycling by dephosphorylating CDKs, and protein kinases that inactivate cyclin/CDK pairs by catalyzing
inhibitory phosphorylations.
Retinoblastoma Proteins and E2F

Cell cycle progression is regulated by proteins whose prototype is the retinoblastoma protein (pRB); these
proteins are sometimes referred to as tumor suppressors because they are substrates for phosphorylations
that regulate cell differentiation and inhibit tumor formation, or pocket proteins because their structure
includes a pocket that can bind transcription factors called E2F. Retinal cells that lack both copies of the
pRB gene undergo malignant transformation and can form metastasizing tumors. Some retinoblastoma
proteins are substrates for cyclin/CDK-mediated phosphorylations that regulate the cell cycle. In the
heart, where the most important pRB is p107, the pRB and E2F regulate DNA repair and apoptosis, and can
activate or inhibit gene expression.
Fig. 9-4: Function of retinoblastoma proteins (pRBs). A: In the hypophosphorylated state, pRBs bind and
inactivate transcription factors such as E2F. B: Phosphorylation of serine and threonine residues reduces the
affinity of pRBs for these transcription factors, which are released in an active form that regulates
proliferative signaling.
In hypophosphorylated pRB, the pocket binds to and inactivates E2F, which arrests the cell cycle in the G1
phase. Hyperphosphorylation of serine and threonine residues causes pRB to activate E2F by “kicking” this
transcription factor out of the pocket (Fig. 9-4). E2Fs can also be inactivated by phosphorylation or direct
binding to the cyclin A/CDK2 pair (Fig. 9-5). Together, pRB and E2F provide a “switch” that controls
passage of the cell cycle through the G1 restriction point.
A variety of signaling systems in addition to CDK/cyclin pairs regulate proliferative signaling by activating
E2F. These include receptor tyrosine kinases, steroid receptors, G protein-coupled receptors, and
cytoskeletal deformation. E2F activity can also be regulated by changes in the rate at which this
transcription factor is synthesized, which allows increased levels of the messenger RNA (mRNA) that
encodes E2F to stimulate proliferative signaling.
Genomic Regulation of Phenotype

Genomic mechanisms, which are among the most highly regulated of all biological functions, depend on
the base pair sequence in chromosomal DNA to control phenotype by directing gene expression (Table 9-
1). In normal hearts, most genes are down-regulated or dormant unless their expression is turned on by
transcription factors. Once activated, genomic DNA serves as the template for the formation of primary
RNA transcripts whose sequences are transcribed in the nucleus to form mRNA; the latter is then
transported to the cytosol where it serves as a template for the amino acid sequences of proteins.
Rna Polymerases and Transcription Factors

Activation of gene expression begins when specific DNA sequences called promoters bind to multiprotein
enzyme aggregates called RNA polymerases (Fig. 9-6). The latter operate in concert with general
transcription factors that, along with the RNA polymerase, bind to the promoter.
P.227
DNA helicases that are included in the RNA polymerase then unwind a short segment of the two strands of
base pairs, which allows the downstream DNA to serve as a template for the primary RNA transcript. The
latter, which is synthesized by the RNA polymerase as it moves downstream along the gene, contains the
information encoded in the DNA. Processing of the transcript then forms mRNA, which is transported to
the cytosol where it determines the amino acid sequence in newly synthesized proteins.
Fig. 9-5: E2F can be inactivated when it forms a complex with the cyclin/CDK pair cyclin A/CDK2
(left), by CDK-catalyzed phosphorylation (center), and when it binds to a hypophosphorylated
retinoblastoma protein (right). CDK, cyclin-dependent kinase.
In addition to the nucleotide sequences that direct protein synthesis, genomic DNA contains sequences
that control the rate at which genes are transcribed, and turn genes on and off. These regulatory
sequences include promoter regions, which are located at the 5′-end of the gene, “upstream” from the
start site where DNA transcription begins, and enhancer and repressor regions that increase and decrease
the rate of gene expression, respectively (Fig. 9-6). One of the most important promoter regions is the
TATA box, so named because its DNA sequence includes thymidine (T) and adenine (A) in the sequence
TATA; this sequence is a weak point in the double-stranded DNA where the helix is most easily unwound
into the separate single strands used for copying. A remarkable discovery that was made possible by the
human genome project is that a large majority of the information encoded in genomic DNA provides
templates for small RNA segments, called miRNA, that control the epigenetic regulation of phenotype (see
below). Basic proteins, called histones, which bind to and stabilize genomic DNA in chromatin, also
participate in epigenetic regulation.
Transcription factors, most of which are proteins and phosphoproteins, turn genes on by binding to a
promoter region, accelerate gene expression by binding to an enhancer sequence, or slow gene
transcription by binding to a repressor sequence (see Fig. 9-6). The ability of transcription factors to
select a specific gene and then regulate its expression is made possible by structural motifs that provide
the tight “fit” needed to identify a specific target DNA sequence. Most operate
P.228
P.229
as homo- or heterodimers in helix-turn-helix, helix-loop-helix, leucine zipper, and zinc finger structures
(Fig. 9-7), where each subunit binds to one of the two strands of genomic DNA. The subunits in these
dimers are held together by non-covalent linkages that include hydrophobic interactions between leucine
residues and divalent zinc atoms. Gene expression can also be regulated by hormones, such as thyroxin
and aldosterone. These hydrophobic molecules cross the
P.230
plasma membrane and bind to intracellular nuclear receptors (also called hormone receptors); the
resulting hormone-receptor complexes then serve as transcription factors that regulate gene expression.
Cooperative interactions between genes and transcription factors can allow a single transcription factor
to generate a coordinated growth response by interacting sequentially with several functionally related
genes, and regulatory regions of single genes to modify the actions of several transcription factors.
Table 9-1 Regulation of Gene Expression and Protein Composition in the Heart
PROTEIN SYNTHESIS
GENOMIC MECHANISMS
DNA transcription
Selection of the DNA sequence to be read
Transcription of the selected DNA sequence as a complementary RNA sequence
RNA processing (alternative splicing)
Selection and arrangement of exons to be spliced to form messenger RNA (mRNA)
RNA export
Control of mRNA transport from the nucleus to the cytoplasm
Translation
Control of mRNA translation into protein
RNA degradation
Control of mRNA breakdown, which turns off protein synthesis
EPIGENETIC MECHANISMS
DNA modification
RNA modification
Chromosomal changes
PROTEIN ACTIVATION AND INACTIVATION (Posttranslational control)
Modifications that increase or decrease the biological activity of a protein, e.g.,

phosphorylation, acetylation
Interactions with other cell components, e.g., myosin-actin interactions, calcium binding
to troponin C
PROTEIN BREAKDOWN (Proteolysis)
Fig. 9-6: Gene transcription. Coding regions of double-stranded genomic DNA that lie downstream (toward
the 3′ end of the DNA sequence) from a “START” sequence provide templates for the primary RNA transcript
that encodes messenger RNA. Gene transcription begins when RNA polymerases actively bind to a regulatory
sequence located upstream from the 5′ end of the gene, called a promoter; the TATA box, which is made up
of thymidine (T) and adenine (A) residues, is among the most important of these promoters. Transcription can
be activated by enhancers and inhibited by repressors; the latter are found not only upstream from the start
site, as shown in the figure, but also downstream and within regulatory sequences in the coding region.
Fig. 9-7: Four types of transcription factor. These regulators, which function as dimers, include
helix-turn-helix (A) and helix-loop-helix structures (B), leucine zipper structures (C) in which the
two subunits of the dimer are linked by hydrophobic forces between leucine residues, and zinc
finger structures (D) where zinc atoms stabilize the structure of the transcription factor. Cylinders
represent α-helices.
Transcription factors can be activated in many ways. Most common is posttranslational modification of
the inactive protein, such as by phosphorylation or dephosphorylation. Transcription factor activity can
also be regulated when another proliferative signaling cascade increases or decreases the rate of its
synthesis, releases the active protein from an inactive complex, or incorporates the transcription factor
into aggregates that are often called scaffolds.
The DNA strands at the ends of each chromosome are capped by telomeres. These structures, which in
humans contain hundreds of copies of the tandem repeat TTAGGG along with members of a family of
protective proteins called shelterins, protect the chromosomes from enzymes that normally degrade
broken DNA strands. The number of tandem repeats in the telomere
P.231
decreases with normal aging, and DNA damage associated with telomere shortening has been suggested to
contribute to the pathogenesis of heart failure.
Fig. 9-8: Alternative splicing. During processing of the primary RNA transcript, the sequences in exons are
encoded in proteins, while those in introns are removed prior to synthesis of messenger RNA (mRNA).
Selection and rearrangement of exons allows a single gene to encode several different mRNAs. In this
example, three exons in the primary transcript form four mRNAs by alternative splicing. These are only a few
of the possible mRNAs in which one, two or all three exons can be combined in various sequences.
Rna Processing and Alternative Splicing

Some segments of the primary transcript, called introns, are eliminated before the remaining nucleotide
sequences, called exons, are assembled into the mRNA that directs protein synthesis. Elimination of
introns and rearrangement of selected exons, called alternative splicing, is an important mechanism that
provides diversity in proliferative signaling. Figure 9-8 shows how a gene containing three exons can
encode four different mRNAs; this illustration is by no means complete as the three exons can be used in
additional combinations. Some genes contain dozens of exons, so that alternative splicing can generate
dozens of protein isoforms from a single DNA sequence.
Epigenetic Regulation
In addition to the genomic mechanisms described above, which select specific genes for expression or
suppression and allow alternative splicing to modify gene transcripts, a recently discovered group of
regulatory mechanisms regulates proliferative signaling. The growing impact of these mechanisms, called
epigenetics, is reminiscent of the late 1950s, when as a research fellow in C. B. Anfinsen's laboratory at
the National Institutes of Health in Bethesda Maryland, I heard Francis Crick describe the beginning of the
DNA revolution (the double helix, genetic code, etc.). Today's emergence of epigenetics promises another
biological revolution that is likely to be even more complex than genetics.
Defined most broadly, epigenetics refers to non-Mendelian mechanisms of inheritance that influence
phenotype without depending directly on the sequence of the base pairs in DNA. Genomics and
epigenetics can be contrasted by viewing the former as using information encoded in genomic
P.232
DNA to determine how cells, organs, and overall body plan can appear, whereas epigenetic mechanisms
determine how this genetic information is used to determine what does appear.
Among the more remarkable features of epigenetics is that regulation by RNA can explain why, although
more than 93% of the DNA in the human genome is transcribed, less than 2% of the RNA transcript encodes
proteins. This, of course, raises the question: “What happens to the non-protein-coding RNA transcript?”
The answer is that many of these RNA transcripts regulate gene expression. It is amazing, although there
is ample precedent in many other fields of science, that for a half century the importance of RNA in
regulating gene expression was overlooked almost completely.
An unexpected feature of epigenetic regulation is that it can allow environmental factors, such as
nutritional state, temperature, toxins, and even maternal nurturing, to induce changes in phenotype that
can be passed down to successive generations (Jirtle and Skinner, 2007). The ability of epigenetic changes
to modify the germ line marks a return toward the 19th century Lamarckian view that environment can
modify inherited traits.
In cardiology, epigenetic mechanisms help regulate impulse conduction, contribute to molecular changes
in failing hearts, cause several inherited cardiomyopathies, and influence the clinical manifestations of
specific gene mutations. The influence of epigenetic mechanisms on phenotypic expression also explains
variations in the severity of the clinical syndromes associated with specific genomic mutations, often
called penetrance (see, for example, Zhou et al., 2006), and how a single genotype can give rise to
different phenotypes (Fraga et al., 2005).
Three types of mechanisms participate in epigenetic control (Table 9-2). The first includes changes in DNA
brought about by mobile genetic elements, covalent modification of genomic
P.233
DNA by cytosine methylation, and interactions between trans-gene alleles. In the second, small sequences
of RNA, called RNAi, inhibit (“silence”) RNA translation; the third, which modifies chromosome and
chromatin structure, includes posttranslational covalent modifications of histone that alter the ability of
genomic DNA to interact with transcriptional regulators.
Table 9-2 Some Epigenetic Mechanisms
Modification of DNA
Modifications by mobile genetic elements (transposons)
Communications between trans-gene alleles and homologous DNA sequences

(paramutations)
Methylation and demethylation
RNA signaling
RNA interference (RNAi)

Chromosomal changes
Histone modifications
Acetylation, methylation, phosphorylation, ubiquitination, ADP-ribosylation, etc., of ε-

amino lysyl groups
Modifications of chromosomes and chromatin
Changes in nuclear conformation
Gene compartmentalization and interactions among chromatin loops
Organization of chromosomes at various sites within the nuclei
Localization of chromosomes within and at the periphery of the nucleus
Altered positioning of chromosomes during anaphase
Modifications of chromatin structure
Dna Modification
More than 50 years ago, McClintock (1950) found that a given gene could appear at various loci in the
maize genome. This genetic instability, which was initially referred to as “jumping genes,” reflects the
movement of DNA sequences between non-homologous regions of the genomic DNA. The mobile DNA
segments, called tranposons, range in size from hundreds to tens of thousands of base pairs. A related
epigenetic mechanism, called parental or genomic imprinting or parent-of-origin regulation, can
inactivate large segments of either X or Y chromosomes.
DNA methylation, which generally “locks” genes in a silent state, is among the most ancient of the
epigenetic mechanisms. It probably appeared as a defense mechanism that allowed unwilling hosts to
eliminate invading parasites by methylating, and so inactivating, parasite genomes. In eukaryotes, DNA
methyl transferases, which catalyze the addition of methyl groups to cytosine, inhibit gene transcription
and apoptosis. Demethylation, which is catalyzed by DNA demethylases, activates proliferative signaling,
while methyl-CpG binding proteins identify and initiate the methylation of specific DNA sequences.
Cytosine methylation has been linked to destabilization of dinucleotide 5′-CpG-3′ (CpG) tandem repeats
and contributes to the pathogenesis of several human diseases (Robertson, 2005; Rodenhiser and Mann,
2007).
Interactions between trans-regions of paired or homologous DNA sequences can silence gene expression by
inhibiting the transcription and/or translation of specific genes. These responses, called paramutation,
are regulated by such epigenetic mechanisms as DNA methylation and RNA inhibition.
Rna Modifications
RNA interference (RNAi), which is emerging as one of the most important epigenetic mechanisms that
determine phenotype, occurs when a short (19 to 25 nucleotide), single-stranded antisense RNA sequence,
called a miRNA, modifies gene expression. At the end of 2009, more than 700 different human miRNAs
were mentioned in the Sanger miRBase (http://www. mirbase.org/). miRNAs regulate many functions in
the heart such as myogenesis, cardiac development, and proliferative responses that include maladaptive
features of the heart's response to hemodynamic overload. miRNAs have been found to be upregulated
and downregulated in failing hearts, and a distinct abnormal miRNA expression pattern has been reported
in human-dilated cardiomyopathies (Naga Prasad et al., 2009).
Synthesis of miRNAs (Fig. 9-9) begins when pri-RNA, one of the primary transcripts that do not encode
proteins (see above), is cropped in the nucleus by an endonuclease called drosha. This forms ∼70
nucleotide pre-miRNAs that are transported to the cytoplasm where they are cleaved by another
endonuclease called dicer. This releases a shortened double-stranded RNA (RNA duplex) that aggregates
with several proteins to form an RNA-induced silencing complex (RISC). One of the proteins that make up
the RISC, called argonaute, hydrolyzes the double-stranded RNA duplex to liberate the single-stranded
miRNAs that bind with high specificity to mRNA sequences. miRNAs can inhibit the synthesis of specific
proteins, catalyze the degradation of other mRNA sequences, and even reenter the nucleus where they
can serve as transcription factors. Artificial miRNAs called small interfering (si)RNAs, which can be
synthesized in vitro, can also silence specific genes.
P.234
Fig. 9-9: microRNA synthesis. A: MicroRNA precursors, called pri-miRNA, are transcribed from
genomic DNA in the nucleus. B: Drosha, an endonuclease, cleaves ∼70 nucleotide double-stranded
pre-MiRNA from the pri-miRNA. C: The pre-miRNA is processed by dicer to form a duplex that
contains a shortened RNA, one strand of which will become the miRNA (dashed line). D: The
double-stranded pre-miRNA forms RNA-induced silencing complexes (RISC) with several regulatory
proteins that include argonaute, which releases the short, ∼22 nucleotide, singe-stranded miRNA.
E: Binding of RISCs to ribosomal RNA sequences (target mRNA) allows the miRNA to silence (inhibit)
translation of the latter. mRNA, messenger RNA.
Chromosomal Changes
Chromosomes are stabilized in structures called nucleosomes in which the two strands of DNA are wound
tightly around histone cores, which causes the genomic DNA to become inaccessible to transcription
factors. Acetylation, methylation, phosphorylation, ubiquitination, and ribosylation of the á-amino lysyl
groups in histone can activate gene transcription by unwinding the nucleosomes. Histone acetylation by
histone acetyltransferases (HATs) stimulates proliferative signaling by exposing active sites on the DNA
(Fig. 9-10). Conversely, removal of the acetyl groups by histone deacetylases (HDACs) reverses this
stimulatory effect by causing DNA to condense in chromatin where it becomes inaccessible to
transcriptional regulation. The activity of HATs and HDACs is controlled by a number of signaling cascades.
For example, phosphorylation of some HDACs by calcium-calmodulin (CAM) kinases and protein kinase C
(PKC) increases their ability to acetylate histones, which inhibits responses to transcription factors that
lead to pathological hypertrophy.
Gene expression is also regulated by epigenetic mechanisms that alter the spatial organization of genes
and transcriptional regulators within the nucleus. This can occur when specific regions of the genomic
DNA are compartmentalized in specialized chromatin loops when chromosomes are arranged in various
degrees of proximity to one another at the periphery of the nucleus, and when chromosome positions
change during anaphase.
P.235
Fig. 9-10: Epigenetic regulation of DNA transcription by histone acetylation. A: Deacetylated

histone forms tightly packed nucleosomes in which DNA transcription sites (closed circles) cannot
interact with transcription factors. B: Acetylation of histone by histone acetyltransferases (HATs)
inhibits chromatin condensation by unwinding histone-bound DNA, which allows the DNA
transcription sites (open circles) to interact with transcription factors. The activated DNA
transcription sites are inactivated when histone is deacetylated by histone deacetylases (HDACs).
Extracellular Growth Factors
Many proliferative signaling mechanisms that operate within cells are controlled by extracellular
messengers. The importance of the latter became apparent in the middle of the 20th century, when adult
cells cultured in artificial media were found to lose their ability to grow and divide even after addition of
all known nutrients, vitamins, and trace elements. The key to these puzzling observations was provided
by the observation that inclusion of fetal calf serum in the artificial media allowed the cultured cells to
proliferate. Efforts to characterize the unknown factors needed to maintain cell cycling identified the
missing ingredients as peptides that came to be called growth factors. Several of these peptides, along
with cytokines and most of mediators of the functional responses described in Chapter 8, are now known
to modify proliferative signaling (Table 9-3).
Peptide Growth Factors

A large number of peptides stimulate proliferative responses. Although initially named for the tissues from
which they were first isolated, or whose growth they were initially found to stimulate, most peptide
growth factors turned out not to be tissue-specific, but instead are able to mediate proliferative signaling
in many cell types throughout the body. These include platelet-derived growth factor (PDGF), epidermal
growth factor (EGF), fibroblast growth factor (FGF), insulin-like growth factor (IGF), and vascular
endothelial growth factor (VEGF), along with transforming growth factors (TGF) and other cytokines
(Table 9-4). All use paracrine, autocrine, and endocrine pathways to reach receptors on their target cells,
most of which contain a latent tyrosine kinase that is activated when the receptors form aggregates with
their ligands. A few, like TGF-β receptors, have latent serine/threonine kinase activity. Cytokine
receptors, which lack an intrinsic protein kinase, activate latent tyrosine kinase activity in other
membrane-associated proteins with which they form aggregates.
Two fundamentally different mechanisms allow binding of a growth factor to its receptor to activate
intracellular signal transduction cascades (Fig. 9-11). In the first, the ligand-receptor
P.236
complex activates a protein kinase or other enzyme that then modifies an intracellular signaling molecule
(Fig. 9-11A). In the second, binding of the growth factor to its receptor initiates the formation of a
multiprotein aggregate containing a docking site (scaffold) that binds to additional intracellular signaling
molecules (Fig. 9-11B). Activation of proliferative signaling by internalized β2 receptors (see Chapter 8) is
one example of the latter mechanism; other examples are described below.
Table 9-3 Some Proliferative Signaling Systems That Operate in the Mammalian
Cardiovascular System
Coupling Second Messenger Intracellular

Extracellular Signal Receptors Proteins or Mediator System Targets
Mediated by G protein-coupled receptors
β-Adrenergic β-adrenergic G Adenylyl Protein

agonists receptors proteins cyclase, kinase A,
cyclic AMP calmodulin,
calcium calcineurin
α-Adrenergic α1-Adrenergic G Phospholipase Protein

agonists receptors proteins C, DAG, InsP3 kinase C
Angiotensin II AT1 receptors G Phospholipase Protein

proteins C, DAG, InsP3 kinase C
Endothelin ETA receptors G Phospholipase Protein

proteins C, DAG, InsP3 kinase C
Mediated by enzyme-linked receptors
Peptide Tyrosine kinase Ras, MAP

growth receptors Rac kinases
factors
Inflammatory Cytokine Rho STATs,

cytokines receptors, SAPKs MAP
gp130 kinases
Mediated by the cytoskeleton
Integrins, LIM Ras Tyrosine

proteins Titin, kinases,
Z-line proteins, MAP
etc. kinases
G protein, heterotrimeric G protein; SAPK, stress-activated protein kinase; MAP

kinase, mitogen-activated protein kinase; DAG, diacylglycerol; AMP, adenosine
monophosphate.
Fibroblast Growth Factor (FGF)

Fibroblast growth factors (FGFs) generally respond to mechanical stresses by stimulating cell growth and
proliferation, favoring expression of fetal genes, and inhibiting differentiation and apoptosis (Table 9-5).
In the heart, FGFs participate in the healing response to injury and the hypertrophic response to
overload, while in blood vessels these peptides modify responses to endothelial damage. FGFs, which are
often divided into acidic and basic peptides, activate FGF tyrosine kinase receptors that contain a latent
tyrosine kinase (Fig. 9-12). Ligand-binding to these receptors causes the latter to form aggregates with
extracellular FGF receptor heparan
P.237
sulfate proteoglycans, which are connective tissue glycoproteins whose glycosoaminoglycan side chains
link cells to one another. Aggregation of the receptors activates the latent tyrosine kinase which
autophosphorylates the receptor (Fig. 9-12B); the latter can then activate a variety of intracellular
signaling systems, including MAP kinases, phospholipase C, and other regulators of proliferative responses
(see below).
Table 9-4 Some Peptide Growth Factors
Tyrosine kinase ligands
Platelet-derived growth factor (PDGF)
Epidermal growth factor (EGF)
Fibroblast growth factor (FGF)
Insulin-like growth factor (IGF)
Vascular endothelial growth factor (VEGF)
Serine/threonine kinase ligands
Transforming growth factors (e.g., TGF-β)
Cytokines
Tumor necrosis factor-α
Interleukins 2–7, 9–13
Growth hormone
FAS ligand
Erythropoietin
Granulocyte colony-stimulating factor (CM-GSF)
Fig. 9-11: Mechanisms of signal transduction by receptor protein kinases. A: Phosphorylation of the ligand-
bound receptor activates a protein kinase or other enzyme that modifies an intracellular signaling molecule
(S). B: Phosphorylation of the ligand-bound receptor creates a docking site which forms an aggregate with a
group of intracellular signaling molecules (S1–S5) that activates a signal transduction pathway.
P.238
Table 9-5 Some Effects of FGF and TGF-β on the Heart
FGF
Stimulation of myocyte growth
Expression of the fetal gene program
Inhibition of differentiation and myogenesis
Inhibition of apoptosis
TGF-β
Stimulation of fibrosis
Expression of the fetal gene program
Stimulation of myocyte differentiation and myogenesis
Stimulation of apoptosis
In patients with heart failure, FGFs released by angiotensin II, endothelin, and other mediators of the
hemodynamic defense reaction activate gene programs that participate in the hypertrophic response to
overload. Interactions of FGF receptor heparan sulfate proteoglycans with the extracellular matrix also
allow cell deformation to initiate proliferative signals that modify cell size, shape, and composition.
Transforming Growth Factor β (TGF-β)

Transforming growth factor-β (TGF-β) plays a major role in tissue responses to injury by stimulating both
fibrosis and proliferative signaling pathways. Although TGF-β inhibits fibroblast proliferation, it increases
the synthesis of matrix proteins. This peptide stimulates myogenesis
P.239
and differentiation in embryonic hearts and, in overloaded adult hearts promotes progressive dilatation
(remodeling), expression of fetal genes, and apoptosis (Table 9-5). Smad3, one of the mediators of TGF-β
signaling (see below), plays a key role in the cardiac response to overload. TGF-β1, the major cardiac
isoform of this peptide, mediates responses to angiotensin II that include fibrosis and stimulation of
maladaptive hypertrophy (see Chapter 18).
Fig. 9-12: Fibroblast growth factor (FGF) signaling. A: When the plasma membrane receptor (FGF receptor
tyrosine kinase) is not bound to FGF, its latent tyrosine kinase activity is inactive. B: Ligand binding to the
FGF receptor causes the latter to bind to a heparan sulfate proteoglycan, which contains glycosoaminoglycan
side chains that are linked to the extracellular matrix. Formation of this aggregate activates the latent
tyrosine kinase in the receptor, which by autophosphorylating one or more tyrosine residues, modifies
downstream intracellular signal transduction cascades.
Fig. 9-13: Transforming growth factor (TGF)-β signaling. A: Two types of TGF-β receptor (I and II) contain
serine/threonine kinase sites that are inactive when the receptors are not bound to their ligand. B: Binding of
a TGF-β homodimer to two type II receptors activates their latent serine/threonine kinase, which
phosphorylates the type I receptors. This activates the latent protein kinase activity in the type I receptors
which, by phosphorylating one or more serine or threonine residues on the receptor, activates Smads and
other intracellular signaling molecules.
Two classes of TGF-β receptors, called type I and type II, mediate TGF-β signaling. Both contain
serine/threonine kinases, but only the type II receptors actually bind TGF-β (Fig. 9-13). Signaling begins
when a TGF-β homodimer binds two of the type II receptors to form a ligand-receptor complex that
activates the latent serine/threonine kinase in these receptors. The type II receptors then form a complex
in which they phosphorylate two type I receptors, which activates their latent protein kinase. The
activated type I receptors propagate the TGF-β signal down intracellular signaling cascades by catalyzing
serine/threonine phosphorylations in regulatory proteins that include transcription factors called Smads (a
conflation of the names of two members of this family: “Sma” in the roundworm Caenorhadbitis elegans
and “Mad” in Drosophila). Phosphorylated Smads, along with ubiquitin ligases (called Smurfs for “Smad
ubiquitination regulator factor”) and signaling proteins like TGF-β-activated kinase (TAK1) regulate MAP
kinase signaling (see below).
Cytokines
Cytokines, an ancient peptide family that includes a number of signaling peptides (Table 9-6), mediate a
variety of responses in cells throughout the body (Table 9-7). Like peptide growth factors, cytokines bind
to plasma membrane receptors; however, cytokine receptors do not contain protein kinase moieties, but
instead interact with other membrane proteins that contain
P.240
this enzyme. Some cytokine receptors contain an amino acid sequence, called a “death domain,” that
stimulates apoptosis (see below).
Table 9-6 Some Members of the Cytokine Family
Cardiotrophin-like cytokine (CT-1)
Erythropoietin
Granulocyte colony-stimulating factor
Growth hormone
Interferons
Interleukins 2–7, 9–13
Leukemia inhibitory factor (LIF)
Oncostatin M
Prolactin
Transforming growth factor-β
Tumor necrosis factor α (TNF-α)
Thrombopoietin
Cytokines have both deleterious and beneficial effects on the heart; in addition to injuring cells, these
inflammatory peptides can stimulate hypertrophy, promote cellular repair and healing, and mediate an
important adaptive hypertrophic response to overload (Uozumi et al., 2001). Elevated cytokine levels
cause a loss of body mass in many chronic diseases, and high circulating levels of tumor necrosis factor-α
(TNF-α or cachectin) and other cytokines contribute to cardiac cachexia, a complication of severe heart
failure characterized by loss of weight and wasting.
TNF-α is not expressed by normal adult mammalian cardiac myocytes, but synthesis of this and other
cytokines can be stimulated when these cells are stressed or overloaded. Overload also causes cardiac
myocytes to increase myocardial cytokine levels by releasing chemotactic and other factors that attract
cytokine-producing monocytes to the heart.
Cytokine signaling begins when these peptides bind to plasma membrane cytokine receptors that are
specific for individual cytokines; however, a single class of cytokines can bind to more than one type of
receptor. Signal transduction begins when the ligand-bound receptors form aggregates that can include
additional coupling proteins called gp130 (Fig. 9-14). Most responses to cytokines are mediated by
phosphorylation of tyrosine, and less frequently serine and threonine; substrates include the cytokine
receptors, gp130, and protein kinases that can interact with both cytokine receptors and gp130.
The JAK/STAT pathway, which is among the most important mediators of cytokine signaling in the heart, is
activated when ligand-bound cytokine receptors activate intracellular tyrosine kinases called JAKs (this
acronym, which originally stood for “just another kinase,” has been redefined to mean Janus kinase
because like Janus—the two-faced Roman god of doorways who looks both outward and inward—JAKs
respond to binding of cytokines outside the cell by phosphorylating proteins within the cell).
JAK phosphorylates the activated receptors, which creates docking sites that form aggregates with a
transcription factor called signal transducer and activator of transcription (STAT). STATs are also
phosphorylated by JAK, after which the phosphorylated STATs dissociate from the
P.241
receptor complex and move to the nucleus where they regulate transcription. The transcriptional activity
of phosphorylated STATs is regulated by SOCS (suppressors of cytokine signaling) and PIAS (protein
inhibitor of STAT) that inhibit the binding of activated STAT to DNA, and SHP-2 (src homology 2) which
dephosphorylates the activated cytokine receptors. STAT-mediated signaling is activated by p300/CREB-
binding protein and CR-6-interacting factor 1 (Crif1), and are turned off when nuclear phosphatases
dephosphorylate the STATs.
Table 9-7 Selected Actions of the Cytokines
Cellular effects
Inflammation
Cell proliferation
Cell transformation
Apoptosis (programmed cell death)
Signal transduction
Activate tyrosine kinases, e.g., Janus kinase (JAK)
Activate protein kinases A and C
Activate stress-activated MAP kinases
Activate phospholipases A2 and C
Increase levels of cyclic AMP and diacylglycerol
Activate signal transducer and activator of transcription (STAT)
Activate NF-κB
Activate immediate-early genes
Induce synthesis of
Inflammatory mediators including other cytokines
Inducible nitric oxide synthase (iNOS)
Growth factors, e.g., PDGF, GM-CSF
Receptors, e.g., EGF receptor, interleukin-2 receptor
Cytoskeletal molecules
Heat shock proteins
Cytokine signaling is also mediated by stress-activated mitogen-activated protein (MAP) kinases (see
below) and signaling cascades that activate a transcription factor called NF-κB (Fig. 9-15). In this
pathway, cytokines bind to a receptor whose cytoplasmic domain contains an amino acid sequence called
a “death domain”; this forms an aggregate with the adaptor protein TNF-associated death-domain protein
(TRADD), TNF-receptor associated factor 2 (TRAF2), and a receptor-interacting protein kinase (RIP). The
latter phosphorylates IκB kinase kinase (IKKK) which phosphorylates IκB kinase (IKK), which
phosphorylates IκB that is in a complex that inactivates the transcription factor NF-κB. Transfer of
ubiquitin (Ub) to IκB by a ubiquitin ligase dissociates the ubiquitinated IκB from NF-κ; this activates NF-
κB and allows the ubiquitinated IκB to be degraded by a proteasome. Translocation of activated NF-κB to
the nucleus allows this transcription factor to regulate gene expression. These cytokine signaling
pathways can be “turned off” by inhibitors of cytokine-induced phosphorylation and phosphatases that
dephosphorylate the activated signaling proteins.
P.242
Fig. 9-14: Cytokine signaling: the JAK-STAT pathway. A: Components include the cytokine receptors,
the coupling protein gp130, tyrosine kinases called JAK, and the transcription factor STAT. JAK/STAT
signaling can be activated when ligand-binding causes cytokine receptors to form aggregates (B) that
may include gp130 (C). Activated cytokine receptors stimulate JAKs to phosphorylate the STAT
transcription factors (1), the receptor (2), gp130 (3), and/or itself (autophosphorylation, 4). These
phosphorylations are controlled by SOCS, which inhibit phosphorylation of both the receptor and JAK;
SHP-2, which inhibits phosphorylation of the receptor, JAK, and STAT; and PIAS, which inhibits binding
of the activated STAT to DNA.
P.243
Fig. 9-15: Cytokine signaling. The NF-κB pathway. Binding of TNF-α and other cytokines to their
receptors (top) forms aggregates with TNF-associated death-domain protein (TRADD), TNF-
receptor-associated factor 2 (TRAF2), and receptor-interacting protein kinase (RIP) that can
phosphorylate IκB kinase kinase (IKKK). The latter phosphorylates IκB kinase (IKK), which then
phosphorylates IκB in a complex with the inactive transcription factor NF-κB. A ubiquitin ligase
transfers ubiquitin (Ub) to IκB which dissociates from NF-κB; this activates NF-κB and allows IκB to
be degraded by proteasomes. The activated NF-κB then moves to the nucleus where it regulates
gene expression.
P.244
Cytokine receptors can be found in a soluble form whose circulating levels, although low in normal
individuals, are elevated in the blood of patients with infectious and autoimmune diseases and some
patients with heart failure. Soluble cytokine receptors can be produced when limited proteolysis of the
membrane-bound receptors releases the extracellular ligand-binding domain of the molecule in a process
called shedding; soluble receptors can also be synthesized de novo from mRNAs that encode only the
ligand-binding portion of the receptor molecule. Complexes formed by soluble cytokine receptors and
their ligands can inhibit cytokine actions; they can also serve regulatory functions similar to those of the
membrane-bound receptor and protect the bound cytokines from proteolysis.
Heterotrimeric G Protein-Coupled Receptor-Activated Pathways

Most of the G protein-coupled receptors that mediate the functional responses described in Chapter 8 also
activate proliferative signaling; these include the receptors that bind β-adrenergic agonists, angiotensin
II, vasopressin, endothelin, and sphingosine-1-phosphate. β-Adrenergic receptor agonists, for example,
activate a number of proliferative signaling pathways when Gαs stimulates cyclic AMP-dependent protein
kinases that phosphorylate transcriptional regulators such as CREB (cyclic AMP receptor element-binding
protein) and CREM (cyclic AMP receptor element modulator), and Gαq stimulates phospholipase C to
release InsP3 and DAG (Fig. 9-16). InsP3-induced calcium release from internal stores activates both CAM
kinases and calcineurin (see below). Other G protein-coupled receptor agonists, including angiotensin II,
endothelin, and vasopressin, also activate Gαq. Binding of sphingosine-1-phosphate to G protein-coupled
receptors releases Gβγ that activates Akt-mediated proliferative signals. Some G protein-activated
receptor agonists stimulate proteases, called sheddases, which activate receptor tyrosine kinases.
Both proliferative and functional signaling cascades use G proteins to activate downstream responses
(Table 9-3). The downstream signal transduction cascades activated by enzyme-linked and G protein-
linked receptors can also be similar; for example, both types of receptors can activate MAP kinases (see
below), and some ligands—like angiotensin II and α-adrenergic agonists—can activate both G protein-
linked and enzyme-linked receptors by a process called transactivation.
Monomeric G Proteins
Most proliferative signal transduction cascades are coupled by monomeric GTP-binding proteins that serve
as molecular switches and timers. These small proteins, whose molecular weights range between 20 and
25 kDa, include subfamilies called Ras, Rho/Rac/cdc42, Rab, Arf, Sar1, and Ran. More than 150
monomeric G proteins are expressed in humans. Like Gα, their heterotrimeric relative (see Chapter 8),
monomeric G proteins are active when bound to GTP and inactivated when an intrinsic GTPase
dephosphorylates the bound GTP to form the inactive GDP-bound state. Unlike heterotrimeric G proteins,
monomeric G proteins do not interact directly with plasma membrane receptors or Gβγ but they can be
activated by Gαq, Gα12/13, Gαs, and Gαi.
The duration of a monomeric G protein-mediated signal is modified by the rates at which the bound GTP
is hydrolyzed and the bound GDP is replaced by GTP. Monomeric G protein-mediated
P.245
signals are generally more long-lasting than those initiated by heterotrimeric G proteins because they
hydrolyze GTP at a slower rate than Gα. Signal duration can be shortened when GTPase-activating
proteins (GAPs) stimulate the intrinsic GTPase activity of the G proteins, which accelerates their
inactivation. Monomeric G protein-mediated signaling can also be activated by guanine nucleotide
exchange factors (GEFs) that increase the rate at which GTP replaces the G protein-bound GDP. Other
mechanisms that regulate signaling by monomeric G proteins include cytoskeletal deformation and
posttranslational modifications such as phosphorylation, methylation, palmitoylation, glutathionylation,
and farnesylation.
Fig. 9-16: Proliferative signaling by β-adrenergic receptors. At least five mechanisms allow β-
adrenergic stimulation to modify proliferative signaling; from left to right these are: (1) cytoskeletal
signaling that is activated when increased contractility causes cell deformation; (2) Gαs-mediated
activation of PKA that phosphorylates transcription factors such as CREB; (3) InsP3-induced calcium
released that activates calcineurin, a phosphatase that dephosphorylates and so activates the
transcription factors GATA4 and MEF2C; (4) CAM kinase-induced activation of transcription factors;
and (5) PKC-mediated phosphorylation that actives transcription factors.
Activated monomeric G proteins generally mediate cell signaling when they are incorporated into
aggregates that activate protein kinases (see above). Major downstream targets for activated Ras, Rac,
and Rho include the MAP kinase pathways that play a central role in proliferative signaling (see below);
Rho signals are also transmitted by Rho kinases (ROCKs). Rac can modify proliferative signaling by
activating a nicotinamide adenine dinucleotide phosphate oxidase that
P.246
increases the production of reactive oxygen species (ROS) which activate the transcription factor NF-κB
(see above), and by activating MAP kinases and calcineurin (see below). Rac also increases myocardial
contractility by activating P21-activated kinase (PAK1), which stimulates a pathway that increases the
calcium sensitivity of the contractile proteins by dephosphorylating troponin I (see Chapter 5). Some of
the clinical benefits of statins have been attributed to attenuation of Rho- and Rac-mediated
hypertrophic and inflammatory responses when these drugs inhibit 3-hydroxy-3-methylglutaryl coenzyme
A (HMGCoA) reductase.
Calcium
Elevated levels of cytosolic calcium are generally linked to functional responses such as contraction and
secretion (see Chapter 7); however, this cation can also mediate proliferative signaling. Cytosolic proteins
that mediate the resulting proliferative responses include protein kinase C (PKC) (see below) and calcium-
calmodulin dependent protein kinases (CAM kinases). The latter can phosphorylate such transcription
factors as cyclic AMP-response element binding protein (CREB), and regulatory proteins that include
histone deacetylases (HDACs) (see above). Calcium also activates calcineurin, a phosphatase that
stimulates cardiac hypertrophy by dephosphorylating the inactive form of a transcription factor called
NFAT (nuclear factor of activated T cells). Calcineurin can generate a maladaptive response when the
increased levels of dephosphorylated NFAT activate transcription factors called GATA4 and MEF2C, which
stimulate pathological hypertrophy.
Protein Kinase C
Many of the signals mediated by G protein-coupled receptor-activated pathways are mediated by PKC, a
family of lipid-dependent enzymes that catalyze serine/threonine phosphorylations. This family of
enzymes include “classical” isoforms (α, βI, βII, and γ) whose activation requires both diacylglycerol
(DAG) and calcium, “novel” isoforms (δ, ε, θ, and η) that can be activated by DAG but do not require
calcium, and “atypical” isoforms (ζ and λ) that are activated by intracellular messengers that are derived
from lipids other than phosphatidylinositol 4,5-bisphosphate.
Mitogen-Activated Protein Kinase Pathways

One of the first mechanisms found to regulate cell growth and proliferation is a tightly coupled cascade of
serine/threonine kinase-catalyzed phosphorylations that allows ligand-bound plasma membrane receptors
to generate downstream transcriptional signals (Egan and Weinberg, 1993; Graves et al., 1997). This
cascade, called a mitogen-activated protein (MAP) kinase, was originally believed to be a single linear
sequence; however, it is now clear that several MAP kinases regulate protein synthesis, cell growth,
differentiation, survival, and, in proliferating cells, control the cell cycle. All have three components: an
upstream MAP kinase kinase kinase or MEK (an abbreviation for MAP kinase/ERK kinase), a MAP kinase
kinase, and a downstream MAP kinase that can phosphorylate a variety of transcription factors and other
regulatory proteins (Fig. 9-18).
MAP kinase pathways were initially found to be activated when peptide growth factors bind to
extracellular receptor kinases (ERKs), and so were called ERK pathways. Additional pathways, called
stress-activated MAP kinases, were subsequently found to be activated by inflammatory
P.247
cytokines, toxic agents, G protein-coupled receptors, cytoskeletal deformation, and other signals (Table
9-8). In the heart, MAP kinases mediate proliferative signals that cause myocytes to hypertrophy and
modify cardiac phenotype; in view of the link between cell growth and cell proliferation (see below), it is
not surprising that MAP kinases also regulate programmed cell death (apoptosis).
Table 9-8 Some Signals That Regulate MAP Kinases in the Heart
Peptide growth factors, e.g., FGF
Circulating neurohumoral mediators, e.g., norepinephrine, angiotensin II, endothelin
Locally released neurohumoral mediators, e.g., angiotensin II
Cytoskeletal deformation, e.g., integrins, cadherins
Cytokines, e.g., TNF-α, interleukin-6, cardiotrophin-like cytokine (CT-1)
The ERK- and stress-activated MAP kinase pathways are activated by monomeric G proteins: the Ras
family for the ERK pathway and the Rho family for stress-activated pathways. MAP kinases can also be
activated by heterotrimeric G protein-mediated signals that promote InsP3-induced calcium release and
activate PKC-catalyzed phosphorylations; these provide crossovers that allow norepinephrine, angiotensin
II, endothelin, and other functional signaling messengers to regulate cell growth and composition. The
switch from functional to proliferative signaling that follows prolonged sympathetic stimulation, when
internalized β2 receptors form scaffolds that activate proliferative signaling pathways (see Chapter 8),
represents a fascinating mechanism that allows a sustained stress to “inform” the heart that a short-term
increase in cardiac output is no longer adequate. This crossover states that in order for the organism to
survive, the heart must grow its way out of trouble!
The “Generic” MAP Kinase Pathway

Signaling by MAP kinase pathways resembles an American square dance in which the signal, like a dancer,
moves gracefully along a series of partners. This dance can begin on a phosphorylated receptor or a
scaffold formed by protein aggregates along the inner surface of the plasma membrane (see above); later,
the dance moves through the cytosol, crosses the nuclear membrane, and concludes when the activated
MAP kinases phosphorylate transcription factors in the nucleus. Common to all MAP kinase pathways is
activation of a MAP kinase kinase kinase (MKKK), a serine/threonine kinase that phosphorylates a second
kinase, called MAP kinase kinase (MKK) (Fig. 9-17). The latter then phosphorylates the MAP kinase. Most
MAP kinases enter the nucleus through pores in the nuclear membrane where they phosphorylate
transcription factors, although some MAP kinase substrates are cytosolic proteins. Signal diversity is
provided by the large number of MKKK, MKK, and MAP kinase isoforms, and the many interactions
between these pathways and other signal transduction cascades.
Extracellular Receptor Kinase Pathways

Extracellular receptor kinase (ERK)-mediated signals are generally initiated when ligand-binding to a
receptor tyrosine kinase causes the latter to form an aggregate that simulates its latent tyrosine kinase
activity (Fig. 9-18). Autophosphorylation of the receptor then creates a “docking site”
P.248
that initiates aggregations with additional signaling proteins, much as the partners in our dance join hands
to form a square. Aggregation begins when an activated receptor tyrosine kinase phosphorylates adaptor
proteins called SHC (from Src-homology because of similarities to the gene src) and Grb2 (growth
receptor binding protein). Scaffolds formed by these multi-protein aggregates interact with a guanine
nucleotide exchange factor called Sos (named after the drosophila mutant son-of-sevenless), which
activates Ras by exchanging its bound GDP for GTP in a reaction that is analogous to activation of Gα by
ligand-bound G protein-coupled receptors (see Chapter 8). In the ERK1/2 pathway, the Ras-GTP complex
activates a MKKK called Raf-1 that phosphorylates and activates the MKKs, which are the next partners in
the dance; the latter include MEK1 and MEK2 that phosphorylate a MAP kinase called extracellular
receptor kinase (ERK) which moves to the nucleus where it can phosphorylate a variety of transcription
factors.
Fig. 9-17: MAP kinase pathways. The “generic” pathway (left) lists key steps that are common to these
signaling pathways, all of which utilize a GTP-binding protein to activate a sequence of serine/threonine
kinases: MAP kinase kinase kinase (MKKK), which phosphorylates a MAP kinase kinase (MKK), which
phosphorylates a MAP kinase (MAPK) which, when phosphorylated, enters the nucleus where it phosphorylates
one or more nuclear transcription factors. Four types of MAP kinase pathway are shown: two extracellular
receptor-activated kinase pathways (ERK1/2 and ERK5) and two stress-activated pathways (JNK and p38). The
ERK pathways can be coupled by monomeric G proteins when activated by receptor tyrosine kinase agonists,
and by heterotrimeric G proteins, such as Gαs, Gαq and Gβγ, when activated by G protein-coupled receptors.
The major MKKK in the ERK1/2 pathway is Raf-1, MKKs include MEK1 and MEK2, and the most important MAP
kinases are ERK1 and ERK2. In the ERK5 pathway, the major MKKKs are MEKK2–3, the MKK is MEK5, and the
MAPK is ERK5. The two stress-activated MAP kinase pathways, which phosphorylate JNK and p38, are
generally coupled by monomeric G proteins of the Rho family and, in the case of the p38 pathway, by Gα11
and Gβγ. The MKKKs in the stress-activated pathways, called MEKKs, phosphorylate MKKs that activate JNK
and p38. When JNK enters the nucleus it phosphorylates a transcription factor called c-jun, while p38
activates nuclear transcription factors and heat shock proteins. MKK4 mediates crosstalk between the stress-
activated JNK and p38 pathways.
P.249
Fig. 9-18: Proliferative signaling by an extracellular receptor kinase (ERK) pathway. Binding of the ligand to
the receptor activates the latent receptor tyrosine kinase which autophosphorylates the receptor. This
initiates the formation of an aggregate, or scaffold, by creating a “docking site” that binds and
phosphorylates the adaptor protein Shc. This creates another docking site in which Shc adds Grb2 to a multi-
protein aggregate assembled along the inner surface of the plasma membrane. The aggregate then activates
Sos, a guanine nucleotide-exchange factor that exchanges Ras-bound GDP for GTP. The activated Ras-GTP
complex stimulates a MAP kinase kinase kinase (MKKK) which phosphorylates and activates a MAP kinase
kinase (MKK), which phosphorylates ERK, a MAP kinase (MAPK). Translocation of the latter to the nucleus
allows the activated MAPK to phosphorylate specific nuclear transcription factors (tc).
P.250
Cytoskeletal deformation can activate ERK1/2 signaling by initiating the formation of scaffolds similar to
those described above, and by activating focal adhesion kinases (see Chapter 5). G protein-coupled
receptors can also activate the ERK pathway, which allows mediators of the hemodynamic defense
reaction, like norepinephrine, angiotensin II, and endothelin, to evoke proliferative responses. ERK
signaling can be initiated when Gαs, Gβγ, and Gαq activate PKA and PKC, and by a GTPase-activating
protein called Rap1 (Fig. 9-17).
The ERK5 pathway, which because of its large size is also called “big MAP kinase,” utilizes a different
complement of kinases (Fig. 9-17). This pathway plays a role in cardiac development, is protective in
ischemic hearts, and has anti-apoptotic effects that favor cell survival. In blood vessels, ERK5 also helps
maintain vascular integrity. However, details regarding the cardiac effects of this MAP kinase are still
scanty.
Stress-Activated MAP Kinase Pathways

Two stress-activated MAP kinase pathways are activated by inflammatory cytokines, cell deformation, and
injury caused by such insults as viral infection, radiation, and toxins. There are many similarities between
stress-activated MAP kinases and the ERK pathways, but the JNK and p38 pathways are regulated
differently and activate c-Jun amino-terminal kinase (Jun kinase or JNK) and p38, respectively. Stress-
activated MKKKs include several MEKK (MEK kinase) isoforms whose function is analogous to that of Raf-1;
the MKKs in the stress-activated pathways are also activated by apoptosis signal-regulating kinase (ASK)
and TGF-β activated kinase (TAK). MEKKs in the stress-activated pathways include MKK7, which mediates
JNK signaling, MKK3 and MKK6, which operate in the p38 pathway, and MKK4, which is active in both.
Phosphorylation of JNK activates the transcription factor c-jun, while p38 is itself a transcription factor
that can activate other transcription factors and heat shock proteins. Responses to stress-activated MAP
kinase include maladaptive hypertrophy and apoptosis, which play an important role in determining
prognosis in patients with heart failure.
Phosphatidylinositol Trisphosphate, Phosphoinositide 3′-Oh Kinase,

and Akt (Protein Kinase B)
A signaling cascade that resembles the MAP kinase and cytokine pathways is controlled by the membrane
phospholipid phosphatidylinositol trisphosphate (PIP3) and a serine/threonine kinase called Akt. The
latter, which is named for a homologous protein in the T8 strain of AKR/J mice, is also called protein
kinase B (PKB) because it is homologous to PKA and PKC. This cascade, abbreviated PI3K-Akt in this text,
can be activated by receptor tyrosine kinase ligands that include insulin, insulin growth factor (IGF), and
growth hormone (GH); by ligands that release the Gβγ subunits of heterotrimeric G proteins; and by
cytoskeletal deformation. Unlike most MAP kinase- and cytokine-activated pathways, which often evoke
maladaptive responses in the heart, PI3K-Akt-mediated proliferative responses are generally adaptive.
Activation of the PI3K-Akt pathway begins when phosphoinositide 3′-OH kinases (PI3-kinases or PI3K)
phosphorylate the sugar moiety of PIP3 (Fig. 9-19). The phosphorylated PIP3 then provides a docking site
that binds to and activates Akt. Akt can also be activated when it is phosphorylated by phosphoinositide-
dependent kinase 1 (PDK1), an additional regulatory protein kinase.
P.251
Downstream responses to Akt phosphorylation include activation of mTOR (mammalian target of
rapamycin), and inhibition of both a serine/threonine kinase called glycogen synthetase kinase 3β (GSK-
3β) and a transcription factor called forkhead (FoxO). This system is turned off when PIP3 is
dephosphorylated by a lipid phosphatase called PTEN.
Fig. 9-19: PI3K/PIP3/Akt Pathways. Ligand binding to receptor tyrosine kinases activates a latent tyrosine
kinase that phosphorylates phosphoinositide 3′-OH kinases (PI3K). The latter then phosphorylate the
membrane phospholipid phosphatidylinositol trisphosphate (PIP3) which, when phosphorylated, provides a
docking site that activates Akt. Akt activity is increased when it is phosphorylated by phosphoinositide-
dependent kinase 1 (PDK1). Activated Akt inhibits glycogen synthetase kinase 3β (GSK-3β), and activates
forkhead (FoxO) and mTOR. Akt can also be activated by heterotrimeric G proteins and cytoskeletal
deformation (not shown).
Proliferative signaling by PI3K-Akt favors beneficial rather than deleterious responses. The former include
the physiological hypertrophy that results from training (the “athlete's heart”), an increase in capillary
density, and signals that favor cell survival; however, this pathway can also evoke deleterious responses,
including pathological hypertrophy (see Chapter 18).
Glycogen Synthase Kinase-3β

Glycogen synthase kinase 3 (GSK-3), which regulates glycogen metabolism (see Chapter 2), also
participates in proliferative signaling. In the heart, GSK-3β promotes apoptosis and inhibits several
transcriptional regulators that mediate maladaptive responses to overload. The latter
P.252
include NFAT (see above), the cytoskeletal protein β-catenin (see Chapter 5) that can be translocated to
the nucleus where it regulates gene expression, and GATA4 which is a zinc-finger transcription factor.
GSK-3β activity is inhibited by the PI3K-Akt pathway, which would be expected to attenuate the
maladaptive responses, but other responses to GSK-3β can be deleterious. The predominant effects of
GSK-3β inhibition on cardiac hypertrophy appear to be beneficial.
Forkhead
Forkhead is a transcription factor that plays a role in cardiac development, increases the expression of
genes that favor cardiac atrophy, reduces cardiac myocyte size, and inhibits hypertrophic responses.
Phosphorylation by the PIP3-Akt pathway, which activates forkhead, can be initiated by the many stimuli
that activate Akt, as well as by atrogin-1, a ubiquitin ligand that is associated with the cytoskeleton (see
Chapter 5). The cardiac effects of forkhead, like those of other mediators of the PIP3-Akt pathway, are
complex, but the overall response appears to be beneficial in chronically overloaded and failing hearts.
Mtor
The mTOR family of protein kinases controls the overall rate of protein synthesis, regulates gene
expression, and coordinates cell growth and nutritional status. In energy-starved cells, the decreased
[ATP]/[AMP] ratio activates an AMP kinase that reduces mTOR activity. Activation of mTOR by the PI3K-Akt
pathway in the heart can lead to both adaptive and maladaptive proliferative responses. The pivotal role
of mTOR in integrating energy production and energy expenditure along with its ability to mediate
adaptive hypertrophic responses probably contribute to the overall beneficial effects of PI3K-Akt signaling
in the heart.
Myocardial Cell Death

The loss of myocytes in the human heart that accompanies normal aging is accelerated by both energy
starvation and many of the proliferative signals generated in chronically overloaded and diseased hearts.
Many of the regulatory pathways that mediate the hypertrophic responses of the terminally differentiated
cardiac myocytes also accelerate myocyte death. Stem cells have been identified in adult mammalian
hearts, but unfortunately they are not able to repopulate the heart.
Myocyte loss creates an especially serious problem in chronically overloaded and damaged hearts because
it adds to the overload on the surviving myocytes. For this reason, cardiac myocyte death can establish a
vicious cycle in which loss of myocytes overloads the surviving myocytes, which intensifies hypertrophic
signaling, accelerates myocyte death, increases overload, etc. Cardiac myocyte death can therefore be a
calamity.
Mechanisms of Cell Death

Cells can die in three ways; they can be killed by extrinsic factors (necrosis or oncosis), or they can be
programmed to die by cellular signaling systems (apoptosis and autophagy). The hallmark of necrosis is
cell swelling and breakdown of the plasma membrane barrier that ends when rupture releases the cell's
contents and evokes an intense inflammatory response that leads to reactive fibrosis. In the heart,
increased plasma membrane permeability allows calcium
P.253
to leak into the cytosol, which exposes the myofilaments to high calcium concentrations. The latter can
initiate explosive interactions between the contractile proteins that cause contraction-band necrosis in
which the myocytes are literally torn apart (see Chapter 17). Apoptosis, a highly regulated process in
which cells first shrink and then vanish when the cell fragments are engulfed by phagocytes, plays an
important role in ridding the body of unneeded or damaged cells. Autophagy, which, like apoptosis, is
highly regulated, allows the cell contents to be reused as a source of energy and as building blocks for the
synthesis of new cells. The distinctions between these forms of cell death are not always sharp: There can
be gradations between necrosis, apoptosis, and autophagy, and cell death that begins as apoptosis can
end as necrosis.
Necrosis (Oncosis)
Necrosis (derived from the Greek word for corpse; also called oncosis from the Greek word for swelling) is
an accidental form of cell death that can occur when cells are damaged by inflammation, toxins,
temperature extremes, mechanical injury, or sudden loss of energy supply. Mechanisms include plasma
membrane abnormalities, proteolysis of cytoskeletal proteins, and severe mitochondrial injury. Necrosis
begins with potentially reversible abnormalities in membrane permeability that lead to cell swelling;
these are followed by the appearance of plasma membrane blebs and membrane rupture that spills the
cell contents into the extracellular space. The latter evokes an exudative inflammatory response that
leads to reactive fibrosis.
Apoptosis
Before describing apoptosis, which is derived from two Greek words, apo (away) and ptosis (falling) that
describe the fall of leaves from a deciduous tree in the autumn, I cannot refrain from commenting on its
pronunciation—and frequent mispronunciation. There are defensible reasons to pronounce or not to
pronounce the second “p” (apo•ptosis′ or apo•tosis′), but there is no basis for the mispronunciation
“a•pop′•tosis,” which combines parts of the two words to create a third, nonsense, syllable. For this
reason, there is no “pop” in apoptosis!
Apoptosis can be initiated by injury and a variety of proliferative signals. The concurrence between
proliferation and apoptosis is especially important in rapidly dividing tissues, where the latter eliminates
cells for which the need has ended; this is essential for embryonic development, where many cell types
appear and then disappear. If removal of unneeded cells in the fetus were to lead to fibrosis, the newborn
would be a mass of scar tissue; for example, as many as half of the neurons that appear in the developing
vertebrate nervous system are eliminated after they form synaptic connections with their target cells
(Raff et al., 1993). Savill et al. (1997) observed that the role of apoptosis is “beautifully demonstrated in
the developing drosophila eye, where to achieve the adult form, thousands of unwanted interommatidial
cells undergo programmed death and phagocytosis without disrupting the delicate architecture of the
organ.” If unneeded cells could not be eliminated without causing fibrosis during early development, we
would all have been born blind—and with gills, tails, and webs between our fingers and toes!
Apoptosis is initiated when the “decision” that a cell must die leads to activation of programs that initiate
programmed cell death, withdrawal of apoptosis-inhibitory factors that allows a pre-programmed death
process to kill the cell, or both. These responses cause the targeted cells to shrink and break up into
membrane-surrounded fragments that often contain bits of condensed chromatin called apoptotic bodies
(Fig. 9-20). Maintenance of plasma membrane integrity
P.254
P.255
until late in the apoptotic process prevents release of reactive cellular contents and allows the fragments
of the dying cell to be engulfed by macrophages. In this way, apoptosis differs from necrosis, in which
plasma membrane rupture releases reactive cell contents into the extracellular space. Another distinction
between necrosis and apoptosis is the way DNA is degraded; in necrosis the DNA is broken down into
randomly sized fragments, whereas DNA breakdown in apoptosis releases regularly sized fragments that
resemble a ladder when they are fractionated on gels. The ability of phagocytes to ingest the cell
fragments formed during apoptosis without provoking an inflammatory response to the “raw” cell
contents that spill into tissues resembles the preparation of an elegant meal in which meats are boned,
tenderized, and cut into bite-sized pieces; shells are removed from lobsters, shrimp, and clams; and fruits
and vegetables are cored and peeled. In necrosis, according to this culinary analogy, the same ingredients
are presented in raw form, often still alive, which causes the diner to experience a stomachache that is
analogous to the inflammation that characterizes necrosis.
Fig. 9-20: Necrosis, apoptosis, and autophagy. Necrosis (left) is generally caused when plasma membrane
damage impairs its ability to serve as a permeability barrier, which causes cells to swell and eventually burst.
The resulting release of cell contents initiates an inflammatory reaction that leads to fibrosis. Apoptosis
(center) is a highly regulated process that causes cell shrinkage and condensation of the cytosol and nucleus;
this yields cell fragments, called apoptotic bodies that, because they are surrounded by plasma membrane,
can be engulfed and digested by phagocytes without evoking an inflammatory reaction. In autophagy (right),
cell organelles are enclosed in membrane-lined vesicles called proteasomes that fuse with lysosomes whose
enzymes degrade the organelle.
Two different signaling pathways can initiate apoptosis. One is activated when extracellular messengers,
such as G protein-coupled receptor agonists, growth factors, and cytokines, bind to plasma membrane
receptors. In the other, apoptosis is initiated by damaged mitochondria within cells.
Extracellular pro-apoptotic signals can be initiated by peptides that are synthesized in response to ligands
that activate cytokine receptors and gp130; functional mediators like norepinephrine, angiotensin II, and
endothelin; cytoskeletal deformation; and many of the proliferative signaling pathways described in this
chapter. A death-receptor pathway is triggered when members of a family of peptides called Fas ligand
(FasL) bind to plasma membrane receptors called Fas (Fig. 9-21). These peptides, which are cytokines
that can be found in both membrane-bound (mFasL) and soluble (sFasL) forms, initiate apoptosis by
activating a sequence of 70 to 90 amino acids in the Fas receptor called a death domain; death domains
are found in other cytokine receptors, caspases, and several regulatory proteins. When bound to their
ligands, Fas receptors form aggregates in which the death domains bind to adaptor proteins, such as FADD
(Fas-associated death domain protein), that can form a death-inducing signaling complex that activates
an enzyme called caspase 8 (see below). Conversely, activators of the PI3K-Akt pathway and a cytosolic
protein called FLIP (Fas ligand inhibitory protein) inhibit these pro-apoptotic pathways.
Intracellular pro-apoptotic pathways (Fig. 9-22), which can be activated in damaged, energy-starved, or
calcium-overloaded cells, open pores in the mitochondrial inner membrane that release cytochrome c, a
respiratory chain enzyme (see Chapter 2). The mitochondrial pathway is controlled by members of a
peptide family called Bcl2 (an acronym for B-cell lymphoma/leukemia 2 gene) that includes both pro- and
anti-apoptotic peptides. Some, including Bak, Bax, Bad, Bid, Bim, Bmf, PUMA, NOXA, and Nix promote
cell death, while Bcl2, Bcl-xl, Bclw, Mcl1, A1, and Boo/Diva inhibit apoptosis. Once in the cytosol,
cytochrome c released from mitochondria binds to procaspase 9 and Apaf-1 (apoptotic protease
activating factor-1), an adaptor protein whose function is similar to that of FADD. This forms an
apoptosome that converts the inactive procaspase 9 to activated caspase 9.
Caspases play a number of different roles in apoptosis. Caspase 8, which is activated by FADD in the
extracellular pathway, and caspase 9, which is activated by the intracellular (mitochondrial) pathway,
activate caspase 3, caspase 6, and caspase 7 that hydrolyze cytoskeletal and nuclear regulatory proteins,
tumor suppressors, and enzymes that participate in RNA splicing, cell division, and DNA repair and
replication. Other caspases contain a death domain that activates apoptosis when it binds to homologous
death domains in adaptor proteins.
P.256
Fig. 9-21: Extracellular pro-apoptotic pathways. Cytokines called Fas ligands (FasL) bind to Fas, a
plasma membrane receptor that contains FasL-binding and death domains. Ligand-bound Fas
receptor aggregates with the death domains and interact with adaptor proteins such as FADD,
which also contain death domains. Inclusion of procaspase 8 in these aggregates releases activated
caspase 8 that, along with other enzymes, break down the cell constituents.
The transcription factor p53, a major regulator of cell survival, has been called the “master watchman”
of the genome because it regulates cell cycling in normally proliferating cells but is pro-apoptotic in
damaged cells (Saini and Walker, 1998). These effects allow p53 to play a central role in determining the
fate of an injured cell. When the damage is mild, p53 favors DNA repair by shutting down the cell cycle,
whereas when an injury is so severe that DNA damage cannot be repaired, its pro-apoptotic effects kill
the cell. In proliferating tissues, the pro-apoptotic effect of p53 helps prevent malignant transformation
by eliminating severely damaged, potentially transformed cells. In terminally differentiated cells like
cardiac myocytes, where malignant transformation is rare, p53 serves mainly to kill badly damaged cells.
Autophagy and the Ubiquitin-Proteasome System

Autophagy (the term derived from the Greek words: auto = self, and phagy = eat) is a form of cell death
in which cells are taken apart in a manner that preserves key constituents for reuse. The classical setting
for autophagy is severe chronic energy starvation, where this process removes and recycles damaged
organelles and proteins. Unlike apoptosis, which is characterized
P.257
by cell shrinkage and the formation of plasma membrane-lined apoptotic vesicles, in autophagy cell
organelles like mitochondria are enclosed within a membrane-lined autophagosomes which then fuse with
lysosomes that contain enzymes that degrade the organelle. The process is regulated by several proteins,
including a PI3 kinase and a member of the Bcl2 family.
Fig. 9-22: Intracellular pro-apoptotic pathways. Opening of pores in the mitochondrial inner
membrane releases cytochrome C, which interacts with Apaf-1 and procaspase 9 to form
apoptosomes that contain activated caspase 9. These pathways are controlled by Bcl-2 and related
peptides, some of which favor cell death (e.g., Bax, Bad, and Bid), while others (e.g., Bcl-2) are
anti-apoptotic.
Damaged and unneeded cell proteins are also removed by ubiquitin-proteasome system in which the
proteins are first covalently bound to ubiquitin, a molecular marker that identifies the proteins that are
to be degraded. This process is regulated by heat shock proteins, members of the Bcl2 class of proteins,
and posttranslational modifications like phosphorylation, hydroxylation, and glycosylation.
Myocardial Responses To Stress

Gene expression is rapidly modified when cardiac myocytes are subjected to stresses such as mechanical
or thermal injury, ischemia, diastolic stretch (increased preload), or high systolic tension (increased
afterload). The first genes to be upregulated, called immediate-early response genes, are not normally
transcribed in the resting (G0) phase of the cell cycle. The rapidity with which these genes become
activated, often within minutes after a cardiac myocyte is overloaded, indicates that most immediate-
early genes are activated by phosphorylations and other posttranslational modifications of previously
inactive transcriptional regulators. If the stress is sustained, additional late-response genes become
activated, in some cases by transcription factors whose synthesis is stimulated by the immediate-early
genes. Both the immediate-early and late responses play a pathogenic role in heart failure, for example,
by stimulating hypertrophy
P.258
and apoptosis, and modifying the size, shape, and protein composition of the cardiac myocytes (see
Chapter 18).
Immediate-Early Response Genes

When the heart becomes damaged or is overloaded, proliferative signaling pathways initiate an
immediate-early response in which more than 100 different genes are activated and then deactivated.
Genes whose expression is increased include ras, which encodes the monomeric GTP-binding protein Ras,
and nuclear transcription factor regulators like c-myc, c-fos, and c-jun, and hsp-70, hypoxia-inducible
factors, and heat shock proteins (see below). Instead of responding in a monotonic fashion, the many
genes that participate in the immediate-early response are activated at different times during the first
minutes and hours after the onset of the stress, and then are inactivated at different rates over the
subsequent hours and days. In addition to these temporal heterogeneities, there are spatial
heterogeneities in the immediate-early response; for example, overload causes mRNAs encoding specific
contractile protein isoforms to be upregulated at different times in various regions of the heart
(Schiaffino et al., 1989).
Hypoxia-Inducible Factor
The heart's defenses against energy starvation are mediated by hypoxia-inducible factors (HIF) that
modify gene transcription in response to changes in oxygen availability. HIF-1α, which is found in the
heart, increases the synthesis of proteins that protect against hypoxic damage; these include
erythropoietin which increases oxygen delivery, VEGF which promotes blood vessel growth, metabolic
enzymes that increase anaerobic ATP production, and signaling molecules that inhibit apoptosis.
Heat Shock Proteins

Damage to stressed cells is minimized by the “heat shock” response, a highly conserved defense
mechanism that received its name because it was first observed in cells exposed to high temperature.
This response activates heat shock proteins (HSPs) which inhibit the denaturation of cellular proteins by
stabilizing hydrophobic surfaces that become exposed in partially denatured proteins and regulate
proliferative signaling by modifying the activity of transcription factors. Because of this protective effect,
they are also called molecular chaperones. Like an elderly relative who protects a susceptible youngster
from associating with undesirable companions, HSPs protect stressed cells from irreversible aggregations
that would otherwise denature their proteins.
HSPs are among the first to be activated by stress. In the heart, upregulation of HSP-70 heat shock
proteins (so named because their molecular weight is about 70 kDa) during the immediate-early response
to pressure overload plays a protective role by inhibiting maladaptive growth; in addition, some smaller
HSPs have an anti-apoptotic effect. The upregulation of the HSPs is so rapid that even a single stretch of
the adult rabbit heart greatly increases their expression (Knowlton et al., 1991).
Two members of this family, HSP-70, which is constitutively expressed, and HSP-27, which is inducible,
favor the appearance of physiological (as opposed to pathological) cardiac hypertrophy, inhibit apoptosis,
counteract some deleterious effects of cytokines, protect against
P.259
mitochondrial damage, and preserve cytoskeletal structure. A highly regulated heat shock transcription
factor-1 (HSF-1) promotes the synthesis of HSPs.
Late Response Genes

The immediate-early response, which is transient, is followed by sustained activation of a different
complement of genes called late response genes. Unlike the immediate-early response, late response
genes encode newly synthesized proteins, including mitochondrial components, cytoskeletal and
myofibrillar proteins, enzymes, and transcriptional regulators like cyclins and CDKs. In response to
overload, many of these newly activated genes encode protein isoforms normally found during
development of the fetal heart (see Chapter 18).
Reversion to the Fetal Phenotype

A remarkable feature of the proliferative response in overloaded hearts is the reappearance of the
patterns of gene expression normally seen in fetal life. This preferential expression of the fetal phenotype
can be viewed simplistically as part of a “failed” effort of these terminally differentiated cells to
proliferate.
The functional consequences of the reversion of the heart to the fetal phenotype are complex. For
example, increased expression of a low ATPase fetal myosin heavy chain, which replaces the higher
ATPase myosin normally found in the adult heart (see Chapter 4), has detrimental effects because it
reduces contractility by slowing the turnover of myosin cross-bridges, thereby worsening the
hemodynamic abnormalities caused by chronic overloading (see Chapter 6). At the same time, however,
this isoform shift slows the rate of ATP hydrolysis by the heart's contractile machinery, which has an
energy-sparing effect that is beneficial in overloaded hearts (see Chapter 18). An important consequence
of the reversion to the fetal phenotype is reduction of the content of sarcoplasmic reticulum which, in
addition to reducing contractility, increases the heart's dependence on calcium derived from the
extracellular fluid (the “extracellular calcium cycle” described in Chapter 7). Because both calcium entry
and calcium efflux across the plasma membrane are accompanied by depolarizing currents, this feature of
the reversion to the fetal phenotype contributes to the arrhythmias and sudden death commonly seen in
end-stage heart failure (see Chapter 18).
Overview
It seems appropriate to conclude this chapter with a few generalizations that highlight the role of
proliferative signaling in cardiac physiology. Most important is that terminally differentiated cardiac
myocytes normally do little except contract and relax; like stolid oxen, they spend their lives pulling a
burden and so do not have time to reproduce. The remarkable durability of these cells, which can survive
for decades—and sometimes for a century—reflects the fact that as long as their activity is restricted to
the tasks of contracting and relaxing, they do not wear out. However, this durability carries a price, which
is that efforts to modify their routine can have fatal consequences. If, for example, a heart is paced
continuously at a rapid rate, the pump begins to fail and cells begin to die; similarly, sustained
mechanical overload shortens cardiac myocyte survival. Thus, even though cardiac myocytes are able to
contract without pause for decades, this activity must remain within limits because overloading the
cardiac pump can trigger proliferative responses that, while initially compensatory, eventually destroy the
heart.
P.260
Another feature of proliferative signaling highlighted in this chapter is that even the simplest intervention
that modifies cardiac function evokes a multiplicity of cellular responses. A stimulus as simple as an
increase in diastolic volume not only causes an immediate increase in performance (Starling's law of the
heart), but because stretch deforms the cytoskeleton, it also activates proliferative responses.
Elucidation of the signaling functions of the cytoskeleton has revealed the intricacy of these regulatory
mechanisms as it is now clear that these proteins allow stretch to generate signals that modify cell size,
shape, and composition.
Another generalization is that because cell signaling rarely proceeds in a “straight line,” most stimuli
generate an impressive array of cellular responses. An important corollary to this principle, which has
been found to be highly relevant to the management of heart disease, is that efforts to modify cell
signaling that are intended to do good almost always do some harm. That the harm can exceed the
benefit is apparent in the counterintuitive results of many of the clinical trials discussed in Chapters 16
and 18. Signal transduction should therefore be viewed as a floodlight, rather than a spotlight. Although
compensatory responses help the patient with heart failure stay on the safest path and avoid the many
pitfalls and dangers that lurk alongside the road, the beam of these “compensatory torches” is often so
broad that it not only helps avoid danger, but also attracts dormant monsters whose awakening can prove
fatal to the cardiac patient. Optimal use of modern therapy, therefore, requires that healthcare providers
understand the mechanisms by which the heart responds to disease so as to maximize the benefits and
minimize the harm that any treatment strategy can cause in a given patient.
Bibliography
General
Akazawa H, Komuro I. Roles of cardiac transcription factors in cardiac hypertrophy. Circ Res 2003;
92:1079–1088.
Alberts B, Johnson A, Lewis J, et al. Molecular Biology of the Cell. 5th ed. New York, NY: Garland,
2008.
Bugaisky L, Gupta M, Gupta MG, et al. Cellular and molecular mechanisms of hypertrophy. In:
Fozzard H, Haber E, Katz A, et al., eds. The Heart and Cardiovascular System. 2nd ed. New York:
Raven Press, 1992:1621–1640.
Buja LM, Vela D. Cardiomyocyte death and renewal in normal and diseased heart. Cardiovasc Pathol
2008;17:349–374.
Chien KR, Knowlton KU, Zhu H, et al. Regulation of cardiac gene expression during myocardial
growth and hypertrophy: molecular studies of an adaptive physiologic response. FASEB J
1991;5:3037–3046.
Dorn GW 2nd, Force T. Protein kinase cascades in the regulation of cardiac hypertrophy. J Clin Invest
2005;115:527–537.
Frey N, Katus HA, Olson EN, et al. Hypertrophy of the heart. A new therapeutic target. Circulation
2004; 109:1580–1589.
Hill JA, Olson EN. Cardiac plasticity. New Engl J Med 2008;358:1370–1380.
Lodish H, Berk A, Kaiser CA, et al. Molecular Cell Biology. 6th ed. New York, NY: Freeman, 2008.
Mercadier JJ. Determinants of left ventricular hypertrophy. Immun Endoc Metab Agents Med Chem
2006;6:343–365.
Molkentin JD, Dorn GW 2nd. Cytoplasmic signaling pathways that regulate cardiac hypertrophy. Annu
Rev Physiol 2001;63:391–426.
Natarajan K, Berk BC. Crosstalk coregulation mechanisms of G protein-coupled receptors and

receptor tyrosine kinases. Methods Mol Biol 2006;332:51–77.
P.261
Novartis Foundation Symposium 274. Heart failure: molecules, mechanisms and therapeutic targets.
Chichester UK: Wiley, 2006.
Olson EN. A decade of discoveries in cardiac biology. Nature Med 2004;10:467–474.

Sekiguchi K, Li X, Coker M, et al. Cross-regulation between the renin-angiotensin system and
inflammatory mediators in cardiac hypertrophy and failure. Cardiovasc Res 2004;63:433–442.
Selvetella G, Hirsch E, Notte A, et al. Adaptive and maladaptive hypertrophic pathways: points of
convergence and divergence. Cardiovasc Res 2004;63:373–380.
Simpson PC. β-protein kinase C and hypertrophic signaling in human heart failure. Circulation
1999;93: 334–337.
Cell Cycle
Dannenberg JH, te Riele HP. The retinoblastoma gene family in cell cycle regulation and suppression
of tumorigenesis. Results Probl Cell Differ 2006;42:183–225.
Field LJ. Modulation of the cardiomyocyte cell cycle in genetically altered animals. Ann N.Y. Acad
Sci 2004;1015:160–170.
Ford HL, Pardee AB. Cancer and cell cycling. J Cell Biochem 1999;75(S32):166–172.
Polager S, Ginsberg D. E2F—At the crossroads of life and death. Trends Cell Biol 2008;18:528–535.
Van den Heuval S, Dyson NJ. Conserved functions of the pRB and E2F families. Nature Mol Cell Biol
2008;9:713–724.
Wong LSM, Oeseburg H, de Boer RA, et al. Telomere biology in cardiovascular disease: the
TERC/mouse as a model for heart failure and aging. Cardiovasc Res 2009;81:244–252.
Zhu W, Hassink RJ, Rubart M, et al. Cell-cycle-based strategies to drive myocardial repair. Pediatr
Cardiol 2009;30:710–715.
Epigenetics
Amaral PP, Dinger ME, Mercer TR, et al. The eukaryotic genome as an RNA machine. Science
2009;319: 1781–1789.
Backs J, Olson EN. Control of cardiac growth by histone acetylation/deacetylation. Circ Res
2006;98:15–24.
Chandler VL, Stam M. Chromatin conversions: mechanisms and implications of paramutation. Nat
Rev Genetics 2004;5:532–544.
Couture JF, Trievel RC. Histone-modifying enzymes: encrypting an enigmatic epigenetic code. Curr
Opinion Struct Biol 2006;16:753–760.
Goldberg AD, Allis CD, Bernstein W. Epigenetics: a landscape takes shape. Cell 2007;128:635–638.
Holmes R, Soloway PD. Regulation of imprinted DNA methylation. Cytogenet Genome Res 2006;113:
122–129.
Kaminsky Z, Wang SC, Petronis A. Complex disease, gender, and epigenetics. Ann Med 2006;38:530–
544.
Klose RJ, Bird AP. Genomic DNA methylation: the mark and its mediators. Trends Biochem Sci
2006;31: 89–97.
Latronico MVG, Condorelli G. MicroRNAs and cardiac pathology. Nat Rev Cardiol 2009;6:418–429.
Mistelli T. Beyond the sequence: cellular organization of genome function. Cell 2007;128:787–800.
Rama TM. Illuminating the silence: understanding the structure and function of small RNAs. Nat Rev
Mol Cell Biol 2007;8:23–36.
Richards EJ. Inherited epigenetic variation—revisiting soft inheritance. Nat Rev Genet 2006;7:395–
401.
Slotkin RK, Martienssen R. Transposable elements and the epigenetic regulation of the genome. Nat
Rev Genet 2007;8:272–285.
Thum T, Catalucci D, Bauersachs J. MicroRNAs: novel regulators in cardiac development and disease.
P.262
Van Rooij E, Marshall WS, Olson EN. Toward microRNA based therapeutics for heart disease: the
sense in antisense. Circ Res 2008;103:919–928.
Zufall RA, Robinson T, Katz LA. Evolution of developmentally regulated genome arrangements in
eukaryotes. J Exp Zool 2005;304B:448–455.
Monomeric and Heterotrimeric G Proteins

Brown JH, Del Re DP, Sussman MA. The Rac and Rho hall of fame: a decade of hypertrophic signaling
hits. Circ Res 2006;98:730–742.
Clerk A, Sugden PH. Ras: the stress and the strain. J Mol Cell Cardiol 2006;41:595–600.
Lezoualc'h F, Mátrich M, Hmitou I, et al. Small GTP-binding proteins and their regulators in cardiac
hypertrophy. J Mol Cell Cardiol 2008;44:623–632.
Wang CY, Liu PY, Liao JK. Pleiotropic effects of statin therapy: molecular mechanisms and clinical
results. Trends Mol Med 2008;14:37–44.
van Biesen T, Luttrell LM, Hawes BE, et al. Mitogenic signaling via G protein-coupled receptors.
Endoc Rev 1996;17:698–714.
Wennerberg K, Rssman KL, Der CJ. The Ras superfamily at a glance. J Cell Sci 2005;118:843–846.
Peptide Growth Factors
Detillieux KA, Sheikh F, Kardami E, et al. Biological activities of fibroblast growth factor-2 in the
adult myocardium. Cardiovasc Res 2003;57:8–19.
ten Dijke P, Hill CS. New insights into TGF-β–Smad signalling. Trends Biochem Sci 2004;29:265–273.
Hill CS. Signaling to the nucleus by members of the transforming growth factor-β (TGF-β)
superfamily. Cell Signal 1996;8:533–544.
Kardami E, Jiang ZS, Jimenez SK, et al. Fibroblast growth factor 2 isoforms and cardiac hypertrophy.
Khan R, Sheppard R. Fibrosis in heart disease: understanding the role of transforming growth factor-
beta in cardiomyopathy, valvular disease and arrhythmia. Immunology 2006;118:10–24.
Rosenkranz S. TGF-β1 and angiotensin networking in cardiac remodeling. Cardiovasc Res 2004;63:
423–432.
Seo D, Hare JM. The transforming growth factor-β/Smad3 pathway: coming of age as a key
participant in cardiac remodeling. Circulation 2007;116:2096–2098.
Cytokines
Boengler K, Hilfiker-Kleiner D, Drexler H, et al. The myocardial JAK/STAT pathway: from protection
to failure. Pharmacol Ther 2008;120:172–185.
Hall G, Hasday JD, Rogers TB. Regulating the regulator: NF-κB signaling in the heart. J Mol Cell
Cardiol 2006;41:580–591.
Mann DL. Stress-activated cytokines and the heart: from adaptation to maladaptation. Annu Rev
Physiol 2003;65:81–101.
McFalls EO, Liem D, Schoonerwoerd K, et al. Mitochondrial function: the heart of myocardial
preservation. J Lab Clin Med 2003;142:141–149.
Stroud RM, Wells JA. Mechanistic diversity of cytokine receptor signaling across cell membranes. Sci
STKE 2004;(231):re7.
P.263
MAP Kinases
Bueno OF, Molkentin JD. Involvement of extracellular signal-regulated kinases 1/2 in cardiac
hypertrophy and cell death. Circ Res 2002;91:776–781.
Boutros T, Chevet E, Metrakos P. Mitogen-activated protein (MAP) kinase/MAP kinase phosphatase

regulation: roles in cell growth, death, and cancer. Pharmacol Rev 2008;60:231–310.
Liang Q, Molkentin JD. Redefining the roles of p38 and JNK signaling in cardiac hypertrophy:
dichotomy between cultured myocytes and animal models. J Mol Cell Cardiol 2003;35:1385–1394.
Obara Y, Nakahata N. The signaling pathway leading to ERK5 activation via G-proteins and ERK5-
dependent neurotrophic effects. Mol Pharmacol 2010;77:10–16.
Ono K, Han J. The p38 signal transduction pathway: activation and function. Cell Signaling
2000;12:1–13.
Petrich BG, Yibin Wang Y. Stress-activated MAP kinases in cardiac remodeling and heart failure new
insights from transgenic studies. Trends CV Med 2004;14:50–55.
Wang Y. Mitogen-activated protein kinases in heart development and diseases. Circulation 2007;116:
1413–1423.
Widmann C, Gibson S, Jarpe MP, et al. Mitogen-activated protein kinase: conservation of a three
kinase module from yeast to human. Physiol Rev 1999;79:143–180.
PI3P Kinases/Akt/GSK/mTOR/Forkhead
Ceci M, Ross J Jr, Condorelli G. Molecular determinants of the physiological adaptation to stress in
the cardiomyocyte: a focus on AKT. J Mol Cell Cardiol 2004;37:905–912.
Downward J. PI 3-kinase, Akt and cell survival. Sem Cell Develop Biol 2004;15:177–182.
Ma XM, Blenis J. Molecular mechanisms of mTOR-mediated translational control. Nat Rev Mol Cell
Biol 2009;10:307–318.
Naga Prasad SV, Perrino C, Rockman HA. Role of phosphoinositide 3-kinase in cardiac function and
heart failure. Trends Cardiovasc Med 2003;13:206–212.
O'Neill BT, Able ED. Akt1 in the cardiovascular system: friend or foe. J Clin Invest 2005;115:2059–
2064.
Oudit GY, Penninger JM. Cardiac regulation by phosphoinositide 3 kinases and PTEN. Cardiovasc Res
2009;82:250–260.
Papanicolaou KN, Izumiya Y, Walsh K. Forkhead transcription factors and cardiovascular biology. Circ
Res 2008;102:16–31.
Shevtsov SP, Haq S, Force T. Activation of β-catenin signaling pathways by classical G-protein-
coupled receptors. Cell Cycle 2006;5:2295–2300.
Sugden PH, Fuller SJ, Weiss SC, et al. Glycogen synthase kinase 3 (GSK3) in the heart: a point of
integration in hypertrophic signalling and a therapeutic target? A critical analysis. Br J Pharmacol
2008; 153(Suppl 1):S137–S153.
Walsh K. Akt signaling and growth of the heart. Circulation 2006;113:2032–2034.
Protein Kinases, Calcium, Stretch
Dorn GW 2nd, Force T. Protein kinase cascades in the regulation of cardiac hypertrophy. J Clin Invest
2005;115:527–537.
Hu H, Sachs F. Stretch-activated ion channels in the heart. J Mol Cell Cardiol 1997;29:1511–1523.
Kudoh S, Akazawa H, Takano H, et al. Stretch-modulation of second messengers: effects on
cardiomyocyte ion transport. Prog Biophys Mol Biol 2003;82:57–66.
Palaniyandi SS, Sun L, Ferreira JCB, et al. Protein kinase C in heart failure: a therapeutic target.
P.264
Yano M, Ikeda Y, Matsuzaki M. Altered intracellular Ca2+ handling in heart failure. J Clin Invest 2005;
115:556–564.
Zhang T, Brown JH. Role of Ca2+/calmodulin-dependent protein kinase II in cardiac hypertrophy and
heart failure. Cardiovasc Res 2004;63:476–486.
Miscellaneous
Gratton JP, Bernatchez P, Sessa WC. Caveolae and caveolins in the cardiovascular system. Circ Res
2004; 94:1408–1417.
Means CK, Brown JH. Sphingosine-1-phosphate receptor signaling in the heart. Cardiovasc Res
2009;82: 193–200.
Rao A, Luo C, Hogan PG. Transcription factors of the NFAT family. Regulation and function. Annu Rev
Immunol 1997;15:707–747.
Steinberg SF. β2-Adrenergic receptor signaling complexes in cardiomyocyte caveoli/lipid rafts. J Mol
Cell Cardiol 2004;37:407–415.
Cell Death: Necrosis, Apoptosis, Autophagy
Crow MT, Mani K, Nam YJ, et al. The mitochondrial death pathway and cardiac myocyte necrosis.
Circ Res 2004;95:957–970.
Foo RSY, Mani K, Kitsis RN. Death begets failure in the heart. J Clin. Invest 2005;115:565–571.
Giordano FJ. Oxygen, oxidative stress, hypoxia, and heart failure. J Clin Invest 2005;115:500–508.
Gustafsson AB, Gottlieb RA. Mechanisms of apoptosis in the heart. J Clin Immunol 2004;23:447–459.
Hotchkiss RS, Strasser A, McDunn JE, et al. Cell death. New Engl J Med 2009;361:1570–1583.
Kidd VJ. Proteolytic activities that mediate apoptosis. Ann Rev Physiol 1998;60:533–573.
Levine B, Kroemer G. Autophagy in the pathogenesis of disease. Cell 2008;132:27–42.
Majno G, Joris I. Apoptosis, oncosis, and necrosis. An overview of cell death. Am J Pathol 1995;146:
3–15.
McFalls EO, Liem D, Schoonerwoerd K, et al. Mitochondrial function: the heart of myocardial
preservation. J Lab Clin Med 2003;142:141–149.
Mowat MRA. p53 in tumor progression: life, death and everything. Adv Cancer Res 1998;74:25–48.
Su H, Wang X. The ubiquitin-proteasome system in cardiac proteinopathy: a quality control

perspective. Cardiovasc Res 2010;85:253–262.
van Empel VPM, De Windt LJ. Myocyte hypertrophy and apoptosis: a balancing act. Cardiovasc Res
2004;63:487–499.
Wyllie AH. Apoptosis: an overview. Br Med Bull 1997;53:451–465.
“Stress” Proteins
Bukau B, Horwich AL. The Hsp70 and Hsp60 chaperone machines. Cell 1998;92:351–366.
Chi NC, Karliner JS. Molecular determinants of responses to myocardial ischemia/reperfusion injury:
focus on hypoxia-inducible and heat shock factors. Cardiovasc Res 2004;61:437–447.
Knowlton AA. The role of heat shock proteins in the heart. J Mol Cell Cardiol 1995;27:121–131.
Pratt WB, Toft DO. Regulation of signaling protein function and trafficking by the hsp90/hsp70-based
chaperone machinery. Exp Biol Med 2003;228:111–133.
Ratcliffe PJ. Understanding hypoxia signaling in cells—a new therapeutic opportunity. Clin Med
2006;6: 573–578.
Sun Y, MacRae TH. The small heat shock proteins and their role in human disease. FEBS J 2005;272:
2613–2627.
P.265
Toko H, Minamoto T, Komoro I. Role of heart shock transcriptional factor 1 and heat shock proteins in
cardiac hypertrophy. Trends Cardiovasc Med 2008;18:88–93.
References
Bourne HR. Team blue sees red. Nature 1995;376:727–729.
Egan SE, Weinberg RA. The pathway to signal achievement. Nature 1993;365:781–783.
Fraga MF, Ballestar E, Paz MF, et al. Epigenetic differences arise during the lifetime of monozygotic
twins. Proc Nat Acad Sci (USA) 2005;102:10604–10609.
Goss RJ. Hypertrophy versus hyperplasia. Science 1966;153:1615–1620.
Graves LM, Bornfeldt KE, Krebs EG. Historical perspectives and new insights involving the MAP kinase
cascades. Adv Second Messenger Phosphoprotein Res 1997;31:49–61.
Jirtle RL, Skinner MK. Environmental epigenomics and disease susceptibility. Nat Rev Genet 2007;8:
253–262.
Knowlton AA, Eberli FR, Brecher P, et al. A single myocardial stretch or decreased systolic fiber
shortening stimulates the expression of the heat shock protein 70 in the isolated, erythrocyte-
perfused rabbit heart. J Clin Invest 1991;88:2018–2025.
McClintock B. The origin and behavior of mutable loci in maize. Proc Nat Acad Sci 1950;36:344–355.
Nadal-Ginard B, Kajstura J, Leri A, et al. Myocyte death, growth, and regeneration in cardiac
hypertrophy and failure. Circ Res 2003;92:139–150.
Naga Prasad SV, Duan ZH, Gupta MK, et al. Unique microRNA profile in end-stage heart failure
indicates alterations in specific cardiovascular signaling networks. J Biol Chem 2009;284:27487–
27499.
Pardee AB. G1 events and regulation of cell proliferation. Science 1989;246:603–608.
Raff MC, Barres BA, Burne JF, et al. Programmed cell death and the control of cell survival: lessons
from the nervous system. Science 1993;262:695–700.
Ring PA. Myocardial regeneration in experimental ischaemic lesions of the heart. J Path Bact
1950;62: 21–27.
Robertson KD. DNA methylation and human disease. Nat Rev Genet 2005;6:597–610.
Rodenhiser D, Mann M. Epigenetics and human disease: translating basic biology into clinical
applications. Canad Med Assn J 2007;174:341–348.
Rumyantsev PP. Interrelations of the proliferation and differentiation of processes during cardiac
myogenesis and regeneration. Int Rev Cytol 1977;51:187–273.
Saini KS, Walker NI. Biochemical and molecular mechanisms regulating apoptosis. Mol Cell Biochem
1998;178:9–25.
Savill J. Recognition and phagocytosis of cells undergoing apoptosis. Br Med J 1997;53:491–508.
Schiaffino S, Samuel JL, Sassoon D, et al. Nonsynchronous accumulation of α-skeletal actin and β-
myosin heavy chain mRNAs during early stages of pressure-overloaded-induced cardiac hypertrophy
demonstrated by in situ hybridization. Circ Res 1989;64:937–948.
Uozumi H, Hiroi Y, Zou Y, et al. gp130 plays a critical role in pressure overload-induced cardiac
hypertrophy. J Biol Chem 2001;276:23115–23119.
Zhou H, Brockington M, Jungbluth H, et al. Epigenetic allele silencing unveils recessive RYR1
mutations in core myopathies. Am J Hum Genet 2006;79:859–868.
> Table of Contents > Part Two - Signal Transduction and Regulation > Chapter 10 - Regulation of Cardiac Muscle Performance
Chapter 10
Regulation of Cardiac Muscle Performance
Regulatory Mechanisms in Cardiac and Skeletal Muscle

Muscle performance can be regulated by the four mechanisms listed in Table 10-1. To meet different
physiological and pathological challenges, these are used differently in cardiac and skeletal muscle. In
most skeletal muscles, where periods of activity alternate with periods of rest, myocyte load is
determined by the mass that must be moved, gravity, and the angles at skeletal joints. In the heart,
which contracts without pause, myocyte load is influenced by chamber size and shape as well as by
circulatory hemodynamics. This is because geometrical laws determine the relationship between myocyte
shortening in the walls of the heart and the volume of blood ejected in each beat, and because wall stress
in the heart is determined by chamber volume, as well as by wall thickness and chamber pressure (the
law of Laplace, described in Chapter 11). Regulation of cardiac muscle performance also differs from that
of most skeletal muscles because the body must meet sustained physiological stresses, such as the need
to increase cardiac output during endurance exercise, by activating the functional responses described in
Chapter 8. Adaptation of cardiac performance to the more prolonged pathological stresses caused by
chronic hemodynamic overload and heart disease is accompanied by activation of the proliferative
responses described in Chapter 9.
Summation of Contractions
The tension developed by a skeletal muscle can be increased when rapid stimulation by motor neurons
causes summation or tetanic contractions (see Chapter 6); however, these mechanisms are of little
physiological importance because most skeletal muscle contractions are brief tetani. Exceptions are the
contractions of extraocular muscles, which are often twitches, but these brief contractile responses,
which resemble the knee jerk reflex, would not be useful for most movements.
Summation plays no role in the heart because cardiac action potentials normally last almost to the end of
the active state. For this reason, an activated cardiac myocyte cannot respond to additional electrical
stimuli until relaxation is virtually complete, which makes it impossible for summation and tetanic
contraction to be used for physiological regulation.
Variations in the Number of Active Motor Units

Skeletal muscles are composed of groups of myocytes, called motor units, each of which is innervated by
a single motor neuron that operates independently of other motor units. This allows performance to be
regulated by the central nervous system, which can vary the number
P.267
of active motor units. For example, when a muscle is called upon to lift a light load, only a small fraction
of its motor units are activated, whereas a greater proportion of the motor units are activated via their
motor neurons when the same muscle lifts a heavy load. This mechanism is the major determinant of
skeletal muscle performance.
Table 10-1 Mechanisms That Regulate the Contractile Performance of Skeletal and
Cardiac Muscle
Role in
Skeletal
Mechanism Muscle Role in Cardiac Muscle
Summation of individual contractile events Minor None

(partial and complete tetanus)
Variations in number of active motor units Major None
Length-dependent changes (length-tension Usually Major in beat-to-beat

relationship, Starling's law of the heart) minor regulation, minor in sustained
responses
Ability to vary intrinsic contractile Minor Major in sustained responses,

properties minor in beat-to-beat
regulation
In the heart, the number of active muscle fibers cannot be varied because the interiors of adjacent
cardiac myocytes are linked by the gap junctions in the intercalated discs (see Chapters 1, 13, and 16).
Because these structures provide a low electrical resistance between adjacent cells, the heart operates
as a functional syncytium, where depolarization of any region normally activates the entire heart.
Changes in Sarcomere Length

The length-tension relationship allows preload to influence the performance of both skeletal and cardiac
muscle (see Chapter 6). In most skeletal muscles, rest length is determined largely by the angles at the
joints; as these angles are generally chosen to optimize leverage, rather than to set sarcomere length
along the length-tension curve, variations in rest length are of little regulatory importance.
Changes in sarcomere length play a key role in matching the heart's output during systole to the return of
blood during diastole. Because cardiac muscle normally operates on the ascending limb of the length-
tension curve (see Chapter 6), an increase in the return of blood during diastole increases the ability of
the heart to eject during systole. These length-dependent changes allow the heart to “fine tune” its
performance to match venous return and cardiac output, to vary cardiac performance in response to
short-term hemodynamic changes, and to equalize the outputs of the right and left ventricles (see
below).
P.268
Inotropic and Lusitropic Changes

Changes in contractility and relaxation play little role in skeletal muscle, where the high content of
sarcoplasmic reticulum allows excitation-contraction coupling to deliver enough calcium to saturate
virtually all of the binding sites on troponin C; as a result, even under basal conditions, skeletal muscle
tension is at or near the maximum that the muscle can generate. Although trains of stimuli normally
maximize tension by generating brief tetani, periods of inactivity allow time for skeletal muscles to relax
completely when their tasks are completed.
The situation in the heart, however, is quite different. Because excitation-contraction coupling does not
provide enough activator calcium to saturate all of the troponin C in myocardial cells operating under
basal conditions, myocardial contractility and relaxation can be regulated by modification of the calcium
cycles described in Chapter 7. These functional responses allow the heart to meet a variety of sustained
physiological stresses, such as the hemodynamic defense reactions described in Chapter 8. Furthermore,
long-term changes in the demands on the heart evoke proliferative responses that change the
architecture and composition of the myocardium (see Chapters 9 and 18).
Functional and Proliferative Responses

Cardiac performance is regulated by three fundamentally different types of response; two are functional
and the third is proliferative (Table 10-2). Changes in organ physiology represent functional responses
that appeared ∼600 million years ago with the evolution of large multicellular animals that required a
circulatory system. A second type of functional response, changes in cell biochemistry and biophysics,
emerged ∼2,000 million years ago, when the evolution of the intracellular membrane systems allowed
single-celled eukaryotes to regulate their internal environment (milieu intárieur). The third type of
response, regulation by altered molecular composition, is the most primitive, having appeared at the
dawn of life more than 3,500 million years ago. Having had the longest time to evolve, it is not surprising
that proliferative responses are the most complex.
The rates at which these responses operate differ. Physiological responses appear most rapidly, often from
beat to beat, while biochemical and biophysical responses generally require several seconds or minutes to
modify cardiac performance. Proliferative responses, which are responsible for changes in cardiac
performance in chronically overloaded and diseased hearts, are brought about by long-term structural
modifications in the heart that can take days, weeks, and even months to develop.
Table 10-2 Mechanisms That Regulate Cardiac Performance
FUNCTIONAL
Physiological: Changes in organ function, e.g., Starling's law of the heart
Biochemical and biophysical: Changes in cellular function, e.g., contractility and

relaxation
PROLIFERATIVE
Molecular: Changes in architecture and composition, e.g., reversion to the fetal phenotype
P.269
Physiological responses adjust cardiac performance to meet short-term changes in hemodynamics like an
increase in venous return, while biochemical and biophysical responses allow cardiac performance to
respond to more sustained hemodynamic challenges like those that occur during exercise. The latter
include changes in myocardial contractility and relaxation brought about by modifications of the calcium
fluxes between the cytosol, extracellular fluid, and sarcoplasmic reticulum (see Chapter 7). Molecular
responses, which depend on proliferative signaling, change the size, shape, and composition of the heart
in response to long-term challenges like chronic hemodynamic overloading and disease.
Functional Responses: Physiological and Biochemical Regulation:

Changes in End-Diastolic Volume and Myocardial Contractility
The two types of functional responses that control cardiac performance are readily understood by
comparing the effects of changing rest length and changing contractility (see Chapter 6). Starling's law of
the heart, an example of physiological regulation, allows changes in end-diastolic volume to match the
output of the heart during systole to the return of blood during diastole (Starling, 1918). This mechanism,
which had been known to physiologists since the middle of the 19th century (Katz, 2002), was generally
viewed as the only determinant of myocardial performance until the mid-1950s, when mechanisms other
than changes in end-diastolic volume were shown to regulate contractile performance (see below). The
latter, which represent changes in contractility, are initiated by biochemical and biophysical mechanisms
that modify the interactions between the heart's contractile proteins.
Regulation by changing end-diastolic volume and changing contractility were initially thought to arise
from entirely different mechanisms. This interpretation was based largely on the now-discarded view that
the length-tension relationship in cardiac muscle is caused when changes in the overlap between the thick
and thin filaments vary the number of potential interactions between actin and myosin (the
“ultrastructural mechanism” described in Chapter 6), whereas variations in contractility are brought
about by changes in the chemistry of these interactions. Even though control by changing end-diastolic
volume, like that by changing contractility, is now known to result from variations in excitation-
contraction coupling, the traditional distinction remains useful because these regulatory mechanisms
serve different physiological functions.
Regulation by Changing End-Diastolic Volume

Starling's law of the heart can be viewed as a mechanism that maintains hemodynamic equilibria by
allowing changes in preload and afterload to modify cardiac performance.
Responses to Changing Preload

Increased venous return to the right atrium, for example, when the legs are elevated, causes an increase
in right atrial pressure that is transmitted immediately across the tricuspid valve to increase right
ventricular end-diastolic volume. According to Starling's law, the higher preload restores the equilibrium
between venous return and right ventricular output by enhancing the
P.270
ability of the right ventricle to eject. This ability of the heart to modify its output in response to changes
in filling represents a positive feedback in which altered blood flow into the heart leads to a
corresponding change in the flow of blood out of the heart.
Matching the Outputs of the Right and Left Ventricles

Operation on the ascending limb of the length-tension curve (see Chapter 6) provides a positive feedback
that plays an essential role in equalizing the outputs of the two ventricles. For example, when leg
elevation increases right ventricular output (see above), the increased flow of blood through the lungs
increases venous return to the left ventricle which, according to Starling's law, increases left ventricular
ejection. Conversely, reduced ejection by the right ventricle decreases left ventricular filling, and so
reduces left ventricular output.
Responses to Changing Afterload

A sudden increase in arterial pressure (afterload), which reduces ventricular ejection (see Chapter 6),
increases the volume of blood remaining in the ventricle at the end of systole (end-systolic volume).
When the venous return during the next diastole is added to the greater end-systolic volume, end-
diastolic volume is increased which, according to Starling's law, compensates for the reduced ejection by
increasing ventricular performance. This leads to an increase in ejection that restores the equilibrium
between the amount of blood ejected during systole to the return of venous blood during diastole.
Regulation by Changing Inotropic Properties (Myocardial

Contractility or Systolic Function)
Although Starling's law plays a central role in beat-to-beat and short-term regulation of cardiac
performance, length-dependent changes are of limited importance when circulatory changes are
sustained. Evidence that factors other than end-diastolic volume regulate cardiac performance is found in
Starling's original papers, which document a positive inotropic effect, sometimes called “homeometric
autoregulation” or the “Anrep effect,” which develops after a sudden increase in left ventricular systolic
pressure. Once thought to be caused by an undiscovered mechanism within cardiac myocytes, this
increase in contractility is now recognized to occur when delayed dilatation of the coronary
microcirculation overcomes the negative inotropic effect of the state of energy starvation that occurs
when the increased systolic wall stress reduces blood flow to the left ventricle. A more important
exception to Starling's law, described more than 75 years ago, is that the heart can become smaller during
upright exercise in humans, even though cardiac output increases several fold (Nylin, 1934). By the mid-
1950s, a substantial body of evidence indicated that regulatory mechanisms other than end-diastolic fiber
length regulate cardiac performance (Katz et al., 1955), but it was not until Sarnoff described the “family
of Starling curves” (1955) that the role of changes in myocardial contractility became clear. Contractility
is often referred to as inotropy; a positive inotropic intervention increases contractility while a negative
inotropic intervention reduces contractility.
What is Myocardial Contractility?

Myocardial contractility is virtually impossible to define because it is a manifestation of all of the factors,
except for preload and afterload (see Chapter 6), that influence the interactions between the contractile
proteins. The simplest definition, the ability of the ventricle to do work
P.271
at a given preload and afterload, reflects the fact that load-dependent changes in performance do not
reflect variations in contractility, but instead are different ways that a muscle operating at a given level
of contractility can perform work. Unfortunately, this definition is of limited practical value because it is
difficult to quantify preload and afterload in patients. A change in myocardial contractility, however, is
easier to define; this is simply any change in cardiac performance that is not caused by altered end-
diastolic volume (preload) or arterial pressure (afterload). If, for example, a drug is given to an isolated
heart when end-diastolic volume and the resistance to ejection are held constant, and ejection and/or
developed pressure increases, then contractility has increased (Fig. 10-1A); conversely, a drug that
reduces the ability of this heart to eject or develop pressure has decreased contractility (Fig. 10-1B).
Even when it is possible to identify a change in myocardial contractility, there remains the question as to
what exactly has been changed. Simple answers fail because contractility is determined by all of the
factors, except for load, that influence the ability of a muscle to do work.
Fig. 10-1: Changing contractility in an isolated heart contracting with constant end-diastolic
volume and resistance to ejection. An increase in myocardial contractility (A) increases the
developed pressure, volume ejected, or both, while a decrease in myocardial contractility (B)
reduces developed pressure, the volume ejected, or both.
Force-velocity curves were once believed to hold the key to quantifying myocardial contractility, which
was thought to be the sole determinant of Vmax (see Chapter 3). This view was based on two assumptions:
that changing preload modifies only Po but not Vmax, and that the influence of afterload can be
eliminated by extrapolating the curves to zero load where shortening velocity is maximal. However, both
assumptions turned out to be incorrect. The first assumption was based on the “ultrastructural
mechanism” for the length-tension relationship, which attributed length-dependent variations in the
ability of the heart to do work entirely to changes in the number of active cross-bridges. Today, however,
several determinants of contractile protein shortening are also known to be modified by changes in
sarcomere length (see Chapter 6), which means that altering preload can modify Vmax as well as Po. The
second assumption, that sarcomere shortening is the only determinant of Vmax, is also incorrect because
of the high resting tension and internal elasticity of cardiac muscle (see Chapter 6) and the spiral
arrangement of the fibers in the walls of the heart (see Chapter 1), all of which contribute elasticities
that modify both tension and shortening velocity as active state increases and
P.272
decreases during each cardiac cycle. Furthermore, those portions of the force-velocity curves that can be
measured in beating hearts are not hyperbolic because the active state in cardiac muscle changes
throughout systole (see Chapter 6); this means that Po and Vmax cannot be accurately extrapolated in
cardiac muscle.
Even though efforts to quantify myocardial contractility are of limited accuracy, the concept has
considerable clinical significance. This is because any effort to determine what might have happened to a
patient who experiences an unexpected change in hemodynamics, such as a fall in blood pressure,
requires an understanding of the potential contributions of changes in preload, afterload, and
contractility. It is only in this way that a therapeutic regimen most appropriate for the individual patient
can be formulated and its effects evaluated.
Regulation by Changes in Lusitropic Properties (Filling or Diastolic

Function)
The clinical importance of lusitropic abnormalities was not generally appreciated until the 1970s, almost
20 years after the role of changing contractility had been recognized. This is largely because the initial
clinical applications of hemodynamic physiology focused on pressures, which were readily measured by
the equipment that was then available. These led to the early use of pressure-based indices, such as the
rate of pressure rise during isovolumic contraction (+dP/dt), to quantify myocardial contractility (see
Chapter 12). Initial efforts to quantify lusitropic state were also based on pressure measurements; in this
case the rate of pressure fall during isovolumic relaxation (-dP/dt). However, the latter is highly
dependent on ventricular systolic pressure, so that use of pressure measurements to define lusitropic
properties requires corrections that are both difficult and imprecise. It was not until echocardiography
and nuclear techniques made it possible to measure changes in the rate and extent of ventricular filling
that diastolic function could be characterized in patients (see Chapter 12).
Regulation by Changes in Coronary Perfusion Pressure (The Garden

Hose Effect)
The finding that changing coronary perfusion pressure can modify the ability of the heart to develop
tension, even after the ventricles are drained of blood and cavity pressure is zero (Salisbury et al., 1960),
revealed the existence of an unexpected regulatory mechanism. Evidence that increasing coronary
arterial pressure has an “erectile” effect that increases cardiac oxygen consumption (Vogel et al., 1982)
demonstrated that the heart's performance is regulated by distension within its walls, much as occluding
the outlet of a garden hose increases its internal diameter. This mechanism, often called the garden hose
effect, was initially explained by postulating that higher intramyocardial pressures increase cardiac
performance by increasing sarcomere length (Starling's law of the heart). However, evidence that this
phenomenon cannot be explained simply on the basis of changes in sarcomere length (Kitakaze and
Marban, 1989; Koretsune et al., 1991; May-Newman et al., 1994; Matsushita et al., 1995) suggests that the
garden hose effect occurs when cytoskeletal deformation modifies the intensity of excitation-contraction
coupling.
P.273
Functional Mechanisms that Regulate Myocardial Contractility and

Relaxation
Most functional changes in contractility and relaxation are initiated by signals that modify calcium fluxes
into and out of the cytosol, and the calcium affinity of the contractile proteins. These physiological
mechanisms generally result from posttranslational changes in the membrane proteins that participate in
excitation-contraction coupling and relaxation, and the contractile proteins. The following discussion
centers on the schematic depiction of the extracellular and intracellular calcium cycles shown in Figure 7-
18.
Regulation by Changing Calcium Fluxes across the Plasma

Membrane
Plasma Membrane L-Type Calcium Channels

Among the most important determinants of myocardial contractility are variations in the amount of
calcium that enters cardiac myocytes from the extracellular fluid via L-type calcium channels
(“dihydropyridine receptors”). This calcium serves three major functions: it binds to the contractile
proteins; it triggers the opening of the intracellular calcium release channels; and it contributes to the
filling of calcium stores within the sarcoplasmic reticulum (Eisner et al., 2000).
The positive inotropic effects of β-adrenergic agonists and phosphodiesterase inhibitors, both of which are
mediated by cyclic AMP, are due in part to phosphorylation of L-type calcium channels by cyclic AMP-
dependent protein kinases (PKA). Because phosphorylation increases the probability of channel opening
(see Chapter 13), more calcium enters the cytosol (larger arrow A in Fig. 10-2) to cause the positive
inotropic response. Conversely, the negative inotropic effect of L-type calcium channel blockers results
from direct inhibition of channel opening (smaller arrow A in Fig. 10-3). β-Adrenergic receptor blockers
also reduce contractility by reducing calcium channel opening, but in this case the inhibition is indirect
and occurs when reduced basal sympathetic activity decreases the probability of L-type calcium channel
opening (see above). Calcium entry is reduced in ischemic and failing hearts, where energy starvation
decreases ATP concentration and attenuates an allosteric effect of this nucleotide that facilitates calcium
channel opening (see below).
The Positive (Bowditch) Staircase

Bowditch (1871) was the first to observe that the pressure developed by the heart increases when the
frequency of stimulation is increased, and decreases when stimulation frequency is slowed (Fig. 10-4).
The stepwise increase in tension seen at the faster rate, called the positive staircase or treppe (the latter
is the German word for staircase), is a manifestation of rate-dependent variations in contractility, known
collectively as the force-frequency relationship. These rate-dependent changes do not violate the “all-or-
none law” (also first described by Bowditch), which states that the magnitude of a response is
independent of the intensity of the stimulus, because the changes in contractility are caused by altered
stimulation frequency, not intensity. The positive staircase occurs when more frequent openings of
plasma membrane calcium channels at higher stimulation frequencies increase calcium influx into cytosol
(Fig. 10-2).
P.274
Fig. 10-2: Increased calcium entry through L-type plasma membrane calcium channels (larger arrow
A) contributes to the positive staircase, post-extrasystolic potentiation, and the inotropic response
caused by sympathetic stimulation. In the staircase, more rapid stimulation increases the frequency
of calcium channel openings, while post-extrasystolic potentiation occurs when premature plasma
membrane depolarization prolongs calcium channel opening. The positive inotropic effect of
sympathetic stimulation is due in part to phosphorylation of L-type calcium channels by PKA, which
increases the probability of channel opening.
Post-Extrasystolic Potentiation
Contractility is generally increased in contractions that follow a premature systole (often referred to as
an “extrasystole”; see Chapter 16); this positive inotropic effect, called post-extrasystolic potentiation,
is among the most intense seen in cardiac muscle (Fig. 10-5). The magnitude of the increase in
contractility is not correlated with the pressure developed by the premature systole; in fact, a large post-
extrasystolic potentiation can occur when the premature beat comes so early as merely to delay
relaxation of the preceding normal beat.
P.275
Fig. 10-3: Calcium entry can be reduced by inhibition of L-type calcium channel opening (smaller
arrow A) both directly (e.g., by calcium channel blockers) and indirectly (e.g., by β-adrenergic
receptor blockers).
Fig. 10-4: The positive (Bowditch) staircase. The pressure that can be developed by an isolated
heart increases in a stepwise manner after stimulation frequency is increased, and decreases when
the frequency is reduced.
P.276
Fig. 10-5: Post-extrasystolic potentiation. The pressure developed after a premature systole (often
called an “extrasystole”) is increased by calcium that enters the cytosol during the premature
depolarization.
Post-extrasystolic potentiation is caused when increased calcium entry through plasma membrane calcium
channels during the premature depolarization is added to the calcium stores within the sarcoplasmic
reticulum. Unlike the positive staircase, where filling of these stores is increased by more frequent
channel openings, post-extrasystolic potentiation occurs when premature stimuli increase the duration of
calcium channel opening. This mechanism was elucidated by Wood et al. (1969), who varied membrane
potential during the absolute refractory period by applying small currents through an intracellular
electrode (Fig. 10-6). Because these currents occurred during the absolute refractory period, when heart
cannot generate an action potential (see Chapter 14), they did not affect tension development during the
beat in which they were applied. However, the currents affected tension developed in the following
systole;
P.277
depolarizing currents that caused membrane potential to become more negative increased tension
developed in the subsequent beat (1, Fig. 10-6), whereas hyperpolarizing currents that shifted membrane
potential toward the resting level reduced tension in the subsequent beat (2, Fig. 10-6). These findings
reflect the ability of the small changes in membrane potential to modify calcium channel opening, which
alters the amount of calcium available for release in the subsequent contraction: the small depolarization
increases calcium stores by prolonging calcium channel opening, while reducing membrane potential
decreases calcium stores by accelerating calcium channel closing.
Fig. 10-6: Effects of changing membrane potential during the action potential plateau on contractile
performance in the following beat. Upper curves: Depolarizing (1) and hyperpolarizing (2) currents applied
during the plateau of the cardiac action potential. Lower curves: Although the applied currents have no
effect on tension developed during the contraction in which the current is applied, the tension developed in
the subsequent contractions is increased by the depolarizing current (1), and decreased by the
hyperpolarizing current (2).

The sodium/calcium (Na/Ca) exchanger, which exchanges sodium and calcium ions in both directions
across the plasma membrane (see Chapter 7), has important effects on both myocardial contractility and
membrane potential. The ability of the exchanger to allow changes in extracellular and intracellular
sodium and calcium concentrations to modify cytosolic calcium content explains why myocardial
contractility is directly proportional to the ratio between extracellular sodium and calcium. Increased
calcium influx via the exchanger accounts for the ability of increased extracellular calcium or decreased
extracellular sodium to increase contractility (larger arrow B2 in Fig. 10-7); conversely, the negative
inotropic effects of decreased extracellular calcium and increased extracellular sodium are caused by
reduced calcium influx. Exchange of each divalent calcium ion in the cytosol for three univalent sodium
ions in the extracellular fluid also generates depolarizing currents (see Chapter 14).
The Sodium Pump

The sodium pump has an important indirect effect in regulating myocardial contractility. Inhibition by
cardiac glycosides (see Chapter 7) reduces sodium efflux and increases intracellular sodium
concentration; this increases intracellular calcium stores by two mechanisms, both of which involve the
Na/Ca exchanger (see above). The first occurs when increased cytosolic sodium reduces calcium efflux by
competing with calcium at the intracellular site of the exchanger (smaller outward arrow B2 in Fig. 10-8);
the second occurs when some of the extracellular sodium that would have been exchanged for
extracellular potassium by the sodium pump is, instead, exchanged for extracellular calcium (larger
inward arrow B2 in Fig. 10-9). Both responses increase intracellular calcium stores, which increases
contractility.
Although sodium pump inhibition increases contractility by reducing calcium efflux and increasing calcium
influx via the Na/Ca exchanger, this response has several adverse consequences. These include the ability
of increased cytosolic calcium to impair relaxation, and a decrease in resting membrane potential that is
caused when reduced potassium influx decreases the Nernst potential for potassium (see Chapter 14).
This provides a substrate for arrhythmias when the decreased resting potential inactivates sodium channel
opening, which slows impulse conduction (see Chapter 16).

Protons generated during anaerobic energy production are transported out of the cytosol in exchange for
sodium by the Na/H exchanger (see Chapter 7). The resulting increase in intracellular sodium has effects
similar to those caused by sodium pump inhibition (see above) that include a positive inotropic response
that is caused by reduced calcium efflux and increased calcium influx via the Na/Ca exchanger.
P.278
Fig. 10-7: Both increased extracellular calcium and decreased extracellular sodium cause a positive
inotropic effect by increasing calcium entry via the Na/Ca exchanger (larger inward arrow B2).
The Plasma Membrane Calcium Pump

The plasma membrane calcium pump, which like the Na/Ca exchanger transports calcium out of the
myocardial cell, is stimulated by calcium-calmodulin-dependent protein kinase (see Chapter 7). By
increasing calcium flux out of the cytosol (larger arrow B1 in Fig. 10-10), this response helps avoid
calcium overload, but also has a negative inotropic effect.
Regulation by Changing Calcium Fluxes across the Sarcoplasmic

Reticulum Membrane
The sarcoplasmic reticulum directly influences both contraction and relaxation in adult human hearts.
The most important determinant of myocardial contractility is the amount of calcium
P.279
released from this internal membrane system, while relaxation depends on the calcium affinity and
turnover rate of the sarcoplasmic reticulum calcium pump (see Chapter 7).
Fig. 10-8: Cardiac glycosides inhibit the sodium pump, which reduces the amount of sodium
exchanged for potassium (thin outward dotted arrow). The resulting increase in cytosolic sodium
concentration inhibits calcium efflux via the Na/Ca exchanger (curved dotted line and smaller
outward arrow B2).
Calcium Release
Calcium is released from the cardiac sarcoplasmic reticulum through intracellular calcium release
channels (“ryanodine receptors”) in the dyads (see Chapter 1). A decrease in the calcium content of the
sarcoplasmic reticulum, slowing of calcium efflux from this membrane system, and attenuation of the
calcium signal that initiates this calcium release have negative inotropic effects (smaller arrow C in Fig.
10-11). Attenuation of the allosteric effect of high ATP concentration that facilitates opening of calcium
release channels (see Chapter 7) also
P.280
reduces the opening of these channels, which contributes to the decreased contractility in energy-starved
hearts. Conversely, increased calcium stores, more rapid calcium release, and greater influx of “trigger”
calcium have positive inotropic effects.
Fig. 10-9: Cardiac glycosides inhibit the sodium pump, which reduces the amount of sodium
exchanged for potassium (thin dotted line). The resulting increase in cytosolic sodium concentration
increases sodium efflux via the Na/Ca exchanger (curved dotted arrow and larger outward dashed
arrow), which increases calcium influx (larger inward arrow B2).
Calcium Uptake
The concentration of ionized calcium in the cytosol is the most important physiological regulator of
calcium uptake into the sarcoplasmic reticulum. Acceleration of the calcium pump (SERCA) by increased
cytosolic calcium plays a key role in determining internal calcium stores
P.281
and helps match the amount of calcium taken up during diastole to that released during systole.
Fig. 10-10: Acceleration of the plasma membrane calcium pump by calcium-calmodulin-dependent
protein kinase reduces intracellular calcium stores by increasing calcium transport out of the cell
(larger outward arrow B1).
Physiological regulation of this calcium pump by cyclic AMP allows β-adrenergic agonists, which stimulate
the synthesis of this intracellular second messenger, and phosphodiesterase inhibitors, which inhibit cyclic
AMP breakdown, to increase both contractility and relaxation. These effects of cyclic AMP are mediated
by PKA-catalyzed phosphorylation of phospholamban, a regulatory protein that in its dephospho form
inhibits the pump (see Chapter 7). Reversal of this effect by phospholamban phosphorylation increases the
calcium sensitivity of the pump, which stimulates calcium uptake into the sarcoplasmic reticulum (larger
arrow D in Fig. 10-12). This response has a direct lusitropic effect, and increasing the amount of calcium
retained in this intracellular membrane system also has a positive inotropic effect.
P.282
Fig. 10-11: Calcium efflux via calcium release channels in the sarcoplasmic reticulum (smaller arrow
C) is reduced when intracellular calcium stores are depleted, when less “trigger” calcium enters via
the plasma membrane calcium channels, and when calcium flux through these channels is slowed.
Energy starvation impairs relaxation by reducing the allosteric effect of ATP that stimulates calcium
uptake by the sarcoplasmic reticulum (smaller arrow D in Fig. 10-13). The calcium pump is also inhibited
when the free energy released during ATP hydrolysis is reduced by the fall in ATP and accompanying rise
in ADP concentration (see below).
The Negative (Woodworth) Staircase

A few decades after Bowditch described the positive staircase, Woodworth (1902) observed another
staircase phenomenon in which increased stimulation frequency leads to a decrease in the pressure
developed by the heart, and decreased stimulation frequency causes pressure to increase (Fig. 10-14).
This feature of the force-frequency relationship, which is most prominent
P.283
at higher stimulation frequencies, is often called the negative or Woodworth staircase. Another
manifestation of the negative staircase, sometimes called the recuperative effect of a pause, is an
increase in the pressure developed by the first beat after a pause (Fig. 10-14).
Fig. 10-12: Activation of the calcium pump of the sarcoplasmic reticulum by phosphorylation of
phospholamban (larger arrow D) favors relaxation by increasing the rate and extent of calcium
uptake from the cytosol. The increased uptake of calcium into the sarcoplasmic reticulum also has
a positive inotropic effect.
The decrease in contractility that follows an increase in stimulation frequency at very high rates of
stimulation is due largely to the shortened diastole, which reduces the calcium stores in the sarcoplasmic
reticulum by abbreviating the time during which activator calcium can be taken up by these internal
membranes. Other mechanisms, such as slow diffusion of calcium from the sarcotubular network to the
calcium release channels in the subsarcolemmal cisternae and incomplete reactivation of the calcium
release channels following the short diastoles, do not appear to play an important role in the negative
staircase.
P.284
Fig. 10-13: A decreased rate of calcium uptake into the sarcoplasmic reticulum (smaller arrow D),
for example, by dephosphorylation of phospholamban, decreased ATP concentration, acidosis, or a
reduced number of calcium pump ATPase molecules, impairs relaxation. By decreasing intracellular
calcium stores, these interventions also reduce contractility.
The negative (Woodworth) and positive (Bowditch) staircases can appear together when stimulation
frequency is changed (Fig. 10-15). The negative staircase, which evolves more rapidly, causes an initial
fall in pressure when stimulation frequency is increased (1, Fig. 10-15), and a transient rise in pressure
when heart rate is decreased (3, Fig. 10-15); the latter is an example of the recuperative effect of a
pause. The initial fall in pressure caused by the negative staircase is overwhelmed in subsequent
contractions by the more slowly developing positive staircase (2, Fig. 10-15); conversely, the initial
increase in pressure seen when stimulation frequency is decreased is followed by a fall in pressure as the
more slowly evolving positive staircase wears off (4, Fig. 10-15). These force-frequency relationships are
altered in failing hearts, where the positive staircase is attenuated and may disappear completely.
P.285
Fig. 10-14: The negative (Woodworth) staircase. The pressure that can be developed by the heart
decreases in the first few beats after stimulation rate is increased, and augmented when
stimulation is slowed. Another manifestation of the negative staircase is the recuperative effect of
a pause, which increases pressure in the first beat after stimulation is interrupted.
Regulation by Changes in the Contractile Proteins

Cardiac performance can be regulated by modifications of the rate of cross-bridge cycling and changes in
the calcium sensitivity of the contractile proteins. Among the most important physiological mechanisms is
the response of the heart to sympathetic stimulation, which is mediated by cyclic AMP-dependent protein
kinases (PKA). These enzymes catalyze myosin light chain phosphorylations that increase contractility by
accelerating cross-bridge cycling, and troponin I phosphorylation that decreases the calcium affinity of
the contractile proteins (smaller arrow E and larger arrow F in Fig. 10-16). The latter facilitates
relaxation, which helps the heart to fill at the higher heart rates caused when sympathetic stimulation
accelerates the SA node pacemaker (see Chapter 14).
The negative inotropic response caused when sympathetic stimulation decreases the calcium affinity of
the troponin complex is outweighed by the positive inotropic effects of increased opening of L-type
calcium channels, which augments the signal that activates the calcium release channels in the
sarcoplasmic reticulum, and accelerated calcium uptake into the cardiac sarcoplasmic reticulum, which
increases the amount of calcium stored in these internal membranes (see above).
Angiotensin II, α1-adrenergic stimulation, and endothelin cause a weak positive inotropic response by
activating signaling pathways that release diacylglycerol (DAG) and inositol
P.286
trisphosphate (InsP3) (see Chapter 8). The latter increases cytosolic calcium levels by opening InsP3-gated
intracellular calcium release channels (see Chapter 9), while DAG activates protein kinase C-catalyzed
phosphorylations that increase the calcium affinity of troponin (larger arrow E and smaller arrow F in Fig.
10-17). This response causes a positive inotropic effect that does not depend on an increase in cytosolic
calcium, and so has the theoretical advantage of avoiding the arrhythmogenic consequences associated
with an increase in the electrogenic calcium efflux via the Na/Ca exchanger (see Chapters 16 and 18).
However, increasing the calcium sensitivity of the contractile proteins has the disadvantage of impairing
relaxation.
Fig. 10-15: Simultaneous operation of the positive and negative staircases. The initial response to
an increase in stimulation frequency is a transient fall in pressure (1, the negative staircase), which
is then overcome by a rise in pressure caused by the more slowly evolving positive staircase (2).
Conversely, the initial response to a decrease in stimulation frequency is a transient rise in pressure
(3, the negative staircase), followed by a fall in pressure as the positive inotropic effect of the
positive staircase wears off (4).
Fig. 10-16: Decreased calcium affinity of the troponin complex, shown as a smaller association arrow
(E) and larger dissociation arrow (F), facilitate relaxation by favoring dissociation of this ion from its
binding site on the contractile proteins.
P.287
Fig. 10-17: Increased calcium affinity of the troponin complex, shown as a larger association arrow
(E) and smaller dissociation arrow (F), increases contractility by increasing the binding of activator
calcium to troponin.
Regulation by Extracellular Potassium, ATP, and Acidosis
In 1975, when the first edition of this text was written, the ability of elevated extracellular potassium to
depress myocardial contractility was believed to play an important physiological role in regulating cardiac
performance. However, the negative inotropic effect of high extracellular potassium is now known to be
the result of a decrease in resting potential across the plasma membrane (see Chapter 14), an indirect
effect that does not play a physiological role in regulating myocardial contractility.
The role of ATP in providing energy for cardiac contraction (see Chapter 7) led early workers to suggest
that changes in ATP concentration could directly depress myocardial contractility
P.288
in energy-starved hearts. However, the effects of energy starvation are more complex because decreased
ATP concentration has many effects on the heart. The most obvious, lack of ATP for binding to the
substrate-binding sites on myosin (see Chapter 4), does not play a physiological role because these
substrate sites are saturated at ATP concentrations around 1 µM, while cytosolic ATP levels are normally
greater than 1 mM. Furthermore, lack of ATP for binding to these substrate sites does not reduce
contractility, but instead increases diastolic stiffness and causes rigor (see Chapter 4). For these reasons,
loss of the substrate effects of ATP is seen only in dying hearts, where complete hydrolysis of ATP leads to
ischemic contracture (see Chapter 17). Attenuation of the allosteric effects of high ATP concentration
(Table 10-3) has little effect on contractility, but increases diastolic stiffness by inhibiting the dissociation
of actin and myosin (see Chapter 4). Added to this negative lusitropic effect is attenuation of the
allosteric effect of ATP that activates ion pumps, ion exchangers, and passive ion fluxes through
membrane channels (see above). However, the most important effects of ATP depletion are due to a
decrease in the free energy of ATP hydrolysis (-ΔG), which reduces the ATP/ADP ratio that determines the
energy that can be made available by hydrolysis of the terminal phosphate bond in ATP. Because the ADP
concentration in cardiac myocytes is much less than that of ATP (see Chapter 2), even a slight fall in ATP
concentration causes a disproportional increase in ADP concentration. The initial effects of a fall in the
free energy of ATP hydrolysis are slowing of the calcium pump of the sarcoplasmic reticulum (Tian and
Ingwall, 1996) and cross-bridge cycling (Tian et al., 1997a,b).
Table 10-3 Effects of Diminished Allosteric Effects of ATP on Myocardial Contraction

and Relaxation
Process Immediate Consequence Mechanical Effect
Actin–myosin Loss of “plasticizing” effect, reduced dissociation of Negative

interactions thick and thin filaments lusitropic,
contracture
Plasma Reduced Ca influx into the cytosol Negative

membrane Ca inotropic
channels
Plasma Reduced Ca efflux from the cytosol, increased Negative

membrane Ca intracellular Ca lusitropic,
pump contracture
Plasma Reduced Na efflux, increased intracellular Na, Negative

membrane Na decreased Ca efflux by the Na/Ca exchanger, lusitropic,
pump increased intracellular Ca contracture
Sarcoplasmic Reduced Ca release during systole, less Ca release for Negative

reticulum Ca binding to contractile proteins inotropic
channels
Sarcoplasmic Reduced Ca uptake during diastole, reduced Ca Negative

reticulum Ca removal from contractile proteins lusitropic,
pump contracture
Acidosis has a profound negative inotropic effect that is due in part to a shift in the calcium sensitivity of
tension development to higher calcium concentrations. This is caused by a
P.289
competition between protons and calcium for the high-affinity calcium-binding sites on troponin. Acidosis
also inhibits excitation-contraction coupling and relaxation because the increased concentration of
protons inhibits most of the pumps, channels, and exchangers that mediate calcium fluxes into and out of
the cytosol (see Chapter 7).
Regulation of Myocardial Contractility by β-Adrenergic

Stimulation: An Integrated Functional Response
Sympathetic stimulation, the most important mediator of the hemodynamic defense reaction, activates
both functional and proliferative signaling pathways in the heart (Table 10-4). The most important
functional responses result from cyclic AMP-stimulated phosphorylations that favor both contraction (e.g.,
increased calcium entry through plasma membrane calcium channels and greater calcium stores in the
sarcoplasmic reticulum) and relaxation (e.g., accelerated calcium uptake into the sarcoplasmic reticulum
and decreased calcium affinity of troponin). Although these responses might, at first glance, appear to
oppose one another, they are components of an integrated response that increases myocardial
contractility and allows the heart to relax completely at faster heart rates.
The response of cardiac myocytes to β-adrenergic stimulation includes increases in the rates of rise of
cytosolic calcium and tension, but both follow abbreviated time courses (Fig. 10-18). These changes are
the result of increases in virtually all of the calcium fluxes involved in excitation-contraction coupling and
relaxation, along with a decrease in the calcium affinity of troponin. The positive inotropic effect is
caused by increased calcium influx across the plasma membrane, which is due mainly to phosphorylation
of the L-type calcium channels, and an increase in the amount of calcium stored in the sarcoplasmic
reticulum; the latter results from both the increased calcium influx into the cytosol via L-type calcium
channels, and the stimulation of calcium uptake into the sarcoplasmic reticulum that results from
phospholamban phosphorylation (see above). Causes of the positive lusitropic effect include troponin I
phosphorylation, which facilitates calcium dissociation from the troponin complex, and the increased rate
of calcium uptake into the sarcoplasmic reticulum caused by phosphorylation of phospholamban. The
positive inotropic effect occurs because the larger amount of calcium that enters the cytosol overcomes
the ability of the facilitated dissociation of calcium from troponin to reduce contractility. However,
relaxation must also be accelerated because sympathetic stimulation increases heart rate. For this
reason, the lusitropic effects that allow the larger amount of activator calcium that enters the cytosol to
be pumped back into the sarcoplasmic reticulum during the shortened diastolic interval are an essential
component of the heart's response to sympathetic stimulation.
Proliferative Responses that Regulate Myocardial Contractility and

Relaxation
The functional responses described above help the heart meet short-term challenges like exercise and
hemorrhage, but rarely persist for more than a few hours or days (see Chapter 8). In chronically
overloaded and failing hearts, where the challenges last much longer, adaptation to stress relies on an
entirely different group of proliferative responses. The examples described below illustrate a few of the
many mechanisms by which proliferative signaling modifies contractility and relaxation.
P.290
Table 10-4 Major Responses of Heart Muscle to β-Adrenergic Stimulation
Response Cellular Effect Major Physiological Role
Increased energy production
Accelerated glycogenolysis Increased ATP regeneration Energy provision for

increased work
Phosphorylation of plasma membrane Ca channels
Atria and ventricles Increased Ca entry, Increased ejection

increased contractility
SA node Accelerated heart rate Increased cardiac

output
AV node Accelerated conduction Maintained AV

velocity conduction
Phosphorylation of the Na pump

Increased Na efflux Increased Ca influx via Increased ejection
Na/Ca exchanger
Phosphorylation of phospholamban
Increased Ca sensitivity of Increased Ca uptake into Increased filling

the SR Ca pump the SR
Increased Ca stores in the Increased ejection

SR
Increased SR Ca pump Increased Ca uptake into Increased filling

activity the SR
Increased Ca stores in the Increased ejection

SR
Phosphorylation of troponin Decreased Ca affinity of Increased filling

I troponin C
Phosphorylation of myosin Accelerated cross-bridge Increased ejection

cycling
Hypertrophy More sarcomeres Increased ejection
Myocyte elongation Progressive dilatation
SR, sarcoplasmic reticulum; SA, sinoatrial; AV, atrioventricular.
Proliferative Responses Involving the Contractile Proteins

One of the first observations suggesting a role for abnormal proliferative signaling in diseased hearts was
published by Alpert and Gordon (1962), who reported a reduction in the ATPase activity of myofibrils
isolated from failing human hearts. This abnormality, which was immediately recognized as a possible
explanation for the depressed myocardial contractility generally
P.291
seen in these patients, is now known to be due in part to isoform shifts involving the myosin heavy chains.
Fig. 10-18: Increased cellular cyclic AMP levels caused by β-adrenergic stimulation increase the
amount of calcium that enters and leaves the cytosol during systole (dotted lines). These effects
increase maximum tension development and shorten the duration of the contractile response (solid
line). (A) Basal conditions; (B) β-adrenergic stimulation.
The pathological hypertrophy caused by a chronic hemodynamic overload is accompanied by replacement

of the high ATPase α-myosin heavy chain isoform normally found in adult hearts with the slower β-isoform
that predominates in the embryonic heart; this isoform shift is an example of reversion to the fetal
phenotype (see Chapters 9 and 18). The opposite response, replacement of slow with fast myosin heavy
chains, accompanies the physiological hypertrophy seen in the “athlete's heart” (Scheuer and Buttrick,
1985). Pathological hypertrophy is also accompanied by isoform shifts in troponin I, which cause allosteric
effects that modify calcium binding by troponin C, along with changes in many other sarcomeric,
cytoskeletal, and membrane proteins (see Chapter 18).
The changes in the cardiac contractile proteins that accompany overload-induced pathological
hypertrophy are part of a highly regulated proliferative response in which the appearance of fetal
isoforms of many, but not all, cellular proteins follows different time courses (Izumo et al., 1988). The
ability of pressure overload and volume overload, like physiological and pathological hypertrophy (see
above), to initiate different molecular phenotypes (Calderone et al., 1995) demonstrates that different
types of overload initiate different proliferative responses, probably by activating load-specific
cytoskeletal signaling pathways. Proliferative signaling is also modified in chronic endocrinopathies; for
example, hyperthyroidism increases the content of the high-ATPase myosin heavy chain isoform, which
aids in the adaptation to the rapid heart rate and low peripheral resistance caused by excessive thyroid
hormone release.
Proliferative Responses Involving the Plasma Membrane and
Several membrane proteins are changed by chronic hemodynamic overloading. The contents of the
sarcoplasmic reticulum calcium pump ATPase and phospholamban are decreased in overloaded and failing
hearts, as are those of the calcium release channels and sodium pump ATPase.
P.292
However, the content of the Na/Ca exchanger increases while the number of L-type calcium channels
appears not to change (see Chapter 18). Together, these molecular changes slow the intracellular calcium
cycle (smaller arrows C, F, and E in Fig. 10-19), which contributes to the depressed contractility and
impaired relaxation generally seen in failing hearts (see Chapter 18).
Fig. 10-19: Reduction of intracellular calcium cycling, which occurs when reversion to the fetal
phenotype in chronically overloaded and failing hearts, reduces the densities of the calcium
release channels (smaller arrow C) and the calcium pump ATPase (smaller arrow D).
Conclusion
The regulatory mechanisms described in this chapter allow the heart to adjust its performance in
response to the changing demands of the circulation. Like the singers and instrumentalists in an opera,
the functional responses to most short-term physiological challenges signals operate in harmony.
Unfortunately, however, the proliferative signals that are initiated by
P.293
long-term pathological challenges, such as chronic overload and heart disease, often lead to cacophonic
responses that worsen patient outcome.
References
Alpert NR, Gordon MS. Myofibrillar adenosine triphosphatase activity in congestive heart failure. Am
J Physiol 1962;202:940–946.
Bowditch HP. áber die Eigenthámlichkeiten der Reizbarkeit, welche die Musklefasern des Herzens
zeigen. Berichte der Kán Sáchs Gesellschaft der Wissenschaften Mathematisch-Physische Classe
1871;23: 652–689.
Calderone A, Takahashi N, Izzo NJ Jr, et al. Pressure- and volume-induced left ventricular
hypertrophies are associated with distinct myocyte phenotypes and differential induction of peptide
growth factor mRNAs. Circulation 1995;92:2385–2390.
Eisner DA, Choi HS, Dáaz ME, et al. Integrative analysis of calcium cycling in cardiac muscle. Circ
Res 2000;87:1087–1094.
Izumo S, Nadal-Ginard B, Mahdavi V. Protooncogene induction and reprogramming of cardiac gene

expression produced by pressure overload. Proc Nat Acad Sci USA 1988;85:339–343.
Katz AM. Ernest Henry Starling, his predecessors, and the “Law of the Heart.” Circulation 2002;106:
2986–2992.
Katz LN, Katz AM, Williams FL. Metabolic adjustments to alterations of cardiac work in hypoxemia.
Am J Physiol 1955;181:539–549.
Kitakaze M, Marban E. Cellular mechanisms for the modification of contractile function by coronary
perfusion pressure in ferret hearts. J Physiol (Lond) 1989;414:455–472.
Koretsune Y, Corretti MC, Kusuoka H, et al. Mechanism of early ischemic contraction failure.
Inexcitability, metabolite accumulation or vascular collapse. Circ Res 1991;68:255–262.
Matsushita T, Takaki M, Fujii W, et al. Left ventricular mechanoenergetics under altered coronary
perfusion in guinea pig hearts. Japan Circ J 1995;45:991–1004.
May-Newman K, Omens JH, Pavelec RS, et al. Three-dimensional transmural mechanical interaction
between the coronary vasculature and passive myocardium in the dog. Circ Res 1994;74:1166–1178.
Nylin G. The relation between heart volume and stroke volume in recumbent and erect positions.
Skand Arch Physiol 1934;69:237–246.
Salisbury PF, Cross CA, Rieben PA. Influence of coronary artery pressure upon myocardial elasticity.
Circ Res 1960;8:794–800.
Sarnoff SJ. Myocardial contractility as described by ventricular function curves: observations on

Starling's Law of the Heart. Physiol Rev 1955;35:107–122.
Scheuer J, Buttrick P. The cardiac hypertrophic response to pathologic and physiologic loads.
Circulation 1985;75(Suppl I, Pt 2):I-63–I-68.
Starling EH. The Linacre Lecture on the Law of the Heart. London: Longmans, Green and Co, 1918.
Physiol 1996;270:H1207–H1216.
Tian R, Nascimben L, Ingwall JS, et al. Failure to maintain a low ADP concentration impairs diastolic
function in hypertrophied rat hearts. Circulation 1997a;96:1313–1319.
Tian R, Christe ME, Spindler M, et al. Role of MgADP in the development of diastolic dysfunction in
the intact beating rat heart. J Clin Invest 1997b;99:745–751.
Vogel WM, Apstein CS, Briggs LL, et al. Acute alterations in left ventricular diastolic chamber
stiffness. Role of the “erectile” effect of coronary arterial pressure and flow in normal and damaged
hearts. Circ Res 1982;51:465–476.
Wood EH, Hepner RL, Weidmann S. Inotropic effects of electrical currents. Circ Res 1969;24:409–
445.
Woodworth RS. Maximal contraction, “staircase” contraction, refractory period, and compensatory
pause of the heart. Am J Physiol 1902;8:213–249.
> Table of Contents > Part Three - Normal Physiology > Chapter 11 - The Heart as a Muscular Pump
Chapter 11
The Heart as a Muscular Pump
Unlike the contraction of a skeletal muscle, which is characterized by changes in tension and length, the beating heart
generates pressure and ejects a volume of blood. For this reason, when describing the work of this hollow muscular
structure, changes in length and tension must be redefined as changes in volume and pressure (Fig. 11-1).
Fiber Length and Chamber Volume: A Geometric Relationship

The relationship between length and volume is determined by the laws of geometry. In a sphere, for example, volume
is defined by the equation:
where V = volume and R = radius. Volume is also related to the third power of circumference, so that doubling
circumference increases volume eightfold (23). Although the left ventricle is not a sphere, its shape can be
approximated as an ellipsoid with three diameters: anterior-posterior diameter (DA), lateral diameter (DL), and
maximal length (LM) (Fig. 11-2). These can be used to estimate left ventricular volume according to the following
equation that describes the volume of an ellipse:
Wall Stress and Chamber Pressure: The Law of Laplace

The pressure developed within a chamber of the beating heart is determ ined by the stress within the chamber walls,
the volume of the chamber, and wall thickness. Although the terms tension and stress are sometimes used
interchangeably, this is incorrect; tension describes a force exerted along a line (e.g., dynes/cm), whereas stress is a
force exerted across an area (e.g., dynes/cm2). The pressure within the chamber, like the stress developed by the
chamber walls, has the units dynes/cm2 because it is a force exerted across an area. The difference is that wall stress
is a force that parallels the circumference of the walls, whereas pressure is a distending force that is exerted at right
angles to the walls. Pressures are usually described as millimeters of mercury (mm Hg) or centimeters of water (cm
H2O); these are the gravitational forces exerted by a column of mercury or water of the stated height. When
converted to centimeter-gram-second (cgs) units, 1 mm Hg = 1330 dyn/cm2 and 1 cm H2O = 980 dyn/cm2.
P.298
Fig. 11-1: Comparison of a skeletal muscle (A), which shortens and develops tension (above), and the heart (B),
whose contraction reduces chamber volume and increases pressure (below). In the resting skeletal muscle, the
load rests on a support; in the resting heart, the load (arterial pressure) is separated from the ventricle by a
semilunar valve.
The relationships between the pressure within a chamber, wall stress, chamber size, and wall thickness are described
by the law of Laplace. In its simplest form, which describes a cylinder with infinitely thin walls (Fig. 11-3), the law of
Laplace states that wall tension is equal to the pressure within the cylinder times its radius:
where T is wall tension (dynes/cm), P is pressure (dynes/cm2), and R is radius (cm). An important corollary of the law
of Laplace is that when the pressure within the thin-walled cylinder is held constant, the tension on its walls increases
with increasing radius, and vice versa.
The law of Laplace is more complex in a thick-walled chamber, where wall stress (θ) is inversely proportional to wall
thickness (h). This is easily understood because when wall thickness is increased, the stress across any area within the
wall decreases. The law of Laplace, as applied to a thick-walled cylinder, is therefore:
Fig. 11-2: Outline of the chamber of the left ventricle viewed in anteroposterior and lateral projections.
Volume can be estimated by assuming the chamber to be an ellipse whose long axis is LM, and whose short axes
are DA and DL.
P.299
Fig. 11-3: The law of Laplace in a thin-walled cylinder relates wall tension (T) to the pressure within the
cylinder (P) and the radius of curvature (R).
A greater wall stress is required to achieve a given pressure in a larger chamber because, when the diameter
increases, a smaller proportion of the force developed by the chamber walls is directed toward the center of the
chamber (Fig. 11-4).
A familiar application of the law of Laplace is seen in the trucks used to transport fluids and gases in which, to
minimize the hazard of bursting, the tanks that carry compressed gasses at high internal pressures are constructed
with smaller radii than tanks that carry liquids at low pressures (Fig. 11-5). The tendency of a tank to burst can also
be reduced by increasing the thickness of its walls simply because the greater wall thickness reduces the stress on
each element in the wall.
The complex geometry of the ventricles makes it impossible to calculate wall stress with precision, but the law of
Laplace states that dilation increases the tension that must be developed by the muscle fibers in the walls of the
heart to generate a given chamber pressure, and that at any given chamber pressure, increasing wall thickness
(hypertrophy) reduces the amount of tension on each muscle fiber. Dilatation of the heart is deleterious because, by
increasing wall stress at any chamber pressure, more work must be done during systole to stretch elastic and viscous
elements in
P.300
the walls of the heart. Conversely, the smaller decrease in wall stress during ejection by a dilated heart reduces its
ability to use energy stored in the stretched elasticities to perform external work (see Chapter 12).
Fig. 11-4: The law of Laplace in a thick-walled cylinder. The wall stress () needed to achieve a given pressure ()
is greater in the larger cylinder because a smaller proportion of the force developed in the walls is directed
toward the center of the cylinder.
Fig. 11-5: A practical application of the law of Laplace. A single tank with a large radius of curvature is used in
a milk truck, where the pressure in the tank is low (left). Trucks used to transport gas under high pressure
contain several cylindrical tanks, each with a short radius of curvature (right).
Distribution of Stress across the Walls of the Heart

Stress is not distributed uniformly throughout the layers of the ventricular walls, but instead is greatest in the inner
(endocardial) regions (Mirsky, 1969; Fenton et al., 1978). This might appear to contradict the law of Laplace, which
states that wall stress increases as chamber diameter increases, but there is no contradiction because the distribution
of stress among the layers of a thick-walled ventricle is not the same as the average amount of stress in the walls of
ventricles that have different chamber sizes. The differences in wall stress in the various layers of the ventricle can be
understood by viewing the ventricle as a series of concentric elastic spheres, where an increase in chamber volume
causes the greatest increase in stress in the innermost layer. These differences are magnified when wall thickness
increases, so that a marked increase in the pressure developed by a thick-walled hypertrophied heart can cause severe
energy starvation, and even necrosis, of the subendocardium. Compression of the muscular branches of the coronary
arteries that penetrate the ventricular wall from their origins in epicardial coronary arteries (see Chapter 1) also
increases the vulnerability of the inner layers of the ventricular wall to energy starvation.
Cycles of Contraction and Relaxation: Pressure-Volume Loops

Table 11-1 lists five major determinants of cardiac performance, of which three are properties of the heart and two
are governed by the circulation. Among the former, heart rate is normally determined by pacemaker currents in the SA
node (see Chapter 13), and inotropy and lusitropy by the biochemical and biophysical properties of the working
cardiac myocytes (see Chapter 10). The key circulatory determinants of cardiac performance are venous return, which
along with lusitropic properties determines preload, and aortic pressure, which along with inotropy determines
afterload.
Pressure-volume loops, which describe changes in ventricular pressure as a function of chamber volume, provide a
valuable way to analyze the hemodynamic abnormalities in cardiovascular disease. To understand these diagrams, it is
useful first to consider a preloaded skeletal muscle that lifts a heavy afterload (Fig. 11-6). As detailed in Chapter 3, a
preload is a small load supported by a muscle before it is stimulated to contract, while an afterload is a heavier load
that is lifted after contraction begins, when the tension developed by the muscle exceeds the weight of the afterload.
P.301
Table 11-1 Major Determinants of the Work of the Heart
The Heart
Heart rate
Contractility (inotropy)
Relaxation (lusitropy)
The Circulation
Venous return (preload)

Arterial pressure (afterload)
Skeletal Muscle
The cycle of contraction and relaxation depicted in Figure 11-6 can be divided into four phases; two of contraction
and two of relaxation. The first phase, isometric contraction (A in Fig. 11-6), begins when the activated muscle starts
to develop tension; however, as long as muscle tension is less than the afterload, shortening cannot occur. The second
phase, isotonic contraction (B in
P.302
Fig. 11-6), begins when active tension equals the afterload, after which the muscle begins to lift the afterload.
Shortening then continues at a constant tension, which is equal to the weight of the afterload (the small preload is
ignored). The third phase, isotonic relaxation (C in Fig. 11-6), begins when the relaxing muscle starts to lengthen
while still bearing the afterload. Lengthening continues at the constant tension of the afterload until the latter
returns to the support and so is removed from the muscle. This marks the beginning of the fourth phase, isometric
relaxation (D in Fig. 11-6), during which tension is dissipated in the unloaded muscle. These four phases can be
plotted as a “work diagram” (Fig. 11-7), which shows how tension and length change during the contraction shown in
Figure 11-6. Because the muscle shortens and lengthens with the same afterload, the curves for isotonic contraction
and isotonic relaxation in Figure 11-7 are superimposed.
Fig. 11-6: Cycle of contraction and relaxation in an afterloaded skeletal muscle. A small preload stretches the
resting muscle, while the heavier afterload rests on a table until after the muscle has started to contract. The
first phase in the cycle is isometric contraction (A), during which tension developed by the muscle increases
until it equals the afterload. In the second phase, isotonic contraction (B), the muscle shortens while lifting the
afterload. Relaxation is initially isotonic, when the afterload is lowered to the tabletop (C). Isometric relaxation
(D) begins when the afterload reaches the table, and continues until muscle tension returns to zero.
Fig. 11-7: Work diagram of the contraction-relaxation cycle depicted in Fig. 11-6. A: Isometric contraction; B:
Isotonic contraction; C: Isotonic relaxation; D: Isometric relaxation. The two isometric curves (A and D) and two
isotonic curves (B and C) should be superimposed but are separated for clarity.
The Left Ventricular Pressure-Volume Loop

The work diagram generated by the left ventricle (Fig. 11-9) resembles the work diagram in Figure 11-8 because the
ventricle, like the skeletal muscle, carries a small preload, in this case the diastolic pressure generated by the venous
return and atrial systole. Furthermore, ventricular afterload, which is related to aortic pressure, does not influence
left ventricular chamber pressure during diastole because the closed aortic valve has “disconnected” the ventricle
from the aorta.
P.303
Fig. 11-8: Work diagram of a muscle from which the afterload is disengaged when relaxation begins. Isometric
contraction (A) and isotonic contraction (B) occur as in the contraction depicted in Figs. 11-6 and 11-7.
However, because the afterload is removed before the muscle begins to relax, tension during isotonic relaxation
(C) is that of the small preload, which stretches the muscle during isometric relaxation (D).
Pressure-volume loops are constrained within two pressure-volume relationships: the end-systolic pressure-volume
relationship, which is determined by the inotropic properties of the contracting ventricle, and the end-diastolic
pressure-volume relationship, which is determined by the lusitropic properties of the relaxed ventricle. Ventricular
end-diastolic volume (EDV), which must lie along the end-diastolic pressure-volume relationship, is determined by
venous return, end-systolic volume (the “residual” volume left behind after the previous cardiac cycle), and the
lusitropic properties of the ventricle (Table 11-2). End-systolic volume, which is determined by EDV, aortic impedance,
and contractility, must lie along the end-systolic pressure-volume relationship.
Table 11-2 Determinants of Ventricular Filling and Ejection
Filling
Venous return: Flow of blood into the heart
End-systolic (residual) volume: Amount of blood left in the ventricle after the previous systole
Lusitropy: Ability of the heart to fill (end-diastolic pressure-volume relationship)
Ejection
Aortic impedance: Ability of the aorta to receive blood from the heart
End-diastolic volume: Amount of blood in the ventricle at the start of systole
Inotropy: Ability of the heart to eject (end-systolic pressure-volume relationship)
P.304
Fig. 11-9: Pressure-volume loop generated by a normal left ventricle. The loop is constrained by the end-
diastolic pressure-volume relationship, which is determined by the lusitropic state of the ventricle, and the
end-systolic pressure-volume relationship, which is determined by the inotropic state. Systole begins at a point
along the end-diastolic pressure-volume relationship that represents the preload, after which the mitral valve
closes (MVC) and pressure increases rapidly. Because both the aortic and mitral valves are closed, this phase is
isovolumic (A). Ejection (B) begins when the aortic valve opens (AVO) and the ventricle meets its afterload, the
aortic pressure. Systole ends when ventricular pressure and volume reach the end-systolic pressure-volume
relationship, which describes the inotropic state of the ventricle. After aortic valve closure (AVC) separates the
afterload (aortic pressure) and the ventricular chamber, blood can neither enter nor leave the ventricle; as a
result, relaxation begins under isovolumic conditions (C). When left ventricular pressure falls below that in the
left atrium, the mitral valve opens (MVO) and blood flows from the atrium into the ventricle during the phase of
filling (D). The cycle ends when ventricular pressure and volume reach the end-diastolic pressure-volume
relationship.
Changes in circulatory hemodynamics modify pressure-volume loops by shifting the end-diastolic and end-systolic
points along the two pressure-volume relationships, which represent limits that cannot be exceeded. In contrast,
changes in inotropic and lusitropic state shift the end-systolic and end-diastolic pressure-volume relationships,
respectively, and so alter these limits.
Pressure-volume loops begin along the end-diastolic pressure-volume relationship and proceed in a counterclockwise
direction. Isovolumic contraction, the first phase of systole, begins when myocyte contraction increases ventricular
wall stress. The rising intraventricular pressure closes the mitral valve, after which pressure increases until it exceeds
that in the aorta, which causes the aortic valve to open. Until this happens, blood can neither enter nor leave the
ventricle, so that the increase of pressure at a constant volume inscribes an upward deflection (A in
P.305
Fig. 11-9). The second phase, ejection, begins when ventricular pressure exceeds that in the aorta, after which blood
is pumped into the aorta; the reduction in ventricular volume causes the pressure-volume loop to turn to the left (B in
Fig. 11-9). Aortic pressure initially rises during ejection because blood flows into the aorta from the ventricle faster
than it flows out to the tissues, and then falls as slowing of ejection allows blood to flow out of the aorta more rapidly
than it enters. Even though aortic pressure rises and falls during ejection, wall stress falls throughout this phase of the
cardiac cycle because of the decreasing chamber volume and thickening of the ventricular walls (the law of Laplace).
Systole ends when the loop reaches the end-systolic pressure-volume relationship.
Diastole begins with isovolumic relaxation (C in Fig. 11-9), which occurs after aortic valve closure; because ventricular
pressure exceeds that in the aorta and the mitral valve remains closed, ventricular volume cannot change. Diastole
therefore proceeds at a constant ventricular volume until left ventricular pressure falls below that in the left atrium,
which allows the mitral valve to open. This initiates the phase of filling (D in Fig. 11-9) during which blood flows
across the mitral valve from the atrium into the relaxing ventricle. Left ventricular pressure and volume increase
gradually during this phase as blood returning from the lungs generates the preload for the next contraction. Diastole
ends when the loop returns to the end-diastolic pressure-volume relationship.
The Cardiac Cycle

The electrical and mechanical events that take place during each cardiac cycle can be depicted as a “Wiggers
diagram” (Fig. 11-10). These are initiated by a wave of electrical depolarization that normally begins in the
pacemaker cells of the SA node, and propagated through the atria, AV junction, and His-Purkinje system to activate
the ventricles (see Chapter 15). The following description of the cardiac cycle is modified from the Carl J. Wiggers'
classic text Physiology in Health and Disease (Philadelphia: Lea and Febiger, 1949:651–654).
The series of superimposed curves which are reproduced in Figure 11-10 unfold at a glance the
story of cardiodynamic events in the left side of the heart which may be briefly summarized as
follows:
At the onset of ventricular systole the pressures are approximately equal in the atrium and
ventricle, and the atrioventricular (AV) valves are in the act of floating into apposition. After
the pressure has risen slightly within the ventricle, the AV valves close completely giving rise to
the first heart sound [S1]. Since the aortic valve is still closed, the ventricle contracts
isovolumically, and the intraventricular pressure rises rapidly. The aortic valve opens when left
ventricular pressure exceeds that in the aorta. As a result, aorta and ventricle become a
common cavity, and the two pressure curves follow one another closely.
With the rapid expulsion of blood during the early moments of ejection–indicated by volume
changes of the ventricles–the pressures in the left ventricle and aorta rise to a summit because
the rate at which blood is expelled into the aorta exceeds that at which it flows from its
branches through the arterioles. The rise is rounded chiefly because, with rather constant
ejection rate, the runoff increases gradually with the progressive rise of aortic pressure. The
rounded summit is reached when ejection and runoff become equal. Since the rate of ejection
diminishes during the latter part of systole while flow out of the aortic branches continues to
be high, aortic and ventricular pressures gradually decline. On the basis of pressure curves it is
possible to separate the period of ejection into two phases, viz., maximum ejection and
reduced ejection. Summarizing, the rise and fall of aortic and ventricular pressures always
represent a balance between the rate at which blood is ejected into the aorta and the rate at
which it leaves by its branches. However, the changes in rate of ventricular ejection normally
dominate the shape of the curves during ejection.
P.306
Fig. 11-10: The cardiac cycle (Wiggers diagram) showing seven phases of left ventricular systole. By
convention, the cycle begins with the onset of ventricular systole. The top three curves represent aortic
pressure (upper dotted line), left ventricular pressure (solid line), and left atrial pressure (lower dashed line).
The solid line below these pressure curves is left ventricular volume, below which are the heart sounds: S4,
fourth (or atrial) sound; S1, first heart sound; S2, second heart sound; S3, third heart sound. The bottom line
shows the timing of the electrocardiogram that records the electrical events during the cardiac cycle.
At the onset of ventricular diastole, aorta and ventricle are still in communication. The first
effect of relaxation consists in a sharp drop in pressure in the ventricle and aorta, the latter
being quickly terminated by the closure of the semilunar valves, after which the aortic curve
declines very gradually for the remainder of diastole. The closure of the semilunar valves is
associated with the second heart sound [S2]. The rate of aortic diastolic pressure decline is
determined chiefly by the rate at which blood flows out of the aortic branches, but is affected
to a variable extent by the increasing distensibility of arteries at different pressure levels.
Within the ventricle, the decline continues rapidly until the AV valves open, and the phase of
isovolumic relaxation terminates. During this phase of ventricular relaxation the atrial pressure
continues to rise slowly. As soon as intraventricular pressure has declined to a level lower than
that in the atrium, the AV valves are opened again by the difference of pressure and a rapid
inflow of blood into the ventricle begins. While this continues, pressures in the atrium and
ventricle decline together, but the atrial pressure remains a trifle higher than the ventricular.
In long cycles this is followed by a phase of slowed filling, or diastasis, during which ventricular
inflow is exceedingly slow, and the pressure rises very gradually both in the atrium and
ventricle. In young normal individuals, and in some pathological states the rapid [fall in tension
in the walls of] the ventricle is associated
P.307
with an audible sound, the third heart sound [S3]. Occasionally, atrial systole also produces an
atrial sound sometimes called the fourth heart sound [S4].
The Phases of the Cardiac Cycle: The succession of atrial and ventricular events constitutes
the cardiac cycle. Since ventricular contraction is dynamically the most important it is fitting
to start the cycle with this event. Accordingly, the cardiac cycle can be divided advantageously
into ventricular systole and diastole, but each of these periods must be further subdivided. For
the sake of clarity these subdivisions are designated as phases of systole and diastole. The
vertical lines of Figure 11-10 serve to demarcate the successive periods and phases of systole
and diastole. The first phase of systole is called isovolumic contraction, for the ventricle
contracts essentially in this manner with all valves closed. The second phase is best referred to
as ejection; it can be further subdivided by reference to the aortic pressure curve alone or with
the aid of the ventricular volume curve into the phase of maximum ejection, and the phase of
reduced ejection.
Diastole begins with closure of the semilunar valves. It is followed by isovolumic relaxation,
which ends as soon as atrial pressure exceeds that in the ventricle. With opening of the AV
valves, rapid filling supervenes, and this is followed by a phase of slowed filling or diastasis,
whose length depends on the heart rate. Finally, atrial systole terminates the period of
ventricular diastole and the cycle begins again.
The durations of these successive phases have been repeatedly studied but with varying degrees
of accuracy. The average values in seconds (Table 11-3) are approximations that are provided to
give an idea as to the relative duration in man of the most commonly used phases.
A schematic drawing showing the relationship of the electrocardiogram to the cardiac cycle is included in Figure 11-
10. The P wave is inscribed during atrial depolarization, while the QRS complex originates from ventricular
depolarization; both of these electrical events precede the corresponding mechanical events. The T wave is due to
ventricular repolarization (see Chapter 15).
The first and second heart sounds (S1 and S2), although initiated by closure of the heart's valves, are generated when
rapid deceleration of the moving stream of blood causes vibrations in the walls of the ventricles and, to a lesser
extent, in the great vessels. The mechanism resembles a
P.308
drum, where sound is initiated by a sudden deceleration of the moving drumstick (analogous to valve closure), but is
actually generated in the head of the drum (analogous to vibrations in the walls of the ventricle). The third heart
sound (S3) is caused by the elastic recoil of the ventricular walls at the end of isovolumic relaxation, and the fourth
heart sound (S4) occurs when atrial systole rapidly distends the ventricle. S3 and S4 give rise to “gallop rhythms”
because each establishes a cadence with S1 and S2; this cadence, also called a “triple rhythm,” resembles the sound
of a galloping horse. The first abnormality noted by an experienced auscultator who listens for S3 and S4 is the triple
rhythm, after which the extra sound can be identified. A useful way to distinguish between S3 and S4, which are
usually softer than the S1 and S2, is that the cadence S1–S2–S3 sounds like “Kentucky,” while S4–S1–S2 sounds like
“Tennessee.”
Table 11-3 Durations of the Phases of the Cardiac Cyclea
Isovolumic contraction 0.05
Maximum ejection 0.09
Reduced ejection 0.17
Total systole 0.31
Isovolumic relaxation 0.08
Rapid inflow 0.11
Diastasis 0.19
Atria systole 0.11
Total diastole 0.49
Approximate durations, in seconds, for the human left ventricle beating at a rate of 75/min.
aModified from Wiggers CJ. Physiology in Health and Disease. Philadelphia: Lea and Febiger, 1949:651–
654.
End-Diastolic Volume as a Determinant of Ventricular Function: Starling's
Law of the Heart (The Frank–Starling Relationship)
The relationship between the volume of blood in ventricles at the moment they begin to contract (end-diastolic
volume or EDV) and cardiac performance is a manifestation of the length-tension relationship (see Chapters 6 and 10).
The operation of this physiological relationship can be measured when a balloon placed within the left ventricle is
filled with increasing volumes of an incompressible fluid (Fig. 11-11). When the volume in the balloon is increased,
end-diastolic pressure increases; the curve relating the changing pressures and volumes at the end of the diastole is,
of course, the end-diastolic PV relationship. The pressure that can be developed during systole is modified by the
volume of the fluid-filled balloon. Over the physiological range of EDV, the heart's ability to generate pressure
increases linearly as the balloon fills. Arrows drawn to connect each end-diastolic pressure with the corresponding
pressure at the end of systole are vertical in Figure 11-11A because the volume in the balloon cannot change, which
means that the heart is contracting under isovolumic conditions. When the lengths of these arrows, which represent
the pressures developed during systole (systolic pressure minus diastolic pressure), are plotted as a function of the
EDV (Fig. 11-11B), the resulting curve (often called a Starling curve) is analogous to the length-tension relationship of
a skeletal muscle (see Chapter 6).
The heart normally operates on the left-hand portion of the Starling curve, which like the left-hand part of the length-
tension curve, is the ascending limb where increasing preload enhances the ability of the heart to empty (see Chapter
6). This allows the heart to respond to an increased EDV by increasing ejection pressure, stroke volume, or both. At
very high, non-physiological, filling pressures the heart can move onto the descending limb; however, the low
compliance of cardiac muscle and the stiff pericardium normally prevent this from occurring. This is very important
because to achieve any steady state, the heart must eject during systole the volume of blood that it receives during
diastole (Katz, 1965). The situation is different in skeletal muscles, which can operate safely on the descending limb
of their length-tension curves because they do not determine their rest length and have opposing muscles that can
reverse an excessive increase in rest length.
Starling (1918) noted that the heart cannot function on the descending limb of the Starling curve because increasing
venous return to such a heart would reduce its ability to eject. This would establish a vicious cycle in which an
increase in EDV would move the heart further down
P.309
the descending limb, which would decrease ejection, which would increase end-systolic volume, which would add to
the EDV, etc. The heart has no way to recover from this vicious cycle, which in theory could cause it to burst like an
overfilled balloon. However, because of the low compliance of cardiac muscle and the stiff pericardium, a heart
operating on the descending limb of the Starling curve does not rupture. Instead, the increasing diastolic pressures
rapidly reach levels where fluid is transudated from the pulmonary capillaries into the alveoli, which results in a
syndrome called acute pulmonary edema (in the 19th century this was referred to as “acute dilatation”).
Fig. 11-11: Starling's law of the heart. A: The difference between systolic and diastolic pressures (vertical
arrows) developed in a series of isovolumic contractions increase with increasing end-diastolic volume (EDV). At
very high EDVs, systolic pressure can decline with increasing volume. The lower dashed line is the end-diastolic
pressure-volume (PV) relationship; the upper dashed line is the end-systolic PV relationship. B: Ascending and
descending limbs of the Starling curve, plotted as developed pressure (vertical arrows in A). The increase in
developed pressure that accompanies an increase in EDV is the ascending limb of the Starling curve; the fall in
developed pressure as the ventricle dilates at very high volumes is the descending limb.
Faced with a patient in whom this disaster is unfolding, efforts must be made to reduce ventricular volume. This was
once done using rotating tourniquets and sometimes phlebotomy to lower preload, along with morphine (which has a
vasodilator effect) to facilitate ejection by reducing afterload; together, these interventions can return the heart to
the steady state that occurs on ascending limb by decreasing ventricular volume. Today's therapy for acute pulmonary
edema also uses diuretics to reduce preload and arteriolar vasodilators to reduce afterload.
P.310
The Pericardium
Pericardial volume can be decreased by chronic inflammation (constrictive pericarditis), or when the pericardial cavity
fills with fluid (pericardial effusion). These impair the heart's ability to fill, and so reduce cardiac output and increase
venous pressures. Pericardial tamponade, which can occur if a pericardial effusion develops rapidly or blood enters
the pericardial cavity from a ruptured ventricle or aortic dissection, can be rapidly fatal.
Although the pericardium limits acute dilatation of the heart, the pericardial sac generally enlarges slowly when
stretched chronically. For this reason, the pericardium plays little role in limiting the progressive dilatation that often
occurs in failing hearts (see Chapter 18).
The Atrium as a Primer Pump

The timing of atrial systole immediately before ventricular systole allows the atria to deliver a small volume of blood
to the ventricles at that instant (end diastole) when ventricular volume determines ventricular performance. This
timing also allows atrial systole to add to EDV without increasing ventricular pressure during all of diastole. The atria,
therefore, operate as a primer pump that increases diastolic volume at the instant (end diastole) when diastolic
volume determines ventricular performance, while avoiding the need to maintain a high atrial pressure throughout
diastole.
Loss of the atrial “kick,” which most commonly occurs when effective atrial contraction ceases in patients with atrial
fibrillation, has two hemodynamic consequences, both bad; ventricular EDV decreases, which reduces cardiac output,
while the higher mean atrial pressure increases venous pressure, which reduces venous return (Fig. 11-12). The fall in
ventricular end-diastolic pressure, according to Starling's law, reduces ejection and decreases cardiac output, while
the rise in mean atrial pressure both impedes the return of blood to the heart and increases venous pressures. These
effects can be well-tolerated in individuals who have a normal heart, but when cardiac function is compromised, as in
patients with heart failure, atrial fibrillation seriously worsens the clinical disorder. Another dangerous complication is
increased risk of emboli caused by the tendency of clots to form in fibrillating atria. Because clots that form in the
left atrium often travel to the cerebral circulation where they can cause a cerebrovascular accident (“stroke”), atrial
fibrillation can have devastating consequences, even when the hemodynamic abnormality is mild. For this reason,
patients with atrial fibrillation are often given anticoagulants.
Conclusion
Two laws govern the relationship between ventricular volume and cardiac pump performance. The first, the law of
Laplace, is a physical law stating that when the ventricle dilates, the wall stress needed to achieve a given
intraventricular pressure is increased. The second, Starling's law of the heart, is a physiological law which states that
when a ventricle dilates, its ability to perform work increases. In a normal heart operating within the normal range of
diastolic pressures, the physiological law predominates: When the ventricle dilates, the increased capacity to
generate pressure is more important than the increased wall stress. However, in patients with heart failure, the
increased wall stress caused by ventricular dilation adds to the energy cost of cardiac contraction, so that the
detrimental effects of increasing ventricular volume can outweigh the beneficial effects.
P.311
Fig. 11-12: Effects of loss of atrial systole. In the normal cardiac cycle, the pressure developed by atrial
contraction (thin dashed line) determines left ventricular pressure (solid line) at the end of diastole. Loss of
atrial systole increases atrial pressure during most of diastole (heavy dashed lines in the upper panel), and
reduces end-diastolic volume and ejection (heavy dashed line in the lower curves).
Bibliography
Baim DS, Grossman W. Grossman's cardiac catheterization, angiography, and intervention. 6th ed. Philadelphia,
PA: Lippincott Williams & Wilkins, 2000.
Burton AC. Physical principles of circulatory phenomena: the physical equilibria of the heart and blood vessels.
In: Hamilton WF, Dow P. Handbook of physiology, section 2: circulation, vol 1. Washington, DC: Am Physiol Soc,
1962:85–106.
Covell JW, Ross J Jr. Systolic and diastolic function (mechanics) of the intact heart. In: Page E, Fozzard HA,
Solaro RJ, ed. Handbook of physiology, section 2: The cardiovascular system, vol 1: The Heart. New York, NY:
Oxford University Press, 2002:741–785.
P.312
Luisada AA, Portaluppi F. The main heart sounds as vibrations of the cardiohemic system: old controversy and
new facts. Am J Cardiol 1983;52:1133–1136.
Suga N. Ventricular energetics. Physiol Rev 1990;70:247–277.
References
Fenton TR, Cherry JM, Klassen GA. Transmural myocardial deformation in the canine left ventricular wall. Am J
Physiol 1978;235:H523–H530.
Katz AM. The descending limb of the Starling curve and the failing heart. Circulation 1965;32:871–875.
Mirsky I. Left ventricular stresses in the intact human heart. Biophys J 1969;9:189–208.
Starling EH. The linacre lecture on the law of the heart. London: Longmans Green & Co, 1918.
> Table of Contents > Part Three - Normal Physiology > Chapter 12 - The Working Heart
Chapter 12
The Working Heart
More than a century ago, Woods (1892) related the different shapes of the right and left ventricles (see Chapter 1) to the law of Laplace. Using
measurements of wall thickness and radii of curvature, and assuming that wall stress is uniform throughout the heart, Woods noted that “the
thickness of the heart at any place bears a direct proportion to the relative tension at that place.” This allowed him not only to predict that left
ventricular pressure is higher than that in the right ventricle, but that the difference is less in the neonatal heart where, as we now know,
pulmonary artery pressure is the same as that in the aorta.
Adaptation of the shape of the heart to the pressure developed in its cavities explains architectural differences in the ventricles of various
animals (Fig. 12-1). For example, in the giraffe left ventricle, where systolic pressure often exceeds 300 mm Hg, wall stress is kept low by thick
walls and a narrow cavity, whereas the amphibian ventricle, which generates low pressures, is thin-walled and almost spherical. The law of
Laplace also explains the different shapes of the right and left ventricles (Fig. 12-2; see also Fig. 1-4). The right ventricle, which develops a
relatively low systolic pressure, is crescentic and has a large radius of curvature, while the more conical left ventricle has a smaller radius of
curvature that allows it to generate high intraventricular pressures with the same wall stress as the right ventricle.
The interventricular septum normally functions as part of the left ventricle, and so moves toward the free wall of the left ventricle as the heart
ejects (Weber et al., 1981). However, when right ventricular pressure is abnormally elevated, for example, in pulmonary hypertension, the
septum contracts “paradoxically” and moves away from the left ventricular cavity to assist in right ventricular ejection.
The Work of the Heart

The heart performs two types of work: external work that propels blood from the left ventricle into the aorta and from the right ventricle into
the pulmonary artery, and internal work that is expended during systole to alter the shape of the heart as it contracts, and to stretch elasticities
and lengthen viscous elements in the walls of the contracting ventricles. Although much of the energy expended to perform internal work is
degraded to heat when the heart relaxes, and so reduces the efficiency of cardiac performance, some of the energy used to stretch elasticities
can help eject blood (see below).
P.314
Fig. 12-1: Ventricular cavity shape. The thin-walled ventricle of the amphibian heart, which develops a low pressure, is almost spherical,
while that of the thicker human left ventricle, which develops a much higher pressure, is conical. In the giraffe, where left ventricular
systolic pressure is extremely high, the cavity of the thick-walled ventricle is almost tubular.
External Work
Stroke Work
The work performed during each cardiac cycle to eject blood under pressure into the aorta and pulmonary artery is the stroke work (sometimes
called pressure–volume work). This can be estimated by multiplying the volume of blood ejected during each stroke (the stroke volume,
abbreviated SV) by the average pressure at which the blood is ejected (P):
The product of pressure (dyn/cm2) and volume (cm3) has the correct cgs units for work (dyn cm). In the left ventricle, when mean ejection
pressure is 105 mm Hg and stroke volume is 70 mL, stroke work is 7,350 mm Hg mL, which in cgs units is ∼9.3 × 106 dyn cm.
Because aortic and pulmonary artery pressures first rise and then fall during ejection (see Chapter 11), stroke work is more accurately calculated
as the integral of pressure and the volume change:
where P is the pressure at which each increment (dV) of the stroke volume is ejected. The total external work is also the area within the
pressure–volume loop (Fig. 12-3). For most purposes, left ventricular stroke work can be estimated simply by multiplying stroke volume by either
peak or mean aortic pressure.
Fig. 12-2: Cross section of the human heart showing the thin-walled, crescentic right ventricle (RV), where systolic pressure is ∼1/5 that
of the thick-walled, narrower left ventricle (LV). The interventricular septum normally functions as part of the LV.
P.315
Fig. 12-3: The external work of the left ventricle during each cardiac cycle is equal to the work performed by the ventricle, which is the
area within the pressure–volume loop (A) plus the work contributed by the inertia of the venous return and by atrial systole (B).
Stroke volume is end-diastolic volume (EDV, the volume in the ventricle when ejection begins) minus end-systolic volume (ESV, the residual
volume at the end of ejection). This allows Equation 12-1 to be expanded as:
Each of the determinants of stroke work is controlled differently. EDV is determined by three variables; two, which together define the preload,
are venous return and ESV, and the third is the lusitropic state of the ventricle. ESV is determined by EDV and the volume of blood that is ejected
(stroke volume), which in turn reflects the inotropic state of the ventricle and the afterload. At any steady state the stroke volumes of the two
ventricles are the same, but because pulmonary artery pressure is approximately one-fifth that of aortic pressure, the stroke work of the right
ventricle is less than that of the left ventricle.
Preload and Afterload

Preload and afterload, which are most accurately defined as wall stresses, are related to but are not the same as systolic and diastolic chamber
pressures. The relationships are determined by the law of Laplace, which states that wall stress is directly proportional to chamber pressure and
chamber diameter, and inversely proportional to wall thickness (see Chapter 11). These variables all change as the heart ejects; chamber size
decreases and wall thickness increases, both of which reduce wall stress; pressure, which initially rises and then falls, has mixed effects. Overall,
the net effect is a marked fall in wall stress during ejection (Fig. 12-4) that has important effects on the efficiency of cardiac performance (see
below).
Resistance and Impedance

Aortic pressure is determined by the amount of blood ejected into the aorta by the left ventricle and aortic impedance. Although impedance is
often equated to resistance, they are not the same, nor are they determined in the same regions of the arterial system.
P.316
Fig. 12-4: Changes in wall stress during ejection by the normal left ventricle. The initial rise in pressure (dashed line) causes wall stress
(solid line) to increase rapidly. However, left ventricular wall thickening (dotted line) and the subsequent fall in pressure and cavity
volume cause a marked fall in wall stress during ejection.
Peripheral resistance, the major determinant of the rate at which blood flows out of the arteries to perfuse the tissues, is regulated largely by
the diameter of small arterioles, often called “resistance vessels,” that control blood flow into the capillaries. Vascular resistance is defined as
the ratio between the pressures at the arterial and venous sides of a vascular bed (ΔP) and the flow through the bed (Q).
Aortic impedance, which determines the peak level of pressure when blood is being ejected from the left ventricle, is influenced by the
elasticities of the aorta and other large arteries as well as by the caliber of the resistance vessels. Unlike resistance, which describes the
relationship between cardiac output and the difference between the mean pressures in the arteries and veins (Equation 12-4), impedance is also
influenced by the changes in pressure and flow velocity that occur throughout the cardiac cycle. For example, systolic blood pressure becomes
higher and diastolic pressure becomes lower when impedance is increased by loss of elasticity in the large arteries, which occurs during normal
aging and is accelerated by arteriosclerotic changes in the large arteries. Because calculation of impedance is complex, resistance is almost
always used for clinical measurements.
The distinction between resistance and impedance has important implications; consider, for example, two patients in whom cardiac output and
venous pressure are the same, but in whom aortic pressures are 110/80 and 170/50 mm Hg (the latter is often called “systolic hypertension”).
Because mean aortic pressure (which can be estimated as aortic diastolic pressure + 1/3 pulse
P.317
pressure) is ∼90 mm Hg in both patients, their peripheral resistances are also the same. However, the left ventricle ejects against a higher
afterload in the patient with systolic hypertension because aortic impedance is greater.
Kinetic Work
The heart performs a small amount of kinetic work when it imparts velocity to the blood as it leaves the ventricles. The kinetic energy of this
moving stream of blood, according to the laws of physics, is proportional to the square of the velocity at which blood leaves the ventricle:
where m is the mass of blood moving out of the left ventricle into the aorta or from the right ventricle to the pulmonary artery, and v is the
velocity at which this blood crosses the semilunar valves.
The kinetic work of the left ventricle is normally less than 5% of the stroke work, but in the right ventricle, where systolic pressure is low, kinetic
work represents a greater proportion of the smaller amount of total work performed. Because stroke volume contributes to both m (cm3) and v2
[(cm3/s)2] in Equation 12-5, kinetic work is proportional to the cube of the stroke volume. For this reason, when stroke volume is abnormally
high (as occurs in aortic insufficiency or severe anemia), kinetic work can represent a significant portion of the work of the left ventricle
(although rarely more than 10%).
Kinetic work contributes to the “useful” work of the heart because most of the kinetic energy is converted to pressure as velocity slows when
blood moves through the circulatory system. For this reason kinetic work, like stroke work, contributes to the “useful” work that pumps blood
through the body.
Atrial Work
The work done when the atria pump blood into the ventricles makes only a negligible contribution to cardiac energy expenditure because atrial
systole propels only a fraction of the stroke volume at a low pressure. Even though it adds little to overall energy utilization, atrial systole is an
important determinant of ventricular EDV because of its timing (see Chapter 11).
Work Done to Fill the Ventricle

The initial phase of ventricular filling, which follows isovolumic relaxation, occurs when the elastic recoil of the ventricular walls literally sucks
blood into the expanding cavity (Nikolic et al., 1988). The energy expended to fill the ventricle later during diastole is provided by the
momentum of the venous return, which is generated by the opposite ventricle; blood pumped out of the right ventricle helps fill the left
ventricle, and blood pumped from the left ventricle helps fill the right ventricle. As noted above, atrial systole normally makes only a small
contribution to filling. Calculations of the work of the ventricle often subtract the energy expended to fill the ventricles, but this is usually a
small fraction of total stroke work (Fig. 12-3).
Internal Work
The heart uses a large amount of energy to perform internal work. This includes energy used to rearrange cytoskeletal structures (see Chapter 5)
and stretch elastic and viscous elements in the myosin cross-bridges other sarcomeric proteins. Energy is also expended to stretch the
P.318
connective tissue that supports the heart and reorient the spiral bundles that make up the muscular architecture of the ventricles (see Chapter
1). Because most of this energy is used to elongate elasticities and viscosities, internal work is proportional to wall stress.
The downward movement of the entire heart during ejection, sometimes referred to as the “descent of the base,” is a form of internal work;
this movement, which is a consequence of Newton's third law (“to every action there is an equal and opposite reaction”), occurs when blood
ejected across the aortic valve toward the head causes the heart to move in the opposite direction, toward the feet. The descent of the base is
the major cause of the x descent in the jugular venous pulse, a downward pulsation seen immediately after the upward a wave caused by atrial
systole. This movement, which explains the bouncing of the needle in synchrony with the heartbeat when one stands quietly on a spring scale,
provided the rationale for ballistocardiography, a now discarded method to estimate stroke volume.
Much of the energy expended as internal work contributes to the heart's inefficiency because it is degraded to heat when the heart relaxes (see
below). However, some of the potential energy stored in stretched elasticities is used to perform external (useful) work when wall stress
decreases during ejection (see below). Conversion of this potential energy to useful work accounts for the energy-sparing effect of interventions
that reduce afterload; conversely, increased release of this potential energy as heat is among the adverse consequences of high wall stress. The
latter is especially deleterious in patients with coronary occlusive disease or heart failure, whose hearts are usually energy-starved.
Minute Work
Clinical descriptions of cardiac work generally refer to the external work performed per minute, or minute work. The latter, which is a power
(work performed over time), can be calculated by multiplying the work per beat (stroke work) by the number of beats per minute (HR = heart
rate):
Because the product of HR × SV in Equation 12-6 is cardiac output (CO), minute work is also the product of pressure times cardiac output:
Each of the variables in Equation 12-6 makes an independent contribution to the minute work; for example, the latter can be increased by
increasing ejection pressure, heart rate, stroke volume, or any combination of these variables. Because stroke volume = EDV - ESV, minute work
has four determinants:
Each of these four variables has a different impact on cardiac energetics, so that minute work provides only a rough index of the energy demands
of the heart.
Work of the Isolated Heart

The variables that determine minute work are readily analyzed in isolated heart preparations, where each can be varied independently. The
preparation illustrated in Figure 12-5, which is similar to a working turtle heart that the author studied as a medical student in 1953, considers
P.319
the work of a single ventricle–this is not a problem in the turtle heart, which has only one ventricle. Filling pressure (preload) can be varied by
adjusting the height of the venous reservoir and ejection pressure (afterload) by raising or lowering the outlet of a flexible tube connected to
the aorta; heart rate is controlled by electrical stimulation. The following discussion examines the effects of changing filling pressure, ejection
pressure, and heart rate while the other determinants of minute work are kept constant.
Fig. 12-5: Isolated turtle heart preparation. Blood flows from the venous reservoir into the left atrium (LA) at a pressure determined by
the height of the reservoir. When the single ventricle (V) contracts, blood is pumped across the aortic valve (AO) into a tube in which the
height of the outlet, relative to the center of the atrium, is the ejection pressure. Stroke volume is the amount of blood ejected during
each beat.
Vary Filling Pressure

Changing the height of the venous reservoir directly modifies ventricular end-diastolic pressure and EDV (preload). The effects on cardiac
performance are predicted by Starling's law of the heart; as long as the heart functions on the ascending limb and ejection pressure is constant,
stroke volume will increase with increasing filling pressure (Fig. 12-6). Because stroke
P.320
work equals stroke volume times ejection pressure, which is held constant, stroke work can be substituted for stroke volume on the ordinate in
Figure 12-6; as heart rate is also kept constant, minute work can be placed on the ordinate of Figure 12-6.
Fig. 12-6: Operation of Starling's law of the heart under conditions where heart rate and ejection pressure are constant, and only filling
pressure is allowed to vary. Under these conditions, stroke volume rises as filling pressure increases. The descending limb shown at high
end-diastolic volumes in this and the next figure is not seen in the normal mammalian heart.
Fig. 12-7: Operation of Starling's law of the heart under conditions where heart rate and stroke volume are constant, so that only ejection
pressure can vary. When ejection pressure is increased, the end-diastolic pressure must also be increased to meet the requirement that
stroke volume remains constant, and vice versa.
Vary Ejection Pressure

Raising or lowering the outlet of the tube connected to the aorta directly affects ejection pressure (afterload). However, stroke volume will also
change because afterload is a major determinant of the ability of a muscle to shorten (see Chapter 3); for this reason, increasing afterload will
reduce stroke volume, and decreasing afterload will increase ejection. To satisfy the requirement that only one variable be changed at any time,
the level of the venous reservoir must also be adjusted to maintain a constant stroke volume. This means that when afterload is increased, the
venous reservoir must be raised to maintain stroke volume; conversely, when reduced ejection pressure allows stroke volume to increase, the
venous reservoir must be lowered. Varying the pressure in the venous reservoir directly alters EDV, which allows Starling's law to maintain a
constant output in the face of the changing afterload (Fig. 12-7). As heart rate is kept constant, stroke work and minute work can replace
ejection pressure on the ordinate in Figure 12-7.
Varying Heart Rate

When heart rate is varied at constant filling and ejection pressures, stroke volume by the turtle heart will be constant (Fig. 12-8); within limits,
therefore, cardiac output and minute work will vary directly with heart rate. However, at very high heart rates diastole becomes too brief to
allow the ventricle to relax completely between beats, so that cardiac output will begin to fall as rate increases further.
The effects of changing heart rate are not the same in the intact animal where, within limits, cardiac output is independent of heart rate. This is
because blood flow through most organs is matched to energy needs by changes in the caliber of the resistance vessels that control tissue
perfusion. This mechanism, called autoregulation, adjusts blood flow to match the needs of the tissue, which means that cardiac output is
normally determined by the needs of the body, rather than the ability of the heart to pump blood (Warner and Toronto, 1960) (Fig. 12-8). At very
low and very high heart rates, however, cardiac output becomes dependent on rate. At very rapid heart rates, cardiac output falls because
diastole does not last long enough to allow the ventricle to fill completely; this is due both to limited flow velocity across the mitral valve and
lack of time for the myocardium to relax completely. In humans, cardiac output begins to fall at heart rates greater than about 160/min and less
when there is heart failure. Cardiac output also falls when heart rate becomes extremely slow, as in patients with complete heart block where
rates can be as low as 20/min. Under these conditions, even though there is ample time for filling, the ability of the heart to fill is limited by its
low diastolic compliance. The adverse clinical effects of these very slow heart rates include profound weakness and clinical evidence of heart
failure, both of which can be relieved by pacing the heart with an electronic pacemaker.
P.321
Fig. 12-8: Effects of heart rate on cardiac output in the isolated turtle heart (dashed line) and in the intact animal (dotted line). In the
former, as long as the duration of diastole is sufficient to allow the heart to relax completely, cardiac output will be directly related to
heart rate. In the intact animal, blood flow through the tissues is normally determined by local metabolic needs, so that cardiac output
will remain constant when heart rate changes; for this reason, stroke volume will vary inversely with heart rate. In both, cardiac output
will begin to fall at heart rates so rapid that the duration of diastole becomes too brief to allow the heart to relax fully. Cardiac output in
intact animals also falls at very low heart rates.
Energy Cost of the Work of the Heart

Cardiac energy utilization can be estimated by measuring the consumption of oxygen by the heart because the myocardium satisfies virtually all
of its energy needs by oxidizing fat, carbohydrates, and to a minor extent protein (see Chapter 2). The energy made available from a given
amount of oxygen is virtually independent of the substrate oxidized because, even though oxidation of fat yields ∼9 calories/g, while
carbohydrate and protein oxidation yields only ∼4 calories/g, more oxygen is consumed by the oxidation of fat. For this reason, energy release
per liter of oxygen consumed is similar for all of these substrates.
Cardiac Efficiency
The efficiency of cardiac contraction can be estimated by dividing the external work of the heart by the energy equivalent of the oxygen
consumed:
Equation 12-9 does not describe a thermodynamic efficiency, which is extremely difficult to calculate. However, this simpler analysis provides a
useful index of the overall economy of the heart. The efficiency of the working heart calculated in this manner ranges between 5% and 20%, the
exact value depending on the nature and amount of work performed.
P.322
Fig. 12-9: Relationship between external work and the energetics of ventricular contraction. Oxygen consumption (solid line) increases
when more work is done while efficiency (dashed line) first rises and then falls. (Based on data from Evans and Matsuoka, 1915).
The importance of hemodynamics on cardiac energetics was first shown by Evans and Matsuoka (1915), who found that oxygen consumption
increases when the heart does more work (Fig. 12-9); as noted in Chapter 3, this is a manifestation of the Fenn effect. Evans and Matsuoka also
found that oxygen consumption and efficiency depend on how work is performed as well as the amount of work (Fig. 12-10). Increases in
pressure and stroke volume both require extra oxygen consumption, but the extra expenditure of energy is greater when the heart contracts
against increasing afterload than when it ejects a larger volume of blood. This means that an increase in pressure generated by the heart is
energetically more costly than a similar increase in work caused by a greater stroke volume.
Effect of Afterload on Efficiency

The high energy cost of increasing afterload reflects the large amount of energy expended to do internal work during isovolumic contraction.
This becomes obvious when afterload is so high that the ventricle cannot develop enough pressure to open the semilunar valves; under these
conditions, where no external work can be performed and the numerator in the equation for cardiac efficiency (Equation 12-9) is zero, energy
must still be expended to perform internal work. When afterload decreases enough to allow even a small amount of blood to be ejected, a finite
P.323
amount of external work is performed and a numerator appears in Equation 12-9. Although extra energy is required to eject this blood,
efficiency initially increases with decreasing afterload because internal work becomes a smaller proportion of the total energy expenditure.
Fig. 12-10: Influence of the type of external work on oxygen consumption (solid lines) and efficiency (dashed lines). Increasing cardiac
work from W1 to W2 by increasing ejection pressure (P) causes a greater increase in oxygen consumption than when work is increased to a
similar extent by increasing cardiac output (CO). For this reason, efficiency is greater when cardiac output is increased than when
ejection pressure is increased. (Based on data from Evans and Matsuoka, 1915).
Fig. 12-11: Effect of afterload on internal work and energy release as heat. Left: The series elasticity in the unloaded muscle is not
stretched. Center: When the muscle lifts a light afterload, a small amount of internal work is expended to increase the length of the
series elasticity. During relaxation, when the afterload is removed from the muscle, a small amount of heat is liberated as the stretched
elasticity shortens. Right: A larger amount of internal work is performed to stretch the series elasticity when the muscle lifts a heavier
afterload. During relaxation, when the afterload is removed from the muscle, shortening of the stretched elasticity liberates a larger
amount of heat.
Increasing afterload adds to the energy cost of cardiac contraction because it increases the amount of energy that must be expended to stretch
internal elasticities. The mechanism is depicted in Figure 12-11, which shows that, when compared to a light afterload, a heavy afterload causes
a greater increase in the tension that stretches the series elasticity when the muscle contracts. Figure 12-11 also shows that when afterload is
increased, more energy is released as heat by the stretched series elasticity when the muscle relaxes and the elasticity returns to its original
length.
The decrease in wall stress that normally occurs when the heart ejects (see Fig. 12-4) has two advantages: it increases cardiac output by
enhancing the heart's ability to empty, and it reduces the amount of energy released as heat during isovolumic relaxation. Both occur because
energy stored in the series elasticity is used to perform external work when decreased chamber volume reduces wall stress (Fig. 12-12). By
allowing the series elasticity to shorten during ejection, the decreased wall stress allows some of the energy in the stretched elasticities to pump
blood rather than being released as heat.
The cost of performing internal work at high wall stress is greater in cardiac muscle than in the frog sartorius, where heat liberation during
contraction is essentially independent of load (Fig. 12-13). More than 40 years ago, this difference was explained to the author by Wallace Fenn,
who had found that not all skeletal muscles contract as efficiently at high loads as the sartorius; in the frog gastrocnemius, for example, total
energy expenditure does not decline appreciably
P.324
with increasing load (Martin, 1928). Fenn pointed out that these differences reflect different amounts of internal work expended to rearrange
fiber bundles when these muscles develop tension. Unlike the sartorius, where the fiber bundles run parallel to one another from one end of the
muscle to the other, the gastrocnemius has an asymmetrical bipinnate arrangement (Fig. 12-14) that allows significant shape changes to occur
even when the ends of the muscle are fixed in an
P.325
isometric contraction. Although the extent of fiber shortening is small, the high tension developed during isometric contraction allows shape
changes in the gastrocnemius to generate a large amount of internal work that, when degraded to heat, increases energy expenditure (Fig. 12-
13B). In the sartorius muscle, where the parallel arrangement of the fiber bundles allows much less shape change to occur during contraction
(Fig. 12-13A), internal work and energy wastage are less at high levels of developed tension (this is one reason that A.V. Hill chose the frog
sartorius as his standard preparation for the classical studies of muscle energetics described in Chapter 5). The complex architecture of the heart
(see Chapter 1), which resembles that of the gastrocnemius much more than the sartorius, is therefore a major cause of the higher energy cost
and low efficiency of pressure work shown in Figure 12-9.
Fig. 12-12: Effect of decreasing afterload on energy expenditure by a contracting muscle. Left: When the muscle lifts a heavy afterload,
internal work is performed to stretch the series elasticity. When the afterload is removed from the relaxing muscle, the stretched
elasticity liberates a large amount of heat. Right: If the afterload is reduced while the muscle contracts, the series elasticity is able to
shorten (dashed arrows); this allows some of the energy in the previously stretched elasticity to help lift the load. This improves
efficiency by allowing the stretched elasticity to perform external work and decreasing the amount of heat liberated as the stretched
elasticity elongates when the muscle relaxes.
Fig. 12-13: Effects of load on work and total energy liberation by the myocardium and two types of skeletal muscle. A: In the frog
sartorius, heat production is independent of load, so that work and total energy expenditure are maximal at intermediate loads. B: In the
myocardium and frog gastrocnemius, heat production increases and total energy expenditure remains disproportionately high when the
muscles contract against a heavy load.
Fig. 12-14: Arrangement of muscle fibers in the frog sartorius muscle, where the fibers are parallel and oriented longitudinally, and the
gastrocnemius, where the nonparallel organization of fiber bundles increases internal work at higher tensions.
Effect of Dilatation on Efficiency

Dilatation, like increasing developed pressure, lowers cardiac efficiency by increasing systolic wall stress and reducing the amount of potential
energy stored in the series elasticities that can be used to perform useful work. The increase in wall stress is due to the law of Laplace, while
the geometry of a dilated heart, where ejection of a given volume is associated with a smaller change in dimensions than in a smaller heart,
contributes to the smaller decrease in wall stress when the dilated heart ejects. These geometrical relationships are apparent when the heart is
modeled as a thin-walled sphere in which dilatation has doubled wall stress at a given cavity pressure (Table 12-1, Fig. 12-15). In this example,
ejection of 70 cm3, by the normal ventricle reduces wall tension by almost 40% (from 3.72 × 105 to 2.26 × 105 dyn/cm), whereas ejection of the
same volume by the dilated ventricle decreases wall tension less than 10% (from 5.98 × 105 to 5.59 × 105 dyn/cm). Dilatation also contributes to
the inefficiency of the dilated heart because increased cavity size reduces the fall in wall stress that normally occurs during ejection (see
above).
Effects of Changing Heart Rate on Efficiency

The high energy cost of internal work helps explain the inefficiency associated with increased heart rate. Because internal work must be done
each time pressure develops during isometric contraction, and because much of this work is converted to heat during relaxation, more
P.326
energy is used to perform internal work when more isovolumic contractions are performed per unit time. Another reason that increasing heart
rate is a costly way to increase the minute work of the heart is that energy must be expended during each cardiac cycle to restore ion gradients
across the plasma membrane and to pump calcium into the sarcoplasmic reticulum.
Table 12-1 Effects of Dilatation on Ventricular Energeticsa
Variable Normal Ventricle Dilated Ventricle
At start of ejection:
Pressure (mm Hgb) 100 100
Volume (cm3) 92 380
Radius (cm) 2.8 4.5

Circumference (cm) 17.5 28
Wall tension (dyn/cm) 3.72 × 105 5.98 × 105
Stroke volume (cm3) 70 70
At end of ejection:
Pressure (mm Hgb) 100 100
Volume (cm3) 22 310
Radius (cm) 1.7 4.2
Circumference (cm) 11 26.5
Wall tension (dyn/cm) 2.26 × 105 5.59 × 105
External stroke work (dyn cm) 9.3 × 106 9.3 × 106
Conditions of external work:
Average wall tension (dyn/cm) 2.99 × 105 5.79 × 105
Change in circumference (cm) 6.5 1.5
As% end-diastolic circumference ∼40 ∼5
Ejection fraction 70% 18%
aVolumes, radii, circumferences, and wall tensions are calculated by assuming that the ventricle is a thin-walled sphere that ejects
at a constant pressure.
b1 mm Hg = 1.33 × 103 dyn/cm2.
Clinical Estimates of the Energy Cost of Cardiac Work

The product of heart rate and aortic pressure correlates closely with cardiac oxygen consumption, and so can be used to estimate the energy
expenditure of the intact heart (Gerola et al. 1956; Katz and Feinberg, 1958):
This product, HR × P, is generally referred to as the double product. A somewhat more elaborate index of myocardial oxygen consumption, the
tension-time index (Sarnoff et al., 1958), multiplies average ejection pressure by the duration of ejection (Fig. 12-16); however, this more
cumbersome index has not been shown to be more useful clinically than the double product.
P.327
Fig. 12-15: Effects of dilatation on the relationships between chamber pressure, wall stress, and the change in circumference during
ejection. A: Normal ventricle. B: Dilated ventricle. Cavity outlines, which for ease in calculating wall stress are assumed to be spherical,
are shown at end-diastole (solid circles) and at end-systole (dashed circles); the shaded areas between each pair of circles is the stroke
volume, which is the same in both A and B. Because the dilated ventricle ejects its stroke volume at a higher wall stress and with a
smaller decrease in circumference than the normal ventricle, dilatation increases average wall stress and reduces the amount of energy
stored in the stretched elasticities that can be used to eject blood. (The data used to construct this figure are found in Table 12-1.)
Even though the double product (Equation 12-10) includes only two of the three terms in the equation for minute work (Equation 12-6), as it
omits stroke volume, the latter is the least important determinant of the energy costs of cardiac work. A more important limitation is that this
index does not take into account the effects of chamber size in determining wall stress. In spite of these limitations, the double product is
widely used to measure cardiac energy consumption, for example, during clinical stress testing.
Effects of Changing Hemodynamics on Pressure–Volume Loops

Pressure–volume (PV) loops are useful in understanding how changes in vascular resistance and venous return modify cardiac performance. These
circulatory variables do not influence the end-systolic and end-diastolic PV relationships, which are determined by the properties of
P.328
the myocardium. Instead, altered hemodynamics modifies the changes in ventricular pressures and volumes that are constrained within these
limits.
Fig. 12-16: Cardiac oxygen consumption can be estimated as the tension-time index (TTI), which is the area under the ejection phase of
the left ventricular pressure–volume loop.
Changes in Afterload
The effects of changes in aortic pressure on afterload are complex because, as noted above, ventricular wall stress changes throughout systole.
Afterload is often equated with left ventricular pressure, to which it is related, but this is an oversimplification. The left ventricle “meets” its
afterload at the end of isovolumic contraction, when left ventricular pressure becomes equal to aortic diastolic pressure; this pressure can
therefore be used as an index of left ventricular afterload. As shown in Figure 12-17, increasing afterload increases left ventricular pressure
throughout ejection but reduces stroke volume, whereas decreasing afterload reduces left ventricular pressure throughout ejection but increases
stroke volume.
Fig. 12-17: Effects of changing the pressure at the end of isovolumic contraction, when the left ventricle “meets” its afterload, on the
left ventricular pressure–volume loop. Increased afterload (closed square) reduces stroke volume and increases systolic pressure (dashed
line) while decreased afterload (closed circle) increases stroke volume and decreases systolic pressure (dotted line). Control curves are
shown as solid lines; end-systolic and end-diastolic pressure–volume relationships are not affected by the changes in afterload.
P.329
Fig. 12-18: Effects of changing end-diastolic pressure and volume, which are directly related to preload, on the left ventricular pressure–
volume loop. An increase in preload (closed square) increases stroke volume and systolic pressure (dashed line) while a decrease in
preload (closed circle) reduces stroke volume and systolic pressure (dotted line). Control curves are shown as solid lines; the end-systolic
and end-diastolic pressure–volume relationships are not affected by the changes in preload.
Changes in Preload
Left ventricular preload, which is most accurately defined as ventricular wall stress at the end of diastole, is related to left ventricular end-
diastolic pressure and volume. Increasing preload increases stroke volume and, to a lesser extent, left ventricular pressure throughout ejection;
conversely, decreasing preload reduces both stroke volume and left ventricular pressure (Fig. 12-18).
Changes in Myocardial Contractility (Inotropy): The Family of Starling Curves

Starling curves, which describe the effect of changing cavity volume on the ability of a ventricle to develop pressure (see Chapter 11), are
similar to end-systolic PV relationships. Each Starling curve defines the effect of changing rest length on myocardial performance at any level of
contractility, whereas a change in contractility generates a new Starling curve. The effects of changing inotropic state can therefore be depicted
as a “family of Starling curves,” each of which describes the effects of changing EDV on cardiac performance at each inotropic state. This
interplay is shown in Figure 12-19, where points C, A, and B lie along a control Starling curve and so define a level of contractility. Changes in
EDV can increase (A → B) or decrease (A → C) the
P.330
ability of the heart to develop pressure. A positive inotropic intervention, which by definition increases the ability of the heart to do work at any
given EDV, shifts the heart to a higher Starling curve (curve containing D); conversely, a negative inotropic intervention causes a shift to a lower
Starling curve (curve containing E). Figure 12-19 therefore illustrates the interplay between changing EDV and changing contractility on cardiac
performance.
Fig. 12-19: Family of Starling curves. The solid line that contains the points C, A, and B is an end-systolic pressure–volume (PV)
relationship (Starling curve) which describes myocardial contractility; point A along this line represents the basal state. The two dashed
lines containing points D and E represent additional Starling curves recorded after contractility has changed. Changes in end-diastolic
volume at a constant level of contractility shift the end-systolic point along a given Starling curve; an increase in venous return increases
the ability of the heart to develop pressure from A to B, and a decrease in venous return decreases developed pressure from A to C. The
work of the heart can also be changed by interventions that modify myocardial contractility, which increase or decrease developed
pressure at any end-diastolic volume; for this reason, positive and negative inotropic interventions cause a shift to a new Starling curve.
Starting from the basal state (point A), a positive inotropic intervention increases developed pressure (A to D) and a negative inotropic
intervention decreases developed pressure from (A to E).
Changes in Myocardial Relaxation (Lusitropy): The Family of Filling Curves

Changes in the lusitropic (relaxation) properties of the heart, which represent an additional mechanism that regulates cardiac performance,
generate a family of filling curves that modify preload (Fig. 12-20). Points V, X, and W, which lie along one such filling curve, show the effects of
changing preload on ventricular pressure and volume at a constant lusitropic state. A change in diastolic properties that causes a shift to a new
filling curve changes EDV at any filling pressure (X → Y or Z in Fig. 12-20), which means that the lusitropic state has changed.
P.331
Fig. 12-20: Family of filling curves. Three end-diastolic pressure–volume (PV) relationships have been added to the end-systolic PV
relationships shown in Figure 12-19. The basal end-diastolic PV relationship contains points V, X, and W. A positive lusitropic intervention,
which increases the ability of the ventricle to fill, shifts the end-diastolic PV relationship to the right and downward (curve containing
point Y), while a negative lusitropic intervention, which reduces ventricular filling, shifts this relationship to the left and upward (curve
containing point Z).
Interplay between Changing Inotropy and Lusitropy Pressure–Volume Loops

The PV relationships depicted in Figure 12-20 are the same as those that enclose the PV loops described in Chapter 11, so that the latter can be
used to illustrate the effects of changing inotropic and lusitropic states. The following discussion describes the effects of four interventions on
PV loops; two alter contractility (inotropic properties), and two modify relaxation (lusitropic properties). For simplicity, only the first beat after
an abrupt change in these properties is shown.
Effects of a Negative Inotropic Intervention

A decrease in contractility shifts the end-systolic PV relationship (the Starling curve) to the right and downward (Fig. 12-21). If the first beat
after the negative inotropic intervention encounters the same aortic diastolic pressure, the initial effect is to reduce stroke volume. Ejection
pressure will also be reduced because ejection of blood into the aorta is slowed.
In subsequent beats, EDV will increase because the decreased stroke volume leaves behind a greater ESV. By shifting the loop to the right along
the end-diastolic PV relationship (not shown), the increased preload allows the operation of Starling's law to return stroke volume toward
normal.
P.332
Fig. 12-21: A negative inotropic intervention shifts the end-systolic pressure–volume relationship to the right and downward (solid lines);
if aortic diastolic pressure remains constant, stroke volume will be reduced. The control loop and control end-systolic and end-diastolic
pressure–volume relationships are shown as dotted lines in this and the following three figures.
Effects of a Positive Inotropic Intervention

When contractility is increased, the end-systolic PV relationship shifts to the left and upward (Fig. 12-22). If aortic diastolic pressure remains the
same in the first beat after the inotropic intervention, the greater ability of the ventricle to eject increases stroke volume. Ejection pressure
also increases because ejection of blood into the aorta is more rapid.
EDV will decrease in subsequent beats because the higher stroke volume reduces ESV. Because of the reduced ESV, subsequent beats will be
shifted to lower volumes along the unchanged end-diastolic PV relationship (not shown). The resulting decrease in EDV will, according to
Starling's law, reduce ejection and so return stroke volume toward the control level.
Effects of a Negative Lusitropic Intervention

An intervention that impairs ventricular filling by decreasing lusitropy shifts the end-diastolic PV relationship to the left and upward, which
raises the pressure needed to achieve a given increment in diastolic volume (Fig. 12-23). If the first beat after the lusitropic change begins at the
same end-diastolic pressure and encounters the same aortic diastolic pressure, stroke volume will be reduced. The fall in stroke volume, like
that which follows the negative inotropic intervention described above, increases EDV which, if venous return remains the same, will increase
preload in subsequent beats (not shown). This causes a rightward shift along the new end-diastolic PV relationship that, by increasing EDV, allows
Starling's law to return stroke volume toward normal.
P.333
Fig. 12-22: A positive inotropic intervention shifts the end-systolic pressure–volume relationship to the left and upward (solid lines); if
aortic diastolic pressure remains constant, stroke volume and ejection pressure will increase.
Fig. 12-23: A negative lusitropic intervention shifts the end-diastolic pressure–volume relationship to the left and upward (solid lines); if
aortic diastolic pressure remains constant, stroke volume will be reduced.
P.334
Fig. 12-24: A positive lusitropic intervention shifts the end-diastolic pressure–volume relationship to the right and downward (solid lines);
if aortic diastolic pressure remains constant, stroke volume will increase.
Effects of a Positive Lusitropic Intervention

A positive lusitropic intervention shifts the end-diastolic PV relationship to the right and downward (Fig. 12-24). If filling pressure and aortic
diastolic pressure do not change, the increased EDV will increase stroke volume. The latter, by reducing EDV, will reduce ejection in subsequent
beats (according to Starling's law) and so tend to return stroke volume toward normal.
Interplay Between Venous Return and Cardiac Output Guyton Diagrams

The fact that blood flows in a circle means that regulatory mechanisms are needed to maintain the circulation at a steady state by equalizing
venous return and cardiac output. Cardiac output is matched to venous return by Starling's law of the heart, which increases stroke volume when
more blood returning to the heart increases ventricular end-diastolic pressures, and decreases stroke volume when reduced venous return lowers
ventricular end-diastolic pressures (Fig. 12-25). Atrial pressure also matches venous return to cardiac output because increased atrial pressure
reduces blood flow into the heart, and decreased atrial pressure increases the return of blood to the heart (Fig. 12-25).
The effects of atrial pressure on venous return are not due to a physiological law, but instead result from a simple hydraulic mechanism by which
increased atrial pressure reduces
P.335
blood flow from the venous system to the heart, and vice versa. This mechanism is depicted in Figure 12-26, which shows how changing the
height of a hose connected to a reservoir affects flow through the hose. Raising the outlet of the hose, like raising atrial pressure, slows flow out
of the reservoir by reducing the pressure gradient between the reservoir and the tip of the hose. Flow stops completely when the levels of the
outlet and the fluid in the reservoir are the same; this corresponds to the situation when the heart is stopped, blood flow ceases, and all
pressures in the cardiovascular system come to equilibrium. The pressure recorded when the heart is stopped (e.g., by causing the ventricle to
fibrillate) is called the mean circulatory filling pressure, and is shown by the arrow labeled “mcfp” in Figure 12-25. Mean circulatory filling
pressure in anaesthetized dogs is ∼7 mm Hg, which is closer to the low pressure in the veins than the higher arterial pressures because there is
much more blood in the venous circulation.
Fig. 12-25: “Guyton diagram” showing how atrial pressure matches venous return (dotted line) and cardiac output (dashed line).
Elevating atrial pressure increases cardiac output (Starling's law of the heart) but decreases venous return, whereas decreasing atrial
pressure has the opposite effects. At any steady state, the two curves intersect at the atrial pressure at which flow into and out of the
heart are the same (A). The pressure recorded when the heart is stopped is the mean circulatory filling pressure (mcfp).
Curves such as those shown in Figure 12-25, which are often called “Guyton diagrams” after Arthur Guyton who devised these plots, show that a
rise in atrial pressure reduces blood flow toward the heart and increases the ejection of blood from the heart. Conversely, a fall in atrial
pressure increases venous return and reduces ejection. The effects of changing atrial pressure on flow into and out of the heart are therefore
opposite to one another; venous return falls with increasing atrial pressure, while cardiac output rises when atrial pressure increases. At any
given steady state, the intersection of the two curves identifies the atrial pressure at which venous return and cardiac output are equal to one
another.
Guyton diagrams are useful for understanding how changes in blood volume modify venous return and cardiac output. A decrease in blood
volume, as occurs after hemorrhage, lowers atrial
P.336
pressure and reduces cardiac output (Fig. 12-27, N → D), whereas increased blood volume, as occurs after a transfusion, elevates atrial pressure
and increases cardiac output (Fig. 12-27, N → I). Guyton diagrams also show how a positive inotropic intervention, which increases ejection and
so lowers atrial pressure, increases venous return (Fig. 12-28, N → I), and how a negative inotropic intervention, which reduces ejection and
increases atrial pressure, lowers venous return (Fig. 12-28, N → D). In all of these examples, shifts of the intersection between the curves
relating atrial pressure to venous return and cardiac output establish new steady states at which venous return and cardiac output are equalized.
Fig. 12-26: Effects of raising and lowering the outlet of a hose that drains a reservoir. Flow out of the reservoir depends on the pressure
gradient between the outlet of the hose, which in the heart represents atrial pressure, and the pressure exerted by the fluid in the
reservoir. When the outlet of the hose is at a low level (A), flow out of the reservoir is rapid. Raising the outlet of the hose reduces flow
(B and C); flow stops completely when the pressure at the outlet of the hose is the same as that in the reservoir (D). The fluid level in the
reservoir is equivalent to the mean circulatory filling pressure.
A composite diagram, based on Figures 12-27 and 12-28, shows how these curves can be used to distinguish between the effects of changes in
blood volume and contractility (Fig. 12-29). If a
P.337
P.338
patient is found to have a low cardiac output (dotted horizontal line in Figure 12-29) and atrial pressure is low (closed square), the low cardiac
output can be attributed to a decrease in blood volume, such as occurs after a hemorrhage. If, on the other hand, atrial pressure is high (closed
circle), the low cardiac output can be attributed to decreased contractility, such as occurs after an acute myocardial infarction.
Fig. 12-27: Curves relating venous return and cardiac output showing effects of changing circulating blood volume. Increased blood
volume increases venous return, whereas decreased blood volume reduces venous return. N, normal; I, increased blood volume; D,
decreased blood volume.
Fig. 12-28: Curves relating venous return and cardiac output showing effects of changing contractility. Increased contractility decreases
right atrial pressure, which allows venous return to increase. Conversely, decreased contractility raises atrial pressure and so lowers
venous return. N, normal; I, increased contractility; D, decreased contractility.
Fig. 12-29: Composite diagram, based on Figures 12-27 and 12-28 showing the interactions between changes in blood volume and
contractility. N: normal, Ibv: increased blood volume, Dbv: decreased blood volume, Ic: increased contractility, Dc: decreased
contractility. The dotted horizontal line shows how these curves can identify the cause of a fall in cardiac output: If atrial pressure is
below normal (closed square), the low cardiac output can be attributed to decreased blood volume (Dbv). If, however, atrial pressure is
above normal (closed circle), the low cardiac output can be attributed to decreased contractility (Dc).
Clinical Indices of Myocardial Contractility (Inotropy)

Efforts to identify clinical indices of myocardial contractility were stimulated in the 1960s by the need to identify the optimal time for surgical
management of valvular heart disease. If a valve is replaced too soon, exposure to such hazards as embolism, infection, and deterioration of the
prosthetic valve is unnecessarily prolonged, whereas if surgery is delayed too long, even a technically perfect operation cannot help the patient
because prolonged overload irreversibly damages the myocardium (see Chapter 18). The following discussion highlights a few indices that
illustrate some clinical applications of the physiological principles described in this text.
Indices Based on Pressure Measurements

Left ventricular diastolic pressure cannot provide an accurate measure of ventricular performance because it is influenced by both peripheral
resistance and venous return. However, several clinically useful indices of myocardial contractility are based on analyses of pressure
measurements obtained
P.339
during isovolumic contraction. A major advantage of these indices is that data are collected before the aortic valve opens, and so are
independent of afterload. The simplest isovolumic index of left ventricular contractility is dP/dtmax, the maximum rate of pressure rise.
However, dP/dtmax is also influenced by EDV, the thickness of the left ventricular wall, regional abnormalities in left ventricular function, a leaky
mitral valve, and loss of the normal synchronicity of contraction. The quotient (dP/dt)/P, obtained by dividing dP/dt continuously by
instantaneous left ventricular pressure, is less influenced by preload than dP/dtmax, but is also less sensitive to changes in contractility.
The velocity of fiber shortening during isovolumic contraction (VCE), which uses pressure measurements to estimate shortening velocities in the
walls of the ventricle, requires the selection of a stiffness constant to convert changing wall stress to changes in the length of the contractile
elements. VCE, along with other pressure-derived indices, is subject to the limitations described for dP/dtmax.
Indices Based on Measurements of Volume and Dimensions

Measurements of ventricular volume and wall motion are very useful in characterizing ventricular architecture, especially in ischemic heart
disease where regional wall motion abnormalities can identify patients in whom occlusion of one or more coronary arteries has caused a
localized contractile abnormality (see Chapter 17). They are also very useful in diagnosing and classifying heart failure (see Chapter 18), but are
of little value in quantifying myocardial contractility.
Ejection fraction (EF), which is commonly used to evaluate ventricular performance, is the fraction of the EDV that is ejected as the stroke
volume:
EF is therefore the ratio between a physiological variable (stroke volume [SV]), and an architectural variable (end-diastolic volume [EDV]). In
humans, left ventricular EF is normally greater than 55%. Fractional wall shortening, which is analogous to EF except that it is calculated using
dimensions measured by echocardiography instead of cavity volumes, has the same limitations as a measure of contractility.
EF is influenced by changes in SV, the numerator in Equation 12-11, and so is highly dependent on hemodynamic variables that include heart
rate, preload, and afterload. However, the major determinant of EF in patients with heart disease is EDV, the denominator. Because of the
limited extent to which cardiac output can fall without killing the patient, the major cause of a low EF in patients with heart failure is not a
decrease in SV; instead, a low EF is due largely to increased EDV. These patients are often said to have systolic heart failure (also called heart
failure with low ejection fraction, or HFlowEF–see Chapter 18). Although EF is a useful predictor of long-term prognosis in these patients, it
provides surprisingly little information about exercise intolerance or the extent of the clinical disability. A slightly reduced, normal, or even
elevated EF in a patient with heart failure is the basis for a diagnosis of diastolic heart failure (also called heart failure with normal ejection
fraction, or HFnEF–see Chapter 18).
Systolic Time Intervals

External carotid pulse recording, phonocardiography, and electrocardiography allow the phases of the cardiac cycle to be timed and related to
the mechanical events occurring in the left ventricle (Fig. 12-30, Table 12-2). The duration of left ventricular systole can be estimated as the
P.340
Q-S2 interval, which begins with the QRS complex on the electrocardiogram and ends with S2, the second heart sound caused by aortic valve
closure. Left ventricular systole can be subdivided into three phases: the Q-S1 interval, measured from the beginning of the QRS complex to
mitral valve closure which generates S1 (the first heart sound); isovolumic contraction time, between S1 and aortic valve opening; and left
ventricular ejection time (LVET), measured from the upstroke of the carotid pulse (the beginning of left ventricular ejection) to the dicrotic
notch (aortic valve closure). Because there is a delay in the transmission of the pulse wave from the aortic root to the carotid artery, isovolumic
contraction time is estimated by subtracting the interval between S2 and the dicrotic notch from the interval between S1 and the carotid
upstroke. The pre-ejection period (PEP), obtained by subtracting LVET from the Q-S2 interval, measures the interval between the onset of
ventricular depolarization and the beginning of ejection. Negative inotropic interventions prolong PEP and shorten LVET, so that the ratio
PEP/LVET can be used to identify patients with depressed myocardial contractility. However, systolic time intervals provide only rough estimates
of contractility because they are influenced by heart rate, preload, afterload, and abnormalities in left ventricular depolarization.
Fig. 12-30: Systolic time intervals can be calculated from the carotid pulse (upper), phonocardiogram (middle), and electrocardiogram
(lower). Delayed transmission of the arterial pulse from the aortic root to the carotid artery causes S2 to precede the dicrotic notch in the
aortic pulse tracing. LVET (left ventricular ejection time) is the interval between the carotid upstroke and the dicrotic notch. The Q-S2
interval represents the total duration of electromechanical systole. Subtraction of LVET from the Q-S2 interval yields the pre-ejection
period (PEP).
P.341
Table 12-2 Systolic Time Intervals
Interval Measurement Physiological Correlation
Q-S2 From beginning of QRS to first high-frequency Total electromechanical systole

vibration of S2
Q-S1 From beginning of QRS to beginning of S1 Excitation-contraction coupling
Isovolumic contraction time From S1 to onset of rise in aortic pressure Excitation-contraction coupling
Left ventricular ejection time From onset of carotid upstroke to the dicrotic notch Total ejection
(LVET)
Pre-ejection period (PEP) Q-S2 minus LVET Isovolumic contraction plus Q-S1
interval
Clinical Indices of Filling (Lusitropy)

Diastolic function can be evaluated clinically by such simple methods as auscultation, where an audible third or fourth heart sound implies that
the ventricle has become stiff. Doppler measurements of blood flow across the mitral valve provide useful noninvasive indices of diastolic
function. These include the E/A ratio, which measures the ratio between early and late maxima of flow velocity; the former (E) is related to
rapid filling, while the latter (A) is due to atrial systole. More elaborate indices of diastolic function include the percentage of end-diastolic
filling that occurs during the first third of diastole, average rapid filling rate, peak filling rate, and the time to peak filling rate. Like the indices
used to estimate contractility, these are generally rather imprecise.
Lusitropic abnormalities can modify the rate of pressure fall during isovolumic relaxation (-dP/dt), the rate of filling during early diastole
(dV/dt), and the slope of the curve relating the increase in ventricular volume to the pressure rise in late diastole. The latter can be expressed
either as compliance (dV/dP, the slope of the curve relating the change in volume for each increment of pressure), or as stiffness, the reciprocal
of compliance (dP/dV, which is the slope of the curve relating the change in pressure to that of volume).
A once commonly used index of diastolic function is the maximum rate of left ventricular pressure fall during isovolumic relaxation, -dP/dtmax.
However, -dP/dtmax is highly dependent on aortic pressure, so that corrections must be made to obtain a pressure-independent index. When the
rate of decline of left ventricular pressure is exponential, a time constant τ can be calculated to describe -dP/dt. However, the fall of left
ventricular pressure during isovolumic relaxation is not usually exponential, so that more elaborate equations are often needed to calculate τ.
The complexity of these determinations generally limits these measurements to the research laboratory.
Conclusions
The hemodynamic principles described in this chapter represent the foundation of modern cardiology. Although conceptual and technological
advances are providing new and useful tools to
P.342
evaluate the many variables that determine cardiac performance, definition of pathophysiology and formulation of a therapeutic plan still
requires a trained observer who can organize and synthesize the vast array of data that can be provided by specialized–and often expensive–
technology. As we learn more about pathophysiology, greater skill is needed to integrate the many lines of data that can be obtained from a
given patient.
It is often impossible to fit all of the clinical data into a coherent pattern, which makes it difficult to come up with an accurate diagnosis. Many
years ago a colleague suggested: “when you can't put everything together, you should begin by throwing out the high technology.” This approach
is also stated in the classical aphorism: “Listen to the patient.”
Bibliography
Krayenbuehl HP, Hess OM, Turina J. Assessment of left ventricular function. Cardiovasc Med 1978;2: 883–910.
Pouleur H, Rousseau MF, van Eyll C, et al. Assessment of left ventricular contractility from late systolic stress-volume relations. Circulation
1982;65:1204–1212.
Weissler AM, Garrard CL Jr. Systolic time intervals in cardiac disease. Mod Concepts Cardiovasc Dis 1971; 40:1–8.
See also the Bibliography for Chapter 11.
References
Evans CL, Matsuoka Y. The effect of various mechanical conditions on the gaseous metabolism and efficiency of the mammalian heart. J
Physiol (Lond) 1915;49:378–405.
Gerola A, Feinberg H, Katz LN. The oxygen cost of cardiac hemodynamic activity. Physiologist 1956;1:31.
Katz LN, Feinberg H. The relation of cardiac effort to myocardial oxygen consumption and coronary flow. Circ Res 1958;6:656–669.
Martin DS. The relation between work performed and heat liberated by the isolated gastrocnemius, semitendinosus and tibialis anticus
muscles of the frog. Am J Physiol 1928;33:543–547.
Nikolic S, Yellin EL, Tamura K, et al. Passive properties of canine left ventricle: diastolic stiffness and restoring forces. Circ Res
1988;62:1210–1222.
Sarnoff SJ, Braunwald E, Welch GH Jr, et al. Hemodynamic determinants of oxygen consumption of the heart with special reference to the
tension-time index. Am J Physiol 1958;192:148–156.
Warner HR, Toronto AF. Regulation of cardiac output through stroke volume. Circ Res 1960;8:549–552.
Weber KT, Janicki JS, Shroff SV, et al. Contractile mechanics and interaction of the right and left ventricles. Am J Cardiol 1981;47:686–695.
Woods RH. A few applications of a physical theorem to membranes in a state of tension in the human body. J Anat Physiol 1892;26:362–370.
> Table of Contents > Part Three - Normal Physiology > Chapter 13 - Cardiac Ion Channels
Chapter 13
Cardiac Ion Channels
Cardiac myocytes are activated by an electrical signal, the action potential, in which changes in electrical
potential across the plasma membrane (Em) result from an elaborate sequence of openings and closing of ion
channels. These action potentials, which differ in various regions of the heart, organize the mechanical
events of the cardiac cycle and help provide the homogeneity of activation needed to optimize the
efficiency of cardiac contraction (see Chapter 6).
Membrane Potential
A microelectrode inserted through the plasma membrane of a resting cardiac myocyte records an
intracellular potential that is more negative than that outside the cell (Fig. 13-1). How this potential
difference is described is determined by convention. When expressed in terms of the potential recorded
within the cell, with the surrounding medium viewed as zero, resting potential is negative. If, on the other
hand, resting potential is described in terms of the potential outside the cell, and intracellular potential
viewed as zero, resting potential is positive.
Much confusion arises because different conventions are used in cardiac electrophysiology and
electrocardiography! Membrane electrophysiologists, who place electrodes inside the cell and measure the
potential of the cell interior relative to that in the surrounding medium, describe resting potential as
negative, whereas clinical electrocardiographers, who place electrodes on the body surface, view the
myocardium from the outside, and describe resting potential as positive. This chapter and Chapter 14, which
describe cellular events, view resting potential as negative, whereas Chapters 15 and 16, which describe
electrocardiograms, view resting cells as positively charged.
A change in membrane potential that decreases the electronegativity inside a resting myocyte is called
depolarization (Fig. 13-2). The return of membrane potential toward its resting level in a depolarized
myocyte is repolarization, whereas an increase in the resting potential of an unexcited cell is
hyperpolarization. Action potentials must include at least two phases: one of depolarization; the other of
repolarization. Action potential amplitude, which is usually expressed in millivolts (mV), defines the extent
to which cellular electronegativity changes from its resting level, including any reversal to electropositivity.
Membrane Currents
The most important physiological currents in the heart are carried by ions that cross the plasma membrane,
and so are ionic currents. By convention, these are described as if they were carried by positively charged
ions. Capacitive currents, on the other hand, are generated by electron fluxes toward and away from the
surfaces of the plasma membrane.
P.344
Fig. 13-1: Resting cardiac muscle cell. The normal resting potential is negative inside the branched
myocytes and positive in the extracellular space.
An inward current, according to electrophysiological convention, is the flux of charge that would occur if a
positive ion moved across the plasma membrane into the cell. Because the interior of the resting cell is
negatively charged, inward currents cause depolarization. Most inward currents in the heart are generated
when positively charged sodium and calcium ions enter the cell; however, the outward movement of a
negative ion has the same effect on membrane potential as the inward movement of a positive ion, so that
the efflux of an anion like chloride also generates an inward current (Table 13-1). An outward current can be
generated when a positively charged ion leaves the cell interior, or when a negatively charged ion enters the
cell. Both increase electronegativity within the cell. In resting cells, outward currents cause
hyperpolarization, while outward currents that follow depolarization, and so return membrane potential
toward the resting negativity, cause repolarization (Fig. 13-2).
Unlike ionic currents, which are generated when ions cross the lipid bilayer through pores in the plasma
membrane, capacitive currents are generated by the movements of electrons relative to the membrane
surface (Fig. 13-3). The latter occur in biological tissues because the phospholipid bilayer is a capacitor
which is made up of an insulator (the hydrophobic core) that is surrounded by two layers of polar molecules
(the phospholipid head groups) (see Chapter 1). A cathode placed outside a resting cell, where the plasma
membrane is positively charged, reduces the positive charge on the extracellular surface of the membrane
and, at the same time, causes
P.345
negative charge to move away from the intracellular surface. These charge movements, which are carried by
electrons, cause the membrane to depolarize by discharging membrane capacitance.
Fig. 13-2: A cardiac action potential showing some conventions of cardiac electrophysiology. The
resting cell is viewed from within, so that resting potential is negative. Hyperpolarization increases
resting potential. Depolarization occurs when resting potential is decreased, while repolarization
occurs when membrane potential returns to its resting level at the end of an action potential. The
amplitude of the action potential is the extent to which activation reduces cellular electronegativity
from its resting level plus any reversal to electropositivity.
Table 13-1 Ions as Charge Carriers Across Cell Membranes
Ion Charge Direction of Current Effect on Membrane

Passive Flux Generated Potential
Calcium Positive Inward Inward Depolarization
Sodium Positive Inward Inward Depolarization
Potassium Positive Outward Outward Repolarization
Chloride Negative Inward Outward Repolarization
Action potentials are normally initiated by capacitive currents that are generated when a wave of
depolarization approaches a region that contains resting cardiac myocytes. These currents depolarize the
resting membrane when the approaching wave of depolarization causes the extracellular surface of the
plasma membrane to become negative and the interior surface to become positive. Action potentials are
propagated along the cell surface when depolarization opens voltage-gated ion channels that carry
additional inward current, which cause this process to become regenerative (see Chapter 14).
Membrane Resistance, Permeability, and Conductance

The ratio between membrane potential and current flow is membrane resistance, which can be described by
Ohm's law:
where R is resistance, E is potential, and I is current flow. Membrane conductance, designated g, is the
reciprocal of membrane resistance, so that
Fig. 13-3: Differences between capacitive and ionic currents. A: A capacitive current is generated when a
cathode is placed outside a resting cell. Movement of electrons from the cathode toward the extracellular
surface of the plasma membrane discharges the positive potential outside the cell, and causes electrons to
move away from the inner surface. B: An inward ionic current is generated when cations move across the
membrane from outside to inside the cell.
P.346
Permeability and conductance both define the ability of a substance to cross a membrane, but they are not
the same. Permeability (P) is the ability of a membrane to allow the movement (flux) of molecules in both
directions, from either side of the membrane to the other, whereas conductance describes the movement of
charge in one direction across the membrane.
Permeability, which is commonly used to describe systems that are at or near equilibrium, is defined by the
relationship:
which states that the flux of a molecule is equal to the permeability coefficient for the molecule (-Pmolecule)
times the difference in the concentration of the molecule at the two sides of the membrane (δcmolecule).
Conductance, which characterizes the electric current generated by the movement of an ion across a
membrane down a pre-existing electrochemical gradient, is defined by the relationship:
which states that the current generated by an ion flux (iion) is determined by the conductance of the
membrane for the ion (gion) and the electromotive force that drives the ion across the membrane (Em - Eion),
which is the difference between the actual transmembrane potential (Em) and the transmembrane potential
at which there would be no net ion flux (Eion). The latter is called the equilibrium potential for the ion (see
Chapter 14). For example, the current carried by potassium ions (iK) is:
where gK is the potassium conductance and (Em - EK) is the electromotive force that moves potassium ions
across the membrane. Depending on the difference between Em and EK, current can flow in either direction
across the membrane, although not in both directions at any given time. Biological currents are expressed in
Siemens, a unit that has the same units as mho, the reciprocal of the ohm that measures resistance.
Changing membrane potential can modify the conductance of a voltage-gated ion channel for a given ion.
This property, called rectification, plays an important role in regulating changes in the potential across
biological membranes (see Chapter 14).
Generation of the Action Potential

In 1902, Bernstein, who knew that mammalian cells contain a high concentration of potassium, proposed
that the plasma membrane is selectively permeable to potassium ions and that fixed negative charges on the
cytosolic proteins establish a Donnan equilibrium which concentrates potassium ions in an electronegative
cytosol (Bernstein, 1902). This hypothesis was confirmed shortly before the outbreak of World War II, when
intracellular potentials were first measured by Hodgkin and Huxley (1945) in England, and Cole and Curtis
(1941) in the United States, who inserted microelectrodes into squid giant axons. Their initial measurements
confirmed Bernstein's prediction that the interior of resting cells was electronegative, and that this
negativity decreased during excitation. However, intracellular recordings carried out after World War II,
when more advanced equipment had become available, yielded a surprise. This was that membrane
potential did not simply decrease to zero during excitation, as would be expected if reduction of potassium
permeability
P.347
alone was responsible for activation. Instead, membrane potential was found to reverse during excitation,
when the cell interior became positive. This meant that action potentials are generated by processes more
complex than dissipation of the potassium gradient.
The Voltage Clamp

In the 1950s, Hodgkin, Huxley, and B. Katz published a series of classical experiments that characterized the
ionic basis for the action potential in squid giant axons. Conceptually, their approach was simple: Instead of
measuring potential changes during an action potential, when ionic currents depolarized and repolarized the
membrane, they measured the currents needed to hold membrane potential at a constant level. This was
accomplished by placing an electrode inside the axon, exciting the membrane, and then applying currents
across the membrane to keep membrane potential from changing. The applied currents, which exactly
matched the currents that would have otherwise caused membrane potential to change, represent the
“voltage clamp.” Measurements of these applied currents made it possible to quantify both the magnitude
and timing of physiological ionic currents that, by flowing in the opposite direction, generated the squid
axon action potential. Interventions such as changing sodium and potassium concentrations inside and
outside the axon demonstrated that depolarization occurs when sodium ions enter the cell, that
repolarization is caused by potassium efflux, and that reversal of membrane potential at the peak of the
action potential is due to the large sodium influx.
General Properties of Plasma Membrane Ionic Currents

Most of the ion fluxes responsible for plasma membrane depolarization and repolarization do not require the
expenditure of energy because the ions move downhill along their electrochemical gradients. These ion
fluxes are mediated by members of an extended family of membrane proteins that contain ion-selective
pores which, when open, favor the passage of a single ion species. In most cardiac channel proteins, changes
in membrane potential open and close these pores, so that these ion channels are generally referred to as
voltage-gated.
Most voltage-gated ion channels, once opened, do not remain in the open state, but instead cycle through at
least two closed states (Fig. 13-4). The physiological transition of the heart's sodium and calcium channels
from the closed (resting) to the open state, called activation, occurs when depolarization increases the
probability of channel opening (see below). At the same time, however, depolarization also increases the
probability that these channels will assume the closed (inactive) state; a process that is called inactivation.
This allows the depolarizing signal that initially opens a resting channel to cause its subsequent closure.
Once closed, the channel assumes a refractory state, called the closed (inactive) state, where the channel
cannot be reopened by additional depolarizing stimuli.
A single depolarization causes sodium and calcium channels to open transiently because activation is faster
than inactivation. This dual response to membrane depolarization is an example of a general principle in
biological regulation, discussed in Chapter 8, that mechanisms which initiate a response also prevent
runaway signaling by initiating slower processes that end the response. In this way, ion channels resemble
the water taps commonly found in public washrooms, where flooding is prevented when the same signal that
starts the flow of water also shuts off the tap, albeit more slowly.
P.348
Fig. 13-4: Schematic diagram showing two closed states and one open state of a voltage-gated ion channel.
Changes in membrane potential initiate transitions between these states that open the channel (activation),
cause the channel to close in a refractory state where it cannot be reopened (inactivation), and reactivate
the channel by ending this refractoriness (recovery). Channel opening is caused by movement of an “m” gate,
and inactivation and reactivation by movements of an “h” gate.
Reactivation (recovery), the transition from the closed (inactive) state to the closed (resting) state where
the channel can again be opened by depolarization, requires an additional signal. In the heart, sodium and
calcium channels are reactivated when membrane potential is returned to its resting level by repolarizing
(outward) currents carried by potassium channels. Once sodium and calcium channels are reactivated and
their refractoriness ends, they can again be opened by depolarizing stimuli.
Ion Channel Gating

The mechanisms that control the transitions between the various states of an ion channel are often called
gating mechanisms. Initially presented as coefficients in mathematical equations that describe the opening
and closing of the channels in the squid axon (Hodgkin and Huxley, 1952), channel gating mechanisms can
now be related to conformational changes in specific regions of ion channel proteins.
The equations developed to characterize the behavior of the sodium current that depolarizes the squid axon
contain two coefficients, m and h, that characterize the regulation of channel conductance. Because m is a
coefficient of channel opening while h is a coefficient of channel closing, sodium current is maximal when m
is 1 (100% probability of being open) and h is 1 (0% probability of being closed). This relationship is described
by the following modification of Equation 13-4:
P.349
Table 13-2 Sodium Channel Gating
Channel State State of the m Gate State of the h Gate
Closed (resting) Closed Open
Open (active) Open Open
Closed (inactive) Open Closed
Equation 13-6 states that the inward sodium current (iNa) is determined by the maximal sodium conductance
(gNa), m and h, and the difference between actual membrane potential (Em) and the sodium equilibrium
potential (ENa). The latter, which is generally about +20 mV, is the potential that would be recorded if the
membrane were permeable only to sodium ions (see Chapter 14). The coefficients m and h are now known to
describe the properties of two regions of the sodium channel, called the “m gate” and “h gate” (Fig. 13-2).
The fact that m appears in the third power initially suggested that each channel contains three m gates; the
correct number, as described below, is now known to be four.
Hodgkin and Huxley predicted that depolarization of the resting membrane rapidly opens the m gate while,
at the same time, causing the h gate to close more slowly. These different time-dependent properties of m
and h explain why the first effect of depolarization is to activate the channel by rapidly opening the m
gates, why slower closure of the h gate inactivates the channel, and why the open state is brief. The
Hodgkin–Huxley equations therefore define three channel states: an open state in which both gates are open,
and two closed states (Table 13-2). In one of the latter, the m gate is closed and the h gate open (closed,
resting); in the other, the m gate is open and the h gate is closed (closed, inactive). The difference between
the two closed states is that in the closed, resting state, the m gate can be opened by membrane
depolarization, while in the closed, inactive state, the h gate can be opened only by repolarizing potassium
currents that return membrane potential to its resting level (Fig. 13-4). Because the m gates remain closed
during reactivation, sodium channels do not reopen unless the membrane is depolarized again.
Hodgkin and Huxley also developed equations to describe the squid axon potassium channel, whose opening
was found to depend on the fourth power of a coefficient that they called n:
One of the more remarkable facts about the Hodgkin–Huxley equations is that they anticipated key features
of the structure of voltage-gated ion channels. For example, each sodium channel contains four α-helical
transmembrane segment rich in positively charged amino acids that are now known to be the activation (m)
gates, while the inactivation (h) gate is a single intracellular peptide chain that responds to depolarization
by assuming a conformation that occludes the inner mouth of the channel pore (see below).
Ion Channel Proteins
Voltage-gated ion channels are members of a family of tetrameric proteins that evolved from a monomeric
ancestor that, through gene duplication and divergence, gave rise to the modern channels (Fig. 13-5).
Around the time that eukaryotes evolved from prokaryotes, the ancestral
P.350
monomeric channel protein gave rise to hyperpolarization-activated channels and potassium channels whose
four subunits are not linked covalently. Calcium channels are the most primitive voltage-gated channels
whose four subunits are covalently linked; later, when multicellular metazoan phyla evolved at the beginning
of the Cambrian period, the gene encoding the calcium channels diverged to give rise to sodium channels,
whose large action potentials conduct much more rapidly than calcium-dependent action potentials. The
rapid conduction made possible by these sodium currents (see Chapter 16) is essential for allowing
multicellular organisms to coordinate the movements of the different regions of their bodies.
Fig. 13-5: Molecular architecture of a potassium channel as represented in “Birth of an Idea,” a

sculpture by Julian Voss-Andrae in 2007 that is based on atomic coordinates determined by Zhou et al.
(2001). The sculpture shows the pore, depicted as contained within a wire mesh, and several
surrounding transmembrane α-helices. (Commissioned by Roderick MacKinnon and reproduced under
the terms of the GNU Free Documentation License).
Pores and Gates

One of the fundamental properties of plasma membrane channels is their selectivity, which allows calcium
channels to conduct calcium ions, sodium channels to conduct sodium ions, etc. Although this selectivity is
not always absolute, the preferences of many channels for a given ion can be very stringent, as in calcium
channels, which under physiological conditions conduct mainly calcium in spite of the almost hundredfold
greater concentration of sodium ions in the extracellular fluid. Selectivity is made possible by ion-binding
sites within the channel that allow only the “correct” ion to enter the pore. For example, the anionic groups
of potassium channels are arranged within the pore in a conformation that recognizes potassium ions (Fig.
13-6). Because other ion species bind with lower affinity to these anionic sites, they are excluded from the
selectivity filter by the bound potassium ions. The energy for the ion flux is provided by the concentration
gradient across the bilayer; ions enter the channel from the side of the membrane containing the higher ion
concentration because these ions are most likely to displace ions already bound within the channel.
P.351
Fig. 13-6: Schematic diagram showing the structure of a voltage-gated potassium channel. A: The
pore through which potassium ions cross the membrane bilayer contains two “vestibules,” a central
cavity and a selectivity filter. The selectivity filter is lined with anionic groups that recognize the
cation that is allowed to pass through the pore when the channel is open. B: Enlarged view of the
selectivity filter showing potassium ions moving in single file through the channel when they interact
with anionic binding sites on the two sides of this region of the pore. The driving force for the ion flux
is the higher concentration of potassium inside the cell than in the extracellular space.
P.352
Gating Currents, Inactivation Gates, and Inactivation Particles

To explain the ability of changing membrane potential to open and close voltage-gated ion channels, Hodgkin
and Huxley predicted that the gating regions contain charged regions that move in response to changing
membrane potential. This led to the prediction that movement of these charged regions in response to
changing membrane potential would generate small “gating currents.” In the 1960s, careful measurements
carried out under conditions that eliminated or corrected for changing membrane capacitance and ion fluxes
were able to identify the gating currents caused by movements of positively charged regions of the sodium
channel that correspond to the activation (m) gates (see below).
Closure of the h gates is now known to occur when depolarization causes a conformational change in an
intracellular region of the sodium channel. This leads to the formation of an “inactivation particle” within
the cytosol that blocks the inner pore of the open channel (Fig. 13-7).
Channel Subunits
Cloning of voltage-dependent cardiac ion channels has identified a large number of channel protein subunits
that are found in different regions of the heart (Gaborit et al., 2007). Most important are the α-subunits,
most of which form a pore that, when open, allows selected ions to cross the bilayer. The majority of these
ion channels operate as tetramers in which amino acid sequences, called P-loops, surround a pore (Fig. 13-
8). α-Subunits that contain six α-helical transmembrane segments (Fig. 13-8A) are found in sodium, calcium,
and outward rectifying potassium channels, and two related channel types, called hyperpolarization-
activated channels (HCN) and cyclic nucleotide-gated channels (CNG), that can carry both sodium and
potassium. Inward rectifying potassium channels are made up of smaller α-subunits that contain only two α-
helical transmembrane segments (Fig. 13-8B), while the α-subunits in two-pore channels are made up of two
P-loops (Fig. 13-8C). Another type of potassium channel called BK (because it interacts with the
dihydropyridine calcium channel agonist BAY K 8644), which is activated by calcium, contains seven α-helical
transmembrane segments (Fig. 13-8D). The structures and pore regions of the major subunits of the CIC
chloride channels (Fig. 13-8E), which mediate anion fluxes across the plasma membrane, differ considerably
from all of the cation channels. Many, but not all, of these channels are regulated by smaller β-subunits.
The four α-subunits in sodium and calcium channels are connected by intracellular linking segments (Fig. 13-
9A), whereas the four α-subunits in the channels that carry transient outward potassium currents, outward
rectifying potassium currents, and hyperpolarization-activated channels are discrete proteins that are not
linked covalently (Fig. 13-9B). The assembly of these α-subunits to form functional channels is constrained
by “identity tags” that favor interactions between some α-subunits, and inhibit interactions between others.
Inward rectifying potassium channels function as tetramers of smaller α-subunits that contain only a P-loop
with its two transmembrane α-helices (Fig. 13-9C). Channels made up of two-pore α-subunits (Fig. 13-9D)
function as dimers; however, these channels resemble the tetrameric channels described above because
each α-subunit contains two P-loops. Most two-pore channels are not voltage-gated, but instead can be
activated by stretch, fatty acids, protons, and other signals (see below).
P.353
Fig. 13-7: Schematic diagram showing the h gate of a sodium channel. A: In the resting state, the h
gate, which is a portion of the intracellular peptide chain of the channel protein, does not occlude the
inner pore. B: Plasma membrane depolarization causes a conformational change in the h gate that
leads to the formation of an inactivation particle that blocks the intracellular opening of the channel
pore.
P.354
Fig. 13-8: Schematic representation of five types of cardiac ion channel. A: The α-subunits in the sodium and
calcium channels, the channels that carry transient outward potassium currents, the outward rectifying
potassium currents, and the hyperpolarization-activated channels, all contain six transmembrane α-helices.
The positively charged S4 transmembrane segment in these domains provides the voltage sensor that
responds to membrane depolarization by opening the channel. In sodium channels this is the “m gate”
described by the Hodgkin–Huxley equations. The channel pore, called the “P-loop,” is made up of the S5 and
S6 transmembrane segments and an intervening peptide chain that “dips” into the membrane bilayer. B: α-
Subunits of single-pore inward rectifier potassium channels are made up largely of a P-loop, which contains
M1 and M2 transmembrane segments and the intervening peptide chain, and lack a charged transmembrane
segment homologous to S4. C: The α-subunits of two-pore channels are made up of two covalently linked
regions that are homologous to the single-pore inward rectifier potassium channels. D: α-Subunit of a BK
channel showing seven transmembrane α-helices. E: α-Subunit of a CIC chloride channel showing 18 α-helices
(labeled a to r). The N- and C-terminal α-helices (a and r) are in the cytosol; the other 14 (b to q) penetrate
the plasma membrane to different depths.
P.355
Fig. 13-9: Schematic representation of four types of voltage-gated ion channel. A: Sodium and calcium
channels are covalently linked tetramers made up of four of the larger α-subunits shown in Figure 13-8
(numbered I to IV), each of which contains six α-helical transmembrane segments. B: The channels that carry
transient outward potassium currents, outward rectifying potassium currents, and hyperpolarization-
activated channels are also made up of the larger α-subunits shown in Figure 13-8, but the α-subunits are not
linked covalently. C: Inward rectifier potassium channels include four of the one-pore subunits shown in
Figure 13-8. D: Two-pore channels are made up of two of the two-pore α-subunits shown in Figure 13-8.
The six α-helical transmembrane segments in the larger α-subunits shown in Figure 13-9A and 13-9B are
organized in the membrane so that the S5 and S6 α-helices, along with the intervening peptide chain, form
the P-loop that allows a given ion species to pass through the open channel, while the S1, S2, and S3 α-helical
segments interact with the surrounding membrane lipids (Fig. 13-10). The voltage sensors in these channels
are the S4 α-helical transmembrane segments, which contain positively charged arginine and lysine residues;
movements of these charged regions, which are responsible for the gating currents described above, allow
membrane depolarization to open the channel (Figs. 13-11). Inactivation gates, on the other hand, are
intracellular peptide loops that form “inactivation particles” which occlude the inner mouth of the pore
(Figs. 13-7 and 13-10). In sodium channels, the inactivation particle is the cytoplasmic peptide chain that
connects the S6 transmembrane α-helix of domain III to the S1 α-helix of domain IV (Fig. 13-9). The high
degree of recognition between the channel pore and an inactivation particle can be documented when the
latter are synthesized as small peptides that, when released into the cytosol, inactivate the channel.
P.356
Fig. 13-10: Schematic three-dimensional representation of the major states of the sodium channel.
The S5 and S6 α-helical transmembrane segments along with the intervening peptide chains of the four
α-subunits surround the pore, while the S1, S2, and S3 α-helical transmembrane segments (not
labeled) allow the channel to interact with the bilayer. A: In the resting state, the pore is closed. B:
The channel is opened when the four charged S4 transmembrane segments (the m gates or voltage
sensors) shift their positions in response to membrane depolarization. C: The channel closes and
becomes refractory when the intracellular peptide chain that connects the III and IV α-subunits (the h
gate or inactivation particle) occludes the inner mouth of the pore.
P.357
Fig. 13-11: Schematic diagram showing the m gate (voltage sensor) of a sodium channel. A: In the
closed (resting) state, where the extracellular surface of the membrane is positively charged and the
interior is negatively charged, the S4 transmembrane segment is in a conformation that closes the
channel. B: Depolarization of the plasma membrane opens the channel by shifting the position of the
S4 transmembrane segment away from the plane of the bilayer.
Sodium and Calcium Channels
A single class of sodium channel α-subunit, called Nav 1.5, along with four different β-subunits are found in
human hearts (Table 13-3). The α-subunit is encoded by the gene SCN5A, while the β-subuits are encoded by
genes called SCN1β, SCN2β, SCN3β, and SCN4β. Two different types of calcium channel, called L- and T-type,
are also found in mammalian hearts; these names reflect the slower inactivation of L-type calcium channels
than T-type channels (see Chapter 14). L-type channels, which play a central role in excitation-contraction
coupling (see Chapter 7) predominate in the t-tubules of working myocytes in the atria and ventricles, while
T-type channels, which mediate pacemaker activity and proliferative signals, are found in atrial myocytes
(but not in humans), nodal cells, and the rapidly conducting cells of the His-Purkinje system. Two α-subunit
isoforms are found in both L- and T-type calcium channels, and the L-type channels, but not the T-type
channels, also contain small subunits (Table 13-3).
P.358
Table 13-3 Sodium and Calcium Channel Subunits in the Heart
Current α-Subunit Gene Small Subunit Gene
iNa Nav1.5 SCN5A β1 SCN1β
β2 SCN2β
β3 SCN3β
β4 SCN4β
ICaL Cav1.2 CACNA1C Ca β2a CACNB2

v
Ca α2Σ1a CACNA2D1
v
Cav1.3 CACNA1D Ca β2a

v
ICaT Cav3.1 CACNA1G None
Cav3.2 CACNA1H None
aThere are several isoforms of these and other small subunits whose interactions with the
different α-subunits have not been resolved.

Potassium Channels
There are many potassium channels whose names, along with those of the currents that they carry (Tables
13-4 and 13-5), can describe the order of their discovery (e.g., iK1), the duration of their open state (e.g.,
ito1), the timing of their opening (iKr and iKs), and substances that open (e.g., iK.Ca, iK. Ach) or close (e.g., iK.
ATP) the channel. Many of the potassium channel α-subunits are regulated by β-subunits and other small
membrane proteins (Table 13-5), all of which can be substrates for regulatory protein kinase-catalyzed
phosphorylations.
Potassium channels that open in depolarized cells are called outward rectifiers (see Chapter 14). An outward
current called ito1, which causes a transient repolarization immediately after cells are depolarized, is caused
by the opening of a channel that carries the slower ito, slow, and two channels that carry a more rapid ito, fast
(Table 13-4). The subsequent opening of several outward rectifier channels generates the delayed rectifier
currents iKr, iKs, and iKur that return membrane potential to its resting level (see Chapter 14). The α-subunits
of the channels that carry these currents, which are members of the Shaker, Shal, KVLQT1, and eag families
of potassium channel (Table 13-4), contain four domains, each of which is made up of six membrane-
spanning α-helices (Fig. 13-8A). The calcium-activated potassium current iK,Ca is carried by BK channels
whose domains contain seven membrane-spanning α-helices (Fig. 13-8D). The inward rectifier potassium
channels, which account for the high potassium permeability of resting cardiac myocytes, are members of
the Kir family, each of whose small α-subunits contain two α-helical transmembrane segments (Fig. 13-8B).
Two-pore channels, which are members of the K2P family (Fig. 13-8C, Table 13-4), generally do not respond
to changing membrane potential, but instead are activated by signals that include changing pH, lipids, and
cell deformation. Members of the TWIK (Tandem of P domains in Weak Inward rectifier K1) family of two-
pore potassium channels mediate electrical responses to stretch in mammalian hearts; these include TREK-1
(TWIK-RElated K1 channel) and TRAAK (TWIK-Related Arachidonic Acid-stimulated K1 channel). Along with a
related two-pore channel called TASK-1 (TWIK-related acid-sensitive K1), these two-pore channels also
participate in proliferative signaling.
P.359
Table 13-4 Potassium Channel α-Subunits in the Heart
α-Subunit Structure Gene Channel Current or Function
Six transmembrane α-helices (Kva)b
Shaker family
Kv1.4 KCNA4 Transient outward (slow) (ito1,slow)
Kv1.5 KCNA5 Delayed rectifier (iKur)

Kv1.7 KCNA7 Delayed rectifier (similar to iKur)
Shal family
Kv4.2 KCND2 Transient outward (fast) (ito1,fast)
Kv4.3 KCND3 Transient outward (fast) (ito1,fast)
KVLQT1 family
Kv7.1 KCNQ1 Delayed rectifier (iKs)
eag family
Kv10.2 (EAG2) KCNH2 Delayed rectifier (iKr)
Kv11.1 (HERG) KCNH2 Delayed rectifier (iKr)
Seven transmembrane α-helices (BK)
KCa1.1 KCNMA1 Calcium-activated, high conductance (iK,Ca)
Two transmembrane α-helices (Kir familyc)
Kir2.1 (IRK1) KCNJ2 Inward rectifier (iK1)
Kir2.2 (IRK2) KCNJ12 Inward rectifier (iK1)
Kir3.1 (GIRK1) KCNJ3 Acetylcholine-regulated (iK. Ach)
Kir3.4 (GIRK4) KCNJ5 Acetylcholine-regulated (iK. Ach)
KCNJ11
Kir6.2 (with SUR2d) ATP/ADP-regulated (iK. ATP)
Two-pore channels (K2P familye)
K 1.1d (TWIK 1) KCNK1 Respond to stretch, lipids, PK-A, PK-C

2P
K2P2.1 (TREK-1) KCNK6 Respond to stretch, lipids, PK-A, PK-C
K2P10.1 (TREK-2) KCNK10 Respond to stretch, lipids, PK-A, PK-C
K2P3.1 (TASK-1) KCNK3 Respond to changes in pH
K2P13.1 (THIK) KCNK13 Respond to changes in pH, oxygen tension
aKv: voltage-gated potassium channel;
bsmall amounts of several Kv channels not listed in this table are also found in normal human
hearts (Nerbonne and Kass, 2005, Gaborit et al., 2007);

cKir: inward rectifier potassium channel;
dSUR: sulfonylurea receptor;
eK : two-pore channel.
2P
Transient Receptor Potential (TRP) Channels

Calcium-selective transient receptor potential (TRP) channels play an important role in mediating stretch-
activated signals. These channels are not voltage-gated and do not play an important role in generating
action potentials, and most lack the high selectivity for calcium seen in the L- and T-type channels described
above. Details regarding the structure of these channels are still emerging, but it is clear that, like the α-
subunits of the voltage-gated ion channels
P.360
described above, TRP channels are tetramers of six-transmembrane α-helices. Their ability to interact with a
variety of regulatory proteins allows TRP channels to serve a number of regulatory functions. In addition to
carrying stretch-activated depolarizing calcium currents, they contribute to arrhythmogenic after-
depolarizations (see Chapter 14) and participate in membrane trafficking and calcium-mediated proliferative
signals (see Chapter 9).
Table 13-5 Potassium Channel β-Subunits in the Heart
Subunit Gene Associated α-Subunit Current
MinK KCNE1 Kv7.1 iKs
MiRP1 KCNE2 Kv11.1 iKr
Kvβ1 (Kvβ3) KCNAB1 Kv1.5 iKur
Kvβ2 KCNAB2 Kv1.5 iKur
SUR2 ABCC9 Kir6.2 iK. ATP
KChIP1 KCNIP1 Kv4.3 ito1
KChIP2 KCNIP2 Kv4.2, Kv4.3 ito1
BK β-subunit KCNMB K1.1 iK,Ca
Chloride Channels
Several types of chloride channels are found in cardiac myocytes (Table 13-6). The major subunits in CIC2
and CIC3 channels contain 18 α-helices (Fig. 13-8E) and so differ structurally
P.361
from the sodium, calcium, and potassium channels described above. These chloride channels function as
homodimers in which each subunit forms a pore. When open, CIC channels can carry either an inward or
outward current, depending on the membrane potential (see Chapter 14); because they are regulated by cell
deformation, they also serve as stretch sensors. Calcium-activated repolarizing currents, called iCl, Ca, have
been attributed to three different classes of chloride channel (Table 13-6); these include CLCA1 channels
that also carry a transient repolarizing current, called ito2, which occurs immediately after the initial
depolarization of atrial, ventricular, and His-Purkinje cells. Another type of chloride channel found in the
heart, the CFTR (cystic fibrosis transmembrane conductance regulator) channels, contain five subunits: two
transmembrane α-helices, two nucleotide-binding domains, and a regulatory subunit. Less is known of the
structures of the other chloride channels listed in Table 13-6.
Table 13-6 Chloride Channel Gene Families in the Heart
Family Current Function
CIC2 iCl, ir Hyperpolarization- and cell swelling-activated inward current in

resting cells
CIC3 iCl, Stretch- and cell swelling-activated outward current in resting cells
volume regulation
swel
iCl, b Small basal current
CLCA1 iCl, Ca Calcium-activated outward current in depolarized cells, contributes

to ito
Bestrophin iCl,Ca Calcium-activated outward current in depolarized cells
TMEM16 iCl,Ca Calcium-activated outward current in depolarized cells
CFTR iCl,PKA PKA, PKC, and purinergic receptor-activated outward current
Unknown iCl,acid Extracellular acidosis-activated current
Channel Regulation and Channel Mutations

The most important physiological regulator of the opening and closing of most plasma membrane ion
channels is changing membrane voltage, but a variety of other signaling mechanisms also modify channel
activity. The latter include cell deformation which, by evoking or modifying electrical signals in the heart,
can play a role in causing arrhythmias and sudden cardiac death; for example in a syndrome called commotio
cordis. This signaling mechanism, often called mechano-electrical feedback, can be mediated by two-pore
channels, chloride channels, and TRP channels.
Many cardiac ion channels are regulated by posttranslational changes, such as phosphorylations of both the
α- and β-subunits. Drugs also modify channel function, often when they bind to the S5 and S6 α-helices and
intervening peptide chain in the P-loop. Mutations in many of the channel proteins are of considerable
clinical importance because they provide a substrate for serious arrhythmias; for this reason, carriers of the
genes that encode these abnormal channel proteins are at risk for sudden cardiac death (see Chapter 16).
Single Channel Recordings

Measurements of single channel opening and closing have revolutionized our understanding of the control of
ion fluxes across biological membranes. For example, recordings from a single calcium channel show that
shortly after the application of a current that depolarizes the membrane from -20 to +50 mV, the channel
begins to flicker into its open state (Fig. 13-12). The slower changes in the membrane currents recorded
from whole cells represent the sum of all the openings and closings of many channels, and so tell us little
about the molecular transitions in single channels. Many functional changes in the magnitude and time
course of cardiac ion currents are now known to be determined by changes in the probability that the
channel will be in its open state. For example, phosphorylation of L-type calcium channels in response to β-
adrenergic stimulation promotes calcium entry by increasing the probability of finding the channel in the
open state after the membrane is depolarized (Fig. 13-13).
The initial view that voltage-sensitive ion channels can exist in only three functional states: open, closed
(resting), and closed (inactivated) (see above), has had to be modified on the basis of analyses of single
channel recordings, which have demonstrated that ion channels can assume a number of additional
functional states (Silva and Rudy, 2005). These analyses explain such puzzling phenomena as the ability of
some calcium channel blocking drugs both to activate and
P.362
inhibit calcium channel opening, and even more remarkably, to do so at the same time! These apparently
paradoxical responses are caused by drug-induced transitions among the many functional states of the
channel (Hess et al., 1984). In addition to the closed state (Mode 0), described earlier as closed (resting), L-
type calcium channels can exhibit brief openings (Mode 1) and long-lasting openings (Mode 2), both of which
are natural states of the channels (Fig. 13-14). The positive inotropic response to sympathetic stimulation is
due in part to phosphorylation of these channels, which increases calcium entry by favoring the appearance
of long-lasting openings (Mode 2). The importance of these additional channel states is highlighted by
evidence that mutations which favor the appearance of abnormal functional states in voltage-gated ion
channels can cause heritable long QT and Brugada syndromes (see Chapter 16).
Fig. 13-12: Response of a single calcium channel to membrane depolarization. Following a change in
membrane potential from -20 to +50 mV (upper tracing), the channel begins to alternate between its
and open states (middle tracing). When depolarization is prolonged, the channel becomes refractory
and opens less frequently; as a result, membrane current (lower tracing), after an initial increase
(downward deflection), begins to decrease.
Fig. 13-13: Response of a single calcium channel to phosphorylation by a cyclic AMP-dependent protein
kinase. Following a change in membrane potential from -20 to +50 mV, the phosphorylated channel spends
more time in its open state (middle tracing); this increases membrane current (lower tracing).
P.363
Fig. 13-14: Three sub-states of a calcium channel. Mode 0: closed; mode 1: brief openings; mode 2:
long-lasting openings.
Ion Channels of the Intercalated Disc: The Gap Junction and

Connexins
Action potentials are propagated by electrical circuits that resemble those in an undersea cable, where
longitudinal currents flow in both the sea water outside the cable and within its copper core (see Chapter
16). In the heart, where longitudinal conduction (parallel to the long axis of cardiac myocytes) is much faster
than transverse conduction (Fig. 13-15), impulse propagation depends on current flow between the interiors
of adjacent cells across an intercalated disc (Fig. 13-16). These longitudinal currents reflect the very low
internal resistance made possible by gap junction channels which differ from the voltage-gated ion channels
that control current flow through the plasma membrane.
Fig. 13-15: Flow of electrical currents in the myocardium. Cardiac myocytes are separated by the plasma
membrane, which has a high electrical resistance (dark lines) and by intercalated discs that contain gap
junctions (small vertical rectangles). The flow of current (white arrows) from a depolarized cell (lighter
myocyte labeled e) is transmitted rapidly in a longitudinal direction through the gap junctions in the
intercalated discs. Much less current flows transversely across the plasma membranes between myocytes.
P.364
Fig. 13-16: Schematic diagram of a strand made up of two cardiac myocytes that are separated by an
intercalated disc (heavy curved line). During depolarization, shown as proceeding from left to right,
the transmembrane potential in the depolarized region (shaded) is reversed compared to that in the
resting region. Longitudinal current, defined as the movement of electrons (dashed arrows), flows
from left to right in the extracellular space, and from right to left within the strand. Current flow
within the strand depends on the low internal electrical resistance made possible by pores in the
intercalated disc.
Non-selective channels found in the gap junction, or nexus, of the intercalated disc (see Chapter 1) contain
pores that account for the low electrical resistance between adjacent cells that is essential for longitudinal
conduction. These gap junction channels are freely permeable to charged molecules; for example,
radioactive potassium injected at one end of a bundle of myocardial cells diffuses across the intercalated
discs from cell to cell as rapidly as this ion would diffuse in an aqueous medium. Much larger organic ions,
such as the anionic dye fluorescene, also pass freely through these channels.
Gap junction channels contain a protein, called connexon (Fig. 13-17), that is found in simple organisms like
coelenterates and ctenophores, as well as in higher animals. Each channel is made up of two connexon
molecules that are aligned to form pores in the plasma membranes of adjacent cells. Connexon contains six
connexin subunits, each of which includes four transmembrane α-helices. Several connexin isoforms are
found in the mammalian myocardium (their names are based on their molecular weight in kD); most
abundant is connexin 43 (Cx43). Connexin 45 (Cx45), which forms lower conductance channels than Cx43, is
the most abundant connexin in the SA and AV nodes and His-Purkinje system. Connexin 40 (Cx40), a high
conductance isoform, is expressed in the rapidly conducting cells of the His-Purkinje system. Isoform shifts
involving the connexins alter conduction velocity in diseased hearts (see Chapter 18), and interactions
between connexins and cytoskeletal proteins participate in proliferative signaling.
Low pH and high cytosolic calcium close the connexon channels in cardiac myocytes. The former limits the
spread of acidosis when a region of the heart is forced to rely on anaerobic energy production, which
generates protons (see Chapter 2). Closure of the gap junctions by calcium prevents the cell death from
spreading after a part of the heart becomes irreversibly damaged, as occurs after coronary artery occlusion
when plasma membrane damage allows uncontrolled calcium entry into the cytosol. In order for the non-
ischemic regions of these damaged hearts to survive, the infarcted cells must be uncoupled from their viable
neighbors. Like the closing of bulkhead doors when a ship hits a reef and begins to take on water, closure of
connexin channels by calcium prevents a situation that is as dangerous to the heart as flooding in an ocean
liner.
P.365
Fig. 13-17: Structure of connexin, connexon, and gap junction channels. A: A single connexin
molecule showing the four transmembrane α-helices. B: Two connexon channels, each made up of six
connexin subunits, in the membranes of adjacent cells in the nexus of an intercalated disc. C: Three-
dimensional view of connexon in one membrane bilayer showing the pore surrounded by six connexin
subunits. D: Three-dimensional view of a gap junction pore that links the interiors of two adjoining
cells.
P.366
Phosphorylation of Cx43 by protein kinase C inhibits connexon channel opening by decreasing their open time
and interfering with the assembly of Cx43 subunits in the plasma membrane. This response plays an
important role in a phenomenon, called preconditioning, that has a protective effect in ischemic hearts (see
Chapter 17).
Conclusions
The relationships between channel structure and channel function described in this chapter are central to
understanding normal and abnormal cardiac electrophysiology. Some of these abnormalities are described in
Chapter 14, which describes how the orchestrated opening and closing of the ion channels generate the
currents that give rise to the cardiac action potential.
Bibliography
General
Abbott GW, Xu X, Roepke TKJ. Impact of ancillary subunits on ventricular repolarization. J

Electrocardiol 2007;40:S42–S46.
Armstrong CM. Sodium channels and gating currents. Physiol Rev 1981;61:644–683.
Baró I, Charpentier F. (Guest Editors). Special issue: Ion channels. J Mol Cell Cardiol 2010;48:1–111.
2001.
Bórjesson S, Ekinder F. Structure, function, and modification of the voltage sensor in voltage-gated ion
channels. Cell Biochem Biophy 2008;52:149–174.
Hille B. Ionic Channels of excitable membranes. 3rd ed. Sunderland MA: Sinauaer, 2001.
Kohl P, Ravens U. Cardiac mechano-electric feedback: past, present, and prospect. Prog Biophys Mol
Biol 2003;82:3–9.
Snyders DJ. Structure and function of cardiac potassium channels. Cardiovasc Res 1999;42:377–390.
Sodium and Calcium Channels
Brette F, Leroy J, Le Guennec JY, et al. Ca2+ currents in cardiac myocytes: old story, new insights. Prog
Biophys Mol Biol 2006;91:1–82.
Clusin WT. Mechanisms of calcium transient and action potential alternans in cardiac cells and tissues.
Am J Physiol Heart Circ Physiol 2008;294:H1–H10.
Tsien RW, Tsien RY. Calcium channels, stores, and oscillations. Ann Rev Cell Biol 1990;6:715–760.
Zaza A, Belardinelli L, Shrylock JC. Pathophysiology and pharmacology of the “late sodium current.”
Pharmacol Therap 2008;119:326–339.
Potassium Channels
Deal KK, England SK, Tamkun MM. Molecular physiology of cardiac potassium channels. Physiol Rev
1996;76:49–67.
Gurney A, Manoury B. Two-pore potassium channels in the cardiovascular system. Eur Biophys J
2009;38:305–318.
Gutman GA, Chandy KG, Grissmer S, et al. International Union of Pharmacology. LIII. Nomenclature and
molecular relationships of voltage-gated potassium channels. Pharmacol Rev 2005;57:473–508.
Hou S, Heinemann SH, Hoshi T. Modulation of BKCa channel gating by endogenous signaling molecules.
Physiology 2008;24:26–36.
P.367
Kane GC, Liu XK, Yamada S, et al. Cardiac KATP channels in health and disease. J Mol Cell Cardiol
2005;38:937–943. (Note: The June and July 2005 issues of J Mol Cell Cardiol contain additional
excellent reviews of KATP channels.)
Koa JH, Ibrahim MA, Park WS, et al. Cloning of large-conductance Ca2+-activated K+ channel à-subunits
in mouse cardiomyocytes. Biochem Biophys Res Comm 2009;389:74–79.
Nerbonne JM. Molecular basis of functional voltage-gated K+ channel diversity in the mammalian
myocardium. J Physiol (Lond) 2000;525:285–298.
Patel SP, Campbell DL. Transient outward potassium current, “ito,” phenotypes in the mammalian left
ventricle: underlying molecular, cellular and biophysical mechanisms. J Physiol (Lond) 2005;569:7–39.
Pourrier M, Schram G, Nattel S. Properties, expression and potential roles of cardiac K+ channel
accessory subunits: MinK, MiRPs, KChIP, and KChAP. J Membrane Biol 2003;194:141–152.
Seino S. ATP-sensitive potassium channels: a model of heteromultimeric potassium channel/receptor

assemblies. Annu Rev Physiol 1999;61:337–362.
Snyders DJ. Structure and function of cardiac potassium channels. Cardiovasc Res 1999;42:377–390.
Tamargo J, Caballero R, Gómez R, et al. Pharmacology of cardiac potassium channels. Cardiovasc Res
2004; 62:9–33.
Yu H, McKinnon D, Dixon JE. et al. Transient outward current, ito1, is altered in cardiac memory.
Circulation 1999;99:1898–1905.
Hyperpolarization-Activated Channels
Biel M. Cyclic nucleotide regulated cation channels. J Biol Chem 2009;284:9017–9021.
Biel M, Wahl-Schott C, Michalakis S, et al. Hyperpolarization-activated cation channels: from genes to

function. Physiol Rev 2009;89:847–885.
Craven KB, Zagotta WN. CNG and HCN channels: two peas, one pod. Ann Rev Phsyiol 2006;68:375–401.
Jackson HA, Marshall CR, Accili EA. Evolution and structural diversity of hyperpolarization-activated
cyclic nucleotide-gate channel genes. Physiol Genomics 2007;29:231–245.
Transient Receptor Potential Channels
Inoue R, Jian Z, Kawarabayashi Y. Mechanosensitive TRP channels in cardiovascular pathophysiology.

Pharmacol Therap 2009;118:371–385.
Moiseenkova-Bell VY, Wensel TG. Hot on the trail of TRP channel structure. J Gen Physiol 2009;133:
239–244.
Stones R, Gilbert SH, Benoist D, et al. Inhomogeneity in the response to mechanical stimulation:
cardiac muscle function and gene expression. Prog Biophys Mol Biol 2008;97:268–281.
Stiber JA, Seth M, Rosenberg PB. Mechanosensitive channels in striated muscle and the cardiovascular
system: not quite a stretch anymore. J Cardiovasc Pharmacol 2009;54:116–122.
Watanabe H, Murakami M, Ohba T, et al. TRP channel and cardiovascular disease. Pharmacol Therap
2009;118:337–351.
Chloride Channels
Baumgarten CM, Clemo HF. Swelling-activated chloride channels in cardiac physiology and
pathophysiology. Prog Biophys Mol Biol 2003;82:25–42.
Duan D. Phenomics of cardiac chloride channels: the systematic study of chloride channel function in
the heart. J Physiol (London) 2009;587:2163–2177.
Dutzler R, Campbell EB, Cadene M, et al. X-Ray structure of a CIC chloride channel at 3.0 ó reveals the
molecular basis of ion selectivity. Nature 2002;415:287–294.
P.368
O'Connell AD, Morton MJ, Hunter M. Two-pore domain K+ channels–molecular sensors. Biochim Biophys
Acta 2002;1566:152–161.
Connexon and Connexin
Jalife J, Morley GE, Vaidya D. Connexins and impulse propagation in the mouse heart. J Cardiovasc
Electrophysiol 1999;10:1649–1663.
Solan JL, Lampe PD. Connexin43 phosphorylation: structural changes and biological effects. Biochem J
2009;419:261–272.
Yeager M. Structure of cardiac gap junction intercellular channels. J Struct Biol 1998;121:231–245.
References
Bernstein J. Untersuchungen zur Thermodynamik der bioelektrischen Ströme. Pflügers Arch ges Physiol
1902;92:521–562.
Cole KS, Curtis HJ. Membrane potential of the squid giant axon during current flow. J Gen Physiol
1941;24:551–563.
Gaborit N, Le Bouter S, Szuts V, et al. Regional and tissue-specific transcript signatures of ion channel
genes in the non-diseased human heart. J Physiol (Lond) 2007;582:675–693.
Hess P, Lansman JB, Tsien RW. Different modes of Ca channel gating behaviour favoured by
dihydropyridine calcium agonists and antagonists. Nature 1984;311:538–544.
Hodgkin AL, Huxley AF. Resting and action potentials in single nerve fibres. J Physiol 1945;104:176–195.
Nerbonne JM, Kass RS. Molecular physiology of cardiac repolarization. Physiol Rev 2005;85:1205–1253.
Silva J, Rudy Y. Subunit interaction determines IKs participation in cardiac repolarization and
repolarization reserve. Circulation 2005;112:1384–1391.
Zhou Y, Morais-Cabral JH, Kaufman A, et al. Chemistry of ion coordination and hydration revealed by a
K+ channel-Fab complex at 2.0 A resolution. Nature 2001;414:1384–1391.
> Table of Contents > Part Three - Normal Physiology > Chapter 14 - The Cardiac Action Potential
Chapter 14
The Cardiac Action Potential
The cardiac action potential, which is generated by the orchestrated opening and closing of the ion channels described in
Chapter 13, is much more complex than the action potentials in nerves and skeletal muscle, where depolarization lasts only a
few milliseconds (Fig. 14-1A). In the heart, action potentials last several hundred milliseconds, consist of several phases, and
vary in characteristics from region to region. Purkinje fiber action potentials, for example, are large, rapidly rising, last over
300 ms, and include five distinct phases (Fig. 14-1B). These include a large rapid upstroke (phase 0) that is followed by two
phases, transient repolarization (phase 1) and a plateau (phase 2), that do not have clear counterparts in nerve and skeletal
muscle. Purkinje fiber action potentials end when repolarization (phase 3) returns membrane potential to its resting level
during diastole (phase 4), after which these myocytes often exhibit spontaneous depolarizations (pacemaker activity). Action
potentials in the working cells of the ventricles are similar to those of Purkinje cells, except that they are somewhat smaller
and lack pacemaker activity; atrial action potentials resemble those of the ventricles except that they are briefer, while
action potentials in the SA and AV nodes are much smaller and lack a plateau.
During each action potential, a small amount of sodium enters working cardiac myocytes and the cells of the His-Purkinje
system, while potassium is lost. It is a frequent misconception that the concentration gradients for sodium and potassium
across the plasma membrane are dissipated during each action potential. However, a single depolarization causes only a very
small change in the chemical composition of the cardiac cell, as evidenced by the large number of action potentials that can
be generated after the Na-K ATPase is poisoned. Although prolonged rapid stimulation does cause cells to gain a measurable
amount of sodium, the increase is small; for example, stimulation of a sheep Purkinje fiber for 3 min at 1 Hz has been
estimated to increase intracellular sodium activity by only 1.9 mM (Cohen et al., 1982).
Resting Potential
Resting potential in myocardial cells is related to the electrochemical gradient for potassium across the plasma membrane,
which is permeable to potassium but impermeable to the anionic proteins and phosphate-containing compounds found in the
cytosol; this establishes a Donnan equilibrium where [K+]i is higher than [K+]o and the cytosol is negatively charged (see
Chapter 13). In this Donnan equilibrium, the tendency for potassium to move down its concentration gradient out of the cell is
balanced by the negative intracellular resting potential that favors potassium influx; in other words, the tendency of the
potassium concentration gradient to cause potassium to leave the cell is countered by an electrical gradient that causes
potassium to enter the cell.
P.370
Fig. 14-1: Skeletal and cardiac muscle action potentials. A: The skeletal muscle action potential is a brief biphasic
event in which rapid depolarization (upward deflection) is quickly followed by repolarization (downward deflection). A
small positive after-potential causes an approximately 10 ms delay prior to return of the membrane potential to its
resting level (dotted arrow). B: The cardiac muscle action potential lasts over 300 ms and consists of five phases. Phase
0 (the upstroke) corresponds to depolarization and phase 3 (repolarization) to repolarization in skeletal muscle. Phases
1 (early repolarization) and 2 (plateau) have no clear counterpart in skeletal muscle, while phase 4 (diastole)
corresponds to the resting potential.
The Nernst and Goldman–Hodgkin–Katz Equations

The plasma membrane in resting cardiac myocytes is not perfectly selective for potassium. There is, for example, a small
permeability to sodium which, because of the higher sodium concentration outside the cell (Table 14-1) favors a membrane
potential opposite in polarity to that associated with the potassium gradient. The membrane potential that results from the
P.371
distributions of these and other ions is described by the Nernst and Goldman–Hodgkin–Katz equations, which define the
equilibrium potential (where there would be no net flux of ions) established by all of the ion concentration gradients across
the plasma membrane. The Nernst equation describes the equilibrium potential established by a single ion, while the
Goldman–Hodgkin–Katz equation describes the potential across a membrane that is permeable to several ions.
Table 14-1 Ion Activities Inside and Outside Mammalian Myocytes
Ion Intracellular Concentration Intracellular Activity Extracellular Concentration Extracellular Activity
Sodium 5–34a 8 140 110
Potassium 104–180a 100 5.4 4
Chloride 4.2b 45 117 88
Calcium 0.0002c 3 1
Values are in mM. Activities are “averages” weighted arbitrarily by the author for use in various equations.
aBased on data from Walker, 1986.
bBased on data from Hille, 2001.
cBased on data from Blinks, 1986.
The Nernst Equation

The membrane potential established by a concentration difference for a single ion across a semi-permeable membrane is
described by the Nernst equation:
where Em is the membrane potential, R the gas constant, T the absolute temperature, z the valence of the ion, F the Faraday
constant, P the permeability to the ion, and ao and ai the activities of the ion outside and inside the membrane, respectively.
The Nernst equations generated at 37°C when alkali metal ions such as sodium and potassium, for which z = 1, are distributed
across a freely permeable membrane (P = 1) can be written for ordinary (base 10) logarithms as:
Equation 14-2 states that a tenfold difference in the activity of a monovalent cation, where ao/ai = 10, generates a potential
difference of +61.5 mV. If the concentration gradient is reversed, and ao/ai = 0.1, Em is -61.5 mV.
In a resting cardiac myocyte, where [K+]o is ∼5.4 mM and [K+]i is ∼120 mM, the corresponding potassium ion activities are
approximately 4 and 100 mM (Table 14-1). If the plasma membrane in resting cardiac myocytes was freely permeable to
potassium, and impermeable to all other ions for which a concentration gradient exists between the inside and outside of the
cell, Em would equal EK:
In most regions of the heart, resting potential is close to that predicted by the Nernst equation for potassium.
Changes in extracellular potassium concentration ([K+]o) have two effects on resting potential. The first is due simply to
changes in the Nernst potential for potassium; according to Equation 14-3, increased [K+]o causes depolarization and reduced
[K+]o hyperpolarizes the plasma membrane. The second occurs because very low [K+]o (<3 mM) reduces potassium permeability
(PK). As a result of this hypokalemia-induced fall in PK at low [K+]o, resting potential does not follow the Nernst equation for
potassium (Fig. 14-2) because the increased influence of the other ions, notably sodium, shifts membrane potential away from
that predicted by the Nernst equation.
P.372
Fig. 14-2: Relationship between extracellular potassium concentration [K+]o and resting membrane potential. Changing
extracellular potassium at normal and high [K+]o causes membrane potential (solid line) to respond in a manner that
closely approximates the predictions of the Nernst equation for potassium (dashed line). However, at low levels of
extracellular potassium membrane potential decreases unexpectedly because potassium permeability falls.
The Goldman–Hodgkin–Katz Equation

More accurate estimates of membrane potential are calculated by the Goldman-Hodgkin-Katz equation, which takes into
account the permeabilities and activities of all ion species for which there is a gradient across the membrane. For a membrane
permeable to sodium and chloride, as well as to potassium, this equation is
Equation 14-4 highlights the fact that the contribution of any ion to membrane potential is determined both by the activity
gradient across the membrane and the permeability of the membrane to that ion. The contribution of any ion to Em disappears
if there is no activity gradient, or if permeability to the ion becomes zero.
Some Features of the Cardiac Action Potential
Threshold
In order for an electrical stimulus to initiate an action potential, depolarization must reach a threshold. Smaller
(subthreshold) depolarizations cause only local responses because they do not open enough depolarizing channels to initiate a
regenerative action potential (Fig. 14-3A and B). The threshold is the lowest membrane potential at which opening of enough
sodium channels—or calcium channels in nodal cells—is able to initiate the sequence of channel openings needed to generate a
propagated action potential. The latter is often called a regenerative response because, once threshold is reached, the extent
of the subsequent depolarization becomes independent of the initial depolarizing stimulus (Fig. 14-3C).
Although small depolarizations that fail to reach threshold do not initiate action potentials, they can have important effects
on the responses to subsequent depolarizing currents. Most important is the ability of subthreshold depolarizations to cause
transient inactivation of sodium
P.373
channels by closing their h gates, which decreases both the rate and extent of depolarization in subsequent action potentials.
Fig. 14-3: Threshold for initiation of a propagated action potential. Small depolarizing stimuli (dotted lines A and B)
that fail to reach threshold (dashed line) are unable to initiate an action potential. A longer depolarization that reaches
threshold (dotted line C) opens enough sodium channels to initiate a regenerative action potential in which the
subsequent depolarization is independent of the initial stimulus.
Rate of Rise and Amplitude of the Action Potential

The rate of depolarization (dV/dt) during an action potential upstroke (phase 0) is determined by the rate at which cations
ions enter the cytosol. In cells whose action potentials depend on sodium channel opening, initial depolarizing currents
develop rapidly, are very large, and last only for a short time; for this reason, sodium currents are often referred to as fast
inward currents. The action potentials in the SA and AV nodes have a lower amplitude and slower rate of rise because
depolarization depends on the slower opening of smaller (lower conductance) calcium channels.
Membrane potential immediately prior to stimulation is an important determinant of sodium channel opening, and so
determines the amplitude and rate of rise of the action potential. In a cardiac myocyte with a normal resting potential of
between -80 and -90 mV, depolarizing stimuli that exceed threshold generate large action potentials with a rapid upstroke,
whereas if the cell is partially depolarized prior to stimulation, the same depolarizing stimulus produces a smaller, more slowly
rising action potential (Fig. 14-4) because the prior subthreshold depolarization inactivates some of the sodium channels.
Sodium channel activation also depends on the rate at which threshold is approached. When threshold is reached quickly,
sodium conductance increases rapidly which causes action potential upstroke to be rapid and the amplitude to be high (Fig.
14-5A), whereas if a longer time elapses before threshold is reached, some of the sodium channels become inactivated, which
generates a lower amplitude action potential that rises more slowly (Fig. 14-5B). These manifestations of voltage-dependent
inactivation of sodium channel opening are important clinically because reduced action potential amplitude and rate of rise
provides a substrate for arrhythmias by slowing impulse conduction (see Chapter 16).
P.374
Fig. 14-4: The rate of depolarization and amplitude of an action potential depend on the resting membrane potential
prior to stimulation. A large, rapidly rising action potential is generated when the resting potential is high (solid line A),
whereas a lower resting potential prior to stimulation causes the same stimulus to produce a smaller, slowly rising
action potential (dashed line B). These changes in the action potential upstroke reflect the voltage-dependent closing
of the inactivation (h) gates of the sodium channels.
The Plateau
When cardiac action potentials were first observed to have a plateau, it was not known whether membrane potential had
stabilized because current flow had ceased, or because currents continued to flow into and out of these cells, but had become
equal. The answer was provided by Weidmann (1951), who found that unlike skeletal muscle and nerve, where membrane
resistance increases sharply a few milliseconds after depolarization, resistance in cardiac muscle remains low early during the
plateau; this means that currents are flowing in both directions across the plasma membrane. Subsequent research has shown
that, during the plateau, outward potassium currents are balanced by an inward calcium current.
Refractory Periods
Like resting depolarization (Fig. 14-4) and a slow rate of activation (Fig. 14-5), stimulation of a ventricular myocyte before it
has recovered from a preceding depolarization results in small, slowly rising action potential (Fig. 14-6). The mechanisms are
similar; all inactivate the sodium channels normally responsible for depolarization. Both resting depolarization and slow
activation reduce the rate of rise and amplitude of the action potential as a result of depolarization-induced channel
inactivation, while the slowly rising, small action potentials seen when
P.375
activation is premature reflect the slow time course of channel reactivation. In cardiac sodium channels, complete recovery of
the h gates can require more than 100 ms after membrane potential has returned to its resting level.
Fig. 14-5: The rate of depolarization and amplitude of an action potential depend in part on the rate at which
membrane potential approaches threshold. When the stimulus reaches threshold rapidly (dotted line A), a large, rapidly
rising action potential is produced (solid line), whereas a stimulus that reaches threshold slowly (dotted line B)
produces a smaller, more slowly rising action potential (dashed line).
The period of depressed excitability that begins when depolarization closes the h gates of sodium channels, called the
refractory period, is traditionally divided into two phases (Fig. 14-7). The first is the absolute refractory period, when no
stimulus, whatever its magnitude, can evoke
P.376
P.377
a propagated response. This is followed by the relative refractory period, when only stimuli that exceed the normal threshold
can initiate a propagated response. The interval between the onset of depolarization and the return of normal excitability,
sometimes called the full-recovery time, encompasses the effective and relative refractory periods, as well as the
supernormal period described below. The long refractory period of myocardial cells is due both to the prolonged plateau of the
action potential, which delays the return of membrane potential to the resting level at which the sodium channels reactivate,
and the slow rate of channel recovery after the membrane has repolarized.
Fig. 14-6: Effect of premature depolarization on the ventricular action potential. When the ventricles are activated
during the relative refractory period, before they have recovered fully from a preceding depolarization, the resulting
action potential (dashed line B) is more slowly rising and smaller than the normal action potentials (solid lines A).
Fig. 14-7: Refractory periods. A: Action potential. During the absolute refractory period, which is caused by closing of
the h gates after membrane depolarization, no stimulus regardless of its strength is able to initiate a propagated action
potential. This is followed by a relative refractory period (RRP) during which only stimuli that exceed the normal
threshold can cause a propagated action potential. The relative refractory period is followed by a supernormal period
(SN), during which stimuli slightly smaller than those that reach the normal threshold can generate a propagated action
potential. Action potentials generated during both the relative refractory and supernormal periods are small and slowly
rising because of incomplete recovery of the sodium channels. Refractoriness ends after the supernormal period, when
threshold returns to normal. B: Threshold. Immediately after depolarization threshold becomes infinite, which marks
the beginning of the absolute refractory period. The relative refractory period begins when threshold starts to fall at
the end of the absolute refractory period, and ends after the supernormal period, when threshold falls briefly below its
resting level.
Action potentials evoked during the heart's relative refractory period depolarize slowly and are of low amplitude (Fig. 14-7),
so that they conduct slowly (see Chapter 16). Stimuli that reach the heart shortly after refractoriness has ended, even after
threshold has returned to normal, may still initiate small, slowly conducting action potentials because of the slow reactivation
of cardiac sodium channels.
Supernormality and Vulnerability

A supernormal period, sometimes seen at the end of the relative refractory period (Fig. 14-7), is characterized by the ability
of stimuli smaller that those needed to reach the normal threshold to produce propagated action potentials. Although
threshold is reduced during supernormality, the action potentials generated during this period are of low amplitude. This
apparent discrepancy reflects the fact that although threshold is low, not all sodium channels have recovered. Supernormality
is well documented in isolated His-Purkinje cells, but is probably absent in the AV node, atria, and ventricles (Spear and
Moore, 1980). Evidence for this phenomenon can occasionally be identified on the clinical electrocardiogram (see Chapter 16).
The supernormal phase occurs at approximately the same time as the vulnerable period, but vulnerability and supernormality
are different phenomena. Supernormality is a lowered threshold, whereas vulnerability describes an increased susceptibility
to ventricular fibrillation caused by heterogeneities in repolarization (see Chapter 16).
Rectification
Ion channels can exhibit a property called rectification, which is the ability of changing membrane potential to modify channel
conductance—the ability to carry current (Fig. 14-8). In an unrectified (Ohmic) current, resistance (R) is independent of
membrane potential (E), so that the current (I) is simply an inverse linear function of voltage
Rectification occurs when changing membrane voltage either increases or decreases channel conductance.
Most of the potassium currents carried by voltage-gated potassium channels across biological membranes are rectified (Fig. 14-
8). Outward rectification occurs when depolarization increases potassium conductance. As a result, outward rectification
increases the flow of repolarizing potassium current when the membrane depolarizes which, by helping depolarized cells to
repolarize, favors the return of membrane potential to the resting level. Outward rectification can therefore be viewed as a
“true” rectification because it “puts things right” by helping to end the
P.378
action potential. Inward rectification, which occurs when depolarization decreases potassium conductance, has the opposite
effect because it responds to depolarization by decreasing the flow of outward current. Inward rectifier potassium currents
therefore tend to maintain the membrane in a depolarized state. Because inward rectification was once considered to be an
anomaly, it was also called “anomalous” rectification. These differences in rectification reflect the different molecular
structures of the α-subunits of the potassium channels that carry outward and inward rectifying currents (see Chapter 13).
Fig. 14-8: Current–voltage relationships showing rectification of potassium currents. The effects of changing membrane
potential (abscissa) on ionic currents (ordinate) are shown for three different types of potassium channel; inward
currents are downward, outward currents are upward. All of the potassium currents are zero at -86 mV, which is the
equilibrium potential for potassium. An unrectified (Ohmic) current (solid line A) is a linear function of membrane
voltage because channel conductance is independent of membrane voltage. Outward (“true”) rectification is seen when
depolarization increases potassium conductance (dotted line B), which allows depolarization to increase the
repolarizing potassium currents that return membrane potential to the resting level. Inward (“anomalous”) rectification
occurs when depolarization decreases potassium conductance (dashed line C), which tends to maintain membrane
potential in the depolarized state by decreasing repolarizing potassium currents.
Sodium Currents
A high sodium permeability favors a membrane potential opposite in polarity to that caused by the potassium gradient because
extracellular sodium activity is higher than that in the cell interior (Table 14-1). This is apparent in a simplified Goldman–
Hodgkin–Katz equation that describes the dependence of membrane potential on both potassium and sodium:
P.379
Extracellular sodium concentration is ∼140 mM, which corresponds to an activity of ∼110 mM, while intracellular sodium
activity is ∼80 mM (Table 14-1), so that the Nernst equation for sodium predicts a membrane potential of +70 mV
Cardiac action potentials do not reach the high positivity predicted by Equation 14-7 because potassium permeability does not
fall to zero (PK > 0), and sodium permeability does not reach its maximum (PNa < 1) at the peak of the action potential.
The major function of the sodium channels in the working myocytes of the atria and ventricles and rapidly conducting cells of
the His-Purkinje system is to carry the fast inward current that depolarizes these cells. These sodium channels can also carry a
much smaller long-lasting inward current, often called the late sodium current or iNaL, which can persist throughout the
action potential plateau. The latter appears not to be due to an additional sodium channel protein, but instead represents a
sub-state of this channel.
Sodium channels are regulated by protein kinases and phosphatases, cytoskeletal deformation, and trafficking proteins. The
protein kinases include PKA, which increases the inward sodium current, and PKC, which reduces sodium conductance. A large
number of sodium channel mutations have been identified, many of which predispose to lethal arrhythmias. These include
mutations that delay repolarization by increasing the late sodium current, which cause a long QT syndrome called LQT3, and
mutations that reduce sodium channel opening, which cause an electrocardiographic abnormality called the Brugada syndrome
(see Chapter 16).
Calcium Currents
The calcium gradient across the plasma membrane, like that for sodium, favors an inward (depolarizing) current because
extracellular ionized calcium activity is ∼1 mM, whereas cytosolic calcium during diastole is ∼0.2 µM (0.0002 mM) (Table 14-1).
This 5,000-fold activity gradient for calcium, according to the Nernst equation, establishes a positive calcium equilibrium
potential
This equilibrium potential decreases during systole, when calcium released from the sarcoplasmic reticulum by excitation-
contraction coupling increases intracellular ionized calcium concentration (see Chapter 7).
The threshold for the opening of L-type calcium channels is higher than that for sodium channels, so that depolarization opens
these calcium channels after the sodium channels have opened. This is why the inward calcium current in working cardiac
myocytes and Purkinje fibers, which is carried by these channels, begins after the sodium current has generated phase 0.
L-Type and T-Type Calcium Channels

Several different types of plasma membrane calcium channels are found in mammalian tissues. Most important in the heart
are the L- and T-type calcium channels, whose names
P.380
reflect the slower inactivation of L-type calcium channels than T-type channels, so that L = long-lasting, T = transient (see
Chapter 13). L-type calcium channels also have a lower threshold for activation than T-type channels, so that “L” can also
stand for “low threshold,” and because T-type channels carry a smaller depolarizing current, “T” can also stand for “tiny.”
L-Type Calcium Channels

The depolarizing current carried by L-type calcium channels, which, because it follows the much larger inward sodium
current, was initially referred to as a secondary or slow inward current. In the working cells of the atria and ventricles and in
Purkinje fibers, this calcium current contributes to the action potential plateau (phase 2). These channels also carry
depolarizing currents in the SA and AV nodes, and so participate in controlling heart rate and AV conduction.
The α1-subunits of L-type calcium channels are activated by posttranslational modifications; most important is phosphorylation
by PKA, which plays an important role in the inotropic response to sympathetic stimulation (see Chapters 8 and 10).
L-type calcium channels in vascular smooth muscle, which control vascular resistance, are related to those found in the heart,
which accounts for the vasodilator effects of most calcium channel-blocking drugs. Both types of L-type calcium channels can
bind to the three major types of calcium channel blocking drugs (dihydropyridines, benzothiazepines, and phenylalkylamines).
Although all can cause smooth muscle to dilate, the calcium channels in cardiac muscle are less sensitive to the
dihydropyridines because these drugs bind with low affinity to the cardiac channels.
L-type calcium channels play a central role in both cardiac and skeletal muscle excitation-contraction coupling, but they do so
in different ways. The cardiac channels, which are concentrated in the t-tubules, form dyads with the terminal cisternae of
the sarcoplasmic reticulum (see Chapter 1) and carry the inward calcium flux that opens the sarcoplasmic reticulum calcium
release channels (“calcium-triggered calcium release,” see Chapter 7). In skeletal muscle, a different L-type calcium channel
isoform opens the intracellular sarcoplasmic reticulum calcium release channels by removing a “plug” (see Chapter 7).
T-Type Calcium Channels

The small inward currents generated by the transient openings of T-type calcium channels do not play an important role in
depolarization of the atria, ventricles, and His-Purkinje fibers because these currents occur at the same time as the much
larger sodium currents. In the SA node, however, the small depolarizing currents carried by T-type calcium channels contribute
to the oscillatory potentials that generate pacemaker activity (see below).
T-type calcium channels play little or a no role in excitation-contraction coupling because they admit only a small amount of
calcium at a slow rate and are not found in close proximity to the calcium release channels of the sarcoplasmic reticulum. The
small calcium signals generated by the T-type calcium channels probably participate in proliferative signaling and may
contribute to resting tension.
P.381
Potassium Currents
Potassium currents serve two important roles in the heart; they generate repolarizing currents that return membrane potential
to the resting level after myocyte depolarization, and they help maintain resting potential (Table 14-2). The inward rectifying
potassium currents responsible for the high potassium permeability that maintains resting potential (phase 4) are carried by
the channels whose smaller α-subunits contain only two transmembrane α-helices (see Chapter 13).
Table 14-2 Some Potassium Currents in the Heart
Small
Current α-Subunit(s) Properties
Subunit(s)
Outward Rectifying Currents
Transient outward Opens immediately after depolarization; regulates action

current (ito1) potential duration
ito1, slow Kv1.4
ito1, fast Kv4.2, KChIP1,

Kv4.3 KChIP2
Rapid delayed Kv10.2, MIRP1 Opens early during plateau; initiates repolarization
rectifier (iKr) Kv11.1
Slow delayed Kv7.1 MinK Opens late during plateau; initiates repolarization
rectifier (iKs)
Ultra-rapid delayed Kv1.5, Kvβ1, Opens very early during plateau; initiates repolarization
rectifier (iKur) Kv1.7 Kvβ2
Calcium-activated KCa1.1 BK β- Activated by calcium; accelerates repolarization

delayed rectifier subunit
(iK.Ca)
Inward Rectifying Currents
Inward (anomalous) Kir2.1, Opened by repolarization, closed by depolarization;

rectifier (iK1) Kir2.2 maintains resting potential; prolongs action potential
plateau
ATP-sensitive Kir6.2 SUR2 Inhibited by ATP, activated by ADP; opens in energy-

current (iK.ATP) starved cells
Acetylcholine- Kir3.1, Activated by Gαi;hyperpolarizes resting cells, slows

activated current Kir3.4
(iK.Ach)
SA node pacemaker, shortens atrial action potential
P.382
Outward Rectifying Potassium Currents

The currents that repolarize cardiac myocytes are carried by five types of potassium channels (see Chapter 13). Four are Kv
channels that are made up of four non-covalently linked α-subunits, each of which contains six transmembrane α-helices;
these include members of four families: Shaker (from the behavior of a drosophila mutant whose appendages shake when the
flies that express this gene are anaesthetized), Shal, KVLQT1, and eag (from the homology to a drosophila mutant that, when
lightly etherized, resembles a go-go dancer). The fifth is a BK channel whose α-subunits contain seven transmembrane α-
helices.
Transient Outward Current (ito1)

Phase 1 (early repolarization) is caused by a repolarizing current, called ito, which shortens action potential duration by
accelerating the opening and closing of the channels that are activated subsequently during the action potential. This current
has at least four components: three are carried by potassium channels that contribute to ito1; the fourth by the chloride
channel that carries ito2 (described below). The potassium currents are carried by Kv1.4, which is responsible for a slower
ito,slow current, and Kv4.2 and Kv4.3, which carry a more rapid ito,fast. Both of the latter are linked to β-subunits called KChIP1
and KChIP2 (see Chapter 13).
Stress-induced changes in the expression of the genes that encode the ito1 channels prolong the action potential by decreasing
the density of the Kv4.3/KChIP1 complex (Özgen and Rosen, 2009). These changes in ito1 channels cause long-lasting
repolarization abnormalities, sometimes called electrical remodeling or cardiac memory, that contribute to post-tachycardia
and pacing-induced T-wave abnormalities, and the T-wave “evolution” that follows an acute myocardial infarction. Activation
of ito1 channels by PKA-catalyzed phosphorylation contributes to the action potential shortening caused by sympathetic
stimulation.
Delayed (Outward) Rectifier Potassium Currents (iKr, iKs, and iKur)

The major currents that repolarize the heart in phase 3 (repolarization) are carried by outward rectifier potassium channels
that, because they open after the initial depolarizing event, are called delayed rectifiers. The major delayed rectifier
currents in Purkinje fibers and the ventricles are iKr and iKs; the subscripts r and s refer to the rate at which these channels
open: r = rapid and s = slow. Activation of iKr begins shortly after depolarization and reaches its peak toward the end of the
plateau while the iKs channels open more slowly. The shorter action potentials seen in atrial myocytes are due to a rapidly
developing delayed rectifier current called iKur (for ultra-rapid) (Table 14-2). Mutations in the α- and β-subunits of these
channel proteins are responsible for long QT syndromes (see Chapter 16).
Calcium-Activated Repolarizing Potassium Current (iK.C a)

High cytosolic calcium concentration activates a high conductivity repolarizing current called iK.Ca. The channels that carry
this current, which are also found in the mitochondrial inner membrane and in vascular smooth muscle, have several names
including BK (see Chapter 13), MaxiK (because of their high conductance), and KCa1.1 (because they are activated by calcium).
The α-subunits of the BK channels (called slo after a drosophila mutant called “slowpoke”) have seven transmembrane α-
helices and so differ from those of the voltage-gated channels described above.
P.383
The ability of calcium to activate the repolarizing currents carried by BK channels, along with those carried by the chloride
channel ito2 described below, shortens the action potential plateau and so reduces the open time of the L-type calcium
channels. This response provides a negative feedback that reduces calcium entry in calcium-overloaded hearts. Activation of
the BK channels by calcium explains why the QT interval (an index of action potential duration, see Chapter 15) is shortened
by cardiac glycosides (which increase cellular calcium) and hypercalcemia, and why hypocalcemia prolongs the QT interval.
BK channels form signaling complexes with a variety of macromolecules, including enzymes and other ion channels. In addition
to regulation by membrane voltage, they can be modified by protons, lipids, steroid hormones, reactive oxygen and nitrogen
species, and several protein kinases.
Inward Rectifier Potassium Currents

Inward rectifier potassium channels, called Kir (K inward rectifier), are tetramers made up of α-subunits that contain two
transmembrane α-helical segments and a P-loop (see Chapter 13). Several classes of inward rectifier channels are found in
human hearts. The inward rectifier current responsible for the high potassium permeability of resting hearts is called iK1;
other Kir currents participate in SA node pacemaker activity, reduce excitability by increasing resting potential in response to
decreased levels of ATP and increased ADP levels (iK.ATP), and mediate responses to parasympathetic stimuli (iK.Ach).
Voltage-Regulated Inward Rectifier Potassium Currents (iK1)

The iK1 channels, which open during diastole, bring membrane potential toward the Nernst potential for potassium, and so are
major determinants of resting potential. Because these channels close in response to depolarization (inward rectification), the
resulting decrease in repolarizing current helps prolong the plateau of the cardiac action potential. Cardiac iK1 channels are
inhibited by increased cytosolic calcium and by acidosis.
Atp-Inhibited Inward Rectifier Potassium Currents (IK.Atp)

The iK.ATP channels are heterotetrameric inward rectifier channels made up of two pore-forming Kir channels and two
molecules of a sulfonylurea receptor called SUR1. (The ability of sulfonylureas to reduce blood sugar in diabetic patients is due
to the role of iK.ATP channels in the pancreas to regulate insulin release.) These ligand-gated inward rectifier channels are
activated by ADP and inhibited by ATP, so that like a contented house cat asleep before the hearth, iK.ATP channels spend most
of their life in a dormant state. Opening of these channels decreases excitability which, by reducing contractility, conserves
energy when the heart becomes energy-depleted.
Acetylcholine-Activated Inward Rectifier Potassium Currents (iK.Ach)

Ligand-gated channels that carry an inward rectifying potassium current called iK.Ach mediate hyperpolarizing responses to
vagal stimulation. The iK.Ach channels are heterotetramers made up of Kir channel α-subunits called GIRK (G protein inward
rectifier K channel). The hyperpolarizing response caused by iK.Ach channel opening in the SA node slows the sinus pacemaker
(see below). In the atria, opening of iK.Ach channels shortens the action potential, which contributes to the negative inotropic
effects of vagal stimulation. These inhibitory effects are seen when acetylcholine binding to cardiac muscarinic receptors
activates the inhibitory heterotrimeric G protein Gi, which releases a Gbg that binds directly to these channels within the
P.384
plasma membrane (see Chapter 8). Binding of adenosine and other purines to P1 purinergic receptors also activates these
inward rectifier channels.
Background Potassium Currents (iK2p)

K2P channels contain α-subunits with two tandem repeat domains (see Chapter 13); their unusual rectification, like that of the
iKir channels, reflects the lack of an S4 transmembrane segment. These two-pore channels help stabilize resting potential, but
make little or no contribution to cardiac action potentials. Their ability to increase “background” or “leak” currents in
response to stretch and other stimuli allows interactions of the two-pore channels with cell components like the cytoskeleton
to mediate proliferative responses to mechanical stress.
Chloride Currents
The Donnan equilibrium responsible for the potassium gradient across the plasma membrane (see above) also establishes a
chloride gradient that allows chloride fluxes to modify membrane potential. In the heart, intracellular chloride activity is ∼4
mM while that outside the cell is ∼80 mM) (Table 14-1), so that according to the Nernst equation, the chloride equilibrium
potential is about -80 mV
Because this potential is more positive than resting potential in working cardiac myocytes and Purkinje fibers, chloride
currents are inward (depolarizing) during diastole and outward when the cells are depolarized. (In nodal cells, which have
lower resting potential, chloride currents are only repolarizing.)
Most chloride currents are small. CLCA2 channels (see Chapter 13), which carry the transient outward current ito2, respond to
increased calcium levels by carrying repolarizing currents that shorten action potential duration. These currents contribute to
the ability of changes in serum calcium levels to modify the QT interval (see above) and, because they are activated by
protein kinase A, they help shorten action potential duration after sympathetic activation. CFTR channels, whose opening is
increased by protein kinases A and C, and by activated P2R purinergic receptors, also shorten cardiac action potentials.
Besides their role in determining membrane potential and excitability, chloride channels regulate cell volume, differentiation
and apoptosis, and chloride channels called VDAC1 participate in redox homeostasis and proliferative signaling.
Membrane Currents Generated by the Sodium Pump and the Sodium/Calcium

Exchanger
The sodium pump generates an outward current because it transports three sodium ions out of the cell in exchange for two
potassium ions (see Chapter 9). While the effect of this repolarizing current on membrane potential is normally small,
increases in this background current can become significant in sodium-overloaded cells, as occurs, for example, at rapid heart
rates.
The sodium/calcium (Na/Ca) exchanger generates both outward and inward membrane currents when it exchanges one
calcium ion for three sodium ions (see Chapter 7). In working cardiac
P.385
myocytes and His-Purkinje fibers, the exchanger generates an outward current immediately after these cell depolarize; this is
because iNa (the fast inward current) increases cytosolic sodium, which increases sodium efflux via the exchanger. The
resulting outward current contributes to early repolarization (phase 1). At the end of systole, however, the Na/Ca exchanger
generates an inward current because it transports calcium released during excitation-contraction coupling out of these cells.
This current, which is increased in calcium-overloaded cells, is maximal toward the end of the refractory period, when the
ventricles are most vulnerable to fibrillation (see below). For this reason, inward currents generated by the Na/Ca exchanger
at the end of systole are important causes of afterdepolarizations, triggered arrhythmias, and potentially lethal arrhythmias
(see below).
Regenerative Aspects of Depolarization and Repolarization

The wave of depolarization that activates the heart is rapidly propagated when electrotonic currents that flow between
depolarized and resting cells initiate regenerative action potentials (see Chapter 16). Repolarization, on the other hand, is not
caused by a propagated wave, but instead is governed largely by local factors, notably action potential duration. To a minor
extent, however, repolarization is regenerative because regions that have already repolarized generate electrotonic currents
that tend to restore resting membrane potential in adjacent depolarized regions of the heart.
The Interval-Duration Relationship

The interval-duration relationship is a physiological mechanism that prolongs cardiac action potential duration when the heart
slows, and shortens the action potential at fast heart rates. This response is especially important when the heart beats rapidly,
which requires that cardiac cycle length be shortened to allow sufficient time for the ventricles to fill. This is apparent from a
simple calculation. The normal ventricular action potential duration is ∼0.35 s when heart rate is 75 beats/min (total cycle
length = 0.80 s), yet trained athletes can achieve heart rates in excess of 180 beats/min, where total cycle length is only 0.33
s. In order for cardiac myocytes to be activated at the more rapid rate, the action potentials must become shorter. This is
made possible by an inverse relationship between heart rate and cycle length, called the interval-duration relationship (Fig.
14-9A and B), which allows a long diastolic interval to prolong the following action potential, and shortens an action potential
that follows a short diastolic interval.
When cardiac rhythm is irregular, as in patients with atrial fibrillation, the influence of the preceding diastolic interval on the
duration of the subsequent action potential is governed by the interval-duration relationship: long diastolic intervals are
followed by longer action potentials and short intervals by shorter action potentials (Fig. 14-9C). These responses explain a
phenomenon called the Ashman phenomenon, in which a stimulus that reaches the heart shortly after an action potential
preceded by a short diastolic interval is able to generate another action potential (Fig. 14-9D1), whereas a stimulus that
reaches the heart at the same time after an action potential that is preceded by a longer diastolic interval does not evoke an
action potential (Fig. 14-9D2). Stated succinctly, a short cycle that follows a short cycle can evoke a response (Fig. 14-9D1),
but when the same short cycle follows a long cycle, it may fail to generate a response (Fig. 14-9D2).
Several mechanisms explain the interval-duration relationship. Abbreviation of the action potential at rapid heart rates, where
diastolic intervals are short, is due in part to incomplete
P.386
decay of the delayed rectifier currents iKr and iKs, which carry outward currents that shorten action potential duration. At
faster heart rates, the plateau is also shortened by the slow recovery of iCaL channels that had been inactivated in the
preceding beat, which decreases the depolarizing calcium current carried by iCaL. The resulting decrease in calcium entry also
shortens action potential duration by reducing the depolarizing current carried by the Na/Ca exchanger. Action potential
duration can also be shortened when the greater frequency of iNa channel openings increases cytosolic sodium concentration,
which increases both early repolarization (see above) and the repolarizing current generated by the sodium pump.
Fig. 14-9: Interval-duration relationship. A: At slow heart rates, where diastolic intervals are long, action potential
duration is long. B: When diastolic intervals are short, action potential duration is also short. C: When the heart's
rhythm is irregular and cycle length varies, action potential duration is determined by the preceding diastolic interval.
D1 and D2: Because the length of the refractory period is correlated with action potential duration, a wave of
depolarization that arrives after a short cycle is less likely to encounter refractoriness left behind by a preceding cycle
if it had been preceded by a short diastolic interval (D1) than by a long diastolic interval (D2).
Afterdepolarizations and Triggered Activity

Afterdepolarizations are spontaneous depolarizations, not caused by external stimuli, that can appear during repolarization
(phase 3) or shortly after the cell has repolarized (phase 4) (Fig. 14-10). When afterdepolarizations are small, they cause only
small oscillations of membrane
P.387
potential; however, afterdepolarizations that exceed threshold can initiate premature depolarizations. These responses, called
triggered activity, can generate propagated action potentials that, when they become repetitive, can initiate life-threatening
tachycardias.
Afterdepolarizations and triggered activity, which are most likely to occur when the heart becomes calcium-overloaded, can
be attributed to the inward current generated by the Na/Ca exchanger when it removes calcium from the cell (see above).
The increased susceptibility of calcium-overloaded hearts to this arrhythmogenic mechanism is a major reason why energy
starvation causes sudden death in cardiac patients. A variety of drugs increase the susceptibility to these arrhythmias; these
include inotropic agents that increase cytosolic calcium, antiarrhythmic drugs, and such non-cardiac drugs as antidepressants
and anticonvulsants. Triggered arrhythmias are common in patients with long QT syndromes (see Chapter 16).
There are two types of afterdepolarization: early afterdepolarizations, which occur before the end of the action potential
when membrane potential ranges between -10 and -30 mV, and delayed afterdepolarizations that appear after membrane
potential has returned to its resting level (Fig. 14-10). Unlike most arrhythmias, which are suppressed when the heart is paced
rapidly (a phenomenon called overdrive suppression), delayed afterdepolarizations and triggered activity are made worse
when heart rate is increased, probably because of the gain in cytosolic calcium caused by the increased frequency of calcium
channel openings (see Chapter 10).
Action Potentials in Specific Regions of the Heart

Differences in the ionic currents that depolarize and repolarize specific regions of the heart, which are responsible for their
different action potential configurations (Fig. 14-11), reflect the specialized electrophysiological functions of the individual
structures.
Purkinje Fibers
The cells of the His-Purkinje system, which are specialized for rapid conduction (see Chapter 1), have very large, rapidly rising
action potentials; resting potentials are between -80 and -90 mV, and the overshoot can reach +25 mV, so that action potential
amplitude can be greater than +110 mV (Fig. 14-1). The rapid upstroke and large amplitude, along with a low internal
resistance, favor rapid conduction (see Chapter 16), while the long duration provides an important safeguard against reentrant
arrhythmias by avoiding impulses transmitted from the His-Purkinje system to the ventricular myocardium from reactivating
the former.
The currents responsible for the Purkinje fiber action potential are depicted in Figures 14-12 and 14-13. Figure 14-12 shows
the changes in total current flow during the action potential, while the major individual currents are shown in Figure 14-13.
Four inward currents are shown: the initial rapid iNa, the small transient iCaT, the larger more prolonged iCaL, and the
hyperpolarizing current ih. A small, long-lasting, inward sodium current that persists after the initial depolarizing current (iNa),
called iNaL, is caused by sodium influx through a partially open sub-state of these sodium channels.
The heart is repolarized by four delayed rectifier potassium currents (ito1, iKr, iKs, and iKCa) and the outward chloride current
ito2. Low levels of a more rapidly developing outward current, called iKur (not shown in Fig. 14-13), are also present in
Purkinje fibers.
P.388
Fig. 14-10: Afterdepolarizations. A: Early afterdepolarizations showing a subthreshold afterdepolarization (1), and
larger afterdepolarizations that generate a single (2) and repetitive (3-4-5) triggered responses. B: Late
afterdepolarizations showing a subthreshold afterdepolarization (1) and afterdepolarizations that generate a single (2)
or series (3-4-5) of triggered responses. (Modified from Wit and Rosen, 1992.)
P.389
Fig. 14-11: Action potential configurations in different regions of the mammalian heart.
Fig. 14-12: Purkinje fiber action potential (upper tracing) and total membrane currents (lower tracing); inward
currents are downward, outward currents are upward. The approximate timing of five different types of ionic current
are shown.
P.390
Fig. 14-13: Major ionic currents responsible for the Purkinje fiber action potential. Zero on the time scale marks the
onset of the action potential. Upper tracing: membrane potential. Lower tracings show 12 different currents; inward
currents are downward, outward currents are upward. Depolarizing ionic currents include iNa, iCaL, iCaT, and ih, while
repolarizing currents include four outward potassium currents (ito1, iKr, iKs, and iKCa) and an outward chloride current
(ito2). Resting potential is maintained by the inward rectifier iK1. The Na/Ca exchanger generates an outward current
immediately after depolarization, and an inward current toward the end of the action potential. The Na-K ATPase
generates a small background outward current throughout the cardiac cycle.
P.391
The Na/Ca exchanger generates a repolarizing current at the beginning of systole, when sodium that has entered the cytosol
via iNa increases the outward current caused when this sodium is exchanged for extracellular calcium by the exchanger. The
exchanger also generates a depolarizing inward current when the increased cytosolic calcium present at the end of systole is
exchanged for extracellular sodium. The Na-K ATPase generates a small outward current throughout the cardiac cycle.
Sinoatrial Node
Heart rate is normally controlled by the sinoatrial (SA) node, a band of spontaneously depolarizing cells located in the right
atrium near its junction with the superior vena cava (Chapter 1). Spontaneous depolarization of the SA node (sometimes called
phase 4 depolarization) provides the pacemaker which initiates the wave of depolarization that normally activates all regions
of the heart (Chapter 15). The firing of the SA node is coordinated by electrical interactions within this band of nodal cells,
and shifts within this pacemaker region can cause a benign arrhythmia called wandering pacemaker (see Chapter 16).
Resting potential in the SA node is low, about -70 mV; the action potentials are small and have a slow upstroke that reflects
the absence of fast sodium channels (Fig. 14-11). At least six currents participate in pacemaker activity by the SA node (Fig.
14-14); four inward currents and two outward potassium currents.
Inward Currents
Spontaneous pacemaker depolarization begins with the opening of the nonspecific cation channels that carry the
hyperpolarization-activated depolarizing current called ih (this current is also called if because of its unusual characteristics—f
stands for “funny”). As the membrane depolarizes, the thresholds for T- and L-type calcium channels are reached. T-type
channels, which have the lower threshold, are first to open, which adds a second small depolarizing current to that carried by
ih. When membrane potential reaches the threshold for the L-type calcium channels, the latter generate a larger depolarizing
current. The inward current generated by the Na/Ca exchanger when each of the calcium ions that entered the cytosol
through the T- and L-type calcium channels is exchanged for three sodium ions (see Chapter 7) also contributes to SA node
depolarization.
Outward Currents
Decreases in two outward potassium currents contribute to spontaneous depolarization by the SA node. The first, iK, is a
delayed rectifier current that ends the action potential. An increase in this repolarizing current accelerates the pacemaker by
shortening the SA node action potential, which decreases the time before the ih channels again begin to depolarize. The
outward current, iK.Ach, which is carried by potassium channels that are opened by acetylcholine and purinergic agonists,
slows heart rate by counteracting the inward current carried by the ih channels.
Autonomic Control of the SA Node Pacemaker

Heart rate is normally under both sympathetic and parasympathetic control. In the resting human heart, the dominant effect
is caused by the parasympathetic tone which slows spontaneous depolarization of the SA node. The high level of basal
parasympathetic tone becomes
P.392
P.393
apparent when a normal person is given the muscarinic receptor blocker atropine (Fig. 14-15), which doubles heart rate from
the basal level of 60 beats/min to about 120 beats/min. The lower level of basal sympathetic tone is evidenced by a much
smaller (∼17%) decrease in heart rate, to about 50 beats/min, following administration of the β-adrenergic receptor blocker
propranolol. When both sympathetic and parasympathetic activities are blocked, heart rate stabilizes at about 100 beats/min,
which represents the “intrinsic” resting heart rate. During exercise, sympathetic stimulation becomes dominant and can
increase heart rate to more than 180 beats/min in a trained athlete.
Fig. 14-14: Individual currents responsible for depolarization of a pacemaker cell in the SA node. Zero on the time scale
marks the onset of the action potential. Upper tracing: membrane potential. Spontaneous diastolic depolarization
caused by the slow opening of hyperpolarization-activated channels (ih) initiates an action potential by increasing two
inward calcium currents (iCaT, and iCaL) that are followed by inward current generated when calcium leaves the cell by
the Na/Ca exchanger. Repolarization, which is caused by an increase in the outward rectifier current iK, ends the
pacemaker cycle. The outward current iK.Ach helps stabilize resting potential; when this current is increased by
parasympathetic stimulation, heart rate slows.
Fig. 14-15: Hypothetical experiment showing the effects of autonomic blockade on heart rate. Abolishing
parasympathetic tone by administration of atropine, a muscarinic receptor blocker, to an individual with a basal heart
rate of 60 beats/min (open circle) doubles heart rate to 120 beats/min (upper shaded circle). However, abolishing
sympathetic tone with the β-adrenergic receptor blocker propranolol causes only a slight slowing of rate, to 50
beats/min (lower shaded circle). Blocking both parasympathetic and sympathetic tone by giving both atropine and
propranolol reveals the intrinsic pacemaker rate of 100 beats/min (closed circle).
Sympathetic stimulation increases heart rate when β-adrenergic receptor agonists activate Gαs. The latter stimulates adenylyl
cyclase to increase cyclic AMP production, which increases ih by direct binding to the hyperpolarization-activated channels.
Cyclic AMP also has indirect effects that increase heart rate when PKA-catalyzed phosphorylation increases iCaT, iCaL, and iK.
Phosphorylation of the channels that carry iCaT and iCaL accelerates the SA node pacemaker by increasing these depolarizing
currents, while phosphorylation of the iK channels increases heart rate when the increased rate of repolarization shortens
action potential duration (Fig. 14-16). SA node depolarization is also accelerated when the additional calcium that enters the
SA node via iCaT and iCaL generates an inward current when it is exchanged for sodium by the Na/Ca exchanger.
Parasympathetic stimulation slows the heart when Gαi activates the inward rectifier current iK.Ach which, by increasing
repolarizing current during diastole, prolongs the time required for ih to bring membrane potential to threshold. Gαi also
decreases heart rate by inhibiting cyclic AMP production.
P.394
Fig. 14-16: Mechanisms by which sympathetic stimulation accelerates the SA node pacemaker. The baseline action
potential is shown as a solid line. Cyclic AMP increases ih by a direct interaction with these cyclic nucleotide-gated
channels that accelerates diastolic depolarization. Cyclic AMP also accelerates the SA node pacemaker indirectly by
activating PKA-catalyzed phosphorylations that increase the depolarizing currents carried by iCaT and iCaL and the
repolarizing current iK, all of which shorten action potential duration. Not shown is a fourth mechanism that increases
heart rate; this is the increased inward current that occurs when the additional calcium that enters the cytosol as the
result of activation of iCaT and iCaL is exchanged for sodium by the Na/Ca exchanger.
Atria
Atrial action potentials are of shorter duration than those of the Purkinje fibers (Fig. 14-11) and usually lack pacemaker
activity. Depolarization (phase 0), which is effected by sodium channel opening, is rapid and is followed by a phase of rapid
repolarization (phase 1) and a brief plateau (phase 2) that often merges into the phase of repolarization (phase 3). The short
duration of the atrial action potential is due to rapid opening of the ultrarapid delayed rectifier current called iKur. Vagal
stimulation shortens both the action potential and refractory period in the atria by activating outward potassium currents; the
resulting abbreviation of the plateau reduces the duration of calcium channel opening, which contributes to the negative
inotropic response of the atria to parasympathetic stimulation.
Atrioventricular Node
The atrioventricular (AV) node can be divided functionally into three regions (Fig. 14-17): an AN (atrionodal) region, an N
(nodal) region, and an NH (nodal-His bundle) region. Resting potential is lowest in the N region while action potentials are
shortest in the AN region,
P.395
P.396
increasing progressively in duration through the N region to the NH region. Spontaneous pacemaker activity, which can appear
in all regions of the AV node, is most rapid in the lower (H and NH) regions and slowest in the N region. The normal delay in
impulse transmission from the atria to the ventricles (see Chapter 15) is caused by slow conduction in the AV node, where
resting potential is approximately -80 mV and membrane potential at the end of depolarization usually does not exceed +5 to
+10 mV.
Fig. 14-17: Atrioventricular (AV) conduction system. The AV node can be divided functionally into three regions: AN
(upper, or atrionodal), N (middle, or nodal), and NH (lower, or nodal-His bundle).
Fig. 14-18: Action potential configurations in different layers of the ventricular wall. A: epicardium; B: M cells of the
midmyocardium; C: endocardium. (Based on data from Liu and Antzelevitch, 1995.)
AV conduction is slow for several reasons. One is the low amplitude and slow rate of rise of the AV node action potentials (Fig.
14-11); both reflect the absence of a fast inward sodium current and the dependence of AV nodal conduction on small, slowly
developing calcium currents. Slow conduction in the AV node is also due to high internal resistance that is caused by the small
diameter of the AV nodal cells, especially in the N region, and a relatively small number of gap junctions. Sympathetic
stimulation accelerates AV conduction by increasing the inward calcium current, while vagal stimulation slows AV conduction
by inhibiting calcium channel opening.
Conduction from atria to ventricle across the AV node is normally precarious because the depolarizing calcium currents are
barely sufficient to generate a propagated action potential; for this reason, the AV node is said to have a low safety factor.
This property, which limits the frequency with which impulses can be transmitted from the atria to the ventricles, can cause
AV conduction to fail. This condition, called AV dissociation, protects against excessive rates of ventricular beating when the
atria are depolarized rapidly, as commonly occurs in patients with atrial flutter and fibrillation. Failure of conduction through
the AV node at normal rates of atrial depolarization is usually abnormal and is one cause of AV block (see Chapter 16).
However, because of the low safety factor and sensitivity of the AV node to parasympathetic inhibition, mild degrees of AV
block can be benign, especially in trained athletes who have high basal vagal tone.
Ventricles
Ventricular action potentials are smaller and more slowly rising than those of the Purkinje fibers, and spontaneous diastolic
depolarization is not normally seen in the ventricles (Fig. 14-11). Action potential duration is shorter than in the His-Purkinje
system; as noted above, this helps prevent reentrant arrhythmias that could occur if impulses transmitted to the ventricular
myocardium were to reactivate the His-Purkinje system. Action potential duration also differs in the different layers of the
ventricular wall, and is longest in midmyocardial cells (called m cells) (Fig. 14-18); this heterogeneity is due largely to the
uneven distribution of repolarizing ito1 and iKs channels.
Conclusions
An understanding of the ionic currents described in this chapter, and their clinical manifestations, which are recorded at the
body surface as the electrocardiogram, are the foundation for understanding electrocardiographic diagnosis, managing clinical
arrhythmias, and preventing sudden cardiac death.
P.397
Bibliography
(See also Bibliography for Chapter 13.)
General
Amin AS, Tan HL, Wilde AAM. Cardiac ion channels in health and disease. Heart Rhythm 2009;7: 118–126.
Clancy CE, Kass RS. Inherited and acquired vulnerability to ventricular arrhythmias: Cardiac Na+ and K+ channels. Physiol
Rev 2005;85:33–47.
Eisner DA, Dibb KM, Trafford AM. The mechanism and significance of the slow changes of ventricular action potential
duration following a change of heart rate. Exp Physiol 2009;94:520–528.
Gaborit N, Le Bouter S, Szuts V, et al. Regional and tissue-specific transcript signatures of ion channel genes in the non-
diseased human heart. J Physiol (Lond) 2007;582:675–693.
Grant AO. Cardiac ion channels. Circ Arrhythm Electrophysiol 2009;2;185–194.
Hille B. Ionic Channels of Excitable Membranes. 3rd ed. Sunderland MA: Sinauer, 2001.
Hoffman BF, Cranefield P. Electrophysiology of the heart. New York, NY: McGraw-Hill, 1960.
Josephson ME. Clinical cardiac electrophysiology. 4th ed. Philadelphia, PA: Lippincott/Williams and Wilkins, 2008.
Katz B. Nerve, Muscle and Synapse. New York, NY: McGraw-Hill, 1966.
Noble D. The initiation of the heartbeat. 2nd ed. Oxford: Clarendon Press, 1979.
Pogwizd SM, Bers DM. Cellular basis of triggered arrhythmias in heart failure. Trends CV Med 2004;14: 61–66.
Pogwizd SM, Schlotthauer K, Li L, et al. Arrhythmogenesis and contractile dysfunction in heart failure: Roles of sodium-
calcium exchange, inward rectifier potassium current, and residual beta-adrenergic responsiveness. Circ Res
2001;88:1095–1096.
Wit AL, Rosen MR. Afterdepolarizations and triggered activity. Distinction from automaticity as an arrhythmogenic
mechanism. In: Fozzard H, Haber E, Katz A, et al., eds. The heart and cardiovascular system. 2nd ed. New York, NY:
Raven Press, 1992:2113–2165.
Zipes DP, Jalife F. Cardiac electrophysiology. From cell to bedside. 5th ed. Philadelphia, PA: WB Saunders, 2009.
References
Blinks JR. Intracellular Ca2+ measurements. In: Fozzard H, Haber E, Katz A, et al., eds. The heart and cardiovascular
system. New York, NY: Raven, 1986:671–701.
Cohen CJ, Fozzard HA, Sheu SS. Increase in intracellular sodium ion activity during stimulation in mammalian cardiac
muscle. Circ Res 1982;50:651–662.
Liu DW, Antzelevitch C. Characteristics of delayed rectifier current (IKr and IKs) in canine ventricular epicardial,
midmyocardial, and endocardial myocytes. A weaker IKs contributes to the longer action potential of the M cell. Circ Res
1995;76:351–365.
Özgen N, Rosen MR. Cardiac memory: a work in progress. Heart rhythm 2009;6:564–571.
Spear JF, Moore EN. Supernormal conduction in the canine bundle of His and proximal bundle branches. Am J Physiol
1980;238:H300–H306.
Walker J. Intracellular inorganic ions in cardiac tissue. In: Fozzard H, Haber E, Katz A, et al., eds. The heart and
cardiovascular system. New York, NY: Raven, 1986:561–572.
Weidmann S. Effect of current flow on the membrane potential of cardiac muscle. J Physiol (London) 1951;115:227–236.
> Table of Contents > Part Four - Pathophysiology > Chapter 15 - The Electrocardiogram
Chapter 15
The Electrocardiogram
The electrocardiogram (abbreviated ECG from the English spelling, or EKG from the Dutch) uses
electrodes placed on the body surface to record the electrical activity of the heart. These tracings
provide measurements of the heart's ability to initiate electrical impulses (chronotropy) and conduct
action potentials (dromotropy), but contain virtually no direct information about contractility (inotropy)
or relaxation (lusitropy).
All adult cardiac myocytes can respond to electrical stimuli and conduct action potentials, but the
working cells of the atria and ventricles generally lack automaticity, and the cells of the sinoatrial (SA)
and atrioventricular (AV) nodes and His-Purkinje system have virtually no ability to contract and relax.
Conduction velocity also differs in various regions of the heart; it is very slow in the SA and AV nodes,
more rapid in the atria and ventricles, and very rapid in the His-Purkinje system. These and other
manifestations of the heart's electrical activity can be evaluated using the ECG, which provides invaluable
pathophysiological information about patients with heart disease.
Chronotropy: Pacemakers and Impulse Formation

The primary pacemaker of the heart is the SA node, which is derived from the sinus venosus, which is the
most rapidly beating portion of the embryonic heart. These tubular hearts, like those of primitive
animals, can be divided into four regions: sinus venosus, atria, ventricles, and truncus arteriosus (Fig. 15-
1). Although embryonic atria and ventricles contain pacemaker cells, their spontaneous depolarizations
are suppressed by the faster sinus venosus pacemaker; this is because the intrinsic rate of pacemaker
depolarization decreases as one proceeds from the higher (upstream) venous to the lower (downstream)
arterial end of the embryonic heart. As a result, the pacemaker cells in the embryonic atria and
ventricles do not have time to reach the threshold needed to initiate a propagated wave of depolarization
unless they are isolated from the more rapidly firing sinus venosus.
A similar hierarchy of pacemaker activity exists in the adult human heart (Table 15-1), where impulses
conducted from the SA node pacemaker normally suppress the slower pacemaker activity in the AV node
and His-Purkinje system. The activity of the lower pacemakers can become manifest only when they fire
more rapidly, when the SA node pacemaker is slowed, or when conduction of impulses generated in the SA
node is blocked (see Chapter 16).
The rate of pacemaker discharge by the normal resting human SA node is 60 to 100 times per minute,
whereas the firing rate of the most rapid lower pacemakers, generally found in the lower (NH) region of
the AV node, is ∼40 to 55 beats/min. Pacemaker cells in His-Purkinje system depolarize spontaneously at
rates of ∼25 to 40 beats/min (Table 15-1). Unfortunately, in spite of their large number, these lower
pacemakers often fail to initiate a propagated wave
P.402
of depolarization when they become isolated electrically from the SA node. The resulting cessation of
ventricular contraction is an important cause of cardiac arrest.
Fig. 15-1: Embryonic or primitive heart. Blood flow in these tubular hearts is from left to right. The intrinsic
rates of the intrinsic pacemaker activity of each chamber decreases as one moves from the sinus venosus to
the truncus arteriosus. This was demonstrated by Stannius, who in 1852 placed tight ligatures at the
sinoatrial junction (first Stannius ligature) and the atrioventricular junction (second Stannius ligature) and
found that elimination of the influence of the more rapid higher (“upstream”) pacemakers revealed the
previously suppressed activity of pacemaker cells in the lower (“downstream”) regions.
Dromotropy: Impulse Propagation Through the Heart

All conduction in the heart is via cardiac muscle; nerves serve only to regulate impulse generation and
conduction. The normal activation sequence, shown in Table 15-1, reflects the heart's embryology and
anatomy. The impulse that originates in the SA node first activates the atria, after which it is conducted
though the AV node, AV bundle, bundle branches, and His-Purkinje
P.403
system before it can activate the ventricles. In the latter, the last regions to be activated are the
posterobasal left ventricle and the right ventricular outflow tract, both of which are derived in part from
the truncus arteriosus. Conduction velocity is most rapid in the AV bundle, bundle branches, and Purkinje
network; less rapid in atrial and ventricular myocardium; and much slower in the SA and AV nodes (Table
15-1).
Table 15-1 Normal Activation in the Human Heart
Sequence Conduction Velocity (m/s)a Pacemaker Rate (min-1)a

SA node ↓ <0.01 60–100
Atrial myocardium ↓ 1.0–1.2 None
AV node ↓ 0.02–0.05 40–55
AV bundle ↓ 1.2–2.0 40–55
Bundle branches ↓ 2.0–4.0 25–40
Purkinje network ↓ 2.0–4.0 25–40
Ventricular myocardium 0.3–1.0 None
aValues are approximations.
Sa Node
The SA node in human hearts is a cigar-shaped collection of specialized cells located at the junction
where the superior vena cava enters the right atrium. Impulses generated in the major pacemaker site,
which lies near the center of the SA node, can follow different pathways before activating contiguous
areas of the right atrium at the periphery of the node. Variations in these pathways can give rise to a
benign arrhythmia called “wandering pacemaker” that is sometimes seen in normal ECGs (see Chapter
16).
Atrial Depolarization
After entering the atria from the SA node, the wave of depolarization reaches the left atrium and AV node
through three preferred conducting pathways that are generally referred to as internodal tracts or ring
bundles (Fig. 15-2). The anterior tract has two branches: one, called Bachmann's bundle, transmits
impulses from the right atrium to the left atrium, while the second links the SA and AV nodes. The others,
called the middle and posterior internodal tracts (named after Wenckebach and Thorel, respectively),
also serve as preferential conduction pathways that link the SA and AV nodes.
Fig. 15-2: Internodal tracts as seen in a posterior view of the heart. The anterior, middle
(Wenckebach), and posterior (Thorel) internodal tracts provide preferential conduction pathways
between the sinoatrial and atrioventricular nodes (cross hatched). A branch of the anterior
internodal tract, sometimes called Bachmann's bundle, provides a conduction pathway that links
the right and left atria. (Redrawn from James, 1967.)
P.404
Fig. 15-3: The atrioventricular (AV) bundle provides the only conduction pathway that normally
transmits impulses across the central fibrous body, which serves as a connective tissue insulator
between the atria (above) and ventricles (below). The AV node controls the access of atrial
impulses to the AV bundle.
Although there is electrophysiological evidence for these rapidly conducting pathways in the atria,
histological studies fail to demonstrate bundles of specialized cells surrounded by a fibrous sheath, which
is the usual anatomical definition of a tract. For this reason, the rapidly conducting internodal tracts
probably represent functional pathways in which preferential conduction is made possible by regions
enriched in rapidly conducting Purkinje-like atrial myocytes, possibly within pectinate muscles.
Atrioventricular Conduction
Key to understanding AV conduction is that the heart's fibrous skeleton serves as an electrical insulator
which separates the atria and ventricles, and that the wave of depolarization transmitted from the SA
node to the atria can normally reach the ventricles only by crossing the central fibrous body through the
AV bundle (also called the bundle of His, His bundle, or common bundle, see Chapter 1). Electrical
communication between the atria and the AV bundle is controlled by the AV node, a group of small nodal
cells located near the coronary sinus in the posterior wall of the right atrium (Fig. 15-3). Slow conduction
through the AV node, which prolongs the interval between atrial and ventricular systole, is critical to the
ability of the atria to serve as a “primer pump” (see Chapter 1).
Abnormal electrical connections can link the atria and ventricles. These include accessory pathways,
sometimes called a bundle of Kent, which are strands of atrial myocardium that cross the central fibrous
body at many locations (Fig. 15-4). Conduction velocity in an accessory pathway is more rapid than the AV
node, which allows these pathways to provide a “short circuit” that bypasses the normal delay caused by
slow conduction in the AV node. Rapid conduction of atrial impulses into the ventricles by an accessory
pathway is responsible for a condition called pre-excitation or the Wolff–Parkinson–White syndrome
(WPW, see Chapter 16). Other abnormal conduction pathways,
P.405
called bypass fibers of James, can link the atria to the upper portion of the AV bundle and so bypass the
normal conduction delay in the AV node. Mahaim fibers, which provide another abnormal pathway, can
transmit impulses from the upper part of the AV bundle to the ventricles. Conduction through these
abnormal pathways is a common cause of supraventricular tachycardias in which impulses go back and
forth between the atria and ventricles—in one direction through the AV node and in the other direction
through the abnormal pathway. Dual conducting pathways in the AV node are an even more common cause
of supraventricular tachycardias (see Chapter 16).
Fig. 15-4: Abnormal conduction pathways that can link the atria and ventricles. These include
accessory pathways (bundle of Kent), whose locations vary; bypass fibers of James, which connect
the atrial myocardium to the upper portion of the atrioventricular (AV) bundle; and Mahaim fibers,
which connect the AV bundle to abnormal sites in the ventricles.
Activation of the Ventricles: Bundle Branches, Fascicles, and

Purkinje Network
The specialized conduction system of the heart, which is made up of rapidly conducting Purkinje fibers,
includes the AV bundle, bundle branches, fascicles, and Purkinje network. The AV bundle divides into the
right and left bundle branches at the top of the interventricular septum (see Chapter 1). The right bundle
branch generally continues in a structure called the moderator band until it reaches the right ventricular
free wall (Fig. 15-5), whereas
P.406
the left bundle branch fans out over the left ventricle (Fig. 15-6). The left bundle branch is
conventionally viewed as dividing into two branches, called the anterior and posterior fascicles, that
conduct impulses to the anterior and posterior walls of the left ventricle. The electrocardiographic
abnormalities seen when conduction in one of these fascicles is interrupted are left anterior fascicular
block (or left anterior hemiblock) when the anterior fascicle is blocked, and left posterior fascicular
block (or left posterior hemiblock) when conduction is blocked in the posterior fascicle (see below).
Although fascicular blocks are well-defined electrocardiographic entities, their anatomic basis is tenuous
because division of the left bundle branch is highly variable and discrete anterior and posterior fascicles
are uncommon (Fig. 15-7).
Fig. 15-5: Conduction system of the human right ventricle viewed from the right after removal of a
portion of the right atrial and ventricular walls. Note the aorta (1); pulmonary artery (2); superior
vena cava (3); inferior vena cava (4); fossa ovale of the interatrial septum (5); Thebesian valve
overlying the coronary sinus (6); false tendon (7); and medial leaflet of the tricuspid valve, which
has been separated from its point of insertion (8). The atrioventricular (AV) node (9), which lies in
the right atrium above the tricuspid valve, receives branches from the atria and continues toward
the ventricles where it merges into the AV bundle at the top of the membranous septum (which has
been partially opened). The latter divides into the right and left bundle branches; the right bundle
branch (11) courses along the right side of the interventricular septum and continues within the
moderator band, usually without branching, to reach the anterior papillary muscle of the right
ventricle (10). (Modified from Wenckebach and Winterberg, 1927.)
Fig. 15-6: The same heart shown in Figure 15-5, viewed from the left after removal of a portion of
the left atrial and left ventricular walls. Note the aorta (1); pulmonary artery (2); left atrium (3);
and two cusps of the aortic valve (4, 5). The left bundle branch, which emerges through an opening
made by removal of a portion of the membranous septum, originates as a broad band (6). The left
bundle branch then fans out into anterior (7) and posterior (8) fascicles, which course toward the
posterior (9) and anterior (not shown) papillary muscles of the left ventricle. (Modified from
Wenckebach and Winterberg, 1927.)
The wave of depolarization that activates the working myocardial cells of the ventricles is propagated by
the Purkinje network, a network of rapidly conducting cardiac myocytes that arise from the bundle
branches and course within the inner third of the ventricular walls. Rapid conduction by these Purkinje
fibers maximizes cardiac efficiency by synchronizing contraction of the ventricular walls. Damage to this
specialized conduction system, which is seen in patients with advanced heart failure, delays activation of
a portion of the left ventricle. The adverse clinical consequences of the resulting inhomogeneities in the
walls of the ejecting heart can be ameliorated by a procedure called cardiac resynchronization therapy,
which uses electrical stimulation to activate regions of the left ventricle where activation is delayed.
P.407
Fig. 15-7: Division of the left bundle branch in 49 human hearts showing that simple bifurcation
into anterior and posterior fascicles is uncommon. (From Demoulin J-C, Thesis. Personal
communication.)
The Electrocardiogram
The electrocardiogram uses electrodes placed on the body surface to record potential differences arising
within the heart (Fig. 15-8). The conventional speed of these recordings is 25 mm/s, so that each large
horizontal division (time) represents 0.20 s, and each small division is 0.04 s. In the normally standardized
ECG, each large vertical division (voltage) equals 0.5 mV and the small divisions are 0.1 mV.
The waves recorded by the ECG were named at the beginning of the 20th century by Einthoven, who
invented the string galvanometer that made possible the first high fidelity records of the electrical
activity of human hearts. To avoid controversies engendered by earlier nomenclature, Einthoven started
in the middle of the alphabet and named these deflections P, Q, R, S, T, and U (Fig. 15-9). By convention,
an interval is the time between the beginning of one deflection and the beginning or end of another,
while segments are inscribed between the end of one deflection and the beginning of the next deflection.
The ECG provides no information regarding the absolute level of membrane potential; a “zero” potential
cannot be measured because these tracings record only potential differences. The fact that the ECG
records only electrical activity and not the mechanical behavior of the heart is sometimes overlooked.
There is a weak correlation between the amplitude of the deflections inscribed during atrial or
ventricular depolarization and the force of contraction, but these correlations are of limited value.
P.408
Fig. 15-8: Normal 12-lead electrocardiogram. The leads in the upper three tracings are recorded
simultaneously; the lower tracing is a continuous lead II. The boxlike deflection at the left is a 1 mV
standardization artifact that causes a 1.0 mm upward shift in the baseline. Paper speed is 25 mm/s, so that
each heavy division is 0.2 s, and each of the smaller divisions is 0.04 s.
The P Wave: Atrial Depolarization

The first deflection recorded by the ECG, named the P wave by Einthoven (Fig. 15-9), is caused by atrial
depolarization. Although depolarization of the SA node precedes atrial depolarization (Fig. 15-10), the
potential differences generated in this structure are very small and cannot be recorded at the body
surface.
Fig. 15-9: Relationship between the electrocardiogram (ECG) and an intracardiac electrogram. A:
Potential differences recorded by the ECG from leads at the body surface showing waves,
complexes, segments, and intervals. B: The same electrical events recorded by an electrode
catheter placed over the AV bundle within the heart. The AH interval is the time between atrial
depolarization (A) and the passage of the impulse through the atrioventricular (AV) bundle (H) and
the HV interval is the time between depolarization of the AV bundle (H) and ventricular activation
(V).
P.409
Fig. 15-10: Tissues depolarized by a wave of activation initiated in the sinoatrial node
superimposed on the deflections of the electrocardiogram. Depolarization of many important
structures during the PR interval does not generate potential differences sufficiently large to be
recorded at the body surface. SA, sinoatrial; AV, atrioventricular.
The width (duration) of the P wave, which reflects the time taken for the wave of depolarization to
spread over the atria, can be prolonged by atrial enlargement or an intra-atrial conduction delay.
Potential differences generated during atrial repolarization, which can be recorded as the TP wave (the
“T of the P”), are rarely seen because they are small and usually “buried” in the much larger QRS
complex. TP waves can be seen when P waves are not followed by QRS complexes, as in AV block (see
Chapter 16).
The PR Interval: Atrioventricular Conduction

The ECG returns to its baseline at the end of atrial depolarization where it normally remains until the
onset of ventricular depolarization. However, a great deal is happening in the heart during this “silent”
period (Fig. 15-10).
The interval between atrial and ventricular activation, called the PR interval, begins with the first
deflection of the P wave and ends with the first deflection of the QRS complex (whether the latter is a Q
wave or an R wave as defined below). During the PR interval, the wave of depolarization is propagated
through the AV node, AV bundle, bundle branches, fascicles of the left bundle branch, and Purkinje
network (Fig. 15-10). Activation of these specialized conducting structures is extremely important, but
does not influence the body surface ECG because of their small size.
Electrode catheters passed into the right heart and placed over the AV bundle can record a His bundle
electrogram whose timing can be used to measure the passage of the wave of depolarization from the
atria to the AV bundle (AH interval), and from the AV bundle to the ventricles (HV interval) (Fig. 15-9).
Prolongation of the PR interval caused by a conduction delay in the AV node prolongs the AH interval,
whereas a delay in AV conduction that is due to slow passage of the wave of depolarization through the AV
bundle prolongs the HV interval (see Chapter 16).
P.410
The QRS Complex: Ventricular Depolarization

The QRS complex is generated by potential differences that appear during the upstrokes (phase 0) of the
action potentials that activate the ventricular myocardium (Fig. 15-11). The relationship between the
electrical events in the ventricles and the potential differences that can be recorded at the body surface
shown in Figure 15-11 is oversimplified because the ECG is influenced by the action potentials in all of the
millions of cells that are depolarized and repolarized during each cardiac cycle.
The “waves” (deflections) of the QRS complex are named according to the following convention: Q, any
initial downward deflection followed by an upward deflection (if there is only a downward deflection, this
is called a QS); R, the first upward deflection whether or not it is preceded by a Q wave; S, any downward
deflection preceded by an R wave. Additional upward deflections after S waves are called Rœ, R–, etc.,
and additional downward deflections after R waves are Sœ, S–, etc.
Loss of the normal synchrony of right and left ventricular depolarization prolongs the QRS complex. This
can occur when conduction is blocked in one of the bundle branches (bundle branch block), or when a
ventricular premature depolarization activates one ventricle before the other (see Chapter 16). The QRS
complex can also be prolonged by slowed conduction in the distal His-Purkinje system; this condition,
sometimes called arborization block, can occur in patients with severe cardiac disease, notably end-stage
heart failure.
Fig. 15-11: Temporal relationships between the electrocardiogram (A) and a ventricular action potential (B).
The QRS complex is produced by the upstrokes (phase 0) of all of the action potentials in the ventricles; the
isoelectric ST segment corresponds to the plateau (phase 2), while the T wave is inscribed during ventricular
repolarization (phase 3). The isoelectric segment between the T wave and the P wave of the subsequent
cardiac cycle, called the TP segment, is inscribed during ventricular diastole (phase 4).
P.411
The ST Segment: Plateau of the Ventricular Action Potential

Following inscription of the QRS complex, the ECG normally returns to, or very nearly to, its baseline,
where it remains until repolarization generates a T wave. This isoelectric phase, called the ST segment,
corresponds to the plateau (phase 2) of the ventricular action potentials (Fig. 15-11), when there are no
potential differences at the body surface because all regions of the ventricle are depolarized. The fact
that the normal ECG is at the baseline during both the ST segment and the TP segment (between the T
wave and the P wave of the subsequent cardiac cycle) might seem to be counterintuitive because the ST
segment is recorded when the ventricles are fully depolarized, whereas the TP segment occurs when the
ventricles are fully repolarized and at their resting potential. The reason that the record is at the baseline
at both times is that there are no potential differences when all regions of the ventricles are depolarized
and when they are all at their resting potential.
When the ST segment is displaced relative to the TP segment, it is not possible to determine whether the
abnormality is due to abnormal potential differences in the resting ventricles (during the TP segment),
abnormal potential differences during depolarization (during the ST segment), or both. By convention, the
TP segment is assumed to represent the baseline (“zero” potential), so that both are described as ST
segment shifts.
The T Wave: Ventricular Repolarization

Repolarization of the ventricles generates the T wave (Fig. 15-11). Unlike the sharp deflections of the QRS
complex, which are due to rapid changes in the direction followed by the wave of depolarization as it is
propagated over the ventricles, the T wave is broad because repolarization is a slower, non-propagated
process. Local factors and heterogeneities in the delayed rectifier potassium channels have a major
influence on repolarization, and are the main reason why the sequence of ventricular repolarization
differs from that of depolarization (see below).
The U Wave
Small deflections, called U waves, are sometimes seen after the end of the T wave. In spite of decades of
effort, the mechanism that causes U waves is not understood. Plausible explanations include delayed
repolarization of the endocardium caused by the long action potentials in the Purkinje fibers that run in
the subendocardium, delayed repolarization of the papillary muscles, the long action potentials in the M
cells that are located in the midmyocardium, and depolarizing currents generated in early diastole by
afterdepolarizations or mechano-electrical feedback.
The QT Interval
The QT interval, which is the time between the onset of the QRS complex and the end of the T wave,
provides a useful index of ventricular action potential duration (Fig. 15-11). QT prolongation and QT
shortening are important clinically because the underlying repolarization abnormalities are substrates for
arrhythmias and sudden cardiac death. Measurement of the QT interval is useful in evaluating potential
arrhythmogenic side effects of drugs that modify ventricular repolarization, and in identifying patients
with ion channel abnormalities that predispose to sudden cardiac death (see Chapter 16).
Table 15-2 Some Normal Values in the Normal Adult Human ECG
Duration (s)
PR interval 0.12–0.20
QT interval 0.35–0.40 (depends on heart rate)
QRS duration <0.12 (values of 0.10 and 0.11 can be abnormal)

Normal Intervals and Durations in the ECG
The durations of the key waves and intervals of the normal adult ECG are listed in Table 15-2. Many are
age-dependent, some vary with heart rate, and the QT interval varies with heart rate and is slightly
longer in women than in men. Formulae have been devised to quantify the relationship between the QT
interval and heart rate, but because the calculations are complex, reference to a table of normal QT
intervals at various heart rates in men and women is the best way to determine whether the QT interval is
abnormal in a given patient.
The Heart as a Dipole in a Volume Conductor

The electrical activity of the heart is routinely measured by 12-lead electrocardiograms that record
potential differences between electrodes placed at standard positions on the body surface (see below).
The pioneers in electrocardiography, in an effort to make their interpretations more scientific, evaluated
the magnitude and direction of the electrical vectors responsible for these potential differences using a
simple model that viewed the heart as a dipole in a volume conductor (Fig. 15-12).
A dipole is an electrical source that can be represented as an asymmetrically distributed electrical

charge. The heart can be viewed as a dipole at any instant during depolarization or repolarization
because the outsides of the depolarized cells in the excited regions are negatively charged relative to the
cytosol, while the outsides of the resting cells in the unexcited regions are positively charged (see
Chapter 14). The dipole in Figure 15-12, which is analogous to a single frame in a motion picture, depicts
the ventricles at one instant during depolarization, when an electrical dipole is created by a resting
(positive) region on the right and a depolarized (negative) region on the left.
Recording of the potential differences generated by cardiac dipoles requires that there be a conducting
medium, called a volume conductor, between the heart and the recording electrodes. The volume
conductor in Figure 15-12 is a dish of saline; in electrocardiography, the body tissues provide the volume
conductor that transmits potentials generated within the heart to the body surface. The isopotential lines
in Figure 15-12 show that the potentials in the half of the volume conductor occupied by the negative
pole are negative, while those in the other half are positive. Projection of the isopotential lines to the
surface of the volume conductor illustrates that the major determinant of the magnitude of the potentials
recorded on the ECG is the angle between the cardiac dipole and the leads on the body surface. Potential
also falls in proportion to the
P.412
square of the distance between an electrode and the dipole, but distance is much less important than the
vector angle and so is usually ignored in the clinical ECG.
Fig. 15-12: The partially depolarized ventricles can be depicted as an electrical dipole in a volume
conductor. The dipole is the rectangle in the center of the figure, which is positively charged to the
right and negatively charged to the left. The volume conductor is shown as a circular dish of saline
(solid line) within which isopotential lines have been drawn (dashed lines). Potential differences
(indicated in millivolts) can be recorded from the surface of the volume conductor using electrodes
such as A and B, which are in regions where the potentials are -1 and +1 mV, respectively. Also
shown is an “indifferent” electrode (V) that is connected to four electrodes placed on the surface
of the volume conductor so that the sum of the potentials is zero.
The model of the heart as a dipole in the center of a homogenous volume conductor shown in Figure 15-12
is greatly simplified. In the first place, the heart is not a single dipole because the complex distribution of
depolarized and repolarized cells within the heart creates multiple simultaneous dipoles. Another
simplification is that the body is not a homogeneous volume conductor; for example, the lungs represent
a region of high electrical resistance, which explains why emphysema or pneumothorax can reduce the
amplitude and distort the ECG.
Electrocardiographic Lead Systems

Two types of lead system can be used to measure the potential difference generated by a dipole in a
volume conductor. The simplest is a bipolar lead, which records the potential difference between two
electrodes, both of which are influenced by the dipole. In the example shown in Figure 15-12, the
potential difference between leads A and B is 2 mV. Whether the recorded potential is +2 mV or -2 mV
depends on convention: if electrode A is defined as “zero,” then B records +2 mV; if B is chosen as “zero,”
then A is -2 mV. Because these are easy to construct, bipolar lead systems were the first to be used in
clinical electrocardiography.
Attempts to make electrocardiography more scientific led to the introduction of unipolar leads in which
one electrode (the “exploring” electrode) measures the potential generated by the dipole, while the
other (the “indifferent” electrode) is assumed not to be influenced by the dipole and so to record “zero”
potential. There are two ways to create an indifferent electrode. One is to place the electrode far away
from the dipole, which takes advantage of the decrease in potential as the distance from the dipole
increases. If, for example, a patient is placed in a corner of a salt water swimming pool, an indifferent
electrode can be used to record from the opposite corner. This method, in addition to being impractical,
records only small potential differences because the indifferent electrode must be far away from the
patient. The second type of indifferent electrode
P.413
is made by connecting several electrodes, all of which are influenced by the dipole, to one another at a
“central terminal.” The latter is assumed to record “zero” potential because the potentials at the many
electrodes tend to cancel one another. If the indifferent electrode (V) in Figure 15-12 is assigned a value
of “zero,” the potential difference recorded between electrode A and the V lead is -1 mV, while electrode
B records a potential of +1 mV relative to the V lead. The indifferent electrode used in clinical
electrocardiography is the V lead, devised by Frank Wilson, which is constructed by connecting electrodes
placed on the two arms and left leg to a central terminal (see below).
Recording of the Dipoles Generated During Depolarization and

Repolarization
The records generated by changing potential differences in the heart are governed by conventions that
were introduced by Einthoven. According to these conventions, an approaching wave of depolarization,
which is in an area of positivity, inscribes an upright deflection, and a wave of depolarization that recedes
from an electrode inscribes a downward deflection because it is in an area of electronegativity. The
operation of these conventions can be understood by examining a bipolar lead that records the electrical
potentials generated during depolarization and repolarization of a strip of myocardium (Fig. 15-13).
When the strip is at rest, its surface is uniformly positive and no potential differences are recorded; for
this reason the record generated by the bipolar lead RA-LA remains at baseline (Fig. 15-13A). Stimulation
of the left side of the strip initiates a wave of depolarization that, as it is propagated from left to right,
causes a potential difference to appear between electrodes RA and LA (Fig. 15-13B). If electrode RA is
chosen to represent “zero” potential, electrode LA is in a region of electropositivity which, according to
Einthoven's convention, causes the recording device to inscribe an upward deflection. As the wave of
depolarization moves along the strip, the potential difference first increases, reaching a peak when half
of the strip is depolarized, and then decreases to zero when the entire strip becomes depolarized.
When the entire strip has become depolarized, there is no potential difference between electrodes RA
and LA because both electrodes are in a region of electronegativity; for this reason, the record returns to
baseline (Fig. 15-13C). This highlights the fact that when there is no potential difference between the two
electrodes, the system cannot distinguish between fully depolarized and fully repolarized tissue.
If repolarization were to begin at the same point on the strip at which the propagated wave of
depolarization began, a potential difference is created that is opposite in polarity to that seen during
depolarization (Fig. 15-13D). Because electrode LA faces an area of electronegativity relative to electrode
RA during repolarization, according to Einthoven's conventions a downward deflection is recorded. If this
wave of repolarization is propagated at the same velocity as that of depolarization, the downward
deflection caused by repolarization will be a mirror image of the upward deflection caused by
depolarization. When the strip is fully repolarized, no potential difference is recorded and the record
returns to baseline (Fig. 15-13E).
Figure 15-13 shows that, according to the conventions of electrocardiography, an electrode that faces an
approaching wave of depolarization records a positive potential and inscribes an upright deflection, and
an electrode that faces a receding wave of depolarization records a negative potential and inscribes a
downward deflection. Had electrode LA rather than electrode RA been chosen to represent zero, Figure
15-13B would have recorded depolarization as a downward deflection.
P.414
Fig. 15-13: Schematic depiction of the depolarization and repolarization in a strip of myocardium. A bipolar
lead system is used to measure the potential difference between electrodes RA (right arm) and LA (left arm),
and a recorder (right) has been wired so that an upward deflection is recorded when LA is more positive than
RA. A: Resting. The entire surface is positively charged (relative to the interior of the cells in the strip) so
that the bipolar lead records no potential difference between electrodes RA and LA; for this reason the
record remains at baseline. B: During depolarization. The strip of myocardium, shortly after it has been
stimulated at its left side (s), is slightly more than half depolarized (shaded area). Because the surfaces of
the myocytes in the depolarized area are negatively charged, electrode LA faces a region of greater positivity
than electrode RA. According to the conventions of electrocardiography, the recorder writes an upright
deflection that reaches its maximum when half of the myocardial strip is depolarized. C: Fully depolarized.
The strip of myocardium is fully depolarized and the action potentials of all cells are in phase 2, so that their
external surfaces are negative. Because there are no potential differences between the cell surfaces in the
strip of myocardium, electrodes RA and LA face a similar degree of negativity; as a result, the recorded
deflection returns to baseline. D: During repolarization. If repolarization of the strip of myocardium begins at
the same point that was first to be depolarized, the cell exteriors in the repolarized region at the left will be
the first to return to the normal resting positivity. As a result, electrode LA will face a region of greater
negativity than electrode RA and the recorder will inscribe a downward deflection. E: Fully repolarized. The
strip of myocardium has returned to resting potential; because there is no potential difference between
electrodes RA and LA, the record returns to baseline.
Concordance of QRS Complexes and T Waves

In the normal ECG, QRS complexes and T waves are normally concordant; that is, they generally point in
the same direction (see Fig. 15-8). This differs from the relationships shown schematically in Figure 15-
13, where the QRS and T are discordant. The reason for this discrepancy is that in the normal heart, the
last areas of the ventricles to depolarize are the first to repolarize. The consequences are illustrated in
Figure 15-14, which differs from Figure 15-13 because repolarization is shown as proceeding in a direction
opposite to that of depolarization.
P.415
Fig. 15-14: Strip of myocardium depicted in Figure 15-13 in which repolarization has begun in the region that
was last to depolarize D. Unlike Figure 15-13D, where both repolarization and depolarization begin at the
same point on the left of the strip, the direction of repolarization is opposite to that of depolarization.
Because the cell exteriors in the repolarized region at the right are the first to return to the normal resting
positivity, electrode LA (left arm) faces a region of greater positivity than electrode RA (right arm) during
repolarization and the recorder inscribes an upward deflection. As a result, the deflections recorded during
depolarization and repolarization are concordant. Because repolarization is not a rapidly propagated wave, it
causes a deflection that is more rounded than that caused by depolarization.
Figure 15-14 provides the more accurate depiction of the normal ECG, where the deflections inscribed
during repolarization (the T waves) are concordant with those inscribed during depolarization (the QRS
complexes). The most likely mechanism is that the direction of the spread of repolarization is influenced
by the action potentials in the Purkinje fibers in the endocardium, which are longer than those of the
ventricular myocytes in the epicardial regions (see Chapter 14). This causes the first regions of the
ventricles to depolarize to be last to repolarize.
Electrical Vectors
The potential differences that appear in the heart (cardiac dipoles) can be represented by vector arrows
whose length and orientation are determined by the magnitude and orientation of the dipole (see below).
In Figure 15-15A, the dipole in Figure 15-14B has been placed in a drawing of the human chest with its
positive side toward the left arm. An electrode on the right arm (RA) will therefore be in a region of
electronegativity, and an electrode on the left arm (LA) will be in a region of electropositivity. This dipole
can be represented by a vector arrow (Fig. 15-15B), which according to yet another convention points to
the region of electropositivity, and its
P.416
length indicates the magnitude of the potential difference. Because of these conventions, cardiac vector
arrows point in the same direction as that followed by a wave of depolarization, and so point to leads that
record an upright deflection. This makes it easy to remember that electrocardiographic vector arrows
point in the direction followed by a wave of depolarization as it passes through the heart.
Fig. 15-15: Cardiac dipole in the human body. A: The partially depolarized ventricles, which are
analogous to the strip of myocardial tissue shown in Figure 15-14B, can be depicted as a dipole
similar to that shown in Figure 15-12. The potential difference generated by this dipole can be
recorded between electrodes on the right (RA) and left (LA) arms. B: The dipole shown in A can be
represented by a vector arrow oriented along the dipole axis. By convention, the head of the arrow
points to the positive pole of the dipole, and so points in the direction of the propagated wave of
depolarization.
Qrs Vectors
Much of the electrical activity of the ventricles is not recorded on the ECG largely because many regions
of the ventricles are activated simultaneously during inscription of the QRS complex (see later). The
electrical activity that does appear at the body surface can be viewed as a sequence of electrical vectors,
each of which is generated at a given instant during depolarization. The vectors generated at any moment
are mean instantaneous QRS vectors, and the sum of all of the instantaneous vectors generated during
ventricular depolarization is the mean QRS vector.
In Figure 15-16, ventricular activation is divided arbitrarily into three phases: septal activation,
activation of the apex, and activation of the base, each of which generates a mean instantaneous QRS
vector. However, these are not discrete events, but are steps in a series of transitions that occur during
ventricular depolarization. Activation of the interventricular septum and anterior portion of the base of
the left ventricle, which begins normal ventricular activation, generates a small electrical vector that is
directed to the right (Fig. 15-16A). In leads facing the left ventricle, this septal vector gives rise to a
small downward deflection, called the “septal Q wave.” As ventricular activation continues, the wave of
depolarization is transmitted through the Purkinje network toward the apex; the much larger mass of the
left ventricle causes leftward electrical forces to predominate over those directed to the right, so that
the second arbitrarily defined vector is directed inferiorly and to the left (Fig. 15-16B). The last regions of
the ventricles to be depolarized are the bases of the left and right ventricles. Again, activation proceeds
from
P.417
endocardium to epicardium and left ventricular forces dominate, so that this vector is directed superiorly
and to the left (Fig. 15-16C).
Fig. 15-16: Spread of the wave of depolarization over the ventricles (shaded) and the resulting instantaneous
QRS vectors (within dotted rectangles). A: Septal activation. The initial portion of the QRS complex is
produced by depolarization of the interventricular septum, which begins at the left ventricular surface of the
septum to produce an initial QRS vector that is directed superiorly and to the right. B: Activation of the apex.
The mid-portions of the QRS complex are produced by depolarization of the apex of the heart, which begins
in the endocardial regions of the ventricles. Because left ventricular forces are dominant, the vector is
directed inferiorly and to the left. C: Activation of the base. The terminal portions of the QRS complex are
produced by depolarization of the base of the heart. The dominant left ventricular forces generate a vector
that is directed superiorly and to the left. D: Because the tails of the instantaneous QRS vector arrows shown
in A, B, and C all represent “zero” potential, they can be superimposed.
The Vector Loop and Mean QRS Vector

The three mean instantaneous QRS vectors, shown in Figure 15-16A, B, and C, share a common zero
potential so that their tails can be superimposed (Fig. 15-16D and Fig. 15-17A). If all of the other
instantaneous vector arrows generated during the QRS complex were to be included, their heads would
inscribe a “vector loop” that starts at zero when the QRS complex begins, is inscribed in a counter
clockwise direction, and returns to zero at the end of the QRS (Fig. 15-17B). All of the mean
instantaneous QRS vectors inscribed during ventricular activation can be added together to generate the
mean QRS vector, which represents the average electrical vector generated during depolarization of the
ventricles (Fig. 15-17C).
P.418
Fig. 15-17: QRS vectors, a vector loop, and the mean QRS vector. A: The instantaneous QRS vector arrows
shown in Figure 15-16D. B: The vector arrows in (A) represent only three of the many instantaneous QRS
vectors generated during ventricular depolarization. Connecting the heads of all of the additional
instantaneous vectors forms a loop that is inscribed in a counter clockwise direction (dotted line). C: The
mean QRS vector is the sum of all instantaneous vectors generated at all times and in all regions during
ventricular depolarization.
The information provided by measurements of the mean QRS vector is limited by the fact that the ECG
normally records less than 10% of the total electrical activity generated during ventricular depolarization.
This is due largely to mutual cancellation of vectors oriented in opposite directions, which occurs because
different regions of the heart are activated at the same time; for example, the left and right ventricles
are depolarized in opposite directions. Mutual cancellation also occurs because the Purkinje fibers that
run in the endocardial regions of the ventricles cause activation to begin within the ventricular wall (Fig.
15-18A). As long as the wave of depolarization remains within the ventricular mass, potential differences
are not recorded by the ECG; it is only after depolarization reaches the surface of the heart that a
potential difference can be recorded at the body surface (Fig. 15-18B).
Bipolar Limb Leads and the Einthoven Triangle

The first systematic approach to evaluating the potential differences recorded by the ECG was devised by
Einthoven, who utilized electrodes placed on the left arm, right arm, and left leg to construct three
bipolar limb leads that he called leads I, II, and III. (The electrode usually placed
P.419
on the right leg when an ECG is taken serves as a ground.) Lead I records the potential difference
between the right arm and left arm, lead II that between the right arm and left leg, and lead III the
potential difference between the left arm and left leg. Einthoven chose to define the potential at the
right arm electrode as “zero” in leads I and II, while the left arm electrode was chosen as “zero” in lead
III (Table 15-3); these assignments of the zero electrodes were based on his desire to obtain upright QRS
complexes in the normal bipolar leads. According to Einthoven's conventions, an upright deflection lead I
is recorded when the left arm is positive relative to the right arm; in lead II when the left leg is positive
to the right arm; and in lead III when the left leg is positive to the left arm. An upright QRS is therefore
recorded in lead I when the wave of depolarization that activates the ventricles is directed to the left,
and in leads II and III when the wave of depolarization is directed inferiorly.
Fig. 15-18: Mutual cancellation of the electrical activity generated during depolarization within
the ventricular wall. A: A wave of depolarization that begins at the end of a Purkinje fiber within
the wall of the ventricle cannot generate a potential difference that can be recorded by electrodes
outside the heart. B: Potential differences can be recorded at the body surface only after the
depolarized area has expanded to include at least one surface of the ventricular wall.
Table 15-3 Standard Limb Leads
Lead Potential Difference Between “Zero”
Bipolar (Einthoven) leads
I Right arm and left arm Right

arm
II Right arm and left leg Right

arm
III Left arm and left leg Left
arm
Unipolar limb leads (augmented)
aVR Right arm and a “V” lead made by connecting the left arm and left V
leg electrodes
aVL Left arm and a “V” lead made by connecting the right arm and left V
leg electrodes
aVF Left leg and a “V” lead made by connecting the left arm and right V
arm electrodes
Einthoven simplified the calculation of electrical vectors using his three bipolar leads by assuming that
the right arm, left arm, and left leg electrodes are at the corners of an equilateral triangle, called the
Einthoven triangle, and that the heart is at the center of this triangle (Fig. 15-19). The mean QRS vector
can be located by projecting the net potential recorded in each of the
P.420
bipolar leads during ventricular depolarization onto the sides of the Einthoven triangle (Fig. 15-20). This is
because bipolar leads record only that portion of the electrical vector that is parallel to the lead axis (see
Fig. 15-12).
Fig. 15-19: Einthoven triangle. The heart, viewed in the frontal plane, is assumed to lie in the
center of an equilateral triangle whose corners are electrodes on the left arm (LA), right arm (RA),
and left leg (LL).
Fig. 15-20: A normal mean QRS vector projected on the Einthoven triangle. The QRS deflections in
leads I, II, and III shown at the bottom of the figure are from the ECG in Figure 15-8. The tail of the
vector, which represents “zero” potential, is the open circle at the center of the triangle, which is
projected by the dotted lines to the midpoints of the three sides. The magnitudes of the three
mean QRS vectors projected to each side of the triangle are determined by subtracting negative
from positive deflections (numbers above the leads at bottom). Perpendiculars drawn from the
heads of the vectors projected along the sides of the triangle (dashed lines) intersect at the head
of the mean QRS vector (dark arrow within the triangle). This vector is approximately +70°. RA,
right arm; LA, left arm; LL, left leg.
The amplitudes of the vectors measured by leads I to III are calculated by subtracting negative from
positive deflections in the recordings shown along the bottom of Figure 15-20, after which the net
deflections are drawn as arrows along the sides of the Einthoven triangle. The largest vector is in lead II
(which has the largest QRS complex) because the angle of this bipolar lead most nearly parallels the axis
of the dipole generated during ventricular depolarization. In contrast, the smallest vector is in lead III
because this lead is most nearly perpendicular to the dipole. These observations illustrate an important
principle, that the smallest QRS complex or the QRS complex in which upward and downward deflections
are most nearly the same is recorded
P.421
in the lead whose axis is oriented most nearly at a right angle to the mean QRS axis; these QRS complexes
are often called transitional because they represent transitions between regions of electronegativity and
electropositivity.
Vectors calculated using the Einthoven triangle are approximations because the triangle defined by the
limb leads in the human body is not equilateral, the heart is not in its center, and the body is not a
homogeneous volume conductor. Other ways to analyze these electrical vectors have been proposed, but
these offer little advantage to the Einthoven triangle.
Unipolar Limb Leads

Three unipolar limb leads are recorded along with Einthoven's bipolar leads in the routine clinical ECG.
Each unipolar limb records the potential difference between an exploring electrode—placed on the right
arm, left arm, or left leg—and an “indifferent” electrode that is assumed to measure zero potential. The
indifferent electrode initially used in electrocardiography was Wilson's central terminal (called V), which
is constructed by connecting electrodes on the right arm (R), left arm (L), and left leg (F = foot) (Fig. 15-
21). The first unipolar limb leads to be used, called VR, VL, and VF, recorded the potential differences
between V and the right arm, left arm, and left leg, respectively. The central terminal was assumed to
measure zero potential, so that an upright deflection is inscribed when the exploring electrode is more
positive that the central terminal. In lead VF, for example, an upright deflection is recorded when the
electrical vector is directed inferiorly, that is, when the left leg electrode faces an approaching wave of
depolarization.
Unfortunately, the potentials recorded in leads VR, VL, and VF are very small; this is readily understood
because in each of these leads one electrode is connected to both sides of the recording device. Lead VF,
for example, records the potential difference between the electrode on the left leg and Wilson's central
terminal, which is also connected to the left leg. To increase the size of these recordings, Goldberger
“augmented” the unipolar limb leads by disconnecting the lead that is on both sides of the circuit from
the central terminal (Table 15-3); for example, in lead
P.422
aVF (augmented VF) the leg electrode is disconnected from the V lead, which ends up containing only the
leads on the two arms. A similar maneuver is used to construct leads aVR and aVL. Although
“augmentation” increases the size of the deflections, these leads are not unipolar; in aVF, for example,
the “central terminal” made by connecting the left and right arm electrodes does not record zero
potential throughout the cardiac cycle. This deviation from theory illustrates the empirical nature of the
interpretation of electrocardiographic contour.
Fig. 15-21: Wilson's central terminal (V) is constructed by connecting the three limb lead
electrodes. Resistances are placed between each electrode and the central terminal to overcome
effects of variable resistances between the electrodes and the skin. RA, right arm; LA, left arm;
LL, left leg.
Fig. 15-22: Cross-section of the chest showing the placement of electrodes used to record leads,
V1 to V6, in the ECG. Electrodes V1 and V2 face the right ventricle, leads V5 and V6 face the
interventricular septum, and leads V3 and V4 face the left ventricle. RV, right ventricle; RA, right
atrium; LV, left ventricle; LA, left atrium; septum, interventricular septum. (Modified from Katz,
1946.)
Chest Leads
The six chest leads in the conventional 12-lead ECG (V1 to V6) are unipolar leads that record the potential
differences between Wilson's central terminal (V) and electrodes placed at six positions on the chest wall
(Fig. 15-22, Table 15-4). As is true for the unipolar limb leads, the
P.423
central terminal is assumed to record “zero” potential, so that an upright deflection is recorded when an
electrode on the chest wall is in an area of electropositivity, which occurs when a wave of depolarization
approaches the electrode.
Table 15-4 Standard Unipolar Chest Leads
Lead Potential Difference Between “Zero”
Fourth intercostal space just to the right of the sternum and the “V” lead V
V1
V2 Fourth intercostal space just to the left of the sternum and the “V” lead V
V3 Midway between V2 and V4 electrodes and the “V” lead V
V4 Fifth intercostal space at the midclavicular line and the “V” lead V
V5 Left anterior axillary line horizontally to the left of V4 and the “V” lead V
V6 Mid-axillary line horizontally to the left of V5 and the “V” lead V
For all leads, the “V” electrode is made by connecting the right arm, left arm, and left leg
electrodes.
The QRS complexes in leads V5 and V6 of the normal ECG (see Fig. 15-8), where the exploring electrodes
are placed over the left side of the chest (Fig. 15-22), are upright because they are determined almost
entirely by left ventricular depolarization, which generates a vector that is directed toward the left.
However, the QRS complexes in leads V1 and V2, where the exploring electrodes are placed over the right
ventricle on the anterior chest wall (Fig. 15-22), are inverted in the normal ECG (see Fig. 15-8). This
finding might seem to be unexpected because the right ventricle generates a wave of depolarization that
is directed anteriorly, and so should inscribe and upright deflection in leads V1 and V2. However, the left
ventricle is much thicker than the right, so that the potential differences generated by left ventricular
depolarization normally dominate the ECG, and often obscure those generated by depolarization of the
right ventricle.
Normal Vectors
All six of the limb leads in the standard ECG are assigned an angle, as shown in Table 15-5. These allow
the mean QRS vector in the frontal plane to be calculated using a convention that assigns an angle of 0°
to a vector directed toward the left arm, +90° to a vector directed inferiorly (toward the foot), -90° to a
vector directed superiorly, and ±180° to a vector directed toward the right arm.
The normal mean QRS axis lies between -30° and +110°; mean QRS vectors greater than +110° represent
right axis deviation; those less than -30° are designated as left axis deviation (Fig. 15-23). Some recent
textbooks have narrowed the range of normal QRS axis to between 0° and +90°; this simplification makes
it easier to estimate “normal” axis, but increases the number of individuals whose ECGs are stated to be
abnormal. It is important to note, however, that individuals with normal hearts can have an abnormal QRS
axis, and the QRS axis is often normal in patients with heart disease.
The chest leads define the position of the cardiac dipole in the horizontal plane of the body, but mean
QRS vectors are not calculated in this plane. Instead, attention is paid to the lead where upward and
downward QRS deflections are most nearly equal, which because it marks
P.424
the transition from inverted to upright complexes is called a “transitional QRS.” The transition from the
inverted QRS complexes normally seen in leads V1 and V2 to the upright QRS complexes normally seen in
leads V5 and V6 generally occurs in lead V3 or V4 (see Fig. 15-8).
Table 15-5 Frontal Plane Angles Conventionally Assigned to the Limb Leads
Lead Vector Angle
I 0°
II +60°
III +120°
aVR -150°
aVL -30°
aVF +90°
Fig. 15-23: Angles assigned to electrical vectors in the frontal plane. A vector directed to the left
is assigned an angle of 0°; one directed to the right is said to have an angle of 180°. A vector
directed inferiorly has an angle of +90°; one directed superiorly has an angle of -90°. Normal mean
QRS vectors range between -30° and +110°. Vectors with angles less than -30° are diagnostic of left
axis deviation; angles greater than +110° indicate right axis deviation.
T wave vectors, which can be calculated according to the same conventions used to estimate QRS vectors,
are sometimes used to estimate the QRS-T angle (the angle between the mean QRS and T wave vectors),
which is sometimes called the ventricular gradient. The QRS-T angle can help determine whether a T
wave is normal or abnormal because T wave vectors are usually concordant with the mean QRS vector
(see above). For this reason, a wide QRS-T angle implies that repolarization is abnormal.
Some Abnormal Electrocardiograms

A few abnormal ECGs are included at this point to illustrate clinical applications of some of the basic
principles presented in this and earlier chapters, and to demonstrate how electrocardiographic
abnormalities can be understood in terms of altered physiology.
Bundle Branch Block

The normal synchrony of right and left ventricular depolarization is lost when conduction in one of the
bundle branches is blocked. Under these conditions, impulses transmitted from the atria still depolarize
both ventricles, but activation of the “blocked” ventricle is delayed. The result is an ECG abnormality
called bundle branch block, in which the QRS complexes are prolonged and abnormal in contour.
Activation of the blocked ventricle is delayed because of the longer path followed by the wave of
depolarization, which must cross the interventricular septum to reach the blocked ventricle.
Furthermore, the delayed wave of depolarization is transmitted through ventricular myocardium, which
conducts more slowly than the His-Purkinje system (Table 15-1). In addition to prolonging the QRS
complex, the abnormal pathway of ventricular activation causes abnormalities in the QRS complex that
differ when the right and left bundle branches are blocked. The T waves are also abnormal in bundle
branch block because the abnormal spread of the wave of depolarization alters the pattern of
repolarization; for this reason, these are often called “secondary” T wave abnormalities. “Primary” T
wave abnormalities, on the other hand, result from local abnormalities in repolarization.
P.425
Fig. 15-24: Right bundle branch block. The diagnosis is established by abnormal prolongation of the
QRS complex in the limb leads and the tall, broad late R wave over the right ventricle (recorded in
lead V1). The broad late S waves in leads V5 and V6 arise from the same late QRS vector, which is
directed away from the left ventricle.
Right Bundle Branch Block
Delayed activation of the right ventricle in right bundle branch block widens the QRS complex and
generates a late wave of depolarization that is directed toward the blocked right ventricle. This is seen in
Figure 15-24, where the duration of the QRS complex (measured in lead II) is 0.14 s, which is significantly
prolonged (see Table 15-2). The late portion of the wave of depolarization that activates the blocked right
ventricle is propagated to the right and anteriorly because the right ventricle lies to the right and anterior
to the left ventricle (Fig. 15-22). The delayed activation of the right ventricle is responsible for the broad
late R' wave in lead V1, which is recorded from an electrode placed on the right side of the chest. Leads
V5 and V6 contain broad S waves because the wave of depolarization responsible for delayed activation of
the right ventricle is directed away from the left side of the chest. The reason that the late electrical
forces generated by right ventricular depolarization are not overwhelmed by left ventricular
depolarization in right bundle branch block is that the delayed conduction through the right ventricle
continues after depolarization of the left ventricle has been completed.
Left Bundle Branch Block

Interruption of the left bundle branch delays activation of the left ventricle so that, as in right bundle
branch block, the QRS complex is widened. In Figure 15-25, the QRS duration in the limb leads is 0.14 s
and the late QRS vector in the precordial leads is directed to the left. This accounts for the late upright
deflection of the notched QRS complex in lead V6, which faces the left ventricle (Fig. 15-22). The
downward deflection of the latter parts of the broad QRS complex in lead V1 also results from delayed
depolarization of the left ventricle, which causes the V1 electrode to lie in an area of electronegativity.
P.426
Fig. 15-25: Left bundle branch block. The diagnosis of left bundle branch block is established by
the prolongation of the QRS complex in the limb leads and the late R wave in lead V6.
Fascicular Blocks (Hemiblocks)

Interruption of one of the fascicles of the left bundle branch does not prolong the QRS complex because
the left ventricle distal to the blocked fascicle can still be activated by the rapidly conducting Purkinje
network. Instead, fascicular blocks are characterized by abnormalities in the mean QRS vector.
Left Anterior Fascicular Block

The late electrical forces in left anterior fascicular block are deviated superiorly, and to the left, toward
the region of the left ventricle whose activation is delayed. The electrocardiographic manifestation of
this conduction delay is a mean QRS axis that is less than -30° (left axis deviation) without prolongation of
the QRS complex; however, some authorities require a mean QRS axis of -45° or less, so that the criteria
for this diagnosis are controversial.
A diagnosis of left anterior fascicular block is suggested in Figure 15-26 by the finding of left axis
deviation; this is because lead II is the transitional lead, where the small positive and negative deflections
are about the same. Lead II is assigned an angle of +60° (Table 15-5), so that if the two deflections were
exactly equal, the mean QRS vector is at right angles to +60°: either -30° (+60° minus 90°) or +150° (+60°
plus 90°). The upright QRS is in lead I and inverted QRS in lead aVF indicates that the mean QRS vector
points toward the left arm and away from the foot, so that the QRS axis must be about -30°. On careful
examination, the downward QRS deflection in lead II is seen to be greater than the upward deflection,
which means that the QRS axis is less than -30°, which meets criteria for abnormal left axis deviation.
Left Posterior Fascicular Block

Left posterior fascicular block is characterized by abnormal right axis deviation because the delayed wave
of depolarization that actives the posterior regions of the left ventricle is directed inferiorly and toward
the right. In Figure 15-27, right axis deviation is suggested by the QRS
P.427
complexes in lead I, where the downward deflection is greater than the upward deflection; this means
that the mean QRS vector is directed toward the right arm. Because the QRS complexes in the inferior
leads (leads II, III, and aVF) are upright, the mean QRS axis must be directed inferiorly (Table 15-5). The
net upward deflection (positive minus negative) in lead III, which is assigned an angle of +120° (Table 15-
5) is ∼14 mm, while that in lead aVF, which is assigned an angle of +90°, is ∼12 mm; for this reason, the
mean QRS vector is more nearly parallel to lead III than to aVF. This tells us that the mean QRS vector is
closer to +120° than to +90°,
P.428
P.429
which places this vector at an angle greater than +105°, which meets the criteria for abnormal right axis
deviation (QRS ≥+110°).
Fig. 15-26: Left anterior fascicular block. Delayed conduction through the anterior fascicle of the
left bundle branch explains the left axis deviation (mean QRS axis less than -40°).
Fig. 15-27: Left posterior fascicular block. Delayed conduction through the posterior fascicle of
the left bundle branch explains the right axis deviation (mean QRS axis ≥ +110°).
Fig. 15-28: Right ventricular hypertrophy. The tall R wave in lead V1 is due to abnormally large
electrical forces directed toward the right ventricle. Unlike right bundle branch block (Fig. 15-24),
the QRS complex is not prolonged.
Ventricular Hypertrophy
Several criteria can be used to make an electrocardiographic diagnosis of ventricular hypertrophy, but all
have a relatively poor predictive accuracy. The examples provided below focus on the QRS abnormalities
and, for simplicity, ignore other important criteria based on P wave, ST
P.430
segment, and T wave configurations. Ventricular hypertrophy generally, but not always, increases the
voltage recorded by leads that “face” the hypertrophied ventricle; this explains tall R waves in lead V1 in
right ventricular hypertrophy (Fig. 15-28) and lead V6 in left ventricular hypertrophy (Fig. 15-29). The
increased height of the QRS complexes in these leads is due to the increased mass and slowed conduction
in the hypertrophied ventricle. Right and left ventricular hypertrophy is often associated with right and
left axis deviation, respectively, but changes in the mean QRS vector are not, by themselves, reliable
criteria for these electrocardiographic diagnoses.
Fig. 15-29: Left ventricular hypertrophy. The tall R waves in the left precordial leads (V5 and V6)
are caused by abnormally large electrical forces directed toward the left ventricle. Unlike left
bundle branch block (Fig. 15-25), the QRS complex is not prolonged.
Conclusion
The abnormal ECGs that conclude this chapter are provided to show how pathophysiological information
obtained from analysis of the QRS complex can be used for the diagnosis of heart disease, and illustrate a
few of the principles used in electrocardiographic interpretation. For more complete guidelines for the
electrocardiographic diagnosis of cardiac disease, the reader is referred to the many excellent textbooks
on clinical cardiology and electrocardiography.
Bibliography
A number of excellent teaching textbooks on electrocardiography have been published; each has it
own strengths and weaknesses. A specific text is not recommended because there are different
learning styles and tastes.
Anderson RH, Ho SY. Anatomy of the atrioventricular junctions with regard to ventricular
preexcitation. Pacing Clin Electrophysiol 1997;20:2071–2076.
Anderson RH, Ho SY. The architecture of the sinus node, the atrioventricular conduction axis, and
the internodal atrial myocardium. J Cardiovasc Electrophysiol 1998;9:1233–1248.
Dobrzynski H, Boyett MR, Anderson RH. New insights into pacemaker activity. Promoting
understanding of sick sinus syndrome. Circulation 2007;115:1921–1932.
Gima K, Rudy Y. Ionic current basis of electrocardiographic waveforms. A model study. Circ Res 2002;
90:889–896.
Ramanathan C, Jia P, Ghanem R, et al. Activation and repolarization of the normal human heart
under complete physiological conditions. Proc Nat Acad Sci (USA) 2006;103:6309–6314.
Riera ARP, Ferreira C, Filho CF, et al. The enigmatic sixth wave of the electrocardiogram: the U
wave. Cardiol J 2008;15:408–421.
See also Bibliographies to Chapters 13 and 14.
References
James TN. Cardiac innervation: anatomic and pharmacologic relations. Bull NY Acad Sci 1967;43:
1041–1086.
Katz LN. Electrocardiography. Philadelphia, PA: Lea and Febiger, 1946.
Wenckebach KF, Winterberg H. Die unregelmässige herstätigkeit. Leipzig: Wilhelm Engelmann, 1927.
> Table of Contents > Part Four - Pathophysiology > Chapter 16 - Arrhythmias
Chapter 16
Arrhythmias
Cardiac arrhythmias were first noted in antiquity, when physicians began to palpate the arterial pulse.
Hippocrates noted that a slow pulse in elderly men heralded sudden death, and Galen predicted the
impending death of an apparently healthy physician who had observed that his pulse was irregular. These
and similar observations were systematized in the late 19th century, when smoked drum kymograph
recordings of arterial and venous pulsations were used to define simple arrhythmias. Knowledge of clinical
arrhythmias advanced rapidly after Einthoven's invention of the string galvanometer at the beginning of
the 20th century, which made it possible to record abnormalities of cardiac rate and rhythm in
experimental animals and humans. By the middle of the 20th century, electrocardiographic analyses
characterized many of the pathophysiological mechanisms that operate in patients with arrhythmias (Katz
and Pick, 1956). Intracardiac recordings, which became available in the 1960s, along with microelectrode
recordings of intracellular potentials and patch clamp measurements of the opening and closing of single
ion channels, greatly advanced our understanding of arrhythmogenic mechanisms. Molecular biology,
which is providing the molecular structures of the voltage-gated ion channels and the cellular components
that regulate their behavior (see Chapters 13 to 15), is now adding a new dimension to our understanding
of arrhythmogenic mechanisms.
A Simple Classification of Arrhythmias

Normal heart rate is defined as 60 to 100 beats/min (for simplicity, only the rates, e.g., “60 to 100,” are
stated in the remainder of this text); this range was chosen because these limits correspond to five and
three large (0.2 s) “boxes” in the standard ECG. Arrhythmias, which can occur when the heart beats too
slowly (bradycardias), too rapidly (tachycardias), and/or irregularly (Table 16-1), are generally classified
on the basis of the structure where the arrhythmia is believed to have originated and the mechanism that
is presumed to have disturbed the rhythm (Table 16-2). Tachycardias, which by definition occur when
heart rate is sustained at rates greater than 100, are generally subdivided into supraventricular and
ventricular tachycardias. The former arise above the bifurcation of the AV bundle, and so are conducted
normally through the two bundle branches into the ventricles, while ventricular tachycardias distort and
prolong the QRS because the ventricles are not depolarized simultaneously.
It might seem logical that the causes of tachycardias differ fundamentally from the causes of
bradycardias, and that mechanisms that cause the heart to beat too rapidly are opposite those that slow
the heart. However, many mechanisms that are responsible for bradyarrhythmias also cause
tachyarrhythmias, so that as is often true, what seems obvious is wrong!
P.432
Table 16-1 A Simple Classification of Clinical Arrhythmias
I. Bradycardias
A. Sinus (SA) node
1. Sinus bradycardia
2. Sinoatrial block
B. Atrioventricular (AV) node and bundle: atrioventricular block:
1. First-degree AV block
2. Second-degree AV block
a. Mobitz I (the Wenckebach phenomenon), usually block in the AV node
b. Mobitz II, usually block in the AV bundle, both bundle branches, or all three
fascicles
3. Third-degree AV block
II. Tachycardias
A. Premature systoles
1. Atrial
2. Junctional
3. Ventricular
B. Tachycardias
1. Supraventricular
a. Sinus
b. Atrial
c. Junctional
2. Ventricular
C. Flutter and fibrillation
1. Atrial
2. Ventricular
Mechanisms Responsible for Bradyarrhythmias

The usual causes of a slow pulse are reduced pacemaker activity (chronotropy) and depressed conduction
(dromotropy). The former is caused by changes in the ionic currents responsible for SA node
depolarization (see Chapter 14), while the latter, generally called block, results from impaired impulse
conduction from the SA node to the ventricles.
Slowed Pacemaker Activity

The most common cause of slowing of the SA node pacemaker, called sinus bradycardia, is excessive
parasympathetic (vagal) tone. Although a heart rate below 60 is defined as abnormal, sinus bradycardia is
often seen in normal individuals, especially when training has increased
P.433
vagal tone. Non-cardiac disease, like hypothyroidism, can also slow the sinus pacemaker, as can drugs like
β-adrenergic blockers and some L-type calcium channel blockers. Sinus bradycardia can also be caused by
a condition, commonly seen in the elderly, that has the sibilant name sick sinus syndrome; because this
syndrome can be accompanied by supraventricular tachycardias, it is sometimes called the bradycardia-
tachycardia syndrome. When severe, sinus slowing can cause syncope, but rarely sudden death.
Table 16-2 A Simple Classification of Arrhythmogenic Mechanisms
I. Bradycardias
A. Slowed SA node pacemaker activity
B. Impaired impulse conduction (block)
1. SA Block
2. AV Block
a. In the AV node
b. In the AV bundle, both bundle branches, or all three fascicles
II. Tachycardias
A. Accelerated pacemaker activity
B. Reentry
1. Abnormal conduction (decremental conduction and unidirectional block)
2. Abnormal action potentials
a. Inhomogeneous resting potential
b. Inhomogeneous depolarization
c. Inhomogeneous repolarization and refractoriness
3. Structural abnormalities
a. Asymmetric tissue damage
b. Abnormal conducting structures

i. Dual AV nodal conduction pathways
ii. Accessory pathway (“bundle of Kent”)
C. Triggered depolarizations (afterdepolarizations)
1. Early afterdepolarizations
2. Delayed afterdepolarizations
Abnormal slowing of “lower” pacemakers in the AV node and His-Purkinje system (AV bundle, bundle
branches, fascicles) cannot cause a bradycardia as long as the ventricles are depolarized by impulses
conducted from a normally functioning SA node. Failure of lower pacemakers is therefore apparent only if
sinus rate is markedly slowed or conduction is blocked.
P.434
Table 16-3 Sites of Conduction Block
Site of Block Usual Cause Effect on the ECG
SA node Structural P waves delayed or absent

and
functional
AV node Functional Prolonged PR interval or P waves

not followed by QRS complexes
His-Purkinje system: (AV bundle, both Structural Prolonged PR interval or P waves

bundle branches; all three fascicles) not followed by QRS complexes
Impaired Conduction (Block)

Impaired impulse conduction from the SA node to the ventricles causes ventricular beats to be absent
(“dropped”). This abnormality, called block, can occur in the SA and AV nodes, where small action
potentials and a low safety factor makes conduction physiologically precarious (Chapter 14), and in the AV
bundle and bundle branches (Table 16-3). In the latter, in spite of the large action potentials, conduction
is anatomically precarious because these conducting structures can be damaged or destroyed by disease.
Slowing of Impulse Conduction
Conduction can be slowed by decreased action potential amplitude, reduced rate of depolarization, a
higher threshold, and increased internal resistance (Table 16-4). Action potential amplitude and rate of
depolarization are proportional to the open probability and number of active plasma membrane channels
that carry the inward currents responsible for the action
P.435
potential upstroke. Threshold reflects the ability of an approaching wave of depolarization to open these
channels, and internal resistance is determined largely by the number of open gap junction channels in
the intercalated discs. Changes in membrane capacitance, electrical resistance across the membrane
bilayer, and the conductivity of the extracellular space also influence conduction velocity, but are of
minor importance in modifying conduction velocity in the heart.
Table 16-4 Major Determinants of Conduction Velocity in the Heart
Determinant Structural and Functional Mechanism
Action potential Number of open sodium or calcium channels in the plasma

amplitude membrane
Rate of rise of the Rate of opening and number of open plasma membrane sodium
action potential or calcium channels
Threshold Amount of depolarization needed to open plasma membrane

sodium or calcium channels
Internal resistance Number of open gap junction (connexon) channels in the

intercalated disc, myocyte diameter
Conduction is accelerated by increasing the rate of opening of depolarizing channels because the more
rapid development of inward current shortens the time needed for an approaching action potential to
bring resting tissue to threshold. Similarly, increased action potential amplitude is able to depolarize
resting cells further ahead of an approaching wave of depolarization (Fig. 16-1A). Opening of gap junction
channels in the intercalated disc, which decreases internal resistance, speeds conduction by facilitating
the propagation of the currents generated by an approaching wave of depolarization (Fig. 16-1B), while a
lower threshold increases the ability of an approaching wave of depolarization to initiate a propagated
action potential (Fig. 16-1C).
The four determinants of conduction velocity listed in Table 16-4 can be likened to a row of falling
dominoes (Fig. 16-2), where increasing the height of the dominoes, like increasing action potential
amplitude, accelerates “impulse” transmission by allowing each falling domino to topple dominoes further
ahead in the row. Increasing the velocity at which each domino falls (as would occur if the dominoes were
moved to the surface of Jupiter), like increasing the rate of depolarization, accelerates transmission.
Placing the row of dominoes in a vacuum, like reducing longitudinal resistance, increases conduction
velocity by accelerating their fall, and if each domino was tipped slightly to reduce the time needed for it
to fall, the increased “tippiness,” like a decrease in threshold, would increase the velocity of
propagation.
These principles explain the different conduction velocities in various regions of the heart. Rapid
conduction in the working cells of the atria and ventricles and the cells in His-Purkinje system reflects
their high content of sodium channels that open rapidly and carry large depolarizing currents, and their
large diameter and high number of connexin channels that reduce internal resistance. Conversely, the
slower conduction in the SA and AV nodes is due to the slower opening, lower conductance, and smaller
number of the calcium channels that depolarize these cells, along with a high internal resistance caused
by their small size and few gap junctions. In diseased hearts, changes in many of these variables play a
major role in producing the arrhythmias that often kill patients with heart disease (see Chapters 17 and
18).
Decremental Conduction and Block

Decremental conduction refers to a progressive decrease in conduction velocity when an impulse enters a
region of slow conduction (Fig. 16-3). This can occur if conduction is slowed by a decrease in action
potential amplitude and rate of rise, as shown in Figure 16-3, and by an increase in threshold or internal
resistance. If the wave of depolarization is able to pass through the area of slow conduction, it can
emerge as a normally sized action potential (Fig. 16-3A) so that conduction will be delayed but not
completely blocked. However, conduction will be interrupted if the ability to generate a propagated
action potential is so severely depressed that the
P.436
P.437
P.438
action potential cannot serve as an effective stimulus to the excitable tissue ahead (Fig. 16-3B).
Decremental conduction is not always abnormal. For example, decremental conduction is a normal
property of the AV node where cells have a small diameter, few gap junctions, and slowly rising calcium-
dependent action potentials.
Fig. 16-1: Determinants of conduction velocity in cardiac muscle. Conduction is from left to right;
depolarized tissue is shaded and resting tissue unshaded. Conduction velocity is accelerated when the
magnitude and rate of rise of the inward currents carried by plasma membrane ion channels is increased (A),
when internal resistance is reduced by opening of connexin channels in the intercalated discs (B), and when
threshold is reduced (shown as a thinner plasma membrane in C).
Fig. 16-2: Propagation of an action potential viewed as a row of dominoes falling from left to right.
Conduction is accelerated when the height of the dominoes is increased and/or each domino falls
more rapidly, when less resistance is encountered by the falling dominoes, and when less inertia is
needed to tip each domino.
Fig. 16-3: Decremental conduction caused when an action potential enters a region where the
amplitude and rate of rise of the action potential are decreased (shaded). A: Conduction of an
action potential is slowed when it becomes smaller and more slowly rising in a region of
decremental conduction. After the impulse has crossed the region of decremental conduction and
has emerged in a region where the amplitude and rate of rise become normal, conduction velocity
returns to normal. B: When an impulse enters a region where the action potential no longer
reaches threshold, and so cannot initiate a propagated action potential, conduction is blocked.
Unidirectional Block
The term unidirectional block means what it implies: block of conduction, but in only one direction.
Although this behavior might seem intuitively to be unusual, it is not. Unidirectional block is common, and
it is a normal property of the AV node where antegrade (forward) conduction from atria to ventricles is
usually more rapid than retrograde (backward) conduction from ventricles to atria. Because unidirectional
block can disorganize impulse propagation and cause conduction to fail, it plays a major role in the
genesis of arrhythmias (see below).
The classical model for unidirectional block is a strand of myocardial tissue that is compressed by a
wedge-shaped wooden block, which causes an asymmetrical impairment of conduction. Depending on how
much pressure is applied to the wedge, the block can be unidirectional (Fig. 16-4B) or complete (Fig. 16-
4C). The mechanism responsible for unidirectional block is shown in Figure 16-5, where the ability of
different regions of a 10 cm long strand of myocardium to initiate a regenerative response is plotted on
the ordinate. Normal action potentials in the uncompressed regions, from 0 to 2 cm and from 8 to 10 cm,
can activate resting tissue up to 2 cm ahead, while between 2 and 3 cm, under the point of the wedge,
the cells are so severely damaged as to have completely lost the ability to initiate a propagated action
potential. Action potentials in the less depressed region, between 3 and 8 cm, can activate resting tissue
but, for only short distances ahead, depending on the degree of compression. The antegrade direction in
this figure is defined as from left to right, and the retrograde direction from right to left.
An action potential initiated by a stimulus on the left side, at 0 cm, is conducted in the antegrade
direction (Fig. 16-5B). Normal impulses in the uncompressed tissue, between 0 and 2 cm, generate a wave
of depolarization that can cross the severely depressed area between 2 and 3 cm.
P.439
Even though propagated impulses cannot be generated in the severely depressed area, electrotonic
spread of current from the normal region can generate a regenerative response 2 cm ahead (arrow a),
which initiates an action potential at 4 cm. Although the tissue at 4 cm is depressed, it is able to
depolarize tissue a short distance (0.4 cm) ahead, at 4.4 cm (arrow b); here the action potentials are less
depressed and so conduct 0.6 cm to the right, reaching the tissue at 5.0 cm (arrow c). At 5.0 cm the
ability to conduct is less depressed, so that the impulse continues to propagate to the right, gaining speed
as it emerges from the depressed area until
P.440
it reaches uncompressed tissue at 8 cm. Because conduction velocity is directly proportional to the
lengths of the arrows in Figure 16-5B, the antegrade impulse reaches normal tissue to the right after a
delay.
Fig. 16-4: Unidirectional block produced by compression of a strand of cardiac muscle by a wedge-shaped
wooden block (darker shading indicates a greater degree of injury). A: Bidirectional conduction in the
normal, uncompressed tissue. B: Moderate compression causes unidirectional block in which conduction can
proceed from left to right (defined here as the antegrade direction) but not from right to left (retrograde
block). C: Severe compression blocks conduction in both directions.
Fig. 16-5: Unidirectional block. A: Distribution of the ability of the moderately compressed strand of cardiac
muscle shown in Figure 16-4B to generate a propagated action potential. Normal tissue (at 0 to 2 and 8 to 10
cm) can activate resting muscle up to 2 cm ahead. The ability of the tissue at 2 to 3 cm to initiate a
propagated action potential is completely lost, and is depressed from 3 to 8 cm. B: Antegrade conduction. An
impulse entering from the left is able to cross the most severely depressed region, where the ability to
initiate a propagated action potential has been lost, although antegrade conduction (left to right) is delayed.
C: Retrograde block. An impulse entering from the right encounters increasingly impaired conduction, so that
the impulse becomes progressively less able to initiate an action potential. As the impulse approaches the
most severely depressed tissue (from 2 to 3 cm), the responses become so small that they cannot excite the
resting tissue ahead; as a result, retrograde conduction (right to left) is blocked. The dotted arrows in B and
C depict the distance ahead that action potentials generated at the tail of the arrow can be propagated.
When an action potential is initiated by stimulating the tissue on the right side, at 10 cm, rather than at 0
cm, retrograde conduction is quite different (Fig. 16-5C). At 8 cm, although the impulse encounters
damaged tissue and conduction begins to slow, it can still be conducted to the left, and so depolarizes the
tissue 2 cm ahead, at 6 cm, where action potentials can propagate only 1.2 cm (arrow a'). The next area
to be depolarized is therefore at 4.8 cm (the head of arrow b'), where the action potential can conduct
only 0.8 cm to reach the tissue at 4 cm (arrow c'). Here, the action potential conducts an even shorter
distance (arrow d'). Propagation continues to slow, as evidenced by the shorter arrows, until the impulse
reaches the severely depressed area at 3 cm where, because propagated action potentials can no longer
be generated, retrograde conduction ceases.
Areas of unidirectional block, such as that shown in Figure 16-5, are common in diseased hearts, where
non-uniform depression of conduction velocity can be caused by asymmetrical decreases in action
potential amplitude, rate of depolarization, number of open gap junction channels, and excitability.
Tachyarrhythmias
Extra beats and tachycardias are generally described in terms of their clinical manifestations (Table 16-1).
A single early beat is a premature systole and a series of premature systoles is a tachycardia. Very rapid
regular depolarization of the atria or ventricles is called flutter, while rapid irregular depolarization,
where there is no effective mechanical response, is fibrillation. Tachyarrhythmias can result from
accelerated pacemaker activity, reentry, and triggered depolarizations (Table 16-2).
Accelerated Pacemaker Activity

Accelerated firing of pacemaker cells in the SA node causes sinus tachycardia, which is often a normal
response to increased sympathetic tone or other extracardiac factors. Accelerated pacemaker discharge
in “lower” pacemakers causes junctional tachycardias and accelerated idioventricular rhythms.
Reentry
Reentrant arrhythmias, which occur when a single impulse gives rise to two or more propagated
responses, can be caused by both pathophysiological and anatomical mechanisms (Table 16-2). The former
include conduction abnormalities, and heterogeneities in resting potential, action potential amplitude,
and refractoriness. Anatomical causes of reentry can result from asymmetric tissue damage and abnormal
conducting structures.
Slow Conduction (Decremental Conduction and Unidirectional Block)

The mechanism by which slow conduction causes a reentrant premature systole or tachycardia can be
understood by examining the junction where an impulse conducted by a Purkinje fiber activates the
ventricular myocardium (Fig. 16-6). When a wave of depolarization (labeled 1 in Fig. 16-6) reaches the
proximal end of an area where unidirectional block prevents antegrade
P.441
conduction (shaded area in Fig. 16-6), the impulse cannot cross the depressed region. However, if
propagation through normal regions of the heart brings this impulse to the distal end of the area of
unidirectional block after a delay that is long enough to allow the depressed tissue to recover its
excitability, the wave of depolarization can be conducted through the depressed area in the retrograde
direction. This can allow the impulse to depolarize (reenter) the proximal region and generate a second
propagated wave of depolarization (labeled 2 in Fig. 16-6). Conduction of the retrograde impulse often
increases the refractoriness of the depressed area so as to block conduction in both directions, which
terminates the reentry. However, repeated conduction through a reentrant circuit can generate a
sustained tachycardia (see below).
Fig. 16-6: Establishment of a reentrant arrhythmia where a Purkinje fiber (left) impinges on the
ventricular myocardium near a region where decremental conduction and unidirectional block
(shaded) prevents antegrade but not retrograde conduction. Although the initial impulse (dashed
line labeled 1) cannot traverse the depressed region in the antegrade direction, propagation
through normal tissue can allow the impulse to reach excitable tissue distal to the region of
decremental conduction and unidirectional block. If the impulse is conducted slowly in a
retrograde direction (dotted arrow), it can reenter the proximal myocardium region after it has
recovered from the initial antegrade impulse so as to initiate a second wave of depolarization
(dashed line labeled 2) that generates a premature systole. The ECG generated by these two beats
is shown schematically at the bottom of the figure.
Two phenomena, called summation and inhibition, are seen in areas of decremental conduction and
unidirectional block (Fig. 16-7). Summation occurs when two subthreshold impulses that by themselves
cannot initiate a response are added to generate a propagated action potential. Inhibition is seen when
an action potential that enters, but cannot cross an area of decremental conduction, blocks conduction of
subsequent impulses; this explains a phenomenon called concealed conduction (see below).
P.442
Fig. 16-7: Summation and inhibition in a branched strand of myocardium A. Summation: (a and b): Impulses 1
and 2 cannot, by themselves, generate a propagated action potential in a region of decremental conduction
(cross-hatched). (c) When impulses 1 and 2 arrive simultaneously from opposite directions, they can be
summated to initiate a propagated action potential in the branch. B: Inhibition: (a) When impulse 1 enters
one side of a region of decremental conduction and unidirectional block (cross-hatched), it can initiate a
propagated action potential. (B) When impulse 2 enters the region of decremental conduction from the other
direction, its conduction is blocked. (c) When impulse 2 (which would have been blocked) enters the region
of decremental conduction immediately before the arrival of impulse 1 (which would have been conducted),
impulse 2 prevents impulse 1 from initiating a propagated wave of depolarization.
Heterogeneities in Resting Potential

Localized areas of resting depolarization are a common substrate for reentrant arrhythmias, and cause
many of the sudden deaths that occur during the first few minutes and hours after an acute coronary
occlusion (see Chapter 17). Heterogeneous resting potential causes arrhythmias when electrotonic
currents, called injury currents, that flow between partially depolarized and normal regions are able to
activate resting cells (Fig. 16-8).
Heterogeneities in Action Potential Duration

Electrotonic current flow between regions of the heart that have repolarized at different rates allow
inhomogeneous action potential duration to initiate tachyarrhythmias (Fig. 16-9). These inhomogeneities
have many causes, including regional ischemia, which can shorten and lengthen refractoriness (see
Chapter 17), end-stage heart failure, ion channel mutations, stress-activated changes in ito, and regional
variations in delayed rectifier potassium channels.
Heterogeneities in Refractoriness
Disruption of the normal reactivation of depolarizing channels often occurs when abnormal refractoriness
causes delayed and inhomogeneous reactivation of sodium channels, which can decrease conduction
velocity and cause unidirectional block (see earlier).
P.443
Fig. 16-8: Inhomogeneous resting potential in which the potential at the surface of a group of partially
depolarized cells (center) is more negative than that of the nearby normal tissue. These differences in resting
potential can allow the action potentials in the depolarized cells to activate surrounding tissue (heavy dotted
arrow).
Abnormal Conducting Structures

Abnormal conducting structures can establish electrical “short circuits” that cause reentrant
tachycardias. Most common are dual AV nodal conduction pathways and accessory pathways, both of
which can initiate reciprocal rhythms in which impulses traveling back and forth between the atria and
ventricles cause supraventricular tachycardias.
Dual AV Nodal Conduction Pathways and AV Nodal Reentrant
Tachycardias
It is not rare for the AV node to contain two pathways that conduct impulses at different velocities; the
fast (more rapidly conducting) pathway, which usually lies in the lower right atrium, has a longer
refractory period than the slow pathway, which is usually in the compact AV node (Fig. 16-10). AV nodal
tachycardias are commonly triggered by a premature atrial
P.444
depolarization that reaches the fast pathway during its refractory period, but after the slow pathway has
recovered from the preceding impulse; when this occurs, a wave of depolarization can be propagated to
the AV bundle through the slow pathway after which it returns to the atria by the fast pathway (Fig. 16-
10B). This can initiate a paroxysmal supraventricular tachycardia in which impulses go back and forth
between the two pathways as a reciprocal rhythm.
Fig. 16-9: Inhomogeneous repolarization. The early return to resting potential in a group of cells
with abnormally short action potentials (s) can allow adjacent normal cells (n) to depolarize the
cells with the shorter action potential (heavy dotted arrow). A similar mechanism allows cells with
abnormally long action potentials to reactivate nearby normal tissue.
Fig. 16-10: Dual AV nodal conduction pathways. A. Anatomy. A fast conducting pathway lies in the
lower right atrium, and a more slowly conducting pathway is in the atrioventricular (AV) node. The
central fibrous body is depicted by the heavily shaded rectangle. B: Premature activation of the
atria can initiate an impulse that travels to the ventricle via the slow pathway, which has the
shorter refractory period, and then returns to the atria after reentering the fast pathway.
Supraventricular tachycardias are usually caused when an atrial premature depolarization
(downward dotted arrow) enters the slow pathway, but not the fast pathway (short horizontal line)
because the latter has a longer refractory period. After reaching the ventricular end of the slow
pathway, the premature impulse can be conducted back to the fast pathway (upward dashed
arrow), and then reenter the slow pathway (downward dashed arrow). If this reciprocal rhythm
continues, the reentrant circuit involving the two pathways can sustain a tachycardia.
Accessory Pathways and Preexcitation (the Wolff–Parkinson–White

Syndrome)
An accessory pathway (bundle of Kent), which is generally a strand of atrial myocardium that can be
located virtually anywhere along the AV junction, can establish an electrical short circuit between the
atria and ventricles (Fig. 16-11A). This pathophysiology explains the clinical features of preexcitation (the
Wolff–Parkinson–White syndrome or WPW) in which conduction from the atria through the accessory
pathway causes a short PR interval (<0.12 s) and distorts
P.445
P.446
the initial deflection of the QRS (called a delta wave) (Fig. 16-12). The PR interval is short because
conduction through the accessory pathway is mediated by myocytes that are depolarized by fast sodium
channels, and so activate the ventricles before the arrival of the impulse conducted through the AV node,
where depolarization depends on the slower opening of calcium channels. The QRS complex begins
abnormally (the delta wave) because impulses conducted down accessory pathways do not enter the
ventricles normally via the His-Purkinje system; instead, ventricular depolarization begins at an abnormal
site, which distorts the initial deflection of the QRS complex.
Fig. 16-11: Preexcitation. A: Anatomy. An accessory pathway (bundle of Kent), which conducts
more rapidly than the pathway that includes the atrioventricular (AV) node, is usually a strand of
atrial tissue that crosses the fibrous skeleton of the heart to provide an abnormal conduction
pathway linking the atria and ventricles. The ventricles are activated by atrial impulses that are
conducted through both pathways, but because conduction is more rapid in the accessory pathway,
ventricular activation begins in an abnormal location (cross-hatched circle). The impulse conducted
by the normal pathway encounters the normal delay in the AV node before it can enter the
ventricles via the AV bundle (smaller shaded circle). The central fibrous body is depicted by the
heavily shaded rectangle. B: Supraventricular tachycardias are usually caused by a reciprocal
rhythm in which impulses travel from the atria to the ventricles via the slower AV nodal pathway,
and from the ventricles to the atria via the accessory pathway. C: The reciprocal rhythm in B is
initiated when an atrial premature depolarization (dotted arrow) enters the slower AV nodal
pathway but not the accessory pathway because the latter has a longer refractory period. After the
premature impulse reaches the ventricles (downward dashed arrow), it can return to the atria in
the retrograde direction through the faster accessory pathway (upward dashed arrow), and then
propagated by antegrade conduction back to the ventricles through the slower AV nodal pathway
(downward dashed arrow). In this way, impulses can be conducted back and forth through the two
pathways to establish a reciprocal rhythm between the atria and ventricles that causes a
supraventricular tachycardia.
Fig. 16-12: Preexcitation (Wolff–Parkinson–White syndrome) recorded in Einthoven bipolar lead I.

The PR interval is abnormally short (0.08 s) because of rapid conduction through an accessory
pathway, while the slurred upstroke of the QRS complex (the delta wave) is caused when this
impulse initiates ventricular depolarization at an abnormal site.
Accessory pathways provide the substrate for supraventricular tachycardias in which impulses usually
travel from the atria to the ventricles via the AV nodal pathway, and from the ventricles to the atria via
the accessory pathway (Fig. 16-11B). These tachycardias begin when an atrial premature depolarization is
transmitted to the ventricles only by the slower AV nodal pathway because the accessory pathway has a
longer refractory period (Fig. 16-11C). After the impulse reaches the ventricles, it returns to the atria via
the accessory pathway to cause a supraventricular tachycardia.
Preexcitation can occur in patients with hypertrophic cardiomyopathies caused by abnormalities in

glycogen metabolism (see Chapter 18). The accessory pathways in these patients consist of myocardial
strands that traverse the central fibrous body in gaps formed by glycogen accumulation.
Circus Movements
In the early 20th century, Mines used a jellyfish mantle to model reentry in a ring of excitable tissue. By
gently cross-clamping the ring before stimulating the ring next to the clamp, he initiated a wave of
depolarization that propagated in only one direction (clockwise in Fig. 16-13)
P.447
because conduction in the other direction was blocked by the clamp. If the clamp was removed before
the impulse returned to the site of block, the wave of depolarization continued to circulate around the
ring as a “circus movement” that lasted until tissue deterioration extinguished the ability to conduct (Fig.
16-14).
Fig. 16-13: Initiation of a circus movement. When a ring of excitable tissue is lightly clamped and
stimulated alongside the clamp, counterclockwise conduction is blocked so that the wave of
depolarization can propagate in only the clockwise direction. If the clamp is removed before the
wave of depolarization returns to the area of block, the impulse can continue to circle the ring in a
clockwise direction.
Fig. 16-14: Initiation of a circus movement in a ring of excitable tissue containing an area of decremental
conduction and unidirectional block created by clamping the tissue near the site of stimulation. Resting areas
are lightly shaded, and activated areas are darkly shaded. Upper row: Stimulation in the absence of block (A)
depolarizes the ring in both directions (B and C) which leads to mutual cancellation of the two impulses (D)
and failure to establish a circus movement (E and F). Lower row: When a clamp blocks counterclockwise
propagation (G) (see Fig. 16-13), the stimulus initiates a wave of depolarization that can propagate only in
the clockwise direction (H, I and J). If the clamp is removed and the impulse allowed to cross the area that
had been blocked, the wave of depolarization continues to circle the ring in a clockwise direction (K and L).
This model was used to define three mechanisms that could interrupt a circus movement (Fig. 16-15). The
first is to slow conduction, for example, by inactivating the channels which carry the inward currents that
propagate the wave of depolarization (Fig. 16-15a). The second is to accelerate conduction so that the
wave of depolarization “catches up” with regions that are still refractory after the prior passage of the
wave front (Fig. 16-15b). The third mechanism is to prolong the refractory period, which can extinguish
the wave of depolarization when it reaches tissue that had not recovered from prior passage of the
impulse (Fig. 16-15c). A classical analogy is to view the impulse traveling around the circular pathway as a
snake which dies if its head catches up with its tail. According to this analogy, slowing conduction is like
paralyzing the snake, while speeding the advance of its head or slowing the retreat of its tail allows the
snake to give itself a fatal bite.
Overdrive Suppression
Most of the tachycardias listed in Table 16-2 (except for triggered activity, see below) can be terminated
by stimulating the heart at a rate more rapid than the tachycardia. Several mechanisms have been
proposed to explain this phenomenon, called overdrive suppression. Frequent depolarization increases
sodium entry which, by elevating cytosolic sodium concentration, can facilitate sodium channel
reactivation when the repolarizing current generated by the
P.448
sodium pump increases resting potential (see Chapter 8). Another explanation is that the more frequent
action potentials increase calcium influx (see Chapter 10) which, by increasing energy utilization, causes
energy starvation and intracellular acidosis, both of which slow conduction and so convert unidirectional
block to bidirectional block.
Fig. 16-15: Three ways to stop the circus movement described in Figure 16-14. Left: The circus movement
can be terminated if conduction is slowed (a), accelerated (B), or if the refractory period is prolonged (c). All
allow the advancing impulse to reach the previously depolarized tissue during its refractory period. Right:
These mechanisms can be viewed as three ways to stop a snake from traveling in a circle. The snake stops
circling if it is paralyzed (a), if its head advances fast enough to bite its tail (B), or if slowing the retreat of
its tail allows the latter to come within biting distance of the advancing head (c).
Triggered Depolarizations
Triggered depolarizations, which can result from early and late afterdepolarizations (see Chapter 14), are
an important cause of premature systoles and tachycardias that can arise in the atria, His-Purkinje
system, and ventricles. Because these arrhythmias are often caused by depolarizing currents generated by
calcium efflux from the cytosol via the Na/Ca exchanger (see Chapter 8), triggered depolarizations occur
in patients with coronary artery disease and heart failure in whom energy starvation causes cytosolic
calcium overload. Rapid electrical stimulation does not suppress calcium overload-induced
afterdepolarizations; instead, these arrhythmias are made worse because calcium flux into the cytosol is
increased by the greater frequency of calcium channel opening.
Clinical Arrhythmias
The following pages describe the mechanisms and electrocardiographic features of several clinical
arrhythmias, along with a few therapeutic principles. These examples are provided to illustrate some
clinical applications of the material covered above, and in Chapters 13 to 15.
P.449
Sinus Node Abnormalities

Cardiac rate and rhythm are normally determined by depolarization of the SA (sinus) node pacemaker. By
convention, a rate faster than 100 that originates in the SA node is a sinus tachycardia, while slowing of
the SA node pacemaker to a rate less than 60 is a sinus bradycardia.
Sinus Tachycardia
The electrocardiographic characteristics of sinus tachycardia (Fig. 16-16) are normal QRS complexes
preceded by normal P waves and normal PR intervals that occur at a rate greater than 100. P waves are
normal because the tachycardia originates in the SA node and is conducted normally through the atria,
while the PR interval and QRS complexes are normal because the ventricles are depolarized normally via
the AV bundle and bundle braches. (When the heart is diseased or heart rate is extremely rapid,
atrioventricular and intraventricular conduction can be abnormal in sinus and other supraventricular
tachycardias.) Sinus tachycardias typically begin and end gradually, and are slowed by vagal stimulation;
these characteristics distinguish this common arrhythmia from most other supraventricular tachycardias,
which start and stop suddenly (see below).
Increased sympathetic activity and reduced vagal tone, which occur normally during physical exercise and
emotional stress, are the most common causes of sinus tachycardia (see Chapter 14); other causes include
fever, anemia, hyperthyroidism, and stretch of the SA node (the Bainbridge reflex). The maximum rate of
normal exercise-induced sinus tachycardia is sometimes stated to be 220 minus the patient's age, so that
a healthy 20-year-old should be able to achieve a sinus rate of 200, while maximum sinus rate in a 70-
year-old should be only 150. This relationship, however, is highly variable.
Sinus Bradycardia
Sinus bradycardia occurs when SA node pacemaker rate is less than 60. The ECG in sinus bradycardia
therefore shows a slow heart rate in which normal P waves are followed by normal QRS complexes after a
normal PR interval (Fig. 16-17). The most common causes of sinus bradycardia are increased
parasympathetic (vagal) activity and decreased sympathetic activity (see Chapter 14). Physical
conditioning increases the tonic parasympathetic tone that normally slows the sinus pacemaker, which
explains the resting sinus bradycardia characteristic of the “athlete's heart.” Severe sinus bradycardia can
cause vasovagal syncope (the “swoon” in 19th-century novels), and can be triggered by reflexes in
patients with acute inferior or posterior myocardial infarction (see Chapter 17). Drugs that cause sinus
bradycardia include β-adrenergic receptor blockers and some L-type calcium channel blockers. The
former inhibit pacemaker
P.450
activity indirectly when reduced synthesis of cyclic AMP decreases PKA-catalyzed phosphorylation of the
channels responsible for several pacemaker currents (see Chapter 14), while the latter act directly to
inhibit calcium channel opening in the SA node. Cardiac glycosides slow the SA node by a central effect
that increases parasympathetic activity.
Fig. 16-16: Sinus tachycardia (lead II). The rates of atrial and ventricular depolarization are approximately
120. Each QRS complex, which is normal, follows a normal P-wave after a normal PR interval of 0.12 s
(retouched).
Fig. 16-17: Sinus bradycardia (lead II). Normal QRS complexes follow normal P-waves after a normal PR
interval, but at a rate of 45.
Sinus Arrhythmia and Wandering Pacemaker
A perfectly regular sinus rhythm is not, as might be expected, a sign of a healthy heart; instead, small
variations in cycle length, called sinus arrhythmia, are characteristically seen in normal individuals. This
variability, which is due largely to physiological changes in sympathetic and parasympathetic tone, is lost
in patients with heart disease, in part because of sustained over-stimulation of the neurohumoral
response (see Chapter 8). For this reason, clocklike regularity of the heartbeat suggests that the normal
fine tuning by the autonomic nervous system has been blunted. Decreased heart rate variability is one of
the first signs of heart disease, so that sinus arrhythmia is generally a sign of good health!
Changes in sinus rate that accompany normal respiration, called respiratory sinus arrhythmia, alter cycle
length without affecting impulse propagation through the atria, AV node, and ventricles; for this reason,
the only changes are in the firing rate of the SA node (Fig. 16-18). Respiratory sinus
P.451
arrhythmia occurs when stimuli arising in stretch receptors in the lung and chest wall alter the balance
between sympathetic and parasympathetic influences on the SA node; heart rate accelerates during
inspiration because of decreased vagal tone and increased sympathetic tone, while the SA node
pacemaker slows when expiration increases vagal tone and decreases sympathetic tone.
Fig. 16-18: Respiratory sinus arrhythmia (lead II). Atrial and ventricular depolarization accelerate during
inspiration and slow during expiration, but each QRS complex is preceded by a normal P wave and a normal
PR interval. The arrhythmia is diagrammed below, using a “Lewis diagram”; the eponym recognizes Sir
Thomas Lewis, a pioneer in electrocardiography. The top line represents the sinoatrial (SA) node, the upper
space (A) the atria, the middle space (AV) the AV node, and the lower space (V) the ventricles. The
downward angle of the lines represents the speed of impulse propagation. Each beat in this arrhythmia begins
in the SA node and is conducted normally through the atria, AV node, and ventricles.
Fig. 16-19: Wandering pacemaker (lead II). The rate in this series of sinus beats is regular, with
cycle lengths between 0.69 and 0.72 s. The marked changes in P wave morphology can be
explained by shifts in the location of the pacemaker within the sinoatrial node.
Another common benign SA node arrhythmia is wandering pacemaker, which is characterized by changes
in P wave morphology and minor variations in cycle length (Fig. 16-19). This arrhythmia, which can be
related to respiration, is caused by shifts in the site of pacemaker activity within the SA node (see
Chapter 15).
Sinoatrial Block
Heart rate can be slowed when impulses arising in the SA node fail to depolarize the atria (Fig. 16-20).
This abnormality, called SA block, can be difficult to distinguish from sinus bradycardia because
depolarization of the SA node pacemaker cannot be recorded by the ECG (see Chapter 15). SA block can
be caused by increased vagal tone, drugs, and other factors that decrease depolarizing currents in the SA
node pacemaker (see above). The latter include occlusion of the SA node artery, apoptosis, fibrosis, and
degeneration of the nodal cells.
Sick Sinus Syndrome

Slowing of the SA node pacemaker and SA block characterize the sick sinus syndrome, a common condition
whose incidence increases with advancing age. Underlying mechanisms include an age-dependent
decrease in the amount of connexin 43 in the intercalated discs that lie between the cells of the SA node,
and mutations in iNa channels, ih channels, and ankyrin-B.
Fig. 16-20: Sinoatrial block (lead II). P waves are regular throughout the record except after the third QRS
complex, where a P wave is missing (“dropped”). The first P wave after the dropped P wave occurs when it
would have been expected had there been no interruption in the regular sinus P waves (dashed arrows),
which indicates that the sinus impulse preceding the missing P wave failed to depolarize the atria because its
exit from the sinoatrial node was blocked.
P.452
Abnormalities of Atrioventricular Conduction

Conduction abnormalities in the atrioventricular (AV) node, AV bundle, and bundle branches can cause
both bradyarrhythmias and tachyarrhythmias. The former, which result from impaired conduction from
the atria to the ventricles (AV block), are discussed below; supraventricular tachycardias caused by
reentry in the AV junction are described later in this chapter.
Even a minor reduction in the amplitude of the calcium-dependent action potentials in the AV node can
impair AV conduction. However, not all AV block occurs in the AV node. Diseases affecting the AV bundle,
bundle branches, and fascicles, where conduction depends on sodium currents, can also inhibit
conduction from the atria to the ventricles. These distinctions are important because the location of the
block has major prognostic implications. AV block caused by depressed conduction in the AV node and AV
block caused by impaired conduction in the His-Purkinje system occur in different patterns, have
different causes, and frequently require different treatment (see below). An intracardiac electrogram
(see Chapter 15) can usually identify the site of block because impaired conduction in the AV node
prolongs the AH interval whereas block in the more distal conduction system prolongs the HV interval (Fig.
16-21).
Fig. 16-21: Intracardiac electrograms in AV block. A: Normal. B: First-degree AV block with
prolonged AH interval and normal HV interval indicating block in the AV node. C: First-degree AV
block with normal AH interval and prolonged HV interval indicating block in the distal His-Purkinje
system. A, atrial depolarization; H, depolarization of the bundle of His; V, ventricular
depolarization.
P.453
The terms AV dissociation and AV block, although often used interchangeably, have different meanings. AV
dissociation describes any condition that impairs impulse transmission from the atria to the ventricles,
whereas AV block refers to AV dissociation that is caused by depressed conduction. For example, during a
supraventricular tachycardia at a rate of 200, the normally low safety factor of the AV node can be
expected to slow ventricular rate to 100, so that the abnormality is 2:1 AV dissociation (two atrial beats
for each ventricular beat). If, on the other hand, atrial rate is 100 and only half of the atrial impulses
depolarize the ventricles, the condition is 2:1 AV block because the normal AV node should be able to
transmit 100 impulses each minute. This distinction, while semantic, is useful in communicating the
significance of an AV conduction abnormality.
First-, Second-, and Third-Degree Atrioventricular Block

Three degrees of AV block are recognized clinically: first-degree AV block, the mildest, is an abnormal
delay in AV conduction that prolongs the PR interval; second-degree AV block is more severe because
some, but not all, atrial impulses fail to activate the ventricles, so that some P waves are not followed by
QRS complexes. In third-degree AV block, no impulses are conducted from the atria to the ventricles so
that no P waves initiate QRS complexes.
First-Degree Atrioventricular Block
First-degree AV block is readily identified by prolongation of the PR interval in an ECG where all P waves
are followed by QRS complexes (Fig. 16-22). Although the delay between atrial and ventricular
depolarization is increased, this abnormality has little effect on the pumping of the heart aside from
reducing the ability of the atria to serve as a “primer pump.” The major clinical significance of this
arrhythmia is that it can herald the development of more severe AV block.
Second-Degree Atrioventricular Block

In second-degree AV block, some but not all atrial impulses depolarize the ventricles so that the obvious
electrocardiographic feature is that some P waves are not followed by a QRS complex. In uncomplicated
second-degree AV block, all ventricular depolarizations are initiated by impulses propagated through the
AV node so that all QRS complexes follow a P wave, but because some atrial impulses fail to reach the
ventricles, there are more P waves than QRS complexes.
Second-degree AV block can be accompanied by a regular or an irregular pulse. If every other atrial
impulse reaches the ventricles and the sinus rate is regular, the pulse will be regular at half of the sinus
rate; this is seen in Figure 16-23, where the atrial rate is regular at 96 and the ventricular rate is 48. This
is called 2:1 AV block because there are twice as many P waves as QRS
P.454
complexes. Other ratios between P waves and QRS complexes are possible; in Figure 16-24, for example,
every fifth P wave fails to reach the ventricles, so that this is an example of 5:4 AV block.
Fig. 16-22: First-degree AV block (lead II). Each P wave is followed by a QRS complex, but the PR
interval is abnormally prolonged to 0.26 s.
Fig. 16-23: Second-degree AV block with 2:1 AV conduction (lead II). There are twice as many P waves (P) as
QRS complexes, so that every other atrial impulse is blocked.
Mobitz I and II Second-Degree Atrioventricular Block

Changes in the PR interval that precede the blocked P waves provide the basis for a classification of
second-degree AV block that helps distinguish block in the AV node from the more dangerous form of AV
block caused by impaired conduction in the His-Purkinje system. The diagnosis of Mobitz I second-degree
AV block is made when the blocked P waves are preceded by progressive prolongation of the PR interval in
a pattern called the Wenckebach phenomenon (Fig. 16-24). In the less common Mobitz II second-degree AV
block, the Wenckebach phenomenon is absent because progressive PR prolongation does not precede the
dropped beats (Fig. 16-25).
Mobitz I second-degree AV block is most often the result of physiological or pharmacological slowing of AV
node conduction; causes include excessive vagal activity and treatment with a β-adrenergic receptor
blocker or calcium channel blocker. For this reason, this form of second-degree AV block is often managed
medically when a physiological or pharmacological cause can be identified and, if appropriate,
eliminated. This strategy is generally safe because, when Mobitz I block worsens, progression is generally
gradual (e.g., 4:3 block progressing to 3:2, then to 2:1 block). Mobitz I AV block generally responds to a
muscarinic receptor blocker like atropine, which can relieve the inhibition of calcium channel opening
caused by parasympathetic effect on the AV node. Sympathetic stimulation, while also logical, can cause
dangerous side effects.
Fig. 16-24: Mobitz type I second-degree AV block demonstrating the Wenckebach phenomenon (lead II).
Beginning with the second QRS complex, which is preceded by a PR interval of 0.25 s, the PR intervals
increase progressively until a P wave is completely blocked; this is the typical Wenckebach phenomenon.
Because the PR prolongation occurs in decreasing increments, the intervals between successive QRS
complexes (RR intervals) shorten before the dropped beat; this causes a slight increase in ventricular rate,
called “group beating.” The pause that follows the completely blocked P wave allows the AV junction to
recover, which accounts for the shortened PR interval of the first cycle after the dropped beat.
P.455
Fig. 16-25: Mobitz type II second-degree AV block (lead V1). In this example of 3:2 AV block, the PR interval
before all of the conducted QRS complexes is 0.18 s and does not increase prior to the dropped beats. The
prolonged QRS complexes (0.14 s) are probably a manifestation of the same conduction system disease that
caused the AV block (retouched).
Mobitz II second-degree AV block is usually caused by structural lesions in the His-Purkinje system, so that
this form of AV block is unlikely to be alleviated by decreasing vagal tone or modifying drug therapy.
Because Mobitz II block implies that conduction in the AV bundle, which depends on sodium-dependent
action potentials, is “hanging by a thread,” this form of AV block can progress suddenly, often without
warning, to third degree (complete) AV block (Fig. 16-26). For this reason, Mobitz II second-degree AV
block is generally an indication for an electronic pacemaker.
The Wenckebach Phenomenon

Progressive lengthening of the PR interval before the dropped beat in second-degree AV block is often
called the Wenckebach phenomenon, after K. F. Wenckebach, a pioneer in the study of clinical
arrhythmias, who described this behavior in 1899 using a smoked drum kymograph to record carotid and
jugular venous pulsations. Sadly, and from a grammatical standpoint inexcusably, Professor Wenckebach
has become a verb, as in: “The patient is Wenckebaching.”
The features of the Wenckebach phenomenon are diagrammed in Figure 16-24, where the first PR interval
after the blocked P wave is 0.25 s. The following three PR intervals are 0.32, 0.37, and 0.40 s, after which
the next P wave is blocked (PR = ∞). The progressive prolongation of the PR interval occurs by decreasing
increments (listed below the ECG in Fig. 16-24), so that when sinus rate (PP interval) is constant, the
decreasing increments in PR interval cause a slight acceleration of ventricular rate. This phenomenon,
called group beating, is documented at the bottom of Figure 16-24, where cycle length (RR interval) in
the group of four QRS complexes is seen to shorten from 0.78 to 0.74 s.
Fig. 16-26: Mobitz type II AV block (monitor lead recorded in an ambulance). The PR interval in the
conducted beats, where the QRS is prolonged, is about 0.20 s. Sudden failure of atrial depolarization to
activate the ventricles after the fifth P wave causes cardiac arrest. This is Mobitz II block because the PR
interval does not increase before cessation of AV conduction.
P.456
Fig. 16-27: Third-degree AV block (complete heart block) (lead II). Although this electrocardiogram
superficially resembles 2:1 AV block, there is no constant relationship between the P waves, which are
regular at a rate of 102, and the QRS complexes, which are regular at a rate of 45. The atria and ventricles
are therefore beating independently. The normal QRS duration (0.06 s) indicates that the ventricles are
activated by an AV junctional pacemaker above the bifurcation of the bundle of His.
There are two explanations for progressive prolongation of the PR interval after the dropped beats. The
first is increasing depression of AV conduction caused when propagation of a series of impulses through
the depressed AV node (or AV bundle) inactivates a progressively increasing fraction of the calcium (or
sodium) channels. This increases the PR interval until, when too few depolarizing channels open to
generate a propagated action potential, conduction fails; this is like paralyzing the snake in Figure 16-
15A. The pause after the dropped QRS allows the depolarizing channels to recover, so that the cycle starts
again. The second explanation highlights the progressive shortening of the RP interval (the interval
between the QRS complex and the following P wave). This is seen in Figure 16-24, where each increase in
PR interval brings the next P wave closer to the preceding QRS complex. As a result, impulses conducted
from the atria begin to reach the AV junction earlier during the relative refractory period left behind
after the preceding beat, which causes these atrial impulses to become blocked. According to the snake
analogy, this would occur if, each time the serpent made its circle, the head came closer to the tail until
the snake bit itself (Fig. 16-15C). Most evidence now favors the second explanation.
Third-Degree Atrioventricular Block

Third-degree AV block, also called complete AV block or simply heart block, occurs when conduction
between the atria and ventricles is completely interrupted. Because the atria and ventricles beat
independently, P waves and QRS complexes bear no relationship to each other in third-degree AV block
(Fig. 16-27). When the atrial rhythm in complete heart block is under the control of the SA node, the
atrial rate is regular and faster than the ventricular rate, which is controlled by a lower pacemaker.
Third-degree AV block can be diagnosed in patients with atrial fibrillation when ventricular rate is slow
and regular (Fig. 16-28).
The site of the lower pacemaker that controls ventricular systole in patients with third-degree AV block is
suggested by the duration of the QRS complex. Impulses generated by a
P.457
wave of depolarization that reaches the ventricles from above the bifurcation of the bundle of His (where
the AV bundle divides into the right and left bundle branches) are conducted simultaneously into the two
ventricles, so that the QRS complex is narrow (Fig. 16-27), whereas QRS complexes initiated by a
pacemaker below the bifurcation of the bundle of His are prolonged because activation of the two
ventricles is no longer simultaneous (Fig. 16-28).
Fig. 16-28: Third-degree AV block in a patient with atrial fibrillation (lead II). Diagnosis of the latter is based
on the absence of P waves and undulations of the baseline. The QRS complexes are wide (∼0.12 s) and appear
at a regular rate of 26, which indicates that the ventricles are depolarized by a pacemaker below the
bifurcation of the bundle of His.
The consequences of third-degree AV block are determined mainly by the rate of ventricular beating. If
the latter is normal or nearly normal, as can occur in congenital heart block, the functional impairment is
usually minor and due mainly to loss of the atrial “primer pump” described in Chapter 11. When
ventricular rate is very slow, however, cardiac output is reduced (see Chapter 12). Patients with
intermittent third-degree AV block often experience syncopal attacks, called the Stokes–Adams syndrome,
when failure of lower pacemakers to fire causes brief episodes of asystolic cardiac arrest. The latter can
progress and lead to the death of the patient.
Pathophysiology of AV Block
The clinical significance of AV block depends in part on whether it is in the AV node or His-Purkinje
system. This is because block in the AV node is usually an exaggeration of the physiologically slow
conduction, which is often reversible, whereas block in the AV bundle, bundle branches, and fascicles is
due most commonly to anatomical lesions that are generally irreversible. The latter can cause two
syndromes, called Lev's disease and Len'gre's disease. Lev's disease is usually due to fibrosis of the
connective tissue skeleton of the heart or calcification of the mitral and aortic valve rings, both of which
can damage the proximal regions of the His-Purkinje system in the upper part of the interventricular
septum. Len'gre's disease was once believed to be caused by degeneration or apoptosis of His-Purkinje
cells in and below the AV bundle; however, several sodium channel mutations are now known to cause this
syndrome. Block in the His-Purkinje system also occurs in ischemic heart disease; because of the
extensive collateral circulation in the interventricular septum, the appearance of this conduction
abnormality in a patient with an acute myocardial infarction usually indicates that multi-vessel coronary
artery occlusion has severely damaged the interventricular septum, which has a dual blood supply (see
Chapter 1). The appearance of block in or below the AV bundle in a patient with acute myocardial
infarction therefore has a bad prognosis not only because of the danger that third-degree AV block might
develop (which can usually be managed by an electronic pacemaker), but because it is a marker for a
large infarction with extensive left ventricular damage.
Bifascicular Block, Trifascicular Block, and Bilateral Bundle Branch

Block
Bundle branch block and fascicular blocks do not prevent atrial impulses from reaching the ventricles (see
Chapter 15), but block in both bundle branches (bilateral bundle branch block) or all three fascicles
(trifascicular block) cause AV conduction to fail. When counting fascicles, the right bundle branch is
viewed as a third fascicle (Table 16-5), so that the combination of right bundle branch block with either
left anterior fascicular block or left posterior fascicular block, called bifascicular block, distorts the QRS,
but because a single fascicle still conducts impulses from the atria to the ventricles, AV conduction is
preserved. Trifascicular block, which occurs when block in the right bundle branch and both fascicles of
the left bundle branch
P.458
impairs AV conduction, can be diagnosed when a patient with known bifascicular block develops an
increasing PR interval or higher degree of AV block.
Table 16-5 Fascicles and Fascicular Blocks
Conduction Blocked in: Remaining Pathways ECG Abnormality
I. Fascicular Blocks
Left anterior fascicle Right bundle branch Left axis deviation

+ left posterior
fascicle
Left posterior fascicle Right bundle branch Right axis deviation

+ left anterior
fascicle
Right bundle branch Left anterior + left Right bundle branch block
posterior fascicles
II. Bifascicular Blocks
Left anterior fascicle + right Left posterior Right bundle branch block +
bundle branch fascicle left axis deviation
Left posterior fascicle + right Left anterior fascicle Right bundle branch block +
bundle branch right axis deviation
Left anterior + left posterior Right bundle branch Left bundle branch block
fascicles
III. Trifascicular Block
Left anterior + left posterior None Increasing second degree AV

fascicle + right bundle block or new third-degree AV
branch block
Arborization Block and Masquerading Bundle Branch Block

Conduction block in the distal Purkinje system, called arborization block, causes QRS complexes to be
markedly prolonged and distorted. An uncommon form of arborization block, called masquerading bundle
branch block, can be diagnosed when the QRS is markedly prolonged and right axis deviation accompanies
left bundle branch block, or left axis deviation accompanies right bundle branch block. Because these
criteria are rather vague and overlap those for bifascicular block, arborization block and masquerading
bundle branch block are rarely diagnosed. However, the concept of arborization block is important
because it identifies patients with heart failure and impaired conduction in the distal Purkinje system in
whom electronic stimulation of regions of the ventricles in which activation is delayed (cardiac
resynchronization therapy) can alleviate symptoms and prolong survival (see Chapter 18).
Premature Systoles
A wave of depolarization that originates in any part of the heart can initiate a premature systole, but the
terminology in this field is arcane (“secret, esoteric, mysterious, obscure, understood by only the
initiated”). Depending on whether one is writing a textbook, examining an
P.459
electrocardiogram, or palpating the pulse, these can be called premature systoles, premature
depolarizations, or premature contractions. The fact that there is no standard word order adds to this
complexity; for example, premature systoles originating in the ventricles can be called ventricular
premature systoles, premature ventricular systoles, ventricular premature depolarizations (VPDs),
premature ventricular depolarizations (PVDs), ventricular premature contractions (VPCs), and premature
ventricular contractions (PVCs). Premature systoles are sometimes referred to as extrasystoles, but this
term is a misnomer when a premature systole replaces a normal systole, and so does not add to the
number of beats. The terms ectopy, ectopic systoles, ectopic contractions, and ectopic depolarizations
are also used because most premature systoles are initiated at abnormal (ectopic) sites.
Fig. 16-29: Atrial premature systole (lead II). The fourth P wave is premature and followed by a normal QRS
complex after a normal PR interval. The intervals between the two sinus P waves immediately before and
after the premature systole (1.72 and 1.76 s) are greater than that between the two sinus P waves that
“enclose” the premature systole (1.54 s), which indicates that the atrial premature systole penetrated and
reset the timing of the sinoatrial node as shown on the Lewis diagram.
Atrial Premature Systoles

Atrial premature systoles are characterized by the early appearance of abnormal P waves followed by QRS
complexes that resemble the QRS complexes in the rest of the ECG (Fig. 16-29); the QRS complexes are
usually not distorted because atrial premature systoles are conducted normally into the ventricles from
above the bifurcation of the bindle of His via the His-Purkinje system. The PR interval after an atrial
premature systole is also normal unless the atrial depolarization occurs so early that it reaches the AV
node before the latter has recovered from the preceding impulse, in which case the PR interval can be
prolonged. If an atrial premature systole reaches the AV node during its absolute refractory period, no
QRS complex will follow the premature P wave; this is called a blocked atrial premature systole (Fig. 16-
30).
Aberrant Conduction and Supernormality

Aberrant conduction or aberrancy is seen in Figure 16-31, where the third QRS complex is prolonged and
abnormal, even though it was initiated by an atrial premature systole (the latter is superimposed on the T
wave of the preceding beat). The reason that this QRS is abnormal is that the wave of depolarization
initiated by the atrial premature systole had arrived at the right bundle branch during the refractory
period that followed the preceding normal beat, and so
P.460
was conducted into the ventricles by an abnormal (aberrant) pathway. Aberrancy is most often the result
of block in the right bundle branch, where the refractory period is especially long. Because long diastolic
intervals prolong both the action potential and refractory period (the interval-duration relationship, see
Chapter 14), atrial or junctional premature systoles are more likely to be conducted with aberrancy when
a short cycle follows a long cycle; this pattern is called the “Ashman phenomenon.”
Fig. 16-30: Blocked atrial premature systoles (lead II). The ECG begins with a diphasic sinus P wave
that is followed, after a PR interval of 0.14 s, by a slightly prolonged QRS complex. An inverted P
wave (P') that is superimposed on the T wave following this QRS is an atrial premature systole that,
because it does not depolarize the ventricles, is blocked. The second QRS complex, which is
initiated by another sinus P wave (P), is followed by a second blocked premature atrial systole (P'),
after which the third and fourth QRS complexes are initiated by sinus P waves. The third QRS
complex resembles the other QRS complexes but, because the PR interval is short (∼0.11 s), it was
not initiated by the sinoatrial node; instead, this is an “escape” beat (see below) caused by
depolarization of a pacemaker in the AV junction. The absence of an inverted deflection on the T
wave of the third QRS, and the regular sinus P waves (which occur at intervals of 0.80 s, and twice
0.78 and 0.75 s) support the interpretation that the downward deflections that follow the first two
QRS complexes (labeled P') are premature P waves.
Supernormal conduction (see Chapter 14) is seen in Figure 16-32, which begins with four wide QRS
complexes that are the result of right bundle branch block. The fifth QRS complex, which follows a short
RR interval because it is initiated by an atrial premature systole, is narrower than the previous QRS
complexes because the premature atrial depolarization reached the right bundle branch during its
supernormal period and so was conducted normally into the ventricles.
Fig. 16-31: Atrial premature systole with aberrant ventricular conduction (lead V2 above, lead II
below). The second QRS complex is followed by a premature P wave that is superimposed on the T
wave of the preceding ventricular beat, which explains the apparent peaking of this T wave
(downward white arrows). The third QRS complex, which follows the premature P wave (black
arrowhead), is prolonged to 0.12 s, and in lead V2 has a contour typical of right bundle branch
block. This prolonged QRS complex was probably caused by aberrant conduction of the atrial
premature depolarization because, although it was initiated by a premature P wave, it arrived at
the bifurcation of the bundle of His before the right bundle branch had recovered from the
preceding depolarization (retouched).
P.461
Fig. 16-32: Supernormality in a patient with rate-dependent bundle branch block (leads V1, V2, and
V3 recorded simultaneously). The first four QRS complexes, which are sinus beats at a cycle length of
0.50 s (P waves are clearly seen in lead V2), are prolonged (QRS duration is 0.11 s) and exhibit a right
bundle branch block pattern. The fifth QRS complex, which has a normal duration of 0.06 s, is
initiated by a premature P wave (the downward deflection on the T wave in lead V2) that shortens
the interval between QRS complexes to 0.46 s. The narrow QRS can be explained if the impulse
responsible for the fifth QRS reached the right bundle branch during its supernormal period. The QRS
remains narrow in the next beat (the sixth), where cycle length is again 0.46 s. A third narrow QRS
(the seventh) then follows a long cycle (0.58 s), which allowed enough time for the right bundle
branch to recover its excitability following the previous beat. The last three QRS complexes, where
cycle length has returned to 0.50 s, again exhibit a right bundle branch block pattern. Narrowing of
the QRS and disappearance of the right bundle branch block pattern after the first two short cycles
can be attributed to supernormality.
Resetting of the SA Node Pacemaker

Atrial premature systoles are often propagated into the sinus node where they reset the timing of the
subsequent sinus P waves (Fig. 16-33). This explains why the first normal P wave following the atrial
premature systole in Figure 16-29 appears sooner than would be expected if the SA node had maintained
its usual firing rate. Atrial premature systoles can also be followed by a compensatory pause (see below).
Junctional Premature Systoles

A wave of depolarization that arises in the AV junction, either in the AV node or the AV bundle above the
bifurcation of the bundle of His, generates a premature QRS complex called a junctional premature
systole (an older term, “nodal premature systole,” is no longer used because not all of these beats
originate in the AV node). The QRS in junctional premature systoles is usually normal in contour because
the ventricles are depolarized synchronously by impulses that are conducted through the two bundle
branches.
P waves generated by junctional premature systoles are called retrograde P waves because they arise
from impulses conducted into the atria from the AV junction, which lies in the right atrium near the top
of the interventricular septum, instead of from the SA node, which is located where
P.462
the superior vena cava enters the right atrium. For this reason, the frontal plane vector generated by a
retrograde P wave is directed superiorly and often to the right, in contrast to normal P waves whose
vector is directed inferiorly and to the left (Fig. 16-34). This is why retrograde P waves are typically
inverted in lead II.
Fig. 16-33: Resetting of the sinoatrial (SA)v node pacemaker by an atrial premature systole. Solid
line: Action potential of a pacemaker cell in the SA node that fires at regular intervals. Dashed line:
Depolarization of the SA node pacemaker by an impulse conducted from an atrial premature systole
(apc) advances the timing of the following SA node cycles.
Junctional premature systoles were once classified according to the relative timing of the retrograde P
waves and QRS complexes; when the retrograde P wave preceded the QRS complex by <0.12 s, the beat
was called an upper nodal beat; when the P wave and QRS complex occurred at the same time, the beat
was a middle nodal beat; and when the retrograde P wave followed the QRS complex, the beat was a
lower nodal beat. These terms have fallen into disuse because it is now clear that the major determinant
of the relationship between the QRS complexes and retrograde P waves in junctional premature systoles is
the timing and relative rates of antegrade and retrograde conduction rather than where in the AV
junction the premature systole originated. Junctional premature systoles that depolarize the atria before
they reach the ventricles generate a retrograde P wave that precedes the premature QRS complex (Fig.
16-35A), whereas premature systoles that depolarize the ventricles ahead of the atria give rise to P waves
that follow the QRS complexes (Fig. 16-35B). P waves are not seen when a premature junctional
depolarization is not conducted into the atria, or when simultaneous depolarization of the atria and
ventricles causes the P wave to be “buried” in the QRS complex.
Fig. 16-34: Normal and retrograde P waves viewed as P wave vectors within the Einthoven triangle.
Left: The normal P wave in the frontal plane propagates to the left and inferiorly, and so inscribes
an upright deflection in lead II. Right: Retrograde P waves propagate superiorly, often to the right,
and inscribe an inverted deflection in lead II.
P.463
Fig. 16-35: Retrograde P waves in junctional beats (lead II). These examples, taken from patients with
accelerated junctional rhythms, show retrograde P waves (arrows) that precede (A) and follow (B) QRS
complexes. Retrograde P waves can be absent if they are buried in the QRS complex (not shown), or if
impulses arising in the AV junction are not conducted to the atria.
Ventricular Premature Systoles
Premature impulses that arise below the bifurcation of the bundle of His give rise to QRS complexes that
are abnormal because the normal synchrony of ventricular depolarization is lost (Fig. 16-36). For this
reason, a premature QRS complex that differs from the normal QRS and is not preceded by a P wave is
usually a ventricular premature systole (Fig. 16-37). However, this diagnosis cannot be made on the basis
of QRS contour alone because the QRS complexes in bundle branch block (see Chapter 15) and aberrant
QRS complexes (see above), like those generated by ventricular premature systoles, are prolonged.
The timing of the P waves that appear after a ventricular premature systole generally differs from the
timing of the P waves that follow atrial and junctional premature systoles. This is because the latter are
often conducted into the SA node where they advance the subsequent train of sinus beats by resetting the
sinus pacemaker (see above), whereas most ventricular premature systoles usually do not reach the SA
node before it has fired spontaneously and so cannot advance the timing of the SA node pacemaker.
When retrograde conduction of a ventricular premature systole into the AV node prevents the next sinus
beat from depolarizing the ventricles, the premature QRS complex is followed by a compensatory pause,
after which the QRS complexes resume their normal schedule (Fig. 16-37). (The term “compensatory
pause” is actually a misnomer because the pause does not compensate for anything, but instead allows
the next sinus beat to occur on the timetable that would have followed had there not been a premature
systole.) Occasionally a ventricular premature systole occurs between two sinus beats, and so is
interpolated (Fig. 16-38); in this case the premature systole is a true extrasystole.
Fig. 16-36: Mechanism of QRS prolongation by a ventricular premature depolarization. A: A normal

impulse conducted from above the bifurcation of the bundle of His (star) depolarizes the ventricles
simultaneously. B: An abnormal impulse that arises in the ventricle below the bifurcation of the
bundle of His (star) prolongs the QRS because of the increased length of the path of depolarization
and slow conduction across the interventricular septum (horizontal dotted arrows).
P.464
Fig. 16-37: Ventricular premature systole (lead II). The fourth QRS complex, which is premature and not
preceded by a P wave, is bizarre and wider than the normal QRS complexes in this record. The pause after
the ventricular premature systole, called a compensatory pause, occurred because the impulse initiated by
the subsequent sinoatrial node depolarization reached the AV junction during its refractory period. The fact
that the interval between the two P waves that “enclose” the ventricular premature systole is almost the
same as that between two conducted P waves demonstrates that the premature systole did not reset the SA
node pacemaker.
Ventricular premature systoles are common in individuals with normal hearts. A single ventricular
premature systole in a standard 12-lead ECG is generally of little consequence, but when they occur
frequently they are more likely to be clinically significant. More important is their complexity. Multifocal
ventricular premature systoles, where premature QRS complexes have different morphologies (Fig. 16-
39), are of more ominous prognostic significance than unifocal ventricular premature systoles that have a
single morphology. Ventricular premature systoles are also of more concern when they occur as bigeminy,
where every other QRS is a ventricular premature systole (Fig. 16-40), or are repetitive; two ventricular
premature systoles in a row (couplets) (Fig. 16-41) are more dangerous, as are three in a row (a triplet or
salvo); while a short run (non-sustained ventricular tachycardia) can indicate a significant risk of sudden
cardiac death. Most dangerous is a run of ventricular premature systoles that lasts longer than 30 s, which
is usually referred to as sustained ventricular tachycardia. Ventricular premature systoles that fall on the
T wave of the preceding cycle (the R on T phenomenon) (Fig. 16-42) are also dangerous because they
occur during the vulnerable period when they can trigger ventricular fibrillation (see below).
Tachycardias
Supraventricular tachycardias can originate in the SA node, atria, and AV junction, which are above the
point at which the AV bundle divides into the bundle branches, whereas ventricular tachycardias arise
below the bifurcation of the bundle of His.
Fig. 16-38: Interpolated ventricular premature systole (lead II). The third QRS, which is premature, is not
preceded by a P wave and is prolonged, so that it is ventricular in origin. As this premature systole occurs
between two sinus beats, it is interpolated.
P.465
Fig. 16-39: Multiform ventricular premature systoles (lead II). The second and fifth QRS complexes, which
are premature, prolonged, bizarre, and not preceded by P waves, are multiform because of their different
contours.
Fig. 16-40: Ventricular bigeminy. QRS complexes alternate between sinus beats and ventricular premature
systoles (retouched).
Fig. 16-41: Ventricular premature systoles appearing as a couplet (lead II). The third and fourth QRS
complexes, which are premature and prolonged, represent a couplet (two ventricular premature systoles in a
row). The underlying rhythm in this patient is atrial fibrillation.
Fig. 16-42: The “R on T” phenomenon (lead II). The third QRS complex, which is prolonged and
differs from the other two QRS complexes, is a ventricular premature systole that falls at the end
of the T wave of the preceding beat. This is confirmed because the QT intervals of the normal
beats (dark bars below the electrocardiogram) are longer than the interval preceding the
ventricular premature depolarization.
P.466
Supraventricular Tachycardias
Tachycardias arising above the bifurcation of the bundle of His can be classified as sinus, atrial, and
junctional on the basis of electrocardiographic criteria such as P wave morphology and the relationship
between P waves and QRS complexes (Table 16-6A). However, this simple classification is becoming
obsolete because optimal clinical management now requires that the mechanism responsible for the
arrhythmia be identified in each patient.
Classification Based on P Wave Morphology and the PR Interval

In sinus tachycardia, P wave morphology is normal, whereas atrial tachycardias are characterized by
abnormal (but not retrograde) P waves that are followed by normal QRS complexes. The PR interval is
normal in both unless the tachycardia is very rapid or there is an underlying AV conduction abnormality. P
waves are not always seen in junctional tachycardias; when present they are retrograde and either occur
after the QRS complexes or precede the QRS with an abnormally short PR interval (<0.12 s).
Classification Based on Mechanism

The older electrocardiographic distinction between sinus, atria, and junctional tachycardias is being
replaced with the mechanistic classification described in Table 16-6B. It is often impossible to distinguish
among these mechanisms using electrocardiographic criteria, so today's “gold standard” has become an
electrophysiological study that collects data from stimulating and recording electrodes placed in the
heart.
Table 16-6 Supraventricular Tachycardias

A. Classification Based on P wave Morphology and the PR Interval
Tachycardia P Waves PR Interval
Sinus Normal Normal, can be prolonged
Atrial Abnormal Normal, can be prolonged
Junctional Retrograde Short or reversed (R-P)
B. Classification Based on Mechanism
Tachycardia Type of Mechanism Specific Mechanism
Sinus tachycardia Increased pacemaker Accelerated sinus pacemaker

activity
Sinus node reentry Reentry SA nodal circuit
Automatic atrial Increased pacemaker Ectopic atrial pacemaker

tachycardia activity
Atrial reentry Reentry Atrial circuit
Triggered atrial tachycardia Triggered Triggered depolarizations in the

depolarizations atria
Automatic AV junctional Increased pacemaker Ectopic AV node or AV bundle

tachycardia activity pacemaker
AV nodal reentry Reentry Dual AV nodal circuit
Accessory pathway reentry Reentry Accessory pathway circuit
P.467
At least eight different mechanisms can give rise to supraventricular tachycardias (Table 16-6B).
Automatic supraventricular tachycardias can be caused by accelerated pacemaker activity in the SA node
(sinus tachycardia), atria (automatic atrial tachycardia), and AV node (automatic AV junctional
tachycardia). Reentry can occur in the SA node, atria, and AV conduction pathways; unlike automatic
tachycardias that often accelerate after they begin (called “warming up”), reentrant tachycardias usually
begin and end suddenly as a paroxysm, and so are called paroxysmal supraventricular tachycardias
(PSVT). A few supraventricular tachycardias are caused by triggered activity in the atria. Clinical
identification of the underlying mechanism can be difficult and may require invasive electrophysiological
testing.
Sinus Node Reentry

This uncommon arrhythmia differs from sinus tachycardia in that it is generated by a reentrant circuit in
the SA node, rather than accelerated depolarization of SA node pacemaker cells. The P waves in SA node
reentry are normal, but this arrhythmia differs from sinus tachycardia in its paroxysmal behavior and the
ability of vagal maneuvers and electrical cardioversion to abolish these tachycardias by interrupting the
reentrant circuit.
Automatic Atrial Tachycardias

Automatic atrial tachycardias are characterized by abnormal P waves that are followed by a normal PR
interval unless the tachycardia is very rapid and normal decremental conduction in the AV node prolongs
the PR interval. These atrial tachycardias are non-paroxysmal because the pacemaker usually accelerates
(“warms up”) after the tachycardia begins. Unlike reentrant atrial tachycardias (see below), automatic
atrial tachycardias are not abolished by vagal stimulation or electrical cardioversion.
Multifocal (or chaotic) atrial tachycardias (Fig. 16-43), which are characterized by irregular variations in
both P wave morphology and PR interval, make up an important subset of the automatic atrial
tachycardias. These tachycardias are often seen in patents with pulmonary disease, and can herald the
appearance of atrial flutter or fibrillation.
Atrial Reentry
Atrial reentry, like other reentrant arrhythmias, is generally paroxysmal. P waves are abnormal and the
reentrant circuit can be interrupted by electrical stimulation of the atria. Unlike most reentrant
supraventricular tachycardias, in which the circuit includes calcium channel-dependent
P.468
conduction through in the AV node (see below), this uncommon tachycardia is usually not influenced by
vagal stimulation or drugs that inhibit calcium channel opening.
Fig. 16-43: Multifocal atrial tachycardia (lead V3). The rate is about 120 and slightly irregular; P
wave morphology changes throughout this record (retouched).
Automatic AV Junctional Tachycardia

Pacemaker cells in the AV junction, whose activity is normally suppressed by the more rapid sinus
pacemaker, can control ventricular beating when the sinus node pacemaker slows or conduction of sinus
node impulses is blocked; under these conditions, AV junctional pacemakers generate escape rhythms at
rates that are usually between 25 and 40 (see Chapter 15). Abnormal acceleration of these junctional
pacemakers to rates greater than 50 causes an automatic AV junctional tachycardia called an accelerated
junctional rhythm; even though ventricular rate may not exceed 100, this is a tachyarrhythmia because
the junctional pacemaker is firing more rapidly than normal. When P waves can be identified in automatic
AV junctional tachycardias, they are retrograde and can precede, follow, or be buried in the QRS
complexes (see Fig. 16-35). These arrhythmias, which are usually non-paroxysmal, can be slowed, but not
often terminated, by vagal stimulation and adenosine.
Cardiac glycosides can accelerate pacemaker activity in the AV junction; when the rates are between 50
and 60, junctional tachycardias are not necessarily a sign of digitalis overdose, but because their rate
increases as the dose of the cardiac glycoside is increased, acceleration of an AV junctional tachycardia is
a warning that too much of this drug is being administered.
Av Nodal Reentry
AV nodal reentry, the most common cause of supraventricular tachycardia, is a reciprocal rhythm that
utilizes the dual AV nodal conduction pathways described above. This arrhythmia, like other reentrant
tachycardias, is usually paroxysmal. The most common AV nodal reentrant tachycardias are triggered by
an atrial premature systole, and conduction from atria to ventricles is via the slow pathway (see Fig. 16-
10); when this occurs, retrograde P waves are often “buried” in the QRS complexes because impulses
reach the atria via the fast pathway at about the same time the ventricles are activated by the slow
pathway. If retrograde P waves are seen, they generally appear immediately after the QRS (Fig. 16-44). AV
nodal reentrant tachycardias can also be triggered when a ventricular premature systole is transmitted to
the atria via the slow pathway and returns to the ventricles via the fast pathway; this causes QRS
complexes to follow retrograde P waves after what appears to be a short PR interval.
Fig. 16-44: Supraventricular tachycardia with retrograde P waves (leads II and III recorded
simultaneously). The narrow QRS complexes, which appear at a rate of 150, are diagnostic of
supraventricular tachycardia. The small downward deflections on the ST segments are retrograde P
waves.
P.469
The “weak” point in AV nodal reentrant circuits is the AV node, which allows these tachycardias to be
terminated when calcium channel-dependent conduction is inhibited by stimulating vagal activity (e.g.,
carotid sinus massage) or administration of a drug that inhibits calcium channel opening (e.g., adenosine,
a purinergic agonist). Calcium channel blockers and β-adrenergic blockers can also be useful in treating
these arrhythmias, as can cardiac glycosides which have a central action that activates vagal outflow.
Because radiofrequency ablation of the fast pathway is generally curative, this procedure has become the
preferred treatment for these tachycardias.
Accessory Pathway Reentry (Wolff–Parkinson–White Syndrome)

Supraventricular tachycardias associated with preexcitation (Wolff–Parkinson–White syndrome or WPW),
like AV nodal reentrant tachycardias, depend on a reciprocal circuit that includes the AV node and the
accessory pathway (see above). These tachycardias, which are typically paroxysmal, are generally
initiated by an atrial premature systole that, because accessory pathways have a longer refractory period
than the AV node, is conducted to the ventricles via the AV node and returns to the atria via the accessory
pathway (Fig. 16-11). This explains why QRS complexes recorded during these supraventricular
tachycardias usually do not begin with a delta wave. Less commonly, when antegrade conduction is
through the accessory pathway, delta waves are present.
Supraventricular tachycardias in patients with accessory pathways depend on both sodium- and calcium
channel-dependent conduction, and so can be terminated by inhibiting either sodium or calcium channel
opening. However, radiofrequency ablation of the accessory pathway has become the preferred treatment
as it eliminates the need for medication and the occasional hospital visit if the arrhythmia recurs.
Successful ablation therapy also eliminates the small, but real, risk of sudden death, which can occur if a
patient with an accessory pathway develops atrial fibrillation. These patients are in serious danger
because unfiltered conduction of impulses from the atria to the ventricles by the accessory pathway can
allow atrial fibrillation, which itself is rarely lethal, to cause ventricular fibrillation by transmitting the
rapid and disorganized electrical activity to the ventricles.
Ventricular Tachycardia
The hallmark of ventricular tachycardia is a rapid series of QRS complexes that are wide (>0.12 s) (Fig.
16-45) and not preceded by P waves; if atrial activity is seen, the P waves bear no fixed relationship to
the prolonged QRS complexes. Although all ventricular tachycardias have wide QRS complexes, not all
wide QRS tachycardias are of ventricular origin because, as noted above, both aberrant conduction caused
by the rapid heart rates in supraventricular tachycardias and preexisting bundle branch block can prolong
the QRS. The difficulty in distinguishing ventricular tachycardia from supraventricular tachycardia with
prolonged QRS (Fig. 16-46) has led to the use of the term wide QRS tachycardia, which has the advantage
of being descriptive without implying mechanism or suggesting therapy. This diagnosis indicates that more
data are needed to distinguish between supraventricular and ventricular tachycardia.
The appearance of fusion beats or captures can establish a diagnosis of ventricular tachycardia in a
patient with a wide QRS tachycardia (Fig. 16-47). “Fusion beats,” in which the QRS is narrower than the
QRS complexes caused by a ventricular tachycardia, are initiated by a wave of depolarization that
activates the ventricles via both the His-Purkinje system and the ectopic focus
P.470
within the ventricles; fusion beats therefore occur when the ventricles are depolarized simultaneously by
impulses conducted via the AV node and from the ectopic ventricular site. A narrow QRS preceded by a P
wave in a wide QRS tachycardia, called a “capture” (Fig. 16-48), demonstrates that a supraventricular
impulse has been conducted normally into the ventricles; this indicates that the tachycardia is ventricular
in origin.
Fig. 16-45: Ventricular tachycardia (leads I, II, and III). The QRS complexes recorded during the tachycardia
on day 1 are wider, begin differently, and have a different morphology compared to the QRS complexes
recorded during sinus rhythm on day 2. For these reasons, the first ECG is probably ventricular tachycardia.
Ventricular tachycardias arise in regions with sodium-dependent action potentials, so that they are not
slowed by vagal stimulation, adenosine, or calcium channel blockers. β-Blockers are useful in preventing
ventricular tachycardias, but the antiarrhythmic effect is due largely to
P.471
reduction of the arrhythmogenic effects of sympathetic stimulation, energy starvation, and calcium
overload, rather than a direct effect on the electrical properties of the heart.
Fig. 16-46: Wide QRS tachycardia (lead II). The rhythm is slightly irregular with an average cycle length of
∼0.38 s, which corresponds to a ventricular rate of ∼160; P waves cannot be identified. The irregular
appearance of QRS complexes suggests that the underlying arrhythmia is atrial fibrillation, but this ECG could
also represent ventricular tachycardia, or atrial tachycardia with preexisting bundle branch block or
aberrancy. Because the underlying mechanism cannot be determined with certainty, the rhythm is best
described as a wide QRS tachycardia.
Fig. 16-47: Ventricular tachycardia with fusion beats and captures (continuous recording of leads III and aVF;
the two vertical lines mark the lead change). The first four QRS complexes (in lead III) are wide (0.12 s) and
occur at a cycle of about 0.56 s, which corresponds to a rate of 107. The fifth QRS is slightly early (the
preceding interval is 0.50 s), and is narrower than the other complexes in this lead. The mechanism for the
narrower QRS is apparent after the lead change, where the sixth QRS complex (now in lead aVF) is again
wide, with a duration in this lead of 0.11 s, and is followed by a beat with a narrow QRS (0.07 s) after an
interval of 0.52 s. This beat, the seventh QRS, represents a “capture” in which a supraventricular impulse
conducted through the AV junction has depolarized the ventricles. The eighth QRS complex, which follows an
interval of 0.40 s, is again wide. The ninth QRS, which follows the preceding QRS by 0.54 s and has a contour
midway between that of the wide QRS beats and the capture in lead aVF, is a fusion beat in which the
ventricles are depolarized by both the ventricular focus and an impulse conducted from above the bifurcation
of the bundle of His. The final (tenth) QRS in this record is another ventricular beat. The appearance of
captures and fusion beats demonstrates that the wide QRS beats are ventricular in origin. (The underlying
ventricular tachycardia is slow enough to be called an accelerated idioventricular rhythm.)
Ventricular tachycardias can be monomorphic, where all of the QRS complexes are similar to one another,
or polymorphic, when the contour of the QRS complexes not all the same; polymorphic ventricular
tachycardias have a more ominous prognosis. Bidirectional tachycardias, where QRS complexes alternate
between upward and downward directions, often herald sudden death in patients with calcium-
overloaded hearts, digitalis overdose, and catecholaminergic polymorphic ventricular tachycardia (see
below). Another polymorphic ventricular tachycardia, called torsades de points (French for “twisting of
the points”), is characterized by QRS complexes which undergo a slow transition between upward- and
downward-directed complexes that resembles the twisting of a sine wave (Fig. 16-49). Torsades, which is
typically triggered by afterdepolarizations, occurs in patients with calcium-overloaded hearts and long QT
syndromes, and can be induced by drugs that prolong the QT interval.
Ventricular tachycardias can be automatic, when the arrhythmia is initiated by a pacemaker in the His-
Purkinje system below the bifurcation of the bundle of His. As ventricular pacemakers usually fire at rates
between 25 and 40, their activity is not seen unless the sinus node slows markedly or impulses conducted
into the ventricles via the His-Purkinje system are blocked.
P.472
When wide QRS complexes appear at rates between 60 and 100, they are often referred to as an
accelerated idioventricular rhythm because this arrhythmia is not associated with the high mortality of
the more rapid ventricular tachycardias.
Fig. 16-48: Ventricular tachycardia with a classical capture (lead III). The first three QRS
complexes, which are wide and not preceded by P waves, are ventricular in origin. The fourth QRS,
which is narrower and preceded by an obvious P wave, is a capture.
Fig. 16-49: Torsades de points (monitor lead). The first QRS is a sinus beat with a very long QT (>0.50 s) that
is interrupted by a premature QRS which falls on the T wave of the preceding cycle (“R on T”); the latter is
followed by a sequence of ventricular premature beats that appears to “twist” around the baseline, which is
the typical appearance of torsades.
A special type of automatic ventricular rhythm, called parasystole, occurs when the slow discharge of a
ventricular pacemaker causes wide QRS complexes to appear at a regular rate, and neither the timing of
the normal QRS complexes nor that of the sinus beats is influenced by the other (Fig. 16-50). Because
sinus impulses do not reset the parasystolic focus in a simple parasystolic rhythm, the wide QRS
complexes appear “on schedule” unless they reach the ventricles during the refractory period that follows
each sinus QRS (Fig. 16-50). This behavior was likened by my father to that of a wicked knight (the
parasystole) who emerges from his castle on a strict schedule to raid the surrounding villages unless his
exit from the castle is blocked by the arrival of the king's men (the sinus beats). Because the king's men
cannot enter the castle, the evil knight is able to continue his strict schedule of attempted raids.
Parasystolic rhythms do not always adhere to this behavior because electrotonic currents generated when
a sinus beat partially depolarizes the myocardium surrounding a parasystolic focus can both accelerate
and delay discharge of the latter; this can cause parasystolic rhythms that, while marvelously complex,
are highly regulated and exhibit a predictable behavior (Jalife and Moe, 1981).
Fig. 16-50: Parasystole (lead II). The intervals between the wide QRS complexes in this record are multiples
of the cycle length of a parasystole that fires at a regular interval of 1.52 to 1.56 s (small upward arrows
below the ECG), but is not depolarized by the sinus beats. Not all parasystolic impulses initiate a QRS
because many occur during the refractory period of preceding sinus beats. The QRS at the end of the upper
strip (small upward and downward arrows) is a fusion beat. (The thick lines in this ECG and those in Figures
16-51 and 16-54 were taken using a string galvanometer that records the shadows cast by a quartz string that
moves up and down in response to small changes in potential.) (Modified from Katz and Pick, 1956.)
P.473
Fig. 16-51: Salvo initiated by an “R on T.” The third of the four ventricular premature systoles initiates a
“salvo” (three ventricular premature systoles in a row); this premature QRS complex follows the preceding
QRS by 0.28 s and so falls during the T wave. The other premature QRS complexes, which do not initiate
repetitive firing, occur later (0.30 to 0.34 s) after the preceding sinus beat. The cycle lengths before the
ventricular premature systoles are given on the Lewis diagram. (Modified from Katz and Pick, 1956.)
The Vulnerable Period

Ventricular premature systoles that fall on the T wave of the preceding cycle (“R on T”; see Fig. 16-42)
can initiate a repetitive ventricular response, such as torsades (Fig. 16-49), a salvo (Fig. 16-51), or
ventricular fibrillation. The later portion of the T wave therefore represents a vulnerable period (Fig. 16-
52; see also Chapter 14) during which ventricular fibrillation threshold reaches its nadir; this allows even
small electrical currents that reach the ventricles to cause sudden death. Vulnerability is due to the
tendency of impulses to become disorganized when they reach the ventricles during their relative
refractory period, when sodium channels are in different phases of recovery. The resulting heterogeneity
of depolarizing currents provides a substrate for decremental conduction and unidirectional block (see
above). Earlier stimuli, which occur during the S-T segment, cannot generate a propagated wave of
depolarization because they reach the ventricles during their absolute refractory period, while stimuli
that arrive after the end of the T wave find the ventricles fully recovered and so able to generate large,
rapidly rising action potentials that do not tend to become disorganized. Although the vulnerable period
occurs at about the same time as the supernormal period, these two phenomena are not directly related.
Flutter and Fibrillation

Regular depolarization of the atria at rates exceeding 250 to 350, or of the ventricles at rates of 150 to
300, causes an arrhythmia called flutter. In fibrillation, totally disorganized depolarization at even faster
rates causes the fibrillating chamber to resemble a bag of worms.
Fig. 16-52: The vulnerable period. Impulses that reach the ventricles during the middle and
terminal portions of the T wave (unshaded) are most likely to initiate ventricular tachycardias and
fibrillation.
P.474
The role of increased heart size in sustaining these arrhythmias was described in 1914 by W. E. Garrey,
who noted that ventricular fibrillation is difficult to produce in a small heart, like that of the cat, but
hard to avoid in a large heart, such as that of the cow. To test the hypothesis that increased heart size
favors fibrillation, Garrey did the simple experiment of causing a cow heart to fibrillate, after which he
cut it into pieces. As the mass of fibrillating tissue decreased, each piece eventually ceased to fibrillate,
but instead either beat synchronously or stopped contracting altogether. This role of heart size
contributes to the increased risk of atrial fibrillation in patients with dilated atria.
Flutter and Fibrillation
Atrial Flutter
Atrial flutter accelerates ventricular rate and makes it impossible for the atria to serve as a primer pump
(Chapter 11). Although the hemodynamic consequences can be minor, especially in patients with a normal
heart in whom AV dissociation slows ventricular beating, this arrhythmia is dangerous because stasis of
blood in fluttering (and fibrillating) atria can lead to the formation of clots that can break off to form
emboli. Clots that travel from the right atrium to a pulmonary artery and from the left atrium to a
peripheral artery can be disastrous, especially systemic emboli that block a cerebral artery and cause a
stroke.
Atrial flutter can be viewed as a “macro” reentry in which a wave of depolarization forms a circus
movement that goes round and round the atria. Although atrial rates in this arrhythmia are usually close
to 300, the normally low safety factor in the AV node protects these patients from excessively rapid
ventricular rates. Common ventricular rates are 150 (2:1 AV dissociation), 100 (3:1 AV dissociation), and
75 (4:1 AV dissociation). The ECG in atrial flutter usually shows a “saw tooth” pattern in leads II, III, and
aVF. These deflections, called F waves, are caused by the atrial circus movement that replaces the normal
P waves (Fig. 16-53). The QRS complexes in atrial flutter are normal unless there is aberrant conduction,
bundle branch block, or another disorder of ventricular depolarization.
The reentrant circuit in atrial flutter typically lies within the right atrium, where impulses loop around
the tricuspid valve ring, inferior vena cava, crista terminalis, and inferior vena cava; other pathways can
also sustain this arrhythmia. Because the circuit involves only atrial myocardium, which depends on
sodium channel-dependent depolarization, vagal stimulation and
P.475
inhibition of calcium channel opening rarely abolish atrial flutter, although both can slow the ventricular
rate by increasing block in the AV node. However, the treatment of choice is becoming ablation therapy,
which because of the variability in the arrhythmogenic pathways requires careful mapping to identify the
reentrant circuit.
Fig. 16-53: Atrial flutter with 4:1 AV dissociation (lead V1, above, and II, below, recorded
simultaneously). Atrial depolarization gives rise to regular saw tooth-like undulations of the
baseline in lead II; these are typical F (flutter) waves. The atrial activity in lead V1 resembles P
waves, but these are the same flutter waves recorded in lead II. Because of the 4:1 AV dissociation,
the ventricular rate is one-fourth that of the atria.
Fig. 16-54: Ventricular flutter (lead II). The electrical activity of the ventricles resembles a sine wave;
neither QRS complexes nor T waves can be made out. (From Katz and Pick, 1956.)
Ventricular Flutter
Ventricular flutter impairs the ability of the ventricles to eject and usually makes it impossible for the
heart to sustain blood pressure or maintain cardiac output. The ECG in ventricular flutter resembles a sine
wave (Fig. 16-54) that is probably caused by a large circus movement around the ventricles.
Atrial Fibrillation
The prevalence of atrial fibrillation, which is the most common of the pathological arrhythmias, increases
with advancing age, due largely to the increasing impact of conditions, notably hypertension and
pulmonary disease, that predispose to this arrhythmia. Depolarization of the fibrillating atria is both rapid
and disorganized; the rate at which impulses pass through any point in the atria generally exceeds 400.
The electrocardiographic manifestations of this activity are undulations in the baseline, called f waves,
that vary in both amplitude and frequency (Fig. 16-55). Ventricular rates in atrial fibrillation, like those in
atrial flutter, are slower than in the atria because of filtering by the AV node; but unlike atrial flutter, the
ventricular rhythm is irregularly irregular. The QRS complexes are generally narrow (unless other
abnormalities modify ventricular depolarization) because the ventricles are depolarized normally by
impulses that arise above the bifurcation of the bundle of His.
A number of mechanisms can cause the atria to fibrillate; these include anatomical and histological
changes associated with inflammation, dilatation, hypertrophy, and fibrosis. Electrophysiological
abnormalities that can cause atrial fibrillation include abnormal automaticity, triggered depolarizations,
chaotic patterns of reentry caused when the breakup of a single wave forms multiple new wavelets, and
spiral waves (see below). Molecular abnormalities that predispose to this arrhythmia include mutations in
the channels that carry iks, ikr, ik1, and iNa, nucleoporin, and a gap junction protein. The recent finding
that atrial fibrillation frequently originates in specialized cells located near the sites where pulmonary
veins enter the left atrium has led to the use of radiofrequency ablation at these regions, which often
abolishes this arrhythmia.
Chronic atrial fibrillation causes the atria to undergo structural and molecular changes, called atrial
remodeling, that are similar to those that occur in failing hearts (see Chapter 18). These changes, which
delay and disorganize atrial conduction, probably explain why it is difficult to abolish long-standing atrial
fibrillation with drugs and electrical cardioversion.
P.476
Fig. 16-55: Atrial fibrillation (lead II). Rapid, disorganized atrial activity causes the irregular
undulations of the baseline (f waves) that in some places occur at intervals of 0.15 s. The QRS
complexes are irregularly irregular, with RR intervals ranging between 0.70 and 1.38 s. These
variations in ventricular cycle length cannot be explained simply by the timing of the arrival of f
waves at the AV node, but instead reflect large changes in the apparent refractory period of the AV
junction caused by concealed conduction. This is shown in the Lewis diagram, where the first three
QRS complexes, labeled V1 to V3, are diagrammed along with five hypothetical atrial impulses,
labeled A1 to A5. (For simplicity, most atrial impulses—which occur much more frequently than
shown here—are not included in this Lewis diagram.) Impulses A1 and A2, which are relatively far
apart, are conducted through the AV node to generate QRS complexes V1 and V2, but A3 occurred
so soon after A2 that its passage through the AV node was blocked (short horizontal line).
Conduction of A3 into the AV node was, however, able to block the next impulse (A4). The following
impulse (A5) entered the AV node much later after A4, so that the AV node was able to recover
from A4; as a result, A5 generated the third QRS complex (V3). This shows that the long pause in
ventricular beating after V2 resulted from concealed conduction of A3 and A4, which entered the
AV node where they delayed the appearance of the next QRS complex (V3), but their conduction
was concealed because they did not generate QRS complexes.
Concealed Conduction
The irregular ventricular rhythm in patients with atrial fibrillation is due largely to the erratic arrival of
atrial impulses at the upper end of the AV node. However, the marked differences in cycle length
commonly seen in these patients are due in part to a mechanism called concealed conduction
(Langendorf, 1948). The operation of this mechanism explains why, in the ECG shown in Figure 16-55, the
intervals between QRS complexes (RR intervals) are as long as 1.38 s, whereas the shortest RR interval
demonstrates that the AV node can conduct at intervals of ∼0.70 s. It is therefore impossible to attribute
the longest cycle in Figure 16-55 to refractoriness in the AV node because the intervals between f waves
average ∼0.15 s, which should have allowed almost 10 atrial impulses to reach the upper end of the AV
node during the long cycle. This indicates that atrial impulses had entered, but did not cross, the AV
junction. The mechanism, which is similar to “inhibition” (see Fig. 16-7), is called concealed conduction
because atrial impulses are conducted into the AV node, as evidenced by their ability to prevent
subsequent impulses from reaching the ventricles, but their conduction is concealed because they do not
give rise to QRS complexes.
Concealed conduction is readily demonstrated by electrical stimulation of a turtle heart impaled with
copper wires. At slow frequencies each stimulus causes a contraction, and when the rate of stimulation is
increased gradually the heart initially contracts in response to each stimulus. As stimulation frequency
increases, some stimuli fail to activate the heart, and at very rapid frequencies all contractile responses
cease. Failure of the heart to respond to rapid stimulation cannot be attributed to irreversible damage
because after stimulation is briefly interrupted, slow stimulation again
P.477
causes contractions. Instead, the rapidly delivered stimuli fail to evoke visible contractions because each
stimulus generates a local subthreshold response that causes a refractory state that prevents subsequent
stimuli from initiating a contraction. Concealed conduction occurred because the rapid stimuli are
conducted into the tissue, as evidenced by their ability to block subsequent contractile responses, but the
conduction is concealed because the responses do not appear.
Fig. 16-56: Ventricular fibrillation. Ventricular depolarization causes the rapid chaotic undulations in the
baseline in this lethal arrhythmia.
Ventricular Fibrillation
Ventricular fibrillation is a lethal arrhythmia because it completely disorganizes ventricular contraction
and so causes cardiac arrest. The ECG in ventricular fibrillation shows chaotic oscillations instead of QRS
complexes (Fig. 16-56).
Spiral Waves and Rotors

There is growing evidence that both atrial and ventricular fibrillation are caused by spiral waves of
depolarization that develop when a wave front moving through the heart begins to slow at its edges (Fig.
16-57) or encounters an obstacle (Fig. 16-58). Slowing of conduction at the edges
P.478
of the spiral causes the movement of the wave front in the curved region to diverge from the initial
direction of impulse propagation. The result is an increase in the curvature of the wave front that leads
first to the formation of spirals (Figs. 16-57 and 16-58), and then rotors that can break off and move
independently through the heart (Fig. 16-59).
Fig. 16-57: Formation of a spiral wave in ventricular fibrillation. When a wave of depolarization moving
through the heart (A) slows at its edges, the front becomes curved (A–C) and eventually forms spirals (D).
When the tips of the spirals move behind tissue that has already been depolarized, and so is refractory
(circled R in (E), the spirals can continue as “rotors,” or break down into multiple disordered waves (F). The
lower series of diagrams is an enlargement of the areas within the dashed rectangle in the upper series.
Fig. 16-58: Formation of a spiral wave in ventricular fibrillation. When a wave of depolarization
moving through the heart (A) encounters an obstacle, such as a scar (triangle), the center of the
wave slows (B and C) and forms spirals (D and E). The lower series of diagrams, which is an
enlargement of the area at the top of the obstacle in the upper series, shows the wave-front as a
dark line and the region of slow conduction as a shaded area.
The formation and propagation of rotors are favored by decremental conduction. Mechanisms that can
decrease conduction velocity include slow reactivation of iNa, prolonged action potential duration caused
by reduction in ito, prolonged refractoriness due to inactivation of delayed rectifier potassium currents,
and/or resting depolarization that can result from a decrease in the inward rectifier current ik1. All of
these changes, which favor the appearance of fibrillation, can be caused by mutations in cardiac ion
channels and when hearts are damaged by disease.
Molecular Abnormalities in Cardiac Ion Channels and Related

Structures
Discovery that mutations in cardiac ion channels and related proteins play a major role in the
pathogenesis of clinical arrhythmias (Table 16-7; see Chapter 14) has opened a new era in the diagnosis
and management of these common, and often lethal, conditions. Molecular abnormalities are now known
to contribute to the pathogenesis of a bradyarrhythmias, such as AV block, sinus bradycardia, and
impaired conduction in the His-Purkinje system, as well as many tachycardias and both atrial and
ventricular fibrillation. Arrhythmias also occur when a molecular abnormality interacts with other
predisposing factors, such as ischemia, drugs, increased sympathetic activity, and even the ringing of an
alarm clock. To provide an insight into the
P.479
diverse causes of these heritable syndromes, the following discussion highlights four disorders, long QT
syndromes, the Brugada syndrome, catecholaminergic polymorphic ventricular tachycardia, and
arrhythmogenic right ventricular cardiomyopathy.
Fig. 16-59: Detail showing the formation of a rotor from the spiral wave shown at the top of Figure
16-58E. Advance of the center of the wave front, called the “core” (indicated by a star), allows a
spiral wave (A–G) to form a rotor that can move through unexcited tissue (H and I). The snake
depicted in Figure 16-15 has been placed in panel I to highlight the similarity between the modern
concept of a rotor to the early 20th-century analogy between a rotating wave of depolarization and
a serpent. The wave-front is shown as a dark line, and the region of slow conduction as a shaded
area.
Long QT Syndromes
The QT interval, which is directly related to the duration of the ventricular action potential (see Chapter
15), can be prolonged by persistence of the inward currents that depolarize the ventricles and reduction
of the outward currents responsible for ventricular repolarization (Table 16-7). It is not surprising,
therefore, that this syndrome can be caused by “gain-of-function” mutations that prolong sodium channel
opening and “loss-of-function” mutations that inhibit the opening of delayed rectifier potassium channels.
Mutations in cytoskeletal and other proteins also cause long QT syndromes. Some sodium and potassium
channel mutations can cause short QT syndromes that, like long QT syndromes, are associated with
ventricular arrhythmias.
P.480
Table 16-7 Heritable Causes of Long QT Syndromes
Abnormal Abnormal
Clinical Syndrome Abnormal Current
Protein Gene
LQT1 ↓iKs Kv7.1* KCNQ1
LQT2 ↓iKr Kv11.1 KCNH2

(HERG)*
HERG mutations can also cause short

QT syndromes.
LQT3 ↑iNa Nav1.5* SCN5A
SCN5A mutations can also cause

Brugada syndrome, progressive cardiac
conduction system disease, and progressive
cardiac dilatation.
LQT4 Cytoskeletal Ankyrin- ANK2

B
LQT5 ↓iKs MinK§ KCNE1
LQT6 ↓iKr MirP1§ KCNE2
LQT7 (Andersen-Tawil Syndrome) ↓iK1 Kir2.1* KCNJ2
KCNJ2 mutations can also cause

short QT syndromes.
LQT8 (Timothy Syndrome) ↑iCaL Cav1.2* CACNA1C
CACNA1C mutations can also cause

short QT syndromes.
LQT9 ↑iNa Caveolin- CAV-3

3
LQT10 ↑iNa Nav1.5§ SCN4β
Jervell–Lang–Nielsen Syndrome (also ↓iKs Kv7.1* KCNQ1

deafness) and and
MinK§ KCNE1
*: ion channel β-subunit,

§: ion channel small subunit; cytoskeletal protein
A long QT syndrome, called LQT3, is caused by gain in function mutations in SCN5A, which encodes the α-
subunits of the sodium channel Nav1.5 (see Chapter 13). Loss-of-function mutations in HERG (responsible
for ikr) cause the long QT syndrome called LQT2, while LQT6 is caused by loss-of-function
P.481
mutations in the β-subunit MiRP1. Similarly, loss-of-function mutations in Kv7.1 (responsible for iks) cause
LQT1, while LQT5 is caused by loss-of-function mutations in the β-subunit MinK. As knowledge of these
heritable syndromes increases, so does their complexity; for example, both LQT4 and LQT10 are caused
by cytoskeletal mutations. Gain-of-function mutations in HERG cause dangerous short QT syndromes,
while other HERG mutations are associated with conduction system disease.
Fig. 16-60: The Brugada syndrome. Single complexes from lead V2 in two patients with this
syndrome show the characteristic pattern of ST segment elevation, which resembles a broad R'.
The sensitivity of iks to activation by β-adrenergic agonists is an important reason why patients with Kv1.5
mutations are prone to sudden death following sympathetic stimulation. The association of the Jervell–
Lang–Nielsen syndrome with congenital deafness reflects the fact that the mutated potassium channels
are also present in the cochlea. The remarkable finding that a ringing alarm clock can cause sudden death
in patients with LQT2 reflects the fact that the potassium channel mutation which prolongs the cardiac
action potential also causes auditory hyperexcitability in the trapezoid body, a structure that links the
middle ear to autonomic centers in the brainstem (Hardman and Forsythe, 2009).
The Brugada Syndrome

The Brugada syndrome, an electrocardiographic abnormality (Fig. 16-60) first identified in 1992, has
emerged as a major cause of sudden death; some estimates put this syndrome second to only accidents as
a cause of mortality in young adults (Antzelevitch et al., 2005). The syndrome can be caused by more
than 80 mutations in the cardiac sodium channel; associated arrhythmias include polymorphic ventricular
tachycardia, ventricular fibrillation, and supraventricular arrhythmias. In most patients, this syndrome is
caused by abnormal persistence of the late depolarizing currents associated with iNa in the outflow tract
of the right ventricle. The finding of anatomical lesions in some of these patients indicates that this
syndrome can involve more an electrophysiological abnormality.
The most common causes of the Brugada syndrome are gain-in-function mutations in Nav1.5, the sodium
channel α-subunit responsible for iNa (Table 16-8). Mutations in two β-subunits and a glycerol-3-phosphate
dehydrogenase-1-like protein that modify sodium channel function also cause the Brugada syndrome, as
can mutations in the α- and β-subunits of L-type calcium channels and a β-subunit associated with iks and
ito.
Catecholaminergic Polymorphic Ventricular Tachycardia

Catecholaminergic polymorphic ventricular tachycardia (CPVT) is a dangerous syndrome characterized by
ventricular tachyarrhythmias that can degenerate to ventricular fibrillation. The tachycardias are
commonly initiated when triggered activity (afterdepolarizations) is induced by
P.482
the depolarizing currents that accompany calcium efflux by the Na/Ca exchanger (see Chapters 7 and 14).
The underlying mutations increase calcium leakage from the sarcoplasmic reticulum by modifying the
genes that encode intracellular calcium release channels (“ryanodine receptors”; see Chapter 7),
calsequestrin, and other proteins of the subsarcolemmal cisternae.
Table 16-8 Heritable Causes of the Brugada Syndrome

Abnormal Protein Abnormal Current Abnormal Gene
Nav1.5 (α subunit) ↓iNa SCN5A
Navβ (β1 subunit) ↓iNa SCN1β
Navβ3 (β3 subunit) ↓iNa SCN3β
GPDIL* (modifies Nav1.5) ↓iNa GPDIL
MirP2 (β subunit) ↑iKs, ↑ito KCNE3
Cav1.2 (α-subunit) ↓iCaL CACNA1C
Cavβ.2 (β-subunit) ↓iCaL CACNβ2
* Glycerol-3-phosphate dehydrogenase-1-like protein.
The potentially lethal consequences of sympathetic stimulation in these patients occur when β-adrenergic
agonists increase calcium entry via L-type calcium channels, which adds to the amount of calcium that
must be transported out of the cytosol by electrogenic Na/Ca exchange. Phosphorylation of the calcium
release channels by PKA, which increases calcium efflux from the sarcoplasmic reticulum, can also play a
role in this syndrome.
Arrhythmogenic Right Ventricular Cardiomyopathy

Lancisi, in the 18th century, described “family of high rank” of which four male members—great-
grandfather, grandfather, father, and son—suffered from palpitations. At autopsy, all three who died had
“aneurysm” of the right ventricle (De subitaneis mortibus 185–195. tr. Jarcho, 1980). This combination of
right ventricular dilatation and arrhythmias represents a syndrome that, because of the prominence of
serious arrhythmias, came to be called arrhythmogenic right ventricular dysplasia. However, the
underlying mechanism is not an ion channel abnormality, but instead is caused by mutations in the genes
that encode three proteins of the desmosome: desmoplakin, plakophilin-2, and plakoglobin (see Chapter
5); calcium release channels; and transforming growth factor β-3 (see Chapter 9). The importance of the
structural abnormalities has led this syndrome to be renamed arrhythmogenic right ventricular
cardiomyopathy.
Antiarrhythmic Drugs
Virtually all antiarrhythmic drugs inactivate voltage-gated ion channels by inhibiting their opening and/or
reactivation, which allows these drugs to slow pacemaker activity and reduce excitability. The latter can
abolish reentrant circuits when depressed conduction converts
P.483
unidirectional block to bidirectional block (see above). However, these drugs inhibit ion channel function
not only in areas where arrhythmias originate, but also in other regions of the heart, a complication that
explains why many antiarrhythmic drugs are dangerous and can cause sudden death.
Classification
The most widely used classification of antiarrhythmic drugs is that of Vaughan Williams (1989) (Table 16-
9); other classifications have been proposed (Members of the Sicilian Gambit, 2001; Knollmann and Roden,
2008), but these have not gained wide acceptance.
Class I antiarrhythmic agents block sodium channels, where the responses to different members of this
class are affected by different channel sub-states (Hondeghem and Katzung, 1977). Differences in the
effects of the many Class I agents have led to their division into three subclasses. Class IA agents, which
include quinidine, procaineamide, and disopyramide, depress sodium channel opening and so slow
conduction in the atria, His-Purkinje system, and ventricles; because they also increase refractoriness and
prolong action potential duration, they are especially useful in rapid tachycardias. Class IB agents, like
lidocaine, diphenylhydantoin, and mexilitine, shorten action potential duration and are potent inhibitors
of sodium-dependent
P.484
P.485
conduction in depolarized areas; the latter explains why these agents are effective in depressing
conduction in ischemic areas of the heart. Class IC agents, such as flecainide, propafenone, and
moricizine, inhibit sodium channel opening but have less effect to prolong refractoriness.
Table 16-9 Classification of the Antiarrhythmic Drugs
Class and Mechanism of Action Examples
I. Sodium channel blockade (inhibits iNa), slows action

potential upstroke
IA. Inhibits iNa, slows depolarization, prolongs Quinidine, procaineamide,

refractoriness disopyramide
IB. Inhibits iNa in depolarized myocytes, shortens Lidocaine, mexilitine,

refractoriness tocainide
IC. Inhibits iNa, slows depolarization, minimal Flecainide, propafenone,
prolongation of refractoriness moricizine
II. β-adrenergic blockade (reduces activation of iCaL), Propranolol, metoprolol,

slows SA pacemaker and AV conduction atenolol, timolol
Also reduces arrhythmogenic responses to β-

adrenergic agonists
III. Potassium channel blockade (inhibits iKr and iKs), Amiodarone, sotalol,
prolongs refractoriness bretylium, ibutilide, dofetilide
IV. Calcium channel blockade (inhibits iCaL), slows SA

pacemaker and AV conduction
Phenylalkylamines Verapamil
Benzothiazepines Diltiazem
Dihydropyridines (minor effects on the heart) Nifedipine, amlodipine,

nitrendipine
Fig. 16-61: Antiarrhythmic and proarrhythmic effects in a region of decremental conduction and
unidirectional block. A: Passage of a single impulse through the depressed tissue generates a single
premature systole (see Fig. 16-6). B: An antiarrhythmic drug can abolish the reentrant circuit by further
depressing conduction so as to convert the unidirectional block to bidirectional (complete) block. C: A
proarrhythmic effect can occur if the drug delays the return of an impulse to the depressed region by slowing
conduction elsewhere in the reentrant circuit (lightly shaded). If the new delay allows sufficient time for the
initially depressed region to recover its ability to propagate additional reentrant impulses, the passage of a
third impulse (dashed arrow labeled 3) can generate a second premature depolarization. This process can
generate a tachycardia if reentry becomes repetitive (circled question mark). ECGs generated by these
responses are shown below each diagram.
Class II antiarrhythmic agents are the β-adrenergic receptor blockers, which indirectly inhibit calcium
channel opening by blocking sympathetic responses (see Chapter 8). Class III agents, which include
amiodarone, sotalol, bretylium, ibutilide, and dofetilide, inhibit repolarizing potassium currents and so
prolong the cardiac action potential; these drugs have little or no effect to inhibit depolarizing currents.
The L-type calcium channel blockers are the Class IV agents; phenylalkylamines (e.g., verapamil) and
benzothiazepines (e.g., diltiazem) slow the SA node pacemaker and inhibit conduction in the AV node,
whereas dihydropyridines (e.g., nifedipine, nitrendipine, and amlodipine) are mainly vasodilators that
have only a minor effect on the heart.
Proarrhythmic Effects
The clinical use of many antiarrhythmic drugs is limited by proarrhythmic effects that can worsen
arrhythmias and cause sudden death. This ability to induce lethal arrhythmias led to the early termination
of major clinical trials (CAST Investigators, 1989; CAST II Investigators, 1992).
Most of the proarrhythmic effects of Class I agents are due to their ability to slow conduction, which like
the antiarrhythmic effect is caused by inhibition of sodium channel opening. These proarrhythmic effects
are especially dangerous in patients with diseased hearts, where depressed conduction is often a major
cause of the arrhythmias.
Figure 16-61 shows how a drug that depresses conduction can both prevent and cause a reentrant
arrhythmia. The ability to convert unidirectional block (Fig. 16-61A) to bidirectional block (Fig. 16-61B),
although eliminating the threat posed by one arrhythmogenic mechanism, can exacerbate other
arrhythmogenic mechanisms by creating new areas of slow conduction (Fig. 16-61C), and so increase
rather than decrease the risk sudden death. One way to view the coexistence of antiarrhythmic and
proarrhythmic effects is to equate the use of these drugs to attempts to survive in a snake pit to shooting
a threatening reptile, only to have the shot awaken several previously dormant serpents (Fig. 16-62).
Fig. 16-62: Allegory showing proarrhythmic effects of antiarrhythmic therapy. An attempt to survive in a
snake pit (left) by shooting one threatening snake (center) can prove fatal if the shot awakens a large
number of previously dormant serpents (right). This situation is reminiscent of administering an
antiarrhythmic drug that, although it eliminates a one site for arrhythmia, increases the likelihood of sudden
death by provoking other arrhythmogenic foci.
P.486
Bibliography
General
Blayney LM, Lai FA. Ryanodine receptor-mediated arrhythmias and sudden cardiac death. Pharmacol
Ther 2009;123:151–177.
Dobrzynski H, Boyett MR, Anderson RH. New insights into pacemaker activity: promoting
understanding of sick sinus syndrome. Circulation 2007;115:1921–1932.
Matthew ST, Patel J, Joseph S. Atrial fibrillation: mechanistic insights and treatment options. Eur J
Int Med 2009;20:672–681.
Olgin JE, Zipes DP. Specific arrhythmias: diagnosis and treatment. Chapter 32 In: Zipes DP, Libby P,
Bonow RO, et al., eds. Braunwald's heart disease. 7th ed. Philadelphia, PA: Elsevier Saunders,
2005:803–863.
Peters NS, Wit AL. Myocardial architecture and ventricular arrhythmogenesis. Circulation 1998;97:
1746–1754.
Vaquero M, Calvo D, Jalife J. Cardiac fibrillation: from ion channels to rotors in the human heart.
Heat Rhythm 2008;5:872–879.
Wagner CD, Persson PB. Chaos in the cardiovascular system: an update. Cardiovasc Res 1998;40:257–
264.
Weiss JN, Chen PS, Wu TJ, et al. Ventricular fibrillation. New insights into mechanisms. Ann NY Acad
Sci 2004;1015:122–132.
Wyse DG, Gersh BJ. Atrial fibrillation: a perspective. Thinking inside and outside the box.
Circulation 2004;109:3089–3095.
Zipes D, Jalife F. Cardiac electrophysiology. From cell to bedside. 5th ed. Philadelphia, PA: WB
Saunders, 2009.
Molecular Syndromes
Dokuparti MVN, Pamuru PR, Thakkar B, et al. Etiopathogenesis of arrhythmogenic right ventricular
cardiomyopathy. J Hum Genet 2005;50:373–381.
Fowler SJ, Prioi SG. Clinical spectrum of patients with a brugada ECG. Curr Opin Cardiol 2008;24:71–
81.
Hedley PL, Jørgensen P, Schlamowitz S, et al. The genetic basis of brugada syndrome: a mutation
update. Hum Mutat 2009;30:1256–1266.
Kontula K, Laitinen PJ, Lehtonen A, et al. Catecholaminergic polymorphic ventricular tachycardia:

recent mechanistic insights. Cardiovasc Res 2005;76:379–387.
Postma AV, Dekker LRC, Soufan AT, et al. Developmental and genetic aspects of atrial fibrillation.
Trends CV Med 2009;19:123–130.
Priori SG. Inherited arrhythmogenic diseases the complexity beyond monogenic disorders. Circ Res
2004;94:140–145.
Roden DM. Long QT syndrome. New Engl J Med 2008;358:169–176.
Zimmer T, Surber R. SCN5A channelopathies—an update on mutations and mechanisms. Prog Biophys
Mol Biol 2008;98:120–136.
For more complete descriptions of the clinical features of the topics discussed in this chapter,
readers are referred to the many textbooks of electrophysiology, cardiology, and
electrocardiography.
References
Antzelevitch C, Brugada P, Borggrefe M, et al. Brugada syndrome. Report of the second consensus
conference. Circulation 2005;111:659–670.
CAST Investigators (The Cardiac Arrhythmia Suppression Trial Investigators). Preliminary report:
effect of encainide and flecainide on mortality in a randomized trial of arrhythmia suppression after
myocardial infarction. New Eng J Med 1989;321:406–412.
P.487
CAST II Investigators (The Cardiac Arrhythmia Suppression Trial II Investigators). Effect of the anti-
arrhythmic agent moricizine on survival after myocardial infarction. New Eng J Med 1992;327: 227–
233.
Garrey WE. The nature of fibrillary contraction of the heart. Its relation to tissue mass and form. Am
J Physiol 1914;33:397–414.
Hardman RM, Forsythe ID. Ether-à-go-go-related gene K+ channels contribute to threshold

excitability of mouse auditory brainstem neurons. J Physiol (London) 2009;587:2487–2497.
Hondeghem L, Katzung BG. Time- and voltage-dependent interaction of interaction of

antiarrhythmic drugs with cardiac sodium channels. Biochim Biophys Acta 1977;472:373–398.
Jalife J, Moe GK. Excitation, conduction, and reflection of impulses in isolated bovine and canine
cardiac Purkinje fibers. Circ Res 1981;49:233–247.
Jarcho S. The concept of heart failure. From avicenna to albertini. Cambridge MA: Harvard
University Press, 1980.
Katz LN, Pick A. Clinical electrocardiography. Part I: The arrhythmias. Philadelphia, PA: Lea and
Febiger, 1956.
Knollmann BC, Roden DM. A genetic framework for improving arrhythmia therapy. Nature 2008;451:
929–936.
Langendorf R. Concealed A-V conduction: The effect of blocked impulses on the formation and
conduction of subsequent impulses. Am Heart J 1948;35:542–552.
Members of the Sicilian Gambit. New approaches to antiarrhythmic therapy. Emerging therapeutic
applications of the cell biology of cardiac arrhythmias. Circulation 2001;104:2865–2873;2990–2994.
Vaughan Williams EM. Classification of antiarrhythmic actions. In: Vaughan Williams EM, Campbell
TJ, ed. Handbook of experimental pharmacology. Antiarrhythmic drugs. Berlin: Springer-Verlag,
1989;89: 45–57.
> Table of Contents > Part Four - Pathophysiology > Chapter 17 - The Ischemic Heart
Chapter 17
The Ischemic Heart
The heart requires an uninterrupted supply of oxygen to meet its high energy needs, which can be
satisfied only by oxidative metabolism (Chapter 2). Coronary artery occlusion causes an almost immediate
loss of function and, within hours, death of the energy-starved cardiac myocytes. Ischemic myocardial
cell death, called myocardial infarction, is a major public health problem in developed countries.
Although this condition is often referred to as ischemic heart disease, the cause lies in the coronary
arteries, so that the myocardium is the victim and not the perpetrator. For this reason, the alternative
names coronary heart disease, arteriosclerotic heart disease, and atherosclerotic heart disease are
commonly used.
Conditions other than coronary occlusion can cause the heart to become energy-starved. These include
aortic stenosis and pulmonary hypertension, in which pressure overload increases oxygen requirements to
levels that exceed the amount that can be supplied by even a normal coronary circulation. Severe
anemia, by reducing the oxygen-carrying capacity of the blood, also causes the heart to become energy-
starved. Angina pectoris (literally strangling in the chest), the typical symptom when the heart's demand
for oxygen exceeds the amount supplied by the coronary circulation, is therefore not diagnostic of
coronary heart disease.
Coronary Occlusive Disease

Coronary atherosclerosis, by far the most common cause of coronary occlusion and myocardial infarction,
is the result of a chronic inflammatory process that evolves over many decades. A sudden decrease in
coronary flow usually occurs when endothelial damage leads to the formation of platelet plugs and
thrombi (clots) that occlude one or more major coronary arteries. These occlusive events, which generally
evolve over periods of hours or days, are most commonly initiated by disruption of a fibrous cap that
separates soft lipid-filled atherosclerotic lesions within the arterial wall from the lumen. This process,
called plaque rupture, exposes collagen and other thrombogenic elements in the atherosclerotic lesion to
the moving column of blood in the coronary artery, which triggers platelet aggregation and thrombus
formation that occlude the diseased vessel. Vasospasm, a less common cause of coronary occlusion, can
be caused when vasoconstrictor peptides are released from activated platelets in diseased arteries that
may lack obvious atherosclerotic lesions.
Total occlusion of a major coronary artery generally causes the entire thickness of the ventricular wall to
become ischemic, a condition called transmural ischemia. However, in a patient with well-developed
coronary artery collaterals, a similar occlusion may cause significant ischemia only in the endocardium,
where the balance between energy supply and energy demand is most precarious. This is because,
compared to the epicardium, the endocardium is less well perfused (see Chapter 1) and its wall stress is
higher (see Chapter 11). Increased myocardial oxygen
P.489
demand in a patient with a partial coronary artery occlusion can also cause the endocardium to become
ischemic; this is the basis for the ECG evidence of subendocardial ischemia in a positive stress test (see
later).
Angina Pectoris
Angina pectoris, the typical symptom caused when cardiac energy demand exceeds energy supply, is a
visceral discomfort that localizes poorly and is often difficult to describe. Most patients experience a
heaviness or squeezing in the left chest, but others note only a vague discomfort that may be mistaken
for indigestion. Angina typically radiates to the inner left arm, neck, or jaw, but may be referred
elsewhere; for example, to the site of an old back or shoulder injury or the socket of a lost tooth. Not all
patients experience typical symptoms, even when total coronary occlusion has caused a large myocardial
infarction. The substance or substances responsible for the chest discomfort remain unclear.
The symptoms of ischemic heart disease appear in patterns that depend on how they are provoked.
Demand angina, which as the name implies occurs when the symptoms are provoked by increased cardiac
work, as occurs during exercise, almost always forces patients to stop exertion and is typically relieved
promptly by rest, usually within a few minutes. The severity of demand angina can be quantified in terms
of the amount of effort needed to provoke symptoms, most precisely by a stress test. In supply angina,
which typically occurs when the patient is at rest, the imbalance between energy supply and energy
demand is caused by a decrease in supply, so that this pattern indicates that coronary flow has been
reduced by a new occlusion or vasospasm.
One of the most important characteristics of angina pectoris is its stability; that is, whether the
discomfort is staying the same, improving, or worsening. Stable angina can remain unchanged for years
because coronary atherosclerosis progresses slowly, often in a stepwise manner. Spontaneous
improvement in the symptoms of cardiac ischemia, which was common before procedures were developed
to revascularize the heart, occurs when blood flow to underperfused regions is increased by enlargement
of partially occluded coronary arteries, called vascular remodeling, and by the development of collateral
vessels. Unstable angina, where the symptom increases in severity, lasts longer, or is provoked by
decreasing effort, indicates that the underlying coronary disease has entered a dangerous period when
acceleration of the underlying vascular disease has increased the risk for a myocardial infarction and
sudden death.
Stable Angina
Symptoms in patients with stable angina pectoris typically follow the pattern of demand angina, in which
the chest discomfort is exacerbated by increased myocardial oxygen demand, for example, exertion or
emotional upset. Stable angina generally has a benign prognosis because the brief episodes of ischemia do
not damage the heart, but instead increase the ability of the myocardium to withstand a fall in blood flow
(see later). For this reason, the natural history of stable angina is determined largely by progression of
the disease in the coronary arteries. The key to improving prognosis in these patients is meticulous
management of the risk factors for atherosclerosis that include high cholesterol, diabetes, hypertension,
obesity, and elimination of smoking. When symptoms become troublesome, patients are usually treated
with β-adrenergic blockers and arteriolar vasodilators to reduce the heart's energy demand, nitrates
which decrease preload and dilate coronary collateral vessels, and drugs that
P.490
inhibit platelet aggregation and clotting in the diseased coronary arteries. Revascularization by coronary
artery bypass surgery or angioplasty usually alleviates symptoms but has little or no ability to prolong
survival.
Vasospastic Angina
Patients with coronary vasospasm typically experience angina when they are at rest, usually without a
clear predisposing cause. This syndrome, which is also called vasospastic angina, variant angina, or
Prinzmetal's angina, is dangerous because, unlike demand angina in which myocardial energy starvation
can be alleviated by rest, ischemia caused by vasospasm persists until the constricted coronary artery
dilates. In some patients, myocardial ischemia lasts long enough to cause a myocardial infarction; in
others, coronary vasospasm can trigger arrhythmias that may be fatal. The obvious goal of treatment is to
relieve the coronary vasoconstriction, which can be done with variable success using combinations of
nitrates, calcium channel blockers, and other vasodilators.
Acute Coronary Syndromes: Unstable Angina and Acute Myocardial

Infarction
Worsening of angina, called acute coronary syndrome or instability, usually follows one of three patterns.
The least severe, called unstable angina, occurs when the symptoms worsen and are provoked more
easily, but there is no evidence of myocardial cell death. The other two, where increasing angina is
accompanied by evidence of necrosis, are non–ST elevation myocardial infarction (NSTEMI) and ST
elevation myocardial infarction (STEMI). These were formerly called subendocardial, non–Q-wave, or non-
transmural myocardial infarction; and transmural or Q-wave myocardial infarction, respectively (see
later). Although all acute coronary syndromes result from increasing coronary occlusion, they differ in
pathophysiology and prognosis, and require different therapy.
Unstable Angina
The pathophysiology of angina and its relationship to energy starvation explains the clinical picture in
patients with unstable angina; the intensity of the symptoms increases, the amount of effort needed to
evoke the chest discomfort decreases, and angina can occur at rest. The appearance of unstable angina
usually indicates that the coronary occlusive disease is worsening but can also be caused when a
complicating condition, like anemia, impairs oxygen delivery to the heart. Unstable angina typically
follows a “stuttering” course because platelet plugs and thrombi that form in the damaged artery break
up or dissolve, and then reappear. The resulting impairment of blood flow is often exacerbated by
coronary vasospasm caused by vasoconstrictor substances that are released when platelets are activated
by the damaged endothelium.
Instability is generally initiated by disruption of an atherosclerotic plaque, where damage to the fibrous
cap in a previously stable coronary artery lesion exposes a thrombogenic surface that leads to arterial
occlusion. The appearance of unstable angina has ominous prognostic implications as it can herald a
myocardial infarction or sudden cardiac death. For this reason, unstable angina requires prompt medical
attention.
P.491
Acute Myocardial Infarction

Acute myocardial infarction occurs when coronary flow is reduced to levels so low as to cause myocardial
cell death. Most clinical infarctions involve the left ventricle; right ventricular infarction can occur, but is
less common. Because the endocardium is especially vulnerable to energy starvation (see earlier),
coronary occlusion causes endocardial myocytes to be severely injured and to die sooner than those in the
epicardium.
The criteria now used to classify patients with myocardial infarction depend on whether the ST segments
in ECG leads facing the infarction are depressed or elevated. The distinction between STEMI and NSTEMI is
determined largely by whether the full thickness of the ventricular wall or just the subendocardial layers
are injured or infarcted (see later).
Prompt restoration of the normal balance between energy delivery and energy utilization is central to
managing any of the acute coronary syndromes. Most important is to increase energy supply by addressing
the underlying disease in the coronary arteries. This can be accomplished using aspirin and other platelet
inhibitors, antithrombotic drugs, and thrombolytic agents. In patients with STEMI, the most effective
treatment is reperfusion of the ischemic myocardium by immediate (primary) angioplasty. Reducing
cardiac energy demands can also alleviate symptoms and improve prognosis, but this approach deals with
the consequences, rather than the cause, of the clinical syndrome. β-Adrenergic blockers administered
immediately after a coronary occlusion have important energy-sparing effects that are especially
beneficial in regions where perfusion is reduced, but not totally abolished. Vasodilators, such as nitrates
and calcium channel blockers, are useful because their ability to dilate arterioles that determine
peripheral vascular resistance reduces afterload and so decreases energy demand. Nitrates also dilate
veins, which has the added benefit of decreasing preload; these drugs also increase blood supply to
ischemic regions of the heart by dilating intracoronary collateral vessels.
Jeopardized Myocardium and “Ischemia at a Distance”

The consequences of coronary occlusion are influenced by collateral vessels that, by connecting large
epicardial coronary arteries, allow more than one major coronary artery to supply a given region of the
heart (Fig. 17-1). Collateral vessels generally develop when partial occlusion of a coronary artery reduces,
but does not totally interrupt, blood flow to a region of the heart. Because the collateral circulation
increases with advancing age, occlusion of a major coronary artery in young patients, who typically have
few collateral vessels, usually causes a large infarct, whereas the same occlusion in an older patient with
an extensive collateral circulation may cause only minor symptoms, and sometimes no symptoms at all.
The role of collaterals can be understood by considering a patient with a large collateral vessel
connecting a normal right coronary artery (RCA) to a partially occluded left anterior descending coronary
artery (LAD), which together provide the anterior wall of the left ventricle with a dual blood supply (Fig.
17-1A). Occlusion of the RCA in this setting not only leads to necrosis of the inferior wall, which had
received its entire blood supply from this vessel, but causes blood flow to the anterior wall to become
jeopardized because the latter has become entirely dependent on the partially occluded LAD for its entire
blood supply (Fig. 17-1B). Because the RCA occlusion has caused ischemia in a region (the anterior wall)
that is outside of its normal distribution, this situation is called ischemia at a distance.
P.492
Fig. 17-1: Ischemia at a distance. A: Anterior view of a heart where partial occlusion of the LAD has
stimulated the development of a collateral vessel linking a normal RCA to the occluded artery. The anterior
region of the left ventricle supplied by the partially occluded LAD can become ischemic when the energy
demands of the left ventricle are increased (light shading). B: A new total occlusion of the RCA causes
infarction of the inferior region of the left ventricle to which it had provided the only source of arterial blood
(black), and in addition decreases coronary flow to the anterior region that had previously received the dual
blood supply (cross-hatched). The new RCA occlusion, therefore, reduces the blood supply to the jeopardized
region that formerly had a dual blood supply, but which must now depend entirely on the partially occluded
LAD. In this way, RCA occlusion caused ischemia at a distance because ischemia appeared in a region (the
anterior wall) that is not normally supplied by the RCA. LAD: left anterior descending coronary artery; RCA,
right coronary artery; CIRC: circumflex coronary artery; RA: right atrium; LA: left atrium; RV right ventricle;
LV: left ventricle; PA: pulmonary artery; Ao: aorta.
Initial Consequences of Coronary Artery Occlusion

Patients who sustain an acute myocardial infarction generally experience severe angina, loss of pump
function, and arrhythmias. The initial hemodynamic consequences depend largely on the amount of the
left ventricle that has been infarcted, but the severity of the arrhythmias and occurrence of sudden death
in patients with an acute myocardial infarction do not correlate well with infarct size.
Two reflexes influence the clinical picture in acute myocardial infarction. The hemodynamic defense
reaction (see Chapter 8), whose major effects in this setting are mediated largely by sympathetic
stimulation, is activated because these patients are generally in pain and terrified; this response becomes
even more powerful when left ventricular damage causes blood pressure to fall. Although the
neurohumoral response helps maintain blood pressure, the ability of β-adrenergic stimulation to increase
heart rate and contractility worsens energy starvation in the ischemic myocardium, as do arteriolar
vasoconstriction and increased afterload caused by α-adrenergic stimulation. Sympathetic stimulation can
also cause arrhythmias and sudden death. The second reflex that is activated in patients with inferior and
posterior left ventricular infarction is the von Bezold–Jarisch reflex, a powerful vagal response that lowers
blood pressure, slows the sinus node pacemaker, and can lead to AV block.
P.493
Hypoxia, like ischemia, causes energy starvation, but the consequences differ because coronary occlusion,
in addition to interrupting oxygen supply to the heart, prevents the removal of metabolites and, by
reducing the pressure within the coronary arteries supplying the left ventricle, attenuates the “garden
hose effect” (see later).
Early Pump Failure

Interruption of coronary flow to the mammalian heart is followed within a few seconds by a decrease in
contractility and profound impairment of ventricular filling (Fig. 17-2). The decrease in contractility,
which is accompanied by abbreviation of systole, causes the ischemic region to bulge outward during
systole because it cannot overcome the intraventricular pressure generated by the normally perfused
myocardium. The resulting decrease in ejection also impairs filling because acute ventricular dilatation
within the non-distensible pericardium reduces the ability of the heart to fill. This negative lusitropic
effect, which is similar to pericardial tamponade, can be of greater hemodynamic importance than the
negative inotropic effect.
The Garden Hose Effect

It was once believed that lack of adenosine triphosphate (ATP) was the major cause of the early pump
failure that follows a coronary artery occlusion. However, the initial decrease in myocardial contractility
precedes a significant fall in ATP content (Fig. 17-2). The most likely explanation for the rapid impairment
in ejection is attenuation of the distending effect of intracoronary pressure (the garden hose effect, see
Chapter 10).
Fig. 17-2: Time courses of the decline in contractility, adenosine triphosphate content, and
phosphocreatine after complete interruption of coronary flow. (Based on data from Williamson,
1966.)
P.494
Metabolic Abnormalities
Oxygen tension within the myocardium falls almost to zero within a minute after complete cessation of
blood flow. This reflects the very high affinity of the respiratory chain for oxygen, which allows the
ischemic myocardial cells to consume all of the available oxygen within a few minutes after the
myocardium loses its blood supply; as a result, oxidative phosphorylation comes to a complete halt.
Anaerobic energy production is increased in part by decreases in ATP and glucose-6-phosphate levels,
increased adenosine diphosphate (ADP), adenosine monophosphate (AMP), and Pi levels (see Chapter 2),
and by the accelerated glycogen breakdown, glucose uptake, and glycolysis caused by sympathetic
stimulation. Sympathetic stimulation also stimulates fatty acid release from triglycerides, but because
these lipids cannot be oxidized, fatty acids and their derivatives accumulate in the ischemic heart (see
later).
Although the rate of anaerobic glycolysis initially increases in the ischemic heart, this response cannot be
sustained because NAD+ levels rapidly decrease to levels below those needed to reduce glyceraldehyde-3-
phosphate. In addition, the ischemic heart rapidly becomes acidotic (see later); this is due initially to the
accumulation of lactate, which cannot be removed because of the decreased coronary flow, and to the
release of protons when ATP is hydrolyzed.
Decreased ATP
ATP content falls rapidly after coronary occlusion, but does not reach levels low enough to deprive the
substrate-binding sites of the contractile proteins and ion pumps of their supply of energy. This is because
the normal cytosolic ATP concentration is 5 to 10 mM, whereas most substrate-binding sites are saturated
at ATP concentrations less than 1 µM.
Attenuation of the allosteric effects of high ATP concentrations increases diastolic stiffness by inhibiting
the dissociation of actin and myosin (see Chapter 4), and reduce contractility and impair relaxation by
slowing ion pumps, ion exchangers, and passive ion fluxes through membrane channels (see Chapters 7
and 10). The sarcoplasmic reticulum calcium pump and plasma membrane sodium pump are also inhibited
by a reduction in the allosteric effect of ATP (see Chapter 7). The former impairs relaxation, as does
sodium pump inhibition, which causes a rise in cytosolic sodium that impairs relaxation by reducing
calcium efflux via the Na/Ca exchanger. Sodium pump inhibition also reduces potassium influx and so
decreases resting membrane potential; the latter, by inactivating sodium channel opening, slows
conduction and provides a substrate for arrhythmias.
ATP depletion also decreases the free energy made available by hydrolysis of the terminal phosphate bond
in ATP, which is proportional to the ATP/ADP ratio. Because even a slight fall in ATP concentration causes
a disproportional increase in ADP concentration, energy starvation can reduce the free energy of ATP
hydrolysis to levels that slow both the calcium pump of the sarcoplasmic reticulum and cross-bridge
cycling (see Chapter 10).
Acidosis
Coronary occlusion causes a rapid fall in the pH within cardiac myocytes. Acidification is due largely to
lactate formation and hydrolysis of ATP, which release weak acids (inorganic phosphate and lactate) that
liberate hydrogen ions at the pH within the cytosol of cardiac myocytes. Some of these protons are
absorbed when phosphocreatine hydrolysis releases creatine, which is a weak base, but the net effect of
ischemia is rapid acidification of the heart.
P.495
Acidosis inhibits many important reactions in the heart; these include glycolysis (see Chapter 2),
contractile protein interactions (see Chapter 4), most of the calcium fluxes that participate in excitation-
contraction coupling and relaxation (see Chapter 7), and the cycling of plasma membrane ion channels
responsible for cardiac action potentials (see Chapter 15). Another reason that acidosis causes
arrhythmias is that protons increase internal electrical resistance by closing gap junctions in the
intercalated disc, which slows conduction (see Chapter 13). Protons also interfere with many of the
reactions involved in contraction and relaxation by competing with calcium for binding sites on troponin,
ion channels, pumps, and exchangers (see Chapters 4 and 7).
Potassium
The large amounts of phosphate released by hydrolysis of ATP and phosphocreatine in the ischemic heart
reduce the calcium sensitivity of the contractile proteins and may form calcium phosphate complexes
that trap calcium within the sarcoplasmic reticulum. Because phosphate, along with lactate, readily
crosses the plasma membrane, anions released in ischemic cells pour into the extracellular space. To
maintain electrical neutrality, the efflux of lactate and phosphate anions is accompanied by the efflux of
potassium, which is the major intracellular cation. The latter depolarizes the ischemic myocardial cells by
reducing the ratio [K+]i/[K+]o, which decreases the Nernst potential for potassium (see Chapter 14). The
resulting decrease in resting potential has several effects, all of them bad; these include sodium channel
inactivation, which slows conduction and depresses excitability (see Chapter 14), and a negative inotropic
effect caused by calcium channel inactivation that contributes to the early pump failure. Membrane
depolarization caused by potassium efflux is among the most dangerous effects of coronary artery
occlusion because the inhomogeneity of the response generates injury currents that are a major cause of
lethal arrhythmias (see later).
Mechanical Abnormalities
The loss of pump function in an ischemic or infarcted ventricle is caused by regional abnormalities, as
opposed to global loss of function, because the cause is occlusion of an artery supplying only a portion of
the heart. The resulting wall motion abnormalities are usually described visually. Mildly impaired
contraction of part of the left ventricle causes asyneresis (reduced inward movement) or asynchrony
(disturbed temporal sequence of contraction), while greater loss of function causes akinesis (failure of a
damaged segment to participate in ejection). The most severe regional wall motion abnormality is
dyskinesis, where the damaged segment bulges outward during systole. A large dyskinetic region in a
patient who has had a large infarction is commonly referred to as an aneurysm. In all cases, the
combination of reduced stroke volume and increased end-diastolic volume causes a fall in ejection
fraction that is directly proportional to the extent of left ventricular damage (see Chapter 12).
Electrophysiological Abnormalities in the Ischemic Heart

Ischemia causes characteristic abnormalities in the ECG, arrhythmias that are often dangerous, and long-
term molecular changes in cardiac ion channels. All of these effects can be influenced by the
neurohumoral response and the von Bezold–Jarisch reflex that can follow an acute myocardial infarction.
P.496
Injury Currents and ST Segment Shifts

Resting depolarization caused by coronary occlusion establishes potential differences that allow currents
to flow between normally perfused and ischemic regions of the heart. These currents, called injury
currents, cause ST segment shifts that distinguish between transmural ischemia, which causes ST
elevation and subendocardial ischemia, which causes ST segment depression.
To understand the mechanisms by which injury currents cause ST segment shifts, it must be remembered
that there is no way to establish a baseline in the clinical ECG. For this reason, ST segment deviations
cannot be distinguished from shifts in the TP segment. By convention, the baseline is assumed to be the
TP segment (see Chapter 15), so that the diastolic injury currents that displace the TP segment are
interpreted as ST segment shifts: TP depression is read as ST elevation, and TP elevation as ST depression
(Fig. 17-3).
St Segment Elevation in Transmural Ischemia

The major cause of ST segment elevation in leads facing a region of transmural injury in a patient with a
STEMI is TP depression caused by a diastolic injury current. Coronary vasospasm, which generally causes
transmural ischemia, also causes ST elevation. In both, the transmural ischemia reduces the
electropositivity at the surface of the ischemic regions during diastole; as a result, leads that face the
ischemic myocardium are in a region of electronegativity. According to electrocardiographic convention
(see Chapter 15) this electronegativity causes a downward shift in the recording that depresses the TP
segment, but because the TP segment is viewed as the baseline this shift is interpreted as ST elevation
(Fig. 17-3A). During systole, when the entire heart is depolarized, the potential difference between the
normally perfused and ischemic regions is decreased; this reduces the injury current and so normalizes
the TP segment (and therefore the ST segment). Other causes for ST elevation in transmural ischemia
include potential differences caused by action potential abbreviation, slow conduction in the ischemic
region, and abnormal motion of the ischemic wall of the ventricle.
The ST segment elevation in patients with a STEMI usually disappears 24 to 48 hours after coronary
occlusion. This can result from two mechanisms: when reperfusion of the ischemic myocardium restores
the normal resting potential, or when the ischemic myocytes become electrically silent. The latter can
occur when acidosis and calcium overload increase internal resistance by closing the connexon channels in
the gap junctions (see Chapter 13), or when these cells die.
St Segment Depression in Subendocardial Ischemia

Subendocardial ischemia causes ST segment depression because a layer of normally perfused myocardium
separates the partially depolarized ischemic endocardium from a lead placed over the epicardial surface
of the heart (Fig. 17-3B). Because the normally perfused epicardial cells are more positively charged
during diastole than cells in the ischemic subendocardium, an ECG lead facing a region of subendocardial
ischemia records TP segment elevation that, for the reasons given earlier, is interpreted as ST segment
depression.
ST segment depression is seen in both NSTEMI and “demand ischemia” because energy starvation is most
severe in the endocardium, where energy utilization is highest and blood supply most precarious. The ST
segment depressions that appear during a positive stress test also occur when exertion causes
subendocardial ischemia.
P.497
Fig. 17-3: Injury currents caused by transmural and subendocardial ischemia. The ECG (right)
records the potential difference between an electrode (E) that faces the surface of the ischemic
region and a central terminal (V) that is defined as “zero.” The ischemic region is shaded, and the
potentials at the surface of the heart during diastole are shown. A: Transmural ischemia, which
depolarizes the ischemic myocardium, causes a diastolic injury current in which the lead facing the
ischemic region is in an area of electronegativity. This causes an electrical vector directed away
from the lead that, by convention, inscribes a downward deflection and so causes TP segment
depression. During systole, when the normal myocardium also becomes depolarized, the injury
current is reduced, which returns the TP segment toward baseline. By convention, TP depression is
interpreted as ST segment elevation. B: Subendocardial ischemia causes a diastolic injury current
in which the lead facing the ischemic region is in an area of electropositivity, which causes an
electrical vector directed toward the lead. By convention, this lead inscribes an upward deflection
that causes TP segment elevation that is interpreted as ST segment depression.
P.498
Fig. 17-4: Abnormal Q wave. A lead facing a transmural infarction (shaded), which is electrically
“silent,” records a downward initial QRS deflection because most of the electrical vectors
generated by depolarization of the non-infarcted myocardium are directed away from this lead.
Abnormal Q Waves and Myocardial Cell Death

The appearance of abnormal Q waves is a useful electrocardiographic marker for transmural infarction in
the left ventricle (Fig. 17-4). These initial downward deflections appear in leads facing the infarct
because the ischemic region, which had normally transmitted a wave of depolarization toward the lead,
becomes electrically silent. A useful way to understand abnormal Q waves is to visualize the electrically
silent infarcted myocardium as a window through which a lead can look into the left ventricular cavity,
where QS complexes are normally recorded because ventricular depolarization begins in the endocardium.
Because all of the electrical vectors in the normal ventricular wall are directed away from the cavity (see
Chapter 15), an electrode placed within the ventricle records only a downward deflection.
During recovery after a small infarction, Q waves often diminish in size and can disappear when the
necrotic area shrinks and forms a scar. If the infarction is large, however, Q waves generally persist. A
useful, but not very precise, index of the extent of left ventricular damage is the number of leads with
abnormal Q waves.
Although abnormal Q waves suggest transmural infarction in patients with acute myocardial infarction,
and the absence of Q waves suggests subendocardial infarction, pathological findings do not always
confirm these anatomic distinctions. This is why these patterns have come to be called ST elevation
myocardial infarction and non-ST elevation myocardial infarction. This distinction is important because
the benefits of reperfusion generally outweigh the risks in STEMI but not NSTEMI, which means that these
two syndromes must be treated differently. Abnormal Q waves are not diagnostic of infarction because
they also occur when an injury causes scar tissue to replace viable myocardium, or when a tumor invades
the wall of the ventricle.
St Segment and T Wave “Evolution”

A key feature of the ECG changes in myocardial infarction are dynamic changes in the ST segments and T
waves, called “evolution,” that usually continue for days, weeks, and sometimes several months after the
acute event. The initial abnormality, which appears within a few minutes after the onset of symptoms, is
a marked increase in the amplitude of the T waves in leads facing the infarction. These tall T waves,
called hyperacute T waves, often merge with the elevated ST segments in a pattern that resembles a
tombstone (Fig. 17-5). The most likely cause of the initial increase in T wave amplitude is the rapid
increase in extracellular potassium concentration caused by potassium efflux from ischemic cells (see
earlier).
P.499
Fig. 17-5: Hyperacute T wave (lead V4). The first record, obtained 2 hours after the onset of
severe chest pain in a patient who subsequently received streptokinase, shows a large hyperacute T
wave that, along with the markedly elevated ST segment, resembles a tombstone. The second
record was obtained 2½ hours after the clot obstructing the coronary artery had been lysed and the
patient had become pain-free.
Hyperacute T waves, which usually last only a few hours, disappear if prompt opening of the infarct
artery reperfuses the ischemic myocardium; however, they also disappear if the heart remains ischemic
and the ischemic myocardial cells die. The ST segment elevations also rapidly return to normal if the
ischemic myocardium is reperfused promptly, but when the infarct artery remains occluded, ST segment
normalization can require several days. If a ventricular aneurysm develops, ST segment elevations can
become permanent.
T wave evolution in ischemic heart disease is not a marker for cell death, but instead is caused by
electrical remodeling in regions of the heart that remain viable after an episode of severe ischemia.
Slowly deepening T wave inversions in leads facing the ischemic region (Fig. 17-6) can progress when
patients are improving clinically (Fig. 17-7). Similar patterns of T wave evolution, called “T wave
memory” (see Chapter 14), are also seen after an episode of rapid tachycardia (post-tachycardia T wave
syndrome), transient bundle branch block, and pacing; these stress-induced T wave changes, like T wave
evolution after infarction, can progress for several weeks after the initiating event.
P.500
Fig. 17-6: Evolution of an acute inferior myocardial infarction (lead aVF). The first record was obtained in
the emergency room while the patient was experiencing chest pain. The subsequent records, taken 1 day
apart, show typical evolution. On the second day, resolution of the ST segment elevation demonstrates that
the injury current had disappeared, while significant deepening of the Q wave indicates that the inferior wall
of the left ventricle had become infarcted. The symmetrically inverted (“coved”) T wave is typical of T wave
evolution. The infarct artery was not opened because neither thrombolytic therapy nor primary angioplasty
was available at the time this patient was seen.
Fig. 17-7: Drawings of ECG evolution after a myocardial infarction (MI). ST segment elevation appears within
minutes after a coronary artery occlusion; Q waves appear generally within a few hours. The time course of
the subsequent T wave changes is highly variable. (Modified from Katz, 1946.)
Electrocardiographic Localization of a Myocardial Infarction

The leads that record abnormal Q waves and ST segment shifts are useful in localizing an infarction (Table
17-1), and so help to identify the infarct artery. Occlusion of the left anterior descending coronary artery
is usually responsible for anterior and anteroseptal infarctions, while occlusion of the right coronary or
circumflex coronary arteries generally causes inferior and posterior infarctions. However,
electrocardiographic criteria for infarct localization often correlate poorly with pathological findings. An
especially important distinction based on the location of
P.501
an infarct is that between anterior infarctions, which tend to be larger, and inferoposterior infarctions,
which are frequently associated with the von Bezold–Jarisch reflex (see later).
Table 17-1 Electrocardiographic Localization of Myocardial Infarctions
Localization Leads in Which Abnormal Q Waves Are Found
ANTERIOR
Anteroseptal V1, V 2
Anterior V2, V 3, V 4
Anterolateral I, aVL, (V4), V5, V6
Extensive anterior I, aVL, V1–V6
INFEROPOSTERIOR
Inferior II, III, aVF
Posterior R in V a
1
Inferolateral II, III, aVF, (V5), V6
Posterolateral R in V a, V (V )
1 6 5
Inferoposterior R in V a, II, III, aVF

1
aThe abnormally tall R wave in lead V would be recorded as a Q wave in posterior leads
1
placed on the patient's back. The latter, called V7 and V8, are rarely used so that the
diagnosis of posterior infarction is usually based on the finding of an abnormally tall initial R
wave in lead V1, which is equivalent to a Q wave recorded in leads V7 and V8.
Arrhythmias
Sudden death, a frequent occurrence in patients following a coronary occlusion, can be caused by several
mechanisms, including cardiac rupture, but the most common mechanism is an arrhythmia. The severity
of these arrhythmias is not closely correlated with the extent of ischemia, and it is clear that patients
with minimal cardiac damage can die suddenly within a few minutes or hours after an acute myocardial
infarction. Most arrhythmias are the direct result of myocardial ischemia, but reflexes often increase the
risk of sudden death during the first few days after an infarction. Norepinephrine released at sympathetic
nerve endings by the hemodynamic defense reaction (see Chapter 8) increases the risk of ventricular
fibrillation by increasing cytosolic calcium, which can cause transient depolarizations, and disorganizing
conduction. The von Bezold–Jarisch reflex, which slows the heart and can cause third-degree AV block by
triggering a powerful vagal response, is activated when ischemia stimulates receptors on the inferior and
posterior walls of the left ventricle.
Arrhythmias that occur weeks, months, and even years after an acute infarction are generally initiated by
reentry in the scarred ventricle (see Chapter 16) or when new ischemic events are caused by progression
of the underlying coronary artery disease. Because of these changes in the arrhythmogenic mechanisms,
use of drugs to prevent sudden death in these patients is like chasing a moving target. This became
apparent many years ago, when patients resuscitated from ventricular fibrillation in the first hours after a
myocardial infarction were found to have only a small increase in risk for a late arrhythmic death; this
indicates that different pathophysiological mechanisms are responsible for early and late arrhythmias.
Arrhythmias Associated with Acute Myocardial Infarction

Potentially lethal arrhythmias begin to appear within a minute after total occlusion of a coronary artery.
Most early tachyarrhythmias are caused when injury currents are generated by diastolic potential
differences between ischemic and non-ischemic regions of the heart. These are sometimes followed by
accelerated idioventricular rhythms that are usually benign. Calcium overload-induced triggered
arrhythmias can appear if the ischemic myocardium is reperfused. Bradyarrhythmias that appear
immediately after an acute inferior or posterior infarct are often caused by the von Bezold–Jarisch reflex,
in which case they are almost always transient. However, bradycardias that result from structural damage
to the AV conduction system usually require an electronic pacemaker.
Tachyarrhythmias
Injury currents between ischemic and well-perfused regions are the most important cause of sudden
death in the first minutes after a coronary artery becomes occluded. These potential differences occur
when the depolarizing effect of potassium efflux from the ischemic myocytes (see earlier) establishes
heterogeneities in resting potential (see Chapter 16). Resting depolarization also inactivates sodium
channels, which results in small, slowly conducting action potentials that provide the substrate for
decremental conduction and unidirectional block (see Chapter 16). The predisposition to reentrant
tachyarrhythmias is increased by local changes in
P.502
action potential duration and refractoriness, both shortening and prolongation, that are common after
coronary occlusion. Reentry is also favored when acidosis and calcium overload slow conduction by closing
gap junction channels in the intercalated discs.
These initial arrhythmogenic mechanisms subside in the first hours after coronary occlusion, when the
ischemic cells die, so that the risk of sudden death decreases rapidly after the acute infarction. However,
reperfusion initiates triggered activity that can generate dangerous arrhythmias following successful
thrombolytic therapy or primary angioplasty, or when a clot in an infarct artery lyses spontaneously. Brief
episodes of an idioventricular rhythm at rates between 60 and 100 are sometimes seen 18 to 36 hours
after an acute myocardial infarction; these benign arrhythmias are caused when accelerated pacemaker
activity in ischemic His-Purkinje fibers is unmasked by slowing of the sinus rate, usually after these
patients fall asleep.
Patients with healed myocardial infarctions are also at risk for tachyarrhythmias, but unlike the early
arrhythmias, these are due largely to heterogeneities caused by scarring and recurrent ischemia in the
damaged heart. The incidence of these late arrhythmias is increased in patients with left ventricular
dysfunction or heart failure (see Chapter 18). Atrial fibrillation can occur in patients with ischemic heart
disease, but this arrhythmia is not common unless left ventricular failure has caused atrial dilatation or
there has been an atrial infarct.
Bradyarrhythmias
Both functional and anatomical mechanisms cause bradyarrhythmias in patients with ischemic heart
disease.
Functional (Reflex) Bradyarrhythmias

The von Bezold–Jarisch reflex is activated when ischemia stretches receptors in the inferior and posterior
walls of the left ventricle. The hallmark of this reflex, which triggers a powerful parasympathetic
response and inhibits sympathetic activity, is the appearance of sinus bradycardia that is often
accompanied by Mobitz I second-degree AV block (see Chapter 16). The von Bezold–Jarisch reflex also
causes hypotension by dilating peripheral resistance vessels, and can evoke a visceral response that leads
to nausea and vomiting. Patients with this reflex can appear to be near death in the first hours after what
may be only a small infarction, but improvement is usually rapid because the reflex soon ends, and
patients often make an excellent recovery.
Appreciation of the importance of the von Bezold–Jarisch reflex (Costantin, 1963) led to a significant
reduction in the mortality of patients with inferoposterior infarction; the once routine use of
vasoconstrictors and sympathomimetic drugs to treat hypotension caused by reflex vasodilatation was
quickly abandoned because these drugs are dangerous, and often ineffective. Instead, volume repletion
can safely return blood pressure to normal, while atropine, a muscarinic blocker, usually restores SA and
AV node functions without the hazards associated with β-adrenergic stimulation. Because this reflex is
transient, and ends when the ischemic tissue either dies or is reperfused, if an electronic pacemaker is
required it is usually needed only temporarily.
Bradyarrhythmias Caused by Structural Abnormalities

Sinus bradycardia and SA block after an acute myocardial infarction can occur when occlusion of the SA
node artery causes infarction of the sinoatrial node. Because the SA node artery is usually a branch of the
right coronary artery (see Chapter 1), these arrhythmias most commonly accompany inferoposterior
infarction. More dangerous bradyarrhythmias result from
P.503
structural damage to the AV conduction system, which can cause Mobitz II second-degree AV block (see
Chapter 16). The appearance of the latter implies a high risk of progression to irreversible third-degree AV
block, and so is an indication for a permanent electronic pacemaker. Because the AV bundle and parts of
the bundle branches run in the interventricular septum, where they receive a dual blood supply from the
left anterior descending and posterior descending coronary arteries (see Chapter 1), infarctions that
damage these conducting structures indicate that the patient has multi-vessel coronary artery disease.
For this reason, conduction block in the His-Purkinje system, including new bundle branch block,
identifies patients with large infarctions who, if they recover, are likely to develop heart failure.
Cell Death
Myocardial cell death begins 15 to 40 minutes after the heart's blood supply is cut off completely and,
about 6 hours later, few viable cells remain in a totally ischemic region. This progression can be viewed as
a wave of necrosis that begins in the endocardium, where energy requirements are greatest, and spreads
outward through the wall of the left ventricle toward the epicardium. The timetable depends on
collateral flow, and so is slower in patients with a well-developed collateral circulation.
The histological appearance of the infarcted myocardium depends on whether or not the tissue has been
reperfused (Fig. 17-8). Regions where blood flow to the dying myocytes has not been restored form pale,
acellular infarcts by a process sometimes called mummification, where the dying cells undergo autolysis.
A much more violent process takes place when irreversibly damaged ischemic cells are reperfused, which
causes a hemorrhagic infarct. Under these conditions, uncontrolled calcium entry causes contraction-band
necrosis, where hypercontracted cardiac myocytes literally tear themselves apart. Because this process
attracts inflammatory cells and leads to the formation of a firm scar, late reperfusion has some benefits
even when injured cells cannot be salvaged.
Infarct Extension, Expansion, and Rupture

Infarct extension is caused when a decrease in the blood supply to a region of jeopardized myocardium
increases the amount of ischemic myocardium, whereas infarct expansion occurs when myofilament
slippage within an infarct causes the wall of the left ventricle to dilate. Cardiac rupture, which is often
preceded by infarct expansion, can occur in the first few days after an acute myocardial infarction.
Rupture, which occurs when dissection of blood under high pressure separates the layers of the left
ventricular wall, can cause tamponade when blood enters the pericardial sac, or a ventricular septal
defect when the rupture is through the interventricular septum. Separation of the head of a papillary
muscle from the ventricular wall, which causes acute mitral insufficiency, is another form of rupture.
Necrosis
The hallmark of necrosis is plasma membrane damage, which allows intracellular proteins, such as
transaminases, creatine phosphokinase, and troponin components, to leak into the bloodstream. This
explains why increased circulating levels of these molecules are useful in the diagnosis of myocardial
infarction. Many factors contribute to membrane damage, including free radicals, fatty acids, increased
lipase and protease activity, osmotic stress, and calcium overload.
P.504
Fig. 17-8: Histology of acute myocardial infarction. A: Pale acellular infarction, approximately 12 hours old,
in a patient whose heart was not reperfused. Inflammatory cells are absent and the myocytes, which are
relaxed, appear to be wavy B: Reperfused infarction showing hypercontracted myocytes surrounded by
inflammatory cells and erythrocytes. The two arrows point to contraction bands. C: Another region of the
heart shown in B; the transverse striations are contraction bands. D: Enlargement of the area in B that is
enclosed by the rectangle. H & E stain. Scale bar = 100 µm. (Photomicrographs provided by Drs. Margaret A.
French and Nora R. Ratcliffe.)
In the normal heart, free radicals generated when electrons pass along the respiratory chain are
neutralized when they are transferred to molecular oxygen and combined with protons to form water (see
Chapter 2). In the ischemic heart, however, oxygen lack leads to the release of highly reactive oxygen
free radicals by preventing electrons from being used to form water. ATP hydrolysis provides another
source of free radicals when ADP is converted to ATP and AMP by adenylyl kinase (see Chapter 2). The
AMP is dephosphorylated to form adenosine, which is deaminated to yield inosine that is converted first to
hypoxanthine, and then to xanthine in a reaction that generates superoxide radicals. Free radicals can
also be generated in the ischemic heart by catecholamine breakdown, arachidonic acid metabolism, and
the catalytic activity of lipases and proteases. Inflammatory cells that are attracted to damaged regions
of the heart release cytokines that also generate free radicals.
P.505
Amphipathic compounds that accumulate in energy-starved hearts have detergent effects that can alter
membrane structure and function. These include long chain fatty acids, their CoA and carnitine
derivatives, and lysophosphatides. It is not clear, however, whether enough of these substances
accumulate in the ischemic myocardium to cause irreversible damage. Osmotic overload can contribute to
plasma membrane rupture when large numbers of small molecules appear in the cytosol of ischemic
myocardial cells; these include ADP and Pi, generated by ATP hydrolysis, creatine and Pi formed from
phosphocreatine, and glucose-1-phosphate released by glycogenolysis.
Calcium overload, an important consequence of energy starvation, inhibits oxidative phosphorylation by

causing calcium to accumulate in mitochondria (see Chapter 2). Calcium overload also activates many
energy-consuming reactions (see Chapter 7), most important of which are contractile protein interactions
that can cause cardiac myocytes to tear themselves apart; this is a major cause of reperfusion injury, and
explains the appearance of the hypercontracted state called contraction band necrosis (see earlier).
A related cause of calcium overload is the calcium paradox, in which reintroduction of calcium into the
fluid surrounding myocytes that had been exposed to very low extracellular calcium concentration, leads
to contracture and cell death. The calcium paradox can result from plasma membrane damage that
follows prolonged exposure to low extracellular calcium, and from excessive calcium uptake that occurs
when this cation enters the cytosol in exchange for the large amount of sodium that had accumulated in
the cells during the time that extracellular calcium was low.
Apoptosis
Although necrosis is the major cause of myocardial cell death after prolonged ischemia, activation of
apoptotic pathways (see Chapter 9) also plays an important role in both ischemic and reperfusion injury.
Stimuli that can cause programmed cell death in ischemic hearts include cytochrome C release into the
cytosol by damaged mitochondria, opening of mitochondrial permeability transition pores, calcium-
activated cysteine proteases called calpains, and activation of matrix metalloproteinases.
Autophagy
This form of myocyte death, in which damaged cell components are sequestered and then digested in
organelles called autophagosomes, allows stressed cells to reuse cellular components (see Chapter 10).
The number of autophagosomes in the heart can be increased by ischemia and, to an even greater extent,
by reperfusion. Stimuli that activate autophagy in ischemic hearts include energy starvation, reactive
oxygen and nitrogen species, and increased mitochondrial pore formation. However, the importance of
autophagy in causing cell death in the infarcted heart is not clear.
Protection of the Ischemic Myocardium and Infarct Size Reduction

Ischemic damage to the heart can be minimized by slowing the rate of energy utilization by cardiac
myocytes. During open heart surgery, when it is often impractical to maintain coronary flow, energy
utilization is usually reduced by cooling the heart after it has been arrested in diastole with a cardioplegic
solution. The latter generally contains potassium, which stops the heart in
P.506
diastole, along with drugs that inhibit calcium entry, metabolic substrates, antioxidants, and other
cardioprotective substances. These measures are generally ineffective after a coronary occlusion,
because the hearts cannot be cooled and cardioprotective substances injected into the bloodstream have
limited access to the ischemic myocardium distal to an occluded coronary artery. Although there is little
or no “border zone” between ischemic and perfused myocardium in the human heart (Hirzel et al., 1977),
collateral vessels can deliver protective agents to jeopardized myocardium in some of these patients.
Attempts to preserve ischemic myocardium in patients who have had a coronary artery occlusion have
met with only limited success (Downey and Cohen, 2009; Ludman et al., 2010), so that the most effective
way to reduce infarct size is to open the occluded artery with thrombolytic agents or primary angioplasty.
Preconditioning, Postconditioning, Stunning, and Hibernation
Most of the myocardium in patients with chronic ischemic heart disease is either adequately perfused, or
has been infarcted and replaced by scar. In some patients, however, adaptive responses called
preconditioning, stunning, and hibernation, can be activated when perfusion is impaired severely enough
to stimulate proliferative signaling, but not cause myocardial cell death. Preconditioning, which occurs
after episodes of ischemia that are too brief to kill myocardial cells, increases the resistance of the
myocardium to subsequent more prolonged reductions of coronary flow. Stunning is a state of depressed
mechanical function that can follow short periods of ischemia, while hibernation describes poorly
functioning regions of viable, but chronically underperfused myocardium that can regain the ability to
contract when coronary flow is reestablished.
Preconditioning and Postconditioning

Preconditioning, in which transient ischemia delays the onset of cell death after subsequent more
prolonged episodes of ischemia, can be caused by episodes of ischemia that are too brief to damage the
myocardium irreversibly. Unlike cardioprotection, where the benefit depends on the presence of a
protective compound when the heart is ischemic, the resistance to injury that occurs when ischemia and
reperfusion cause preconditioning is long-lasting and remains after the protective intervention has
dissipated. Like electrical remodeling and cardiac memory (see Chapter 12) and T wave evolution (see
earlier), preconditioning involves stress-induced changes in the composition of the myocardium.
The protective effects of preconditioning occur in two phases: classical or early preconditioning, which
appears within minutes, and a delayed or late process. The latter, sometimes called the second widow of
protection (or SWOP), begins after ∼24 hours and lasts ∼3 days.
Classical preconditioning can be initiated when adenosine, norepinephrine, bradykinin, opioids, and other
extracellular messengers bind to G-protein coupled receptors that activate Gi-mediated responses, and by
non-receptor-mediated responses to free radicals, calcium, hypothermia, stretch, and other stimuli. Most,
if not all, of these signals exert their preconditioning effects by activating protein kinases, including
protein kinase C, MAP kinases, tyrosine kinases, and PI3-kinases (see Chapter 9). Delayed preconditioning
is also complex; triggers include adenosine, bradykinin, nitric oxide, prostanoids, opioids, and other
substances that activate signal transduction systems that are mediated by protein kinase C, tyrosine
kinases, and MAP kinases. Effector mechanisms include increased expression of heat shock proteins,
antioxidant enzymes, and mitochondrial iK.ATP channels.
P.507
Postconditioning, which describes mechanisms that can attenuate the detrimental effects following
reperfusion, is mediated by functional and proliferative signaling pathways that are similar to those which
mediate preconditioning.
Stunning and Hibernation

Stunning refers to a state of depressed contractile function that, like preconditioning, follows brief
episodes of severe ischemia. This negative inotropic response, which does not involve cell death and so is
completely reversible, can last as long as several weeks. The reversibility distinguishes stunning from
infarction, where function cannot be recovered. Hibernation is loss of contractile function in a heart that
has been chronically underperfused, but received enough blood to maintain its viability. The result is a
wall motion abnormality that can be difficult to distinguish from infarction. Heart failure caused by
hibernation is uncommon, but hibernation is important because revascularization can sometimes restore
virtually normal cardiac function in patients with chronic heart failure whose prognosis had been viewed
as hopeless.
The mechanisms responsible for stunning and hibernation are not well understood. The depressed
contractility has been attributed to abnormal calcium fluxes and decreased calcium sensitivity of the
contractile proteins. Initiating mechanisms may include oxygen free radicals and damage to the
myofilaments, mitochondria, and/or extracellular matrix, but these syndromes can also result from
stress-induced changes in myocardial composition similar to those responsible for preconditioning.
Conclusion
The focus of this chapter on the effects of ischemia on the heart should not obscure the fact that
ischemic heart disease is caused by disease of the coronary arteries. For this reason, although patients
gain considerable benefit from appropriate management of the cardiac abnormalities, in the last analysis
a good outcome depends on prevention and treatment of the underlying coronary artery occlusive
disease.
Bibliography
General
Bouchardy B, Manjo G. A new approach to the histological diagnosis of early myocardial infarcts.
Cardiology 1971/1972;56:327–332.
Davies MJ. Stability and instability: two faces of coronary atherosclerosis. The Paul Dudley White
Lecture. Circulation 1996;94:2013–2020.
Gettes LS, Cascio WE. Effect of acute ischemia on cardiac electrophysiology. In: Fozzard H, Haber E,
Katz A, et al., eds. The heart and cardiovascular system, 2nd ed. New York, NY: Raven Press, 1991:
2021–2054.
Rader DJ, Daugherty A. Translating molecular discoveries into new therapies for atherosclerosis.
Nature 2008;451:905–913.
Ross R. Atherosclerosis—an inflammatory disease. New Eng J Med 1999;340:115–126.
Rosenbaum MB, Blanco HH, Elizari M, et al. Electronic modulation of the T wave and cardiac
memory. Am J Cardiol 1982;50:213–222.
P.508
For more complete discussions of ischemic heart disease and coronary atherosclerosis, readers
should consult textbooks of medicine and cardiology.
Cell Death
Armstrong SC. Protein kinase activation and myocardial ischemia/reperfusion injury. Cardiovasc Res
2004;61:427–436.
Gustafsson Å, Gottlieb RA. Autophagy in ischemic heart disease. Circ Res 2009;108:150–158.
Hamacher-Brady A, Brady NR, Gottlieb RA. The interplay between pro-death and pro-survival
signaling pathways in myocardial ischemia/reperfusion injury: apoptosis meets autophagy.
Cardiovasc Drugs Ther 2006;20:445–462.
Reimer KA, Jennings RB. Myocardial ischemia, hypoxia, and infarction. In: Fozzard H, Haber E, Katz
A, et al., eds. The heart and cardiovascular system, 2nd ed. New York, NY: Raven Press, 1991:1875–
1973.
Whelan RS, Kaplinskiy V, Kitsis RN. Cell death in the pathogenesis of heart disease: mechanisms and
significance. Ann Rev Physiol 2009;72:19–44.
Willis MS, Townley-Tilson WHD, Yang EK, et al. Sent to destroy: the ubiquitin proteasome system
regulates cell signaling and protein quality control in cardiovascular development and disease. Circ
Res 2010;106:463–478.
Preconditioning, Postconditioning, Stunning, Hibernation
Bolli R. Preconditioning: a paradigm shift in the biology of myocardial ischemia. Am J Physiol Heart
Circ Physiol 2007;292:H19–H27.
Bolli R, Marban E. Molecular and cellular mechanisms of myocardial stunning. Physiol Rev 1999;79:
609–634.
Burley DS, Baxter GF. Pharmacological targets revealed by myocardial postconditioning. Curr Opin
Pharmacol 2009;9:177–188.
Camici PG, Prasad SK, Rimoldi OE. Stunning, hibernation, and assessment of myocardial viability.
Circulation 2008;117:103–114.
Eisen A, Fisman EZ, Rubenfire M, et al. Ischemic preconditioning: nearly two decades of research. A
comprehensive review. Atherosclerosis 2004;172:201–210.
Heusch G. Hibernating myocardium. Physiol Rev 1998;78:1055–1085.

Kandasamy AD, Chow AK, Ali MAM, et al. Matrix metalloproteinase-2 and myocardial oxidative stress
injury: beyond the matrix. Cardiovasc Res 2010;85:413–423.
Kloner RA, Bolli R, Marban E, et al. Medical and cellular implications of stunning, hibernation and
preconditioning. Circulation 1998;97:1848–1867.
Kloner RA, Jennings RB. Consequences of brief ischemia: stunning, preconditioning, and their
clinical implications Part 1. Circulation 2001;104:2981–2989.
Kloner RA, Jennings RB. Consequences of brief ischemia: stunning, preconditioning, and their
clinical implications Part 2. Circulation 2001;104:23158–23167.
Rajabi M, Kassiotis C, Razeghi P, et al. Return to the fetal gene program protects the stressed heart:
a strong hypothesis. Heart Fail Rev 2007;12:131–343.
Slezak J, Tribulova N, Okruhlicova L, Dhingra R, Bajaj A, Freed D, Singal P. Hibernating myocardium:

pathophysiology, diagnosis, and treatment. Can J Physiol Pharmacol 2009;87: 252–265.
Yellon DM, Downey JM. Preconditioning the myocardium: from cellular physiology to clinical
cardiology. Physiol Rev 2003;83:1113–1151.
P.509
References
Costantin L. Extracardiac factors contributing to hypotension during coronary occlusion. Am J
Cardiol 1963;11:205–217.
Downey JM, Cohen MV. Why do we still not have cardioprotective drugs? Circ J 2009;73:1171–1177.
Hirzel HO, Sonnenblick EH, Kirk ES. Absence of a lateral border zone of intermediate creatine
phosphokinase depletion surrounding a central infarct 24 hours after acute coronary artery occlusion
in the dog. Circ Res 1977;41:673–683.
Katz LN. Electrocardiography. Philadelphia, PA: Lea and Febiger, 1946.
Ludman AJ, Yellon DM, Hausenloy DJ. Cardiac preconditioning for ischaemia: lost in translation. Dis
Model Mech 2010;3:35–38.
Physiol 1996;270:H1207–H1216.
Williamson JR. Glycolytic control mechanisms. II. Kinetics of intermediate changes during the
aerobic-anoxic transition in perfused rat heart. J Biol Chem 1966;241:5026–5036.
> Table of Contents > Part Four - Pathophysiology > Chapter 18 - Heart Failure
Chapter 18
Heart Failure
Heart failure is a common, progressive, and usually lethal syndrome that represents a final common
pathway by which a variety of disease processes impair cardiac function. In the United States, heart
failure can be diagnosed in almost 10% of individuals over the age of 65, and 1 in 8 death certificates at
all ages mentions heart failure (Heart disease and stroke statistics—2010 Update). Heart failure is deadly,
with a worse prognosis than most malignancies; for example, mean survival for patients with this
syndrome (Levy et al., 2006) is about the same as that for stage 3B breast cancer (chest wall metastases
and/or extensive lymph node metastases) (Woodward et al., 2003). Because the most obvious abnormality
in heart failure is impaired pump function, definitions traditionally focus on the abnormal hemodynamics,
highlighting the clinical signs and symptoms that result from the impaired pumping of blood from the
veins to the arteries. However in the late 1980s, when long-term clinical trials began to reveal the dismal
prognosis in these patients, definitions began to recognize the progressive deterioration of failing hearts
and the shortened life expectancy, and so included changes that damage the failing heart. This text
defines heart failure as a clinical syndrome in which heart disease reduces cardiac output, increases
venous pressures, and generally causes progressive deterioration of the heart muscle.
Hemodynamic Abnormalities
The heart is a biological pump that moves blood from a region of low pressure (the veins) to one at higher
pressure (the arteries), and so can be compared to a mechanical pump that moves water from a leaky
basement into a garden hose (Fig. 18-1). Failure of the pump causes the basement to flood and reduces
flow out of the hose, which in heart failure are analogous to increasing venous pressure and decreasing
cardiac output, respectively. To make a more “realistic” comparison to clinical heart failure, as it is
understood today, the defective mechanical pump must also deteriorate rapidly.
A simple way to classify the hemodynamic abnormalities in heart failure is to define the reduced ejection
of blood into the aorta and pulmonary artery as forward failure, and the reduced return of blood from
the veins to the heart as backward failure. This classification is complicated by the fact that the
underlying abnormality can involve mainly the left or the right ventricle, so that there can be four
distinct manifestations of this syndrome (Table 18-1). However, the two sides of the heart operate in
series, so that when blood backs up behind one ventricle it impairs ejection by the other ventricle, and
when less blood is pumped out of one ventricle, less returns to the other. Furthermore, the heart is a
reciprocating pump, where phases of filling alternate with phases of ejection. As a result, when ejection
is reduced, the increased volume of blood remaining in the heart at the end of systole reduces the heart's
ability to fill during the next diastole; conversely, when filling is reduced the heart cannot eject a normal
stroke volume (SV).
P.511
Fig. 18-1: The failing heart as a defective pump in a leaky basement. A: The heart can be viewed as a pump
that moves water from a leaky basement into a hose. B: Failure of the pump can cause inadequate emptying
of the basement, which then floods (high venous pressure or “backward failure”), and inadequate flow of
water into the hose (low cardiac output or “forward failure”), or both.
Left and Right Heart Failure

The clinical picture in most patients with heart failure is dominated by the signs and symptoms of
impaired left ventricular function. This is especially true in developed countries, where the major
etiologies are ischemic and hypertensive heart disease. Symptoms of left heart failure also dominate the
clinical picture in most patients with dilated, hypertrophic, and infiltrative cardiomyopathies.
Tachycardia-induced cardiomyopathy, which is rare but important because it is among the few curable
forms of heart failure, usually causes left heart failure. Primary right heart failure, which is less common,
can be caused by cor pulmonale that is a complication of chronic lung disease, multiple pulmonary
emboli, and primary pulmonary hypertension. Valvular heart disease can cause either left or right heart
failure, depending on the anatomic abnormalities, while congenital heart disease often causes right heart
failure, for example, when pulmonic stenosis or pulmonary hypertension associated with an intracardiac
shunt overloads the right ventricle. Right heart failure can come to dominate the clinical picture in
patients in whom the primary abnormality involves the left heart (see later).
Table 18-1 Hemodynamic Abnormalities in Heart Failure
Site of Failure Type of Failure Initial Hemodynamic Consequence
Right heart Forward Reduced ejection into pulmonary artery—low cardiac output
Right heart Backward Increased systemic venous pressure
Left heart Forward Reduced ejection into aorta—low cardiac output
Left heart Backward Increased pulmonary venous pressure
P.512
Fig. 18-2: Effect of fluid retention. If the failing heart is viewed as a defective pump in a leaky basement
(Fig. 18-1), fluid retention worsens the flooding.
Backward and Forward Failure

The concepts of forward and backward failure are useful in understanding the signs and symptoms of
heart failure (see later), but are of no value in describing the abnormal hemodynamics because a heart
that cannot fill normally cannot eject a normal stroke volume, and a heart that cannot eject normally
cannot accept a normal venous return. For this reason, heart failure must limit ejection and filling to the
same extent. Forward failure is sometimes equated with depressed myocardial contractility (decreased
inotropy), and backward failure with impaired relaxation (decreased lusitropy), but this is not correct
because at any steady state ejection and filling must be equal.
The major determinant of whether a patient with heart failure has a high venous pressure and is
edematous (backward failure) or suffers from low cardiac output (forward failure) is the neurohumoral
response (see Chapter 8). Fluid retention by the kidneys, a major cause of increased preload and venous
congestion, worsens backward failure (Fig. 18-2), while the increased afterload caused by peripheral
vasoconstriction reduces ejection and so worsens forward failure (Fig. 18-3).
Therapy is also an important determinant of the manifestations of forward and backward failure (see
later). Vasodilator drugs, by reducing afterload, increase stroke volume and alleviate
P.513
the low cardiac output that characterizes forward failure, while diuretics reduce blood volume and
decrease venous pressures, and so alleviate the manifestations of backward failure. Excessive
vasodilatation, however, can worsen forward failure by lowering blood pressure, and excessive diuresis
can worsen forward failure by reducing preload. For these reasons, the severity of forward and backward
failures, defined as reduced cardiac output and increased venous pressure, respectively, provides little
information regarding the extent to which the abnormal pump function is caused by reduced filling or
reduced ejection, or whether contractility is depressed (decreased inotropy) or relaxation is impaired
(decreased lusitropy).
Fig. 18-3: Effect of vasoconstriction. If the failing heart is viewed as a defective pump in a leaky basement
(Fig. 18-1), vasoconstriction increases the work of the pump and decreases its forward output.
Systolic and Diastolic Dysfunction

Systolic and diastolic dysfunction, which describe different pathophysiological mechanisms, are not the
same as systolic and diastolic heart failure, which refer to different clinical syndromes (see later).
Systolic dysfunction describes ejection abnormalities that slow the rate of pressure rise during isovolumic
contraction (+dP/dt) and the rate and extent of ejection (Fig. 18-4A), whereas diastolic dysfunction
impairs ventricular filling by slowing the rate of pressure fall during isovolumic relaxation (-dP/dt),
decreasing the rate of filling, or increasing stiffness throughout diastole (Fig. 18-4B).
Causes of systolic dysfunction include left ventricular damage caused by myocardial infarction, dilated
cardiomyopathies, viral myocarditis, and toxic and metabolic abnormalities.
P.514
Diastolic dysfunction can be caused by wall thickening and reduced cavity volume in patients with
hypertension or hypertrophic cardiomyopathies, and restrictive diseases like amyloidosis, fibroelastosis,
and pericardial disease. A number of biochemical and molecular abnormities also cause systolic and
diastolic dysfunction by impairing cardiac myocyte contraction and relaxation, respectively.
Fig. 18-4: Pathophysiological mechanisms that can impair cardiac performance. A: Systolic dysfunction can
result from: (a) slowed rate of pressure rise during isovolumic contraction (+dP/dt), (b) slowed ejection,
and/or (c) reduced peak systolic wall stress. B: Diastolic dysfunction can result from: (a) Slowed rate of
pressure fall during isovolumic relaxation (-dP/dt), (b) slowed filling during early diastole, and/or (c)
increased stiffness throughout diastole.
Measurements of stroke volume and venous and arterial pressures cannot distinguish between systolic and
diastolic dysfunction because aortic pressure and cardiac output are decreased, and ventricular diastolic
pressures increased in most of these patients. Systolic dysfunction can be suggested by a marked
reduction in +dP/dt, ejection rate, or systolic compliance, while diastolic dysfunction is suggested by a
low -dP/dt, slowed ventricular filling, or reduced diastolic compliance. However, systolic and diastolic
dysfunction commonly occur together, which makes it difficult for any hemodynamic measurement to
identify which of these mechanisms is responsible for impaired pump function.
Systolic and Diastolic Heart Failure

Systolic and diastolic heart failure are clinical syndromes that are both easy and difficult to define. The
simple definitions are that systolic heart failure is caused by abnormalities in ejection, while diastolic
heart failure occurs when filling is impaired. However, these definitions are not valid because, as noted
earlier, a ventricle that ejects poorly cannot fill normally, and vice versa. For this reason, the distinction
has come to rely on ventricular architecture: the ventricle is dilated in patients with systolic heart
failure, whereas ventricular cavity volume can be normal, slightly reduced, or minimally increased in
diastolic heart failure. The measurement usually used to distinguish between systolic and diastolic heart
failure is the ejection fraction (EF) (see Chapter 12), which is the ratio between stroke volume and end-
diastolic volume (EDV):
The reason that EF has become the major basis for this distinction is that absolute values for stroke
volume and EDV are difficult to measure clinically whereas the ratio, which is dimensionless, is readily
determined using noninvasive methods, notably echocardiography. The dependence on EF in distinguishing
between these two types of heart failure is evidenced by current terminology, which describes systolic
heart failure as HFlowEF (heart failure with low ejection fraction), and diastolic heart failure as HFnEF
(heart failure with normal ejection fraction) or HFpEF (heart failure with preserved ejection fraction).
The new terminology reflects the central importance of ventricular architecture in making this
distinction.
EF distinguishes systolic and diastolic heart failure because EDV (the denominator) is increased when
ejection is impaired and usually reduced when impaired filling is the underlying abnormality.
Abnormalities in stroke volume (the numerator) are less important because the extent to which cardiac
output can decrease is limited in both types of heart failure. This is especially true in the elderly, where a
decrease in stroke volume is limited by the normally low cardiac output; for example, cardiac index
(cardiac output normalized for body surface area) at age 70 is ∼2.5 l/min/m2, which is only slightly
greater than in moderate to severe heart failure, where cardiac index is 1.6 to 2.3 l/min/m2 (Baim and
Grossman, 2000). A major exception, high output failure, is uncommon and is readily identified by clinical
evidence of a high cardiac output.
P.515
Table 18-2 Systolic and Diastolic Heart Failure
Systolic Heart Failure (Heart Failure with Low EF):
Eccentric hypertrophy (increased cavity volume)
Global: Dilated cardiomyopathies, viral or toxic myocarditis
Regional: Myocardial infarction
Diastolic Heart Failure (Heart Failure with Preserved EF):
Concentric hypertrophy (reduced cavity volume)

Hypertrophic cardiomyopathy, hypertensive heart disease
Systolic heart failure, which is characterized by left ventricular dilatation, is usually caused by diseases
that destroy, damage, or weaken the myocardium (Table 18-2). The left ventricular wall motion
abnormalities in systolic heart failure can be either regional or diffuse (global). Regional wall motion
abnormalities are seen when occlusion of a major coronary artery has destroyed part of the left ventricle
(myocardial infarction), whereas global wall motion abnormalities result from dilated cardiomyopathies
and myocarditis, in which depression of ventricular function is more uniform.
Diastolic heart failure is seen in patients with concentric left ventricular hypertrophy and stiff ventricles.
Etiologies include hypertension and the decreased aortic impedance that accompanies aging, which
explains why diastolic heart failure is commonly diagnosed in the elderly. Sudden episodes of severe
pulmonary edema (called flash pulmonary edema) are often seen in patients with diastolic heart failure
who have only mild symptoms under basal conditions. These episodes are caused by the reduced
ventricular compliance, which allows even a small increase in preload or afterload to cause a sharp rise in
end-diastolic pressure that “tips” the patient onto the descending limb of the Starling curve (see later).
Signs and Symptoms of Heart Failure

The clinical manifestations of heart failure include signs, which are objective manifestations of depressed
cardiac performance, and symptoms, which are abnormalities perceived by the patient. The
consequences of increased venous pressure behind a failing ventricle are readily understood because
upstream transmission of this pressure elevates capillary pressure, which increases the hydrostatic forces
that cause fluid to be transudated across the capillary endothelium into the tissues. The result is edema
in the lungs of patients with left heart failure and peripheral edema, ascites (fluid in the peritoneal
cavity), and pleural effusions in right heart failure. Fatigue, the major symptom associated with
decreased cardiac output, is due mainly to a skeletal muscle myopathy (see later).
None of the signs and symptoms described earlier are specific for heart failure. Hemodynamic
abnormalities resembling those caused by increased venous pressure in heart failure occur in patients
with normal hearts when increased blood volume, renal failure, hepatic disease, protein deficiency or
infusion of large amounts of blood or saline elevates systemic and pulmonary
P.516
venous pressures. Conversely, decreased blood volume, which can be caused by hemorrhage, extensive
burns, cholera, or toxic shock syndrome, reduces cardiac output. Fatigue, a major symptom in virtually
all patients with heart failure, also has many causes.
Table 18-3 Some Common Causes of Heart Failure
Right/Left Systolic/Diastolic Heart Global/Regional Wall

Etiology
Heart Failure Failure Motion Abnormality
Myocardial infarction Left Systolic Regional
Dilated Left Systolic Global

cardiomyopathy/myocarditis
Hypertensive heart disease Left Diastolic Global
Hypertrophic Left Diastolic Often regional

cardiomyopathy
Valvular/congenital Depends on Global

structures affected
Cor pulmonale Right Both Global
Tachycardia-induced Left Diastolic Global
Right and Left Heart Failure

The clinical picture in most patients with heart failure is dominated by the signs and symptoms of left
ventricular failure; this is especially true in developed countries, where the major etiologies of heart
failure are ischemic and hypertensive heart disease (Table 18-3). Patients with dilated cardiomyopathies
also suffer mainly from left heart failure. Right heart failure, which is less common, occurs most often in
patients with congenital heart disease and cor pulmonale; the latter can also come to dominate the
clinical picture in patients with left heart failure. The latter occurs when chronic elevation of left atrial
pressure increases pulmonary arterial pressure by causing reactive pulmonary vasoconstriction and
proliferative changes that obliterate small pulmonary arteries. Although this response reduces blood flow
through the lungs, and so alleviates pulmonary congestion, it replaces one problem (left heart failure) for
another (right heart failure) (Wood, 1954).
Backward Failure of the Left Heart

Backward failure of the left heart causes shortness of breath (dyspnea) largely by increasing the work
required to ventilate the congested lungs. In contrast to normal breathing, which is rarely perceived, it is
impossible to ignore the increased respiratory effort seen when elevated pulmonary venous pressure
causes the lungs to become stiff and inelastic, like a water-logged sponge. Cardiac dyspnea is
exacerbated by arterial hypoxia, which occurs when pulmonary interstitial edema impairs oxygen
exchange. Ventilation-perfusion mismatch and weakness of the respiratory muscles (see later) can also
contribute to hypoxia in these patients.
P.517
Dyspnea caused by both left heart failure and pulmonary disease is worsened by exertion, but cardiac
dyspnea generally becomes more severe when the patient lies down (orthopnea) because elevation of the
lower extremities increases central blood volume by draining blood from the leg veins. Paroxysmal
nocturnal dyspnea, the sudden onset of severe shortness of breath in a patient who has been recumbent
for several hours, occurs when blood volume is increased more slowly by resorption of interstitial fluid
from the edematous lower extremities.
High pulmonary venous pressure causes fluid to be transudated across the pulmonary capillaries; this fluid
appears first in the interstitium, from which it is carried to the systemic veins via lymphatic vessels. If
transudation occurs more rapidly than the rate at which the fluid can be removed by the lymphatics, the
interstitium becomes edematous. Interstitial edema can be seen on an ordinary chest x-ray as thin
horizontal lines, called Kerley B lines that, because of the effects of gravity, initially appear in the lower
lung fields. Left heart failure also results in a radiological pattern called “cephalization,” in which the
pulmonary veins draining the upper lobes appear larger than those that drain the lower lobes. However,
clinical estimation of left ventricular diastolic pressure is difficult, and so may require measurements
made by pulmonary artery catheterization.
Accumulation of interstitial fluid interferes with gas exchange between the pulmonary capillaries and the
alveoli; most prominent is arterial hypoxia because oxygen is much less soluble in water than carbon
dioxide. When backward failure of the left heart becomes severe, fluid entering the small bronchi causes
râles, which are crackling sounds described by Hippocrates as like the “seething of vinegar.” Because of
gravity, râles appear initially at the lung bases. In severe left heart failure, when pulmonary capillary
pressure (which drives fluid into the interstium) greatly exceeds plasma protein oncotic pressure (which
causes fluid resorption), flooding of the airspaces leads to pulmonary edema that can literally drown a
patient.
Backward Failure of the Right Heart

Backward failure of the right heart elevates jugular venous pressure, which provides an invaluable
bedside measurement for quantifying the severity of right heart failure. Right heart failure also causes
fluid transudation into the soft tissues of the periphery (edema) and the pleural, pericardial, and
abdominal spaces (anasarca). Massive fluid accumulation, once called dropsy, caused horrible suffering in
patients before the development of modern diuretics which, along with salt (sodium) restriction, has
changed the clinical manifestations of this syndrome. Today, fatigue, rather than fluid retention, is the
most troublesome problem in these patients. It is for this reason that heart failure is now replacing the
older designation congestive heart failure (CHF).
Forward Failure
Fatigue, which has emerged as a major problem in patients with heart failure, has several causes, the
most important of which is a skeletal muscle myopathy. Mechanisms that contribute to this myopathy
include reduced skeletal muscle perfusion, disuse caused by inactivity, malnutrition, cytokines and other
inflammatory mediators, apoptosis, and molecular abnormalities in the skeletal myocytes. Mitochondrial
dysfunction in the latter impairs oxidative adenosine triphosphate (ATP) regeneration and increases
anaerobic lactate production that accelerates the appearance of systemic acidosis during exercise. The
ability of
P.518
exercise training both to alleviate the symptoms associated with this myopathy and improve prognosis
argues strongly against the once widely held view that heart failure should be treated with prolonged
rest.
Interplay Between the Failing Heart and the Peripheral Circulation

Heart failure reduces blood flow into the arteries and causes blood to accumulate in the veins behind the
failing ventricle (Fig. 18-1) at the same time, increases in peripheral resistance and blood volume
exacerbate the clinical manifestations of heart failure. Understanding of these and other hemodynamic
changes in patients with heart failure is facilitated by examining pressure–volume loops, which focus on
the heart to show how changing preload and afterload alter cardiac performance (see Chapter 11), and
Guyton diagrams which describe the interplay between blood flow into and out of the heart by showing
the effects of atrial pressure on venous return and cardiac output (see Chapter 12). Both provide valuable
insights into the way that changes in blood volume and arteriolar resistance modify cardiac performance.
Effects of Heart Failure on Pressure–Volume Loops

Pressure–volume loops depict the interplay between lusitropy, which determines the end-diastolic
pressure–volume relationship; inotropy, which determines the end-systolic pressure–volume relationship;
and preload and afterload, which determine the positions of the end-diastolic and end-systolic points
along the corresponding pressure–volume relationships. The following discussion describes the events that
occur when a sudden decrease in contractility impairs the heart's ability to eject, and when a rapid
increase in diastolic stiffness impairs the ability of the heart to fill.
Impaired Ejection
A sudden decrease in contractility, such as occurs after coronary artery occlusion, shifts the end-systolic
pressure–volume relationship (the Starling curve) to the right and downward (Fig. 18-5A). Because
ejection is reduced, end-systolic (residual) volume increases; as a result, addition of a normal venous
return to the greater end-systolic volume increases EDV in the next cardiac cycle (Fig. 18-5B). This
hemodynamic adjustment increases preload by moving the end-diastolic point upward and to the right
along the end-diastolic pressure–volume relationship which, according to Starling's law of the heart,
increases stroke volume.
The neurohumoral response (see Chapter 8) causes two important hemodynamic adjustments to the
impaired cardiac performance. The first is arteriolar vasoconstriction, which increases afterload. This
response, which initially results from the α-adrenergic effects of sympathetic activation, and later from
the effects of angiotensin II, endothelin, and vasopressin, increases aortic pressure and reduces stroke
volume (Fig. 18-5C). Because vasoconstriction reduces cardiac output and increases energy expenditure
(see Chapter 12), this response is often detrimental in patients with heart failure. Salt and water
retention by the kidneys, the second major component of the neurohumoral response, increases preload,
and so shifts the end-diastolic point upward and to the right along the end-diastolic pressure–volume
relationship (Fig. 18-5D). According to Starling's law, this increases stroke volume, but at the expense of
increasing venous pressure.
P.519
Fig. 18-5: Left ventricular pressure–volume loops after a sudden decease in the heart's ability to eject; for
clarity, the responses to the fall in the end-systolic pressure–volume relationship (ESPVR) are shown
sequentially. Hemodynamic and other changes are identified by dotted arrows (indicated by asterisks) that
point to open circles that represent the new condition. In all six panels, the heart operating under baseline
conditions is depicted using dashed lines; in panels B–F, the pressure–volume loop in the preceding panel is
shown using dotted lines. A: Impaired ejection shifts the ESPVR to the right and downward (solid curve)
which, if afterload remains constant, reduces stroke volume by shifting the end-systolic point (open circle) to
a lower volume. B: Reduced ejection increases end-diastolic pressure and volume (open circle), which
according to Starling's law increases stroke volume; this is apparent when the solid curves in A and B are
compared. C: Vasoconstriction increases afterload (open circle), which moves the end-systolic point upward,
thereby increasing arterial pressure but reducing stroke volume; this is apparent when the solid loops in B
and C are compared. D: Fluid retention increases preload (open circle), which increases stroke volume but at
the expense of higher end-diastolic pressure and volume; this is apparent when the solid loops in C and D are
compared. E: Increased contractility (dotted line) increases stroke volume, which is apparent when the solid
loops in D and E are compared. F: Increased lusitropy (dotted line) increases stroke volume, which is
apparent when the solid loops in E and F are compared. ESPVR: end-systolic pressure–volume relationship;
EDPVR: end-diastolic pressure–volume relationship.
P.520
The neurohumoral response affects the heart when β-adrenergic stimulation increases contractility and
facilitates filling. The inotropic response, which shifts the end-systolic pressure–volume relationship
upward and to the left (Fig. 18-5E), along with the chronotropic response to β-stimulation, increases the
ability of the failing heart to eject at a given EDV and afterload. The lusitropic response shifts the end-
diastolic pressure–volume relationship downward and to the right (Fig. 18-5F), which improves backward
failure by reducing end-diastolic pressure; this response also alleviates forward failure when the increased
EDV increases ejection (Starling's law of the heart).
Impaired Filling
A lusitropic abnormality that decreases the ability of the heart to fill, which can appear rapidly when the
heart becomes energy-starved, shifts the end-diastolic pressure–volume relationship to the left and
upward (Fig. 18-6A). Because ejection is reduced, end-systolic (residual) volume increases so that
addition of the venous return to the greater end-systolic volume increases EDV in the next beat.
According to Starling's law of the heart, the latter increases stroke volume (Fig. 18-6B). Vasoconstriction
and fluid retention, which are initiated by the neurohumoral response, increase both afterload and
preload. The higher afterload caused by arteriolar constriction increases aortic pressure but reduces
stroke volume (Fig. 18-6C), while the higher EDV caused by venoconstriction and fluid retention increases
stroke volume, but at the expense of a further increase in diastolic pressure (Fig. 18-6D). β-Adrenergic
stimulation of the heart increases contractility, which increases ejection (Fig. 18-6E) and, by stimulating
the biochemical mechanisms that relax the heart, reduces end-diastolic pressure and increases ejection
(Fig. 18-6F).
Combination of Impaired Ejection and Impaired Filling

Impairment of both ejection and filling “compresses” the pressure–volume loop (Fig. 18-7). The smaller
area within the pressure–volume loop reflects the reduced ability of the failing heart to perform external
work, which decreases both stroke volume and the ability of the ventricle to develop pressure.
Effects of Systolic and Diastolic Dysfunction on the Response to

Increased Venous Return
The hemodynamic consequences of increasing venous return differ when cardiac function is impaired by
an inotropic and a lusitropic abnormality (Fig. 18-8). Systolic dysfunction reduces stroke volume at any
given preload (Fig. 18-8B) and causes venous pressure to rise in response to an increase in venous return
that restores stroke volume (Fig. 18-8D). The steeper end-diastolic pressure–volume relationship in a
heart with diastolic dysfunction (Fig. 18-8C) causes a similar increase in venous return to generate a much
greater rise in venous pressure (Fig. 18-8E). This adverse effect of diastolic dysfunction is amplified
because the steep end-diastolic pressure–volume relationship reduces the ability of the Starling's law to
increase stroke volume by impairing the ability of the ventricle to dilate. Similar differences are seen in
chronic systolic and diastolic heart failure, and explain the rapid onset of severe pulmonary congestion
(often called “flash pulmonary edema”) commonly seen in patients with diastolic heart failure.
Changes in Ventricular Architecture

The pressure–volume loops in Figure 18-5 through 18-8, which illustrate the consequences of acute
changes in the inotropic and lusitropic properties of the heart, do not provide an accurate picture of what
happens to patients when ventricular architecture is modified in chronic
P.521
P.522
heart failure (Fig. 18-9). Cavity enlargement in eccentric hypertrophy, which characterizes systolic heart
failure, shifts the pressure–volume loops to higher volumes (Fig. 18-10C), whereas reduction in cavity size
in concentric hypertrophy (Fig. 18-10A), which characterizes diastolic heart failure, shifts the pressure–
volume loops to lower volumes.
Fig. 18-6: Left ventricular pressure–volume loops after a sudden decease in the heart's ability to fill; for
clarity, the responses to an upward shift in the end-diastolic pressure–volume relationship (EDPVR) are shown
sequentially. Hemodynamic and other changes are identified by dotted arrows (indicated by asterisks) that
point to open circles which represent the new condition. In all six panels, the heart operating under baseline
conditions is depicted using dashed lines; in panels B–F, the pressure–volume loop in the preceding panel is
shown using dotted lines. A: Impaired relaxation shifts the EDPVR to the left and upward (solid curve), which
causes the end-diastolic point (open circle) to move to a lower volume at a higher pressure; if afterload
remains constant, stroke volume will be reduced. B: Reduced ejection leads to a further increase in end-
diastolic pressure and volume (open circle), which according to Starling's law increases stroke volume; this is
apparent when the solid curves in A and B are compared. C: Vasoconstriction increases afterload (open
circle), which moves the end-systolic point upward, thereby increasing ejection pressure but reducing stroke
volume; this is apparent when the solid loops in B and C are compared. D: Fluid retention increases preload
(open circle), which increases stroke volume but at the expense of increasing end-diastolic pressure and
volume; this is apparent when the solid loops in C and D are compared. E: Increased contractility (dotted
line) increases stroke volume, which is apparent when the solid loops in D and E are compared. F: Increased
lusitropy (dotted line) increases stroke volume, which is apparent when the solid loops in E and F are
compared. ESPVR, end-systolic pressure–volume relationship; EDPVR, end-diastolic pressure–volume
relationship.
Fig. 18-7: Left ventricular pressure–volume loop after a sudden decease in the heart's ability both
to eject and to fill (solid lines). Hemodynamic and other changes are identified using dotted arrows
that point to open circles that represent the new condition (indicated by asterisks). The pressure–
volume loop for the heart operating under baseline conditions is depicted using a dashed line.
ESPVR, end-systolic pressure–volume relationship; EDPVR, end-diastolic pressure–volume
relationship.
Although systolic heart failure is characterized by the inability to eject normally, filling is also impaired
because dilatation increases wall stress at any diastolic pressure (the law of Laplace; see Chapter 11). In
diastolic heart failure, where the underlying problem is inability to fill normally, the heart's ability to
eject is impaired when concentric hypertrophy reduces cavity volume. The steepness of the end-diastolic
pressure–volume relationship in both types of heat failure is also increased by fibrosis that generally
accompanies ventricular hypertrophy.
Effects of Atrial Pressure on Cardiac Output and Venous Return

(Guyton Diagrams)
The interplay between venous return and cardiac output depicted in Guyton diagrams provides an
additional way to view the effects of heart failure on cardiac performance and circulatory hemodynamics.
These are illustrated in Figure 18-11, which describes the sequence of events that follows a decrease in
ejection caused by depressed myocardial contractility. The initial effect is a shift in the intercept
between the curves relating cardiac output to venous return that decreases stroke volume and increases
atrial (venous) pressure (Fig. 18-11B). The latter then increases cardiac output according to Starling's law
of the heart (Fig. 18-11C). These changes do not initially affect blood volume, so that the curve relating
venous return to atrial pressure does not change.
P.523
Fig. 18-8: Differences in the hemodynamic effects of the same increase in venous return in hearts
with systolic and diastolic dysfunction. In all five panels, the heart operating under baseline
conditions is depicted by dashed lines. Pressure–volume loops are shown under basal conditions (A),
after a decrease in inotropy that causes systolic dysfunction (B, D), and after a decrease in lusitropy
that causes diastolic dysfunction (C, E). B: The first beat after a reduction in inotropy that shifts the
end-systolic pressure–volume relationship to the right and downward (solid arrow) reduces stroke
volume by 50% (arrow labeled ↓SV); note that end-diastolic pressure and volume are unchanged. C:
The first beat after a reduction in lusitropy that reduces stroke volume by 50% (arrow labeled ↓SV);
because the decrease in diastolic compliance shifts the end-diastolic pressure–volume relationship to
the left and upward (solid arrow), end-diastolic pressure increases from 10 to 20 mm Hg. D: The
decrease in stroke volume shown in B increases end-systolic volume which, when added to the
venous return causes end-diastolic volume to increase in subsequent beats. If venous return
remained unchanged, the increased end-diastolic volume would restore stroke volume to normal by
increasing end-diastolic pressure to ∼23 mm Hg (dotted arrow marked by an asterisk). E: The
increased end-diastolic pressure shown in C causes end-diastolic volume to increase which, if the
venous return in the next beat restored stroke volume to normal, would cause end-diastolic pressure
to increase to ∼40 mm Hg (dotted arrow marked by an asterisk). The much higher diastolic pressures
in diastolic dysfunction (C and E) than in systolic dysfunction (B and D) are caused by the greater
shift of the end-diastolic pressure–volume relationship upward and to the left. ESPVR, end-systolic
pressure–volume relationship; EDPVR, end-diastolic pressure–volume relationship; SV, stroke volume.
P.524
Fig. 18-9: Architectural patterns of cardiac hypertrophy. Normal left ventricle (B). The eccentric hypertrophy
seen in the “athlete's heart” (C), an example of physiologic hypertrophy, differs from the pathological
eccentric hypertrophy in dilated cardiomyopathy, where ejection is decreased (F), and that caused by
chronic volume overload where ejection is increased (E). Myocardial infarction, which causes a regional wall
motion abnormality (A), leads to eccentric hypertrophy of the noninfarcted regions of the ventricle. All differ
from the concentric hypertrophy caused by chronic pressure overload (D).
Two components of the neurohumoral response (see Chapter 8) modify these curves. The first is fluid
retention, which by increasing blood volume, shifts the curve relating venous return to atrial pressure
upward and to the right (Fig. 18-11D); this response increases cardiac output (Starling's law of the heart),
but causes a further rise in venous pressure. The second component of the neurohumoral response, an
increase in myocardial contractility (Fig. 18-11E), causes an upward shift of the curve relating atrial
pressure to cardiac output which, by shifting the intercept upward and to the left, increases cardiac
output and lowers atrial pressure.
P.525
Fig. 18-10: Pressure–volume loops (above) in different architectural phenotypes of left ventricular
hypertrophy (below). A: Concentric hypertrophy; B: normal; C: eccentric hypertrophy. (Based on Kass, 1988.)
Effects of Therapy on Hemodynamics

One of the major goals in treating the abnormal hemodynamics in patients with heart failure is to
alleviate the adverse consequences of the neurohumoral response. Most important are to reverse fluid
retention by the kidneys, which is the major cause of edema (Fig. 18-2), and reduce peripheral arteriolar
vasoconstriction, which lowers cardiac output (Fig. 18-3) and decreases cardiac efficiency (see Chapter
12). Fluid retention is treated with diuretics, which improve backward failure by reducing blood volume
(Fig. 18-12A), and vasodilators increase cardiac output by reducing afterload (Fig. 18-12B).
Diuretics, by lowering venous pressure, are generally of enormous benefit to patients with heart failure;
diuretics are also energy sparing because they reduce wall stress. However, because reducing preload also
decreases ejection (Starling's law of the heart), diuretics can lower both cardiac output and blood
pressure, while excessive diuresis can replace backward failure with forward failure (Table 18-4).
Vasodilators decrease arterial pressure and increase cardiac output. Reduced afterload decreases energy
expenditure, although excessive lowering of blood pressure can decrease perfusion of the brain and heart
and, by activating the baroreceptor response, can further intensify the neurohumoral response (Table 18-
4).
P.526
Fig. 18-11: Effects of decreased myocardial contractility on curves relating atrial pressure to venous return
and cardiac output. A: Control curves showing the intercept that defines the steady state where blood flow
into and out of the heart are equal (open circle A). In B–E, hemodynamic changes are identified by dotted
arrows (indicated by asterisks), new steady states by open circles, and previous steady states by closed
circles. B: Reduced ejection decreases cardiac output and so shifts the intercept to a lower cardiac output
and higher atrial pressure (open circle B); the shift to a lower Starling curve is apparent when the solid curve
is compared to the dashed curve obtained under baseline conditions. C: Impaired ejection increases end-
diastolic volume, and so leads to a rise in atrial pressure that moves the intercept upward and to the right
(open circle C). According to Starling's law, the upward shift of the intercept increases cardiac output. D:
Fluid retention shifts the curve relating venous return and atrial pressure upward and to the right. The
upward shift in the intercept increases cardiac output, and the rightward shift increases atrial pressure (open
circle D). E: Increased contractility (solid line) shifts the intercept upward and to the left, which increases
cardiac output and reduces atrial pressure (open circle E).
P.527
Fig. 18-12: Treatment of the hemodynamic disorders in heart failure. A: Diuretics, which inhibit
fluid retention, improve “backward failure.” B: Vasodilators, which reduce afterload, improve
“forward failure.”
β-Adrenergic blockers have a negative inotropic effect that, along with their ability to slow the heart,
reduces cardiac output. At the same time, however, these drugs are energy-sparing and have an
important beneficial effect to reduce maladaptive proliferative signaling (Table 18-4). Cardiac glycosides
such as digitalis are generally viewed as inotropic agents whose beneficial effects are due to increased
myocardial contractility; however, their most important effect is to slow ventricular rate in patients with
atrial fibrillation. For this reason, these drugs are of little benefit in patients who are in sinus rhythm.
There is substantial evidence that a centrally mediated counterregulatory response that slows ventricular
rate and reduces afterload is responsible for many of the beneficial effects of digitalis. However, like
inotropic agents whose effects are mediated by a cyclic adenosine monophosphate (AMP)-induced
increase in iCaL (see Chapter 10), the ability of cardiac glycosides to increase cytosolic calcium can cause
triggered depolarizations and dangerous arrhythmias.
Biochemical and Biophysical Abnormalities in the Failing Heart

Heart failure is associated with a number of abnormalities that impair excitation-contraction coupling,
contraction, and relaxation by the cardiac myocytes. These are initiated by functional responses that
modify cell chemistry, and by proliferative responses that alter the size, shape, and composition of failing
hearts.
Energy Starvation
The question as to whether the failing heart is in an energy-starved state goes back at least to the 1930s,
but it is only recently that tools like nuclear magnetic resonance spectroscopy were
P.528
able to show conclusively that ATP and phosphocreatine levels are significantly reduced in overloaded and
failing hearts. Energy starvation is especially marked in the subendocardial regions of the left ventricle
where a combination of high wall stress and low perfusion can cause myocyte necrosis.
Table 18-4 Some Beneficial and Deleterious Effects of Therapy in Heart Failure
Therapy Beneficial Effect Deleterious Effect Consequence
Diuretics Lower venous Decreased venous

pressure congestion
Decreased Decreased cardiac

preload output
Decreased wall Reduced energy

stress expenditure
Vasodilators Decreased afterload Increased cardiac output
Lower blood Decreased perfusion of

pressure brain, heart
Lower blood Neurohumoral

pressure stimulation
Decreased wall Reduced energy

stress expenditure
β-Adrenergic Decreased Decreased cardiac

blockade inotropy output
lusitropy output

chronotropy output
Reduced energy Slowed deterioration of

expenditure heart
Less maladaptive Slowed deterioration of

hypertrophy heart
Antiarrhythmic Less sudden death

effects
Energy starvation in failing hearts results from an imbalance between energy consumption, which is
generally increased, and energy production, which is usually decreased. In systolic heart failure (heart
failure with low EF), energy utilization is increased when dilation of the ventricle and thinning of its walls
increase systolic wall stress and reduce cardiac efficiency. Cardiac energy demands are also increased in
diastolic heart failure (heart failure with normal EF), especially when left ventricular afterload is
increased by arterial hypertension and decreased aortic compliance. Energy starvation in ischemic heart
disease is caused by reduced oxygen delivery to the myocardium and prior myocardial infarction; the
latter, by reducing the number of viable myocytes, increases the work that must be done by the
remaining myocardial cells. Cardiac energy expenditure in most forms of valvular and congenital heart
disease is increased by high levels of cardiac work.
Energy production is also reduced in failing hearts by a decrease in capillary density relative to
ventricular mass, and by increased intercapillary spacing that impairs the diffusion of substrates to
hypertrophied cardiac myocytes. The greater diameter of hypertrophied cardiac myocytes also causes the
core of the enlarged fibers to become hypoxic by increasing the distance over which oxygen must diffuse.
P.529
Fig. 18-13: Some metabolic abnormalities in failing hearts. Decreased fatty acid oxidation and
reduced oxidative phosphorylation increase glycolysis, which partly compensates for these
abnormalities in early heart failure. Impaired oxidative phosphorylation is due to reduced delivery
of oxygen to working cardiac myocytes, decreased respiratory chain and ATP synthase activities,
and uncoupling of oxidation and ATP regeneration. Energy transfer by the phosphocreatine shuttle
is reduced in failing hearts by decreases in phosphocreatine and creatine kinase; consequences
include a slight fall in cytosolic ATP and a more marked increase in cytosolic ADP; together these
decrease the free energy of ATP hydrolysis, which is proportional to the ratio between ATP and ADP
concentrations. (Based on Neubauer, 2007.)
A number of metabolic abnormalities impair energy production in failing hearts (Fig. 18-13). However,
tight coupling between various metabolic pathways makes it difficult to isolate and evaluate the
importance of a given abnormality. Impaired mitochondrial function is especially detrimental because of
the heart's dependence on ATP regeneration by oxidative phosphorylation, which is also reduced when
calcium overload inhibits both the respiratory chain and ATP synthase. Mitochondrial dysfunction causes
the release of cytochromes and reactive oxygen species, which initiates apoptosis, and reduced fatty acid
oxidation leads to the accumulation of lipids that can cause membrane damage. Decreased synthesis of
oxidative enzymes, notably mitochondrial creatine phosphokinase, slows the phosphocreatine shuttle and
so impairs the transfer of high-energy phosphates from the mitochondria, where they are regenerated, to
cytosolic energy-utilizing systems such as the myofibrillar proteins and ion pumps. An isoform switch that
replaces the M isoform of creatine phosphokinase with the B isoform facilitates adenosine diphosphate
(ADP) rephosphorylation; this provides a limited compensation for the reduced content of this enzyme,
but the overall effect remains a decrease in energy transfer. The heart's limited capacity for glycolysis
makes it impossible for anaerobic pathways to provide enough ATP for the contractile machinery when
mitochondrial energy production is impaired. Acidosis, which occurs when glycolysis is accelerated,
inhibits many processes involved in excitation-contraction coupling, contraction, and relaxation (see
Chapters 4 and 7).
High-energy phosphate content falls in severe heart failure, but levels of ATP do not become low enough
to deprive the substrate-binding sites of the contractile proteins, ion pumps, and other energy-consuming
systems of their energy source. More detrimental is attenuation of the allosteric effects of ATP, which
allow a modest fall in ATP concentration to increase diastolic stiffness, depress
P.530
myocardial contractility, and decrease resting membrane potential (see Chapter 10). One of the most
important consequences of ATP depletion in failing hearts is a decrease in the free energy available from
ATP hydrolysis (-ΔG) that slows the sarcoplasmic reticulum calcium pump.
Fig. 18-14: Energy starvation, by increasing cytosolic calcium, contributes to several vicious cycles that can
cause cardiac myocyte necrosis. Energy starvation impairs calcium removal from the cytosol when
attenuation of the allosteric effects of ATP inhibits the Na/Ca exchanger, the sodium pump, and the calcium
pumps of the sarcoplasmic reticulum and plasma membrane. The resulting increase in cytosolic calcium
inhibits actomyosin dissociation, which in addition to impairing relaxation, increases energy utilization and so
worsens the energy starvation. Oxidative phosphorylation is also inhibited when cytosolic calcium is taken up
by the mitochondria; the result is a worsening of energy starvation that amplifies these vicious cycles, which
can cause cardiac myocyte necrosis.
The ability of energy starvation to increase cytosolic calcium can initiate a number of vicious cycles that
lead to myocyte necrosis (Fig. 18-14). These can occur when high cytosolic calcium accelerates energy-
consuming interactions, which impairs calcium removal from the cytosol, which causes a further increase
in cytosolic calcium, etc.
Molecular Abnormalities in the Failing Heart

Many of the maladaptive consequences of chronic overloaded and heart failure reflect the fact that adult
human cardiac myocytes are terminally differentiated cells with little or no capacity to proliferate (see
Chapter 9). Although these myocytes can divide when they are subjected to severe stress, their attempts
to proliferate are unsuccessful because the new cells rarely participate in ejection (see Chapter 9).
Myocytes in overloaded and failing hearts are able to enlarge, but this is an unnatural response. For this
reason, cardiac hypertrophy is initially adaptive, but has maladaptive long-term consequences that
eventually become a major problem in patients with heart failure.
The clinical importance of overload-induced cardiac hypertrophy was noted in the 19th century, when
both eccentric and concentric hypertrophy were observed to have deleterious as well as beneficial
consequences. Osler (1892) described three phases in the hypertrophic response of the heart to overload
(Table 18-5). The first phase, development, is adaptive because when the heart enlarges it becomes
better able to meet the increased load. Clinical improvement
P.531
continues in the second phase, which Osler called full compensation, but ends with broken compensation,
a third, maladaptive, phase where degeneration and weakening of the heart muscle worsen symptoms and
cause the patients to die.
Table 18-5 Three Stages in the Response to a Sudden Increase in Load
Phase 1: Osler: Development; Meerson: Transient breakdown
Clinical: Symptomatic left ventricular dysfunction
Pathophysiology: Left ventricular dilatation, pulmonary congestion, low cardiac output,

early hypertrophy
Phase 2: Osler: Full compensation; Meerson: Stable hyperfunction

Clinical: Class I–II heart failure
Pathophysiology: Improved symptoms, resolved pulmonary congestion, increased cardiac

output, established myocardial hypertrophy
Phase 3: Osler: Broken compensation; Meerson: Progressive cardiosclerosis
Clinical: Class III–IV heart failure
Pathophysiology: Worsening congestion, hemodynamic deterioration, progressive ventricular

dilatation, myocardial cell death, fibrosis
Meerson (1961), who was the first to use modern methods to study overload-induced cardiac hypertrophy
in animal models, called attention to the three phases described by Osler (Table 18-5), but it was not
until the late 1980s that progressive dilatation of some overloaded hearts (now called “remodeling”) was
recognized as a major cause of clinical deterioration. These observations indicated that the deleterious
effects of hypertrophy represent a “cardiomyopathy of overload” (Katz, 1990a), and drew attention to
the architectural, cellular, and molecular consequences of maladaptive cardiac hypertrophy. As described
later, it is now clear that the hypertrophic response can be both beneficial and deleterious, often at the
same time!
Myocardial Cell Death

Cardiac myocyte death, which is among the most devastating consequences of a chronic hemodynamic
overload, is a calamity because the heart has little or no ability to replace its terminally differentiated
myocytes. Furthermore, myocyte loss increases the work that must be done by the surviving myocytes,
and so contributes to a vicious cycle in which cell death increases overload, which intensifies the
hypertrophic response, which accelerates cell death, which increases overload, etc.
Necrosis, autophagy, and apoptosis all occur in failing hearts. Causes of necrosis include energy
starvation, calcium overload, and possibly membrane damage caused by fatty acid accumulation. Energy
starvation is especially important because it increases cytosolic calcium and so can establish vicious
cycles that increase energy demand and impair energy production (see earlier). The role of autophagy is
less clear as the consequences of the increased number of autophagosomes seen in failing hearts could be
part of the process of cell destruction, an effort to recycle and reuse cellular components, or both.
Apoptosis is increased in end-stage heart failure, so that this normally infrequent cause of cell death
contributes to progression in this syndrome.
P.532
Several pro-apoptotic mechanisms are activated in failing hearts. Stimuli include oxygen free radicals and
cytochromes released from abnormal mitochondria, along with activation of hypertrophic signaling by
cytoskeletal deformation, increased cytosolic calcium, and elevated cytokine levels (see Chapter 9).
Activation of the heterotrimeric G protein Gαq by the neurohumoral response increases apoptosis by
stimulating pathways controlled by pro-apoptotic Bcl-2 peptides (see Chapter 9); the latter include Nix
(Nip3 [19 kD interacting protein-3]-like protein X) and Bnip3 (Bcl-2/adenovirus E1B 19 kD interacting
protein 3). Pro-apoptotic signaling also mediates progressive dilatation (“remodeling”).
Architectural Changes
The adaptive nature of the hypertrophic response to overload was clearly shown in the 1960s, when
overload-induced hypertrophy was found to normalize wall stress in the hearts of patients with valvular
disease (Sandler and Dodge, 1963; Hood et al., 1968; Grossman et al., 1975). In 1985, a landmark paper
showed that angiotensin converting enzyme (ACE) inhibitors can slow the progressive dilatation that
follows experimental myocardial infarction, which came to be called remodeling (Pfeffer et al., 1985).
The clinical importance of this observation became clear when ACE inhibitors were found to prolong
survival in patients with heart failure, and the ability of angiotensin II to stimulate proliferative signaling
suggested that inhibition of maladaptive hypertrophy by ACE inhibitors could explain some of the clinical
benefits of these drugs (Katz, 1990b).
Many architectural phenotypes are found in human hearts (Fig. 18-9, Table 18-6). Those that appear and
then disappear during embryonic development are not abnormal, nor is the
P.533
physiological hypertrophy that is induced by exercise training. However, the phenotypes that appear in
chronically overloaded hearts are pathological because they are accompanied by maladaptive changes
that include progressive dilatation, reversion to the fetal phenotype, myocyte death, and fibrosis. Unlike
physiological hypertrophy, pathological hypertrophy is associated with arrhythmias.
Table 18-6 Examples of Different Cardiac Myocyte Phenotypes
Normal embryonic phenotypes
Normal adult phenotypes
Working myocardial cells
Atrial myocardium
Ventricular myocardium
Specialized cells
Nodal cells
His-Purkinje cells
Physiologic hypertrophy phenotype
Exercise-Induced Hypertrophy (the “Athlete's heart”)
Pathological phenotypes
Eccentric hypertrophy
Myocyte elongation caused by volume overload (e.g., aortic or mitral regurgitation)
Myocyte elongation caused by myocardial infarction
Dilated cardiomyopathies
Myocyte damage (e.g., viral or toxic myocarditis)
Concentric hypertrophy
Myocyte thickening caused by pressure overload (e.g., aortic stenosis, hypertension)
Hypertrophic cardiomyopathies
Pathological hypertrophy was initially divided into two phenotypes: concentric hypertrophy, where cavity
volume is reduced and wall thickness is increased; and eccentric hypertrophy, where the heart is dilated
and its walls thinned. These architectural patterns, however, fail to tell the complete story. For example,
the pathological eccentric hypertrophy caused by a chronic volume overload is different from the
eccentric hypertrophy in familial dilated cardiomyopathies (see later), and both differ from the
physiological dilatation seen in the athlete's heart. Similarly, the concentric hypertrophy caused by a
chronic pressure overload, such as hypertension and aortic stenosis, differs from that in patients with
hypertrophic cardiomyopathies. These and other differences reflect the diverse changes in cellular
structure and molecular composition that can occur when the heart hypertrophies.
Cellular Changes
More than 40 years ago Grant et al., (1965), who were studying valvular heart disease in humans,
postulated that eccentric hypertrophy (dilatation) occurs when sarcomeres are added to cardiac myocytes
in series, which causes the cells to become longer, whereas parallel addition of sarcomeres causes the
myocytes in concentrically hypertrophy hearts to become thicker (Fig. 18-15). This insight has been
confirmed by studies of the size and shape of cardiac myocytes isolated from hypertrophied human
hearts; in concentric hypertrophy, myocyte thickness is increased, whereas myocyte length is increased in
eccentrically hypertrophied hearts (Gerdes et al., 1994). The appearance of these different myocyte
phenotypes, which can result from activation of different signaling pathways (Wollert et al., 1996),
reflects the ability of diastolic stretch to increase myocyte length by causing sarcomeres to be added at
the ends
P.534
of the cells, and increased systolic stress to cause new sarcomeres to be added throughout the myocytes
(Russell et al., 2000). Evidence that different mitogen-activated protein (MAP) kinases are activated when
cardiac myocytes stretched during diastole and during systole (Yamamoto et al., 2001), and that different
patterns of focal adhesion kinase (FAK) (Senyo et al., 2007) and LIM protein (Boateng et al., 2009)
activation lead to the appearance of different cellular phenotypes, helps explain why pressure overload
and volume overload induce specific architectural phenotypes.
Fig. 18-15: Cellular basis for eccentric and concentric hypertrophy. Eccentric hypertrophy (dilation) occurs
when sarcomeres are added in series, at the ends of cells, whereas sarcomere addition occurs throughout the
cell in concentric hypertrophy.
Molecular Changes
The possibility that heart failure is accompanied by molecular abnormalities in the contractile proteins
was suggested almost 50 years ago, when Alpert and Gordon (1962) found that myofibrils isolated from
failing hearts have a low ATPase activity. This abnormality, which provided an explanation for the
depressed contractility in these patients, opened a new field of research that led to the discovery of
molecular changes in most of the myofibrillar proteins, as well as changes in several metabolic enzymes
and many of the membrane proteins that participate in both the extracellular and intracellular calcium
cycles (Table 18-7). However, not all of these changes in protein content, isoforms, and activities have
been found in human heart failure, and published findings disagree because of differences in
experimental models and the etiology and severity of the syndromes studied.
Myofibrillar Proteins
The low myofibrillar ATPase in failing hearts, which decreases maximal shortening velocity (see Chapters 4
and 6), is due largely to increased expression of the slow (β) myosin heavy chain isoform (Table 18-7). This
isoform shift is part of a reversion to the fetal phenotype in pathological hypertrophy that reduces
myocardial contractility, and so has an energy-sparing effect. In contrast, cardiac enlargement in
exercise-induced physiological hypertrophy (the “athlete's heart”) causes an opposite response—greater
expression of the fast (α) myosin heavy chain isoform that increases contractility (Scheuer and Buttrick,
1985). The extent of the myosin heavy chain isoform shift in pathological hypertrophy depends on the
severity, duration, and even the nature of the overload; for example, overexpression of the slow β-
isoform in chronic volume overload is less than in chronic pressure overload (Calderone et al., 1995).
Chronic overload and heart failure are also accompanied by isoform shifts in troponin T and actin, but the
more highly conserved tropomyosin, troponin I, and troponin C appear not to undergo isoform shifts. The
troponin T isoform shifts, which result from changes in alternative splicing, increase the content of a fetal
isoform that slows cross-bridge turnover. In some heart failure models, small amounts of α-cardiac actin,
the normal isoform, is replaced with α-skeletal actin, the fetal isoform, but the functional consequences
of this change are not clear.
The Cytoskeleton
Isoform shifts in several cytoskeletal proteins, including titin, α-actinin, myosin-binding protein C,
microtubules, and fibronectin have been found in overloaded and failing hearts. Changes in titin, which is
an important determinant of diastolic compliance, may be clinically significant because different titin
isoforms have been found to be expressed preferentially in systolic and diastolic heart failure; the less
stiff N2BA isoform is increased in dilated hearts whereas more of the stiffer N2B is found in diastolic heart
failure (van Heerebeck et al., 2006).
P.535
Table 18-7 Some Molecular Alterations in Hypertrophied or Failing Hearts§
A. Contractile Proteins
Protein Heart Failure

Myosin heavy chain Expression of low ATPase (β) isoform
Myosin light chains Isoform shift
Actin Isoform shift
Troponin I No change or decreased expression
Troponin T Isoform shift
Troponin C No change
Tropomyosin No change
B. Sarcoplasmic Reticulum Proteins
Protein Heart Failure
Calcium pump ATPase (SERCA2a) Decreased content
Phospholamban Decreased content (less than for SERCA2a)
Calcium release channel (ryanodine Probably decreased, altered function

receptor)
KKBP12.6 (calstabin) Impaired function (calcium leak)
Calsequestrin and calreticulin No change
Calcium/calmodulin-dependent Increased content

protein kinase II
C. Plasma Membrane Proteins
Protein (current) Heart Failure
Sodium channels (iNa) Decreased content and altered substates

Decreased iNa current
Increased iNaL current
Transient outward potassium Decreased content

channels (ito1)
Inward rectifying potassium channels Decreased content

(iK1)
Delayed rectifying potassium No change or decreased content

channels (iKr)
Delayed rectifying potassium Decreased content

channels (iKs)
Hyperpolarization-gated cation Increased content

channels (ih)
L-type calcium channels (iCaL) No change or decreased content Increased or

normal iCaL current
Na/Ca exchanger Increased content
Sodium pump Decreased content, isoform shift
§The changes listed in this table represent an arbitrary overview of the molecular
abnormalities that have been reported in overload-induced hypertrophy and heart failure;
not all have been found in humans, and all experimental models. Published findings disagree
in part because of differences in etiology, the species studied, the age of the animals, and
the severity of the syndromes.
P.536
Gap Junctions
Several mechanisms reduce current flow through gap junctions in the failing heart. In addition to acidosis
and increased cytosolic calcium, which result from energy starvation, internal resistance is increased by a
decrease in connexin phosphorylation and by downregulation of connexin expression by a stress-activated
c-Jun. These changes in the gap junctions slow conduction by increasing internal resistance, and so play
an important role in causing arrhythmias in patients with heart failure.
Plasma Membrane
Important abnormalities of plasma membrane function in failing hearts include action potential
prolongation and resting depolarization (Fig. 18-16), both of which increase the likelihood of reentrant
arrhythmias. Major causes of action potential prolongation are reduced expression of ito1 channels, an
increase in a sodium channel substrate that increases iNaL, and reduced expression of iKs channels.
Decreased expression of the α-subunit of ito (Fig. 18-16A) prolongs the opening of iCaL channels and delays
the opening of the delayed rectifier potassium channels that end the cardiac action potential. Increased
iNaL and decreased expression of the α-subunit of the delayed rectifier channel iKs (Fig. 18-16A), and
possibly also that of iKr, contribute to action potential prolongation in failing hearts. Reduced density of
the inward rectifying potassium channels that carry iK1 decreases resting potential, which slows
depolarization and decreases action potential amplitude by inactivating sodium channels (Fig. 18-16B).
Energy starvation contributes to resting depolarization when increased cellular levels of ADP and
decreased ATP levels reduce iK.ATP. The content of L-type plasma membrane calcium channels is probably
decreased in failing hearts, but iCaL undergoes little change because phosphorylation activates the
remaining channels.
An increase in the inward current generated by the Na/Ca exchanger when it transports calcium from the
cytosol into the extracellular fluid (see Chapter 7) appears to be the most important cause of sudden
death in patients with heart failure. This arrhythmogenic current, which is maximal during the vulnerable
period at the end of the action potential, can initiate after depolarizations, triggered activity, torsades de
pointes, and ventricular fibrillation. The arrhythmogenic calcium efflux by the exchanger is increased
further in failing hearts when cytosolic calcium is increased by energy starvation and inotropic agents that
elevate cytosolic calcium. The latter is the major reason that β-adrenergic agonists and
phosphodiesterase inhibitors, which increase calcium influx via L-type channels, increase mortality in
patients with heart failure.
Reduced expression of the sarcoplasmic reticulum calcium pump ATPase (SERCA2a) slows relaxation in
hypertrophied and failing hearts by decreasing the rate of calcium uptake into this internal membrane
system. Contractility is also depressed because this response reduces intracellular calcium stores (Table
18-7). Decreased phospholamban phosphorylation further slows relaxation and, by decreasing the ability
of cardiac myocytes to replenish their internal calcium stores, adds to the negative inotropic effect of the
decrease in SERCA2a. The density of the intracellular calcium release channels appears either to be
slightly reduced or not to change in heart failure.
Fibrosis
Severe heart failure is accompanied by fibrosis of the walls of the heart. This response impairs filling and,
by slowing impulse conduction, provides a substrate for arrhythmias. However, connective tissue
proliferation can be beneficial because it can slow progressive dilatation (remodeling).
P.537
Fig. 18-16: Major effects of changes in plasma membrane ion channels in the failing heart. A: A
decrease in the transient inward current (ito) slows the cycling of calcium channels and delayed
rectifier potassium channels. The former increases calcium entry and, along with delayed onset of
repolarizing potassium currents, prolongs the action potential. B: Resting depolarization caused by
decreased iK1 inactivates sodium channels, which results in smaller, more slowly rising action
potentials. C: A decrease in iKs, and possibly iKr, prolongs the action potential and contributes to
inhomogeneous repolarization.
P.538
Causes of fibrosis include myocyte necrosis (see earlier) and increased production of matrix proteins by
activated fibroblasts and other connective tissue cells (see Chapter 9). The type of collagen that appears
after the heart is overloaded changes with time. The more elastic embryonic type III collagen is
synthesized preferentially during the initial response, whereas when overload becomes chronic type III
collagen is replaced by type I collagen, which has a higher tensile strength (Chapman et al., 1990;
Marijianowski et al., 1995). Physiological hypertrophy, which lacks adverse long-term effects, is not
accompanied by abnormal fibrosis (Weber and Brilla, 1991).
Stimuli That Cause Hypertrophy

Many different signaling mechanisms mediate cardiac hypertrophy (Fig. 18-17). These include seven
signaling cascades: (1) PI3K/PIP3/Akt pathways, (2) G protein-coupled receptor (GPCR) pathways, (3)
cytoskeletal pathways, (4) peptide growth factor-activated monomeric G protein-mediated pathways, (5)
cytokine-activated pathways, (6) histone deacetylases (HDACs), and (7) microRNA pathways. Because of
overlap and crosstalk between proliferative signaling cascades (see Chapter 9), a given pathway often
mediates both adaptive and maladaptive growth responses, and so can induce both physiological or
pathological hypertrophy.
Pi3K/Pip3/Akt Pathways
Insulin, insulin growth factor (IGF), and growth hormone (GH) evoke proliferative responses when they
bind to receptor tyrosine kinases (RTKs) that activate phosphoinositide 3′-OH kinases (PI3Ks). These
serine/threonine kinases phosphorylate a membrane lipid called phosphatidylinositol tris phosphate (PIP3)
that, when phosphorylated, activates a serine/threonine kinase called Akt (also called protein kinase B or
PKB). The latter then activates mTOR (mammalian target of rapamycin), a regulator of protein synthesis,
and inhibits glycogen synthetase kinase 3 (GSK-3), an inhibitory serine/threonine kinase. This signaling
pathway generally induces physiological hypertrophy, but can also cause pathological hypertrophy.
G Protein-Coupled Receptor (GPCR) Pathways

Mediators of the hemodynamic defense reaction stimulate cellular responses when they bind to GPCRs
that activate heterotrimeric G proteins (see Chapter 8). Binding of β-adrenergic agonists to Gαs activates
adenylyl cyclase to form cyclic AMP (cAMP), an intracellular messenger that activates protein kinase A
(PKA). In addition to its many functional effects (see Chapter 8), PKA stimulates proliferative signaling by
phosphorylating nuclear transcription factors. Binding of angiotensin II, endothelin, vasopressin, and other
extracellular messengers to their GPCRs stimulates proliferative signaling by activating Gαq, which, along
with Gbg, activates phospholipase C (PLC). The latter releases two intracellular messengers:
diacylglycerol (DAG) and inositol tris phosphate (InsP3). DAG activates protein kinase C (PKC), which
phosphorylates nuclear transcription factors, while InsP3-induced calcium release from intracellular stores
activates calcium-calmodulin kinases (CAMKs) that phosphorylate nuclear transcription factors. Calcium
also activates a protein phosphatase, called calcineurin, which dephosphorylates and activates nuclear
factor of activated T cell. The latter then activates transcription factors called MEF2C and GATA4 whose
predominant effect is to stimulate pathological hypertrophy.
P.539
Cytoskeletal Pathways
Cytoskeletal signaling, which is emerging as one of the most important regulators of cardiac hypertrophy,
can be viewed as providing the “local” control of growth needed to maintain homogeneity of contraction
by matching myocyte size and shape to local stresses (Katz and Katz, 1989). The remarkable precision of
the responses to local stresses is seen when cardiac myocytes are cultured on surfaces that limit their
ability to grow. The shape of these surfaces determines not only the outlines of the cultured cells, but
also their internal architecture (Parker et al., 2008) (Fig. 18-18). In overloaded and failing hearts,
cytoskeletal signaling evokes changes in cell size and shape that can be both adaptive and maladaptive.
Peptide Growth Factor-Activated Monomeric G Protein-Mediated

Pathways
Monomeric G protein-mediated signaling be activated when peptide growth factors activate receptor
tyrosine kinases; many of these pathways are mediated when monomeric G proteins, such as Ras, activate
MAPK pathways. The four major MAPK pathways described in Chapter 9, extracellular receptor-mediated
kinases (ERK), c-Jun kinase (JNK), p38 kinase (p38 K), and extracellular receptor-mediated kinase-5 (ERK-
5) can all be activated in failing hearts. Like the GPCR-activated pathways, the major response to MAPK
activation is pathological hypertrophy.
Cytokine Pathways
Ligand-bound cytokine receptors regulate proliferative signaling by dissociating an inhibitory protein
called IkB from the transcription factor NFkB; this allows the latter to move to the nucleus where it
stimulates gene transcription (see Chapter 9). Cytokines also activate MAP kinase-mediated proliferative
responses when gp130-mediated signaling pathways stimulate janus kinase (Jak), which activates both
MAP kinase pathways and a transcription factor called signal transducer and activator of transcription
(STAT). Cytokine-activated proliferative signaling pathways participate in both physiological and
pathological hypertrophy; it is not clear which is the predominant response, but adverse effects in gp130
knockout mice and the disappointing results of anti-cytokine therapy in humans may be due in part to
inhibition of adaptive hypertrophy.
Histone Deacetylase (HDAC) Pathways

Histone acetyltransferases (HATs) mediate an epigenetic mechanism that regulates cardiac hypertrophy
when acetylation of histones exposes genes for transcription by “unwinding” double-stranded DNA.
Reversal of this activation reaction by histone deacetylation inhibits the activation of proliferative
signaling. Class II HDACs, which catalyze histone deacetylation, are activated in the heart when they are
phosphorylated by CAMK and protein kinase D (PKD) after the latter is activated by PKC-catalyzed
phosphorylations. The major effect of histone deacetylation is to stimulate pathological hypertrophy.
MicroRNA Pathways
Dozens of micro-RNAs (miRNAs) are now known to regulate proliferative signaling in the heart. These
small RNA fragments combine with regulatory proteins to form RNA-induced silencing complexes that bind
specifically to messenger RNA (mRNA) where they inhibit (“silence”) mRNA transcription. Because this
field is growing very rapidly (Naga Prasad et al., 2009), it is not possible now to characterize the role of
miRNAs in heart failure.
P.540
Fig. 18-17: Major proliferative signaling pathways that mediate cardiac hypertrophy are divided into seven
groups: (1) PI3K/PIP3/Akt pathways, (2) G protein-coupled receptor (GPCR)-mediated pathways, (3)
cytoskeletal pathways, (4) peptide growth factor-monomeric G protein-mediated pathways, (5) cytokine-
activated pathways, (6) histone deacetylases (HDACs), and (7) micro-RNAs (miRNAs). The major response to
the PI3K/PIP3/Akt pathways is physiological hypertrophy, whereas the major response to GPCR and
monomeric G protein-mediated signaling, and HDACs is pathological hypertrophy. Cytoskeletal signaling,
which mediates local changes that adapt form to function, and cytokines mediate both physiological and
pathological hypertrophy. The PI3K/PIP3/Akt pathway is activated when insulin, insulin growth factor (IGF),
and growth hormone (GH) bind to receptor tyrosine kinases (RTKs). The latter activates phosphoinositide 3′-
OH kinases (PI3K) that phosphorylate phosphatidylinositol tris phosphate (PIP3), which activates a protein
kinase called Akt. The latter then activates a proliferative signaling regulator called mammalian target of
rapamycin (mTOR) and inhibits an inhibitory serine/threonine kinase called glycogen synthetase kinase 3
(GSK-3); the major response to this pathway is physiological hypertrophy. Mediators of the hemodynamic
defense reaction, such as β-adrenergic receptor agonists, angiotensin II (Ang II), and endothelin, bind to
GPCRs that interact with heterotrimeric G proteins to activate Gα and Gbg; these pathways preferentially
activate pathological hypertrophy. Binding of β-adrenergic agonists to Gα activates adenylyl cyclase (AC) to
form cyclic AMP (cAMP) that stimulates protein kinase A (PKA) which phosphorylates several transcription
factors. Other Gα isoforms, notably Gαq, along with Gbg, activate phospholipase C (PLC) to form
diacylglycerol (DAG) and inositol tris phosphate (InsP3). DAG activates PKC, which phosphorylates additional
transcription factors. Calcium released from intracellular stores by InsP3 also activates PKC and calcium-
calmodulin kinase (CAMK), both of which phosphorylate transcription factors. Calcium also activates
calcineurin, a protein phosphatase that dephosphorylates and activates nuclear factor of activated T cell
(NFAT), which activates the transcription factors MEF2C and GATA4 that stimulate pathological hypertrophy.
Cell deformation activates cytoskeletal signaling pathways that mediate local responses which can be both
adaptive and maladaptive. Monomeric G protein-mediated pathways are activated when peptide growth
factors bind to RTKs in reactions that are mediated by Ras and other monomeric G proteins. The latter
stimulate mitogen-activated protein kinase (MAPK) pathways that include extracellular receptor-mediated
kinases (ERK), c-Jun kinase (JNK), p38 kinase (p38 K), and extracellular receptor-mediated kinase-5 (ERK5).
The JNK and p38 pathways are also activated by a variety of stresses. ERK, JNK, and p38 K phosphorylate
transcription factors that generally favor pathological hypertrophy; less is known of the role of ERK5-
mediated signals. Cytokine-bound receptors regulate proliferative signaling by releasing an inhibitory effect
of IkB on the transcription factor NFkB; cytokine receptors also activate gp130-mediated signaling pathways
that stimulate Janus kinase (Jak) to activate transcription factors called signal transducer and activator of
transcription (STAT). Cytokines appear to induce both physiological and pathological hypertrophy. Two
epigenetic pathways also mediate the hypertrophic response. Calcium and PKC can stimulate hypertrophy by
activating a CAMK and protein kinase D (PKD), respectively. Both of the latter catalyze phosphorylations that
inactivate Class II histone deacetylases (HDACs), which increases histone acetylation. This causes pathological
hypertrophy when unwinding of histones exposes DNA to transcription factors. Micro-RNAs (miRNA) aggregate
with regulatory proteins to form RNA-inducing silencing complexes (RISC) that inhibit transcription of
messenger RNA (mRNA). KEY: Thin solid arrows indicate signaling pathways, dashed arrows with solid
arrowheads indicate phosphorylations (P), dephosphorylations (deP), and acetylations (Ac) that activate
transcription factors. Transcription factors are indicated by dashed arrows with shaded arrowheads.
P.541
Fig. 18-18: Effects of different mechanical stresses on the growth of Japanese watermelons and
cardiac myocytes. A: Left to right: Japanese watermelons grown without constraint, and in a
pyramidal and a square box. B: Cardiac myocytes cultured on differently shaped adherent islands
that direct myocyte shape. a: Freshly isolated myocyte. b: A myocyte placed on a large adherent
surface that allows free growth. c: Myocyte placed on a rectangular island. d: Myocyte placed on a
triangular island. e: Myocyte placed on a star-shaped island. (Images in B are modified from Parker
et al., 2008.)
P.542
Arrhythmias and Sudden Death
Arrhythmias, which account for more than half of the deaths in patients with heart failure, are caused by
several mechanisms. These include architectural abnormalities such as cardiac enlargement and fibrosis,
which favor reentry by slowing and disorganizing impulse conduction (see Chapter 16). Arrhythmias are
also initiated when conduction is slowed by physiological abnormalities like resting depolarization, which
is due in part to sodium pump inhibition by energy starvation (see Chapter 7) and closure of gap junction
connexin channels by acidosis and elevated intracellular calcium (see Chapter 13). Many of the molecular
abnormalities described in this chapter also increase the susceptibility of heart failure patients to sudden
cardiac death. Decreased iNa slows conduction, which favors reentry, as does resting depolarization
caused by the reduced number of iK1 channels and decreased content of the sodium pump. Increased iNaL
and decreased ito, iKs, and possibly iKr, increase the heterogeneity of repolarization by prolonging the
action potential (see Chapter 16), while the increased content of hyperpolarization-gated cation channels
(ih) accelerates lower pacemakers. The most important cause of sudden death in patients with heart
failure appears to be triggered arrhythmias caused by the inward current generated by the Na/Ca
exchanger as the result of electrogenic calcium efflux during the vulnerable period (see Chapters 7 and
10). This arrhythmogenic ion flux is increased in heart failure by calcium overload (see earlier) and by
reversion to the fetal phenotype, which reduces the content of the sarcoplasmic reticulum calcium pump
and increases the heart's dependence on the extracellular calcium cycle (Louch et al., 2010).
Familial Cardiomyopathies
In 1990, the report of myosin heavy chain mutations in familial hypertrophic cardiomyopathies
(Geisterfer-Lowrance et al., 1990) opened a new era in understanding the pathophysiology of heart
failure. Today, hundreds of mutations affecting genes that encode dozens of proteins are known to cause
both hypertrophic (HCM) and dilated (DCM) cardiomyopathies (Table 18-8). Although a single missense
mutation or a small deletion can lead to severe disease, not all individuals who carry a specific gene
abnormality will exhibit the typical phenotype and some may have few, if any, cardiac abnormalities.
Several architectural phenotypes are seen in HCM; most common is asymmetrical hypertrophy of the
interventricular septum, but both apical and diffuse hypertrophy are also seen. The typical histological
picture is myocyte hypertrophy, with loss of the parallel orientation of the myofibrils within and among
adjacent myocytes, called “myofibrillar disarray.” This disorganization leads to heterogeneities in
mechanical performance that reduce cardiac efficiency, and may also generate proliferative stimuli that
lead to concentric hypertrophy. In contrast, the hearts of most patients with DCM are uniformly dilated,
and histology shows myocyte degeneration and fibrosis.
Most mutations that cause HCM involve the contractile proteins; abnormalities in other proteins also cause
this syndrome, including myosin-binding protein C, a sarcomeric protein, and a number of other proteins,
including AMP-activated protein kinase, galactosidase A, lysosome-associated membrane protein 2,
vinculin, and several cytoskeletal proteins. A variety of mutations also cause DCM (Table 18-8); many of
these are cytoskeletal, but mutations in other proteins also cause this syndrome along with a related
syndrome, called left ventricular noncompaction, in which the left ventricular wall is thickened but
spongy because it contains deep recesses. In some
P.543
P.544
cases, different mutations in the same protein can cause either a hypertrophic or a dilated
cardiomyopathy, so that the link between specific mutations and different patterns of hypertrophy is not
clear. Some of these architectural differences may occur when subtle abnormalities in deforming stresses
activate different cytoskeletal signaling pathways.
Table 18-8 Some Molecular Causes of Human Familial Cardiomyopathies
Hypertrophic Dilateda
Myofibrillar proteins
β-Myosin heavy chain x x
α-Myosin heavy chain x
Regulatory myosin light chain x
Essential myosin light chain x
Troponin T x x
Troponin I x x
Troponin C x x
α-Tropomyosin x x
Cardiac α-actin x x
Cytoskeletal proteins Sarcomere related
Myosin-binding protein C x x
Titin x x
Z-line-related
α-Actinin x
Muscle LIM proteins x x
Cypher/ZASP x
Vinculin x
Metavinculin x
Obscurin x
Calsarcin 1 (myozenin 2) x
T-cap x x
Telethonin x
Dystrophin-related
Sarcoglycans (α, β, γ, σ) x
Dystrophin x
Dystrobrevin x
Syntrophin x
Caveolin-3 x
Desmosome-related, intercalated disc
Desmin x
Plakoglobin x
Desmoplakinb x
Desmoglein-2b x
Desmocollin-2b x
Plakophilin-2b x
Vinculin x
Nuclear
Emerin x
Lamin A/C x
Membrane proteins
Calcium release channel (ryanodine receptor)b x x
Phospholamban x
Sodium channel (iNa) x
ATP-regulated potassium channel (iK,ATP) x
Metabolic proteins
AMP-activated protein kinasec x
Lysosome-associated membrane protein-2 x x
(Danon's disease)c
α-Galactosidase A (Fabry's disease)c x
α-1,4 Glucosidase (Pompe's disease)c x
Polysaccharide metabolism x
(Hurler, Hunter, and Morquio syndromes)
Carnitine transporter (carnitine deficiency) x
Tafazzin (phospholipid acyltransferase) x x
(Barth syndrome)
Frataxin (heme synthesis) x
(Friedreich's ataxia)
Acyl-CoA deficiency x
Mitochondrial DNA (point mutations) x x
Other proteins
Tyrosine phosphatase SHP-2 (Noonan syndrome) x
Dystrophia myotonica protein kinase (Myotonic dystrophy) x
Nkx2.5 (homeobox) x
aLeft ventricular noncompaction is included in this category.

bAssociated with arrhythmogenic right ventricular cardiomyopathy.
cGlycogen storage disease.
Some Therapeutic Implications
Until the late 1980s, treatment of patients with heart failure centered on correction of the hemodynamic
abnormalities; diuretics were given to reduce preload, arteriolar vasodilators to reduce afterload, and
inotropes to increase contractility. During the past 30 years, however,
P.545
clinical trials showed that most vasodilators and virtually all inotropes worsen prognosis. Cardiac
glycosides, once the mainstay of therapy in these patients, have been largely abandoned because of their
toxicity, while β-adrenergic blockers are now viewed as among the most effective way to prolong survival.
The reason why direct-acting arteriolar vasodilators fail to improve long-term survival is probably that
they cause a fall in blood pressure that activates the neurohumoral response. This explanation is
consistent with evidence that nitrates, ACE inhibitors, and angiotensin II receptor blockers, all of which
are vasodilators that improve prognosis in heart failure, directly inhibit maladaptive proliferative
signaling pathways that lead to progressive dilatation (remodeling). Similarly, the long-term beneficial
effect of β-blockers, which because of their negative inotropic effect had been almost universally viewed
as contraindicated in heart failure, appears to be due largely to inhibition of remodeling. Aldosterone
antagonists, which had been used for decades as potassium-sparing diuretics, were tested for their ability
to improve prognosis in patients with heart failure because they inhibit fibrosis. This class of drugs did
prolong survival, but it now appears that the survival benefits are due at least in part to their ability to
slow progressive dilatation. Cardiac resynchronization therapy, which was introduced to minimize the
heterogeneities caused by delayed depolarization in some regions of the left ventricle in patients with
heart failure, and left ventricular assist devices, which are used as a “bridge to transplant” in dying
patients, can also reverse ventricular dilatation in end-stage heart failure. Common to all of these
successful treatments is that they slow, and sometimes temporarily reverse, ventricular remodeling.
It is important to realize that the trials which established the long-term therapeutic benefits described
earlier included mainly patients with a low EF (systolic heart failure), where progressive dilatation is a
major cause of progression and premature death. These findings therefore raise an important question,
whether similar benefits will be seen in the growing population of patients, mostly elderly, who have
diastolic heart failure in which, by definition, remodeling does not occur. It seems unlikely that therapy
that improves prognosis in systolic heart failure by inhibiting or reversing progressive dilatation will have
the same benefit in diastolic heart failure. Until this and other questions about diastolic heart failure are
answered by appropriately designed clinical trials, optimal management of this syndrome will remain
among the most important unresolved challenges in treating heart failure.
Some Final Comments and A Speculation About the Future

There is no better way to conclude this text than with a chapter on heart failure, as this documents the
practical importance of virtually all of the material covered in preceding pages. Energy production and
energy utilization, the contractile proteins and cytoskeleton, the extracellular and intracellular calcium
cycles, functional and proliferative signaling, the electrical activity of the heart, and cardiovascular
hemodynamics are all of practical importance in understanding and treating this syndrome. The increasing
clinical relevance of these fields of study, which until a few years ago were viewed as arcane basic
science that was of little or no help in managing patients, clearly demonstrates the rapidity with which
modern biology is closing the “gap” between bench and bedside (Katz, 2008a, 2008b).
An analogy that the author has found to be useful in discussing the management of heart failure is to view
the failing heart as a sick, tired horse pulling a load up a hill (Fig. 18-19). This analogy is becoming
increasingly relevant as the flood of new information about the
P.546
molecular causes of this syndrome is bringing the “ideal” solution, to heal the horse, within our grasp.
Fig. 18-19: View of the failing heart as a sick, tired horse pulling a heavily loaded wagon up a steep hill.
Although application of a whip (inotropes) encourages the horse to move faster, this can kill the animal.
Unloading the wagon (vasodilators) would seem to be advantageous, but in heart failure, this approach can
do harm by activating deleterious neurohumoral responses. Slowing the horse (β-blockers), while delaying the
journey, can be beneficial, especially if this also helps to heal the horse. Replacing the horse (cardiac
transplantation) is useful as long as there are enough spare horses, while getting a tractor will be a
reasonable solution when reliable and inexpensive machines become available. The ideal solution, of course,
is to learn what ails the animal, and then use this information to heal the horse.
A clear example of the practical importance of molecular biology is seen in the variety of proteins that,
when mutated, can cause the heart to fail—and in different ways. Even benign polymorphisms influence
clinical outcome in this syndrome. For example, β-blockers have been found to be of much less benefit in
African-American patients than those of European descent. This difference is caused by a genetic variant
in a G protein-coupled receptor kinase (GRK) that is common in the African-American population. Because
this GRK variant desensitizes β-adrenergic receptors to their agonists (Liggett et al., 2008), it protects
against the adverse effects of the high circulating levels of β-adrenergic agonists seen in heart failure.
This is a major reason why the response of African-Americans to β-blocker therapy is much less than in
the European-American population, in whom this GRK variant is rare.
These findings, along with the improving ability to identify patients at high risk for adverse reactions to
specific drugs, indicate that we are now in an era of “individualized” medicine, where therapy can be
tailored to the pathophysiology in the individual patient rather than the average patient in a large
population. It is a remarkable coincidence that the same
P.547
week that the manuscript of this chapter was completed, a special issue of Heart Failure Reviews that
was devoted to the pharmacogenetics of heart failure therapy (Lanfear, 2010) was delivered to the
author's mailbox.
The concept of individualized therapy is not new, however. More than 2,000 years ago, when discussing
the “absolutes,” Aristotle wrote in his Nicomachean Ethics (I.vi.14–16):
it is not easy to see … how anybody will be a better physician … for having
contemplated the absolute idea. In fact, it does not appear that the physician
studies even health in the abstract: he studies the health of the human being—or
rather of some particular human being, for it is individuals he has to cure.
Bibliography
General
Bugaisky L, Gupta M, Gupta MG, et al. Cellular and molecular mechanisms of hypertrophy. In:
Fozzard H, Haber E, Katz A, et al., eds. The heart and cardiovascular system. 2nd ed. New York:
Raven Press, 1992:1621–1640.
Collucci WS. Atlas of heart failure. Cardiac function and dysfunction. 4th ed. Philadelphia: Current
Med, 2005.
Katz AM, Konstam MA. Heart failure. Pathophysiology, molecular biology and clinical management.
2nd ed. Philadelphia: Lippincott Williams & Wilkins, 2006.
Mann DL, ed. Heart failure: A companion to braunwald's heart disease 1st ed. Philadelphia:
Saunders, 2004.
Miller AJ. Lymphatics of the Heart. New York, NY: Raven Press, 1982.
Rumyantsev PP. Interrelations of the proliferation and differentiation of processes during cardiac
myogenesis and regeneration. Int Rev Cytol 1977;51:187–273.
Swynghedauw B. Molecular mechanisms of myocardial remodeling. Physiol Rev 1999;79:215–262.

Yano M, Ikeda Y, Matsuzaki M. Altered intracellular Ca2+ handling in heart failure. J Clin Invest
2005;115:556–564.
Additional discussions are found in textbooks of Medicine and Cardiology.
Cell Death
Dorn GW II. Apoptotic and non-apoptotic programmed cardiomyocyte death in ventricular

remodelling. Cardiovasc Res 2009;81:465–473.
Foo RS-Y, Mani K, Kitsis RN. Death begets failure in the heart. J Clin. Invest 2005;115:565–571.
Gustafsson AB, Gottlieb RA. Mechanisms of apoptosis in the heart. J Clin Immunol 2004;23:447–459.
van Empel VP, De Windt LJ. Myocyte hypertrophy and apoptosis: a balancing act. Cardiovasc Res
2004;63:487–499.
Whelan RS, Kaplinskiy V, Kitsis RN. Cell death in the pathogenesis of heart disease: mechanisms and
significance. Ann Rev Physiol 2009;72:19–44.
Hypertrophic Signaling
Akazawa H, Komuro I. Roles of cardiac transcription factors in cardiac hypertrophy. Circ Res
2003;92:1079–1088.
Bueno OF, Molkentin JD. Involvement of extracellular signal-regulated kinases 1/2 in cardiac
hypertrophy and cell death. Circ Res 2002;91:776–781.
Dorn GW II, Force T. Protein kinase cascades in the regulation of cardiac hypertrophy. J Clin Invest
2005;115:527–537.
P.548
Frey N, Katus HA, Olson EN, et al. Hypertrophy of the heart. A new therapeutic target. Circulation
2004;109:1580–1589.
Hoshijima M, Chien KR. Mixed signals in heart failure: cancer rules. J Clin Invest 2002;1098:849–855.
Kelly DP. Peroxisome proliferator-activated receptor α as a genetic determinant of cardiac

hypertrophic growth. Culprit or innocent bystander. Circulation 2002;105:1025–1027.
Lips DJ, deWindta LJ, van Kraaij DJW, et al. Molecular determinants of myocardial hypertrophy and
failure: alternative pathways for beneficial and maladaptive hypertrophy. Eur Heart J 2003;24:883–
896.
McKinsey TA, Olson EN. Toward transcriptional therapies of the failing heart: chemical screens to
modulate genes. J Clin Invest 2005;115:538–546.
Architectural Changes
Gerdes AM. Cardiac myocyte remodeling in hypertrophy and progression to failure. J Card Fail
2002;8(Suppl):S264–S264.
Goldsmith EC, Borg TK. The dynamic interaction of the extracellular matrix in cardiac remodeling. J
Card Fail 2002;8(Suppl):S314–S318.
Li F, Wang X, Yi XP, et al. Structural basis of ventricular remodeling: role of the myocyte. Curr Heart
Fail Rep 2004;1:5–8.
Ross RS. The extracellular connections: the role of integrins in cardiac remodeling. J Card Fail
2002;8(Suppl):S326–S331.
Saffitz JE, Klâber AG. Effects of mechanical forces and mediators of hypertrophy and remolding of
gap junctions in the heart. Circ Res 2004;94:585–591.
Selvetella G, Hirsch E, Notte A, et al. Adaptive and maladaptive hypertrophic pathways: points of
convergence and divergence. Cardiovasc Res 2004;63:373–380.
Energetics
Finck BN, Kelly DP. Peroxisome proliferator-activated receptor α (PPARα) signaling in the gene
regulatory control of energy metabolism in the normal and diseased heart. J Mol Cell Cardiol
2002;34:1249–1257.
Giordano FJ. Oxygen, oxidative stress, hypoxia, and heart failure. J Clin Invest 2005;115:500–508.
Huss JM, Kelly DP. Mitochondrial energy metabolism in heart failure: a question of balance. J Clin
Invest 2005;115:547–555
Ingwall JS. Energy metabolism in heart failure and remodelling. Cardiovasc Res 2009;81:412–419.
Mettauer B, Zoll J, Garnier A, et al. Heart failure: a model of cardiac and skeletal muscle energetic
failure. Pflugers Arch 2009;452:653–666.
Neubauer S. The failing heart—an engine out of fuel. New Engl J Med 2007;356:1140–1151.
Ion Channels
Amin AS, Tan HL, Wilde AAM. Cardiac ion channels in health and disease. Heart Rhythm 2009;7:118–
126.
Biel M. Cyclic nucleotide regulated cation channels. J Biol Chem 2009;284:9017–9021.
Brette F, Leroy J, Le Guennec JY, et al. Ca2+ currents in cardiac myocytes: old story, new insights.
Biophys Mol Biol 2006;91:1–82.
Nattel S, Maguy A, Le Bouter S, et al. Arrhythmogenic ion-channel remodeling in the heart: heart
failure, myocardial infarction, and atrial fibrillation. Physiol Rev 2007;87:425–445.
Vander Heyden MAG, Wijnhoven TJM, Opthof T. Molecular aspects of adrenergic modulation of the
transient outward current. Cardiovasc Res 2006;71:430–442.
Zaza A, Belardinelli L, Shrylock JC. Pathophysiology and pharmacology of the “late sodium current.”
Pharmacol Ther 2008;119:326–339.
P.549
Familial Cardiomyopathies
Arad M, Maron BJ, Gorham JM, et al. Glycogen storage diseases presenting as hypertrophic
cardiomyopathy. New Engl J Med 2005;352:362–372.
Arad M, Seidman JG, Seidman CE. Phenotypic diversity in hypertrophic cardiomyopathy. Hum Mol
Genet 2002;11:2499–2506.
Gomes AV, Potter JD. Molecular and cellular aspects of troponin cardiomyopathies. Ann N Acad Sci
2004;1015:214–224.
Hershberger RE, Cowan J, Morales A, Siegfried JD. Progress with genetic cardiomyopathies
screening, counseling, and testing in dilated, hypertrophic, and arrhythmogenic right ventricular
dysplasia/cardiomyopathy. Cir Heart Fail 2009;2:253–261.
Klaassen S, Probst S, Oechslin E, et al. Mutations in sarcomeric protein genes in left ventricular non-
compaction. Circulation 2008;117:2893–2901.
Morimoto S. Sarcomeric proteins and inherited cadiomyopathies. Cardiovasc Res 2008;77:659–666.
Morita H, Seidman J, Seidman CE. Genetic causes of human heart failure. J Clin Invest
2005;115:518–526.
Wang L, Seidman JG, Seidman CE. Narrative review: harnessing molecular genetics for the diagnosis
and management of hypertrophic cardiomyopathy. Ann Int Med 2010;152:513–520.
Skeletal Muscle Abnormalities
Fitts RH. Cellular mechanisms of muscle fatigue. Physiol Rev 1994;74:49–94.
Libera LD, Vescovo G. Muscle wastage in chronic heart failure, between apoptosis, catabolism and
altered anabolism: a chimaeric view of inflammation? Curr Opin Clin Nutr Metab Care 2004;7:435–
441.
Mancini DM, LaManca J, Henson D. The relation of respiratory muscle function to dyspnea in patient
with heart failure. Heart Fail 1992;8:183–189.
Niebauer J, Clark AJ, Webb-Peploe KM, et al. Exercise training in chronic heart failure: effects on
proinflammatory markers. Eur J Heart Fail 2005;7:189–193.
Wilson JR. Exercise intolerance in heart failure. Importance of skeletal muscle. Circulation 1995;91:
559–661.
References
Alpert NR, Gordon MS. Myofibrillar adenosine triphosphatase activity in congestive heart failure. Am
J Physiol 1962;202:940–946.
Baim DS, Grossman W. Grossman's Cardiac Catheterization, Angiography and Intervention.

Philadelphia, PA: Lippincott/Williams & Wilkins, 2000.
Boateng SY, Senyo SE, Qi Y, et al. Myocyte remodeling in response to hypertrophic stimuli requires
nucleocytoplasmic shuttling of muscle LIM protein. J Mol Cell Cardiol 2009;47:426–535.
Calderone A, Takahashi N, Izzo NJ Jr, et al. Pressure- and volume-induced left ventricular
hypertrophies are associated with distinct myocyte phenotypes and differential induction of peptide
growth factor mRNAs. Circulation 1995;92:2385–2390.
Chapman D, Weber KT, Eghbali M. Regulation of fibrillar collagen types I and III and basement
membrane type IV collagen gene expression in pressure overloaded rat myocardium. Circ Res
1990;67:787–794.
Geisterfer-Lowrance AAT, Kass S, Tanigawa G, et al. A molecular basis for familial hypertrophic
cardiomyopathy: a β-cardiac myosin heavy chain gene missense mutation. Cell 1990;62:999–1006.
Gerdes AM, Kellerman SE, Malec KB, et al. Transverse shape characteristics of cardiac myocytes from
rats and humans. Cardioscience 1994;5:31–36.
Grant C, Greene DG, Bunnell IL. Left ventricular enlargement and hypertrophy. A clinical and
angiocardiographic study. Am J Med 1965;39:895–904.
Grossman W, Jones D, McLaurin LP. Wall stress and patterns of hypertrophy in the human left
ventricle. J Clin Invest 1975;56:56–64.
P.550
Heart disease and stroke statistics—2010 update: a report from the American Heart Association.
Circulation 2010;121:e46–e215.
Hood WP Jr, Rackley CE, Rolett EL. Wall stress in the normal and hypertrophied human left ventricle.
Am J Cardiol 1968;22:5550–5558.
Kass DA. Evaluation of left-ventricular systolic function. Heart Fail 1988;4:198–205.
Katz AM. Cardiomyopathy of overload. A major determinant of prognosis in congestive heart failure.
New Engl J Med 1990a;322:100–110.
Katz AM. Angiotensin II: hemodynamic regulator or growth factor? J Mol Cell Cardiol 1990b;22: 739–
747.
Katz AM. The “gap” between bench and bedside: widening or narrowing. J Card Fail 2008a;14:91–94.
Katz AM. The “modern” view of heart failure: how did we get here? Circ Heart Fail 2008b;1:63–71.
Lanfear DE, ed. Special issue on pharmacogenetic of heart failure therapies. Heart Fail Rev 2010;15:
183–248.
Levy WC, Mozaffarian D, Linker DT, et al. The Seattle Heart Failure Model: prediction of survival in
heart failure. Circulation 2006;113:1424–1433.
Liggett SB, Cresci S, Kelly RJ, et al. A GRK5 polymorphism that inhibits β-adrenergic receptor
signaling is protective in heart failure. Nature Med 2008;14:510–517.
Louch WE, Hougen K, Mârk HK, et al. Sodium accumulation promotes diastolic dysfunction in end-
stage heart failure following Serca2 knockout. J Physiol 2010;588:465–478.
Marijianowski MMH, Teeling P, Mann J, et al. Dilated cardiomyopathy is associated with an increase
in the type I/type III collagen ratio: a quantitative assessment. J Am Coll Cardiol 1995;25:1263–
1272.
Meerson FZ. On the mechanism of compensatory hyperfunction and insufficiency of the heart. Cor
Vasa 1961;3:161–177.
Naga Prasad SV, Duan ZH, Gupta MK, et al. Unique microRNA profile in end-stage heart failure
indicates alterations in specific cardiovascular signaling networks. J Biol Chem 2009;284:27487–
27499.
Osler W. The principles and practice of medicine. NY: Appleton, 1892:634.
Parker KK, Tan J, Chen CS, et al. Myofibrillar architecture in engineered cardiac myocytes. Circ Res
2008;103:340–342.
Pfeffer JM, Pfeffer MA, Braunwald E. Influence of chronic captopril therapy on the infarcted left
ventricle of the rat. Circ Res 1985;57:84–95.
Russell B, Motlagh GHG, Ashley WW. Form follows function: how muscle shape is regulated by work.
J Appl Physiol 2000;88:1127–1132.
Sandler H, Dodge HT. Left ventricular tension and stress in man. Circ Res 1963;13:91–104.
Scheuer J. Buttrick P. The cardiac hypertrophic responses to pathologic and physiologic loads.
Circulation 1985;75(1 Pt 2):I63–I68.
Senyo SE, Koshman YE, Russell B. Stimulus interval, rate and direction differentially regulate
phosphorylation for mechanotransduction in neonatal cardiac myocytes. FEBS Lett 2007;581:4241–
4247.
van Heerebeck L, Borbâly A, Niessen HWM, et al. Myocardial structure and function differ in systolic
and diastolic heart failure. Circulation 2006;113:1966–1973.
Weber KT, Brilla CG. Pathological hypertrophy and cardiac interstitium. Fibrosis and the renin-
angiotensin-aldosterone system. Circulation 1991;83:1849–1865.
Wollert KC, Taga T, Saito M, et al. Corticotrophin-1 activates a distinct form of cardiac muscle cell
hypertrophy. Assembly of sarcomeric units in series via gp130/leukemia inhibitory factor receptor-
dependent pathways. J Biol Chem 1996;271:9535–9545.
Wood P. An appreciation of mitral stenosis. Br Med J 1954;1:1051–1063, 1113–1124.
Woodward WA, Strom EA, Tucker SL, et al. Changes in the 2003 American Joint Committee on Cancer
Staging for breast cancer dramatically affect stage-specific survival. J Clin Oncol 2003;21:3244–
3248.
Yamamoto K, Dang QN, Maeda Y, et al. Regulation of cardiac myocyte mechanotransduction by the
cardiac cycle. Circulation 2001;103:1459–1464.

Fisiologia de Katz

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Fisiologia de Katz

Uploaded by

Copyright:

Available Formats

Authors: Katz, Arnold M.

Title: Physiology of the Heart, 5th Edition

Lippincott Williams & Wilkins

© 2011 by LIPPINCOTT WILLIAMS & WILKINS, a WOLTERS KLUWER business

Acquisitions Editor: Frances DeStefano

Product Manager: Leanne McMillan

Production Manager: Alicia Jackson

Senior Manufacturing Manager: Benjamin Rivera

Marketing Manager: Kimberly Schonberger

Design Coordinator: Stephen Druding

Production Service: Aptara, Inc.

Library of Congress Cataloging-in-Publication Data

Physiology of the heart / Arnold M. Katz. — 5th ed.

Includes bibliographical references and index.

1. Heart—Physiology. 2. Heart—Pathophysiology. I. Title.

[DNLM: 1. Heart—physiology. 2. Heart Diseases—physiopathology. WG 202 K19p 2011]

> Front of Book > Authors

> Front of Book > Foreword

> Front of Book > Preface to the First Edition

Preface to the First Edition

> Front of Book > Dedication

> Front of Book > Preface

> Front of Book > Acknowledgments

—William Harvey (1628).

Exercitatio Anatomica de Moto Cordis et Sanguinis in Animalibus.

Architecture of the Walls of the Heart

Major Epicardial Coronary Arteries

Blood Supply to the Ventricular Myocardium

Blood Supply to the Conduction System

Coronary Venous Drainage

Fractal Anatomy of the Heart

Table 1-1 Components of a Working Myocardial Cell (Rat Left Ventricle)

Component Percentage of Cell Volume

Sarcoplasmic reticulum 3.5

Subsarcolemmal cisternae 0.35

Sarcotubular network 3.15

Other (mainly cytosol) 11.5

Modified from Page (1978).

Membrane µm2 Membrane Area Per µm3 Cell Volume

Plasma membrane 0.465

Total sarcoplasmic reticulum 1.22

Sarcotubular network 1.03

Modified from Page (1978).

The Plasma Membrane and Transverse Tubules

Intracellular Membrane Structures

Table 1-3 Cell-to-Cell Communication Across the Intercalated Disc

Fascia Mechanical N-cadherin β-Catenin Microfilament

Gap Electrical Connexin 43 Ion channel

Membrane Structure and Function

Benninghoff A. Lehrbuch der Anatomie des Menschen. Munich: JF Lehmanns, 1944.

Brutsaert DL. The endocardium. Ann Rev Physiol 1989;51:263–273.

Katz AM, Katz PB. Homogeneity out of heterogeneity. Circulation 1989;79:712–717.

Lower R. Tractus de Corde. London: Allestry, 1669.

Page E. Quantitative ultrastructural analysis in cardiac membrane physiology. Am J Physiol 1978;

Yacoub MH. Two hearts that beat as one. Circulation 1995;92:160–161.

Table 2-1 Biochemical Differences between Red and White Muscle

Biochemical Characteristic Red Muscle White Muscle

Substrates Lipid, carbohydrate Carbohydrate

Metabolites CO2, H2O Lactic acid

Glycolytic enzymes Sparse Abundant

Mitochondria Abundant Sparse

Phosphocreatine stores Minor Significant

Dependence on oxygen Marked Little