Zinc Finger Proteins From Atomic Contact To Cellular Function
Zinc Finger Proteins From Atomic Contact To Cellular Function
INTELLIGENCE
UNIT
Shiro Iuchi
Department of Cell Biology
Harvard Medical School
Boston, Massachusetts, U.S.A.
Natalie Kuldell
Biological Engineering Division
Massachusetts Institute of Technology
Cambridge, Massachusetts, U.S.A.
Copyright ©2005 Landes Bioscience / Eurekah.com and Kluwer Academic / Plenum Publishers
Kluwer Academic / Plenum Publishers, 233 Spring Street, New York, New York, U.S.A. 10013
http://www.wkap.nl
Zinc Finger Proteins: From Atomic Contact to Cellular Function, edited by Shiro Iuchi and Natalie Kuldell, Landes / Kluwer
dual imprint / Landes series: Molecular Biology Intelligence Unit
ISBN: 0-306-48229-0
While the authors, editors and publisher believe that drug selection and dosage and the specifications and usage of equipment and
devices, as set forth in this book, are in accord with current recommendations and practice at the time of publication, they make no
warranty, expressed or implied, with respect to material described in this book. In view of the ongoing research, equipment
development, changes in governmental regulations and the rapid accumulation of information relating to the biomedical sciences,
the reader is urged to carefully review and evaluate the information provided herein.
Zinc finger proteins : from atomic contact to cellular function / [edited by] Shiro Iuchi, Natalie Kuldell.
p. ; cm. -- (Molecular biology intelligence unit)
Includes bibliographical references and index.
ISBN 0-306-48229-0
1. Zinc-finger proteins. I. Iuchi, Shiro. II. Kuldell, Natalie. III. Series: Molecular biology intelligence unit (Unnumbered)
[DNLM: 1. Zinc Fingers--physiology. QU 55 Z767 2005]
QP552.Z55Z56 2005
572.8'645--dc22
2004023431
CONTENTS
Preface xiii 12. Wig-1, a p53-Induced Zinc Finger Protein
that Binds Double Stranded RNA 76
Abbreviations xiv Cristina Mendez-Vidal, Fredrik Hellborg,
Margareta T. Wilhelm, Magdalena Tarkowska
1. The Discovery of Zinc Fingers and Their Practical and Klas G. Wiman
Applications in Gene Regulation:
13. Tandem CCCH Zinc Finger Proteins
A Personal Account 1
Aaron Klug
in mRNA Binding 80
Perry J. Blackshear, Ruth S. Phillips and Wi S. Lai
4. GAGA: Structural Basis for Single Cys2His2 15. LIM Domain and Its Binding
Zinc Finger-DNA Interaction 20 to Target Proteins 99
G. Marius Clore and James G. Omichinski Algirdas Velyvis and Jun Qin
5. The DNA-Binding Domain of GATA 16. RING Finger-B Box-Coiled Coil (RBCC)
Transcription Factors—A Prototypical Proteins As Ubiquitin Ligase
Type IV Cys2-Cys2 Zinc Finger 26 in the Control of Protein Degradation
Angela M. Gronenborn and Gene Regulation 106
Kazuhiro Ikeda, Satoshi Inoue
6. MutM: Single C2C2 Zinc and Masami Muramatsu
Finger-DNA Interaction 31
Ryoji Masui, Noriko Nakagawa 17. Structure and Function of the CBP/p300
and Seiki Kuramitsu TAZ Domains 114
Roberto N. De Guzman,
7. Homing Endonuclease I-TevI: Maria A. Martinez-Yamout,
An Atypical Zinc Finger H. Jane Dyson and Peter E. Wright
with a Novel Function 35
Patrick Van Roey, Marlene Belfort
18. A Zinc Ribbon Motif Is Essential
and Victoria Derbyshire for the Formation of Functional
Tetrameric Protein Kinase CK2 121
8. Zinc Finger Interactions with Metals Odile Filhol, Maria José Benitez
and Other Small Molecules 39 and Claude Cochet
Jay S. Hanas, Jason L. Larabee
and James R. Hocker BINDING OF ZINC FINGERS TO SMALL MOLECULES
9. Synthetic Zinc Finger Transcription Factors 47 19. The FYVE Finger:
Nicoletta Corbi, Valentina Libri
A Phosphoinositide Binding Domain 128
and Claudio Passananti
Harald Stenmark
BIOLOGY OF ZINC FINGER PROTEINS 32. The Role of XPA in DNA Repair 239
Takahisa Ikegami and Masahiro Shirakawa
23. Sp1 and Huntington’s Disease 168
Dimitri Krainc 33. MOF, an Acetyl Transferase Involved
in Dosage Compensation in Drosophila,
24. The Role of WT1 in Development Uses a CCHC Finger
and Disease 174 for Substrate Recognition 247
Sean Bong Lee, Hongjie Li and Ho-Shik Kim Asifa Akhtar and Peter B. Becker
25. Yin Yang 1 182 34. MDM2: RING Finger Protein and Regulator
Huifei Liu and Yang Shi of p53
Liqing Wu and Carl G. Maki
26. The Multiple Cellular Functions of TFIIIA 195
Natalie Kuldell
BIOLOGY OF ZINC ION
27. The Role of the Ikaros Gene Family
in Lymphocyte Development 200 35. The Zip Family of Zinc Transporters 261
Pablo Gómez-del Arco, Taku Naito, John Seavitt, David J. Eide
Toshimi Yoshida, Christine Williams
and Katia Georgopoulos 36. Apoptosis by Zinc Deficiency 265
Kirsteen H. Maclean
28. Basonuclin:
A Zinc Finger Protein of Epithelial Cells Index 273
and Reproductive Germ Cells 207
Howard Green and Hung Tseng
EDITORS
Shiro Iuchi
Department of Cell Biology
Harvard Medical School
Boston, Massachusetts, U.S.A.
E-mail: [email protected]
Chapters 2, 21
Natalie Kuldell
Biological Engineering Division
Massachusetts Institute of Technology
Cambridge, Massachusetts, U.S.A.
E-mail: [email protected]
Chapter 26
CONTRIBUTORS
K. Farid Ahmad Maria José Benitez
Department of Medical Biophysics Departamento de Quimica Fisica Aplicada
University of Toronto & Division of Molecular Universidad Autonoma de Madrid
and Structural Biology Cantoblanco
Ontario Cancer Institute Madrid, Spain
Toronto, Ontario, Canada E-mail: [email protected]
Chapter 20 Chapter 18
Noriko Nakagawa
RIKEN Harima Institute
Hyogo, Japan
Chapter 6
Tara L. Sander Hung Tseng
Division of Pediatric Surgery Department of Dermatology
Medical College of Wisconsin University of Pennsylvania School of Medicine
Cardiovascular Research Center Cell and Developmental Biology
Milwaukee, Wisconsin, U.S.A. and Center for Research on Reproduction
Chapter 22 and Women’s Health
Philadelphia, Pennsylvania, U.S.A.
John Seavitt Chapter 28
Cutaneous Biology Research Center
Massachusetts General Hospital Patrick Van Roey
and Harvard Medical School Wadsworth Center
Charlestown, Massachusetts, U.S.A. New York State Department of Health
Chapter 27 Albany, New York, U.S.A.
E-mail: [email protected].
Yang Shi Chapter 7
Department of Pathology
Harvard Medical School Algirdas Velyvis
Boston, Massachusetts, U.S.A. Department of Pharmacology
E-mail: [email protected] School of Medicine
Chapter 25 Case Western Reserve University
Cleveland, Ohio, U.S.A.
Masahiro Shirakawa Chapter 15
Graduate School of Integrated Science
Yokohama City University Lucas Waltzer
Yokohama, Kanagawa, Japan Centre de Biologie du Développement
E-mail: [email protected] CNRS UMR
Chapter 32 Toulouse, France
Chapter 30
Harald Stenmark
Department of Biochemistry Margareta T. Wilhelm
Institute for Cancer Research Department of Oncology and Pathology
The Norwegian Radium Hospital Karolinska Institute
Oslo, Norway Cancer Center Karolinska (CCK)
E-mail: [email protected] Stockholm, Sweden
Chapter 19 Chapter 12
In the early 1980s, a few scientists started working on a Xenopus transcription factor, TFIIIA. They soon discovered a novel
domain associated with zinc, and named this domain “zinc finger.” The number of proteins with similar zinc fingers grew quickly
and these proteins are now called C2H2, Cys2His2 or classical zinc finger proteins. To date, about 24,000 C2H2 zinc finger proteins
have been recognized. Approximately 700 human genes, or more than 2% of the genome, have been estimated to encode C2H2
finger proteins. From the beginning these proteins were thought to be numerous, but no one could have predicted such a huge
number. Perhaps thousands of scientists are now working on C2H2 zinc finger proteins from various viewpoints. This field is a
good example of how a new science begins with the insight of a few scientists and how it develops by efforts of numerous
independent scientists, in contrast to a policy-driven scientific project, such as the Human Genome Project, with goals clearly set
at its inception and with work performed by a huge collaboration throughout the world.
As more zinc finger proteins were discovered, several subfamilies, such as C2C2, CCHC, CCCH, LIM, RING, TAZ, and
FYVE emerged, increasing our understanding of zinc fingers. The knowledge was overwhelming. Moreover, scientists began
defining the term “zinc finger” differently and using various names for identical zinc fingers. These complications may explain
why no single comprehensive resource of zinc finger proteins was available before this publication.
This book adopts a broad definition of zinc finger as a peptide domain with a special tertiary structure stabilized by Zn 2+
coordination. These tertiary structured fingers confer specific binding activities to various molecules such as DNA, RNA, pro-
teins or small molecules. Some groups of scientists exploited the binding specificity to develop tools that target zinc finger
proteins to any DNA and RNA segments at will. Other groups studied additional domains that are required for executing func-
tions of zinc finger protein molecules. Some of these domains, such as BTB, KRAB and SCAN, are often present in C2H2 zinc
finger proteins, while other associated domains are less common but may also play important roles. Functions of zinc finger
proteins include gene expression, signal transduction, cell growth, differentiation and development. Given the broad role zinc
finger proteins play in the cell, it is logical that mutations in some zinc finger proteins have been found to cause diseases such as
cancer and neurological disorders. Likewise, zinc ion deficiency leads to cell death, probably because of the essential nature of
zinc finger proteins.
This book systematically describes features of various zinc finger proteins, presenting specific activities of their domains in
the earlier chapters and cellular activities of the protein molecules in the later chapters. The chapters are written by authors
outstanding in the field. These chapters are grouped according to their binding function and typically include several illustrative
examples of zinc finger proteins for each function. The arrangement reflects this book’s intention to encompass the principles
and the paradigms of zinc finger proteins at levels from atomic through cellular to organismal. This book should be a useful
resource for experts in the field, but also valuable for scientists and graduate students of biological science.
We thank Drs. J.M. Baraban, D.A. Haber, A.G. Jochemsen, F.L. Rauscher III and B. Schulz for their suggestions regarding
experts of research fields, Ms. K. Easley for her assistance, and Dr. H. Green for his support.
Shiro Iuchi
Natalie Kuldell
ABBREVIATIONS
aACRYBP1 aA-crystallin binding protein dn dominant negative
AAV adeno associated virus DN double negative
ADPRT poly(ADP-ribosyl) transferase DNase deoxyribonuclease
AF-1 activation function 1 DP double positive
AF-2 activation function 2 Dpp decapentaplegic
AFC antibody forming cell ds double-stranded
AGM aorta gonad mesonephros DTT dithiothreitol
AIF apoptosis-inducing factor E6AP viral E6-associated protein
Aio Aiolos ED effector domain
AR androgen receptor Edn endothelin
ARE androgen responsive element EED embryonic ectoderm development
AT-BP2 a1-antitrypsin promoter binding protein Efp estrogen-responsive finger protein
ATM ataxia telangiectasia mutated EGR1 early growth response gene 1
ATR ATM and Rad3-related ERCC excision repair cross-complementing rodent repair
BAC bacterial artificial chromosome deficiency
BAF brg1 associated factors ES cells embryonic stem cells
BLAST basic local alignment search tool EWS Ewings sarcoma gene
BM bone marrow EXAFS extended X-ray absorption fine structure
BMI-1 B-cell-specific Moloney murine leukemia virus EZF-2 endothelial zinc finger protein-2
integration site 1 EZH human homolog of enhancer of Zeste
BMP bone morphogenetic protein FA focal adhesion
Bn1 Basonuclin 1 FKBP25 FK506-binding protein
Bn2 Basonuclin 2 FL(2)D female-lethal(2)-d
bp base pairs FS Frasier syndrome
BRCA1 breast cancer predisposition gene 1 FSHD facioscapulo-humeral muscular dystrophy
BTB broad-complex, tramtrack, and bric-a-brac FYVE conserved in Fab1, YOTB Vac1 and EEA1
CARD caspase recruitment domain GATA family GATA sequence-binding zinc finger protein
Cbl casitas B-lineage lymphoma protein GEF GDP/GTP exchange factor
CBP cAMP response element binding protein GlcNAc N-acetylglucosamine
CD circular dichroism; chromo domain; chromosome Gly glycine
modifier domain GO 8-oxoguanine
Cdk cyclin-dependent kinase GST glutathione transferase
CFTR cystic fibrosis trans-membrane conductance GTF general transcription factor
regulator H2TH helix-two turns-helix
CHD congenital heart defect HAT histone acetyl transferase
ChIP chromatin immunoprecipitation HD Huntington’s disease
CHX cyclohexamide HDAC histone deacetylase
CK2 protein kinase CK2 HIF-1a hypoxia-inducible factor-1a
(formerly known as casein kinase II) HIV human immunodeficiency virus
Cys cysteine HIV-1 human immunodeficiency virus type I
CKII casein kinase II HIV-EP human immunodeficiency virus type 1 enhancer
CLP common lymphoid progenitor binding protein
CoA coenzyme A HMG high-mobility group
Col2a1 type II collagen gene HmL37ae Haloarcula marismortui ribosomal protein L37ae
CpG cytidine-guanidine dinucleotide pairs HMQC heteronuclear multiple quantum coherence
CREB cAMP response element binding HMTs histone methytransferases
CRIP cysteine-rich intestinal protein HP1a heterochromatin protein 1 alpha
CRP cysteine-rich protein HPC human homolog of Polycomb
CS cleavage site HPH human homolog of Polyhomeotic
CSD chromo shadow domain HPV human papillomavirus
CtBP c-terminal binding protein HSC hemopoietic stem cell
5S DNA 5S ribosomal RNA gene HSED Heteronuclear Spin-Echo Difference
DBD DNA binding domain HSP heat shock protein
DCC dosis compensation complex HSQC heteronuclear single quantum correlation
DDS Denys-Drash syndrome ICAD inhibitor of calcium-activated DNAase
DHFR dihydrofolate reductase ICR internal control region
DHS Dnase I hypersensitivity site ICR imprinted-control region
DHT 5a-dihydrotestosterone Ig immunoglobulin
DMD Duchenne muscular dystrophy IGF-1 insulin growth factor-1
IL-2 interleukin-2 PMA phorbol myristate acetate
Ileu isoleucine PML promyelocytic leukemia protein
ILF3 interleukin enhancer binding factor 3 PML-Nbs promyelocytic leukemia protein-nuclear bodies
ILK integrin-linked kinase pol III RNA polymerase III
IS insertion site POZ poxvirus and zinc finger
KAP-1 KRAB-associated protein 1 PPARγ peroxisome proliferator-activated receptor gamma
kb kilobase ppm parts per million
kD (KDa) kilodalton PRDII-BF1 positive regulatory domain II binding factor
KLF Krüppel-like factor PRE polycomb response element
KRAB Krüppel-associated box Pro proline
LBD ligand binding domain R DNA degenerate alphabet A or G
LBP-1 leader binding protein-1 5S RNP 5S ribonucloeprotein particle
Ldb LIM domain binding 5S rRNA 5S ribosomal RNA
LDL low density lipoprotein 7S RNP 7S storage particle;
LeR leucine rich region 7S ribonucleoprotein particle
Leu leucine RAG recombination activating gene
LFSE ligand field stabilization energy RAR retinoic acid receptor
Lhx LIM homeobox RAZ1 SCAN-related protein associated with MZF1B
LID LIM interaction domain Rb retinoblastoma protein
LIM-HD LIM homeodomain RBCC RING finger-B box-coiled coil
LMO LIM only protein RE response element
LSF late SV40 factor Rex-1 reduced expression 1
LTR long terminal repeat RFC replication factor C
Lys lysine RFP ret finger protein
MBP major histocompatability complex enhancer RING really interesting new gene
binding protein; maltose binding protein RMSD root-mean-square-deviation
MBT midblastula transition RNA pol II RNA polymerase II
MDM murine double minute RNAi RNA interference
Mdm2 murine double minute 2 RNP ribonucleoprotein particle
MEFs mouse embryo fibroblasts rp ribosomal protein
MGI mouse genome informatics RPA replication protein A
MIS Müllerian inhibiting substance RPB9 RNA polymerase II subunit 9
MMLV Moloney murine leukemia virus RRM RNA recognition motif
MMPs matrix metalloproteinases rRNA ribosomal RNA
MOF males-absent-on-the-first RSS recombination signal sequences of V(D)J
MOZ monocytic leukemia zinc finger protein recombination
MTA2 metastasis-associated protein 2 RYBP ring1- and YY1- binding protein
MTF metalloregulatory transcription factor SAA1 serum amyloid A1
MVE multivesicular endosome SAS something about silencing
MZF1 myeloid zinc finger gene 1 SBMA spinal and bulbar muscular atrophy
NB nuclear body SCAN SRE-ZBP, Ctfin51, AW-1 (ZNF 174),
NCBI National Center for Biotechnology Information and Number 18
NER nucleotide excision repair SCAN-ZFP SCAN domain-containing C2H2 zinc finger
NES nuclear export signal protein
NLS nuclear localization signal Scm sex combs on midleg
NMR nuclear magnetic resonance SDP1 SCAN-domain containing protein 1
NO nitric oxide SELEX systematic evolution of ligands by exponential
NOE nuclear Overhauser effect enrichment
nt nucleotide SET SuVar3-9, Enhancer of Zeste, Trithorax
NuRD nucleosome remodeling deacetylase SF1 steroidogenic factor 1
OS Opitz syndrome SKIP ski-interacting protein
oTFIIIA oocyte form of TFIIIA SLC39 solute carrier family 39
PAR4 prostate apoptosis response gene 4 SNOC S-nitrosocysteine
Pc polycomb SOX9 SRY-related HMG box
PCAF p300-CBP associated factor SP single positive
PcG polycomb group SPRY SPla and the RYanodine Receptor
Pcl polycomb-like SRC-1 steroid receptor coactivator-1
PCNA proliferating cell nuclear antigen SRE serum response element
Peg3 paternally expressed gene 3 SREBP serum response element binding protein
PGC-2 PPARg coactivator-2 SRF serum response factor
Ph polyhomeotic SRY sex-determining region Y gene
PHA phytohemagglutinin ss single-stranded
PHAX phosphorylated adaptator for RNA export SSTR2 somatostatin receptor type II
PHD plant homeobox domain protein; plant sTFIIIA somatic form of TFIIIA
homeodomain SUMO small ubiquitin-related modifier
Pho pleiohomeotic TBP TATA-box binding protein;
PI 3-kinase phosphatidylinositol 3-kinase TATA-binding protein
PI(3,5)P2 phosphatidylinositol 3,5-bisphosphate td thymidylate synthase gene
PI3P phosphatidylinositol 3-phosphate TF transcription factor
TFIIA transcription factor IIA VEGF vascular endotherial growth factor
TFIIH transcription factor IIH VHL von Hipple Lindau
TFIIIA transcription factor IIIA W DNA degenerate alphabet A or T
RRE Rev response element WAGR Wilms tumor, aniridia, genitourinary
RRE-IIB Rev response element stem loop IIB malformations, and mental retardation
TTP tristetraprolin wt wild type
TGF-β transforming growth factor beta WT1 Wilms tumor suppressor gene 1
TIF1 transcriptional intermediary factor 1 WTAP WT1-associating protein
TIF1β transcriptional intermediary factor 1 beta X. laevis Xenopus laevis
Tip TAT-interacting protein Xi X chromosome inactivation
tkv thick veins Xlo Xenopus oocyte type
TM transmembrane domain Xls Xenopus somatic type
TNF tumor necrosis factor XP xeroderma pigmentosum
TPEN N,N,N’,N’-tetrakis- XPA Xeroderma pigmentosum A
(2-pyridylmethyl)ethylenediamine YAF2 YY1-associated factor 2
TRAF tumor necrosis factor receptor-associated factor YL37a yeast ribosomal protein L37a
TRIM tripartite motif YY1 Yin Yang 1
ts temperature sensitive ZBRK1 zinc finger and BRCA1-interacting proteins with
TSA trichostatin A KRAB domain 1
Tyr tyrosine ZF zinc finger
URR upstream regulatory region ZFP zinc finger protein
UTR untranslated region ZIP Zrt-, Irt-like protein
V(D)J variable, diversity, and joining ZNF zinc finger
Val valine
CHAPTER 1
Abstract
A
n account is given of the discovery of the classical Cys2His2 tute of Washington in Baltimore on the 5S RNA genes of Xeno-
(C2H2) zinc finger, arising from biochemical studies on pus laevis, which are transcribed by RNA polymerase III (reviewed
the protein transcription factor IIIA found in Xenopus in Brown).3 They discovered that the correct initiation of tran-
ooctyes, and of subsequent structural studies on its 3D structure scription requires the binding of a 40 kD protein factor, vari-
and its interaction with DNA. Each finger is a self-contained ously called factor A or transcription faction IIIA (TFIIIA), which
domain stabilized by a zinc ion ligated to a pair of cysteines and had been purified from oocyte extracts. By deletion mapping it
a pair of histidines, and by an inner structural hydrophobic core. was found that this factor interacts with a region about 50 nucle-
This work showed not only a novel protein fold but also a novel otides long within the gene, called the internal control region.
principle of DNA recognition. Whereas other DNA binding pro- Immature oocytes store 5S RNA molecules in the form of 7S
teins generally make use of the symmetry of the double helix, ribonucleoprotein particles,4 each containing a single 40 kD pro-
zinc fingers can be linked linearly in tandem to recognize nucleic tein which was later shown5 to be identical with transcription
acid sequences of different lengths. This modular design offers a factor IIIA. TFIIIA therefore binds both 5S RNA and its cognate
large number of combinatorial possibilities for the specific rec- DNA and it was therefore suggested that it may mediate auto-
ognition of DNA (or RNA). It is therefore not surprising that regulation of 5S gene transcription.5 Whether this autoregula-
this zinc finger is found widespread in nature, in 3% of the genes tion occurred in vivo or not, the dual interaction provided an
of the human genome. interesting structural problem with could be approached because
of the presence of large quantities of the protein TFIIIA in im-
It had long been the goal of molecular biologists to design
mature Xenopus oocytes.
DNA binding proteins for specific control of gene expression. It
In the autumn of 1982, I therefore proposed to a new gradu-
has been demonstrated that the zinc finger design is ideally suited
ate student, Jonathan Miller, that he begin studies on TFIIIA.
for such purposes, discriminating between closely related se- This led to the discovery of a remarkable repeating motif within
quences both in vitro and in vivo. The first example of the poten- the protein, which we later, in laboratory jargon, called zinc fin-
tial of the method was in 1994 when a three-finger protein was gers, because they contained zinc and gripped or grasped the DNA.
constructed to block the expression of an oncogene transformed The full story of the experiments is told in our first paper6 and I
into a mouse cell line. By fusing zinc finger peptides to repression will only summarize it here. I should however emphasize that the
or activation domains, genes can be selectively switched off and repeating structure was discovered through biochemistry, not as
on. Several recent applications are described. some reviews have stated, by computer sequence analysis. When
After the initial discovery, other types of zinc-binding domains the sequence was published I looked for, and found by eye, a
which fold and interact with DNA or RNA in a different way repeating pattern, which was then confirmed and aligned as a
were found, and these have become loosely grouped under the motif of 30 amino acids by Andrew McLachlan’s computer
name of zinc finger proteins. analysis.6
Zinc Finger Proteins: From Atomic Contact to Cellular Function, edited by Shiro Iuchi and Natalie Kuldell.
©2005 Landes Bioscience/Eurekah.com and Kluwer Academic/Plenum Publishers.
2 Zinc Finger Proteins: From Atomic Contact to Cellular Function
together by disulfide bridges, and that a metal might be involved. amino acid units folds around a Zn ion to form a small indepen-
When the particle was incubated with a variety of chelating agents, dent structural domain, the “finger.” 25 of the 30 amino acid
particle dissociation could be prevented only by prior addition of residues form a loop around the central Zn ion and the five inter-
Zn2+, and not by a variety of other metals. Analysis of a partially vening amino acids provide the linkers between consecutive fin-
purified 7S preparation by atomic absorption spectroscopy also gers. The Zn ion forms the basis of the folding by being tetrahe-
revealed a significant concentration of Zn, with at least 5 mol Zn drally coordinated to the two invariant pairs of cysteines and
per mol particle. histidines. Each repeat also contained besides this unique con-
While these experiments were in progress, Hanas et al7 re- served pattern of Cys-Cys …. His-His, three other conserved
ported the presence of Zn in the 7S particle, at a ratio of two per amino acids, namely Tyr6 (or Phe6), Phe17 and Leu23, all of
particle. We thought this must be an underestimate since their which are large hydrophobic residues. The whole of the 30 amino
buffers contained 0.5mM or 1mM DTT, which has a high bind- acid repeat is rich in basic and polar residues but the largest num-
ing constant for Zn of about 1010. We therefore repeated the ber are found concentrated in the region between the second cys-
analysis with pure and undissociated particle preparations, tak- teine and the first histidine, implicating this region in particu-
ing great care to ensure no contamination. We concluded that lar in nucleic acid binding.
the native 7S particle contains between 7 and 11 zinc ions. This Formally, when indexed on a 30 amino acid repeat, the re-
result was consistent with the fact that the protein contains large peating structure could be written at
numbers of histidine and cysteine residues, the commonest ligands 1–5 6 8 13 17 23 26 30
for zinc in enzymes and other proteins. This hinted at some kind linker h X1 C X2,4 C X3 h X5 h X2,3 H X3,4 H
of internal substructure.
A natural step was therefore to see if any such substructure where h represents a conserved (large) hydrophobic residue.
could be revealed by proteolytic digestion, and Miller had already The proposal that each 30 amino acid unit formed an indepen-
begun such studies. He found two products, an intermediate 33 dently folded, Zn-stabilized domain soon gained support from
kD fragment, and a limit 23 kD. At about that time Brown’s two lines of research. First, we carried out a study using EXAFS
group8 also showed that, on treatment with proteolytic enzymes, (extended X-ray absorption fine structure) confirming that the
the 40 kD TFIIIA protein breaks down to a 30 kD product, which Zn ligands are two cysteines and two histidines. Secondly it was
is then converted to a 20 kD product. They proposed that TFIIIA found by Tso et al12 that, in the DNA sequence of the gene for
consists of three structure domains which they identified as bind- TFIIIA, the position of the intron-exon boundaries mark most
ing to different parts of the 50 base-pair internal control region of the 9 proposed finger domains.
of the 5S RNA gene. In evolutionary terms, the multi-fingered TFIIIA may have
Carrying on these proteolytic studies, we found that on pro- arisen by gene duplication of an ancestral domain comprising
longed proteolysis the TFIIIA protein breaks down further, fi- about 30 amino acids. Because one such self-contained small
nally to a limit digest of about 3 kD. In the course of this break- domain would have had the ability to bind to nucleic acids, and
down, periodic intermediates differing in size by about 3 kD could could be passed on by exon shuffling, we suggested that these
be seen. The correspondence in size between these last two values domains might occur more widely in gene control proteins than
suggested that the 30 3kD domain of TFIIIA might contain a in just this case of TFIIIA. The extent to which this prediction
periodic arrangement of small, compact domains of kD. If each has been borne out (3% of the genes of the human genome, at
such domain contained one Zn atom, then the observed high Zn the latest count) still, on occasion, astonishes me. Indeed, within
content would be accounted for. months of the publication of our paper, I received word of se-
This novel idea of small Zn-stabilized domains was strength- quences homologous to the zinc finger motif of TFIIIA. The first
ened by the timely publication by Roeder’s group9 of the sequence two were from Drosophila, the serendipity gene from Roshbash’s
of TFIIIA derived from a cDNA clone. By inspection, it could be group13 and Kruppel from Jackle’s group.14
seen that the large number of cysteines and histidines present in The key point that emerged from our first paper was that, not
the protein appeared to occur in more or less regular patterns. A only had there emerged a novel protein fold for nucleic acid bind-
rigorous computer analysis showed that, of the 344 amino acids ing, but also a novel principle of DNA recognition. The overall
of the TFIIIA sequence residues, numbers 13-276 form a con- design for specific DNA recognition was distinctly different from
tinuous run of nine tandemly repeated, similar units of about 30 that of the helix-turn-helix motif found in the first DNA binding
amino acids, each containing two invariant pairs of histidines proteins to be described. The latter binds to DNA as a symmetric
and cysteines.6 Repeating patterns in the sequence were also no- dimer to a palindromic sequence on the DNA, thus making use
ticed by R S Brown, Sander and Argos10 who concluded, how- of both the 2-fold symmetry of the DNA helix backbones and
ever, that the whole protein was divided into twelve repeats, in- also the nucleotide sequence. (Later heterodimeric variations of
dexed on a 39 amino acid unit (although their abstract states “about this and related designs were found, but they still make use of the
30”). helix symmetry).
In contrast, the zinc finger is a module that can be used singly
or linked tandemly in a linear fashion to recognize DNA (or RNA)
A Repeating Structure for TFIIIA sequences of different lengths. Each finger domain has a similar
From the three different lines of evidence described above, structural framework but can achieve chemical distinctiveness
namely (1) a 30 amino acid repeat in the sequence, which (2) through variations in a number of key amino acid residues. This
corresponds in size to the observed periodic intermediates and modular design offers a large number of combinational possibili-
the limit-digest product of 3 kD, and (3) the measured Zn con- ties for the specific recognition of DNA (or RNA). It was not
tent of 7-11 atoms, we proposed 6 that most of the TFIIIA surprising that it is found widespread throughout so many differ-
protein has a repeating structure in which each of the nine 30 ent types of organisms.
The Discovery of Zinc Fingers and their Practical Applications in Gene Regulation: A Personal Account 3
The Structure of the Zinc Finger and Its In our laboratory, my colleagues and I adopted a different ap-
Interaction with DNA proach. The reason was that the “rules” did not take into account
the fact that real DNA structures are not fixed in the canonical B
We had noted,6 that in addition to the characteristic arrange-
form but there are wide departures, depending on the DNA se-
ments of conserved cysteines and histidines which are fundamental
quence.20,21 This was further brought home to us by the struc-
in the folding of the finger by the coordinating Zn, there are
ture of the second zinc finger DNA complex to be solved, that
several other conserved amino acids, notably Tyr6, Phe17 and
with the Drosophila tramtrack protein, by Fairall et al22 in Daniela
Leu23, and that were likely to form a hydrophobic structural core Rhodes’s group our laboratory. Here the helical position used for
of the folded structure. In other words, the seven conserved amino the primary contact with the 3'-most base of one of the triplets
acids in each unit provide the framework of tertiary folding, (thymine) is not -1, but 2. The reason was that the DNA struc-
whereas some of the variable residues determine the specificity of ture was much distorted from the B form, with the thymine fol-
each domain. Jeremy Berg15 built on these original observations lowed by the adenine at a helical rotation angle of 39˚ rather than
by fitting known structural motifs from other metallo-proteins the canonical 36˚. The reason is that the T-A step in DNA is
to the consensus sequence of the TFIIIA finger motifs. His pro- unstable as I had noted long ago.20
posed model consisted of an antiparallel β-sheet, which contains
the loop formed by the two cysteines and an α-helix containing
the His-His loop. The two structural units are held together by Affinity Selection from a Library of Zinc
the Zn atom. In analogy with the way in which the bacterial Fingers by Phage Display
helix-turn-helix motif binds DNA, DNA recognition was postu- The alternative to this rational but biased design of proteins
lated to reside mostly in the helical region of the protein structure. with new specificities is the isolation of desirable variants from a
Berg’s model was soon confirmed by the NMR studies of Pe- large pool or library. A powerful method of selecting such pro-
ter Wright’s group16 on a single zinc finger in solution, and sub- teins is the cloning of peptides23 or protein domains24 as fusions
sequently by David Neuhaus in our laboratory17,18 on a two-finger to the minor coat protein (pIII) of bacteriophage fd, which leads
peptide. Our work took longer to solve but had the merit of show- to their expression on the tip of the capsid. Phage displaying the
ing that adjacent zinc fingers are structurally independent in so- peptides of interest can then be affinity purified by binding to
lution, being joined by flexible linkers. the target and then amplified for use in further rounds of selec-
The question remained of the precise pattern of amino acid tion and for DNA sequencing of the cloned gene. We applied
interactions of zinc fingers with DNA. The breakthrough came this technology to the study of zinc finger-DNA interactions,
in 1991 when Nikola Pavletich and Carl Pabo,19 both then at after my colleague, Yen Choo, demonstrated that functional zinc
Johns Hopkins, solved the crystal structure of a complex of a finger proteins could be displayed on the surface of fd phage, and
DNA oligonucleotide specifically bound to the three-finger DNA that such engineered phage could be captured on a solid support
binding domain of the mouse transcription factor Zif268, an early coated with the specific DNA.25,26 The phage display method
response gene. The primary contacts are made by the α-helix was also adopted by other groups working on zinc fingers, in-
which binds in the DNA major groove through primary cluding those of Carl Pabo and Carlos Barbas.
hydrogen-bond interactions from helical positions –1, 3 and 6 to We created phage display libraries comprising about 107 vari-
three successive bases (a triplet) on one strand of the DNA, and ants of the middle finger from the DNA-binding domain of
through a secondary interaction from helical position 2 on the Zif268. A DNA oligonucleotide of fixed sequence was used to
other strand. This is the canonical docking arrangement, but there bind and hence purify phage from this library over several rounds
are however some wide variations from this arrangement in the of selection, returning a number of different but related zinc fin-
family of zinc finger – DNA complexes now known. There are gers which bind the given DNA. By comparing similarities in the
also, of course, other minor secondary interactions. amino acid sequence of functionally equivalent fingers, we could
deduce the likely mode of interaction of these fingers with DNA.
Zinc Finger Peptides for the Regulation of Remarkably we found that many base contacts occur from three
Gene Expression primary positions on the α-helix of the zinc finger, correlating
with the implications of the crystal structure of Zif268 bound to
The mode of DNA recognition by a finger is principally a
DNA. This demonstrated ability to select zinc fingers with de-
one-to-one interaction between amino acids from the recogni-
sired specificity meant that DNA-binding proteins containing
tion helix and the DNA bases. Moreover, because the fingers func-
zinc fingers could be made to measure, and that some general
tion as independent modules, fingers with different triplet speci-
rules could be devised for a crude recognition code.
ficities are combined to give specific recognition of longer DNA
sequences. For this reason, the zinc finger motifs are ideal natural
building blocks for the de novo design of proteins for recognizing Use of Engineered Zinc Fingers to Repress
any given sequence of DNA. Indeed the first protein engineering Gene Expression in a Mouse Cell Line
experiments by Berg and others, using site-directed mutagenesis, In the course of the above work, we showed that a zinc finger
showed that it is possible to alter rationally the DNA-binding mini-domain could discriminate between closely related DNA
characteristics of individual zinc fingers when one or more of the triplets, and proposed that they could be linked together to form
α-helical positions are varied in a number of proteins. As a col- longer peptides for the specific recognition of longer DNA se-
lection of these mutants accumulated, it became possible to pro- quences. One interesting possibility for the use of such protein
pose some rules relating amino acids on the recognition α-helix peptides is to target selectively genetic differences in pathogens
to corresponding bases in the bound DNA sequence. or transformed cells. In December 1994 we reported the first
4 Zinc Finger Proteins: From Atomic Contact to Cellular Function
such application,27 in which we built a protein which recognized tially independent modules that bind three base-pair subsites to
a specific DNA sequence both in vitro and in vivo. This was a a model that considers functional synergy between adjacent in-
crucial test of our understanding of the mechanism of zinc-finger dependently folded zinc fingers. In this revised model,
DNA recognition. The proof of principle stimulated ourselves, Zif268-like zinc fingers potentially bind four base-pair overlap-
and later others, to devote our future studies to potential applica- ping subsites. We therefore redesigned our method of phage
tions in gene regulation for research purposes or for therapeutic library construction to take account of this refinement,29 which
correction. also has the merit of being more widely applicable and rapid.
In summary, we created a three finger peptide able to bind
(ii) An important step forward has been to increase the length of
site-specifically to a unique nine base-pair region of the p190
the DNA sequence targeted and hence its degree of rarity. Three
bcr-abl cDNA: this is a transforming oncogene which arises by
zinc fingers recognize nine base pairs, a sequence which would
translocation between the tips of chromosomes 9 and 22, of which
occur randomly several times in a large genome. However six
one product is the Philadelphia chromosome. The latter contains
fingers linked together would recognize a DNA sequence 18
a novel DNA sequence at the junction of two exons, one each
base pairs in length, sufficiently long to constitute a rare address
from the two genomic parent bcr and abl genes. Our engineered
in the human genome. One cannot simply go on adding fin-
peptide discriminated in vitro against like regions of the parent
gers, because the periodicity of packed fingers does not quite
bcr and c-abl genes, differing in only a single base, by factors greater
match the DNA periodicity, so that they get out of register. We
than one order of magnitude.
have learned how to engineer longer runs of zinc fingers which
Our peptide also contained a nuclear localization signal fused
can target longer DNA sequences.30,31 By fusing functional
to the zinc finger domain so that the peptide could accumulate in
groups to the engineered DNA binding domains, for example
the nucleus. Consequently, stably transformed mouse cells, made
silencing or activation domains, highly specific transcription
interleukin-3 independent by the action of the oncogene, were
factors can be generated to up- or down-regulate expression of a
found to revert to IL-3 dependence on transient transfection with
target gene.
a vector expressing the peptide. Our construct was also engineered
to contain a c-myc epitope, which enabled us to follow by immu-
nofluorescence the localization of the peptide to the nuclei of the Some Applications of Engineered Zinc
transfected cells. When IL-3 is subsequently withdrawn from cell
culture, over 90% of the transfected p190 cells become apoptotic Fingers
(that is, showing chromosome degradation) within 24 hours. Our Examples of some recent applications by ourselves and others,
experiments were repeated on cells transformed by another re- using either three finger or six finger peptides are:
lated oncogene p210 bcr-abl, which served as a control. All trans- (i) inhibition of HIV-1 expression;32
fected p210 cells maintained their IL-3 dependence, and remained (ii) the disruption of the infective cycle of infection by herpes sim-
intact on entry of the engineered peptide. plex virus;33
Measurements of the levels of p190 bcr-abl mRNA extracted
from cells treated with the peptide showed that the repression of (iii) activating the expression of VEGF-A in a monkey kidney cell
oncogenic expression by the zinc finger peptide was due to a tran- line;34
scriptional block imposed by the sequence-specific binding of the (iv) activating the expression of vascular endothelial growth factor
peptide, which presumably obstructed the path of the RNA (VEGF) in a human cell line, and in an animal model; and35
polymerase.
(v) regulation of zinc finger expression by small molecules.36
Promoter-Specific Activation by Zinc Finger
These experiments showed that a zinc finger peptide could be Other Classes of Zinc Fingers and
engineered to switch off gene expression in vivo. In the same pa-
per27 we described other experiments on a different cell system
Zinc-Binding Domains
(cultured mouse fibroblasts) to show that a gene could also be Shortly after the classical C2H2 zinc finger was discovered in
switched on in a similar way. We used the same nine base pair TFIIIA, sequence motifs that appeared to be related were found
sequence, but this time as a promoter for a CAT reporter gene in several other protein or cDNA sequences of molecules which
bound DNA. It was therefore at first thought that these might
contained in a plasmid. The peptide, which recognized the pro-
have a rather similar structure to the TFIIIA type finger domains.
moter, was fused to a VP16 activation domain and, on transient
The most important and widespread examples are those from
transfection, stimulated expression of the reporter gene by a factor
members of the superfamily of hormone-activated nuclear recep-
of 30-fold above controls.
tors which play a central role in the control of eukaryotic gene
expression, and are indeed transcription factors. 37,38 The
Improving Zinc Finger Specificity DNA-binding domains (DBDs) of such receptors all include two
(i) Our more recent work has focused on improving the specificity motifs in tandem, each about 30 amino acids long, but each motif
of recognition by zinc fingers of the DNA target. While the contains two pairs of cysteines rather than a pair of cysteines and
main source of specificity lies in the amino acids at positions – a pair of histidines as in the first class. They do indeed bind Zn2+,
1, 3 and 6 of the recognition α-helix of a zinc finger for succes- but the three-dimensional structure of two such DBDs, deter-
sive bases lying on one strand of a DNA triplet, we found that mined in solution using 2-D NMR spectroscopy, showed that
the “cross-stand” interaction described above from helical posi- the receptor DBD is structurally distinct from the TFIIIA type of
tion 2 to the neighboring base pair on the adjacent triplet can zinc finger.39,40 The two motifs in each domain each fold up into
significantly influence the specificity.28 Therefore it has been an irregular loop followed by an α-helix, but the two together
necessary to revise the simple model that zinc fingers are essen- form a single structural unit with their helices crossing at right
The Discovery of Zinc Fingers and their Practical Applications in Gene Regulation: A Personal Account 5
angles, so that the DNA recognition helix (from the first motif) The fourth class of zinc fingers discovered was constituted by
is supported by the helix from the second motif. the Cys-X2-Cys-X4-His-X4-Cys sequences found in the nucleo-
Hormone receptors bind to palindromic sites (response ele- capsid proteins of retroviruses, which form ‘stubby’ fingers.
ments, RE) on the DNA as dimers, and the DBDs alone also Thereafter several more types were found, which are described in
form dimers, the dimer interface arising from a region of the loop later chapters in this book.
of the second motif of each receptor. These different roles for the This general use of zinc became increasingly clear, even though
two motifs within on structural unit, namely helix recognition the structural information was often still limited. A diverse set of
and dimerization, were deduced by mapping on to the families of proteins which bind Zn2+ and interact with nucleic
three-dimensional structures the site-directed mutagenesis data acids were uncovered and became loosely grouped under the name
from a number of laboratories, particularly those of P. Chambon, of zinc finger proteins. The term was not inappropriate since the
R. Evans and G. Ringold. This combination of structural analy- zinc-binding domains in all these cases do grip or grasp the double
sis and biochemical and genetic experiments pointed toward a helix. Nor would it be inappropriate if the domain does not bind
mechanism of interaction with DNA and a general model was DNA or RNA but another protein, as was foreshadowed,46 and
proposed by Härd et al39 and Schwabe et al.40 as has been increasingly found.
The detailed chemistry of the interactions at the protein-DNA However in more recent years the term zinc finger has begun
interface, however, had to await a crystal structure of a complex. to be used even when the zinc-binding motif or amino acid se-
The first to be determined was that at 2.9Å resolution by Luisi quence does not form an independent, self-folding domain or
and Sigler41 of the DBD of the glucocorticoid receptor (GR) minidomain. I personally would prefer the term to be restricted
complexed with a DNA segment 18 bp long. However, the DNA to the latter cases. The cases where a Zn ion, say, is merely used to
segment used was composed of two half sites (each of 6 bp) sepa- link together two regions of a protein, belong to the field of the
rated by a nonnative spacing, with four (rather than three) inter- bioinorganic chemistry of zinc. Its widespread use in proteins for
vening base pairs. As a consequence of this, the two DBDs did such structural purposes must be attributed to the fact that zinc
not bind equivalently to DNA. is what I would call a “safe” metal, in that it has only one main
The crystal structure of a second hormone-receptor-DNA oxidation state and hence has no redox chemistry, as for instance
complex with a correct cognate binding site was solved (at 2.4Å copper and iron do. As a structural element in a protein it has the
advantage over a disulfide bridge in bringing together two parts
resolution) by Schwabe et al42 in our laboratory. This was of the
of a protein, because it cannot be reduced in the reducing atmo-
DBD of the estrogen receptor which recognizes a different DNA
sphere inside a cell.
half-site from GR but with the same native separation of 3 base
pairs between half sites (EREs3). The protein binds as a symmetrical
dimer in an equivalent manner to both half-sites. The interac- Postscript
tions seen in this complex are characteristic in number and type While, as described in the Introduction, it has long been known
of those later found in specific interactions in other families of that TFIIIA binds RNA as well as DNA, there is now increasing
protein-DNA complexes. evidence that zinc fingers are widely used to recognize RNA.
Clearly the second class of zinc finger DNA-binding domains However the molecular basis of the recognition has remained elu-
is not a simple variant of the first TFIIIA class. They differ both sive. In this laboratory, we had not forgotten that TFIIIA binds
in their structure and the way in which they interact with DNA. 5S RNA. After our failure to crystallize the native 7S RNP par-
Above all the GR and ER receptors operate as dimers which bind ticle itself (or a reconstituted version) we set out47 to find rel-
to palindromic DNA sites, whereas the binding of zinc fingers of evant subcomplexes which could show the main interactions be-
the first C2H2 class makes no use of the symmetry of the DNA tween TFIIIA and the RNA. We succeeded and have recently
structure nor of base sequence. The latter function as indepen- determined the X-ray structure of a zinc finger RNA complex
dent modules which can be strung together in a directly repeat- which reveals two modes of zinc finger binding, both different
ing (tandem) fashion with no restriction on their number. from that for DNA.48
It should however be added that some members of the nuclear
receptor family (for example thyroid, vitamin D, retinoic acid) References
also bind as dimers, but as nonsymmetrical dimers to a DNA 1. Kornberg RD. Structure of chromatin. Ann Rev Biochem 1977;
46:931-954.
binding site made of two directly repeated identical “half-sites.” 2. Klug A. From macromolecules to biological assemblies. In: The Nobel
Here clearly another interface is brought into play, but again the Foundation 1983; 93-125 Les Prix Nobel en 1982.
discrimination depends entirely on the separation between the 3. Brown DD. The role of stable complexes that repress and activate
repeats in the DNA sequence.43 eukaryotic genes. Cell 1984; 37:359-365.
The structure of a member of a third class of zinc binding 4. Picard B, Wegnez M. Isolation of a 7S particle from Xenopus laevis
domain of a distinct structural type emerged soon afterwards.44 oocytes: A 5S RNA-protein complex. Proc Nat Acad Sci USA 1979;
This was the GAL4 transcriptional activator, representative of a 76:241-245.
5. Pelham HRB, Brown DD. A specific transcription factor that can
small family found only in yeast. Here two zinc ions and six cys- bind either the 5S RNA gene or 5S RNA. Proc Nat Acad Sci USA
teines in the DBD form a binuclear cluster with each Zn2+ coor- 1980; 77:4170-4174.
dinated by four cysteines, so that two of the cysteines are shared. 6. Miller J, McLachlan AD, Klug A. Repetitive zinc-binding domains
This cluster holds together two short helices, related by a quasi in the protein transcription factor III A from Xenopus oocytes.
dyad, one of which inserts into the major groove of the DNA as EMBO J 1985; 4:1609-1614.
in the hormone receptors; we thus see the recognition helix sup- 7. Hanas JS, Hazuda DJ, Bogenhagen DF et al. Xenopus transcription
factor A requires zinc for binding to the 5S RNA gene. J Biol Chem
ported by a second helix. Again, the molecule binds as a dimer to 1983; 258(14):120-125.
a palindromic DNA binding site, with two short half-sites sepa- 8. Smith DR, Jackson IJ, Brown DD. Domains of the positive
rated by approximately 1.5 turns of the DNA helix, so that the transcription factor specific for the Xenopus 5S RNA gene. Cell
DBDs bind on opposite faces of the DNA. 1984; 37:645-652.
6 Zinc Finger Proteins: From Atomic Contact to Cellular Function
9. Ginsberg AM, King BO, Roeder RG. Xenopus 5S gene transcription 29. Isalan M, Klug A, Choo Y. A rapid, generally applicable method to
factor, TFIIIA: characterisation of a cDNA clone and measurement engineer zinc fingers illustrated by targeting the HIV-1 promoter.
of RNA levels throughout development. Cell 1984; 39:479-489. Nat Biotechnol 2001; 19:656-660.
10. Brown RS, Sander C, Argos P. The primary structure of transcription 30. Moore M, Choo Y, Klug A. Design of polyzinc finger peptides with
factor TFIIIA has 12 consecutive repeats. FEBS Lett 1985; structured linkers. PNAS 2001; 98:1432-1436.
186:271-274. 31. Moore M, Klug A, Choo Y. Improved DNA binding specificity from
11. Diakun GP, Fairall L, Klug A. EXAFS study of the zinc-binding polyzinc finger peptides by using strings of two-finger units. PNAS
sites in the protein transcription factor IIIA. Nature 1986; 2001; 98:1437-1441.
324:698-699. 32. Reynolds L, Ullman C, Moore M et al. Repression of the HIV-1 5'
12. Tso JY, van den Berg DJ, Korn LT. Structure of the gene for Xenopus LTR promoter and inhibition of HIV-1 replication by using
transcription factor TFIIIA. Nucl Acids Res 1986; 14:2187-2200. engineered zinc-finger transcription factors. Proc Natl Acad Sci USA
13. Vincent A, Colot HV, Rosbash M. Sequence and structure of the 2003; 100(4):1615-1620.
serendipity locus of Drosophila melanogaster: A densely transcribed 33. Papworth M, Moore M, Isalan M et al. Inhibition of herpes simplex
region including a blastoderm-specific gene. J Mol Biol 1985; virus 1 gene expression by designer zinc-finger transcription factors.
185:146-166. Proc Natl Acad Sci USA 2003; 100(4):1621-1626.
14. Rosenberg UB, Schröder C, Preiss A et al. Structural homology of 34. Liu PQ, Rebar EJ, Zhang L et al. Regulation of an endogenous
the product of the Drosophila Krüppel gene with Xenopus locus using a panel of designed zinc finger proteins targeted to
transcription factor IIIA. Nature 1986; 319:336-339. accessible chromatin regions. Activation of vascular endothelial growth
15. Berg JM. Proposed structure for the zinc binding domains from factor A. J Biol Chem 2001; 276(14):11323-11334.
transcription factor IIIA and related proteins. Proc Nat Acad Sci 35. Rebar E, Hung Y, Hickey R et al. Induction of angiogenesis in a
USA 1988; 85:99-102. mouse model using engineered transcription factors. Nature Med
16. Lee MS, Gippert GP, Soman KV et al. Three-dimensional solution 2002; 8:1427-1432.
structure of a single zinc finger DNA-binding domain. Science 1989; 36. Beerli RR, Schopfer U, Dreier B et al. Chemically regulated zinc
245:635-637. finger transcription factors. J Biol Chem 2000;
17. Nakaseko Y, Neuhaus D, Klug A et al. Adjacent zinc finger motifs 275(42):32617-32627.
in multiple zinc finger peptides from SW15 form structurally 37. Evans RM. The steroid and thyroid hormone receptor superfamily.
independent flexibly linked domains. J Mol Biol 1992; 228:619-636. Science 1988; 240:889-898.
18. Neuhaus D, Nakaseko Y, Schwabe JW et al. Solution structures of 38. Green S, Chambon P. Nuclear receptors enhance our understanding
two zinc-finger domains from SWI5 obtained using two-dimensional of transcription regulation. Trends Genet 1988; 4:309-315.
1H nuclear magnetic resonance spectroscopy. A zinc-finger structure 39. Hard T, Kellenbach E, Boelens R et al. Solution structure of the
with a third strand of beta-sheet. J Mol Biol 1992; 228(2):637-651. glucocorticoid receptor DNA-binding domain. Science 1990;
19. Pavletich NP, Pabo CO. Zinc finger-DNA recognition: Crystal 249(4965):157-160.
structure of a Zif268-DNA complex at 2.1Å. Science 1991; 40. Schwabe JWR, Neuhaus D, Rhodes D. Solution structure of the
252:809-817. DNA-binding domain of the oestrogen receptor. Nature 1990;
20. Klug A, Jack A, Viswamitra MA et al. A hypothesis on a specific 348:458-461.
sequence-dependent conformation of DNA and its relation to the 41. Luisi BF, Xu WX, Otwinowski Z et al. Crystallographic analysis of
binding of the lac-repressor protein. J Mol Biol 1979; the interaction of the glucocorticoid receptor with DNA. Nature
131(4):669-680. 1991; 352(6335):497-505.
21. Rhodes D, Klug A. Sequence dependent helical periodicity of DNA. 42. Schwabe JW, Chapman L, Finch JT et al. The crystal structure of
Nature 1981; 292:378-380. the estrogen receptor DNA-binding domain bound to DNA: How
22. Fairall L, Schwabe JW, Chapman L et al. The crystal structure of a receptors discriminate between their response elements. Cell 1993;
two zinc-finger peptide reveals an extension to the rules for 75(3):567-578.
zinc-finger/DNA recognition. Nature 1993; 366(6454):483-487. 43. Umesono K, Murakami KK, Thompson CC et al. Direct repeats as
23. Smith GP. Filamentous fusion phage: Novel expression vectors that selective response elements for the thyroid hormone, retinoic acid,
display cloned antigens on the virion surface. Science 1985; and vitamin D3 receptors. Cell 1991; 65(7):1255-1266.
228:1315-1317. 44. Marmorstein R, Carey M, Ptashne M et al. DNA recognition by
24. McCafferty J, Griffiths AD, Winter G et al. Phage antibodies: GAL4: Structure of a protein-DNA complex. Nature 1992;
Filamentous phage displaying antibody variable domains. Nature 356(6368):408-414.
1990; 348(6301):552-554. 45. Summers MF, Henderson LE, Chance MR et al. Nucleocapsid zinc
25. Choo Y, Klug A. Towards a code for the interactions of zinc fingers fingers detected in retroviruses: EXAFS studies of intact viruses and
with DNA: Selection of randomised fingers displayed on phage. Proc the solution-state structure of the nucleocapsid protein from HIV-1.
Natl Acad Sci USA 1994; 91:11163-11167. Protein Sci 1992; 1(5):563-574.
26. Choo Y, Klug A. Selection of DNA binding sites for zinc fingers 46. Klug A, Rhodes D. ‘Zinc fingers’: A novel protein motif for nucleic
using rationally randomised DNA reveals coded interactions. Proc acid recognition. Trends Biochem 1987; 12:464-469.
Natl Acad Sci USA 1994; 91:11168-11172. 47. Searles MA, Lu D, Klug A. The role of the central zinc fingers
27. Choo Y, Sánchez-García I, Klug A. In vivo repression by a transcription factor IIIA in binding to 5S RNA. J Mol Biol 2000;
site-specific DNA-binding protein designed against an oncogenic 301:47-60.
sequence. Nature 1994; 372:642-645. 48. Lu D, Searles MA, Klug A. Crystal structure of a zinc finger RNA
28. Isalan M, Choo Y, Klug A. Synergy between adjacent zinc fingers in complex reveals two modes of molecular recognition. Nature 2003;
sequence-specific DNA recognition. Proc Natl Acad Sci USA 1997; 426:96-100.
94:5617-5621.
CHAPTER 2
A
great number of C2H2 zinc finger proteins selectively bind Requirement of zinc ion for transcription factors to bind their
to specific DNA sequences and play a critical role in con- cognate DNA was found in Xenopus TFIIIA (Transcription Fac-
trolling transcription of genes. The specific binding is tor IIIA) first.8 The requirement was due to the ion coordination
achieved by zinc finger domains with ββα structure that is formed with a small peptide domain, named zinc finger, that
by tetrahedral binding of Zn2+ ion to the canonical cysteine and contains two canonical cysteine and histidine residues.1,4-6 Now,
histidine residues. Two to three tandem zinc fingers are thousands of proteins are known to have C2H2 zinc fingers and
necessary and sufficient for the specific binding without the majority of the fingers are thought to bind to DNA. Zinc
participation of any other domains. Zinc fingers bind in the major finger proteins can take three states: unfolded, folded and
groove of the DNA, wrapping around the strands, with specificity DNA-bound forms (Fig. 1 and for details of folded zinc finger
conferred by side chains of several amino acid on the α helices. domain see ref. 1 and 3 as well as Fig. 1 of Chapter 8). Unfolded
Some zinc finger proteins undergo homodimerization by zinc fingers do not bind to target DNA, but folded fingers bind
hydrophobic interactions or by finger-finger binding and to the cognate DNA duplexes. The protein molecules of the DNA
reinforce the specific binding to DNA. Conserved linkers complex are usually associated with other transcription factors
between tandem fingers are necessary for stabilizing the DNA that bind to different domains on the zinc finger protein, or to
complex. Regulatory mechanisms of zinc finger binding to DNA different parts of the zinc fingers. The C2H2 zinc finger motif
are emerging. Some cellular factors are found to acetylate and contains all the information necessary for its folding but folds
phosphorylate zinc fingers and the linkers of a few proteins. These properly only when Zn2+ binds to the canonical residues, two
modifications alter the binding activity of the zinc finger cysteines and two histidines.9,10 The change in Gibbs free energy
proteins and hence control expression of their target genes. Other (∆G) of the folding is enthalpy-driven. The amount, about -8.8
factors can methylate promoter regions of genes. This modification kcal per mole (Table 2),11 indicates that the folded fingers are
alters affinity of zinc finger proteins for the DNA segments and very stable. C 2 H 2 zinc fingers contain three conserved
hence controls expression of their target genes. hydrophobic amino acids at position -12, -3 and +4 in addition
to the two canonical cysteines and histidines (see bolded amino
Introduction acids in the sequence below where the first amino acid residue of
The C2H2 zinc finger consists of twenty to thirty amino acid the α helix is designated as position 1). It has been shown that
residues that have a special secondary structure stabilized by zinc these seven amino acid residues are necessary and sufficient to
tetrahedral binding to two cysteine and two histidine residues.1-6 fold peptides properly by using a designed-synthetic peptide,
Proteins with these zinc fingers are called C2H2 zinc finger K(-13)-YACAACAAAFAAKAALAAHAAAHA-K13. 9 This
proteins. The C2H2 zinc finger protein family is the largest group peptide binds Co2+, a substitute for Zn2+, to the cysteines with a
of all zinc finger protein families (Table 1) and the second largest higher affinity than to the histidines, and the proper folding
group of all protein classes after the envelope glycoprotein GP120 occurs only when the ratio of the ion to the peptide is one or
family. The proteins are present in prokaryotes as well as eukaryotes higher. Similar binding of Zn2+ to the zinc finger motif has also
and are abundant in mammals. More than 700 human genes, or been observed in the synthetic peptide of the zif268 third finger
greater than 2% of the total human genes, encode C2H2 zinc [F(-12)-ACDICGRKFARSDERKRHTKIHLRQ-K15].10,12 The
finger proteins.7 Not surprisingly, C2H2 zinc finger proteins authors of this work have proposed that the zinc finger folding
participate in a variety of cellular activities including develop- begins with binding of Zn2+ to the canonical cysteines and then
ment, differentiation, and tumor suppression. Among C2H2 zinc establishes the tetrahedral structure involving the histidines. They
finger proteins, many bind to DNA duplexes in a finger-mediated have also suggested that the α helix emerges at the
specific manner and participate in controlling expression of the S(1)-DERKRHTKI-H11 sequence as the metal ion binding
target genes. C2H2 zinc fingers, which are often described as proceeds.
X2CX2-4CX12HX2-8H to show the intervals between the zinc
binding residues, contain two β strands and one α helix. All the Tandem C2H2 Zinc Fingers
primary, secondary and tertiary structures are important for C2H2 zinc finger proteins often contain the fingers as tandem
binding to DNA duplexes. In this chapter the general features of repeats connected by short oligopeptides, called linkers. Based
C2H2 zinc fingers-DNA binding are described. on the number and repeat pattern of the fingers, C2H2 zinc
Zinc Finger Proteins: From Atomic Contact to Cellular Function, edited by Shiro Iuchi and Natalie Kuldell.
©2005 Landes Bioscience/Eurekah.com and Kluwer Academic/Plenum Publishers.
8 Zinc Finger Proteins: From Atomic Contact to Cellular Function
C2H2 23,989 23
CCHC (C2HC) 5,215 17
RING 2,010 41
LIM 1,246 58
C2C2 (2x)* 1,243 65
CCCH (C2CH) 1,019 26
TAZ 51 81
Data are taken from Pfam 9.0 at Washington Uninversity in St.
Louis. *The number is for proteins containing two C2C2 fingers.
finger proteins can be divided into four classes (Fig. 2), (A) single
C2H2, (B) triple C2H2, (C) multiple-adjacent C2H2, and (D)
separated-paired C2H2 zinc finger proteins.13 This classification
is useful to predict how the zinc fingers of the proteins exert their
binding activity. The single zinc finger differs from the other zinc
fingers in that it requires an additional, non-zinc finger domain
to establish the binding to the target DNA.14 The other classes,
that is the triple, multiple-adjacent and separated-paired C2H2
fingers, bind to the specific DNA sequence without the aid of
other domains. Both the triple C2H2 and the multiple-adjacent
C2H2 zinc fingers bind to the cognate DNA at the three consecu-
tive fingers.13 Another experiment has shown that a finger pep-
tide with four-tandem identical repeats binds to the target DNA
sequence at the three consecutive fingers only.15 Furthermore,
the separated-paired C2H2 finger can also specifically bind to the Figure 1. Schematic representation of three states of C2H2 zinc finger
cognate DNA, often at the one pair finger. Taking all these proteins. A) Primary structure of proteins. Polypeptide has three tan-
results into account, it may be concluded that two to three dem fingers (ZF1, ZF2 and ZF3), each of which has α helix region
successive C2H2 zinc fingers are the most suitable unit to (thick line) and is connected by linker (L). Other domains of the polypep-
specifically bind to the cognate DNA. tide are presented by A, B, -----, Z. B) Folded zinc finger proteins. Zn2+
Multiple-adjacent C2H2 zinc fingers bind to the DNA, based tetrahedrally binds to the canonical cysteine and histidine residues and
on the rule that two to three successive fingers are responsible for stabilizes the zinc finger’s tertiary structure. (For details of folded C2H2
the specific DNA binding, but the fingers have additional DNA zinc fingers, see ref. 1 and 3 as well as Fig. 1 of Chapter 8). C) Zinc finger
contacts. For example, TFIIIA with nine zinc fingers establishes protein-DNA complex. The α helices of three zinc fingers bind to the
the DNA binding at fingers 1-3, but briefly touches the DNA at DNA duplex in the major groove. Various domains of zinc finger pro-
finger 5 and weakly binds to the DNA at fingers 7-9.16-20 teins, including a part of zinc fingers, interact with other transcription
Another example is Zac. This has seven zinc fingers and binds to factors.
a GC rich DNA duplex. Biochemical and genetic analysis have
shown that Zac binds to the DNA at the two consecutive fingers linker sequences flanking the C2H2 zinc finger DNA may allow a
(finger 2-3) and also at the three consecutive fingers zinc finger sequence to duplicate together with the linker itself
(finger 5-7), without involving finger 4.21 Accordingly, Zac (Fig. 3). The duplication would in turn stimulate the gene
contacts the DNA at more than three fingers, keeping the two to duplication further by using the linkers again or the zinc finger
three finger-DNA binding rule. It is curious that extra fingers are itself. The multiple zinc fingers can gain a better affinity for a
present in multiple-adjacent zinc finger and separated-paired zinc DNA sequence and give the zinc finger protein molecules a
finger proteins, but it is becoming increasingly clear that the ex- selective advantage over the original finger proteins to function
tra fingers are engaged in other interactions not only with a as transcription factors. This speculation is consistent with the
secondary locus of DNA but also with distinct molecules such as fact that many of the separated-paired C2H2 and triple C2H2
RNA and proteins.13,22-24 zinc fingers conserve amino acid residues between tandem zinc
It is interesting to speculate how these tandem zinc fingers fingers.13 After the duplication, point mutations would further
have evolved from a single finger. The Escherichia coli gene, arcA, improve its affinity for the target. Alternative splicing within the
encodes a helix-turn-helix DNA binding repressor protein for the gene26,27 or recombination with another zinc finger gene28 would
genes directing aerobic respiration enzymes.25 This gene has an allow the finger peptide to acquire tandem hybrid zinc fingers
identical eleven-nucleotide sequence flanking a short region, and with a very different specificity. It should not be difficult to show
this genetic organization prompts the region to duplicate sponta- that gene duplication indeed causes the evolution in living cells
neously (Iuchi and Lin, unpublished). Similarly, the identical using well-established bacterial and yeast genetic systems.
C2H2 Zinc Fingers As DNA Binding Domains 9
Overall Features of the Triple-C2H2 Zinc finger-DNA complex formation. Phosphate contacts, linking to
the DNA backbone, may not be significant in determining the
Fingers Binding specificity but appear to be important for strengthening the
The change in Gibbs free energy of zinc finger-DNA binding is
binding.37 Through further analysis of various zinc finger–DNA
similar to or higher than that of the zinc finger folding (Table complexes, it was found that the overall DNA binding mode of
2)11,29-33 and the binding requires no enzyme action. ∆G is
Zif268 is shared with other tandem C2H2 zinc fingers. These
constant over a biologically-meaningful temperature range, 5 to
fingers include TFIIIA, SP1, GL1, WT1 and Tramtrack17,18,38-41
45 °C.31 The reactions are mostly enthalpy- and entropy-driven as well as the single finger GAGA.14 In addition to these GC-rich
reactions, but some are only enthalpy- or only entropy-driven.
DNA binding transcription factors, an AT-rich DNA recognizing
The ∆G values suggest that the C2H2 zinc fingers bind to DNA
transcription factor, CF2II, is also thought to take the same
as strongly as some antibodies, whose Ka for their antigens is 105 binding mode.27 However, one AT-rich DNA binding transcrip-
to 1012 M-1.34 These zinc finger-DNA complexes are of such high
tion factor, Nmp4, is proposed to associate with minor groove of
affinity that they routinely display an electrophoretic mobility
the DNA.42
shift on native polyacrylamide gels.
When an amino acid and guanine are mixed in water, the
amino acid starts to associate with the nucleotide base and the
reaction soon reaches equilibrium. Tendency of the reaction is
described by the affinity constant, Ka = 1/Kd = [aa-guanine] / [aa]
[guanine]. The Ka of amino acids for guanine is in the order
arginine > lysine > glutamine > glutamate > glycine.35 Amino
acids also have an inherent Ka for the three other bases. Accord-
ingly, it is predictable that the amino acid-base associations are
key for zinc finger peptides to specifically bind to the DNA du-
plexes. Indeed, the arginine-guanine contact, whose association is
the greatest of all the combinations, is quite often present in zinc
finger-DNA complexes (Fig. 4B and C). However, each amino
acid residue of the zinc finger peptides does not have free
mobility to access to the favorite bases due to the rigid zinc finger
structure. How the zinc finger peptides recognize and bind to the
cognate DNA duplex is a big question for the zinc finger-DNA
interactions and the answer has come mostly from structure-ori-
ented and genetic oriented analyses of zinc finger-DNA
complexes. In particular, the structure of Zif268-DNA and
TFIIIA-DNA complexes contributed enormously to the
understanding of how the triple fingers align with the DNA du-
plex and how each finger interacts with the nucleotides of the
DNA.17,18,36 Three α helices of the Zif268 triple zinc fingers bind
in the major groove of the target DNA duplex antiparallel to the
primary strand (defined as the strand to which zinc finger con-
tacts most, Fig. 4A and for details of the binding see ref. 2), mak- Figure 2. Schematic representation of four classes of C2H2 zinc finger
ing hydrogen bonds and forming hydrophobic interactions with proteins. Only one example is shown for each class. Some
nucleotide bases and wrapping around the DNA for almost one multiple-adjacent C2H2 zinc finger proteins contain more than thirty
turn. In addition to hydrogen bonds and hydrophobic zinc fingers. The number and the pattern of C2H2 zinc fingers indicate
interactions, phosphate contacts also participate in the zinc how the zinc fingers are involved in the DNA binding (see text).
10 Zinc Finger Proteins: From Atomic Contact to Cellular Function
Figure 3. A model of intramolecular duplication of the zinc finger region. ZF1 flanked by two identical linker sequences for (L) would form a loop
overlapping at the identical sites, or form a hairpin structure complementing within the single strand at the identical sites. When DNA polymerase
reads the loop or the hairpin structure twice, then the zinc finger region is duplicated together with the linker sequence.
DNA of the Zif268-DNA complex takes a slightly unwound factors involved in specific zinc finger–DNA binding, including
B form so that the major groove is still wide and can be deep.2 amino acid residues within the subsite,47 linkers, and adjacent
This form contains 11.3 bp per turn, which is slightly more bases fingers.2,37
than the B-form itself (10.5 bp per turn). Absence of a strict rule in the side chain-base contacts makes
it impossible to predict the side chain-base contact with 100%
Side Chain-Base Contacts in C2H2 Zinc
Finger-DNA Complexes
Amino acid side chains on the surface of α helices in Zif268
are exposed to the cognate DNA duplex in the major groove, and
the side chains at position 6, 3, 2 and -1 contact selectively with
four successive bases (subsite) (Fig. 4B). Residues at position 6, 3
and -1 bind to three successive bases of the primary strand, the
strand contacted most by the side chains, and the residue at posi-
tion 2 binds to the fourth base present in the complementary
strand. In this way, Zif268 finger 3, 2 and 1 recognize the pri-
mary strand’s subsite, 5’-GCGT-3’, 5’-TGGG-3’ and
5’-GCGT-3’, respectively. Of the four base pairs in the subsites,
the base pair(s) at the end is shared by adjacent fingers. Conse-
quently, the Zif268 triple zinc fingers bind to the ten nucleotide
base pair, 5’-GCGTGGGCGT-3’. Based on statistics, one can
predict with a high accuracy which side chain-base contacts can
happen at the key positions of the α helices in the Zif268 finger
context.43 Stereochemistry between amino acids and bases can
also predict the contacts with a similar accuracy or somewhat less
accuracy.44
The rule that side chains positioned at 6, 3, 2 and -1 contact
bases is well preserved throughout the family of C2H2 zinc fin-
gers, but additional contacts can occur in other fingers. Such an
example has been observed in the TFIIIA-DNA complex (Fig.
4C).17,18 There is an additional contact of the side chain at posi-
tion 10. Furthermore, some side chains at the regular positions 6,
3, 2 and -1 reach bases out of the subsites. When Zif268 was
engineered to bind to AT-rich DNA duplexes, many irregular
contacts occurred.2,45 In fact, many of these irregularities occurred Figure 4. A) Antiparallel binding of zinc fingers to DNA. The orienta-
in the complementary strand but not in the primary strand.46 tion of the zinc fingers and the primary strand are antiparallel. B) and
These facts, together with results obtained by mutational analy- C) show amino acid side chain-nucleotide base contacts. Boxes show
ses, have led to the conclusion that there is no simple code for subsites. The side chain-DNA contacts are drawn, based on the results
side chain-base contacts. 2,47 Superimposition of several obtained with the crystallized Zif268-DNA complex B)2, 36 and with the
finger-DNA complex images showed that the irregular contacts TFIIIA-DNA complex in solution C).17 Similar results of the side
coincide with the slight difference in docking angle of the α helix chain-DNA contact for the TFIIIA-DNA complex has been obtained
to DNA. The difference may reveal the influences of all of the with the crystallized complex.18
C2H2 Zinc Fingers As DNA Binding Domains 11
accuracy, but it rather indicates that C2H2 zinc fingers are able to
bind to almost any DNA duplexes. The versatility of the zinc
finger binding relies on hydrogen bonds that can (i) make long
distance contact to bases with and without participation of a water
molecule (>2.75 angstroms),35,48 (ii) make contacts to more than
one base, and (iii) make contacts to bases with flexible angles.35
Zinc fingers’ binding also relies on nonspecific hydrophobic in-
teractions. Although it is difficult to predict exactly which zinc
finger peptide sequence specifically binds to a DNA sequence,
one can obtain desired zinc fingers by the phage display methods
and manipulate expression of the target genes with the obtained
finger proteins.2,49-53
Linkers Figure 5. Modification of zinc finger proteins and the target DNA.
About half of zinc finger proteins have a well-conserved linker,
TGEKP, between adjacent fingers.37 The importance of the linker
specific finger-finger interaction and hence for the Ikaros
in zinc finger-DNA binding has been revealed by analyzing the
homodimerization.62 These amino acids correspond to those of
effect of mutation on DNA affinity after making substitutions
C2H2 zinc fingers with DNA binding activity, and differ from
for the conserved residues54,55 and by analyzing structures of
those needed for the hydrophobic dimerization.
finger-DNA complexes.2,3,56,57 The linker is flexible in solution
without DNA. Upon forming the finger-DNA complex,
however, the conserved lysine residue of the TGEKP sequence Regulation of Zinc Finger Binding to DNA
(Fig. 1) interacts with the phosphate backbone. Moreover, In vitro zinc finger-DNA binding proceeds as long as both
linkers contact the C terminus of the preceding α helix by substances are present in a suitable buffer, but the in vivo
involving threonine and glycine residues so that the zinc binding does not happen automatically and proceeds only under
finger-DNA complex becomes more stable (C capping).17,57,58 certain conditions. The regulation is mediated by modification
Alternative splicing of the WT 1 message disrupts the conserved of either the zinc finger protein itself or the target DNA (Fig. 5).
linker between finger 3 and 4 by inserting the sequence KTS. The former can be executed by acetylation and also by phospho-
The change from TGEKP to sequence TGKTSEP is accompa- rylation, whereas the latter is mediated by methylation. All these
nied by a severe decrease in DNA binding. An NMR study of the modifications increase or decrease the zinc finger protein-DNA
finger peptide-DNA complex has shown that the insertion in- complex formation. The direction in which the regulation pro-
creases the flexibility between finger 3 and 4 and abrogates bind- ceeds, repression or activation, depends on the fingers and the
ing of finger 4 to its cognate site.41 These findings have demon- DNA, as well as the regulatory signal involved.
strated that the conserved linker is not only necessary to promote
fingers to fit completely into the DNA major groove, but also Acetylation of Zinc Fingers
necessary to strengthen the DNA binding. A separated-paired YY1 (Yin Yang 1), with four tandem C2H2 zinc fingers, binds
zinc finger class, ZAS family, also contains the TGEKP linker,59 to various genes and represses or activates their target gene
but fingers belonging to a subtype of the same class, expression.65,66 PCAF (p300/CBP associated factor) acetylates
basonuclin-type fingers, have a distinct linker, LR(K)MHK.13,60 YY1 at the zinc finger domain and inhibits DNA binding. PCAF
Tramtrack also has a distinct linker sequence, KRNVKVYP.39 also acetylates members of the KLF (Krüppel-like factor) family
belonging to the triple C2H2 zinc finger protein class.67,68 It
Dimerization of Zinc Finger Protein acetylates KLF13 at two lysines of the TGEKK linker between
DNA binding proteins often bind to the target DNA duplex finger 2 and 3 and results in stimulating the fingers to bind to
as dimers in order to increase their binding affinity and modulate DNA. Another coactivation factor CBP/p300 acetylates a total
their regulatory activities.61 This is true for C2H2 zinc finger pro- of six lysines including the linker’s two lysines, but it prevents the
teins. Many multiple-adjacent C2H2 zinc finger proteins, such as fingers from binding to DNA. Of the six residues, the lysine of
Ikaros, Roaz, GL1, SW15, TRPS-1 and Zac, form homodimers finger 1 directly contacts bases. However, acetylation of this lysine
on the target DNA duplex using their zinc fingers.23,24,62-64 GL1 is not enough to interrupt the DNA binding and acetylation of
forms a homodimer through the hydrophobic surface of zinc fin- all six lysines is required. Moreover in vitro experiments showed
ger 1 that is not involved in the DNA binding. Similarly, SWI5 that KLF2 zinc finger cotransfected with PCAF increases the tar-
dimerizes through finger 1 at the hydrophobic surface of both get γ globulin gene expression in vivo while the zinc finger
the β strand and the α helix’s C-terminal half. This hydrophobic cotransfected with p300 decreases gene expression.68 Therefore,
dimerization was applied to make the homodimer of an engi- it can be concluded that acetylation of KLF2 regulates the target
neered two-finger peptide on the DNA.63 However, the hydro- gene expression both positively and negatively. Because lysine resi-
phobic binding is not the only dimerization mechanism for zinc dues are abundant in a variety of zinc fingers, acetylation of C2H2
fingers. Ikaros contains six zinc fingers. The N-terminal four fin- zinc fingers is likely to be a common mechanism to regulate gene
gers participate in the specific binding to the DNA, and the expression by modulating the zinc fingers’ DNA binding activity.
C-terminal two fingers, separated from the four and making a Another domain of YY1, the central glycine-lysine-rich
pair of fingers, are responsible for the homodimerization.23 Re- domain, is acetylated by p300 and PCAF. This acetylation has
cently, a mutational analysis of zinc fingers 5 and 6 demonstrated nothing to do with the DNA binding but is required for fully
that amino acid residues on the α helix are responsible for the repressing the target gene transcription.69
12 Zinc Finger Proteins: From Atomic Contact to Cellular Function
Phosphorylation of Zinc Fingers 4. Miller J, McLachlan AD, Klug A. Repetitive zinc-binding domains
in the protein transcription factor IIIA from Xenopus oocytes.
During mitosis, Ikaros is phosphorylated on threoine/serine EMBO J 1985; 4(6):1609-1614.
residues of the three linkers and interrupted for its DNA bind- 5. Brown RS, Sander C, Argos P. The primary structure of transcription
ing.70 Phosphoryaltion of linkers also occurs to Sp1. This sug- factor TFIIIA has 12 consecutive repeats. FEBS Lett 1985;
gests that phosphorylation is a global inhibitory mechanism to 186(2):271-274.
keep C2H2 zinc fingers out of DNA during mitosis. Interestingly, 6. Diakun GP, Fairall L, Klug A. EXAFS study of the zinc-binding
phosphorylation modifies the linkers as the favored sites. A few sites in the protein transcription factor IIIA. Nature 1986;
324(6098):698-699.
interpretations can be given. First, modification of the conserved
7. Lander ES, Linton LM, Birren B et al. Initial sequencing and analysis
linker is the most effective way to modulate zinc finger binding of the human genome. Nature 2001; 409(6822):860-921.
activity because the linker is essential for the high affinity zinc 8. Hanas JS, Hazuda DJ, Bogenhagen DF et al. Xenopus transcription
finger-DNA binding. Second, the modification sites, theonines/ factor A requires zinc for binding to the 5 S RNA gene. J Biol
serines, are always available in the conserved linkers. Finally, linkers Chem 1983; 258(23):14120-14125.
may be accessible even in the finger-DNA complexes so that zinc 9. Michael SF, Kilfoil VJ, Schmidt MH et al. Metal binding and folding
finger peptides can be separated from the DNA as soon as the properties of a minimalist Cys2His2 zinc finger peptide. Proc Natl
Acad Sci USA 1992; 89(11):4796-4800.
regulatory signal is delivered to the fingers. The same arguments 10. Cox EH, McLendon GL. Zinc-dependent protein folding. Curr Opin
can be applied to regulation by acetylation since it also modifies Chem Biol 2000; 4(2):162-165.
linkers. 11. Lachenmann MJ, Ladbury JE, Phillips NB et al. The hidden
A serine residue of the C terminus immediately after a zinc thermodynamics of a zinc finger. J Mol Biol 2002; 316(4):969-989.
finger can be phosphorylated in the case of Cre1 (catabolite 12. Miura T, Satoh T, Takeuchi H. Role of metal-ligand coordination
repression) of Hypocrea jecorina, perhaps by caseinkinase-II-like in the folding pathway of zinc finger peptides. Biochim Biophys
enzyme.71 Surprisingly, this phosphorylation increases the finger’s Acta 1998; 1384(1):171-179.
13. Iuchi S. Three classes of C2H2 zinc finger proteins. Cell Mol Life
DNA biding activity, suggesting that the unphosphorylated serine Sci 2001; 58(4):625-635.
residue may mask the finger’s DNA binding activity directly by 14. Omichinski JG, Pedone PV, Felsenfeld G et al. The solution structure
interacting with the finger or indirectly by interacting with other of a specific GAGA factor-DNA complex reveals a modular binding
intramolecular domains. mode [see comments]. Nat Struct Biol 1997; 4(2):122-132.
15. Nagaoka M, Kaji T, Imanishi M et al. Multiconnection of identical
zinc finger: implication for DNA binding affinity and unit
Methylation of DNA modulation of the three zinc finger domain. Biochemistry 2001;
Engineered zinc fingers have been shown to distinguish 40(9):2932-2941.
methylated DNA duplex from unmethylated.72 It is becoming 16. Liao XB, Clemens KR, Tennant L et al. Specific interaction of the
increasingly clear that natural zinc fingers differentially recognize first three zinc fingers of TFIIIA with the internal control region of
methylated and unmethylated CpG (cytidine-guanidine dinucleo- the Xenopus 5 S RNA gene. J Mol Biol 1992; 223(4):857-871.
tide pairs) and play important regulatory roles in the expression of 17. Wuttke DS, Foster MP, Case DA et al. Solution structure of the
first three zinc fingers of TFIIIA bound to the cognate DNA
target genes in vivo. CTCF (CCCTC-binding factor) has eleven sequence: determinants of affinity and sequence specificity. J Mol
C2H2 zinc fingers and binds to an element present between a Biol 1997; 273(1):183-206.
promoter and an enhancer in order to block the enhancer activ- 18. Nolte RT, Conlin RM, Harrison SC et al. Differing roles for zinc
ity.73 The CTCF binding sites are widely distributed in verte- fingers in DNA recognition: structure of a six-finger transcription
brates and several similar CpG-rich sequences are present in ICR factor IIIA complex. Proc Natl Acad Sci USA 1998;
(imprinted-control region) of the ifg2/H19 as well. Electrophoretic 95(6):2938-2943.
19. Clemens KR, Zhang P, Liao X et al. Relative contributions of the
mobility shift experiments have demonstrated that CTCF binds zinc fingers of transcription factor IIIA to the energetics of DNA
to the CpG-rich elements of ICR but only when these are binding. J Mol Biol 1994; 244(1):23-35.
unmethylated.74 Although more experiments have to be done to 20. Fairall L, Rhodes D, Klug A. Mapping of the sites of protection on
obtain the CTCF function in vivo, the zinc finger a 5 S RNA gene by the Xenopus transcription factor IIIA. A model
proteins are likely to participate in the genomic imprinting pro- for the interaction. J Mol Biol 1986; 192(3):577-591.
cess by differentially binding to the methylated- and unmethylated 21. Hoffmann A, Ciani E, Boeckardt J et al. Transcriptional activities of
CpG-rich elements. the zinc finger protein Zac are differentially controlled by DNA
binding. Mol Cell Biol 2003; 23(3):988-1003.
Kaiso, with triple C2H2 zinc fingers, is another example of 22. Fan CM, Maniatis T. A DNA-binding protein containing two widely
this type of DNA binding. However, in this case, the zinc fingers separated zinc finger motifs that recognize the same DNA sequence.
recognize the methylated CpG-rich sequence but not the Genes Dev 1990; 4(1):29-42.
unmethylated sequence.75,76 In vivo transfection experiments have 23. Sun L, Liu A, Georgopoulos K. Zinc finger-mediated protein
shown that kaiso represses expression of genes with methylated interactions modulate Ikaros activity, a molecular control of
binding sites in the promoter.75 lymphocyte development. EMBO J 1996; 15(19):5358-5369.
24. Mackay JP, Crossley M. Zinc fingers are sticking together. Trends
Biochem Sci 1998; 23(1):1-4.
Acknowledgements 25. Iuchi S, Lin EC. Adaptation of Escherichia coli to respiratory
I thank Dr. R. Freeman for information on the Pfam database. conditions: regulation of gene expression. Cell 1991; 66(1):5-7.
26. Hsu T, Gogos JA, Kirsh SA et al. Multiple zinc finger forms resulting
from developmentally regulated alternative splicing of a transcription
References factor gene. Science 1992; 257(5078):1946-1950.
1. Klug A, Schwabe JW. Protein motifs 5. Zinc fingers. FASEB J 1995; 27. Gogos JA, Hsu T, Bolton J et al. Sequence discrimination by
9(8):597-604. alternatively spliced isoforms of a DNA binding zinc finger domain.
2. Pabo CO, Peisach E, Grant RA. Design and selection of novel Science 1992; 257(5078):1951-1955.
Cys2His2 zinc finger proteins. Annu Rev Biochem 2001; 70:313-340. 28. Greisman HA, Pabo CO. A general strategy for selecting high-affinity
3. Laity JH, Lee BM, Wright PE. Zinc finger proteins: new insights zinc finger proteins for diverse DNA target sites. Science 1997;
into structural and functional diversity. Curr Opin Struct Biol 2001; 275(5300):657-661.
11(1):39-46.
C2H2 Zinc Fingers As DNA Binding Domains 13
29. Hamilton TB, Borel F, Romaniuk PJ. Comparison of the DNA 53. Isalan M, Choo Y. Engineering nucleic acid-binding proteins by
binding characteristics of the related zinc finger proteins WT1 and phage display. Methods Mol Biol 2001; 148:417-429.
EGR1. Biochemistry 1998; 37(7):2051-2058. 54. Wilson TE, Day ML, Pexton T et al. In vivo mutational analysis of
30. Liggins JR, Privalov PL. Energetics of the specific binding interaction the NGFI-A zinc fingers. J Biol Chem 1992; 267(6):3718-3724.
of the first three zinc fingers of the transcription factor TFIIIA with 55. Choo Y, Klug A. A role in DNA binding for the linker sequences
its cognate DNA sequence. Proteins 2000; Suppl 4:50-62. of the first three zinc fingers of TFIIIA. Nucleic Acids Res 1993;
31. Houbaviy HB, Burley SK. Thermodynamic analysis of the interaction 21(15):3341-3346.
between YY1 and the AAV P5 promoter initiator element. Chem 56. Elrod-Erickson M, Rould MA, Nekludova L et al. Zif268
Biol 2001; 8(2):179-187. protein-DNA complex refined at 1.6 A: a model system for
32. Matsushita K, Sugiura Y. Effect of arginine mutation of alanine-556 understanding zinc finger-DNA interactions. Structure 1996;
on DNA recognition of zinc finger protein Sp1. Bioorg Med Chem 4(10):1171-1180.
2001; 9(9):2259-2267. 57. Laity JH, Dyson HJ, Wright PE. DNA-induced alpha-helix capping
33. Miller JC, Pabo CO. Rearrangement of side-chains in a Zif268 in conserved linker sequences is a determinant of binding affinity in
mutant highlights the complexities of zinc finger-DNA recognition. Cys(2)-His(2) zinc fingers. J Mol Biol 2000; 295(4):719-727.
J Mol Biol 2001; 313(2):309-315. 58. Foster MP, Wuttke DS, Radhakrishnan I et al. Domain packing and
34. Harlow E, Lane D. Antibodies. New York: Cold Spring Harbor dynamics in the DNA complex of the N-terminal zinc fingers of
Labortory; 1988. TFIIIA. Nat Struct Biol 1997; 4(8):605-608.
35. Saenger W. Principles of nucleic acid structure. New York: 59. Wu LC. ZAS: C2H2 zinc finger proteins involved in growth and
Springer-Verlag; 1983. development. Gene Expr 2002; 10(4):137-152.
36. Pavletich NP, Pabo CO. Zinc finger-DNA recognition: crystal 60. Tseng H, Green H. Basonuclin: A keratinocyte protein with multiple
structure of a Zif268-DNA complex at 2.1 A. Science 1991; paired zinc fingers. Proc Natl Acad Sci USA 1992;
252(5007):809-817. 89(21):10311-10315.
37. Wolfe SA, Nekludova L, Pabo CO. DNA recognition by Cys2His2 61. Ptashne M. A genetic switch: Gene control and phage λ. Cambridge:
zinc finger proteins. Annu Rev Biophys Biomol Struct 2000; Cell Press & Blackwell Scientific Publications; 1986.
29:183-212. 62. McCarty AS, Kleiger G, Eisenberg D et al. Selective dimerization of
38. Pavletich NP, Pabo CO. Crystal structure of a five-finger GLI-DNA a C2H2 zinc finger subfamily. Mol Cell 2003; 11(2):459-470.
complex: new perspectives on zinc fingers. Science 1993; 63. Wang BS, Grant RA, Pabo CO. Selected peptide extension contacts
261(5129):1701-1707. hydrophobic patch on neighboring zinc finger and mediates
39. Fairall L, Schwabe JW, Chapman L et al. The crystal structure of a dimerization on DNA. Nat Struct Biol 2001; 8(7):589-593.
two zinc-finger peptide reveals an extension to the rules for 64. Tsai RY, Reed RR. Identification of DNA recognition sequences and
zinc-finger/DNA recognition. Nature 1993; 366(6454):483-487. protein interaction domains of the multiple-Zn-finger protein Roaz.
40. Yokono M, Saegusa N, Matsushita K et al. Unique DNA binding Mol Cell Biol 1998; 18(11):6447-6456.
mode of the N-terminal zinc finger of transcription factor Sp1. 65. Shi Y, Lee JS, Galvin KM. Everything you have ever wanted to
Biochemistry 1998; 37(19):6824-6832. know about Yin Yang 1. Biochim Biophys Acta 1997;
41. Laity JH, Dyson HJ, Wright PE. Molecular basis for modulation of 1332(2):F49-66.
biological function by alternate splicing of the Wilms’ tumor 66. Thomas MJ, Seto E. Unlocking the mechanisms of transcription
suppressor protein. Proc Natl Acad Sci USA 2000; factor YY1: are chromatin modifying enzymes the key? Gene 1999;
97(22):11932-11935. 236(2):197-208.
42. Torrungruang K, Alvarez M, Shah R et al. DNA binding and gene 67. Song CZ, Keller K, Chen Y et al. Functional interplay between CBP
activation properties of the Nmp4 nuclear matrix transcription and PCAF in acetylation and regulation of transcription factor KLF13
factors. J Biol Chem 2002; 277(18):16153-16159. activity. J Mol Biol 2003; 329(2):207-215.
43. Benos PV, Lapedes AS, Stormo GD. Probabilistic code for DNA 68. Song CZ, Keller K, Murata K et al. Functional interaction between
recognition by proteins of the EGR family. J Mol Biol 2002; coactivators CBP/p300, PCAF, and transcription factor FKLF2. J
323(4):701-727. Biol Chem 2002; 277(9):7029-7036.
44. Suzuki M, Brenner SE, Gerstein M et al. DNA recognition code of 69. Yao YL, Yang WM, Seto E. Regulation of transcription factor YY1
transcription factors. Protein Eng 1995; 8(4):319-328. by acetylation and deacetylation. Mol Cell Biol 2001;
45. Choo Y, Klug A. Designing DNA-binding proteins on the surface 21(17):5979-5991.
of filamentous phage. Curr. Opin Biotechnol 1995; 6(4):431-436. 70. Dovat S, Ronni T, Russell D et al. A common mechanism for mitotic
46. Wolfe SA, Grant RA, Elrod-Erickson M et al. Beyond the inactivation of C2H2 zinc finger DNA-binding domains. Genes Dev
“recognition code”: structures of two Cys2His2 zinc finger/TATA 2002; 16(23):2985-2990.
box complexes. Structure (Camb) 2001; 9(8):717-723. 71. Cziferszky A, Mach RL, Kubicek CP. Phosphorylation positively
47. Gogos JA, Jin J, Wan H et al. Recognition of diverse sequences by regulates DNA binding of the carbon catabolite repressor Cre1 of
class I zinc fingers: asymmetries and indirect effects on specificity in Hypocrea jecorina (Trichoderma reesei). J Biol Chem 2002;
the interaction between CF2II and A+T-rich elements. Proc Natl 277(17):14688-14694.
Acad Sci USA 1996; 93(5):2159-2164. 72. Choo Y. Recognition of DNA methylation by zinc fingers. Nat Struct
48. Tsui V, Radhakrishnan I, Wright PE et al. NMR and molecular Biol 1998; 5(4):264-265.
dynamics studies of the hydration of a zinc finger-DNA complex. J 73. Bell AC, West AG, Felsenfeld G. The protein CTCF is required for
Mol Biol 2000; 302(5):1101-1117. the enhancer blocking activity of vertebrate insulators. Cell 1999;
49. Choo Y, Isalan M. Advances in zinc finger engineering. Curr Opin 98(3):387-396.
Struct Biol 2000; 10(4):411-416. 74. Hark AT, Schoenherr CJ, Katz DJ et al. CTCF mediates
50. Beerli RR, Barbas CF, 3rd. Engineering polydactyl zinc-finger methylation-sensitive enhancer-blocking activity at the H19/Igf2
transcription factors. Nat Biotechnol 2002; 20(2):135-141. locus. Nature 2000; 405(6785):486-489.
51. Choo Y, Klug A. Toward a code for the interactions of zinc fingers 75. Prokhortchouk A, Hendrich B, Jorgensen H et al. The p120 catenin
with DNA: selection of randomized fingers displayed on phage partner Kaiso is a DNA methylation-dependent transcriptional
[published erratum appears in Proc Natl Acad Sci USA 1995; repressor. Genes Dev 2001; 15(13):1613-1618.
92(2):646]. Proc Natl Acad Sci USA 1994; 91(23):11163-11167. 76. Daniel JM, Spring CM, Crawford HC et al. The p120(ctn)-binding
52. Rebar EJ, Pabo CO. Zinc finger phage: affinity selection of fingers partner Kaiso is a bi-modal DNA-binding protein that recognizes
with new DNA-binding specificities. Science 1994; both a sequence-specific consensus and methylated CpG
263(5147):671-673. dinucleotides. Nucleic Acids Res 2002; 30(13):2911-2919.
CHAPTER 3
T
ranscription factor IIIA (TFIIIA) is widely regarded as Expression of a fluorescent green protein/TFIIIA fusion pro-
the archetypal zinc finger protein. It is a member of a tein shows that TFIIIA is localized as expected in the nucleolus6
very large multigene family of eukaryotic DNA-binding where rRNA genes are transcribed and ribosomes are assembled.
proteins. More than two decades of research have been dedi- Two other pol III transcription factors are also required for initia-
cated to understanding its interaction with the 5S ribosomal RNA tion of transcription of 5S DNA, namely TFIIIB and TFIIIC. In
gene (5S DNA). TFIIIA has nine tandem C2H2 zinc fingers along contrast to TFIIIA these factors contain a number of subunits.
the peptide sequence. The three-dimensional structure of the TFIIIB, which binds upstream of TFIIIA to 5S DNA, consists of
N-terminal 6 zinc fingers bound to 31 base pairs of DNA shows three subunits, the TATA box-binding protein TBP, a single zinc
that not all fingers are equal. Four of them make contacts lo- finger protein Brf and a third subunit B. TFIIIC contains at least
cated on both DNA strands while two fingers act as spacers. six subunits, some of which bind to sequences essential for tran-
Individual fingers can recognize overlapping and interlocking scription within the ICR (for a review see ref. 7).
base pair quartets. Side chains in the short alpha helices of fin- 5S rRNA is the product of transcription from 5S DNA. It is a
gers contact bases in the major groove. It is likely that the linker component of the ribosome 60S subunit. The 120 nucleotide
sequences connecting adjacent zinc fingers evolved to dictate sequence of 5S rRNA is highly conserved among vertebrates. There
which fingers bind to DNA. Signals essential for transcription are two major types of 5S rRNA genes in eukaryotes. Xenopus
initiation of 5S DNA, nuclear localization, and nucleocytoplas- oocyte type (Xlo) and the somatic type (Xls) genes have sequences
mic transport of 5S ribosomal RNA (5S rRNA) are located in which differ at eight nucleotide positions between them.8 Six of
the C-terminal part of TFIIIA. these sequence positions lie within the ICR region. The differ-
ences in sequence do not appear to significantly affect TFIIIA
Introduction binding to 5S DNA in vitro.
An RNA-binding protein accumulates in the oocytes of the
ovaries of the African clawed frog (Xenopus laevis). This protein has TFIIIA Structure and Function
a molecular mass of 40 kilodaltons (kD). About 1010 molecules of
the protein are estimated to be present per oocyte and using fluo-
rescent antibodies it can be seen that this protein is localized mainly Internal Domains
outside of the nucleus.1 Its partner in the cytoplasm was identified In addition to the oocyte form of TFIIIA (oTFIIIA) there is a
as 5S rRNA. About half of the 5S rRNA in the oocyte is associated larger somatic protein, sTFIIIA, which has 22 additional amino
with the 40 kD protein in a stable complex called the 7S storage acids at the N-terminus originating from translation at an up-
particle (7S RNP).2 Similar 7S RNPs are also present in the oo- stream AUG codon.9,10 The existence of sTFIIIA has been con-
cytes of other amphibians, teleost fish and maize embryos.3 firmed by precipitation with antibodies. Cellular levels of the two
In 1980 a transcription factor that is essential for RNA poly- different forms of TFIIIA are probably related to switching from
merase III (pol III) to transcribe Xenopus oocyte 5S rRNA genes the production of oocyte 5S rRNA entirely to that of somatic 5S
was isolated. It was shown to be identical to the 40 kD RNA-binding rRNA in the developing embryo.
protein. The protein isolated from the 7S RNP binds specifically The internal organization of oTFIIIA has been probed by lim-
to radioactively labeled 5S DNA.4,5 This transcription factor was ited digestion of 7S RNP with proteases such as elastase, trypsin
named TFIIIA. In fact this was the first eukaryotic transcription or papain.11 These enzymes can readily cut in exposed regions of
factor to be purified and used to transcribe an isolated gene. Sur- a protein that would be expected to lie between the relatively
prisingly the promoter region that TFIIIA recognizes was found to resistant folded domains. In each case, proteolytic fragments are
be located inside the gene itself. Even more impressive was the dis- released from the C-terminus of the 40 kD protein in a step-wise
covery that a single TFIIIA protein binds without interruption to a fashion to give a 35 kD product first and then a 30 kD core
long sequence of about 55 base pairs. This region of the 5S DNA fragment. The same oTFIIIA digestion pattern is obtained from
was named the internal control region (ICR). either the 7S RNP or a TFIIIA-5S DNA complex. The 30 kD core
Zinc Finger Proteins: From Atomic Contact to Cellular Function, edited by Shiro Iuchi and Natalie Kuldell.
©2005 Landes Bioscience/Eurekah.com and Kluwer Academic/Plenum Publishers.
TFIIIA: A Sophisticated Zinc Finger Protein 15
Zinc Fingers
Addition of a metal chelating agent lowers the DNA-binding
activity of TFIIIA. However, the protein function can be restored
by addition of zinc.13,14 The presence of integral bound zinc in
the 7S RNP was shown by extended X-ray absorption fine struc-
ture (EXAFS).15 Atomic absorption spectroscopy of the 7S RNP12
and chemical analysis of purified oTFIIIA16 established that there
are about 9-11 moles of zinc in the protein. At the time, it was
very surprising to find that a transcription factor with no enzy-
matic activity was a metalloprotein containing multiple zinc ions.
The amino acid sequence of the 344 residue X. laevis oTFIIIA
was published in 1984.17 Its organization, 9 alternating pairs of
cysteines and pairs of histidines always separated by 12 residues, Figure 1. N-terminal zinc fingers 1-6 of TFIIIA. The crystal structure
was unique. Such a repeating pattern can be seen particularly of a fragment of X. laevis oTFIIIA consisting of residues 10-188 is shown
well in a DotPlot analysis,18 in which a TFIIIA sequence is com- above.48 Individual zinc fingers are numbered from zf1 to zf6. The
pared with itself. The nine internal repeats are revealed as regu- secondary structure of the protein, showing the short alpha helices and
larly spaced, parallel, diagonal lines of scored points (see Fig. 1 in beta strands, is represented as a ribbon and colored green. Zinc atoms
ref. 18). It is striking that repeat 6 has the lowest similarity to the in the fingers are represented as small spheres in violet. An endwise view
others. The nine sequence repeat segments can be aligned using which shows the right-handed helical conformation adopted by fingers
the invariant cysteines and histidines to reveal a consensus 30-34 1-3 (residues 10-99) is seen on the lower left. In contrast fingers 4-6
amino acid motif.12,18 (residues 100-188) have a different, irregular structure that is shown on
It was correctly proposed that consecutive pairs of cysteines the lower right. This figure was made with Swiss-PdbViewer and
and of histidines most likely combine together as four ligands Pov-Ray™.62,63
binding a zinc ion.12 Similar arrangements of cysteine and histi-
dine ligands are common in metalloprotein enzymes. This Comparison of TFIIIA Sequences
Cys2His2 motif, often described as C2H2, has now come to be
The genomic DNA of X. laevis TFIIIA contains nine exons.23,24
known as the TFIIIIA-like zinc finger. In contrast C2H2 zinc fin- Some of these encode single zinc fingers suggesting that the pro-
ger motifs connected with a highly conserved five residue H/C tein was created through gene duplication. TFIIIA sequences con-
linker sequence (TGEKP) are often called Kruppel-like zinc fin- taining nine zinc fingers are presently known for several amphib-
gers.19 The nine zinc fingers in tandem that occupy nearly all of ians; Xenopus laevis,17 Xenopus borealis,25 Rana catesbeiana,25 Bufo
the N-terminal part of TFIIIA (residues 13-276) are required for americanus and Rana pipiens.26 TFIIIA sequences have also been
transcriptional activity, together with the adjacent C-terminal reported for channel catfish27 and some mammmals; mouse,28
region (residues 277-344). rat28 and human.29,30 Bakers’ yeast TFIIIA31 has a long sequence
inserted between fingers 8/9 and the fission yeast protein32 has
C-Terminal Signals 10 zinc fingers with a long insert between fingers, 9/10. A plant
The C-terminal part of oTFIIIA contains several amino acid sequence from Aradopsis thaliana33 also has nine zinc fingers, but
sequences that are required for the different functions of the pro- has long inserts between fingers 1/2 and between 4/5. Proteins
tein. A lysine-rich nuclear localization signal (NLS), EKRKLKEK with TFIIIA-like activity have also been identified in other plants,
(residues 274-286) is situated after finger 9. The presence of an such as, tulip34 and maize,35 and in the amoeboid protozoon
Acanthamoeba castellanii.36
NLS allows TFIIIA to migrate to the nucleus where it acts as a
Although TFIIIA is perhaps the best known example of a C2H2
transcription factor. Domain swapping has demonstrated the
zinc finger protein, it is actually quite atypical. Sequences from
presence of a functional leucine-rich nuclear export signal (NES), different species contain additional N-terminal and C-terminal
SLVLDKLTIQ (residues 335-344) close to the C-terminus of extensions of varying lengths. There is little internal sequence
Xenopus oTFIIIA.20 These signals suggest that TFIIIA might be similarity between any of the nine zinc fingers. The number of
a nuclear shuttle protein. However, although TFIIIA is certainly intervening residues found in a cysteine pair varies from 2 to 5,
involved in the nuclear export of 5S rRNA, reentry of 5S rRNA and from 3 to 5 for a histidine pair, in contrast to the more preva-
into the nucleus takes place via a 5S RNP with ribosomal pro- lent Cx2C and Hx3H that is typical of Kruppel-type zinc finger
tein L5.21 A further 15 amino acid sequence, KRSLASRLTGYIP proteins. Sequence divergence within the protein has led to the
(residues 291-304), has been identified as the transcription acti- proposal that TFIIIA may have evolved two differently adapted
vating signal.22 Deletion of the latter sequence inhibits transcrip- sets of zinc fingers, 1-3 that recognize DNA, and 4-7 that bind to
tion initiation. 5S rRNA.37,38
16 Zinc Finger Proteins: From Atomic Contact to Cellular Function
(for a review see ref. 39). Such approaches cannot provide the
details of the atomic contacts between amino acid side chains of
TFIIIA and individual base pairs in the target DNA. Neverthe-
less, they have established that the ICR contains three sequences
essential for transcription of 5S DNA. These are called the box A
sequence (+50 to +60), the intermediate element (+69 to +71)
and the box C sequence (+80 to +89) (see Fig. 2).40
Protection of DNA by bound proteins can be probed with
chemicals that methylate purine bases or ethylate phosphates and
with hydroxyl radical reagents that cleave exposed regions of the
phosphodiester chain. Accessibility can also be tested by diges-
tion with DNase I, II and micrococcal nuclease. In the case of
Xenopus TFIIIA, the DNase I footprint spans positions +42 to
+96 (54 base pairs) of the plus (noncoding) strand of somatic 5S
DNA. A similar span of the minus (coding strand) is also pro-
tected. This region corresponds closely to the ICR. A similar foot-
print from +42 to +96 can also be produced by binding of a trun-
cated version of TFIIIA (fingers 1-9) synthesized in and purified
from Escherichia coli.41
observations indicate that the limits of protected 5S DNA are helix, the remaining fingers 4-6 have a different, extended con-
+62/+63 to +89 for fingers 1-7 whereas intact TFIIIA influences figuration that runs along the DNA. Fingers 1, 2, 3 and 5 bind in
a region extending from +44 to +94. the major groove in the classical way that is found in the crystal
structures of other zinc finger proteins. In contrast, fingers 4 and
DNA Interaction of C-Terminal Fingers 6 are positioned above the minor groove and act as spacers be-
The binding site of the C-terminal part of TFIIIA can be de- tween their neighboring DNA-binding fingers. The two nonbind-
duced by subtracting the footprints of a series of C-terminal trun- ing fingers allow the protein to change direction and cross the
cated mutants from each other.41 By comparing the DNase I foot- minor groove twice. This results in a customized fit in which
print of TFIIIA with those of fragments 1-9 and 1-8, zinc fingers fingers 1-3 bind to the box C sequence and finger 5 binds to the
8-9 can be assigned to the box A sequence, base pairs +40 to +61 intermediate element of the ICR.
of the ICR.44 The existence of direct contacts with base pairs is In an X-ray structure at a modest resolution of 3.1 angstroms
suggested by the TFIIIA-induced protection from methylation it is not possible to be absolutely certain of the protein-DNA
of guanine bases G53, G57 (coding strand) and G51, G59, G60 interactions. However, consideration of atom positions in the
(noncoding strand).45 The methylation of G51, G52 and G57 model, the distances and angles between atoms, together with
also interferes with TFIIIA binding to DNA.46 Similarly, protec- their chemical nature, identifies some if not all of the bonds. The
tion from hydroxyl radical cleavage occurs between noncoding DNA contacts span base pairs +68 to +93. Fingers 1-3 bind in
positions +53 to +57.47 This is further supported by resistance to the major groove to bases in the box C sequence (+81 to +91)
exonuclease III digestion on coding strand +44 to +62. All these xxAxGGGxxxx/xxTxCxCxxTG, in which only bases recognized
results are consistent with a hypothetical model in which zinc by the fingers are specified, and finger 5 binds to bases in the
fingers 7-9 are in contact with certain guanine bases in the major intermediate element (+70 to +74) GGxxx/xxxxT. There are ad-
groove, perhaps binding in a similar way to fingers 1-3.46 ditional contacts to the 5'-phosphates of nucleotides 69, 79, 83,
84, 86, 87, 88 and to riboses 68, 81, 84 on the noncoding strand,
and contacts to the 5'-phosphates of nucleotides 74, 86, 91, 92,
X-Ray Structure Determination 93 and riboses 75, 89, 92 on the coding strand.
in TFIIIA and connect fingers 1/2 and 2/3. As fingers 1-3 wrap References
around the major groove of B-form DNA, their particular linkers 1. Mattaj IW, Lienhard S, Zeller R et al. Nuclear exclusion of
may have evolved to support this kind of helical side-by-side transcription factor IIIA and the 42S particle transfer RNA-binding
packing. protein in Xenopus oocytes: A possible mechanism for gene control?
J Cell Biol 1983; 97(4):1261-1265.
The abrupt turn in the polypeptide chain before the linker is 2. Picard B, Wegnez M. Isolation of a 7S particle from Xenopus laevis
stabilized by hydrogen bonding interactions.53 There is a H-bond oocytes: A 5S RNA-protein complex. Proc Natl Acad Sci USA 1979;
between the main chain NH groups of Glu40 and Glu70 and 76(1):241-245.
Gln132 and the side chain gamma oxygens of threonines Thr38 3. Rincon-Guzman A, Beltran-Pena E, Ortiz-Lopez A et al.
and Thr86 and Thr130. Main chain H-bonds also exist between Ribonucleoprotein particles of quiescent maize embryonic axes. Plant
the carbonyl oxygen of the +5 residue of the preceding alpha he- Mol Biol 1998; 38(3):357-364.
4. Pelham HR, Brown DD. A specific transcription factor that can
lix and NH groups in the adjoining linker. As is the case in linker
bind either the 5S RNA gene or 5S RNA. Proc Natl Acad Sci USA
1/2 where the carbonyl oxygen of Cys35 makes H-bonds to Gly39 1980; 77(7):4170-4174.
and to Glu40 main chain NH groups. Similarly, in the 2/3 linker 5. Honda BM, Roeder RG. Association of a 5S gene transcription factor
the Leu65 carbonyl oxygen makes H-bonds to Gly69 NH and to with 5S RNA and altered levels of the factor during cell
Glu70 NH groups. These interactions may stabilize the linker differentiation. Cell 1980; 22(1 Pt 1):119-126.
conformation so that a pair of fingers is correctly positioned for 6. Mathieu O, Yukawa Y, Prieto J-L et al. Identification and
major groove DNA binding. The sequence variation in other characterization of transcription factor IIIA and ribosomal protein
L5 from Arabidopsis thaliana. Nucleic Acids Res 2003; 31(9):
TFIIIA linkers may have evolved to allow alternative spatial ar- 2424-2433.
rangements, or to dictate which of the fingers actually binds to 7. Huang Y, Maraia RJ. Comparison of the RNA polymerase III
RNA and DNA. transcription machinery in Schizosaccharomyces pombe,
Saccharomyces cerevisiae and human. Nucleic Acids Res 2001;
29(13):2675-2690.
Mutagenesis and TFIIIA Activity 8. Xing YY, Worcel A. The C-terminal domain of transcription factor
While there has been a greater effort concentrated in probing IIIA interacts differently with different 5S RNA genes. Mol Cell
the 5S DNA, the activity of Xenopus TFIIIA has also been inves- Biol 1989; 9(2):499-514.
tigated by mutagenesis (see Fig. 2). The effects of amino acid 9. Pelham HR, Wormington WM, Brown DD. Related 5S RNA
substitutions and consequences of “finger swap” experiments upon transcription factors in Xenopus oocytes and somatic cells. Proc Natl
DNA affinity have been investigated.54 Replacement of linker Acad Sci USA 1981; 78(3):1760-1764.
10. Kim SH, Darby MK, Joho KE et al. The characterization of the
sequences or mutations at the T, G, K or P positions within the TFIIIA synthesized in somatic cells of Xenopus laevis. Genes Dev
1/2 and 2/3 linkers cause a drastic decrease in DNA binding.55,56 1990; 4(9):1602-1610.
Substituting residue +6 in the alpha helix of finger 2 reduces the 11. Smith DR, Jackson IJ, Brown DD. Domains of the positive
binding activity.57 Mutations at positions +2, +5 or +6 in the transcription factor specific for the Xenopus 5S RNA gene. Cell
alpha helix of finger 3,58 at the +3 position of finger 4 or the -1 1984; 37(2):645-652.
position of finger 6 are also detrimental to the binding activity.41 12. Miller J, McLachlan AD, Klug A. Repetitive zinc-binding domains
in the protein transcription factor IIIA from Xenopus oocytes.
The exchange of finger 5 or finger 7 with their counterparts EMBO J 1985; 4:1609-1614.
from another nine-zinc finger protein, p43, results in a decrease 13. Hanas JS, Hazuda DJ, Bogenhagen DF et al. Xenopus transcription
of TFIIIA affinity for DNA.54 Many loss-of-function mutations factor A requires zinc for binding to the 5S RNA gene. J Biol Chem
in fingers 7-9, the C-terminus and zinc binding residues have 1983; 258(23):14120-14125.
been identified using a genetic assay in yeast cells.59 The function 14. Wingender E, Dilloo D, Seifart KH. Zinc ions are differentially
of TFIIIA demands a highly extended and modular protein struc- required for the transcription of ribosomal 5S RNA and tRNA in a
HeLa-cell extract. Nucleic Acids Res 1984; 12(23):8971-8985.
ture. Thus it is not surprising that mutations in the linkers, alpha 15. Diakun GP, Fairall L, Klug A. EXAFS study of the zinc-binding
helices and metal binding residues that may affect zinc finger struc- sites in the protein transcription factor IIIA. Nature 1986;
ture also interfere with DNA-binding activity. 324(6098):698-699.
16. Brown RS. Structural investigation of the Xenopus 7S
TFIIIA: A Sophisticated Protein Ribonucleoprotein complex. PhD Thesis 1990. London University,
Great Britain.
TFIIIA is a gene-specific transcription factor that has evolved 17. Ginsberg AM, King BO, Roeder RG. Xenopus 5S gene transcription
with a number of very different cellular functions. It plays an factor, TFIIIA: characterization of a cDNA clone and measurement
important role in the formation of pol III transcription initiation of RNA levels throughout development. Cell 1984; 39(3 Pt
complexes, the nuclear export and cytoplasmic storage of riboso- 2):479-489.
mal 5S RNA in oocytes, and also interacts with other proteins.60 18. Brown RS, Sander C, Argos P. The primary structure of transcription
factor TFIIIA has 12 consecutive repeats. FEBS Lett 1985;
Expression of the Xenopus TFIIIA gene is developmentally regu- 186(2):271-274.
lated, giving rise to short and longer forms of the protein specific 19. Rosenberg UB, Schroeder C, Preiss A et al. Structural homology of
to oocyte and somatic cells. the product of the Drosophila Krueppel gene with Xenopus
It is an unusual modular protein containing multiple zinc fin- transcription factor IIIA. Nature 1986; 319:336-339.
gers that are highly adapted to recognize elements in a long pro- 20. Fridell RA, Fischer U, Luhrmann R et al. Amphibian transcription
moter sequence, rather than bind to a short symmetrical DNA factor IIIA proteins contain a sequence element functionally
palindrome. Surprisingly it is not necessary for all of its fingers to equivalent to the nuclear export signal of human immunodeficiency
virus type 1 Rev. Proc Natl Acad Sci USA 1996; 93(7):2936-2940.
exhibit DNA-binding activity although they are virtually indis- 21. Rudt F, Pieler T. Cytoplasmic retention and nuclear import of 5S
tinguishable in structure.48 TFIIIA is a member of a group of ribosomal RNA containing RNPs. EMBO J 1996; 15(6):1383-91.
fascinating proteins that “have it both ways” through a dual abil- 22. Mao X, Darby MK. A position-dependent transcription-activating
ity to interact with either double stranded DNA or folded RNA.61 domain in TFIIIA. Mol Cell Biol 1993; 13(12):7496-7506.
TFIIIA: A Sophisticated Zinc Finger Protein 19
23. Tso JY, Van Den Berg DJ, Korn LJ. Structure of the gene for 44. Vrana KE, Churchill MEA, Tullius TD et al. Mapping functional
Xenopus transcription factor TFIIIA. Nucleic Acids Res 1986; regions of transcription factor TFIIIA. Mol Cell Biol 1988;
14(5):2187-2200. 8(4):1684-1696.
24. Taylor W, Jackson IJ, Siegel N et al. The developmental expression 45. Fairall L, Rhodes D, Klug A. Mapping of the sites of protection on
of the gene for TFIIIA in Xenopus laevis. Nucleic Acids Res 1986; a 5 S RNA gene by the Xenopus transcription factor IIIA. A model
14(15):6185-6195. for the interaction. J Mol Biol 1986; 192(3):577-591.
25. Gaskins CJ, Hanas JS. Sequence variation in transcription factor 46. Clemens KR, Liao X, Wolf V et al. Definition of the binding sites
IIIA. Nucleic Acids Res 1990; 18(8):2117-2123. of individual zinc fingers in the transcription factor IIIA-5S RNA
26. Gaskins CJ, Smith JF, Ogilvie MK et al. Comparison of the sequence gene complex. Proc Natl Acad Sci USA 1992; 89(22):10822-10826.
and structure of transcription factor IIIA from Bufo americanus and 47. Hayes JJ, Tullius TD. Structure of the TFIIIA-5 S DNA complex. J
Rana pipiens. Gene 1992; 120(2):197-206. Mol Biol 1992; 227(2):407-417.
27. Ogilvie MK, Hanas JS. Molecular biology of vertebrate transcription 48. Nolte RT, Conlin RM, Harrison SC et al. Differing roles for
factor IIIA: Cloning and characterization of TFIIIA from channel zinc fingers in DNA recognition: Structure of a six-finger
catfish oocytes. Gene 1997; 203(2):103-112. transcription factor IIIA complex. Proc Natl Acad Sci USA 1998;
28. Hanas JS, Hocker JR, Cheng YG et al. cDNA cloning, DNA 95(6):2938-2943.
binding, and evolution of mammalian transcription factor IIIA. Gene 49. Wuttke DS, Foster MP, Case DA et al. Solution structure of the
2002; 282(1-2):43-52. first three zinc fingers of TFIIIA bound to the cognate DNA
29. Arakawa H, Nagase H, Hayashi N et al. Molecular cloning, sequence: determinants of affinity and sequence specificity. J Mol
characterization, and chromosomal mapping of a novel human gene Biol 1997; 273(1):183-206.
(GTF3A) that is highly homologous to Xenopus transcription factor 50. Brown RS, Argos P. Fingers and helices. Nature 1986;
IIIA. Cytogenet Cell Genet 1995; 70(3-4):235-238. 324(6094):215.
30. Drew PD, Nagle JW, Canning RD et al. Cloning and expression 51. Pavletich NP, Pabo CO. Zinc finger-DNA recognition: Crystal
analysis of a human cDNA homologous to Xenopus TFIIIA. Gene structure of a Zif268-DNA complex at 2.1 A. Science 1991;
1995; 159(2):215-218. 252(5007):809-817.
31. Archambault J, Milne CA, Schappert KT et al. The deduced 52. Wolfe SA, Nekludova L, Pabo CO. DNA recognition by Cys2His2
sequence of the transcription factor TFIIIA from Saccharomyces zinc finger proteins. Annu Rev Biophys Biomol Struct 2000;
cerevisiae reveals extensive divergence from Xenopus TFIIIA. J Biol 29:183-212.
Chem 1992; 267(5):3282-3288. 53. Laity JH, Dyson HJ, Wright PE. DNA-induced alpha-helix capping
32. Schulman DB, Setzer DR. Identification and characterization of in conserved linker sequences is a determinant of binding affinity in
transcription factor IIIA from Schizosaccharomyces pombe. Nucleic Cys(2)-His(2) zinc fingers. J Mol Biol 2000; 295(4):719-727.
Acids Res 2002; 30(13):2772-2781. 54. Hamilton TB, Turner J, Barilla K et al. Contribution of individual
33. Mathieu O, Yukawa Y, Prieto JL et al. Identification and amino acids to the nucleic acid binding activities of the Xenopus
characterization of transcription factor IIIA and ribosomal protein zinc finger proteins TFIIIIA and p43. Biochemistry 2001;
L5 from Arabidopsis thaliana. Nucleic Acids Res 2003; 40(20):6093-6101.
31(9):2424-2433. 55. Choo Y, Klug A. A role in DNA binding for the linker sequences
34. Wyszko E, Barciszewska M. Purification and characterization of of the first three zinc fingers of TFIIIA. Nucleic Acids Res 1993;
transcription factor IIIA from higher plants. Eur J Biochem 1997; 21(15):3341-3346.
249(1):107-112. 56. Ryan RF, Darby MK. The role of zinc finger linkers in p43 and
35. Wyszko E, Radlowski M, Bartkowiak S et al. Maize TF IIIA—the TFIIIA binding to 5S rRNA and DNA. Nucleic Acids Res 1998;
first transcription factor IIIA from monocotyledons. Purification and 26(3):703-709.
properties. Acta Biochim Pol 1997; 44(3):579-589. 57. Hanas JS, Koelsch G, Moreland RM et al. Differential requirements
36. Polakowski N, Paule MR. Purification and characterization of of basic amino acids in transcription factor IIIA-5S gene interaction.
transcription factor IIIA from Acanthamoeba castellanii. Gene 1997; Biochim Biophys Acta 1998; 1398(3):256-264.
203(2):103-112. 58. Zang WQ, Veldhoen N, Romaniuk PJ. Effects of zinc finger
37. Clemens KR, Wolf V, McBryant SJ et al. Molecular basis for specific mutations on the nucleic acid binding activities of Xenopus
recognition of both RNA and DNA by zinc finger protein. Science transcription factor IIIA. Biochemistry 1995; 34(47):15545-15552.
1993; 260(5107):530-533. 59. Bumbulis MJ, Wroblewski G, McKean D et al. Genetic analysis of
38. Searles MA, Lu D, Klug A. The role of the central zinc fingers of Xenopus transcription factor IIIA. J Mol Biol 1998;
transcription factor IIIA in binding to 5 S RNA. J Mol Biol 2000; 284(5):1307-1322.
301(1):47-60. 60. Moreland RJ, Dresser ME, Rodgers JS et al. Identification of a
39. Shastry BS. Transcription factor IIIA (TFIIIA) in the second decade. transcription factor IIIA-interacting protein. Nucleic Acids Res 2000;
J Cell Sci 1996; 109(Pt 3):535-539. 28(9):1986-1993.
40. Pieler T, Hamm J, Roeder RG. The 5S gene internal control region 61. Cassiday LA, Maher LJ 3rd. Having it both ways: Transcription
is composed of three distinct sequence elements, organized as two factors that bind DNA and RNA. Nucleic Acids Res 2002;
functional domains with variable spacing. Cell 1987; 48(1):91-100. 30(19):4118-4126.
41. Clemens KR, Zhang P, Liao X et al. Relative contributions of the 62. Guex N, Peitsch MC. SWISS-MODEL and the Swiss-PdbViewer:
zinc fingers of transcription factor IIIA to the energetics of DNA An environment for comparative protein modelling. Electrophoresis
binding. J Mol Biol 1994; 244(1):23-35. 1997; 18(15):2714-2723.
42. Wu C. An exonuclease protection assay reveals heat-shock element 63. Pov-Ray™ (http://www.povray.org).
and TATA box DNA-binding proteins in crude nuclear extracts. 64. Bernstein HJ. Recent changes to RasMol, recombining the variants.
Nature 1985; 317(6032):84-87. Trends Biochem Sci 2000; 25(9):453-455.
43. Nikolaev LG, Glotov BO, Belyavsky AV et al. Identification of
sequence-specific DNA-binding factors by label transfer: Application
to the adenovirus-2 major late promoter. Nucleic Acids Res 1988;
16(2):519-535.
CHAPTER 4
Abstract
nM.12 The GAGA-DBD extends from residues 310 to 372 in
T
he structural basis of sequence specific binding of the
Cys2His2 single zinc finger DNA binding domain of the the intact full length GAGA protein, and consists of a single clas-
transcription factor GAGA is explored on the basis of sical (TFIIIA-like) Cys2-His2 zinc finger13,14 preceded by two
the three-dimensional structure of a complex between the mini- highly basic regions (BR1 and BR2) (Fig. 1A). (Note that there is
mal DNA binding domain (DBD) of the GAGA factor also a basic region C-terminal to the zinc finger domain, but its
(GAGA-DBD) and an oligonucleotide containing its GAGAG presence or absence has no effect on DNA binding affinity12).
consensus binding site. The GAGA-DBD comprises a single clas- While the genes for several proteins containing only a single
sical Cys2His2 zinc finger core, and an N-terminal extension con- classical zinc finger have been identified,12,15,16 the GAGA-DBD
taining two highly basic regions, BR1 and BR2. The zinc finger was the first single classical zinc finger containing protein for which
core binds in the major groove and recognizes the first three sequence specific DNA binding by a Cys2His2 zinc finger region
GAG bases of the consensus in a manner very similar to that has been experimentally demonstrated. In other reported
seen in other classical Cys2His2 zinc finger/DNA complexes. TFIIIA-like zinc finger proteins, a tandem array comprising a mini-
Unlike the latter which require tandem zinc finger repeats with a mum of two zinc binding domains is required for sequence spe-
minimum of two units for high affinity binding, the GAGA-DBD cific high-affinity DNA binding.13,14 The ability of the single
makes use of only a single finger complemented by the two Cys2His2 zinc finger containing protein GAGA to specifically bind
N-terminal basic regions BR1 and BR2. BR2 forms a helix that DNA suggests that all proteins containing a single classical zinc
interacts in the major groove recognizing the last G of the con- fingers are potential DNA-binding transcription factors. Over the
sensus, while BR1 wraps around the DNA in the minor groove last several years, several single zinc-finger domain proteins have
and recognizes the A in the fourth position of the consensus. been identified in a variety of plant species.17 In each case, these
plant proteins contain the highly conserved sequence QALGGH
Introduction in their putative recognition helix within the single zinc-finger
The GAGA factor of Drosophila melanogaster, so-called be- domain. In addition, the SUPERMAN protein from Arabidopsis
cause it binds to (GA)n rich sites, is a TFIIIA-like zinc finger thalania has been recently shown to be a sequence specific
protein which was originally identified on the basis of its ability DNA-binding protein.18 As with the DBD of the GAGA protein,
to bind to (GA)n rich sites in the Ultrabithorax promoter.1,2 The the DBD of the SUPERMAN protein is composed of a single
GAGA factor is encoded by the Thrithorax-like gene,3 is required zinc-finger domain flanked by two basic rich regions. However,
for normal expression of homeotic genes, and acts as a modifier the DBD of SUPERMAN differs from the DBD of the GAGA
of position-effect variegation.4 Both in vivo5,6 and in vitro7-11 protein in that one of the basic regions is C-terminal to the zinc
experiments suggest that the GAGA factor acts as an anti-repressor finger while the second basic region is N-terminal to the zinc fin-
by helping to disrupt nucleosomes associated with gene regula- ger. This arrangement of a DBD containing a single zinc-binding
tory sequences. Putative Drosophila target genes for the GAGA domain flanked by basic regions on both termini has also been
factor include genes such as the heat shock (hsp26 and hsp70), demonstrated with the N-terminal zinc-binding domains of
histone h3/h4, homeotic and housekeeping/constitutive genes.4 chicken GATA-2 and GATA-3.19 These DBDs consisting of a
The GAGA factor is 519 residues in length and comprises three single zinc-binding domains flanked by basic rich regions repre-
domains: an N-terminal POZ/PTB protein interaction domain, sents a very versatile mode of specific DNA recognition that could
a central DNA binding domain, and a polyglutamine-rich be easily exploited in a number of different applications.
carboxy-terminus.4 The minimal DNA binding domain (DBD) In this chapter, adapted from our previously published paper
of the GAGA factor has been delineated and shown to bind spe- describing the three-dimensional solution NMR structure of the
cifically to DNA derived from the h3/h4 promoter and contain- GAGA-DBD/DNA complex (PDB accession code 1YUI),20 we
ing the sequence GAGAGAG with a dissociation constant of 5 summarize the structural basis for sequence specific recognition
Zinc Finger Proteins: From Atomic Contact to Cellular Function, edited by Shiro Iuchi and Natalie Kuldell.
©2005 Landes Bioscience/Eurekah.com and Kluwer Academic/Plenum Publishers.
GAGA: Structural Basis for Single Cys2His2 Zinc Finger-DNA Interaction 21
each other via protein-protein contacts, may be required to achieve 3. Farkas G, Gausz J, Galloni M et al. The trithorax-like gene encodes
efficient disruption of the nucleosome. It also suggests that the the Drosophila GAGA factor. Nature 1994; 371:806-808.
GAGA-factor may have the possibility of migrating along the 4. Granok H, Leibovitch BA, Shaffer CD et al. Ga-ga over GAGA
factor. Curr Biol 1995; 5:238-241.
DNA, diffusing along a linear lattice from one site to the next 5. Lu Q, Wallrath LL, Elgin SCR. (CT)n.(GA)n repeats and heat shock
adjacent site. In the in vivo situation, the presence of such a large elements have a distinct roles in chromatin structure and
number of adjacent GAGA factor binding sites is likely to ensure transcriptional activation of the Drosophila hsp26 gene. Mol Cell
that at least some sites would be located in a linker region on a Biol 13:2802-2814.
nucleosome array, serving as an anchor point for GAGA factor 6. O’Brien T, Wilkins RC, Giardina C et al. Distribution of GAGA
binding. Consistent with that mechanism is the observation that protein on Drosophila genes in vivo. Genes & Dev 1995;
9:1098-1110.
the GAGA factor binds weakly to mononucleosomes, and even 7. Croston GE, Kerrigan LA, Lira M et al. Sequence specific
in the presence of the nucleosome remodeling factor NURF does antirepression of histone H1-mediated inhibition of basal RNA
not cause complete disruption of nucleosome structure under these polymerase II transcription. Science 1991; 251:643-649.
conditions.27 In contrast, the GAGA factor binds tightly to a nu- 8. Raff JW, Kellum R, Alberts B. The Drosophila GAGA transcription
cleosome array containing multiple GAGA factor sites, even in factor is associated with specific regions of heterochromatin
the absence of NURF.9,27 throughout the cell cycle. EMBO J 1994; 13:5977-5983.
9. Tsukiyama T, Becker PB, Wu C. ATP-dependent nucleosome
The GAGA factor is able to alleviate the repression of tran-
disruption at a heat-shock promoter mediated by binding of GAGA
scription by linker histones (H1 or H5).7 Conversely, the linker transcription factor. Nature 1994; 367:525-532.
histone H1 decreases NURF/ATP dependent stimulation of 10. Wall G, Varga-Weisz PD, Sandaltzopoulos R et al. Chromatin
GAGA factor binding.9,29 The globular domain of the linker hi- remodeling by GAGA factor and heat shock factor at the
stones specifically recognizes and binds to the nucleosome core hypersensitive Drosophila hsp26 promoter in vivo. EMBO J 1995;
and is thought to be located asymetrically inside the superhelical 14:1727-1736.
gyre of DNA, just inside the nucleosome core region.30,31 The 11. Shopland LS, Hirayoshi K, Fernandes M et al. HSF access to heat
shock elements in vivo depends critically on promoter architecture
globular domain of the linker histones belongs to the winged defined by GAGA factor, TFIID, and RNA polymerase II binding
helix-turn-helix family of DNA binding proteins,32 exemplified sites. Genes & Dev 1995; 9:2756-2769.
by HNF-3/fork head. 33 While the structure of a linker 12. Pedone PV, Ghirlando R, Clore GM et al. The single Cys2-His2
histone-DNA complex has not yet been determined, presumably zinc finger domain of the GAGA protein flanked by basic residues
because linker histones do not bind naked DNA in a sequence is sufficient for high-affinity specific DNA binding. Proc Natl Acad
specific manner, it seems likely that the mode of binding of the Sci USA 1996; 93:2822-2826.
13. Klug A, Schwabe JWR. Zinc fingers. FASEB J 1995; 9:597-604.
linker histones would be similar to that seen in the complex of 14. Berg JM, Shi Y. The galvanization of biology: A growing appreciation
HNF-3/fork head with DNA. In the latter complex, the recogni- for the roles of zinc. Science 1996; 271:1081-1085.
tion helix of the helix-turn-helix motif binds in the major groove, 15. Sakai H, Medrano LJ, Meyerowitz EM. Role of SUPERMAN in
and the two basic wings contact the adjacent minor grooves on maintaining Arabdopsis floral whorl boundaries. Nature 1995;
either side of the interacting major groove.33 The C-terminal wing 378:199-203.
also makes an arginine mediated base specific contact in the mi- 16. Tague BW, Goodman HM. Characterization of a family of
Arabidopsis zinc finger protein cDNAs. Plant Mol Biol 1995;
nor groove.33 Thus, it is possible that the combination of major
28:267-279.
and minor groove binding by the GAGA-DBD, coupled with its 17. Isernia C, Bucci E, Leone M et al. NMR structure of the single
significantly higher DNA sequence specificity relative to the linker QALGGH zinc finger domain from the Arabidopsis thaliana
histones, permits the GAGA-DBD to directly displace a linker SUPERMAN protein. ChemBioChem 2003; 4:171-180.
histone under certain circumstances (e.g., when the linker his- 18. Dathan N, Zaccaro L, Esposito S et al. The Arabidopsis
tone is directly bound to a GAGA factor site). SUPERMAN protein s able to specifically bind DNA through its
As the GAGA-DBD effectively clamps the DNA by binding single Cys2-His2 zinc finger motif. Nucl Acid Res 2002;
30:4845-4951.
in the major and minor grooves, there are also limitations on the 19. Pedone P, Omichinksi JG, Nony P et al. The N-terminal fingers of
location of the N-terminal POZ domain and the chicken GATA-2 and GATA-3 are independent sequence-specific
polyglutamine-rich C-terminal tail in the DNA complex with DNA binding domains. EMBO J 1997; 16:2874-2882.
the full-length GAGA factor. These must be free to interact with 20. Omichinski JG, Pedone PV, Felsenfeld G et al. The solution structure
other components of the transcription machinery, including the of a specific GAGA factor-DNA complex reveals a modular binding
mode. Nature Struct Biol 1997; 4:122-132.
nucleosome remodeling factor NURF which facilitates the com- 21. Fairall L, Schwabe JWR, Chapman L et al. The crystal structure of
plete displacement of the histone octamer from the complex in a two zinc-finger peptide reveals an extension to the rules of
an ATP-dependent manner.9 zinc-finger/DNA recognition. Nature 1993; 366:483-487.
22. Pavletich NP, Pabo CO. Zinc finger-DNA recognition: Crystal
structure of a Zif268-DNA complex at 2.1 Å. Science 1991;
Acknowledgements 252:809-816.
We thank Gary Felsenfeld, Paolo Pedone and Angela 23. Pavletich NP, Pabo CO. Crystal structure of a five-finger GLI-DNA
Gronenborn for a highly fruitful collaboration on the structure complex: New perspectives on zinc fingers. Science 1993;
of the GAGA-DBD/DNA complex. 261:1701-1707.
24. Dutnall EN, Neuhaus D, Rhodes D. The solution structure of the
first zinc finger domain of SWI5: A novel structural extension to a
References common fold. Structure 1996; 4:599-611.
1. Biggin MD, Tjian R. Transcription factors that activate the 25. Omichinski, JG, Clore GM, Schaad O et al. NMR structure of a
Ultrabithorax promoter in developmentally staged extracts. Cell 1988; specific DNA complex of a Zn-containing DNA binding domain of
53:699-711. GATA-1. Science 1993; 261:438-446.
2. Soeller WC, Oh CE, Kornberg TB. Isolation of cDNAs encoding 26. Luisi B, Xu WX, Otwinowski Z et al. Crystallographic analysis of
the Drosophila GAGA transcription factor. Mol Cell Biol 1993; the interaction of the glucocorticoid receptor with DNA. Nature
13:7961-7970. 1991; 352:497-505.
GAGA: Structural Basis for Single Cys2His2 Zinc Finger-DNA Interaction 25
Abstract helix with respect to the remainder of the first antiparallel β sheet.
The metazoan GATA DBDs always have a 17-residue stretch
T
he highly conserved DNA-binding domains (DBDs) of
eukaryotic GATA factors comprise one or two zinc bind- between the two cysteine pairs (Cys-X2-Cys-X17-Cys-X2-Cys) and
ing modules with four cysteines embedded in the se- a leucine (an alanine in some C. elegans fingers) in the seventh
quence Cys-X2-Cys-X17/18-Cys-X2-Cys and an adjacent basic re- position after the second cysteine. In vertebrates, two adjacent,
gion. The fold has been defined as a class IV zinc finger motif homologous DNA-binding domains are present in tandem, sepa-
and belongs to the superfamily of glucocorticoid receptor-like rated by a few amino acids, with the carboxy-terminal domain
DNA binding domains. Members of the GATA family are found being the physiological important DBD. The amino-terminal
in a wide range of organisms ranging from slime molds to fungi domain can modulate DNA binding specificity and/or partici-
and plants to vertebrates and exhibit differing and complex roles pates in interactions with other proteins.5,6 In invertebrates both
in transcription regulation mediated by binding to regulatory two-domain and one-domain GATA factors occur. Plant GATA
DNA sequences of the form (A/T)GATA(A/G). Fungal GATA factors contain only one DBD and 18 residues between the two
factors control nitrogen metabolism, light induction, siderophore cysteine pairs. Putative plant GATA factors contain either leu-
biosynthesis and mating-type switching, while the prototypical cine or glutamine in the seventh position. Fungal GATA factors
animal factor GATA-1 is involved in the regulation of all eryth- come in two flavors: one is “animal-like” and the second is
roid cell-specific genes. A unique feature of the GATA DBDs is “plant-like.” The residue at position seven after the second cys-
their ability to interact with DNA as well as with proteins, lead- teine of plant-like fungal GATA factors is glutamic acid.7 With
ing to a multitude of intricate functionalities. One important one exception, fungal GATA factors contain only one DBD that
role is associated with dramatic chromatin rearrangement. resembles most closely the carboxy-terminal DBD of the verte-
brate GATA factors.
GATA DBDs are most likely very ancient eukaryotic motifs
Introduction and a sequence clearly related to GATA factors is present in the
The GATA transcription factors represent a major class of stalky gene of Dictyostelium discoideum.8 A evolutionary relation-
zinc-containing, regulatory proteins that are found in a wide range ship of GATA-binding motifs has been published9 and a sequence
of organisms (for reviews see refs. 1 and 2). They are named after comparison for GATA domains of different species is shown in
the consensus DNA sequence that they recognize, containing a Figure 1.
six base-pair sequence with the invariant GATA core.3,4 The amino
acid sequence of the DNA binding domain (DBD) comprises a
Cys2-Cys2 (C2C2) type IV zinc finger motif, followed by a basic Structures of GATA DNA Binding Domains
terminal tail. The core of the structure consists of about 50 resi- and DNA Recognition
dues. At the N terminus the chain starts with a turn that is fol- High-resolution three-dimensional NMR structures have been
lowed by two short, irregular antiparallel β sheets, an α helix and determined for two GATA-binding domains complexed with their
a long loop that contains a helical turn. At the C terminus of the DNA recognition sequence: the chicken GATA-1 carboxy-
core a Ω loop is found. β strands 1 and 2 form the first β sheet, terminal domain and the Aspergillus nidulans AreA domain.10,11
and β strands 3 and 4 form the second β sheet. Following the Ω The domain contains a compact core formed by two pairs of β
loop is a long, irregular tail which, in the free protein, is unstruc- sheets followed by an α helix and a loop (Fig. 2).
tured and highly flexible. The core is dominated by the Zn ion The structure is dominated by the tetrahedrally coordinated
that is tetrahedrally coordinated to the Sγ atoms of the four cys- zinc via the cysteines and numerous additional hydrophobic con-
teines. The first two coordinating cysteine residues are located on tacts that determine the relative orientation of the two β sheets
β strand 1, and the turn between β1 and β2, while the second and the helix. The α helix and the loop connecting β strands β2
pair resides on the α helix. The geometric constraints imposed by and β3 interact with the major groove of the DNA, mostly via
the Zn and its ligands therefore determine the orientation of the hydrophobic amino acids. Methyl-groups of three leucines and
Zinc Finger Proteins: From Atomic Contact to Cellular Function, edited by Shiro Iuchi and Natalie Kuldell.
©2005 Landes Bioscience/Eurekah.com and Kluwer Academic/Plenum Publishers.
The DNA-Binding Domain of GATA Transcription Factors—A Prototypical Type IV Cys2-Cys2 Zinc Finger 27
Figure 4. Comparison between the wild type and mutant AreA com-
plexes with DNA. A) Tube representation of the protein backbones and
space-filling representation of the DNA. Wild type AreA, red tube;
Figure 2. A) Structure of the C-terminal DBD of chicken GATA-1 mutant AreA, orange tube. The major and minor groove surfaces of the
complexed with the DNA target site. The protein backbone is shown in DNA are colored blue and orange, respectively. B) Detailed view of the
green and the color coding for the bases is as follows: A, red; T, lilac; G, interactions around the sites of mutation (residues 22-24). The wild
dark blue; C, light blue. B) Core module of the DBD of AreA. The two type DBD backbone and side chains are shown in red and blue, respec-
anti-parallel β sheets are shown in red, the α helix in blue and the zinc tively, while the mutant AreA backbone and side chains are shown in
ion as a purple sphere surrounded by the four coordinating cysteine gold and green, respectively. The GCG element of the GCGATAG site
residues in orange. Selected side-chains are shown in green. To view is shown in red and the GTG element of the GTGATAG site is shown
online color version, go to http://www.eurekah.com/chapter.php? in tan. To view online color version, go to http://www.eurekah.com/
chapid=1881&bookid=124&catid=30. chapter.php?chapid=1881&bookid=124&catid=30.
28 Zinc Finger Proteins: From Atomic Contact to Cellular Function
database is, however, considerably smaller than that for other tran- “Plant-Like” GATA Factors: Light Regulation
scription factors such as the classical zinc fingers. No functional
conservation across the fungal and metazoan kingdoms or be-
in N. crassa
tween vertebrates and invertebrates seems to have occurred. Al- Many functions in N. crassa are inducible by blue light, in-
though the DBD sequence motif and DNA binding specificity cluding conidiation, mycelial carotenoid biosynthesis and reset-
are conserved, the rest of the protein sequences are widely differ- ting of the circadian rhythm. The two loci involved, WC1 and
ent. Thus, the same motif and binding specificity can serve dif- WC2, encode “plant-like” GATA factors. These two factors con-
ferent purposes in different contexts. tain typical PAS dimerization domains and act as dimers. A rapid
In ascomycetes the use of nitrogen sources involves transcrip- light-induced turnover of WC1 in the WC1/WC2 complex is
tional control at two levels: induction of specific genes and nitro- correlated with the induction of light-inducible promoters.
gen metabolite repression.18 For example, in order to induce ni-
trate and nitrite reductases, both the presence of nitrate and the A Noncanonical Factor Involved in
absence of ammonium and glutamine are required. Significant Mating-Type Switching
chromatin rearrangement was observed that was independent of Although several unique features are present in its finger loop,
transcription and of the specific factor NirA, but strictly depen- Ash1p of budding yeast, clearly seems to be a GATA factor. Dele-
dent on the GATA factor AreA.19 The transcription of genes in- tion of the gene results in massive daughter mating-type switch-
volved in the utilization of most nitrogen sources requires two ing, whereas overexpression prevents mother switching.
transcription factors and each pathway is regulated specifically by
one transcription factor of the zinc binuclear cluster class. The
general regulator is always a GATA factor. The binding (or acti- The Leucine to Valine Mutant of AreA:
vation) of the specific factor is thought to require the presence of A Paradigm for the Structural Basis of the
a specific inducer, the GATA factor, and a low intracellular con- Associated Phenotype
centration of glutamine. The overall process is very well under- The Arst laboratory has mapped and characterized a myriad
stood and similar in Aspergillus nidulans and Neurospora crassa. of mutations in the AreA DBD. As these mutations usually affect
The master gene and GATA factor of nitrogen regulation is the growth of A. nidulans on all or some nitrogen sources, second
NIT2 in N. crassa and areA in A. nidulans.18 The accepted model site revertants can be isolated. Thus, the A. nidulans AreA system
is that a negative-acting protein forms a complex with NIT2 (and is a treasure trove, in which structural data can be uniquely corre-
AreA) in the presence of the repressing nitrogen sources, most lated with functional studies both in vivo and in vitro.
likely mediated by glutamine. This negative regulator interacts The areA102 mutant grows exceptionally well on some amides
both with residues in the DBD and in the carboxy terminus of but not at all on uric acid. The mutation responsible is a leucine
AreA and NIT2, thereby preventing NIT2 and AreA binding to to valine change at position 7 after the second cysteine (labeled
DNA. In the case of AreA, two other levels of regulation are L22V in the structures) (Fig. 4). This leucine in the seventh posi-
present: first, transcription of areA is autoregulated and second, tion of the zinc finger loop is conserved in almost all “animal-like”
the stability of the areA mRNA is lower in mycelia grown on GATA factors (Fig. 1). Revertants with a valine to methionine
repressing nitrogen sources.22 change exhibit a “mirror image” mutant phenotype, thus they
In S. cerevisiae the basic pattern of regulation is the same but grow better on uric acid and worse on amides.25 Specific suppres-
in this case four GATA factors are involved, all of which are inter- sors map to the promoters of the uric acid permease genes. A
connected in a complex regulatory network: Gln3p and Gat1p number of these are mutations of CGATAG or AGATAG to
are transcriptional activators, while Dal80p and Gzf3p are re-
TGATAG. All sites in the promoter region that are involved in
pressors. The haploid genome of S. cerevisiae is most likely to be
utilization of amides are TGATAG sites. A detailed structural
derived from an ancestral diploid and thus there have been two
comparison of the interactions between the amino acids leucine
sets of genes able to diverge in the course of the evolution of the
and valine and their respective target bases for the wild type AreA/
organism. The two filamentous fungi A. nidulans and N. crassa
CGATG and the L22Vmutant/TGATAG complexes was carried
are rather ‘wild’ and represent probably the general situation in
out.26 The global fold and the overall binding mode was found to
the ascomycetes, whereas baker’s yeast is a very domesticated or-
be very similar for both DBDs (Fig. 4A). The positions for all
ganism, therefore exhibiting the more complex regulatory mecha-
sugars and bases are almost identical, with the exception of those
nism.
directly surrounding the C/T mutation. The most significant per-
turbations to the protein structure occur around the leucine to
“Vertebrate-Like” GATA Factors: Regulation valine mutation site. The protein backbone shifts slightly, exhib-
of Siderophore Synthesis iting the largest displacement relative to the wild type coordi-
The urbs-1 gene of Ustilago maydis is involved in siderophore nates. Also, the positioning of the valine side chain relative to the
biosynthesis. This gene encodes a GATA factor with two typical base is clearly altered and different χ1 angles are observed. As a
GATA-DBDs. Unlike in vertebrates where the DNA-binding result, the γ2 methyl group of valine occupies a very similar posi-
motifs of GATA factors are separated by only a few residues, the tion to that of the δ1 methyl of the wild type leucine. In the
two motifs of Urbs-1 are separated by 92 amino acid residues.23 structure of the wild type AreA DBD with the wild type AGATAG
As in the vertebrate factors, the DNA-binding finger is the DNA, the leucine side chain is involved in a hydrophobic con-
carboxy-terminal domain. Regulators with similar sequences and tact with the first base, whether that be A, C, or T. In the mutant
functions have since been found in N. crassa, A. nidulans and P. protein, the shorter valine side chain makes van der Waals con-
chrysogenum24 (but not in the genome of S. cerevisiae) implying a tacts with the bulky methyl of the T. Thus, the missing methyl
conservation of siderophore biosynthesis control. group of valine compared to leucine is compensated for by a
The DNA-Binding Domain of GATA Transcription Factors—A Prototypical Type IV Cys2-Cys2 Zinc Finger 29
methyl group now provided by the DNA (Fig. 4B). In another Acknowledgements
mutant, the longer methionine can interact well with an A or C, I am grateful to H. N. Arst Jr, C. Scazzocchio and members of
but clashes with a T. Therefore, the introduction of a T preceding the Felsenfeld laboratory for sharing results and/or helpful dis-
the GATA sequence effectively compensates for the shorter va- cussions.
line side chain in the mutant, restoring a closed-packed hydro-
phobic interface. Based on these structural results with AreA, the References
DBDs of C. elegans and STALKY that contain an alanine and a 1. Orkin SH. GATA-Binding transcription factors in hematopoietic
threonine in the equivalent position should only recognize cells. Blood 1992; 80:575-581.
TGATAR sequences. 2. Scazzocchio C. The fungal GATA factors. Curr Op Microbiol 2000;
3:126-131.
3. Merika M, Orkin SH. DNA-binding specificity of GATA family
The GATA DBD Is Capable of More Than transcription factors. Mol Cell Biol 1993; 13:3999-4010.
Just DNA Binding 4. Ko LJ, Engel JD. DNA-binding specificities of the GATA
Given the similarity of GATA DBDs, it is not easy to under- transcription factor family. Mol Cell Biol 1993; 13:4011-4022.
5. Yang HY, Todd E. Distinct roles for the two cGATA-1 finger
stand how promoter specificity is achieved. It has been estab-
domains. Mol Cell Biol 1992; 12:4562-4570.
lished that GATA domains of mammalian and fungal proteins 6. Fox AH, Kowalski K, King GF et al. Key residues characteristic of
are able to interact with other proteins, including other zinc fin- GATA N-fingers are recognized by FOG. J Biol Chem 1998;
ger containing ones.27 AreA/NIT2 can interact via its DBD with 273:33595-33603.
the NMR protein and the specific regulator NirA/NIT4. Whether 7. Ballario P, Vittorioso P, Magrelli A et al. White collar-1, a central
further interactions with other specific regulators occur and regulator of blue light responses in Neurospora, is a zinc finger
whether the determinants for specificity reside in the promoters protein. EMBO J 1996; 15:1650-1657.
8. Chang WT, Newell PC, Gross JD. Identification of the cell fate
has not been established yet and requires further study. gene stalky in Dictyostelium. Cell 1996; 87:471-481.
9. Teackle GR, Gilmartin PM. Two forms of type IV zinc-finger motif
GATA Factors and Chromatin Rearrangement and their kingdom-specific distribution between the flora, fauna and
fungi. Trends Biochem Sci 1998; 23:100-102.
GATA-1 is stringently required in the maintenance of DNase
10. Omichinski JG, Clore GM, Schaad O et al. NMR structure of a
I hypersensitive sites in the chicken- and human globin gene cluster specific DNA complex of Zn-containing DNA binding domain of
enhancers28,29 and it was demonstrated that GATA factors are GATA-1. Science 1993; 261:438-446.
capable of destabilizing nucleosomes in vitro.30 It therefore seemed 11. Starich MR, Wikström M, Arst Jr HN et al. The solution structure
reasonable to suspect involvement of GATA factors in nucleo- of a fungal AREA protein-DNA complex: An alternative binding
some maintenance. The nitrite and nitrate reductase bidirectional mode for the basic carboxyl tail of GATA factors. J Mol Biol 1998;
promoter of A. nidulans contains six positioned nucleosomes. In 277:605-620.
12. Gao X, Sedgwick T, Shi YB et al. Distinct functions are implicated
the presence of nitrate and the absence of ammonium, these nu- for the GATA-4,-5 and-6 transcription factors in the regulation of
cleosomes are no longer observed, independent of transcription intestine epithelial cell differentiation. Mol Cell Biol 1998;
or the presence of the NirA-specific factor. AreA is absolutely 18:2901-2911.
required in this positioning.20 Mutations in the AreA-binding 13. Zhu J, Hill RJ, Heid PJ et al. end-1 encodes an apparent GATA
domain result in constitutive chromatin rearrangement. Given factor that specifies the endoderm precursor in Caenorhabditis elegans
that GATA-1 itself is acetylated,31 it may well be that the verte- embryos. Genes Dev 1997; 11:2883-2896.
14. Petersen UM, Kadalayil L, Rehorn KP et al. Serpent regulates
brate GATA-1 DBDs influence directly or indirectly histone acety- Drosophila immunity genes in the larval fat body through an essential
lation. GATA motif. EMBO J 1999; 18:4013-4022.
15. Lossky M, Wensink PC. Regulation of Drosophila yolk protein genes
Conclusions by an ovary-specific GATA factor. Mol Cell Biol 1995;
15:6943-6952.
GATA factors are widespread eukaryotic regulators most likely 16. Haenlin M, Cubadda Y, Blondeau F et al. Transcriptional activity
evolved from an ancestral GATA protein containing only a single of Pannier is regulated negatively by heterodimerization of the GATA
type IV zinc finger. In vertebrates, evolution of GATA factors DNA-binding domain with a cofactor encoded by the u-shaped gene
involved gene duplication whereas additional modular evolution of Drosophila. Genes Dev 1997; 11:3096-3108.
in nonvertebrates has occurred. Vertebrate GATA factors stimu- 17. Rastogi R, Bate NJ, Sivasankar S et al. Footprinting of the spinach
late gene transcription in hematopoietic development while in nitrite reductase gene promoter reveals the preservation of nitrate
regulatory elements between fungi and higher plants. Plant Mol Biol
fungi they act either as transcription factors or repressors in a 1997; 34:465-476.
number of different processes, ranging from nitrogen source uti- 18. Arst Jr HN , Cove DJ. Nitrogen metabolite repression in Aspergilus
lization to mating-type switching. The DBDs recognize DNA in nidulans. Mol Gen Genet 1973; 126:111-141.
a rather special fashion: most of the specific interactions with the 19. Muro-Pastor MI, Strauss J, González R et al. The GATA factor AreA
bases in their respective target sites are hydrophobic. Mutational isessential for chromatin remodelling in an eucaryotic bidirectional
studies based on the powerful selection techniques possible in promoter. EMBO J 1999; 18:1584-1597.
20. Marzluf GA. Genetic regulation of nitrogen metabolism in the fungi.
fungi, coupled with high-resolution structural work, provided a
Microbiol Mol Biol Rev 1997; 61:17-32.
paradigm for the study of this type of side chain-base interac- 21. Xiao X, Fu Y-H, Marzluf GA. The negative-acting NMR regulatory
tions. In addition, it was possible to elucidate interactions with protein of Neurospora crassa binds to and inhibits the DNA-binding
other proteins and of the role of transcription factors in chroma- activity of the positive-acting nitrogen regulatory protein NIT2.
tin remodeling. Biochem 1995; 34:8861-8868.
30 Zinc Finger Proteins: From Atomic Contact to Cellular Function
22. Platt A, Langdon T, Arst Jr HN et al. Nitrogen metabolite signalling 27. Mackay JP, Crossley M. Zinc fingers are sticking together. Trends
involves the C-terminus and the GATA domain of the Aspergillus Biochem Sc 1998; 23:1-4.
transcription factor AREA and the 3' untranslated region of its 28. Stamatoyannopoulos JA, Goodwin A, Joyce T et al. NF-E2 and
mRNA. EMBO J 1996; 15:2791-2801. GATA binding motifs are required for the formation of Dnase I
23. Voisard C, Wang J, McEvoy JL et al. urbs1, a gene regulating hypersensitive site 4 of the human –globin locus control region.
siderophore biosynthesis in Ustilago maydis, encodes a protein similar EMBO J 1995; 14:106-116.
to the erythroid transcription factor GATA-1. Mol Cell Biol 1993; 29. Boyes J, Felsenfeld G. Tissue-specific factors additively increase the
13:7091-7100. probability of the all-or-none formation of a hypersensitive site.
24. Haas H, Zadra I, Stoffler G et al. The Aspergillus nidulans GATA EMBO J 1996; 15:2496-2507.
factor SREA is involved in regulation of siderophore biosynthesis 30. Boyes J, Omichinski J, Clark D et al. Perturbation of nucleosome
and control of iron uptake. J Biol Chem 1999; 274:4613-4619. structure by the erythroid transcription factor GATA-1. J Mol Biol
25. Kudla B, Caddick MX, Langdon T et al. The regulatory gene areA 1998; 279:529-544.
mediating nitrogen metabolite repression in Aspergillus, nidulans. 31. Boyes J, Byfield P, Nakatani Y et al. Regulation of activity of the
Mutations, affecting specific gene activation alter a loop residue of a transcription factor GATA-1 by acetylation. Nature 1998;
putative zinc finger. EMBO J 1990; 9:1355-1364. 396:594-598.
26. Starich MR, Wikström M, Schumacher S et al. The solution structure
of the Leu22Val mutant AREA DNA binding domain complexed
with a TGATAG core element defines a role for hydrophobic packing
in the determination of specificity. J Mol Biol 1998; 277:621-634.
CHAPTER 6
Abstract
activities identical to those of MutM. As MutY and MutT
T
he prokaryotic MutM protein is a trifunctional DNA base
excision repair enzyme that removes a wide range of homologues have already been found in many eukaryotes, this
oxidatively damaged bases, especially 8-oxoguanine, (N- finding demonstrates the generality of the GO system as a de-
glycosylase activity) and cleaves both the 3’- and 5’- fense against oxidative damage.
phosphodiester bonds of the resulting apurinic/apyrimidinic site
(AP lyase activity). This enzyme possesses a zinc finger motif Zinc in MutM
(-Cys-X2-Cys-X16-Cys-X2-Cys-) at the C terminus, which forms Atomic absorption spectrophotometric analysis has
a β-hairpin loop. The positively charged Arg247 and Arg253 on demonstrated that each MutM protein molecule possesses one
the β-hairpin loop interact with the phosphate groups to pinch zinc atom.7 Extended X-ray absorption fine structure spectra show
the backbone of the lesion-containing strand. This feature directly that the metal is coordinated to the sulfur atoms of four
establishes that the zinc finger motif is essential for binding of cysteine residues. The zinc atom can be displaced by other
MutM to DNA as well as its enzymatic activities. divalent metal cations. Cd2+, Cu2+and Hg2+ inhibit the activity
of MutM,8 although Co2+-substituted MutM is still active.9 The
Introduction importance of the zinc finger motif for the activity has been
In aerobic organisms, cellular DNA is frequently damaged by indicated by site-directed mutagenesis and chemical
activated oxygen species produced as a result of aerobic energy modification of the cysteine residues within the zinc finger motif
metabolism or oxidative stress. In particular, the 8-oxoguanine of the E. coli MutM protein.8 Mutant MutM proteins with a
(GO) lesion is one of the most stable products of oxidative DNA single substitution of each cysteine residue by glycine show
damage.1 GO can base pair with adenine as well as with cytosine, decreased capacity to bind DNA as well as decreased enzymatic
resulting in a G:C to T:A transversion.2 The GO repair system activity; they still have a zinc ion, but bind the ion with much
composed of the MutM protein, also called formamidopyrimidine lower affinity than the wild type. The double mutant (C244S,
DNA glycosylase (Fpg), accompanied by MutY and MutT, C247S) lacks zinc without significant alteration of the secondary
prevents this mutation (Fig. 1). 3 The MutM protein is a structure.10 The cysteine residues within the zinc finger motif of
trifunctional DNA base excision repair enzyme that removes a the MutM protein are the primary sites of NO interaction, which
wide range of oxidatively damaged bases (N-glycosylase activity) irreversibly inhibits its activity.11
and cleaves both the 3’- and 5’-phosphodiester bonds of the
resulting apurinic/apyrimidinic site (AP lyase activity) (Fig. 2).4 3D Structure of MutM
The MutM protein is highly conserved across a wide range of Precise coordination to a zinc ion in MutM proteins has been
organisms. Bacterial MutM proteins (Mr 30 kD) possess the revealed by studies of the three-dimensional structure of Thermus
invariant N-terminal sequence, PELPEV, two strictly conserved thermophilus MutM.12 The crystal structure was determined at
lysine residues, and a zinc finger motif (-Cys-X2-Cys-X16-Cys- 1.9 Å resolution with multiwavelength anomalous diffraction
X2-Cys-) at the C terminus (Fig. 3). As MutM protein functions phasing using the intrinsic zinc ion of the zinc finger. MutM is
as a monomer, this enzyme belongs to the group possessing a composed of two domains connected by a flexible hinge (Fig.
single zinc finger motif of the C2C2-type. The length of the 4A). The N-terminal domain is a two-layered β-sandwich
sixteen residues between two cysteines clusters is relatively short structure composed of nine antiparallel β strands. The C-termi-
in comparison with other zinc finger motifs. This region is rela- nal domain has four helices and two β strands that form a zinc
tively rich in basic residues. finger motif. Most of the conserved residues including the
Recently, MutM homologues have been identified in the catalytic residues are clustered in a large, electrostatically positive
genome database of higher eukaryotes, including man.5,6 These cleft between the domains. The two DNA-binding motifs, the
gene products are thought to constitute an alternative pathway helix-two turns-helix (H2TH) and zinc finger motifs, are also
for the repair of GO and other oxidative lesions, in addition to located in this region (Fig. 4C). The β-hairpin loop formed by
Ogg1, another eukaryotic 8-oxoguanine DNA glycosylase with the zinc finger motif is directed to the domain interface. The zinc
Zinc Finger Proteins: From Atomic Contact to Cellular Function, edited by Shiro Iuchi and Natalie Kuldell.
©2005 Landes Bioscience/Eurekah.com and Kluwer Academic/Plenum Publishers.
32 Zinc Finger Proteins: From Atomic Contact to Cellular Function
Figure 3. Aligned sequences of the zinc finger motifs from bacterial MutM proteins. The residue number of the T. thermophilus MutM sequence
is shown in the uppermost line. Conserved residues are marked with asterisks.
MutM: Single C2C2 Zinc Finger-DNA Interaction 33
Figure 4. Overall structure of the MutM-DNA complex. A) Ribbon diagram of the complex. The DNA duplex is drawn by wireframes. A zinc
atom is shown as a sphere. B) Wireframe diagram of the complex. DNA backbones are shown as thick wire-frames. A β-hairpin loop formed by
the zinc finger motif is also shown as a thick wire-frame. Side chains of cysteine residues are diplayed as sticks. The letters N and C represent the
N- and C-termini, respectively. C) Same as in A but viewed from a different perspective. The DNA interfaces besides the zinc finger motif are shown
by arrows. Images were produced using RasMol.15
suitable for structure determination. In the resolved structure of side of the residue neighboring GO (Fig. 5). Although Arg247 is
the MutM-DNA complex, the N- and C-terminal domains mainly not completely conserved among the MutM protein family, the
latch the backbone of the GO-containing DNA strand of the corresponding lysine residue of Bacillus stearothermophilus MutM
duplex (Fig. 4B). The DNA helix is sharply bent (~60°) at the similarly interacts with the backbone phosphate (Fig. 3).13 These
locus of the GO site, but holds an almost canonical B-form interactions demonstrate that the zinc finger motif plays a direct
conformation outside the lesion. role in the binding of MutM to substrate DNA.
The substrate GO residue is fully extruded from the helix and Eukaryotic MutM homologues have no zinc finger motif at
is inserted deeply into an active site in the cleft on the enzyme the C terminus, although they retain the same activities as MutM.
(Fig. 4A). The catalytically important residues, Pro1, Glu2, and Despite the lack of sequence homology in the C-terminal region,
Lys52, are located near the GO residue in the active site. The the eukaryotic MutM proteins possess a conserved arginine
space vacated by GO is filled with the side chains of Met70, Arg99, residue at the position corresponding to Arg253. This arginine
and Phe101 inserted from the minor groove side of the DNA. may play a role similar to that of Arg253. Although no zinc
These residues interact with the opposite and neighboring bases,
thus helping to stabilize the complex. The distorted conformation
of the DNA is also enforced with extensive interactions between
the protein and the backbone of the DNA. His67, Asn161,
Tyr231, Arg247, and Arg253 form hydrogen bonds with
nonbridging oxygens of phosphate moieties in the GO-containing
strand, whereas His81 and Arg100 do so in the complementary
strand.
finger motif has been demonstrated in other DNA glycosylases, a 8. O’Connor TR, Graves RJ, de Murcia G et al. Fpg protein of
β-hairpin loop resembling the zinc finger motif of MutM has Escherichia coli is a zinc finger protein whose cysteine residues have
been reported.14 a structural and/or functional role. J Biol Chem 1993; 268:9063-
9070.
9. Buchko GW, Hess NJ, Bandaru V et al. Spectroscopic studies of
References zinc(II)- and cobalt(II)-associated Escherichia coli
1. Dizdaroglu M. Formation of an 8-oxoguanine moiety in formamidopyrimidine-DNA glycosylase: extended X-ray absorption
deoxyribonucleic acid on γ-irradiation in aqueous solution. fine structure evidence for a metal-binding domain. Biochemistry
Biochemistry 1985; 24:4476-4481. 2000; 39:12441-12449.
2. Wood ML, Dizdaroglu M, Gajewski E et al. Mechanistic studies of 10. Tchou J, Michaels ML, Miller JH et al. Function of the zinc finger
ionizing radiation and oxidative mutagenesis: genetic effects of a in Escherichia coli Fpg protein. J Biol Chem 1993; 268:26738-
single 8-hydroxyguanine (7-hydro-8-oxoguanine) residue inserted at 26744.
a unique site in a viral genome. Biochemistry 1990; 290:7024-7032. 11. Wink DA, Laval J. The Fpg protein, a DNA repair enzyme, is
3. Michaels ML, Miller JH. The GO system protects organisms from inhibited by the biomediator nitric oxide in vitro and in vivo.
the mutagenic effect of the spontaneous lesion 8-hydroxyguanine Carcinogenesis 1994; 15:2125-2129.
(7,8-dihydro-8-oxoguanine). J Bacteriol 1992; 174:6321-6325. 12. Sugahara M, Mikawa T, Kumasaka T et al. Crystal structure of a
4. Bhagwat M, Gerlt JA. 3'- and 5'-strand cleavage reactions catalyzed repair enzyme of oxidatively damaged DNA, MutM (Fpg), from an
by the Fpg protein from Escherichia coli occur via successive beta- extreme thermophile, Thermus thermophilus HB8. EMBO J 2000;
and delta-elimination mechanisms, respectively. Biochemistry 1996; 19:3857-3869.
35:659-665. 13. Fromme JC, Verdine GL. Structural insights into lesion recognition
5. Hazra TK, Izumi T, Boldogh I et al. Identification and and repair by the bacterial 8-oxoguanine DNA glycosylase MutM.
characterization of a human DNA glycosylase for repair of modified Nat Struct Biol 2002; 9:544-552.
bases in oxidatively damaged DNA. Proc Natl Acad Sci USA 2002; 14. Scharer OD, Jiricny J. Recent progress in the biology, chemistry
99:3523-3528. and structural biology of DNA glycosylases. Bioessays 2001;
6. Morland I, Rolseth V, Luna L et al. Human DNA glycosylases of 23:270-281.
the bacterial Fpg/MutM superfamily: an alternative pathway for the 15. Sayle R, Milner-White EJ. RASMOL: biomolecular graphics for all.
repair of 8-oxoguanine and other oxidation products in DNA. Trends Biochem. Sci 1995; 20:374.
Nucleic Acids Res 2002; 30:4926-4936.
7. Boiteux S, O’Connor TR, Lederer F et al. Homogeneous Escherichia
coli FPG protein. A DNA glycosylase which excises imidazole ring-
opened purines and nicks DNA at apurinic/apyrimidinic sites. J Biol
Chem 1990; 265:3916-3922.
CHAPTER 7
Summary
I
-TevI is a site-specific, sequence-tolerant homing endonu- propagated. While a certain level of sequence specificity in both
clease encoded by the td intron of bacteriophage T4. The binding and cleavage is important to minimize random cleavage,
enzyme consists of an N-terminal catalytic domain and a precise cleavage is not essential, since the host’s recombination
C-terminal DNA-binding domain that are connected by a long, and replication functions will repair the cleavage event as part of
flexible linker. The crystal structure of the DNA-binding do- the homing process. In fact, the ability to cleave variant substrates
main of I-TevI, residues 130 to 245, complexed with the 20-bp is advantageous to both the endonuclease and the host intron.
primary binding region of its DNA target, reveals the presence GIY-YIG family members contain up to five conserved se-
of a zinc finger, comprising residues 151 to 167, that makes back- quence motifs that make up the GIY-YIG module with essen-
bone contacts with the DNA from the minor groove. Biochemi- tially no similarity among the proteins beyond that.3 In addition
cal data have shown that the zinc finger does not contribute to to the GIY-(9 to 10 residue)-YIG sequence, this module includes
the DNA-binding affinity or to the specificity of the enzyme, several other highly conserved residues, some of which have been
but rather that it has a novel function and acts as a distance shown to be critical for catalytic activity. These include Tyr1 Arg27,
determinant that controls the relative positions of the catalytic Glu75, and Asn90 (I-TevI sequence numbers).
and DNA-binding domains. I-TevI is encoded by a group I intron in the thymidylate syn-
thase (td) gene of bacteriophage T4, and is the best characterized
Introduction of the GIY-YIG enzymes and the first for which structural infor-
Group I introns are intervening sequences within the coding mation is available (Fig. 1). I-TevI specifically recognizes its 37-bp
sequence of genes that are excised by RNA splicing. They can DNA substrate, or homing site, as a monomer. Despite nanomolar
also be mobile genetic elements at the DNA level.1 If an affinity, the enzyme exhibits a high degree of sequence tolerance.4
intron-containing gene is in close proximity to an intron-less ver-
sion of the same gene, there will be a gene conversion event
whereby the intron-less copy of the gene becomes intron-plus.
This process is called intron homing because the intron “goes
home”. Group I intron homing is brought about by the action of
an endonuclease that is encoded by an ORF located in the RNA
intron. The endonuclease recognizes the exon:exon junction of
the intron-less allele—a contiguous sequence that is not present
in the intron-plus version because of the presence of the intron.
The endonuclease binds to the intron-less cognate allelle and in-
troduces a double-stranded break. Then, the intron-plus gene
aligns with the cleaved version, and recombination and repair
events lead to copying of the intron, resulting in two intron-plus
alleles. Figure 1. Overview of three-dimensional structural information cur-
Homing endonucleases are grouped into at least four differ- rently available for I-TevI. The enzyme consists of an N-terminal cata-
ent families that are defined on the basis of conserved sequence lytic domain joined to the C-terminal DNA-binding domain by a long,
elements as the GIY-YIG, LAGLIDADG, H-N-H and His-Cys flexible linker that remains to be structurally characterized. The
box families.1,2 They are site-specific DNA endonucleases that DNA-binding domain itself has an elongated structure and could only
recognize long DNA targets of 14-40 bp, with some degree of be crystallized in the presence of its DNA substrate. To view online color
sequence tolerance. The biological function of homing endonu- versions of images, go to http://www.eurekah.com/eurekahlogin.php?
cleases is to cleave intronless sites so that their host intron can be chapid=1734&bookid=124&catid=30.
Zinc Finger Proteins: From Atomic Contact to Cellular Function, edited by Shiro Iuchi and Natalie Kuldell.
©2005 Landes Bioscience/Eurekah.com and Kluwer Academic/Plenum Publishers.
36 Zinc Finger Proteins: From Atomic Contact to Cellular Function
No single nucleotide in the 37-bp target is essential for binding to 167); a minor groove-binding α-helix (residues 183 to 194);
and cleavage, and many multiple substitutions are well tolerated.5 and, a helix-turn-helix subdomain (residues 204 to 245). There
Consistent with this sequence tolerance, ethylation and methyla- are phosphate backbone contacts throughout the full length of
tion interference studies indicated that most of the protein-DNA the complex. The elongated segments, which are inserted in the
contacts are via the minor groove and the phosphate backbone.4 minor groove, provide the only hydrogen bonding interactions
The primary binding region of the enzyme is approximately 20 with bases. However, due to the poor information content of the
bp in length, spanning the intron insertion site (IS), with a sec- minor groove, these are of limited value with respect to substrate
ond region of contact close to the cleavage site (CS), which is specificity. In contrast, the helix-turn-helix subdomain, inserted
23-25 bp upstream of the IS. In addition, I-TevI can tolerate into the major groove, makes hydrophobic contacts with the
insertions or deletions between the CS and IS, and still effect C5-methyls of thymines and thus is selective for an AT-rich
cleavage.5 region.
I-TevI consists of two functionally distinct domains, an The biochemical and structural data present a picture of I-TevI
N-terminal catalytic domain and a C-terminal DNA-binding as a catalytic GIY-YIG domain, common to all family members,
domain, that are separated by a long, flexible linker.6 The cata- connected by a long linker of more than 50 amino acids to a
lytic domain contains the GIY-YIG module common to family DNA-binding domain that itself is composed of several
members and forms a discrete structural domain contained within sub-domains. Sequence analysis indicates that the linker and the
the first 92 amino acids of the protein.3,7 This domain has an α/ DNA-binding domain vary greatly among the different mem-
β-fold with a central three-stranded β-sheet surrounded by three bers of the family, with the exception of the C-terminal
α-helices. The GIY and YIG sequences are located on the first helix-turn-helix subdomain which appears to be more conserved,
and second β-strands, respectively, and are important structur- at least among a subset of family members. It is likely that the
ally as well as functionally. A surface composed primarily of the variations in the composition (and structure) of the linkers and
first and third α-helices and the second β-strand is thought to of the DNA-binding domains are important for the recognition
provide the main substrate interaction site. Three residues that of the large and highly variable substrates. Furthermore, the
are known to be important for catalytic activity, Tyr17, Arg27, GIY-YIG homing endonucleases appear to have different separa-
and Glu75, are exposed on this surface. Evidence of a tions between their primary recognition sites and their cleavage
metal-binding site between Tyr17 and Glu75, as well as similar- sites, requiring differences in the relative positioning of the cata-
ity in the arrangement of the three residues with that of the cata- lytic and DNA-binding domains, which can be achieved through
lytically important residues of the His-Cys box homing endonu- alternative linker structures and domain interactions.
clease I-PpoI8,9 (despite completely different protein folds) are
indicative of similarities in the DNA cleavage activities of the Structure of the Zinc Finger Subdomain
GIY-YIG and His-Cys box families.7 The determination of the crystal structure of the DNA-binding
The DNA-binding domain of I-TevI was initially defined to domain of I-TevI revealed the presence of a zinc finger.10 This
be contained between residues 130 and the C terminus of the zinc finger was not anticipated from sequence analysis,3 because
245-amino acid enzyme, based on footprinting and limited pro- it does not occur in any other GIY-YIG endonucleases and be-
teolysis experiments6 The crystal structure10 of this moiety, in cause of its unusual sequence, CXCX10CX2C. Sequence database
complex with a 20-bp duplex DNA that corresponds to the pri-
searches indicate the presence of Cys4-zinc fingers with similar
mary binding region of the enzyme, shows that residues 151 to
separations of the cysteines in the nuclear factor NHR-63 of
245 consist of an extended modular structure that interacts with
Caenorhabditis elegans11 and in the β’-subunits of bacterial RNA
the full length of the DNA duplex (Fig. 2). The molecule consists
polymerases. However, none of these zinc fingers have been struc-
of three distinct subdomains that are connected by elongated seg-
turally characterized. The structure of NHR-63 has not yet been
ments that lack secondary structure: a zinc finger (residues 151
reported while the zinc finger sequence lies just outside the re-
solved three-dimensional structure of the RNA polymerase of
Thermus aquaticus.12 The fact that I-TevI is the only GIY-YIG
protein characterized to date that contains a zinc finger suggests
that this element plays a specific role in the function of I-TevI,
beyond mere DNA recognition.
The I-TevI zinc finger (Fig. 3) consists of two type-I β-turns,
followed by a small loop and a single-turn α-helix. The first two
cysteines, 151 and 153, are the first and third residues in the first
β-turn, while the last two cysteines, 164 and 167, are the first
and last residues of the helix. The zinc finger of I-TevI has been
classified among the treble clef group,11 although it is the small-
est known member of the group and is quite degenerate. The
treble clef zinc finger typically consists of a β-hairpin, a loop, and
an α-helix. The first two cysteines are contributed by the zinc
knuckle, a unique turn with a CPXCG consensus sequence that
Figure 2. Structure of the DNA-binding domain of I-TevI. The joins the two β-strands. In I-TevI, the corresponding sequence,
DNA-binding domain has an extended fold that consists of three CKCG, forms a tighter β-turn, bringing the two short (three resi-
subdomains: a zinc finger (Zn), an α-helix inserted into the minor dues each) β-strands closer together. Krishna and coworkers11
groove (α) and a helix-turn-helix (H-T-H). Segments that lack second- further suggested that, due to the shortness of the α-helix, the
ary structure (orange) connect the three subdomains but are important I-TevI zinc finger could also be classified as a zinc ribbon, albeit
for DNA-binding as they are the only sites of direct hydrogen bonding one that is lacking a β-strand. Interestingly, several Cys4-zinc fin-
contacts with the bases. gers of the treble clef and zinc ribbon groups are known to be
Homing Endonuclease I-TevI: An Atypical Zinc Finger With a Novel Function 37
Acknowledgements
This research was supported by NIH grants GM56966 (PVR),
GM39422 and GM44844 (MB).
References
1. Belfort M, Roberts RJ. Homing endonucleases: Keeping the house
in order. Nucl Acids Res 1997; 25:3379-3388.
2. Chevalier BS, Stoddard BL. Homing endonucleases: Structural and
functional insight into the catalysis of intron/intein mobility. Nucl
Acids Res 2001; 29:3757-3774.
3. Kowalski JC, Belfort M, Stapleton MA et al. Configuration of the
catalytic GIY-YIG domain of intron endonuclease I-TevI:
Coincidence of computational and molecular findings. Nucl Acids
Res 1999; 27:2115-2125.
4. Bryk M, Quirk SM, Mueller JE et al. The td intron endonuclease
I-TevI makes extensive sequence-tolerant contacts across the minor
groove of its DNA target. EMBO J 1993; 12:2141-2149.
5. Bryk M, Belisle M, Mueller JE et al. Selection of a remote cleavage
site by I-TevI, the td intron-encoded endonuclease. J Mol Biol 1995;
247:197-201.
6. Derbyshire V, Kowalski JC, Dansereau JT et al. Two-domain structure
of the td intron-encoded endonuclease I-TevI correlates with the
Figure 5. The I-TevI zinc finger acts as a distance determinant. The two two-domain configuration of the homing site. J Mol Biol 1997;
different scenarios for cleavage preference of wild-type I-TevI at dis- 265:494-506.
tance (top), and the zinc-finger mutant at sequence (bottom), are pre- 7. Van Roey P, Meehan L, Kowalski JC et al. Catalytic domain structure
and hypothesis for function of GIY-YIG intron endonuclease I-TevI.
sented. Scissors represent site of preferred cleavage as described in the
Nature Struct Biol 2002; 9:806-811.
text. Linker organizer (A) and catalytic clamp (B) models for distance 8. Galburt EA, Chevalier B, Tang W et al. A novel endonuclease
determination are designated on the wild-type protein, in the top panel. mechanism directly visualized for I-PpoI. Nature Structural Biology
1999; 6:1096-1099.
site, even if that is displaced far from the IS. It seems that loss of 9. Galburt EA, Chadsey MS, Jurica MS et al. Conformational changes
the zinc finger correlates with loss of a distance determinant for and cleavage by the homing endonuclease I-PpoI: A critical role for
a leucine residue in the active site. J Mol Biol 2000; 300:877-887.
cleavage. There are currently two models for how the zinc finger 10. Van Roey P, Waddling CA, Fox KM et al. Intertwined structure of
might act as a distance determinant that both involve putative the DNA-binding domain of intron endonuclease I-TevI with its
protein-protein interactions between the zinc finger and other substrate. EMBO J 2001; 20:3631-3637.
parts of the enzyme. In one model, the zinc finger acts as a “linker 11. Krishna SS, Majumdar I, Grishin NV. Structural classification of
organizer”, interacting with other parts of the lengthy linker re- zinc fingers. Nucl Acids Res 2003; 31:532-550.
gion to reduce the potential for the catalytic domain to reach out 12. Campbell EA, Korzheva N, Mustaev A, et al. Structural mechanism
for rifamycin inhibition of bacterial DNA polymerase. Cell 2001;
to seek the natural cleavage sequence, and thus making it more 104:901-912.
likely to cleave at distance (“A” in top panel of Fig. 5). In the 13. Dean AB, Stanger M, Dansereau J et al. The zinc finger as a distance
second model, the zinc finger is postulated to interact directly determinant in the flexible linker of intron endonuclease I-TevI. Proc
with the catalytic domain as a “catalytic clamp”, again constrain- Natl Acad Sci USA 2002; 99:8554-8561.
CHAPTER 8
Abstract
posed by Brown.4 Klug termed this modular domain a “zinc fin-
Z
inc fingers encompass a wide variety of compact protein
domains that are stabilized by a structural zinc ion which ger” and proposed that each zinc-containing DNA binding mod-
minimally interacts with a cysteine-rich coordination ule coordinated zinc via chelation with two cysteine and two his-
sphere. The selectivity for zinc ion binding is governed by coor- tidine residues.3 This cysteine-histidine coordination sphere for
dinating amino acid side chains and by thermodynamic param- zinc in TFIIIA was later confirmed by EXAFS technology.5 In
eters. Since metal coordination spheres in zinc finger proteins 1989 it was demonstrated that the zinc in TFIIIA was respon-
are susceptible to chemical attack (principally at thiolates) and sible for holding the protein in the proper conformation for DNA
because zinc finger proteins have prominent roles in many cellu- binding as the removal of the metal resulted in an unfolding of
lar processes including the regulation of gene expression and sig- the structure and loss of DNA binding specificity.6
nal transduction, an underlying mechanism for a number of cel- In the original description of the zinc finger model for TFIIIA
lular dysfunctions is likely to be the disruption of zinc coordina- it was suggested this modular cysteine and histidine-rich struc-
tion spheres by a variety of metals and other small molecules. ture folded around a zinc ion may be present in a wide variety of
For instance, a number of toxicity mechanisms are likely to be regulatory proteins as its compactness seemed an ideal structure
the consequence of zinc replacement by xenobiotic metals re- for binding nucleic acids and for evolutionary duplication.3 Berg
sulting in changes in polypeptide conformation and the con- identified this modular structure in a wide variety of bacterial,
comitant loss of protein function. Zinc finger disruption could viral, and eukaryotic protein sequences.7 The term “zinc finger”
also occur by oxidation and modification of critical cysteine and is now commonly used to identify a wide variety of compact pro-
histidine amino acids in the zinc coordination sphere resulting tein domains stabilized by a structural zinc ion.8-11 Zinc finger
in zinc release and alteration of conformation. The chemical re- proteins encompass the largest class of proteins found in eukary-
activity of metal coordination spheres of zinc finger proteins are otic cells and have diverse roles in most cell processes including
utilized in normal physiological processes by providing regula- transcription and translation, DNA replication and repair, me-
tory sites for signal transduction via small molecules like nitric tabolism, signaling, cell proliferation, and apoptosis.11 In most
oxide and oxygen and their reactive intermediates. In addition, cases zinc ions are likely coordinated by both cysteine and histi-
zinc finger proteins and their metal binding sites are promising dine residues but some major families like nuclear receptors and
targets for specific drug design to help ameliorate major diseases. GATA family members have only cysteine coordination. Zinc fin-
ger proteins are classified by the arrangement of the zinc coordi-
Introduction nating amino acids (cysteines and histidines) and the secondary
The discovery that zinc has a structural role in eukaryotic gene structures (fold groups) in which these amino acids are present.
regulatory proteins (as opposed to a functional role in enzyme About 8 fold groups are currently specified.11 One of the more
catalysis) originated from studies which demonstrated the common fingers and fold groups includes the C 2H 2 fold
cysteine-rich transcription factor IIIA (TFIIIA) contained zinc (ββ-hairpin followed by an α-helix) characteristic of TFIIIA, the
and required the metal for specific binding to DNA.1 That TFIIIA prototypical zinc finger protein. This fold contains about 30 amino
had zinc-dependent DNA binding domains was a novel observa- acids. When the zinc coordinating cysteines are present in the
tion at the time as most DNA binding proteins were believed to sequence CPXCG, this structure is referred to as a “knuckle”.
have helix-turn-helix motifs for DNA interaction similar to the This zinc knuckle structure is found in many zinc finger protein
bacterial repressor proteins. The elucidation of the amino acid folds including the retroviral Gag (nucleocapsid) zinc finger fold.12
sequence of TFIIIA2 led Klug and colleagues to propose that re- The treble clef finger fold which consists of a zinc knuckle and a
petitive cysteine and histidine residues present in the sequence shortened α-helix is found in a wide variety of zinc finger pro-
could form modular domains of about 30 amino acids each, folded teins including RING fingers and the nuclear receptors.11 Zinc
around a zinc atom.3 A similar structure for TFIIIA was also pro- ribbon folds which may represent the most diverse group of zinc
Zinc Finger Proteins: From Atomic Contact to Cellular Function, edited by Shiro Iuchi and Natalie Kuldell.
©2005 Landes Bioscience/Eurekah.com and Kluwer Academic/Plenum Publishers.
40 Zinc Finger Proteins: From Atomic Contact to Cellular Function
Figure 1. Zinc interaction induces the proper DNA binding conformation of the C2H2 finger polypeptide. By coordinating with the two cysteine
side chains and the two histidine side chains, the zinc ion induces formation of the ββ turn structure in the N-terminal half of the peptide and
the α helical structure of the C-terminal half of the peptide. Six water molecules are released in the peptide coordination reaction. Specific amino
acids in the α helical structure form specific contacts with the DNA bases.
finger proteins are characterized by zinc coordination between with the two coordinating cysteines followed by interaction with
two zinc knuckles separated by about 40 amino acids. 13 the histidines.15,18,19
Cysteine-rich zinc clusters as well as zinc loop folds are character- The anti-parallel ββ sheet plays a role in making nonspecific
ized by the yeast Gal4 regulator and metallothioneins.8 In gen- contacts on the phosphodiester backbone; a conserved basic resi-
eral, zinc finger folds or domains function as interaction modules due located two amino acids from the second coordinating cys-
that bind to many large molecules including nucleic acids and teine in the second β strand makes a phosphodiester oxygen con-
other proteins and to a variety of small molecules including met- tact as does the protonated nitrogen on the first coordinating
als, redox agents, and therapeutic drugs. A number of X-ray and histidine.17 NMR analysis using carboxymethylation of cysteines
NMR structures are available for zinc fingers complexed with a and protonation of imidazole nitrogens indicates zinc cation bind-
variety of nucleic acid templates.11 ing is blocked under these conditions. 20 In addition,
carboxymethylation of cysteines in the absence of imidazole pro-
Regulation of Zinc Finger Conformation by tonation did not result in proper α helical formation, again in-
dicative of the importance of cysteine coordination in the overall
Zinc Ion Binding finger folding process.20 Loss of either cysteine still results in α
Studies with TFIIIA correlated the loss of zinc with a signifi- helical formation but the helix is a looser and less extensive struc-
cant change in protein conformation as well as with a concomi- ture; loss of first histidine results in no helical structure whereas
tant loss of specific DNA binding ability.1,6 These observations loss of the second histidine results in a wild-type CD spectrum.15
indicated that the role of zinc in TFIIIA was to hold the protein It is suggested that the role of zinc is to bring the α helix and β
in the proper conformation allowing specific amino acid side sheet closer together to form a hydrophobic core which is essen-
chains to make specific contacts on the DNA binding site. Much tial for finger stability and that the first coordinating histidine is
has been learned about the role of zinc in maintaining zinc finger essential for hydrophobic core formation.15 In vitro mutagenesis
conformation from studying individual zinc fingers of about 30 studies on TFIIIA demonstrated the importance of the first coor-
amino acids. However, the specific DNA binding ability by these dinating histidine residue for protein-DNA interactions.21 Zinc
individual fingers in the presence of zinc has been difficult to binding to multiple finger domains appears to be more complex
demonstrate. In the absence of zinc, the finger polypeptide is in a than binding to single finger domains. For example, zinc binding
random coil conformation as evidenced by circular dichroism to the C2C2 DNA binding domain of the vitamin D receptor
(CD).14,15 Added zinc induces a prominent α-helix in the finger which contains two such fingers results in a minimal conforma-
structure as evidenced by the two negative CD Cotton peaks at tion change to the first finger and a much more significant con-
208 and 222 nm.15 This α-helix is a prominent structural feature formational change to the second finger as assayed by CD and
in the TFIIIA-type zinc finger (C2H2) as evidenced by NMR and electrospray ionization mass spectrometry.22
X-ray crystallography.16,17 The two zinc coordinating histidine
residues located in this helix are three amino acids apart, except
in the rare cases when they are four residues apart. This spacing is Zinc Coordination Structures in Zinc Finger
critical because it places these amino acids on the same face of the Proteins
helix in an ideal position for zinc coordination. This helical struc- For a number of reasons, zinc is an ideal cation for structural
ture allows specific noncoordinating amino acids located in in- or catalytic roles in proteins. The zinc ion (Zn2+) contains a filled
variant positions to be involved in formation of hydrogen bonds d orbital (d10) and therefore the metal cation does not participate
with specific nucleotide bases in the major groove of DNA.17 in redox reactions. Rather it functions as a Lewis acid and will
Another prominent structure in the C2H2 zinc finger is the accept pairs of electrons.23 This lack of redox behavior makes the
N-terminal anti-parallel ββ loop structure which is induced by zinc ion very stable and an ideal metal cofactor for enzymatic
zinc interaction with the two cysteine residues that are usually reactions that require a stable ion to function as a Lewis acid in
2-4 amino acids apart (Fig. 1). The ββ structure is essential for catalytic mechanisms like proteolysis (carboxypeptidases) and
overall zinc finger stability. There is some evidence that when carbon dioxide hydration (carbonic anhydrases).24 As a structural
zinc binds to the unfolded apo-finger the metal ion first interacts cofactor in zinc finger proteins, which to a large degree interact
Zinc Finger Interactions with Metals and Other Small Molecules 41
affinity.39 The β domain appears to release one zinc atom much ligands (for example, carboxylates). The “hardness” has to do with
more readily than the α domain.39 It is generally believed that a lack of polarizability whereas “softness” refers to ease of polariz-
the free zinc ion concentration in cells is very low and hence zinc ability. Soft acids have greater numbers of d electrons than hard
metalloproteins do not acquire the metal from a pool of free metal acids and d-rich metals such as cadmium, lead, and mercury most
ions.40 Therefore, zinc uptake and transfer in the cell is possibly readily interact with soft ligands like sulfhydryls. The hard-soft
maintained by metal trafficking proteins like metallothioneins classification is not a sharp divide but rather borderline metals
and ZIP proteins in higher organisms.40,41 Redox reactions which and ligands exist between these two extremes. Zinc ions are in
have the potential to release zinc from a donor protein as well as this borderline classification and interact with nitrogen-based
direct transfer of zinc from a coordination sphere of the donor to ligands like imidazoles but will also interact with sulfhydryls. The
a coordination sphere of the receptor protein are believed to be Irving-Williams series of metals refers to the stabilities of chela-
some of the mechanisms by which zinc metalloproteins includ- tion complexes where chelation is defined as the coordination of
ing zinc finger proteins are populated with zinc in the cell. These two or more donor atoms from a single ligand to a central metal
observations suggest cells and metalloproteins have evolved com- ion.25 In general, for any oxygen, nitrogen, or sulfur chelate, the
plex mechanisms for zinc binding and zinc release, which are stabilities of the metal-chelate complex will increase with decreas-
undoubtedly under regulatory control. The elucidation of these ing cation diameter. Therefore large diameter calcium and
regulatory mechanisms will provide valuable information for magnisium cation chelates are less stable than smaller diameter
understanding how cells function in normal and disease environ- d-rich metals like cobalt, nickel, copper, and zinc.
ments. A third stabilization factor helping to explain the zinc prefer-
ence for the cysteine and histidine-rich metal coordination spheres
Zinc Finger Selectivity for Metal Ion Binding characteristic of zinc finger proteins is ligand field stabilization
The initial study demonstrating zinc dependency for the pro- energy (LFSE) which arises because of incompletely filled d or-
totypical zinc finger protein, TFIIIA, also demonstrated the zinc bitals in the transition metal series.26 Because zinc has a com-
specificity for this protein.1 Brief EDTA treatment was used to pletely filled d10 electronic structure, it does not lose any LFSE
abolish specific DNA binding by native TFIIIA and only zinc when it goes from an octahedral coordination sphere with water
from the transition metal series at micromolar concentrations that molecules in solution to a tetrahedral coordination sphere with
included iron (Fe2+), cobalt (Co2+), nickel (Ni2+), manganese sulfur and/or nitrogen ligands in proteins. The other transition
(Mn2+), and copper (Cu2+) was able to restore DNA binding abil- metals (for example, nickel, cobalt, iron) experience a loss of LFSE
ity.1 This observation is significant as it indicates that the C2H2 when going from the octahedral hexaaqua form to the protein
zinc coordination sphere exhibits exquisite specificity for zinc ions. chelated form.42 This loss of energy contributes to the lower af-
This is especially important with respect to the redox active iron finity binding observed for these transition metals when binding
and cooper ions which if presented close to DNA in a finger struc- zinc coordination spheres in proteins. The extra stability of the
ture could result in oxidative damage to DNA. Actual affinity zinc-ligand interaction is also driven by the increased entropy
measurements of zinc and nickel for TFIIIA indicated that two resulting from water release from the zinc hydration shell when
classes of metal binding sites exists on TFIIIA and that zinc binds the water molecules are replaced by the stronger metal-ligand
with higher affinity; zinc ions bind with dissociation constants of bonds (Fig. 1). In the case of zinc (Zn2+), the metal is in the
1.0 x10-8 M and 2.3 x10-5 M whereas nickel ions bind with dis- hexaaqua form and therefore six water molecules are available for
sociation constants of 2.6 x 10-5 M and 5.2 x 10-4 M.38 Because release and entropic increase is necessary to substantially help offset
of their ease of use and interpretation, many more metal binding the entropy loss upon zinc-dependent peptide folding. This tight
studies are being performed with synthetic zinc finger polypep- binding by zinc to the finger structure is necessary for switching
tides than with intact zinc finger-containing proteins.42 Using the apofinger into the correct conformation for specific DNA
synthetic polypeptides, Berg and colleagues have demonstrated binding.
that individual C2H2 zinc finger peptides readily bound zinc ions
but do not readily bind iron, cobalt, nickel, or manganese ions.43 Alterations in Zinc Finger Structure and
The dissociation constants (Kd) for binding of this transition se- Function by Xenobiotic Metals
ries of metals to a prototypical consensus C2H2 sequence was 2 x Most disease states ultimately result from alterations in gene
10-12 M for Zn2+, 5.0 x 10–8 M for Co2+, 1.6 x 10-6 M for Ni2+, expression. Zinc finger proteins have prominent roles in gene
2.5 x 10-6 M for Fe2+, and >10-5 M for Mn2+ (also see Table 1).43,44 expression mechanisms in mammalian cells and up to 1% of
At this time it is not clear why the affinities of the various transi- mammalian genomes code for zinc finger proteins.45 Since metal
tion metals are significantly lower for the intact TFIIIA protein coordination spheres in zinc finger proteins are susceptible to
than the individual synthetic C2H2 zinc fingers but it is possible chemical attack (principally at thiolates) and because zinc finger
the intact protein has a three-dimensional structure that limits proteins have prominent roles in the regulation of gene expres-
access of metals to the coordination spheres. sion, it was reasoned that zinc finger proteins may represent tar-
There are relatively straight forward explanations as to why gets for a variety of toxic agents that could lead to different dis-
some of the more prevalent metal ions in cells (Mg, Ca, Na, and ease states including cancer.46 Such toxic mechanisms could be
K) do not interact with the coordination spheres of cysteine and the consequence of zinc replacement by other metals resulting in
histidine-rich zinc finger proteins. These have to do with the changes in conformation or result from chemical oxidation/modi-
“hard-soft acid-base” concepts as well as the Irving-Williams ef- fication of critical cysteine and histidine amino acids in the zinc
fect for chelation stabilities.26 In general “hard” acids like sodium, coordination sphere. The elucidation of these toxic mechanisms
potassium, magnesium, and calcium (metals are considered acids is important since toxic and carcinogenic metal compounds are
in that they are positive and will accept a negative species, that is known to interfere with DNA metabolic processes like transcrip-
a base) interact with “hard” bases or ligands like oxygen-rich tion and repair at low concentrations.33,47-51
Zinc Finger Interactions with Metals and Other Small Molecules 43
Xenobiotic metals that disrupt zinc finger structure and func- coordinate histidine residues. These structural differences could
tion include cadmium, lead, mercury, gold, and arsenic. Cad- account for the misfolding of zinc fingers when coordinated by
mium ions (Cd2+) have a known specificity for cysteine residues lead ions. Therefore, disruption of zinc finger domains may be a
and are an obvious toxic metal to examine for inhibitory effects common target mechanism that underlies many of the deleteri-
on cysteine-rich zinc finger proteins. Cadmium and arsenite ions, ous effects of lead.60
which also have specificity for reacting with thiol groups, were Other xenobiotic metals that were shown to alter the DNA
shown to block steroid hormone binding as well as DNA bind- binding functions of zinc finger proteins include mercuric and
ing by the glucocorticoid receptor.52 This protein is a member of gold salts. Micromolar concentrations of mercury (Hg2+) were
the nuclear receptor superfamily and is characterized by two C2C2 found to inhibit specific DNA binding by TFIIIA and Sp1 but,
zinc fingers. Micromolar concentrations of Cd2+ but not arsenite in the case of TFIIIA, possibly at additional sites besides the pro-
were shown to inhibit TFIIIA binding to the 5S RNA gene and tein zinc fingers.61 This metal was also shown to inhibit tran-
the target of action was determined to be the zinc finger structure scription factor AP2 and restriction enzymes neither of which
of the protein.47 Cadmium ions did not exhibit any affinity for a have zinc finger sites. Gold salts are known to have
synthetic peptide corresponding to zinc finger 3 of TFIIIA but anti-inflammatory properties and exhibit high affinity for pro-
exhibited high affinity for the third finger from activator protein tein thiols. A thiomalate derivative of gold (Au1+) was found to
Sp1 as well as a consensus C2H2 zinc finger.53 The tumor sup- inhibit the C2C2 zinc finger-dependent DNA binding of the
pressor protein p53, which is mutated in a large proportion of progesterone receptor in vitro and in vivo.62 Thiomalate alone
human carcinomas, requires zinc for specific DNA binding and exhibited no such effects. Selenite, an oxyanion of the
has a CHC2 metal binding domain.54,55 Significantly, micromo- nonxenobiotic selenium, interacted with TFIIIA and Sp1 zinc
lar levels of cadmium ions (levels found in the blood of heavy fingers in rapid fashion and inhibited their DNA binding prop-
smokers) were shown to inhibit specific DNA binding by tumor erties.63 This inhibition was zinc finger specific as the AP2 tran-
suppressor p53 in vitro and in vivo.50 Through the use of scription factor was not affected. Selenium compounds are pro-
conformation-specific monoclonal antibodies against p53, it was posed to be involved in redox coupling between thiol and
demonstrated that micromolar amounts of cadmium caused a zinc-bound states of zinc metalloproteins in cells.64 This coupling
conformational change in the protein in vivo which correlated may contribute to the anti-cancer and anti-inflammatory prop-
with loss of specific DNA binding ability as well as loss of erties of selenium.
transactivation of target genes and loss of p53-dependent G1 ar-
rest in the cell cycle.50 These results are significant since it is known Interactions of Zinc Fingers with Signaling
that cadmium is carcinogenic and that normal p53 function in
cells is a major impediment to tumor development. NMR results Molecules and Therapeutic Drugs
indicate that cadmium ions when bound to C2H2 zinc fingers in Zinc finger proteins play such prominent roles in gene expres-
the human metalloregulatory transcription factor MTF-1, induce sion and signal transduction that it is likely evolution has taken
an altered conformation in the zinc fingers that can be restored to advantage of their chemical properties and evolved mechanisms
the ββα structure by added zinc ions.56 DNA repair enzymes for tuning these regulatory processes via interactions with pro-
like the XPA protein (xeroderma pigmentosum A which contains teins, small molecules, and other possible effectors. In addition,
a single C4 zinc finger), poly (ADP-ribose) polymerase (which zinc finger proteins remain promising targets for drug design to
contains two C2HC zinc fingers), are both inhibited in their DNA help ameliorate major diseases. Protein interactions with zinc fin-
recognition mechanisms by cadmium and arsenite ions.51 Since gers are best exemplified with members of the nuclear receptor
DNA repair is so vital for maintaining a normal cell phenotype, superfamily (steroid, glucocorticoid, retinoic acid, vitamin D, etc.).
these results suggest additional mechanisms by which cadmium These receptors all have two C2C2 zinc fingers. The first finger
and arsenic may be exerting their carcinogenic activities. region is responsible for specific DNA contacts and the second
Lead is the most prevalent toxic metal in our environment, finger is responsible for protein-protein interactions forming
and its propensity for interaction with protein sulfhydryl groups homo- and heterodimer contacts with the second finger of the
makes it likely to interact with zinc coordination spheres in zinc various classes of nuclear receptors. Such dimerization enables
finger proteins. Using gel shift assays and cell extracts, lead ions these transcription factors to “crosstalk” with a wide variety of
(Pb2+) were initially shown to alter the specific DNA binding receptor family members and greatly increases the regulatory ca-
ability of Sp1 in vitro and in vivo and were thus proposed to alter pacity of this important family of regulatory proteins.
Sp1-dependent transcription in vivo.57 Lead ions at micromolar Further discussion will focus on zinc finger interactions with
concentrations were shown to inhibit the purified zinc finger small molecules like nitric oxide and oxygen and their reactive
domains of TFIIIA as well as Sp1 as assayed by DNase I protec- intermediates, and with drugs of therapeutic interest. These in-
tion.58 The inhibition kinetics were rapid which suggested a di- teractions are based on the unique coordination chemistry of the
rect involvement of lead with the DNA binding domains of these metal binding spheres of zinc finger proteins. Nitric oxide (NO)
proteins. Thermodynamic studies reveal that lead ions bind zinc is a major cellular mediator of signal transduction. NO release
finger domains with dissociation constants ranging from 10-9 to can S-nitrosate sulfhydryl groups under physiological conditions
10-14 M depending upon the number of cysteine residues present and cysteine-rich zinc finger domains were hypothesized to be
in the domain.59 In spite of this tight binding, circular dichroism molecular targets for this molecule and its derivatives.65 Under
and NMR studies indicate that the zinc finger domains are aerobic conditions NO was demonstrated to release zinc from
misfolded when bound to lead which likely explains the inhibi- metallothionein by S-nitrosothiol modification of the
tory nature of the lead-zinc finger interaction.58,59 The lead ion zinc-chelating cysteine residues in the protein followed by disul-
has a much larger radius than the zinc ion and prefers fide bond formation.66 In addition, NO was demonstrated to
six-coordinate ligation spheres rather than the four-coordinate release zinc from metallothionein in endothelial cells.67 Incubation
spheres preferred by zinc. In addition, lead ions do not appear to of lymphocyte nuclear extracts with the NO donor molecule
44 Zinc Finger Proteins: From Atomic Contact to Cellular Function
S-nitrosocysteine (SNOC) resulted in the inhibition of the spe- correct protein conformation are very redox active. The cell is in
cific DNA binding ability of the RNA polymerase II activator a constant balancing act with respect to enabling these centers to
zinc finger protein Sp1 which activates the IL-2 gene in these bind zinc as well as restoring that binding once oxidation and
cells.68 This inhibition of Sp1 DNA binding paralleled the loss of zinc release has occurred.
IL-2 gene transcription in activated lymphocytes. If lymphocytes The reversibility and reactivity of the zinc binding and zinc
were preactivated before NO treatment, no loss of Sp1 DNA bind- coordination spheres in zinc finger proteins lends itself to the
ing was observed as well as no loss of IL-2 gene transcription.68 design of specific inhibitors that will oxidize these sites and/or
These results collectively indicate that Sp1 once bound to the eject zinc resulting in protein conformational changes and loss of
activator site on the IL-2 gene is resistant to NO-mediated dis- protein function. The specificity of such agents can come from
ruption of the Sp1 zinc fingers. Sp1 also acts as a repressor of drug interactions with nonconserved amino acids in close prox-
TNF-α gene expression and in this case NO was found to induce imity to the zinc coordinating center as well as from secondary
TNF-α by disrupting the repressor binding to the promoter of and tertiary protein structure determinants. One human viral
this gene.69 These results provide sound evidence zinc finger tran- protein targeted for selective inactivation of a crucial zinc finger
scription factors are physiological targets for NO-derived is the E6 oncoprotein from the human papillomavirus (HPV)
nitrosative stress in living cells. which is the causative agent for cervical cancer. This protein has
Oxidation-reduction reactions at critical protein thiol groups two C2C2 zinc fingers which are unique to this virus and have a
play a major role in cellular responses to oxidative stress through role in forming a ternary complex with the viral E6-associated
alterations in gene expression.70 Zinc finger proteins are prime protein (E6AP) and the cellular p53 tumor suppressor. This com-
targets for redox regulation since they contain cysteine-rich zinc plex results in the degradation of p53 via the ubiquitination-
coordination spheres. Evidence is presented that oxidative stress proteosome pathway. The uniqueness of these zinc fingers to the
inhibits DNA binding by Sp1 in vitro and inhibits the in vivo HPV virus makes them attractive candidates for therapeutic in-
transcription of the Sp1-dependent gene, dihydrofolate reduc- tervention. A large number of organic disulfide, azoic, and nitroso
tase (DHFR). This gene is a glycolytic housekeeping gene and is aromatic compounds with known sulfhydryl redox activity and
the type of gene believed to be initially shut down by a cell when zinc ejecting properties were screened for inhibitory activities
encountering oxidative stress.70 The tumor suppressor p53 con- against the human papillomavirus (HPV) oncoprotein E6, its
tains a single CHC2 zinc finger as well as four additional con-
zinc-dependent interaction with E6AP, and for growth inhibi-
served cysteines which could serve as a secondary metal binding
tion of HPV transformed cells.76 The organic disulfide 4,4'
site.71 Reactive oxygen intermediates (ROI) have a number of
dithiodimorpholine was found to eject zinc from the E6
roles in the p53 anti-proliferative pathway including inducing
oncoprotein, inhibit its interaction with E6AP, and also activate
DNA breaks, regulating DNA binding through oxidation of criti-
p53 in an HPV transformed cell and induce apoptosis.76 This
cal cysteines, and controlling several redox effector genes. With
respect to the regulation of p53 DNA binding, oxidation of agent is believed to form disulfides with the zinc–coordinating
zinc-coordinating cysteines leads to loss of DNA binding.72 In cysteine residues thus ejecting zinc and placing the protein in an
addition, thioredoxin which reduces protein thiols, activates p53 inactive conformation. Another zinc finger protein being targeted
function in vivo which indicates that redox activation of p53 is for therapeutic intervention is the nucleocapsid 7 protein of the
physiologically relevant.73 One of the highly conserved cysteines human immunodeficiency virus (HIV). Specific inactivation of
(Cys277) in p53 that does not have a direct role in zinc binding is the C2HC zinc finger of this protein was accomplished using
believed to have a direct role in DNA binding.74 Oxidation of specific disulfide compounds as well as other organics like
this cysteine results in a protein that is unable to recognize a p53 azodicarbonamide which are capable of undergoing redox reac-
binding site containing cytosines in the 3' end of the DNA rec- tions with zinc coordinating thiols resulting in zinc ejection and
ognition pentamer but will specifically bind to a pentamer with a loss of activity.77,78
thymine in this position.74 Thus redox reactions are able to Although technically not a zinc finger structure because the
fine-tune the DNA binding specificity of p53. zinc ion selected for drug design is involved in catalysis, the search
Another example of a zinc finger protein whose activity is regu- for matrix metalloproteinases (MMPs) inhibitors is instructive in
lated by redox reactions is the heat shock protein (HSP) 33 which the discussion of zinc-targeted intervention therapies. Human
is a bacterial chaperone whose activity is stimulated by cellular macrophage elastase plays an important role in inflammatory pro-
oxidative stress.75 This protein has a unique C2C2 zinc finger that cesses and is believed to be involved in the destruction of elastase
binds zinc in the inactivated state. When a cell is subjected to in lung alveolar tissues in emphysema. The crystal structure of a
oxidative stress, the zinc coordinating cysteines become oxidized hydroxamic acid inhibitor with the zinc catalytic site of this en-
resulting in zinc ejection. This cysteine oxidation and loss of zinc zyme reveals a number of structural requirements for such
results in a protein conformational change that converts the pro- zinc-targeted therapies.79 This structure reveals that positioning
tein to its activated chaperone conformation where it can now of the hydroxamic acid zinc chelator group of the inhibitor is
help maintain the proper conformation of other critical proteins dependent upon precise hydrogen bonding of the inhibitor polar
necessary for cell survival. Zinc finger reactivities are a direct re- groups with elastase amino acid side chains as well as interactions
sult of the unique coordination chemistry present at the confor- between a hydrophobic pocket in the enzyme active site with an
mational active sites of these proteins. These active centers repre- aromatic group on the inhibitor.79 This combination of unique
sent a paradoxical state for the cell. Although the zinc ions protein-inhibitor interactions contributes to the strength and
themselves are not redox active, the cysteine residues and espe- specificity of inhibitor binding and such information gleaned from
cially zinc-bound thiolates responsible for metal coordination and this study is applicable to other zinc-targeted therapies as well.
Zinc Finger Interactions with Metals and Other Small Molecules 45
References 26. Huheey JE, Keiter EA, Keiter JL. Inorganic chemistry: Principals of
1. Hanas JS, Hazuda DJ, Bogenhagen DF et al. Xenopus transcription structure and reactivity. 4th ed. New York: HarperCollins College
factor A requires zinc for binding to the 5 S RNA gene. J Biol Publishers, 1993:578:346-355.
Chem 1983; 258:14120-14125. 27. Dudev T, Lim C. Tetrahedral vs octahedral zinc complexes with
2. Ginsberg AM, King BO, Roeder RG. Xenopus 5S gene transcription ligands of biological interest: A DFT/CDM study. J Am Chem Soc
factor, TFIIIA: Characterization of a cDNA clone and measurement 2000; 122:11146-11153.
of RNA levels throughout development. Cell 1984; 39:479-489. 28. Roe RR, Pang Y-P. Zinc’s exclusive tetrahedral coordination governed
3. Miller J, McLachlan AD, Klug A. Repetitive zinc-binding domains by its electronic structure. J Mol Model 1999; 5:134-140.
in the protein transcription factor IIIA from Xenopus oocytes. 29. Blasie CA, Berg JM. Structurebased thermodynamic analysis of a
EMBO J 1985; 4:1609-1614. coupled metal binding-protein folding reaction involving a zinc finger
4. Brown RS, Sander C, Argos P. The primary structure of transcription peptide. Biochemistry 2002; 41:15068-15073.
factor TFIIIA has 12 consecutive repeats. FEBS Lett 1985; 30. Smith JN, Shirin Z, Carrano CJ. Control of thiolate nucleophilicity
186:271-274. and specificity in zinc metalloproteins by hydrogen bonding: Lessons
5. Diakun GP, Fairall L, Klug A. EXAFS study of the zinc-binding from model compound studies. J Am Chem Soc 2003; 125:868-869.
sites in the protein transcription factor IIIA. Nature 1986; 31. Simonson T, Calimet N. CysxHisy-Zn2+ interactions: Thiol vs. thiolate
324:698-699. coordination. Proteins 2002; 49:37-48.
6. Hanas JS, Duke AL, Gaskins C. Conformational states of Xenopus 32. Dudev T, Lim C. Factors governing the protonation state of cysteines
transcription factor IIIA. Biochemistry 1989; 28:4083-4088. in proteins: An Ab initio/CDM study. J Am Chem Soc 2002;
7. Berg JM. Potential metal-binding domains in nucleic acid binding 124:6759-6766.
proteins. Science 1986; 232:485-487. 33. Hartwig A. Zinc finger proteins as potential targets for toxic metal
8. Vallee BL, Coleman JE, Auld DS. Zinc fingers, zinc clusters, and ions: Differential effects on structure and function. Antiox Redox
zinc twists in DNA-binding protein domains. Proc Natl Acad Sci Signal 2001; 3:625-634.
USA 1991; 88:999-1003. 34. Berg JM, Shi Y. The galvanization of biology: A growing appreciation
9. Hanas JS, Gaskins CJ, Smith JF et al. Structure, function, evolution for the roles of zinc. Science 1996; 271:1081-1085.
of transcription factor IIIA. Prog Nucleic Acid Res Mol Biol 1992; 35. Krizek BA, Merkle DL, Berg JM. Ligand variation and metal ion
43:205-239 binding specificity in zinc finger peptides. Inorg Chem 1993;
10. Klug A, Schwabe JW. Protein motifs 5. Zinc fingers. FASEB J 1995; 32:937-940.
9:597-604. 36. Berg JM, Merkle DL. On the metal ion specificity of “zinc finger”
11. Krishna SS, Majumdar I, Grishin NV. Structural classification of proteins. J Am Chem Soc 1989; 111:3759-3761.
zinc fingers. Nucleic Acids Res 2003; 31:532-550. 37. Posewitz MC, Wilcox DE. Properties of the Sp1 zinc finger 3
12. Klein DJ, Johnson PE, Zollars ES et al. The NMR structure of the peptide: Coordination chemistry, redox reactions, and metal binding
nucleocapsid protein from the mouse mammary tumor virus reveals competition with metallothionein. Chem Res Toxico 1995;
unusual folding of the C-terminal zinc knuckle. Biochemistry 2000; 8:1020-1028.
39:1604-1612. 38. Makowski GS, Sunderman Jr.FW. The interactions of zinc, nickel,
13. Zhu W, Zeng Q, Colangelo CM et al. The N-terminal domain of and cadmium with Xenopus transcription factor IIIA, assessed by
TFIIB from Pyrococcus furiosus forms a zinc ribbon. Nature Struct equilibrium dialysis. J Inorg Biochem 1992; 48:107-119.
Biol 1996; 3:122-124. 39. Jiang L-J, Vasak M, Vallee BL et al. Zinc transfer potentials of the
14. Frankel AD, Berg JM, Pabo CO. Metal-dependent folding of a single alpha- and beta- clusters of metallothionein are affected by domain
zinc finger from transcription factor IIIA. Proc Natl Acad Sci USA interactions in the whole molecule. Proc Natl Acad Sci USA 2000;
1987; 84:4841-4845. 97:2503-2508.
15. Nomura A, Sugiura Y. Contribution of individual zinc ligands to 40. Finney LA, O’Halloran TV. Transition metal speciation in the cell:
metal binding and peptide folding of zinc finger peptides. Inorg Insights from the chemistry of metal ion receptors. Science 2003;
Chem 2002; 41:3693-3698. 300:931-936.
16. Lee MS, Gippert GP, Soman KV et al. Three-dimensional solution 41. Jacob C, Maret W, Valee BL. Control of zinc transfer between
structure of a single zinc finger DNA-binding domain. Science 1989; thionein, metallothionein, and zinc proteins. Proc Natl Acad Sci USA
245:635-637. 1998; 95:3489-3494.
17. Pavletich NP, Pabo CO. Zinc finger-DNA recognition: Crystal 42. Berg JM, Godwin HA. Lessons from zinc-binding peptides. Annu
structure of a Zif268-DNA complex at 2.1 A. Science 1991; Rev Biophys Biomol Struct 1997; 26:357-371.
252:809-817. 43. Krizek BA, Berg JM. Complexes of zinc finger peptides with Ni2+
18. Shi Y, Berger RD, Berg JM. Metal binding properties of single amino and Fe2+. Inorg Chem 1992; 31:2984-2986.
acid deletion mutants of zinc finger peptides: Studies using cobalt 44. Krizek BA, Amann BT, Kilfoil VJ et al. A consensus zinc finger
(II) as a spectroscopic probe. Biophys J 1993; 64:749-753. peptide: Design, high-affinity metal binding, a pH-dependent
19. Miura T, Satoh T, Takeuchi H. Role of metal-ligand coordination structure, and a His to Cys sequence variant. J Am Chem Soc
in the folding pathway of zinc finger peptides. Biochim Biophys 1991;113:4518-4523.
Acta 1998; 1384:171-179. 45. Mackay JP, Crossley M. Zinc fingers are sticking together. Trends
20. Parraga G, Horvath S, Hood L et al. Spectroscopic studies of Biochem 1998; 23:1-4.
wild-type and mutant “zinc finger” peptides: Determinants of domain 46. Sunderman FW, Barber AM. Finger-loops, oncogenes, and metals.
folding and structure. Proc Natl Acad Sci USA 1990; 87:137-141. Claude Passmore Brown Memorial Lecture. Ann Clin Lab Sci 1988;
21. Smith JF, Hawkins J, Leonard RE et al. Structural elements in the 18:267-288.
N-terminal half of transcription factor IIIA required for factor 47. Hanas JS, Gunn CG. Inhibition of transcription factor IIIA-DNA
binding to the 5S RNA gene internal control region. Nucl Acids interactions by xenobiotic metal ions. Nucleic Acids Res 1996;
Res 1991; 19:6871-6876. 24:924-930.
22. Veenstra TD, Johnson KL, Tomlinson AJ et al. Zinc-induced 48. Beyersmann D, Hechtenberg S. Cadmium, gene regulation, and
conformational changes in the DNA-binding domain of the vitamin cellular signalling in mammalian cells. Toxicol Appl Pharmacol 1997;
D receptor determined by electrospray ionization mass spectrometry. 144:247-261.
J Am Soc Mass Spectrom 1998; 9:8-14. 49. Hartwig A. Carcinogenicity of metal compounds: Possible role of
23. Williams RJP. The biochemistry of zinc. Polyhedron 1987; 6:61-69. DNA repair inhibition. Toxicol Lett 1998; 102-103:235-239.
24. McCall KA, Huang C-c, Fierke CA. Function and mechanism of 50. Meplan C, Mann K, Hainaut P. Cadmium induces conformational
zinc metalloenzymes. J Nutr 2000; 130:1437S-1446S. modifications of wild-type p53 and suppresses p53 response to DNA
25. Lippard SJ, Berg JM. Principals of Bioinorganic Chemistry. University damage in cultured cells. J Biol Chem 1999; 274:31663-31670.
Science Books. Mill Valley, CA: 1994:21-39.
46 Zinc Finger Proteins: From Atomic Contact to Cellular Function
51. Hartwig A, Asmuss M, Blessing H et al. Interference by toxic metal 65. Kroncke K-D. Zinc finger proteins as molecular targets for nitric
ions with zinc-dependent proteins involved in maintaining genomic oxide-mediated gene regulation. Antiox Redox Signal 2001;
stability. Food Chem Toxicol 2002; 40:1179-1184. 3:565-575.
52. Simons Jr SS, Chakraborti PK, Cavanaugh AH. Arsenite and 66. Kroncke KD, Fehsel K, Schmidt T et al. Nitric oxide destroys
cadmium(II) as probes of glucocorticoid receptor structure and zinc-sulfur clusters inducing zinc release from metallothionein and
function. J Biol Chem 1990; 265:1938-1945. inhibition of the zinc finger-type yeast transcription activator LAC9.
53. Petering DH, Huang M, Moteki S et al. Cadmium and lead Biochem Biophys Res Commun 1994; 200:1105-1110.
interactions with transcription factor IIIA from Xenopus laevis: A 67. Pearce LL, Gandley RE, Han W et al. Role of metallothionein in
model for zinc finger protein reactions with toxic metal ions and nitric oxide signalling as revealed by a green fluorescent protein fusion
metallothionein. Mar Environ Res 2000; 50:89-92. protein. Proc Natl Acad Sci USA 2000; 97:477-482.
54. Pavletich NP, Chambers KA, Pabo CO. The DNA-binding domain 68. Berendji D, Kolb-Bachofen V, Zipfel PF et al. Zinc finger
of p53 contains the four conserved regions and the major mutation transcription factors as molecular targets for nitric oxide-mediated
hot spots. Genes Dev 1993; 7:2556-2564. immunosuppression: Inhibition of IL-2 gene expression in murine
55. Cho Y, Gorina S, Jeffrey PD et al. Crystal structure of a p53 tumor lymphocytes. Mol Med 1999; 5:721-730.
suppressor-DNA complex: Understanding tumorigenic mutations. 69. Wang S, Wang W, Wesley RA et al. A Sp1 binding site of the tumor
Science 1994; 265:346-355. necrosis factor promoter functions as a nitric oxide repressor element.
56. Giedroc DP, Chen X, Pennella MA et al. Conformational J Biol Chem 1999; 274:33190-33193.
heterogeneity in the C-terminal zinc fingers of human MTF-1. J 70. Wu X, Bishopric NH, Discher DJ et al. Physical and functional
Biol Chem 2001; 276:42322-42332. sensitivity of zinc finger transcription factors to redox change. Mol
57. Zawia NH, Sharan R, Brydie M et al. Sp1 as a target site for Cell Biol 1996; 16:1035-1046.
metal-induced pertubations of transcriptional regulation of 71. Meplan C, Richard M-J, Hainaut P. Redox signalling and transition
developmental brain gene expression. Brain Res Dev Brain Res 1998; metals in the control of the p53 pathway. Biochem Pharmacol 2000;
107:291-298. 59:25-33.
58. Hanas JS, Rodgers JS, Bantle JA et al. Lead inhibition of 72. Parks D, Bolinger R, Mann K. Redox state regulates binding of p53
DNA-binding mechanism of Cys2His2 zinc finger proteins. Mol to sequence-specific DNA, but not to nonspecific or mismatched
Pharmacol 1999; 56:982-988. DNA. Nucl Acids Res 1997; 25:1289-1295.
59. Payne JC, ter Horst MA, Godwin HA. Lead fingers: Pb2+ binding 73. Casso D, Beach D. A mutation in a thioredoxin reductase homolog
to structural zinc-binding domains determined directly by monitoring suppresses p53-induced growth inhibition in the fission yeast
lead-thiolate charge-transfer bands. J Am Chem Soc 1999; Schizosaccharomyces pombe. Mol Gen Genet 1996; 252:518-529.
121:6850-6855. 74. Buzek J, Latonen L, Kurki S et al. Redox state of tumor suppressor
60. Zawia NH, Crumpton T, Reddy GR et al. Disruption of the zinc p53 regulates its sequence-specific DNA binding in DNA-damaged
finger domain: A common target that underlies many of the effects cells by cysteine 277. Nucl Acids Res 2002; 30:2340-2348.
of lead. Neurotoxicol 2000; 2:1069-1080. 75. Jakob U, Eser M, Bardwell JCA. Redox switch of Hsp33 has a novel
61. Rodgers JS, Hocker JR, Hanas RJ et al. Mercuric ion inhibition of zinc-binding motif. J Biol Chem 2000; 275:38302-38310.
eukaryotic transcription factor binding to DNA. Biochem Pharmacol 76. Beerheide W, Bernard H-U, Tan Y-J et al. Potential drugs against
2001; 61:1543-1550. cervical cancer: Zinc-ejecting inhibitors of the human papillomavirus
62. Handel ML, deFazio A, Watts CKW et al. Inhibition of DNA type 16 E6 oncoprotein. J Natl Cancer Inst 1999; 91:1211-1220.
binding and transcriptional activity of a nuclear receptor transcription 77. Tummino PJ, Scholten JD, Harvey PJ et al. The in vitro ejection of
factor by aurothiomalate and other metal ions. Mol Pharmacol 1991; zinc from human immunodeficiency virus (HIV) type I nucleocapsid
40:613-618. protein by disulfide benzamides with cellular anti-HIV activity. Proc
63. Larabee JL, Hocker JR, Hanas RJ et al. Inhibition of zinc finger Natl Acad Sci USA 1996; 93:969-973.
protein-DNA interactions by sodium selenite. Biochem Pharmacol 78. Rice WG, Supko JG, Malspeis L et al. Inhibitors of HIV
2002; 64:1757-1765. nucleocapsid protein zinc fingers as candidates for the treatment of
64. Chen Y, Maret W. Catalytic selenols couple the redox cycles of AIDS. Science 1995; 270:1194-1197.
metallothionein and glutathione. Eur J Biochem 2001; 79. Nar H, Werle K, Bauer MMT et al. Crystal structure of human
268:3346-3353. macrophage elastase (MMP-12) in complex with a hydroxamic acid
inhibitor. J Mol Biol 2001; 312:743-751.
CHAPTER 9
Abstract
the amino-terminal part of the α helix.9 In particular, a “code”
T
he possibility of using designed transcription factors to
control gene expression is highly appealing in view of a that relates the amino acids of a single zinc finger to its associated
wide range of promising applications in research and bio- DNA target has been proposed for a variety of zinc finger do-
medicine. In the last decade, the efforts of several research groups mains.10-12 Using and expanding this “code” several research
have clarified the molecular interactions between zinc finger pro- groups have recently engineered and/or selected peptides con-
teins and DNA, generating a “recognition code” that relates taining zinc finger domains able to recognize and to bind specific
amino acids in specific positions of the finger domain to its DNA DNA sequences. These peptides represent DNA binding mod-
target sequence. This chapter describes the methods to design ules that can act as artificial transcription factors when fused to
artificial zinc finger transcription factors using the “code” and an appropriate effector domain.13 Therefore, synthetic ZFP TFs
recent novel selection approaches. Efficacy and possible applica- represent the feasibility of “gene regulation on demand” in higher
tions of artificial transcription factors are discussed. eukaryotes, opening great opportunities for genome analysis and
control. Moreover, zinc finger peptides can act as RNA-binding
modules, enlarging the number of the their possible applications
Introduction in biotechnologies and molecular medicine.14-17
Despite the sequencing and the analysis of the genome of dif-
ferent organisms, the mechanisms by which a peculiar cell type
coordinates gene regulation remain largely unknown. The com- Evolution of the Recognition Code Concept
plex process of eukaryotic gene expression requires integration of Despite the great variety of structural domains, the specificity
various events, accurately selected in million of years of evolu- of interactions observed between protein and DNA is most fre-
tion. These events include: the assembly of general transcription quently due to the complementarity of the surfaces of a protein α
factors, the recruitment of theh transcription machinery to the helix and the major groove of DNA.18 With regards to the recur-
promoter, utilization of cofactor-complexes involved in activa- ring physical interaction of α helix and major groove, it has seemed
tion or repression and chromatin remodeling.1-4 Transcription feasible to describe contacts between particular amino acids and
regulation is mainly achieved through the action of proteins known DNA bases by a simple set of rules that can be combined to form
as transcription factors (TF), commonly organized in two sepa- a stereochemical recognition code. Among the variety of DNA
rable modules: a DNA binding domain (DBD) capable of target- binding domains, the C2H2 zinc finger motif has represented an
ing specific DNA sequences and an effector domain (ED) ca- attractive motif as it is structurally well characterized and it pre-
pable of regulating at different levels the multiprotein machinery, sents distinct DNA binding properties.19-21 ZFP domains dis-
that directs transcription. Transcription factors represent 6% of play only minimal cooperativity, so each finger recognizes a DNA
all the proteins in the human genome.5 In the large class of DNA sequence with relative independence. An understanding of the
binding molecules the Cys2His2 (C2H2) zinc finger domain, with interactions between ZFP domain and its DNA target sequence
the 4,500 examples so far identified, represents the most frequently was initially guided by the analysis of sequence homology be-
encoded motif in the human genome. In particular, C2H2 zinc tween individual fingers from the same protein, or from differ-
finger proteins (ZFP) provide a simple and versatile framework ent, though related, proteins. Site directed mutagenesis experi-
favorable to design and select novel DNA binding proteins that ments, that substituted the non homologous α helix residues,
recognize desired target sequences.6-8 The C2H2 zinc finger do- were able to change the specificity of the mutant finger.6,22-25
main is a compact motif of 28-30 amino acids, that comprises an Thus, the analysis of the available large database of zinc finger
α helix containing two invariant histidine residues coordinated sequence and the rational mutagenesis revealed some
through a zinc atom to two cysteines of a single β turn. The x-ray base-contacting positions even before the solution of the Zif268
crystal structures of the three zinc finger domains of the tran- crystal structure. The determination of Zif268 protein/DNA com-
scription factor Zif268 bound to its DNA target site, reveal that plex at 2.1Å was the basis for many subsequent studies, and Zif268
an individual finger domain binds essentially three base pairs of has proved to be a very useful model system for understanding
double stranded DNA sequence with specific contacts through how C2H2 ZFPs recognize DNA.21 The crystal structure shows
Zinc Finger Proteins: From Atomic Contact to Cellular Function, edited by Shiro Iuchi and Natalie Kuldell.
©2005 Landes Bioscience/Eurekah.com and Kluwer Academic/Plenum Publishers.
48 Zinc Finger Proteins: From Atomic Contact to Cellular Function
Parallel Selection
This approach (Fig. 4a) preselects individual monomers by
phage display and subsequently combines them into three-ZFPs.
In particular, two Zif268 fixed fingers (anchor fingers), the first
and the third one, ensure the correct positioning on target site,
Figure 3. Table of nondegenerate recognition code. The table allows while the middle finger is randomized in the crucial positions.29
identification of one specific amino acid at each of positions: -1, +2, +3, To date, using parallel selection, Barbas group described zinc fin-
and +6 of the α-helical region of the zinc-finger domain from the over- ger domains that recognize 30 different trinucleotides compris-
lapping 4-bp target sequence in a given DNA target. The depicted cor- ing the GNN and ANN subsets (where N is any one of the four
respondences between amino acid and base are derived from observed nucleotides).38,39,58-60 These selected “monomers” supply prefab-
X-ray crystal structures contacts and from potential amino acid-base ricated building blocks that can be “stitched” together.59 This
interactions. The correspondences suggested exclusively in nondegenerate method had several applications in the construction of artificial
code (not reported in the degenerate code table) are indicated with transcription factors that modulate the expression of diverse tar-
asterisks. get genes even at the endogenous chromosomal site (Table
1).7,13,39,58,61,62 The following two approaches address the prob-
published two papers in which ZFPs were rationally designed to lems dealing with target sites overlap and cooperativity in zinc
bind arbitrary DNA sequences.45,46 These ZFPs bind to the pre- finger-DNA recognition.
dicted sequence with a fairly strong affinity. As a proof of speci-
ficity, the proteins selected strictly related sequences among ran- Sequential Selection
domized ones in SELEX experiments. The results were In this system (Fig. 4b) critical amino acid positions of indi-
encouraging as they validate some of the rules derived from the vidual zinc finger domains are randomized sequentially in the
code. The design based on the code is a valid alternative to selec- context of two “anchor” fingers.32 Thus, finger three of Zif268
tion approach, even considering further level of complexity due protein is substituted by the library represented by randomized
to context-dependent effects such as synergy between adjacent finger one (F1R). After the primary selection against the first three
fingers.36,50-53 Sera and Uranga proposed a rapid creation scheme, nucleotides, the N-terminal anchor finger is removed and the
based on a nondegenerate code (Fig. 3) to develop new ZFPs library of finger two (F2R) is cloned at the C-terminus of the
with satisfactory binding properties in a “high-through-put man- newly selected finger one (F1S), which now is in the central posi-
ner”.54 This method will enable manipulation of gene expres- tion. The F2R library is selected to bind the next three nucle-
sions by screening multiple ZFP TFs for multiple sites in a given otides in the DNA sequence. After one additional round of
Target Effector
Gene Reporter Endogenous ZFP-DBD Used Domain Activation Repression References
exchanges, a new three-finger protein is created in which all of lections the two selected zinc finger portions are spliced together
the fingers (F1S F2S F3S) have been selected in the context of the to make a novel three-ZFP.64
finger next to it.
Alternative Strategies
Bipartite Selection Alternative strategies for selecting ZFPs have used a yeast
This method (Fig. 4c) combines the advantages of the two one-hybrid and a bacterial two hybrid selection systems.65,66 These
preceding methods and overcome incompatibilities at the inter- systems show two useful features: the selection is performed in
face between adjacent fingers.63 According to this method selec- the context of living cells and the best proteins can be obtained
tions are carried out in parallel from two premade “half ” librar- through a one-step selection. In addition, a novel selection sys-
ies, in which one-and-a-half fingers of the three-finger Zif268 are tem using mammalian cultured cells has been recently described.67
randomized. Selection of these two libraries is carried out at the Efforts in synthetic ZFPs design have also been focused on the
same time against DNA sequences, in which half of the Zif268 assembly of ZFPs with more than three fingers.14,57,68 In order to
target site is replaced by the sequence of interest. After these se- assembly such polydactyl ZFPs two strategies have been described:
Synthetic Zinc Finger Transcription Factors 51
overexpressed in human cancers.13,38,61 Significantly, regulation In addition, artificial ZFPs could be selected by ZFP libraries
of the two genes was highly specific, despite the fact that bind- on the basis of an induced improved phenotype. Such ZFPs could
ing sites targeted by erB-2 and erbB-3 share 15 of 18 nucle- also be used to regulate the expression of uncharacterized genes
otides. Moreover, they also obtained an inducible expression of leading to identifying new genes and determining their role.67
the artificial ZFPs, which is particularly useful in both gene
therapy and understanding of biological mechanisms. ZFPs and RNA
Liu et al synthesized a panel of ZFPs to target the endogenous The relevance of RNAs in many biological processes has gen-
locus of vascular endothelial growth factor (VEGF-A), a specific erated great interest in designing sequence-specific RNA-binding
inducer of new blood vessel growth, involved in a variety of medi- molecules.46,103,104 In particular, ZFPs selected to bind RNA, fused
cal conditions.96 They demonstrated that combinations of to appropriate modules, can regulate gene expression at different
three-ZFPs targeted to distinct-adjacent chromosomal sites linked levels, interfering with alternative splicing, RNA trafficking and
to either VP16 or p65 were acting synergistically on VEGF-A stability.105
activation. Moreover, they compared the behaviour of ZFPs tar- Using a phage display approach Friesen et al reported on the
geted to accessible regions (identified by DNAase I hypersensi- amino acid requirements for the binding of TFIIIA to 5S RNA,
tivity assay) with ZFPs targeted to inaccessible regions. While no demonstrating that some RNA/α helix contacts are identical to
evident differences were detected using naked reporter constructs, DNA contacts made by C2H2 zinc fingers.106 Their results indi-
a clear difference in the activity of the endogenous VEGF-A gene cated that zinc finger recognition of RNA might be similar to
was seen, thus emphasizing the importance of chromatin struc- DNA. However more than DNA, RNAs, on the basis of their
ture status. A similar approach was followed to activate the en- rich repertoire of possible tertiary structures, display surfaces that
dogenous erythropoietin gene (EPO), suggesting the possibility resemble protein motifs. It has been observed that proteins and
of recruiting designed ZFPs to a site of interest in order to re- RNA can undergo peculiar conformational changes only in the
model the chromatin and regulate transcription.95 context of the complex RNA/protein.107 This characteristic gives
Peroxisome proliferator activated receptor-γ2 (PPARγ2) has the opportunity to design zinc fingers using structurebased strat-
been used as an interesting example of knockdown expression. egies. McColl used this structure-based approach to design
PPARγ2 is nuclear hormonal receptor, whose expression was se- RNA-binding zinc fingers that recognize the HIV- REV response
lectively repressed by a designed ZFP. This ZFP distinguishes element (RRE).108 An arginine-rich α helix from HIV REV was
between the two isoforms γ1 and γ2, leading to determine an inserted into the middle zinc finger of Zif268 framework. The
unique role of PPARγ2 isoform for a correct adipogenesis.97 This resulting hybrid zinc finger-REV was able to fold in a zinc de-
kind of study could be functional for the design of a specific drug pendent manner and binds specifically to the RRE. NMR ex-
able to affect only the desired isoform.14 Customized ZFP TFs periments data provide evidences that monomeric zinc fingers
were also able to inhibit transcription of viral DNA and therefore can recognize specific nucleic acids demonstrating once again the
viral replication. Klug and colleagues designed ZFPs targeted to modular nature of the zinc finger motif.109 In particular, it seems
the viral immediate-early promoter (IE) of HSV-1 that were able to contact RNA into major groove through an extensive protein
to reduce the viral titer by about 90%.99 Similarly, they obtained interface, involving probably not only the α helix, but also the β
repression of the HIV-1 LTR promoter and inhibition of HIV-1 sheet of the zinc finger-binding surface. Moreover several studies
replication.100 These results are promising for biotherapies to face have demonstrated that tethering of multiple binding modules
viral infections. can enhance RNA binding affinity and specificity, a strategy simi-
Juliano et al reported data on an artificial transcription factor larly used in DNA recognition.104
able to stimulate the expression of the endogenous pro-apoptotic
gene bax, one of the p53 targets.101 They propose the induction
of bax gene as valuable tool in cancer chemotherapy, by dimin- ZFPs and Plants
ishing survival of p53-deficient tumor cells. The manipulation of plants is an important issue with appli-
All these data regarding the possibility to reprogram the ex- cations in basic plant biology research and agricultural biotech-
pression of an endogenous gene target seem to predict that engi- nologies.110 Transgenesis has been the most common method to
neered transcription factors could be effective in a whole organ- overexpress a gene in plants. Engineered ZFP TFs represent a
ism model. Indeed, expression of an artificial ZFP TF in a mouse novel promising tool to alter the expression of specific target genes,
ear model (delivered by adeno-vectors) led to induction of the in order to get healthier or more nutritious crop plants.111,112
VEGF-A, demonstrating for the first time the feasibility of regu- Genes related to disease/stress resistance could be turned on, while
lating, in vivo, crucial biological processes as angiogenesis.98 These genes related to pathologic status or to the production of
data underline the potential therapeutic effects of designer TFs. anti-nutritive proteins could be turned off.113 Importantly, to re-
In this context, Kim group proposed the elegant “GeneGrip” duce or eliminate a gene expression in plants is not easy, since to
method based on screening of “natural” zinc fingers from human date no natural modular repressors have been identified. How-
genome, in order to use them as modular building blocks in the ever, foreign repressor domains may function in plants. The mam-
generation of novel ZFPs.102 A naturally occurring linker sequence malian repressor domain KRAB appears to be a poor candidate
was used to connect individual zinc fingers. Microarrays with cell for gene regulation in plants, since it is not found in organisms
lines stable transfected with engineered ZFPs permit to monitor such as Drosophila melanogaster, Caenorhabditis elegans, Saccharo-
the specificity of their action on a genome wide scale. More im- myces cerevisiae. By contrast, the Sid domain (Sin3A interaction
portantly, ZFPs constructed by shuffling naturally occurring hu- domain) interacts with the highly conserved Sin3A and is there-
man zinc fingers may be optimal in therapeutic applications in fore a promising effector domain to regulate endogenous plant
terms of reducing host immune response. genes.111
Synthetic Zinc Finger Transcription Factors 53
Several promising studies in the regulation of plant genes by 16. Urnov FD, Rebar EJ, Reik A et al. Designed transcription factors as
ZFP TFs have recently reported.114 A ZFP TF was able to acti- structural, functional and therapeutic probes of chromatin in vivo.
vate a β-glucuronidase (GUS) reporter stably integrated in to- Fourth in review series on chromatin dynamics. EMBO Rep 2002;
3(7):610-615.
bacco plants. Guan et al reported of ZFP TFs able to target the 17. Urnov FD, Rebar EJ. Designed transcription factors as tools for
Apetala 3 (Ape3) promoter in Arabidopsis for both activation and therapeutics and functional genomics. Biochem Pharmacol 2002;
repression. Importantly, in both cases activation and repression 64(5-6):919-923.
were inherited in subsequent generations.79 In the future, librar- 18. Pabo CO, Sauer RT. Transcription factors: Structural families and
ies of ZFP TFs could be delivered using Agrobacterium transfor- principles of DNA recognition. Annu Rev Biochem 1992;
mation to obtain large libraries that could be screened for novel 61:1053-1095.
19. Rhodes D, Klug A. An underlying repeat in some transcriptional
plant phenotypes. control sequences corresponding to half a double helical turn of
DNA. Cell 1986; 46(1):123-132.
Perspective 20. Berg JM. Zinc finger domains: Hypotheses and current knowledge.
The design of efficient TFs, with a wide range of applications Annu Rev Biophys Biophys Chem 1990; 19:405-421.
21. Pavletich NP, Pabo CO. Zinc finger-DNA recognition: Crystal
in both basic and applied sciences across many types of biotech- structure of a Zif268-DNA complex at 2.1 A. Science 1991;
nologies, remains a focal point of research in the post-genomic 252(5007):809-817.
era.14,115 At the moment, the delivery of artificial TFs into organ- 22. Desjarlais JR, Berg JM. Redesigning the DNA-binding specificity of
isms and the ability to maintain the desired expression level of a zinc finger protein: A data base-guided approach. Proteins 1992;
target genes are challenging objectives.15 12(2):101-104.
23. Desjarlais, Berg JM. Redesigning the DNA-binding specificity of a
zinc finger protein: A data base-guided approach. Proteins 1992;
Acknowledgements 13(3):272.
This work was supported by Telethon A160 and Firb 24. Nardelli J, Gibson TJ, Vesque C et al. Base sequence discrimination
“Epigenetica e cromatina”. N. Corbi is recipient of a FIRC fel- by zinc-finger DNA-binding domains. Nature 1991;
lowship. 349(6305):175-178.
25. Nardelli J, Gibson T, Charnay P. Zinc finger-DNA recognition:
Analysis of base specificity by site-directed mutagenesis. Nucleic Acids
References Res 1992; 20(16):4137-4144.
1. Ptashne M, Gann A. Transcriptional activation by recruitment. 26. Desjarlais JR, Berg JM. Use of a zinc-finger consensus sequence
Nature 1997; 386(6625):569-577. framework and specificity rules to design specific DNA binding
2. Sauer F, Tjian R. Mechanisms of transcriptional activation: proteins. Proc Natl Acad Sci USA 1993; 90(6):2256-2260.
Differences and similarities between yeast, Drosophila, and man. Curr 27. Rebar EJ, Pabo CO. Zinc finger phage: Affinity selection of fingers
Opin Genet Dev 1997; 7(2):176-181. with new DNA-binding specificities. Science 1994;
3. Hampsey M. Molecular genetics of the RNA polymerase II general 263(5147):671-673.
transcriptional machinery. Microbiol Mol Biol Rev 1998; 28. Desjarlais JR, Berg JM. Length-encoded multiplex binding site
62(2):465-503. determination: Application to zinc finger proteins. Proc Natl Acad
4. Cremer T, Cremer C. Chromosome territories, nuclear architecture Sci USA 1994; 91(23):11099-11103.
and gene regulation in mammalian cells. Nat Rev Genet 2001; 29. Choo Y, Klug A. Designing DNA-binding proteins on the surface
2(4):292-301. of filamentous phage. Curr Opin Biotechnol 1995; 6(4):431-436.
5. Venter JC, Adams MD, Myers EW et al. The sequence of the human 30. Wu H, Yang WP, Barbas III CF. Building zinc fingers by selection:
genome. Science 2001; 291(5507):1304-1351. Toward a therapeutic application. Proc Natl Acad Sci USA 1995;
6. Desjarlais JR, Berg JM. Toward rules relating zinc finger protein 92(2):344-348.
sequences and DNA binding site preferences. Proc Natl Acad Sci 31. Jamieson AC, Wang H, Kim SH. A zinc finger directory for
USA 1992; 89(16):7345-7349. high-affinity DNA recognition. Proc Natl Acad Sci USA 1996;
7. Isalan M, Choo Y. Rapid, high-throughput engineering of 93(23):12834-12839.
sequence-specific zinc finger DNA-binding proteins. Methods 32. Greisman HA, Pabo CO. A general strategy for selecting high-affinity
Enzymol 2001; 340:593-609. zinc finger proteins for diverse DNA target sites. Science 1997;
8. Pabo CO, Peisach E, Grant RA. Design and selection of novel 275(5300):657-661.
Cys2His2 zinc finger proteins. Annu Rev Biochem 2001; 70:313-340. 33. Isalan M, Choo Y. Engineering nucleic acid-binding proteins by
9. Pavletich NP, Pabo CO. Zinc finger-DNA recognition: Crystal phage display. Methods Mol Biol 2001; 148:417-429.
structure of a Zif268-DNA complex at 2.1 A. Science 1991; 34. Wolfe SA, Nekludova L, Pabo CO. DNA recognition by Cys2His2
252(5007):809-817. zinc finger proteins. Annu Rev Biophys Biomol Struct 2000;
10. Jamieson AC, Kim SH, Wells JA. In vitro selection of zinc fingers 29:183-212.
with altered DNA-binding specificity. Biochemistry 1994; 35. Kim CA, Berg JM. A 2.2 A resolution crystal structure of a designed
33(19):5689-5695. zinc finger protein bound to DNA. Nat Struct Biol 1996;
11. Choo Y, Klug A. Toward a code for the interactions of zinc fingers 3(11):940-945.
with DNA: Selection of randomized fingers displayed on phage. Proc 36. Elrod-Erickson M, Benson TE, Pabo CO. High-resolution structures
Natl Acad Sci USA 1994; 91(23):11163-7. of variant Zif268-DNA complexes: Implications for understanding
12. Choo Y, Klug A. Selection of DNA binding sites for zinc fingers zinc finger-DNA recognition. Structure 1998; 6(4):451-464.
using rationally randomized DNA reveals coded interactions. Proc 37. Choo Y, Klug A. Physical basis of a protein-DNA recognition code.
Natl Acad Sci USA 1994; 91(23):11168-11172. Curr Opin Struct Biol 1997; 7(1):117-125.
13. Beerli RR, Dreier B, Barbas III CF. Positive and negative regulation 38. Segal DJ, Dreier B, Beerli RR et al. Toward controlling gene
of endogenous genes by designed transcription factors. Proc Natl expression at will: Selection and design of zinc finger domains
Acad Sci USA 2000; 97(4):1495-1500. recognizing each of the 5'-GNN-3' DNA target sequences. Proc Natl
14. Jamieson AC, Miller JC, Pabo CO. Drug discovery with engineered Acad Sci USA 1999; 96(6):2758-2763.
zinc-finger proteins. Nat Rev Drug Discov 2003; 2(5):361-368. 39. Dreier B, Beerli RR, Segal DJ et al. Development of zinc finger
15. Ansari AZ, Mapp AK. Modular design of artificial transcription domains for recognition of the 5'-ANN-3' family of DNA sequences
factors. Curr Opin Chem Biol 2002; 6(6):765-772. and their use in the construction of artificial transcription factors. J
Biol Chem 2001; 276(31):29466-29478.
54 Zinc Finger Proteins: From Atomic Contact to Cellular Function
40. Elrod-Erickson M, Rould MA, Nekludova L et al. Zif268 63. Isalan M, Klug A, Choo Y. Comprehensive DNA recognition through
protein-DNA complex refined at 1.6 A: A model system for concerted interactions from adjacent zinc fingers. Biochemistry 1998;
understanding zinc finger-DNA interactions. Structure 1996; 37(35):12026-12033.
4(10):1171-1180. 64. Isalan M, Klug A, Choo Y. A rapid, generally applicable method to
41. Pavletich NP, Pabo CO. Crystal structure of a five-finger GLI-DNA engineer zinc fingers illustrated by targeting the HIV-1 promoter.
complex: New perspectives on zinc fingers. Science 1993; Nat Biotechnol 2001; 19(7):656-660.
261(5129):1701-1707. 65. Bartsevich VV, Juliano RL. Regulation of the MDR1 gene by
42. Fairall L, Schwabe JW, Chapman L et al. The crystal structure of a transcriptional repressors selected using peptide combinatorial
two zinc-finger peptide reveals an extension to the rules for libraries. Mol Pharmacol 2000; 58(1):1-10.
zinc-finger/DNA recognition. Nature 1993; 366(6454):483-487. 66. Joung JK, Ramm EI, Pabo CO. A bacterial two-hybrid selection
43. Houbaviy HB, Usheva A, Shenk T et al. Cocrystal structure of YY1 system for studying protein-DNA and protein-protein interactions.
bound to the adeno-associated virus P5 initiator. Proc Natl Acad Proc Natl Acad Sci USA 2000; 97(13):7382-7387.
Sci USA 1996; 93(24):13577-13582. 67. Blancafort P, Magnenat L, Barbas CF. Scanning the human genome
44. Kim CA, Berg JM. Serine at position 2 in the DNA recognition with combinatorial transcription factor libraries. Nat Biotechnol
helix of a Cys2-His2 zinc finger peptide is not, in general, responsible 2003; 21(3):269-274.
for base recognition. J Mol Biol 1995; 252(1):1-5. 68. Imanishi M, Hori Y, Nagaoka M et al. Design of novel zinc finger
45. Corbi N, Perez M, Maione R et al. Synthesis of a new zinc finger proteins: Towards artificial control of specific gene expression. Eur J
peptide; comparison of its ‘code’ deduced and ‘CASTING’ derived Pharm Sci 2001; 13(1):91-97.
binding sites. FEBS Lett 1997; 417(1):71-74. 69. Liu Q, Segal DJ, Ghiara JB et al. Design of polydactyl zinc-finger
46. Corbi N, Libri V, Fanciulli M et al. Binding properties of the artificial proteins for unique addressing within complex genomes. Proc Natl
zinc fingers coding gene Sint1. Biochem Biophys Res Commun 1998; Acad Sci USA 1997; 94(11):5525-5530.
253(3):686-692. 70. Kim JS, Pabo CO. Getting a handhold on DNA: Design of poly-zinc
47. Corbi N, Libri V, Fanciulli M et al. The artificial zinc finger coding finger proteins with femtomolar dissociation constants. Proc Natl
gene ‘Jazz’ binds the utrophin promoter and activates transcription. Acad Sci USA 1998; 95(6):2812-2817.
Gene Ther 2000; 7(12):1076-1083. 71. Kamiuchi T, Abe E, Imanishi M et al. Artificial nine zinc-finger
48. Choo Y, Sanchez-Garcia I, Klug A. In vivo repression by a site-specific peptide with 30 base pair binding sites. Biochemistry 1998;
DNA-binding protein designed against an oncogenic sequence. 37(39):13827-13834.
Nature 1994; 372(6507):642-645. 72. Nagaoka M, Sugiura Y. Artificial zinc finger peptides: Creation, DNA
49. McNamara AR, Ford KG. A novel four zinc-finger protein targeted recognition, and gene regulation. J Inorg Biochem 2000;
against p190(BcrAbl) fusion oncogene cDNA: Utilisation of 82(1-4):57-63.
zinc-finger recognition codes. Nucleic Acids Res 2000; 73. Imanishi M, Hori Y, Nagaoka M et al. DNA-bending finger:
28(24):4865-4872. Artificial design of 6-zinc finger peptides with polyglycine linker and
50. Isalan M, Choo Y, Klug A. Synergy between adjacent zinc fingers in induction of DNA bending. Biochemistry 2000; 39(15):4383-4390.
sequence-specific DNA recognition. Proc Natl Acad Sci USA 1997; 74. Moore M, Klug A, Choo Y. Improved DNA binding specificity from
94(11):5617-5621. polyzinc finger peptides by using strings of two-finger units. Proc
51. Wolfe SA, Greisman HA, Ramm EI et al. Analysis of zinc fingers Natl Acad Sci USA 2001; 98(4):1437-1441.
optimized via phage display: Evaluating the utility of a recognition 75. Nagaoka M, Kaji T, Imanishi M et al. Multiconnection of identical
code. J Mol Biol 1999; 285(5):1917-1934. zinc finger: Implication for DNA binding affinity and unit
52. Pabo CO, Nekludova L. Geometric analysis and comparison of modulation of the three zinc finger domain. Biochemistry 2001;
protein-DNA interfaces: Why is there no simple code for recognition? 40(9):2932-2941.
J Mol Biol 2000; 301(3):597-624. 76. Nagaoka M, Nomura W, Shiraishi Y et al. Significant effect of linker
53. Wolfe SA, Grant RA, Elrod-Erickson M et al. Beyond the sequence on DNA recognition by multi-zinc finger protein. Biochem
“recognition code”: Structures of two Cys2His2 zinc finger/TATA Biophys Res Commun 2001; 282(4):1001-1007.
box complexes. Structure (Camb) 2001; 9(8):717-23. 77. Moore M, Choo Y, Klug A. Design of polyzinc finger peptides with
54. Sera T, Uranga C. Rational design of artificial zinc-finger proteins structured linkers. Proc Natl Acad Sci USA 2001; 98(4):1432-1436.
using a nondegenerate recognition code table. Biochemistry 2002; 78. Liu Q, Xia Z, Zhong X et al. Validated zinc finger protein designs
41(22):7074-7081. for all 16 GNN DNA triplet targets. J Biol Chem 2002;
55. Choo Y, Isalan M. Advances in zinc finger engineering. Curr Opin 277(6):3850-3856.
Struct Biol 2000; 10(4):411-416. 79. Guan X, Stege J, Kim M et al. Heritable endogenous gene regulation
56. Segal DJ, Barbas III CF. Design of novel sequence-specific in plants with designed polydactyl zinc finger transcription factors.
DNA-binding proteins. Curr Opin Chem Biol 2000; 4(1):34-39. Proc Natl Acad Sci USA 2002; 99(20):13296-13301.
57. Beerli RR, Barbas III CF. Engineering polydactyl zinc-finger 80. Arora PS, Ansari AZ, Best TP et al. Design of artificial transcriptional
transcription factors. Nat Biotechnol 2002; 20(2):135-141. activators with rigid poly-L-proline linkers. J Am Chem Soc 2002;
58. Dreier B, Segal DJ, Barbas III CF. Insights into the molecular 124(44):13067-13071.
recognition of the 5'-GNN-3' family of DNA sequences by zinc 81. Imanishi M, Sugiura Y. Artificial DNA-bending six-zinc finger
finger domains. J Mol Biol 2000; 303(4):489-502. peptides with different charged linkers: Distinct kinetic properties
59. Segal DJ. The use of zinc finger peptides to study the role of specific of DNA bindings. Biochemistry 2002; 41(4):1328-1334.
factor binding sites in the chromatin environment. Methods 2002; 82. Nomura W, Nagaoka M, Shiraishi Y et al. Influence of TFIIIA-type
26(1):76-83. linker at the N- or C-terminal of nine-zinc finger protein on
60. Segal DJ, Beerli RR, Blancafort P et al. Evaluation of a modular DNA-binding site. Biochem Biophys Res Commun 2003;
strategy for the construction of novel polydactyl zinc finger 300(1):87-92.
DNA-binding proteins. Biochemistry 2003; 42(7):2137-2148. 83. Pomerantz JL, Wolfe SA, Pabo CO. Structure-based design of a
61. Beerli RR, Segal DJ, Dreier B et al. Toward controlling gene dimeric zinc finger protein. Biochemistry 1998; 37(4):965-970.
expression at will: Specific regulation of the erbB-2/HER-2 promoter 84. Wang BS, Pabo CO. Dimerization of zinc fingers mediated by
by using polydactyl zinc finger proteins constructed from modular peptides evolved in vitro from random sequences. Proc Natl Acad
building blocks. Proc Natl Acad Sci USA 1998; 95(25):14628-14633. Sci USA 1999; 96(17):9568-9573.
62. Jouvenot Y, Ginjala V, Zhang L et al. Targeted regulation of 85. Wolfe SA, Ramm EI, Pabo CO. Combining structure-based design
imprinted genes by synthetic zinc-finger transcription factors. Gene with phage display to create new Cys(2)His(2) zinc finger dimers.
Ther 2003; 10(6):513-522. Structure Fold Des 2000; 8(7):739-750.
Synthetic Zinc Finger Transcription Factors 55
86. Wang BS, Grant RA, Pabo CO. Selected peptide extension contacts 101. Falke D, Fisher M, Ye D et al. Design of artificial transcription
hydrophobic patch on neighboring zinc finger and mediates factors to selectively regulate the pro-apoptotic bax gene. Nucleic
dimerization on DNA. Nat Struct Biol 2001; 8(7):589-593. Acids Res 2003; 31(3):e10.
87. Berg JM. Sp1 and the subfamily of zinc finger proteins with 102. Bae KH, Do KY, Shin HC et al. Human zinc fingers as building
guanine-rich binding sites. Proc Natl Acad Sci USA 1992; blocks in the construction of artificial transcription factors. Nat
89(23):11109-11110. Biotechnol 2003; 21(3):275-280.
88. Yaghmai R, Cutting GR. Optimized regulation of gene expression 103. Friesen WJ, Darby MK. Specific RNA binding proteins constructed
using artificial transcription factors. Mol Ther 2002; 5(6):685-694. from zinc fingers. Nat Struct Biol 1998; 5(7):543-546.
89. Segal DJ, Barbas III CF. Custom DNA-binding proteins come of 104. Cheng AC, Calabro V, Frankel AD. Design of RNA-binding proteins
age: Polydactyl zinc-finger proteins. Curr Opin Biotechnol 2001; and ligands. Curr Opin Struct Biol 2001; 11(4):478-484.
12(6):632-627. 105. Cartegni L, Krainer AR. Correction of disease-associated exon
90. Ansari AZ. Fingers reach for the genome. Nat Biotechnol 2003; skipping by synthetic exon-specific activators. Nat Struct Biol 2003;
21(3):242-243. 10(2):120-125.
91. Thoma C, Hasselblatt P, Kock J et al. Generation of stable mRNA 106. Friesen WJ, Darby MK. Phage display of RNA binding zinc fingers
fragments and translation of N-truncated proteins induced by from transcription factor IIIA. J Biol Chem 1997;
antisense oligodeoxynucleotides. Mol Cell 2001; 8(4):865-872. 272(17):10994-10997.
92. Hutvagner G, Zamore PD. RNAi: Nature abhors a double-strand. 107. Frankel AD. Fitting peptides into the RNA world. Curr Opin Struct
Curr Opin Genet Dev 2002; 12(2):225-232. Biol 2000; 10(3):332-340.
93. Harris S, Foord SM. Transgenic gene knock-outs: Functional 108. McColl DJ, Honchell CD, Frankel AD. Structure-based design of
genomics and therapeutic target selection. Pharmacogenomics 2000; an RNA-binding zinc finger. Proc Natl Acad Sci USA 1999;
1(4):433-443. 96(17):9521-9526.
94. Kang JS, Kim JS. Zinc finger proteins as designer transcription 109. Friesen WJ, Darby MK. Specific RNA binding by a single C2H2
factors. J Biol Chem 2000; 275(12):8742-8748. zinc finger. J Biol Chem 2001; 276(3):1968-1973.
95. Zhang L, Spratt SK, Liu Q et al. Synthetic zinc finger transcription 110. Frommer WB, Beachy R. Plant biotechnology. A future for plant
factor action at an endogenous chromosomal site. Activation of the biotechnology? Naturally!. Curr Opin Plant Biol 2003; 6(2):147-149.
human erythropoietin gene. J Biol Chem 2000; 275(43):33850-33860. 111. Stege JT, Guan X, Ho T et al. Controlling gene expression in plants
96. Liu PQ, Rebar EJ, Zhang L et al. Regulation of an endogenous using synthetic zinc finger transcription factors. Plant J 2002;
locus using a panel of designed zinc finger proteins targeted to 32(6):1077-1086.
accessible chromatin regions. Activation of vascular endothelial growth 112. Sanchez JP, Ullman C, Moore M et al. Regulation of gene expression
factor A. J Biol Chem 2001; 276(14):11323-11334. in Arabidopsis thaliana by artificial zinc finger chimeras. Plant Cell
97. Ren D, Collingwood TN, Rebar EJ et al. PPARgamma knockdown Physiol 2002; 43(12):1465-1472.
by engineered transcription factors: Exogenous PPARgamma2 but 113. Segal DJ, Stege JT, Barbas CF. Zinc fingers and a green thumb:
not PPARgamma1 reactivates adipogenesis. Genes Dev 2002; Manipulating gene expression in plants. Curr Opin Plant Biol 2003;
16(1):27-32. 6(2):163-168.
98. Rebar EJ, Huang Y, Hickey R et al. Induction of angiogenesis in a 114. Ordiz MI, Barbas III CF, Beachy RN. Regulation of transgene
mouse model using engineered transcription factors. Nat Med 2002; expression in plants with polydactyl zinc finger transcription factors.
8(12):1427-1432. Proc Natl Acad Sci USA 2002; 99(20):13290-13295.
99. Papworth M, Moore M, Isalan M et al. Inhibition of herpes simplex 115. Falke D, Juliano RL. Selective gene regulation with designed
virus 1 gene expression by designer zinc-finger transcription factors. transcription factors: Implications for therapy. Curr Opin Mol Ther
Proc Natl Acad Sci USA 2003; 100(4):1621-1626. 2003; 5(2):161-166.
100. Reynolds L, Ullman C, Moore M et al. Repression of the HIV-1 5'
LTR promoter and inhibition of HIV-1 replication by using
engineered zinc-finger transcription factors. Proc Natl Acad Sci USA
2003; 100(4):1615-1620.
CHAPTER 10
Abstract
T
FIIIA and p43 are multifunctional zinc finger proteins genes encoding the large rRNAs undergo 1000X amplification
that share a specific affinity for 5S ribosomal RNA. In in order to provide the necessary transcriptional capacity. The
this chapter I summarize over 25 years of research that expression of these amplified genes begins during the later,
highlights the similarities and differences in the RNA binding vitellogenic phase of oogenesis and provides the quantity of large
activity of these two proteins. rRNA required for ribosome assembly. The genes encoding the
5S rRNA component do not undergo amplification, but instead
Introduction are completely derepressed and transcribed throughout oogen-
esis. During the previtellogenic stages of oogenesis, when expres-
In the immature oocytes of Xenopus laevis, other amphibia and
sion of the large rRNA genes is low the excess 5S rRNA that
teleost fish, 5S ribosomal RNA (rRNA) is transcribed in large
accumulates is sequestered in the cytoplasm (Fig. 1).2 The accu-
amounts and stored in the cytoplasm as ribonucleoprotein par-
mulated 5S rRNA is stored in two RNPs, one which sediments at
ticles (RNPs). In the case of Xenopus, approximately half of the
7S3 and one which sediments at 42S.4 As the vitellogenic phase
cytoplasmic 5S rRNA is stored in a 7S RNP bound in a 1:1 com-
of oogenesis proceeds, the 5S rRNA is imported to the nucleolus
plex with the protein TFIIIA, while the other half of the cyto-
where it is assembled into the large ribosomal subunit.
plasmic 5S rRNA is stored in a 42S RNP bound in a 1:1 complex
The 7S RNP consists of a bimolecular complex of 5S rRNA
with the protein p43. Although TFIIIA and p43 both bind to 5S
bound to TFIIIA, the 5S rRNA gene-specific transcription fac-
rRNA and each has 9 tandem zinc fingers, there is a remarkable
tor.5 Expression of the TFIIIA gene is under developmental regu-
lack of amino acid sequence homology between the two proteins
lation.5-8 The levels of TFIIIA produced in immature oocytes (ap-
outside of the conserved elements required to fold the zinc finger
proximately 1012 molecules/oocyte) are sufficient to sequester the
domains.
5S rRNA in the cytoplasm. In the very late stages of oogenesis,
In addition to forming a 5S rRNA storage particle, TFIIIA
the stored 5S rRNA is used for ribosome assembly and the re-
also acts as an RNA polymerase III transcription factor by bind-
leased TFIIIA is degraded. After fertilization, the level of mRNA
ing to the internal control region of the 5S rRNA gene. It there-
fore has a specific DNA binding activity as well as a specific RNA
binding activity. The p43 protein has no known DNA binding
activity. In the 42S RNP, p43 interacts with the p48 protein com-
ponent of this particle as well as the 5S rRNA.
One of the questions that has occupied my research lab is “what
are the similarities and differences in the mechanisms by which
TFIIIA and p43 interact with nucleic acids?” Ours has been a
very detailed approach to this question, using quantitative equi-
librium binding assays and targeted mutagenesis of the nucleic
acids and proteins in order to identify critical residues required
for the formation of these specific protein-RNA complexes. In
the case of TFIIIA, we have also been interested in understanding
the similarities and differences in the binding of the protein to
DNA and RNA. Before describing twenty years of research into
these intriguing questions, I’ll begin by describing the biology of
the system we are investigating.
Zinc Finger Proteins: From Atomic Contact to Cellular Function, edited by Shiro Iuchi and Natalie Kuldell.
©2005 Landes Bioscience/Eurekah.com and Kluwer Academic/Plenum Publishers.
TFIIIA and p43: Binding to 5S Ribosomal RNA 57
References
1. Denis H, le Maire M. Thesaurisomes, a novel kind of nucleoprotein
particle. Subcell Biochem 1983; 9:263-297.
2. Ford PJ. Noncoordinated accumulation and synthesis of 5S
ribonucleic acid by ovaries of Xenopus laevis. Nature 1971;
233:561-564.
3. Picard B, Wegnez M. Isolation of a 7S particle from Xenopus laevis
oocytes: A 5S RNA-protein complex. Proc Natl Acad Sci USA 1979;
76:241-245.
4. Denis H, Mairy M. Recherches biochimiques sur l’oogenese. Eur J
Biochem 1972; 25:524-534.
5. Honda BM, Roeder RG. Association of a 5S gene transcription factor
with 5S RNA and altered levels of the factor during cell
differentiation. Cell 1980; 22:119-126.
6. Taylor W, Jackson IJ, Siegel N et al. The developmental expression
of the gene for TFIIIA in Xenopus laevis. Nucl Acids Res 1986;
14:6185-6195.
7. Shastry BS, Honda BM, Roeder RG. Altered levels of a 5S
gene-specific transcription factor (TFIIIA) during oogenesis and
embryonic development of Xenopus laevis. J Biol Chem 1984;
Figure 12. Comparison of the effects of site directed mutations in 5S 259:11373-11382.
rRNA on the binding of TFIIIA and p43. See legend to Figure 3 for 8. Pelham HRB, Wormington WM, Brown DD. Related 5S RNA
details on the symbols used. transcription factors in Xenopus oocytes and somatic cells. Proc Natl
Acad Sci USA 1981; 78:1760-1764.
two proteins. Another open question concerns what aspects of 9. Rollins MB, Del Rio S, Galey AL et al. Role of TFIIIA zinc fingers
protein structure determine which fingers in these proteins par- in vivo - analysis of single-finger function in developing Xenopus
embryos. Mol Cell Biol 1993; 13:4776-4783.
ticipate in binding to 5S rRNA. It is particularly intriguing that
10. Pittman RH, Andrews MT, Setzer DR. A feedback loop coupling
fingers 1 to 3 in TFIIIA can bind specifically to 5S rRNA when 5S rRNA synthesis to accumulation of a ribosomal protein. J Biol
expressed as a peptide, but apparently do not contribute signifi- Chem 1999; 274:33198-33201.
cantly to the RNA binding activity of full length TFIIIA. An- 11. Andrews MT, Brown DD. Transient activation of oocyte 5S RNA
other pressing question is why the TWT motif found in finger 6 genes in Xenopus embryos by raising the level of the trans-acting
of both TFIIIA and p43 is required for 5S rRNA binding only in factor TFIIIA. Cell 1987; 51:445-453.
TFIIIA. The model proposed by David Setzer of unfavorable 12. Mattaj IW, Lienhard S, Zeller R et al. Nuclear exclusion of
finger-finger interactions determining the role of individual zinc transcription factor IIIA and the 42S particle transfer RNA-binding
fingers in nucleic acid binding deserves considerably more study. protein in Xenopus oocytes: A possible mechanism for gene control?
Investigating these questions may lead to important insights into J Cell Biol 1983; 97:1261-1265.
how proteins with more than 3 or 4 zinc fingers form specific 13. Guddat U, Bakken AH, Pieler T. Protein-mediated nuclear export
complexes with one or more ligands. of RNA: 5S rRNA containing small RNPs in Xenopus oocytes. Cell
1990; 60:619-628.
14. Picard B, le Maire M, Wegnez M et al. Biochemical research on
Acknowledgements oogenesis. Composition of the 42-S storage particles of Xenopus
During the time I’ve been involved in studying the RNA– laevis oocytes. Eur J Biochem 1980; 109:359-368.
protein interactions of TFIIIA and p43, I have had the great for- 15. Viel A, Djé MK, Mazabraud A et al. Thesaurin a, the major protein
tune to work with a talented group of graduate students, and I of Xenopus laevis previtellogenic oocytes, present in the 42 S
thank them for their enthusiasm and dedication. Together, we particles, is homologous to elongation factor EF-1 alpha. FEBS Lett
also had the support of a number of excellent research assistants, 1987; 223:232-236.
whose contributions have been recognized by co-authorship on 16. Viel A, Le Maire M, Philippe H et al. Structural and functional
our primary publications. I would like to thank my collaborators properties of thesaurin a (42Sp50), the major protein of the 42-S
for their generous contributions to our joint work: Lizabeth particles present in Xenopus laevis previtellogenic oocytes. J Biol
Chem 1991; 266:10392-10399.
64 Zinc Finger Proteins: From Atomic Contact to Cellular Function
17. Joho KE, Darby MK, Crawford ET et al. A finger protein structurally 37. Baudin F, Romaniuk PJ. A difference in the importance of bulged
similar to TFIIIA that binds exclusively to 5S RNA in Xenopus. nucleotides and their parent base pairs in the binding of transcription
Cell 1990; 61:293-300. factor IIIA to Xenopus 5S RNA and 5S RNA genes. Nucl Acids
18. Kloetzel P-M, Whitfield W, Sommerville J. Analysis and Res 1989; 17:2043-2056.
reconstitution of an RNP particle which stores 5S RNA and tRNA 38. Baudin F, Romaniuk PJ, Romby P et al. Involvement of hinge
in amphibian oocytes. Nucl Acids Res 1981; 9:605-621. nucleotides of Xenopus laevis 5S rRNA in the RNA structural
19. Steitz JA, Berg C, Hendrick JP et al. A 5S rRNA/L5 complex is a organization and in the binding of transcription factor TFIIIA. J
precursor to ribosome assembly in mammalian cells. J Cell Biol 1988; Mol Biol 1991; 218:69-81.
106:545-556. 39. In: Romaniuk PJ, de Stevenson IL, You Q, eds. The specificity of
20. Wormington WM. Developmental expression and 5S rRNA binding the RNA binding activity of Xenopus transcription factor IIIA. New
activity of Xenopus laevis ribosomal protein L5. Mol Cell Biol 1989; York: Alan R. Liss, Inc.; 1989. Cech TR, ed. Molecular Biology of
9:5281-5288. RNA.
21. Allison LA, North MT, Murdoch KJ et al. Structural requirements 40. Veldhoen N, You QM, Setzer DR et al. Contribution of individual
of 5S rRNA for nuclear transport, 7S ribonucleoprotein particle base pairs to the interaction of TFIIIA with the Xenopus 5S RNA
assembly, and 60s ribosomal subunit assembly in Xenopus oocytes. gene. Biochemistry 1994; 33:7568-7575.
Mol Cell Biol 1993; 13:6819-6831. 41. You Q, Romaniuk PJ. The effects of disrupting 5S RNA helical
22. Allison LA, Romaniuk PJ, Bakken AH. RNA-protein interactions structures on the binding of Xenopus transcription factor IIIA. Nucl
of stored 5S RNA with TFIIIA and ribosomal protein L5 during Acids Res 1990; 18:5055-5062.
Xenopus oogenesis. Dev Biol 1991; 144:129-144. 42. You Q, Veldhoen N, Baudin F et al. Mutations in 5S DNA and 5S
23. Allison LA, North MT, Neville LA. Differential binding of RNA have different effects on the binding of Xenopus transcription
oocyte-type and somatic-type 5S rRNA to TFIIIA and ribosomal factor IIIA. Biochemistry 1991; 30:2495-2500.
protein L5 in Xenopus oocytes: Specialization for storage versus 43. Romaniuk PJ. The role of highly conserved single-stranded
mobilization. Dev Biol 1995; 168:284-295. nucleotides of Xenopus 5S RNA in the binding of transcription factor
24. Romaniuk PJ. Characterization of the RNA binding properties of IIIA. Biochemistry 1989; 28:1388-1395.
transcription factor IIIA of Xenopus laevis oocytes. Nucl Acids Res 44. Setzer DR, Menezes SR, Del Rio S et al. Functional interactions
1985; 13:5369-5387. between the zinc fingers of Xenopus transcription factor IIIA during
25. Romaniuk PJ. Characterization of the equilibrium binding of 5S rRNA binding. RNA 1996; 2:1254-1269.
Xenopus transcription factor IIIA to the 5S RNA gene. J Biol Chem 45. Romaniuk PJ, de Stevenson IL, Ehresmann C et al. A comparison
1990; 265:17593-17600. of the solution structures and conformational properties of the
26. Sakonju S, Bogenhagen DF, Brown DD. A control region in the somatic and oocyte 5S rRNAs of Xenopus laevis. Nucl Acids Res
center of the 5S RNA gene directs specific initiation of transcription: 1988; 16(5):2295-2312.
I. The 5' border of the region. Cell 1980; 19:13-25. 46. Westhof E, Romby P, Romaniuk PJ et al. Computer modeling from
27. Sakonju S, Brown DD, Engelke D et al. The binding of a solution data of spinach chloroplast and of Xenopus laevis somatic
transcription factor to deletion mutants of a 5S rRNA gene. Cell and oocyte 5S rRNAs. J Mol Biol 1989; 207:417-431.
1981; 23:665-669. 47. Romby P, Baudin F, Brunel C et al. Ribosomal 5S RNA from
28. Sakonju S, Brown DD. Contact points between a positive Xenopus laevis oocytes: Conformation and interaction with
transcription factor and the Xenopus 5S RNA gene. Cell 1982; transcription factor IIIA. Biochimie 1990; 72:437-452.
31:395-405. 48. Shen Z, Hagerman PJ. Conformation of the central, three-helix
29. Pieler T, Hamm J, Roeder RG. The 5S gene internal control region junction of the 5S ribosomal RNA of Sulfolobus acidocaldarius. J
is composed of three distinct sequence elements, organized as two Mol Biol 1994; 241:415-430.
functional domains with variable spacing. Cell 1987; 48:91-100. 49. Ha T, Zhuang X, Kim HD et al. Ligand-induced conformational
30. Pieler T, Oei S-L, Hamm J et al. Functional domains of the Xenopus changes observed in single RNA molecules. Proc Natl Acad Sci USA
laevis 5S gene promoter. EMBO J 1985; 4:3751-3756. 1999; 96:9077-9082.
31. Romaniuk PJ, de Stevenson IL, Wong H-HA. Defining the binding 50. de Stevenson IL, Romby P, Baudin F et al. Structural studies on
site of Xenopus transcription factor IIIA on 5S RNA using truncated site-directed mutants of domain 3 of Xenopus laevis oocyte 5S
and chimeric 5S RNA molecules. Nucl Acids Res 1987; ribosomal RNA. J Mol Biol 1991; 219:243-255.
15:2737-2755. 51. Theunissen O, Rudt F, Pieler T. Structural determinants in 5S RNA
32. Darsillo P, Huber PW. The use of chemical nucleases to analyze and TFIIIA for 7S RNP formation. Eur J Biochem 1998;
RNA-protein interactions - the TFIIIA-5S rRNA complex. J Biol 258:758-767.
Chem 1991; 266:21075-21082. 52. Rawlings SL, Matt GD, Huber PW. Analysis of the binding of
33. Andersen J, Delihas N, Hanas JS et al. 5S RNA structure and Xenopus transcription factor IIIA to oocyte 5 S rRNA and to the 5
interaction with transcription factor A. 2. Ribonuclease probe of the S rRNA gene. J Biol Chem 1996; 271:869-877.
7S RNP from Xenopus laevis immature oocytes and RNA exchange 53. Wimberly B, Varani G, Tinoco I. The conformation of loop E of
properties of the 7S particle. Biochemistry 1984; 23:5759-5766. eukaryotic 5S Ribosomal RNA. Biochemistry 1993; 32:1078-1087.
34. Andersen J, Delihas N, Hanas JS et al. 5S RNA structure and 54. Sands MS, Bogenhagen DF. TFIIIA binds to different domains of
interaction with transcription factor A. 1. Ribonuclease probe of the 5S RNA and the Xenopus borealis 5S RNA gene. Mol Cell Biol
7S particle from Xenopus laevis immature oocytes and RNA exchange 1987; 7:3985-3993.
properties of the 7S particle. Biochemistry 1984; 23:5752-5759. 55. Wuttke DS, Foster MP, Case DA et al. Solution structure of the
35. Huber PW, Wool IG. Identification of the binding site on 5S rRNA first three zinc fingers of TFIIIA bound to the cognate DNA
for the transcription factor IIIA: Proposed structure of a common sequence: Determinants of affinity and sequence specificity. J Mol
binding site on 5S rRNA and on the gene. Proc Natl Acad Sci USA Biol 1997; 273:183-206.
1986; 83:1593-1597. 56. Nolte RT, Conlin RM, Harrison SC et al. Differing roles for zinc
36. Christiansen J, Brown RS, Sproat BS et al. Xenopus transcription fingers in DNA recognition: Structure of a six finger transcription
factor IIIA binds primarily at junctions between double helical stems factor IIIA complex. Proc Natl Acad Sci USA 1998; 95:2938-2943.
and internal loops in oocyte 5S RNA. EMBO J 1987; 6:453-460. 57. Vrana KE, Churchill MEA, Tullius TD et al. Mapping functional
regions of transcription factor TFIIIA. Mol Cell Biol 1988;
8:1684-1696.
TFIIIA and p43: Binding to 5S Ribosomal RNA 65
58. Fiser-Littell RM, Duke AL, Yanchick JS et al. Deletion of the 70. Del Rio S, Menezes SR, Setzer DR. The function of individual zinc
N-terminal region of Xenopus transcription factor IIIA inhibits fingers in sequence specific DNA recognition by transcription factor
specific binding to the 5S RNA gene. J Biol Chem 1988; IIIA. J Mol Biol 1993; 233:567-579.
263:1607-1610. 71. Kehres DG, Subramanyan GS, Hung VS et al. Energetically
59. Theunissen O, Rudt F, Guddat U et al. RNA and DNA binding unfavorable interactions among the zinc fingers of transcription factor
zinc fingers in xenopus TFIIIA. Cell 1992; 71:679-690. IIIA when bound to the 5S rRNA gene. J Biol Chem 1997;
60. Clemens KR, Liao XB, Wolf V et al. Definition of the binding sites 272:20152-20161.
of individual zinc fingers in the transcription factor IIIA-5S RNA 72. Zang WQ, Veldhoen N, Romaniuk PJ. Effects of zinc finger
gene complex. Proc Natl Acad Sci USA 1992; 89:10822-10826. mutations on the nucleic acid binding activities of Xenopus
61. Clemens KR, Wolf V, McBryant SJ et al. Molecular basis for specific transcription factor IIIA. Biochemistry 1995; 34:15545-15552.
recognition of both RNA and DNA by a zinc finger protein. Science 73. Bogenhagen DF. Proteolytic footprinting of transcription factor
1993; 260:530-533. TFIIIA reveals different tightly binding sites for 5S RNA and 5S
62. Clemens KR, Zhang PH, Liao XB et al. Relative contributions of DNA. Mol Cell Biol 1993; 13:5149-5158.
the zinc fingers of transcription factor IIIA to the energetics of DNA 74. McBryant SJ, Veldhoen N, Gedulin B et al. Interaction of the RNA
binding. J Mol Biol 1994; 244:23-35. binding fingers of Xenopus transcription factor IIIA with specific
63. Darby MK, Joho KE. Differential binding of zinc fingers from regions of 5S ribosomal RNA. J Mol Biol 1995; 248:44-57.
Xenopus TFIIIA and p43 to 5S RNA and the 5S RNA Gene. Mol 75. Friesen WJ, Darby MK. Phage display of RNA binding zinc fingers
Cell Biol 1992; 12:3155-3164. from transcription factor IIIA. J Biol Chem 1997; 272:10994-10997.
64. Liao XB, Clemens KR, Tennant L et al. Specific interaction of the 76. Searles MA, Lu D, Klug A. The role of the central zinc fingers of
first three zinc fingers of TFIIIA with the internal control region of transcription factor IIIA in binding to 5S RNA. J Mol Biol 2000;
the Xenopus 5S RNA gene. J Mol Biol 1992; 223:857-871. 301:47-60.
65. Liggins JR, Privalov PL. Energetics of the specific binding interaction 77. Ryan RF, Darby MK. The role of zinc finger linkers in p43 and
of the first three zinc fingers of the transcription factor TFIIIA with TFIIIA binding to 5S rRNA and DNA. Nucl Acids Res 1998;
its cognate DNA sequence. Proteins 2000; Suppl 4:50-62. 26:703-709.
66. Hansen PK, Christensen JH, Nyborg J et al. Dissection of the DNA 78. Hamilton TB, Turner J, Barilla K et al. Contribution of individual
binding domain of Xenopus laevis TFIIIA: Quantitative DNase I amino acids to the nucleic acid binding activities of the Xenopus
footprinting analysis of specific complexes between a 5S RNA gene zinc finger proteins TFIIIA and p43. Biochemistry 2001;
fragment and N-terminal fragments of TFIIIA containing 3 zinc 40:6093-6101.
finger, 4 zinc finger or 5 zinc finger domains. J Mol Biol 1993; 79. Giel-Pietraszuk M, Barciszewska MZ. Additivity of interactions of
233:191-202. zinc finger motifs in specific recognition of RNA. J Biochem 2002;
131:571-578.
67. Churchill MEA, Tullius TD, Klug A. Mode of interaction of the
80. Neely LS, Lee BM, Xu J et al. Identification of a minimal domain
zinc finger protein TFIIIA with a 5S RNA gene of Xenopus. Proc of 5S ribosomal RNA sufficient for high affinity interactions with
Natl Acad Sci USA 1990; 87:5528-5532. the RNA-specific zinc fingers of transcription factor IIIA. J Mol
68. Hayes JJ, Clemens KR. Locations of contacts between individual Biol 1999; 291:549-560.
zinc fingers of Xenopus laevis transcription factor IIIA and the 81. Sands MS, Bogenhagen DF. Two zinc finger proteins from Xenopus
internal control region of a 5S RNA gene. Biochemistry 1992; laevis bind the same region of 5S RNA But with different nuclease
31:11600-11605. protection patterns. Nucl Acids Res 1991; 19:1797-1803.
69. Hayes JJ, Tullius TD. Structure of the TFIIIA-5S DNA complex. J 82. Zang WQ, Romaniuk PJ. Characterization of the 5S RNA binding
Mol Biol 1992; 227:407-417. activity of Xenopus zinc finger protein p43. J Mol Biol 1995;
245:549-558.
CHAPTER 11
Abstract
founded on these synthetic proteins.11,12 RNA-zinc finger com-
R
NA-protein complexes have important functions in gene
expression and regulation. Zinc fingers of the plexes are yet to be solved at an atomic level, yet some of the
Cys-Cys-His-His (C2H2) class that bind RNA do so via features characteristic of DNA-zinc finger recognition appear to
contacts with amino acid side chains in the α-helical portion of apply to the interaction of zinc fingers with RNA and have been
the zinc finger, similar to their interaction with DNA. In gen- successfully exploited to isolate RNA binding zinc fingers with
eral, two or more tandem zinc fingers are present in naturally novel sequence specificity.13 Like their DNA binding counter-
occurring zinc finger proteins. In vitro selection and recombina- parts, therapeutic or agricultural applications for such
tion techniques have isolated single zinc fingers that bind com- RNA-binding proteins, particularly as anti-viral agents, are envi-
plex RNA structures with high affinity and specificity. These zinc sioned. However, there is unlikely to be a straightforward parallel
fingers may ultimately find use as pharmaceutical or agricultural between the nature of zinc finger interaction with DNA and RNA,
agents designed specifically to modify the function of cellular or since there are fundamental differences in structure that appear
viral RNA. to influence the principles governing protein binding to RNA
rather than DNA. DNA is frequently bound by protein struc-
Introduction tures, including the α helix of C2H2 zinc fingers, that recognize
the unique pattern of hydrogen bond donors and acceptors found
RNA is key to gene expression. It serves important functions
within the major groove of the DNA chain.14 In contrast, the
as a template, catalyst and amino-acid acceptor for protein syn-
major groove of duplex RNA is too narrow to accommodate an α
thesis, a cofactor for mRNA splicing, a sequence guide for RNA
helix and in addition, the rich tertiary structure of RNA contrib-
editing and modification, a structural component for the ribo-
utes significantly to the formation of protein-RNA complex. Fur-
some and spliceosome, and also serves as the genome of many
thermore, most RNA complexes upon formation exhibit substan-
plant and animal viruses.1 In all of these varied functions, com-
tial conformational changes in both RNA and protein (Fig. 1).15,16
plexes of RNA with proteins are essential. The structures of pro-
This chapter explores efforts to develop synthetic RNA bind-
teins that bind RNA are as diverse as the functions of RNA, but
ing C2H2 zinc fingers that bind specific RNAs, from early obser-
can be classified through a few characteristic protein motifs. The
vations that determined the role of zinc fingers in RNA binding,
most common of these are the RNA recognition motif (RRM),2
to contemporary phage display approaches for constructing zinc
the KH domain,3 the dsRNA binding domain,4 the arginine-rich
fingers for developing future applications.
motif,5 and three of the major classes of zinc finger.6 These zinc
finger classes are distinguished structurally by the arrangement of
cysteine and histidine zinc ligands, and with the examples known RNA Binding C2H2 Zinc Finger Proteins
at present, also fall into functional classes based on their predomi- The RNA binding properties of C2H2 zinc fingers have been
nant roles in living cells (see Table 1 for descriptions). The first known since the identification of the repeated zinc finger sequence,
class, C2H2 zinc fingers, are by far the most abundant nucleic yet the number of C2H2 zinc finger proteins shown to bind RNA
acid binding motif in the human genome, and have the unique since that time remains small in comparison to the number shown
ability among zinc finger proteins to mediate binding to both to bind DNA (Table 1). Why are there so many confirmed DNA
RNA and DNA. A second class, the CCHC zinc fingers of binding C2H2 zinc fingers and so few RNA specific? The answer
retroviral nucleocapsids bind and package genomic RNA through may lie in the lack of an identifying sequence characteristic other
tandem zinc fingers or “knuckles”, and the third class, CCCH than the zinc finger consensus. A search of the PIR annotated
zinc fingers, are present in proteins that regulate mRNA stability protein sequence database identified more than 4000 protein
through AU-rich elements. matches to a consensus C 2 H 2 zinc finger sequence
The function of C2H2 zinc finger proteins in DNA binding (Cys-X2-5-Cys-X12-His-X3-6-His) and approximately one third of
has been illuminated by detailed structural studies of DNA-zinc these were connected with the amino acids TGEKP (TGERP) or
finger complexes at the atomic level, and subsequent application Krüppel-like linkers,17 characteristic of DNA binding zinc fin-
of the principles of protein-DNA interaction learned from these gers. Given that about half of the currently confirmed DNA bind-
structures has led to the design of specific DNA binding proteins ing C2H2 proteins have the Krüppel linker, up to two-thirds of
and transcription factors.7-10 Hopes for future gene therapies, the consensus matches may be DNA binding proteins. The re-
based on the controlled manipulation of gene expression, are maining third may well be RNA binding proteins that await
Zinc Finger Proteins: From Atomic Contact to Cellular Function, edited by Shiro Iuchi and Natalie Kuldell.
©2005 Landes Bioscience/Eurekah.com and Kluwer Academic/Plenum Publishers.
RNA Binding by Single Zinc Fingers 67
C2H2 TFIIIA 9 zinc fingers, binds 5S RNA (7S RNP) and 5S RNA gene. Transcription factor.
Cys-X3-5-Cys-X12-His-X4-6-His P43 9 zinc fingers, binds 5S RNA isolated from 42S RNP in Xenopus oocytes
WT1 4 zinc fingers, +KTS isoform associated with splicing factors.
Transcription factor.
Wig-1 3 zinc fingers, p53 induced, localized to nucleoli
dsRBP-Zfa 7 zinc fingers, binds dsRNA through 3 N-terminal fingers or 3 C-terminal zinc
fingers
JAZ 4 spaced zinc fingers, binds dsRNA or DNA/RNA hybrids.
MOK2 2 zinc fingers, binds RNA and DNA (brain and testis)
NUFIP 1 zinc finger associates with FMR RNA binding protein
XFO-5, XFG-6 5 – 15 zinc fingers, members of FAR family proteins
ZNF74 12 zinc fingers, deleted in DiGeorge syndrome, binds poly(U)
PEP 4 zinc fingers, co-precipitates in complex including hnRNPK
SL100 18 zinc fingers, binds to SLBP/SL RNA complex using fingers 2-8
Xfin 31 zinc fingers, binds poly(G) preferentially
discovery. Perhaps more likely, the conformational flexibility of factor.21 In the Xenopus oocyte, where many proteins and RNAs
RNA may account for the lack of consensus amino acid sequences are stockpiled to fuel the rapid cycles of cell division characteris-
within zinc fingers that define an RNA-binding zinc finger and tic of early embryonic development, TFIIIA was found in abun-
there may not be a sequence motif, but rather a structural motif. dance complexed with 5S rRNA in a 7S RNP particle.20,22 Thus
Of the C2H2 proteins listed in Table 1 only TFIIIA, p43 and the notion of the zinc finger protein as an RNA binding protein
WT1 exhibit sequence specific interaction with RNA. RNA se- was established early in the infancy of zinc finger biochemistry.
lection experiments in vitro such as SELEX have been used to The role of the zinc fingers in DNA binding was established later
isolate specific RNA aptamers that bind to zinc finger proteins.18,19 through deletion mutagenesis23 and their involvement in RNA
However, given the extraordinary power of RNA aptamers to dis- binding was suggested (retrospectively) by a residual N-terminal
tinguish even small chemical modifications, it is not surprising 20 kD RNA bound proteolytic fragment, including the zinc fin-
that RNAs specific for zinc fingers can be isolated in this manner. ger region, of the 7S RNP.24
Support for RNA binding by C2H2 zinc fingers in some cases is Identification of a TFIIIA-related protein from another oo-
cyte storage particle further implicated zinc fingers in RNA bind-
indirect and inferred from subcellular location in the cytoplasm
ing.25 The 42S ribonucleoprotein particle accumulates in imma-
with splicing complexes or association with other RNA binding
ture Xenopus oocytes prior to ribosome biogenesis. This particle
proteins. comprises two proteins, p43 and p50, together with 5S rRNA
and amino acyl-tRNAs.26 Similar to TFIIIA, p43 is complexed
TFIIIA and p43 with 5S rRNA in a 7S sub-particle, but unlike TFIIIA, p43 fails
TFIIIA and p43 are the clearest examples of sequence-specific to stimulate 5S RNA gene transcription or to bind DNA, sug-
C2H2 RNA-binding zinc finger proteins. The archetypal C2H2 gesting that it interacts only with 5S RNA.25 Thus, the role of the
zinc finger protein, TFIIIA, was demonstrated to be an RNA C2H2 zinc finger as a motif that could interact exclusively with
binding protein,20 soon after its identification as a transcription RNA was firmly established.
68 Zinc Finger Proteins: From Atomic Contact to Cellular Function
Figure 1. Comparison of RNA and DNA structure. Duplex forms of Phage Display of RNA Binding Zinc Fingers
RNA (pdb: 1SRA) and DNA (pdb: 1BNA) are shown with backbone
The role of individual amino acids in RNA binding by zinc
phosphodiester chains in gray and white. The major groove (M) of the
fingers has been sought through a variety of mutagenesis experi-
RNA duplex is deeper and narrower than that of DNA and prohibits
ments in which individual amino acids or groups thereof have
access by a protein α-helix unless distorted. In contrast, the minor groove
been changed from wild type. Displaying a library of mutated
(m) is accessible, but possible H-bond contacts limit sequence distinc-
zinc fingers on the surface of bacteriophage screens the effects of
tion to AT or GC base-pairs.
many amino acid changes simultaneously. Using this approach
important amino acid contacts between RNA and TFIIIA zinc
RNA Binding by Sub-Groups of Zinc Fingers finger 4 were identified and the utility of a phage display ap-
in Multi-Finger Proteins proach was established.37
Much of our understanding of the interaction of C2H2 zinc Phage display presents multiple copies of a single mutant on
fingers with RNA derives from deletion and site-directed mu- the surface of a filamentous phage such as fd or M13 (Fig. 3).38
tagenesis experiments and reveals that subgroups of zinc fingers Binding of a phage library to an RNA substrate (selector RNA) is
within multi-finger proteins can perform independent functions carried out in solution and complexes formed between phage and
such as RNA or DNA binding, although not always with affinity biotinylated RNA are recovered by affinity purification on
comparable to the intact protein (Fig. 2).27 The close structural streptavidin-coated microtiter plates. Phage that do not form com-
relationship between TFIIIA and p43, in contrast to their differ- plexes with RNA are removed by washing under conditions of
ing nucleic acid binding properties, has made this pair of pro- varied stringency. In principle, a single round of affinity selection
teins a favorite choice for comparative structure-function stud- could yield specific RNA-binding phage. However, in practice,
ies.25,27-31 By analyzing groups of three or four zinc fingers from initial selections are impure and the phage are amplified in bacte-
ria after each round of selection and then in an iterative process
TFIIIA, DNA binding activity was localized principally to the
reselected against the same RNA. Successful isolation of
amino terminal zinc fingers, specifically fingers 1-3 (Fig. 2 light
RNA-binding phage is judged by the fraction of input phage re-
gray boxes).29,32,33 These zinc fingers in TFIIIA are connected by
covered from each round of selection. DNA sequencing of recov-
the Krüppel linkers and this linker sequence is notably absent ered phage is used to reveal the amino acid sequence of the zinc
from p43, which binds only RNA. However, conversion of p43 fingers that bind the selector RNA.
linkers to those in TFIIIA does not confer DNA binding activity, Alanine scanning mutagenesis of each of the central zinc fin-
suggesting that the Krüppel linker sequence, while indicative of a gers of TFIIIA (fingers 4 to 7), identified zinc fingers 4 and 6 as
DNA binding protein, is insufficient.29,31,34 Central zinc fingers important RNA binding fingers. In this experiment, four amino
from TFIIIA (fingers 4 to 7) and amino terminal zinc fingers acid positions previously shown to make contact with DNA in
from p43 have the highest RNA affinity of isolated zinc finger crystal structures -1, +2, +3 and +6 with respect to the
fragments (Fig. 2, blue boxes). 29,32,33,35,36 Single finger amino-terminal end of the α helix, were substituted by alanine.
alanine-scanning revealed an important role for zinc fingers 4 Since alanine at these positions does not support RNA binding,
and 6 of TFIIIA when expressed as a four finger fragment (fin- screening all amino acid substitutions at these positions through
gers 4 to 7).37 Finger 6 of both TFIIIA and p43 contains a con- phage display reveals any amino acid combinations that can sup-
served amino acid triplet, Thr-Trp-Thr, which plays a significant port 5S RNA binding (Fig. 4). A cDNA library representing all
role for TFIIIA interaction with 5S rRNA, but not for p43.27 It is possible permutations of amino acids at these four positions in
worth noting that single zinc fingers derived from either protein TFIIIA zinc finger 4 was constructed using degenerate oligonucle-
have no measurable affinity for either nucleic acid and that pro- otides and cloned into a phage display vector that expressed the
teins containing a single finger are a rarity within any class of zinc TFIIIA zinc fingers 4 through 7 as an N-terminal fusion of fd
finger protein found in nature. coat protein III. Phage recovery of 3% of input was observed
RNA Binding by Single Zinc Fingers 69
Acknowledgements
I would like to acknowledge the important contributions of
Drs. Westley Friesen, Robert Ryan and Michele Shields for the
work done in my laboratory. Dr. Markus Germann determined
the model for a single zinc finger. My laboratory was supported
by grants from the National Science Foundation (MCB-9206873)
Figure 10. Rev binding helix embedded in a zinc finger framework. A) and the National Institutes of Health (GM50846 and GM/
An alignment of Rev RNA recognition helix with zinc finger 2 of the AI47459).
DNA binding protein zif268. Amino acids essential for zinc finger for-
mation are indicated in bold type, while those important for Rev binding References
to RNA are given in bold-italic. The zinc finger recognition helix is 1. Simons RW, Grunberg-Manago M. RNA structure and function.
boxed. B) Overlap between structures for zif268 (white, pdb 1ZAA) and Plainview. NY: Cold Spring Harbor Laboratory Press, 1998.
Rev peptide (gray, pdb 1ETF). C) RRE-IIB gel shift with indicated 2. Query CC, Bentley RC, Keene JD. A common RNA recognition
concentrations of ZF2-Rev peptide. Adapted with permission from ref- motif identified within a defined U1 RNA binding domain of the
70K U1 snRNP protein. Cell Apr 7 1989; 57(1):89-101.
erence 62. National Academy of Sciences, U.S.A. © 1999. 3. Siomi H, Matunis MJ, Michael WM et al. The premRNA binding
K protein contains a novel evolutionarily conserved motif. Nucleic
from either a known protein or one that has been selected against Acids Res 1993; 21(5):1193-1198.
the RNA target. This ‘founder’ zinc finger is ligated to a random- 4. St Johnston D, Brown NH, Gall JG et al. A conserved
ized zinc finger and selected again against the target RNA. The double-stranded RNA-binding domain. Proc Natl Acad Sci USA
two finger proteins recovered from this selection are ligated to a 1992; 89(22):10979-10983.
5. Calnan BJ, Tidor B, Biancalana S et al. Arginine-mediated RNA
randomized finger and so on. In this manner a multi-finger pro- recognition: The arginine fork. Science 1991; 252(5010):1167-1171.
tein that binds large RNA targets could be constructed. A similar 6. Laity JH, Lee BM, Wright PE. Zinc finger proteins: New insights
approach for DNA binding protein design has been described.64 into structural and functional diversity. Curr Opin Struct Biol 2001;
11(1):39-46.
7. Pavletich NP, Pabo CO. Zinc finger-DNA recognition: Crystal
CCCH Zinc Fingers: Recognition of Short Motifs structure of a Zif268-DNA complex at 2.1 A. Science 1991;
Other finger structures may be more versatile for construction 252(5007):809-817.
of RNA binding proteins, most notably the CCHC nucleocapsid 8. Rebar EJ, Pabo CO. Zinc finger phage: Affinity selection of fingers
zinc fingers and CCCH zinc finger of ARE-binding proteins. with new DNA-binding specificities. Science 1994;
The latter is of particular interest, because ARE-binding proteins 263(5147):671-673.
recognize very short nucleotide sequences. Tristetraprolin (TTP/ 9. Choo Y, Klug A. Physical basis of a protein-DNA recognition code.
Curr Opin Struct Biol 1997; 7(1):117-125.
NUP475/GOS24) binds a short RNA sequence UUAUUUAUU, 10. Choo Y, Sanchez-Garcia I, Klug A. In vivo repression by a site-specific
that is the core of the AU-rich element (ARE) found in the 3' DNA-binding protein designed against an oncogenic sequence.
untranslated region of a number of mRNAs, and leads to the Nature 1994; 372(6507):642-645.
degradation of the mRNA (see Chapter 13).65 High affinity RNA 11. Jamieson AC, Miller JC, Pabo CO. Drug discovery with engineered
binding (dissociation constant 19 nM for a 24 residue RNA) is zinc-finger proteins. Nat Rev Drug Discov 2003; 2(5):361-368.
mediated through tandem CCCH zinc fingers and a single finger 12. Ordiz MI, Barbas 3rd CF, Beachy RN. Regulation of transgene
expression in plants with polydactyl zinc finger transcription factors.
is unable to bind RNA in a gel-shift assay.66 However, using a Proc Natl Acad Sci USA 2002; 99(20):13290-13295.
more sensitive assay, weak binding by a single CCCH zinc finger 13. Friesen WJ, Darby MK. Specific RNA binding proteins constructed
to the subsequence, UUUAUUU, can be detected (dissociation from zinc fingers. Nat Struct Biol 1998; 5(7):543-546.
constant of 5 µM).67 An NMR spectrum has recently been ac- 14. Steitz T. Similarities and Differences between RNA and DNA
quired for a peptide representing the tandem zinc fingers and Recognition by Proteins. In: Gesteland R, Atkins J, eds. The RNA
World. Vol Monograph 24. Cold Spring Harbor: CSHL Press,
1993:219-237.
74 Zinc Finger Proteins: From Atomic Contact to Cellular Function
15. Leulliot N, Varani G. Current topics in RNA-protein recognition: 39. Nardelli J, Gibson TJ, Vesque C et al. Base sequence discrimination
Control of specificity and biological function through induced fit by zinc-finger DNA-binding domains. Nature 1991;
and conformational capture. Biochemistry 2001; 40(27):7947-7956. 349(6305):175-178.
16. Williamson JR. Induced fit in RNA-protein recognition. Nat Struct 40. Wolfe SA, Greisman HA, Ramm EI et al. Analysis of zinc fingers
Biol Oct 2000;7(10):834-837. optimized via phage display: Evaluating the utility of a recognition
17. Schuh R, Aicher W, Gaul U et al. A conserved family of nuclear code. J Mol Biol 1999; 285(5):1917-1934.
proteins containing structural elements of the finger protein encoded 41. Pabo CO, Peisach E, Grant RA. Design and selection of novel
by Kruppel, a Drosophila segmentation gene. Cell 1986; Cys2His2 zinc finger proteins. Annu Rev Biochem 2001; 70:313-340.
47(6):1025-1032. 42. Liu Q, Segal DJ, Ghiara JB et al. Design of polydactyl zinc-finger
18. Bardeesy N, Pelletier J. Overlapping RNA and DNA binding proteins for unique addressing within complex genomes. Proc Natl
domains of the wt1 tumor suppressor gene product. Nucleic Acids Acad Sci USA 1997; 94(11):5525-5530.
Res 1998; 26(7):1784-1792. 43. Christiansen J, Brown RS, Sproat BS et al. Xenopus transcription
19. Zhai G, Iskandar M, Barilla K et al. Characterization of RNA factor IIIA binds primarily at junctions between double helical stems
aptamer binding by the Wilms’ tumor suppressor protein WT1. and internal loops in oocyte 5S RNA. Embo J 1987; 6(2):453-460.
Biochemistry 2001; 40(7):2032-2040. 44. Pieler T, Erdmann VA, Appel B. Structural requirements for the
20. Pelham HR, Brown DD. A specific transcription factor that can interaction of 5S rRNA with the eukaryotic transcription factor IIIA.
bind either the 5S RNA gene or 5S RNA. Proc Natl Acad Sci USA Nucleic Acids Res 1984; 12(22):8393-8406.
1980; 77(7):4170-4174. 45. Theunissen O, Rudt F, Pieler T. Structural determinants in 5S RNA
21. Engelke DR, Ng SY, Shastry BS et al. Specific interaction of a and TFIIIA for 7S RNP formation. Eur J Biochem 1998;
purified transcription factor with an internal control region of 5S 258(2):758-767.
RNA genes. Cell 1980; 19(3):717-728. 46. Setzer DR, Menezes SR, Del Rio S et al. Functional interactions
22. Picard B, Wegnez M. Isolation of a 7S particle from Xenopus laevis between the zinc fingers of Xenopus transcription factor IIIA during
oocytes: A 5S RNA-protein complex. Proc Natl Acad Sci USA 1979; 5S rRNA binding. RNA 1996; 2(12):1254-1269.
76(1):241-245. 47. Darsillo P, Huber PW. The use of chemical nucleases to analyze
23. Vrana KE, Churchill ME, Tullius TD et al. Mapping functional RNA-protein interactions. The TFIIIA-5 S rRNA complex. J Biol
regions of transcription factor TFIIIA. Mol Cell Biol 1988; Chem 1991; 266(31):21075-21082.
8(4):1684-1696. 48. Kjems J, Calnan BJ, Frankel AD et al. Specific binding of a basic
24. Smith DR, Jackson IJ, Brown DD. Domains of the positive peptide from HIV-1 Rev. EMBO Journal 1992; 11(3):1119-1129.
transcription factor specific for the Xenopus 5S RNA gene. Cell 49. Pollard VW, Malim MH. The HIV-1 Rev protein. Annu Rev
1984;37(2):645-652. Microbiol 1998;52:491-532.
25. Joho KE, Darby MK, Crawford ET et al. A finger protein structurally 50. Sodroski J, Goh WC, Rosen C et al. A second post-transcriptional
similar to TFIIIA that binds exclusively to 5S RNA in Xenopus. trans-activator gene required for HTLV- III replication. Nature 1986;
Cell 1990;61(2):293-300. 321(6068):412-417.
26. Picard B, le Maire M, Wegnez M et al. Biochemical Research on 51. Hope T, Pomerantz RJ. The human immunodeficiency virus type 1
oogenesis. Composition of the 42-S storage particles of Xenopus Rev protein: A pivotal protein in the viral life cycle. Curr Top
laevix oocytes. Eur J Biochem 1980; 109(2):359-368. Microbiol Immunol 1995; 193:91-105.
27. Hamilton TB, Turner J, Barilla K et al. Contribution of individual 52. Malim MH, Cullen BR. HIV-1 structural gene expression requires
amino acids to the nucleic acid binding activities of the Xenopus the binding of multiple Rev monomers to the viral RRE: Implications
zinc finger proteins TFIIIIA and p43. Biochemistry 2001; for HIV-1 latency. Cell 1991; 65(2):241-248.
40(20):6093-6101. 53. Pasquinelli AE, Ernst RK, Lund E et al. The constitutive transport
28. Sands MS, Bogenhagen DF. Two zinc finger proteins from Xenopus element (CTE) of Mason-Pfizer monkey virus (MPMV) accesses a
laevis bind the same region of 5S RNA but with different nuclease cellular mRNA export pathway. Embo J 1997; 16(24):7500-7510.
protection patterns. Nucleic Acids Res 1991; 19(8):1797-1803. 54. Tan R, Chen L, Buettner JA et al. RNA recognition by an isolated
29. Darby MK, Joho KE. Differential binding of zinc fingers from alpha helix. Cell 1993; 73(5):1031-1040.
Xenopus TFIIIA and p43 to 5S RNA and the 5S RNA gene. Mol 55. Wuttke DS, Foster MP, Case DA et al. Solution structure of the
Cell Biol 1992; 12(7):3155-3164. first three zinc fingers of TFIIIA bound to the cognate DNA
30. Zang WQ, Romaniuk PJ. Characterization of the 5 S RNA binding sequence: Determinants of affinity and sequence specificity. J Mol
activity of Xenopus zinc finger protein p43. J Mol Biol 1995; Biol 1997; 273(1):183-206.
245(5):549-558. 56. Stemmer WP. DNA shuffling by random fragmentation and
31. Ryan RF, Darby MK. The role of zinc finger linkers in p43 and reassembly: In vitro recombination for molecular evolution. Proc Natl
TFIIIA binding to 5S rRNA and DNA. Nucleic Acids Res 1998; Acad Sci USA 1994; 91(22):10747-10751.
26(3):703-709. 57. Stemmer WP. Rapid evolution of a protein in vitro by DNA
32. Theunissen O, Rudt F, Guddat U et al. RNA and DNA binding shuffling. Nature 1994; 370(6488):389-391.
zinc fingers in Xenopus TFIIIA. Cell 1992;71(4):679-690. 58. Friesen WJ, Darby MK. Specific RNA binding by a single C2H2
33. Clemens KR, Wolf V, McBryant SJ et al. Molecular basis for specific zinc finger. J Biol Chem 2001;276(3):1968-1973.
recognition of both RNA and DNA by a zinc finger protein. Science 59. Southgate C, Zapp ML, Green MR. Activation of transcription by
1993; 260(5107):530-533. HIV-1 Tat protein tethered to nascent RNA through another protein.
34. Choo Y, Klug A. A role in DNA binding for the linker sequences Nature 1990; 345(6276):640-642.
of the first three zinc fingers of TFIIIA. Nucleic Acids Res 1993; 60. Brigati C, Giacca M, Noonan DM et al. HIV Tat, its TARgets and
21(15):3341-3346. the control of viral gene expression. FEMS Microbiol Lett 2003;
35. Searles MA, Lu D, Klug A. The role of the central zinc fingers of 220(1):57-65.
transcription factor IIIA in binding to 5 S RNA. J Mol Biol 61. Darby MK, Germann MW. Unpublished.
2000;301(1):47-60. 62. McColl DJ, Honchell CD, Frankel AD. Structure-based design of
36. Neely LS, Lee BM, Xu J et al. Identification of a minimal domain an RNA-binding zinc finger. Proc Natl Acad Sci USA 1999;
of 5 S ribosomal RNA sufficient for high affinity interactions with 96(17):9521-9526.
the RNA-specific zinc fingers of transcription factor IIIA. J Mol 63. Blancafort P, Steinberg SV, Paquin B et al. The recognition of a
Biol 1999; 291(3):549-560. noncanonical RNA base pair by a zinc finger protein. Chem Biol
37. Friesen WJ, Darby MK. Phage display of RNA binding zinc fingers 1999; 6(8):585-597.
from transcription factor IIIA. J Biol Chem 1997; 64. Greisman HA, Pabo CO. A general strategy for selecting high-affinity
272(17):10994-10997. zinc finger proteins for diverse DNA target sites. Science 1997;
38. Smith GP, Scott JK. Libraries of peptides and proteins displayed on 275(5300):657-661.
filamentous phage. Methods Enzymol 1993; 217:228-257.
RNA Binding by Single Zinc Fingers 75
65. Blackshear PJ. Tristetraprolin and other CCCH tandem zinc-finger 78. Grondin B, Bazinet M, Aubry M. The KRAB zinc finger gene
proteins in the regulation of mRNA turnover. Biochem Soc Trans ZNF74 encodes an RNA-binding protein tightly associated with the
2002; 30(Pt 6):945-952. nuclear matrix. J Biol Chem 1996; 271(26):15458-15467.
66. Lai WS, Carballo E, Strum JR et al. Evidence that tristetraprolin 79. Amero SA, Elgin SC, Beyer AL. A unique zinc finger protein is
binds to AU-rich elements and promotes the deadenylation and associated preferentially with active ecdysone-responsive loci in
destabilization of tumor necrosis factor alpha mRNA. Mol Cell Biol Drosophila. Genes Dev 1991; 5(2):188-200.
1999; 19(6):4311-4323. 80. Hovemann BT, Reim I, Werner S et al. The protein Hrb57A of
67. Michel SL, Guerrerio AL, Berg JM. Selective RNA binding by a Drosophila melanogaster closely related to hnRNP K from vertebrates
single CCCH zinc-binding domain from Nup475 (Tristetraprolin). is present at sites active in transcription and coprecipitates with four
Biochemistry 2003; 42(16):4626-4630. RNA-binding proteins. Gene 2000; 245(1):127-137.
68. Blackshear PJ, Lai WS, Kennington EA et al. Characteristics of the 81. Dominski Z, Erkmann JA, Yang X et al. A novel zinc finger protein
interaction of a synthetic human tristetraprolin tandem zinc finger is associated with U7 snRNP and interacts with the stem-loop
peptide with AU-rich element-containing RNA substrates. J Biol binding protein in the histone premRNP to stimulate 3'-end
Chem 2003 (Please add the missing information) processing. Genes Dev 2002; 16(1):58-71.
69. Lai WS, Carballo E, Thorn J et al. Interactions of CCCH zinc finger 82. Koster M, Kuhn U, Bouwmeester T et al. Structure, expression and
proteins with mRNA. Binding of tristetraprolin-related zinc finger in vitro functional characterization of a novel RNA binding zinc
proteins to Au-rich elements and destabilization of mRNA. J Biol finger protein from Xenopus. Embo J 1991; 10(10):3087-3093.
Chem 2000; 275(23):17827-17837. 83. Gorelick RJ, Henderson LE, Hanser JP et al. Point mutants of
70. Shastry BS. Transcription factor IIIA (TFIIIA) in the second decade. Moloney murine leukemia virus that fail to package viral RNA:
J Cell Sci 1996; 109 ( Pt 3):535-539. Evidence for specific RNA recognition by a “zinc finger-like” protein
71. Caricasole A, Duarte A, Larsson SH et al. RNA binding by the sequence. Proc Natl Acad Sci USA 1988; 85(22):8420-8424.
Wilms tumor suppressor zinc finger proteins. Proc Natl Acad Sci 84. Curtis D, Treiber DK, Tao F et al. A CCHC metal-binding domain
USA 1996; 93(15):7562-7566. in Nanos is essential for translational regulation. Embo J 1997;
72. Mendez-Vidal C, Wilhelm MT, Hellborg F et al. The p53-induced 16(4):834-843.
mouse zinc finger protein wig-1 binds double-stranded RNA with 85. Worthington MT, Amann BT, Nathans D et al. Metal binding
high affinity. Nucleic Acids Res 2002; 30(9):1991-1996. properties and secondary structure of the zinc-binding domain of
73. Finerty Jr PJ, Bass BL. A Xenopus zinc finger protein that specifically Nup475. Proc Natl Acad Sci USA 1996; 93(24):13754-13759.
binds dsRNA and RNA-DNA hybrids. J Mol Biol 1997; 86. Barabino SM, Hubner W, Jenny A et al. The 30-kD subunit of
271(2):195-208. mammalian cleavage and polyadenylation specificity factor and its
74. Yang M, May WS, Ito T. JAZ requires the double-stranded yeast homolog are RNA-binding zinc finger proteins. Genes Dev
RNA-binding zinc finger motifs for nuclear localization. J Biol Chem 1997; 11(13):1703-1716.
1999; 274(39):27399-27406. 87. Ogura K, Kishimoto N, Mitani S et al. Translational control of
75. Arranz V, Harper F, Florentin Y et al. Human and mouse MOK2 maternal glp-1 mRNA by POS-1 and its interacting protein SPN-4
proteins are associated with nuclear ribonucleoprotein components in Caenorhabditis elegans. Development 2003; 130(11):2495-2503.
and bind specifically to RNA and DNA through their zinc finger 88. Tabara H, Hill RJ, Mello CC et al. pos-1 encodes a cytoplasmic
domains. Mol Cell Biol 1997; 17(4):2116-2126. zinc-finger protein essential for germline specification in C. elegans.
76. Bardoni B, Schenck A, Mandel JL. A novel RNA-binding nuclear Development 1999; 126(1):1-11.
protein that interacts with the fragile X mental retardation (FMR1) 89. Plambeck CA, Kwan AH, Adams DJ et al. The structure of the
protein. Hum Mol Genet 1999; 8(13):2557-2566. zinc finger domain from human splicing factor ZNF265 fold. J Biol
77. Klocke B, Koster M, Hille S et al. The FAR domain defines a new Chem 2003; 278(25):22805-22811.
Xenopus laevis zinc finger protein subfamily with specific RNA
homopolymer binding activity. Biochim Biophys Acta 1994;
1217(1):81-89.
CHAPTER 12
Abstract
T
he wig-1 gene (for Wild type p53-Induced Gene 1) is a To learn more about the p53 tumor suppressor pathway, it is
direct transcriptional target of wild type p53. Wig-1 is a important to identify and study the function of p53-regulated
highly conserved unusual nuclear zinc finger protein that genes. wig-1 is a novel p53 target gene that encodes a zinc finger
binds double stranded RNA. Overexpression of Wig-1 inhibits protein with unusual characteristics and unknown function.
cell growth. Wig-1 has been implicated in neuronal apoptosis
and stem cell proliferation/differentiation. Elucidation of the bio- Discovery of Mouse wig-1 and Its Rat
logical function of Wig-1 should provide new insights into the Ortholog pag608
p53 tumor suppressor pathway. The mouse wig-1 gene was identified in J3D mouse T lym-
phoma cells carrying a temperature sensitive Val135 mutant p53
Introduction construct (tsp53). This construct is expressed as mutant p53 at
The p53 tumor suppressor protein is activated by different 37oC but temperature shift to 32°C induces wild type p53 ex-
stress signals such as irradiation, genotoxic drugs, hypoxia, and pression, triggering cell cycle arrest and massive apoptosis. wig-1
oncogenic activation. This leads to p53 accumulation and induc- was identified by differential display analysis as one of several
tion of cell cycle arrest, differentiation, apoptosis and/or senes- mRNAs expressed at 32oC but not 37oC in J3D-tsp53 cells. Wig-1
cence. p53 orchestrates these biological responses through tran- is a 290 amino acid protein with an apparent molecular weight of
scriptional regulation of a set of target genes. This involves binding 32 kD. Northern blotting using the wig-1 coding region as a probe
to a conserved sequence site that consists of two copies of the revealed two main transcripts of 7.6 kb and 2.2 kb. Both tran-
motif 5'-PuPuPuC(A/T) (T/A)GPyPyPy-3' separated by 0-13 scripts are induced by gamma irradiation in mouse fibroblasts in
nucleotides. Induction of the p21 and GADD45 genes triggers a wild type p53-dependent manner. Moreover, wig-1 is strongly
cell cycle arrest whereas transactivation of proapoptotic genes, induced in brain, testis, kidney, lungs and spleen of irradiated
for instance Bax, Fas, KILLER/DR5, Noxa, and PUMA, induces mice, and also expressed at detectable levels in unirradiated brain
apoptosis. p53-mediated repression of the antiapoptotic Bcl-2 and and testis.3
IGF1 receptor genes may also contribute to the apoptotic response. The rat homolog of wig-1, PAG608 (p53-Activated Gene 608)
The p53-induced MDM2 protein regulates p53 stability by bind- was independently identified by Israeli et al.4 At the protein level,
ing and ubiquitinating p53, resulting in proteasome-mediated PAG608 is 97% identical to Wig-1 and is induced in
degradation of p53 in the cytoplasm. This ensures low levels of tsp53-carrying cells in the same manner as wig-1. The protein
p53 in the absence of p53-activating stimuli. p53 induces several localizes to the nucleus and possibly to nucleoli. A weak apoptotic
proteins that regulate MDM2, including cyclin G, PTEN and effect of PAG608 was observed in transient transfection experi-
WISP1 (reviewed by Vousden and Lu ).1 ments.
p53 is inactivated by point mutation or deletion in around Wig-1/PAG608 is a zinc finger protein that contains three
50% of all human tumors. Most point mutations cluster in the Cys2-His2 (C2H2) zinc finger motifs and a nuclear localization
p53 core domain that interacts specifically with DNA, leading to signal between the second and the third zinc finger (Fig. 1). The
inactivation of p53 DNA binding and failure to transactivate two N-terminal zinc fingers show strong homology to each other,
p53-regulated genes. This makes mutant p53 an interesting tar- whereas the third zinc finger differs significantly.
get for cancer therapy. Several novel strategies for p53-based can-
cer therapy have been designed and in some cases tested clini- Human wig-1
cally, including p53 gene therapy and reactivation of mutant p53 Human Wig-1 shows 87% amino acid sequence identity to
by small molecules.2 mouse Wig-1.5 The zinc fingers are almost perfectly conserved,
Zinc Finger Proteins: From Atomic Contact to Cellular Function, edited by Shiro Iuchi and Natalie Kuldell.
©2005 Landes Bioscience/Eurekah.com and Kluwer Academic/Plenum Publishers.
Wig-1, a p53-Induced Zinc Finger Protein that Binds Double Stranded RNA 77
Figure 2. Alignment of human, mouse and rat Wig-1 protein sequences. Zinc fingers are indicated in bold. The nuclear localization signal between
zinc finger 2 and 3 is underlined.
78 Zinc Finger Proteins: From Atomic Contact to Cellular Function
and lack the consensus linker sequence TG(Q/E)KP found in been shown to exhibit sequence similarities (22-32%) with
DNA-binding zinc finger proteins. It has been suggested that dsRBP-ZFa and Wig-1.
proteins harboring widely spaced fingers may show greater versa- In conclusion, dsRNA binding may be a property of any pro-
tility in its binding specificity, strength and span.12,13 Further- tein containing zinc fingers of the type and with the distribution
more, zinc finger proteins normally have four amino acids sepa- found in Wig-1, JAZ and dsRBP-ZFa. The possible functional
rating the histidines involved in zinc coordination. In contrast, relationship between proteins in this group needs further investi-
these three proteins have inter-histidine spaces of five amino ac- gation.
ids. The linker regions show no sequence similarities. Interest-
ingly, Wig-1, dsRBP-ZFa and JAZ all show a characteristic bind- Wig-1 and Neuronal Apoptosis
ing to DNA-RNA hybrids and lack DNA binding activity. Like Rat wig-1 (PAG608) has been reported to be constitutively
the binding to dsRNA through the consensus dsRBM, the bind- expressed throughout the hippocampus, and has also been impli-
ing of this novel group of proteins to dsRNA and DNA-RNA cated in neuronal cell death following cerebral ischemia. In a rat
hybrids is apparently sequence-independent. However, they show model for transient global cerebral ischemia, increased wig-1/
a clear preference for binding to an A-form helix versus the B-form pag608 mRNA and protein levels were observed in the vulner-
helix recognized by most DNA-binding proteins. No function able pyramidal neurons of the hippocampal CA1 region.19 This
has so far been attributed to dsRBP-ZFa, JAZ and Wig-1, al- region is known to undergo apoptosis in response to ischemia,
though JAZ and Wig-1 have been implicated in inhibition of cell suggesting a role for wig-1 in neuronal cell death. Moreover, p53
growth and/or apoptosis. and Wig-1 expression were increased by 6-hydroxydopamine
In addition, it is worth mentioning the identification of two (6-OHDA) treatment in catecholaminergic neuronal cell lines.20
other proteins sharing structural similarities to Wig-1, JAZ and 6-OHDA is a known neurotoxin that causes degeneration of ty-
dsRBP-ZFa but these proteins are less well characterized regard- rosine hydroxylase-positive neurons, and has been used to gener-
ing their nucleic acids binding properties. ZFR (zinc finger RNA ate animal models of degenerative diseases such as Parkinson’s
binding) is a nuclear protein highly expressed in testis, ovary and disease. Interestingly, the p53 induction following 6-OHDA treat-
brain that was isolated in a screen for RNA binding proteins ex- ment and the subsequent apoptosis-related morphological changes
pressed during mouse spermatogenesis.14 ZFR also exists in could be inhibited with antisense wig-1 cDNA. In addition, a
worms, flies and humans and contains three widely spaced C2H2 Wig-1 zinc finger 1 deletion-mutant also inhibited cell death.
zinc finger motifs, two putative nuclear localization signals and a This implies not only that Wig-1 is an effector of neuronal
314 amino acids region conserved in several dsRNA-binding pro-
apoptosis but that the dsRNA-binding domain of Wig-1 is cru-
teins harboring dsRBMs. Human zfr is mostly expressed in brain.15
cial for the apoptotic effect. Furthermore, Wig-1 was observed to
In nucleic acids binding assays, recombinant mouse ZFR protein
increase p53 protein expression in neuronal cells, suggesting a
displayed affinity for RNA known to contain many duplexed re-
positive feedback loop in which 6-OHDA induces p53 that in
gions but also for DNA.14 Homozygous deletion of the zfr gene
turn induces Wig-1 that increases p53 levels, triggering neuronal
in the mouse leads to both increased programmed cell death and
apoptosis.
decreased mitotic index and ultimately to early (E-9.5-10.5) em-
bryonic lethality.16 Hzf is another protein containing three zinc
finger domains similar to the zinc finger domains found in Wig-1, Wig-1 and Stem Cell Proliferation
JAZ and dsRBP-ZFa. It was isolated as a novel gene regulated Another important clue as to the function of wig-1 has come
during hematopoietic development from embryonic stem (ES) from the study of mechanisms that regulate self-renewal of
cells containing a random gene trap insertion and induced to haematopoitetic stem cells (HSC). Bmi-1 is a member of the PcG
differentiate into hematopoietic cells.17 Hzf is predominantly family of transcriptional repressors that control development by
expressed in megakaryocytes suggesting a role in the differentia- the regulation of cell growth and differentiation genes.
tion and/or maturation of this cell lineage. Indeed, studies on Bmi-1-deficient mice developed hypocellular bones and died less
Hzf-deficient mice showed an essential function of Hzf in than 2 months after birth due to no self-renewal of adult HSC.21
megakaryopoiesis.18 Importantly, the Hzf putative peptide has When comparing gene expression profiles of bone marrow mono-
Wig-1, a p53-Induced Zinc Finger Protein that Binds Double Stranded RNA 79
nuclear cells obtained from wild type and Bmi-1 deficient mice, 10. Yang M, May WS, Ito T. JAZ requires the double-stranded
upregulation of wig-1 mRNA was observed in HSC lacking RNA-binding zinc finger motifs for nuclear localization. J Biol Chem
Bmi-1. Moreover, analysis of transcriptional profiles for three dif- 1999; 274(39):27399-27406.
11. Mendez-Vidal C, Wilhelm MT, Hellborg F et al. The p53-induced
ferent types of stem cells, i.e., embryonic stem cells, neural stem mouse zinc finger protein wig-1 binds double-stranded RNA with
cells, and haematopoietic stem cells, showed that wig-1 mRNA high affinity. Nucleic Acids Res 2002; 30(9):1991-1996.
was enriched in all three types of stem cells.22 These results sug- 12. Reuter G, Giarre M, Farah J et al. Dependence of position-effect
gest that wig-1 has an important role in the control of stem cell variegation in Drosophila on dose of a gene encoding an unusual
proliferation and/or differentiation. zinc-finger protein. Nature 1990; 344(6263):219-223.
13. Fasano L, Roder L, Core N et al. The gene teashirt is required for
the development of Drosophila embryonic trunk segments and
Acknowledgments encodes a protein with widely spaced zinc finger motifs. Cell 1991;
The authors are supported by the Swedish Cancer Society 64(1):63-79.
(Cancerfonden), the King Gustaf V Jubilee Fund, the Ingabritt 14. Meagher MJ, Schumacher JM, Lee K et al. Identification of ZFR,
& Arne Lundberg Research Foundation, the Robert Lundberg an ancient and highly conserved murine chromosome-associated zinc
finger protein. Gene 1999; 228(1-2):197-211.
Foundation, and the Karolinska Institute Funds. 15. Kleines M, Gartner A, Ritter K et al. Cloning and expression of the
human single copy homologue of the mouse zinc finger protein zfr.
References Gene 2001; 275(1):157-162.
1. Vousden KH, Lu X. Live or let die: The cell’s response to p53. Nat 16. Meagher MJ, Braun RE. Requirement for the murine zinc finger
Rev Cancer 2002; 2(8):594-604. protein ZFR in perigastrulation growth and survival. Mol Cell Biol
2. Lane DP, Lain S. Therapeutic exploitation of the p53 pathway. Trends 2001; 21(8):2880-2890.
Mol Med 2002; 8(4 Suppl):S38-42. 17. Hidaka M, Caruana G, Stanford WL et al. Gene trapping of two
3. Varmeh-Ziaie S, Okan I, Wang Y et al. Wig-1, a new p53-induced novel genes, Hzf and Hhl, expressed in hematopoietic cells. Mech
gene encoding a zinc finger protein. Oncogene 1997; Dev 2000; 90(1):3-15.
15(22):2699-2704. 18. Kimura Y, Hart A, Hirashima M et al. Zinc finger protein, Hzf, is
4. Israeli D, Tessler E, Haupt Y et al. A novel p53-inducible gene, required for megakaryocyte development and hemostasis. J Exp Med
PAG608, encodes a nuclear zinc finger protein whose overexpression 2002; 195(7):941-952.
promotes apoptosis. EMBO J 1997; 16(14):4384-4392. 19. Tomasevic G, Shamloo M, Israeli D et al. Activation of p53 and its
5. Hellborg F, Wang Q, Méndez-Vidal C et al. Human wig-1, a target genes p21(WAF1/Cip1) and PAG608/Wig-1 in ischemic
p53-target gene that encodes a growth inhibitory zinc finger protein. preconditioning. Brain Res Mol Brain Res 1999; 70(2):304-313.
Oncogene 2001; 20:5466-5474. 20. Higashi Y, Asanuma M, Miyazaki I et al. The p53-activated gene,
6. Wilhelm MT, Mendez-Vidal C, Wiman KG. Identification of PAG608, requires a zinc finger domain for nuclear localization and
functional p53-binding motifs in the mouse wig-1 promoter. FEBS oxidative stress-induced apoptosis. J Biol Chem 2002;
Lett 2002; 524(1-3):69-72. 277(44):42224-42232.
7. Fierro-Monti I, Mathews MB. Proteins binding to duplexed RNA: 21. Park IK, Qian D, Kiel M et al. Bmi-1 is required for maintenance
One motif, multiple functions. Trends Biochem Sci 2000; of adult self-renewing haematopoietic stem cells. Nature 2003;
25(5):241-246. 423(6937):302-305.
8. Iuchi S. Three classes of C2H2 zinc finger proteins. Cell Mol Life 22. Ramalho-Santos M, Yoon S, Matsuzaki Y et al. “Stemness”:
Sci 2001; 58(4):625-635. Transcriptional profiling of embryonic and adult stem cells. Science
9. Finerty Jr PJ, Bass BL. A Xenopus zinc finger protein that specifically 2002; 298(5593):597-600.
binds dsRNA and RNA-DNA hybrids. J Mol Biol 1997;
271(2):195-208.
CHAPTER 13
Summary
rat intestinal epithelial cells. Gomperts et al2 were apparently the
A
small family of mammalian zinc finger proteins contain-
ing an unusual putative tandem zinc finger motif was first to describe the internal cysteine-rich repeat sequence that we
identified approximately 13 years ago. The tandem zinc now refer to as the tandem zinc finger (TZF) domain; they noted
finger domain was characterized by two hypothetical fingers with that this was found both in their clone of cMG1 and also in a
identical Cx8Cx5Cx3H spacing, with exactly 18 amino acids partial clone of TIS11 (tristetraprolin or TTP) previously pub-
between the carboxyl terminal H of the first zinc finger and the lished by Herschman and colleagues.3 As stated by Gomperts et
amino terminal C of the second zinc finger. The two fingers also al2 “… it is noteworthy that each of the repeated elements in
shared a characteristic amino-terminal lead-in sequence of TIS11 and cMG1 contains three cysteines and one histidine resi-
RYKTEL or a close variant. Although first thought to be tran- due, suggesting a possible coordination of these four amino acids
scription factors, these proteins are becoming better understood with a metal ion in a manner analogous to zinc-finger structures.”
as the result of experiments with knockout mice for tristetraprolin It is this TZF domain that will be the focus of this review.
(TTP), currently the best-studied member of the family. These We became involved in the study of the CCCH TZF proteins
mice developed a systemic inflammatory syndrome found to be more than 10 years ago, when one of us (WSL) cloned an
secondary to elevations of tumor necrosis factor alpha (TNF) insulin-induced gene from a library from an insulin-sensitive fi-
and possibly granulocyte-macrophage colony-stimulating factor broblast cell line by differential hybridization. This gene was rap-
(GM-CSF). These elevations were found to be due to stabilized idly induced by insulin, serum and other growth factors, and en-
mRNAs for these cytokines, and subsequent work found that coded a hypothetical protein that we called tristetraprolin or TTP
TTP could bind to the AU-rich elements within the because of three PPPPG motifs in its primary sequence. We de-
3'-untranslated region of these mRNAs and destabilize them, scribed this sequence and the “immediate early response gene”
apparently by initiating a process of 3'-5' deadenylation. This characteristics of its induction in the September 25, 1990 issue
chapter will summarize some of our current thinking about this of the Journal of Biological Chemistry.4 A partial clone of this cDNA
small but interesting protein family, including binding site and (TIS11) had been identified earlier in a screen for phorbol
binding domain characterization, and recent developments in ester-responsive genes, and was described in the January, 1989
mutagenesis and structure determination. issue of Oncogene;3 these authors described the corrected sequence
in the July, 1991 issue of Oncogene.5 The same gene was also cloned
Introduction and Background nearly simultaneously as a serum-induced gene (Nup475) by
Zinc fingers are small, independently folding protein motifs DuBois et al;6 this clone was described in the November 5, 1990,
that require one or more zinc ions to maintain structural stabil- issue of the Journal of Biological Chemistry.
ity. Zinc finger-containing proteins are among the most abun- A third mammalian cDNA was cloned by the Herschman
dant in the eukaryotic genome and are most commonly classified group in a study seeking to identify cDNAs with similar TZF
as transcription factors containing the well-described Cys2His2 domains; this protein was labeled TIS11d, and was described in
zinc finger structure (reviewed in ref. 1). Recently, a more un- the March, 1991, issue of Molecular and Cellular Biology.7 The
usual class of zinc finger proteins has become recognized. In these protein translated from their original TIS11d sequence (GenBank
Cys3H or CCCH zinc finger proteins, the zinc ion is coordinated accession number M58564.1) was truncated in the amino termi-
by three cysteines and one histidine. nus and had a reading frame shift in the carboxyl terminus when
To our knowledge, the first description of a novel family of compared to the current human protein RefSeq (NP_008818),
tandem CCCH zinc finger proteins was made by Brown and col- as we have discussed previously.8,9 We have recently confirmed
leagues in their initial description of the cMG1 protein, pub- that the extended amino terminus is expressed as part of the pro-
lished in Oncogene in July, 1990.2 The cDNA encoding this pro- tein in the mouse (S.B. Ramos and P.J. Blackshear, unpublished
tein was cloned as a gene induced by epidermal growth factor in data).
Zinc Finger Proteins: From Atomic Contact to Cellular Function, edited by Shiro Iuchi and Natalie Kuldell.
©2005 Landes Bioscience/Eurekah.com and Kluwer Academic/Plenum Publishers.
Tandem CCCH Zinc Finger Proteins in mRNA Binding 81
The Human Genome Nomenclature Committee has recently main peptides to bind to the ARE probe with high affinity.19
formalized the nomenclature for these three genes in man as Similar conclusions about a physiological role for TTP in the
ZFP36 (encoding TTP, TIS11, Nup475 or G0S24), ZFP36L1 regulation of granulocyte macrophage colony stimulating factor
(for cMG1, TIS11b, Berg 36, ERF1 and BRF1), and ZFP36L2 (GM-CSF) have been drawn from studies in bone-marrow
(for TIS11D, ERF2 and BRF2). Other aspects of the genomics derived stromal cells from TTP knockout mice and their wild-type
of the human genes, their chromosomal localization and com- littermates.20,21 Although TTP can bind to AREs in the mRNAs
mon variants, can be found in a recent review.10 We will refer to encoding interleukin 3 and other mRNAs, the only mRNAs
the proteins in this chapter as TTP, ZFP36L1 and ZFP36L2. shown to be affected so far in the knockout mice or cells derived
As far as we know, these are the three mammalian family mem- from them are the mRNAs encoding TNF and GM-CSF. Thus,
bers of the CCCH TZF proteins. A fourth protein has been iden- we will focus this review on the TZF domains of these proteins as
tified in frogs8 and fish.11 This transcript appears to be expressed ARE binding domains.
only in oocytes, ovary and eggs in these species, and contains not
only the characteristic tandem CCCH zinc fingers but also two Binding Sites for TTP
more degenerate fingers of this type. Although there has been re- To our knowledge, the only physiologically relevant binding
cent speculation that this protein, known as C3H-4 in Xenopus, is sites for TTP that have been proven from experiments in cells
actually a family of related proteins,12 the recent completion of the derived from knockout mice are those in the mRNAs encoding
small Fugu genome suggests that there is only one representative of TNF and GM-CSF. TTP can bind to other AU-rich elements in
this gene in this species, as well as the Fugu representative of the mRNAs, and can be shown to exert its usual deadenylating abil-
three mammalian CCCH TZF proteins (see below). ity on these mRNAs, in cell transfection experiments; examples
Because of its purported nuclear localization, immediate early include the mRNA for IL-3, and for TTP itself. However, we will
response gene induction characteristics, and presumed zinc fin- focus this discussion on the two known physiological substrates.
ger structure, TTP, the best studied of the mammalian CCCH In early experiments,13 we showed by UV light-induced
TZF proteins, was presumed to be a transcription factor. In fact, crosslinking that TTP could bind to the ARE-containing por-
in addition to its well-known “squelching” effect on overall gene tion of the TNF mRNA. We then showed that this could be nar-
transcription,13 TTP has been implicated as a transcription fac- rowed down to a core ARE of 24 b, that was still compatible with
tor in more recent studies. That this remains a possibility is sug- TTP binding and TTP-dependent destruction in cotransfection
gested by the fact that it can shuttle between nucleus and cyto- experiments.16,17 However, when one or more of the key A resi-
plasm.9 However, the best-documented function of the protein, dues within this core ARE was changed to C or G residues, bind-
in our view, is its ability to bind to AU-rich elements in certain ing could be almost completely abrogated (although this depended
mRNA molecules and destabilize those mRNAs, apparently by somewhat on the placement of these residues within the core –
initiating a process of deadenylation. These studies were based on see below).
the fact that TTP knockout mice chronically over-expressed tu- The naturally occurring TNF ARE contains a number of se-
mor necrosis factor alpha (TNF), due in part to their chronic quences resembling so-called class II AREs that may be capable of
over-expression of TNF mRNA.14,15 This was found to be stabi- TTP binding, raising the possibility of multiple TTP binding sites
lized in macrophages derived from the TTP-deficient mice.13 In within the same ARE. To explore the natural variation within the
transfection systems, TTP (and its other family members) can TNF ARE, we performed an alignment of the ARE portion of the
destabilize this mRNA after binding directly to AU-rich elements known mammalian TNF mRNAs using ClustalW (Fig. 1). This
in its 3'-untranslated region (3’UTR).16-18 That the tandem zinc was done by searching known mRNA sequences with the human
finger (TZF) domain is responsible for the binding was demon- ARE portion of the TNF ARE using the nr database from GenBank
strated by the ability of single amino acid mutations in any of the and DNA-DNA blast with the repetitive elements filter disabled;
canonical cysteines or histidines to greatly decrease binding16 as we also searched the mammalian EST collection using the same
well as in the ability of both biosynthetic and synthetic TZF do- program. As will be apparent from this figure, the general gestalt of
Table 1. Summary of point mutant effects on TTP binding to an ARE probe and destabilizing
an ARE-containing mRNA
Interval Amino Acid Change ARE Binding Decrease TNF mRNA? Increase TNF mRNA?
CX8C
R110E(L) ++ ++++
H148E(L) ++++ ++++
F112Y ++++ ++++
F112Q ++ ++
F150Y ++++ ++++
G116E +++ ++++
G154K ++++ ++++
R117E(L) ++++ ++++
R155L ++++ ++++
C147R 0 0 Yes
CX5C
Y120F ++++ ++++
Y158E 0 0 Yes
Y158Q 0 0 Yes
G121T(R) ++ ++
G159A(E)(K) ++ ++
K123D ++++ ++++
R161L + +
C124H(R) 0 0 Yes
C162H 0 0 Yes
CX3H
F126Y ++++ ++++
F126N 0 0 Yes
F164Y ++++ ++++
F164L 0 0 Yes
H128K 0 0 Yes
H166L 0 0 Yes
These data summarize the results from three to four individual experiments. “Interval” represents the distance between each of two
key amino acids in the two CCCH motifs, and X represents any amino acid. “Amino acid change” refers to the results of mutations
encoding conserved amino acids of the CCCH motifs, as illustrated in Figure 1. “ARE binding” means the ability of expressed hTTP
mutants to bind to a GM-CSF mRNA ARE probe as assessed by gel mobility shift assays. The binding activity of wild-type TTP is indicated
by ++++, whereas no binding activity is indicated by 0. ”Decrease TNF mRNA?” represents the ability of a mutant TTP construct to cause
breakdown of the TNF mRNA in cotransfection experiments in 293 cells, indicated as 0 (no effect) to ++++ (equivalent to wild-type
TTP). “Increased TNF mRNA?” refers to the ability of a given mutant to cause an increase in the steady-state concentration of TNF or
other ARE containing mRNAs in the 293 cell cotransfection studies. Modified from ref. 26 with permission.
table does not include mutations within the highly conserved domain (see below). Similar studies are underway for a variety of
lead-in sequences for both fingers or the interfinger spacer which other of the conserved residues within the zinc fingers themselves;
also has a number of conserved residues. Some of these results within the lead-in sequences; within the intra-finger spacer; and
were somewhat surprising. For example, mutation of the perfectly finally, “spacers” are being introduced into the highly conserved
conserved R110 to an acidic E residue had only a modest effect inter-finger and intra-finger distances, to determine how well
on RNA binding, and did not seem to affect the ability of TTP to changes in these distances can be tolerated. This is of particular
promote TNF mRNA degradation in the co-transfection studies. importance when evaluating the numerous members of this
Similarly, mutation of G116 or G154 also had only a modest extended protein family in C. elegans (see below), where there is
effect on TTP binding activity. However, mutation of other con- considerable variability in the distances between key residues in
served residues, such as Y158 to either E or Q, completely abol- the two zinc fingers.
ished RNA binding ability. Another example of this type of pro- In addition to the RNA binding properties of the TZF do-
found change was the F residue in the final intra-finger sequence main, it has emerged within the past few years that the TZF do-
in either zinc finger: mutation of F126 or F164 to Y had no effect main contains sequences that seem to be capable of nuclear local-
on binding, whereas mutation of F126 to N or F164 to L abol- ization. TTP has long known to be capable of moving between
ished binding. the nuclear and cytosolic compartments,27 and recent studies from
Many of these highly conserved residues that are critical for our laboratory have identified nuclear export sequences within
binding have been found to be in significant positions within the the amino terminus of TTP and the carboxyl-termini of ZFP36L1
recently described three-dimensional structure of this binding and L2.9 However, the nature of the nuclear localization signals
Tandem CCCH Zinc Finger Proteins in mRNA Binding 87
was not clear. Recently, Murata and colleagues demonstrated that cussion, vertebrate genomes seem to express only four members
two arginine residues within the inter-finger spacer were impor- of the TTP family of TZF CCCH finger proteins, using the fairly
tant for nuclear localization of the protein.28 In our own work, strict spacing definition used here. We recently searched GenBank
the TZF domain alone seems to be entirely nuclear, i.e., when for all known members of this family in vertebrates, in general
lacking the nuclear export sequences. Mutation of the same two using tblastn to search nr, EST and genome databases for the
arginine residues in the inter-finger linker of the whole TTP and various vertebrates. The results of this overall alignment using
ZFP36L1 proteins caused the expressed proteins to be completely ClustalW are shown in Figure 9. In this alignment, the prototype
cytosolic, compatible with an important role for these two resi- TZF domain belonged to mouse TTP, and the distance from
dues as a nuclear localization motif. However, when these mu- mouse was directly proportional to the difference in sequence
tant proteins were expressed in 293 cells and used in a gel shift within the TZF domain. The representation of these differences
assay, they were unable to bind to an ARE probe, suggesting that in a phylogenetic tree are shown in Figure 10. Interestingly, Fig-
these conserved basic residues within the inter-finger linker may ure 9 demonstrates that all three of the Xenopus orthologues of
be critical for the three dimensional configuration of the protein TTP, ZFP36L1 and ZFP36L2, known as C3H-1, C3H-2, and
and are necessary for binding to occur. This in turn suggests that C3H-3, respectively, are closer to mouse TTP than any of the
these residues may not constitute true nuclear localization sig- fish members of the family. In fact, the fish members are often
nals, since the fact that the mutant protein can no longer bind difficult to assign to an orthologous family member, and so were
RNA suggests that a fairly fundamental change in structure has arbitrarily numbered 1-4. In particular, the boundaries between
occurred that may prevent the nuclear import of the protein, in- the fish orthologues of ZFP36L1 and ZFP36L2 were often
dependent of true primary sequence nuclear localization motifs. blurred, so it seems possible that these proteins will be found to
Distinguishing among these possibilities will require further ex- be interchangeable in the fish genome, with biological activity
perimentation. possibly more dependent upon expression pattern than primary
sequence.
Conserved Residues within Vertebrate CCCH This alignment allows for the identification of amino acids
TZF Domain Proteins that are conserved among these various disparate groups of verte-
As mentioned above, a fourth member of this family of pro- brate animals, as shown at the bottom of Figure 9. Remarkably,
teins was identified in frogs and fish.8,11 Although an analysis of many of the conserved residues found in the mammalian align-
the Xenopus and Silurana tropicalis ESTs encoding this fourth ment are still conserved among all vertebrates, and occasional
protein, known as Xenopus C3H-4, suggested that it might repre- differences, such as the equivalent of E145 to Q change in zebrafish
sent a family of closely related sequences,12 we recently deter- “TTP” as the only vertebrate with this change, are possibly due
mined by blasting the Fugu genome that there probably is only a to EST sequencing errors or polymorphisms. Nonetheless, these
single representative of this protein in the Fugu genome, as well conserved residues will continue to represent mutation targets
as three additional proteins representing the Fugu versions of TTP, for understanding the RNA binding domain and nuclear local-
ZFP36L1 and ZFL36L2. Therefore, for the purposes of this dis- ization functions of the proteins.
peptide.31 Thus, we can make the assumption that a nearly iden- 14. Taylor GA, Carballo E, Lee DM et al. A pathogenetic role for TNF
tical situation will exist for all the mammalian family members in alpha in the syndrome of cachexia, arthritis, and autoimmunity
complex with the RNA binding site. Further challenges will in- resulting from tristetraprolin (TTP) deficiency. Immunity 1996;
4(5):445-454.
clude additional structure function studies on many conserved 15. Carballo E, Gilkeson GS, Blackshear PJ. Bone marrow
and variable residues within the different TZF domains; the elu- transplantation reproduces the tristetraprolin-deficiency syndrome in
cidation of the mechanisms by which the intact proteins pro- recombination activating gene-2 (-/-) mice. Evidence that monocyte/
mote deadenylation after this binding complex has formed; the macrophage progenitors may be responsible for TNFalpha
identity of the bona fide physiological binding partners for the overproduction. J Clin Invest 1997; 100(5):986-995.
other family members; the numerous regulatory possibilities af- 16. Lai WS, Carballo E, Strum JR et al. Evidence that tristetraprolin
binds to AU-rich elements and promotes the deadenylation and
forded by phosphorylation, nuclear-cytoplasmic shuttling, tran- destabilization of tumor necrosis factor alpha mRNA. Mol Cell Biol
scriptional regulation and protein binding events; and finally, the 1999; 19(6):4311-4323.
possibility of using this novel structural information to develop 17. Lai WS, Carballo E, Thorn JM et al. Interactions of CCCH zinc
new types of therapies for diseases caused in part by elevated lev- finger proteins with mRNA. Binding of tristetraprolin-related zinc
els of circulating TNF. finger proteins to Au-rich elements and destabilization of mRNA. J
Biol Chem 2000; 275(23):17827-17837.
18. Lai WS, Blackshear PJ. Interactions of CCCH zinc finger proteins
References with mRNA: Tristetraprolin-mediated AU-rich element-dependent
1. Laity JH, Lee BM, Wright PE. Zinc finger proteins: New insights mRNA degradation can occur in the absence of a poly(A) tail. J
into structural and functional diversity. Curr Opin Struct Biol 2001; Biol Chem 2001; 276(25):23144-23154.
11(1):39-46. 19. Blackshear PJ, Lai WS, Kennington EA et al. Characteristics of the
2. Gomperts M, Pascall JC, Brown KD. The nucleotide sequence of a interaction of a synthetic human tristetraprolin tandem zinc finger
cDNA encoding an EGF-inducible gene indicates the existence of a peptide with AU-rich element-containing RNA substrates. J Biol
new family of mitogen-induced genes. Oncogene 1990; Chem 2003; 278(22):19947-19955.
5(7):1081-1083. 20. Carballo E, Lai WS, Blackshear PJ. Evidence that tristetraprolin is a
3. Varnum BC, Lim RW, Sukhatme VP et al. Nucleotide sequence of physiological regulator of granulocyte-macrophage colony-stimulating
a cDNA encoding TIS11, a message induced in Swiss 3T3 cells by factor messenger RNA deadenylation and stability. Blood 2000;
the tumor promoter tetradecanoyl phorbol acetate. Oncogene 1989; 95(6):1891-1899.
4(1):119-120. 21. Carballo E, Blackshear PJ. Roles of tumor necrosis factor-alpha
4. Lai WS, Stumpo DJ, Blackshear PJ. Rapid insulin-stimulated receptor subtypes in the pathogenesis of the tristetraprolin-deficiency
accumulation of an mRNA encoding a proline-rich protein. J Biol syndrome. Blood 2001; 98(8):2389-2395.
Chem 1990; 265(27):16556-16563. 22. Lai WS, Kennington EA, Blackshear PJ. Tristetraprolin and its family
5. Ma Q, Herschman HR. A corrected sequence for the predicted members can promote the cell-free deadenylation of AU-rich
protein from the mitogen-inducible TIS11 primary response gene. element-containing mRNAs by poly(A) ribonuclease. Mol Cell Biol
Oncogene 1991; 6(7):1277-1278. 2003; 23(11):3798-3812.
6. DuBois RN, McLane MW, Ryder K et al. A growth factor-inducible 23. Worthington MT, Pelo JW, Sachedina MA et al. RNA binding
nuclear protein with a novel cysteine/histidine repetitive sequence. J properties of the AU-rich element-binding recombinant Nup475/
Biol Chem 1990; 265(31):19185-19191. TIS11/tristetraprolin protein. J Biol Chem 2002;
7. Varnum BC, Ma QF, Chi TH et al. The TIS11 primary response 277(50):48558-48564.
gene is a member of a gene family that encodes proteins with a 24. Garcia-Castillo J. Interactions between tristetraprolin and the TNFa
highly conserved sequence containing an unusual Cys-His repeat. mRNA of the Marine Fish Gilthead Seabream. Paper presented at:
Mol Cell Biol 1991; 11(3):1754-1758. International Society for Developmental and Comparative
8. De J, Lai WS, Thorn JM et al. Identification of four CCCH zinc Immunology (ISDCI) - 9th International Congress. Scotland:
finger proteins in Xenopus, including a novel vertebrate protein with University of St Andrews, 2003.
four zinc fingers and severely restricted expression. Gene 1999; 25. Michel SL, Guerrerio AL, Berg JM. Selective RNA binding by a
228(1-2):133-145. single CCCH zinc-binding domain from Nup475 (Tristetraprolin).
9. Phillips RS, Ramos SB, Blackshear PJ. Members of the tristetraprolin Biochemistry 2003; 42(16):4626-4630.
family of tandem CCCH zinc finger proteins exhibit 26. Lai WS, Kennington EA, Blackshear PJ. Interactions of CCCH zinc
CRM1-dependent nucleocytoplasmic shuttling. J Biol Chem 2002; finger proteins with mRNA: Nonbinding tristetraprolin mutants exert
277(13):11606-11613. an inhibitory effect on degradation of AU-rich element-containing
10. Blackshear PJ. The CCCH tandem zinc finger domain as a novel mRNAs. J Biol Chem 2002; 277(11):9606-9613.
nucleic acid binding element. In: Iuchi S, KN, ed. Zinc Finger 27. Taylor GA, Thompson MJ, Lai WS et al. Mitogens stimulate the
Proteins. 2003:in press. rapid nuclear to cytosolic translocation of tristetraprolin, a potential
11. te Kronnie G, Stroband H, Schipper H et al. Zebrafish CTH1, a zinc-finger transcription factor. Mol Endocrinol 1996; 10(2):140-146.
C3H zinc finger protein, is expressed in ovarian oocytes and embryos. 28. Murata T, Yoshino Y, Morita N et al. Identification of nuclear import
Dev Genes Evol 1999; 209(7):443-446. and export signals within the structure of the zinc finger protein
12. Blackshear PJ. Xenopus laevis genomic biomarkers for environmental TIS11. Biochem Biophys Res Commun 2002; 293(4):1242-1247.
toxicology studies. In: WsaS WA, ed. Biomarkers of environmentally 29. Thompson MJ, Lai WS, Taylor GA et al. Cloning and
associated diseae. LLC: CRC Press, 2002:339-353. characterization of two yeast genes encoding members of the CCCH
13. Carballo E, Lai WS, Blackshear PJ. Feedback inhibition of class of zinc finger proteins: Zinc finger-mediated impairment of
macrophage tumor necrosis factor-alpha production by tristetraprolin. cell growth. Gene 1996; 174(2):225-233.
Science 1998; 281(5379):1001-1005. 30. Hudson BP, Martinez-Yamout MA, Dyson HJ et al. Recognition of
the mRNA AU-rich element by the zinc finger domain of TIS11d.
Nat Struct Mol Biol 2004; 11(3):257-264.
CHAPTER 14
Abstract
Z
inc finger motifs are common in ribosomal proteins: they
are widely distributed in nature, having been found
amongst the proteins of both subunits of the ribosomes
of all species examined in the three kingdoms; the motif is al-
ways of the C2C2 variety and occurs only once in a protein. De-
spite wide distribution there is neither strict conservation of the
ribosomal proteins with the motif nor of the entire motif in ho-
Figure 1. Zinc finger motifs in S. cerevisiae ribosomal proteins. The
mologous proteins. A comprehensive genetic, biochemical, and
numbers in the second column are the position in the amino acid se-
structural analysis has been made of the contribution of the zinc
quence of the first cysteine residue in the motif.
finger to the function of yeast ribosomal protein YL37a, to date
the only study of its kind. Replacement, one at a time, of the
cysteines with serines in the motif in YL37a revealed that all analyses by atomic absorption spectroscopy indicated that
four cysteines are required for the binding of zinc; nonetheless, mammalian ribosomes have approximately six moles of zinc.8
cells with mutations in three of the four cysteines do not suffer a In rps the zinc finger motif occurs only once; this is in contrast
significant impairment of growth, nor is the binding to rRNA of to DNA binding proteins where there are often tandem repeats
the mutant proteins materially affected. It is possible that the of the motif.9
zinc finger motif in ribosomal proteins are the vestiges, biologi- There is a crystal structure of the large ribosomal subunit (50S)
cal fossils if you will, of a former function, and that the motif from archaebacterium Haloarcula marismortui (Hm) resolved to
has been preserved despite the ribosomal proteins having come 2.4 Å;10 the structure confirms the presence of zinc finger motifs
to use alternate amino acid sequences and/or structures to bind in rps:1,2 HmL37e (a homolog of RL37); HmL37ae (RL37a); Hm
to rRNA as has been shown to be the case for YL37a. L24e (RL24); and HmL44e (RL36a). The rat homolog of
HmL24e (RL24) lacks the internal two cysteines and, therefore,
Introduction is unlikely to coordinate zinc. The rat homolog of HmL44e
Ribosomes are large ribonucleoprotein complexes that medi- (RL36a) has an unusually large insert (56 amino acids) between
ate protein synthesis in all organisms in the biosphere; they func- the two internal cysteines of the motif and was overlooked in the
tion to link the genotype to the phenotype. The amino acid se- original analysis.2 In sum, of the four zinc finger proteins in the
quences of the 79 mammalian1,2 and of the 78 yeast3 ribosomal large subunit, one is unique to rat, RL40, and one to H.
proteins (rps) have been determined. The sequences are needed marismortui, HmL24e.
for a solution of the structure of eukaryotic ribosomes, and they In the crystal structure of the 50S subunit of the ribosomes of
should help to understand the evolution of rps, to unravel their the eubacterium Deinococcus radiodurans (Dr) there are two rps
function, and to define the rules that govern their interaction with zinc finger motifs, DrL32 and DrL36;11 both lack eukary-
with ribosomal RNAs (rRNAs). An unexpected finding was that otic homologs. In the structure of the 30S subunit from the
some rps had amino acid sequences that were first associated with eubacterium Thermus thermophilus (Tt),12 a zinc atom is coordi-
binding to DNA.1,2,4-6 These motifs include helix-turn-helix,7 nated by four cysteine residues in rps TtS4 (a homolog of RS9;
basic region leucine-zippers,5,6 and zinc fingers.1,2,5 RS9 lacks the zinc coordinating residues and thus the motif ) and
Zinc finger motifs are present in the rps of species from the in TtS14 (homolog of RS29).
three kingdoms.4 Seven rat (R) rps have zinc finger motifs of Thus, in archaebacterial and eukaryotic ribosomes, the large
the C2C2 variety; three are in the small (S) and four are in the subunit contains four rps with zinc finger motifs; in contrast, the
large (L) ribosomal subunit; they are RS27, RS27a, RS29, eubacterial 50S subunit has only two. The similarity of the zinc
RL36a, RL37, RL37a, and RL40 (ref. 2) (Fig. 1); moreover, finger rps of eukaryotic and of archaebacterial ribosomes accords
Zinc Finger Proteins: From Atomic Contact to Cellular Function, edited by Shiro Iuchi and Natalie Kuldell.
©2005 Landes Bioscience/Eurekah.com and Kluwer Academic/Plenum Publishers.
92 Zinc Finger Proteins: From Atomic Contact to Cellular Function
MBP – 0.15
with the finding that, in general, the amino acid sequences of MBP-YL37a-C57S 106 0.16
their rps are close, closer than either are to bacteria. Nonetheless,
it is important to note that there is neither strict conservation of MBP-YL37a-C59S 102 0.93
the rps with zinc finger motifs nor of the motif in homologous
rps (see later as well). MBP-YL37a-C60S 141 0.20
Little is known of the role, if any, of the motif or of the metal a YPD liquid media: 1% yeast extract, 2% bactopeptone, 2% glucose.
in the structure and the function of the zinc finger rps. But, the
importance of metals in general and of zinc in particular for many
cellular processes inclines one to take seriously the possibility that The Phenotype of Cells with Mutations in the
the rps with zinc fingers might be critical for the assembly or for Cysteine Residues of the YL37a Zinc Finger
the function of the ribosome, perhaps because the motif has a Motif
role in interactions with rRNA. There is a great deal of data that suggests that the tetrahedral
coordination of zinc initiates the folding of the finger motif.16-18
The YL37a Zinc Finger Motif In general, mutations of the amino acids that coordinate zinc
The sequence of nucleotides (nts) in the two genes encoding abolish the binding of the metal and parri passu lead to the loss of
yeast (Y) rp YL37a (Fig. 2) had been determined,13,14 and the function of the protein. Hence, the initial prediction was that
amino acid sequence derived from the nt sequence had estab- each cysteine residue in the YL37a zinc finger motif was essential
lished that YL37a has a zinc finger motif of the C2C2 type. What for tetrahedral coordination of the metal, for the folding of the
had not been determined at the time was whether YL37a binds module, and for the function of the protein. These assumptions
zinc and, if it does, whether the structure formed by the coordi- were tested.14 Single replacements of individual cysteine residues
nation of the metal is necessary for the contribution of the pro- with serine were constructed in a YL37a-B gene in a high copy
tein to the function of the ribosome. plasmid. The mutations were: C39S, C42S, C57S, C59S, and
YL37a is essential for cell viability.14 The lethal phenotype of a C60S (Fig. 2).
strain in which both copies of the YL37a gene were disrupted A plasmid bearing a mutation of the cysteine at position 39
could be complemented with a copy of either the YL37a-A or the (C39S) in a YL37a-B gene was unable to rescue the y137a null
YL37a-B gene.15 Complementation of the null phenotype of the strain (Table 1), i.e., cysteine 39 is essential for the function of
yeast strain yl37a with the open reading frame encoding the ho- YL37a. However, in contradistinction to expectation, plasmids
mologous rat RL37a indicated that the portions of the amino
with mutations C42S, C57S, C59S, or C60S in YL37a-B were
acid sequences of the yeast and of the rat proteins most critical
able to complement the null strain. There are at least two expla-
for function are conserved.14
To address the question of the importance of the zinc finger nations of these results. First, that the coordination of zinc is not
motif for YL37a function, both chromosomal copies of the YL37a be required for the function of YL37a nor of ribosomes and, hence,
gene were disrupted. The efficiency of the complementation of for the growth of yeast cells. This explanation assumes that all
the lethal yl37a null phenotype with plasmid-encoded copies of a four cysteine residues are required for the coordination of zinc;
YL37a gene in which the cysteine residues of the zinc finger mo- i.e., that mutation of any one of the cysteines in the motif abol-
tif were replaced, one at a time, with the codons for serine was ishes metal binding. A second explanation is that YL37a with
assessed. The effect of the mutations on the binding of zinc to three cysteines and a serine at the fourth position is able to coor-
YL37a was also determined.14 Later the role in the function of dinate zinc. Serine differs from cysteine only in that the γ-oxygen
YL37a of the amino acids in the intervening region of the zinc of the former is replaced by a sulfur atom in the latter. Serine
finger motif and of the residues at the amino-terminus of the does participate in the coordination of zinc in several enzymes.19
protein was evaluated.15 In addition, the YL37a binding sites in However, canonical zinc fingers are not known to use amino ac-
26S rRNA were identified and the role of the zinc finger motif ids other than cysteine or histidine to coordinate the metal. To
and of residues in the amino-terminal region on the interaction investigate the two possibilities an in vitro assay for the binding
of the protein with the nucleic acid was investigated.15 of zinc to YL37a was developed.
Ribosomal Zinc Finger Proteins: The Structure and the Function of Yeast YL37a 93
Figure 3. The binding of the fusion protein MBP-YL37a to the indi- YL37a Zinc Finger Binding to 26S rRNA
vidual domains of 26S rRNA. The binding of 1 µM of MBP-YL37a to Domains II and III
1 nM of each of the six radioactive 26S rRNA domains was determined To delineate the contribution of the zinc finger to the interac-
by spotting the reaction mixture (10 µl) in a single drop on a nitrocel- tion of YL37a with domains II and III of 26S rRNA, the effect of
lulose filter. In (a), a histogram of the fraction of the input of the separate mutations in the cysteines and in the intervening sequence of the
RNA domains in complexes with MBP-YL37a. In (b), autoradiographs motif was assessed. In short, there is a close correspondence be-
of the complexes of MBP-YL37a and the radioactive 26S rRNA domain tween the effect of the mutations on the growth of cells and on
fragments captured on nitrocellulose filters. the binding to 26S rRNA. A substitution of serine for the first
cysteine in the motif (MBP-YL37a-C39S), a mutation that is le-
YL37a Binding Sites in 26S rRNA thal for yeast cells,14 all but abolished binding to domains II and
It is all but certain that it is the rRNAs that are most critical III (Fig. 4). More importantly, for the YL37a-C57S and
for the chemistry of protein synthesis: mRNA, aminoacyl- and YL37a-C60S mutations, there is a clear dissociation between their
peptidyl-tRNA, and the accessory initiation, elongation, and ter- inability to bind zinc on the one hand, and the lack of an affect
mination factors bind to sites in rRNA; it is likely that nts in on cell growth or on binding to rRNA on the other. The C57S
rRNA catalyze peptide bond formation; and changes in the con- and C60S mutations had abolished the binding of zinc to YL37a,
formation of the rRNAs may underlie translocation. However, but the former did not affect the growth of yeast cells and the
the importance of rps is not to be gainsaid since functional ribo- latter caused only a small increase in the doubling time (Table 1).
somes cannot be formed in their absence. Rps serve as chaper- The same mutations either did not affect binding of the protein
ones facilitating the folding of the rRNAs during biogenesis and to domain II or III or caused only a small decrease (Fig. 4). There
stabilize an optimal conformation in the mature particle.10,20 was negligable binding of the double mutant
Mutations in rps that produce a phenotype often are in rRNA MBP-YL37a-K48A-R49A to either domain II or III (Fig. 4). Fi-
binding sites.23 Thus, the chemistry of RNA-protein recognition nally, there was no binding to 26S rRNA of MBP-YL37a-W55A,
is critical to a determination of the structure of the particle and a mutation that was lethal for yeast cells (Table 2 and Fig. 4).
to the unraveling of the molecular details of its function.
It was assumed that the function of YL37a, and hence the Interaction of 26S rRNA with YL37a
phenotype of mutations in the protein, must derive from its in- Amino-Terminal Region
teraction with RNA. For that reason it seemed important to first Once again, it was possible that the side chains of conserved
identify the rRNA binding sites for YL37a and then to evaluate amino acids in YL37a apart from the zinc finger motif, contrib-
the effects of mutations in YL37a on binding to the sites. To- uted to the binding to rRNA. This interpretation is supported
wards this end, a set of six rDNA fragments that encode the do- by the observation that K16 near (but not in) the zinc finger
mains in 26S rRNA was generated by PCR amplification from a motif in Bacillus stearothermophilus S14 can be cross-linked to
plasmid that had a Saccharomyces cerevisiae 35S rDNA transcrip- 16S rRNA.25 Regions rich in basic amino acids are a common
tion unit.24 The rDNA domain fragments were subcloned in ap- feature of RNA-binding proteins. In the amino-terminal region
propriate plasmid vectors for the preparation of radioactive tran- of YL37a six of the fifteen residues between positions 3 and 17
scripts using T7 RNA polymerase.15 An in vitro assay to quantitate are basic and there are, as well, two tyrosines (Fig. 2). In all
binding was developed and the affinity of MBP-YL37a for each eukaryotic organisms for which an amino acid sequence of YL37a
of the six 26S rRNA domains was determined (Fig. 3). Domains is available, six of these eight residues (K6, K7, K13, Y14, R17,
II and III have the primary sites to which YL37a associates.15 and Y18) are either invariant or the changes are conservative.
Ribosomal Zinc Finger Proteins: The Structure and the Function of Yeast YL37a 95
Structural Data
The genetic and the biochemical experiments do not unam-
biguously define the nature of the participation of individual
amino acids in YL37a in the function of the protein, since the
results do not distinguish between a direct contribution of a resi-
due to the free energy of binding to rRNA and an indirect contri-
bution because the residue is critical for the maintenance of the
three-dimensional structure. A decision between these alterna-
tives requires correlation of the results of the genetic and bio-
chemical experiments with the three-dimensional structure of the
protein-RNA complex. On the other hand, hydrogen bonds,
nonpolar contacts, and electrostatic interactions between amino
acids and nts can only be inferred from the X-ray structure;21
biochemistry and genetics are needed to define the subset of
proximities that have functional significance.
An atomic structure of the 50S subunit of H. marismortui ri-
bosomes that has the archaebacterial homolog of YL37a,
HmL37ae, has been determined by X-ray diffraction of crystals
at 2.4 Å resolution.10 The C2C2 motif in HmL37ae (Fig. 5) coor-
dinates a zinc atom and the finger is folded into a form similar to
a zinc ribbon motif. There is conservation of the secondary struc-
tures of H. marismortui 23S and of S. cerevisiae 26S rRNAs and
of the amino acid sequences of HmL37ae and YL37a (Fig. 2).
The proteins have 41% amino acid identity (51% similarity) al-
beit HmL37ae lacks 10 of the amino acids that are near the
carboxyl-terminus of YL37a. Thus, the structure of HmL37ae
can be used, with some confidence, to interpret the biochemical
and genetic results obtained with YL37a.
HmL37ae (Fig. 5) has a globular body (residues 35 to 73) Figure 5. The structure of the ribosomal protein HmL37ae taken from
with two α-helical extensions. The globular body is formed by the crystal structure of the 50S ribosomal subunit.10 In (a), a ribbon
the folded zinc finger; the motif is comprised of four anti-parallel diagram of HmL37ae. α-Helices are red; β-strands are dark blue; loops
β-strands and five loops. One α-helical extension is near the are green; and side chains of amino acids are light blue. Two aromatic
amino-terminus (residues 21 to 34) and the other is at the residues, F14 and W55, are represented by van der Waals spheres. The
carboxyl-terminus (residues 74 to 81). There is also a 310-helix four cysteines (yellow) that coordinate a zinc atom (grey-red ball) in the
(residues 11 to 13) closer to the amino-terminus. One surface of finger motif are at the bottom. In (b), the interactions of amino acids at
the carboxyl-terminal helix is packed against HmL2 (YL8 in yeast the amino-terminus of HmL37ae. In (c), interactions between amino
ribosomes). Basic and aromatic amino acids form an extended acids in HmL37ae and nucleotides in 23S rRNA. Domain II nucleotides
surface on the side of HmL37ae that is in contact with rRNA. (817-822) are lavender; domain III nucleotides are either orange
These residues include K46, K48, and R49 in the intervening (1600-1605) or magenta (1648-1653).
region of the zinc finger; R13, R17, and Y18 at the
amino-terminus; and K34 and K36 as well. of a nt (the closest nt is more than 10 Å away). In support of this
That YL37a binds to domains II and III of 26S rRNA15 ac- interpretation, it should be noted that there is not a single in-
cords with the crystal structure of the H. marismortui 50S sub- stance, in a database of the structures of 45 protein-RNA com-
unit where HmL37ae is in proximity to domains II, III, IV of plexes, of a cysteine being implicated in the recognition of a nt.22
23S rRNA.10 None of the cysteines in the zinc finger motif of The results suggest that the two basic residues, K44 and K45
HmL37ae are within hydrogen-bonding distance (2.4 to 3.4 Å) in YL37a, combine to create a region of positive charge, since the
of a nt in 23S rRNA. For the mutations of three of the cysteines double mutation, YL37a-K44A-K45A, slowed the growth of yeast
in YL37a (C42, C57, and C60) this is compatible with a lack of cells and the binding to rRNA domains more than either muta-
a significant growth phenotype or an effect on binding to rRNA. tion alone. Of the residues in the intervening region of the zinc
The lethal phenotype of the C39S mutation can only be ascribed finger that were mutated, K45 (K44 in YL37a) is not within
to a perturbation of the structure of YL37a, since the mutation hydrogen-bonding distance of a nt; whereas the side chain of K46
abolishes binding to rRNA even though it is unlikely that the (K45 in YL37a) makes contacts to the O1 phosphate of a gua-
amino acid makes a contact to any nt. Indeed, the homologous nosine and the O2 phosphate of an adjacent cytosine in domain
cysteine in HmL37ae is not within hydrogen-bonding distance III (Table 3).
96 Zinc Finger Proteins: From Atomic Contact to Cellular Function
Table 3. Atomic contacts between amino acids in HmL37ae and nucleotides in 23S rRNA of H. marismortui
The YL37a-K48A-R49A double mutation increased the dou- Because alanine is smaller than tryptophan it is unlikely that the
bling time of yeast cells six-fold and markedly affected its affinity former would fill the cavity ordinarily occupied by the latter and
for rRNA. Once again, the phenotype of the double mutation the mutation might well cause the collapse of the fold. Thus, the
was far more severe than the sum of the separate K48A or R49A contribution of W55 to the function of YL37a is likely to derive
mutations. The ionic nature of the interaction is supported by from a role in the maintenance of the structure of the protein
the observation that arginine or lysine to alanine substitutions at rather than from participation in binding to rRNA.
positions K48 and R49 respectively reduced binding to rRNA. Among the several basic residues at the amino-terminus of
Indeed, the side chain of K48 in HmL37ae makes a contact to an YL37a that were surveyed, a K13A mutation slowed growth. The
O1 phosphate of a guanosine in domain III of 23S rRNA; whereas, side chain of R13 (the homolog in HmL37ae of K13 in YL37a)
the side chain amino groups of R49 make multiple contacts to makes a van der Waals contact to the 2' oxygen of the sugar of an
O2 phosphates and to a sugar oxygen of adjacent cytosines also adenine in domain II (Table 3). The side chain of a second basic
in domain III, as well as a van der Waals contact to an O2 phos- residue, R17, makes contacts to three consecutive nts in domain
phate of the second cytosine (Table 3). Despite the multiple con- II: van der Waals contacts to the O1 phosphate of a guanosine
tacts, the functional importance of R49 is in some way amplified and to a sugar carbon of a uridine; two electrostatic contacts to
by K48. Of course, the relationship between the two residues can the O1 phosphate of the guanosine; and finally a contact to the
be put the other way around. In any case, the two basic amino 3' oxygen of a sugar in an adenosine (Table 3). The R17A muta-
acids either form an essential region for electrostatic interaction tion in YL37a caused a marked decrease in the growth of yeast
or the fine structure of the region is dependent on the two. It may cells, presumably because these contacts were compromised; in-
be that we have here an illustration of a theme for the chemistry deed, the binding of YL37a-R17A to 26S rRNA was all but abol-
of RNA-protein recognition: the participation of clusters of two ished. The mutation YL37a-Y18 retards the growth of yeast cells,
or more basic amino acids in the recognition of RNA binding and this could be explained by the observation that the hydroxyl
sites. Clusters of basic residues are common in proteins that bind of the aromatic ring of Y18 in HmL37ae makes a contact to an
to RNA, especially in rps.1 The results suggest that it is the over- O1 phosphate of a guanosine.
all charge density of the intervening sequence that is important The closest nt to F14 in HmL37a (Y14 in YL37a) is more
for binding to 26S rRNA rather than specific residues. than 10 Å away; hence, the presumption has to be that the Y14A
A W55A mutation in YL37a was lethal. In HmL37ae, W55, mutation was lethal because the residue is essential to the mainte-
the only solvent inaccessible residue in the protein, occupies a nance of the stability of the amino-terminal region, thereby, pre-
cavity in the hydrophobic core of the globular body formed by serving the orientation of the contacts of nearby amino acids,
the β-strands of the zinc finger motif (Fig. 5); W55, however, is particularly R17 and Y18, to rRNA. The amino-terminal region
not close to an rRNA nt (the closest is more than 10 Å away). of HmYL37ae forms a 310-helix-turn-helix (residues 10-34)
Ribosomal Zinc Finger Proteins: The Structure and the Function of Yeast YL37a 97
(Fig. 5). R13 forms two hydrogen bonds with E30 converting a Biological Significance of YL37a Zinc Finger
hook into a closed loop. The side chains of R17 and Y18 are on
the outer surface of the loop where they make contacts to nts
Motif
(Table 3). The interior of the loop has a network of interactions It is a tenet of the faith of ribosomologists that the
between protein backbone carbonyls and amides and side chains “proto-ribosome” had only RNA. This is a belief that is supported
of conserved amino acids in that way forming a stable fold hav- by the mounting evidence that the rRNAs are responsible for the
ing a cavity filled by the bulky side chain of F14; 84 % of F14 is basic biochemistry of protein synthesis. The rps are now viewed
buried. Presumably, alanine in the YL37a-Y14A mutant would as a later evolutionary embellishment and are deemed to facili-
not fill the cavity and, hence, would not maintain the integrity tate the folding of the rRNA and the maintenance of an optimal
of the structure; this in turn might prevent the proper presenta- configuration, in this way endowing protein synthesis with speed
tion of the side chains of R17 and Y18 for binding to rRNA, a and accuracy. This may be too severe a restriction on the role of
failure that could rationalize the lethal phenotype of the Y14A the proteins in ribosome function, nonetheless, it is likely that
mutation. RNA preceded the proteins and hence it is pertinent to ask from
The contacts presumed to have been altered by mutations in whence the latter came.
YL37a, and to have affected binding to rRNA, are predomi- The occasion for the transition during evolution from a ribo-
nantly between amino acid side chains and nt oxygen phos- some that had only RNA to a RNP machine may have coincided
phates or sugars. That the main amino acid contacts are not to with, or been a response to, the appearance of nucleases, which
nt bases suggests that recognition by YL37a is of the shape of its would have put an RNA ribosome at risk. There are at least two
rRNA binding site rather than a readout of sequence. Inspec- possible story lines for the origin of ribosomal proteins. That they
tion of the three-dimensional structure of the relevant regions were designed specifically for the ribosome, or that they were
of 23S rRNA supports this interpretation. The conformation coopted from amongst a set of preexistent proteins that already
of the rRNA binding pocket for HmL37ae is complex with sur- had defined functions. The two possibilities are by no means ex-
faces that have many deviations from A-form. Guanosine 820, clusive nor is it likely that all the proteins were added at one time.
which has three separate contacts to R17 and one to Y18, is If the latter conjecture has substance, the preexistent proteins most
bulged out of a helix distorting the A-form geometry (Fig. 5). likely to have been recruited would have been those that already
In a like manner, cytosine 1652, which has three contacts to had the capacity to bind to nucleic acids. It is possible then that
R49, is also bulged out of a helix, once again distorting the zinc finger motifs in rps are the vestiges of a former function,
A-form geometry (Fig. 5). Nts that distort A-form geometry, perhaps, of binding to DNA or to other RNAs, and that the motif
usually because they form noncanonical pairs, can widen the has been preserved despite the rps having come to use alternate
otherwise inaccessible major groove, or they can reveal distinctly structures or alternate amino acids for associating with rRNA.4-6
shaped features in the minor groove that proteins can recog- This interpretation fits the phenotype of the cysteine mutations
nize. 21,26 It seems most likely that R17 and Y18 in the in the zinc finger motif in YL37a: their replacement with serines
amino-terminal region of YL37a, and R49 in the zinc finger yields variants that lack the capacity to bind zinc but, nonethe-
motif, is recognizing the unique shape of the rRNA in its bind- less, support growth and bind to rRNA; instead basic and aro-
ing site, and that G820 and C1652 are the identity elements in matic residues in the intervening sequence and at the
rRNA for the binding of the rp. amino-terminus of the protein are important in binding to rRNA.
The contacts of the zinc finger motif residues K45, K48,
It has not yet been determined, however, whether the dispens-
and R49 are to nts in domain III; whereas, the contacts of resi-
ability of zinc binding is a common feature of rps with finger
dues at the amino-terminus, K13, R17, and Y18, are predomi-
motifs, or whether it is unique to YL37a.
nantly to domain II. Yet, mutations in these six residues affect
Mutations to serine of three of the four cysteines in the zinc
binding to both domains more or less equally (Fig. 4). Binding
to the two domains is specific since a forty-fold excess of nonra- finger motif of YL37a (C42S, C57S, and C60S) abolish tetrahe-
dioactive domain II rRNA did not affect binding of MBP-YL37a dral coordination of zinc; nonetheless, cells with the mutations
to radioactive domain III rRNA; and nonradioactive domain do not suffer a significant impairment of growth; and binding to
III rRNA did not inhibit binding to domain II.15 Still the origi- rRNA of the mutant proteins is not affected. The results force one
nal observation is a paradox that implies cooperativity in the to conclude, either that the mutations do not materially affect the
binding of YL37a to the two rRNA domains. The possibility structure of the motif despite not binding zinc, or that the motif
has been tested and there is cooperativity but it is not recipro- fold is not essential for the function of YL37a. Whether the motif
cal. An excess of nonradioactive domain III RNA increased the can fold into its characteristic three-dimensional conformation in
binding of MBP-YL37a to radioactive domain II RNA 1.7 fold. the absence of the coordination of the metal is not known. It is
In the reciprocal experiment domain II RNA had no effect on possible that zinc only stabilizes the folded state and that binding
binding of MBP-YL37a to radioactive domain III.15 It appears to rRNA is an alternate way to maintain the conformation of the
that binding of residues in the zinc finger motif to domain III protein, a conformation determined in the first instance by the
alters the conformation of the amino-terminus of YL37a so as amino acid sequence. Based on the results of the experiments with
to favor association with domain II. YL37a and of the natural experiments during the evolution of, for
If we assume that the rps only condition the folding and the example, the S14 family of rps (see earlier), we favor the second-
end state conformation of rRNA, then the corollary that de- ary possibility: that the fold of the zinc finger motif in YL37a and
rives from this study is that the sites in domains II and III to perhaps in other rps is not essential for function.
which YL37a binds must contribute directly or indirectly to
one or another of the partial reactions of protein synthesis.
98 Zinc Finger Proteins: From Atomic Contact to Cellular Function
Acknowledgements 13. Cherry JM, Adler C, Ball C et al. SGD: Saccharomyces Genome
Database. Nucleic Acids Res 1998; 26:73-79.
The work was supported by a grant from the National Insti- 14. Rivlin AA, Chan YL, Wool IG. The contribution of a zinc finger
tutes of Health (GM 33702). motif to the function of yeast ribosomal protein YL37a. J Mol Biol
1999; 294:909-919.
References 15. Dresios J, Chan YL, Wool IG. The role of the zinc finger motif and
1. Wool IG, Chan YL, Glück A. Structure and evolution of mammalian of the residues at the amino-terminus in the function of yeast
ribosomal proteins. Biochem Cell Biol 1995; 73:933-947. ribosomal protein YL37a. J Mol Biol 2002; 316:475-488.
2. Wool IG, Chan YL, Glück A. Mammalian ribosomes: The structure 16. Boysen RI, Hearn MT. The metal binding properties of the CCCH
and the evolution of the proteins. In: Hershey JWB, Mathews MF, motif of the 50S ribosomal protein L36 from Thermus thermophilus.
Sonnenberg N, eds. Translational Control. Cold Spring Harbor, New J Pept Res 2001; 57:19-28.
York: Cold Spring Harbor Press, 1996:685-732. 17. Fowle DA, Stillman MJ. Comparison of the structures of the
3. Mager WH, Planta RJ, Ballesta JPG et al. A new nomenclature for metal-thiolate binding site in Zn(II)-, Cd(II)-, and
the cytoplasmic ribosomal proteins of Saccharomyces cerevisiae. Hg(II)-metallothioneins using molecular modeling techniques. J
Nucleic Acids Res 1997; 25:4872-4875. Biomol Struct Dyn 1997; 14:393-406.
4. Wool IG. The bifunctional nature of ribosomal proteins and 18. Parraga G, Horvath SJ, Eisen A et al. Zinc-dependent structure of a
speculations on their origins. In: Nierhaus KH, Franceschi F, single-finger domain of yeast ADR1. Science 1988; 241:1489-1492.
Subramanian AR, Erdmann VA, Wittmann-Liebold B, eds. The 19. Vallee BL, Auld DS. Cocatalytic zinc motifs in enzyme catalysis.
Translational Apparatus. New York: Plenum Press, 1993:727-737. Proc Natl Acad Sci USA 1993; 90:2715-2718.
5. Wool IG. Extraribosomal functions of ribosomal proteins. In: Green 20. Stern S, Powers T, Changchien LM et al. RNA-protein interactions
R, Schroeder R, eds. Ribosomal RNA and Group I Introns. Austin, in 30S ribosomal subunits: Folding and function of 16S rRNA.
TX: RG Landes, 1996:153-178. Science 1989; 244:783-790.
6. Wool IG. Extraribosomal functions of ribosomal proteins. Trends 21. Draper DE. Themes in RNA-protein recognition. J Mol Biol 1999;
Biochem Sci 1996; 21:164-165. 293:255-270.
7. Rice PA, Steitz TA. Ribosomal protein L7/L12 has a helix-turn-helix 22. Allers J, Shamoo Y. Structurebased analysis of protein-RNA
motif similar to that found in DNA-binding regulatory proteins. interactions using the program ENTANGLE. J Mol Biol 2001;
Nucleic Acids Res 1989; 17:3757-3762. 311:75-86.
8. Chan Y-L, Suzuki K, Olvera J, Wool IG. Zinc finger-like motifs in 23. Ramakrishnan V, White SW. Ribosomal protein structures: Insights
rat ribosomal proteins S27 and S29. Nucleic Acids Res 1993; into the architecture, machinery and evolution of the ribosome.
21:649-655. Trends Biochem Sci 1998; 23:208-212.
9. Schwabe JWR, Rhodes D. Beyond zinc fingers: Steroid hormone 24. Nogi Y, Yano R, Nomura M. Synthesis of large rRNAs by RNA
receptors have a novel structural motif for DNA recognition. Trends polymerase II in mutants of Saccharomyces cerevisiae defective in
Biochem Sci 1991; 16:291-296. RNA polymerase I. Proc Natl Acad Sci USA 1991; 88:3962-3966.
10. Ban N, Nissen P, Hansen J et al. The complete atomic structure of 25. Urlaub H, Kruft V, Bischof O et al. Protein-rRNA binding features
the large ribosomal subunit at 2.4 Å resolution. Science 2000; and their structural and functional implications in ribosomes as
289:905-920. determined by cross-linking studies. The EMBO J 1995;
11. Harms J, Schluenzen F, Zarivach R et al. High resolution structure 14:4578-4588.
of the large ribosomal subunit from a mesophilic eubacterium. Cell 26. Misra VK, Hecht JL, Sharp KA et al. Salt effects on protein-DNA
2001; 107:679-88. interactions. The cI repressor and Eco RI endonuclease. J Mol Biol
12. Wimberly BT, Brodersen DE, Clemons WM et al. Structure of the 1994; 238:264-280.
30S ribosomal subunit. Nature. 2000; 407:327-348.
CHAPTER 15
L Proteins
IM domain is a unique double-zinc finger motif found in
a variety of proteins such as homeodomain transcription Proteins containing LIM domains are involved in a variety of
factors, kinases, and adaptors. The LIM-containing pro- biological processes including regulation of gene transcription,
teins are involved in diverse biological processes including cy- cytoskeleton organization, cell lineage specification and organ de-
toskeleton organization, cell lineage specification and organ de- velopment. LIM-containing proteins can have significantly
velopment. Dysfunctions of LIM domains induce pathological
different functions even in the same groups. Three commonly
effects including muscle detachment, embryonic lethality, and
oncogenesis. Acting as a protein-protein interaction motif, the known subclasses of group 1 LIM-containing proteins are: (i)
LIM domain has a conserved scaffold but highly variable mode LIM-homeodomain (LIM-HD) transcription factors, which par-
in recognizing diverse target proteins. This chapter describes the ticipate in activation of transcription and development of the
structure and function of LIM domain proteins and discusses nervous system.6 (ii) LMO proteins that are thought to act as
the molecular basis by which the domain mediates molecular adapters involved in development and oncogenesis.7
protein-protein interactions. The LMO-1 and LMO-2 were first discovered at translocation
breakpoints of T-cell leukemia patients.8,9 The LMO family has
Introduction been now extended to LMO-3 and LMO-4.10 (iii) LIM kinases,
which are involved in cytoskeleton establishment and regulation.
LIM domains were discovered as a novel double zinc finger
sequence motif, C-X2-C-X17-19-H-X2-C-X2-C-X2-C-X15-19-C, LIM kinases (Limk1 and Limk2) are known to phosphorylate
with conserved distribution of cysteine and histidine residues in and thereby inhibit cofilin.11,12 Cofilin functions by depolymer-
lin-11, Isl-1 and mec-3 (hence - LIM) gene products.1,2 LIM do- izing F-actin, hence activation of LIM-kinases by the upstream
mains are found in proteins from varied branches of eukaryotes: small GTPase dependent PAK1 or ROCK kinases leads to accu-
plants, animals, fungi and mycetozoa (D. discoideum) and have mulation of actin filaments.13
been classified as A, B, C or D types based on their sequence A functionally well-characterized subclass of group 2 LIM pro-
similarity.3,4 Several domains fall outside of these classes due to teins is cysteine-rich proteins (CRP1 - CRP3), which are promi-
pronounced sequence divergence. nent in myogenesis and muscle structure. Expression of all three
LIM domains have been also classified into three groups, 1, 2 is upregulated at distinct stages of embryo development and
and 3. LIM domains from classes A and B are most frequently myogenesis.13 Antisense disruption of CRP3 blocks terminal dif-
found fused to other functional domains such as kinase domain. ferentiation in myoblasts.14 Interestingly, in CRP3 knock-out the
But nuclear LIM-only (LMO) proteins, which contain two LIM development appears to be normal, but the arrangement of stri-
domains only, are also important members of this group, called ated muscle myofibrils is disrupted, arguing for a structural or
group 1.5 Class C domains are called group 2. This group’s pro- structureregulating role of CRP3. A structural function for all
teins often contain two copies of LIM domain per protein mol- three CRPs in fibroblast cells is implied by robust intracellular
ecule where the domains are more similar to each other than those staining at actin filaments15 and in focal adhesions (at least for
from classes A and B. There are no other domains in association
CRP1).16 Given that these proteins bind to α-actinin (a well
with the C-class LIM domains. It was suggested that class
C-containing proteins arose via internal duplication of LIM do- known actin bundle cross-linking protein) and to zyxin (a focal
mains.4 Class D serves as a sorting bin for LIM domains lacking adhesion marker protein), such intracellular localization is ex-
homology to other classes and with little similarity among them- pected.13 CRP3 is also reported to bind to βI-spectrin in muscle
selves. Given that sequence similarity of LIM domains in classes cells and colocalizes with spectrin at the plasma membrane where
A - C predicts rather well the global architecture of the entire it overlies Z- and M- lines of striated muscle.17 The composition
protein, perhaps it is not surprising that the motley collection of and function of these regions of muscle cells is rather similar to
LIM domains in class D appears in dissimilar proteins. These are focal adhesions, therefore CRP proteins have similar structural
called group 3 and harbor from one to five LIM domains, either roles in different cell types.
with or without additional functional domains or motifs.
Zinc Finger Proteins: From Atomic Contact to Cellular Function, edited by Shiro Iuchi and Natalie Kuldell.
©2005 Landes Bioscience/Eurekah.com and Kluwer Academic/Plenum Publishers.
100 Zinc Finger Proteins: From Atomic Contact to Cellular Function
The class D LIM domains that belong to group 3 are exempli- group. The cysteine to aspartic acid mutant folded and 113Cd
fied by zyxin (3 LIM domains), paxillin (4 LIM domains), and NMR and Co UV-visible spectra are consistent with conversion
PINCH (5 LIM domains) - three adaptor proteins that are classi- of the S 4 to the (apparently perfectly functional) S 3 O
cal markers of focal adhesion (FA) plaques.18,19 All are thought to Zn2+-coordination sphere. Finally, the arrangements of metal co-
act as adaptor proteins that orchestrate assembly of FAs. These ordinating residues in the LIM domains were derived from NMR
proteins regulate cell shape change and spreading via distinct structure determination.23,24
LIM-mediated protein-protein interactions.
LIM Domain As Protein Interaction Motif
Zinc Coordination in LIM Domains Currently, the only function ascribed to LIM domains is
The presence of a conserved pattern of cysteine and histidine protein-protein recognition. Two recent reviews10,13 present an
residues in LIM domains suggested that LIM domains likely con- impressive list of about 30 known protein-protein interaction pairs
tain structural Zn2+ ions and this is indeed so. From the current for LIM domains. Given that the LIM domain count is in the
data, the consensus pattern appears to be C - X2 - C - X16-23 - H/ hundreds and the consensus opinion in the field that LIM do-
C - X2 - C/H - X2 - C - X2 - C - X15-30 - C - X1-3 - C/H/D (/ mains function by recognizing target proteins, it is clear that this
indicates alternative amino acid residue). Metal coordination by is but the tip of the iceberg. The following sections present some
LIM domains was extensively studied on two members of the well-characterized examples of LIM domain-mediated
family, CRP1 and CRIP proteins. These proteins contain two protein-protein interactions.
and one copies of LIM domains respectively, with metal coordi-
nation consensus residues CCHCCCCC, which all belong to class LIM-HD and LMO Proteins
C. It was established that these LIM domains, expressed in heter- Well-known partners for LIM domains, discovered simulta-
ologous bacterial hosts or isolated from natural sources, contain neously for LIM-HD25 and LMO26 proteins, are proteins from
approximately 2 equivalents of zinc per equivalent of protein.20 the Ldb (LIM domain binding) family. For LIM-HD proteins,
Bound zinc was also demonstrated for native preparations of in nearly all cases protein-protein interactions depend on LIM
zyxin.16 Metal ions can be stripped from LIM domains by dena- domains.6 It is the LIM-Ldb interaction mediated formation of
turation with low pH or guanidinium hydrochloride. A number different heterodimers of LIM-HD proteins that is thought to
of other metals- Cd2+, Co2+, Cu(I),- can be reconstituted into determine distinct identities for motor neurons during develop-
LIM domains upon refolding, again at two ions per domain as ment.27 Ldb1 specifically binds to nuclear group 1 LIM proteins:
established by atomic absorption spectroscopy16 and by titration.21 LMO1, LMO2, Lmx1, Lhx1, Isl1, Mec3, but not LIM kinase,
UV spectroscopy shows bands consistent with formation of paxillin, zyxin, enigma or CRP1.26 Deletion mutagenesis was used
S-metal bonds in the complex.21 A strong CoS4 band dominates to determine the LIM-interacting part of Ldb1. A 39-residue
the Co2+ spectrum. Interestingly, the two metal sites appear to stretch within Ldb1 (residues 300-338 of mouse protein) was
have different affinities for metal ions. This observation was craftily identified that is sufficient to bind LIM domains.28,29 This re-
exploited by preparing a LIM domain with the high affinity site gion is termed “LID” – LIM interaction domain. The LID re-
occupied by Zn2+ and the second site by Co2+.21 This prepara- gion contains both highly hydrophobic and hydrophilic stretches
tion yielded a spectrum consistent with CoS3N type complex. of residues and encompasses a 34 residue sequence where human
1D 113Cd NMR spectra of Cd2+-substituted LIM domains pro- and C. elegans proteins share 23 identical residues (additionally, 7
vided further information on the coordination. Chemical shifts residues are conservatively substituted), which implies strong evo-
of two signals per LIM domain observed in Cd2+ NMR are con- lutionary pressure to preserve the LIM - LID interaction. It is not
sistent with Cd-S4 coordination.21 The chemical shift of one of clear so far what features of LIM domains are responsible for rec-
the signals, however, is in the region consistent with both S4 and ognition of LID: in some instances it is LIM1 (first LIM), while
in others the LIM2 (second LIM) domain of the LIM-HD pro-
S3N coordination. The possibility for S3N coordination was
tein is implicated. Also, some conflicting data exist for X. laevis
strongly supported by observation of J-coupling between Cd2+
Xlim-1 where both LIM1 and LIM2 appear to bind to Ldb1.29
and hydrogen atoms in the approximately 7 parts per million
LID of Ldb1 appears to be functionally independent from the
region, which can be explained only by the presence of Cd2+ co- Ldb1 dimerization domain,28 and monomeric behaviour of LID
ordinated to a histidine imidazole ring.21 However, none of the – LIM (of LMO2 or LMO4 proteins) fusion proteins suggests
above studies determined which of the 7 conserved cysteines form 1:1 binding stoichiometry. 30 The importance of LIM
the S4 and the S3N site respectively. Based on a generally valid domain-mediated interactions for LIM-HD proteins is under-
assumption that individual C-X2-C/H units coordinate the same scored by the finding that Y116C mutation in a human LIM-HD
metal ion, Beckerle and coworkers used mutagenesis to address protein (Lhx3) causes a genetic disease, combined pituitary hor-
this question.22 Results of Co2+ UV-visible and 113Cd NMR spec- mone deficiency.31 Residues equivalent to Tyr116 play a central
troscopy on the mutant proteins indicated that the N-terminal 4 role in hydrophobic cores of known LIM domain structures, hence
residues in the protein sequence (CCHC) form the S3N site, while this mutation is predicted to cause unfolding and loss of function
the S4 site is composed of the 4 C-terminal cysteines. On a side of the LIM domain. Indeed, mutant Lhx3 behaves consistently
note, among the mutants tested was a cysteine to aspartic acid with such prediction.32 While binding to Ldb by LIM-HD and
substituted domain, in which the C-terminal CCCC zinc finger LMO proteins is important for the activation of transcription, a
was transformed into CCCD. Aspartic acid was predicted to be a recently discovered RLIM protein seems to have a negative regu-
zinc binding residue in many LIM domains among LIM-HD latory role33 by binding to Lhx3 via its basic region of about 100
proteins, but generally it is not common as a Zn2+-coordinating residues.34
LIM Domain and Its Binding to Target Proteins 101
Zyxin ing, with LIM3 being absolutely necessary for detectable bind-
A number of binding partners are known for zyxin: CRP pro- ing, while the presence of LIM2 and LIM4 clearly augments the
teins,35,16 α-actinin,36 proto-oncogene Vav,37 p130Cas,38 and mem- extent of binding.55 Given that the LIM domains of paxillin and
bers of the Ena/VASP family of proteins.39 LIM1 of zyxin is nec- His-5 share 68% sequence similarity,51 it is not obvious whether
essary and sufficient for binding to CRP, as demonstrated by blot the protein interface is the same for both members of the family,
overlay and GST pull-down assays.40 Interestingly, although two or only the net result, formation of the complex with PTP-PEST,
zinc fingers in LIM domains form a single domain, individual is conserved.
fingers appear to have some functional independence.22,41 This is
also implied by the study that the N-terminal and not the Enigma Protein
C-terminal zinc finger of zyxin LIM1 is responsible for CRP1 A relatively well-characterized system is Enigma protein that
recognition.42 The C-terminal zinc finger in LIM1 of zyxin could contains three LIM domains at its C terminus. Yeast two-hybrid
be swapped with the C-terminal zinc finger from another LIM screening revealed that enigma binds to the insulin receptor
domain (but not deleted!) with no apparent effect on binding to (InsR) internalization motif. 56 Point mutants in two
CRP1. In CRP1, LIM1 and LIM2 appear to cooperate in bind- internalization-driving tetrapeptides within this segment some-
ing to zyxin, so that deletion of either reduces the extent of bind- what diminished (weaker peptide) and thoroughly blocked (stron-
ing more than twenty-fold in blot overlay assay.43 The entire LIM1 ger peptide) interaction with full-length enigma and LIM3 con-
is required, but the C-terminal zinc finger in LIM2 is disposable. struct. The interaction is specific since the internalization segments
The linker between the two LIM domains is inert. Taken together of IGF1, EGF, sequence-optimized EGF, Transferrin and LDL
these studies present an interesting picture—LIM domains in receptors failed to bind. However, the paired sequence-optimized
CRP1 are linked by a long disordered segment and structurally EGF receptor motif (the tandem repeat of the internalization motif
seem to be entirely independent.44 Zyxin was also suggested to of EGF that has been optimized to display high activity in the
bind to actin filaments via α-actinin and play a regulatory rather internalization assay) did bind, albeit to a lesser extent. The rel-
than a structural role in FAs. Ample biochemical data support evance of the last observation is not clear, since LIM2 in Mec-3,
this notion. Interaction with Ena/VASP proteins is important in a nuclear LIM-HD protein which has no known function in re-
targeting zyxin to lamellipodia,45 although under normal cellular ceptor endocytosis also binds to the paired sequence-optimized
situations the vast majority of zyxin resides in focal adhesions.46 EGFR motif. Furthermore, this paired sequence-optimized mo-
Ena/VASP proteins are known to bind profilin—an actin poly- tif binds equally well to LIM2, LIM3 and the C-terminally trun-
merizing protein. In an important development, it was shown cated LIM3 of enigma.56 Yeast two-hybrid screening was also in-
that zyxin-VASP complex is able to initiate actin polymerization.47 strumental in discovering that the LIM domain fragment of
enigma also binds to Ret Receptor tyrosine kinase.57 Mutating
Paxillin the most C-terminal tyrosine at position 1062 to phenylalanine
or deleting twenty-three C-terminal residues (including Tyr1062)
Paxillin is one of the central proteins in FAs acting as a scaf-
from Ret abolished the interaction. In fact, a 61 C-terminal resi-
fold for a myriad of binding partners. 48 The LIM
due segment of Ret is perfectly capable of binding to enigma58 in
domain-containing C-terminal half of paxillin is known to bind
a GST pull-down assay. Their association is not
to protein tyrosine phosphatase – PEST49,50 and this binding is phosphorylation-dependent, and is specific for enigma, as zyxin
postulated to bring PTP-PEST into the vicinity of its target, phos- and CRP1 did not bind. Enigma-Ret binding was confirmed by
phorylated p130Cas.51 PTP-PEST is required for disassembly of the observation that coinjection of DNA for the LIM
FAs possibly by dephosphorylating p130Cas and paxillin52 and domain-containing part of enigma with the Ret construct into
therefore modulating their FA targeting regions. The LIM do- fibroblasts abolished the mitogenic effect of Ret in a
main region is responsible for targeting paxillin to Fas,53 via a yet dominant-negative style.57 Unlike InsR, Ret peptide is specifi-
unknown binding event. It was recently shown that paxillin LIM cally recognized by enigma LIM2,58 while the LIM domains from
domains bind to α- and γ-tubulin as well,54 which led to a pro- LIM-HD proteins, zyxin, CRP1 or paxillin do not bind Ret or
posal that paxillin is responsible for the interplay between actin InsR peptides. Which residues are important for Ret - LIM2 rec-
filaments and microtubules. A fragment containing both the ognition was further assessed with a competition assay.58 Wild
C-terminal LIM3 and LIM4 of paxillin is competent to bind to type (NKLY1062) and mutant (AKLA and NKLF) peptides
PTP-PEST, while LIM3 or LIM4 in isolation are not.51 Dele- within an otherwise intact 20-residue segment of Ret Tyr1062
tions of LIM3 or LIM4 from an otherwise intact paxillin or its region were used as competitors for Ret in GST-pull down. While
fragments abolish binding, as does disruption of the N- or AKLA peptide and InsR LIM3-recognizing peptide failed to in-
C-terminal zinc fingers within LIM3 and the N-terminal finger hibit the interaction, NKLY and NKLF peptides were inhibitors.
in LIM4 with point mutagenesis. Comprehensive deletion of The capacity of NKLF peptide to act as an inhibitor, however,
PTP-PEST showed that a part of the protein that centers on the contradicts previous data57 that the Y1062F mutation disrupts
2nd (of five total) proline-rich region of PTP-PEST is the binding binding. Also, the GST pull-down titration experiment showed
epitope. The structure of this PTP-PEST peptide is not known, that lower amounts of Y1062F mutant Ret are retained by
but the 50% content of proline likely precludes formation of a GST-LIM2. The apparent discrepancy may arise from the fact
globular fold. It is interesting that a close homologue of paxillin, that the amount of observed binding in GST-pull down assay
Hic-5 protein, also binds to PTP-PEST.55 Broadly the same re- depends on times and volumes of the binding and washing steps,
gion of the phosphatase, rich in prolines, is involved. Here, in in addition to the affinity of interaction. Also, 20-residue pep-
contrast to paxillin, LIM2, LIM3 and LIM4 are involved in bind- tides may be suboptimal for binding, since high concentrations
102 Zinc Finger Proteins: From Atomic Contact to Cellular Function
Figure 3. Hydrophobic core in cCRP2 LIM2. Backbone is shown as a Figure 4. Overlay of CRIP LIM domain (gray, PDB ID 1IML) with
tube. The hydrophobic residues in the core are labeled. cCRP1 LIM2 (black), illustrating the relative twist of the C-terminal
Zn-finger in CRIP. Backbone residues 1-23 of CRIP were best fit super-
tively. The importance of these residues is reflected in their conser- imposed onto corresponding residues of LIM2. a) Overall view; b and
vation: among currently known LIM domain sequences, the posi- c) – top view of individual fingers. b) N-terminal Zn-fingers, used for
tion equivalent to equivalent to Phe143 is occupied by phenylala- superimposition. c) C-terminal fingers, showing rotation in CRIP rela-
nine (80.4%), leucine (10.8%), or tyrosine, valine, tryptophan tive to LIM2.
or isoleucine (6.9%) (total 98.1%). It is quite possible that the
preference for aromatic residues at these two positions is due to binding residues are the key structural determinants, which orches-
strengthening of the hydrophobic cluster by additional aromatic π trate the fold of this fragment. Other residues in Lasp-1 also adopt
- π interaction. Other hydrophobic residues (also highly conserved) conformations close to those observed in the N-terminal halves of
pile on the sides of these two: Val131 from the N-terminal zinc full-length LIM domains.
finger and Leu152, Leu157, Ieu164 from the C-terminal finger. A key issue about the LIM domain is its structural basis in
Interestingly, Perez-Alvarado et al63 also discovered a rather good recognizing target proteins. Recent chemical shift mapping stud-
superposition between the C-terminal zinc finger of LIM2 and the ies on PINCH LIM domains have provided important insights
DNA-binding zinc fingers in the glucocorticoid receptor and in into this issue: 23,24 For PINCH LIM1, it was shown that
the GATA-1 transcription factor (backbone rmsds 2.0 and 1.7 Å, C-terminal hydrophobic patch and several adjoining residues are
respectively), which suggested a possible a role for LIM domains as most perturbed upon binding to ankyrin repepat (ANK) domain
DNA-binding molecules. However, so far this hypothesis is not of ILK (Fig. 5a) whereas residues from N-terminal basic region
supported by experimental evidence.
Since the report of cCRP LIM2 structure, numerous structures
of other LIM domains have been determined by NMR spectros-
copy. These include all three groups of LIM domains: (i) LIM1
domains of LMO2 and LMO4,64 which belong to group 1. (ii)
LIM domains in group 2 family including LIM2 of cCRP1 as
mentioned above, CRIP protein (one LIM domain),65 full-length
cCRP1 (LIM1 and LIM2),43 LIM1 and LIM2 of qCRP2.66,67 In
addition, the structure of an isolated N-terminal zinc finger of LIM
domain in Lasp-1 has been determined.41 (iii) LIM1 and LIM4
domains of PINCH, which belong to group 3.23,24 The main struc-
tural features of these LIM domains were found to be the same as
cCRP LIM2: four short β hairpins with a C-terminal α helix; two
“rubredoxin knuckles”; small hydrophobic core contributed by both
zinc fingers. Although the global fold is similar among all known
LIM domain structures, there are some differences, notably the
relative orientation of the two zinc fingers varies about the long
axis of the domain (Fig. 4). Different orientations between the fin-
gers in LIM domains may be important for protein-protein recog-
nition. Each zinc finger appears to be an independent folding unit Figure 5. Comparison of PINCH LIM1 and LIM4 binding modes to
despite the interactions between the two fingers. This is evidenced target proteins: a) the major binding surface (shaded area with labeled
by the determination of the structure of isolated N-terminal CCHC residues) of LIM1 (PDB ID 1G47) to ILK ANK domain; b) the major
half of the LIM domain of Lasp-1 protein.41 Clearly, four zinc binding surface (shaded area with labeled residues)
104 Zinc Finger Proteins: From Atomic Contact to Cellular Function
of PINCH LIM4 appear to be responsible for binding to SH3 19. Wang Y, Gilmore TD. Zyxin and paxillin proteins: Focal adhesion
domain (Fig. 5b). The structure of LMO-Ldb1 complex revealed plaque LIM domain proteins go nuclear. Biochim Biophys Acta 2003;
a rather extensive binding interface involving both zinc finger 1 1593(2-3):115-20.
20. Michelsen JW, Schmeichel KL, Beckerle MC et al. The LIM motif
and zinc finger 2.64 Hence, LIM domains do not appear to have defines a specific zinc-binding protein domain. Proc Natl Acad Sci
a single, well defined binding epitope, but rather present on their USA 1993; 90:4404-4408.
surfaces clusters of different functional groups, individually tai- 21. Kosa JL, Michelsen JW, Louis HA et al. Common metal ion
lored for diverse binding partners. Such diverse recognition mode coordination in LIM domain proteins. Biochemistry 1994;
has also been observed in PTB and SH2 domains.68,69 Clearly, 33:468-477.
more structures of LIM domains in complex with target proteins 22. Michelsen JW, Sewell AK, Louis HA et al. Mutational analysis of
the metal sites in an LIM domain. J Biol Chem 1994;
are necessary for a thorough understanding of this fascinating 269:11108-11113.
protein interaction motif. 23. Velyvis A, Yang Y, Wu C et al. Solution structure of the focal
adhesion adaptor PINCH LIM1 domain and characterization of its
Acknowledgments interaction with the integrin-linked kinase ankyrin repeat domain. J
Biol Chem 2001; 276:4932-4939.
The work was supported by NIH grants to J.Q. (HL58758, 24. Velyvis A, Vaynberg J, Yang Y et al. Structural and functional insights
GM62823). into PINCH LIM4 domain-mediated integrin signaling Nat Struct
Biol 2003; 10:558-564.
References 25. Agulnick AD, Taira M, Breen JJ et al. Interactions of the
1. Karlsson O, Thor S, Norberg T et al. Insulin gene enhancer binding LIM-domain-binding factor Ldb1 with LIM homeodomain proteins.
protein Isl-1 is a member of a novel class of proteins containing Nature 1996; 384:270-272.
both a homeo- and a Cys-His domain. Nature 1990; 344:879-882. 26. Jurata LW, Kenny DA, Gill GN. Nuclear LIM interactor, a
2. Freyd G, Kim SK, Horvitz HR. Novel cysteine-rich motif and rhombotin and LIM homeodomain interacting protein, is expressed
homeodomain in the product of the Caenorhabditis elegans cell early in neuronal development. Proc Natl Acad Sci USA 1996;
lineage gene lin-11. Nature 1990; 344:876-879. 93:11693-11698.
3. Taira M, Evrard JL, Steinmetz A et al. Classification of LIM proteins. 27. Jurata LW, Pfaff SL, Gill GN. The nuclear LIM domain interactor
Trends Genet 1995; 11:431-432. NLI mediates homo- and heterodimerization of LIM domain
4. Dawid IB, Toyama R, Taira M. LIM domain proteins. CR Acad Sci transcription factors. J Biol Chem 1998; 273:3152-3157.
III 1995; 318:295-306. 28. Jurata LW, Gill GN. Functional analysis of the nuclear LIM domain
5. Dawid IB, Breen JJ, Toyama R. LIM domains: Multiple roles as interactor NLI. Mol Cell Biol 1997; 17:5688-5698.
adapters and functional modifiers in protein interactions. Trends 29. Breen JJ, Agulnick AD, Westphal H et al. Interactions between LIM
Genet 1998; 14:156-162. domains and the LIM domain-binding protein Ldb1. J Biol Chem
6. Hobert O and Westphal H. Functions of LIM-homeobox genes. 1998; 273:4712-4717.
Trends Genet 2000; 16:75-83. 30. Deane JE, Sum E, Mackay JP et al. Design, production and
7. Rabbitts, T.H. LMO T-cell translocation oncogenes typify genes characterization of FLIN2 and FLIN4: The engineering of
activated by chromosomal translocations that alter transcription and intramolecular ldb1: LMO complexes. Protein Eng 2001;
developmental processes. Genes Dev 1998; 12(17):2651-7 14:493-499.
8. Larson RC, Osada H, Larson TA et al. The oncogenic LIM protein 31. Netchine I, Sobrier ML, Krude H et al. Mutations in LHX3 result
Rbtn2 causes thymic developmental aberrations that precede in a new syndrome revealed by combined pituitary hormone
malignancy in transgenic mice. Oncogene 1995; 11(5):853-62 deficiency. Nat Genet 2000;25:182-186.
9. Larson RC, Lavenir I, Larson TA et al. Protein dimerization between 32. Howard PW, Maurer RA. A Point Mutation in the LIM Domain of
Lmo2 (Rbtn2) and Tal1 alters thymocyte development and Lhx3 Reduces Activation of the Glycoprotein Hormone alpha
potentiates T cell tumorigenesis in transgenic mice. EMBO J 1996; -Subunit Promoter. J Biol Chem 2001; 276:19020-19026.
15(5):1021-7 33. Bach I, Rodriguez-Esteban C, Carriere C et al. RLIM inhibits
10. Bach I. The LIM domain: Regulation by association. Mech. Dev functional activity of LIM homeodomain transcription factors via
2000; 91:5-17. recruitment of the histone deacetylase complex. Nat Genet 1999;
11. Arber S, Barbayannis FA, Hanser H et al. Regulation of actin 22:394-399.
dynamics through phosphorylation of cofilin by LIM-kinase. Nature 34. Ostendorff HP, Peirano RI, Peters MA et al.
1998; 39:805-809. Ubiquitination-dependent cofactor exchange on LIM homeodomain
12. Yang,N, Higuchi,O, Ohashi,K et al. Cofilin phosphorylation by transcription factors. Nature 2002; 416:99-103.
LIM-kinase 1 and its role in Rac-mediated actin reorganization. 35. Crawford AW, Pino JD, Beckerle MC. Biochemical and molecular
Nature 1998; 393:809-812. characterization of the chicken cysteine-rich protein, a
13. Khurana,T, Khurana, B, Noegel, AA. LIM proteins: Association with developmentally regulated LIM-domain protein that is associated with
the actin cytoskeleton. Protoplasma 2002; 219:1-12. the actin cytoskeleton. J Cell Biol 1994; 124:117-127.
14. Arber S, Halder G, Caroni P. Muscle LIM protein, a novel essential 36. Crawford AW, Michelsen JW, Beckerle MC. An interaction between
regulator of myogenesis, promotes myogenic differentiation. Cell zyxin and alpha-actinin. J Cell Biol 1992; 116:1381-1393.
1994; 79:221-231. 37. Hobert O, Schilling JW, Beckerle MC et al. SH3 domain-dependent
15. Louis HA, Pino JD, Schmeichel KL et al. Comparison of three interaction of the proto-oncogene product Vav with the focal contact
members of the cysteine-rich protein family reveals functional protein zyxin. Oncogene 1996; 12:1577-1581.
conservation and divergent patterns of gene expression. J Biol Chem 38. Yi J, Kloeker S, Jensen CC et al. Members of the Zyxin family of
1997; 272:27484-27491. LIM proteins interact with members of the p130Cas family of signal
16. Sadler I, Crawford AW, Michelsen JW et al. Zyxin and cCRP: Two transducers. J Biol Chem 2002; 277:9580-9589.
interactive LIM domain proteins associated with the cytoskeleton. J 39. Drees B, Friederich E, Fradelizi J et al. Characterization of the
Cell Biol 1992; 119:1573-1587. interaction between zyxin and members of the Ena/
17. Flick MJ, Konieczny SF. The muscle regulatory and structural protein vasodilator-stimulated phosphoprotein family of proteins. J Biol
MLP is a cytoskeletal binding partner of betaI-spectrin. J Cell Sci Chem 2000; 275:22503-22511.
2000; 113(Pt 9):1553-1564. 40. Schmeichel KL, Beckerle MC. The LIM domain is a modular
18. Wu C, Dedhar S. Integrin-linked kinase (ILK) and its interactors: protein-binding interface. Cell 1994; 79:211-219.
A new paradigm for the coupling of extracellular matrix to actin
cytoskeleton and signaling complexes. J Cell Biol 2001; 155:505-510.
LIM Domain and Its Binding to Target Proteins 105
41. Hammarstrom A, Berndt KD, Sillard R et al. Solution structure of 56. Wu RY, Gill GN. LIM domain recognition of a tyrosine-containing
a naturally-occurring zinc-peptide complex demonstrates that the tight turn. J Biol Chem 1994; 269:25085-25090.
N-terminal zinc-binding module of the Lasp-1 LIM domain is an 57. Durick K, Wu RY, Gill GN et al. Mitogenic Signaling by Ret/ptc2
independent folding unit. Biochem 1996; 35:12723-12732. Requires Association with Enigma via a LIM Domain. J Biol Chem
42. Schmeichel KL, Beckerle MC. Molecular dissection of a LIM 1996; 271:12691-12694.
domain. Mol Biol Cell 1997; 8:219-230. 58. Wu RY, Durick K, Songyang Z et al. Specificity of LIM Domain
43. Schmeichel KL, Beckerle MC. LIM domains of cysteine-rich protein Interactions with Receptor Tyrosine Kinases. J Biol Chem 1996;
1 (CRP1) are essential for its zyxin-binding function. Biochem J 271:15934-15941.
1998; 331(Pt 3):885-892. 59. Rearden A. A new LIM protein containing an autoepitope
44. Yao X, Perez-Alvarado GC, Louis HA et al. Solution structure of homologous to “senescent cell antigen.” Biochem Biophys Res
the chicken cysteine-rich protein, CRP1, a double-LIM protein Commun 1994; 201:1124-1131.
implicated in muscle differentiation. Biochem 1999; 38:5701-5713. 60. Tu Y, Li F, Goicoechea S et al. The LIM-only protein PINCH
45. Nix DA, Fradelizi J, Bockholt S et al. Targeting of zyxin to sites of directly interacts with integrin-linked kinase and is recruited to
actin membrane interaction and to the nucleus. J Biol Chem 2001; integrin-rich sites in spreading cells. Mol Cell Biol 1999;
276:34759-34767. 19:2425-2434.
46. Beckerle MC. Zyxin: Zinc fingers at sites of cell adhesion. Bioessays 61. Li F, Zhang Y, Wu C. Integrin-linked kinase is localized to cell-matrix
1997; 19:949-957. focal adhesions but not cell-cell adhesion sites and the focal adhesion
47. Fradelizi J, Noireaux V, Plastino J et al. ActA and human zyxin localization of integrin-linked kinase is regulated by the
harbour Arp2/3-independent actin-polymerization activity. Nat Cell PINCH-binding ANK repeats. J Cell Sci 1999; 112(Pt
Biol 2001; 3:699-707. 24):4589-4599.
48. Turner CE. Paxillin and focal adhesion signaling. Nat Cell Biol 2000; 62. Tu Y, Li F, Wu C. Nck-2, a novel Src homology2/3-containing
2:E231-E236. adaptor protein that interacts with the LIM-only protein PINCH
49. Cote JF, Turner CE, Tremblay ML. Intact LIM 3 and LIM 4 and components of growth factor receptor kinase-signaling pathways.
domains of paxillin are required for the association to a novel Mol Biol Cell 1998; 9:3367-3382.
polyproline region (Pro 2) of protein-tyrosine phosphatase-PEST. J 63. Perez-Alvarado GC, Miles C, Michelsen JW et al. Structure of the
Biol Chem 1999; 274:20550-20560. carboxy-terminal LIM domain from the cysteine rich protein CRP.
50. Shen Y, Schneider G, Cloutier JF et al. Direct association of Nat Struct Biol 1994; 1:388-398.
protein-tyrosine phosphatase PTP-PEST with paxillin. J Biol Chem 64. Deane JE, Mackay JP, Kwan AH et al. Structural basis for the
1998; 273:6474-6481. recognition of ldb1 by the N-terminal LIM domains of LMO2 and
51. Cote JF, Turner CE, Tremblay ML. Intact LIM 3 and LIM 4 LMO4. EMBO J 2003, 22(9):2224-33
domains of paxillin are required for the association to a novel 65. Perez-Alvarado GC, Kosa JL, Louis HA et al. Structure of the
polyproline region (Pro 2) of protein-tyrosine phosphatase-PEST. J cysteine-rich intestinal protein, CRIP. J Mol Biol 1996; 257:153-174.
Biol Chem 1999; 274:20550-20560. 66. Konrat R, Weiskirchen R, Krautler B et al. Solution structure of the
52. Shen Y, Lyons P, Cooley M et al. The noncatalytic domain of carboxyl-terminal LIM domain from quail cysteine-rich protein
protein-tyrosine phosphatase-PEST targets paxillin for CRP2. J Biol Chem 1997; 272:12001-12007.
dephosphorylation in vivo. J Biol Chem 2000; 275:1405-1413. 67. Kontaxis G, Konrat R, Krautler B et al. Structure and intramodular
53. Brown MC, Perrotta JA, Turner CE. Identification of LIM3 as the dynamics of the amino-terminal LIM domain from quail cysteine-
principal determinant of paxillin focal adhesion localization and and glycine-rich protein CRP2. Biochem 1998; 37:7127-7134.
characterization of a novel motif on paxillin directing vinculin and 68. Li SC, Zwahlen C, Vincent SJ et al. Structure of a Numb PTB
focal adhesion kinase binding. J Cell Biol 1996; 135:1109-1123. domain-peptide complex suggests a basis for diverse binding
54. Herreros L, Rodriguez-Fernandez JL, Brown MC et al. Paxillin specificity. Nat Struct Biol 1998; 5:1075-83.
localizes to the lymphocyte microtubule organizing center and 69. Mallis RJ, Brazin KN, Fulton DB et al. Structural characterization
associates with the microtubule cytoskeleton. J Biol Chem 2000; of a proline-driven conformational switch within the Itk SH2
275:26436-26440. domain. Nat Struct Biol 2002; 9:900-5.
55. Nishiya N, Iwabuchi Y, Shibanuma M et al. Hic-5, a Paxillin
Homologue, Binds to the Protein-tyrosine Phosphatase PEST
(PTP-PEST) through Its LIM 3áDomain. J Biol Chem 1999;
274:9847-9853.
CHAPTER 16
Abstract
T
he protein family harboring the RING finger motif,
defined as a linear array of conserved cysteines and histidines,
has grown enormously in the last decade. The members of
the family are involved in various biological processes including
growth, differentiation, apoptosis, transcription and also in dis-
eases and oncogenesis. It has been postulated that the RING finger
domains have crucial roles in these phenomena themselves, in some
cases, working with other domains in other proteins, although the
precise mechanisms and common features of RING finger func-
tion have not been fully elucidated. However, most recently, an
accumulating body of evidence has revealed that some of the RING
finger proteins work as E3 ubiquitin ligases in ubiquitin-mediated
specific protein degradation pathway. In this review, we focus on
the RING finger protein with special reference to E3 ligase.
Zinc Finger Proteins: From Atomic Contact to Cellular Function, edited by Shiro Iuchi and Natalie Kuldell.
©2005 Landes Bioscience/Eurekah.com and Kluwer Academic/Plenum Publishers.
RBCC Proteins As Ubiquitin Ligase in the Control of Protein Degradation and Gene Regulation 107
NCL-1, HT2A, and LIN-41.47 NCL is involved in rRNA me- B-box and coiled coil domains.43 This strongly suggests that the
tabolism. HT2A was identified as an interacting partner of the overall architecture of the RBCC/TRIM motif is highly conserved,
HIV Tat protein and Lin41 is involved in posttranscriptional regu- perhaps relating to the motif acting as a scaffold for higher-order
lation of mRNA. The NHL motif has a slight homology with protein-protein interactions.57 Molecular modeling suggests that
WD40 domain, suggesting a protein-protein interaction. The the position and orientation of the B-box (adjacent to the coiled
C-terminal PHD fingers and bromodomains are found in TIF1α coil) would be critical for the correct alignment of the α-helices
and KAP1/TIF1β. The PHD domain is a motif characteristically that form the coiled coil. Interestingly, unlike the RING and coiled
defined by seven cysteines and a histidine that are highly homolo- coil motifs, the B-box is only found in RBCC/TRIM family mem-
gous to the RING motif and is contained in some transcription bers suggesting that it is a critical determinant of the overall motif
factors. The PHD domain of MEKK1 (MEK kinase 1) exhibited and its function.43
E3 ubiquitin ligase activity toward ERK (extracellular As mentioned above, the latest findings of RING fingers in
signal-regulated protein kinase) 2, suggesting a negative regula- E3 ubiquitin ligases imply that the members of this RBCC/TRIM
tory mechanism for decreasing ERK1/2 activity. 48 The subfamily are potential candidates for specific regulators/adopt-
bromodomain is also found in transcription factors, can bind hi- ers in ubiquitin-dependent protein degradation. In fact, some
stones with acetylated lysines and appears to be involved in chro- genes belonging to the subfamily has been proven to act as E3
matin remodeling.49 KAP1/TIF1β and TIF1α are involved in ligase. We next discuss such genes focusing on the recent findings.
transcriptional regulation. Genes belonging to this RBCC/TRIM
family are implicated in a variety of processes such as develop- Ring Fingers That Act As E3 Ligases
ment and cell growth and are involved in several human diseases.
PYRIN,50 MID1 (Midline 1)51 and MUL (for mulibrey nanism
proteins)52 are mutated in familial Mediterranean fever, X-linked Efp
Opitz/GBBB syndrome and mulibrey nanism, respectively, Estrogen-responsive finger protein (Efp) is a member of the
whereas PML, RFP (ret finger protein) and TIF1α acquire onco- RING-finger, B1 and B2-boxes, coiled coil and SPRY
genic activity when fused to RARα, RET or B-raf, respectively. (RBCC-SPRY) subfamily in the RING finger family. Efp was
It has been shown that the RBCC/TRIM proteins can isolated as an estrogen-responsive gene by genomic binding-site
oligomerize through their coiled coil domains. In cloning using a recombinant estrogen receptor (ER) protein.58
homodimerization of RFP proteins, the coiled coil region with the The estrogen-responsive element (ERE) to which ER can bind is
B-box but not the RING finger is required.53 In this case, while found at the 3'-untranslated region (UTR) in the Efp gene and
the B-box is not an interacting interface itself, the mutation of the gene’s expression is predominantly detected in female repro-
conserved cysteine residues within the B-box affects the ability of ductive organs including uterus, ovary and mammary gland59 and
RFP to multimerize, suggesting that its structural integrity is nec- in breast and ovarian cancers.60 Estrogen-induced expression is
essary for this interaction to occur.53 The coiled coil domain of found in the uterus, brain and mammary gland cells. Efp knock-
RFP is also necessary for interaction with Enhancer of polycomb out mice have an underdeveloped uterus and estrogen responses
protein (EPC) to repress gene transcription. 54 The of uterin cells from knockout mice are markedly attenuated, sug-
homodimerization and binding with EPC occurs with the proxi- gesting that Efp is necessary for estrogen-induced cell growth.61
mal coil in RFP protein. RFP also directly interacts and colocalizes Moreover, tumor growth of breast cancer MCF7 cells implanted
with PML in a subset of the PML NBs (nuclear bodies).55 This in female athymic mice has been demonstrated to be reduced by
interaction is mediated by the RFP B-box and the distal two coils. treatment with antisense Efp oligonucleotide. In contrast,
The association of RFP with the PML NBs is altered by mutations Efp-overexpressing MCF7 cells in ovariectomized athymic mice
that affect RFP/PML interaction and in APL patients-derived cells. generate tumors in the absence of estrogen.33 These results indi-
These results indicate that RFP have an important role in regulat- cate that Efp mediates estrogen-dependent growth in breast can-
ing cellular growth and differentiation. MID1 protein, which is cer cells. We identified 14-3-3σ which is responsible for reduced
mutated in patients with Opitz GBBB syndrome, and the highly cell growth, as a binding factor to Efp and found an accumula-
related gene MID2 also make both homo- and hetero-dimers me- tion of 14-3-3σ in Efp knockout mouse embryonic fibroblasts.
diated by the coiled coil motifs. The dimerization is a prerequisite Furthermore, it has been revealed that Efp is an ubiquitin ligase
for the association of MID and Alpha 4 (a regulatory subunit of (E3) that targets proteolysis of 14-3-3σ. Specifically, the RING
PP2-type phosphatases) and the complex formation with micro- which preferentially bound to ubiquitin-conjugating enzyme
tubules which seems important for normal midline development.56 UbcH8 has been shown to be essential for the ubiquitination of
In contrast, it has been shown that the entire RBCC/TRIM do- 14-3-3σ. Our findings provide an insight into the cell-cycle ma-
main is required for hetero-oligomerization or binding natural chinery and tumorigenesis of breast cancer by identifying 14-3-3σ
ligands. The RBCC region of KAP1/TIF1β associates with the as a target for proteolysis by Efp, leading to cell proliferation. The
KRAB (Krüppel-associated box) transcriptional repressor domain degradation of 14-3-3σ is subsequently followed by dissociation
of KOX-1.57 From extensive studies of the interaction, it has been of the protein from cyclin–Cdk complexes, leading to cell cycle
revealed that the interaction is specific for the KAP1 RBCC/TRIM progression and tumor growth (Fig. 5).
domain. Namely, when each RBCC/TRIM motifs of KAP1 was
swapped with other corresponding ones of MID1, KAP1 did not MID1, MID2
bind the KRAB domain any more. Therefore, each domain of the Opitz GBBB syndrome (OS; Opitz syndrome) is a genetically
RBCC may function as an independent functional unit and have and phenotypically complex disorder defined by characteristic fa-
important roles in the specific recognition of interacting partners cial anomalies, structural heart defects, as well as anal and genital
or oligomers formation. In other RBCC/TRIM proteins, only one anomalies.62,63 A positional cloning approach has revealed a can-
copy of the B-box motif is present, but inspection of the whole didate gene designated MID151 which is a member of the
family reveals a conserved residue spacing between the RING, RBCC-SPRY family. Most of the mutations identified so far in
110 Zinc Finger Proteins: From Atomic Contact to Cellular Function
25. Kakizuka A, Miller Jr WH, Umesono K et al. Chromosomal 48. Lu Z, Xu S, Joazeiro C et al. The PHD domain of MEKK1 acts as
translocation t(15;17) in human acute promyelocytic leukemia fuses an E3 ubiquitin ligase and mediates ubiquitination and degradation
RAR alpha with a novel putative transcription factor, PML. Cell of ERK1/2. Mol Cell 2002; 9(5):945-956.
1991; 66(4):663-674. 49. Dhalluin C, Carlson JE, Zeng L et al. Structure and ligand of a
26. de The H, Lavau C, Marchio A et al. The PML-RAR alpha fusion histone acetyltransferase bromodomain. Nature 1999;
mRNA generated by the t(15;17) translocation in acute promyelocytic 399(6735):491-496.
leukemia encodes a functionally altered RAR. Cell 1991; 50. Ancient missense mutations in a new member of the RoRet gene
66(4):675-684. family are likely to cause familial Mediterranean fever. The
27. Goddard AD, Borrow J, Freemont PS et al. Characterization of a International FMF Consortium. Cell 1997; 90(4):797-807.
zinc finger gene disrupted by the t(15;17) in acute promyelocytic 51. Quaderi NA, Schweiger S, Gaudenz K et al. Opitz G/BBB syndrome,
leukemia. Science 1991; 254(5036):1371-1374. a defect of midline development, is due to mutations in a new RING
28. Miki Y, Swensen J, Shattuck-Eidens D et al. A strong candidate for finger gene on Xp22. Nat Genet 1997; 17(3):285-291.
the breast and ovarian cancer susceptibility gene BRCA1. Science 52. Avela K, Lipsanen-Nyman M, Idanheimo N et al. Gene encoding a
1994; 266(5182):66-71. new RING-B-box-Coiled-coil protein is mutated in mulibrey nanism.
29. Le Douarin B, Zechel C, Garnier JM et al. The N-terminal part of Nat Genet 2000; 25(3):298-301.
TIF1, a putative mediator of the ligand-dependent activation function 53. Cao T, Borden KL, Freemont PS et al. Involvement of the rfp
(AF-2) of nuclear receptors, is fused to B-raf in the oncogenic protein tripartite motif in protein-protein interactions and subcellular
T18. EMBO J 1995; 14(9):2020-2033. distribution. J Cell Sci 1997; 110(Pt 14):1563-1571.
30. Hershko A, Ciechanover A. The ubiquitin system. Annu Rev 54. Shimono Y, Murakami H, Hasegawa Y et al. RET finger protein is
Biochem 1998; 67:425-479. a transcriptional repressor and interacts with enhancer of polycomb
31. Joazeiro CA, Wing SS, Huang H et al. The tyrosine kinase negative that has dual transcriptional functions. J Biol Chem 2000;
regulator c-Cbl as a RING-type, E2-dependent ubiquitin-protein 275(50):39411-39419.
ligase. Science 1999; 286(5438):309-312. 55. Cao T, Duprez E, Borden KL et al. Ret finger protein is a normal
32. Lorick KL, Jensen JP, Fang S et al. RING fingers mediate component of PML nuclear bodies and interacts directly with PML.
ubiquitin-conjugating enzyme (E2)-dependent ubiquitination. Proc J Cell Sci 1998; 111(Pt 10):1319-1329.
Natl Acad Sci USA 1999; 96(20):11364-11369. 56. Trockenbacher A, Suckow V, Foerster J et al. MID1, mutated in
33. Urano T, Saito T, Tsukui T et al. Efp targets 14-3-3 sigma for Opitz syndrome, encodes an ubiquitin ligase that targets phosphatase
proteolysis and promotes breast tumour growth. Nature 2002; 2A for degradation. Nat Genet 2001; 29(3):287-294.
417(6891):871-875. 57. Peng H, Begg GE, Schultz DC et al. Reconstitution of the
34. Zhang Y, Xiong Y. Control of p53 ubiquitination and nuclear export KRAB-KAP-1 repressor complex: A model system for defining the
by MDM2 and ARF. Cell Growth Differ 2001; 12(4):175-186. molecular anatomy of RING-B box-coiled-coil domain-mediated
35. Seol JH, Feldman RM, Zachariae W et al. Cdc53/cullin and the protein-protein interactions. J Mol Biol 2000; 295(5):1139-1162.
essential Hrt1 RING-H2 subunit of SCF define a ubiquitin ligase 58. Inoue S, Orimo A, Hosoi T et al. Genomic binding-site cloning
module that activates the E2 enzyme Cdc34. Genes Dev 1999; reveals an estrogen-responsive gene that encodes a RING finger
13(12):1614-1626. protein. Proc Natl Acad Sci USA 1993; 90(23):11117-11121.
36. Galisteo ML, Dikic I, Batzer AG et al. Tyrosine phosphorylation of 59. Orimo A, Inoue S, Ikeda K et al. Molecular cloning, structure, and
the c-cbl proto-oncogene protein product and association with expression of mouse estrogen-responsive finger protein Efp.
epidermal growth factor (EGF) receptor upon EGF stimulation. J Colocalization with estrogen receptor mRNA in target organs. J Biol
Biol Chem 1995; 270(35):20242-20245. Chem 1995; 270(41):24406-24413.
37. Thien CB, Walker F, Langdon WY. RING finger mutations that 60. Ikeda K, Orimo A, Higashi Y et al. Efp as a primary
abolish c-Cbl-directed polyubiquitination and downregulation of the estrogen-responsive gene in human breast cancer. FEBS Lett 2000;
EGF receptor are insufficient for cell transformation. Mol Cell 2001; 472(1):9-13.
7(2):355-365. 61. Orimo A, Inoue S, Minowa O et al. Underdeveloped uterus and
38. Haupt Y, Maya R, Kazaz A et al. Mdm2 promotes the rapid reduced estrogen responsiveness in mice with disruption of the
degradation of p53. Nature 1997; 387(6630):296-299. estrogen-responsive finger protein gene, which is a direct target of
39. Kubbutat MH, Jones SN, Vousden KH. Regulation of p53 stability estrogen receptor alpha. Proc Natl Acad Sci USA 1999;
by Mdm2. Nature 1997; 387(6630):299-303. 96(21):12027-12032.
40. Geyer RK, Yu ZK, Maki CG. The MDM2 RING-finger domain is 62. Opitz JM. G syndrome (hypertelorism with esophageal abnormality
required to promote p53 nuclear export. Nat Cell Biol 2000; and hypospadias, or hypospadias-dysphagia, or “Opitz-Frias” or
2(9):569-573. “Opitz-G” syndrome)—perspective in 1987 and bibliography. Am J
41. Chen Y, Chen CF, Riley DJ et al. Aberrant subcellular localization Med Genet 1987; 28(2):275-285.
of BRCA1 in breast cancer. Science 1995; 270(5237):789-791. 63. Robin NH, Opitz JM, Muenke M. Opitz G/BBB syndrome: Clinical
42. Lupas A. Coiled coils: New structures and new functions. Trends comparisons of families linked to Xp22 and 22q, and a review of
Biochem Sci 1996; 21(10):375-382. the literature. Am J Med Genet 1996; 62(3):305-317.
43. Reymond A, Meroni G, Fantozzi A et al. The tripartite motif family 64. Gaudenz K, Roessler E, Quaderi N et al. Opitz G/BBB syndrome
identifies cell compartments. EMBO J 2001; 20(9):2140-2151. in Xp22: Mutations in the MID1 gene cluster in the
44. Tissot C, Mechti N. Molecular cloning of a new interferon-induced carboxy-terminal domain. Am J Hum Genet 1998; 63(3):703-710.
factor that represses human immunodeficiency virus type 1 long 65. Jerome LA, Papaioannou VE. DiGeorge syndrome phenotype in mice
terminal repeat expression. J Biol Chem 1995; 270(25):14891-14898. mutant for the T-box gene, Tbx1. Nat Genet 2001; 27(3):286-291.
45. Der SD, Zhou A, Williams BR et al. Identification of genes 66. Cainarca S, Messali S, Ballabio A et al. Functional characterization
differentially regulated by interferon alpha, beta, or gamma using of the Opitz syndrome gene product (midin): Evidence for
oligonucleotide arrays. Proc Natl Acad Sci USA 1998; homodimerization and association with microtubules throughout the
95(26):15623-15628. cell cycle. Hum Mol Genet 1999; 8(8):1387-1396.
46. Orimo A, Tominaga N, Yoshimura K et al. Molecular cloning of 67. Short KM, Hopwood B, Yi Z et al. MID1 and MID2 homo- and
ring finger protein 21 (RNF21)/interferon-responsive finger protein heterodimerise to tether the rapamycin-sensitive PP2A regulatory
(ifp1), which possesses two RING-B box-coiled coil domains in subunit, alpha 4, to microtubules: Implications for the clinical
tandem. Genomics 2000; 69(1):143-149. variability of X-linked Opitz GBBB syndrome and other
47. Slack FJ, Ruvkun G. A novel repeat domain that is often associated developmental disorders. BMC Cell Biol 2002; 3(1):1.
with RING finger and B-box motifs. Trends Biochem Sci 1998; 68. Borden KL. RING fingers and B-boxes: Zinc-binding protein-protein
23(12):474-475. interaction domains. Biochem Cell Biol 1998; 76(2-3):351-358.
RBCC Proteins As Ubiquitin Ligase in the Control of Protein Degradation and Gene Regulation 113
69. Kastner P, Perez A, Lutz Y et al. Structure, localization and 83. Kamitani T, Kito K, Nguyen HP et al. Identification of three major
transcriptional properties of two classes of retinoic acid receptor alpha sentrinization sites in PML. J Biol Chem 1998;
fusion proteins in acute promyelocytic leukemia (APL): Structural 273(41):26675-26682.
similarities with a new family of oncoproteins. EMBO J 1992; 84. Borden KL, Lally JM, Martin SR et al. In vivo and in vitro
11(2):629-642. characterization of the B1 and B2 zinc-binding domains from the
70. Grignani F, Ferrucci PF, Testa U et al. The acute promyelocytic acute promyelocytic leukemia protooncoprotein PML. Proc Natl Acad
leukemia-specific PML-RAR alpha fusion protein inhibits Sci USA 1996; 93(4):1601-1606.
differentiation and promotes survival of myeloid precursor cells. Cell 85. Lallemand-Breitenbach V, Zhu J, Puvion F et al. Role of
1993; 74(3):423-431. promyelocytic leukemia (PML) sumolation in nuclear body
71. Grignani F, Testa U, Rogaia D et al. Effects on differentiation by formation, 11S proteasome recruitment, and As2O3-induced PML
the promyelocytic leukemia PML/RARalpha protein depend on the or PML/retinoic acid receptor alpha degradation. J Exp Med 2001;
fusion of the PML protein dimerization and RARalpha DNA binding 193(12):1361-1371.
domains. EMBO J 1996; 15(18):4949-4958. 86. Le XF, Yang P, Chang KS. Analysis of the growth and transformation
72. Weis K, Rambaud S, Lavau C et al. Retinoic acid regulates aberrant suppressor domains of promyelocytic leukemia gene, PML. J Biol
nuclear localization of PML-RAR alpha in acute promyelocytic Chem 1996; 271(1):130-135.
leukemia cells. Cell 1994; 76(2):345-356. 87. Chen XP, Losman JA, Rothman P. SOCS proteins, regulators of
73. Koken MH, Puvion-Dutilleul F, Guillemin MC et al. The t(15;17) intracellular signaling. Immunity 2000; 13(3):287-290.
translocation alters a nuclear body in a retinoic acid-reversible fashion. 88. Haque SJ, Harbor PC, Williams BR. Identification of critical residues
EMBO J 1994; 13(5):1073-1083. required for suppressor of cytokine signaling-specific regulation of
74. Dyck JA, Maul GG, Miller Jr WH et al. A novel macromolecular interleukin-4 signaling. J Biol Chem 2000; 275(34):26500-26506.
structure is a target of the promyelocyte-retinoic acid receptor 89. Terstegen L, Maassen BG, Radtke S et al. Differential inhibition of
oncoprotein. Cell 1994; 76(2):333-343. IL-6-type cytokine-induced STAT activation by PMA. FEBS Lett
75. Szostecki C, Guldner HH, Netter HJ et al. Isolation and 2000; 478(1-2):100-104.
characterization of cDNA encoding a human nuclear antigen 90. Toniato E, Chen XP, Losman J et al. TRIM8/GERP RING finger
predominantly recognized by autoantibodies from patients with protein interacts with SOCS-1. J Biol Chem 2002;
primary biliary cirrhosis. J Immunol 1990; 145(12):4338-4347. 277(40):37315-37322.
76. Koken MH, Reid A, Quignon F et al. Leukemia-associated retinoic 91. Niikura T, Hashimoto Y, Tajima H et al. A tripartite motif protein
acid receptor alpha fusion partners, PML and PLZF, heterodimerize TRIM11 binds and destabilizes Humanin, a neuroprotective peptide
and colocalize to nuclear bodies. Proc Natl Acad Sci USA 1997; against Alzheimer’s disease-relevant insults. Eur J Neurosci 2003;
94(19):10255-10260. 17(6):1150-1158.
77. Chen Z, Brand NJ, Chen A et al. Fusion between a novel 92. Guo B, Zhai D, Cabezas E et al. Humanin peptide suppresses
Kruppel-like zinc finger gene and the retinoic acid receptor-alpha apoptosis by interfering with Bax activation. Nature 2003;
locus due to a variant t(11;17) translocation associated with acute 423(6938):456-461.
promyelocytic leukaemia. EMBO J 1993; 12(3):1161-1167. 93. Sibilia J. Ro(SS-A) and anti-Ro(SS-A): An update. Rev Rhum Engl
78. Ishov AM, Maul GG. The periphery of nuclear domain 10 (ND10) Ed 1998; 65(1):45-57.
as site of DNA virus deposition. J Cell Biol 1996; 134(4):815-826. 94. Itoh Y, Reichlin M. Autoantibodies to the Ro/SSA antigen are
79. Kamitani T, Nguyen HP, Kito K et al. Covalent modification of conformation dependent. I: Anti-60 kD antibodies are mainly
PML by the sentrin family of ubiquitin-like proteins. J Biol Chem directed to the native protein; anti-52 kD antibodies are mainly
1998; 273(6):3117-3120. directed to the denatured protein. Autoimmunity 1992; 14(1):57-65.
80. Fagioli M, Alcalay M, Tomassoni L et al. Cooperation between the 95. Billaut-Mulot O, Cocude C, Kolesnitchenko V et al. SS-56, a novel
RING + B1-B2 and coiled-coil domains of PML is necessary for its cellular target of autoantibody responses in Sjogren syndrome and
effects on cell survival. Oncogene 1998; 16(22):2905-2913. systemic lupus erythematosus. J Clin Invest 2001; 108(6):861-869.
81. Mu ZM, Chin KV, Liu JH et al. PML, a growth suppressor disrupted 96. Fukuda-Kamitani T, Kamitani T. Ubiquitination of Ro52
in acute promyelocytic leukemia. Mol Cell Biol 1994; autoantigen. Biochem Biophys Res Commun 2002; 295(4):774-778.
14(10):6858-6867.
82. Regad T, Saib A, Lallemand-Breitenbach V et al. PML mediates the
interferon-induced antiviral state against a complex retrovirus via its
association with the viral transactivator. EMBO J 2001;
20(13):3495-3505.
CHAPTER 17
Abstract
T
he TAZ (transcriptional adapter zinc binding) domains region forms a zinc-binding domain similar to the PHD
in the transcriptional coactivators CBP/p300 are zinc-finger motif, and the CH3 region forms two
zinc-binding motifs that are primarily involved in independently-folded zinc-binding domains, the ZZ domain, and
protein-protein recognition. The activation domains of more than the TAZ2 domain.
30 transcription factors have been reported to bind to the TAZ The two TAZ domains are important sites of protein-protein
domains, and each TAZ domain generally binds a different sub- interaction in CBP/p300, and to date, about 30 different tran-
set of transcription factors. Structurally, the TAZ domains con- scription factors have been reported to bind to TAZ1 and TAZ2
tain three zinc binding sites and four alpha helices that are packed (Fig. 1). The zinc-binding motifs of TAZ1 and TAZ2 share a
against each other to form a hydrophobic core. Each zinc atom high degree of homology, each consisting of one histidine and
is bound to one histidine and three cysteine residues that are three cysteine residues (Fig. 2). Sequence homology searches in-
located at the end of one helix, in the interconnecting loop, and dicate that the TAZ domains are currently found only in CBP/
at the beginning of a second helix, forming a characteristic p300-related proteins, and each protein contains two TAZ do-
HCCC-type zinc binding motif. In the absence of zinc, the pro- mains. The TAZ2 domain is always associated with the much
teins are unfolded. The role of the zinc atoms is to organize and larger acetyltransferase module of CBP/p300, an arrangement
stabilize the four amphipathic helices into a global fold, which evolutionarily conserved even in plant proteins.15,16 Each TAZ
then serves as the site of protein-protein interaction. The TAZ1 domain contains three zinc-binding sites with the consensus se-
domains of CBP and p300 binds other proteins mostly by hy- quence His1-X3-Cys2-X4-12-Cys3-X2-4-Cys4, where X is any
drophobic interactions in the grooves and canyons created by
the helical interfaces. The first three helices of TAZ1 and TAZ2
are structurally homologous; the fourth helix, however, is in a
different relative orientation. This major structural difference
between the two domains appears to play an important role in
determining the protein-binding specificity of the TAZ domain.
Introduction
The protein CBP (cyclic-AMP response element binding pro-
tein (CREB) binding protein)1 and its paralog p3002 are large
multifunctional transcriptional coactivator proteins that bind to
the activation domains of numerous transcription factors.3-8 Struc-
tural and biological studies of CBP/p300 paint a picture of a large
protein (2441 residues for human CBP) containing several globu- Figure 1. CBP/p300 contains four zinc-binding domains that map to
lar domains that are independently capable of interacting with the three cyteine/histidine-rich regions, CH1, CH2, and CH3. The
different sets of transcription factors. Three dimensional struc- first cysteine/histidine-rich region, CH1, is the TAZ1 domain, and the
tures have been reported for the KIX domain9-11 and the nuclear CH3 region is made of two zinc-binding domains, ZZ and TAZ2. CH2
coactivator binding domain (CABD).12 In addition, CBP and forms another zinc-binding domain that is homologous to the PHD
p300 contain four zinc-binding domains that map to three cys- zinc finger motif. The relative location of the CREB-binding domain
teine/histidine-rich regions, initially designated in the literature (KIX), Bromodomain (BR), glutamine-rich region, and the C-terminal
as CH1, CH2, and CH3.13 An alternative nomenclature was pro- activator binding domain (CABD) of CBP/p300 are indicated. Shown
posed by Ponting et al.14 The CH1 region corresponds to the below are the protein factors that have been reported to interact with
TAZ1 (for transcriptional adaptor zinc-binding) domain, the CH2 TAZ1 and the CH3 region.
Zinc Finger Proteins: From Atomic Contact to Cellular Function, edited by Shiro Iuchi and Natalie Kuldell.
©2005 Landes Bioscience/Eurekah.com and Kluwer Academic/Plenum Publishers.
Structure and Function of the CBP/p300 TAZ Domains 115
Figure 2. Sequences of TAZ1 and TAZ2 aligned at the three zinc binding clusters (Zn1, Zn2, and Zn3). Secondary structures are indicated as wavy
lines for helices and solid lines for loops. The abbreviations and accession numbers are: Mm. (Mus musculus/mouse AAL87531); Hs. (Homo sapiens/
human Q92793 (CBP) / Q09472 (P300)); Xl. (Xenopus laevis/frog AAH44677); Ac. (Aplysia californica/sea hare AAL54859); Ce. (Caenorhabditis
elegans/worm P34545); Ag (Anopheles gambiae/mosquito EAA06516); Dm. (Drosophila melanogaster/fly AAB53050); Gg. (Gallus gallus/chicken
TC65031); At. (Arabidopsis thaliana/plant AAF79331).
amino acid and the subscript denotes the number of residues be- The Zinc-Binding Site
tween the zinc ligands. The basic building block of a TAZ domain is a single
Although the zinc-binding motifs of TAZ domains are ho- zinc-binding site, as illustrated in Figure 3. Each zinc binding mo-
mologous, TAZ1 and TAZ2 domains have been reported to bind tif consists of two helices joined by a connecting loop, with one
a different subset of transcription factors, as indicated in Figure 1 histidine and three cysteine ligands that are arranged sequentially
and Tables 1 and 2. Sequence analyses of the activation domains to form an HCCC-type zinc-binding motif.21 The first two zinc
of their binding partners also do not reveal an obvious sequence ligands (His1 and Cys2) are located at the end of a helix, the third
homology. CBP and p300 do not bind nucleic acids, and no cata- ligand (Cys3) is in the loop following the helix, and the fourth
lytic activity has been reported for regions containing the TAZ ligand (Cys4) is at the beginning of the next helix. The zinc atom is
domains. Thus, their primary role appears to be in protein-protein coordinated to the histidine Nε2 atom; the other possible
recognition, a common function for a growing list of zinc-binding metal-binding site, Nδ1, was excluded based on chemical shift
domains. Structural results from NMR spectroscopy reveal that analysis. Packing interactions between the two helices are stabilized
TAZ117-20 and TAZ221 form a novel zinc-binding motif and show by hydrophobic contacts mediated by nonpolar residues.
how they participate in protein-protein interaction. Comparison
of the structures of TAZ1 and TAZ2 (see later discussion) illus-
trates how a structural change in the orientation of one helix al-
TAZ1 and TAZ2 Structures
lows them to discriminate between two different subsets of bind- The basic zinc-binding cluster in Figure 3 is repeated three
ing partners. times to form a complete TAZ domain, as shown in Figure 4.
The NMR structure of the TAZ2 domain from CBP21 shown in
Figure 4 is currently the only available structure of a free TAZ
domain. The structure of the free CBP TAZ1 has not yet been
reported; the structure shown in Figure 4 was taken from the
TAZ1/HIF-1α complex.17 The free TAZ1 domain from p300
was recently reported to be only partially folded in the presence
of Zn2+and to behave like a molten globule. Binding of the free
p300 TAZ1 to HIF-1α was required for it to form a stable three
dimensional structure.22 This conclusion was based on the poor
dispersion of the NMR proton-nitrogen correlation spectra (2D
HSQC) of p300 TAZ1 in the presence of Zn2+. We have found,
however, that the free TAZ1 domain from CBP has a
well-dispersed 2D HSQC spectra in the presence of Zn2+ ions,
indicating that the CBP domain is folded even in the absence of
binding partners. Preliminary NMR structure calculations for the
free CBP TAZ1 domain indicate a folded structure that is very
similar to that found in the TAZ1/HIF-1α complex (R. N. De
Guzman, S.A. Dames, M. A. Martinez-Yamout, J. Wojciak, H.J.
Dyson, P.E. Wright, unpublished data).
In TAZ1 and TAZ2, the three zinc-binding motifs, Zn1-Zn3,
are distributed approximately at the apices of an irregular triangle.
Figure 3. A zinc-binding motif of the TAZ domains consists of four This arrangement maximizes the hydrophobic interactions be-
zinc-binding ligands His1, Cys2, Cys3, and Cys4, located at the tween the four helices (α1-α4), forming a hydrophobic core in
C-terminal end of a helix, the interconnecting loop, and the N-terminal the middle of the domain, and directing the connecting loops
end of the next helix. outwards. The four helices are generally amphipathic in character,
116 Zinc Finger Proteins: From Atomic Contact to Cellular Function
Table 1. List of proteins that have been reported to bind the TAZ1 domain
with the hydrophobic residues facing the core of the domain, and dithiothreitol at pH 7.5-8.5, and folded with addition of three
the nonpolar residues solvent exposed. The connecting loops are equivalents of Zn2+ ions.17,20 The NMR 15N-HSQC spectra for
mostly polar. The first three helices, α1-α3, are structurally ho- refolded TAZ1 and for TAZ1 purified under native conditions
mologous in the TAZ domains. Although the TAZ1 α1 helix is are identical, indicating that the refolding procedure does not
longer than the corresponding helix in TAZ2 α1, the relative po- alter the biologically relevant structure.17 In the absence of Zn2+
sition of the zinc atom in both domains is very similar. or with an excess of EDTA, TAZ1 and TAZ2 are unstructured in
The major structural difference between TAZ1 and TAZ2 is solution, and circular dichroism (CD) spectra indicate that there
in the orientation of the fourth helix, α4 (red in Fig. 4). In TAZ1, is little or no residual helical character.
both α1 and α4 are on the same face of α3, while in TAZ2, α3 is Bacterially expressed TAZ2 is less soluble than TAZ1, form-
sandwiched between α1 and α4. The difference in the orientation ing inclusion bodies 21 that require solubilization in 6 M
of the fourth helix also results in a major shift in the position of guanidinium hydrochloride or 8 M urea. Coexpression with its
the zinc atom in Zn3, relative to Zn1 and Zn2.17 This structural binding partners E1A or p53, also resulted in the formation of
difference is also reflected in the number of inter-helical contacts: inclusion bodies, but some soluble protein is formed. As for TAZ1,
there are more inter-helical NOEs between α1 and α4 in TAZ1 three equivalents of Zn2+ were needed to fold TAZ2 under re-
than in TAZ2. In TAZ2, the arrangement of the four helices re- ducing conditions.21 A large excess of Zn2+ resulted in a decrease
sults in a more globular structure, approximating a pyramidal-like of the helical content in the CD spectra of TAZ2. This probably
overall shape. In TAZ1, the overall shape is more of a flatter tri- indicates nonspecific metal binding since TAZ2 contains addi-
angle, or a pie-shaped structure, due to the positioning of α4. tional metal binding residues (four cysteines and two histidines).
This structural difference will be more evident below in the dis-
cussion of the protein-protein complexes formed by TAZ1. Protein-Protein Interactions of the TAZ1
Domains
Preparation of TAZ Domains Recent NMR structures of the complexes formed by CBP and
Some bacterially expressed TAZ1 aggregates in the insoluble p300 TAZ1 reveal the structural determinants of protein recog-
fraction, but there is a significant amount (up to ~ 60%) that nition by TAZ1.17-20 Figure 5 shows the structures of complexes
remains in the soluble fraction. This soluble fraction increases to formed by the CBP TAZ1 with the C-terminal activation do-
more than 90% when TAZ1 is coexpressed with binding part- mains (CAD) of HIF-1α17 and CITED2.20 The heterodimeric
ners such as HIF-1α or CITED2.17,20 TAZ1 can be expressed transcription factor HIF-1 (Hypoxia Inducible Factor) plays an
and purified under denaturing conditions, reduced in 10-20 mM important role in the cellular response to low oxygen conditions,
Structure and Function of the CBP/p300 TAZ Domains 117
Table 2. List of proteins that have been reported to bind the TAZ2 domain
and thus, is a major protein involved in the pathology of cancer, region is capable of tight binding to TAZ1 (Kd = 46 nM)17,20 (and
heart disease and stroke.23-25 Under conditions of oxygen depri- indeed this region was not included in the structure determination
vation, the HIF-1α subunit recruits CBP/p300 to direct expres- of the p300/HIF-1α complex),18 these additional residues increase
sion of genes involved in vascularization and glycolysis to ensure the affinity by about five-fold (Kd = 10 nM).17,20 The poorly struc-
cellular survival.26 HIF-1 also controls the expression of another tured N-terminal αA helical region is followed by an extended re-
nuclear protein, CITED2 (CBP/p300 Interacting Transactivator gion containing the sequence LPQL, which binds tightly at the
with glutamic acid (E) and aspartic acid (D) rich Tail) at the tran- junction of the TAZ1 α1/α2/α3 interface. Following the LPQL
scriptional level. CITED2 competes with HIF-1α for binding to loop, is a short helix, αB, which binds at the TAZ1 α1/α3 interface.
the TAZ1 domain, effectively blocking HIF-1 transcriptional ac- This helix contains the so-called hypoxic switch, an asparagine resi-
tivity.27 This suggests regulation during hypoxia by a feedback due in position 803 that when hydroxylated, impairs the interac-
loop mediated by protein-protein interactions of HIF-1α and tion of Hif-1α with CBP or p300.28 Amino acid analysis and NMR
CITED2with TAZ1. The structures of TAZ1/HIF-1α17,18 and has defined the site of hydroxylation at the pro-S position of the
TAZ1/CITED219,20 complexes provide a detailed picture of the Cβ atom of Asn803.29 Finally, the HIF-1α CAD forms the
protein-protein interactions mediated by a zinc-binding domain C-terminal αC, which binds tightly at the TAZ1 α1/α4 interface.
during hypoxia. The TAZ1/CITED2 Complex.
The NMR structures of the p30019 and CBP20 TAZ1/CITED2
Structure of the TAZ1/HIF-1α Complex complexes provide another detailed picture of how TAZ1 serves as
The structure of the CBP TAZ1 domain in complex with the a scaffold for protein-protein interaction. The structural features,
HIF-1α CAD is shown in Figure 5A. The HIF-1α CAD is un- four helices (α1-α4), three zinc binding clusters, and the overall
structured in isolation, but, upon complex formation, folds and global fold of TAZ1 are preserved in the two complexes. There are
wraps around TAZ1, forming three helices, αA, αB, and αC, that minor differences in the TAZ1 structure between the Hif-1α and
bind in the grooves formed by the interfaces of TAZ1 helices.17,18 CITED2 complexes: a slight bending of α1 in the CITED2 com-
The αA helix is poorly defined and flexible, and interacts mainly plex,19,20 and a slight increase in the length of the α4 in the CBP
with TAZ1 α1. Although the HIF-1α CAD lacking this N-terminal TAZ1/CITED2 complex.20 The main difference between the
118 Zinc Finger Proteins: From Atomic Contact to Cellular Function
45. Huang SM, Qiu Y, Stein RW et al. P300 functions as a 61. Swanson DJ, Adachi M, Lewis EJ. The homeodomain protein Arix
transcriptional coactivator for the TAL1/SCL oncoprotein. Oncogene promotes protein kinase A-dependent activation of the dopamine
1999; 18:4958-4967. beta -hydroxylase promoter through multiple elements and interaction
46. Zeng X, Li X, Miller A et al. The N-terminal domain of p73 interacts with the coactivator cAMP-response element-binding protein-binding
with the CH1 domain of p300/CREB binding protein and mediates protein. J Biol Chem 2000; 275:2911-2923.
transcriptional activation and apoptosis. Mol Cell Biol 2000; 62. Cho S, Tian Y, Benjamin TL. Binding of p300/CBP coactivators by
20:1299-1310. polyoma large T antigen. J Biol Chem 2001; 276:33533-33539.
47. Abraham SE, Lobo S, Yaciuk P et al. p300, and p300-associated 63. Sisk TJ, Gourley T, Roys S et al. MHC class II transactivator inhibits
proteins, are components of TATA-binding protein (TBP) complexes. IL-4 gene transcription by competing with NF-AT to bind the
Oncogene 1993; 8:1639-1647. coactivator CREB binding protein (CBP)/p300. J Immunol 2000;
48. Dallas PB, Yaciuk P, Moran E. Characterization of monoclonal 165:2511-2517.
antibodies raised against p300: Both p300 and CBP are present in 64. Suzuki T, Yoshida M. Tax protein of HTLV-1 inhibits CBP/
intracellular TBP complexes. J Virol 1997; 71:1726-1731. p300-mediated transcription by interfering with recruitment of CBP/
49. Kwok RPS, Lundblad JR, Chrivia JC et al. Nuclear protein CBP is p300 onto DNA element of E-box or p53 binding site. Oncogene
a coactivator for the transcription factor CREB. Nature 1994; 1999; 18:4137-4143.
370:223-226. 65. Hottiger MO, Nabel GJ. Interaction of human immunodeficiency
50. Yang C, Shapiro LH, Rivera M et al. A role for CREB binding virus type 1 Tat with the transcriptional coactivators p300 and CREB
protein and p300 transcriptional coactivators in Ets-1 transactivation binding protein. J Virol 1998; 72:8252-8256.
functions. Mol Cell Biol 1998; 18:2218-2229. 66. Ito M, Yu RN, Jameson JL. Steroidogenic factor-1 contains a
51. Jayaraman G, Srinivas R, Duggan C et al. p300/cAMP-responsive carboxy-terminal transcriptional activation domain that interacts with
element-binding protein interactions with Ets-1 and Ets-2 in the steroid receptor coactivator-1. Mol Endocrinol 1998; 12:290-301.
transcriptional activation of the human stromelysin promoter. J Biol 67. Trouche D, Kouzarides T. E2F1 and E1A(12S) have a homologous
Chem 1999; 274:17342-17352. activation domain regulated by RB and CBP. Proc Natl Acad Sci
52. Gerritsen ME, Williams AJ, Neish AS et al. CREB-binding protein USA 1996; 93:1439-1442.
p300 are transcriptional coactivators of p65. Proc Natl Acad Sci USA 68. Lee JS, See RH, Deng T et al. Adenovirus E1A downregulates cJun-
1997; 94:2927-2932. and JunB-mediated transcription by targeting their coactivator p300.
53. Zhong H, Voll RE, Ghosh S. Phosphorylation of NF-κB p65 by Mol Cell Biol 1996; 16:4312-4326.
PKA stimulates transcriptional activity by promoting a novel bivalent 69. Rossow KL, Janknecht R. Synergism between p68 RNA helicase and
interaction with the coactivator CBP/p300. Mol Cell Biol 1998; the transcriptional coactivators CBP and p300. Oncogene 2003;
1:661-671. 22:151-156.
54. Kishimoto M, Okimura Y, Yagita K et al. Novel function of the 70. Nakajima T, Uchida C, Anderson S et al. RNA helicase A mediates
transactivation domain of a pituitary-specific transcription factor, association of CBP with RNA polymerase II. Cell 1997;
Pit-1. J Biol Chem 2002; 277:45141-45148. 90:1107-1112.
55. Bhattacharya S, Eckner R, Grossman S et al. Cooperation of Stat2 71. Mink S, Haenig B, Klempnauer KH. Interaction and functional
and p300/CBP in signalling induced by interferon-α. Nature 1996; collaboration of p300 and C/EBPbeta. Mol Cell Biol 1997;
383:344-347. 17:6609-6617.
56. Dell H, Hadzopoulou-Cladaras M. CREB-binding protein is a 72. Qiu Y, Sharma A, Stein R. p300 mediates transcriptional stimulation
transcriptional coactivator for hepatocyte nuclear factor-4 and by the basic helix-loop-helix activators of the insulin gene. Mol Cell
enhances apolipoprotein gene expression. J Biol Chem 1999; Biol 1998; 18:2957-2964.
274:9013-9021. 73. Sato S, Roberts K, Gambino G et al. CBP/p300 as a cofactor for
57. Doucas V, Tini M, Egan DA et al. Modulation of CREB binding the Microphthalmia transcription factor. Oncogene 1997;
protein function by the promyelocytic (PML) oncoprotein suggests 14:3083-3092.
a role for nuclear bodies in hormone signaling. Proc Natl Acad Sci 74. Price ER, Ding HF, Badalian T et al. Lineage-specific signaling in
USA 1999; 96:2627-2632. melanocytes - c-Kit stimulation recruits p300/CBP to
58. Zimmermann H, Degenkolbe R, Bernard HU et al. The human microphthalmia. J Biol Chem 1998; 273:17983-17986.
papillomavirus type 16 E6 oncoprotein can down-regulate p53 75. Felzien LK, Farrell S, Betts JC et al. Specificity of cyclin E-Cdk2,
activity by targeting the transcriptional coactivator CBP/p300. J Virol TFIIB, and E1A interactions with a common domain of the p300
1999; 73:6209-6219. coactivator. Mol Cell Biol 1999; 19:4241-4246.
59. Eckner R, Ludlow JW, Lill NL et al. Association of p300 and CBP 76. Nakajima T, Fukamizu A, Takahashi J et al. The signal-dependent
with simian virus 40 large T antigen. Mol Cell Biol 1996; coactivator CBP is a nuclear target for pp90RSK. Cell 1996;
16:3454-3464. 86:465-474.
60. Waltzer L, Bienz M. A function of CBP as a transcriptional
coactivator during Dpp signalling. EMBO J 1999; 18:1630-1641.
CHAPTER 18
Abstract
P
rotein kinase CK2 plays an essential role in the regulation tional overlap of both CK2 isoforms. RNA-mediated interference
of many cellular functions. The enzyme is an in Caenorhabditis elegans has shown that inhibition of the CK2β
heterotetrameric complex formed by the association of two gene leads to embryonic lethality.8 Similarly, disruption of the CK2β
catalytic α/α’ subunits with two regulatory β subunits. subunit in mice leads to a cell-autonomous defect and early em-
High-resolution structure of the CK2β subunit revealed the pres- bryonic lethality.9 Thus, the production of both the α and β sub-
ence of a zinc binding motif made by three-stranded antiparallel units of CK2 appears to be required for cell survival.
β sheets and two “knuckles” (Cys-X4-Cys and Cys-X2-Cys) con- Mounting evidence suggest that CK2 is a component of regu-
tained in the invariant motif CPX3C-X22-CPXC. This zinc bind- latory protein kinase networks that are involved in various as-
ing motif belongs to the sub-family of zinc ribbon domains. pects of transformation and cancer. Targetted overexpression of
CK2β exist as a dimer in which the zinc ribbon motif makes CK2α in T cells and mammary glands of transgenic mice leads
many hydrophobic interactions with the zinc ribbon motif of to lymphomagenesis and mammary tumorigenesis, respec-
the other monomer forming the protein-protein interface. Im- tively.10,11 A number of studies have also shown that in all human
portantly, functional and biochemical studies have indicated that cancers that have been examined, CK2 activity is consistently
the integrity of the zinc binding motif which is pivotal in the enhanced suggesting that the kinase plays a role in cell prolifera-
formation of the CK2β homodimer, is also instrumental for the tion but also participates in the transduction of cell survival sig-
regulatory functions of this important protein. nals.12 Finally, the involvement of CK2 in viral infection and on-
cogenesis has been demonstrated.3
Protein phosphorylation is a major mechanism for the regula-
tion of fundamental cellular processes.1 Among the hundred of Structural and Regulatory Features
protein kinases encoded by the genome of eukaryotic cells, the Ser/ Several recent reports have described specific structural aspects
Thr protein kinase CK2 (formely called casein kinase II) represents of the CK2 subunits and their functional properties providing
an essential component of this family of regulatory enzymes.2-5 long-awaited insights into a potential modus operandi of this piv-
The broad spectrum of protein substrates which are phosphory- otal protein kinase.
lated by CK2 underscores the functional importance of this pro- In many organisms as well as in humans several isoforms of
tein kinase in the regulation of many cellular functions.5 the catalytic subunit of CK2 have been identified.13-16 Although
Protein kinase CK2 exibits a high degree of conservation of closely related, they are in fact the products of different genes.
amino acid sequence between evolutionarily distant organisms such The crystal structure of the catalytic α subunit from Zea mays has
as yeast, Drosophila and human.2 When purified from many shown the common bilobal architecture described for protein
sources, CK2 appears as a tetrameric enzyme (with a molecular kinases. More importantly, this study provided clues for the ap-
mass of 130 kDa), composed of two types of structurally analo- parent constitutive activity of the enzyme and for the fact that
gous catalytic α (42 to 44 kDa) and α’(38 kDa) subunits and two this kinase can use GTP as well as ATP as phosphate donors.17
regulatory β (27 kDa) subunits which associate to form stable α2β2, Distinct isoforms of CK2 regulatory subunits exist in S.
αα’β2, or α’2β2 structures. The lethal phenotype resulting from cerevisiae, A. thaliana and in D. melanogaster. Surprisingly, in
disruption of the α subunit in yeast can be rescued by expression of mammals, this subunit is encoded by a single gene. On the basis
Drosophila α showing that the catalytic subunit is functionnaly of sequence comparisons, CK2β does not share similarity with
interchangeable between distantly related species.6 In male mice, any known protein. However, CK2β exhibits remarkable conser-
CK2α’ is preferencially expressed in late stages of spermatogenesis vation between species suggesting that this subunit plays impor-
and disruption of its gene results in infertility owing to defective tant roles in cellular functions. This contention is illustrated by
spermatogenesis.7 However, the apparently normal embryonic de- the demonstration that the expression of CK2β is essential for
velopment of CK2α’ knockout mice argues again for a strong func- cell proliferation and survival in D. melanogaster18 and in mice.9
Zinc Finger Proteins: From Atomic Contact to Cellular Function, edited by Shiro Iuchi and Natalie Kuldell.
©2005 Landes Bioscience/Eurekah.com and Kluwer Academic/Plenum Publishers.
122 Zinc Finger Proteins: From Atomic Contact to Cellular Function
Figure 3. Structure-based sequence alignement of zinc ribbon domains in representative CK2β proteins and in members of the topoisomerase
family. 1) Protein name: CK2β= CK2β regulatory subunit; RPB9= DNA-directed RNA polymerase II subunit 9; TFIIS= transcriptional elon-
gation factor SII; TFIIB= transcription initiation factor IIB; TOP3= topoisomerase III. 2) Organism name: At= Arabidopsis thaliana; Ce=
Caenorhabditis elegans; Dm= Drosophila melanogaster; Hs= Homo sapiens; Sc=, Saccharomyces cerevisiae; Sp= Schizosaccharomyces pombe. 3) Sequence
number of the first residue shown in the alignment. 4) Sequence number of the last residue shown in the alignement. Cys residues in Zn-binding
sites are highlighted in red and are shown in white. Uncharged residues in hydrophobic sites of β-strands are highlighted in yellow. Secondary
structure consensus for β-strand residues is shown below the alignment by black arrows.
unlike the classical zinc finger, no α helix is observed. The Zn2+ bon motif of the other monomer forming the protein-protein in-
binding site of CK2β is well ordered and contains two non-canonical terface. Most of the interactions between two CK2β monomers are
zinc finger loops termed “knuckles” (Cys-X4-Cys and Cys-X2-Cys) contacts between residues that are localized at the edge of the
contained in the invariant motif CPX3C-X22-CPXC. The two knuck- Cys109-X4-Cys114 element and on the β3 strand. In this core
les are placed at almost 90˚ to each other, resulting in tetrahedral region, several conserved hydrophobic residues make van der Waals
geometry of zinc ion. Based on pronounced structural similarity interactions that form several important contacts between the two
and residual sequence similarity, the zinc binding motif present in monomers (Fig. 5). The most significant contacts are Pro 110, Tyr
CK2β belongs to the sub-family of zinc ribbon domains. This motif
resembles the Zn-ribbon structure of transcription factors TFIIS,25
TFIIB, 26 the RNA polymerase II subunit 9 RPB9,27 and
Topoisomerase I and III28 (Fig. 3). Interestingly, the three β strands
and the P loop that form the core of the Zn2+ binding motif of
CK2β show a striking similar topology to the structure depicted
for TFIIS (Fig. 4). The knuckles of the Zn2+ binding site of CK2β
are similar to those described for TFIIS and overall, the zinc ribbon
motif of CK2β aligns structurally with an RMSD of 1.13 Å (96
atoms) with TFIIS. The β sheet surface of the Zn ribbon motif in
CK2β is remarkable for its hydrophobicity, with 13 hydrophobic
side chains defining the central surface. Unlike TFIIS which is a
highly soluble nucleic-acid binding domain, the zinc ribbon motif
of CK2β makes many hydrophobic interactions with the zinc rib-
Figure 7. Overall shape of the human CK2 holoenzyme showing that the CK2β homodimer is the building block of this hetero-complex, bridging
the two catalytic subunits. The view is perpendicular to the C2 axis. The two CK2β chains are drawn in blue and red, the two CK2α subunits in yellow
and grey. The non-hydrolysable analog AMPPNP bound to the ATP binding site of one catalytic subunit is illustrated in red. (Adapted from ref. 36).
A Zinc Ribbon Motif Is Essential for the Formation of Functional Tetrameric Protein Kinase CK2 125
Figure 9. Stimulation of CK2α subunit catalytic activity by wt and mutated CK2β subunits. Known CK2 protein substrates (Topoisomerase II,
Engrailed, PHAX) were phosphorylated by CK2α in the presence of increasing amount of wt CK2β ( ), CK2βP110D (), or CK2βV143D ().
Phosphorylated proteins were separated by SDS-PAGE electrophoresis and 32P incorporation was determined by autoradiography. (unpublished
results from our group).
Figure 10. Nuclear import of mutant CK2β subunits. NIH3T3 cells were transiently transfected with different GFP-CK2β constructs as illustrated
in the diagram. Transfected cells were observed for protein localization. (Adapted from ref. 40).
A Zinc Ribbon Motif Is Essential for the Formation of Functional Tetrameric Protein Kinase CK2 127
feature of this regulatory protein. The integrity of this motif ap- 19. Chantalat L, Leroy D, Filhol O et al. Crystal structure of the human
pears crucial not only for the correct molecular architecture of this protein kinase CK2 regulatory subunit reveals its zinc finger-mediated
important protein, but also for its regulatory functions. dimerization. EMBO J 1999; 18(11):2930-40.
20. Leroy D, Heriche JK, Filhol O et al. Binding of polyamines to an
autonomous domain of the regulatory subunit of protein kinase CK2
Acknowledgements induces a conformational change in the holoenzyme. A proposed
We acknowledge L. Chantalat and O. Dideberg for help in role for the kinase stimulation. J Biol Chem 1997; 272(33):20820-7.
21. Boldyreff B, Meggio F, Pinna LA et al. Reconstitution of normal
generating structural diagrams. We thank A. Larsen for provid- and hyperactivated forms of casein kinase-2 by variably mutated
ing us with Topoisomerase II, A. Joliot for Engrailed, and C. Mazza beta-subunits. Biochemistry 1993; 32(47):12672-7.
for the PHAX protein. Work on CK2 in our group is supported 22. Krishna SS, Majumdar I, Grishin NV. Structural classification of
by grants from the INSERM, the CNRS (contrat 8BC06G), the zinc fingers: survey and summary. Nucleic Acids Res 2003;
Commissariat à l’Energie Atomique, the Association pour la Re- 31(2):532-50.
cherche contre le Cancer (réseau ARECA), and the Ligue 23. Laity JH, Lee BM, Wright PE. Zinc finger proteins: new insights
Nationale contre le Cancer. into structural and functional diversity. Curr Opin Struct Biol 2001;
11(1):39-46.
24. Mackay JP, Crossley M. Zinc fingers are sticking together. Trends
References Biochem Sci 1998; 23(1):1-4.
1. Hunter T. Signaling-2000 and beyond. Cell 2000; 100(1):113-27. 25. Qian X, Gozani SN, Yoon H et al. Novel zinc finger motif in the
2. Allende JE, Allende CC. Protein kinases. 4. Protein kinase CK2: an basal transcriptional machinery: three-dimensional NMR studies of
enzyme with multiple substrates and a puzzling regulation. FASEB the nucleic acid binding domain of transcriptional elongation factor
J 1995; 9(5):313-23. TFIIS. Biochemistry 1993; 32(38):9944-59.
3. Guerra B, Issinger OG. Protein kinase CK2 and its role in cellular 26. Zhu W, Zeng Q, Colangelo CM et al. The N-terminal domain of
proliferation, development and pathology. Electrophoresis 1999; TFIIB from Pyrococcus furiosus forms a zinc ribbon. Nat Struct
20(2):391-408. Biol 1996; 3(2):122-4.
4. Litchfield DW. Protein kinase CK2: structure, regulation and role in 27. Wang B, Jones DN, Kaine BP et al. High-resolution structure of an
cellular decisions of life and death. Biochem J 2003; 369(Pt 1):1-15. archaeal zinc ribbon defines a general architectural motif in eukaryotic
5. Meggio F, Pinna LA. One-thousand-and-one substrates of protein RNA polymerases. Structure 1998; 6(5):555-69.
kinase CK2? FASEB J 2003;17(3):349-68. 28. Grishin NV. C-terminal domains of Escherichia coli topoisomerase
6. Padmanabha R, Chen-Wu JL, Hanna DE et al. Isolation, sequencing, I belong to the zinc-ribbon superfamily. J Mol Biol 2000;
and disruption of the yeast CKA2 gene: casein kinase II is essential for 299(5):1165-77.
viability in Saccharomyces cerevisiae. Mol Cell Biol 1990; 10(8):4089-99. 29. Filhol O, Cochet C, Wedegaertner P et al. Coexpression of both
7. Xu X, Toselli PA, Russell LD et al. Globozoospermia in mice lacking alpha and beta subunits is required for assembly of regulated casein
the casein kinase II alpha’ catalytic subunit. Nat Genet 1999; kinase II. Biochemistry 1991; 30(46):11133-40.
23(1):118-21. 30. Marin O, Meggio F, Sarno S et al. Physical dissection of the structural
8. Fraser AG, Kamath RS, Zipperlen P et al. Functional genomic elements responsible for regulatory properties and intersubunit
analysis of C. elegans chromosome I by systematic RNA interference. interactions of protein kinase CK2 beta-subunit. Biochemistry 1997;
Nature 2000; 408(6810):325-30. 36(23):7192-8.
9. Buchou T, Vernet M, Blond O et al. Disruption of the regulatory 31. Meggio F, Boldyreff B, Marin O et al. Role of the beta subunit of
Beta subunit of protein kinase CK2 in mice leads to a casein kinase-2 on the stability and specificity of the recombinant
cell-autonomous defect and early embryonic lethality. Mol Cell Biol reconstituted holoenzyme. Eur J Biochem 1992; 204(1):293-7.
2003; 23(3):908-15. 32. Chen M, Cooper JA. The beta subunit of CKII negatively regulates
10. Landesman-Bollag E, Song DH, Romieu-Mourez R et al. Protein Xenopus oocyte maturation. Proc Natl Acad Sci USA 1997;
kinase CK2: signaling and tumorigenesis in the mammary gland. 94(17):9136-40.
Mol Cell Biochem 2001; 227(1-2):153-65. 33. Sugano S, Andronis C, Ong MS et al. The protein kinase CK2 is
11. Seldin DC, Leder P. Casein kinase II alpha transgene-induced murine involved in regulation of circadian rhythms in Arabidopsis. Proc Natl
lymphoma: relation to theileriosis in cattle. Science 1995; Acad Sci U S A 1999; 96(22):12362-6.
267(5199):894-7. 34. Luscher B, Litchfield DW. Biosynthesis of casein kinase II in
12. Ahmed K, Gerber DA, Cochet C. Joining the cell survival squad: lymphoid cell lines. Eur J Biochem 1994; 220(2):521-6.
an emerging role for protein kinase CK2. Trends Cell Biol 2002; 35. Graham KC, Litchfield DW. The regulatory beta subunit of protein
12(5):226-30. kinase CK2 mediates formation of tetrameric CK2 complexes. J Biol
13. Lozeman FJ, Litchfield DW, Piening C et al. Isolation and Chem 2000; 275(7):5003-10.
characterization of human cDNA clones encoding the alpha and the 36. Niefind K, Guerra B, Ermakowa I et al. Crystal structure of human
alpha’ subunits of casein kinase II. Biochemistry 1990; 29(36):8436-47. protein kinase CK2: insights into basic properties of the CK2
14. Maridor G, Park W, Krek W et al. Casein kinase II. cDNA holoenzyme. EMBO J 2001; 20(19):5320-31.
sequences, developmental expression, and tissue distribution of 37. Meggio F, Ruzzene M, Sarno S et al. pCMB treatment reveals the
mRNAs for alpha, alpha’, and beta subunits of the chicken enzyme. essential role of cysteinyl residues in conferring functional competence
J Biol Chem 1991; 266(4):2362-8. to the regulatory subunit of protein kinase CK2. Biochem Biophys
15. Shi X, Potvin B, Huang T et al. A novel casein kinase 2 alpha-subunit Res Commun 2000; 267(1):427-32.
regulates membrane protein traffic in the human hepatoma cell line 38. Canton DA, Zhang C, Litchfield DW. Assembly of protein kinase
HuH-7. J Biol Chem 2001; 276(3):2075-82. CK2: investigation of complex formation between catalytic and
16. Xu X, Rich ES, Jr, Seldin DC. Murine protein kinase CK2 alpha’: regulatory subunits using a zinc-finger-deficient mutant of CK2beta.
cDNA and genomic cloning and chromosomal mapping. Genomics Biochem J 2001; 358(Pt 1):87-94.
1998; 48(1):79-86. 39. Martel V, Filhol O, Nueda A et al. Dynamic localization/association
17. Niefind K, Guerra B, Pinna LA et al. Crystal structure of the catalytic of protein kinase CK2 subunits in living cells: a role in its cellular
subunit of protein kinase CK2 from Zea mays at 2.1 A resolution. regulation? Ann N Y Acad Sci 2002; 973:272-7.
EMBO J 1998; 17(9):2451-62. 40. Filhol O, Nueda A, Martel V et al. Live-cell fluorescence imaging
18. Jauch E, Melzig J, Brkulj M et al. In vivo functional analysis of reveals the dynamics of protein kinase CK2 individual subunits. Mol
Drosophila protein kinase casein kinase 2 (CK2) beta-subunit. Gene Cell Biol 2003; 23(3):975-87.
2002; 298(1):29-39. 41. Filhol O, Martiel JL, Cochet C. Protein kinase CK2: A new view
of an old molecular complex. EMBO Reports 2004; 4:351-355.
CHAPTER 19
Abstract
T
he FYVE finger is an evolutionarily conserved double-zinc subunit (named Vps15 and p150 in yeast and mammals, respec-
binding domain with structural similarity to RING and tively) which has serine/threonine protein kinase activity.7,8 To-
PHD fingers. It consists of two β-hairpins that are stabi- gether with several other proteins, the catalytic and regulatory
lized by a C-terminal α-helix and the coordination of two Zn2+ subunits are found in high-molecular weight complexes.9 Unlike
ions. The most characteristic feature of the FYVE finger is a R(R/ class I and II PI 3-kinases, class III PI 3-kinases are active even in
K)HHCR motif associated with the first β-strand. This motif unstimulated cells, and the levels of PI3P in most mammalian
mediates coordination of the ligand, phosphatidylinositol 3-phos- cell types remain fairly constant, typically amounting to about
phate (PI3P), via contacts with the phosphate and inositol hy- 0.2% of the amount of phosphatidylinositol.10 In certain cell types,
droxyl groups of PI3P. PI3P, a rare lipid formed on endosomes such as phagocytes, PI3P levels may increase more than two-fold
and phagosomes by phosphorylation of phosphatidylinositol, is upon cell stimulation, and this is likely to have important func-
crucial for phagosome maturation and endocytic trafficking; and tional implications.11,12 PI3P is mainly associated with limiting
FYVE finger proteins are important effectors of PI3P as it has membranes of early (sorting) endosomes and intraluminal vesicles
been shown that several FYVE finger proteins among the 27 of multivesicular endosomes.13 This highly specific localization
human proteins play central roles in endocytic and phagocytic probably results from the fact that the early-endosomal GTPase
membrane trafficking. Rab5 interacts with p150 and thereby directs formation of PI3P
on early endosomes only.14,15
Introduction Metabolism of PI3P can occur via three pathways.16 Quanti-
The FYVE finger was originally discovered as a double-zinc tatively, the most important metabolic pathway appears to be deg-
finger required for the association of the early-endosome radation by lipases within the lumen of multivesicular
autoantigen EEA1 with endosome membranes. Its name derives endosomes.13,17 However, it is also clear that PI3P can be de-
from four of the first proteins shown to contain this motif (Fab1, phosphorylated by PI3P 3-phosphatases of the myotubularin fam-
YOTB, Vac1, EEA1).1 Genomic analyses show that the FYVE ily18-20 or phosphorylated by the PI3P 5-kinase Fab1/PIKfyve (Fig.
finger is highly conserved in yeast, nematodes, flies, plants and 1).21,22
mammals, and that typically about 0.1% of the genes encode Evidence from genetic and pharmacological studies indicates
FYVE finger proteins.2 Thus, Saccharomyces cerevisiae contains 5 that PI3P is crucial for endocytic membrane trafficking and pha-
FYVE finger genes, whereas the human genome contains 27 such gosome maturation.23,24 Indeed, the catalytic and regulatory sub-
genes. With few exceptions, FYVE fingers occur in only one copy units of the class III PI 3-kinase in yeast, Vps34 and Vps15, were
per protein. While most zinc fingers bind to large ligands (nucleic both isolated in a screen for vacuolar protein sorting mutants
acids and proteins), the FYVE finger highly specifically binds to (the vacuole is the yeast counterpart of mammalian lysosomes).7,25
a membrane lipid, phosphatidylinositol 3-phosphate (PI3P).3-5 Yeast cells devoid of Vps34 or Vps15 do not display any overt
We will here review the structural basis and functional implica- defects in processes other than vacuolar protein sorting and au-
tions of this interaction. tophagy, indicating that in this organism PI3P is exclusively re-
quired for these processes. The picture is a little more compli-
cated in mammalian cells, in which the role of PI3P has been
Formation and Functions of studied with PI 3-kinase inhibitors and with microinjected anti-
Phosphatidylinositol 3-Phosphate bodies against hVps34. Since pharmacological inhibitors of PI
Phosphatidylinositol 3-phosphate (PI3P) is formed by 3'-phos- 3-kinase also inhibit class I PI 3-kinases (which produce
phorylation of the abundant membrane phospholipid, phosphoinositides other than PI3P),6 pharmacological studies
phosphatidylinositol (Fig. 1). The reaction is catalyzed by alone do not unequivocally establish a role for PI3P in cellular
phosphatidylinositol 3-kinases (PI 3-kinases), notably the class processes. However, when studies using microinjected
III enzymes.6 The latter are the most conserved among PI 3-ki- anti-hVps34 or synthetic PI3P are also taken into account, we
nases and are found both in yeast (Vps34) and in humans can conclude that the following membrane trafficking processes
(hVps34). The catalytic subunit is associated with a regulatory in mammalian cells are controlled by PI3P: Endocytic membrane
Zinc Finger Proteins: From Atomic Contact to Cellular Function, edited by Shiro Iuchi and Natalie Kuldell.
©2005 Landes Bioscience/Eurekah.com and Kluwer Academic/Plenum Publishers.
The FYVE Finger: A Phosphoinositide Binding Domain 129
Figure 1. Formation and metabolism of PI3P. PI3P is formed by phosphorylation of phosphatidylinositol by a class III PI 3-kinase. Turnover of
PI3P may occur by lysosomal lipases (not shown), dephosphorylation by members of the myotubularin family, or 5-phosphorylated by Fab1/PIKfyve.
fusion,26-28 endosomal invagination,29,30 phagosome matura- to the structural integrity of the FYVE finger.38 Only two muta-
tion12,31 and autophagy.32,33 For all these processes, FYVE finger tions were found to inhibit ligand binding without causing struc-
containing effectors of PI3P have already been identified (see below tural distortion, those of the underlined arginine residues in the
and Table 1). RRHHCR motif. The equivalent residues of other FYVE fingers
should therefore be regarded as primary candidates for mutagen-
Structure and Ligand Binding of the FYVE esis whenever a functional analysis that investigates the role of
PI3P binding is required.
Finger Monomeric FYVE fingers bind the PI3P headgroup with too
The FYVE finger typically consists of 60-70 amino acid resi- low affinity (KD values in the high micromolar range) to explain
dues and shares both sequence and structural similarities with the targeting of these proteins to endosomal membranes.39 Bind-
RING and PHD fingers.2 The structure is built around a pair of ing to such membranes is strengthened by several mechanisms.
β-hairpins which are stabilized by a C-terminal α-helix and two First, a conserved hydrophobic “turret” loop preceding the
Zn2+-binding clusters.34 Each Zn2+-binding cluster consists of two β1-strand is predicted to insert into the lipid bilayer,34-36 (see
pairs of CXXC motifs (in single-letter amino acid code; X is any Fig. 2A) and biophysical studies using purified FYVE fingers and
amino acid), in which the cysteines coordinate Zn2+. In a few liposome membranes support this idea.37,40 Thus, the KD value
cases, the fifth cysteine is replaced by a histidine residue. Similar for binding of a monomeric Hrs FYVE domain to
to RING and PHD fingers, the two cysteine rich clusters bind PI3P-containing membranes has been estimated to be about 2.5
two Zn2+ ions in a cross-braced fashion. Beside the 8 conserved µM, which is a more than 50-fold increase in affinity over the
cysteines, 3 short sequence motifs distinguish the FYVE finger: isolated PI3P headgroup.39 Second, avidity for PI3P-containing
An N-terminal WXXD motif, a R(R/K)HHCR motif associated membranes can be enhanced through multimerization. This has
with the β1-strand, and a C-terminal RVC motif.2 Three been well documented in the case of EEA1, which forms a paral-
high-resolution crystal structures and one solution structure of lel dimer due to coiled-coil interactions N-terminal to the FYVE
FYVE fingers are available,34-37 and the crystal structure of the finger (Fig. 2B) and which significantly increases the avidity for
EEA1 FYVE finger bound to inositol(1,3)bisphosphate,36 reveals PI3P-containing membranes.36 Likewise, while the isolated FYVE
that all the three conserved sequence motifs contribute to ligand finger of Hrs is cytosolic, a tandem Hrs FYVE finger is efficiently
binding (Fig. 2A). In particular, the R(R/K)HHCR motif creates targeted to endosomes.13 Third, the association of FYVE finger
a shallow basic pocket that mediates PI3P binding (Fig. 2B). The containing proteins with membranes may be strengthened through
EEA1 FYVE structure also explains why the FYVE finger only interactions with additional molecules. For example, a domain
binds PI3P and not related phosphoinositides phosphorylated in adjacent to the FYVE finger of EEA1 binds to the endosomal
the 4'- and/or 5'- positions of the inositol ring. Binding of the GTPase Rab5,41 and whereas the monomeric EEA1 FYVE finger
latter ligands would cause sterical problems and disrupt hydro- is not targeted to endosome membranes, constructs that contain
gen bonds formed with inositol ring hydroxyl groups. additional dimerization/Rab5-binding domains are.1
Alanine mutagenesis of conserved residues of the EEA1 FYVE Even though sequences in addition to the FYVE finger are
finger has shown that most of the conserved residues contribute required for endosomal targeting of EEA1 and Hrs, there are a
130 Zinc Finger Proteins: From Atomic Contact to Cellular Function
A B
Figure 2. Structural basis of the interaction of the EEA1 FYVE domain with PI3P. A) The indicated side groups of the EEA1 FYVE domain interact
with inositol 1,3-bisphosphate, the headgroup of PI3P. Note that all the three conserved motifs of the FYVE domain (bottom line; residue
nubmering according to the published sequence of EEA177) are involved in ligand binding. Black= carbon; red= oxygen; blue= nitrogen; yellow=
sulfur. Based on the published crystal structure of the ligand bound FYVE domain of EEA136 and adapted from ref. 2. B) model for the binding
of dimeric EEA1 to a PI3P-containing membrane. The molecular surfaces are shown in the same orientation as the schematic drawing and are
colored blue (basic), red (acidic), green (hydrophobic), and white (uncharged polar). The two Zn2+ ions are shown as light blue spheres. The
structure is docked to a membrane, with the polar interface and hydrophobic core regions of the membrane shown to scale. Residues predicted
to be involved in nonspecific hydrophobic and electrostatic interactions with membrane lipids other than PI3P are shown. Myristoyl tails on PI3P
have been modeled. Based on the published crystal structure of the ligand bound FYVE domain of EEA136 and reproduced with kind permission from
J. Hurley.78
The FYVE Finger: A Phosphoinositide Binding Domain 131
few examples of FYVE fingers that by themselves are sufficient EEA1 inhibit phagosome maturation.54 The requirement for
for endosomal targeting. The FYVE fingers of SARA, Endofin EEA1 in phagosomal maturation could possibly reflect the need
and FENS-1 are efficiently targeted to endosomal membranes for Golgi- or endosome-derived vesicles to fuse with the phago-
without the need for additional domains.42-45 It is not known if some as part of the maturation process.24
this reflects an inherent propensity of these FYVE fingers to dimer- In addition to EEA1 and Rabenosyn-5, several other FYVE
ize or whether they interact with additional endosomal molecules. finger containing proteins serve as regulators of endocytic traf-
ficking, albeit at distinct steps. Rabip4/RUFY is an effector of
Cellular Functions of FYVE Finger Proteins the early-endosomal Rab GTPase Rab4, which, in contrast to
Most of the FYVE finger proteins expressed in yeast and hu- Rab5, regulates recycling of endocytosed molecules to the plasma
mans have been functionally characterized to some extent (see membrane. Overexpression studies suggest that Rabip4 indeed
examples in Table 1). Since PI3P is mainly localized to functions as a regulator of endocytic recycling, possibly by regu-
endosomes,13 it is hardly surprising that many FYVE finger pro- lating vesicle formation on endosomal membranes.55,56 Other
teins are found on these organelles. However, certain FYVE fin- endosomal FYVE finger proteins are involved in cargo sorting
ger proteins, such as the Cdc42 GDP/GTP exhange factor Fgd1 and endosomal involution. Hrs (hepatocyte growth factor regu-
and the myotubularin-related PI3P phosphatase MTMR3 (which lated tyrosine kinase substrate)57 is a clathrin-associated protein58
has been implicated in autophagy) are localized to other organelles. which binds to ubiquitinated membrane proteins and targets these
The reason why these proteins are not localized to endosomes to lysosomes.59-62 Besides its sorting function, this protein is also
could be that their FYVE domains bind PI3P with low affintiy or required for endosomal invagination.21,62 Interestingly, another
that they contain additional targeting domains that override the FYVE finger protein, PIKfyve also seems to play a role in endo-
endosomal targeting conferred by the FYVE domain.20,46 some invagination, although its possible relationship with Hrs is
Several FYVE finger proteins bind to the GTP-bound form not known.63 Unlike most other FYVE finger proteins, PIKfyve
(and thus serve as effectors) of endosomal Rab GTPases (Table has enzymatic activity -it is a PI3P 5-kinase which phosphory-
1). Both EEA1 and Rabenosyn-5 are FYVE finger containing lates PI3P into PI(3,5)P2. Evidence primarily derived from stud-
Rab5 effectors that are required for Rab5-dependent membrane ies of Fab1, the yeast homolog of PIKfyve, have indicated that
fusion in the endocytic pathway.41,47,48 This includes homotypic PI(3,5)P2 is crucial for the formation of intraluminal endosomal
fusion between early endosomes as well as heterotypic fusion be- vesicles, and for membrane trafficking from the vacuole (lyso-
tween endocytic or Golgi-derived vesicles and early endosomes. some) to other compartments.21,64,65 Surprisingly, while EEA1,
Yeast cells defective of Vac1, the yeast homolog of Rabenosyn-5, Rabenosyn-5, Hrs and PIKfyve are targeted to endosomes by vir-
have impaired Golgi to endosome trafficking, presumably because tue of their FYVE domains.1,47,66,67 endosomal targeting of Rabip4
Golgi-derived vesicles do not fuse with early endosomes.49,50 In appears to depend on an N-terminal RUN domain. The FYVE
addition to binding to Rab5, EEA1 and Rabenosyn-5 also bind domain of this protein might nevertheless play a role in its re-
to other endosomal Rab GTPases. EEA1 has a C-terminal Rab5 cruitment into specific membrane microdomains.68
binding domain (adjacent to the FYVE finger) and an N-terminal Although several FYVE finger containing proteins serve as regu-
C2H2 zinc finger which can bind both Rab5 and Rab22, an lators of endocytic membrane trafficking, at least one FYVE fin-
endosomal GTPase with unknown function.51 Since the two ger protein functions in signal transduction. SARA (Smad an-
Rab binding domains are spaced apart by a long coiled-coil re- chor for receptor activation) localizes to early endosomes via PI3P
gion, it is possible that simultaneous binding to Rab GTPases on binding by its FYVE domain, and here it serves as a scaffold for
two apposing membranes may serve to tether vesicles/endosomes Smad2, a crucial signal transducer downstream of the activated
prior to fusion. Both Rabenosyn-5 and Vac1 contain an transforming growth factor-β (TGFβ) receptor.42,69 Phosphory-
N-terminal C2H2 zinc finger similar to that found in EEA1. Sur- lation of Smad2 by the ligand-bound receptor causes Smad2 to
prisingly, while the C2H2 finger of Rabenosyn-5 binds to Rab5, dissociate from endosomes, and to associate with Smad4. The
the similar zinc finger of Vac1 does not bind to the yeast Rab5 Smad2/Smad4 complex then enters the nucleus in order to acti-
homolog Vps21.52 In addition to the N-terminal Rab5 binding vate transcription of TGFβ–responsive genes. Interestingly, Smad2
domain, Rabenosyn-5 also has a C-terminal Rab5 binding do- phosphorylation and nuclear translocation depend on endocyto-
main.53 Thus, like EEA1, this protein might potentially serve as sis of the receptor/ligand complex and its localization to EEA1/
a tether between two Rab5 positive membranes, even though it SARA-positive early endosomes.43 This provides a clear example
does not contain the long coiled-coil regions found in EEA1. of compartmentalized receptor signalling, and SARA plays a cru-
However, Rabenosyn-5 in addition contains a Rab4 binding do- cial role in this compartmentalization by restricting Smad2 phos-
main in the central part of the molecule, just C-terminal to the phorylation to early endosomes. An alternative internalization
FYVE finger. Since overexpression of Rabenosyn-5 causes the pathway of TGFβ receptors into caveolin-positive (but
apparent merging of otherwise separate Rab4 and Rab5 contain- EEA1-negative) vesicles (“caveosomes”) causes the association of
ing endosomal microdomains, it has been proposed that the receptor with a ubiquitin ligase and results in its rapid turn-
Rabenosyn-5 may serve as an organizer of such microdomains.53 over.70 Thus, segregation of TGFβ receptors into distinct
It is not yet understood how the same protein may serve to regu- endocytic compartments regulates Smad activation and receptor
late both membrane tethering/fusion and domain organization. turnover. It is worth noting that SARA may not exclusively func-
Studies using microinjected antibodies indicate that hVPS34 tion in signal transduction: Similar to several other FYVE finger
and hence PI3P is required for maturation of phagosomes, a pro- proteins, this protein also appears to play a role in the regulation
cess that is crucial for efficient inactivation of phagocytosed mi- of endocytic membrane traffic.71 One might conclude that the
croorganisms.12,54 Interestingly, EEA1 associates with early most common function of FYVE finger proteins is to regulate
phagosomes in a PI3P dependent manner, and antibodies against membrane trafficking into or out of the endosome.
132 Zinc Finger Proteins: From Atomic Contact to Cellular Function
Conclusion 18. Taylor GS, Maehama T, Dixon JE. Myotubularin, a protein tyrosine
phosphatase mutated in myotubular myopathy, dephosphorylates the
The FYVE finger is a rare but important zinc finger that serves lipid second messenger, phosphatidylinositol 3-phosphate. Proc Natl
to recruit a small subset of proteins to endosomal membranes by Acad Sci USA 2000; 97:8910-8915.
binding to PI3P. Such recruitment ensures specificity in endocytic 19. Laporte J, Liaubet L, Blondeau F et al. Functional redundancy in
membrane fusion and receptor sorting and creates an environ- the myotubularin family. Biochem Biophys Res Commun 2002;
ment for compartmentalized receptor signaling. Some variant 291:305-312.
FYVE fingers lack several of the residues that mediate PI3P bind- 20. Walker DM, Urbé S, Dove SK et al. Characterization of MTMR3:
an inositol lipid 3-phosphatase with novel substrate activity. Curr
ing.2,72 These FYVE fingers bind PI3P with too low affinity to be Biol 2001; 11:1600-1605.
recruited to endosomes, and their functions will have to be re- 21. Odorizzi G, Babst M, Emr SD. Fab1p PtdIns(3)P 5-kinase function
vealed by future studies. Even though the FYVE finger is princi- essential for protein sorting in the multivesicular body. Cell 1998;
pally a lipid binding domain, some FYVE fingers may well be 95:847-858.
involved in protein-protein interactions as well.73 22. Sbrissa D, Ikonomov OC, Shisheva A. PIKfyve, a mammalian
ortholog of yeast Fab1p lipid kinase, synthesizes 5-phosphoinositides.
J Biol Chem 1999; 274:21589-21597.
References 23. Gillooly DJ, Simonsen A, Stenmark H. Cellular functions of
1. Stenmark H, Aasland R, Toh BH et al. Endosomal localization of phosphatidylinositol 3-phosphate and FYVE domain proteins.
the autoantigen EEA1 is mediated by a zinc-binding FYVE finger. J Biochem J 2001; 355:249-258.
Biol Chem 1996; 271:24048-24054. 24. Gillooly DJ, Simonsen A, Stenmark H. Phosphoinositides and
2. Stenmark H, Aasland R, Driscoll PC. The phosphatidylinositol phagocytosis. J Cell Biol 2001; 155:15-17.
3-phosphate-binding FYVE finger. FEBS Lett 2002; 513:77-84. 25. Schu PV, Takegawa K, Fry MJ et al. Phosphatidylinositol 3-kinase
3. Gaullier J-M, Simonsen A, D’Arrigo A et al. FYVE fingers bind encoded by yeast VPS34 gene essential for protein sorting. Science
PtdIns(3)P. Nature 1998; 394:432-433. 1993; 260:88-91.
4. Patki V, Lawe DC, Corvera S et al. A functional PtdIns(3)P-binding 26. Jones AT, Clague MJ. Phosphatidylinositol 3-kinase activity is
motif. Nature 1998; 394:433-434. required for early endosome fusion. Biochem J 1995; 311:31-34.
5. Burd CG, Emr SD. Phosphatidylinositol(3)-phosphate signaling 27. Li G, D’Souza-Schorey C, Barbieri MA et al. Evidence for
mediated by specific binding to RING FYVE domains. Mol Cell phosphatidylinositol 3-kinase as a regulator of endocytosis via
1998; 2:157-162. activation of Rab5. Proc Natl Acad Sci USA 1995; 92:10207-10211.
6. Vanhaesebroeck B, Leevers SJ, Ahmadi K et al. Synthesis and 28. Siddhanta U, McIlroy J, Shah A et al. Distinct roles for the p110α
function of 3-phosphorylated inositol lipids. Annu Rev Biochem and hVPS34 phosphatidylinositol 3'- kinases in vesicular trafficking,
2001; 70:535-602. regulation of the actin cytoskeleton, and mitogenesis. J Cell Biol
7. Stack JH, DeWald DB, Takegawa K et al. Vesicle-mediated protein 1998; 143:1647-1659.
transport: Regulatory interactions between the Vps15 protein kinase 29. Fernandez-Borja M, Wubbolts R, Calafat J et al. Multivesicular body
and the Vps34 PtdIns 3-kinase essential for protein sorting to the morphogenesis requires phosphatidylinositol 3-kinase activity. Curr
vacuole in yeast. J Cell Biol 1995; 129:321-334. Biol 1999; 14:55-58.
8. Volinia S, Dhand R, Vanhaesebroeck B et al. A human 30. Futter CE, Collinson LM, Backer JM et al. Human VPS34 is
phosphatidylinositol 3-kinase complex related to the yeast required for internal vesicle formation within multivesicular bodies.
Vps34p-Vps15p protein sorting system. EMBO J 1995; J Cell Biol 2001; 155:1251-1263.
14:3339-3348. 31. Fratti RA, Backer JM, Gruenberg J et al. Role of phosphatidylinositol
9. Kihara A, Noda T, Ishihara N et al. Two distinct Vps34 3-kinase and Rab5 effectors in phagosomal biogenesis and
phosphatidylinositol 3-kinase complexes function in autophagy and mycobacterial phagosome maturation arrest. J Cell Biol 2001;
carboxypeptidase Y sorting in Saccharomyces cerevisiae. J Cell Biol. 154:631-644.
2001 ;152:519-530. 32. Blommaart EFC, Krause U, Schellens JPM et al. The
10. Stephens L, McGregor A, Hawkins P. Phosphoinositide 3-kinases: phosphatidylinositol 3-kinase inhibitors wortmannin and LY294002
Regulation by cell-surface receptors and function of 3-phosphorylated inhibit autophagy in isolated rat hepatocytes. Eur J Biochem 1997;
lipids. In: Cockcroft S, ed. Biology of phosphoinositides. Oxford: 243:240-246.
Oxford University Press, 2000:32-108. 33. Petiot A, Ogier-Denis E, Blommaart EFC et al. Distinct classes of
11. Ellson CD, Anderson KE, Morgan G et al. Phosphatidylinositol phosphatidylinositol 3'-kinases are involved in signaling pathways
3-phosphate is generated in phagosomal membranes. Curr Biol 2001; that control macroautophagy in HT-29 cells. J Biol Chem 2000;
11:1631-1635. 275:992-998.
12. Vieira OV, Botelho RJ, Rameh L et al. Distinct roles of class I and 34. Misra S, Hurley JH. Crystal structure of a phosphatidylinositol
III phosphatidylinositol 3-kinases in phagosome formation and 3-phosphate-specific membrane-targeting motif, the FYVE domain
maturation. J Cell Biol 2001; 155:19-25. of Vps27p. Cell 1999; 97:657-666.
13. Gillooly DJ, Morrow IC, Lindsay M et al. Localization of 35. Mao Y, Nickitenko A, Duan X et al. Crystal structure of the VHS
phosphatidylinositol 3-phosphate in yeast and mammalian cells. and FYVE tandem domains of Hrs, a protein involved in membrane
EMBO J 2000; 19:4577-4588. trafficking and signal transduction. Cell 2000; 100:447-456.
14. Christoforidis S, Miaczynska M, Ashman K et al. 36. Dumas JJ, Merithew E, Sudharshan E et al. Multivalent endosome
Phosphatidylinositol-3-OH kinases are Rab5 effectors. Nature Cell targeting by homodimeric EEA1. Mol Cell 2001; 8:947-958.
Biol 1999; 1:249-252. 37. Kutateladze T, Overduin M. Structural mechanism of endosome
15. Murray JT, Panaretou C, Stenmark H et al. Role of Rab5 in the docking by the FYVE domain. Science 2001; 291:1793-1796.
recruitment of hVps34/p150 to the early endosome. Traffic 2002; 38. Gaullier J-M, Rønning E, Gillooly DJ et al. Interaction of the EEA1
3:416-427. FYVE finger with phosphatidylinositol 3-phosphate and early
16. Stenmark H, Gillooly DJ. Intracellular trafficking and turnover of endosomes. Role of conserved residues. J Biol Chem 2000;
phosphatidylinositol 3-phosphate. Semin Cell Dev Biol 2001; 275:24595-24600.
12:193-199. 39. Sankaran VG, Klein DE, Sachdeva MM et al. High-affinity binding
17. Wurmser AE, Emr SD. Phosphoinositide signaling and turnover: of a FYVE domain to phosphatidylinositol 3-phosphate requires
Ptdlns(3)P, a regulator of membrane traffic, is transported to the intact phospholipid but not FYVE domain oligomerization.
vacuole and degraded by a process that requires lumenal vacuolar Biochemistry 2001; 40:8581-8587.
hydrolase activities. EMBO J 1998; 17:4930-4942. 40. Stahelin RV, Long F, Diraviyam K et al. Phosphatidylinositol
3-phosphate induces the membrane penetration of the FYVE
domains of Vps27p and Hrs. J Biol Chem 2002; 277:26379-26388.
The FYVE Finger: A Phosphoinositide Binding Domain 133
41. Simonsen A, Lippé R, Christoforidis S et al. EEA1 links PI(3)K 60. Shih SC, Katzmann DJ, Schnell JD et al. Epsins and Vps27p/Hrs
function to Rab5 regulation of endosome fusion. Nature 1998; contain ubiquitin-binding domains that function in receptor
394:494-498. endocytosis. Nat Cell Biol 2002; 4:389-393.
42. Panopoulou E, Gillooly DJ, Wrana JL et al. Early endosomal 61. Bilodeau PS, Urbanowski JL, Winistorfer SC et al. The Vps27p
regulation of Smad-dependent signaling in endothelial cells. J Biol Hse1p complex binds ubiquitin and mediates endosomal protein
Chem 2002; 277:18046-18052. sorting. Nat Cell Biol 2002; 4:534-539.
43. Hayes S, Chawla A, Corvera S. TGF beta receptor internalization 62. Lloyd TE, Atkinson R, Wu MN et al. Hrs regulates endosome
into EEA1-enriched early endosomes: Role in signaling to Smad2. J invagination and receptor tyrosine kinase signaling in Drosophila.
Cell Biol 2002; 158:1239-1249. Cell 2002; 108:261-269.
44. Seet LF, Hong W. Endofin, an endosomal FYVE domain protein. J 63. Ikonomov OC, Sbrissa D, Shisheva A. Mammalian cell morphology
Biol Chem 2001; 276:42445-42454. and endocytic membrane homeostasis require enzymatically active
45. Ridley SH, Ktistakis N, Davidson K et al. FENS-1 and DFCP-1 phosphoinositide 5-kinase PIKfyve. J Biol Chem 2001;
are FYVE-domain containing proteins with distinct functions in the 276:26141-26147.
endosomal and Golgi compartments. J Cell Sci 2001; 64. Yamamoto A, DeWald DB, Boronenkov IV et al. Novel PI(4)P
114:3991-4000. 5-kinase homologue, Fab1p, essential for normal vacuole function
46. Nagata K, Driessens M, Lamarche N et al. Activation of G1 and morphology in yeast. Mol Biol Cell 1995; 6:525-539.
progression, JNK mitogen-activated protein kinase, and actin filament 65. Gary GD, Wurmser AE, Bonangelino CJ et al. Fab1p is essential
assembly by the exchange factor FGD1. J Biol Chem 1998; for PtdIns(3)P 5-kinase activity and the maintenance of vacuolar
273:15453-15457. size and membrane homeostasis. J Cell Biol 1998; 143:65-79.
47. Nielsen E, Christoforidis S, Uttenweiler-Joseph S et al. Rabenosyn-5, 66. Raiborg C, Bremnes B, Mehlum A et al. FYVE and coiled-coil
a novel Rab5 effector, is complexed with hVPS45 and recruited to domains determine the specific localisation of Hrs to early
endosomes through a FYVE finger domain. J Cell Biol 2000; endosomes. J Cell Sci 2001; 114:2255-2263.
151:601-612. 67. Sbrissa D, Ikonomov OC, Shisheva A. Phosphatidylinositol
48. Rubino M, Miaczynska M, Lippé R et al. Selective membrane 3-phosphate-interacting domains in PIKfyve. Binding specificity and
recruitment of EEA1 suggests a role in directional transport of role in PIKfyve. Endomenbrane localization. J Biol Chem 2002;
clathrin-coated vesicles to early endosomes. J Biol Chem 2000; 277:6073-6079.
275:3745-3748. 68. Mari M, Macia E, Marchand-Brustel Y et al. Role of FYVE-finger
49. Weisman LS, Wickner W. Molecular characterization of VAC1, a and the run domain for the subcellular localization of RABIP4. J
gene required for vacuole inheritance and vacuole protein sorting. J Biol Chem 2001; 276:42501-42508.
Biol Chem 1992; 267:618-623. 69. Tsukazaki T, Chiang TA, Davison AF et al. SARA, a FYVE domain
50. Tall GG, Hama H, DeWald DB et al. The phosphatidylinositol protein that recruits Smad2 to the TGFβ receptor. Cell 1998;
3-phosphate binding protein Vac1p interacts with a Rab GTPase 95:779-791.
and a Sec1p homologue to facilitate vesicle-mediated vacuolar protein 70. Di Guglielmo GM, Le Roy C, Goodfellow AF et al. Distinct
sorting. Mol Biol Cell 1999; 10:1873-1889. endocytic pathways regulate TGF-β receptor signalling and turnover.
51. Kauppi M, Simonsen A, Bremnes B et al. The small GTPase Rab22 Nat Cell Biol 2003; 5:410-421.
interacts with EEA1 and controls endosomal membrane trafficking. 71. Hu Y, Chuang JZ, Xu K et al. SARA, a FYVE domain protein,
J Cell Sci 2002; 115:899-911. affects Rab5-mediated endocytosis. J Cell Sci 2002; 115:4755-4763.
52. Merithew E, Stone C, Eathiraj S et al. Determinants of Rab5 72. Stenmark H, Aasland R. FYVE-finger proteins - effectors of an
interaction with the N terminus of early endosome antigen 1. J inositol lipid. J Cell Sci 1999; 112:4175-4183.
Biol Chem 2003; 278:8494-8500. 73. Lawe DC, Patki V, Heller-Harrison R et al. The FYVE domain of
53. de Renzis S, Sonnichsen B, Zerial M. Divalent Rab effectors regulate early endosome antigen 1 is required for both phosphatidylinositol
the sub-compartmental organization and sorting of early endosomes. 3-phosphate and Rab5 binding. J Biol Chem 2000; 275:3699-3705.
Nat Cell Biol 2002; 4:124-133. 74. Christoforidis S, McBride HM, Burgoyne RD et al. The Rab5
54. Fratti RA, Backer JM, Gruenberg J et al. Role of phosphatidylinositol effector EEA1 is a core component of endosome docking. Nature
3-kinase and Rab5 effectors in phagosomal biogenesis and 1999; 397:621-626.
mycobacterial phagosome maturation arrest. J Cell Biol 2001; 75. Schaletzky J, Dove SK, Short B et al.
154:631-644. Phosphatidylinositol-5-phosphate activation and conserved substrate
55. Cormont M, Mari M, Galmiche A et al. A FYVE-finger-containing specificity of the myotubularin phosphatidylinositol 3-phosphatases.
protein, Rabip4, is a Rab4 effector involved in early endosomal Curr Biol 2003; 13:504-509.
traffic. Proc Natl Acad Sci USA 2001; 98:1637-1642. 76. Pasteris NG, Cadle A, Logie LJ et al. Isolation and characterization
56. Yang J, Kim O, Wu J et al. Interaction between tyrosine kinase Etk of the faciogenital dysplasia (Aarskog-Scott syndrome) gene: a putative
and a RUN domain- and FYVE domain-containing protein RUFY1. Rho/Rac guanine nucleotide exchange factor. Cell 1994; 79:669-678.
A possible role of ETK in regulation of vesicle trafficking. J Biol 77. Mu FT, Callaghan JM, Steele-Mortimer O et al. EEA1, an early
Chem 2002; 277:30219-30226. endosome-associated protein. EEA1 is a conserved alpha-helical
57. Komada M, Kitamura N. Growth factor-induced tyrosine peripheral membrane protein flanked by cysteine “fingers” and
phosphorylation of Hrs, a novel 115-kilodalton protein with a contains a calmodulin-binding IQ motif. J Biol Chem 1995;
structurally conserved putative zinc finger domain. Mol Cell Biol 270:13503-13511.
1995; 15:6213-6221. 78. Misra S, Miller GJ, Hurley JH. Recognizing phosphatidylinositol
58. Raiborg C, Bache KG, Mehlum A et al. Hrs recruits clathrin to 3-phosphate. Cell 2001; 107:559-562.
early endosomes. EMBO J 2001; 20:5008-5021.
59. Raiborg C, Bache KG, Gillooly DJ et al. Hrs sorts ubiquitinated
proteins into clathrin-coated microdomains of early endosomes. Nat
Cell Biol 2002; 4:394-398.
CHAPTER 20
Abstract
contain a MATH domain,29 while in D. melanogaster and the
T
he BTB/zinc finger proteins have a wide range of func-
tions in development and homeostasis and a wide range vertebrates, most BTB domains are most often found in proteins
of interactions. The BTB domain appears essential for that contain either kelch repeats,30 voltage-gated ion channels31
these proteins to dimerize, facilitate DNA looping, form (where the BTB domain is commonly known as the T1 domain),
specific multi-protein structures in the nucleus and interact with or Krüppel-type C2H2 zinc finger domains (Fig. 1).
co-repressor molecules. The BTB domains of the proteins differ The human genome encodes 205 BTB proteins, and these are
in their affinity for co-repressors and contribution to the roughly equally distributed among the three subclasses of
transcriptional activity of the proteins. The BTB domain also proteins. There are 47 BTB/zinc finger proteins, 53 BTB/kelch
allows interaction with other BTB proteins perhaps by forming proteins and 57 BTB(T1)/ion channels. The remaining ~20% of
higher order multimers and a network of BTB interactions likely human BTB domain proteins are associated with a variety of other
exists. All of the BTB/zinc finger proteins have other important domains, or with no identified domain such as the Skp1/ElonginC
functional domains. PLZF and Bcl6 have a second repression proteins. One recently identified subfamily are RhoBTB proteins
domain, while Miz-1 has an important activation domain consisting of a GTPase domain, a proline rich region and a
internal to the protein. In addition, the zinc fingers of these tandem arrangement of two BTB domains at the C-terminus (Fig.
proteins can interact with co-factors implicated in transcriptional 1). This family of BTB proteins is largely uncharacterized, but
activity as well as nuclear cytoplasmic shuttling. Lastly given the other Rho GTPase containing proteins have been shown to be
recent information indicating the importance of the BTB involved in the regulation of a number of cellular processes,
domain in ubiquitylation pathways the BTB/zinc finger proteins including cytoskeletal remodelling, cell cycle progression,
may also play a role in degradation of specific proteins in the cell. morphogenesis and apoptosis.32 There is also evidence linking
Whether this is related to or distinct from their transcriptional Rho-regulated signal transduction pathways with tumorigenesis
functions remains to be discovered. and metastasis. For instance, DBC2 (deleted in breast cancer) is a
recently identified RhoBTB protein homozygously deleted in
3.5% of breast tumours.33
Introduction The subclass of BTB/zinc finger proteins are generally thought
The BTB (Broad Complex, Tramtrack and Bric à Brac) to be transcription factors, and several of these have received
domain, also known as POZ (Poxvirus and Zinc finger) domain, considerable attention because of their involvement in human
is a highly conserved, widely distributed motif first identified in cancers. Two of the most studied human BTB domain proteins
Drosophila and poxvirus proteins.1-5 Examples of genes encoding are PLZF and BCL-6, which are transcriptional repressors
BTB domain proteins have since been found in all eukaryotes implicated in t(11:17)-associated acute promyelocytic leukemia
studied to date. The domain consists of 90-120 residues, and is (APL) and in diffuse large cell lymphoma (DLCL), respectively.
typically present as a single copy at the extreme N-terminus of The crystal structure of residues 6-126 of PLZF serves to identify
proteins that often contain other motifs. BTB domains are the structural features of the BTB domain in the zinc finger and
protein-protein interaction motifs, and were shown to mediate kelch repeat proteins.6 Briefly, the structure is a butterfly-shaped
homo-oligomerization, 6-8 hetero-oligomerization 5,9-12 and homodimer with a tightly interwound peptide chain (Fig. 2). The
interactions with non-BTB containing proteins.13-27 first ~30 residues form strand β1 and helix α1, and these ele-
While there are only four examples of BTB genes in S. cerevisiae, ments make up an integral part of the dimer interface. The re-
the number is increased dramatically in plants and in each of the maining core of the domain is mostly α-helical, with a short 3
major animal lines (nematodes, arthropods and vertebrates)28 stranded β-sheet exposed to the surface. The dimer interface is
(Table 1). The expansion of BTB domain proteins in multi-cellular hydrophobic and makes up one quarter of the monomer surface,
lineages suggests that these proteins many play suggesting that the PLZF BTB domain is an obligate homodimer.
crucial roles in cellular differentiation and development, and these Structural alignments of BTB-fold proteins reveal a core
expansions appear to be species-specific. For example, in the C. region of 90 amino acids common to all examples of the domain,
elegans genome, the majority of BTB domain proteins also including the T1 and Skp1 subfamiles of the fold. The pairwise
Zinc Finger Proteins: From Atomic Contact to Cellular Function, edited by Shiro Iuchi and Natalie Kuldell.
©2005 Landes Bioscience/Eurekah.com and Kluwer Academic/Plenum Publishers.
The BTB Domain Zinc Finger Proteins 135
H. sapiens 205 12
C. elegans 183 8
D. melanogaster 147 6
S. cerivisiae 5 > 100
Data compiled at http://xtal.uhnres.utoronto.ca/prive/
btb_database.html. Rank indicates the prevalence in the given
genome in the Interpro database.200
modules. In many BTB transcription factors, the domain is homodimer-homodimer contacts, generating much larger BTB
separated from the C-terminal DNA-binding region by several oligomers.
hundred residues, which are predicted to have little fixed
structure. This leads to a model in which an oligomeric chain of Interactions of the BTB Domain
BTB domains is linked to a series of DNA reading heads by long
flexible linkers, which can in turn interact with multiple DNA
with Co-repressors
recognition sites. It has been demonstrated that the ability PLZF The BTB domain is strongly implicated in the repression of
to form a high molecular weight DNA-protein complex is gene expression through the local control of chromatin
dependent on its BTB domain. Furthermore, the BTB domain is conformation. Several BTB domains were shown to interact with
necessary for directing PLZF,39 BCL-643 and ZID5 to a punctate components of a histone deacetylase complex, such as the
expression pattern in the nucleus, suggesting that these proteins co-repressors silencing mediator of retinoid and thyroid hormone
are part of large assemblies. receptor (SMRT), the nuclear receptor co-repressor (N-CoR),
In the case of Drosophila GAGA factor, polyvalent binding to Sin3A and histone deacetylase 1 (HDAC-1).13-22 Repression of
natural operator sequences was also dependent on the formation gene transcription by BTB containing transcription factors such
of higher order protein complexes mediated by the N-terminal as Bach2,53 PLZF,14-18,21,22 FAZF,54 ZID13 and BCL-613,17,19,20
BTB domain.44,45 GAGA factor is a constitutively expressed is mediated by the recruitment of this complex to the chromatin,
Drosophila transcription factor encoded by the essential where the removal of acetyl groups from the histone lysine tails
Trithorax-like gene46 and it binds to GA rich DNA sequences leads to changes in chromatin structure. Furthermore, transcrip-
through its single zinc finger.47 Unlike many BTB domain tional repression mediated by these particular BTB domains is
proteins, however, GAGA factor has been shown to activate reversible with trichostatin A, a specific inhibitor of HDAC
transcription.48 Electron microscopy (EM) experiments have activity.55 Disruption of the BTB fold of PLZF and BCL-6 by
directly shown that GAGA factor binds multiple DNA elements point mutation abrogates the interaction of the BTB domain with
these co-repressors, further verifying the critical role of BTB
as a large oligomer and that the DNA in this complex is bent and
dimerization.40,54 Recruitment of a histone deacetylase complex,
might be wrapped around the surface of a GAGA factor
however, is not a universal property of BTB domain containing
multimer.45 It is thought that, after transient chromatin remod-
transcriptional repressors. The HIC-1 (hypermethylated in
eling by NURF,49 the association of a GAGA factor oligomer
cancer) gene encodes a BTB domain with 13 amino acid alanine
with multiple binding sites within a promoter constrains the DNA
rich insert in the region analogous to an alpha helical region of
trajectory, consequently reorganizing the promoter DNA and
the solved BTB domain structures.56 This domain does not bind
maintaining it in an open conformation.45 This is correlated with
to N-Cor and SMRT and repression by this domain is not
in vivo evidence that have shown that GAGA-binding elements
inhibited by HDAC inhibitors.55 Furthermore the BTB domains
coincide with DNase-hypersensitive sites at promoters.50 Similar
of Bric à Brac 1 (BAB1) and Bric à Brac 2 (BAB2) interact with
higher order complexes have also been observed for the murine
BIP2, a Drosophila TATA-box protein associated factor (TAFII),57
BTB-containing transcription factor Bach1, for which polyvalent raising the possibility that some BTB domains may repress tran-
binding to natural operator sequences was dependent on the scription by interfering with the basal transcriptional machinery.
presence of the BTB domain.51
Chain formation may also be a mechanism for heteromeric
associations between different BTB proteins. Some examples of The Emerging Role of BTB Domains
heteromeric pairs formed between BTB domains include PLZF in Ubiquitylation
and BCL-6,9,10 PLZF and FAZF,11 BCL-6 and BAZF12 and Very recently the BTB domains was identified as a substrate
GAGA factor and TTK. 5,52 However, there appears to be a strict adaptor for cullin 3 (Cul3) E3 ligase complexes.23,24,26,27 Several
specificity to heterodimer formation, as it was shown that the groups found that BTB domain of the C. elegans protein Mel-26
following pairs of domains do not interact: PLZF and LRF,9 interacts directly with cullin 3 (Cul3) and targets the
BCL-6 and LRF,9 BCL-6 and kaiso,10 TTK and ZID,5 TTK and microtubule-severing protein MEI-1 for degradation by the Cul3
ZF5.5 It is important to note that such interactions have been E3-ubiquitin ligase complex.23,24,27 In yeast, S. pombe Cul3
observed by yeast two-hybrid, co-immunoprecipitation and in (Pcu3p) forms a complex with all three BTB proteins in this
vitro co-expression experiments only. In fact, when non-tagged species.26 Complexes of human Cul3 with BTB/kelch and BTB/
and tagged ZID derivatives of differing sizes were mixed together MATH proteins, as well as with the BTB/zinc finger protein PLZF
rather than co-translated, co-immunoprecipitation was inefficient, were found, strongly suggesting that a large number of BTB
indicating that limited exchange took place.5 Therefore the proteins may be regulated by the Cul3/ubiqutin ligase complex
notion that BTB heterodimers may form is an attractive but as or regulate the stability of other proteins through this complex.27
yet unproven possibility. For the DNA-binding members of the Skp1 and Elongin C are small proteins that consist almost
BTB family, heterodimerization might increase the number of exclusively of the core BTB fold, and these are part of the related
gene promoters that proteins family binds can bind to, increasing SCF (Skp1-Cul1-F-box)34,58 and ECS (Elongin C-Cul2-SOCS
the potential for functional variation. However, the biological box)59 E3-ligases, respectively. The latter complex is well known
relevance of the observed heterodimers remains to be clarified. because of the interactions with the VHL tumor suppressor, a
Based on their three-dimensional crystal structures, we have SOCS-box protein associated with the von Hippel-Lindau
proposed that the PLZF and BCL-6 BTB domains form obligate cancer syndrome and kidney cancer. Skp1 and Elongin C inter-
homodimers.6 This is due to the extensive hydrophobic interface act with F-box or SOCS-box adaptor proteins such as Cdc4 or
between monomers, and on thermal denaturation studies on the VHL to recognize target substrates. In the newly described
PLZF BTB domain showing a single step transition between Cul3-E3 ligase complexes, however, BTB domain proteins may
folded dimers and unfolded monomers. 6-8 It is possible, have a “built-in” substrate-recognizing functionality in their
nonetheless, that BTB heteromers can form through C-terminal downstream domains.25 For example, the N-terminal
The BTB Domain Zinc Finger Proteins 137
Figure 4. Domain structure of PLZF and the PLZF-RARa fusion proteins of acute promyelocytic leukemia.
BTB domain of Mel-26 associates with Cul3, while the C-terminal a poor prognosis and lack of response to all-transretinoic acid
kelch repeats bind directly to the target MEI-1.24,27 Thus, BTB (ATRA), in vivo and in vitro. 62 However, more recently
domain proteins may carry out both the cullin-interacting and several cases responsive to combination therapies such as ATRA
substrate-recognizing functions that are achieved by two proteins and G-CSF 65 or ATRA and chemotherapy,63,67 were reported
in the SCF and ECS systems. Given that zinc fingers are also indicating that the resistance of this syndrome is not absolute.
protein-protein interaction domains it also offers potential However in one of these cases only PLZF-RARα and not the
additional non-transcriptional functions for the BTB/zinc finger reciprocal fusion gene was expressed.67 This and another case64
proteins. suggested that PLZF-RARα may be sufficient to generate disease
and suggesting that the reciprocal RARα-PLZF protein might
The Role of PLZF in Leukemia play a role in the refractoriness of t(11;17)-APL.
The PLZF Gene and Its Fusion with RARα The PLZF Protein
The promyelocyticleukemia zinc finger (PLZF) gene was identi- The PLZF zinc finger transcription factor60 contains nine
fied by its rearrangement in a case of APL with t(11;17)(q23;q21) C-terminal Krüppel-like C2H2 zinc finger motifs and a N-terminal
translocation.60,61 Since then 16 cases of APL with a fusion BTB domain (Fig. 4). PLZF is a 673 amino acid nuclear serine/
between the PLZF and RARα genes were described,62-65 repre- threonine phosphorylated protein68 that binds to DNA specifi-
senting less than 1% of cases of APL. The PLZF gene is located cally to a consensus sequence of GTACT/AGTAC. The last three
on chromosome 11q23 about 1 MB centromeric to the MLL zinc fingers of PLZF which are retained in the RARα-PLZF
(mixed lineage leukemia) gene.66 The PLZF gene was completely fusion are sufficient to recognize the PLZF cognate binding
sequenced and found to be contained in six exons spread over site.36,68,69 In fact RARα-PLZF can bind to a site within the Hoxb2
201 kb of DNA. The resulting 8kb mRNA has a long 6kB 5’ promoter more strongly than PLZF itself70 suggesting that the
untranslated region of unknown function. In (t:11;17) N-terminal BTB domain, as demonstrated originally by
associated APL, the PLZF gene is rearrange with the RARα gene. Bardwell71 may negatively regulate DNA binding. The highly
The breakpoint in PLZF is between exons two and three except related BTB zinc finger FAZF72 also known as TZFP73 or ROG74
in one case where the breakpoint was more 3' in the PLZF gene also has a N-terminal BTB domain and only three C-terminal
between exons three and four. The chromosomal rearrangement zinc fingers which are very similar to those of PLZF. The three
is reciprocal, leading to two fusion gene products PLZF-RARα FAZF zinc fingers recognize the PLZF binding site.72,75 The FAZF
and RARα-PLZF. PLZF RARα contains the N-terminal portion gene is localized on chromosome 19q13.1 adjacent to the MLL2
of PLZF, which contains the transcriptional effector domains of gene. This information suggests that the genes encoding PLZF
PLZF lined to domains B-F of RARα including the DNA and MLL may have arisen from an ancient duplication of genetic
binding and ligand binding domain of RARα. The reciprocal material and the more simple FAZF protein may have been the
fusion RARα-PLZF links the A1 or A2 domain of RARα precursor of the larger PLZF. What the function of the more
resulting from the alternative promoter’s of the RARα gene to the N-terminal zinc fingers of PLZF might be is not yet certain. One
last 7 of nine zinc fingers of PLZF. The PLZF-RARα fusion is an publication suggested that these fingers might be important for
aberrant form of the retinoic acid receptor while RARα-PLZF is transcriptional repression function of the PLZF- RARα fusion.76
an abnormal form of PLZF (Fig. 4). In contrast to the more fre- The first three zinc fingers of PLZF can mediate interaction with
quent t(15;17) APL, this translocation is generally associated with the RAR as well as estrogen receptor in vitro and overexpression
138 Zinc Finger Proteins: From Atomic Contact to Cellular Function
of these zinc fingers or full-length PLZF can inhibit the as suspected and found for many otherleukemia-associated
transcriptional activity of those receptors and interfere with proteins that HOX genes may be major targets of PLZF and its
RARα/RXR interaction.77 Whether this is a significant interac- fusion protein with RARα.
tion in vivo remains unclear however. A recent paper showed that PLZF may play roles in other tissues as well. Recently PLZF
zinc fingers 5-8 interacted with the intracellular fragment of was identified as a gene whose expression was stimulated in
proHB-EGF and mediated nuclear-cytoplasmic shuttling of response to glucocorticoid and progesterone treatment of endome-
PLZF.78 trial stromal and myometrial cells.91 In addition PLZF was
The PLZF protein is localized to the nucleus79,80 in a pattern identified as a gene up-regulated in breast cancer cells in response
of ~50 small nuclear speckles,80 which are dependent on the to glucocorticoids and less so in response to progestins.92 PLZF
presence of the BTB domain.76 PLZF is co-localized or adjacent may in part mediate growth suppression by these agents. PLZF
to the PML protein which is disrupted in the more common was detected in the uterus in the mid and late secretory phase of
t(15;17) associated form of APL in a large multi-protein complex the menstrual cycle but not in the proliferative phase, consistent
called the PML oncogenic domain (POD) or nuclear body with the general negative effect of PLZF on cell growth. These
(NB)81,82 suggesting a functional interplay. In t(15;17) and t(5;17) results warrant examination of the reproductive tract of PLZF
APL, PLZF is de-localized into a microspeckled pattern similar null animals for possible developmental changes.
to that of PML-RARα21,81,83 suggesting that delocalization of
PLZF might be important in the pathogenesis of APL.83,84 Biological Effects of PLZF and PLZF Target Genes
Enforced expression of PLZF was achieved in myeloid 32D
PLZF in Development cells and U937 cells as well as non-hematopoetic cells NIH 3T3
PLZF is expressed in undifferentiated myeloid cell lines and cells. In all cases expression of PLZF led to G1 arrest and a block
at lower levels in more differentiated erythroleukemia, in the ability of cells to traverse the S-phase.38,78,86,93,94 Similarly
promyelocytic and monocytic cell lines as well as in peripheral expression of the highly related FAZF protein induces growth
blood mononuclear cells.60,79,80 PLZF is down-regulated during arrest virtually identical to PLZF in U937 cells.85 Prolonged
differentiation of NB4 and HL60 cells60 but is up-regulated in expression of PLZF or FAZF eventually leads to apoptosis,85,94
the MDS cell line after treatment with calcium ionophore, however somewhat paradoxically PLZF expressing cells could
perhaps recapitulating some aspect of monocyte development.80 survive two weeks of withdrawal from IL3.93 The ability of 32D
CD34+ human progenitor cells could be immunostained with cells to differentiate in the presence of G-CSF or GM-CSF was
PLZF antisera in a distinct nuclear speckled pattern.79 When such blocked by PLZF. This was not simply due to the fact that PLZF
cells were placed into culture and allowed to differentiate, PLZF blocks cell division, since even before CSF treatment, PLZF
levels declined with terminal differentiation.85 PLZF expression expressing cells showed elevated expression of the primitive Sca1
may be important for the maintenance or survival of hematopoietic marker and reduced expression of the mature granulocyte marker
stem cells and or early progenitors, down-regulated with Gr1. PLZF expression in monocytic U937 cells inhibited the
differentiation and re-expressed in monocytes where it plays a ability of vitamin D3 (VD3) to stimulate monocytic differentia-
role in regulating monocytic differentiation through direct tion of these cells while leaving differentiation in response to
interaction with the vitamin D3 receptor.86 PLZF expression retinoic acid intact. 86 Similarly enforced PLZF expression
increases through megakaryocytic development and overexpression inhibited VD3-mediated differentiation in HL60 cells.95 The
of PLZF in an erythro-megakaryocytic cell line induced ability of PLZF into inhibit the action of the VDR was associated
megakaryocytic markers and the thrombopoietin receptor with the ability of the BTB domain of PLZF to bind and inhibit
suggesting a role of PLZF in megakaryocytic development.87 This
transcriptional activation by the Vitamin D receptor.
appeared to be mediated through interaction between PLZF and
PLZF can repress cell growth by binding to the promoters
the GATA-1 transcription factor. PLZF may also modulate
and inhibiting the expression of cell cycle regulators such as cyclin
hematopoietic development through interaction with the GATA-2
factor.88 The BTB domain of PLZF mediates interaction with A and c-myc.38,93,94 Enforced expression of cyclinA or c-myc could
the zinc finger domains of GATA-2 and PLZF as well as reverse growth suppression by PLZF, and both of these promoters
PLZF-RARα can block trans-activation by GATA-2, potentially bound PLZF in vivo as demonstrated by chromatin immunopre-
acting to modulate hematopoetic cell division. cipitation, validating them as PLZF targets.94 Furthermore about
During mouse embryogenesis, PLZF is expressed in a segmental 25% of genes identified as regulated by PLZF in a microarray
pattern89 in the nervous system in a pattern reciprocal to the Krox analysis are known or suspected c-myc targets. PLZF regulation
20 zinc finger gene and the Hoxb2 homeobox gene. In of Hox genes96 may also play a role in regulation of cell growth
co-transfection experiments PLZF can repress the Hoxb2 and differentiation. Through a combination of microarray analyses,
promoter through a cognate PLZF binding site as well as an AT bioinformatics and chromatin immunoprecipitation studies we
rich region.70 PLZF expression is also notable in the limb buds plan to more comprehensively identify PLZF targets important
where Hox genes help to guide the development89 and mice in development and differentiation. Currently our working model
lacking PLZF 90 display homeotic limb defects. This phenotype suggests that PLZF, by direct repression of these genes as well as
was associated with a more anterior pattern of expression of some potentially through its action on GATA-2 is associated with
HoxD genes in the developing hind limb bud, suggesting that relative cell quiescence. In addition through interaction with other
PLZF could inhibit the expression of this class of pattern formation transcription factors such as the VDR, PLZF inhibits differentia-
genes. Moreover, Barna et al showed that PLZF directly binds to tion.86 Furthermore the BTB domain of PLZF mediates interac-
and represses HoxD genes. PLZF null mice have not exhibited an tion with that of FAZF72 and additional BTB dimer/dimer inter-
obvious hematopoetic phenotype, nor have they developed actions by PLZF and the PLZF-RARα fusion can be envisioned
leukemia or other tumors. These data do indicate, however, that that may lead to regulation of additional sets of target genes.
The BTB Domain Zinc Finger Proteins 139
Scheduled down-regulation of PLZF during hematopoiesis receptor co-repressors N-COR, SMRT and sin3A as well as
may allow for controlled cell division and differentiation. Hence histone deacetylases.15, 16,18,22,103,104 PLZF co-localizes to the same
PLZF might be considered a tumor suppressor disrupted by the nuclear sub-location as the SMRT co-repressor.104 Dimerization
t(11;17) translocation in APL. The PLZF-RARα fusion proteins by the BTB domain appears to be required for its ability to
resulting from this translocation may act as dominant negative mediate repression however our group performed a detailed
inhibitors of PLZF. In fact, heterodimerization between mutational analysis of the BTB domain of PLZF and found that
PLZF-RARα and PLZF seems preferential to the formation of a charged pocked at the “top” of the domain was necessary for
PLZF homodimeric complexes, in vitro.76 Therefore, high-level repression. These mutations in the isolated BTB domain
expression of PLZF-RARα in t(11;17)(q23;q21) blasts might tethered to GAL4 changes the domain from a repressor to an
sequester PLZF from binding to its natural target genes and/or activator and when the mutations are made in full-length PLZF
bind to limiting quantities of PLZF transcriptional co-factors.97 the protein is inactive for transcriptional repression and growth
Growth factor-mediated signalling may also modulate suppression. This has led us to investigate the charged pocket
PLZF-mediated gene repression. We found that transcriptional and other features of the BTB domain as therapeutic targets.105,106
repression by PLZF was inhibited by co-expression of a mutant, A second repression domain localized more central in the protein
constitutively active form of the tyrosine kinase receptor Flt3. bears no similarity to other transcriptional effects domains and
Flt3 is the most common mutation identified in humanleukemia binds to ETO (Eight-twenty-one oncoprotein). ETO, the fusion
and mediates signalling through the ras-MAP kinase pathway as partner of AML1 in t(8;21) associated M2 leukemia binds to
well as through the JAK-STAT pathway.98 Co-expression of mSin3a and the N-Cor and SMRT co-repressors. In contrast the
constitutively activated ras or stress kinases a well as mutant Flt3 AML1-ETO fusion protein blocks repression by PLZF, in part
blocks transcriptional repression by PLZF.99,100 Signaling by by inhibiting DNA binding by PLZF and by displacement of
active Flt3 leads to the export of the critical PLZF co-repressor PLZF from the nuclear matrix and possibly from the deacetylase
SMRT from the nucleus to the cytoplasm. While signalling structures.107 This finding suggested that PLZF function might
decreases PLZF-SMRT interaction it increases interaction between be disrupted in other forms of leukemia. We found that deletion
SMRT and the p65 subunit of NFκB, which can then shuttle of either the BTB domain or the second repression domain of
SMRT to the cytoplasm (Takahashi and Licht Submitted). This PLZF severely impaired the ability of PLZF to repress transcrip-
may represent a general way in which signalling may influence
tion and to cooperate with ETO or other co-repressors.108 We
multiple transcriptional repressors that utilize the SMRT co-factor.
hence believe that PLZF forms a dimerization/multimerization
A mechanism more specific to PLZF was recently identified,
platform, which may bind to DNA cooperatively and forms a
potentially important in skin development. Upon proteolytic
multi-protein complex with co-repressors and histone deacetylase.
cleavage of the heparin-binding EGF-like growth factor
In accordance with this model histone deacetylase inhibitors block
(ectodomain shedding) the EGF-like ligand to bind to its
repression by PLZF and the absence of PLZF expression was
cognate receptor. Meanwhile a small ~7kD fragment found within
associated with up-regulation of the endogenous HoxD11 gene
the cytoplasm of the cell translocates to the nucleus where it can
bind to PLZF through the N-terminal zinc fingers of PLZF which in mice, an effect associated with hyperacetylation of that
are not critical for sequence specific DNA binding, leading to the promoter.96
export of PLZF from the cell nucleus. This interaction was The nature of the endogenous PLZF repressor complex
verified to occur naturally in keratinocytes. Nuclear export of remains incompletely understood although several other proteins
PLZF was associated with reversal of PLZF-mediated suppres- have been found in complexes with PLZF. For example the PLZF
sion of cell growth and cyclin A expression.78 The phorbol ester interacts with Bmi-1 a member of the Polycomb protein family
TPA stimulates HB-EGF cleavage and treatment of murine skin implicated in chromatin remodelling.96 PLZF which has a
with this agent induces keratinocytes hyperplasia, possibly by predicted molecular weight of 70 kD, binds to DNA in native
inhibiting the ability of PLZF to control cell growth. Hence gels as a high molecular weight complex of nearly 600 kD.68
signalling through receptor tyrosine kinases may, through two Deletion of the BTB domain of PLZF reduces this complex to
different mechanisms defeat PLZF-mediated transcriptional and about 150 kD suggesting that the BTB domain may allow a
growth regulation. PLZF can associate with another cytoplasmic tetramer of PLZF to form on a DNA binding site. In accordance
factor Epsin1, that is associated with clathrin-coated pits.101 with these studies atomic force microscopy experiments of PLZF
Epsin1 has a structure similar to β-catenin a well-known binding to the Hoxd11 promoter concluded that PLZF bound
transcriptional co-factor. As noted below Kaiso, another BTB-zinc to DNA as a dimer, forming loops when allowed to bind to
finger protein interacts with a catenin-like factor, suggesting a naked DNA.96 The BTB multimer may serve as a platform upon
more general interplay between nuclear BTB-zinc finger proteins which the repressor complex can assemble. DNA looping through
and cytoplasmic co-factors.102 PLZF interacts with Epsin1 the formation of PLZF dimer/multimers might allow repression
through the zinc finger motifs as well as sequences N-terminal to by PLZF to be propagated at distant sites.
the zinc fingers and can recruit Epsin1 to the nucleus. The effect of The identity of the other PLZF partners in the 600 kD
Epsin1 on PLZF transcriptional function and nuclear localization complex is uncertain. One protein we did identify in the 150 kD
remains to be defined. complex is the cdc2 protein, which may play a role in the
phosphorylation of PLZF. Phosphorylation of PLZF may be
critical for its function as treatment of PLZF with a phosphatase
Transcriptional Repression by PLZF and PLZF leads to inhibition of DNA binding activity. In contrast the
Partner Proteins RARα-PLZF protein which is devoid of the BTB or other
Two different portions of PLZF can act as repression domains repression domains of the protein binds to DNA as a 40 kD
when tethered to the GAL4 DNA binding domain. The first complex, exactly as predicted from its molecular weight, does not
corresponding to the BTB/POZ domain interacts with nuclear bind to cdc2 and is unaffected by phosphatase treatment. These
140 Zinc Finger Proteins: From Atomic Contact to Cellular Function
data suggest that RARα-PLZF is not subjected to the same mal RXR/RAR heterodimers.117 However, due to the ability of
regulation as wild-type PLZF and is consistent with the notion the PLZF BTB domain to bind the co-repressors in an ATRA
that this leukemia-associated protein is a potent dominant insensitive manner, while the PML-RARα protein of typical APL
negative inhibitor of PLZF. releases these factors in the presence of 10 -6 M ATRA,
Several other proteins were recently found to complex with PLZF-RARα does not.15,16,21,22,104 In addition, ETO can bind
PLZF. VDUP1 also known as TRX2 (thioredoxin binding and augment repression by PLZF-RARα but not PML-RARα
protein 2) is a nuclear and cytoplasmic protein mutated in and could contribute to the retinoid resistance of t(11;17)-APL.107
hyperlipidemic syndromes. PLZF was identified in a 2-hybrid HDAC inhibitors, such as trichostatin A and sodium butyrate,
screen as a VDUP1 partner95 and although overexpressed PLZF allowed cells harboring PLZF-RARα to respond to ATRA,
and VDUP1 can interact it is not yet certain if this is a significant presumably by inactivating the co-repressor complexes bound to
in vivo interaction. However VDUP1 can augment repression by the BTB domain.21,22,104 However, there remains a paradox in
PLZF in reporter gene assays. Since VDUP1 binds thioredoxin is the lack of response of PLZF-RARα associated APL. When APL
intriguing to speculate that it could affect the redox state of PLZF. cells are treated with ATRA the PML-fusion protein degrades
PLZF can also interact with the LIM domain protein DRAL/ which may help account for clinical response. A study with
FHL2, an interaction that required both the BTB domain and leukemic blasts from a t(11;17) patient showed that ATRA,
the second repression domain of PLZF. Co-expression of DRAL although unable to induce clinical response did lead to
potentiated transcriptional repression by PLZF. DRAL could be PLZF-RARα degradation.118 This suggests that the PLZF-RARα
localized in both the nucleus and cytoplasm and serum stimula- may leave a permanent, epigenetic mark on the leukemic cell si-
tion of cells increased nuclear accumulation of DRAL and lencing target genes in a long lasting manner in a way recently
further augmented repression by PLZF. This is in accordance with described for the PML-RARα fusion protein through the recruit-
increasing data that PLZF-mediated repression may be subjected ment of DNA methyl transferases to target genes.119
to control through growth factors. In a recent paper one group Indeed recent evidence indicates that there are significant
found that PLZF could be modified by SUMO (small ubiquitin differences in the genes regulated by PML-RARα and
like molecule). The modification occurred in the second repres- PLZF-RARα. Upon induction of these fusion proteins in the
sion domain of PLZF and curiously appeared to inhibit both DNA myeloid U937 cell line, 57 genes were down regulated in
binding and transcriptional activity of PLZF.109 The conditions common by the two proteins but PLZF-RARα down-regulated
under which PLZF may be modified by SUMO and the roles of 31 genes that were unaffected by PML-RAR and PML-RAR
other potential post-translational modifications such as phospho- exclusively regulated a distinct set of genes.120 This group
rylation and acetylation is currently under study. identified a subset of genes down-regulated by both fusion
proteins, that might explain the similar block in differentiation
The PLZF Fusion Proteins in APL at the promyelocytic stage observed in t(15:17) and
The t(11;17)(q23;q21) fusion yields two reciprocal transcripts t(11;17)-associated APL. However, some of these commonly
encoding PLZF-RARα and RARα-PLZF.61,62 The PLZF-RARα regulated genes were only regulated in the presence of ATRA in
fusion contains the entire N-terminal transcriptional effector PML-RARα-expressing cells. For example, the gene encoding a
region of PLZF and the first two zinc fingers of PLZF. In one tumor necrosis factor receptor (TNFR1) was repressed by both
case, due to an alternative breakpoint in PLZF, three zinc fingers fusions and only re-activated by PML-RARα. In addition ATRA
are included in the fusion. The reciprocal RARα-PLZF transcript treatment of PML-RARα cells led to decreased expression of the
links the ligand independent activation domain of RARα110 to antiapoptotic Bcl2 gene but similar treatment of PLZF-RARα
the last 7 PLZF zinc fingers αPLZF-RARα is an aberrant retinoic cells did not regulate Bcl2. These differences in gene expression,
acid receptor which unlike the wild-type retinoic acid receptor potentially due to the differences in the repressive complexes
can bind to target sites as a homodimer76,80,111,112 α due to the brought to target genes by the two different fusion proteins could
ability of the PLZF BTB domain to mediate homodimerization. help explain the differential sensitivity of the two syndromes to
In combination with RXR, PLZF-RARα forms multiple ATRA and other therapies.
DNA-protein complexes, but the PLZF-RARα/RXR heterodimer Animal models also indicate differences between PLZF-RARα
binds to retinoic acid response elements (RAREs) with higher and PML-RARα. Transgenic animals expressing PLZF-RARα
affinity than PLZF-RARα homodimers.76,80 The PLZF-RARα/ developed a CML-like syndrome rather than APL.21,121 Unlike
RXR heterodimer bound to the RARE less efficiently than PML-RARα transgenic mice, the PLZF-RARα mice did not
wild-type RXR/RARα perhaps due to the ability of the BTB achieve durable remission after ATRA treatment, although leu-
domain to multimerize and preclude efficient DNA binding.71 kemic cells from the mice treated with ATRA differentiated ex
Hence, PLZF-RARα might sequester RXRα, an essential co-factor vivo.21 The relative insensitivity of the disease correlates with the
for RARα function. Furthermore, PLZF-RAR/RXR heteromers markedly decreased ability of PLZF-RARα to transactivate due
also bound to non-classical RAREs, potentially contributing to binding of co-repressors. This idea was confirmed by the result
toleukemogenesis by misdirecting the chimeric protein to novel that the combination of a histone deacetylase inhibitor and ATRA
genes.111,113 PLZF-RARα is relatively weak trans-activator76,80,114 synergistically induced differentiation of these leukemic cells.67
and is a dominant negative inhibitor of wild-type RARα. In ad- Though explaining the poor response of this disease to ATRA,
dition PLZF-RARα may contribute to theleukemic phenotype these animal models do not explain the poor response of some
by blocking vitamin D3-induced hematopoietic differentiation t(11;17)(q23;q21) patients to chemotherapy and the complete
by sequestering the RXR co-factor.115 Repression by PLZF-RARα resistance of fresh APL cells to high doses of ATRA in vitro.62,122
is predominantly due to its high affinity for the HDAC containing These findings implicate the reciprocal RARα-PLZF protein in
co-repressor complexes.15,16,21,22,103,104,116 This could be due in the aggressive nature of PLZF-RARα associated APL.
part to the presence of four docking sites for co-repressors on The reciprocal transcript encoding RARα-PLZF was expressed
each oncoprotein homodimer rather than two sites found in nor- in nearly all of the t(11;17)(q23;q21) patients.62,123 The three
The BTB Domain Zinc Finger Proteins 141
zinc fingers of RARα-PLZF can bind to the same site as full-length We performed extensive mutagenesis of the PLZF and Bcl-6
PLZF. 36,124 While PLZF represses gene transcription, BTB domain and tested these products in both the full length
RARα-PLZF activates it.36 Whereas PLZF is a growth suppres- and isolated BTB context in vitro and in vivo.6,40,106,137 These
sor, RARα-PLZF activates expression of cyclin A2 and enhances results allowed us to assign functions to its various structural
cell growth.38 Hence t(11;17)(q23;q21) may be an ATRA and motifs. For example, conserved residues in the dimer interface
chemotherapy resistant disease due to the presence of two are required for proper folding, indicating that dimerization is an
oncogenes. PLZF-RARα blocks the differentiation inducing architectural function of BTB domains that is invariant throughout
action of ATRA, while RARα-PLZF may activate cell cycle evolution. While the oligomerization surface was required for high
regulators, blocking the anti-proliferative effects of ATRA. In molecular weight complex formation by PLZF, which is known
accordance with this idea, mice engineered to express the to form such oligomers, there was no evidence that Bcl-6 forms
RARα-PLZF protein develop a myeloproliferative syndrome and similar oligomers through this motif. Therefore, it is possible that
splenomegaly.125 When PLZF-RARα mice were crossed with the Bcl-6 hydrophobic face is involved in other protein-protein
RARα-PLZF mice there was a phenotypic change in the resulting interactions.6,40,106,137 As in the case of PLZF, mutation of the
leukemia with the appearance of more primitive, more prolifera- charged pocket of Bcl6 inhibited both transcriptional repression
tive cells that were highly resistant to apoptosis and morphologi- and interaction with the co-repressors N-Cor and SMRT.
cally more similar to human APL.125 Strikingly, an identical Comparing the PLZF, Bcl6 and PLZF-related FAZF BTB
phenotype was seen when PLZF-RARα transgenic mice were bred domains a single amino acid difference was noted. When this
with PLZF null mice. This information is in accordance with single residue of these three BTB domains was altered to that of
data suggesting that RARα-PLZF is a dominant negative form of one of the other two proteins, the strength of transcriptional
PLZF and a growth activator. It is possible that through aberrant repression and interaction with N-Cor or SMRT was modulated.
expression of PLZF targets such as HOX genes, cyclins, c-myc, This supports the notion that the charged pocket of the BTB is
and other cell cycle regulators, the presence of both fusion genes an important contact site for co-repressors.
lead to a more aggressive and undifferentiated form of leukemia. Further analysis of the Bcl6 BTB domain indicated that the
Though t(11;17)-APL is a very rare disease, the comparison charged pocket could not completely explain repression by the
of contrast of this illness with the more typical form of APL BTB domain. The Bcl6 BTB domain directly bound N-Cor and
solidified the notion of aberrant transcriptional repression as a SMRT much more tightly than PLZF and a small region of N-Cor
root cause of leukemia. In addition studies from our group and or SMRT mediated this direct binding. The co-crystal structure
others have implicated PLZF as an important negative regulator of the Bcl6 BTB domain with a segment of SMRT (aa 1353-1526)
of cell growth with roles in differentiation, development and showed that this portion of SMRT tightly bound to the lateral
homeostasis. face of the BTB domain.138 Subtle mutations in either the SMRT
peptide or in the BCL6 BTB domain abrogated SMRT/Bcl6
The Role of Bcl-6 Protein in Lymphoma interaction and transcriptional repression.
The Bcl-6 gene was first cloned by virtue of its frequent
mutations in B-cell lymphomas, where overexpression or lack of Bcl-6 As a Transcriptional Repressor: Other
downregulation of Bcl-6 occurs when translocated to a
Functional Domains of the Bcl-6 Protein
heterologous promoter or mutations introduced by somatic
The other major structural motif of Bcl-6 is the array of six
hypermutation damage Bcl-6 regulatory elements.126-128 Such
tumors represent approximately 40% of human B-cell lympho- C2H2 zinc fingers at the C-terminal end of the protein, which
mas and are believed to be biologically dependent on Bcl-6 bind to a specific DNA consensus sequence (TTCCT(A/
expression.129,130 Bcl-6 was subsequently shown to also play a criti- C)GAA). 139 The zinc fingers also mediate protein-protein
cal role in the development of the immune system.131-133 The BTB interactions. For example, the zinc fingers interact with class II
domain of Bcl-6 was shown to interact with the SMRT/N-CoR HDACs as well as with the zinc fingers of other BTB proteins
corepressors and histone deacetylases (HDACs).15,16,20,134-136 As in such as PLZF and LRF.140-143 The zinc finger domain also
the case of PLZF there is a second repression domain in the protein interacted with the c-jun transcription factor, leading to negative
localized to the mid-portion of the protein. regulation of AP1 transcriptional activation by Bcl-6.144 Recently,
we found that the ETO protein which was first cloned as a
member of the acute myeloid leukemia (8;21) translocation
Bcl-6 As a Transcriptional Repressor: fusion product, also interacts with Bcl-6 through the zinc
Structure-function Studies of the BTB Domain fingers.145 We had previously demonstrated that ETO is a novel
Along the lines of our studies with PLZF, we performed in corepressor much like N-CoR and SMRT, which interacts with
depth structure-function analysis of the Bcl-6 BTB transcriptional repressors and enhances transcriptional repression
domain.6,40,106,137 The PLZF and Bcl-6 dimer crystals revealed by amplifying histone deacetylation.142,146,147 We found that ETO
that the two structures are very similar. BTB monomers interact and Bcl-6 are co-expressed in normal and malignant lymphoid
via an extensive interface to form obligate dimers. The most highly tissue, where they interact and colocalize in nuclear speckles. ETO
conserved surface feature is a charged pocket. The charged pocket binds to the fourth zinc finger of Bcl-6, enhances Bcl-6 repres-
is structurally consistent with an interaction site for peptides from sion of artificial and endogenous genes in an HDAC-dependent
putative partner proteins. However, there is a region of confor- manner, and forms a complex with Bcl-6 on the promoters of its
mational variability between the PLZF and Bcl-6 pockets endogenous target genes in B-cell lymphoma cells.145 Therefore,
conferred by an α3/β4 loop, where the BCL-6 BTB domain has ETO is a bona fide corepressor for Bcl-6 and is the only corepres-
a unique 4 amino acid insertion relative to PLZF that might sor conclusively demonstrated to bind outside of the BTB
obstruct access to this motif in the case of Bcl-6. domain in the context of physiological transcriptional repression.
142 Zinc Finger Proteins: From Atomic Contact to Cellular Function
Although the zinc finger domain of Bcl-6 was often used as a once again it is unknown whether the BTB contributes architec-
dominant negative construct to antagonize Bcl-6 repression, tural or corepressor function.144
multiple lines of evidence indicate that it may also function as an Bcl-6 was recently shown to induce down regulation of the
autonomous repression motif albeit with much less potency than GATA-3 inducer of Th2 differentiation. This is consistent with
the BTB domain. Although Seyfert et al found that the zinc the increased levels of Th2 cells in Bcl-6 null animals.
finger domain did not repress transcription of a Bcl-6 binding Surprisingly, only GATA-3 protein levels were reduced, while the
site containing reporter,148 the same group also reported that the mRNA level didn’t change.161 This implicates Bcl-6 in the
zinc finger domain functioned as a weak dominant negative when regulation of protein stability, an attractive idea in light of
expressed in Bcl-6 positive Raji cells, indicating that it either does emerging evidence of the role of the BTB domain in
not repress or that it represses less efficiently than full-length ubiquitylation. Finally, Bcl-6 was found in the proximity of DNA
Bcl-6. 149 In contrast, a GAL4-zinc finger fusion construct replicative foci raising the possibility of a role in DNA replication
inhibited transcriptional activation of a reporter gene containing as well.162 Therefore, Bcl-6 is a multifunctional regulatory factor
LexA and GAL4 sites when it was activated by a LexA-VP16 that represses expression of target genes and proteins through
fusion while in reciprocal experiments, deletion of the zinc several different mechanisms.
fingers caused a loss in potency of Bcl-6 repression.141 A zinc
finger only construct was able to partially mediate Bcl-6 induced Bcl-6 and the Immune System
growth and cell cycle arrest in U2OS cells although the BTB
Bcl-6 is a critical regulator of specific components of the
domain was required to induce apoptosis.150 Finally, the Bcl-6
immune system and its hierarchy in this process is underlined by
zinc fingers alone were able to partially inhibit AP1 transactivation,
the severity of the Bcl-6 null phenotype and by the fact that
although full length Bcl-6 as much more powerful.144 These
de-regulation of Bcl-6 expression causes cancer.130 Accordingly,
experiments illustrate the difficulty of determining the functional
expression of Bcl-6 is tightly regulated and limited to specific
contribution of the zinc finger domain to transcriptional repres-
subsets of immune system cells.
sion, which is unlikely to be resolved by studying deletion
mutants of Bcl-6. Taken together these results seem to indicate
that the zinc finger and BTB repression domains both contribute B-Cells
to gene silencing by Bcl-6, although the gene loci they regulate In the B-cell lineage, Bcl-6 is expressed solely in
may not fully overlap. Finally, the central region of Bcl-6 may germinal center (GC) cells. 163,164 Upon T-cell dependent
also have autonomous repression activity by virtue of recruiting antigen challenge, subsets of activated B-cells (pre-GC cells)
HDACs and possibly the Sin3A corepressor.151 enter lymphoid follicles and up-regulate Bcl-6, which induces
cells to proliferative and survive and at the same time to, resist
Transcriptional Mechanisms of Action of Bcl-6 further differentiation.131-133,155,165 Simultaneously, induction of
the somatic hypermutation machinery leads to affinity maturation
Bcl-6 mediates transcriptional repression of a number of
of the immunoglobulin genes. When this process is complete Bcl-6
endogenous target genes that include surface antigens (for
is downregulated.131-133,155,165,166 B-cell clones with productive
example CD23b, CD44 and CD69), cell cycle mediators (for
immunoglobulin rearrangements differentiate further into
example Cyclin D2, and p27KIP), cytokines, chemokines and their
receptors (for example IL2Rα chain, IL5, IL18, MCP-1, MCP-3, post-GC memory or plasma cells while clones with
MIP1α, and CXCR4), and mediators of B-Cell differentiation non-productive rearrangements undergo apoptosis.155,167
such as blimp-1.132,149,152-156 In most cases Bcl-6 binds directly Regulation of Bcl-6 expression at the mRNA level is poorly
to its consensus binding site located in its target loci and understood, although Bcl-6 itself can repress transcription of its
mediates repression in large part through the BTB domain as own message.168 From the protein standpoint, MAP kinase leads
discussed above. However, Bcl-6 also represses downstream to phosphorylation of Bcl-6 via the ERK1/2 kinases with subse-
transcriptional activation by the IL-2, IL-4 and CD40 signal quent ubiquitylation and proteolysis.166,169 This process is under
transduction pathways that mediate lymphocyte differentia- the control of the B-cell receptor in pre-GC cells and possibly
tion.132,155,157-160 A subset of DNA binding sites for mediators under the control of CD40 in post-GC cells.166
such as STAT6, STAT3 and IRF-4 are similar to the Bcl-6 The GC specific role of Bcl-6 is underscored by the pheno-
consensus sequence.132,155,157,160 Bcl-6 can bind to the same sites type of Bcl-6 knockout mice. Such animals are born at the
as these transactivators and repress expression of their target genes. expected frequencies and with all the lymphoid organs fully formed
For example, Bcl-6 inhibits IgE class switching by repressing IL-4 and containing the expected primary lymphoid follicles.131-133
induced activation of the IgE promoter through an overlapping All lymphoid cell subsets are present in normal numbers, although
DNA element with STAT6.160 Consistent with this observation, baseline antibody levels are slightly low.131,132 However, Bcl-6 null
Bcl-6-/- mice demonstrated increased IgE class switching and mice exhibited a complete failure to generate GCs in all
tissue infiltration with IgE+ B-lymphocytes, while Bcl-6-/- lymphoid tissues when challenged by T-cell dependent antigens.
STAT6-/- mice had neither of these features.160 Bcl-6 represses T-cell dependent antibody generation was deficient, with
activation by these transcriptional messengers in a BTB-dependent antibody levels of all classes (Except IgM) lower than control.131-133
manner, although it is not known whether the BTB domain is Even more remarkably, these antibodies had not undergone the
required for an architectural or corepressor recruiting function. (GC-dependent) process of affinity maturation as demonstrated
A third mechanism of Bcl-6 transcriptional repression was by the absence of high affinity binding in antigen binding
recently described in which Bcl-6 acts as a corepressor for AP-1 assays.131 Not surprisingly, memory B cells in these animals
by virtue of binding to c-Jun through the C2H2 zinc finger express only low affinity antibodies after booster shots of T-cell
domain. In this way Bcl-6 can behave as a corepressor by dependent antigens, indicating that these cells had not under-
repressing AP-1 target genes such as p27kip through c-Jun. Full gone somatic hypermutation as they did not transit through GCs
repression also required the BTB domain in this case, although during the course of their terminal differentiation.167
The BTB Domain Zinc Finger Proteins 143
The defect in GC formation was entirely B-cell autonomous IL4 animals developed progressive infections by Leishmania
as demonstrated by transplantation experiments in RAG1 major, since Th2 reactions are inhibitory to macrophages.159
deficient mice in which the immune system was reconstituted Therefore, Bcl-6 must normally inhibit Th2 formation through
with normal T cells plus Bcl-6 deficient B cells or vice versa.133 an alternate pathway independent of IL4/STAT6 signaling. One
This was confirmed in experiments were chimeric mice were such pathway is mediated through GATA-3 activation of Th2
generated by injecting Bcl-6-/- ES cells into RAG1, IgM or TCRβδ cytokines in the absence of STAT6.173 Interestingly, Bcl-6 single
null blastocysts.153 In the resulting RAG1-/- mice all B and T cells and double knock-out with STAT6 mice have high levels of
derive from the Bcl-6-/- cells, while in IgM-/- only the B-cells are GATA-3.161 Re-introduction leads to downregulation of GATA-3
exclusively Bcl-6-/-, and in the TCRβδ-/- cells only T cells are and downregulation of Th2 cytokine secretion (IL4 & IL10).161
Bcl-6-/-.153 In concordance with the transplantation experiment, As mentioned above, the GATA-3 gene did not have Bcl-6 binding
the Bcl-6-/- RAG1 and IgM null mice were unable to form GCs, sites in its promoter and Bcl-6 seemed to downregulate only the
while GCs were able to form in the Bcl-6-/- TCRβδ chimeric protein but not the mRNA levels of this protein, indicating a
mice.153 potentially novel regulatory mechanism.161 Finally, these results
Only a few of the Bcl-6 target genes relevant for its biological indicate that loss of Bcl-6 results in both humoral and cellular
activity in the GC were identified so far. For example, Bcl-6 immunodeficiency.
directly represses transcription of the Blimp1 transcriptional
repressor gene, which in turn represses expression of genes Macrophages
involved in GC processes such as class switching (Ku70), somatic A role for Bcl-6 regulation of macrophage
hypermutation (AID), proliferation (myc) and B-cell signaling function was suggested by the fact that Bcl-6 was shown to be
(STAT6).149,170 As noted, Bcl-6 binding sites overlap in several expressed in monocytic cells174 and the finding that RAG1 null
loci with binding sties for the STAT3, STAT6 and IRF4 mice transplanted with Bcl-6-/- bone marrow or RAG-/- Bcl-6-/-
transcriptional activators as shown by EMSA and DNA chimeric mice failed to develop eosinophilic infiltrates, indicat-
footprinting.132,155,157-160 Bcl-6 also physically interacts with these ing that full blown Th2 inflammation requires the participation
transcriptional activators and represses promoters with overlap- of additional non-lymphoid cells.153 A Bcl-6-/- TCRα-/- chimeric
ping Bcl-6 and STAT sites such as the CD23b promoter.155,157 mouse specifically unable to generate Th2 cells was used to study
In this way GC cells expressing Bcl-6 could resist differentiation macrophage function.153 Interestingly several chemokines involved
signaling from the IL2, IL4 and CD40 pathways. in inflammatory reactions and eosinophil chemotaxis were
upregulated and some of these were direct Bcl-6 target genes.153
T-Cells Re-introduction of Bcl-6 suppressed expression of these genes.
Bcl-6 plays important regulatory roles in T cells, and is nor- Therefore, macrophage products such as MCP-1 contribute to Th2
mally expressed in germinal center T cells and in a subset of CD30 inflammation possibly by promoting antigen presenting cell stimu-
positive T cells present in interfollicular regions of lymph lation of Th2 and though inhibition of the Th1
nodes.163,171 Bcl-6 is also expressed in cortical and medullary T response by virtue of inhibiting Il-12 production. The triggering
cells of the thymus.172 Bcl-6 null animals appear similar to event that leads to this inflammatory picture remains obscure,
control littermates when born, but rapidly become sickly and although one of the Bcl-6 null mice models had a high incidence
growth retarded.131-133 At least 50% of these mice die by 5 weeks of bacterial infections even though they were housed in a barrier
of age due in large part to an inflammatory infiltrate of facility.
monocytes/macrophages and eosinophils, affecting the myocar-
dium, lungs, gut, skin and other mucosas.131-133 Taken together
with observed increases in IgE expressing B-cells, these findings
Other Tissues
suggested that a powerful Th2-type reaction was occurring in these Bcl-6 is highly expressed in terminally differentiated tissues
animals leading to their demise through systemic inflammatory such as the heart, skin and skeletal muscle.175-177 For example,
reactions. Accordingly, cells isolated from Bcl-6-/- lymph nodes Bcl-6 is upregulated in cardiac myocytes within two weeks of
expressed high levels of (Th2 related) IL4, IL5 and IL13 but low birth. Bcl-6 null mice develop subtle signs of cardiac degenera-
levels of (Th1 related) IFNγ at mRNA and protein levels.131-133 tion within four weeks of birth and apparently prior to infiltra-
These results implicated Bcl-6 in the skewing of T-cells towards tion by the Th2 inflammatory disease (as detected by
the Th1 lineage or in the inhibition of differentiation into Th2 electron microscopy).175 In these tissues Bcl-6 might be required
cells. Since both IL4 and STAT6 null mice lack Th2 cells and to maintain a quiescent, differentiated state. In contrast, Bcl-6 is
given that Bcl-6 is an inhibitor of IL4-STAT6 signaling, it was also highly expressed in spermatocytes undergoing meiosis, which
anticipated that generating mice null for Bcl-6 and deficient of like the somatic hypermutation and class switching of GC B-cells,
either IL4 or STAT6 might resolve the Th2 skewing. 158 leads to DNA breakage in the setting of rapidly proliferating
However, both Bcl-6-/- STAT6-/- and Bcl-6-/- IL4-/- mice cells.178 Interestingly, Bcl-6 protects spermatocytes from apoptosis
exhibited Th2 skewing when mice were challenged with TNP/ and Bcl-6 null animals are infertile.158,178 Therefore, Bcl-6
KHL antigen/hapten. The defect was not due to loss of Th1 physiological effects may vary according to cell type and available
differentiation since Bcl-6 single or double knock-out mice were partner proteins.
able to generate Th1 cells in response to IL12.158
The immunological significance of this phenomenon was Bcl-6 and B-Cell Lymphomas
illustrated by the following experiment. The 129sv and C57BL/ In spite of the frequent involvement of the Bcl-6 locus in B-cell
6 strains from which the Bcl-6 null animals were derived are lymphomas, no definitive gain or loss of function study has
normally able to mount a Th1 immune reaction that heals demonstrated that deregulated Bcl-6 expression causes lymphoma.
infections by Leishmania major through macrophage stimulation However, several lines of evidence are consistent with potential
by IFNγ.159 In contrast, both single and double knock-out Bcl-6/ oncogenic effects of Bcl-6.
144 Zinc Finger Proteins: From Atomic Contact to Cellular Function
Most studies of Bcl-6 functions in primary cells and lymphoma Like many BTB domains both the HIC-1 and γFBP-B (an
cells indicate that Bcl-6 protects cells from apoptosis. For example, avian homologue) BTB domains can function as autonomous
Bcl-6 deficiency induces apoptosis in spermatocytes, skeletal transcriptional repression domains. However, they do not
muscle cells, and causes degenerative changes in cardiac interact with components of the histone deacetylase complex in
myocytes.175,178,179 In B-cell lymphoma cell lines, a dominant vitro or in vivo. Furthermore, transcriptional repression by HIC-1
negative form of Bcl-6 that releases repression of Bcl-6 target genes was not alleviated with the HDAC inhibitor trichostatin A,
induced apoptosis as well.149 Furthermore, Bcl-6 may prevent further substantiating the idea of an alternate mechanism of
apoptosis by directly repressing the PDCD2 pro-apoptotic transcriptional repression by these particular BTB domains.55 This
gene.180 Most recently, Bcl-6 was discovered in an unbiased screen was confirmed when it was determined that HIC-1 could complex
of cDNA libraries to bypass p53 mediated cellular senescence in with the co-repressor CtBP through a motif more internal within
mouse embryonal fibroblasts.181 When Bcl-6 was co-expressed the protein.185 The BTB domain did not directly interact with
with an activated allele of Ras, these cells became fully transformed CtBP, yet deletion of the BTB domain of Hic-1 severely impaired
and formed colonies in soft agar and tumors in nude mice.181 transcriptional repression by the protein. This and evidence from
Induction of cyclinD1 was required for these effects. the study of PLZF and Bcl6 indicates that dimerization/
Furthermore, primary human tonsilar B cells infected with a Bcl-6 multimerization through the BTB domain is a critical function
expressing retrovirus were able to grow indefinitely in culture in of the motif in transcriptional repressors, but the BTB domain
the presence of CD40L, IL2 and IL4.181 However, Bcl-6 can also itself need not be required for direct interaction with co-repressor
induce apoptosis, although this usually occurs in the context of complexes.
forced overexpression studies in non-physiologically relevant cell
lines such as CV1 and U2OS.150,182 Kaiso
The ability of Bcl-6 to block differentiation induced by IL2, The Kaiso BTB domain protein was identified in a two-hybrid
IL4 and CD40L pathways explains in part how Bcl-6 mediates screen as an interactor with p120Ctn a protein similar to catenin
the B-cell GC life cycle.155 Accordingly, expression of Bcl-6 in with multiple armadillo repeats. Kaiso interaction with p120Ctn
WIL2 and SUDHL5 cells blocked differentiation while in Raji was specific and no interaction with E-cadherin, β-catenin or
cells, a dominant negative Bcl-6 zinc finger construct triggered α-catenin could be detected. Kaiso is ubiquitously expressed and
incomplete plasmacytic differentiation.149 contains a N-terminal BTB domain and three C-terminal zinc
Since Bcl-6 establishes a proliferative cohort of cell in finger motifs.102 Kaiso/p120Ctn interaction was mediated by a
germinal centers, it would be expected to mediate cell cycle region of the protein encompoassing the zinc finger motifs and
progression. Accordingly, in Raji cells, dominant negative Bcl-6 including a region immediately N-terminal to these motifs.102,186
caused cell cycle arrest in G1 and in BCL-1 cells, expression of Kaiso was localized mostly to the nucleus but in some cells
Bcl-6 overcomes the G2/M phase cell cycle arrest mediated by cytoplasmic localization was noted suggesting that the Kaiso/
IL-2 and IL-5.149,155 Bcl-6 regulates a number of cell cycle p120Ctn interaction may mediate a nuclear/cytoplasmic shuttling.
mediators and inhibitors including cyclin D2 gene, cyclin A2, Kaiso was subsequently identified as part of a methyl CpG DNA
p27 and cyclin D1.149,154,181 Therefore, it is possible that Bcl-6 binding activity known as MeCP1. Kaiso itself bound to DNA
may differentially affect cell cycle progression according the specifically in a methylation dependent manner.186,187 Zinc
milieu of the cell type. Thus, while Bcl-6 may mediate cell cycle fingers 2 and 3 of the protein were sufficient for specific DNA
progression in GC B cells, it might inhibit cell cycle in terminally binding to two symmetrically methylated CpG dinucleotides in
differentiated cells such as myocytes and keratinocytes. the core sequence. Kaiso could also recognize with high specificity a
Taken together, these data suggest that in germinal center non-methylated DNA sequence implying a role for the protein
B-cells, de-regulated expression of Bcl-6 causes cells to survive in in epigenetic silencing as well as broader transcriptional
the face of DNA damage and evade p53 mediated senescence, as processes.186 Kaiso repressed transcription of a reporter gene
well as to transit the cell cycle, block differentiation and undergo containing a consensus Kaiso binding sequence in a cell line
continuous mutagenesis through somatic hypermutation charac- otherwise defective for methylation-dependent silencing due to
teristic of germinal center B cells. deletion of another methyl-DNA binding protein. Repression by
Kaiso required the BTB domain as well as the zinc fingers of
Kaiso.
Other BTB/Zinc Finger Proteins The function of Kaiso is conserved in evolution as Xenopus
Kaiso, tethered to the Gal4-DNA binding domain could represses
HIC-1 (Hypermethyalted in Cancer-1) a reporter gene in frog oocytes.188 This information suggested
Recruitment of a histone deacetylase complexes not a universal that Kaiso might play a part in the silencing of methylated genes
property of BTB domain containing transcriptional repressors. by recruitment of co-repressors complexes through the BTB and
The HIC-1 gene encodes a protein with five C2H2 zinc fingers other domains of the protein. Indeed this notion was confirmed
and an N-terminal BTB domain. 183 Transcription of this when Kaiso was found as a component of a high molecular weight
ubiquitously expressed gene is activated by p53, but HIC-1 is complex containing the co-repressor N-Cor (but not the related
not expressed in several common types of human cancers due to SMRT protein) and HDAC3.189 As in the case of PLZF and
hypermethylation of its promoter. These facts make HIC-1 a Bcl6 the BTB domain mediated interaction of Kaiso with N-Cor
strong candidate for a tumour suppressor gene.183 In accordance and a histone deacetylase inhibitor reversed repression by the Kaiso
with this notion it was recently shown that a heterozygous BTB domain. Chromatin immunoprecipitation demonstrated
disruption of HIC-1 in mice led to a variety of sex-dependent that N-Cor and Kaiso were present on a methylated promoter
tumors arising after a year. In most tumor tissue, it was found at and were released from the gene when cells were treated with the
least one promoter of the wild-type HIC-1 allele was de-methylating agent 5-aza-2-deoxycytidine. The presence of
hypermethylated.184 Kaiso on an endogenous promoter was associated with
The BTB Domain Zinc Finger Proteins 145
de-acetylation histone 3 lysine tail residue 9 and methylation of domain mediates p300 interaction. Myc-Miz-1 interaction may
this residue, a set of changes associated with gene repression.190 also release a latent repression activity of Miz-1, possibly involv-
Knockdown of Kaiso by siRNA led to reactivation of a ing the BTB domain but to date no mechanism of active repres-
methylated gene and loss of co-repressors from the gene. These sion by the Miz-1 has been elucidated. The ability of TGFβ to
studies indicate a major role of Kaiso in epigenetic gene inhibit cell growth was related to its ability to decrease c-myc
silencing. However the importance of kaiso in regulation of other levels and inhibit myc-Miz-1-mediated repression of p15INK4b.194
genes through non-methylated DNA sequences and the role of In addition activation and nuclear migration of SMAD3 and
p120Ctn in Kaiso function in vivo and gene silencing is as yet SMAD4 proteins allowed these transcription actors to complex
undetermined. In turn these studies raise the idea that other BTB/ with Miz-1 through its zinc finger motifs and further stimulate
zinc finger proteins such as PLZF and Bcl6 may also bind to the p15INK4b promoter. The myc-Miz interaction is also an
methylated DNA and play a role in both adjustment of gene important regulator of p53 action. In the presence of myc,
expression in a dynamic manner as well as in more permanent repression of the p21 promoter through Miz-1 prevents the
inhibition of methylated genes. ability of p53 to inhibit the cell cycle and favors the pro-apoptotic
actions of p53.195
Miz-1 The Miz-1 BTB domain does not appear to play a direct role
The Miz-1 (myc interacting zinc finger-1) BTB protein has in transcriptional activation but does mediate important
come under increasing study, as it appears to be an important protein-protein interactions. An interacting partner for this BTB
mediator of the function of the c-myc oncoprotein. Miz-1 was domain, topoisomerase IIβ binding protein (TOBP1), a BRCT
identified in a two hybrid screen for proteins that interact with domain protein implicated in DNA damage response was recently
c-myc.191 Miz-1, like other BTB/zinc finger proteins, contains isolated.197 Intriguing in light of recent data implicating the BTB
an N-terminal BTB domain and has 12 zinc finger motifs domain in protein degradation, the other interacting protein
localized in the middle of the protein and a lone zinc finger motif isolated with Miz-1 was an E3 ligase. The TOPBP1 protein did
near the C-terminus of the protein. Two regions flanking the 12 not interact with the Bcl6 BTB domain, which we20,138 and
zinc fingers mediated interaction with the helix-loop-helix motif others showed to be a potent transcriptional repressor domain.
of c-myc as well as the related N-myc proteins. Miz-1 bound The TOPBP1 protein inhibited activation by Miz-1 but was found
specifically to a site within the adenovirus major late promoter to be down-regulated in response to UV irradiation, leading to
and unlike the other BTB/zinc finger proteins described activated de-repression of Miz-1 target promoters p21 waf1/Cip1 and
this promoter. Co-expression of myc impaired transcriptional p15INK4b.197 Host cell factor (HCF) a protein previously impli-
activation by Miz-1. Curiously, although the BTB domain did cated in viral gene regulation also interacts with Miz-1 to inhibit
not interact with c-myc, the BTB domain was required for the its ability to activate transcription. This interaction can occur
ability of c-myc to inhibit trans-activation by Miz-1. Unlike other independently of the BTB domain but HCF can also interact in
BTB domain proteins like PLZF, Miz-1 itself was not localized to vitro with the Miz-1 BTB domain.198 Like c-myc, HCF blocks
discrete nuclear foci and was readily extracted from the cell. activation by Miz-1 by interfering with Miz-1/p300 interaction.
However, when co-expressed with c-myc, Miz-1 did localize to As noted with many of the other BTB/ zinc finger proteins
punctate domains and appears to make the Miz-1 protein described above, a common theme appears to be emerging in
insoluble in the cell and extractable from the nucleus only under that many of these proteins may shuttle to the cytoplasm and
high-salt conditions. In part this appears to be due to the ability interact with cytoplasmic proteins. Miz-1 was originally found to
of myc to donate a functional nuclear localization signal to Miz-1. co-purify with tubulin191 and this was confirmed in a chemical
Whether the Miz-1 nuclear domains overlap with those of other screen for activators of the gene encoding the low density
BTB/zinc finger proteins or co-localize with co-repressors and receptor.199 Addition of a small chemical to cells released Miz-1
histone deacetylases is unknown. from microtubules and allows Miz-1 to bind and stimulate
Miz-1 is a growth suppressor and co-expression of c-myc transcription of this gene. Miz-1 bound to tubulin through its
relieves its ability to suppress growth. Deletion of the BTB most C-terminal zinc finger motifs, while the BTB domain of
domain only partially inhibited trans-activation and growth Miz-1 was required for the protein to migrate to the nucleus after
suppression by Miz-1 implying that domains within Miz-1 other chemical treatment. The physiological role of Miz-1 sequestra-
than the BTB domain were also critical for its transcriptional tion and what natural stimuli encourage its nuclear import are
function. However the deletion of the BTB domain made growth not yet known.
suppression by Miz-1 insensitive to co-expression of c-myc. Miz-1
can augment transcriptional activation by c-Myc and its obligate References
partner Max. This correlates with the ability of Miz-1 to stabilize 1. Koonin EV, Senkevich TG, Chernos VI. A family of DNA virus
genes that consists of fused portions of unrelated cellular genes.
c-Myc and inhibit its proteosome-mediated degradation.192 Trends Biochem Sci 1992; 17:213-214.
It has become clear that transcriptional repression by c-myc is 2. Godt D, Couderc JL, Cramton SE et al. Pattern formation in the
essential for its ability to transform cells. Miz-1 is a major limbs of Drosophila: bric a brac is expressed in both a gradient and
mediator of myc-mediated repression. Miz-1 binds to the a wave-like pattern and is required for specification and proper
initiator region of both the promoters of the cell cycle regulators segmentation of the tarsus. Development 1993; 119:799-812.
p21waf1/Cip1 and p15INK4b to activate expression of these genes, 3. Numoto M, Niwa O, Kaplan J et al. Transcriptional repressor ZF5
identifies a new conserved domain in zinc finger proteins. Nucleic
accounting for its activity as a growth suppressor. c-myc activates
Acids Res 1993; 21:3767-3775.
cell growth and inhibits cell senescence least in part due to its 4. Zollman S, Godt D, Prive GG et al. The BTB domain, found
ability to bind to Miz-1 on these promoters and lead to the primarily in zinc finger proteins, defines an evolutionarily conserved
repression of transcription of these cell cycle regulators.193-196 Myc family that includes several developmentally regulated genes in
inhibits Miz-1 mediated activation by preventing p300 binding Drosophila. Proc Natl Acad Sci USA 1994; 91:10717-10721.
to Miz-1.193 A C-terminal region of Miz-1 and not the BTB
146 Zinc Finger Proteins: From Atomic Contact to Cellular Function
5. Bardwell VJ, Treisman R. The POZ domain: a conserved 28. Aravind L, Koonin EV. Fold Prediction and Evolutionary Analysis
protein-protein interaction motif. Genes Dev 1994; 8:1664-1677. of the POZ Domain: Structural and Evolutionary Relationship with
6. Ahmad KF, Engel CK, Prive GG. Crystal structure of the BTB the Potassium Channel Tetramerization Domain. J Mol Biol 1999;
domain from PLZF. Proc Natl Acad Sci USA 1998; 95:12123-12128. 285:1353-1361.
7. Li X, Lopez-Guisa JM, Ninan N et al. Overexpression, purification, 29. Uren AG, Vaux DL. TRAF proteins and meprins share a conserved
characterization, and crystallization of the BTB/POZ domain from domain. Trends Biochem Sci 1996; 21:244-245.
the PLZF oncoprotein. J Biol Chem 1997; 272:27324-27329. 30. Xue F, Cooley L. Kelch encodes a component of intercellular bridges
8. Li X, Peng H, Schultz DC et al. Structure-function studies of the in Drosophila egg chambers. Cell 1993; 72:681-693.
BTB/POZ transcriptional repression domain from the promyelocytic 31. Kreusch A, Pfaffinger PJ, Stevens CF et al. Crystal structure of the
leukemia zinc finger oncoprotein. Cancer Res 1999; 59:5275-5282. tetramerization domain of the Shaker potassium channel. Nature
9. Davies JM, Hawe N, Kabarowski J et al. Novel BTB/POZ domain 1998; 392:945-948.
zinc-finger protein, LRF, is a potential target of the LAZ-3/BCL-6 32. Ramos S, Khademi F, Somesh BP et al. Genomic organization and
oncogene. Oncogene 1999; 18:365-375. expression profile of the small GTPases of the RhoBTB family in
10. Daniel JM, Reynolds AB. The catenin p120(ctn) interacts with Kaiso, human and mouse. Gene 2002; 298:147-157.
a novel BTB/POZ domain zinc finger transcription factor. Mol Cell 33. Hamaguchi M, Meth JL, von Klitzing C et al. DBC2, a candidate
Biol 1999; 19:3614-3623. for a tumor suppressor gene involved in breast cancer. Proc Natl
11. Hoatlin ME, Zhi Y, Ball H et al. A novel BTB/POZ transcriptional Acad Sci USA 2002; 99:13647-13652.
repressor protein interacts with the Fanconi anemia group C protein 34. Schulman BA, Carrano AC, Jeffrey PD et al. Insights into SCF
and PLZF. Blood 1999; 94:3737-3747. ubiquitin ligases from the structure of the Skp1-Skp2 complex.
12. Okabe S, Fukuda T, Ishibashi K et al. BAZF, a novel Bcl6 homolog, Nature 2000; 408:381-386.
functions as a transcriptional repressor. Mol Cell Biol 1998; 35. Stebbins CE, Kaelin WG, Jr, Pavletich NP. Structure of the
18:4235-4244. VHL-ElonginC-ElonginB complex: implications for VHL tumor
13. Huynh KD, Bardwell VJ. The BCL-6 POZ domain and other POZ suppressor function. Science 1999; 284:455-461.
domains interact with the co-repressors N-CoR and SMRT. 36. Li JY, English MA, Ball HJ et al. Sequence-specific DNA binding
Oncogene 1998; 17:2473-2484. and transcriptional regulation by the promyelocytic leukemia zinc
14. Lin RJ, Nagy L, Inoue S et al. Role of the histone deacetylase finger protein. J Biol Chem 1997; 272:22447-22455.
complex in acute promyelocytic leukaemia. Nature 1998; 37. Ball HJ, Melnick A, Shaknovich R et al. The promyelocytic leukemia
391:811-814. zinc finger (PLZF) protein binds DNA in a high molecular weight
15. Hong SH, David G, Wong CW et al. SMRT corepressor interacts complex associated with cdc2 kinase. Nucleic Acids Res 1999;
with PLZF and with the PML-retinoic acid receptor alpha 27:4106-4113.
(RARalpha) and PLZF-RARalpha oncoproteins associated with acute 38. Yeyati PL, Shaknovich R, Boterashvili S et al. Leukemia translocation
promyelocytic leukemia. Proc Natl Acad Sci USA 1997; protein PLZF inhibits cell growth and expression of cyclin A.
94:9028-9033. Oncogene 1999; 18:925-934.
16. Guidez F, Ivins S, Zhu J et al. Reduced retinoic acid-sensitivities of 39. Dong S, Zhu J, Reid A et al. Amino-terminal protein-protein
nuclear receptor corepressor binding to PML- and PLZF-RARalpha interaction motif (POZ-domain) is responsible for activities of the
underlie molecular pathogenesis and treatment of acute promyelocytic promyelocytic leukemia zinc finger- retinoic acid receptor-alpha fusion
leukemia. Blood 1998; 91:2634-2642. protein. Proc Natl Acad Sci USA 1996; 93:3624-3629.
17. Wong CW, Privalsky ML. Components of the SMRT corepressor 40. Melnick A, Ahmad KF, Arai S et al. In-depth mutational analysis of
complex exhibit distinctive interactions with the POZ domain the promyelocytic leukemia zinc finger BTB/POZ domain reveals
oncoproteins PLZF, PLZF-RARalpha, and BCL-6. J Biol Chem motifs and residues required for biological and transcriptional
1998; 273:27695-27702. functions. Mol Cell Biol 2000; 20:6550-6567.
18. David G, Alland L, Hong SH et al. Histone deacetylase associated 41. Kim CA, Phillips ML, Kim W et al. Polymerization of the SAM
with mSin3A mediates repression by the acute promyelocytic domain of TEL in leukemogenesis and transcriptional repression.
leukemia-associated PLZF protein. Oncogene 1998; 16:2549-2556. EMBO J 2001; 20:4173-4182.
19. Dhordain P, Lin RJ, Quief S et al. The LAZ3(BCL-6) oncoprotein 42. Kim CA, Gingery M, Pilpa RM, Bowie JU. The SAM domain of
recruits a SMRT/mSIN3A/histone deacetylase containing complex polyhomeotic forms a helical polymer. Nat Struct Biol 2002;
to mediate transcriptional repression. Nucleic Acids Res 1998; 9:453-457.
26:4645-4651. 43. Dhordain P, Albagli O, Ansieau S et al. The BTB/POZ domain
20. Dhordain P, Albagli O, Lin RJ et al. Corepressor SMRT binds the targets the LAZ3/BCL6 oncoprotein to nuclear dots and mediates
BTB/POZ repressing domain of the LAZ3/BCL6 oncoprotein. Proc homomerisation in vivo. Oncogene 1995; 11:2689-2697.
Natl Acad Sci USA 1997; 94:10762-10767. 44. Espinas ML, Jimenez-Garcia E, Vaquero A et al. The N-terminal
21. He LZ, Guidez F, Tribioli C et al. Distinct interactions of POZ domain of GAGA mediates the formation of oligomers that
PML-RARalpha and PLZF-RARalpha with co- repressors determine bind DNA with high affinity and specificity. J Biol Chem 1999;
differential responses to RA in APL. Nat Genet 1998; 18:126-135. 274:16461-16469.
22. Grignani F, De Matteis S, Nervi C et al. Fusion proteins of the 45. Katsani KR, Hajibagheri MA, Verrijzer CP. Co-operative DNA
retinoic acid receptor-alpha recruit histone deacetylase in binding by GAGA transcription factor requires the conserved BTB/
promyelocytic leukaemia. Nature 1998; 391:815-818. POZ domain and reorganizes promoter topology. EMBO J 1999;
23. Pintard L, Willis JH, Willems A et al. The BTB protein MEL-26 is 18:698-708.
a substrate-specific adaptor of the CUL-3 ubiquitin-ligase. Nature 46. Farkas G, Gausz J, Galloni M et al. The Trithorax-like gene encodes
2003; 425:311-316. the Drosophila GAGA factor. Nature 1994; 371:806-808.
24. Xu L, Wei Y, Reboul J et al. BTB proteins are substrate-specific 47. Pedone PV, Ghirlando R, Clore GM et al. The single Cys2-His2
adaptors in an SCF-like modular ubiquitin ligase containing CUL-3. zinc finger domain of the GAGA protein flanked by basic residues
Nature 2003; 425:316-321. is sufficient for high-affinity specific DNA binding. Proc Natl Acad
25. Krek W. BTB proteins as henchmen of Cul3-based ubiquitin ligases. Sci USA 1996; 93:2822-2826.
Nat Cell Biol 2003; 5:950-951. 48. Granok H, Leibovitch BA, Shaffer CD et al. Chromatin. Ga-ga over
26. Geyer R, Wee S, Anderson S et al. BTB/POZ domain proteins are GAGA factor. Curr Biol 1995; 5:238-241.
putative substrate adaptors for cullin 3 ubiquitin ligases. Mol Cell 49. Tsukiyama T, Becker PB, Wu C. ATP-dependent nucleosome
2003; 12:783-790. disruption at a heat-shock promoter mediated by binding of GAGA
27. Furukawa M, He YJ, Borchers C et al. Targeting of protein transcription factor. Nature 1994; 367:525-532.
ubiquitination by BTB-Cullin 3-Roc1 ubiquitin ligases. Nat Cell
Biol 2003; 5: 1001-1007.
The BTB Domain Zinc Finger Proteins 147
50. Shopland LS, Hirayoshi K, Fernandes M et al. HSF access to heat 68. Ball HJ, Melnick A, Shaknovich R et al. The promyelocytic leukemia
shock elements in vivo depends critically on promoter architecture zinc finger (PLZF) protein binds DNA in a high molecular weight
defined by GAGA factor, TFIID, and RNA polymerase II binding complex associated with cdc2 kinase. Nucleic Acids Res 1999;
sites. Genes Dev 1995; 9:2756-2769. 27:4106-4113.
51. Yoshida C, Tokumasu F, Hohmura KI et al. Long range interaction 69. Sitterlin D, Tiollais P, Transy C. The RAR alpha-PLZF chimera
of cis-DNA elements mediated by architectural transcription factor associated with Acute Promyelocytic Leukemia has retained a
Bach1. Genes Cells 1999; 4:643-655. sequence-specific DNA-binding domain. Oncogene 1997;
52. Pagans S, Ortiz-Lombardia M, Espinas ML et al. The Drosophila 14:1067-1074.
transcription factor tramtrack (TTK) interacts with Trithorax-like 70. Ivins S, Pemberton K, Guidez F et al. Regulation of Hoxb2 by
(GAGA) and represses GAGA-mediated activation. Nucleic Acids Res APL-associated PLZF protein. Oncogene 2003; 22:3685-3697.
2002; 30:4406-4413. 71. Bardwell VJ, Treisman R. The POZ domain: a conserved
53. Muto A, Hoshino H, Madisen L et al. Identification of Bach2 as a protein-protein interaction motif. Genes Dev 1994; 8:1664-1677.
B-cell-specific partner for small maf proteins that negatively regulate 72. Hoatlin ME, Zhi Y, Ball H et al. A Novel BTB/POZ Transcriptional
the immunoglobulin heavy chain gene 3' enhancer. EMBO J 1998; Repressor Protein Interacts With the Fanconi Anemia Group C
17:5734-5743. Protein and PLZF. Blood 1999; 94:3737-3747.
54. Melnick A, Carlile G, Ahmad KF et al. Critical residues within the 73. Lin W, Lai CH, Tang CJ et al. Identification and gene structure of
BTB domain of PLZF and Bcl-6 modulate interaction with a novel human PLZF-related transcription factor gene, TZFP.
corepressors. Mol Cell Biol 2002; 22:1804-1818. Biochem Biophys Res Commun. 1999; 264:789-795.
55. Deltour S, Guerardel C, Leprince D. Recruitment of SMRT/ 74. Miaw SC, Choi A, Yu E et al. ROG, repressor of GATA, regulates
N-CoR-mSin3A-HDAC-repressing complexes is not a general the expression of cytokine genes. Immunity 2000; 12:323-333.
mechanism for BTB/POZ transcriptional repressors: the case of 75. Tang CJ, Chuang CK, Hu HM et al. The zinc finger domain of
HIC-1 and gammaFBP-B. Proc Natl Acad Sci USA 1999; Tzfp binds to the tbs motif located at the upstream flanking region
96:14831-14836. of the Aie1 (aurora-C) kinase gene. J Biol Chem 2001;
56. Guerardel C, Deltour S, Leprince D. Evolutionary divergence in 276:19631-19639.
the broad complex, tramtrack and bric a brac/poxviruses and zinc 76. Dong S, Zhu J, Reid A et al. Amino-terminal protein-protein
finger domain from the candidate tumor suppressor gene interaction motif (POZ-domain) is responsible for activities of the
hypermethylated in cancer. FEBS Lett 1999; 451:253-256. promyelocytic leukemia zinc finger- retinoic acid receptor-alpha fusion
57. Pointud JC, Larsson J, Dastugue B et al. The BTB/POZ domain of protein. Proc Natl Acad Sci USA 1996; 93:3624-3629.
the regulatory proteins Bric a brac 1 (BAB1) and Bric a brac 2 77. Martin PJ, Delmotte MH, Formstecher P et al. PLZF is a negative
(BAB2) interacts with the novel Drosophila TAF(II) factor BIP2/ regulator of retinoic acid receptor transcriptional activity. Nucl Recept
dTAF(II)155. Dev Biol 2001; 237:368-380. 2003; 1:6.
58. Zheng N, Schulman BA, Song L et al. Structure of the 78. Nanba D, Mammoto A, Hashimoto K et al. Proteolytic release of
Cul1-Rbx1-Skp1-F boxSkp2 SCF ubiquitin ligase complex. Nature the carboxy-terminal fragment of proHB-EGF causes nuclear export
2002; 416:703-709. of PLZF. J Cell Biol 2003; 163:489-502.
59. Stebbins CE, Kaelin WG, Jr., Pavletich NP. Structure of the 79. Reid A, Gould A, Brand N et al. Leukemia translocation gene, PLZF,
VHL-ElonginC-ElonginB complex: implications for VHL tumor is expressed with a speckled nuclear pattern in early hematopoietic
suppressor function. Science 1999; 284:455-461. progenitors. Blood 1995; 86:4544-4552.
60. Chen Z, Brand NJ, Chen A et al. Fusion between a novel 80. Licht JD, Shaknovich R, English MA et al. Reduced and altered
Kruppel-like zinc finger gene and the retinoic acid receptor-alpha DNA-binding and transcriptional properties of the PLZF-retinoic
locus due to a variant t(11;17) translocation associated with acute acid receptor-alpha chimera generated in t(11;17)-associated acute
promyelocytic leukaemia. EMBO J 1993; 12:1161-1167. promyelocytic leukemia. Oncogene 1996; 12:323-336.
61. Chen SJ, Zelent A, Tong JH et al. Rearrangements of the retinoic 81. Koken MH, Reid A, Quignon F et al. Leukemia-associated retinoic
acid receptor alpha and promyelocytic leukemia zinc finger genes acid receptor alpha fusion partners, PML and PLZF, heterodimerize
resulting from t(11;17)(q23;q21) in a patient with acute and colocalize to nuclear bodies. Proc Natl Acad Sci USA 1997;
promyelocytic leukemia. J Clin Invest 1993; 91:2260-2267. 94:10255-10260.
62. Licht JD, Chomienne C, Goy A et al. Clinical and molecular 82. Ruthardt M, Orleth A, Tomassoni L et al. The acute promyelocytic
characterization of a rare syndrome of acute promyelocytic leukemia leukaemia specific PML and PLZF proteins localize to adjacent and
associated with translocation (11;17). Blood 1995; 85:1083-1094. functionally distinct nuclear bodies. Oncogene 1998; 16:1945-1953.
63. Culligan DJ, Stevenson D, Chee Y et al. Acute promyelocytic 83. Hummel J, Wells R, Dubé I et al. Deregulation of NPM and PLZF
leukaemia with t(11;17)(q23;q12-21) and a good initial response to in a variant t(5;17) case of acute promyelocytic leukemia. Oncogene
prolonged ATRA and combination chemotherapy. Br J Haematol 1999; 18:633-641.
1998; 100:328-330. 84. Melnick A, Licht J. Deconstructing a disease: RARα, its fusion
64. Grimwade D, Biondi A, Mozziconacci MJ et al. Characterization of proteins and their roles in the pathogenesis of acute promyelocytic
acute promyelocytic leukemia cases lacking the classic t(15;17): results leukemia. Blood 1999; 93:3167-3215.
of the European Working Party. Groupe Francais de Cytogenetique 85. Dai MS, Chevallier N, Stone S et al. The effects of the Fanconi
Hematologique, Groupe de Francais d’Hematologie Cellulaire, UK anemia zinc finger (FAZF) on cell cycle, apoptosis, and proliferation
Cancer Cytogenetics Group and BIOMED 1 European are differentiation stage-specific. J Biol Chem 2002;
Community-Concerted Action “Molecular Cytogenetic Diagnosis in 277:26327-26334.
Haematological Malignancies”. Blood 2000; 96:1297-1308. 86. Ward JO, McConnell MJ, Carlile GW et al. The acute promyelocytic
65. Jansen JH, de Ridder MC, Geertsma WM et al. Complete remission leukemia-associated protein, promyelocytic leukemia zinc finger,
of t(11;17) positive acute promyelocytic leukemia induced by all-trans regulates 1,25-dihydroxyvitamin D(3)-induced monocytic
retinoic acid and granulocyte colony-stimulating factor. Blood 1999; differentiation of U937 cells through a physical interaction with
94:39-45. vitamin D(3) receptor. Blood 2001; 98:3290-3300.
66. Zhang T, Xiong H, Kan LX et al. Genomic sequence, structural 87. Labbaye C, Quaranta MT, Pagliuca A et al. PLZF induces
organization, molecular evolution, and aberrant rearrangement of megakaryocytic development, activates Tpo receptor expression and
promyelocytic leukemia zinc finger gene. Proc Natl Acad Sci USA interacts with GATA1 protein. Oncogene 2002; 21:6669-6679.
1999; 96:11422-11427. 88. Tsuzuki S, Enver T. Interactions of GATA-2 with the promyelocytic
67. Petti MC, Fazi F, Gentile M et al. Complete remission through blast leukemia zinc finger (PLZF) protein, its homologue FAZF, and the
cell differentiation in PLZF/RARalpha-positive acute promyelocytic t(11;17)-generated PLZF-retinoic acid receptor alpha oncoprotein.
leukemia: in vitro and in vivo studies. Blood 2002; 100:1065-1067. Blood 2002; 99:3404-3410.
148 Zinc Finger Proteins: From Atomic Contact to Cellular Function
89. Cook M, Gould A, Brand N et al. Expression of the zinc-finger 110. Nagpal S, Saunders M, Kastner P et al. Promoter context-and
gene PLZF at rhombomere boundaries in the vertebrate hindbrain. response element-dependent specificity of the transcriptional
Proc Natl Acad Sci USA 1995; 92:2249-2253. activation and modulating functions of retinoic acid receptors. Cell
90. Barna M, Hawe N, Niswander L et al. Plzf regulates limb and axial 1992; 70:1007-1019.
skeletal patterning. Nat Genet 2000; 25:166-172. 111. Hauksdottir H, Privalsky ML. DNA recognition by the aberrant
91. Fahnenstich J, Nandy A, Milde-Langosch K et al. Promyelocytic retinoic acid receptors implicated in human acute promyelocytic
leukaemia zinc finger protein (PLZF) is a glucocorticoid- and leukemia. Cell Growth Differ 2001; 12:85-98.
progesterone-induced transcription factor in human endometrial 112. Perez A, Kastner P, Sethi S et al. PMLRAR homodimers: distinct
stromal cells and myometrial smooth muscle cells. Mol Hum Reprod DNA binding properties and heteromeric interactions with RXR.
2003; 9:611-623. EMBO J 1993; 12:3171-3182.
92. Wan Y, Nordeen S. Overlapping but Distinct Gene Regulation 113. So CW, Dong S, So CK et al. The impact of differential binding of
Profiles by Glucocorticoids and Progestins in Human Breast Cancer wild-type RARalpha, PML-, PLZF- and NPM-RARalpha fusion
Cells. Mol Endo 2002; 16:1204–1214. proteins towards transcriptional co-activator, RIP-140, on retinoic
93. Shaknovich R, Yeyati PL, Ivins S et al. The promyelocytic leukemia acid responses in acute promyelocytic leukemia. Leukemia 2000;
zinc finger protein affects myeloid cell growth, differentiation, and 14:77-83.
apoptosis. Mol Cell Biol 1998; 18:5533-5545. 114. Chen Z, Guidez F, Rousselot P et al. PLZF-RAR alpha fusion
94. McConnell M, Chevallier N, Berkofsky-Fessler W et al. Growth proteins generated from the variant t(11;17)(q23;q21) translocation
Suppression by Acute Promyelocytic Leukemia-Associated Protein in acute promyelocytic leukemia inhibit ligand-dependent
PLZF is Mediated by Repression of c-Myc Expression. Mol Cell transactivation of wild-type retinoic acid receptors. Proc Natl Acad
Bio. 2003; 23:9375-9388. Sci USA 1994; 91:1178-1182.
95. Han SH, Jeon JH, Ju HR et al. VDUP1 upregulated by TGF-beta1 115. Puccetti E, Obradovic D, Beissert T et al. AML-associated
and 1,25-dihydorxyvitamin D3 inhibits tumor cell growth by Translocation Products Block Vitamin D(3)-induced Differentiation
blocking cell-cycle progression. Oncogene 2003; 22:4035-4046. by Sequestering the Vitamin D(3) Receptor. Cancer Res 2002;
96. Barna M, Merghoub T, Costoya JA et al. Plzf Mediates 62:7050-7058.
Transcriptional Repression of HoxD Gene Expression through 116. Tomita A, Buchholz DR, Obata K et al. Fusion protein of retinoic
Chromatin Remodeling. Dev Cell 2002; 3:499-510. acid receptor alpha with promyelocytic leukemia protein or
97. Ruthardt M, Testa U, Nervi C et al. Opposite effects of the acute promyelocytic leukemia zinc finger protein recruits N-CoR-TBLR1
promyelocytic leukemia PML-retinoic acid receptor alpha (RAR corepressor complex to repress transcription in vivo. J Biol Chem
alpha) and PLZF-RAR alpha fusion proteins on retinoic acid 2003; 278:30788-30795.
signalling. Mol Cell Biol 1997; 17:4859-4869. 117. Lin RJ, Evans RM. Acquisition of oncogenic potential by RAR
98. Gilliland DG, Griffin JD. Role of FLT3 in leukemia. Curr Opin chimeras in acute promyelocytic leukemia through formation of
Hematol 2002; 9:274-281. homodimers. Mol Cell 2000; 5:821-830.
99. Hong SH, Wong CW, Privalsky ML. Signaling by tyrosine kinases 118. Koken MH, Daniel MT, Gianni M et al. Retinoic acid, but not
negatively regulates the interaction between transcription factors and arsenic trioxide, degrades the PLZF/RARalpha fusion protein, without
SMRT (silencing mediator of retinoic acid and thyroid hormone inducing terminal differentiation or apoptosis, in a RA-therapy
receptor) corepressor. Mol Endocrinol 1998; 12:1161-1171. resistant t(11;17)(q23;q21) APL patient. Oncogene 1999;
100. Zhou Y, Gross W, Hong SH et al. The SMRT corepressor is a target 18:1113-1118.
of phosphorylation by protein kinase CK2 (casein kinase II). Mol 119. Di Croce L, Raker VA, Corsaro M et al. Methyltransferase
Cell BioChem 2001; 220:1-13. recruitment and DNA hypermethylation of target promoters by an
101. Hyman J, Chen H, Di Fiore PP et al. Epsin 1 undergoes oncogenic transcription factor. Science 2002; 295:1079-1082.
nucleocytosolic shuttling and its eps15 interactor NH(2)-terminal 120. Park DJ, Vuong PT, De Vos S et al. Comparative analysis of genes
homology (ENTH) domain, structurally similar to Armadillo and regulated by PML/RAR{alpha} and PLZF/RAR{alpha} in response
HEAT repeats, interacts with the transcription factor promyelocytic to retinoic acid using oligonucleotide arrays. Blood 2003;
leukemia Zn(2)+ finger protein (PLZF). J Cell Biol 2000; 102:3727-3736.
149:537-546. 121. Cheng GX, Zhu XH, Men XQ et al. Distinct leukemia phenotypes
102. Daniel JM, Reynolds AB. The catenin p120(ctn) interacts with Kaiso, in transgenic mice and different corepressor interactions generated
a novel BTB/POZ domain zinc finger transcription factor. Mol Cell by promyelocytic leukemia variant fusion genes PLZF-RARalpha and
Biol 1999; 19:3614-3623. NPM-RARalpha. Proc Natl Acad Sci USA 1999; 96:6318-6323.
103. David G, Alland L, Hong SH et al. Histone deacetylase associated 122. Guidez F, Huang W, Tong JH et al. Poor response to all-trans retinoic
with mSin3A mediates repression by the acute promyelocytic acid therapy in a t(11;17) PLZF/RAR alpha patient. Leukemia 1994;
leukemia-associated PLZF protein. Oncogene 1998; 16:2549-2556. 8:312-317.
104. Lin RJ, Nagy L, Inoue S et al. Role of the histone deacetylase 123. Grimwade D, Gorman P, Duprez E et al. Characterization of cryptic
complex in acute promyelocytic leukaemia. Nature 1998; rearrangements and variant translocations in acute promyelocytic
391:811-814. leukemia. Blood 1997; 90:4876-4885.
105. Melnick A, Ahmad KF, Arai S et al. In-depth mutational analysis of 124. Ball HJ, Melnick A, Shaknovich R et al. The promyelocytic leukemia
the promyelocytic leukemia zinc finger BTB/POZ domain reveals zinc finger (PLZF) protein binds DNA in a high molecular weight
motifs and residues required for biological and transcriptional complex associated with cdc2 kinase. Nucleic Acids Res 1999;
functions. Mol Cell Biol 2000; 20:6550-6567. 27:4106-4113.
106. Melnick A, Carlile G, Ahmad KF et al. Critical residues within the 125. He L, Bhaumik M, Tribioli C et al. Two critical hits for
BTB domain of PLZF and Bcl-6 modulate interaction with promyelocytic leukemia. Mol Cell 2000; 6:1131-1141.
corepressors. Mol Cell Biol 2002; 22:1804-1818. 126. Kerckaert JP, Deweindt C, Tilly H et al. LAZ3, a novel zinc-finger
107. Melnick A, Carlile GW, McConnell MJ et al. AML-1/ETO fusion encoding gene, is disrupted by recurring chromosome 3q27
protein is a dominant negative inhibitor of transcriptional repression translocations in human lymphomas. Nat Genet 1993; 5:66-70.
by the promyelocytic leukemia zinc finger protein. Blood 2000; 127. Baron BW, Nucifora G, McCabe N et al. Identification of the gene
96:3939-3947. associated with the recurring chromosomal translocations
108. Melnick AM, Westendorf JJ, Polinger A et al. The ETO protein t(3;14)(q27;q32) and t(3;22)(q27;q11) in B-cell lymphomas. Proc
disrupted in t(8;21)-associated acute myeloid leukemia is a corepressor Natl Acad Sci USA 1993; 90:5262-5266.
for the promyelocytic leukemia zinc finger protein. Mol Cell Biol 128. Ye BH, Lista F, Lo Coco F et al. Alterations of a zinc finger-encoding
2000; 20:2075-2086. gene, BCL-6, in diffuse large-cell lymphoma. Science 1993;
109. Kang S, Chang W, Cho S et al. Modification of promyelocytic 262:747-750.
leukemia zinc finger protein (PLZF) by SUMO-1 conjugation
regulates its transcriptional repressor activity. J Biol Chem 2003;In
Press.
The BTB Domain Zinc Finger Proteins 149
129. Capello D, Vitolo U, Pasqualucci L et al. Distribution and pattern 154. Hosokawa Y, Maeda Y, Seto M. Target genes downregulated by the
of BCL-6 mutations throughout the spectrum of B-cell neoplasia. BCL-6/LAZ3 oncoprotein in mouse Ba/F3 cells. Biochem Biophys
Blood 2000; 95:651-659. Res Commun 2001; 283:563-568.
130. Ye BH. BCL-6 in the pathogenesis of non-Hodgkin’s lymphoma. 155. Reljic R, Wagner SD, Peakman LJ et al. Suppression of signal
Cancer Invest 2000; 18:356-365. transducer and activator of transcription 3-dependent B lymphocyte
131. Ye BH, Cattoretti G, Shen Q et al. The BCL-6 proto-oncogene terminal differentiation by BCL-6. J Exp Med 2000; 192:1841-1848.
controls germinal-centre formation and Th2-type inflammation. Nat 156. Arima M, Toyama H, Ichii H et al. A putative silencer element in
Genet 1997; 16:161-170. the IL-5 gene recognized by Bcl6. J Immunol 2002; 169:829-836.
132. Dent AL, Shaffer AL, Yu X et al. Control of inflammation, cytokine 157. Gupta S, Jiang M, Anthony A et al. Lineage-specific modulation of
expression, and germinal center formation by BCL-6. Science 1997; interleukin 4 signaling by interferon regulatory factor 4. J Exp Med
276:589-592. 1999; 190:1837-1848.
133. Fukuda T, Yoshida T, Okada S et al. Disruption of the Bcl6 gene 158. Dent AL, Hu-Li J, Paul WE et al. T helper type 2 inflammatory
results in an impaired germinal center formation. J Exp Med 1997; disease in the absence of interleukin 4 and transcription factor
186:439-448. STAT6. Proc Natl Acad Sci USA 1998; 95:13823-13828.
134. Dhordain P, Lin RJ, Quief S et al. The LAZ3(BCL-6) oncoprotein 159. Dent AL, Doherty TM, Paul WE et al. BCL-6-deficient mice reveal
recruits a SMRT/mSIN3A/histone deacetylase containing complex an IL-4-independent, STAT6-dependent pathway that controls
to mediate transcriptional repression. Nucleic Acids Res 1998; susceptibility to infection by Leishmania major. J Immunol 1999;
26:4645-4651. 163:2098-2103.
135. Huynh KD, Bardwell VJ. The BCL-6 POZ domain and other POZ 160. Harris MB, Chang CC, Berton MT et al. Transcriptional repression
domains interact with the co-repressors N-CoR and SMRT. of Stat6-dependent interleukin-4-induced genes by BCL-6: specific
Oncogene 1998; 17:2473-2484. regulation of iepsilon transcription and immunoglobulin E switching.
136. Wong CW, Privalsky ML. Components of the SMRT corepressor Mol Cell Biol 1999; 19:7264-7275.
complex exhibit distinctive interactions with the POZ domain 161. Kusam S, Toney LM, Sato H et al. Inhibition of Th2 differentiation
oncoproteins PLZF, PLZF-RARalpha, and BCL-6. J Biol Chem and GATA-3 expression by BCL-6. J Immunol 2003;
1998; 273:27695-27702. 170:2435-2441.
137. Privé 6. Unpublished results 2003. 162. Albagli O, Lindon C, Lantoine D et al. DNA replication progresses
138. Ahmad K, Melnick A, Lax S et al. Mechanism of SMRT Corepressor on the periphery of nuclear aggregates formed by the BCL6
Recruitment by the BCL6 BTB Domain. Molecular Cell 2003; transcription factor. Mol Cell Biol 2000; 20:8560-8570.
551-564. 163. Cattoretti G, Chang CC, Cechova K et al. BCL-6 protein is
139. Kawamata N, Miki T, Ohashi K et al. Recognition DNA sequence expressed in germinal-center B cells. Blood 1995; 86:45-53.
of a novel putative transcription factor, BCL6. Biochem Biophys 164. Onizuka T, Moriyama M, Yamochi T et al. BCL-6 gene product, a
Res Commun 1994; 204:366-374. 92- to 98-kD nuclear phosphoprotein, is highly expressed in germinal
140. Dhordain P, Albagli O, Honore N et al. Colocalization and center B cells and their neoplastic counterparts. Blood 1995;
heteromerization between the two human oncogene POZ/zinc finger 86:28-37.
proteins, LAZ3 (BCL6) and PLZF. Oncogene 2000; 19:6240-6250. 165. Allman D, Jain A, Dent A et al. BCL-6 expression during B-cell
141. Lemercier C, Brocard MP, Puvion-Dutilleul F et al. Class II histone activation. Blood 1996; 87:5257-5268.
deacetylases are directly recruited by BCL6 transcriptional repressor. 166. Niu H, Ye BH, Dalla-Favera R. Antigen receptor signaling induces
J Biol Chem 2002; 277:22045-22052. MAP kinase-mediated phosphorylation and degradation of the BCL-6
142. Melnick A. Unpublished results. 2003. transcription factor. Genes Dev 1998; 12:1953-1961.
143. Davies JM, Hawe N, Kabarowski J et al. Novel BTB/POZ domain 167. Toyama H, Okada S, Hatano M et al. Memory B cells without
zinc-finger protein, LRF, is a potential target of the LAZ-3/BCL-6 somatic hypermutation are generated from Bcl6-deficient B cells.
oncogene. Oncogene 1999;18:365-375. Immunity 2002; 17:329-339.
144. Vasanwala FH, Kusam S, Toney LM et al. Repression of AP-1 168. Wang X, Li Z, Naganuma A, Ye BH. Negative autoregulation of
function: a mechanism for the regulation of Blimp-1 expression and BCL-6 is bypassed by genetic alterations in diffuse large B cell
B lymphocyte differentiation by the B cell lymphoma-6 lymphomas. Proc Natl Acad Sci USA 2002; 99:15018-15023.
protooncogene. J Immunol 2002; 169:1922-1929. 169. Moriyama M, Yamochi T, Semba K et al. BCL-6 is phosphorylated
145. Chevallier N, Corcoran C, Lennon C et al. The ETO Protein of at multiple sites in its serine- and proline-clustered region by
t(8;21) AML is a Corepressor for the Bcl-6 B-Cell Lymphoma mitogen-activated protein kinase (MAPK) in vivo. Oncogene 1997;
Oncoprotein. Blood 2004; in press. 14:2465-2474.
146. Melnick A, Carlile GW, McConnell MJ et al. AML-1/ETO fusion 170. Shaffer AL, Lin KI, Kuo TC et al. Blimp-1 orchestrates plasma cell
protein is a dominant negative inhibitor of transcriptional repression differentiation by extinguishing the mature B cell gene expression
by the promyelocytic leukemia zinc finger protein. Blood 2000; program. Immunity 2002; 17:51-62.
96:3939-3947. 171. Carbone A, Gloghini A, Gaidano G et al. BCL-6 protein expression
147. Melnick AM, Westendorf JJ, Polinger A et al. The ETO protein in human peripheral T-cell neoplasms is restricted to CD30+
disrupted in t(8;21)-associated acute myeloid leukemia is a corepressor anaplastic large-cell lymphomas. Blood 1997; 90:2445-2450.
for the promyelocytic leukemia zinc finger protein. Mol Cell Biol 172. Hyjek E, Chadburn A, Liu YF. BCL-6 protein is expressed in
2000; 20:2075-2086. precursor T-cell lymphoblastic lymphoma and in prenatal and
148. Seyfert VL, Allman D, He Y et al. Transcriptional repression by the postnatal thymus. Blood 2001; 97:270-276.
proto-oncogene BCL-6. Oncogene 1996; 12:2331-2342. 173. Ouyang W, Lohning M, Gao Z et al. Stat6-independent GATA-3
149. Shaffer AL, Yu X, He Y et al. BCL-6 represses genes that function autoactivation directs IL-4-independent Th2 development and
in lymphocyte differentiation, inflammation, and cell cycle control. commitment. Immunity 2000; 12:27-37.
Immunity 2000; 13:199-212. 174. Yamochi T, Kitabayashi A, Hirokawa M et al. Regulation of BCL-6
150. Albagli O, Lantoine D, Quief S et al. Overexpressed BCL6 (LAZ3) gene expression in human myeloid/monocytoid leukemic cells.
oncoprotein triggers apoptosis, delays S phase progression and Leukemia 1997; 11:694-700.
associates with replication foci. Oncogene 1999; 18:5063-5075. 175. Yoshida T, Fukuda T, Hatano M et al. The role of Bcl6 in mature
151. Zhang H, Okada S, Hatano M et al. A new functional domain of cardiac myocytes. Cardiovasc Res 1999; 42:670-679.
Bcl6 family that recruits histone deacetylases. Biochim Biophys Acta 176. Yoshida T, Fukuda T, Okabe S et al. The BCL6 gene is
2001; 1540:188-200. predominantly expressed in keratinocytes at their terminal
152. Takeda N, Arima M, Tsuruoka N et al. Bcl6 is a transcriptional differentiation stage. Biochem Biophys Res Commun 1996;
repressor for the IL-18 gene. J Immunol 2003; 171:426-431. 228:216-220.
153. Toney LM, Cattoretti G, Graf JA et al. BCL-6 regulates chemokine
gene transcription in macrophages. Nat Immunol 2000; 1:214-220.
150 Zinc Finger Proteins: From Atomic Contact to Cellular Function
177. Albagli-Curiel O, Dhordain P, Lantoine D et al. Increased expression 188. Kim SW, Fang X, Ji H et al. Isolation and characterization of XKaiso,
of the LAZ3 (BCL6) proto-oncogene accompanies murine skeletal a transcriptional repressor that associates with the catenin Xp120(ctn)
myogenesis. Differentiation 1998; 64:33-44. in Xenopus laevis. J Biol Chem 2002; 277:8202-8208.
178. Kojima S, Hatano M, Okada S et al. Testicular germ cell apoptosis 189. Yoon HG, Chan DW, Reynolds AB et al. N-CoR mediates DNA
in Bcl6-deficient mice. Development 2001; 128:57-65. methylation-dependent repression through a methyl CpG binding
179. Kumagai T, Miki T, Kikuchi M et al. The proto-oncogene Bc16 protein Kaiso. Mol Cell 2003; 12:723-734.
inhibits apoptotic cell death in differentiation-induced mouse 190. Jenuwein T, Allis CD. Translating the histone code. Science 2001;
myogenic cells. Oncogene 1999; 18:467-475. 293:1074-1080.
180. Baron BW, Anastasi J, Thirman MJ et al. The human programmed 191. Peukert K, Staller P, Schneider A et al. An alternative pathway for
cell death-2 (PDCD2) gene is a target of BCL6 repression: gene regulation by Myc. EMBO J 1997; 16:5672-5686.
implications for a role of BCL6 in the down-regulation of apoptosis. 192. Salghetti SE, Kim SY, Tansey WP. Destruction of Myc by
Proc Natl Acad Sci USA 2002; 99:2860-2865. ubiquitin-mediated proteolysis: cancer-associated and transforming
181. Shvarts A, Brummelkamp TR, Scheeren F et al. A senescence rescue mutations stabilize Myc. EMBO J 1999; 18:717-726.
screen identifies BCL6 as an inhibitor of anti-proliferative 193. Staller P, Peukert K, Kiermaier A et al. Repression of p15INK4b
p19(ARF)-p53 signaling. Genes Dev 2002; 16:681-686. expression by Myc through association with Miz-1. Nat Cell Biol
182. Yamochi T, Kaneita Y, Akiyama T et al. Adenovirus-mediated high 2001; 3:392-399.
expression of BCL-6 in CV-1 cells induces apoptotic cell death 194. Seoane J, Pouponnot C, Staller P et al. TGFbeta influences Myc,
accompanied by down-regulation of BCL-2 and BCL-X(L). Miz-1 and Smad to control the CDK inhibitor p15INK4b. Nat
Oncogene 1999; 18:487-494. Cell Biol 2001; 3:400-408.
183. Wales MM, Biel MA, el Deiry W et al. p53 activates expression of 195. Seoane J, Le HV, Massague J. Myc suppression of the p21(Cip1)
HIC-1, a new candidate tumour suppressor gene on 17p13.3. Nat Cdk inhibitor influences the outcome of the p53 response to DNA
Med 1995; 1:570-577. damage. Nature 2002; 419:729-734.
184. Chen WY, Zeng X, Carter MG et al. Heterozygous disruption of 196. Wu S, Cetinkaya C, Munoz-Alonso MJ et al. Myc represses
Hic1 predisposes mice to a gender-dependent spectrum of malignant differentiation-induced p21CIP1 expression via Miz-1-dependent
tumors. Nat Genet 2003; 33:197-202. interaction with the p21 core promoter. Oncogene 2003; 22:351-360.
185. Deltour S, Pinte S, Guerardel C et al. The human candidate tumor 197. Herold S, Wanzel M, Beuger V et al. Negative regulation of the
suppressor gene HIC1 recruits CtBP through a degenerate mammalian UV response by Myc through association with Miz-1.
GLDLSKK motif. Mol Cell Biol 2002; 22:4890-4901. Mol Cell 2002; 10:509-521.
186. Daniel JM, Spring CM, Crawford HC et al. The p120(ctn)-binding 198. Piluso D, Bilan P, Capone JP. Host cell factor-1 interacts with and
partner Kaiso is a bi-modal DNA-binding protein that recognizes antagonizes transactivation by the cell cycle regulatory factor Miz-1.
both a sequence-specific consensus and methylated CpG J Biol Chem 2002; 277:46799-46808.
dinucleotides. Nucleic Acids Res 2002; 30:2911-2919. 199. Ziegelbauer J, Shan B, Yager D et al. Transcription factor MIZ-1 is
187. Prokhortchouk A, Hendrich B, Jorgensen H et al. The p120 catenin regulated via microtubule association. Mol Cell 2001; 8:339-349.
partner Kaiso is a DNA methylation-dependent transcriptional 200. Mulder NJ, Apweiler R, Attwood TK et al. The InterPro Database,
repressor. Genes Dev 2001; 15:1613-1618. 2003 brings increased coverage and new features. Nucleic Acids Res
2003; 31:315-318.
CHAPTER 21
Abstract
K
rüppel-associated box (KRAB) domain, present at the N details of the mechanisms have not been revealed, binding of HP1
terminus of mammalian C2H2 zinc finger proteins, binds is thought to render the DNA region into condensed states and
KAP-1 (also named TIF1β or KRIP-1) that binds HP1α result in gene silencing.4,5 Highly structured DNA and the com-
and SETDB1 (H3-K9 methyltransferase) as well as NuRD com- partmentalization of the heterochromatin may make promoters
plexes (histone deacetylases). The KRAB domain tethers these within inaccessible to RNA polymerase II and III complexes.6
binding proteins to a specific gene recognized by zinc fingers, This repression is called “gene silencing” or “position effect varie-
resulting in a heterochromatin-associated, strong, long-term re- gation (PEV)” after its discovery as the gene-shut-off phenom-
pression of the gene. The importance of such repression is re- enon associated with chromosome rearrangement in Drosophila.7
flected in the large number (two hundred ninety) of the KRAB Only two years after KRAB was found to be a repressor domain,8
zinc finger protein genes in the human genome and revealed by two groups independently identified a KRAB binding protein
early stage-development failure of mouse embryos lacking the and named it KAP-1 or KRIP1,9,10 a third group found a HP1
common corepressor, KAP-1 (TIF1β). Many reports have de- binding protein and named it TIF1β,11 and a fourth group found
scribed KRAB zinc finger proteins that participate in a variety of a KRAB binding protein recognizing it to be TIF1β and KAP-1.12
cellular processes including development of organs and cell types These findings firmly established that KRAB domains of zinc
such as heart, bone, sperm and hematopoietic cells. Some KRAB finger proteins interact with KAP-1 (TIF1β) which in turn binds
zinc finger proteins have an additional repression mechanism HP1 to repress gene expression. This repression turned out to in-
that is KRAB domain-independent and involves histone volve several other proteins that affect chromosomal structure (See
deactylases or NSD1 histone lysine methyltrasferase. below).
Zinc Finger Proteins: From Atomic Contact to Cellular Function, edited by Shiro Iuchi and Natalie Kuldell.
©2005 Landes Bioscience/Eurekah.com and Kluwer Academic/Plenum Publishers.
152 Zinc Finger Proteins: From Atomic Contact to Cellular Function
Figure 1. A) A schematic representation of various KRAB domains present in zinc finger proteins. B) A consensus sequence of KRAB AB domain.
The sequence was previously described.26
about 75-amino acids, is thought to be the origin of all and has sub-domain A is the principal repressor.8,25 Sub-domain B forti-
diverged to the other types by means of duplication, deletion and fies the role of sub-domain A to some extent but the sub-domain
mutation. One strong argument for this suggestion has come from b barely plays any role.25 Sub-domain C has been suggested to
the fact that as many as two hundred sixty two genes of zinc strengthen the repression by sub-domain A but it has not been
finger proteins cluster in the human chromosome 19 and as many demonstrated.20 Perhaps, it may be concluded that KRAB A
as twenty-one genes of KRAB zinc finger proteins cluster in the sub-domain is necessary and sufficient to repress gene expression
700-kb region of human chromosome 19q13.2.19,23,24 There is as originally stated.8
no perfect consensus sequence for the KRAB domains among KAP-1 (TIF1β) binds to the KRAB domain through the
zinc finger proteins, but KRAB A with 42-amino acids is the N-terminal RING, B-boxes, coiled coil (RBCC) (Fig. 2, see Chap-
longest and best conserved of all the sub-domains (Fig. 1B) This ter 16).25-27 The affinity constant is 7x106 M-1.28 It is roughly
sub-domain contains hydrophobic as well as charged amino ac- comparable to the Ka for zinc fingers-DNA binding (see Chapter
ids and is supposed to fold into an amphiphilic α-helix.14 KRAB 2) and the KRAB-RBCC binding has been well demonstrated by
B with a 32-amino acid long sequence is also well conserved and electrophoretic mobility shift assay27,28 and yeast two hybrid bind-
supposed to fold into an α-helix. KRAB b is, however, highly ing assay.25 The KRAB-RBCC binding is very specific as it has
diverged from KRAB B and less conserved within the been shown that the KRAB domains of KOX1, MZF22, MTZ1
sub-domain.18 KRAB C with a 21-amino acid long sequence is and MZF13 bind RBCC from KAP-1 (TIF1β) but not any other
also less conserved.20 homologues including RBCC derived from TIF1α, TIF1γ and
MID.25,27 This strict specificity is consistent with the observa-
Functions of KRAB and the Protein tion that TIF1β -/- mouse embryos are affected at an early devel-
opmental stage and die by E8.5.29 Interestingly, one KRAB binds
Complexes a homotrimer of RBCC in vitro27 and binding of a KRAB zinc
finger protein to KAP-1 (TIF1β) oligomers has been observed in
KRAB and the Interaction with KAP-1(TIF1β) living cells.30 Thus, a KRAB zinc finger protein is able to amplify
Because no natural target genes had been identified, studies the repression through KAP-1 (TIF1β).
on the function of KRAB domains began with KRAB AB do- The RING finger of KAP-1 (TIF1β) contains many canoni-
main (from KOX1 zinc finger protein) fused to GAL4. GAL4 cal cysteine residues that are supposed to chelate zinc ions and
alone activated transcription of a reporter gene in cells but the thereby form the stable tertiary structure (Fig. 2 and see Chapter
expression was abolished by fusing the KRAB domain to the ac- 16). A double mutation of two cysteine residues of the RING
tivator, giving the first insight into KRAB domain’s function.8 finger completely destroys its binding ability to KRAB,27 and a
Since then, the GAL4-KRAB fusion protein and similar constructs similar mutational effect occurs in the second B box. The coiled
have been almost exclusively used for in vivo and in vitro experi- coil domain does not contain conserved cysteine residues but
ments. Analyzing by this method, scientists now know that al- contains many leucine or other hydrophobic residues, which are
though KRAB domain structures has several variations (Fig. 1A), periodically included in the peptide separated by six hydrophobic
KRAB Zinc Finger Proteins 153
Figure 2. A schematic representation of protein-protein interaction in a KRAB zinc finger protein complex. Domain-domain interactions are shown
by dotted lines. The N terminus of proteins is presented at the left and the C terminus at the right, but the KRAB zinc finger protein is drawn in
the opposite direction. Numbers (1), (3) and (6) next to the C terminus of proteins indicate molar ratios in the protein complex.
and hydrophilic amino acid residues and thought to compose a that have a chromosome with a reporter gene at one locus and a
primary pole of a leucine zipper-like structure. Mutagenesis of a repressor gene, encoding a KRAB domain fused to a
leucine of each coil domain intermediately or completely abol- reporter-gene-binding domain whose activity is regulated by a
ished the KRAB binding. These results suggest the predicted leu- hormone, at another.35 In these cells the reporter gene was present
cine zipper-like coiled coil structure is present. in euchromatin without the active repressor, but it was located
within heterochromatin upon activation of the repressor. Fur-
KAP-1 (TIF1β) and the Interactions thermore, the silenced gene was physically bound with KAP-1
(TIF1β), HP1α and SETDB1. This silencing was maintained
with the Downstream Proteins over fifty cell divisions, suggesting that once heterochromatin is
From the middle of KAP-1 (TIF1β) protein to the C termi- made, the structure persists.35 Thus, a KRAB zinc finger protein
nus, there are three conserved domains, HP1 binding box (HP1 silences expression of the target gene by establishing the hetero-
BD), PHD finger and bromodomain (Fig. 2).9-12 These are in- chromatin through the concerted action of KRAB, KAP-1
dispensable for repression. HP1 BD binds HP1α31,32 with 1: 2 at (TIF1β), HP1α and SETDB135 as well as NuRD complexes.33
a Ka value of 1.6 x 107 M-1,32 and PHD finger plus bromodomain It was also shown that the gene silencing is variegated from one
bind NuRD complex as well as SETDB1.33,34 The NuRD com- cell to another. The observation is consistent with the idea that
plexes remove acetyl groups from the N terminus of histones and the more tightly folding heterochromatin gives rise to the stron-
allow histones to interact with DNA33 while SETBD1 methy- ger and longer silencing. Since the promoter of the silence re-
lates the lysine residue at position 9 of histone 3 and stimulates porter gene was hypermethylated at the CpG rich region, DNA
HP1α to interact with DNA.34 The three different KAP1 (TIF1β) methylation may also participate in the gene silencing.35
binding proteins together work toward forming heterochroma- KAP-1 (TIF1β) and HP1α are present in undifferentiated
tin. As KAP-1 (TIF1β) is present as a trimer,27 HP1 can bind six mouse embryonic carcinoma F9 cells, but only HP1α is
molecules per KRAB zinc finger complex and both NuRD and localized within heterochromatin. When the cells are differenti-
SETDB1 can bind up to three molecules (Fig. 2). It is likely that ated by retinoic acid, however, KAP-1 (TIF1β) is compartmented
the three executers would be associated simultaneously through with HP1 within heterochromatin.36 This suggests that some fac-
KAP-1 (TIF1β) to mediate the severe repression. tor, which is likely to be a specific KRAB zinc finger protein,
initiates specific gene’s heterochromatin formation resulting in
Gene Silencing by Heterochromatin Formation silencing of gene expression. Presence of the KRAB protein with
A majority of results on KRAB domain that have been col- an active form of the zinc fingers would be a critical step to ini-
lected to date measure either repression or heterochromatin for- tiate such gene silencing (see Chapter 2). KAP-1 (TIF1β) under-
mation, and the relation between the two has not been clear. Re- goes autophosphorylation and transphosphorylates HP1.31 This
cently, however, a series of experiments examined both gene may have also something to do with regulation of heterochroma-
silencing and chromatin structure together by constructing cells tin formation and gene silencing.
154 Zinc Finger Proteins: From Atomic Contact to Cellular Function
22. Collins T, Stone JR, Williams AJ. All in the family: the BTB/POZ, 36. Cammas F, Oulad-Abdelghani M, Vonesch JL et al. Cell
KRAB, and SCAN domains. Mol Cell Biol 2001; 21(11):3609-3615. differentiation induces TIF1beta association with centromeric
23. Shannon M, Hamilton AT, Gordon L et al. Differential expansion heterochromatin via an HP1 interaction. J Cell Sci 2002; 115(Pt
of zinc-finger transcription factor loci in homologous human and 17):3439-3448.
mouse gene clusters. Genome Res 2003; 13(6A):1097-1110. 37. Zheng L, Pan H, Li S et al. Sequence-specific transcriptional
24. Dehal P, Predki P, Olsen AS et al. Human chromosome 19 and corepressor function for BRCA1 through a novel zinc finger protein,
related regions in mouse: conservative and lineage-specific evolution. ZBRK1. Mol Cell 2000; 6(4):757-768.
Science 2001; 293(5527):104-111. 38. Tan W, Zheng L, Lee WH et al. Functional dissection of
25. Abrink M, Ortiz JA, Mark C et al. Conserved interaction between transcription factor ZBRK1 reveals zinc fingers with dual roles in
distinct Kruppel-associated box domains and the transcriptional DNA-binding and BRCA1-dependent transcriptional repression. J
intermediary factor 1 beta. Proc Natl Acad Sci USA 2001; Biol Chem 2004; 279(8):6576-6587.
98(4):1422-1426. 39. Iuchi S. Three classes of C2H2 zinc finger proteins. Cell Mol Life
26. Agata Y, Matsuda E, Shimizu A. Two novel Kruppel-associated Sci 2001; 58(4):625-635.
box-containing zinc-finger proteins, KRAZ1 and KRAZ2, repress 40. Nielsen AL, Jorgensen P, Lerouge T et al. Nizp1, a novel multitype
transcription through functional interaction with the zinc finger protein that interacts with the NSD1 histone lysine
corepressor KAP-1 (TIF1beta/KRIP-1). J Biol Chem 1999; methyltransferase through a unique C2HR motif. Mol Cell Biol
274:16412-16422. 2004; 24(12):5184-5196.
27. Peng H, Begg GE, Schultz DC et al. Reconstitution of the 41. Yi Z, Li Y, Ma W et al. A novel KRAB zinc-finger protein, ZNF480,
KRAB-KAP-1 repressor complex: a model system for defining the expresses in human heart and activates transcriptional activities of
molecular anatomy of RING-B box-coiled-coil domain-mediated AP-1 and SRE. Biochem Biophys Res Commun 2004;
protein-protein interactions. J Mol Biol 2000; 295(5):1139-1162. 320(2):409-415.
28. Peng H, Begg GE, Harper SL et al. Biochemical analysis of the 42. Hering TM, Kazmi NH, Huynh TD et al. Characterization and
Kruppel-associated box (KRAB) transcriptional repression domain. J chondrocyte differentiation stage-specific expression of KRAB
Biol Chem 2000; 275(24):18000-18010. zinc-finger protein gene ZNF470. Exp Cell Res 2004;
29. Cammas F, Mark M, Dolle P et al. Mice lacking the transcriptional 299(1):137-147.
corepressor TIF1beta are defective in early postimplantation 43. Ganss B, Kobayashi H. The zinc finger transcription factor Zfp60
development. Development 2000; 127(13):2955-2963. is a negative regulator of cartilage differentiation. J Bone Miner Res
30. Germain-Desprez D, Bazinet M, Bouvier M et al. Oligomerization 2002; 17(12):2151-2160.
of transcriptional intermediary factor 1 regulators and interaction 44. Jheon AH, Ganss B, Cheifetz S et al. Characterization of a novel
with ZNF74 nuclear matrix protein revealed by bioluminescence KRAB/C2H2 zinc finger transcription factor involved in bone
resonance energy transfer in living cells. J Biol Chem 2003; development. J Biol Chem 2001; 276(21):18282-18289.
278(25):22367-22373. 45. Looman C, Mark C, Abrink M et al. MZF6D, a novel KRAB
31. Nielsen AL, Ortiz JA, You J et al. Interaction with members of the zinc-finger gene expressed exclusively in meiotic male germ cells.
heterochromatin protein 1 (HP1) family and histone deacetylation DNA Cell Biol 2003; 22(8):489-496.
are differentially involved in transcriptional silencing by members of 46. Takashima H, Nishio H, Wakao H et al. Molecular cloning and
the TIF1 family. Embo J 1999; 18(22):6385-6395. characterization of a KRAB-containing zinc finger protein, ZNF317,
32. Lechner MS, Begg GE, Speicher DW et al. Molecular determinants and its isoforms. Biochem Biophys Res Commun 2001;
for targeting heterochromatin protein 1-mediated gene silencing: 288(4):771-779.
direct chromoshadow domain-KAP-1 corepressor interaction is 47. Gebelein B, Urrutia R. Sequence-specific transcriptional repression
essential. Mol Cell Biol 2000; 20(17):6449-6465. by KS1, a multiple-zinc-finger-Kruppel-associated box protein. Mol
33. Schultz DC, Friedman JR, Rauscher FJ, 3rd. Targeting histone Cell Biol 2001; 21(3):928-939.
deacetylase complexes via KRAB-zinc finger proteins: the PHD and 48. Peng H, Zheng L, Lee WH et al. A common DNA-binding site for
bromodomains of KAP-1 form a cooperative unit that recruits a SZF1 and the BRCA1-associated zinc finger protein, ZBRK1. Cancer
novel isoform of the Mi-2alpha subunit of NuRD. Genes Dev 2001; Res 2002; 62(13):3773-3781.
15(4):428-443. 49. Iuchi S, Green H. Basonuclin, a zinc finger protein of keratinocytes
34. Schultz DC, Ayyanathan K, Negorev D et al. SETDB1: a novel and reproductive germ cells, binds to the rRNA gene promoter. Proc
KAP-1-associated histone H3, lysine 9-specific methyltransferase that Natl Acad Sci USA 1999; 96(17):9628-9632.
contributes to HP1-mediated silencing of euchromatic genes by 50. Tseng H, Biegel JA, Brown RS. Basonuclin is associated with the
KRAB zinc-finger proteins. Genes Dev 2002; 16(8):919-932. ribosomal RNA genes on human keratinocyte mitotic chromosomes.
35. Ayyanathan K, Lechner MS, Bell P et al. Regulated recruitment of J Cell Sci 1999; 112 Pt 18:3039-3047.
HP1 to a euchromatic gene induces mitotically heritable, epigenetic 51. Roberts SG. Mechanisms of action of transcripton activation and
gene silencing: a mammalian cell culture model of gene variegation. repression domains. Cell Mol Life Sci 2000; 57:1149-1160.
Genes Dev 2003; 17(15):1855-1869.
CHAPTER 22
Abstract
T
he SCAN domain is a highly conserved 84 residue motif Accompanying the C2H2 zinc finger elements in some of the
that is found near the N-terminus of a subfamily of C2H2 transcription factors are a series of extended sequence motifs.
zinc finger proteins. The SCAN domain, which is also These structural modules regulate subcellular localization, DNA
known as the leucine rich region (LeR), functions as a protein binding, and gene expression by controlling selective associa-
interaction domain, mediating self-association or selective asso- tion of the transcription factors with each other, or with other
ciation with other proteins. Bioinformatic approaches were used cellular components. In the C2H2 class of zinc fingers, these
to identify 71 SCAN domains in the human genome and to associated modules include the Kruppel-associated box (KRAB),3
define the structures of the members in the human SCAN do- the poxvirus and zinc finger (POZ) domain,4 which is also
main family. In addition to a single SCAN domain, the mem- known as the BTB domain (Broad-Complex, Tramtrack, and
bers of the family can have a variable number of zinc fingers Bric-a-brac),5 and the SCAN domain.6 These domains define
(2-22), a KRAB domain, as well as a novel N-terminal motif. subgroups within the C2H2 family and may provide insights
The genes encoding SCAN domains are clustered, often in tan- into the functions of the members of this large family of zinc
dem arrays, in both the human and mouse genomes and are finger transcription factors. Here we will place the SCAN do-
capable of generating isoforms that may affect the function of main family of transcription factors in perspective.
family members. Twenty-three members of the mouse SCAN
family appear to be orthologous with human family members, The SCAN Domain
and human-specific cluster expansions were observed. Although The SCAN domain was originally identified in the C2H2 zinc
the function of most of the family members is unknown, an finger transcription factor, ZNF174.6 The name was based on
overview of selected members of this group of transcription fac- the first letters found in some of the founding members of the
tors suggests that the SCAN domain family is involved in the family (SREZBP, Ctfin51, AW-1 (ZNF174), and Number 18).
regulation of growth factor gene expression, genes involved in
lipid metabolism, as well as other genes involved in cell survival Definition of the SCAN Domain
and differentiation. Analysis of the SCAN domain family using
Based on the analysis of a limited number of SCAN genes, the
phylogenetic and comparative genomics approaches reveals that
definition of the SCAN domain has varied, resulting in several
the SCAN family is vertebrate -specific. Remarkably, the SCAN
forms of the domain in online databases.7 Following a more com-
domains in lower vertebrates are not associated with C2H2 zinc
plete analysis of the SCAN domains found in the human ge-
finger genes, but are contained in large retrovirus-like
nome, the definition of the SCAN domain has been refined.8 As
polyproteins. Collectively, these studies define a large family of
will be discussed below, the human genome has at least 70 genes
transcriptional regulators that have rapidly expanded during re- containing the SCAN domain.8 An alignment of a select number
cent evolution. of these members is illustrated in Figure 1. A more extensive align-
ment of the highly homologous SCAN domains (as well as other
Introduction associated domains) can be found online at the Project Website
Zinc finger proteins are a large class of regulatory proteins. (http://www.scanfamily.org). Based on the sequence alignment,
The zinc finger is a small peptide domain with a secondary struc- as well as the biochemical properties of the recombinant protein
ture stabilized by a zinc ion interacting with the cysteine and detailed below, the SCAN domain is defined as 84 residues, be-
histidine residues of the finger.1 Although several types of zinc ginning with E-43 and ending with R-126 of ZNF174. In many
finger motifs have been identified,2 the C2H2 finger has emerged sequences, there are proline residues both before and after the
as the classical zinc finger and is described as CX2-4CX12HX2-6H domain, helping to delineate the boundaries of the predicted sec-
(reviewed in Chapter 1 of this book). ondary structural elements. Of the 84 residues that comprise the
Zinc Finger Proteins: From Atomic Contact to Cellular Function, edited by Shiro Iuchi and Natalie Kuldell.
©2005 Landes Bioscience/Eurekah.com and Kluwer Academic/Plenum Publishers.
The Superfamily of SCAN Domain Containing Zinc Finger Transcription Factors 157
with some, but not all, SCAN domains. Third, there is signifi-
cant variation in the relative affinities of the SCAN domains.
Although these studies with nine SCAN family members repre-
sent only a small fraction of the members of the family, they
may provide some initial guidelines for the assembly of a SCAN
protein network.
In addition to the findings from the mammalian
two-hybrid system, various SCAN domains have been used
as bait in the yeast two-hybrid system. These studies have
identified SCAN containing proteins, revealing new interac-
tions among family members. For example, a region of MZF1
(ZNF42) containing a SCAN domain interacted with RAZ1
(SCAN-related protein associated with MZF1B),13 which is
also known as SDP1 (SCAN-domain containing protein 1)
or SCAND1. Similarly, RAZ1 interacted with the SCAN Figure 2. A model network of SCAN family members. A summary of the
domain of ZNF202 and inhibited its repressive function. 14,15 reported interactions between SCAN family members is shown. Each
A summary of the existing interactions between SCAN fam- SCAN member is represented by a different color: ZNF174 (orange),
ily members is shown in the accompanying model (Fig. 2). ZNF192 (teal), ZNF24 (yellow), ZNF197 (red), MZF1 (blue), SCAND1
The majority of these observations outlined above were based (pink), ZNF202 (green). The SCAN domain of each protein is illus-
on studying isolated SCAN-SCAN associations. Whether trated with a circle; the C2H2 zinc finger motifs are indicated with a box.
most of these interactions occur in the context of full-length Note that some of the interactions were not studied within the context
proteins is uncertain. Additionally, understanding the nature of the full-length protein.
of the partnerships occurring in vivo will be key to defining
the family. diversity from a relatively small number of proteins.16 With a
Transcription factor dimerization can increase the selectivity group of regulators the size of the SCAN family (see below), the
of protein-DNA interactions and generate a large amount of range of potential dimers is tremendous. It will be important to
Figure 3. Structural features of the human SCAN domain family. The names and predicted structures of the 64 members of the human SCAN
domain family are schematically shown in order of their gene locations, starting with chromosome 1. The conserved domains are indicated with
a shaded box: novel region (red), SCAN (orange), KRAB-A (green), KRAB-B (yellow), integrase core domain (black), and C2H2 zinc finger domain
(blue). Zinc fingers displayed as an open white box are atypical and deviate from the C2H2 consensus. A dotted line in some of the structures
indicates that the predicted C-terminal domains are questionable because of discrepancies between the databases. The official gene symbols assigned
by NCBI (as of November 1, 2003) are shown. In instances where an official (or interim) name was not available, the genomic DNA clone
containing the SCAN sequence was given. Note that the SCAN domains of ZFP29 and SCAND1 are truncated. Modified from Sander et al, 2003.
The Superfamily of SCAN Domain Containing Zinc Finger Transcription Factors 159
determine the structural rules that regulate SCAN dimerization (e.g., ZNF197, ZFP95), while other members contain single or
specificity and obtain experimental data both in vitro and in duplicate pairs of fingers (e.g., FLJ12895, ZNF215). Since
vivo that verifies some of the partners. Based on this informa- Krüppel-type factors are frequently involved in DNA-binding, this
tion it may be possible to classify many of the SCAN family variation in zinc finger number and spacing may have an effect
members based on dimerization properties, as has been done on nucleotide base pair recognition.21 Application of the pro-
for the B-ZIP proteins.17 posed DNA-zinc finger rules,22,23 suggests that most of the
C-terminal zinc fingers of the family members can participate in
Superfamily of SCAN-Domain Containing DNA binding. Notably, members of the SCAN family with mul-
tiple adjacent C2H2 zinc fingers may have more than one DNA
Proteins binding activity, as is seen with other proteins with more than
The SCAN field was limited by an incomplete description four zinc fingers.24 Additionally, since C2H2 fingers have been
and definition of the members within the SCAN domain family. implicated in protein-protein interactions,25 it is possible that some
We reasoned that the release of the human genome could provide of the zinc fingers in the SCAN family members may interact
the key to defining the entire human SCAN domain family of with other members of the family, or with other target proteins.
transcription factors. Therefore, a bioinformatics approach was
used to determine the total number of SCAN domains in the
human genome and to define the structure and chromosome lo- KRAB Domain
cation of each human SCAN family member.8 Twenty-four members of the human SCAN domain family
To determine the total number of genes predicted to encode also contain a KRAB domain (Fig. 3). As outlined elsewhere in
SCAN domains in the human genome, both public and private this book, the KRAB domain is found at the N-terminus of ap-
human genome databases were screened with a representative proximately 200 C2H2 zinc finger proteins19,20 and mediates re-
human SCAN domain (ZNF174). Once the SCAN domains were pression of transcription.26,27 The KRAB domain spans about 75
identified, DNA sequences adjacent to the motif were annotated amino acids and has been divided into two subregions, desig-
to predict the cDNA structures.18 Analysis of expressed sequence nated A and B.3 It has been demonstrated that the KRAB-A do-
tag (EST) databases and serial analysis of gene expression (SAGE) main contains transcriptional repression activity. This domain is
libraries confirmed that most of the predicted SCAN family mem- predicted to form an amphipathic helix that interacts with KAP1
bers are actively expressed. (KRAB-associated protein-1, also known as TIF-1β, transcrip-
This screen revealed the presence of 71 SCAN domain- tion intermediary factor-1β). 28-34 KAP1 can enhance
containing genes in the human genome. Twenty-four of the SCAN KRAB-A-mediated repression by recruitment of a histone
domains were known, while 47 were previously unidentified. deacetylase complex containing N-CoR (nuclear receptor core-
Thus, the SCAN family constitutes approximately 10% of the pressor),35 or the Mi-2α subunit of the NuRD complex.33 KAP1
estimated 700 C2H2 zinc finger genes present in the human ge- can also associate with members of the heterochromatin protein
nome.19,20 Sixty-four of the 71 SCAN domains (90%) were found 1 (HP1) family, a family of nonhistone heterochromatin-associated
in complete open reading frames, suggesting that they are con- proteins with an established gene-silencing function.31
tained within functional genes. The remaining seven SCAN do- A sequence alignment of the KRAB domains from the SCAN
mains have termination codons that prematurely truncate the open family members is found on the project web site. As with other
reading frame, suggesting that these domains are contained in KRAB domain containing genes, the KRAB domain is located at
pseudogenes. The DNA sequences adjacent to the 64 SCAN do- the N-terminus of SCAN family members, usually C-terminal to
mains were computationally analyzed and manually annotated the SCAN domain. KRAB-containing proteins can be classified
to predict the cDNA structure and protein coding-potential for into three types:36 those that contain a KRAB-A domain alone,
each member of the human SCAN family. A schematic diagram those with both –A and –B domains, or those that have an –A
of the structural features of the human SCAN family, excluding domain and a divergent –B domain. Twelve members of the hu-
potential pseudogenes, is presented in Figure 3. This approach man SCAN family contain both KRAB-A and -B domains, while
successfully predicted the structures for all of the 24 previously another twelve members have a KRAB-A domain only. It is likely
reported members of the human SCAN family. that the SCAN domain will influence the function of the KRAB
domain, although this has not been extensively studied.
Components of SCAN Family Members
Novel Domain
in Addition to the SCAN Domain Thirty-four members of the SCAN family contain a novel re-
The SCAN family members were predicted to contain con- gion of homology that was found at the very N-terminus of the
served modular motifs lying outside of the SCAN domain. predicted protein (Fig. 3). A sequence alignment of a select num-
ber of the members is illustrated in Figure 4 and a full alignment
Zinc Fingers of all the members is found on the project web site. This region is
The majority of SCAN domains (91%) are associated with 13 residues in length, with the consensus EqEGLLiVKvEEe
C2H2 zinc finger domains. These 58 members contain a variable (where capital letters correspond to amino acids conserved over
number (2-22) of C2H2 zinc fingers, defined by the consensus 51%, and lowercase letters to amino acids conserved over 35%),
sequence CX2-4CX12HX2-6H (with X being any amino acid). All and may be similar to the previously described N-terminal acidic
of the C2H2 zinc finger motifs are located at the C-terminal end domain of the murine myeloid-specific zinc finger gene, MZF-2.37
of the SCAN proteins (Fig. 3). Most of these C2H2 zinc fingers A search of the human genome with the consensus sequence in-
are Krüppel-type, as defined by the conserved link (TGEKP(Y/ dicates that this region may be found only in members of the
F)X) between the histidine of the preceding finger with the cys- SCAN family.
teine of the next finger (H-C link). Many members contain The core of this conserved N-terminal domain in the SCAN
multiple-adjacent fingers that are arranged in clusters of 3 or more family members contains a sequence for conjugation to the small
160 Zinc Finger Proteins: From Atomic Contact to Cellular Function
SCAN-(C2H2)x
Figure 4. The conserved region N-terminal of the SCAN domain. The The first group of SCAN domain proteins consists of family
amino acid sequences of the N-terminal novel region from selected members with just a SCAN domain and a variable number of
members of the human SCAN family were aligned by consensus ClustalW C-terminal C2H2 zinc fingers (Fig. 3). There are at least 40 of
sequence alignment using MacVector 7.1.1 (Genetics Computer Group) these genes in the SCAN family, but only a few have been charac-
with a gap penalty setting of 10.0. Residues that share > 51% identity terized. Some of the better known representatives from this class
are outlined with dark shading, conserved amino acid differences are of genes are indicated in Table 1.
indicated by lighter shading. The consensus sequence for the novel The myeloid zinc finger gene (MZF1, also known as ZNF42)
region of the 34 human SCAN family members is shown below the is expressed preferentially in hematopoietic progenitor cells of
alignment, where capital letters correspond to amino acids conserved the myeloid lineage.42 Inactivation of the murine MZF1 gene
> 51% and lowercase letters to amino acids conserved > 35%. The results in a striking increase in hematopoetic progenitors, with
putative SUMO acceptor site is underlined. the eventual development of lethal myeloid neoplasias,43 suggest-
ing that MZF1 is involved in growth, differentiation and tum-
ubiquitin-related modifier, SUMO. Like ubiquitin, SUMO is a origenesis of myeloid progenitors. In both human and mouse
small polypeptide that is covalently attached to lysine residues in cells, several isoforms of MZF proteins are produced from a single
substrate proteins.38 SUMO is attached to substrate proteins by a gene by the use of alternative promoters and alternative splic-
series of enzymatic reactions similar to those involved in ing.37,44 Of the isoforms, MZF1B, and its murine counterpart
ubiquitination. Following cleavage to expose a C-terminal glycine, MZF-2A, seems to represent the major product of the gene.
SUMO is attached to a heterodimeric E1 SUMO activating en- MZF1B shares identity to the carboxy-terminus of MZF1, a
zyme (Aos1/Uba2), then transferred to an E2 SUMO conjugating C-terminal DNA binding domain consisting of 13 C2H2 zinc
enzyme (Ubc9), before being covalently attached to a lysine in the finger modules, but MZF1B encodes an additional amino termi-
substrate protein. A consensus SUMO acceptor site has been iden- nal extension that contains a novel N-terminal domain, a SCAN
tified consisting of the sequence ΨKXE, where Ψ is a large hydro- domain and a transactivation domain.
phobic amino acid and K is the site of SUMO conjugation.39 Con- To further characterize this member of the SCAN family, pro-
tained within the core of the novel N-terminal region teins that interact with several of the domains of MZF1B
(EqEGLLiVKvEEe) is the sequence VKxE, which fits the consen- (MZF-2A) have been defined. The SCAN domain of MZF1B
sus for a SUMO acceptor site. Additionally, a glycine residue is interacts with RAZ1,13 a member of the SCAN family that lacks
located five amino acids upstream from the acceptor lysine, consis- zinc fingers, as discussed below. The transactivation domain of
tent with the novel domain containing a sumoylation site (Fig. 4). MZF-2A works specifically in myeloid cells and can function as
an autonomous transactivation motif when fused to a heterolo-
Post-translational modification of the SCAN family member
gous DNA binding domain.45 In studies exploring the regula-
by SUMO may have a substantial effect on transcription factor
tion of the activity of the transactivation domain, three serine
activity. By analogy to other proteins, SUMO modification of
residues were found to be phosphorylated by ERK and p38 MAP
the SCAN family member could prevent modification by kinases.46 This suggests that the activity of the transactivation
ubiquitin at that position, potentially preventing degradation of region of MZF-2A is negatively regulated through phosphoryla-
the family member by the proteasome. SUMO may also com- tion by MAP kinases. To understand the mechanism of MZF-2A
pete for post-translational modification of lysines by acetylation. dependent transcriptional activation, a yeast two-hybrid Ras re-
Additionally, SUMO modification of the SCAN family member cruitment screen was done to look for interacting proteins.47 A
may disrupt protein-protein interactions with some proteins, while novel SWI2/SNF2-related protein, termed mammalian Domino,
promoting interaction with other proteins. In addition to play- was identified as a candidate MZF-2A interacting partner. Nota-
ing a role in regulating transcription factor activity, SUMO modi- bly, mDomino contains a SWI2/SNF2-type ATPase/helicase do-
fication also affects transcription factor subcellular localization. main, a SANT domain, and a glutamine rich domain. The
For example, SUMO-1 modified Sp3 accumulates at the nuclear C-terminal Q-rich domain physically associated with the
periphery and in nuclear dots whereas the transcriptionally active transactivation domain of MZF-2A and overexpression of
form lacking SUMO-1 has a more diffuse nuclear localization.40 mDomino enhanced MZF-2A mediated activation of a reporter
Although experimental verification will be required, the potential gene.47 Collectively, these findings suggest that an ATF-dependent
sumoylation site N-terminal to the SCAN domain may have an chromatin-remodeling complex interacts with MZF-2A to regu-
important role in the function of some of the family members. late gene expression in myeloid cells.
The Superfamily of SCAN Domain Containing Zinc Finger Transcription Factors 161
Table 1. Summary of representative SCAN family members from the three largest classes
SCAN-(C2H2)X
ZNF24 (ZNF191, KOX17, mZF-12) ZNF174, ZNF192, Binding to TCAT repeat in tyrosine
ZNF202, SCAND1 hydroxylase gene
ZNF42 (MZF1, MZF1B, mMZF-2A) RAZ1, ERK, p38, mDomino Regulator of myeloid progenitor
differentiation/proliferation
ZNF174 (AW-1) ZNF24, ZNF192, ZNF197 Negative regulator of PDGF-B chain,
TGF-β1 expression
ZNF444 (EZF-2) ND Positive regulator of SREC gene
expression
SCAN-KRAB-(C2H2)X
ZNF192(LD5-1) ZNF24, ZNF174, ZNF197 ND
ZNF197 (ZnF20) ZNF174, ZNF192, VHL Negative regulator of HIF-1α transactivation
ZNF202 ZNF24, SCAND1 Negative regulator of genes involved in
lipid metabolism
ZNF263 (FPM315, mNT2) ND Negative regulator of α2 (XI) collagen
gene (Col11α-2)
SCAN
SCAND1 (SDP1, RAZ1, mPGC-2) MZF1 , ZNF24, ZNF202, Co-activator, adipocyte differentiation
PPARγ, NF-κB
SCAND2 ND ND
The official human gene symbol assigned by NCBI is given; alternate human and mouse (m) gene names are in parentheses. ND=
not determined.
To address the key issue of defining genes regulated by MZF1, domain and lacking the zinc fingers.53 The ZNF202 gene prod-
a DNA binding site was determined for the zinc fingers of MZF1 uct is a transcriptional repressor that binds to elements found
and consensus elements were found in the promoters of several predominantly in genes that participate in lipid metabolism, such
hematopoietic cell-specific genes, such as CD34, c-myb, lactoferin as the group of genes that comprise the apoA-I/C-III/A-IV/A-V
and myeloperoxidase.48,49 The MZF1 binding motif is also found gene cluster on chromosome 11, the apoE/C-I/C-IV/C-II gene
in the promoter of the telomerase reverse transcriptase gene,50 cluster on chromosome 19, phospholipid transfer protein, and a
and in an intron in the gene for the high affinity IgE receptor,51 series of enzymes involved in lipid processing, including lipopro-
where MZF1 is likely to function as a negative regulator for gene tein lipase, hepatic triglyceride lipase and lecithin cholesteryl es-
expression. Interestingly, MZF1 activates gene expression in cells ter transferase.53 Additional targets for repression by ZNF202
of the hematopoietic lineage, while in nonhematopoietic cells it include the ATP-binding cassette A1 (ABCA1) and ABCG1.15
functions as a transcriptional repressor.42 It is possible that other ABCA1 is a key regulator of plasma HDL levels, while ABCG1
SCAN family members may be both positive and negative regu- supports lipid efflux in human macrophages. ZNF202, which
acts as a transcriptional repressor of both genes, is able to reduce
lators of target gene expression, depending upon the cell type.
phospholipid and cholesterol efflux in transiently transfected
macrophages, demonstrating the functional relevance of the
SCAN-KRAB-(C2H2)x SCAN family member.15 Additionally, ZNF202 expression is in-
A second class of SCAN family members consists of a versely regulated with respect to the expression of its target genes
SCAN-KRAB-(C2H2)x modular architecture (Fig. 3). There are ABCA1 and apoE during macrophage differentiation and foam
at least 24 genes in this class of SCAN family members. Some of cell formation.54 Collectively, these findings demonstrate that
the better known members of this class of SCAN family mem- ZNF202 regulates reverse-cholesterol transport and are consis-
bers are listed in Table 1. tent with the proposal that ZNF202 controls the balanced ex-
An interesting example of this subgroup of SCAN family mem- pression of genes involved in lipid metabolism.
bers is the candidate hypoalphalipoproteinemia susceptibility gene,
ZNF202. A low HDL cholesterol locus on chromosome 11q23 Isolated SCAN Domain Proteins
was identified that is distinct from the apoA-I/C-III/A-IV/AV For six members of the human SCAN family there is no cur-
cluster of genes.52 This new familial susceptibility locus for rent evidence for associated C2H2 zinc finger motifs (Fig. 3). These
hypoalphalipoproteinemia contains the SCAN family member, SCAN domain-only sequences might represent novel genes with-
ZNF202. The gene for ZNF202 encodes a protein predicted to out zinc fingers, or splice forms of larger transcripts that contain
contain a SCAN domain, an intact KRAB domain and eight C2H2 zinc finger motifs. Indeed, there is cDNA evidence that some
zinc-finger motifs. Two splice forms have been identified for SCAN-containing genes are alternatively spliced generating tran-
ZNF202, the m1 form encoding the full-length protein, and the scripts that lack the SCAN, KRAB, or zinc finger domains, as
m3 form encoding a smaller version containing only the SCAN discussed below.
162 Zinc Finger Proteins: From Atomic Contact to Cellular Function
Isoforms Lacking Zinc Fingers target, cis-acting regulatory element.62 Several members of the
Other SCAN family member genes also generate forms that SCAN family have been identified by this approach. EZF-2 (en-
lack the C-terminal zinc fingers. Thus in a general context, a dothelial zinc finger protein-2 or ZNF444) is a member of the
SCAN domain-only protein may dimerize with another member family with four zinc fingers and specifically targets the scavenger
of the family and alter nucleotide binding, since the resulting receptor expressed by endothelial cells.63 ZNF263 has nine zinc
heterodimer would lack the pair of DNA binding motifs (Fig. 6). fingers and a KRAB domain and interacts with an element in the
The expression of these alternate transcripts from SCAN con- promoter of the α2(XI) collagen gene.64 ZNF24 (ZNF191) in-
taining genes would, therefore, increase the available number of teracts with an intronic polymorphic TCAT repeat in the tyrosine
protein combinations within the family. Such forms may help hydroxylase gene, the rate-limiting enzyme in the synthesis of
regulate the function of the SCAN family members. catecholamines.65
One interesting example of this type of isoform is derived from
ZNF197 because it interacts with the von Hipple Lindau tumor Binding Site Selection
suppressor (pVHL). pVHL is a component of an E3 ubiquitin A second experimental approach to determine the genes
ligase and targets hypoxia-inducible factor-1α (HIF-1α) for regulated by a SCAN family member is to determine its DNA
ubiquitination and degradation under normoxic conditions.59 binding site and then utilize this information to identify can-
pVHL also directly inhibits HIF-1α transactivation by recruit- didate target genes. As outlined above, based on the binding
ing histone deacetylases. In a yeast two-hybrid screen for proteins site determination, ZNF202 has been shown to interact with
interacting with pVHL, an isoform of ZNF197 was identified and control the expression of a series of genes involved in lipid
which was given the name pVHL-associated KRAB-A metabolism, including the ATP binding cassette transporter
domain-containing protein (VHLaK).60 It contains a SCAN do- A1.15 Binding sites for MZF1 (ZNF42) were found in CD34,
main and a KRAB A-domain, but lacks the 22 zinc fingers present c-myb and myeloperoxidase, genes that are expressed in he-
in ZNF197 (Fig. 5). The KRAB A domain in VHLaK mediates matopoietic cells.
pVHL binding and functions as a transcriptional repression mod-
ule,60 consistent with previous reports that the KRAB domain of Overexpression
ZNF197 could repress transcription of a reporter gene.9 The In these studies, members of the SCAN family are
SCAN domain in VHLaK mediates homo-oligomerization and overexpressed with promoter-reporter constructs of candidate
enhances VHLaK repressive activity. pVHL can recruit VHLaK target genes. In this type of approach, ZNF174 selectively re-
to repress HIF-1α transcriptional activity and HIF-1α-induced pressed the expression of PDGF-B chain and TGF-β1
VEGF expression.60 Additionally, the ZNF197 isoform can re- promoter-reporters.6 Establishing the biological relevance of the
cruit both KAP1 and pVHL simultaneously, indicating that KAP1 findings is always the challenge with this type of experimental
may participate in pVHL mediated transcriptional repression of approach.
HIF-1α.60 To date, chromatin immunoprecipitation (ChIP) has not been
In summary, SCAN family members can generate isoforms used to verify the interaction of SCAN family members with DNA
lacking the SCAN domain as well as isoforms lacking the zinc binding sites in an authentic gene or to clone SCAN family mem-
fingers. Expression of these isoforms may be static, but more likely, ber target promoters.66 Additionally, it should be possible to iden-
the production of these isoforms may be biologically relevant. tify physiologic targets of SCAN family members by looking at
For example, it is possible that in tumors, relative to the corre- altered gene expression profiles using microarrays following
sponding normal tissues, the levels of the various isoforms of the overexpression of a SCAN family member.
SCAN family members may change.61
Genomics of the SCAN Family Members
Identification of Target Genes for SCAN
Family Members Human SCAN Family Genes
The biological function of most of the SCAN family mem- The genes for the human SCAN domain family have been
bers is not known, although the function of a few of the mem- mapped to specific human chromosome locations.8 Of the 71
bers has been inferred following targeted mutagenesis of the cor- SCAN-containing genes, 14 are isolated single genes. The major-
responding mouse gene, or following the identification of ity of genes (80%) are found in clusters on human chromosomes
interacting protein partners. Another way to determine the func- 3p21, 6p21.3, 7q22, 15q25, 16p13.3, 17p11.2, 18q12, 19q13.4
tion of a SCAN family member is to identify candidate target and Xq26, and some of these genes are arrayed in tandem. Some
genes that the member might regulate. Several experimental ap- of these locations are sites frequently disrupted and associated
proaches have been used to identify candidate target genes for with cytogenetic abnormalities. For example, ZNF197 is located
SCAN family members at 3p21, a region frequently involved in cytogenetic abnormali-
ties associated with epithelial malignancies of the kidney, lung,
One-Hybrid Approach thyroid and breast, as well as other tumors.67 The gene for MZF1
Perhaps most compelling are results from groups that have is located at the extreme end of the long arm of chromosome 19.
functionally characterized a specific regulatory element in a gene Since telomeres are known to shorten as cells age, the location of
of interest and then used that site in a “one-hybrid” screen to MZF1 at the telomere may make it vulnerable to dysregulation
identify the cognate regulatory protein. The yeast one-hybrid sys- or disruption during aging.68
tem is based on the principle of the yeast two-hybrid system and When the amino acid sequences of human SCAN domains
is used for isolating novel genes encoding proteins that bind to a were analyzed using a cladistics program, the majority of SCAN
164 Zinc Finger Proteins: From Atomic Contact to Cellular Function
Mouse SCAN Family Genes 16 and 17 (Fig. 7). These findings provide evidence for
An analysis of conserved segments between human and mouse human-specific cluster expansions of SCAN family members. This
chromosomes becomes a useful approach in the identification of argues that some genes within the SCAN family are lineage-specific
likely gene orthologs. To establish potential SCAN-containing and may have been selected independently since the divergence
gene orthologs, two approaches were taken. First, BLAST searches of primate and rodent lineages.
were performed against available mouse genome databases using Similar to the human SCAN family, a large percentage (> 75%)
representative mouse SCAN domains (Zfp38 or Skz1-pending). of the mouse SCAN domain-containing genes are predicted to
Second, conserved reference sets of genes that flank the genomic encode one SCAN domain and a variable number of C2H2 zinc
segment of the human SCAN gene(s) were located in the mouse fingers. Where orthologous relationships could be established, pre-
genome by Mouse Genome Informatics (MGI) and Ensembl. dicted structures of the human and mouse family members were
The mouse DNA region between the markers was then searched; highly similar in that they had comparable amino acid lengths and
significant matches to DNA sequences corresponding to SCAN the same number of conserved domains (KRAB and zinc fingers).
domains were identified.8 For example, the human SCAN gene, ZNF287, and its mouse
A reciprocal comparison of the human and mouse ortholog, Zfp287/Skat-2, encode predicted proteins of 754 and 758
SCAN-containing genes within regions of conserved synteny iden- amino acids, respectively, containing one SCAN domain, one KRAB
tified likely orthologs. Based on the SCAN-containing genes rep- domain, and 14 zinc fingers. Interestingly, in some pairs there is
resented in the Ensembl database, at least 23 of these members substantial conservation of the amino acids in the region of the
are represented by putative orthologs on conserved segments of C2H2 zinc fingers that determine the recognition specificity, sug-
the human chromosomes.8 In several cases, homologous SCAN gesting that the contact sites in target genes may be conserved.
family members within the human clusters were indistinguish-
able from each other when compared to the mouse and, as a re-
sult, an ortholog assignment was not possible. Interestingly, we SCAN Family Members Are Vertebrate
found human SCAN clusters that are represented by a smaller Specific
number of SCAN genes in the conserved syntenic regions of the Initial reports of the human genome suggested that SCAN
mouse. For example, the six clustered SCAN genes on human domain-containing C2H2 zinc finger proteins (SCAN-ZFP) are
chromosome 16p13.3 are represented by two clusters of only four unique to vertebrates.19,20 As expected, the genomes of inverte-
SCAN genes in the conserved segment of mouse chromosomes brate species, such as the fly (D. melanogaster), worm (C. elegans),
The Superfamily of SCAN Domain Containing Zinc Finger Transcription Factors 165
Figure 8. SCAN C2H2 proteins are vertebrate specific. The apparent number of SCAN domains found in databases of the worm, fly, zebrafish,
frog, chicken, mouse, and human are indicated in the first column. The apparent number of SCAN-containing retrovirus-like polyproteins,
compared to the apparent number of SCAN-containing C2H2 zinc finger proteins (ZFP), in the represented species is indicated in the second and
third columns, respectively (ND, not determined). Searches were limited by the availability of less complete sequence databases for the chicken
and frog. The predicted structure of a representative SCAN-containing protein is illustrated to the right of each number. The following conserved
regions are indicated: SCAN domain (gray), CCHC zinc knuckle (smallest box), retrovirus-like sequences (black), and C2H2 zinc finger (shaded).
The representative SCAN proteins are: human polyprotein KIAA1925 (Accession XP_166375), human zinc finger protein MGC4161 (Accession
NP_077279), mouse zinc finger protein Zfp29 (Accession NP_033579), frog polyprotein (Xenopus laevis, Accession S33978), zebrafish polyprotein
(Sanger Centre Accession dZ37K09). A color version of this figure can be viewed at http://www.eurekah.com/chapter.php?chapid=
1765&bookid=124&catid=30.
and yeast (S. cerevisiae), do not contain SCAN domains. Updated A number of interesting questions remain about the SCAN
databases were searched for predicted SCAN domains outside of domain, including a detailed understanding of its structure,
the human and mouse genomes. The search was limited by the mechanism of partner choice and the nature of interacting pro-
content available in large-scale sequence databases that are less com- teins. Additionally, identifying the target genes regulated by this
plete than their human and mouse counterparts. Nevertheless, ap- family will be useful in determining the function of this group of
parent SCAN domains were found in the genomes of a number of zinc finger transcription factors. Like the POZ/BTB and KRAB
vertebrate organisms including monkey, cow, pig, mouse, rat, frog domains, the presence of a SCAN domain may provide some
and several types of fish. Representative examples of SCAN do- mechanistic insights into the large number of C2H2 type zinc
mains from each species are provided in Sander et al 2003. finger proteins.
7. Stone JR, Maki JL, Blacklow SC et al. The SCAN domain of 31. Ryan RF, Schultz DC, Ayyanathan K et al. KAP-1 corepressor protein
ZNF174 is a dimer. J Biol Chem 2002; 277:5448-5452. interacts and colocalizes with heterochromatic and euchromatic HP1
8. Sander TL, Stringer KL, Maki JL et al. The SCAN domain defines proteins: A potential role for Krüppel-associated box-zinc finger
a large family of zinc finger transcription factors. Gene 2003; proteins in heterochromatin-mediated gene silencing. Mol Cell Biol
310:29-38. 1999; 19:4366-4378.
9. Pengue G, Calabrò V, Bartoli PC et al. Repression of transcriptional 32. Lechner MS, Begg GE, Speicher DW et al. Molecular determinants
activity at a distance by the evolutionary conserved KRAB domain for targeting heterochromatin protein 1-mediated gene silencing:
present in a subfamily of zinc finger proteins. Nucleic Acids Res Direct chromoshadow domain-KAP-1 corepressor interaction is
1994; 22:2908-2914. essential. Mol Cell Biol 2000; 20:6449-6465.
10. Bertin J, Nir WJ, Fischer CM et al. Human CARD4 protein is a 33. Schultz DC, Friedman JR, Rauscher III FJ. Targeting histone
novel CED-4/Apaf-1 cell death family member that activates deacetylase complexes via KRAB-zinc finger proteins: The PHD and
NF-kappaB. J Biol Chem 1999; 274:12955-12958. bromodomains of KAP-1 form a cooperative unit that recruits a
11. Schuster JM, Nelson PS. 2000. Toll receptors: An expanding role novel isoform of the Mi-2a subunit of NuRD. Genes Dev 2001;
in our understanding of human disease. J Leukoc Biol 2000; 15:428-443.
67:767-73. 34. Schultz DC, Ayyanathan K, Negorev D et al. SETDB1: A novel
12. Williams AJ, Blacklow SC, Collins T. The zinc finger-associated KAP-1-associated histone H3, lysine 9-specific methyltransferase that
SCAN box is a conserved oligomerization domain. Mol Cell Biol contributes to HP1-mediated silencing of euchromatic genes by
1999; 19:8526-8535. KRAB zinc-finger proteins. Genes Dev 2002; 16:919-932.
13. Sander TL, Haas AL, Peterson MJ et al. Identification of a novel 35. Underhill C, Qutob MS, Yee SP et al. A novel nuclear receptor
SCAN box-related protein that interacts with MZF1B: The corepressor complex, N-CoR, contains components of the
leucine-rich SCAN box mediates hetero- and homoprotein mammalian SWI/SNF complex and the corepressor KAP-1. J Biol
associations. J Biol Chem 2000; 275:12857-12867. Chem 2000; 275:40463-40470.
14. Schumacher C, Wang H, Honer C et al. The SCAN-domain 36. Mark C, Åbrink M, Hellman L. Comparative analysis of KRAB
mediates selective oligomerization. J Biol Chem 2000; zinc finger proteins in rodents and man: Evidence for several
275:17173-17179. evolutionarily distinct subfamilies of KRAB zinc finger genes. DNA
15. Porsch-Özcürümez M, Langmann T, Heimer S et al. The zinc finger Cell Biol 1999; 18:381-396.
protein 202 (ZNF202) is a transcriptional repressor of ATP binding 37. Murai K, Murakami H, Nagata S. A novel form of the
cassette transporter A1 (ABCA1) and ABCG1 gene expression and a myeloid-specific zinc finger protein (MZF-2). Genes to Cells 1997;
modulator of cellular lipid efflux. J Biol Chem 2001; 276:12427-12433. 2:581-591.
16. Lamb P, McKnight SL. Diversity and specificity in transcriptional 38. Gill G. Post-translational modification by the small ubiquitin-related
regulation: The benefits of heterotypic dimerization. Trends Biochem modifier SUJO has big effects on transcription factor activity. Curr
Sci 1991; 11:417-422. Opin Genetics Dev 2003; 12:108-113.
17. Vinson C, Myakishev M, Acharya A et al. Classification of human 39. Rodriguez MS, Dargemont C, Hay RT. SUMO-1 conjugation in
B-ZIP proteins based on dimerization properties. Mol Cell Biol 2002; vivo requires both a consensus modification motif and nuclear
22:6321-6335. targeting. J Biol Chem 2001; 276:12654-12659.
18. Zhang MQ. Computational predication of eukaryotic protein-coding 40. Ross S, Best JL, Zon LI et al. SUMO-1 modification represses Sp3
genes. Nature Rev Genet 2002; 3:698-702. transcriptional activation and modulates its subnuclear localization.
19. Venter JC, Adams MD, Myers EW et al. The sequence of the human Mol Cell 2002; 10:831-842.
genome. Science 2001; 291:1305-1351. 41. Collins T, Stone JR, Williams AJ. All in the family: The BTB/POZ,
20. Lander ES, Linton LM, Birren B et al. Initial sequencing and analysis KRAB, and SCAN domains. Mol Cell Biol 2001; 21:3609-3615.
of the human genome. Nature 2001; 409:860-921. 42. Hromas R, Davis B, Rauscher 3rd FJ et al. Hematopoietic
21. Nagaoka M, Nomura W, Shiraishi Y et al. Significant effect of linker transcriptional regulation by the myeloid zinc finger gene, MZF-1.
sequence on DNA recognition by multi-zinc finger protein. Biochem Curr Top Microbiol Immunol 1996; 211:159-64.
Biophys Res Comm 2001; 282:1001-1007. 43. Gaboli M, Kotsi PA, Gurrieri C et al. Mzf1 controls cell proliferation
22. Desjarlais JR, Berg JM. Toward rules relating zinc finger protein and tumorigenesis. Genes Dev 2001; 15:1625-1630.
sequences and DNA binding site preferences. Proc Natl Acad Sci 44. Peterson MJ, Morris JL. Human myeloid zinc finger gene MZF
USA 1992; 89:7345-7349. produces multiple transcripts and encodes a SCAN box protein. Gene
23. Suzuki M, Gerstein M, Yagi N. Stereochemical basis of DNA 2000; 254:105-18.
recognition by Zn fingers. Nucleic Acids Res 1994; 22:3397-3405. 45. Murai K, Murakami H, Nagata S. Myeloid-specific transcriptional
24. Iuchi S. Three classes of C2H2 zinc finger proteins. Cell Mol Life activation by murine myeloid zinc-finger protein 2. Proc Natl Acad
Sci 2001; 58:625-635. Sci USA 1998; 95:3461-3466.
25. Makay JP, Crossley M. Zinc fingers are sticking together. Trends 46. Ogawa H, Murayama A, Nagata S et al. Regulation of myeloid zinc
Biochem Sci 1998; 23:1-4. finger protein 2A (MZF-2A) transactivation activity through
26. Margolin JF, Friedman JR, Meyer WK et al. Krüppel-associated boxes phosphorylation by MAP kinases. J Biol Chem 2003a;
are potent transcriptional repression domains. Proc Natl Acad Sci 278:2921-2927.
USA 1994; 91:4509-4513. 47. Ogawa H, Ueda T, Aoyama T et al. A SW12/SNF2-type ATPase/
27. Witzgall R, O’Leary E, Leaf A et al. The Kruppel-associated box-A helicase protein, mDomino, interacts with myeloid zinc finger protein
(KRAB-A) domain of zinc finger proteins mediates transcriptional 2A (MZF-2A) to regulate its transcriptional activity. Genes to Cells
repression. Proc Natl Acad Sci USA 1994; 91:4514-4518. 2003b; 8:235-329.
28. Friedman JR, Fredericks WJ, Jensen DE et al. KAP-1, a novel 48. Morris JF, Hromas R, Rauscher FJ. Characterization of the
corepressor for the highly conserved KRAB repression domain. Genes DNA-binding properties of the myeloid zinc finger protein MZF1:
Dev 1996; 10:2067-2078. Two independent DNA-binding domains recognize two DNA
29. Kim SS, Chen YM, O’Leary E et al. A novel member of the RING consensus sequences with a common G-rich core. Mol Cell Biol
finger family, KRIP-1, associates with the KRAB-A transcriptional 1994; 14:1786-1795.
repressor domain of zinc finger proteins. Proc Natl Acad Sci USA 49. Morris JF, Rauscher FJ, Davis B et al. The myeloid zinc finger gene,
1996; 93:15299-15304. MZF-1, regulates the CD34 promoter in vitro. Blood 1995;
30. Moosmann P, Georgiev O, Le Douarin B et al. Transcriptional 86:3640-3647.
repression by RING finger protein TIF1β that interacts with the 50. Fujimoto K, Kyo S, Takakura M et al. Identification and
KRAB repressor domain of KOX1. Nucleic Acids Res 1996; characterization of negative regulatory elements of the human
24:4859-4867. telomerase catalytic subunit (hTERT) gene promoter: Possible role
of MZF-2 in transcriptional repression of hTERT. Nucleic Acids
Res 2000; 28:2557-2562.
The Superfamily of SCAN Domain Containing Zinc Finger Transcription Factors 167
51. Takahashi K, Nishiyama C, Hasegawa M et al. Regulation of the 61. Gonsky R, Knauf JA, Elisei R et al. Identification of rapid turnover
high affinity IgE receptor β-chain gene expression via an intronic transcripts overexpressed in thyroid tumors and thyroid cancer cell
element. J Immunol 2003; 171:2478-2484. lines: Use of a targeted differential RNA display method to select
52. Kort EN, Ballinger DG, Ding W et al. Evidence of linkage of familial for mRNA subsets. Nucleic Acids Res 1997; 25:3823-3831.
hypoalphalipoproteinemia to a novel locus on chromosome 11q23. 62. Wang MM, Reed RR. Molecular cloning of the olfactory neuronal
Am J Hum Genet 2000; 66:1845-1856. transcription factor Olf-1by genetic selection in yeast. Nature 1993;
53. Wagner S, Hess MA, Ormonde-Hanson P et al. A broad role for 364:121-126.
the zinc finger protein ZNF202 in human lipid metabolism. J Biol 63. Adachi H, Tsujimoto M. Characterization of the human gene
Chem 2000; 275:15685-15690. encoding the scavenger receptor expressed by endothelial cell and its
54. Langmann T, Schumacher C, Morham SG et al. ZNF202 is inversely regulation by a novel transcription factor, endothelial zinc finger
regulated with its target genes ABCA1 and apoE during macrophage protein-2. J Biol Chem 2002; 277:24014-240212.
differentiation and foam cell formation. J Lipid Res 2003; 64. Tanaka K, Tsumaki N, Kozak C et al. A kruppel-associated box-zinc
44:968-977. finger protein, NT2, represses cell-type-specific promoter activity of
55. Castillo G, Brun RP, Rosenfield JK et al. An adipogenic cofactor the α2(XI) collagen gene. Mol Cell Biol 2002; 22:4256-4267.
bound by the differentiation domain of PPARgamma. EMBO J 65. Albanese V, Biguet NF, Kiefer H et al. Quantitative effects of gene
1999; 18:3676-3687. silencing by allelic variation at a tetranucleotide microsatellite. Hum
56. Babb R, Bowen BR. SDP1 is a peroxisome-proliferator-activated Mol Genet 2001; 10:1785-1792.
receptor gamma 2 coactivator that binds through its SCAN domain. 66. Weinmann AS, Farnham PJ. Identification of unknown target genes
Biochem J 2003; 270:719-727. of human transcription factors using chromatin immunoprecipitation.
57. Suzawa M, Takada I, Yanagisawa J et al. Cytokines suppress Methods 2002; 26:37-47.
adipogenesis and PPAR-gamma function through the TAK1/TAB1/ 67. Imreh S, Klein G, Zabarovsky ER. Search for unknown
NIK cascade. Nat Cell Biol 2003; 5:224-230. tumor-antagonizing genes. Genes Chromosomes Cancer 2003;
58. Sander TL, Morris JF. Characterization of the SCAN box encoding 38:307-321.
RAZ1 gene: Analysis of cDNA transcripts, expression, and cellular 68. Hoffman SM, Hromas R, Amemiya C et al. The location of MZF-1
localization. Gene 2002; 293:53-64. at the telomere of human chromosome 19q makes it vulnerable to
59. Safran M, Kaelin WG. HIF hydroxylation and the mammalian degeneration in aging cells. Leuk Res 1996; 20:281-283.
oxygen-sensing pathway. J Clin Invest 2003; 111:779-783. 69. Dehal P, Predki P, Olsen AS et al. Human chromosome 19 and
60. Li Z, Wang D, Na X et al. The VHL protein recruits a novel related regions in mouse: Conservation and lineage-specific evolution.
KRAB-A domain protein to repress HIF-1alpha transcriptional Science 2001; 293:104-111.
activity. EMBO J 2003; 22:1857-1867.
CHAPTER 23
Abstract
S
p1, a triple C2H2 zinc finger protein, has been studied for polyglutamine-rich regions are transcription factors. Therefore,
more than two decades and biochemical features of the one possibility is that this tract plays a role in gene transcription.
protein have been well documented. However, its biological Through the use of two-hybrid screens, a number of proteins
roles in humans are still ambiguous. Recent studies on have been identified that interact with proteins containing
Huntington’s disease have shown that the mutant huntingtin polyglutamines; some of these interactions are sensitive to
proteins with expanded polyglutamine tracts interact with tran- polyglutamine-tract length. A recent survey of all S. cerevisiae pro-
scription factors, such as TFIID and Sp1, and have provided teins containing polyglutamine tracts showed that a significant
new insights into mechanisms of gene transcription as well as majority of these tracts are in proteins involved in transcription.
Huntington’s disease. Here, I describe aspects of Sp1 function in Thus, when the mutant form of the HD gene expresses an ab-
the context of Huntington’s disease research. normally long polyglutamine moiety, it might resemble a tran-
scription factor.14-16
Huntington’s Disease
Huntington’s disease (HD) is an autosomal, dominantly in- Nuclear Localization of Mutant Huntingtin
herited disorder characterized phenotypically by chorea, invol- Normal huntingtin is localized in the cytoplasm,17,18 but
untary movements, dystonia, intellectual impairment and emo- mutant huntingtin in addition to being found in the cytoplasm
tional disturbances.1 The gene, huntingtin, which is mutated in is also found localized in the nucleus. The processes governing
HD patients, was mapped to chromosome 4p16.3 in 1983 and nuclear translocation are unclear, although higher numbers of
cloned a decade later.2 The mutation is an expanded polyglutamine polyglutamine residues produce more nuclear localization in a
repeat, (CAG)n, within exon 1 of the gene. The number of dis- variety of experimental systems.19-25 Although polyglutamine re-
eases identified to be caused by (CAG)n (in general n=35 repeats) peat length seems to be one feature governing nuclear entry of
continues to grow and a common mechanism could underlie these huntingtin, another is the length of the whole molecule. Shorter
diseases. Most of the diseases caused by expanded CAG repeats, fragments of huntingtin are translocated to the nucleus more ef-
(CAG) n , share common features, which include ficiently, with longer fragments tending to form aggregates in the
neurodegeneration, a dominant pattern of inheritance and ge- cytoplasm.19-24 In transgenic mouse models, lines that synthesize
netic anticipation.3-6 The CAG trinucleotide repeats in all these truncated forms of huntingtin26,27 demonstrate more nuclear in-
genes are found in the coding region and are translated into a clusion formation and more abnormal phenotypic behavior than
string of glutamines. Despite the widespread tissue distribution transgenic lines that synthesize full-length huntingtin.25,28,29 Simi-
of the protein for each of these genes, the affected region is pri- lar results have been obtained in cellular models and in Droso-
marily the brain, and the regions of neuronal loss are somewhat phila models.29,30 It is not clear if huntingtin possesses active
selective and specific for each given disease.7 nuclear localization signals (NLS). However, in cellular experi-
In HD, the number of CAG repeats ranges from 6-35 whereas mental models, addition of nuclear export signals greatly reduces
in individuals affected by HD the repeat length ranges from nuclear accumulation of huntingtin. Nuclear localization of
40-121. The age of onset of HD is inversely correlated with huntingtin corresponded to increased cellular toxicity, although
CAG-repeat length.1 Although the gene encoding huntingtin is the mechanism by which this causes cellular damage is not yet
expressed ubiquitously,8 selective cell loss and fibrillary astrocytosis clear.31-33 It has been demonstrated that the proteolytic cleavage
is observed in the brain, particularly in the caudate and putamen of a glutathione-S-transferase-huntingtin fusion protein leads to
of the striatum, and to a lesser extent in hippocampus and sub- the formation of insoluble high molecular weight protein aggre-
thalamus. In addition, cerebellar degeneration was also observed gates only when the polyglutamine expansion is in the pathogenic
in post-mortem examination of juvenile-onset cases.9-13 range. 34 Furthermore, electron microscopic studies of
polyglutamine aggregates revealed a fibrillary structure that is remi-
niscent of scrapie prions and -amyloid fibrils in Alzheimer’s dis-
Mutant Huntingtin As “a Transcription ease. So far, similar aggregates have been found in human brains
Factor” affected with CAG-repeat disorders, in animal models for
The functional significance of the expanded polyglutamine tract CAG-repeat-associated diseases and in cell-culture models for these
is not well understood. Many other proteins containing disorders. It is becoming evident that the formation of intranuclear
Zinc Finger Proteins: From Atomic Contact to Cellular Function, edited by Shiro Iuchi and Natalie Kuldell.
©2005 Landes Bioscience/Eurekah.com and Kluwer Academic/Plenum Publishers.
Sp1 and Huntington’s Disease 169
and perinuclear aggregates is a characteristic feature of CAG-repeat symptoms or observable cell loss, receptor decreases are not merely
disorders, and might act as a marker of the disease process. How- epiphenomena of disease progression; supporting the notion that
ever, the direct link to the disease process is still elusive. On the neuronal dysfunction is just as relevant for the etiology of symp-
basis of HD mouse-model studies,31 SCA1 mouse-model stud- toms as neuronal death. Especially prominent are early decreases
ies35 and the above in vitro culture studies,36,37 evidence is accu- in the levels of dopamine D1 and D2 receptors, both of which
mulating to suggest that the formation of aggregates is not re- regulate gene expression, suggesting that certain alterations in
quired for the initiation of cell death. On the contrary, it has mRNA levels observed in HD are secondary to the loss of neu-
been proposed that aggregate formation might actually have a rotransmitter receptors.50,51 Although decreased mRNA levels
protective role.37 However, it remains to be determined what the suggest that receptor alteration occurs at the level of gene tran-
true role of inclusions is in all CAG-repeat-associated disorders, scription, these observations cannot exclude the possibility that
particularly HD. these changes occur at the level of mRNA degradation. However,
the observation that many mRNA transcripts are unchanged ar-
Interactions of Huntingtin with Transcription gues against a general defect with mRNA processing. In HD mice
crossed with caspase-1 dominant negative mice, delayed symp-
Factors tom onset and prolonged survival correspond not only to delayed
The combination of aberrant nuclear localization and altered nuclear inclusion formation, but also to delayed loss of neurotrans-
protein-protein interactions of mutant huntingtin suggests that mitter receptor binding and a delayed decrease of the correspond-
mutant huntingtin might interact with nuclear proteins. Mutant ing receptor mRNAs.52
huntingtin can recruit transcription factors into aggregates. For Gene expression arrays on DNA microchips have recently been
example, recruitment of TATA-box binding protein (TBP) into used to examine the scope of transcriptional changes in transgenic
aggregates has been shown in vitro as well as in human HD HD mice.53 Using the murine Mu6500 chip (Affymetrix, Santa
post-mortem brains.38 Mutant huntingtin has also been shown Clara, CA, USA) mRNA transcripts from striata of 6- and
to recruit CREB-binding protein (CBP) into aggregates in vitro39 12-week-old R6/2 mice were compared with striatal transcripts
and CBP is recruited into aggregates with the androgen receptor from age-matched controls. The results showed that the genes
in cell culture, in a mouse model of spinobulbar muscular atro- whose expression levels were altered were limited to several key
phy (SBMA), and in SBMA samples from humans.40 Similarly, molecular systems. The affected systems included neurotransmitter
atrophin, the abnormal protein in the polyglutamine disease, receptors, intracellular signaling mechanisms, retinoic acid recep-
dentatorubropallidoluysian atrophy (DRPLA), can recruit Eto. tor machinery and calcium homeostasis systems. Overall, only
Eto is a transcriptionally active molecule, which is part of the 1.2% of the genes examined had significantly altered levels of
histone deacetylase complex involved in transcriptional repres- expression. The pattern of affected genes shows that a single mu-
sion.41 In addition, an interaction between the N-terminal patho- tant protein affects the expression of many different groups of
genic region of huntingtin and p53 has been shown in vitro by genes.
coimmunoprecipitation and observation of coaggregation within Interestingly, when we examined the known regulatory se-
inclusions generated in cell culture.42 Interactions of mutant quences of the genes identified by the DNA microchips it be-
huntingtin with transcriptional repressors have also been shown. came apparent that all of them contained binding sites for tran-
A yeast two-hybrid system was used to identify an interaction of scription factor Sp1 in their TATA-less promoters. Huntingtin
huntingtin with the transcription factor, nuclear corepressor physically interacts with Sp1 and its cognate coactivator TAF130
(N-CoR)43 More recently, huntingtin has been shown to inhibit and inhibits Sp1-mediated transcription of the D2 dopamine
acetylation of histones via its association with CBP,44 and inter- receptor gene (Fig. 1).54,55
fere with CBP- and BDNF-mediated transcription.45-47
These interactions support the possibility that huntingtin and Transcription Factors Associated with
other CAG-repeat disease proteins might affect cellular function-
ing via alteration of the normal expression pattern of genes, Huntingtin and Sp1
through abnormal association, intracellular redistribution and/ A major target of Sp1 is the general transcription factor TFIID
or sequestration of crucial transcription factors. If huntingtin is that is recruited to DNA by Sp156-63 Once bound to DNA, TFIID
involved in regulating gene transcription it is important to deter- mediates initiation and elongation of transcription. TFIID is a
mine which may be target genes affected by normal and/or mu- multisubunit complex made up of the TATA box-binding pro-
tant huntingtin. Some obvious choices are genes altered in pa- tein and multiple64 TATA box-binding protein-associated factors
tients or in animal models of HD. (TAFs). Previous in vitro studies have pointed to the role of TAFs
as coactivators that interact with and mediate activation by se-
lected transcriptional regulators65-68 In addition, TAFIIs interact
Changes in Gene Expression in HD with core promoter elements and play a role in promoter recog-
Neurotransmitter alterations have been described in early-stage nition. The recent discovery of a subset of TAFs shared by TFIID
human HD autopsy material; many of these changes have been and chromatin remodeling complexes further points to multiple
confirmed in transgenic mouse models of HD.48,49 In transgenic functions of TAFs in regulating eukaryotic gene expression69-71
HD mice, neurotransmitter receptors are affected selectively; To determine contribution that TAFs make to transcriptional
downregulation of specific receptors argues against a generalized activation, several studies have examined the interplay between
problem with receptor production. This receptor downregulation transcriptional regulators and TAFs. A number of TAFs from
occurs only in transgenic lines with abnormal CAG-repeat num- Drosophila, yeast, and humans have been isolated and shown to
bers49 Decreases in receptor binding are corroborated by decreases interact with distinct activators in vitro72 In vivo studies in yeast
in receptor mRNA levels, indicating that affected receptors de- have indicated that TAFs are not required for the transcription of
crease in number, rather than changing their binding characteris- all genes73 but that certain TAF’s are required for the transcrip-
tics. Because receptor downregulation occurs before the onset of tion of distinct groups of genes 76-79 Genome-wide expression
170 Zinc Finger Proteins: From Atomic Contact to Cellular Function
profiling confirmed these findings; 16% of the genes in the yeast tiator element and its intrinsic enzymatic activities that
genome were affected by the inactivation of yTAFII145 (yeast post-translationally modify general transcription factors and his-
homologue of hTAF(II)250)80 Thus, TAFs serve important func- tones.86 It has recently been shown that huntingtin inhibits his-
tions for a subset of genes. tone acetyltransferase (HAT) activity of CBP.45 Therefore, it was
of interest to determine if huntingtin interferes with HAT activ-
TAF(II)130 and TAF(II)250 ity of TAFII250 and directly modulates the activity of TFIID
The involvement of one of the human TAFs (hTAF130) in complex.
activator-TAF interactions has been examined in detail. The essential nature of TAFII250 can be attributed to its broad
hTAFII130 has been shown to interact with cellular transcrip- requirement during RNA-pol-II-dependent transcription. The
tional activators such as Sp1, the cAMP-response element bind- signal instructing TFIID where to bind the genome comes from
ing protein (CREB), CCAAT-binding factor CBF, as well as activators, such as Sp1, that bind both cis-regulatory regions of
nuclear receptors, viral transcriptional regulators, and a transcrip- target genes and TFIID to recruit TFIID to particular genes.
tional repressor. Sp1 and CREB interact with one or more of the TAFII250 plays a fundamental role in this recruitment process,
four central glutamine-rich domains Q1-Q4 of hTAFII130 as serving as a binding partner for activators and as a scaffold for
determined by in vitro binding and yeast two-hybrid assays.81-85 other TAF subunits of TFIID. The arsenal of TAFII250 enzy-
Since both CREB/CBP40,42,45 and Sp154,55 have been shown matic functions, such as histone acetylation (HAT) and phos-
to interact with huntingtin, it was postulated that TAF(II)130 phorylation, may be employed at this point to provide accessibil-
might play an important role in huntingtin-mediated transcrip- ity to nucleosome-embedded core promoters. Requirements for
tional regulation54 It was found that both Sp1 and TAFII130 are individual TAFII250 activities could be gene specific, since im-
required for maximal protection from the effects of mutant pairment of phosphorylation or HAT activity affects the expres-
huntingtin on gene transcription and cell toxicity. sion of different genes.87 In yeast, genome-wide microarray analysis
We examined whether other TAFs, focusing on TAF(II)250 indicates that expression of 14-27% of RNA-pol-II-transcribed
may also be involved in huntingtin-mediated gene transcription. genes is downregulated more than twofold upon TAFII250 inac-
Recent excitement about TAF250 stems from its involvement in tivation.88 In ts13 hamster cells that harbor a temperaturesensitive
regulating the association of TFIID with the core promoter ini- allele of TAFII250 (mutated HAT domain of TAFII250), similar
analysis indicates that the transcription of 18% of genes tran-
scribed by RNA pol II is affected more than twofold at the
nonpermissive temperature.87 Mizzen et al using in-gel activity
assays showed that TAFII250 has HAT activity.89 The HAT do-
main maps to the central, most conserved portion of metazoan
TAFII250. Many studies have established a correlation between
acetylation of highly conserved lysine residues in the N-terminal
tails of histones and transcriptional activation.90 One model to
explain how histone acetylation affects gene expression proposes
that the extent of chromatin condensation is directly related to
the level of histone acetylation. Accordingly, hyperacetylation re-
duces the affinity of histone tails for DNA, resulting in less com-
pact chromatin and increased accessibility of transcription fac-
tors to DNA. TAFII250-mediated structural changes within
chromatin could allow initiation and elongation of transcription
of a subset of target genes.86,91
Figure 1. Mutant huntingtin disrupts transcriptional activation by Sp1
and TAFII130 in HD. The transcription factor Sp1 binds to DNA Mutant Huntingtin Inhibits Sp1 Acetylation by
elements called GC boxes in cellular promoters. A specific protein-protein TAF250
interaction between the glutamine-rich regions of Sp1 and the TAFII130 These detailed in vitro and in vivo analyses have uncovered a
subunit of TFIID is required for recruitment of the general transcrip- plethora of TAFII250 activities that together underscore the im-
tional machinery, which includes transcription factors TFIIA, B, D, E, portance of TAFII250 as a regulator of RNA pol II transcription.
F, and H. This glutamine interface serves to bridge Sp1 to the machinery In our most recent work, we showed that mutant huntingtin-
required to recruit RNA pol II. Once correctly targeted to the target mediated repression of gene transcription is rescued by an
gene, RNA pol II initiates transcription of an mRNA copy of the overexpression of normal TAF250 but not mutant forms of hu-
target gene. In HD, the glutamine expansion in huntingtin (Htt) dis- man TAF250 with mutated histone acetyltransferase domain
rupts transcriptional activation by Sp1 and TAFII130. An (HAT), suggesting that the effect of TAF250 on
amino-terminal fragment of mutant huntingtin, containing an expanded huntingtin-mediated repression depends on the HAT activity of
polyglutamine tract, accumulates in the nucleus. Here, this mutant TAF250. Furthermore, we found that mutant huntingtin inhib-
protein associates with Sp1 and TAFII130, preventing Sp1 from bind- its acetylation of Sp1 by TAF250, which in turn leads to decreased
ing to the GC box, and ultimately disrupting the ability of Sp1 and binding of Sp1 to DNA and decreased expression of Sp1-driven
TAFII130 to interact. Without proper targeting by the general tran- genes. These findings are particulary intriguing in light of recent
scriptional machinery, RNA pol II cannot properly locate the target reports that Sp1 can be acetylated in vivo and that HDAC in-
promoter region and the gene cannot be transcribed. The hibitors can augment Sp1 acetylation and interfere with Sp1/DNA
deregulated gene expression was noted in HD brain samples isolated binding.92,93
from presymptomatic HD patients suggesting that it may represent an These data suggests that huntingtin interacts with transcrip-
early step in the neurodegenerative process. tional coactivators in the context of TFIID complex and that Sp1
Sp1 and Huntington’s Disease 171
transcription plays a central role in the process. Therefore, we a decrease in Sp1 binding, suggesting the ratio change of the Sp
hypothesize that mutant huntingtin may alter the composition family of transcription factors may also play an important role in
and DNA contacts of transcriptional complexes in HD. It will of regulating transcription in post mitotic neurons. Since the zinc
interest to identify individual components in multiprotein com- finger motif in Sp1 that has been shown to interact with the
plexes in HD and examine the complexes in different brain re- huntingtin protein are conserved in Sp3 and Sp4, the zinger fin-
gions and brain samples from presymptomatic patients as well as ger motifs in Sp3 and Sp4 may also interact with huntingtin,
transgenic animals. For example, striatum may be compared with leading to interruption of Sp3 and Sp4 transcriptional function
hippocampus, cerebellum and cortex in presymptomatic and as well. Because of the distinct role for each Sp family of tran-
symptomatic mice or postmortem human brain samples. In ad- scription factors in neurons, the contribution of these interaction
dition, other Sp1 gene targets, such as Sp1-genes that were al- between huntingtin and different factors to the pathogenesis may
tered in HD (i.e., D2) and genes that were not altered (i.e., vary.
NMDAR1) should be analyzed. Such regional and gene-specific It is worth noting that in those cases that have key residues
approach may identify specific molecular alterations in transcrip- differing between members of the family, the DNA-binding speci-
tional complexes in presymptomatic and symptomatic HD. ficity is frequently altered. For example, Sp2, which has a leucine
In addition to Sp1, neuronal cells also express transcription residue within the first zinc-finger motif in place of the histidine
factors Sp3 and Sp4 that are structurely and functionally related found in the corresponding region of Sp1, preferentially recog-
to Sp1.94 Sp1, Sp3 and Sp4 comprise most of the GC box bind- nizes the GT box (5'-GGTGTGGGG-3'), found in many differ-
ing activity in post-mitotic neurons, suggesting that the Sp fam- ent promoters, rather than the GC box. Previous studies also sug-
ily of transcription factors is of importance for the normal func- gested that the linkers between the zinc-finger motifs contribute
tion and activity of post-mitotic neurons. Sp1, Sp3 and Sp4 to high-affinity binding of certain zinc-finger proteins. The linker
recognize and bind GC box with similar affinities dependent on regions of members of the Sp1-like family have several poten-
three highly conserved zinc finger motifs in the carboxyl termi- tially relevant differences, but it is currently not known if all of
nus. In addition, each of these Sp family proteins contains two these contribute to differences in DNA-binding activity.
glutamine-rich regions in the amino terminus that contribute to By regulating the expression of a large number of genes that
transcriptional activation. have GC-rich promoters, Sp1 and other zinc-finger transcription
regulators may take part in virtually all facets of cellular function,
Genetic Approaches for Understanding of including cell proliferation, apoptosis, differentiation, and neo-
Function of Sp1 and the Sp Protein Family plastic transformation. Individual members of the Sp1-like fam-
Research on Sp1 and other Sp proteins has made rapid ad- ily can function as activators or repressors depending on which
vances in the last few years, with more and more family members promoter they bind and the coregulators with which they inter-
being identified. Most of the knowledge is derived from biochemi- act. A long-standing research aim has been to define the mecha-
cal analyses, in vitro protein-DNA interaction studies and trans- nisms by which Sp1-like factors regulate gene expression and cel-
fection experiments. The knockouts published to date show that lular function in a cell- and promoter-specific manner. Gene arrays
individual family members do have specific functions in the studies in HD have shown that transcription of many Sp1 target
mouse. However, even knockout mice may not reflect the whole genes is altered by mutant huntingtin. Therefore, it will be of
truth because so little is known about the extent of overlapping interest to determine which are the specific cellular functions that
functions. For instance, inactivation of Sp1 was expected to have are altered by the dysfunction of Sp1-like factors in the context
dramatic effects at the cellular level since it was generally believed of HD. Moreover, it will be of interest to examine whether Sp1-like
that Sp1 is the essential transcription factor for expression of house- zinc finger proteins play a more general role in the pathogenesis
keeping genes. Possibly, Sp1 is indeed engaged in the expression of other polyglutamine disorders. For example, androgen recep-
of these genes in vivo, but functionally replaced by Sp3 under tor and Sp1 have been shown to interact with one another and
knockout conditions. In addition, since deletion of the Sp1 gene cooperatively regulate gene transcription suggesting that Sp1 tran-
leads to lethality early in development, the role of Sp1 in fully scription may be deregulated in spinobulbar muscular atrophy
developed tissues remains an open question. In this regard, it is (SBMA).
interesting to note that mutations of Sp1-binding sites in the LDL Conditional disruption of the Sp1 gene in specific tissues at
receptor and collagen type alpha genes have been associated with any given stage of development will be an important step to fur-
hypercholesterolaemia and osteoporosis in humans. ther unravel the physiological role of Sp1 in vivo. The identifica-
Although Sp1, Sp3 and Sp4 are highly related proteins, they tion of target genes presents another formidable challenge. Re-
may have distinct biological roles. Sp3, unlike Sp1 and Sp4 con- cent advances in the detection of differentially expressed genes
tains an inhibitory domain that suppresses transcriptional activa- and the development of DNA microchip arrays provide very use-
tion and may contribute to active repression. Whether Sp3 func- ful tools for high throughput analysis of target genes. This should
tions as activator or a repressor depends on promoter architecture be of great help in understanding the downstream effects that
and cell type. Although both Sp1 and Sp3 are ubiquitously ex- result in the knockout phenotypes.
pressed, Sp4 mRNA and protein are highly expressed in the cen- To fully understand the physiological function of the Sp pro-
tral nervous system and the brain. Gene disruption of Sp1, Sp3 teins, it will be equally important to understand the mechanisms
or Sp4 in mice produces different phenotypes. Interestingly, of their mode of action. Detailed characterization of the indi-
knockout of Sp4 genes in mice produces a complex phenotype vidual transcription factors, identification of specific interaction
that includes behavioral defects, supporting a role for Sp4 in regu- partners, careful analysis of protein modifications and the identi-
lating transcription of genes in the nervous system. In addition, fication of signals and transduction pathways by which these pro-
the relative contribution of Sp family of transcription factors to teins are regulated will be essential for a mechanistic understand-
GC box binding activity was observed to change during neu- ing of transcriptional control by this growing family of
ronal differentiation, with a increase in Sp3 and Sp4 binding and transcription factors. Another important area of research pertains
172 Zinc Finger Proteins: From Atomic Contact to Cellular Function
to the transcriptional regulatory function of the amino-terminal 19. Yang W et al. Aggregated polyglutamine peptides delivered to nuclei
domains of these proteins. What are the molecular mechanisms are toxic to mammalian cells. Hum Mol Genet 2002;
that regulate how these domains interact with coregulatory com- 11(23):2905-2917.
20. Chen S et al. Huntington’s disease age-of-onset linked to
plexes and thereby repress or activate gene expression? Further polyglutamine aggregation nucleation. Proc Natl Acad Sci USA 2002;
insights into the functions of different Sp1-like proteins have the 99(18):11884-11889.
potential to change the partially informative classification of these 21. Kalchman MA et al. Huntingtin is ubiquitinated and interacts with
proteins, which is currently based on primary structure. Lastly, be- a specific ubiquitin-conjugating enzyme. J Biol Chem 1996;
cause these proteins are important in morphogenesis, it is likely that 271:19385-19394.
they may play a significant role in the mechanisms underlying hu- 22. Burke JR et al. Huntingtin and DRPLA proteins selectively interact
with the enzyme GAPDH. Nat Med 1996; 2:347-350.
man diseases that are characterized by aberrant growth and differ- 23. Sittler A et al. SH3GL3 associates with the Huntingtin exon 1
entiation. Future studies of the Sp1-like proteins have a large poten- protein and promotes the formation of polygln-containing protein
tial for defining the machinery that not only regulates physiological aggregates. Mol Cell 1998; 2:427-436.
processes but may also modulate human diseases such as HD. 24. Faber PW et al. Huntingtin interacts with the WW domain proteins.
Hum Mol Genet 1998; 7:1463-1474.
25. Nasir J et al. Targeted disruption of the Huntington’s disease gene
Acknowledgement results in embryonic lethality and behavioral and morphological
I thank Naoko Tanese, Anne Young, Anthone Dunah, Hyun changes in heterozygotes. Cell 1995; 81:811-823.
Jeong, Libin Cui, Edith Wang, Maral Mouradian, Steven Hersch 26. Zeitlin SZ et al. Increased apoptosis and early embryonic lethality
and Robert Tjian for their contributions to the experiments de- in mice nullizygous for the Huntington’s disease gene homologue.
scribed in this chapter. Nat Genet 1995; 11:155-163.
27. Goldberg YP et al. Absence of the disease phenotype and
intergenerational stability of the CAG repeat in transgenic mice
References expressing the human Huntington disease transcript. Hum Mol
1. Vonsattel JP, DiFiglia M. Huntington’s disease. J Neuropathol Exp Genet 1996; 5:177-185.
Neurol 1998; 57:369-384. 28. Mangiarini L et al. Exon 1 of the HD gene with an expanded CAG
2. Huntington’s disease collaborative research group. A novel gene repeat is sufficient to cause a progressive neurological phenotype in
containing a trinucleotide repeat that is unstable in Huntington’s transgenic mice. Cell 1996; 87:493-506.
disease chromosomes. Cell 1993; 72:971-983. 29. Burright EN et al. SCA1 transgenic mice: A model for
3. Albin RL, Tagle DA. Genetics and molecular biology of Huntington’s neurodegeneration caused by an expanded CAG trinucleotide repeat.
disease. Trends Neurosci 1995; 18:11-14. Cell 1996; 82:937-948.
4. Warren ST. The expanding world of trinucleotide repeats. Science 30. Kawaguchi Y et al. CAG expansions in a novel gene for
1996; 271:1374-1375. Machado-Joseph disease at chromosome 14q32.1. Nat Genet 1996;
5. Zoghbi HY, Orr HT. Glutamine repeats and neurodegeneration. 8:221-228.
Annu Rev Neurosci 2000; 23:217-47. 31. Reddy PH et al. Bahavioural abnormalities and selective neuronal
6. Reddy PS, Housman DE. The complex pathology of trinucleotide loss in HD transgenic mice expressing mutated full-length HD
repeats. Curr Opin Cell Biol 1997; 9:364-372. cDNA. Nat Genet 1998; 20:198-202.
7. MacDonald ME et al. Huntington’s disease. Neuromolecular Med 32. Hodgson JG et al. Human huntingtin derived from YAC transgenes
2003; 4(1-2):7-20. compensates for loss of murine huntingtin by rescue of the embryonic
8. Strong TV et al. Widespread expression of the human and rat lethal phenotype. Hum Mol Genet 1996; 5:1875-1885.
Huntington’s disease gene in brain and nonneural tissues. Nat Genet 33. Davies SW et al. Formation of neuronal intranuclear inclusions
1995; 5:259-263. underlies the neurological dysfunction in mice transgenic for the
9. Spargo E et al. Neuronal loss in the hippocampus in Huntington’s HD mutation. Cell 1997; 90:537-548.
disease: Acomparison with HIV infection. J Neurol Neurosurg 34. Scherzinger E et al. Huntingtin-encoded polyglutamine expansions
Psychiatry 1993; 56:487-491. form amyloid-like protein aggregates in vitro and in vivo. Cell 1998;
10. Byers RK et al. Huntington’s disease in children. Neuropathologic 90:549-558.
study of four cases. Neurology 1973; 23:561-569. 35. Klement S et al. Ataxin-1 nuclear localization and aggregation: Roll
11. Lange HG et al. Morphometric studies of the neuropathological in polyglutamine-induced disease in SCA1 transgenic mice. Cell
changes in choreatic diseases. J Neurol Sci 1976; 28:401-425. 1998; 95:41-53.
12. Landwehrmeyer GB et al. Huntington’s disease gene: Regional and 36. Kim M et al. Mutant huntingtin expression in clonal striatal cells:
cellular expression in brain of normal and affected individuals. Ann Dissociation of inclusion formation and neuronal survival by caspase
Neurol 1995; 37:218-230. inhibition. J Neurosci 1998; 19:964-973.
13. Ferrante RJ et al. Heterogeneous topographic and cellular distribution 37. Saudou F et al. Huntingtin acts in the nucleus to induce apoptosis
of huntingtin expression in the normal human neostriatum. J but death does not correlate with the formation of intranuclear
Neurosci 1997; 17:3052-3063. inclusions. Cell 1998; 95:55-66.
14. Jou YS, Myers RM. Evidence from antibody studies that the CAG 38. Huang CC et al. Amyloid formation by mutant huntingtin:
repeat in the Huntington disease gene is expressed in the protein. Threshold, progressivity and recruitment of normal polyglutamine
Hum Mol Genet 1995; 4:465-469. proteins. Somat Cell Mol Genet 1998; 24:217-233.
15. Trottier Y et al. Cellular localization of the Huntington’s disease 39. Kazantsev A et al. Insoluble detergent-resistant aggregates form
protein and discrimination of the normal and mutated form. Nat between pathological and nonpathological lengths of polyglutamine
Genet 1995; 10:104-110. in mammalian cells. Proc Natl Acad Sci USA 1999; 96:11404-11409.
16. Cha J. Transcriptional dysregulation in Huntington’s disease. Trends 40. McCampbell ARV et al. Colocalization of CBP with expanded
Neurosci 2000; 23:87-392. polyglutamine containing androgen receptor. Am J Hum Genet 1999;
17. DiFiglia M et al. Huntington is a cytoplasmic protein associated 65 (Suppl:A):106.
with vesicles in human and rat brain neurons. Neuron 1995; 41. Ross CA et al. Polyglutamine pathogenesis. Philos Trans R Soc
14:1075-1081. London B Biol Sci 1999; 354:1005-1011.
18. Wood JD et al. Partial characterization of the murine huntingtin 42. Steffan JS et al. The Huntington’s disease protein interacts with p53
and apparent variations in the subcellular localization of huntingtin and CBP and represses transcription. Proc Natl Acad Sci USA 2000;
in human, mouse and rat brain. Hum Mol Genet 1996; 5:481-487. 97:6763-6768.
Sp1 and Huntington’s Disease 173
43. Boutell JM et al. Aberrant interactions of transcriptional repressor 67. Lee TI, Young, RA. Regulation of gene expression by TBP-associated
proteins with the Huntington’s disease gene product, huntingtin. proteins. Genes Dev 1998; 12:1398-1408.
Hum Mol Genet 1999; 8:1647-1655. 68. Verrijzer CP, Tjian R. TAFs mediate transcriptional activation and
44. Jones AL. The localization and interactions of huntingtin. Philos promoter selectivity. Trends Biochem Sci 1996; 21:338-342.
Trans R Soc Lond B Biol Sci 1999; 354:1021-1027. 69. Struhl K, Moqtaderi Z. The TAFs in the HAT. Cell 1998; 94:1-4.
45. Steffan JS et al. Histone deacetylase inhibitors arrest 70. Bjorklund S et al. Global transcription regulators of eukaryotes. Cell
polyglutamine-dependent neurodegeneration in Drosophila. Nature 1999; 96:759-767.
2001; 413:739-743. 71. Hampsey M. Molecular genetics of the RNA polymerase II general
46. Nucifora FC et al. Interference by huntingtin and atrophin-1 with transcriptional machinery. Microbiol Mol Biol Rev 1998; 62:465-503.
CBP-mediated transcription leading to cellular toxicity. Science 2001; 72. Sauer F, Tjian R. Mechanisms of transcriptional activation:
291:2423-2428. Differences and similarities between yeast, Drosophila, and man. Curr
47. Zuccato C et al. Loss of Huntingtin-mediated BDNF gene Opin Genet Dev 1997; 7:176-181.
transcription in Huntington’s disease. Science 2001; 293:493-498. 73. Moqtaderi Z et al. TBP-associated factors are not generally required
48. Augood SJ et al. Dopamine D1 and D2 receptor gene expression in for transcriptional activation in yeast. Nature 1996; 383:188-191.
the striatum in Huntington’s disease. Ann Neurol 1997; 42:215-221. 76. Walker SS et al. Yeast TAFII145 required for transcription of G1/S
49. Cha J-HJ et al. Altered brain neurotransmitter receptors in transgenic cyclin genes and regulated by the cellular growth state. Cell 1997;
mice expressing a portion of an abnormal human Huntington disease 90:607-614.
gene. Proc Natl Acad Sci USA 1998; 95:6480-6485. 77. Wang, EH, Tjian R. Promoter-selective transcriptional defect in cell
50. Berke JD et al. A complex program of striatal gene expression cycle mutant ts13 rescued by hTAFII250. Science 1994;
induced by dopaminergic stimulation. J Neurosci 1998; 263:811-814.
18:5301-5310. 78. Wang, EH et al. TAFII250-dependent transcription of cyclin A is
directed by ATF activator proteins. Genes Dev 1997; 11:2658-2669.
51. Gerfen CR et al. D1 and D2 dopamine receptor-regulated gene
79. Suzuki-Yagawa Y et al. The ts13 mutation in the TAF(II)250 subunit
expression of striatonigral and striatopallidal neurons. Science 1990;
(CCG1) of TFIID directly affects transcription of D-type cyclin genes
250:1429-1432. in cells arrested in G1 at the nonpermissive temperature. Mol Cell
52. Ona VO et al. Inhibition of caspase-1 slows disease progression in a Biol 1997; 17:3284-3294.
mouse model of Huntington’s disease. Nature 1999; 399:263-267. 80. Holstege FC et al. Dissecting the regulatory circuitry of a eukaryotic
53. Luthi-Carter R et al. Decreased expression of striatal signalling genes genome. Cell 1998; 95:717-728.
in a mouse model of Huntington’s disease. Hum Mol Genet in press. 81. Tanese N et al. Molecular cloning and analysis of two subunits of
54. Dunah AW, Jeong H, Griffin A et al. Sp1 and TAF130 transciptional the human TFIID complex: HTAFII130 and hTAFII100. Proc Natl
activity disrupted in early Huntington’s Disease. Science 2002; Acad Sci USA 1996; 93:13611-13616.
296(5576):2238-2243. 82. Hoey T et al. Molecular cloning and functional analysis of Drosophila
55. Li SH, Cheng AL, Zhou H et al. Interaction of Huntington disease TAF110 reveal properties expected of coactivators. Cell 1993;
protein with transcriptional activator Sp1. Mol Cell Biol 2002; 72:247-260.
22(5):1277-1287. 83. Rojo-Niersbach E et al. Genetic dissection of hTAF130 defines a
56. Kadonaga JT et al. Isolation of cDNA encoding transcription factor hydrophobic surface required for interaction with glutamine-rich
Sp1 and functional analysis of the DNA binding domain. Cell 1987; activators. J Biol Chem 1999; 274:33778-33784.
51:1079-1090. 84. Saluja D et al. Distinct subdomains of human TAFII130 are required
57. Dynan WS, Tjian R. Control of eukaryotic messanger RNA synthesis for interactions with glutamine-rich transcriptional activators. Mol
by sequence-specific DNA binding proteins. Nature 1985; Cell Biol 1998; 18:5734-5743.
316:774-778. 85. Ferreri K et al. The cAMP-regulated transcription factor CREB
58. Black AR et al. Sp1 and Kruppel-like factor family of transcription interacts with a component of the TFIID complex. Proc Natl Acad
factors in cell growth regulation and cancer. J Cell Physiol 2001; Sci USA 1994; 91:1210-1213.
188:143-160. 86. Wassarman DA et al. TAFII250: A transcription toolbox. J Cell Sci
59. Jackson SP, Tjian R. O-Glycosylation of eukaryotic transcription 2001; 114:2895-2902.
factors: Implications for mechanisms of transcriptional regulation. 87. O’Brien T et al. Different functional domains of TAFII250 modulate
Cell 1988; 55:125-133. expression of distinct subsets of mammalian genes. Proc Natl Acad
60. Hagen G et al. Cloning by recognition site screening of two novel Sci USA 2000; 97:2456-2461.
GT box binding proteins: A family of Sp1 related genes. Nucleic 88. Lee TI et al. Redundant roles for the TFIID and SAGA complexes
Acids Res 1992; 20:5519-5525. in global transcription. Nature 2000; 405:701-704.
61. Suske G. The Sp-family of transcription factors. Gene 1999; 89. Mizzen CA et al. Transcription. New insights into an old
238:291-300. modification. Science 2000; 289:2290-2291.
62. Sjottem E et al. The promoter activity of long terminal repeats of 90. Strahl BD et al. The language of covalent modification. Nature 2000;
403:41-45.
the HERV-H family of human retrovirus-like elements is critically
91. Dunphy EL et al. Requirement for TAFII250 acetyltransferase activity
dependent on Sp1 family proteins interacting with a GC/GT box
in cell cycle progression. Mol Cell Biol 2000; 20:1134-1139.
located immediately 3' to the TATA box. J Virol 1996; 70:188-198. 92. Ryu H et.al. Histone deacetylase inhibitors prevent oxidative neuronal
63. Dennig J et al. An inhibitor domain in Sp3 regulates its death independent of expanded polyglutamine repeats via an
glutamine-rich activation domains. EMBO J 1996; 15:5659-5667. Sp1-dependent pathway. Proc Natl Acad Sci USA 2003;
64. Burley SK et al. Biochemistry and structural biology of transcription 100(7):4281-6.
factor IID (TFIID). Annu Rev Biochem 1996; 65:769-799. 93. Ferrante R et al. Histone deacetylase inhibition by sodium butyrate
65. Chen JL et al. Assembly of recombinant TFIID reveals differential chemotherapy ameliorates the neurodegenerative phenotype in
coactivator requirements for distinct transcriptional activators. Cell Huntington’s disease mice. J Neurosci 2003; 23(28):9418-9427.
1994; 79:93-105. 94. Kaczynski J et al. Sp1- and Kruppel-like transcription factors.
66. Hahn S. The role of TAFs in RNA polymerase II transcription. Cell Genome Biol 2003; 4(2):206-210.
1998; 95:579-582.
CHAPTER 24
Abstract
W
ilms tumor suppressor gene product WT1 is a by Frasier syndrome,8 where decreased production of the +KTS
Cys2His2 (C2H2) zinc finger protein, one of the largest leads to severe developmental defects (see below).
protein families in higher eukaryotes. Since the cloning
of WT1 in 1990, we have learned a great deal about the func- Evolutionary Conservation of WT1
tions of WT1. Now, even greater molecular details of WT1 are The high degree of evolutionary conservation of WT1 is re-
beginning to emerge with a possibility of the unexpected role of markable throughout vertebrate species.9 The overall sequence
WT1 in the post-transcriptional processes. This chapter does similarity between the fish (Fugu) and human WT1 is 68% (57%
not include all aspects of WT1, but is intended to focus on and identical) but exclusive comparison of all vertebrate WT1 in the
highlight the biological function of WT1 (for detailed reviews zinc finger domain alone indicates the greatest conservation in
see refs. 1-4). The chapter consists largely of two parts, the bio- the DNA binding domain (more than 90-95% similar across spe-
chemical characterization of WT1 and the genetics of WT1 in cies). Furthermore, the alternative KTS splicing (KPS in Fugu)
both the human and the mouse, and ends with future perspec- has been conserved in all WT1 orthologs, whereas the alternative
tives on WT1 research. splice I (17 aa) is absent in chicken, reptile, and fish.9 This sug-
gests that the first alternative splicing occurred rather recently
Biochemical Characterization of WT1 with the appearance of placental mammals, while the KTS splic-
ing arose with the earliest forms of WT1 and thus likely to serve
Alternative Splicing and Generation of Multiple a more fundamental role in WT1 function. WT1 transcripts of
lower vertebrates also do not utilize multiple translation initia-
WT1 Isoforms tions and give rise to only two isoforms:10 WT1-KTS and the
WT1 was cloned more than a decade ago,5,6 and the sequence +KTS. Thus, the function of non-KTS isoforms may be mam-
revealed a Drosophila Krüppel-like zinc finger protein with four
mal specific or serve minor or redundant functions of WT1. In-
C2H2 zinc fingers at its carboxyl terminus as the DNA binding
deed, the wt1 knockout mice, which disrupt only the first alter-
domain (Fig. 1). The amino-terminal domain contains a
native splice (exon 5)11 or the upstream CUG-initiation,12 do
proline-glutamine rich region, which is frequently found in many
transcriptional activators. WT1 transcripts undergo numerous not display any gross abnormalities (see below).
modifications such as two alternative splicing events, RNA edit- The remarkable conservation of the KTS sequence through-
ing at codon 280 and alternative usage of either the upstream out vertebrates may indicate a special role of the KTS residues
CUG or the internal in-frame AUG initiations that result in 24 (such as a phosphorylation site). However, a recent study sug-
possible isoforms.3 The exact in vivo functions of the RNA-edited, gests that the KTS can be altered in both its sequence and length
upstream CUG- and internal-AUG initiated gene products are without affecting its apparent function.13 Then, why is the KTS
not clear and represent minor species relative to the 4 major so well conserved from fish to man? The most likely answer comes
isoforms, which arise as the result of two alternative splicing events from the studies of Davies et al13 who showed that the nucle-
(Fig. 1). The first alternative splice event includes or excludes 17 otides that code for the KTS as well as the immediate surround-
amino acids (aa) encoded by exon 5 while the second alternative ing sequences are extremely well conserved throughout evolu-
splice involves differential usage of two splice donor sites located tion. Because maintaining the KTS alternative splicing is central
9 nucleotides apart at the end of exon 9;7 this alternative splice II to the WT1 function as exemplified by the Frasier syndrome8
results in the exclusion or insertion of three amino acids, termed (see below), any changes in the nucleotides within or surround-
KTS (lysine, threonine, and serine), between the zinc fingers three ing the KTS during evolution would have led to defects in KTS
and four. The KTS-containing (+KTS) transcripts are the pre- splicing and subsequent demise of the organism.13 Thus, KTS
dominant form of WT1, at least two to three times more abun- conservation exemplifies a situation where the nucleotides rather
dant than the –KTS, throughout development.7 Maintaining this than amino acid sequences served as the determining factor be-
ratio is critical to the normal function of WT1 as demonstrated hind strict evolutionary conservation.
Zinc Finger Proteins: From Atomic Contact to Cellular Function, edited by Shiro Iuchi and Natalie Kuldell.
©2005 Landes Bioscience/Eurekah.com and Kluwer Academic/Plenum Publishers.
The Role of WT1 in Development and Disease 175
the zinc finger 1 abolished the RNA binding, which is in direct dition to WAGR, two other rare syndromes are also affected by
contrast to WT1-DNA interaction where the zinc finger 1 ap- germ line heterozygous WT1 mutations: Denys-Drash syndrome
pears to contribute minimally.20 A recent study has also demon- (DDS) and Frasier syndrome (FS). Symptoms of DDS patients
strated the importance of zinc finger 1 in RNA binding in vivo.42 include defects in genito-urinary development (such as diffuse
Another study using systematic evolution of ligands by exponen- mesangial sclerosis and male-female sex reversal phenotype) and
tial enrichment (SELEX) has identified a group of RNA sequences high incidents of Wilms tumor.56 The majority of WT1 muta-
bound by the WT1-KTS;41 however, this approach did not yield tions found in DDS patients are point mutations within the critical
a consensus RNA binding sequences nor any common secondary residues of the zinc fingers 2 and 3, which abolish DNA binding
structures were evident in the identified RNA. WT1 protein has in vitro.1 Wilms tumors arising in DDS, like WAGR, show a loss
also been isolated from the nuclear poly(A)+ ribonucleoprotein of the remaining WT1 allele.56 In contrast, most FS mutations
fraction using oligo(dT) chromatography.43 Together, a body of are found in the non-coding, splice junction of intron 9, effec-
evidence is starting to emerge in favor of WT1-RNA interaction tively disrupting the production of the alternatively spliced
and a possible role in RNA processing in vivo. The major chal- WT1+KTS isoform.57 This imbalance of the +/-KTS isoforms in
lenge now is to identify the in vivo RNA targets and the exact FS leads to severe genito-urinary defects as many of FS patients
role of WT1 in the post-transcriptional processes. The dual prop- suffer from male-female sex reversal phenotype (as seen in some
erty of WT1 in DNA and RNA binding is not unique to WT1 as DDS patients) and end-stage renal failure resulting from focal
another C2H2 zinc finger protein TFIIIA in Xenopus laevis also segmental glomerular sclerosis.8 FS patients, unlike DDS patients,
binds specifically to the intragenic region of the 5S rRNA gene44 however, are not predisposed for development of Wilms tumor,
and the 5S rRNA.45 but they suffer from a greater risk of gonadoblastoma.57 The role
of WT1 in gonadoblastoma is not known. Thus, these human
Expression of WT1 in Development studies highlight the importance of WT1 function in renal and
Expression of human WT1 has been documented in fetal kid- gonadal development and the involvement of WT1 inactivation
neys and gonads,46 but more detailed knowledge about WT1 in Wilms tumor. Of note, a homozygous point mutation of WT1
expression is available from the mouse studies.47-49 Expression of is also found in non-asbestos-related mesothelioma,58 a tumor
WT1 in developing mouse embryos is restricted to specific tis- derived from the peritoneal lining, although further reports of
sues, most of which undergo mesenchyme to epithelial transition WT1 mutations in mesothelioma are warranted.
during organogenesis.48 These tissues include kidneys, gonads, WT1 mutations are also found in about 10-15% of leuke-
mesothelial lining of the abdominal organs including the epicar- mias,59-61 mostly in adult acute myeloid leukemia. But unlike
dium, spleen, spinal cord, brain stem, ganglion cells of develop- Wilms tumors, many but not all of the WT1 mutations are het-
ing retina and digits. In adults, Wt1 (mouse WT1) expression erozygous in leukemias,61 raising the possibility that the muta-
becomes even more restricted and is maintained only in very few tions act in a dominant or dominant-negative fashion, or per-
cells such as the podocytes of glomeruli, Sertoli cells in testes, haps these mutations represent secondary effects of leukemia.
granulosa cells of the ovary, uterine wall,47 and the immature There are also conflicting results regarding the role of WT1 in
CD34+/CD38- hematopoietic precursor cells in the bone mar- leukemia as some data argue for an oncogenic role of WT1 (re-
row.50 Thus, wt1 (mouse WT1 gene) encodes a transcription fac- viewed in refs 62, 63). In fact, based on the clinical correlation of
tor (and also a RNA-binding protein) whose expression is devel- high WT1 expression and poor prognosis in some leukemias,
opmentally regulated and likely has a limited but specialized immunocancer therapy using anti-WT1-specific cytotoxic T-cells
functions in only a few organs after organogenesis is complete. is being explored for treatment of leukemia.64,65 The paradoxical
effects of WT1 in the hematopoietic system may be partially rec-
onciled by the observation of Ellisen et al.66 In this study, the
Genetic Characterization of WT1 ectopic expression of WT1 in primary CD34+cells isolated from
human umbilical cord blood, which include very rare multipotent
Human Genetics of WT1 stem cells and lineage-committed but undifferentiated progeni-
Identification of diseases affected by the mutations of the WT1 tor cells, resulted in two different populations of cells in culture:
and their symptoms has provided a great deal of knowledge about (1) undifferentiated, but lineage-committed progenitor cells rap-
the possible functions of WT1 (Table 1). A rare human disease idly underwent spontaneous differentiation with ectopic expres-
WAGR syndrome (Wilms tumor, aniridia, genitourinary malfor- sion of WT1-KTS; (2) WT1-KTS expression in a subset of
mations, and mental retardation) provided the first clue about CD34+/CD38- cells which represent the most primitive hemato-
the location of WT1. WAGR patients are distinguished by con- poietic precursor cells led to persistent quiescence of these cells in
stitutional hemizygous deletion of the chromosome 11p13 culture. This may be a reflection of the biphasic nature of WT1
band,51,52 and these children are at high risk for Wilms tumor, a expression during normal hematopoiesis, as low level expression
childhood kidney cancer and one of the most common solid tu- of WT1 is detected in a subset of very primitive CD34+/CD38-
mors among young children. Thus, a tumor suppressor gene, as cells50 and is dramatically downregulated as these cells begin to
well as genes responsible for other defects associated with WAGR, differentiate. 67 Expression of WT1 reappears as the
was thought to reside in 11p13. This has led to the positional lineage-committed cells begin to differentiate into granulocyte,
cloning of a zinc finger-containing gene WT1,5,6 which was sub- monocyte and B-cell lineages.66 Thus, it can be postulated that
sequently found to harbor mutations in about 10-15% of all spo- WT1 is involved in both maintaining the cells in multipotent
radic Wilms tumors.53 In most cases, Wilms tumor samples from progenitor state as well as promoting differentiation in later stages,
WAGR patients lose the remaining wild type WT1 allele,54,55 partially explaining the anti- and pro-differentiation properties
suggesting that complete loss of WT1 is required to develop Wilms of WT1, respectively.
tumor. WT1 hemizygosity is also responsible for the gonadal de- WT1 is also involved in a reciprocal chromosomal transloca-
fects and renal failures associated with WAGR patients.53 In ad- tion t(11;22)(p13;q12) in a rare tumor called Desmoplastic small
178 Zinc Finger Proteins: From Atomic Contact to Cellular Function
Wilms tumor Childhood kidney cancer Point mutations, frameshift (small insertions
or deletions), homozygous, somatic
WAGR Wilms tumor Hemizygous interstitial deletion of 11p13
Genitourinary defects (includes WT1) in the germ line, somatic in the
tumor (homozygous)
Denys-Drash Wilms tumor Point mutations usually in the zinc finger
Genitourinary defects (i.e., Arg 394 to Trp), heterozygous in the
Diffuse mesangial sclerosis germ line, somatic in the tumor (homozygous)
Male-female sex reversal phenotype
Frasier Genitourinary defects KTS splice donor site point mutation
Focal segmental glomerular sclerosis (i.e., position +2 in intron 9: TG to TC)
Male-female sex reversal phenotype heterozygous in the germ line
Gonadoblastomaa (reduction of +KTS form)
Leukemia Acute myeloid leukemia Frameshifts, usually heterozygous,
(few homozygous), somatic
Desmoplastic small Small round cells in the abdomen Reciprocal chromosome translocation
round cell tumor (propensity for young adolescent males) t(11;22) (p13;q12), heterozygous, somatic
a Status of WT1 mutation is not known
round cell tumor,68 which fuses the N terminus of Ewings sar- spinal cord. Thus, during kidney development, Wt1 is required
coma gene (EWS) with the last three zinc fingers of WT1. The for the survival of the mesenchymal cells, as well as promoting
EWS/WT1 fusion transcript retains the KTS splicing and thus outgrowth of ureteric bud and responding to the signals gener-
produces two isoforms: EWS/WT1 -/+KTS. EWS/WT1–KTS ated by ureteric bud. In addition to the early stages of renal devel-
is thought to function as a novel transcription factor and some of opment, continuous expression of Wt1 from both alleles is re-
its target genes have been identified (ref. 69 and the references quired for the homeostasis of mature kidneys as demonstrated by
therein). The function of the +KTS version of the fusion protein, frequent renal failures observed in wt1+/- mice and in individuals
however, remains unclear. EWS/WT1-KTS but not the +KTS with only one functional WT1 allele (Table 1). The hematopoi-
can transform NIH3T3 cells in vitro demonstrating the onco- etic system of wt1-null embryos appeared to develop normally,
genic potential of the fusion gene product.70 although early lethality of the mutant animals prevented a more
detailed study.71
Mouse Genetics of wt1 Out-breeding the mixed background of the original wt1+/-
The initial description of wt1 heterozygotes reported no ab- strain to a pure M1 strain background led to prolonged survival
normalities,71 but upon prolonged examination of the wt1+/- strain of the wt1-null embryos, but the animals still lacked kidneys and
showed that 80% of the mice die of terminal renal failure.72 Au- gonads, and died just before or soon after birth.74 However, this
topsy results revealed that these mice suffered from severe glom- allowed the investigators to uncover additional defects within
erular sclerosis. Renal dysfunction and premature death of the wt1-null embryos as these animals lacked spleen and adrenal
wt1+/- strain were also reported by another study.73 These obser- gland.74 The absence of spleen was shown to be due to 4 to 5-fold
vations are consistent with the human studies of WAGR and DDS increase in apoptosis within the spleen primordium. Another
syndromes where WT1 hemizygosity leads to end-stage renal fail- defect was uncovered more recently in the eyes of wt1-/- embryos
ure. Unlike WAGR and DDS patients, however, wt1+/- strain do which were smaller than the wild type littermates.75 Specifically,
not develop Wilms tumor.72,73 wt1-/- retina showed less retinal cell proliferation and increased
Mice with homozygous inactivation of wt1 are embryonic le- apoptosis, resulting in about 40% less cell number in the gan-
thal at around E13.5.71 Examination of the wt1-/- embryos showed glion cell layer of the mutant retina. This was correlated with the
that these animals lack kidneys and gonads in addition to defects absence of the POU-domain transcription factor, pou4f2, which
in diaphram, epicardium and mesothelium lining the abdominal is required for the survival of the retinal ganglion cells. Recurrent
organs. The cause of death was likely due to malformation of observations of decreased proliferation and differentiation along
epicardium, a tissue that surrounds the heart, leading to excessive with increased apoptosis in various wt1-/- organs suggest that one
bleeding and heart failure. Renal agenesis was due to a develop- of the primary functions of wt1 is in regulating genes that are
mental arrest during the early stage of kidney development as critical for cell proliferation, differentiation and survival, either
metanephric mesenchymes, where Wt1 is expressed, die of mas- transcriptionally or post-transcriptionally.
sive apoptosis. In vitro kidney organ culture assay demonstrated A more detailed insight into specific functions of the WT1
that wt1-/- mesenchymal cells also fail to differentiate in response isoforms was provided by the generation of either the –KTS or
to signals from strong inducers of kidney differentiation such as a +KTS-specific knockout mice by introducing mutations in the
The Role of WT1 in Development and Disease 179
splice donor sites at the end of exon 9.35 The first surprise came time, questions that remain unanswered from the past efforts,
when these mice were born with incomplete kidneys and gonads. such as identifying the physiological transcriptional target genes,
As expected, both—and +KTS knockout mice died soon after delineating the transcriptional regulation mechanisms and the
birth due to renal dysfunction. Because the wt1-null animals com- mechanism of tissue-specific regulation of WT1, will continue to
pletely lack kidneys and gonads,71 this suggests that both the fuel and reshape our efforts on WT1. Combining the biochemi-
–KTS and the +KTS isoforms have limited redundant functions cal, cellular and genetical approaches with new tools in genomics
at least during the initial steps of renal and gonadal development. and proteomics will continue to unravel the fascinating stories of
This is both remarkable and unexpected given the dramatic dif- WT1 in development and cancer.
ferences in the DNA binding and the distinct nuclear distribu-
tion displayed by the two isoforms. However, there are also clear Acknowledgments
functional differences between the two isoforms as neither knock- Due to space limitation, many important findings on WT1
out animals fully develop in the absence of the other isoform.35 could not be discussed here. Thus, we would like to acknowledge
The –KTS-null kidneys and gonads are much more reduced in all the investigators in the WT1 field for their many contribu-
size and undifferentiated compared to the +KTS-null counter- tions and apologize to many of our colleagues whose important
parts, suggesting that the –KTS has more fundamental role in work could not be mentioned. We thank Drs. Rick Proia, Connie
proliferation, differentiation and survival. The +KTS-null ani- Noguchi and Barbara Christensen for critical reading and com-
mals, on the other hand, display complete male to female sex ments on the manuscript. SBL is indebted to Dr. Dan Haber for
reversal phenotype. This is reminiscent of the Frasier syndrome his encouragement and unwavering support.
in humans although in Frasier syndrome, the symptoms are mani-
fested by a heterozygous mutation8 while in mouse it takes ho-
mozygous KTS splice mutations to reveal the phenotype.35 The
References
1. Reddy JC, Licht JD. The WT1 Wilms’ tumor suppressor gene: how
lack of Wt1+KTS expression in the +KTS-null gonads resulted in much do we really know? Biochim Biophys Acta 1996; 1287(1):1-28.
a drastically decreased expression of two important male-specific 2. Little M, Holmes G, Walsh P. WT1: what has the last decade told
genes, the Y-chromosome associated mSry (sex-determining re- us? Bioessays 1999; 21(3):191-202.
gion Y gene) and mSox9 (Sry-related HMG box). Both of these 3. Lee SB, Haber DA. Wilms tumor and the WT1 gene. Exp Cell Res
genes encode transcription factors that are critical for the male 2001; 264(1):74-99.
4. Wagner KD, Wagner N, Schedl A. The complex life of WT1. J
sex determination pathway.4 Another feature of Frasier syndrome Cell Sci 2003; 116(Pt 9):1653-1658.
that is recapitulated by the +KTS mice is that 70% of the wt1+KTS 5. Call KM, Glaser T, Ito CY et al. Isolation and characterization of a
heterozygous mice die of renal dysfunction due to glomerular zinc finger polypeptide gene at the human chromosome 11 Wilms’
nephropathy. No abnormalities are reported in the kidneys and tumor locus. Cell 1990; 60(3):509-520.
gonads of wt1-KTS heterozygous mice. These observations dem- 6. Gessler M, Poustka A, Cavenee W et al. Homozygous deletion in
onstrate the importance of Wt1+KTS during a male-specific dif- Wilms tumours of a zinc-finger gene identified by chromosome
jumping. Nature 1990; 343(6260):774-778.
ferentiation of gonads, and the severe consequences of reducing 7. Haber DA, Sohn RL, Buckler AJ et al. Alternative splicing and
the gene dosage of the +KTS but not the –KTS in maintaining genomic structure of the Wilms tumor gene WT1. Proc Natl Acad
proper glomerular function. Interestingly, homozygous inactiva- Sci U S A 1991; 88(21):9618-9622.
tion of two mammal-specific wt1 isoforms, the alternative, up- 8. Barbaux S, Niaudet P, Gubler MC et al. Donor splice-site mutations
stream CUG-initiated isoform12 and the exon 5-containing in WT1 are responsible for Frasier syndrome. Nat Genet 1997;
isoform,11 had no effect on the mouse development, survival and 17(4):467-470.
9. Kent J, Coriat AM, Sharpe PT et al. The evolution of WT1 sequence
fertility. Thus, these two isoforms of Wt1 probably have func- and expression pattern in the vertebrates. Oncogene 1995;
tions that are redundant with the major Wt1 isoforms. 11(9):1781-1792.
10. Schedl A, Hastie N. Multiple roles for the Wilms’ tumour suppressor
Conclusion and Future Perspectives gene, WT1 in genitourinary development. Mol Cell Endocrinol
1998; 140(1-2):65-69.
Over the years, WT1 research has contributed greatly to our 11. Natoli TA, McDonald A, Alberta JA et al. A mammal-specific exon
general concept of how zinc finger proteins work as well as to our of WT1 is not required for development or fertility. Mol Cell Biol
understanding of development and cancer. This chapter briefly 2002; 22(12):4433-4438.
discussed the structure and molecular details of WT1 zinc fingers 12. Miles CG, Slight J, Spraggon L et al. Mice lacking the
and aimed to focus on the biology of WT1 in the context of 68-amino-acid, mammal-specific N-terminal extension of WT1
develop normally and are fertile. Mol Cell Biol 2003;
development and disease. Biochemical studies of WT1 illumi-
23(7):2608-2613.
nate its role in both transcriptional and post-transcriptional pro- 13. Davies RC, Bratt E, Hastie ND. Did nucleotides or amino acids
cesses while the genetic studies underlie the importance of WT1 drive evolutionary conservation of the WT1 +/-KTS alternative splice?
in development and in different types of human disease. It is evi- Hum Mol Genet 2000; 9(8):1177-1183.
dent that future studies on WT1 will provide even greater in- 14. Richard DJ, Schumacher V, Royer-Pokora B et al. Par4 is a
sights to our knowledge of zinc finger proteins, and continue to coactivator for a splice isoform-specific transcriptional activation
domain in WT1. Genes Dev 2001; 15(3):328-339.
serve as a paradigm for studying the link between development
15. Johnstone RW, See RH, Sells SF et al. A novel repressor, par-4,
and cancer. Current advances in WT1 have raised a new set of modulates transcription and growth suppression functions of the Wilms’
questions and opened new fields of research, and the latest ad- tumor suppressor WT1. Mol Cell Biol 1996; 16(12):6945-6956.
ventures of WT1 research will take us into the world of 16. Moffett P, Bruening W, Nakagama H et al. Antagonism of WT1
post-transcriptional mechanisms. There is also a growing interest activity by protein self-association. Proc Natl Acad Sci U S A 1995;
in the role of WT1 in hematopoiesis and leukemia. In this re- 92(24):11105-11109.
17. Kennedy D, Ramsdale T, Mattick J et al. An RNA recognition motif
gard, a provocative and intriguing research in the development of
in Wilms’ tumour protein (WT1) revealed by structural modelling.
WT1-based immunocancer therapy is worth noting. At the same Nat Genet 1996; 12(3):329-331.
180 Zinc Finger Proteins: From Atomic Contact to Cellular Function
18. Bruening W, Moffett P, Chia S et al. Identification of nuclear 42. Ladomery M, Sommerville J, Woolner S et al. Expression in Xenopus
localization signals within the zinc fingers of the WT1 tumor oocytes shows that WT1 binds transcripts in vivo, with a central
suppressor gene product. FEBS Lett 1996; 393(1):41-47. role for zinc finger one. J Cell Sci 2003; 116(Pt 8):1539-1549.
19. Rauscher FJ, 3rd, Morris JF, Tournay OE et al. Binding of the Wilms’ 43. Ladomery MR, Slight J, Mc Ghee S et al. Presence of WT1, the
tumor locus zinc finger protein to the EGR-1 consensus sequence. Wilm’s tumor suppressor gene product, in nuclear poly(A)(+)
Science 1990; 250(4985):1259-1262. ribonucleoprotein. J Biol Chem 1999; 274(51):36520-36526.
20. Nakagama H, Heinrich G, Pelletier J et al. Sequence and structural 44. Engelke DR, Ng SY, Shastry BS et al. Specific interaction of a
requirements for high-affinity DNA binding by the WT1 gene purified transcription factor with an internal control region of 5S
product. Mol Cell Biol 1995; 15(3):1489-1498. RNA genes. Cell 1980; 19(3):717-728.
21. Wang ZY, Qiu QQ, Enger KT et al. A second transcriptionally active 45. Honda BM, Roeder RG. Association of a 5S gene transcription factor
DNA-binding site for the Wilms tumor gene product, WT1. Proc with 5S RNA and altered levels of the factor during cell
Natl Acad Sci U S A 1993; 90(19):8896-8900. differentiation. Cell 1980; 22(1 Pt 1):119-126.
22. Pavletich NP, Pabo CO. Zinc finger-DNA recognition: crystal 46. Pritchard-Jones K, Fleming S, Davidson D et al. The candidate
structure of a Zif268-DNA complex at 2.1 A. Science 1991; Wilms’ tumour gene is involved in genitourinary development.
252(5007):809-817. Nature 1990; 346(6280):194-197.
23. Laity JH, Dyson HJ, Wright PE. Molecular basis for modulation of 47. Pelletier J, Schalling M, Buckler AJ et al. Expression of the Wilms’
biological function by alternate splicing of the Wilms’ tumor tumor gene WT1 in the murine urogenital system. Genes Dev 1991;
suppressor protein. Proc Natl Acad Sci USA 2000; 5(8):1345-1356.
97(22):11932-11935. 48. Armstrong JF, Pritchard-Jones K, Bickmore WA et al. The expression
24. Wang ZY, Qiu QQ, Deuel TF. The Wilms’ tumor gene product of the Wilms’ tumour gene, WT1, in the developing mammalian
WT1 activates or suppresses transcription through separate functional embryo. Mech Dev 1993; 40(1-2):85-97.
domains. J Biol Chem 1993; 268(13):9172-9175. 49. Moore AW, Schedl A, McInnes L et al. YAC transgenic analysis reveals
25. Reddy JC, Hosono S, Licht JD. The transcriptional effect of WT1 Wilms’ tumour 1 gene activity in the proliferating coelomic epithelium,
is modulated by choice of expression vector. J Biol Chem 1995; developing diaphragm and limb. Mech 1998; 79(1-2): 169-184.
270(50):29976-29982. 50. Baird PN, Simmons PJ. Expression of the Wilms’ tumor gene (WT1)
26. Maheswaran S, Park S, Bernard A et al. Physical and functional in normal hemopoiesis. Exp Hematol 1997; 25(4):312-320.
interaction between WT1 and p53 proteins. Proc Natl Acad Sci U 51. Riccardi VM, Sujansky E, Smith AC et al. Chromosomal imbalance
S A 1993; 90(11):5100-5104. in the Aniridia-Wilms’ tumor association: 11p interstitial deletion.
27. Scharnhorst V, Dekker P, van der Eb AJ et al. Physical interaction Pediatrics 1978; 61(4): 604-610.
between Wilms tumor 1 and p73 proteins modulates their functions. 52. Francke U, Holmes LB, Atkins L et al. Aniridia-Wilms’ tumor
J Biol Chem 2000; 275(14):10202-10211. association: evidence for specific deletion of 11p13. Cytogenet Cell
28. Wang W, Lee SB, Palmer R et al. DA. A functional interaction Genet 1979; 24(3):185-192.
with CBP contributes to transcriptional activation by the Wilms 53. Hastie ND. The genetics of Wilms’ tumor-a case of disrupted
tumor suppressor WT1. J Biol Chem 2001; 276(20):16810-16816. development. Annu Rev Genet 1994; 28:523-558.
29. Dallas PB, Yaciuk P, Moran E. Characterization of monoclonal 54. Baird PN, Groves N, Haber DA et al. Identification of mutations
antibodies raised against p300: both p300 and CBP are present in in the WT1 gene in tumours from patients with the WAGR
intracellular TBP complexes. J Virol 1997; 71(2):1726-1731. syndrome. Oncogene 1992; 7(11):2141-2149.
30. Nachtigal MW, Hirokawa Y, Enyeart-VanHouten DL et al. Wilms’ 55. Brown KW, Watson JE, Poirier V et al. Inactivation of the remaining
tumor 1 and Dax-1 modulate the orphan nuclear receptor SF-1 in allele of the WT1 gene in a Wilms’ tumour from a WAGR patient.
sex-specific gene expression. Cell 1998; 93(3):445-454. Oncogene 1992; 7(4):763-768.
31. Du X, Hublitz P, Gunther T et al. The LIM-only coactivator FHL2 56. Pelletier J, Bruening W, Kashtan CE et al. Germline mutations in
modulates WT1 transcriptional activity during gonadal the Wilms’ tumor suppressor gene are associated with abnormal
differentiation. Biochim Biophys Acta 2002; 1577(1):93-101. urogenital development in Denys-Drash syndrome. Cell 1991; 67(2):
32. Bannister AJ, Kouzarides T. The CBP co-activator is a histone 437-447.
acetyltransferase. Nature 1996; 384(6610):641-643. 57. Koziell A, Charmandari E, Hindmarsh PC et al. Frasier syndrome,
33. Larsson SH, Charlieu JP, Miyagawa K et al. Subnuclear localization part of the Denys Drash continuum or simply a WT1 gene associated
of WT1 in splicing or transcription factor domains is regulated by disorder of intersex and nephropathy? Clin Endocrinol (Oxf ) 2000;
alternative splicing. Cell 1995; 81(3):391-401. 52(4):519-524.
34. Englert C, Vidal M, Maheswaran S et al. Truncated WT1 mutants 58. Park S, Schalling M, Bernard A et al. The Wilms tumour gene WT1
alter the subnuclear localization of the wild-type protein. Proc Natl is expressed in murine mesoderm-derived tissues and mutated in a
Acad Sci U S A 1995; 92(26):11960-11964. human mesothelioma. Nat Genet 1993; 4(4):415-420.
35. Hammes A, Guo JK, Lutsch G et al. Two splice variants of the 59. King-Underwood L, Renshaw J, Pritchard-Jones K. Mutations in
Wilms’ tumor 1 gene have distinct functions during sex the Wilms’ tumor gene WT1 in leukemias. Blood 1996;
determination and nephron formation. Cell 10 2001; 87(6):2171-2179.
106(3):319-329. 60. Miyagawa K, Hayashi Y, Fukuda T et al. Mutations of the WT1
36. Davies RC, Calvio C, Bratt E et al. WT1 interacts with the splicing gene in childhood nonlymphoid hematological malignancies. Genes
factor U2AF65 in an isoform-dependent manner and can be Chromosomes Cancer 1999; 25(2):176-183.
incorporated into spliceosomes Oct 15 1998; 12(20):3217-3225. 61. King-Underwood L, Pritchard-Jones K. Wilms’ tumor (WT1) gene
37. Little NA, Hastie ND, Davies RC. Identification of WTAP, a novel mutations occur mainly in acute myeloid leukemia and may confer
Wilms’ tumour 1-associating protein. Hum Mol Genet 2000; drug resistance. Blood 1998; 91(8):2961-2968.
9(15):2231-2239. 62. Pritchard-Jones K, King-Underwood L. The Wilms tumour gene
38. Granadino B, Campuzano S, Sanchez L. The Drosophila WT1 in leukaemia. Leuk Lymphoma 1997; 27(3-4):207-220.
melanogaster fl(2)d gene is needed for the female-specific splicing 63. Algar E. A review of the Wilms’ tumor 1 gene (WT1) and its role
of Sex-lethal RNA. Embo J 1990; 9(8):2597-2602. in hematopoiesis and leukemia. J Hematother Stem Cell Res 2002;
39. Zhou Z, Licklider LJ, Gygi SP et al. Comprehensive proteomic 11(4):589-599.
analysis of the human spliceosome. Nature 2002; 419(6903):182-185. 64. Bellantuono I, Gao L, Parry S et al. Two distinct
40. Caricasole A, Duarte A, Larsson SH et al. RNA binding by the HLA-A0201-presented epitopes of the Wilms tumor antigen 1 can
Wilms tumor suppressor zinc finger proteins. Proc Natl Acad Sci U function as targets for leukemia-reactive CTL. Blood 2002;
S A 1996; 93(15):7562-7566. 100(10):3835-3837.
41. Bardeesy N, Pelletier J. Overlapping RNA and DNA binding 65. Scheibenbogen C, Letsch A, Thiel E et al. CD8 T-cell responses to
domains of the wt1 tumor suppressor gene product. Nucleic Acids Wilms tumor gene product WT1 and proteinase 3 in patients with
Res 1998; 26(7):1784-1792. acute myeloid leukemia. Blood 2002; 100(6):2132-2137.
The Role of WT1 in Development and Disease 181
66. Ellisen LW, Carlesso N, Cheng T et al. The Wilms tumor suppressor 71. Kreidberg JA, Sariola H, Loring JM et al. WT-1 is required for
WT1 directs stage-specific quiescence and differentiation of human early kidney development. Cell 1993; 74(4):679-691.
hematopoietic progenitor cells. Embo J 2001; 20(8):1897-1909. 72. Menke AL, Clarke AR, Leitch A et al. Genetic interactions between
67. Maurer U, Brieger J, Weidmann E et al. The Wilms’ tumor gene is the Wilms’ tumor 1 gene and the p53 gene. Cancer Res 2002;
expressed in a subset of CD34+ progenitors and downregulated early 62(22):6615-6620.
in the course of differentiation in vitro. Exp Hematol 1997; 73. Guo JK, Menke AL, Gubler MC et al. WT1 is a key regulator of
25(9):945-950. podocyte function: reduced expression levels cause crescentic
68. Ladanyi M, Gerald W. Fusion of the EWS and WT1 genes in the glomerulonephritis and mesangial sclerosis. Hum Mol Genet 2002;
desmoplastic small round cell tumor. Cancer Res 1994; 11(6):651-659.
54(11):2837-2840. 74. Herzer U, Crocoll A, Barton D et al. The Wilms tumor suppressor
69. Palmer RE, Lee SB, Wong JC et al. Induction of BAIAP3 by the gene wt1 is required for development of the spleen. Curr Biol 1999;
EWS-WT1 chimeric fusion implicates regulated exocytosis in 9(15):837-840.
tumorigenesis. Cancer Cell 2002; 2(6):497-505. 75. Wagner KD, Wagner N, Vidal VP et al. The Wilms’ tumor gene
70. Kim J, Lee K, Pelletier J. The desmoplastic small round cell tumor Wt1 is required for normal development of the retina. Embo J 2002;
t(11;22) translocation produces EWS/WT1 isoforms with differing 21(6):1398-1405.
oncogenic properties. Oncogene 1998; 16(15):1973-1979.
CHAPTER 25
Yin Yang 1
Huifei Liu and Yang Shi*
Abstract
T
he transcription factor Yin Yang 1 (YY1) (also known as Table 1. YY1 Homologues
NF-E1, δ, CP-1 and UCRBP) is a GLI-Kruppel
zinc finger protein.1-4YY1 is highly conserved from Xe- Chromosome
nopus to mammalian YY1 and has been demonstrated to play an Name Species # aa Localization Refs
essential role in mouse embryonic development and other physi-
ological and pathological conditions.5,6 As one of the first mam- YY1 Human 414 14q32 16
malian transcription factors that have been identified to possess YY2 Human 372 Xq22. 11
both repression and activation activities, YY1 has served as a 1-22.2
model for understanding how a single molecule can conduct Delta Mouse 414 12F1 17
opposite transcriptional functions, that is, repression and activa- YY1 Rat 411 6q32 10,18
tion. Recent studies suggest that YY1 is a mammalian counter- cYY1 Chicken 415
part of the Drosophila Polycomb Group (PcG) protein FIII/xYY1 Xenopus 373 9
Pleiohomeotic (Pho),7,8 providing a new framework in which to YY1 Danio rerio 357
further explore YY1’s biological functions and mechanism of (Zebrafish)
action in vivo. In this chapter, we discuss progress on YY1 since Pho Drosophila 520 4 8
its discovery. Pho-like Drosophila 669 3 19
Rex-1 Mouse 288 8 1,12,20
Introduction
amino acids (aa) (human YY1 aa205-226, Pho aa148-169) is con-
YY1 Family of Proteins served as well with 80% sequence similarity, suggesting that this
YY1 is evolutionarily conserved from human to Xenopus with region may be important for yet to be identified, evolutionarily
significant sequence homology spanning the entire protein (Table conserved functions.8
1). The four zinc fingers that are responsible for sequence-specific
DNA-binding are identical among Xenopus, rat, mouse and hu- YY1 Structure
man YY1, while the N-terminal regions are more divergent.2,5,9,10 YY1 protein contains modular features characteristic of many
Recently, a YY1-related protein, which we term YY2, has been transcription factors: a DNA binding motif and domains impor-
predicted based on the genome sequence.11 YY2 shares 55% ho- tant for transcriptional repression and activation (Fig. 1). The
mology with YY1 and 86% homology in the zinc finger regions. four-zinc-finger DNA-binding domain is located at its C-terminus
Preliminary data suggest that it is likely to be an expressed gene (aa 298-320, 327-347, 355-377 and 385-407), which overlaps
(Shi and Seto labs, unpublished results). However, the function with one of the two repression domains, and is also the most
and mechanism of action of YY2 remain unclear. A third mam- conserved motif among YY1 across species (from Drosophila to
malian protein related to YY1 is Rex-1 (Reduced Expression 1).1,12 human). The second repression domain is located in the central
Sequence homology between YY1 and Rex-1 is limited to the region (aa 170-200). Both repression domains have been shown
four zinc fingers (75% homology).1 Rex-1 was isolated as a pro- to recruit histone deacetylases (HDACs) for repression.21 The
tein whose expression decreases during F9 cell differentiation in- activation domain of YY1 lies at the N-terminus and is generally
duced by retinoic acid.12 Recently Rex-1 has been classified as believed to be bipartite. It contains an acidic activation domain
one of the definitive markers for embryonic stem cells.13-15 (aa16-29 and 43-53) and a glutamine-proline rich region (aa
Two Drosophila PcG proteins, Pho and Pho-like share signifi- 80-100).22 The acidic, glutamine- and proline-rich regions are
cant homology with YY1 within the four zinc fingers.8,19 The common features to many transcription activators.1,23-26 The fact
overall zinc finger sequence similarity between Pho and YY1 is as that both repression and activation domains are present in YY1
high as 96%, with zinc fingers 2 and 3 being completely identi- provides the structural basis for the dual repression and activa-
cal.8 Interestingly, outside the zinc finger region, a stretch of 22 tion activities of YY1. Between the two activation domains, there
Zinc Finger Proteins: From Atomic Contact to Cellular Function, edited by Shiro Iuchi and Natalie Kuldell.
©2005 Landes Bioscience/Eurekah.com and Kluwer Academic/Plenum Publishers.
Yin Yang 1 183
Histone Acetylases Interact with YY1 Directly and Histone Deacetylases Are Required for Repression by
Regulate the Transcriptional Activities of YY1 YY1 in Several Promoters
The YY1 interacting proteins p300 and CBP are two closely YY1 can repress transcription through direct interactions with
related proteins that have HAT activity and function as transcrip- HDACs or indirect recruitment of HDACs (Fig. 2b). YY1 binds
tional coactivators by acetylating histone tails (Fig. 2a).21,44 How- HDAC1 and 2 directly and this interaction is critical for
ever, their role in YY1-mediated transcription is not completely YY1-mediated transcription repression at several promoters such
understood. An earlier study showed that adenovirus E1A could as the HIV-1 long terminal repeat (LTR) and MHC class II
relieve YY1-mediated repression via interaction with p300, sug- gene.49,55,56 YY1 may recruit HDACs indirectly through bridge
gesting that p300 might play a role in YY1-mediated repression.54 proteins as well. Several YY1 binding proteins, including
Consistent with this idea, a recent study showed that upon acety- SAP30, FKBP25 (FK506-binding protein) and MTA2
lation by p300/PCAF (p300/CBP associated factor), the acety- (metastasis-associated protein 2), have been demonstrated to
lated YY1 has decreased DNA binding ability and increased af- bind HDACs. 57-59 The interaction between YY1 and HDACs is
finity to HDACs.21 Taken together, these findings suggest that regulated by post-translational modification of both YY1 and
interactions of YY1 with p300 may in some cases help maintain HDACs. Acetylated YY1 has increased affinity for HDACs,21
YY1 as a repressor. On the other hand, a positive role of CBP has while phosphorylation of HDAC1 disrupts its interaction with
been shown in the 25-hydroxyvitamin D (3)-24-hydroxylase pro- YY1.60
moter where it relieves YY1-mediated repression by preventing
YY1 from contacting and inhibiting TFIIB function.34 However, YY1 May Recruit Histone Methyltransferases to
this action of CBP does not appear to require its HAT activity Regulate Transcription
since a HAT defective CBP mutant is still able to relieve Histone methylation has been demonstrated to play a role in
YY1-mediated repression. In contrast, a CBP mutant defective both transcriptional repression and activation.50,61-64 YY1 has re-
for YY1 interaction failed to relieve YY1 repression, highlighting cently been shown to activate transcription by recruiting the his-
the importance of protein-protein interactions.34 It will be inter- tone H4 (Arg 3)-specific methyltransferase, PRMT1, through
esting to use a combination of ChIP and RNA interference (RNAi) DRBP76, an alternative splicing form of ILF3 (interleukin en-
technology to determine the histone acetylation status at these hancer binding factor 3) which binds both YY1 and PRMT1
YY1-repressed promoters, which may provide insight into whether directly (Fig. 2c).50 PRMT1 has also been demonstrated to facili-
p300/CBP HAT activities play an active role in transcriptional tate transcriptional activation by nuclear hormone receptor.64
regulation at these promoters.
Figure 2. a) YY1 regulates transcription through DNA bending or chromatin remodeling. YY1 recruits coactivators such as p300 and CBP to
acetylate histone tails or YY1 itself; b) YY1 interacts with HDACs either directly or indirectly through bridge proteins. HDACs then deacetylate
histone tails and result in transcriptional repression; c) YY1 tethers HMT through intermediate molecules, HMT then regulates transcription by
methylating histone tails. CBP= cAMP response element binding protein; HDACs= histone deacetylases; HMT= histone methyltransferase; Ac=
acetyl; Me= methyl.
Yin Yang 1 185
Recent studies suggest that YY1 may also recruit YY1 May Regulate Transcription by Directly
HMT-containing protein complexes to regulate transcription. Contacting the General Transcription Factors (GTFs)
Elegant genetic studies in Drosophila showed that YY1 repres- Several YY1-interacting proteins are members of the RNA
sion requires the transcriptional corepressor CtBP.7 CtBP was polymerase II associated basal transcription machinery including
initially identified as an E1A-binding protein and has been shown RNA polymerase II, TBP, TFIIB and TAFII55 (TBP-associated
to repress transcription by recruiting both HDACs and factor II) (Fig. 3b).22,67,72 The interaction between YY1 and these
HMTs.51,65 Thus, YY1-mediated repression may occur through GTFs and the subsequent impact on transcriptional regulation
the recruitment of the HMT-containing CtBP complex. remain largely unclear. It was shown that binding to TFIIB sig-
In summary, recruitment of histone modifiers appears to be nificantly increases the binding affinity of YY1 to its cognate sites
an important mechanism for YY1-mediated transcriptional regu- while YY1 may repress transcription by sequestration of TFIIB to
lation. The role of HATs in YY1-mediated transcription appears prevent it from interacting with transcription activators.34,67
more complicated and will require further investigation.
Competition for Promoter Occupancy and Physical
YY1 Regulates Transcription by Affecting the Basal
Interactions with Other Transcription Factors
Transcription Machinery An earlier model for YY1-mediated repression is competition
In addition to the chromatin-remodeling model, YY1 has been with other activators for overlapping DNA binding sites in a pro-
shown to regulate transcription by influencing the basal transcrip- moter (Fig. 3c).5 This model is supported by the convergence of
tion machinery. YY1 may substitute the function of TBP recognition sites between YY1 and several transcription factors
(TATA-box binding protein) and directly recruit TFIIB (tran- such as serum response factor (SRF) on the skeletal α-actin and
scription factor IIB) and RNA polymerase II to the promoter c-fos promoter, NF-κB on the serum amyloid A 1 (SAA1) pro-
region to initiate transcription. YY1 has also been shown to di- moter. The competitive binding between YY1 and these tran-
rectly interact with several components of the basal transcription scription factors to these cognate sequences are regulated in dif-
machinery. YY1 may also regulate the basal transcription machin- ferent cellular conditions, contributing to differential gene
ery indirectly by interacting with other transcription factors, expression.73-75
coactivators and corepressors.
A second model is a combinatorial regulation of gene tran-
scription by multiple transcription factors binding to their re-
YY1 May Function As an Initiator Element Binding spective consensus sites in the same promoter region. In this case,
Protein to Initiate Transcription in Vitro the YY1 recognition sequence is close to but not overlapping with
Several studies suggest a potential role of YY1 to initiate tran- the DNA binding sites of other transcription factors. The close
scription. YY1 binds to the transcription initiation sites of several proximity of YY1 with other transcription factors facilitates their
promoters such as human PCNA, mouse Surf-1 and the AAV P5 physical interactions at the promoter, which result in new pat-
promoter.5,66 YY1 directly interacts with TFIIB and RNA poly- terns of gene expression. For example, binding of YY1 or E2F2/3
merase II.67 It was demonstrated that the combination of YY1, to their consensus sites in the cdc6 promoter had little effect on
TFIIB and RNA polymerase II is sufficient to initiate transcrip- gene transcription by themselves.76 However, the coordination
tion from the AAV P5 promoter in vitro.67,68 The ability of YY1
of YY1 and E2F2/3 through the cofactor RYBP allows the syner-
to initiate transcription is augmented by the presence of a TATA
gistic activation of cdc6 transcription (Fig. 3d). On the other
motif or binding sites for the transcription factor Sp1, which in-
teracts with YY1 directly.69,70 Finally, the cocrystal structure of hand, it was shown that binding of LexA-YY1 fusion protein to
YY1 zinc fingers with the AAV P5 initiator element provided struc- the LexA-binding sites is important to the YY1-mediated repres-
tural basis for understanding the YY1-dependent, sion on several activators binding to the nearby GAL4 binding
TBP-independent transcription initiation.31 The zinc fingers of sites in a reporter plasmid.77 Although direct protein-protein in-
YY1 bind to both strands of DNA upstream of the transcription teraction between YY1 and these transcription factors may occur,
initiation site while only the template strand downstream of the it is not necessary for at least some, if not all, combinatorial regu-
initiation site binds to YY1, allowing access of RNA polymerase lation.77 The exact mechanism for the combinatorial regulation
II to the transcription initiation sites. This YY1-DNA binding is also not clear. It is possible that this may allow differential re-
permits transcription to occur only from the downstream of the cruitment of histone modifiers and/or interactions with the basal
YY1 binding sites. Transcription started upstream or in the op- transcription machinery. It is often the case that multiple
posite direction will be blocked by YY1-DNA binding which pre- nonoverlapping binding sites are present in a single promoter,
vents the double strand separation. Since all these studies were making this model a potentially common mechanism for tran-
conducted in vitro with the AAV P5 initiator, their in vivo sig- scriptional control involving YY1.
nificance remains to be investigated. Nevertheless, YY1 has been YY1 may also activate or repress transcription through direct
identified to exist in a large RNA polymerase II complex isolated interactions with other transcription factors without binding to
from HeLa nuclear extracts that contains several general tran- DNA itself (Fig. 3e).78 For example, YY1 inhibits the serum re-
scription factors but not TFIIB, TBP and TBP-associated fac- sponse element binding protein (SREBP)-mediated activation of
tors,71 suggesting a potential function for YY1 to initiate tran- the low-density lipoprotein (LDL) receptor gene in a
scription in vivo (Fig. 3a) and providing a new angle to consider dose-dependent manner.78 This inhibition does not require DNA
transcription initiation from TATA-less promoters. binding of YY1 but relies on the direct interaction between YY1
186 Zinc Finger Proteins: From Atomic Contact to Cellular Function
Figure 3. a) YY1 regulates transcription through protein-protein interaction. YY1 functions as an initiator element binding protein to assemble
the basal transcription machinery. b) YY1 regulates transcription by contacting general transcription factors. c) YY1 regulates transcription by
competing DNA binding with other transcription factors. d) Combinatorial regulation of transcription by YY1 and other transcription factors
binding to the same promoter. e) YY1 either activates or represses transcription by interacting with other transcription factors without binding
to DNA. GTF= general transcription factor; RNApolII= RNA polymerase II.
and Sp1, which may subsequently inhibit the Sp1-SREBP inter- in vivo significance of these models under different physiological
action necessary for transcription activation.78 This model is fur- and pathologic conditions awaits further studies.
ther supported by another study showing YY1 repression of sev-
eral activator- mediated transcription of a reporter gene containing Regulation of YY1’s Transcriptional Activity
no YY1 binding sites.77 This DNA-binding independent activity Previous studies suggest that YY1 can be regulated at multiple
of YY1 is expected to increase the repertoire of genes that can be levels. Activity of YY1 can be modified by its DNA binding sites
regulated by YY1. and its-interacting proteins. In addition, YY1 function also de-
pends on its cellular localization and post-translational modifica-
DNA Bending Plays a Role in Some YY1-Regulated tions. These different factors may help achieve cell and develop-
Promoters mental specific control of YY1-mediated transcription.
DNA bending induced by DNA-binding transcription fac-
tors such as TBP has been suggested to be an important tran- DNA Binding Sites May Regulate the
scription regulatory mechanism. Previous studies suggested that Transcriptional Activity of YY1
YY1 might regulate promoters such as c-fos by inducing DNA YY1 Binding to its consensus sites is likely to be influenced by
bending.79,80 However, the absence of DNA bending in the several features of the DNA binding sites including the flanking
cocrystal structure of the YY1 zinc fingers and the AAV P5 nucleotides, CpG methylation and local chromatin structure. The
initiator element argue against DNA bending as a general fea- differential binding of YY1 to a selected set of promoters in a
ture of YY1-mediated transcription.31 specific cellular environment is critical for its tissue and develop-
In summary, YY1 mediates transcriptional regulation through mental specific control of gene expression.
multiple nonexclusive mechanisms. Current studies of YY1 tran- The local chromatin structure is important in determining
scription mechanisms have mostly been carried out in vitro. The whether a putative binding site is available for YY1 binding. For
Yin Yang 1 187
conducted with nuclear extracts. Therefore the observed effects for YY1 in cell proliferation and differentiation, implicating a
on YY1 could also be due to dephosphorylation of YY1-interacting potential role for YY1 during development of other systems and
proteins. For instance, the dephosphorylated retinoblastoma pro- at a later developmental stage. Consistent with the above hypoth-
tein (Rb) has been shown to bind YY1 and prevents it from bind- esis, YY1 is likely to be the mammalian homolog of the Droso-
ing to DNA while phosphorylated Rb releases YY1, permitting phila PcG protein Pho. The multiple phenotypic defects associ-
subsequent YY1 binding to DNA.110 ated with the Pho mutants suggest a similar role for YY1 during
YY1 can also be modified by glycosylation through enzymatic mammalian development through its PcG function.118 Below we
addition of N-acetylglucosamine (GlcNAc) to serine and threo- discuss the role of YY1 in cell proliferation, cell differentiation,
nine residues by nucleocytoplasmic glycosyltransferase in HeLa neural development and its PcG function.
cells.111 Glycosylation appears to disrupt the interaction between
YY1 and Rb, adding a potential mechanism by which YY1 regu- Cell Proliferation and Differentiation
lates cell cycle.111 Cell culture studies suggest that YY1 promote cell growth and
Transient poly(ADP-ribosyl)ation of YY1 has been detected proliferation by activating transcription of genes whose products
in HeLa cells immediately after genotoxic treatment.112,113 are required for cell cycle transition and progression. For instance,
Poly(ADP-ribosylation) is catalyzed by the enzyme previous studies showed that YY1 activates transcription of the
poly(ADP-ribosyl) transferase (ADPRT) and is believed to be a proto-oncogene, c-myc.119 Increased c-myc protein stimulates G1
cellular protective response to DNA damage. 114
to S transition by activating transcription of E2F2 and E2F3 which
Poly(ADP-ribosylation) decreased the binding affinity of YY1 to
subsequently activate several other molecules required for S phase
DNA and thus may prevent transcription from damaged
entry.76 YY1 also collaborates with E2F2 and E2F3 to activate
DNA.112,115 Furthermore, ADP-ribosylated YY1 may expedite
transcription of genes essential for DNA replication such as cdc6.76
the process of DNA repair mediated by ADPRT by increasing its
In addition, YY1 has been shown to activate transcription of such
enzymatic activity through protein-protein interaction.112-116
Thus, YY1 may play a role in DNA repair. genes as ribonucleotide reductase, proliferating cell nuclear anti-
YY1 can also be acetylated by p300 and PCAF (p300/CBP gen (PCNA) and the replication-dependent histones.109,120-122
associated factor) and deacetylated by HDAC1 and HDAC2.21 The protein products of these genes are either essential for S phase
Sequence analysis reveals that there are 6 lysine residues that are DNA synthesis or important for packaging newly synthesized
potential substrates for acetylation within one of the two YY1’s DNA. Consistent with a role for YY1 in cell proliferation, YY1
HDAC interaction domains (residues 170-200). Acetylation of heterozygous cells have been shown to display significantly im-
YY1 residues 170-200 might be important for YY1-mediated tran- paired cell proliferation, supporting a dosage-dependent role for
scription repression, perhaps due to an increased recruitment of YY1 in cell growth control (Shi lab, unpublished result). It should
the repressive HDACs.21 Acetylation of YY1 at residues 261-333 be noted that although the above-mentioned genes are mostly
overlaps with the DNA binding domains (residues 298-414) and engaged in G1/S transition and S phase progression, it has not
decreases YY1’s DNA-binding ability.21 In addition, HDACs do been well characterized as to how YY1 regulates different phases
not deacetylate YY1 at residues 261-333 in vitro.21 The question of cell cycle.
is how these two sets of enzymes with antagonistic functions regu- Numerous studies show that YY1 is important in the tran-
late YY1 in vivo on different promoters. scriptional control of many differentiation-specific genes in cell
Several studies suggest that YY1 can be ubiquitinated and culture, suggesting a potential role for YY1 in cell differentiation.
ubiquitin-mediated proteasomal degradation allows the cells to However, the exact role for YY1 in differentiation remains largely
rapidly reduce YY1 protein level in response to various signals. It unclear, and sometimes controversial. It appears that YY1 may
was demonstrated that purified YY1 could be readily promote cell differentiation in some systems by activating the
polyubiquitinated in reticulocyte lysates.105 Treatment with the transcription of differentiation associated gene expression. These
proteasome inhibitors MG132 and lactacystin stabilizes YY1 dur- include peripheral B lymphocyte differentiation, erythroid dif-
ing in vitro muscle cell differentiation. 105 In addition, ferentiation and myeloid differentiation.2,81,110,123-128 In contrary,
down-regulation of YY1 protein level during serum deprivation it has been shown that YY1 inhibits cell differentiation in several
induced apoptosis in cultured mouse muscle satellite cells and is cultured systems including myeloid differentiation, myogenic dif-
accompanied by the accumulation of ubiquitinated YY1.117 ferentiation, neuronal differentiation and mesenchymal differen-
In summary, it is possible that all the above-discussed mecha- tiation.73,75,102-104,129-133 This idea is supported by the observa-
nisms contribute to the regulation of YY1 activities, allowing this tion that transient down-regulation of YY1 protein level and/or
ubiquitously expressed stable protein to achieve tissue- and activity is correlated with cell differentiation,73,75,102-104,129-132
developmental-specific control of gene expression. Future studies while over-expression of YY1 promotes cell proliferation and in-
will provide us with a better understanding on how these differ- hibits cell differentiation.
ent mechanisms function and cooperate at the molecular levels. It must be pointed out that these cell culture studies may or
may not reflect the actual in vivo function of YY1. Direct genetic
The Role of YY1 During Development evidence is necessary and crucial to establishing a role for YY1 in
YY1 probably plays important roles in multiple stages of de- cell proliferation and differentiation.
velopment. The embryonic death of the constitutive YY1 knock-
out mice provided the first evidence of a crucial function for YY1 Neural Development
during embryogenesis.6 However, the molecular basis for this early Two lines of evidence strongly support a role for YY1 in the
lethality is currently unclear. The neurulation defects observed in induction and patterning of embryonic neural development. First,
a subset of YY1 heterozygous mice and the studies of YY1 on some YY1 heterozygous mouse embryos exhibit neurulation de-
Xenopus development suggest a critical role for YY1 during em- fects resembling exencephaly.6 Second, the role for YY1 in Xeno-
bryonic neural development.6 Cell culture studies support a role pus neural development is shown by both over-expression and
Yin Yang 1 189
YY1 As a PcG Protein YY1 may regulate development through its PcG mediated ac-
tivities. Knockout mice of YY1, EED and EZH2 all die at early
Both sequence analysis and functional studies suggest that YY1
embryonic stages, suggesting related developmental func-
is a mammalian counterpart of the Drosophila PcG protein Pho.
tions.6,145,146 As discussed earlier, YY1 is probably essential for
The DNA binding domain of YY1 and Pho share 96% sequence
embryonic neural development and its function relies on the in-
homology and binds to the same consensus DNA sites. 8
terplay with EED.52,134 Mutant animal studies have also demon-
Over-expression of exogenous Xenopus YY1 or Pho in Xenopus strated a critical role for several PcG proteins such as EED, EZH2
embryos caused similar phenotypes that are dependent on an- and Bim1 in mouse lymphopoiesis.135,147-149 It is possible that
other PcG protein EED.52 YY1 also partially rescue the Pho mu- YY1 may play a role in lymphocyte development through its in-
tants.7 Furthermore, in vivo YY1-mediated repression seems to teractions with these proteins.
depend on other PcG proteins including Pc (polycomb), Pcl
(polycomb-like), Scm (Sex combs on midleg), E(z) and Esc.7
PcG proteins function as high molecular weight complexes YY1 and Diseases
and regulate transcriptional repression of target genes, mainly YY1 may play a role in a myriad of pathological conditions
homeotic genes, by binding to the DNA sequence called PcG including viral infection and cancer. The YY1-related pathology
response element (PRE).135-137 There are two multimeric PcG can be the result of loss of function as demonstrated by decreased
complexes identified thus far, the human HPC-HPH complex YY1 protein level or elimination of YY1 binding sites, or a gain
and the EED-EZH2 complex, corresponding to the Drosophila of function due to YY1 over-expression or a newly created YY1
binding site. Although it is currently unclear whether the
PRC 1-complex and E(z)/Esc complex, respectively.135 It is largely
disease-associated YY1 changes are the cause of the pathogenesis,
unclear how PcG complexes are recruited to their target genes
abnormal YY1 activities are likely to contribute to the develop-
since most of the PcG proteins do not have sequence-specific
ment, progress and complications of diseases.
DNA binding ability except for the Drosophila Pho and the mam-
malian YY1 and Mel-18.7,19,138 Previous studies demonstrated
that Pho can link both PcG complexes to PREs by a direct inter- Viral Infection
action with Esc from E(z)/Esc complex and Pc, Ph from PRC YY1 has been suggested to influence infection by viruses such
1-complex.52,139 Pho-binding sites are found in most PREs such as HPV, Epstein Bar (EB) virus, human Immunodeficiency virus
as those from the engrailed, the Ultrabithorax and the Abdominal-B type I (HIV-1).48,150,151 It appears that YY1 may play a positive
promoter regions and are necessary for the PRE function in PcG role in limiting acute viral infection and inhibit viral oncogenesis
repression.8,19,136,139,140 Furthermore, immunolocalization stud- by activating multiple protective mechanisms in the host against
ies show that pho and pho-like proteins collaborate to anchor Pc, viral infection and by inhibiting viral gene transcription. For in-
stance, YY1 induces transcription of β-interferon, one of the first
Psc, and Scm to some polytene chromosomal sites.19 YY1 also
responses of cell to viral infection.152 YY1 may help decrease viral
interacts with EED directly and may also interact with the
infection by down-regulating cell surface expression of viral re-
HPC-HPH complexes indirectly through bridging proteins such
ceptors as demonstrated in the HIV infection.153-155 YY1 may
as CtBP or RYBP.7,52,76,93,141 Thus, one attractive hypothesis is reduce virion production by repressing viral immediate early gene
that YY1, the mammalian homologue of Pho, may also tether expression and by interfering with viral protein function through
PcG proteins to target genes as illustrated in Figure 5. It is re- protein-protein interaction.48,55,150,156-159 All these different
ported that YY1 binds the promoters of several homeobox genes host-responses mediated by YY1 may contribute to viral latency
such as gp91, Msx2, Hoxb4 and Hoxb7.128,142-144 However, YY1 control. YY1 may also inhibit virus-induced malignant transfor-
seems to activate, rather than repress, transcription of gp91 and mation by repressing viral oncogene expression such as HPV E6
Msx2 and therefore contributes to myeloid differentiation and and E7.151,160,161 An effective mechanism developed by the virus
early mouse craniofacial and limb morphogenesis.21,142 Future is to create mutations at YY1 binding sites in the viral pro-
experiments that examine the function of YY1 in mammalian moter region as identified in many HPV 16-positive cervical
PcG repression will provide more direct evidence to resolve this tumors.160,162-164 These mutations decrease or eliminate YY1
issue. binding and are expected to release the YY1-mediated repression,
190 Zinc Finger Proteins: From Atomic Contact to Cellular Function
22. Austen M, Luscher B, Luscher-Firzlaff JM. Characterization of the 41. Chan PK, Chan FY, Morris SW et al. Isolation and characterization
transcriptional regulator YY1. The bipartite transactivation domain of the human nucleophosmin/B23 (NPM) gene: Identification of
is independent of interaction with the TATA box-binding protein, the YY1 binding site at the 5' enhancer region. Nucleic Acids Res
transcription factor IIB, TAFII55, or cAMP-responsive 1997; 25:1225-32.
element-binding protein (CPB)-binding protein. J Biol Chem 1997; 42. Ficzycz A, Eskiw C, Meyer D et al. Expression, activity, and
272:1709-17. subcellular localization of the Yin Yang 1 transcription factor in
23. Xiao H, Friesen JD, Lis JT. A highly conserved domain of RNA Xenopus oocytes and embryos. J Biol Chem 2001; 276:22819-25.
polymerase II shares a functional element with acidic activation 43. Ficzycz A, Ovsenek N. The Yin Yang 1 transcription factor associates
domains of upstream transcription factors. Molecular & Cellular with ribonucleoprotein (mRNP) complexes in the cytoplasm of
Biology 1994; 14:7507-16. Xenopus oocytes. J Biol Chem 2002; 277:8382-7.
24. Brou C, Wu J, Ali S et al. Different TBP-associated factors are 44. Thomas MJ, Seto E. Unlocking the mechanisms of transcription
required for mediating the stimulation of transcription in vitro by factor YY1: Are chromatin modifying enzymes the key? Gene 1999;
the acidic transactivator GAL-VP16 and the two nonacidic activation 236:197-208.
functions of the estrogen receptor. Nucleic Acids Research 1993; 45. Peterson CL. Chromatin remodeling enzymes: Taming the machines.
21:5-12. Third in review series on chromatin dynamics. EMBO Reports 2002;
25. Defossez PA, Baert JL, Monnot M et al. The ETS family member 3:319-22.
ERM contains an alpha-helical acidic activation domain that contacts 46. Nielsen AL, Oulad-Abdelghani M et al. Heterochromatin formation
TAFII60. Nucleic Acids Research 1997; 25:4455-63. in mammalian cells: Interaction between histones and HP1 proteins.
26. Kim TK, Roeder RG. Proline-rich activator CTF1 targets the TFIIB Molecular Cell 2001; 7:729-39.
assembly step during transcriptional activation. Proceedings of the 47. Galarneau L, Nourani A, Boudreault AA et al. Multiple links between
National Academy of Sciences of the United States of America 1994; the NuA4 histone acetyltransferase complex and epigenetic control
91:4170-4. of transcription. Molecular Cell 2000; 5:927-37.
27. Bushmeyer S, Park K, Atchison ML. Characterization of functional 48. He G, Margolis DM. Counterregulation of chromatin deacetylation
domains within the multifunctional transcription factor, YY1. J Biol and histone deacetylase occupancy at the integrated promoter of
Chem 1995; 270:30213-20. human immunodeficiency virus type 1 (HIV-1) by the HIV-1
28. Lee JS, See RH, Galvin KM et al. Functional interactions between repressor YY1 and HIV-1 activator Tat. Mol Cell Biol 2002;
YY1 and adenovirus E1A. Nucleic Acids Res 1995; 23:925-31. 22:2965-73.
29. Yant SR, Zhu W, Millinoff D et al. High affinity YY1 binding motifs: 49. Yang WM, Inouye C, Zeng Y et al. Transcriptional repression by
Identification of two core types (ACAT and CCAT) and distribution YY1 is mediated by interaction with a mammalian homolog of the
of potential binding sites within the human beta globin cluster. yeast global regulator RPD3. Proc Natl Acad Sci USA 1996;
Nucleic Acids Res 1995; 23:4353-62. 93:12845-50.
30. Hyde-DeRuyscher RP, Jennings E, Shenk T. DNA binding sites for 50. Rezai-Zadeh N, Zhang X, Namour F et al. Targeted recruitment of
the transcriptional activator/repressor YY1. Nucleic Acids Res 1995; a histone H4-specific methyltransferase by the transcription factor
23:4457-65. YY1. Genes Dev 2003; 17:1019-29.
31. Houbaviy HB, Usheva A, Shenk T et al. Cocrystal structure of YY1 51. Shi Y, Sawada J, Sui G et al. Coordinated histone modifications
bound to the adeno-associated virus P5 initiator. Proc Natl Acad mediated by a CtBP co-repressor complex. Nature 2003; 422:735-8.
Sci USA 1996; 93:13577-82. 52. Satijn DP, Hamer KM, den Blaauwen J et al. The polycomb group
32. Ma SL, Lovmand J, Sorensen AB et al. Triple basepair changes within protein EED interacts with YY1, and both proteins induce neural
and adjacent to the conserved YY1 motif upstream of the U3 tissue in Xenopus embryos. Mol Cell Biol 2001; 21:1360-9.
enhancer repeats of SL3-3 murine leukemia virus cause a small but 53. Ogawa H, Ishiguro K, Gaubatz S et al. A complex with chromatin
significant shortening of latency of T-lymphoma induction. Virology modifiers that occupies E2F- and Myc-responsive genes in G0 cells.
2003; 313:638-44. Science 2002; 296:1132-6.
33. Sitwala KV, Adams K, Markovitz DM. YY1 and NF-Y binding sites 54. Lee JS, Galvin KM, See RH et al. Relief of YY1 transcriptional
regulate the transcriptional activity of the dek and dek-can promoter. repression by adenovirus E1A is mediated by E1A-associated protein
Oncogene 2002; 21:8862-70. p300. Genes Dev 1995; 9:1188-98.
34. Raval-Pandya M, Dhawan P, Barletta F et al. YY1 represses vitamin 55. Coull JJ, Romerio F, Sun JM et al. The human factors YY1 and
D receptor-mediated 25-hydroxyvitamin D(3)24-hydroxylase LSF repress the human immunodeficiency virus type 1 long terminal
transcription: Relief of repression by CREB-binding protein. Mol repeat via recruitment of histone deacetylase 1. J Virol 2000;
Endocrinol 2001; 15:1035-46. 74:6790-9.
35. Satyamoorthy K, Park K, Atchison ML et al. The intracisternal 56. Osborne A, Zhang H, Yang WM et al. Histone deacetylase activity
A-particle upstream element interacts with transcription factor YY1 represses gamma interferon-inducible HLA-DR gene expression
to activate transcription: Pleiotropic effects of YY1 on distinct DNA following the establishment of a DNase I-hypersensitive chromatin
promoter elements. Mol Cell Biol 1993; 13:6621-8. conformation. Mol Cell Biol 2001; 21:6495-506.
36. L’Honore A, Lamb NJ, Vandromme M et al. MyoD distal regulatory 57. Huang NE, Lin CH, Lin YS et al. Modulation of YY1 activity by
region contains an SRF binding CArG element required for MyoD SAP30. Biochemical & Biophysical Research Communications 2003;
expression in skeletal myoblasts and during muscle regeneration. Mol 306:267-75.
Biol Cell 2003; 14:2151-62. 58. Yao YL, Yang WM. The metastasis-associated proteins 1 and 2 form
37. McNeil S, Guo B, Stein JL et al. Targeting of the YY1 transcription distinct protein complexes with histone deacetylase activity. J Biol
factor to the nucleolus and the nuclear matrix in situ: The Chem 2003; 278:42560-8.
C-terminus is a principal determinant for nuclear trafficking. J Cell 59. Yang WM, Yao YL, Seto E. The FK506-binding protein 25
Biochem 1998; 68:500-10. functionally associates with histone deacetylases and with transcription
38. Guo B, Odgren PR, van Wijnen AJ et al. The nuclear matrix protein factor YY1. Embo J 2001; 20:4814-25.
NMP-1 is the transcription factor YY1. Proc Natl Acad Sci USA 60. Galasinski SC, Resing KA, Goodrich JA et al. Phosphatase inhibition
1995; 92:10526-30. leads to histone deacetylases 1 and 2 phosphorylation and disruption
39. Inouye CJ, Seto E. Relief of YY1-induced transcriptional repression of corepressor interactions. J Biol Chem 2002; 277:19618-26.
by protein-protein interaction with the nucleolar phosphoprotein 61. Sewalt RG, Lachner M, Vargas M et al. Selective interactions between
B23. J Biol Chem 1994; 269:6506-10. vertebrate polycomb homologs and the SUV39H1 histone lysine
40. Bushmeyer SM, Atchison ML. Identification of YY1 sequences methyltransferase suggest that histone H3-K9 methylation contributes
necessary for association with the nuclear matrix and for to chromosomal targeting of Polycomb group proteins. Molecular
transcriptional repression functions. J Cell Biochem 1998; 68:484-99. & Cellular Biology 2002; 22:5539-53.
192 Zinc Finger Proteins: From Atomic Contact to Cellular Function
62. Kuzmichev A, Nishioka K, Erdjument-Bromage H et al. Histone 86. Costa M, Grant PJ, Rice GI et al. Human endothelial cell-derived
methyltransferase activity associated with a human multiprotein nuclear proteins that recognise polymorphic DNA elements in the
complex containing the Enhancer of Zeste protein. Genes & von Willebrand factor gene promoter include YY1. Thromb Haemost
Development 2002; 16:2893-905. 2001; 86:672-9.
63. Cao R, Wang L, Wang H et al. Role of histone H3 lysine 27 87. Field JM, Tate LA, Chipman JK et al. Identification of functional
methylation in Polycomb-group silencing. Science 2002; regulatory regions of the connexin32 gene promoter. Biochim
298:1039-43. Biophys Acta 2003; 1628:22-9.
64. Wang H, Huang ZQ, Xia L et al. Methylation of histone H4 at 88. Hines RN, Luo Z, Hopp KA et al. Genetic variability at the human
arginine 3 facilitating transcriptional activation by nuclear hormone FMO1 locus: Significance of a basal promoter yin yang 1 element
receptor. Science 2001; 293:853-7. polymorphism (FMO1*6). J Pharmacol Exp Ther 2003; 306:1210-8.
65. Sewalt RG, Gunster MJ, van der Vlag J et al. C-Terminal binding 89. de Souza AP, Trevilatto PC, Scarel-Caminaga RM et al. Analysis of
protein is a transcriptional repressor that interacts with a specific the TGF-beta1 promoter polymorphism (C-509T) in patients with
class of vertebrate Polycomb proteins. Molecular & Cellular Biology chronic periodontitis. J Clin Periodontol 2003; 30:519-23.
1999; 19:777-87. 90. Rao A. New functions for DNA binding domains. Sci STKE 2001;
66. Seto E, Shi Y, Shenk T. YY1 is an initiator sequence-binding protein 2001:E1.
that directs and activates transcription in vitro. Nature 1991; 91. Lefstin JA, Yamamoto KR. Allosteric effects of DNA on
354:241-5. transcriptional regulators. Nature 1998; 392:885-8.
67. Usheva A, Shenk T. YY1 transcriptional initiator: Protein interactions 92. Austen M, Cerni C, Luscher-Firzlaff JM et al. YY1 can inhibit c-Myc
and association with a DNA site containing unpaired strands. Proc function through a mechanism requiring DNA binding of YY1 but
Natl Acad Sci USA 1996; 93:13571-6. neither its transactivation domain nor direct interaction with c-Myc.
68. Usheva A, Shenk T. TATA-binding protein-independent initiation: Oncogene 1998; 17:511-20.
YY1, TFIIB, and RNA polymerase II direct basal transcription on 93. Garcia E, Marcos-Gutierrez C, del Mar Lorente M et al. RYBP, a
supercoiled template DNA. Cell 1994; 76:1115-21. new repressor protein that interacts with components of the
69. Lee JS, Galvin KM, Shi Y. Evidence for physical interaction between mammalian Polycomb complex, and with the transcription factor
the zinc-finger transcription factors YY1 and Sp1. Proc Natl Acad YY1. Embo J 1999; 18:3404-18.
Sci USA 1993; 90:6145-9. 94. Lee JS, Zhang X, Shi Y. Differential interactions of the CREB/ATF
70. Seto E, Lewis B, Shenk T. Interaction between transcription factors family of transcription factors with p300 and adenovirus E1A. J
Sp1 and YY1. Nature 1993; 365:462-4. Biol Chem 1996; 271:17666-74.
71. Maldonado E, Shiekhattar R, Sheldon M et al. A human RNA 95. Sawa C, Yoshikawa T, Matsuda-Suzuki F et al. YEAF1/RYBP and
polymerase II complex associated with SRB and DNA-repair YAF-2 are functionally distinct members of a cofactor family for
proteins.[erratum appears in Nature 1996 Nov 28;384(6607):384]. the YY1 and E4TF1/hGABP transcription factors. J Biol Chem 2002;
Nature 1996; 381:86-9. 277:22484-90.
72. Chiang CM, Roeder RG. Cloning of an intrinsic human TFIID 96. Kalenik JL, Chen D, Bradley ME et al. Yeast two-hybrid cloning of
subunit that interacts with multiple transcriptional activators. Science a novel zinc finger protein that interacts with the multifunctional
1995; 267:531-6. transcription factor YY1. Nucleic Acids Res 1997; 25:843-9.
73. Lee TC, Shi Y, Schwartz RJ. Displacement of BrdUrd-induced YY1 97. Bauknecht T, Angel P, Royer HD et al. Identification of a negative
by serum response factor activates skeletal alpha-actin transcription regulatory domain in the human papillomavirus type 18 promoter:
in embryonic myoblasts. Proc Natl Acad Sci USA 1992; 89:9814-8. Interaction with the transcriptional repressor YY1. Embo J 1992;
74. Lu SY, Rodriguez M, Liao WS. YY1 represses rat serum amyloid 11:4607-17.
A1 gene transcription and is antagonized by NF-kappa B during 98. Bauknecht T, Jundt F, Herr I et al. A switch region determines the
acute-phase response. Mol Cell Biol 1994; 14:6253-63. cell type-specific positive or negative action of YY1 on the activity
75. Gualberto A, LePage D, Pons G et al. Functional antagonism between of the human papillomavirus type 18 promoter. J Virol 1995;
YY1 and the serum response factor. Mol Cell Biol 1992; 12:4209-14. 69:1-12.
76. Schlisio S, Halperin T, Vidal M et al. Interaction of YY1 with E2Fs, 99. Bauknecht T, See RH, Shi Y. A novel C/EBP beta-YY1 complex
mediated by RYBP, provides a mechanism for specificity of E2F controls the cell-type-specific activity of the human papillomavirus
function. Embo J 2002; 21:5775-86. type 18 upstream regulatory region. J Virol 1996; 70:7695-705.
77. Galvin KM, Shi Y. Multiple mechanisms of transcriptional repression 100. Yevtodiyenko A, Carr MS, Patel N et al. Analysis of candidate
by YY1. Mol Cell Biol 1997; 17:3723-32. imprinted genes linked to Dlk1-Gtl2 using a congenic mouse line.
78. Bennett MK, Ngo TT, Athanikar JN et al. Co-stimulation of Mamm Genome 2002; 13:633-8.
promoter for low density lipoprotein receptor gene by sterol 101. Safrany G, Perry RP. Characterization of the mouse gene that encodes
regulatory element-binding protein and Sp1 is specifically disrupted the delta/YY1/NF-E1/UCRBP transcription factor. Proc Natl Acad
by the yin yang 1 protein. J Biol Chem 1999; 274:13025-32. Sci USA 1993; 90:5559-63.
79. Natesan S, Gilman MZ. DNA bending and orientation-dependent 102. Erkeland SJ, Valkhof M, Heijmans-Antonissen C et al. The gene
function of YY1 in the c-fos promoter. Genes Dev 1993; 7:2497-509. encoding the transcriptional regulator Yin Yang 1 (YY1) is a myeloid
80. Kim J, Shapiro DJ. In simple synthetic promoters YY1-induced DNA transforming gene interfering with neutrophilic differentiation. Blood
bending is important in transcription activation and repression. 2003; 101:1111-7.
Nucleic Acids Res 1996; 24:4341-8. 103. Lee TC, Zhang Y, Schwartz RJ. Bifunctional transcriptional properties
81. Roque MC, Smith PA, Blasquez VC. A developmentally modulated of YY1 in regulating muscle actin and c-myc gene expression during
chromatin structure at the mouse immunoglobulin kappa 3' enhancer. myogenesis. Oncogene 1994; 9:1047-52.
Mol Cell Biol 1996; 16:3138-55. 104. Pizzorno MC. Nuclear cathepsin B-like protease cleaves transcription
82. Kim J, Kollhoff A, Bergmann A et al. Methylation-sensitive binding factor YY1 in differentiated cells. Biochim Biophys Acta 2001;
of transcription factor YY1 to an insulator sequence within the 1536:31-42.
paternally expressed imprinted gene, Peg3. Hum Mol Genet 2003; 105. Walowitz JL, Bradley ME, Chen S et al. Proteolytic regulation of
12:233-45. the zinc finger transcription factor YY1, a repressor of
83. Gaston K, Fried M. CpG methylation has differential effects on the muscle-restricted gene expression. J Biol Chem 1998; 273:6656-61.
binding of YY1 and ETS proteins to the bi-directional promoter of 106. Bovolenta C, Camorali L, Lorini AL et al. In vivo administration of
the Surf-1 and Surf-2 genes. Nucleic Acids Res 1995; 23:901-9. recombinant IL-2 to individuals infected by HIV down-modulates
84. Hobbs K, Negri J, Klinnert M et al. Interleukin-10 and transforming the binding and expression of the transcription factors ying-yang-1
growth factor-beta promoter polymorphisms in allergies and asthma. and leader binding protein-1/late simian virus 40 factor. J Immunol
Am J Respir Crit Care Med 1998; 158:1958-62. 1999; 163:6892-7.
85. Pulleyn LJ, Newton R, Adcock IM et al. TGFbeta1 allele association
with asthma severity. Hum Genet 2001; 109:623-7.
Yin Yang 1 193
107. Becker KG, Jedlicka P, Templeton NS et al. Characterization of 130. Patten M, Hartogensis WE, Long CS. Interleukin-1beta is a negative
hUCRBP (YY1, NF-E1, delta): A transcription factor that binds transcriptional regulator of alpha1-adrenergic induced gene expression
the regulatory regions of many viral and cellular genes. Gene 1994; in cultured cardiac myocytes. J Biol Chem 1996; 271:21134-41.
150:259-66. 131. Liu SH, Peng BH, Ma JT et al. Serum response element associated
108. Patten M, Wang W, Aminololama-Shakeri S et al. IL-1 beta increases transcription factors in mouse embryos: Serum response factor, YY1,
abundance and activity of the negative transcriptional regulator yin and PEA3 factor. Dev Genet 1995; 16:229-40.
yang-1 (YY1) in neonatal rat cardiac myocytes. J Mol Cell Cardiol 132. Chen CY, Schwartz RJ. Competition between negative acting YY1
2000; 32:1341-52. versus positive acting serum response factor and tinman homologue
109. Eliassen KA, Baldwin A, Sikorski EM et al. Role for a YY1-binding Nkx-2.5 regulates cardiac alpha-actin promoter activity. Mol
element in replication-dependent mouse histone gene expression. Mol Endocrinol 1997; 11:812-22.
Cell Biol 1998; 18:7106-18. 133. Kurisaki K, Kurisaki A, Valcourt U et al. Nuclear factor YY1 inhibits
110. Gordon SJ, Saleque S, Birshtein BK. Yin Yang 1 is a transforming growth factor beta- and bone morphogenetic
lipopolysaccharide-inducible activator of the murine 3' Igh enhancer, protein-induced cell differentiation. Molecular & Cellular Biology
hs3. J Immunol 2003; 170:5549-57. 2003; 23:4494-510.
111. Hiromura M, Choi CH, Sabourin NA et al. YY1 is regulated by 134. Kwon HJ, Chung HM. Yin Yang 1, a vertebrate polycomb group
O-linked N-acetylglucosaminylation (O-glcNAcylation). J Biol Chem gene, regulates antero-posterior neural patterning. Biochem Biophys
2003; 278:14046-52. Res Commun 2003; 306:1008-13.
112. Oei SL, Griesenbeck J, Schweiger M et al. Regulation of RNA 135. Raaphorst FM, Otte AP, Meijer CJ. Polycomb-group genes as
Polymerase II-dependent Transcription by Poly(ADP-ribosyl)ation of regulators of mammalian lymphopoiesis. Trends Immunol 2001;
Transcription Factors. J Biol Chem 1998; 273:31644-31647. 22:682-90.
113. Oei SL, Shi Y. Poly(ADP-ribosyl)ation of transcription factor Yin 136. Brock HW, van Lohuizen M. The Polycomb group—no longer an
Yang 1 under conditions of DNA damage. Biochem Biophys Res exclusive club? Current Opinion in Genetics & Development 2001;
Commun 2001; 285:27-31. 11:175-81.
114. Oei SL, Griesenbeck J, Schweiger M et al. Interaction of the 137. Gould A. Functions of mammalian Polycomb group and trithorax
transcription factor YY1 with human poly(ADP-ribosyl) transferase. group related genes. Current Opinion in Genetics & Development
Biochem Biophys Res Commun 1997; 240:108-11. 1997; 7:488-94.
115. Griesenbeck J, Ziegler M, Tomilin N et al. Stimulation of the 138. Kanno M, Hasegawa M, Ishida A et al. mel-18, a Polycomb
catalytic activity of poly(ADP-ribosyl) transferase by transcription group-related mammalian gene, encodes a transcriptional negative
factor Yin Yang 1. FEBS Lett 1999; 443:20-4. regulator with tumor suppressive activity. EMBO Journal 1995;
116. Oei SL, Shi Y. Transcription factor Yin Yang 1 stimulates 14:5672-8.
poly(ADP-ribosyl)ation and DNA repair. Biochem Biophys Res 139. Mohd-Sarip A, Venturini F, Chalkley GE et al. Pleiohomeotic can
Commun 2001; 284:450-4. link polycomb to DNA and mediate transcriptional repression.
117. Mampuru LJ, Chen SJ, Kalenik JL et al. Analysis of events associated Molecular & Cellular Biology 2002; 22:7473-83.
with serum deprivation-induced apoptosis in C3H/Sol8 muscle 140. Mishra RK, Mihaly J, Barges S et al. The iab-7 polycomb response
satellite cells. Exp Cell Res 1996; 226:372-80. element maps to a nucleosome-free region of chromatin and requires
118. Girton JR, Jeon SH. Novel embryonic and adult homeotic both GAGA and pleiohomeotic for silencing activity. Mol Cell Biol
phenotypes are produced by pleiohomeotic mutations in Drosophila. 2001; 21:1311-8.
Dev Biol 1994; 161:393-407. 141. Trimarchi JM, Fairchild B, Wen J et al. The E2F6 transcription
119. Riggs KJ, Saleque S, Wong KK et al. Yin-yang 1 activates the c-myc factor is a component of the mammalian Bmi1-containing polycomb
promoter. Mol Cell Biol 1993; 13:7487-95. complex. Proceedings of the National Academy of Sciences of the
120. Wu F, Lee AS. YY1 as a regulator of replication-dependent hamster United States of America 2001; 98:1519-24.
histone H3.2 promoter and an interactive partner of AP-2. J Biol 142. Tan DP, Nonaka K, Nuckolls GH et al. YY1 activates Msx2 gene
Chem 2001; 276:28-34. independent of bone morphogenetic protein signaling. Nucleic Acids
121. Johansson E, Hjortsberg K, Thelander L. Two YY-1-binding proximal Res 2002; 30:1213-23.
elements regulate the promoter strength of the TATA-less mouse 143. Meccia E, Bottero L, Felicetti F et al. HOXB7 expression is regulated
ribonucleotide reductase R1 gene. J Biol Chem 1998; 273:29816-21. by the transcription factors NF-Y, YY1, Sp1 and USF-1. Biochim
122. Labrie C, Lee BH, Mathews MB. Transcription factors RFX1/EF-C Biophys Acta 2003; 1626:1-9.
and ATF-1 associate with the adenovirus E1A-responsive element of 144. Gilthorpe J, Vandromme M, Brend T et al. Spatially specific
the human proliferating cell nuclear antigen promoter. Nucleic Acids expression of Hoxb4 is dependent on the ubiquitous transcription
Res 1995; 23:3732-41. factor NFY. Development 2002; 129:3887-99.
123. Hehlgans T, Strominger JL. Activation of transcription by binding 145. O’Carroll D, Erhardt S, Pagani M et al. The polycomb-group gene
of NF-E1 (YY1) to a newly identified element in the first exon of Ezh2 is required for early mouse development. Mol Cell Biol 2001;
the human DR alpha gene. J Immunol 1995; 154:5181-7. 21:4330-6.
124. Zabel MD, Wheeler W, Weis JJ et al. Yin Yang 1, Oct1, and NFAT-4 146. Faust C, Lawson KA, Schork NJ et al. The Polycomb-group gene
form repeating, cyclosporin-sensitive regulatory modules within the eed is required for normal morphogenetic movements during
murine CD21 intronic control region. J Immunol 2002; gastrulation in the mouse embryo. Development 1998;
168:3341-50. 125:4495-506.
125. Raich N, Clegg CH, Grofti J et al. GATA1 and YY1 are 147. Lessard J, Schumacher A, Thorsteinsdottir U et al. Functional
developmental repressors of the human epsilon-globin gene. Embo J antagonism of the Polycomb-Group genes eed and Bmi1 in
1995; 14:801-9. hemopoietic cell proliferation. Genes Dev 1999; 13:2691-703.
126. Peters B, Merezhinskaya N, Diffley JF et al. Protein-DNA 148. Su IH, Basavaraj A, Krutchinsky AN et al. Ezh2 controls B cell
interactions in the epsilon-globin gene silencer. J Biol Chem 1993; development through histone H3 methylation and Igh rearrangement.
268:3430-7. Nat Immunol 2003; 4:124-31.
127. Wandersee NJ, Ferris RC, Ginder GD. Intronic and flanking 149. Lessard J, Sauvageau G, Kwon HJ et al. Polycomb group genes as
sequences are required to silence enhancement of an embryonic epigenetic regulators of normal and leukemic hemopoiesis. Exp
beta-type globin gene. Mol Cell Biol 1996; 16:236-46. Hematol 2003; 31:567-85.
128. Jacobsen BM, Skalnik DG. YY1 binds five cis-elements and 150. Zalani S, Coppage A, Holley-Guthrie E et al. The cellular YY1
trans-activates the myeloid cell-restricted gp91(phox) promoter. J Biol transcription factor binds a cis-acting, negatively regulating element
Chem 1999; 274:29984-93. in the Epstein-Barr virus BRLF1 promoter. J Virol 1997; 71:3268-74.
129. Galvagni F, Cartocci E, Oliviero S. The dystrophin promoter is
negatively regulated by YY1 in undifferentiated muscle cells. J Biol
Chem 1998; 273:33708-13.
194 Zinc Finger Proteins: From Atomic Contact to Cellular Function
151. Dong X, Liu H, Pfister H. YY1 and its repressive effect on human 166. Dong XP, Stubenrauch F, Beyer-Finkler E et al. Prevalence of
papillomavirus 16 early promoter P97 existed widely among human deletions of YY1-binding sites in episomal HPV 16 DNA from
epithelial cell lines. Chinese Journal of Experimental & Clinical cervical cancers. Int J Cancer 1994; 58:803-8.
Virology 1998; 12:217-22. 167. May M, Dong XP, Beyer-Finkler E et al. The E6/E7 promoter of
152. Weill L, Shestakova E, Bonnefoy E. Transcription factor YY1 binds extrachromosomal HPV16 DNA in cervical cancers escapes from
to the murine beta interferon promoter and regulates its cellular repression by mutation of target sequences for YY1. Embo J
transcriptional capacity with a dual activator/repressor role. J Virol 1994; 13:1460-6.
2003; 77:2903-14. 168. Parija T, Das BR. Involvement of YY1 and its correlation with c-myc
153. Moriuchi M, Moriuchi H, Margolis DM et al. USF/c-Myc enhances, in NDEA induced hepatocarcinogenesis, its prevention by
while Yin-Yang 1 suppresses, the promoter activity of CXCR4, a d-limonene. Mol Biol Rep 2003; 30:41-6.
coreceptor for HIV-1 entry. J Immunol 1999; 162:5986-92. 169. Nayak BK, Das BR. Differential binding of NF1 transcription factor
154. Moriuchi M, Moriuchi H. YY1 transcription factor down-regulates to P53 gene promoter and its depletion in human breast tumours.
expression of CCR5, a major co-receptor for HIV-1. J Biol Chem Mol Biol Rep 1999; 26:223-30.
2003; 5:5. 170. Kalantari M, Blennow E, Hagmar B et al. Physical state of HPV16
155. Cristillo AD, Bierer BE. Regulation of CXCR4 expression in human and chromosomal mapping of the integrated form in cervical
T lymphocytes by calcium and calcineurin. Mol Immunol 2003; carcinomas. Diagn Mol Pathol 2001; 10:46-54.
40:539-53. 171. Yengi L, Inskip A, Gilford J et al. Polymorphism at the glutathione
156. Margolis DM, Somasundaran M, Green MR. Human transcription S-transferase locus GSTM3: Interactions with cytochrome P450 and
factor YY1 represses human immunodeficiency virus type 1 glutathione S-transferase genotypes as risk factors for multiple
transcription and virion production. J Virol 1994; 68:905-10. cutaneous basal cell carcinoma. Cancer Res 1996; 56:1974-7.
157. Romerio F, Gabriel MN, Margolis DM. Repression of human 172. Rose BR, Thompson CH, Zhang J et al. Sequence variation in the
immunodeficiency virus type 1 through the novel cooperation of upstream regulatory region of HPV 18 isolates from cervical cancers.
human factors YY1 and LSF. J Virol 1997; 71:9375-82. Gynecol Oncol 1997; 66:282-9.
158. Lee D, Kim H, Lee Y et al. Identification of sequence requirement 173. Shrivastava A, Saleque S, Kalpana GV et al. Inhibition of
for the origin of DNA replication in human papillomavirus type transcriptional regulator Yin-Yang-1 by association with c-Myc.
18. Virus Res 1997; 52:97-108. Science 1993; 262:1889-92.
159. Lee KY, Broker TR, Chow LT. Transcription factor YY1 represses 174. Austen M, Cerni C, Henriksson M et al. Regulation of cell growth
cell-free replication from human papillomavirus origins. J Virol 1998; by the Myc-Max-Mad network: Role of Mad proteins and YY1. Curr
72:4911-7. Top Microbiol Immunol 1997; 224:123-30.
160. Park JS, Hwang ES, Lee CJ et al. Mutational and functional analysis 175. Shrivastava A, Yu J, Artandi S, Calame K. YY1 and c-Myc associate
of HPV-16 URR derived from Korean cervical neoplasia. Gynecol in vivo in a manner that depends on c-Myc levels. Proc Natl Acad
Oncol 1999; 74:23-9. Sci USA 1996; 93:10638-41.
161. O’Connor MJ, Tan SH, Tan CH et al. YY1 represses human 176. Wang J, Mager J, Chen Y et al. Imprinted X inactivation maintained
papillomavirus type 16 transcription by quenching AP-1 activity. J by a mouse Polycomb group gene. Nat Genet 2001; 28:371-5.
Virol 1996; 70:6529-39. 177. Klenova EM, Fagerlie S, Filippova GN et al. Characterization of the
162. Schmidt M, Kedzia W, Gozdzicka-Jozefiak A. Intratype HPV16 chicken CTCF genomic locus, and initial study of the cell
sequence variation within LCR of isolates from asymptomatic carriers cycle-regulated promoter of the gene. J Biol Chem 1998;
and cervical cancers. J Clin Virol 2001; 23:65-77. 273:26571-9.
163. Veress G, Murvai M, Szarka K et al. Transcriptional activity of human 178. Furlong EE, Rein T, Martin F. YY1 and NF1 both activate the
papillomavirus type 16 variants having deletions in the long control human p53 promoter by alternatively binding to a composite
region. Eur J Cancer 2001; 37:1946-52. element, and YY1 and E1A cooperate to amplify p53 promoter
164. Stephen AL, Thompson CH, Tattersall MH et al. Analysis of activity. Mol Cell Biol 1996; 16:5933-45.
mutations in the URR and E6/E7 oncogenes of HPV 16 cervical 179. Gabellini D, Green MR, Tupler R. Inappropriate gene activation in
cancer isolates from central China. Int J Cancer 2000; 86:695-701. FSHD: A repressor complex binds a chromosomal repeat deleted in
165. Kozuka T, Aoki Y, Nakagawa K et al. Enhancer-promoter activity of dystrophic muscle. Cell 2002; 110:339-48.
human papillomavirus type 16 long control regions isolated from 180. Romey MC, Pallares-Ruiz N, Mange A et al. A naturally occurring
cell lines SiHa and CaSki and cervical cancer biopsies. Jpn J Cancer sequence variation that creates a YY1 element is associated with
Res 2000; 91:271-9. increased cystic fibrosis transmembrane conductance regulator gene
expression. J Biol Chem 2000; 275:3561-7.
CHAPTER 26
Abstract
T
ranscription factor IIIA (TFIIIA) is a single polypeptide protein with a single function in a simple organism can have
with several distinct functions in the cell. In this chapter multiple functions in the cells of more complex organisms. This
I will review the role of TFIIIA in transcription model runs counter to the classic “one gene, one protein
initiation of the 5S rRNA gene. I will also describe the model by function” idea from the Beadle and Tatum experiments in Neuro-
which TFIIIA associates with the 5S rRNA itself to regulate spora, but there are several examples of “multitasking” proteins,8,9
ribosome biosynthesis and assembly. Finally, I will compare these including TFIIIA.
functions of TFIIIA in various eukaryotic cells, from yeast to TFIIIA was the first eukaryotic transcription factor to be
vertebrate. In all cases, the part played by the protein’s zinc purified and it is required for the in vitro transcription of 5S
fingers will be emphasized. rRNA gene.10 A second, distinct function was subsequently
assigned to TFIIIA, namely rRNA localization and storage.11 Both
Introduction functions of TFIIIA modulate 5S rRNA availability. TFIIIA, in
its role as a DNA-binding transcription factor, modulates the
Although it is notoriously difficult to identify genes from
synthesis of 5S rRNA and then, in its role as an RNA-binding
sequencing data,1,2 whole-genome analyses have led to the
nuclear/cytoplasmic shuttle, controls the assembly of 5S rRNA
unexpected conclusion that an organism’s absolute number of
into ribosomes. The zinc fingers of TFIIIA (Fig. 1) are critical for
genes is not directly related to its morphological, developmental
both its DNA and RNA binding activities.
or behavioral complexity. Humans, whose genes are estimated at
30,000 or fewer3,4 seem more than twice as complex as flies or
worms, whose genes are estimated at fewer than 14,000 for D. DNA Binding by TFIIIA and Transcription
melanogaster5 and 20,000 for C. elegans.6 Not only are there of the 5S rRNA Gene
unexpectedly few protein-encoding genes in metazoan genomes The work of transcription in a eukaryotic cell is divided among
but also their sequences represent only a small fraction of the three RNA polymerases. In general, RNA polymerase I is used
total. A mere 2% of the human genome is thought to encode for transcription of rRNAs, while RNA polymerase II is the
proteins.7 enzyme that transcribes protein-encoding genes, and RNA
If an organism’s complexity does not arise from the absolute polymerase III primarily transcribes tRNAs and small nuclear
number of proteins in its genome or the amount of genomic “real RNAs. Each multisubunit polymerase requires additional
estate” that is dedicated to them, then perhaps it is their proteins for accurate initiation of transcription. These proteins,
regulation that accounts for interspecies diversity. Perhaps increas- collectively called “basal” or “general” transcription factors, tether
ing complexity arises from a limited number of proteins with and position the polymerase at a gene’s promoter for accurate
increasingly complex expression patterns. Indeed, the more and efficient transcription initiation. TFIIIA is one such basal
complex an organism, the more transcription regulatory proteins transcription factor.
are encoded by its genome,7 and since regulatory proteins often Unlike most rRNAs, the 5S rRNA is transcribed by RNA
multimerize with other regulatory proteins, a small increase in polymerase III in conjunction with the transcription factors
the number of regulatory proteins could give rise to a significant TFIIIA, TFIIIB, and TFIIIC (Fig. 2). Transcription begins with
number of alternative regulatory molecules. Also consistent with the recognition of a 5S DNA sequence by TFIIIA. This recogni-
this “few proteins, greater regulation” model is the observation tion event is unusual in several ways. First, TFIIIA is used
that, in complex organisms, the regulatory sequences upstream exclusively for transcription of the 5S rRNA gene. No other genes
of protein-encoding genes are more dense with binding sites for transcribed by RNA polymerase III are known to require TFIIIA
regulatory factors and extend further upstream from the gene’s for expression, thus the term “general transcription factor” may
start site. For example the regulatory sequence for a yeast gene is not strictly apply. The designation may be more related to the
typically within 200 basepairs of the transcription start site while history of TFIIIA since one of the first eukaryotic genes to be
human genes can have multiple regulatory factors binding transcribed in vitro required purified TFIIIA.10 Second, the
sequences 50 or even 100 kb upstream or downstream from the sequence recognized by TFIIIA is within the 5S gene itself as
gene.7 opposed to upstream of the gene where transcription initiation
An alternative explanation for the gene counting conundrum events traditionally occur. The structural details of this
is the “few proteins, multiple functions” model. Perhaps a protein-DNA interaction have been characterized12-14 and
Zinc Finger Proteins: From Atomic Contact to Cellular Function, edited by Shiro Iuchi and Natalie Kuldell.
©2005 Landes Bioscience/Eurekah.com and Kluwer Academic/Plenum Publishers.
196 Zinc Finger Proteins: From Atomic Contact to Cellular Function
suggesting defects in ribosome assembly.53 In Xenopus embryos, bound the 5S rRNA gene as well as its rRNA product in vitro.
overexpression of L5 results in increased expression of 5S rRNA, Furthermore, a TFIIIA-GFP fusion protein localized to the nucleo-
presumably because L5 displaces TFIIIA for the existing pool of lus, suggesting the presence of a nuclear localization signal though
5S rRNA, allowing TFIIIA to re-enter the nucleus and initiate none could be identified by sequence gazing.
new transcription of the 5S rRNA gene.54 These studies provide Given the variety of TFIIIA arrangements it is interesting that
compelling evidence for a TFIIIA/L5/5S rRNA network. the sequence of the 5S rRNA gene, and consequently the rRNA
Further study is needed to understand the mechanisms that itself, is highly conserved. The gene is relatively short, only 120
regulate this network and coordinate it with the synthesis and nucleotides in most organisms, and the minor sequence
assembly of the other ribosome components. variations that occur are primarily in the ICR.
Recent experiments show TFIIIA can be phosphorylated by
casein kinase II,57 a tantalizing observation given that this modi- Summary
fication is so often associated with regulation of protein activity. Studies of TFIIIA have revealed multiple cellular functions
The site of phosphorylation appears to be the first zinc finger. for this protein. Its ability to bind both DNA and RNA has made
This serine/threonine kinase is associated with RNA-binding this protein a wonderful tool for exploration of distinct cellular
proteins in Xenopus oocytes58 and can phosphorylate TFIIIA in activities, providing insight into important regulatory mechanisms.
vitro.59 Indeed, early purifications of TFIIIA were “contaminated” The zinc fingers of TFIIIA play critical roles in the protein’s
with kinase activity.60 Interestingly, human L5 protein was functions. Recent evidence indicates the zinc fingers are essential
isolated in a yeast two-hybrid screen for proteins that interact for yet another cellular function, namely recognition of damaged
with the regulatory subunit of casein kinase II.61 The phosphory- DNA and targeting of DNA repair.66 Studies of TFIIIA continue
lation of L5 decreases its affinity for 5S rRNA in vitro,62 an to generate much interest and excitement.
observation consistent with the idea that phosphorylation
regulates the nuclear/cytoplasmic shuttle of 5S rRNA. References
The physiological significance of TFIIIA phosphorylation is 1. Snyder M, Gerstein M. Defining Genes in the Genomics Era. Science
still unclear. Most TFIIIA that is bound to 5S rRNA is not 2003; 300:258-260.
phosphorylated, but it is certainly possible that the modification 2. Pennisi E. Gene Counters Struggle to Get the Answer Right. Science
is transient or occurs on a small but active fraction of the 7S 2003; 301:1040-1041.
RNP. Alternatively, phosphorylation of TFIIIA could influence 3. Lander ES, Linton LM, Birren B et al. Initial sequencing and analysis
of the human genome. Nature 2001; 409:860-921.
its DNA binding activity or transcription activation function. 4. Venter JC, Adams MD, Myers EW et al. The sequence of the human
Another potentially relevant observation is the ATP-ase activity genome. Science 2001; 291:1304-1351.
detected for TFIIIA,63 an activity that is dependent on DNA. 5. Adams MD, Celniker SE, Holt RA et al. The genome sequence of
How these observations connect and fit into existing models for Drosophila melanogaster. Science 2000; 287:2185-2195.
TFIIIA function and regulation awaits further investigation. 6. Ruvkun G, Hobert O. The taxonomy of developmental control in
Caenorhabditis elegans. Science 1998; 282:2033-2041
7. Levine M, Tjian R. Transcription regulation and animal diversity.
TFIIIA in Yeast, Plants and Complex Nature 2003; 424:147-151.
Eukaryotes 8. Pieler T, Theunissen O. TFIIIA: nine fingers-three hands? Trends
Biochem Sci 1993; 18:226-230.
The sequences of TFIIIA from a wide range of organisms has 9. Cassiday LA, Maher LJ 3rd. Having it both ways: transcription
been determined and are widely divergent (reviewed in reference factors that bind DNA and RNA. Nucleic Acids Res 2002;
64). The presence of nine zinc fingers is perhaps the only 30:4118-4126.
common feature of the TFIIIA sequences described so far, and 10. Engelke DR, Ng S-Y, Shastry BS et al. Specific Interaction of a
even this commonality is violated by TFIIIA from the yeast S. Purified Transcription Factor with an Internal Control Region of 5S
RNA Genes. Cell 1980; 19:717-728.
pombe which bears a tenth zinc finger. The yeast S. cerevisiae also 11. Pelham HRB, Brown DD. A specific transcription factor that can
has some noteworthy differences in the arrangement of its TFIIIA bind either the 5S rRNA gene or 5S RNA. Proc Natl Acad Sci
zinc fingers. The transcription activation domain and protein’s USA 1980; 77:4170-4174.
localization signals are found between the eighth and ninth zinc 12. Wuttke DS, Foster MP, Case DA et al. Solution structure of the
fingers in this protein (compare to Fig. 1). TFIIIA from first three zinc fingers of TFIIIA bound to the cognate DNA
vertebrate organisms is similarly varied. Among frog species, X. sequence: determinants of affinity and sequence specificity. J Mol
Biol 1997; 273:183-206.
laevis is only 84% and 63% identical to X. borealis and R. 13. Foster MP, Wuttke DS, Radhakrishnan I et al. Domain packing and
catesbiana, respectively.20 dynamics in the DNA complex of the N-terminal zinc fingers of
The recent cloning of TFIIIA from the plant A. thaliana TFIIIA. Nat Struct Biol 1997; 4:605-608.
further emphasizes the flexibility of this protein.65 The authors 14. Nolte RT, Conlin RM, Harrison SC et al. Differing roles for zinc
identified the A. thaliana TFIIIA homolog despite its limited fingers in DNA recognition: structure of a six-finger transcription
sequence identity with the S. cerevisiae and X. laevis proteins, factor IIIA complex. Proc Natl Acad Sci USA 1998; 95:2938-2943.
15. Brown RS. TFIIIA: A Sophisticated Zinc Finger Protein. In Iuchi, S
reported as 17 and 26% respectively. The most significant and Kuldell N, eds. Zinc Finger Proteins: from atomic contacts to
alignment of these proteins was found in the sequences of the … Landes Bioscience, 2004: pgs XXX.
zinc fingers themselves. Outside of these motifs, however, there 16. Theunissen O, Rudt F, Guddat U et al. RNA and DNA binding
was little to indicate that the cloned cDNA was, in fact, TFIIIA. zinc fingers in Xenopus TFIIIA. Cell 1992; 71:679-690
The clone has two unusual spacer sequences separating zinc 17. Mao X and Darby MK. A position-dependent transcription-activating
fingers 1 and 2, and zinc fingers 4 and 5. In addition, there are no domain in TFIIIA. Mol Cell Biol 1993; 13: 7496-7506.
18. Moreland RJ, Dresser ME, Rodgers JS et al. Identification of a
sequences reminiscent of the nuclear import or export signals or transcription factor IIIA-interacting protein. Nucleic Acids Res 2000;
the transcription activation domains found in TFIIIA from other 28:1986-1993.
organisms. Nevertheless, the identified cDNA was used to make 19. Schulman DB, Setzer DR. Functional analysis of the novel
a recombinant protein and that protein sequence-specifically C-terminal domains of S pombe transcription factor IIIA. J Mol
Biol 2003; 331:321-330.
The Multiple Cellular Functions of TFIIIA 199
20. Paule MR, White RJ. Transcription by RNA polymerases I and III. 43. Fridell RA, Fischer U, Luhrmann R et al. Amphibian transcription
Nucleic Acids Res 2000; 28:1283-1298. factor IIIA proteins contain a sequence element functionally
21. Del Rio S, Menezes SR, Setzer DR. The function of individual zinc equivalent to the nuclear export signal of human immunodeficiency
fingers in sequence-specific DNA recognition by transcription factor virus type 1 Rev. Proc Natl Acad Sci USA 1996; 93:2936-2940.
IIIA. J Mol Biol 1993; 233:567-579. 44. You QM, Romaniuk PJ. The effects of disrupting 5S RNA helical
22. Veldhoen N, You Q, Setzer DR et al. Contribution of individual structures on the binding of Xenopus transcription factor IIIA.
base pairs to the interaction of TFIIIA with the Xenopus 5S RNA Nucleic Acids Res 1990; 18:5055-5062.
gene. Biochemistry 1994; 33:7568-7575. 45. Theunissen O, Rudt F, Pieler T. Structural determinants in 5S RNA
23. Hanas JS, Bogenhagen DF, Wu CW. Cooperative model for the and TFIIIA for 7S RNP formation. Eur J Biochem.
binding of Xenopus transcription factor A to the 5S RNA gene. 1998;258:758-767.
Proc Natl Acad Sci USA 1983; 80:2142-2145. 46. Searles MA, Lu D, Klug A. The role of the central zinc fingers of
24. Braun BR, Bartholomew B, Kassavetis GA et al. Topography of transcription factor IIIA in binding to 5 S RNA. J Mol Biol 2000;
transcription factor complexes on the Saccharomyces cerevisiae 5 S 301:47-60.
RNA gene. J Mol Biol 1992; 228:1063-1077. 47. Lu D. Searles MA, Klug A. Crystal structure of a zinc-finger-RNA
25. Kassavetis GA, Braun BR, Nguyen LH et al. TFIIIB is the complex reveals two modes of molecular recognition. Nature 2003;
transcription initiation factor proper of RNA polymerase III, while 426:96-100.
TFIIIA and TFIIIC are assembly factors. Cell 1990; 60:235-245. 48. You QM, Veldhoen N, Baudin F et al. Mutations in 5S DNA and
26. Kassavetis GA, Joazeiro CA, Pisano M et al. The role of the 5S RNA have different effects on the binding of Xenopus
TATA-binding protein in the assembly and function of the transcription factor IIIA. Biochemistry 1991; 30:2495-2500.
multisubunit yeast RNA polymerase III transcription factor, TFIIIB. 49. Bumbulis MJ, Wroblewski G, McKean D et al. Genetic analysis of
Cell 1992; 71:1055-1064. Xenopus transcription factor IIIA. J Mol Biol 1998; 284: 1307-1322.
27. Bogenhagen DF, Wormington WM, Brown DD. Stable transcription 50. Darby MK and Joho KE. Differential binding of zinc fingers from
complexes of Xenopus 5S RNA genes: a means to maintain the Xenopus IIIA and p43 to 5S RNA and the 5S RNA gene. Mol Cell
differentiated state. Cell 1982; 28:413-421. Biol 1992; 12:3155-3164.
28. Engelke DR, Gottesfeld JM. Chromosomal footprinting of 51. Guddat U, Bakken AH, Pieler T. Protein-mediated nuclear export
transcriptionally active and inactive oocyte-type 5S RNA genes of of RNA: 5S rRNA containing small RNPs in xenopus oocytes. Cell
Xenopus laevis. Nucleic Acids Res 1990; 18:6031-6037. 1990; 60:619-628.
29. Bardeleben C, Kassavetis GA, Geiduschek EP. Encounters of 52. Scripture JB, Huber PW. Analysis of the binding of Xenopus
Saccharomyces cerevisiae RNA polymerase III with its transcription ribosomal protein L5 to oocyte 5 S rRNA. The major determinants
factors during RNA chain elongation. J Mol Biol 1994; of recognition are located in helix III-loop C. J Biol Chem 1995;
235:1193-1205. 270:27358-27365.
30. Weser S, Riemann J, Seifart KH et al. Assembly and isolation of 53. Deshmukh M, Stark J, Yeh L-C C et al. Multiple Regions of Yeast
intermediate steps of transcription complexes formed on the human Ribosomal Protein L1 are Important for its Interaction with 5S rRNA
5S rRNA gene. Nucleic Acids Res 2003; 31:2408-2416. and Assembly into Ribosomes. J Biol Chem 1995; 270:30148-30156.
31. Wolffe AP, Brown DD. Developmental regulation of two 5S 54. Pittman RH, Andrews MT, Setzer DR. A feedback loop coupling 5
ribosomal RNA genes. Science 1988; 241:1626-1632. S rRNA synthesis to accumulation of a ribosomal protein. J Biol
32. Hansen JC, Wolffe AP. A role for histones H2A/H2B in chromatin Chem 1999; 274:33198-33201.
folding and transcriptional repression. Proc Natl Acad Sci U S A 55. Rudt F, Pieler T. Cytoplasmic retention and nuclear import of 5S
1994; 91:2339-2343. ribosomal RNA containing RNPs. EMBO J 1996; 15:1383-1391.
33. Hansen JC, Wolffe AP. Influence of chromatin folding on 56. Andrews MT, Brown DD. Transient activation of oocyte 5S RNA
transcription initiation and elongation by RNA polymerase III. genes in Xenopus embryos by raising the level of the trans-acting
Biochemistry 1992; 31: 7977-7988. factor TFIIIA. Cell 1987; 51:445-453.
57. Westmark CJ, Ghose R, Huber PW. Phosphorylation of Xenopus
34. Hayes JJ, Wolffe AP. Histone H2A/H2B inhibit the interaction of
transcription factor IIIA by an oocyte protein kinase CK2. Biochem
TFIIIA with a nucleosome including the Xenopus borealis somatic
J 2002; 362: 375-382.
5S RNA gene. Proc Natl Acad Sci USA 1992; 89:1229-1233.
58. Kandror KV, Stepanov AS. RNA-binding protein kinase from
35. Tremethick D, Zucker K, Worcel A. The transcription complex of
amphibian oocytes is a casein kinase II. FEBS Lett 1984; 170: 33-37.
the 5 S RNA gene, but not transcription factor IIIA alone, prevents
59. Stepanov AS, Kandror KV, Elizarov SM. Protein kinase activity in
nucleosomal repression of transcription. J Biol Chem 1990;
RNA-binding proteins of Amphibia oocytes. FEBS Lett 1982; 141:
265:5014-5023.
157-160.
36. Tse C, Fletcher TM, Hansen JC. Enhanced transcription factor access
60. Bieker JJ, Roeder RG. Physical properties and DNA-binding
to arrays of histone H3/H4 tetramer:DNA complexes in vitro:
stoichiometry of a 5 S gene-specific transcription factor. J Biol Chem
implications for replication and transcription. Proc Natl Acad Sci 1984; 259:6158-6164.
USA 1998; 95:12169-12173. 61. Kim JM, Cha JY, Marshak DR et al. Interaction of the beta subunit
37. Vitolo JM, Thiriet C, Hayes JJ. The H3-H4 N-terminal tail domains of casein kinase II with the ribosomal protein L5. Biochem Biophys
are the primary mediators of transcription factor IIIA access to 5S Res Commun 1996; 226: 180-186.
DNA within a nucleosome. Mol Cell Biol 2000; 20:2167-2175. 62. Park JW, Bae YS. Phosphorylation of ribosomal protein L5 by protein
38. Yang Z, Hayes JJ. Xenopus transcription factor IIIA and the 5S kinase CKII decreases its 5S rRNA binding activity. Biochem Biophys
nucleosome: development of a useful in vitro system. Biochem Cell Res Commun 1999; 263: 475-481.
Biol 2003; 81:177-184. 63. Hazuda DJ, Wu CW. DNA-activated ATPase activity associated with
39. Romaniuk PJ. TFIIIA and p43: Binding to 5S Ribosomal RNA. In: Xenopus transcription factor A. J Biol Chem 1986;
Iuchi S, Kuldell N, eds. Zinc Finger Proteins: From Atomic Contacts 261:12202-12208.
to Cellular Function. Landes Bioscience and Kluwer Academic Press, 64. Huang Y, Maraia RJ. Comparison of the RNA polymerase III
2004. transcription machinery in Schizosaccharomyces pombe,
40. Setzer DR, Menezes SR, Del Rio S et al. Functional interactions Saccharomyces cerevisiae and human. Nucleic Acids Res 2001;
between the zinc fingers of Xenopus transcription factor IIIA during 29:2675-2690.
5S rRNA binding. RNA 1996; 2:1254-1269. 65. Mathieu O, Yukawa Y, Prieto JL et al. Identification and
41. Rollins MB, Del Rio S, Galey AL et al. Role of TFIIIA zinc fingers characterization of transcription factor IIIA and ribosomal protein
in vivo: analysis of single-finger function in developing Xenopus L5 from Arabidopsis thaliana. Nucleic Acids Res 2003;
embryos. Mol Cell Biol 1993; 13:4776-4783. 31:2424-2433.
42. Hanas JS, Hocker JR, Cheng Y-G et al. cDNA cloning, DNA 66. Kwon Y and Smerdon MJ Binding of Zinc Finger Protein TFIIIA
binding, and evolution of mammalian transcription factor IIIA. Gene to its Cognate DNA Sequence with Single UV Photoproducts at
2002; 282:43-52. Specific Sites and its Effect on DNA Repair J Biol Chem 2003;
278:45451-45459.
CHAPTER 27
Abstract
I
n many developmental systems, nuclear regulators have been
implicated in coupling key events in gene expression with
specific cell fate and lineage decisions. In the hemo-lymphoid
system, the Ikaros gene family of zinc finger DNA binding fac-
tors controls lymphocyte specification and homeostasis from the
hemopoietic stem cell (HSC) throughout development. The de-
pendence of hemo-lymphoid differentiation on Ikaros DNA
binding activity together with the presence of Ikaros proteins
within higher order chromatin remodeling complexes supports
the hypothesis that Ikaros plays a key role in the lineage-specific
remodeling of chromatin. Association of Ikaros and its remodel-
ing partners with the chromatin of key lineage-specific genes,
and the dependence of these genes on Ikaros complexes for their
expression supports this hypothesis and provides unique para-
digms to study chromatin regulation of differentiation.
The Ikaros-Gene Family Figure 1. Ikaros family proteins. Schematic representation of functional
domains in Ikaros family member proteins and various splicing isoforms.
Functional Domains of Ikaros Gene Family Exons are illustrated individually with zinc fingers shown as boxes.
The Ikaros gene family encodes a unique group of Regions required for DNA-binding and multimerization are indicated.
DNA-binding proteins characterized by the presence of two func-
tionally distinct domains containing Krüppel-type zinc fingers.1-3 esis studies indicate that the function and presence of the core
Initially identified as a regulator of lymphoid-specific genes, Ikaros pair of zinc fingers, F2 and F3, are critical for high affinity bind-
was subsequently determined to be the archetype for a family of ing to the core motif 5'-c/TGGGAAT/c-3'.2,11 The presence of
proteins including Aiolos, Helios, and Eos, expression of which at least one additional flanking zinc finger, F1 or F4, further in-
extends from hemo-lymphoid to neuro-epithelial tissues (Fig. fluences binding to this motif.2 The amino-terminal zinc finger
1).4-7 cDNAs for the Ikaros family members have been isolated domain is also required for nuclear localization.2,11 All recognized
from human, mouse, chicken, amphibian, fish, and protochordate splicing isoforms retain the carboxy-terminal pair of zinc fingers.
species and comparison of the protein sequences indicates a strong This region may represent an evolutionarily conserved interac-
conservation of domain function.8,9 tion module, as it shares homology with the Drosophila hunch-
The full length Ikaros protein is characterized by an back and mammalian Pegasus and tricho-rhino-phalangeal TRPS1
amino-terminal zinc finger-dependent DNA binding domain and proteins.1,12-14 This multimerization domain allows assembly
a carboxy-terminal zinc finger domain required for multimer as- between Ikaros proteins and other family members.
sembly (Fig. 1).2,3 Ikaros is comprised of two alternatively used Multimerization between Ikaros isoforms that can and cannot
noncoding exons followed by seven coding exons.10 Alternative bind DNA compromises the ability of the resulting complex to
splicing of exons four through seven gives rise to ten recognized bind DNA. This suggests that the nonbinding isoforms may rep-
isoforms. These isoforms vary in the inclusion of the four resent a naturally occurring dominant negative mechanism to
amino-terminal zinc fingers required for DNA binding. Mutagen- regulate the activity of this family of proteins.15 Nevertheless, the
Zinc Finger Proteins: From Atomic Contact to Cellular Function, edited by Shiro Iuchi and Natalie Kuldell.
©2005 Landes Bioscience/Eurekah.com and Kluwer Academic/Plenum Publishers.
The Role of the Ikaros Gene Family in Lymphocyte Development 201
Expression
Ikaros is highly expressed in the earliest sites of hemopoiesis in
the developing embryo including the blood islands of the yolk
sac at E8, the Aorta Gonad Mesonephros (AGM) and the fetal
liver at E9.5. Ikaros mRNA is subsequently detected in the fetal
thymus from E10.5, concomitant with the onset of its popula-
tion with fetal lymphoid precursors.1 In the adult, Ikaros is ex-
pressed in hemopoietic progenitor populations enriched for the
pluripotent and self-renewing HSC, and in myelo-erythroid pre-
cursors.4,5,16 Upon differentiation of these precursors to mono-
cytes, macrophages and erythrocytes, Ikaros message level is
downregulated, but maintained at significant levels in granulo-
cytes.6 In contrast, Ikaros is upregulated in T and B cell precur- Figure 2. Modes of transcriptional regulation by Ikaros. A) Through
sors as they expand and differentiate and remains expressed in chromatin remodeling. Chromatin remodeling can result in both acti-
mature lymphocytes in the fetus and adult. Among the lymphoid vation and repression by facilitating the binding of a transcriptional
populations, Ikaros expression is highest in double positive thy- activator or a repressor. B) Through histone modifications. The Ikaros/
mocytes and mature T cells, populations that display strong NuRD and Ikaros/Sin3 complexes contain histone deacetylases
haplo-insufficiency phenotypes in mice heterozygous for Ikaros (HDACs). Thus Ikaros can affect transcription by changing the histone
mutations. acetylation status in the vicinity of its target sites. C) Through direct
In contrast to Ikaros, expression of the closely related family perturbation of basal transcription factors. Both CtBP and CtIP can
member Aiolos is restricted to the lymphoid system.4 It is first interact with basal transcription factors. It has been proposed that these
detected at low levels in the common lymphoid progenitor (CLP) interactions can directly inhibit the activity or affect the stability of the
and is progressively upregulated during B and T cell differentia- preinitiation complex, thus contributing an HDAC-independent re-
tion.4,17 Aiolos expression peaks in mature and recirculating pe- pression mechanism.
ripheral B cells, populations that are most affected by Aiolos de-
ficiency. The Helios pattern of expression in early hemopoietic The Ikaros interactors described above reveal that there are at
progenitors is similar to that of Ikaros but is detected at much least three different mechanisms that are potentially utilized by
lower levels.5 Unlike Ikaros, Helios is additionally expressed in a Ikaros to alter gene expression (Fig. 2).
variety of epithelial tissues. Finally, Eos is expressed in a variety of a. Targeting of ATP-dependent chromatin remodeling to catalyze
tissues outside the hemopoietic system in very low amounts.12 chromatin fluidity, conceptualized as a dynamic equilibrium
between closed and open chromatin states. As many transcrip-
Mechanisms of Action tion factors cannot bind to their recognition sequences in the
To gain insight into the mechanisms by which Ikaros exerts its context of a restrictive nucleosomal structure, nucleosome re-
modeling targeted by an Ikaros complex can facilitate binding
effects on gene expression and to identify associated proteins, bio-
of downstream transcription factors. This may promote or in-
chemical purification of Ikaros was carried out. Purification of
hibit gene expression, depending on the nature of the factors
Ikaros from the thymus of a Flag-tagged Ikaros-expressing
recruited to the region (panel A). Alternatively, Ikaros mediated
transgenic line revealed that it is associated with two different chromatin remodeling may prohibit binding of transcription
classes of chromatin remodeling complexes.18 A major fraction factors and limit the potential for regulation.
of Ikaros associates with components of a previously described
NuRD (nucleosome remodeling and deacetylating) complex that b. The authoring of a “histone code” by Ikaros complexes. Core
contains the ATPase Mi-2 and HDAC1/2. A smaller fraction of histones are biochemically modified at specific residues in a va-
Ikaros was found associated with Swi/Snf, another chromatin re- riety of ways, including acetylation, methylation, phosphoryla-
modeling complex containing the ATPase Brg1 and BAFs. Ikaros tion and ubiquitination. It is widely theorized that a combina-
complexes were also purified from an erythroleukemia cell line tion of specific histone modifications dictates a code establish-
with both classes of chromatin remodelers.19 This group suggested ing a downstream event. Indeed, several conserved protein mo-
that Ikaros may form a mega-complex containing both NuRD tifs are known to recognize specific histone modifications. The
and Swi/Snf components. Outside the NuRD complex, a small Ikaros/NuRD and Ikaros/Sin3 complexes contain HDAC1/2
fraction of Ikaros can also associate with another that can remove acetyl groups from histone tails and establish a
repressive histone code (panel B).
HDAC-interacting factor, Sin3.20 Ikaros can therefore recruit
HDAC activity using two independent pathways. Ikaros has also c. Ikaros may repress gene expression in an HDAC-independent
been shown to repress gene expression in an HDAC-independent mechanism, which involves CtBP and CtIP. Repression by these
manner through CtBP and CtIP.21 molecules is thought to be achieved through a direct inhibition
of basal transcription factors21 (panel C).
202 Zinc Finger Proteins: From Atomic Contact to Cellular Function
expression during development. Activation of Rag-1 and Rag-2 Of the three major areas of Ikaros phosphorylation detected
gene expression is also dependent on Ikaros. Rag gene expression during the G1-S transition, phosphorylation of the S/TRR resi-
is abolished in the HSC compartment (S. Ng, unpublished data) dues in exon 8 negatively regulates the ability of the protein to
of Ikaros null mice and Rag BAC transgenes are downregulated arrest the cell cycle. Ikaros phosphorylation mutants (Serine to
in these cell types (T. Naito, unpublished data). Alanine) at the S/TRR residues act as stronger cell cycle inhibi-
TdT is another potential Ikaros gene target during develop- tors than the wild type protein and increase Ikaros’ DNA bind-
ment. An Ikaros binding site in the TdT promoter overlaps that ing activity. Casein Kinase II (CKII) was found to be responsible
of Elf-1, a putative activator of TdT. In vitro binding of Ikaros for the Ikaros phosphorylation events during the G1-S transi-
and Elf-1 to the site is mutually exclusive, suggesting that Ikaros tion. During mitosis, Ikaros becomes further phosphorylated in
may down-regulate TdT by competing away Elf-1.34 Ikaros has the linker region of the four DNA-binding zinc fingers which
also been implicated in the repression of the lambda5 gene.35 further reduces Ikaros DNA binding.37 Thus, there appears to be
Here again, competition with E2A at a common binding site has a gradual inactivation of Ikaros DNA binding activity during the
been suggested as a repression mechanism. cell cycle that is mediated by phosphorylation. Further studies
are required to determine how these post-transcriptional modifi-
Regulation of Ikaros cations modulate Ikaros’ activity as a transcriptional regulator of
cell cycle control genes.
Post-Translational
A progressive reduction of Ikaros protein levels in T cells fa- Transcriptional Regulation of Ikaros
cilitates their activation and causes the rapid development of leu- Ikaros expression in the hemopoietic system is controlled by a
kemias and lymphomas, suggesting Ikaros functions as a potent large number of regulatory regions that are scattered throughout
tumor suppressor. Indeed, Ikaros provides for critical control over its 120kb locus on mouse Chromosome 11.10 The Ikaros gene
cell cycle progression by negatively regulating the G1-S transi- consists of two untranslated exons (1a and 1b) and seven trans-
tion through its DNA binding activity.36 lated exons (exon 2 to 8). The two untranslated exons are inde-
In search of post-translational modifications that regulate Ikaros pendently spliced to exon 2. Several hemopoietic-specific tran-
activity during the cell cycle, three regions of phosphorylation scription start sites were detected in their vicinity that are utilized
were identified in Ikaros proteins.36,37 Phosphorylation impacts at different frequencies indicating the existence of two promoter
Ikaros’ DNA binding ability at different stages of the cell cycle regions upstream of exon 1a and 1b (Fig. 6). These regions do
(Fig. 5). In the G1 phase, Ikaros is expressed in a not contain any TATA or CAAT box canonical promoter ele-
de-phosphorylated state, becomes phosphorylated during the rep- ments and are GC rich in composition, like other hemopoietic
licative phase (S), and remains so through M phase. Ikaros phos- and lymphoid-specific promoters.
phorylation is lost in the G1 phase of the next cycle and is rees- Ten clusters of hemopoietic tissue-specific DNase I hypersen-
tablished during S phase (Fig. 5B). sitivity sites (DHS) in the Ikaros locus were identified which de-
marcate regions of chromatin accessibility found in the vicinity
of active regulatory regions (Fig. 6). Some of these DHS are de-
tected in both spleen and thymus and others are more restricted
to the thymus. Studies on cross species sequence conservation
have identified eight conserved regions, most of which are found
in the proximity of the DHS clusters described for the mouse
Ikaros genomic locus (Fig. 6).
The activity of the two genomic fragments that contain the
Ikaros promoter regions has been reported in mouse transgenic
studies. The DHS-C2 (A) region containing the promoter up-
stream of exon 1a drives expression of the enhanced green fluo-
rescent protein (EGFP) reporter in granulocytes (Mac-1+/Gr-1+)
(Fig. 6, A-p-GFP). The DHS-C3 (B) region containing the pro-
moter upstream of exon 1b provides reporter activity predomi-
nantly in B cells and in some myeloid cells (Mac-1+/Gr-1-) (Fig.
6, B-p-GFP). In both cases, the reporter is subject to position
effect variegation (PEV) indicating that critical regulatory ele-
ments that counteract the restrictive effects of heterochromatin
in the Ikaros locus are missing. The DHS-C3 region is also active
in the hemopoietic stem cell compartment (HSCs). DHS-C3
activity is maintained throughout B cell development and in
mature B cells; however, during T cell development, it starts to
decline at the DN3 stage (CD44-CD25+) and minimal activity is
detected at the DP stage. Thus additional elements are required
Figure 5. Phosphorylation controls Ikaros’ ability to negatively regulate to maintain DHS-C3 activity past the DN3 stage of T cell differ-
the G1 to S phase transition. A) Schematic representation of Ik-1 show- entiation.
ing translated exons (horizontal rectangles) and its 6 Zinc Fingers (F). The DHS-C3 regulatory region, in combination with the
The asterisk and grey rectangle represent residues where Ikaros is phos- intronic DHS-C6 (C) region, provides for almost 100% reporter
phorylated during the G1-S transition. B) Model for Ikaros’ control of activity throughout T cell development in the thymus (Fig. 6,
the G1 to S transition and its regulation by Ikaros phosphorylation. B-p-GFP-C). This combination of regulatory elements in part
The Role of the Ikaros Gene Family in Lymphocyte Development 205
25. Avitahl N, Winandy S, Friedrich C et al. Ikaros sets thresholds for 32. Cortes M, Georgopoulos K. Aiolos is required for the generation of
T cell activation and regulates chromosome propagation. Immunity high affinity bone marrow plasma cells responsible for long-term
1999; 10:333-343. immunity. J Exp Med 2004; 199(2):209-219.
26. Harker N, Naito T, Cortes M et al. The CD8alpha Gene Locus Is 33. Chi TH, Wan M, Zhao K et al. Reciprocal regulation of CD4/
Regulated by the Ikaros Family of Proteins. Mol Cell 2002; CD8 expression by SWI/SNF-like BAF complexes. Nature 2002;
10(6):1403-1415. 418(6894):195-199.
27. Winandy S, Wu P, Georgopoulos K. A dominant mutation in the 34. Trinh LA, Ferrini R, Cobb BS et al. Down-regulation of TDT
Ikaros gene leads to rapid development of leukemia and lymphoma. transcription in CD4(+)CD8(+) thymocytes by Ikaros proteins in
Cell 1995; 83:289-299. direct competition with an Ets activator. Genes Dev 2001;
28. Brown KE, Guest SS, Smale ST et al. Association of transcriptionally 15(14):1817-1832.
silent genes with Ikaros complexes at centromeric heterochromatin. 35. Sabbattini P, Lundgren M, Georgiou A et al. Binding of Ikaros to
Cell 1997; 91:845-854. the lambda5 promoter silences transcription through a mechanism
29. Wang J-H, Avitahl N, Cariappa A et al. Aiolos regulates B cell that does not require heterochromatin formation. EMBO J 2001;
activation and maturation to effector state. Immunity 1998; 20(11):2812-2822.
9:543-553. 36. Gomez-del Arco P, Maki K, Georgopoulos K. Phosphorylation
30. Cariappa A, Tang M, Parng C et al. The follicular versus marginal controls Ikaros’ ability to negatively regulate the G1-S transition.
zone B lymphocyte cell fate decision is regulated by Aiolos, Btk, MCB 2004; 24(7):2797-2807.
and CD21. Immunity 2001; 14(5):603-615. 37. Dovat S, Ronni T, Russell D et al. A common mechanism for mitotic
31. Sun J, Matthias G, Mihatsch MJ et al. Lack of the transcriptional inactivation of C2H2 zinc finger DNA-binding domains. Genes Dev
coactivator OBF-1 prevents the development of systemic lupus 2002; 16(23):2985-2990.
erythematosus-like phenotypes in Aiolos mutant mice. J Immunol
2003; 170:1699-1706.
CHAPTER 28
Abstract
keratinocytes with a cDNA encoding the HLH domain of Myf-5.2
B
asonuclin is a C2H2 zinc finger protein first discovered in
the keratinocytes of stratified squamous epithelia and in Keratinocyte cDNAs selected by the screening contained a se-
certain keratinocytes of the hair follicles. It was later de- quence of only 26 nucleotides similar to those of the Myf-5 cod-
tected in the reproductive germ cells of the testis and ovary and ing region; these nucleotides encoded the so-called Ω loop3,4 con-
in the cells of the ocular lens epithelium. The only known func- served in the myogenic proteins as NPNQRL.5 The selected
tion of basonuclin is that of a transcription factor in the synthe- keratinocyte cDNAs did not encode an HLH protein but rather
sis of ribosomal RNA, but the interesting question remains why a zinc finger protein containing the loop sequence NPNPRL be-
this function of basonuclin should be required by a small but tween the two histidines of the second zinc finger. The cDNA
diverse group of epithelial cell types. The recent discovery of a was sequenced and found to contain 4611 nucleotides encoding
second basonuclin with strong resemblances in structure and an unknown protein with six typical C2H2 zinc fingers, arranged
function to the first, but with some important differences in cel- in three separate pairs.2 Aside from the short loop sequence, there
lular distribution and in evolutionary behavior, adds to the in- was no similarity between basonuclin and the HLH proteins.
terest of basonuclin as a regulatory protein of epithelial cell types.
Structural Features of Basonuclin
The Discovery of Basonuclin
Basonuclin was discovered during a search for keratinocyte
transcription regulators containing a helix-loop-helix (HLH) The Zinc Fingers
domain. This domain is found in the myogenic family of pro- The C-terminal two-thirds of basonuclin contain the three
teins such as MyoD and Myf-5, which cooperate to initiate pairs of zinc fingers, each pair separated from its neighbors by
myogenesis in myoblasts by inducing transcription of genes spe- 150-320 amino acids (Fig. 1). The first finger of each pair (Fig.
cifically expressed in muscle.1 On the possibility that the control 2) has a short spacer between the two cysteines (C-X2-C) and
of keratinocyte differentiation might rely on similar proteins, we between the two histidines (H-X4-H), whereas the second finger
screened a cDNA library prepared from cultured human has a longer spacer in each location (C-X4-C and H-X6-8-H). The
Figure 1. Principal features of basonuclin 1. Below the line are the numbers of the amino acid sequence beginning or ending each structural element.
Each zinc finger ends on the second histidine. * indicates C-terminus of the protein.
Zinc Finger Proteins: From Atomic Contact to Cellular Function, edited by Shiro Iuchi and Natalie Kuldell.
©2005 Landes Bioscience/Eurekah.com and Kluwer Academic/Plenum Publishers.
208 Zinc Finger Proteins: From Atomic Contact to Cellular Function
Table 1. Comparison of murine with human basonuclin (first cysteine to second histidine of each finger)
regions of different lengths. The 5' end of the longest mRNA was The Cell Types Containing Basonuclin
located approximately 350 nucleotides upstream of the 5' end of
the original cDNA clone.15 This discrepancy might be explained
by the fact that the sequence near the 5' end of the cDNA clone
Keratinocytes of Stratified Squamous Epithelia
Basonuclin was first detected in the nucleus of basal
(-250 to +187) has a GC content of over 80%,13 in contrast to
keratinocytes of plantar epidermis.17 At the tip of the rete ridges,
the average of 40-50% in the flanking sequences. RNA with high
where the epidermis is thickest, some cells in the first supra-basal
GC content usually contains stable secondary structure that may layer also contained basonuclin. Beyond this layer, basonuclin
hinder reverse transcription. Indeed, evidence from computer was usually absent from the terminally differentiating cells; a typi-
analysis showed that a large and stable (∆G = 210 kcal/mol) cal protein of terminal differentiation, involucrin, was never found
stem-loop structure resembling a cruciform is likely formed by in the same cells as basonuclin.17 Basonuclin of the basal cells
the sequence between –212 to –59 and may explain interference may be either nuclear or cytoplasmic; the latter location seemed
with reverse transcription by the avian enzyme.15 Study of a to be more common in the epidermis of elderly humans.8
coupled transcription-translation system in vitro, from DNA in- The expression pattern of basonuclin in the corneal epithe-
creasingly deleted into the stem-loop region, showed that full lium has been studied in detail in the mouse. In adults, basonuclin
deletion of the stem-loop sequence was required to produce the was present and nuclear in virtually all cells of the basal layer of
full-length basonuclin protein. It was suggested that the stem-loop the epithelium.18 The basonuclin was sometimes present in small
region might be the basis for regulation of basonuclin synthesis intranuclear aggregates or dots, suggesting association of the pro-
at the level of translation.15 tein with a sub-nuclear structure, possibly the nucleolus. This
It has been reported16 that in both human and mouse there is was also true of esophageal and vaginal keratinocytes.
an alternative exon (BSN1b) located downstream from the During corneal development in the mouse, basonuclin first
basonuclin exon I reported earlier.13 Although the human sequence appeared in a few cells of the epithelium at post-natal day 4 and
BSN1b is present in a bacterial artificial chromosome derived from was present in all basal cells by day 20. Curiously, basonuclin was
human chromosome 15, we do not know of any closely related consistently absent from the limbal region, where the stem cells
murine sequence present in GenBank. of the corneal epithelium reside.
The basonuclin promoter was found to be TATA-less.13,15 Tran-
scription from such a promoter usually requires Sp1 as an activa- Hair Follicle Cells
tor protein. Many Sp1 binding sites were found within the GC In the anagen or growth phase of the hair cycle in adult hu-
rich region, but all were located downstream of the start sites of mans and mice, basonuclin-containing cells were found within
the longer mRNAs.13,15 the outer root sheath throughout the length of the follicle includ-
Despite the heterogeneity at their 5' ends, all human basonuclin ing the bulge region, in the adjacent sebaceous gland, and in those
mRNAs shared the same two neighboring AUG codons at +89 cells of the bulb in contact with the mesenchymal cells of the
and +179 of which the first is the initiator codon.15 As basonuclin follicular papilla. In these regions, the basonuclin appeared strictly
mRNA of the mouse lacked the AUG orthologous to the initia- nuclear, whereas in the basal layer of the epidermis and adjacent
tor AUG of human basonuclin, the mouse protein had a slightly outer region of the follicle, basonuclin was located in the cyto-
greater mobility in SDS-PAGE electrophoresis than the human plasm.8,19 The protein also was cytoplasmic in regions associated
protein.11 It is likely that the nucleotide sequence orthologous to with terminal differentiation such as the supramatrical region of
that located between the first two human AUGs is not translated. the bulb.17,19
210 Zinc Finger Proteins: From Atomic Contact to Cellular Function
In the mouse, strong immunofluorescent staining of basonuclin An unusual aspect of basonuclin in the anterior lens epithe-
was first detected in the nuclei of precursor keratinocytes of the lium is that the protein was always located only in the cytoplasm,
hair follicles on embryonic day 14.19 This shows that the synthe- concentrated in its apical part.18 Even when the cells were prolif-
sis of basonuclin is an early part of the program of follicular gen- erating, the protein was never detected in the nuclei.
esis. As the newly formed hair peg grew into the dermis,
basonuclin-containing cells were concentrated in the nuclei of Location of Intranuclear Basonuclin
the leading tip of the peg, the outer root sheath and the bulb Knowledge of the structure of the gene for ribosomal RNA
close to the follicular papilla. In catagen, basonuclin remained and the role of the transcription factors UBF and SL owes much
nuclear in the retreating epithelial strand and adjacent outer root to the work of R. Tjian22-25 and R. H. Reeder.23,26 This form of
sheath. In telogen, the resting phase of the cycle, basonuclin per- regulation of ribosomal RNA synthesis is thought to be present
sisted only in the peg and adjacent outer root sheath of the club. in all cell types.
This is the region from which the new follicle regenerates during Basonuclin is associated with human mitotic chromosomes at
the next anagen. Basonuclin therefore seems to remain in cells eight highly specific sites corresponding to the short arms of the
with proliferative potential for a long period during which they acrocentric chromosomes.14 These sites are known to contain the
are not yet proliferating.17,19 genes for rRNA. This was confirmed by demonstration of
colocalization on those chromosomes of basonuclin and UBF1.14
Germ Line Cells of Testis and Ovary Cultured keratinocytes, like other cell types, contain UBF1
Surveys of mouse and human RNA of different tissues exam- (S. Iuchi and H. Green, unpublished experiments). Unlike UBF1,
ined by Northern analysis and RT-PCR revealed that basonuclin which is entirely nucleolus-associated, basonuclin is present in
mRNA was present not only in epidermis and related stratified the nucleoplasm as well. In tissue sections through the mouse
squamous epithelia but also in the testis,16 where it was even more ovary, basonuclin of the oocyte, like that of the keratinocyte, was
abundant.20 When the sensitivity of detection was increased by not confined to the nucleolus and instead appeared uniformly
hybridizing the RT-PCR products with labeled basonuclin cDNA, distributed in the nucleoplasm and nucleolus.21 But when grow-
bands of basonuclin mRNA were detected in mammary gland ing oocytes were isolated and permeabilized with Triton X100
and thymus at perhaps one order of magnitude lower abundance.20 under conditions that favored Pol I but not Pol II transcription,
This may be explained by the presence in these tissues of some nucleolar basonuclin became predominant. Under these condi-
keratinocytes or related cell types. Placenta and spleen gave sig- tions, active Pol I transcription could be simultaneously detected.21
nals of intensity possibly two orders of magnitude lower than
those of epidermis and testis, but the responsible cell types re- Basonuclin and The Synthesis of Ribosomal
main unknown.
As basonuclin of mouse testis was absent from the Sertoli cells,
RNA
the protein was restricted to the germ cell lineage which forms The presence of zinc fingers in basonuclin suggested that it
the seminiferous tubules. At first, basonuclin mRNA was found was a transcriptional regulator with ability to interact with a spe-
in the spermatogonia, spermatocytes and spermatids, but for some cific DNA sequence. This led to a search for the target gene. Two
reason the protein was not detected.16 Later, use of a different groups, independently and by different routes, reached the con-
antiserum revealed the presence of basonuclin protein at all stages clusion that the zinc fingers of basonuclin interact with sites within
of spermatogenesis, including the earliest precursors, the sper- the promoter region of human rDNA, where, like UBF1, the
matogonia, where it was located in the nuclei.20 By late stages of protein acts as a transcription factor for the synthesis of rRNA.14,27
differentiation in the spermatozoa, the basonuclin had moved to In order to isolate human genomic sequences that interact with
an extra-nuclear location and was found associated with the basonuclin, a human genomic library was incubated with bacterial
midpiece, the centrioles and the acrosomes.16,20 inclusions containing human basonuclin and selected for genomic
Basonuclin was shown to be present not only in the male germ sequences with high affinity for the protein.27 After five rounds of
line but also in developing oocytes, the only ovarian cell type that selection, a 2 kb DNA fragment became predominant. By electro-
stained intensively for basonuclin.20 In tissue sections made phoresis mobility shift assay, a 258-bp sequence located within this
through growing primary oocytes and secondary oocytes that had fragment was found to interact with the first pair of human
undergone the first meiotic division, basonuclin was found dis- basonuclin zinc fingers but not with the second. This DNA frag-
tributed evenly through the oocyte nucleoplasm and nucleolus.21 ment contained the rDNA promoter region and all the cis-elements:
the up-stream control element or UCE, the CORE, and the tran-
scription start site. Within the rDNA promoter, basonuclin inclu-
Epithelium of the Ocular Lens sion bodies protected two regions, BBS-1 and BBS-2, from DNase
Basonuclin was shown to be present in the lens epithelium at I digestion. BBS-1 was largely located within a palindrome con-
birth, considerably earlier than in the corneal epithelium. The taining short repeats and both BBS-1 and 2 were also part of a
protein was confined to the preequatorial epithelium and was larger dyad sequence (5'-CGCGTCGCCX18GGCGACGCG).
absent from equatorial cells that contained p57KIP2, an early Independent studies by electrophoretic mobility shift assay also
differentiation marker for these cells.18 Although the first ap- revealed that the DNA-binding sites of basonuclin were located
pearance of basonuclin in the lens epithelium coincided with a within the promoter sequence of rDNA.14 A highly purified prepa-
period of proliferative activity, the protein remained in the cells ration containing the recombinant N-terminal pair of basonuclin
during a long subsequent period of mitotic inactivity. In adults, zinc fingers (designated B10B) protected against DNase I diges-
only 1% of the cells in the preequatorial region of the lens epithe- tion and generated footprints on each strand of the DNA, corre-
lium could be labeled by BrdU, but basonuclin remained in all sponding to two binding sites, A and B.14 The location of bind-
the cells. This observation suggests that the presence of basonuclin ing site A agrees well with that of BBS-1 (nucleotides -95 to -135
in the lens epithelial cells is a reflection of their potential for fu- vs. -88 to 138) and that of B with BBS-2 (-64 to -79 vs -61 to
ture proliferation, for these cells are known to be capable of greatly -74). A third binding site C was located within the transcribed
increasing their rate of multiplication in response to injury. sequence between nucleotides +16 and +30.
Basonuclin: A Zinc Finger Protein of Epithelial Cells and Reproductive Germ Cells 211
Figure 3. Affinity of basonuclin for the promoter region of the ribosomal RNA gene. Shown above is a summary of DNase I footprint experiments
on the promoter of ribosomal DNA by intact basonuclin27 and by zinc fingers 1 and 2.14 By both methods, two regions of protection (boxes) were
found overlapping and adjacent to the region protected by UBF1.25 Both protected regions contain sequence with dyad symmetry.27 In these
regions are numerous copies of a hexanucleotide conforming to the consensus sequence 5’ (G/C) G (C/T) G (A/T) C 3’ (heavy black arrows). There
are five such sequences in the RNA-like strand and three in the antisense strand.
The binding site A/BBS1 (Fig. 3) is of particular interest be- 3. There are six exons in the gene for Bn2, compared with five for
cause it has the highest affinity for the N-terminal zinc finger Bn1, but in both cases most of the protein is encoded by two
pair and overlaps with the binding site of UBF1. Both A/BBS1 exons: 4 and 5 of Bn1, 5 and 6 of Bn2. The splice point be-
and B/BBS2 contain numerous repeats of a short sequence (Fig. tween these exons is identical, interrupting zinc finger 4 at the
3) within the large dyad.27 A consensus sequence for these short same point. It is therefore clear that the genes encoding Bn1
repeats was proposed: 5'-(G/C)G(C/T)G(A/T)C, in which the and Bn2 have a common evolutionary origin. Nevertheless, the
G at the second and the fourth positions and the C at the sixth gene for Bn2 is located on human chromosome 9, not chromo-
position are invariable.14 some 15, which contains the gene for Bn1.
The N-terminal pairs of zinc fingers of human and mouse
basonuclin are nearly identical, but the rDNA promoter sequences 4. The evolutionary origins of Bn1 and Bn2 are ancient, since both
of human and mouse are quite different (only 30% identical). genes are present not only in mammals, but also in birds and
Nevertheless, there are similar binding sites for this pair of zinc fishes. The amino acid sequences of Bn1 of human and mouse
fingers on the rDNA promoter of the mouse.21 This region con- are 88% identical, but the sequences of Bn2 of those species are
tains four copies of a hexanucleotide corresponding to a more 97.5% identical. Moreover, the Bn2 of the mouse, the chicken
relaxed consensus 5' (A/G/C) G (C/T) G (G/A/T) C 3'. Further- and the zebrafish are over 99% identical. This indicates highly
more, the first pair of zinc fingers of basonuclin can inhibit Pol I unusual evolutionary stability of the Bn2 sequence, comparable
transcription by a dominant-negative effect on the mouse gene, to that of p63, a transcription factor essential for the keratinocyte
as detected by a run-on assay.21 This demonstrates that there is lineage.30
functional interaction of basonuclin and the rDNA promoter in 5. Conservation of zinc fingers, the NLS and the serine stripe by
the mouse as well as the human. Bn1 and Bn2 and divergence of the rest of their amino acid
sequence suggests that the recognition system is similar or iden-
The Discovery of Basonuclin 2 tical, but that resulting functions may be different. Since the
It has recently been discovered independently in the labora- sequence of Bn2 is conserved over a broad evolutionary range,
tory of S. Sinha, by Romano et al28 and by Vanhoutteghem and it is possible that its gene is the older of the two and after its
Djian29 that there exists a gene encoding a second basonuclin duplication to produce Bn1, the latter was free to evolve in other
(Bn2). The relation between Bn2 and the original basonuclin (now directions, while Bn2 remained practically invariant.
Bn1) is of great interest. The following is a brief summary of
their findings, some of which were discovered by both laborato- 6. Like Bn1, Bn2 binds to the promoter region of the gene for
ries, others by one laboratory and not yet confirmed by the other. rRNA.
There were no serious inconsistencies between the results of the 7. Bn2 mRNA is expressed in the same cell types that possess abun-
two laboratories. The reader is referred to the original articles for dant Bn1 mRNA: keratinocytes and the germ cells of ovary and
more precise attribution. testis. But there is evidence supporting wider expression of Bn2
1. The amino acid sequences of Bn1 and Bn2 of the mouse are mRNA in epithelial cell types: mouse kidney and lung, human
only 44% identical; but the gene for Bn2 encodes all the princi- embryonic kidney cells and HeLa cells, small intestine, uterus
pal features of Bn1 described above. The sequence identity is and even (to a small extent) in brain.
higher in the zinc fingers, the region of the NLS and the region 8. In most of these tissues or cell types, Bn2 mRNA is more abun-
of the serine stripe than in the rest of the molecule. dant than that of Bn1. But this relation may depend on the
2. The gene encoding Bn2 is considerably larger than that of Bn1 circumstances. In mouse epidermis, whose Bn1 is cytoplasmic,
(300 - 400 kb compared with 29 kb for Bn1). there is rather more mRNA encoding Bn2 than Bn1. But in
212 Zinc Finger Proteins: From Atomic Contact to Cellular Function
cultured human epidermal cells, in which Bn1 is strongly nuclear, 8. Iuchi S, Easley K, Matsuzaki K et al. Alternative subcellular locations
the amount of Bn1 mRNA greatly exceeds the amount of Bn2. of keratinocyte basonuclin. Exp Dermatol 2000; 9(3):178-184.
9. Heilig JS, Freeman M, Laverty T et al. Isolation and characterization
9. Previous work on immunocytological detection of Bn1 was car- of the disconnected gene of Drosophila melanogaster. Embo J 1991;
ried out with polyclonal antisera prepared against antigen con- 10(4):809-815.
10. Mahaffey JW, Griswold CM, Cao QM. The Drosophila genes
taining most or all of the Bn1 protein. In view of the similarities disconnected and discorelated are redundant with respect to larval
between Bn1 and Bn2, we believe that these antisera are not likely head development and accumulation of mRNAs from deformed
to distinguish between the two basonuclins. In order to detect target genes. Genetics 2001; 157(1):225-236.
the protein of each independently, it will be necessary to obtain 11. Matsuzaki K, Iuchi S, Green H. Conservation of human and mouse
antibodies to specific determinants not shared by Bn1 and Bn2. basonuclins as a guide to important features of the protein. Gene
1997; 195(1):87-92.
12. Makalowski W, Zhang J, Boguski MS. Comparative analysis of 1196
orthologous mouse and human full-length mRNA and protein
Possible Functions of Basonuclin sequences. Genome Res 1996; 6(9):846-857.
Although the evidence is by no means conclusive, it seems to 13. Teumer J, Tseng H, Green H. The human basonuclin gene. Gene
us that the function of basonuclin is concerned with the poten- 1997; 188(1):1-7.
tial for cell proliferation.31 Since all studies on which this hy- 14. Tseng H, Biegel JA, Brown RS. Basonuclin is associated with the
pothesis is based were performed prior to the discovery of ribosomal RNA genes on human keratinocyte mitotic chromosomes.
J Cell Sci 1999; 112 Pt 18:3039-3047.
basonuclin 2, it will be necessary to understand to what extent 15. Tang W, Tseng H. A GC-rich sequence within the 5' untranslated
the functions of the two proteins overlap and to what extent they region of human basonuclin mRNA inhibits its translation. Gene
may differ. 1999; 237(1):35-44.
While some cell types containing basonuclin remain to be iden- 16. Yang Z, Gallicano GI, Yu QC et al. An unexpected localization of
tified, the cell types that have already been identified are all epi- basonuclin in the centrosome, mitochondria, and acrosome of
developing spermatids. J Cell Biol 1997; 137(3):657-669.
thelial cells or reproductive germ cells. Basonuclin has not been 17. Tseng H, Green H. Association of basonuclin with ability of
found in any cells of mesenchymal origin. keratinocytes to multiply and with absence of terminal differentiation.
The one identified function of basonuclin is that of a tran- J Cell Biol 1994; 126(2):495-506.
scription factor in the synthesis of ribosomal RNA. It is not clear 18. Tseng H, Matsuzaki K, Lavker RM. Basonuclin in murine corneal
why epithelial cell types and reproductive germ cells should re- and lens epithelia correlates with cellular maturation and proliferative
quire a specific transcription factor for this purpose, while other ability. Differentiation 1999; 65(4):221-227.
19. Weiner L, Green H. Basonuclin as a cell marker in the formation
cell types appear to use only the widely distributed UBF1-SL1 and cycling of the murine hair follicle. Differentiation 1998;
system. It seems possible that basonuclin also has a nucleoplas- 63(5):263-272.
mic function in the regulation of transcription of genes transcribed 20. Mahoney MG, Tang W, Xiang MM et al. Translocation of the zinc
by Pol II. It is unlikely that cytoplasmic basonuclin has any func- finger protein basonuclin from the mouse germ cell nucleus to the
tion except to constitute a pool that can be quickly mobilized for midpiece of the spermatozoon during spermiogenesis. Biol Reprod
1998; 59(2):388-394.
transport into the nucleoplasm, but in this case it would be unex- 21. Tian Q, Kopf GS, Brown RS et al. Function of basonuclin in
plained why basonuclin in cells of the ocular lens epithelium has increasing transcription of the ribosomal RNA genes during mouse
so far been found only in the cytoplasm.18 Perhaps basonuclin oogenesis. Development 2001; 128(3):407-416.
might have RNA- or protein-binding properties, such as have 22. Learned RM, Cordes S, Tjian R. Purification and characterization
been found for some other zinc finger proteins.6 of a transcription factor that confers promoter specificity to human
Finally, although basonuclin is a very early marker in the re- RNA polymerase I. Mol Cell Biol 1985; 5(6):1358-1369.
23. Bell SP, Pikaard CS, Reeder RH et al. Molecular mechanisms
productive germ line, it is not a very early marker in the develop- governing species-specific transcription of ribosomal RNA. Cell 1989;
ment of the keratinocyte lineage from embryonic stem cells, for 59(3):489-497.
the transcription factor ∆Np63 appears in this lineage consider- 24. Haltiner MM, Smale ST, Tjian R. Two distinct promoter elements
ably earlier.32 in the human rRNA gene identified by linker scanning mutagenesis.
Mol Cell Biol 1986; 6(1):227-235.
25. Learned RM, Learned TK, Haltiner MM et al. Human rRNA
References transcription is modulated by the coordinate binding of two factors
1. Lassar AB, Weintraub H. The Myogenic Helix-Loop-Helix Family: to an upstream control element. Cell 1986; 45(6):847-857.
Regulators of Skeletal Muscle Determination and Differentiation. In: 26. Reeder RH. Regulation of transcription by RNA polymerase I.
McKnight S, Yamamoto K, eds. Transcriptional Regulation. Cold Transcription Regulation. Cold Spring Harbor: Cold Spring Harbor
Spring Harbor Laboratory Press, 1992:2:1037-1061. Press, 1992.
2. Tseng H, Green H. Basonuclin: A keratinocyte protein with multiple 27. Iuchi S, Green H. Basonuclin, a zinc finger protein of keratinocytes
paired zinc fingers. Proc Natl Acad Sci USA 1992; and reproductive germ cells, binds to the rRNA gene promoter. Proc
89(21):10311-10315. Natl Acad Sci USA 1999; 96(17):9628-9632.
3. Leszczynski JF, Rose GD. Loops in globular proteins: A novel 28. Romano R, Li H, Ramakumar T et al. Molecular cloning and
category of secondary structure. Science 1986; 234(4778):849-855. characterization of Basonuclin2, a DNA-binding Zinc Finger protein
4. Murre C, McCaw PS, Baltimore D. A new DNA binding and expressed in germ tissues and skin keratinocytes. Genomics 2004;
dimerization motif in immunoglobulin enhancer binding, 83(5):821-833.
daughterless, MyoD, and myc proteins. Cell 1989; 56(5):777-783. 29. Vanhoutteghem A, Djian P. Basonuclin 2: An extremely conserved
5. Murre C, Baltimore D. The Helix-Loop-Helix Motif: Structure and homologue of the zinc finger protein basonuclin. Proc Natl Acad
Function. In: McKnight S, Yamamoto K, eds. Transcriptional Sci USA 2004; 101(10):3468-3473.
Regulation. Cold Spring Harbor Laboratory Press, 1992:2:861-879. 30. Yang A, Kaghad M, Wang Y et al. p63, a p53 homolog at 3q27-29,
6. Iuchi S. Three classes of C2H2 zinc finger proteins. Cell Mol Life encodes multiple products with transactivating, death-inducing, and
Sci 2001; 58(4):625-635. dominant-negative activities. Mol Cell 1998; 2(3):305-316.
7. Iuchi S, Green H. Nuclear localization of basonuclin in human 31. Tseng H. Basonuclin, a zinc finger protein associated with epithelial
keratinocytes and the role of phosphorylation. Proc Natl Acad Sci expansion and proliferation. Front Biosci 1998; 3:D985-988.
USA 1997; 94(15):7948-7953. 32. Green H, Easley K, Iuchi S. Marker succession during the
development of keratinocytes from cultured human embryonic stem
cells. Proc Natl Acad Sci USA 2003; 100(26):15625-15630.
CHAPTER 29
Abstract
40.4 Mouse ZAS1 has been named αA-crystallin binding pro-
T
he ZAS gene family includes ZAS1, ZAS2, and ZAS3 in
mammals as well as Schnurri in Drosophila melanogaster. tein, αACRYBP1.5,6 Rat ZAS1 has been named α1-antitrypsin
The ZAS genes encode large proteins with two separate promoter binding protein, AT-BP2.7
C2H2 zinc finger pairs that independently bind to specific The ZAS nomenclature recognizes the unique protein domain
DNA sequences, including the κB motif. In vitro experiments shared by this family. A ZAS domain is a modular protein struc-
have shown that ZAS proteins bind to target sequences and ture consisting of a pair of C2H2 zinc fingers with an acidic-rich
regulate transcription of a variety of genes involved in cell region and a serine/threonine-rich sequence.8 Currently, the hu-
growth, development, and metastasis. ZAS proteins also func- man and mouse ZAS genes are named HIVEP-1, -2, and -3 in the
tion in signal transduction: Schnurri interacts with Smad ho- Unigene database.9 We feel that the ZAS name more accurately
mologs in TGF-β-related pathways and ZAS3 interacts with conveys a distinctive family moniker as the HIV enhancer is only
TNF receptor-associated factors in TNF-α pathways. In vivo one of the many ZAS protein targets. Over the past 15 years, more
experiments have shown that disruption of ZAS expression than 20 cDNA sequences corresponding to nine ZAS genes have
alters normal cell growth as well as proliferation and results been cloned from humans, mice and rats (Table 1). An inverte-
in spontaneous tumor formation and impaired lymphocyte brate protein, Schnurri in Drosophila, with zinc fingers similar to
development. those of vertebrate ZAS proteins has also been identified.10-12
Introduction Structure
Three ZAS genes, ZAS1, ZAS2 and ZAS3, have each been
characterized from human, mouse and rat species. A related gene, Gene Organization
Schnurri, has been identified in Drosophila melanogaster. Typi- Human ZAS1, ZAS2, and ZAS3 (hZAS1, hZAS2, and hZAS3)
cally, a ZAS protein is enormous (ranging from 2348 to 2717 have been mapped to chromosomal loci on 6p24, 6q24, and 1p34,
amino acid residues) and contains four to five zinc fingers. Two respectively.13-15 All three ZAS genes are linked to endothelin
sets of the C2H2 zinc fingers pairs form the conserved proximal (EDN) and/or transcription factor AP2 (TFAP2) genes, suggest-
and distal DNA binding domains, ZAS-N and ZAS-C, respec- ing that a single -ZAS-EDN-TFAP2 ancestral gene cluster was
tively. Functionally, these proteins regulate transcription through expanded by tandem duplications.16
binding of specific DNA targets as well as by influencing signal The gene structures of the hZAS1, hZAS2, and hZAS3 are
transduction pathways. They have emerged as a protein family largely similar (Fig. 1). Each gene spans over 150 kb and contains
that may play significant roles in the regulation of immunity as at least 10 exons. The 5' ends of the ZAS genes have not yet been
well as normal cell growth, differentiation and proliferation. reported, though the 5'-untranslated region (UTR) for hZAS3 is
at least 888 nucleotides and the 5'-UTR for rZAS2 (rat ZAZ2) is
Nomenclature at least 766 nucleotides. Both 5'-UTRs are much larger than the
We propose naming this gene family ZAS. The ZAS genes average, 210 nucleotides for human and 186 nucleotides for ro-
have been characterized by many independent laboratories and dent.17 The most notable feature of the ZAS genes is an excep-
have been named according to target genes or motifs, which has tionally large exon, ranging from 5582 to 5978 nucleotides. For
resulted in several names for the same protein. For example, hu- comparison, the average vertebrate exon size is only 137 bp and
man ZAS1 has been called major histocompatability complex few are greater than 600 bp. The putative translation initiation
enhancer binding protein, MBP1,1 human immunodeficiency codon for hZAS1 is located at exon 2. Therefore, its largest exon,
virus type I enhancer binding protein, HIV-EP1,2 positive regu- exon 4, contains exclusively protein-coding sequences. However,
latory domain II binding factor, PRDII-BF1,3 and zinc finger the corresponding exons of ZAS2 and ZAS3 contain UTRs at the
Zinc Finger Proteins: From Atomic Contact to Cellular Function, edited by Shiro Iuchi and Natalie Kuldell.
©2005 Landes Bioscience/Eurekah.com and Kluwer Academic/Plenum Publishers.
214 Zinc Finger Proteins: From Atomic Contact to Cellular Function
Protein Species Name CDNA Origin Binding Site Binding Sequence References
A.
ZAS1 Human MBP-1 B cell MHC class I enhancer TGGGGA-TTCCCCA 1,44
Igκ enhancer AGGGGACTTTCCC
SV40 enhancer TGGGGACTTTCCA
HIV LTR-1 enhancer AAGGGACTTTCCG
HIV LTR-2 enhancer TGGGGACTTTCCA
β2-microglobulin AAGGGACTTTCCC
promoter
HIV-EP1 B cell HIV LTR-1 enhancer TAGGGACTTTCCG 2
HIV LTR-2 enhancer TGGGGACTTTCCA
PRDII-BF1 osteosarcoma IFN-β promoter GGGAAATTCC 3
ZnF-40 4
Mouse αACRYBP1 lens epithelium αA crystallin promoter GGGAAATCCC 6
Rat AT-BP2 thyroid α1-antitrypsin promoter TGGGGATTCACCA 7
ZAS2 Human MBP-2 retinal cells, preB, MHC Class I enhancer TGGGG(N14)ATTCCCCA 78,79
osteosarcoma Igκ enhancer AGGGGACTTTCCC
β2-microglobulin AAGGGACTTTCCC
promoter
HIV-EP2 T cell HIV LTR-1 enhancer TAGGGACTTTCCG 54
HIV LTR-2 enhancer TGGGGACTTTCCA
Mouse Schnurri-2 77
Rat MIBP-1 liver, brain c-myc intron I GGGTAGGCC 80
AT-BP1 thyroid, liver α1-antitrypsin promoter TGGGG(N14)ATTCACCA 7
ZAS3 Human KBP-1 B cell lymphoma MHC Class I enhancer TGGGGATTCCCCA 79
Igκ enhancer AGGGGACTTTCCC
β2-microglobulin AAGGGACTTTCCC
promoter
HIV-EP3 brain-stem, fetal 15
brain
Mouse Rc thymocyte, brain Igκ enhancer GGGACTTTCCC 31
MHC Class I enhancer
KRC adenocarcinomas, S100A4/mts-1 enhancer GGGGTTTTTCCAC 8,47
thymus, brain,
B cells
Schnurri Fly Schnurri Ultrabithorax enhancer B GGGGGGAGCCA 10,11,50
CGGGTGCACCC
CONSENSUS GGGGNNNNCC
B.
ZAS1 Mouse limb bud α1(II) Collagen TTGAGAAAAGCC 33
intron I
ZAS2 Mouse Brain Somatostatin receptor TTCCTCTTTTCC 34
type II promoter
ZAS3 Mouse thymocyte Recombination CACAGTG GGTTTTTTG 31,45,72
signal sequences
5' region. In addition, the largest exons of ZAS1 and ZAS3 have exon, exon 5b, between exon 5 and exon 6.15 The divergence of the
been shown to be involved in alternative splicing, which results upstream exons and the acquisition of a novel exon in hZAS3 re-
in truncated proteins without ZAS-N.18,19 The Schnurri gene flect the evolution dynamics of the ZAS gene family.
structure is different from the mammalian ZAS genes, suggesting
divergence of exons and introns during evolution from inverte- Protein Domains and Motifs
brates to vertebrates (Fig. 1). The ZAS proteins are enormous and contain four C2H2 zinc
The three exons following the largest exon are similar among fingers, which divided into two pairs. A solitary medial C2HC
the human ZAS genes with respect to size, sequence, and exon-intron zinc finger is present in ZAS1, ZAS3 and Schnurri, but not in
boundaries. One caveat to the faithful conservation of the 3' re- ZAS2. Schnurri has an additional C-terminal zinc fingers cluster,
gion of the ZAS genes is that hZAS3 has acquired an additional giving it a total of eight zinc fingers (Fig. 2).
ZAS Zinc Finger Proteins: The Other κB-Binding Protein Family 215
The C2H2 zinc fingers are common nucleic acid binding mo- ZAS-N domain, one contains only the ZAS-C domain, and one
tifs, but their sequence and organization is unique in the ZAS has no ZAS domain at all. Some alternative splicing events of
proteins. There are three classes of zinc fingers: multiple-adjacent, ZAS3 transcripts occur in a tissue-specific manner. For example,
triple, and separate-paired. The ZAS, basonuclin, and tramtrack whereas the 176 bp exon is present in most ZAS3 transcripts in
proteins constitute the separate-paired class.20 Sequence compari- the thymus, it is spliced out in brain transcripts.19 The latter tran-
son of zinc finger families across species found the ZAS family scripts encode ZAS3 proteins that lack the carboxyl one third of
zinc fingers to be distinct.21 The overall structure of the ZAS the protein including the ZAS-C domain. Therefore, ZAS3 may
domain, with a tandem zinc finger pair, acidic domain, and serine/ have distinct function between lymphoid and neuronal tissues.
threonine-rich region to form a composite structure, is also unique. Variations in the number of the DNA binding domains should
Sequence identities among the corresponding paralogous and render truncated proteins with functional characteristics distinct
orthologous ZAS zinc fingers are very high, including the back- from full-length proteins. Functional control via RNA process-
bone residues and the linker regions, suggesting conserved ing is supported by the finding that full-length ZAS1 activates
DNA-binding function (Fig. 2B). the HIV-LTR enhancer,32 whereas two truncated isoforms lack-
Excluding the zinc finger pairs, there are eight conserved re- ing either ZAS-N or ZAS-C domains repress transcription.18
gions scattered within the vertebrate ZAS proteins that are not Alternative splicing of mZAS3 transcripts has been observed
present in Schnurri. The largest conserved region is a stretch of in the largest exon of 5487 bp and a 176 bp exon. These exons
~100 amino acid residues between the ZAS domains (region 4 encode sequences including the ZAS-N domain and the ZAS-C
in Fig. 2A). It begins with a glutamate- and proline-rich region, domain, respectively. Additionally, a unique alternative splicing
followed by a putative nuclear localization signal, and ends with event within that enormous exon has been identified.19 Some
a serine stripe, which is also found in basonuclin.15 As afore- mZAS3 transcripts lack a sequence of 459 nucleotides within this
mentioned, basonuclin and the ZAS proteins belong to the exon, which encodes the ZAS-N domain. This intra-exon de-
separate-paired zinc finger class. Basonuclin is expressed in pro- leted sequence starts with “GC” and ends with “AG”, which
liferating cells and appears to be involved in regulation of cell matches splice donor and acceptor dinucleotides. This 459 nucle-
growth.22,23 Three other conserved regions have notable fea- otide sequence may be regarded as an unusual intron which, when
tures. Region 2 contains 27 residues, 40% of which are serine retained in mRNA, encodes a DNA binding domain. Such an
or proline. Region 3 contains 14 residues and is bipolar with alternative splicing event has not been demonstrated for ZAS1,
are three lysines and three arginines at the N-terminus, and two ZAS2, or Schnurri.
glutamic acids and two aspartic acids at the other end. Finally, Human ZAS3 produces a shorter 7 kb transcript by transcrip-
region 10 contains 27 residues, and is named the linker region. tion termination at an upstream exon. As discussed above, hZAS3
Those conserved region do not have obvious resemblance to has an additional 310 nucleotide exon, exon 5b.15 Exon 5b en-
known protein domains.7,15 However, their high degree of con- codes 69 amino acid residues, a 3' UTR of 103 nucleotides, and
servation in the ZAS proteins suggests they may have functional contains a hexanucleotide AAUAAA located 39 nucleotides from
importance. the 3' end of the exon. This AAUAAA sequence is the canonical
Throughout the ZAS proteins there are many potential sites polyadenlyation signal. Therefore, hZAS3 is unique among the
for protein modification. There are three or four tyrosine kinase ZAS genes in that its transcription may termination at either exon
sites clustered within or near the ZAS-C domain. Phosphoryla- 5b, resulting in a 7 kb truncated transcript, or at exon 9, resulting
tion of these sites has been shown to affect DNA binding affinity in the full-length 9.5 kb transcript.
of ZAS proteins.24,25 There are 7 to 12 copies of the serine/
threonine-proline-X-arginine/lysine (SPXK) motif in each ZAS
protein.8 The SPXK motif, frequently found in gene regulatory Tissue Distribution of ZAS Transcripts
proteins,26 can bind AT-rich DNA sequences,27 and may be phos- In adult mouse tissues, ZAS1 and ZAS2 are ubiquitously ex-
phorylated.28 pressed,5,6 where ZAS3 is expressed only in lymphoid and neural
There are protein motifs found only in some ZAS proteins. tissues.8,19,31 ZAS genes are all expressed in developing mouse
Leucine zipper motifs, which may confer protein-protein inter- embryos, though the embryonic stage and patterns of expression
actions, are found in mZAS2 at residues 751-772 and in hZAS3 differ: ZAS1 is first detected at E8.5,33 ZAS2 is detected at E10.5,34
at residues 2024-2052. Three conserved GTPase motifs are found and ZAS3 is detected at E13.5.35 ZAS1 and ZAS2 are expressed
in ZAS3, suggesting the use of GTP as an energy source or allos- relatively uniformly in developing tissues, where ZAS3 expres-
teric effector. These and other unique domains may confer spe- sion is limited to developing trigeminal ganglion, dorsal root gan-
cific functions to individual ZAS proteins. glia, cerebral cortex, and thymus.35 The distinct spatial and tem-
poral expression patterns of ZAS genes suggest that they may
perform general functions as well as specific functions during lym-
Protein Isoforms and Alternative RNA Processing phoid and neural development.
The largest open reading frame of each ZAS is >250 kD (Fig.
2A). Smaller proteins have been identified by immunoblot analysis
using ZAS129 or ZAS325,30 antibodies. These proteins are likely
Function
ZAS isoforms generated by alternative splicing and/or protein
processing. Multiple alternative splicing events have been charac- DNA Binding
terized in ZAS1 and ZAS3 transcripts.5-7,18,19,31 Significantly, al- The Rel/NF-κB and ZAS both bind to the κB motif. Sen and
ternative splicing events involve sequences encoding the ZAS-N Baltimore first described NF-κB as a constitutively active protein
or ZAS-C regions8,18,19,31 and could lead to the production of at in the nuclei of B cells which binds a specific site in the immuno-
least three truncated protein isoforms: One contains only the globulin enhancer.36 Since then, the NF-κB/Rel family has been
ZAS Zinc Finger Proteins: The Other κB-Binding Protein Family 217
expression of global regulators, such as brinker, Schnurri illus- duced by disruption of ZAS3 in cell and animal models suggest
trates the potential for the ZAS proteins as master regulators of that ZAS3 may have clinical relevance as a tumor suppressor.
gene transcription. Chromosomal analysis of multiple tumors identify mutations
around the ZAS3 locus, 1p34: Allelic loss at 1p34 has been
Signal Transduction by ZAS3 and TRAF Proteins reported in multiple tumors including breast cancer,62-65 liver
A role for the ZAS proteins in mammalian TGF-β pathways cancers,66-68 pheochromocytomas,69 and B cell lymphomas.70
has not yet been established. However, biochemical properties Furthermore, changes in expression of ZAS1 and ZAS2 pro-
of ZAS3 suggest the ZAS proteins to be well-suited to mediate teins have been associated with poor prognosis in chronic lym-
signal transduction as DNA binding is regulated by posttrans- phocytic leukemia patients.71
lational modification: ZAS3 is a substrate for epidermal growth
factor kinase and p34cdc2 in vitro, and phosphorylation on Immunity
serine and tyrosine residues increases DNA binding affinity of a Review of known ZAS target genes (Table 1A) implicates the
ZAS3 fusion protein.24 Recent evidence suggests that ZAS3 may ZAS family in transcriptional regulation of genes involved in
function in signal transduction. The result of a yeast two-hybrid immune response as well as cell proliferation. In addition to the
screen suggests that ZAS3 may interact with TRAF proteins κB motif, ZAS3 has been shown to specifically bind the recom-
which transduces TFN receptor signaling, resulting in bination signal sequences (RSS) which flank gene segments eli-
down-regulation of TRAF-dependent JNK activation.53 The gible for somatic recombination in the variable regions of immu-
ZAS proteins, therefore, may have transcriptional as well as noglobulin and T cell receptor genes.31,45,72 These conserved
nontranscriptional functions. motifs are cis-acting gene elements essential for recombination
activating gene (RAG-1 and RAG-2) mediated recombination.73,74
Proliferation and Carcinogenesis In fact, ZAS3 is the most abundant RSS-binding species in the
While molecular studies continue to identify ZAS DNA tar- thymus.35 What a transcription factor might be doing binding
gets and explore the role of ZAS proteins in signal transduction the RSS is not yet clear, however the DNA-binding ability of
pathways, functional studies demonstrate that the ZAS proteins ZAS3 does correlate with activation of V(D)J recombination.25
have important roles in the regulation of cellular proliferation In the context of the accessibility model, ZAS3 may potentially
and growth. Expression of all three ZAS genes can be induced activate transcription at specific target loci, rendering them ac-
by mitogens, such as fetal bovine serum, PHA/PMA and li- cessible to the V(D)J recombinase.75,76 Additionally, ZAS3 could
popolysaccharide.1,35,53,54 In Drosophila, expression of Schnurri theoretically recruit other transcriptional proteins to RSS to ac-
in cyst cells is required to restrict proliferation and mitosis of tivate recombination or block the access of RAGs. Ultimately,
neighboring germ cells.55 In the case of ZAS3, expression is ZAS3 does not appear to be essential for V(D)J recombination
normally restricted to lymphoid and neuronal tissues. However, as RAG2-/-;ZAS3-/- chimeric mice do develop mature thymocytes.
ZAS3 transcripts are detected in every cultured cell line exam- However, there is a progressive depletion of the double-positive
ined, regardless of origin (unpublished observation). When (CD4+/CD8+) T cell compartment over time.57 Similarly, tar-
ZAS3 expression was decreased by an antisense construct in a geted disruption of ZAS2 also results in T cell maturation de-
preB cell line, 38B9, cell proliferation accelerated.30 Conversely, fects. Thymi of ZAS2-/- mice contain few single positive (CD4+/
when ZAS3 protein expression was increased, cell proliferation CD8- and CD4-/CD8+) thymocytes but with normal levels of
slowed. Furthermore, decreased ZAS3 protein expression in double positive (CD4+/CD8+) thymocytes, reflecting defects
HeLa cells conferred anchorage independence and induced for- in positive selection.77 These data suggest that individual ZAS
mation of terminal multinucleated giant cells.30 A similar phe- proteins may be important at different stages of thymocyte
notype was observed with p53 anti-sense expression in HeLa development.
cells.56 These findings support a role for ZAS3 in regulation of
normal cell growth, proliferation, and cell death. Future Directions
The generation of embryonic stem cells in which endogenous From Drosophila to mammals, the ZAS C2H2 zinc finger pairs
ZAS3 is disrupted has provided clues the physiological function specifically bind the κB motif. The snapshots of ZAS-mediated
of ZAS3 protein. In a RAG2-/-;ZAS3-/- complementation assay, transcription collected to date suggest a model in which ZAS
a tumor spontaneously formed in the axilla of one chimera which proteins may function as adaptors that recruit or inhibit other
was composed of > 90% ZAS3-/- cells. The tumor was first noted transcription factors to gene regulatory sites, resulting in activa-
at 3-weeks as a 5 mm mass which grew to more than 20% of the tion or repression of transcription of target genes. Binding of the
total body mass by 5 weeks. DNA analysis showed that the tu- κB motif by the Rel/NF-κB proteins has been extensively stud-
mor was derived entirely from ZAS3-/- cells. It was diagnosed as ied over the past decade (reviewed in 38-41). While ZAS and
a malignant teratoma due to the presence of poorly-differentiated Rel/NF-κB share the same DNA target, little is known about
carcinoma cells coupled with multiple tissues of diverse origin, their competition or interaction. It is reasonable to hypothesize
including hyaline cartilage, skeletal muscle, fat, lymphocytes, that ZAS proteins may stimulate or inhibit NF-κB-mediated tran-
and primitive neural elements. Permanent cell lines were subse- scription.
quently derived from teratoma cells, which continue to grow The ZAS proteins are subject to post-translational regulation,
without viral transformation for over 2 years.57 The spontane- a key feature of proteins involved in signal transduction. In Droso-
ous development of an extragonadal teratoma is very rare. In phila, Schnurri functions as a master-regulator of transcription
the past 25 years, only one case of such an event has been re- by repressing transcription of a super-repressor gene, brinker.
ported in mice.58 Teratomas, however, have been induced in While no brinker homolog has yet been identified in vertebrates,
animals with p53 mutations,59 double p53/atm mutations,60 and it will be interesting to investigate a role for ZAS proteins in Smad
double mdm2/p53 mutations.61 The changes in cell growth in- interactions in TGF-β signal transduction.
ZAS Zinc Finger Proteins: The Other κB-Binding Protein Family 219
Disruption of ZAS expression in cell models suggests a funda- 15. Hicar MD, Liu Y, Allen CE et al. Structure of the human zinc
mental role for this protein family in regulation of cell growth finger protein HIVEP3: Molecular cloning, expression, exon-intron
and proliferation. Targeted disruption of ZAS genes in animal structure, and comparison with paralogous genes HIVEP1 and
HIVEP2. Genomics 2001; 71(1):89-100.
models resulted in spontaneous tumor formation as well as aber- 16. Wu LC. ZAS: C2H2 Zinc finger proteins involved in growth and
rant thymocyte differentiation. Clearly, ZAS proteins individu- development. Gene Expression 2002; 10(4):137-152.
ally influence fundamental cellular processes. However, it is not 17. Mignone F, Gissi C, Liuni S et al. Untranslated regions of mRNAs.
yet clear to what extent individual ZAS functions are redundant. Genome Biol 2002; 3(3):REVIEWS0004.
In Drosophila, disruption of Schnurri is lethal, while disruption 18. Muchardt C, Seeler JS, Nirula A et al. Regulation of human
of mammalian ZAS is not. It will be intriguing to compare the immunodeficiency virus enhancer function by PRDII-BF1 and c-rel
gene products. J Virol 1992; 66(1):244-250.
phenotypes of individual ZAS knock-out animals with 19. Mak CH, Li Z, Allen CE et al. KRC transcripts: Identification of
double-knockouts and, ultimately, a triple knock-out. Ultimately, an unusual alternative splicing event. Immunogenetics 1998;
as ZAS functions become further defined, post-translational 48(1):32-39.
modifications of ZAS proteins or pharmacology strategies based 20. Iuchi S. Three classes of C2H2 zinc finger proteins. Cell Mol Life
on ZAS zinc finger structure may become useful in modifying Sci 2001; 58(4):625-635.
ZAS protein function in human disease. 21. Knight RD, Shimeld SM. Identification of conserved C2H2
zinc-finger gene families in the Bilateria. Genome Biol 2001;
2(5):RESEARCH0016.
References 22. Tian Q, Kopf GS, Brown RS et al. Function of basonuclin in
1. Baldwin AS, LeClair KP, Singh H et al. A large protein containing increasing transcription of the ribosomal RNA genes during mouse
zinc finger domains binds to related sequence elements in the oogenesis. Development 2001; 128(3):407-416.
enhancers of the class I major histocompatibility complex and kappa 23. Tseng H, Matsuzaki K, Lavker RM. Basonuclin in murine corneal
immunoglobulin genes. Mol Cell Biol 1990; 10(4):1406-1414. and lens epithelia correlates with cellular maturation and proliferative
2. Maekawa T, Sakura H, Sudo T et al. Putative metal finger structure ability. Differentiation 1999; 65(4):221-227.
of the human immunodeficiency virus type 1 enhancer binding 24. Bachmeyer C, Mak CH, Yu CY et al. Regulation by phosphorylation
protein HIV-EP1. J Biol Chem 1989; 264(25):14591-14593. of the zinc finger protein KRC that binds the kappaB motif and
3. Fan CM, Maniatis T. A DNA-binding protein containing two widely V(D)J recombination signal sequences. Nucleic Acids Res 1999;
separated zinc finger motifs that recognize the same DNA sequence. 27(2):643-648.
Genes Dev 1990; 4(1):29-42. 25. Wu LC, Hicar MD, Hong J et al. The DNA-binding ability of
4. Fountain JW, Dracopoli NC, Housman DE et al. MspI RFLP HIVEP3/KRC decreases upon activation of V(D)J recombination.
detected by a ZNF-40 cDNA sequence. Nucleic Acids Res 1991; Immunogenetics 2001; 53(7):564-571.
19(9):2514. 26. Suzuki M. SPXX, a frequent sequence motif in gene regulatory
5. Nakamura T, Donovan DM, Hamada K et al. Regulation of the proteins. J Mol Biol 1989; 207(1):61-84.
mouse alpha A-crystallin gene: Isolation of a cDNA encoding a 27. Brzeski J, Grycuk T, Lipkowski AW et al. Binding of SPXK- and
protein that binds to a cis sequence motif shared with the major APXK-peptide motifs to AT-rich DNA. Experimental and theoretical
histocompatibility complex class I gene and other genes. Mol Cell studies. Acta Biochim Pol 1998; 45(1):221-231.
Biol 1990; 10(7):3700-3708. 28. Kennelly PJ, Krebs EG. Consensus sequences as substrate specificity
6. Brady JP, Kantorow M, Sax CM et al. Murine transcription factor determinants for protein kinases and protein phosphatases. J Biol
alpha A-crystallin binding protein I. Complete sequence, gene Chem 1991; 266(24):15555-15558.
structure, expression, and functional inhibition via antisense RNA. 29. Kantorow M, Becker K, Sax CM et al. Binding of tissue-specific
J Biol Chem 1995; 270(3):1221-1229. forms of alpha A-CRYBP1 to their regulatory sequence in the mouse
7. Mitchelmore C, Traboni C, Cortese R. Isolation of two cDNAs alpha A-crystallin-encoding gene: Double-label immunoblotting of
encoding zinc finger proteins which bind to the alpha 1-antitrypsin UV-crosslinked complexes. Gene 1993; 131(2):159-165.
promoter and to the major histocompatibility complex class I 30. Allen CE, Wu LC. Downregulation of KRC induces proliferation,
enhancer. Nucleic Acids Res 1991; 19(1):141-147. anchorage independence, and mitotic cell death in HeLa cells. Exp
8. Wu LC, Liu Y, Strandtmann J et al. The mouse DNA binding Cell Res 2000; 260(2):346-356.
protein Rc for the kappa B motif of transcription and for the V(D)J 31. Wu LC, Mak CH, Dear N et al. Molecular cloning of a zinc finger
recombination signal sequences contains composite DNA-protein protein which binds to the heptamer of the signal sequence for V(D)J
interaction domains and belongs to a new family of large recombination. Nucleic Acids Res 1993; 21(22):5067-5073.
transcriptional proteins. Genomics 1996; 35(3):415-424. 32. Seeler JS, Muchardt C, Suessle A et al. Transcription factor
9. Boguski MS, Schuler GD. ESTablishing a human transcript map. PRDII-BF1 activates human immunodeficiency virus type 1 gene
Nat Genet 1995; 10(4):369-371. expression. J Virol 1994; 68(2):1002-1009.
10. Arora K, Dai H, Kazuko SG et al. The Drosophila schnurri gene 33. Tanaka K, Matsumoto Y, Nakatani F et al. A zinc finger transcription
acts in the Dpp/TGF beta signaling pathway and encodes a factor, alpha A-crystallin binding protein 1, is a negative regulator
transcription factor homologous to the human MBP family. Cell of the chondrocyte-specific enhancer of the alpha1(II) collagen gene.
1995; 81(5):781-790. Mol Cell Biol 2000; 20(12):4428-4435.
11. Grieder NC, Nellen D, Burke R et al. Schnurri is required for 34. Dorflinger U, Pscherer A, Moser M et al. Activation of somatostatin
Drosophila Dpp signaling and encodes a zinc finger protein similar receptor II expression by transcription factors MIBP1 and SEF-2 in
to the mammalian transcription factor PRDII-BF1. Cell 1995; the murine brain. Mol Cell Biol 1999; 19(5):3736-3747.
81(5):791-800. 35. Hicar MD, Robinson ML, Wu LC. Embryonic expression and
12. Staehling-Hampton K, Laughon AS, Hoffmann FM. A Drosophila regulation of the large zinc finger protein KRC. Genesis 2002;
protein related to the human zinc finger transcription factor PRDII/ 33(1):8-20.
MBPI/HIV-EP1 is required for dpp signaling. Development 1995; 36. Sen R, Baltimore D. Multiple nuclear factors interact with the
121(10):3393-3403. immunoglobulin enhancer sequences. Cell 1986; 46(5):705-716.
13. Sudo T, Ozawa K, Soeda EI et al. Mapping of the human gene for 37. Baldwin AS. The NF-kappa B and I kappa B proteins: New
the human immunodeficiency virus type 1 enhancer binding protein discoveries and insights. Annu Rev Immunol 1996; 14:649-683.
HIV-EP2 to chromosome 6q23-q24. Genomics 1992; 38. Ghosh S, May MJ, Kopp EB. NF-kappa B and Rel proteins:
12(1):167-170. Evolutionarily conserved mediators of immune responses. Annu Rev
14. Olavesen MG, Bentley E, Mason RV et al. Fine mapping of 39 Immunol 1998; 16:225-260.
ESTs on human chromosome 6p23-p25. Genomics 1997;
46(2):303-306.
220 Zinc Finger Proteins: From Atomic Contact to Cellular Function
39. Perkins ND. The Rel/NF-kappa B family: Friend and foe. Trends 60. Westphal CH, Rowan S, Schmaltz C et al. atm and p53 cooperate
Biochem Sci 2000; 25(9):434-440. in apoptosis and suppression of tumorigenesis, but not in resistance
40. Pahl HL. Activators and target genes of Rel/NF-kappaB transcription to acute radiation toxicity. Nat Genet 1997; 16(4):397-401.
factors. Oncogene 1999; 18(49):6853-6866. 61. McDonnell TJ, Montes de Oca Luna R, Cho S et al. Loss of one
41. Chen F, Castranova V, Shi X et al. New insights into the role of but not two mdm2 null alleles alters the tumour spectrum in p53
nuclear factor-kappaB, a ubiquitous transcription factor in the null mice. J Pathol 1999; 188(3):322-328.
initiation of diseases. Clin Chem 1999; 45(1):7-17. 62. Emi M, Yoshimoto M, Sato T et al. Allelic loss at 1p34, 13q12,
42. Karin M, Cao Y, Greten FR et al. NF-kappaB in cancer: From 17p13.3, and 17q21.1 correlates with poor postoperative prognosis
innocent bystander to major culprit. Nat Rev Cancer 2002; in breast cancer. Genes Chromosomes Cancer 1999; 26(2):134-141.
2(4):301-310. 63. Tsukamoto K, Ito N, Yoshimoto M et al. Allelic loss on chromosome
43. Gerondakis S, Grossmann M, Nakamura Y et al. Genetic approaches 1p is associated with progression and lymph node metastasis of
in mice to understand Rel/NF-kappaB and IkappaB function: primary breast carcinoma. Cancer 1998; 82(2):317-322.
Transgenics and knockouts. Oncogene 1999; 18(49):6888-6895. 64. Utada Y, Emi M, Yoshimoto M et al. Allelic loss at 1p34-36 predicts
44. Singh H, LeBowitz JH, Baldwin AS et al. Molecular cloning of poor prognosis in node-negative breast cancer. Clin Cancer Res 2000;
an enhancer binding protein: Isolation by screening of an 6(8):3193-3198.
expression librar y with a recognition site DNA. Cell 65. Hirano A, Utada Y, Haga S et al. Allelic losses as prognostic markers
1988;52(3):415-423. for breast cancers. Int J Clin Oncol 2001; 6(1):6-12.
45. Allen CE, Mak CH, Wu LC. The kappaB transcriptional enhancer 66. Li SP, Wang HY, Li JQ et al. Genome-wide analyses on loss of
motif and signal sequences of V(D)J recombination are targets for heterozygosity in hepatocellular carcinoma in Southern China. J
the zinc finger protein HIVEP3/KRC: A site selection amplification Hepatol 2001; 34(6):840-849.
binding study. BMC Immunology 2002; 3(1):10. 67. Parada LA, Hallen M, Tranberg KG et al. Frequent rearrangements
46. Fukuda S, Yamasaki Y, Iwaki T et al. Characterization of the of chromosomes 1, 7, and 8 in primary liver cancer. Genes
biological functions of a transcription factor, c-myc intron binding Chromosomes Cancer 1998; 23(1):26-35.
protein 1 (MIBP1). J Biochem (Tokyo) 2002; 131(3):349-357. 68. Sun M, Eshleman JR, Ferrell LD et al. An early lesion in hepatic
47. Hjelmsoe I, Allen CE, Cohn MA et al. The kappaB and V(D)J carcinogenesis: Loss of heterozygosity in human cirrhotic livers and
recombination signal sequence binding protein KRC regulates dysplastic nodules at the 1p36-p34 region. Hepatology 2001;
transcription of the mouse metastasis-associated gene S100A4/mts1. 33(6):1415-1424.
J Biol Chem 2000; 275(2):913-920. 69. Benn DE, Dwight T, Richardson AL et al. Sporadic and familial
48. Massague J. TGF-beta signal transduction. Annu Rev Biochem 1998; pheochromocytomas are associated with loss of at least two discrete
67:753-791. intervals on chromosome 1p. Cancer Res 2000; 60(24):7048-7051.
49. Raftery LA, Sutherland DJ. TGF-beta family signal transduction in 70. Liu J, Sen R, Rothstein TL. Abnormal kappa B-binding protein in
Drosophila development: From Mad to Smads. Dev Biol 1999; 2 the cytoplasm of a plasmacytoma cell line that lacks nuclear
10(2):251-268. expression of NF-kappa B. Mol Immunol 1993; 30(5):479-489.
50. Dai H, Hogan C, Gopalakrishnan B et al. The zinc finger protein 71. Aalto Y, El Rifa W, Vilpo L et al. Distinct gene expression profiling
schnurri acts as a Smad partner in mediating the transcriptional in chronic lymphocytic leukemia with 11q23 deletion. Leukemia
response to decapentaplegic. Dev Biol 2000; 227(2):373-387. 2001; 15(11):1721-1728.
51. Marty T, Muller B, Basler K et al. Schnurri mediates Dpp-dependent 72. Mak CH, Strandtmann J, Wu LC. The V(D)J recombination signal
repression of brinker transcription. Nat Cell Biol 2000; sequence and kappa B binding protein Rc binds DNA as dimers
2(10):745-749. and forms multimeric structures with its DNA ligands. Nucleic Acids
52. Torres-Vazquez J, Park S, Warrior R et al. The transcription factor Res 1994; 22(3):383-390.
Schnurri plays a dual role in mediating Dpp signaling during 73. Oettinger MA, Schatz DG, Gorka C et al. RAG-1 and RAG-2,
embryogenesis. Development 2001; 128(9):1657-1670. adjacent genes that synergistically activate V(D)J recombination.
53. Oukka M, Kim ST, Lugo G et al. A mammalian homolog of Science 1990; 248(4962):1517-1523.
Drosophila schnurri, KRC, regulates TNF receptor-driven responses 74. Schatz DG, Oettinger MA, Baltimore D. The V(D)J recombination
and interacts with TRAF2. Mol Cell 2002; 9(1):121-131. activating gene, RAG-1. Cell 1989; 59(6):1035-1048.
54. Nomura N, Zhao MJ, Nagase T et al. HIV-EP2, a new member of 75. Yancopoulos GD, Alt FW. Developmentally controlled and
the gene family encoding the human immunodeficiency virus type tissue-specific expression of unrearranged VH gene segments. Cell
1 enhancer-binding protein. Comparison with HIV-EP1/PRDII-BF1/ 1985; 40(2):271-281.
MBP-1. J Biol Chem 1991; 266(13):8590-8594. 76. Schlissel MS, Stanhope-Baker P. Accessibility and the developmental
55. Matunis E, Tran J, Gonczy P et al. punt and schnurri regulate a regulation of V(D)J recombination. Semin Immunol 1997;
somatically derived signal that restricts proliferation of committed 9(3):161-170.
progenitors in the germline. Development 1997; 124(21):4383-4391. 77. Takagi T, Harada J, Ishii S. Murine Schnurri-2 is required for positive
56. Iotsova V, Stehelin D. Antisense p53 provokes changes in HeLa cell selection of thymocytes. Nat Immunol 2001; 2(11):1048-1053.
growth and morphology. Eur J Cell Biol 1995; 68(2):122-132. 78. van ‘t Veer LJ, Lutz PM, Isselbacher KJ et al. Structure and expression
57. Allen CE, Muthusamy N, Weisbrode SE et al. Developmental of major histocompatibility complex-binding protein 2, a 275-kDa
anomalies and neoplasia in animals and cells deficient in the large zinc finger protein that binds to an enhancer of major
zinc finger protein KRC. Genes Chromosomes Cancer 2002; histocompatibility complex class I genes. Proc Natl Acad Sci USA
35(4):287-298. 1992; 89(19):8971-5.
58. Artzt K, Damjanov I. Spontaneous extragonadal teratocarcinoma in 79. Rustgi AK, Van ‘t Veer LJ, Bernards R. Two genes encode factors
a mouse. Lab Anim Sci 1978; 28(5):584-586. with NF-kappa B- and H2TF1-like DNA-binding properties. Proc
59. Jacks T, Remington L, Williams BO et al. Tumor spectrum analysis Natl Acad Sci USA 1990; 87(22):8707-10.
in p53-mutant mice. Curr Biol 1994; 4(1):1-7. 80. Makino R, Akiyama K, Yasuda J et al. Cloning and characterization
of a c-myc intron binding protein (MIBP1). Nucleic Acids Res 1994;
22(25):5679-5685.
CHAPTER 30
Abstract In Vertebrates
M
embers of the GATA transcription factor family are During vertebrate development, hematopoiesis occurs in two
found throughout eukaryotes, including plants, fungi, successive waves (reviewed in ref. 17). In mammals, the blood
invertebrates and vertebrates. In this review, we islands of the yolk sac produce transient embryonic red blood
discuss some of the roles of GATA factors during vertebrate and cells (primitive hematopoiesis). At midgestation, definitive
invertebrate development. We place particular emphasis on their hematopoietic stem cells (HSC), specified in the AGM
function in hematopoiesis, in heart and in endoderm (aorta-gonad-mesonephros) region, colonise foetal and definitive
development, phenomena in which several different organisms hematopoietic sites, such as the embryonic liver and the bone
have shown striking molecular and developmental similarities. marrow. Subsequent hematopoietic differentiation is usually
assumed to proceed in a hierarchical fashion: HSC give rise to
multipotent progenitors and then to committed progenitors that
Introduction will differentiate into one or several specialised blood cell types
Members of the GATA transcription factor family are so called (Fig. 1). “Hematopoietic” GATA factors display distinct, but
because they bind to the DNA consensus site WGATAR.1, 2 They overlapping, expression profiles during hematopoietic
all contain one or two conserved zinc finger DNA-binding development, in which they participate from the emergence of
domains, with the characteristic consensus sequence the HSC to terminal differentiation of blood cells.
Cys-X2-Cys-X17-18-Cys-X2-Cys followed by a basic region. GATA GATA2 is expressed in the primitive and definitive HSC as
factors are found throughout eukaryotes, from yeast to well as in mast cells and megakaryocytes and at lower levels in
vertebrates,3 where they regulate a variety of developmental erythroid cells.18-21 Homozygous GATA2 mutant mice die by
processes. In yeast, fungi and plants, they participate in nitrogen embryonic day 10 (E10) after having undergone severely reduced
metabolism, blue-light-regulated morphogenesis and circadian primitive hematopoiesis.22 Furthermore, GATA2 null embryonic
rhythm.4 In animals, GATA factors have been implicated in stem cells (ES) fail to contribute to definitive hematopoiesis in
endoderm formation, hematopoiesis and heart development in chimeric animals.23 Thus GATA2 is essential for the development
numerous species,5 in adipogenesis,6 urogenital development7 and of the blood cell progenitors. One key target of GATA2 in the
neurogenesis8,9 in vertebrates, and in patterning and ectoderm HSC is probably SCL/Tal1, a basic-Helix-Loop-Helix
differentiation in invertebrates.10-14 Here we shall concentrate on (bHLH)-encoding gene, required for all blood cell development,
the conserved functions of GATA factors in the animal kingdom that contains functionally important GATA2 binding sites in its
during hematopoiesis, heart formation and endoderm enhancer.24 The expression pattern of GATA2 in zebrafish and
development and we shall discuss the implication of GATA genes Xenopus suggests that the role of this factor in hematopoietic
in human diseases. progenitor development has been conserved.25-27 Interestingly,
GATA2 can either promote proliferation at the expense of
differentiation or induce differentiation, depending on the
Hematopoieisis hematopoietic cell type in which it is expressed and on its
The founding member of the GATA family, GATA1, was first expression level. Indeed, GATA2 is required at high levels to
identified as an erythroid-specific transcription factor that binds promote proliferation of early hematopoietic progenitors and/or
to functionally important cis-regulatory sequences in the globin survival,23,28 but it has to be downregulated to allow erythroid
genes.15,16 Subsequent cloning of the other vertebrate GATA differentiation to occur.29,30 In addition, GATA2 is required for
factors revealed that three of them, GATA1, 2 and 3, were mast cell differentiation.23
prominently expressed in blood cells. Targeted disruptions of each GATA1 is expressed at a high level in erythroid cells, mast
of these three genes in mouse demonstrate that they play critical cells, megakaryocytes, and eosinophils, and at a low level in
roles in hematopoiesis. Furthermore, a growing body of evidence multipotent progenitors.15,16 Several lines of evidence show that
shows that GATA factor function during hematopoiesis has been GATA1 plays a central role in erythroid differentiation. In
conserved through evolution. chimeric mice, GATA1-/- ES cells contribute to white blood cells
Zinc Finger Proteins: From Atomic Contact to Cellular Function, edited by Shiro Iuchi and Natalie Kuldell.
©2005 Landes Bioscience/Eurekah.com and Kluwer Academic/Plenum Publishers.
222 Zinc Finger Proteins: From Atomic Contact to Cellular Function
represses Th1 development by inhibiting Th1-specific genes and transcription factors involved in the cardiogenic program
induces Th2 cytokine expression by direct transactivation of the (reviewed in ref. 89). Of these, the GATA factors appear to play a
IL5 and IL13 gene, as well as through remodelling/demethylation critical role.
of the IL4 locus.76-79 Conversely, IL12 signalling inhibits GATA3
expression/function and induces Th1 development by activating In Drosophila
STAT4 and T-bet expression.78 Thus, the precise regulation of During Drosophila embryogenesis, the GATA factor-encoding
GATA3 expression at several stages is essential for proper T cell gene pannier (pnr) is expressed in the dorsal mesoderm from which
differentiation. the cardiac cells develop.90 Its expression overlaps that of the Nk2
type homeodomain protein-encoding gene tinman (tin), a key
In Drosophila regulator of heart development.91 The expression of both genes
Despite the physiological differences between Drosophila and in the dorsal mesoderm depends on the signalling molecule Dpp,
vertebrate blood cells, various aspects of hematopoietic a member of the bone morphogenetic protein (BMP) family,
development have been evolutionarily conserved.80,81 As in which is expressed in the overlaying dorsal ectoderm. In embryos
vertebrates, Drosophila hematopoiesis (from mesodermally-derived deficient for pnr function, there results a dramatic reduction in
progenitors) occurs in two waves: one during embryogenesis and the number of cardiac progenitors. Moreover, in pnr mutant
the second during the larval stages.82,83 Blood cells (hemocytes) embryos, tin expression is greatly diminished along the dorsal
differentiate into two major classes: plasmatocytes and crystal edge of mesoderm, where the heart precursors form. In tin
cells.82 Plasmatocytes act as macrophages, whereas crystal cells loss-of-function mutants, no heart cells develop at all, which
participate in melanisation, an insect-specific process involved in suggests that the extensive loss of heart cells in pnr mutant is
the encapsulation of foreign bodies and in wound healing. Most probably a consequence of this strong reduction of tin
strikingly, it appeared recently that transcription factors of the expression. The few cells that still express tin develop normally.92
GATA, FOG and RUNX families, which play critical roles Interestingly, whereas pnr expression requires tin, pnr may also
during hematopoiesis in vertebrates, are also involved in regulate tin expression, since GATA binding sites have been
controlling blood cell development in Drosophila. Like GATA2, identified in its promotor region. Besides these transcriptional
the Drosophila GATA factor Serpent (Srp) is expressed in blood cross-regulations, Tin and Pnr have also been shown to interact
cell precursors and is required for the determination, to activate the MADs box transcription factor dMef2, another
proliferation and maintenance of this population. 84,85 gene essential for cardiac myogenesis that is conserved in
Furthermore, Srp is still detected in mature plasmatocytes and, at vertebrates.93,94 Finally, the evolutionary conservation of the
lower levels, in crystal cells. Reminiscent of GATA2 regulation structure and function of GATA factors is strikingly demonstrated
during erythropoiesis, down-regulation of srp expression in by the fact that mouse GATA4 can substitute for pnr during
crystal cells may be important, as suggested by the fact that Drosophila cardiogenesis.90
overexpression of Srp in these cells inhibits their differentiation.86
Nonetheless, Srp is required for crystal cell differentiation in In Vertebrates
combination with the RUNX factor Lozenge (Lz), with which it The vertebrate heart develops from a bilaterally symmetrical
synergistically interacts to trigger the crystal cell-specific genetic region of the anterior lateral plate mesoderm, the cardiac
program (Waltzer et al, in press). Interestingly, the interaction crescent, which adopts a cardiac fate in response to BMP
and cooperation between GATA and RUNX factors appear to be signalling from the adjacent endoderm.95,96 In mouse, three GATA
conserved in vertebrates and it has been suggested that GATA1 factors, GATA4, 5 and 6, are expressed at various stages during
cooperates with RUNX1 during megakaryopoiesis. 87 heart development and can regulate cardiac-specific gene expres-
Furthermore, the Drosophila FOG factor U-shaped (Ush) represses sion in cell culture.97-100 At E7.5, GATA4 and GATA6 are both
crystal cell fate by interacting with Srp,86,88 suggesting that a expressed in the mesoderm that gives rise to the cardiogenic plate
competition between Ush and Lz may modulate Srp activity and whereas GATA5 is expressed in a restricted manner within the
thereby influence the subsequent development into plasmatocytes precardiac mesoderm. Subsequently, GATA4, 5 and 6 are
rather than crystal cells or vise versa. Such cross-interactions
expressed throughout the primitive heart tube, but only GATA4
between GATA factors and their partners may constitute a
and 6 continue to be expressed in the heart after E17 and during
conserved means of regulating lineage choice and differentiation
postnatal development. Expression of GATA5 becomes restricted
during hematopoiesis.
to the atrial endocardium and is extinguished at E16.5. These
three GATA factors are among the earliest markers of the cardiac
Heart Formation myocyte lineage and are expressed as early as the vertebrate
Recent studies have shown that heart formation is controlled ortholog of tin, Nkx2-5. 101 In the mesoderm, Nkx2-5 is expressed
by a developmental program in which many molecular and in the cardiogenic crescent on each side of the embryo in a
developmental aspects are conserved between fly and man. pattern very similar to that of GATA4.
Whereas heart morphology in Drosophila melanogaster is simpler The role of GATA factors in cardiogenesis has been analysed
than the complex multichambered heart of mammals, both are by gene inactivation studies. In GATA4 knockout mice, the heart
constituted of spontaneous contractile myocardial cells. In both tube fails to form, but this does not appear to be a consequence
vertebrates and Drosophila, the heart develops from bilaterally of defective cardiomyogenesis. Instead, it appears to result from
symmetrical regions of the mesoderm, where heart progenitors anomalous rostral-to-caudal and lateral-to-ventral folding of the
migrate to the midline and fuse to form a linear, beating heart embryo, which reflects a generalized disruption of the ventral body
tube (Fig. 2). In vertebrates, this tube undergoes further specific pattern. 102,103 Notably, in homozygous mutant embryos,
morphogenetic movements that lead to the definitive looped and differentiated cardiomyocytes form and markers of terminal
multichambered heart. In parallel with these developmental cardiac differentiation are expressed normally. This could well be
similarities, there is a remarkable evolutionary conservation of due to functional redundancy among the GATA factors.
224 Zinc Finger Proteins: From Atomic Contact to Cellular Function
Figure 2. Formation of the heart in vertebrates and Drosophila. Schematic cross sections of human embryos aged 17 days (A), 18 days (B), 21 days
(C), 22 days (D), and a Drosophila embryo aged 20 hrs (E). F) Drosophila embryos aged 11 hrs (a) and 20 hrs (b) respectively. a) cardioblast
precursors on both sides of the dorsal mesoderm, are stained. b) they meet at the dorsal midline and form a tube that includes the heart.
Consistent with this suggestion, the level of GATA6 transcripts is and double knockouts, should shed light on the precise role of
increased in GATA4 knockout embryos. GATA5 knockout mice these factors during cardiogenesis.
are viable, but females display an abnormal urogenital tract.104 Contrary to what is observed in zebrafish, it has been shown
GATA6 null mice die during early development (stage E5.5-E7.5) in mice that neither GATA4 nor Nkx2-5 alone can initiate
as a result of extraembryonic defects (see below),105,106 which pre- cardiogenesis, although either protein can potentiate it in
cludes analysis of cardiogenesis. committed cells. Studies on the atrial natriuretic factor (ANF)
Interestingly, in zebrafish, the mutation faust (fau), which maps promoter have revealed a physical interaction and functional
to the GATA5 gene, is embryonic lethal and leads to a cardiac cooperation between GATA4 and Nkx2-5.108-110 Thus, as in
phenotype similar to that oberserved in GATA4 null mice.107 Drosophila, these factors may cooperate in regulating cardiogenesis
Analysis of fau mutants indicates that GATA5 is required early in in vertebrates.
embryogenesis for the production of normal numbers of
myocardial precursors and the expression of several myocardial Control of Endoderm Formation
genes, including Nkx2.5. Furthermore, overexpression of zebrafish All metazoan phyla contain an endoderm and as this germ
GATA5 activates ectopic expression of Nkx2.5 and induces the layer was probably invented only once during evolution, at least
formation of ectopic regions of rhythmically contracting tissue.107 some elements of the gene regulatory network controlling its
These observations suggest a role for zebrafish GATA5 in the formation may have been conserved during evolution. It has
cardiogenic program, including control of growth, morphogenesis become clear over the last decade that GATA factors are among
and differentiation. the most pervasive members of this regulatory network. Indeed,
Further experiments in mouse using the Cre-lox technique to GATA factors play a crucial role in endoderm development in C.
generate tissue-specific mutants of GATA4 and 6, both as single elegans, Drosophila, sea urchin, Xenopus, zebrafish and mouse.
Role of GATA Factors in Development 225
that expresses higher levels of GATA6.102,103 Thus GATA6 may mutations do indeed affect GATA-mediated regulation of these
compensate for GATA4 loss of function. Furthermore, although genes in vivo remains to be shown. Of note, disruption of a GATA
the definitive endoderm forms normally in GATA4 null motif in the Duffy gene promoter was observed in individuals
embryos, 102,103 analysis of chimeras indicates that with Duffy-negative blood group antigen.156 As this gene code
GATA4-deficient cells cannot differentiate correctly into for the erythrocyte receptor for the Plasmodium vivax malaria
endodermally-derived gastric epithelial cells. 144 GATA6 parasite, this mutation may be linked to the resistance to
knockout mice die in the early postimplantation period and infection by this parasite in West African populations.157
exhibit a specific block in visceral endoderm differentiation.105 The second category of human genetic diseases linked to GATA
However, GATA6-deficient cells contributed to gut-derived factor dysfunction comprises mutations directly affecting GATA
organs, indicating that GATA6 is not required for definitive genes. The first evidence for the implication of a GATA gene mu-
endoderm specification.105,145 Interestingly though, GATA6 is tation in human disease was obtained after a study of a
required for the differentiation of endodermal cells into a congenital heart defect (CHD), found to be linked to a deletion
specialised lung cell type as well as for lung branching morpho- in human chromosome 8p23.1, which contains GATA4.158 Re-
genesis.145,146 Finally, GATA5 knockout mice are viable, as cently, two different heterozygous point mutations in GATA4 were
mentioned earlier, and do not show defects in the endoderm.104 found in two families with CHD. 159 The missense
Since GATA4, 5 and 6 are coexpressed in the prospective mutation G296S results in diminished GATA4 DNA binding
primitive endoderm in the mouse, the fact that endoderm and transcriptional activity, whereas the frameshift mutation
specification is normal after knockout of each of the individual E359del gives rise to a shorter, transcriptionally inactive GATA4
genes may reflect functional redundancy among them. protein. Strikingly, the interaction bewteen GATA4 and the T
Interestingly, it is in zebrafish that the role for vertebrate GATA box factor TBX5, which is also mutated in a subset of cardiac
factors in endoderm formation has been most clearly septal defects, was disrupted by missense mutations in GATA4 or
demonstrated. The fau mutation specifically affects the zebrafish in TBX5 that cause heart defects. Thus, consistent with genetic
GATA5 gene, which is expressed in all endodermal and in some analysis in model organisms, GATA4 plays an essential role,
mesodermal progenitors. fau mutants present major defects in possibly in association with TBX5, during human heart formation.
myocardial differentiation and gut morphogenesis.107 Notably, Haploinsufficient mutations in other GATA factors have been
there is a substantial decrease in the number of endodermal cells, associated with human congenital disease. Several point
which do not differentiate normally into hepatic, pancreatic and mutations affecting GATA1 have been identified in patients with
pharyngeal derivatives.147 Since the fau mutation is not a null X-linked thrombocytopenia.160-164 These missense mutations are
mutation in GATA5, the extent of the endoderm defect due to located in the N-terminal zinc finger and several of them disrupt
total loss of zebrafish GATA5 may be even stronger (see for the interaction with the cofactor FOG1. The phenotype obserevd
example the phenotype of GATA4 -/- in mice). It has also been are consistent with a role for the GATA1/FOG complex during
shown that GATA5 expression depends on the endoderm-inducing megakaryocytic differentiation in human. Furthermore, GATA3
signalling of Nodal,148 and overexpression of GATA5 in zebrafish haploinsufficiency causes human HDR (Hypoparathyroidism,
embryos induced endodermal genes even when Nodal signalling sensorineural Deafness, Renal anomaly) syndrome.165 Surprisingly
was reduced. 147 Thus GATA5 is required for endoderm though, whereas human and mouse GATA3 are expressed in the
formation and differentiation in zebrafish. developing kidney, otic vesicle and parathyroid, no HDR-related
GATA factors thus seem to be part of a cascade of genes phenotype was observed in GATA3 knockout or heterozygous
controlling endoderm specification that has been be conserved in mice. Finally, it was shown that the tricho-rhino-phalangeal
Xenopus, zebrafish and mouse (for review see ref. 149). In these syndrome type I (TRPS1), characterized by craniofacial and
different species, Nodal-related signalling contributes to the ex- skeletal abnormalities, was due to monoallelic nonsense
pression of GATA factors as well as of Mix-type homeodomain mutations in the TRPS1 gene.166 TRPS1 codes for an atypical
proteins in endoderm precursors and both of these factors GATA protein containing only one C-terminal GATA zinc
regulate the expression of Sox17-related factors. Finally, finger and other unrelated zinc fingers. Interestingly, in a severe
members of the Forkhead family of transcription factors, like form of TRPS (type III), mutations in TRPS1 are confined to
GATA factors, have been implicated in endoderm formation, not the GATA domain.167 Consistent with these phenotypes, targeted
only in vertebrates but also in Drosophila and C. elegans. Thus, disruption of the GATA-finger encoding exon in mice mimicked
several aspects of endodermal development have been conserved several features of human TRPS as well as more profound and
during evolution. generalized skeletal dysplasia.168
Finally, somatically acquired mutations leading to GATA gene
GATA Factors and Diseases dysregulation may participate in cancer development. Indeed,
Although GATA factors play crucial roles during normal GATA gene expression has been shown to be upregulated in a
development, direct evidence for their involvement in human variety of primary tumors and cancer cell lines. Moreover, GATA2
diseases has been relatively scarce until recently. The first category may be implicated in leukemia, since it is located near the 3q21
of human inherited disorders linked to GATA factor dysfunction breakpoint found in some myeloid leukemias and is overexpressed
consists of point mutations of GATA binding sites located in the in these leukemias.169 Recently, acquired somatic mutations in
cis-regulatory regions of GATA1 target genes. For example, three GATA1 have been specifically associated with megakaryocytic
independent single nucleotide transitions affecting GATA1 leukemia and transient myeloproliferative disorder in children
binding were found in the ∂-globin gene of patients with with Down’s syndrome.170,171 These mutations result in the
delta-thalassemia, a hereditary disease characterised by reduced expression of a truncated GATA1 protein that lacks the N-terminal
synthesis of ∂-globin.150-152 Mutations in GATA binding sites transactivation domain and they are probably initiating events in
potentially affecting the expression of a relevant gene have been these leukemia. Since RUNX1, the gene coding for which is
observed in other congenital diseases.153-155 Whether these located on chromosome 21, interacts and cooperates with GATA1
Role of GATA Factors in Development 227
to induce megakaryocytic differentiation in vitro, it is tempting 9. Nardelli J, Thiesson D, Fujiwara Y et al. Expression and genetic
to speculate that the interaction between GATA1 and RUNX1 interaction of transcription factors GATA-2 and GATA-3 during
participates in the development of these megakaryocytoses in development of the mouse central nervous system. Dev Biol 1999;
210(2):305-321.
human. 10. Ramain P, Heitzler P, Haenlin M et al. pannier, a negative regulator
of achaete and scute in Drosophila, encodes a zinc finger protein
Conclusions with homology to the vertebrate transcription factor GATA-1.
Development 1993; 119(4):1277-1291.
During the last decade, much has been learned concerning 11. Calleja M, Herranz H, Estella C et al. Generation of medial and
the functions of GATA factors in animal development. In lateral dorsal body domains by the pannier gene of Drosophila.
vertebrates, based on protein sequence, exon/intron structure, Development 2000;127(18): 3971-3980.
expression pattern and functions, GATA factors can be assigned 12. Herranz H, Morata G. The functions of pannier during Drosophila
to two different families (GATA1/2/3 and GATA4/5/6) that arose embryogenesis. Development 2001; 128(23):4837-4846.
by duplications of two ancestral deuterostome GATAs.3 The 13. Koh K, Peyrot SM, Wood CG et al. Cell fates and fusion in the C.
precise evolutionary relationship between GATA factors from elegans vulval primordium are regulated by the EGL-18 and ELT-6
GATA factors-apparent direct targets of the LIN-39 Hox protein.
invertebrates and vertebrates remains unclear. However, it is Development 2002; 129(22):5171-5180.
striking that members of the GATA family play evolutionarily 14. Page BD, Zhang W, Steward K et al. ELT-1, a GATA-like
conserved roles in heart, blood and endoderm development from transcription factor, is required for epidermal cell fates in
invertebrates to mammals. Moreover, some of their partners and Caenorhabditis elegans embryos. Genes Dev 1997;
of their target genes have also been functionally conserved in these 11(13):1651-1661.
processes. Altogether, GATA factors appear to be critical regulators 15. Tsai SF, Martin DI, Zon LI et al. Cloning of cDNA for the major
DNA-binding protein of the erythroid lineage through expression
of proliferation, differentiation and survival of various cell in mammalian cells. Nature 1989; 339(6224):446-451.
lineages in the three germ layers. They achieve these diverse 16. Evans T, Felsenfeld G. The erythroid-specific transcription factor
functions either through the successive use of different sets of Eryf1: a new finger protein. Cell 1989; 58(5):877-885.
GATA factors harbouring intrinsic specificity or through the 17. Orkin SH, Zon LI. Hematopoiesis and stem cells: plasticity versus
reiterative use of the same factor but at different thresholds or in developmental heterogeneity. Nat Immunol 2002; 3(4):323-328.
combination with various cofactors. 18. Leonard M, Brice M, Engel JD et al. Dynamics of GATA
transcription factor expression during erythroid differentiation. Blood
In the future, the use of cell-type-specific targeted gene
1993; 82(4):1071-1079.
disruption in mice and the analysis of human inherited disease 19. Jippo T, Mizuno H, Xu Z et al. Abundant expression of transcription
caused by GATA dysfunction should help understand the factor GATA-2 in proliferating but not in differentiated mast cells
physiological roles and mechanisms of action of GATA genes. in tissues of mice: demonstration by in situ hybridization. Blood
Furthermore, genetic screens in zebrafish, Drosophila, C. elegans 1996; 87(3):993-998.
and other model organisms may also provide unexpected insights 20. Harigae H, Takahashi S, Suwabe N et al. Differential roles of
into GATA functions and regulations during development. GATA-1 and GATA-2 in growth and differentiation of mast cells.
Genes Cells 1998; 3(1):39-50.
21. Minegishi N, Ohta J, Yamagiwa H et al. The mouse GATA-2 gene
Acknowledgements is expressed in the para-aortic splanchnopleura and aorta-gonads and
We are grateful to J. Smith for critically reading the mesonephros region. Blood 1999; 93(12):4196-4207.
22. Tsai FY, Keller G, Kuo FC et al. An early haematopoietic defect in
manuscript. We thank B. Savelli for art work. This work was
mice lacking the transcription factor GATA-2. Nature 1994;
supported by the Centre National de Recherche Scientifique 371(6494):221-226.
(CNRS) and grants from the Association pour la Recherche sur 23. Tsai FY, Orkin SH. Transcription factor GATA-2 is required for
le Cancer (ARC). proliferation/survival of early hematopoietic cells and mast cell
formation, but not for erythroid and myeloid terminal differentiation.
References Blood 1997; 89(10):3636-3643.
24. Gottgens B, Nastos A, Kinston S et al. Establishing the transcriptional
1. Martin DI, Orkin SH. Transcriptional activation and DNA binding
programme for blood: the SCL stem cell enhancer is regulated by a
by the erythroid factor GF-1/NF-E1/Eryf 1. Genes Dev 1990;
multiprotein complex containing Ets and GATA factors. EMBO J
4(11):1886-1898.
2002; 21(12):3039-3050.
2. Merika M, Orkin SH. DNA-binding specificity of GATA family
25. Kelley C, Yee K, Harland R et al. Ventral expression of GATA-1
transcription factors. Mol Cell Biol 1993; 13(7):3999-4010.
and GATA-2 in the Xenopus embryo defines induction of
3. Lowry JA, Atchley WR. Molecular evolution of the GATA family
hematopoietic mesoderm. Dev Biol 1994; 165(1):193-205.
of transcription factors: conservation within the DNA-binding
26. Walmsley ME, Guille MJ, Bertwistle D et al. Negative control of
domain. J Mol Evol 2000; 50(2):103-115.
Xenopus GATA-2 by activin and noggin with eventual expression
4. Scazzocchio C. The fungal GATA factors. Curr Opin Microbiol
in precursors of the ventral blood islands. Development 1994;
2000; 3(2):126-131.
120(9):2519-2529.
5. Patient RK, McGhee JD. The GATA family (vertebrates and
27. Detrich HW, 3rd, Kieran MW, Chan FY et al. Intraembryonic
invertebrates). Curr Opin Genet Dev 2002; 12(4):416-422.
hematopoietic cell migration during vertebrate development. Proc
6. Tong Q, Dalgin G, Xu H et al. Function of GATA transcription
Natl Acad Sci USA 1995;92(23):10713-10717.
factors in preadipocyte-adipocyte transition. Science 2000;
28. Kitajima K, Masuhara M, Era T et al. GATA-2 and GATA-2/ER
290(5489):134-138.
display opposing activities in the development and differentiation of
7. Tevosian SG, Albrecht KH, Crispino JD et al. Gonadal
blood progenitors. EMBO J 2002; 21(12):3060-3069.
differentiation, sex determination and normal Sry expression in mice
29. Persons DA, Allay JA, Allay ER et al. Enforced expression of the
require direct interaction between transcription partners GATA4 and
GATA-2 transcription factor blocks normal hematopoiesis. Blood
FOG2. Development 2002;129(19):4627-4634.
1999; 93(2):488-499.
8. Xu RH, Kim J, Taira M et al. Differential regulation of neurogenesis
30. Kumano K, Chiba S, Shimizu K et al. Notch1 inhibits differentiation
by the two Xenopus GATA-1 genes. Mol Cell Biol 1997;
of hematopoietic cells by sustaining GATA-2 expression. Blood 2001;
17(1):436-443.
98(12):3283-3289.
228 Zinc Finger Proteins: From Atomic Contact to Cellular Function
31. Pevny L, Simon MC, Robertson E et al. Erythroid differentiation 51. McDevitt MA, Fujiwara Y, Shivdasani RA, Orkin SH. An upstream,
in chimaeric mice blocked by a targeted mutation in the gene for DNase I hypersensitive region of the hematopoietic-expressed
transcription factor GATA-1. Nature 1991; 349(6306):257-260. transcription factor GATA-1 gene confers developmental specificity
32. Fujiwara Y, Browne CP, Cunniff K et al. Arrested development of in transgenic mice. Proc Natl Acad Sci USA 1997;
embryonic red cell precursors in mouse embryos lacking transcription 94(15):7976-7981.
factor GATA-1. Proc Natl Acad Sci USA 1996; 93(22):12355-12358. 52. Kobayashi M, Nishikawa K, Yamamoto M. Hematopoietic regulatory
33. Takahashi S, Komeno T, Suwabe N et al. Role of GATA-1 in domain of gata1 gene is positively regulated by GATA1 protein in
proliferation and differentiation of definitive erythroid and zebrafish embryos. Development 2001; 128(12):2341-2350.
megakaryocytic cells in vivo. Blood 1998; 92(2):434-442. 53. Grass JA, Boyer ME, Pal S et al. GATA-1-dependent transcriptional
34. Lyons SE, Lawson ND, Lei L et al. A nonsense mutation in zebrafish repression of GATA-2 via disruption of positive autoregulation and
gata1 causes the bloodless phenotype in vlad tepes. Proc Natl Acad domain-wide chromatin remodeling. Proc Natl Acad Sci USA 2003;
Sci USA 2002; 99(8):5454-5459. 100(15):8811-8816.
35. Shivdasani RA, Fujiwara Y, McDevitt MA et al. A lineage-selective 54. Weiss MJ, Keller G, Orkin SH. Novel insights into erythroid
knockout establishes the critical role of transcription factor GATA-1 development revealed through in vitro differentiation of GATA-1
in megakaryocyte growth and platelet development. EMBO J 1997; embryonic stem cells. Genes Dev 1994; 8(10):1184-1197.
16(13):3965-3973. 55. Mead PE, Deconinck AE, Huber TL et al. Primitive erythropoiesis
36. Vyas P, Ault K, Jackson CW et al. Consequences of GATA-1 in the Xenopus embryo: the synergistic role of LMO-2, SCL and
deficiency in megakaryocytes and platelets. Blood 1999; GATA-binding proteins. Development 2001; 128(12):2301-2308.
93(9):2867-2875. 56. Wadman IA, Osada H, Grutz GG et al. The LIM-only protein Lmo2
37. Tsang AP, Visvader JE, Turner CA et al. FOG, a multitype zinc is a bridging molecule assembling an erythroid, DNA-binding
finger protein, acts as a cofactor for transcription factor GATA-1 in complex which includes the TAL1, E47, GATA-1 and Ldb1/NLI
erythroid and megakaryocytic differentiation. Cell 1997; proteins. EMBO J 1997; 16(11):3145-3157.
90(1):109-119. 57. Visvader JE, Elefanty AG, Strasser A et al. GATA-1 but not SCL
38. Haenlin M, Cubadda Y, Blondeau F et al. Transcriptional activity induces megakaryocytic differentiation in an early myeloid line.
of pannier is regulated negatively by heterodimerization of the GATA EMBO J 1992; 11(12):4557-4564.
DNA-binding domain with a cofactor encoded by the u-shaped gene 58. Zhang P, Behre G, Pan J et al. Negative cross-talk between
of Drosophila. Genes Dev 1997; 11(22):3096-3108. hematopoietic regulators: GATA proteins repress PU.1. Proc Natl
39. Cubadda Y, Heitzler P, Ray RP et al. u-shaped encodes a zinc finger Acad Sci USA 1999; 96(15):8705-8710.
protein that regulates the proneural genes achaete and scute during 59. Zhang P, Zhang X, Iwama A et al. PU.1 inhibits GATA-1 function
the formation of bristles in Drosophila. Genes Dev 1997; and erythroid differentiation by blocking GATA-1 DNA binding.
11(22):3083-3095. Blood 2000; 96(8):2641-2648.
40. Lu JR, McKinsey TA, Xu H et al. FOG-2, a heart- and 60. Rekhtman N, Radparvar F, Evans T et al. Direct interaction of
brain-enriched cofactor for GATA transcription factors. Mol Cell hematopoietic transcription factors PU.1 and GATA-1: functional
Biol 1999; 19(6):4495-4502. antagonism in erythroid cells. Genes Dev 1999; 13(11):1398-1411.
41. Svensson EC, Tufts RL, Polk CE et al. Molecular cloning of FOG-2: 61. Nerlov C, Querfurth E, Kulessa H et al. GATA-1 interacts with the
a modulator of transcription factor GATA-4 in cardiomyocytes. Proc myeloid PU.1 transcription factor and represses PU.1-dependent
Natl Acad Sci USA 1999; 96(3):956-961. transcription. Blood 2000; 95(8):2543-2551.
42. Tevosian SG, Deconinck AE, Cantor AB et al. FOG-2: A novel 62. Querfurth E, Schuster M, Kulessa H et al. Antagonism between C/
GATA-family cofactor related to multitype zinc-finger proteins Friend EBPbeta and FOG in eosinophil lineage commitment of multipotent
of GATA-1 and U-shaped. Proc Natl Acad Sci USA 1999; hematopoietic progenitors. Genes Dev 2000; 14(19):2515-2525.
96(3):950-955. 63. Kulessa H, Frampton J, Graf T. GATA-1 reprograms avian
43. Deconinck AE, Mead PE, Tevosian SG et al. FOG acts as a repressor myelomonocytic cell lines into eosinophils, thromboblasts, and
of red blood cell development in Xenopus. Development 2000; erythroblasts. Genes Dev 1995; 9(10):1250-1262.
127(10):2031-2040. 64. Cantor AB, Orkin SH. Transcriptional regulation of erythropoiesis:
44. Crispino JD, Lodish MB, MacKay JP et al. Use of altered specificity an affair involving multiple partners. Oncogene 2002;
mutants to probe a specific protein-protein interaction in 21(21):3368-3376.
differentiation: the GATA-1:FOG complex. Mol Cell Feb 1999; 65. Ho IC, Vorhees P, Marin N et al. Human GATA-3: a
3(2):219-228. lineage-restricted transcription factor that regulates the expression of
45. Chang AN, Cantor AB, Fujiwara Y et al. GATA-factor dependence the T cell receptor alpha gene. EMBO J 1991; 10(5):1187-1192.
of the multitype zinc-finger protein FOG-1 for its essential role in 66. George KM, Leonard MW, Roth ME et al. Embryonic expression
megakaryopoiesis. Proc Natl Acad Sci USA 2002; 99(14):9237-9242. and cloning of the murine GATA-3 gene. Development 1994;
46. Shimizu R, Takahashi S, Ohneda K et al. In vivo requirements for 120(9):2673-2686.
GATA-1 functional domains during primitive and definitive 67. Lieuw KH, Li G, Zhou Y et al. Temporal and spatial control of
erythropoiesis. EMBO J 2001; 20(18):5250-5260. murine GATA-3 transcription by promoter-proximal regulatory
47. Takahashi S, Shimizu R, Suwabe N et al. GATA factor transgenes elements. Dev Biol 1997; 188(1):1-16.
under GATA-1 locus control rescue germline GATA-1 mutant 68. Oosterwegel M, Timmerman J, Leiden J et al. Expression of GATA-3
deficiencies. Blood 2000;96(3):910-916. during lymphocyte differentiation and mouse embryogenesis. Dev
48. Vyas P, McDevitt MA, Cantor AB et al. Different sequence Immunol 1992; 3(1):1-11.
requirements for expression in erythroid and megakaryocytic cells 69. Pandolfi PP, Sonati F, Rivi R et al. Targeted disruption of the
within a regulatory element upstream of the GATA-1 gene. housekeeping gene encoding glucose 6-phosphate dehydrogenase
Development 1999; 126(12):2799-2811. (G6PD): G6PD is dispensable for pentose synthesis but essential
49. Onodera K, Takahashi S, Nishimura S et al. GATA-1 transcription for defense against oxidative stress. EMBO J 1995;
is controlled by distinct regulatory mechanisms during primitive and 14(21):5209-5215.
definitive erythropoiesis. Proc Natl Acad Sci USA 1997; 70. Ting CN, Olson MC, Barton KP et al. Transcription factor GATA-3
94(9):4487-4492. is required for development of the T-cell lineage. Nature 1996;
50. Nishimura S, Takahashi S, Kuroha T et al. A GATA box in the 384(6608):474-478.
GATA-1 gene hematopoietic enhancer is a critical element in the 71. Hendriks RW, Nawijn MC, Engel JD et al. Expression of the
network of GATA factors and sites that regulate this gene. Mol Cell transcription factor GATA-3 is required for the development of the
Biol 2000; 20(2):713-723. earliest T cell progenitors and correlates with stages of cellular
proliferation in the thymus. Eur J Immunol 1999; 29(6):1912-1918.
Role of GATA Factors in Development 229
72. Hernandez-Hoyos G, Anderson MK, Wang C et al. GATA-3 96. Schultheiss TM, Burch JB, Lassar AB. A role for bone morphogenetic
expression is controlled by TCR signals and regulates CD4/CD8 proteins in the induction of cardiac myogenesis. Genes Dev 1997;
differentiation. Immunity 2003; 19(1):83-94. 11(4):451-462.
73. Zheng W, Flavell RA. The transcription factor GATA-3 is necessary 97. Grepin C, Dagnino L, Robitaille L et al. A hormone-encoding gene
and sufficient for Th2 cytokine gene expression in CD4 T cells. identifies a pathway for cardiac but not skeletal muscle gene
Cell 1997; 89(4):587-596. transcription. Mol Cell Biol 1994; 14(5):3115-3129.
74. Zhang DH, Cohn L, Ray P et al. Transcription factor GATA-3 is 98. Ip HS, Wilson DB, Heikinheimo M et al. The GATA-4 transcription
differentially expressed in murine Th1 and Th2 cells and controls factor transactivates the cardiac muscle-specific troponin C
Th2-specific expression of the interleukin-5 gene. J Biol Chem 1997; promoter-enhancer in nonmuscle cells. Mol Cell Biol 1994;
272(34):21597-21603. 14(11):7517-7526.
75. Murphy KM, Reiner SL. The lineage decisions of helper T cells. 99. Lyons I, Parsons LM, Hartley L et al. Myogenic and morphogenetic
Nat Rev Immunol 2002; 2(12):933-944. defects in the heart tubes of murine embryos lacking the homeo
76. Lee GR, Fields PE, Flavell RA. Regulation of IL-4 gene expression box gene Nkx2-5. Genes Dev 1995; 9(13):1654-1666.
by distal regulatory elements and GATA-3 at the chromatin level. 100. Molkentin JD, Kalvakolanu DV, Markham BE. Transcription factor
Immunity 2001; 14(4):447-459. GATA-4 regulates cardiac muscle-specific expression of the
77. Ouyang W, Ranganath SH, Weindel K et al. Inhibition of Th1 alpha-myosin heavy-chain gene. Mol Cell Biol 1994;
development mediated by GATA-3 through an IL-4-independent 14(7):4947-4957.
mechanism. Immunity 1998; 9(5):745-755. 101. Harvey RP. NK-2 homeobox genes and heart development. Dev Biol
78. Usui T, Nishikomori R, Kitani A et al. GATA-3 suppresses Th1 1996; 178(2):203-216.
development by downregulation of Stat4 and not through effects on 102. Kuo CT, Morrisey EE, Anandappa R et al. GATA4 transcription
IL-12Rbeta2 chain or T-bet. Immunity 2003; 18(3):415-428. factor is required for ventral morphogenesis and heart tube formation.
79. Lee HJ, Takemoto N, Kurata H et al. GATA-3 induces T helper Genes Dev 1997; 11(8):1048-1060.
cell type 2 (Th2) cytokine expression and chromatin remodeling in 103. Molkentin JD, Lin Q, Duncan SA et al. Requirement of the
committed Th1 cells. J Exp Med 2000; 192(1):105-115. transcription factor GATA4 for heart tube formation and ventral
80. Evans CJ, Banerjee U. Transcriptional regulation of hematopoiesis morphogenesis. Genes Dev 1997; 11(8):1061-1072.
in Drosophila. Blood Cells Mol Dis 2003; 30(2):223-228. 104. Molkentin JD, Tymitz KM, Richardson JA et al. Abnormalities of
81. Fossett N, Schulz RA. Functional conservation of hematopoietic the genitourinary tract in female mice lacking GATA5. Mol Cell
factors in Drosophila and vertebrates. Differentiation 2001; Biol 2000; 20(14):5256-5260.
69(2-3):83-90. 105. Morrisey EE, Tang Z, Sigrist K et al. GATA6 regulates HNF4 and
82. Lebestky T, Chang T, Hartenstein V et al. Specification of Drosophila is required for differentiation of visceral endoderm in the mouse
hematopoietic lineage by conserved transcription factors. Science embryo. Genes Dev 1998; 12(22):3579-3590.
2000; 288(5463):146-149. 106. Koutsourakis M, Langeveld A, Patient R et al. The transcription
83. Tepass U, Fessler LI, Aziz A et al. Embryonic origin of hemocytes factor GATA6 is essential for early extraembryonic development.
and their relationship to cell death in Drosophila. Development 1994; Development 1999; 126(4):723-732.
120(7):1829-1837. 107. Reiter JF, Alexander J, Rodaway A et al. Gata5 is required for the
84. Sam S, Leise W, Hoshizaki DK. The serpent gene is necessary for development of the heart and endoderm in zebrafish. Genes Dev
progression through the early stages of fat-body development. Mech 1999; 13(22):2983-2995.
Dev 1996; 60(2):197-205. 108. Lee Y, Shioi T, Kasahara H et al. The cardiac tissue-restricted
85. Rehorn KP, Thelen H, Michelson AM et al. A molecular aspect of homeobox protein Csx/Nkx2.5 physically associates with the zinc
hematopoiesis and endoderm development common to vertebrates finger protein GATA4 and cooperatively activates atrial natriuretic
and Drosophila. Development 1996; 122(12):4023-4031. factor gene expression. Mol Cell Biol 1998; 18(6):3120-3129.
86. Waltzer L, Bataillé L, Peyrefitte S et al. Two isoforms of Serpent 109. Durocher D, Charron F, Warren R et al. The cardiac transcription
containing either one or two GATA zinc fingers have different roles factors Nkx2-5 and GATA-4 are mutual cofactors. EMBO J 1997;
in Drosophila haematopoiesis. EMBO J 2002; 21(20):5477-5486. 16(18):5687-5696.
87. Elagib KE, Racke FK, Mogass M et al. RUNX1 and GATA-1 110. Charron F, Nemer M. GATA transcription factors and cardiac
coexpression and cooperation in megakaryocytic differentiation. Blood development. Semin Cell Dev Biol 1999; 10(1):85-91.
2003; 101(11):4333-4341. 111. Maduro MF, Rothman JH. Making worm guts: the gene regulatory
88. Fossett N, Tevosian SG, Gajewski K et al. The Friend of GATA network of the Caenorhabditis elegans endoderm. Dev Biol 2002;
proteins U-shaped, FOG-1, and FOG-2 function as negative 246(1):68-85.
regulators of blood, heart, and eye development in Drosophila. Proc 112. Maduro MF, Meneghini MD, Bowerman B et al. Restriction of
Natl Acad Sci USA 2001; 98(13):7342-7347. mesendoderm to a single blastomere by the combined action of
89. Cripps RM, Olson EN. Control of cardiac development by an SKN-1 and a GSK-3beta homolog is mediated by MED-1 and -2
evolutionarily conserved transcriptional network. Dev Biol 2002; in C. elegans. Mol Cell 2001; 7(3):475-485.
246(1):14-28. 113. Zhu J, Hill RJ, Heid PJ et al. end-1 encodes an apparent GATA
90. Gajewski K, Fossett N, Molkentin JD et al. The zinc finger proteins factor that specifies the endoderm precursor in Caenorhabditis elegans
Pannier and GATA4 function as cardiogenic factors in Drosophila. embryos. Genes Dev 1997; 11(21):2883-2896.
Development 1999; 126(24):5679-5688. 114. Zhu J, Fukushige T, McGhee JD et al. Reprogramming of early
91. Azpiazu N, Frasch M. tinman and bagpipe: two homeo box genes embryonic blastomeres into endodermal progenitors by a
that determine cell fates in the dorsal mesoderm of Drosophila. Genes Caenorhabditis elegans GATA factor. Genes Dev 1998;
Dev 1993; 7(7B):1325-1340. 12(24):3809-3814.
92. Klinedinst SL, Bodmer R. Gata factor Pannier is required to establish 115. Fukushige T, Hawkins MG, McGhee JD. The GATA-factor elt-2 is
competence for heart progenitor formation. Development 2003; essential for formation of the Caenorhabditis elegans intestine. Dev
130(13):3027-3038. Biol 1998; 198(2):286-302.
93. Lilly B, Zhao B, Ranganayakulu G et al. Requirement of MADS 116. Fukushige T, Hendzel MJ, Bazett-Jones DP et al. Direct visualization
domain transcription factor D-MEF2 for muscle formation in of the elt-2 gut-specific GATA factor binding to a target promoter
Drosophila. Science 1995; 267(5198):688-693. inside the living Caenorhabditis elegans embryo. Proc Natl Acad Sci
94. Ranganayakulu G, Zhao B, Dokidis A et al. A series of mutations USA 1999; 96(21):11883-11888.
in the D-MEF2 transcription factor reveal multiple functions in larval 117. Britton C, McKerrow JH, Johnstone IL. Regulation of the
and adult myogenesis in Drosophila. Dev Biol 1995; 171(1):169-181. Caenorhabditis elegans gut cysteine protease gene cpr-1: requirement
95. Schultheiss TM, Xydas S, Lassar AB. Induction of avian cardiac for GATA motifs. J Mol Biol 1998; 283(1):15-27.
myogenesis by anterior endoderm. Development 1995;
121(12):4203-4214.
230 Zinc Finger Proteins: From Atomic Contact to Cellular Function
118. Hawkins MG, McGhee JD. elt-2, a second GATA factor from the 140. Kelley C, Blumberg H, Zon LI et al. GATA-4 is a novel transcription
nematode Caenorhabditis elegans. J Biol Chem 1995; factor expressed in endocardium of the developing heart.
270(24):14666-14671. Development 1993; 118(3):817-827.
119. Moilanen LH, Fukushige T, Freedman JH. Regulation of 141. Fujikura J, Yamato E, Yonemura S et al. Differentiation of embryonic
metallothionein gene transcription. Identification of upstream stem cells is induced by GATA factors. Genes Dev 2002;
regulatory elements and transcription factors responsible for 16(7):784-789.
cell-specific expression of the metallothionein genes from 142. Narita N, Bielinska M, Wilson DB. Wild-type endoderm abrogates
Caenorhabditis elegans. J Biol Chem 1999; 274(42):29655-29665. the ventral developmental defects associated with GATA-4 deficiency
120. Reuter R. The gene serpent has homeotic properties and specifies in the mouse. Dev Biol 1997; 189(2):270-274.
endoderm versus ectoderm within the Drosophila gut. Development 143. Soudais C, Bielinska M, Heikinheimo M et al. Targeted mutagenesis
1994; 120(5):1123-1135. of the transcription factor GATA-4 gene in mouse embryonic stem
121. Lin WH, Huang LH, Yeh JY et al. Expression of a Drosophila GATA cells disrupts visceral endoderm differentiation in vitro. Development
transcription factor in multiple tissues in the developing embryos. 1995; 121(11):3877-3888.
Identification of homozygous lethal mutants with P-element insertion 144. Jacobsen CM, Narita N, Bielinska M et al. Genetic mosaic analysis
at the promoter region. J Biol Chem 1995; 270(42):25150-25158. reveals that GATA-4 is required for proper differentiation of mouse
122. Davidson EH, Rast JP, Oliveri P et al. A genomic regulatory network gastric epithelium. Dev Biol 2002; 241(1):34-46.
for development. Science 2002; 295(5560):1669-1678. 145. Keijzer R, van Tuyl M, Meijers C et al. The transcription factor
123. Arceci RJ, King AA, Simon MC et al. Mouse GATA-4: a retinoic GATA6 is essential for branching morphogenesis and epithelial cell
acid-inducible GATA-binding transcription factor expressed in differentiation during fetal pulmonary development. Development
endodermally derived tissues and heart. Mol Cell Biol 1993; 2001; 128(4):503-511.
13(4):2235-2246. 146. Yang H, Lu MM, Zhang L et al. GATA6 regulates differentiation
124. Laverriere AC, MacNeill C, Mueller C et al. GATA-4/5/6, a of distal lung epithelium. Development 2002; 129(9):2233-2246.
subfamily of three transcription factors transcribed in developing heart 147. Reiter JF, Kikuchi Y, Stainier DY. Multiple roles for Gata5 in
and gut. J Biol Chem 1994; 269(37):23177-23184. zebrafish endoderm formation. Development 2001; 128(1):125-135.
125. Morrisey EE, Ip HS, Lu MM et al. GATA-6: a zinc finger 148. Rodaway A, Takeda H, Koshida S et al. Induction of the
transcription factor that is expressed in multiple cell lineages derived mesendoderm in the zebrafish germ ring by yolk cell-derived
from lateral mesoderm. Dev Biol 1996; 177(1):309-322. TGF-beta family signals and discrimination of mesoderm and
126. Morrisey EE, Ip HS, Tang Z et al. GATA-5: a transcriptional endoderm by FGF. Development 1999; 126(14):3067-3078.
activator expressed in a novel temporally and spatially-restricted 149. Stainier DY. A glimpse into the molecular entrails of endoderm
pattern during embryonic development. Dev Biol 1997; formation. Genes Dev 2002; 16(8):893-907.
183(1):21-36. 150. Angelotti T, Krishna G, Scott J et al. Nodular invasive
127. Jiang Y, Evans T. The Xenopus GATA-4/5/6 genes are associated tracheobronchitis due to Aspergillus in a patient with systemic lupus
with cardiac specification and can regulate cardiac-specific erythematosus. Lupus 2002; 11(5):325-328.
transcription during embryogenesis. Dev Biol 1996; 174(2):258-270. 151. Moi P, Loudianos G, Lavinha J et al. Delta-thalassemia due to a
128. Gove C, Walmsley M, Nijjar S et al. Over-expression of GATA-6 in mutation in an erythroid-specific binding protein sequence 3' to the
Xenopus embryos blocks differentiation of heart precursors. EMBO delta-globin gene. Blood 1992; 79(2):512-516.
J 1997; 16(2):355-368. 152. Matsuda M, Sakamoto N, Fukumaki Y. Delta-thalassemia caused
129. Huggon IC, Davies A, Gove C et al. Molecular cloning of human by disruption of the site for an erythroid-specific transcription factor,
GATA-6 DNA binding protein: high levels of expression in heart GATA-1, in the delta-globin gene promoter. Blood 1992;
and gut. Biochim Biophys Acta 1997; 1353(2):98-102. 80(5):1347-1351.
130. Suzuki E, Evans T, Lowry J et al. The human GATA-6 gene: 153. Ludlow LB, Schick BP, Budarf ML et al. Identification of a mutation
structure, chromosomal location, and regulation of expression by in a GATA binding site of the platelet glycoprotein Ibbeta promoter
tissue-specific and mitogen-responsive signals. Genomics 1996; resulting in the Bernard-Soulier syndrome. J Biol Chem 1996;
38(3):283-290. 271(36):22076-22080.
131. Tamura S, Wang XH, Maeda M et al. Gastric DNA-binding proteins 154. Manco L, Ribeiro ML, Maximo V et al. A new PKLR gene mutation
recognize upstream sequence motifs of parietal cell-specific genes. in the R-type promoter region affects the gene transcription causing
Proc Natl Acad Sci USA 1993; 90(22):10876-10880. pyruvate kinase deficiency. Br J Haematol 2000; 110(4):993-997.
132. Liu C, Glasser SW, Wan H et al. GATA-6 and thyroid transcription 155. Solis C, Aizencang GI, Astrin KH et al. Uroporphyrinogen III
factor-1 directly interact and regulate surfactant protein-C gene synthase erythroid promoter mutations in adjacent GATA1 and CP2
expression. J Biol Chem 2002; 277(6):4519-4525. elements cause congenital erythropoietic porphyria. J Clin Invest
133. Shaw-White JR, Bruno MD, Whitsett JA. GATA-6 activates 2001; 107(6):753-762.
transcription of thyroid transcription factor-1. J Biol Chem 1999; 156. Tournamille C, Colin Y, Cartron JP et al. Disruption of a GATA
274(5):2658-2664. motif in the Duffy gene promoter abolishes erythroid gene expression
134. Bruno MD, Korfhagen TR, Liu C et al. GATA-6 activates in Duffy-negative individuals. Nat Genet 1995; 10(2):224-228.
transcription of surfactant protein A. J Biol Chem 2000; 157. Parasol N, Reid M, Rios M et al. A novel mutation in the coding
275(2):1043-1049. sequence of the FY*B allele of the Duffy chemokine receptor gene
135. Gao X, Sedgwick T, Shi YB et al. Distinct functions are implicated is associated with an altered erythrocyte phenotype. Blood 1998;
for the GATA-4, -5, and -6 transcription factors in the regulation 92(7):2237-2243.
of intestine epithelial cell differentiation. Mol Cell Biol 1998; 158. Pehlivan T, Pober BR, Brueckner M et al. GATA4 haploinsufficiency
18(5):2901-2911. in patients with interstitial deletion of chromosome region 8p23.1
136. Maeda M, Kubo K, Nishi T et al. Roles of gastric GATA and congenital heart disease. Am J Med Genet 1999; 83(3):201-206.
DNA-binding proteins. J Exp Biol 1996; 199 (Pt 3):513-520. 159. Garg V, Kathiriya IS, Barnes R et al. GATA4 mutations cause human
137. Cirillo LA, Lin FR, Cuesta I et al. Opening of compacted chromatin congenital heart defects and reveal an interaction with TBX5. Nature
by early developmental transcription factors HNF3 (FoxA) and 2003; 424(6947):443-447.
GATA-4. Mol Cell 2002; 9(2):279-289. 160. Mehaffey MG, Newton AL, Gandhi MJ et al. X-linked
138. Bossard P, Zaret KS. GATA transcription factors as potentiators of thrombocytopenia caused by a novel mutation of GATA-1. Blood
gut endoderm differentiation. Development 1998; 2001; 98(9):2681-2688.
125(24):4909-4917. 161. Nichols KE, Crispino JD, Poncz M et al. Familial dyserythropoietic
139. Weber H, Symes CE, Walmsley ME et al. A role for GATA5 in anaemia and thrombocytopenia due to an inherited mutation in
Xenopus endoderm specification. Development 2000; GATA1. Nat Genet 2000; 24(3):266-270.
127(20):4345-4360.
Role of GATA Factors in Development 231
162. Freson K, Devriendt K, Matthijs G et al. Platelet characteristics in 167. Ludecke HJ, Schaper J, Meinecke P et al. Genotypic and phenotypic
patients with X-linked macrothrombocytopenia because of a novel spectrum in tricho-rhino-phalangeal syndrome types I and III. Am
GATA1 mutation. Blood 2001; 98(1):85-92. J Hum Genet 2001; 68(1):81-91.
163. Freson K, Matthijs G, Thys C et al. Different substitutions at residue 168. Malik TH, Von Stechow D, Bronson RT et al. Deletion of the GATA
D218 of the X-linked transcription factor GATA1 lead to altered domain of TRPS1 causes an absence of facial hair and provides new
clinical severity of macrothrombocytopenia and anemia and are insights into the bone disorder in inherited tricho-rhino-phalangeal
associated with variable skewed X inactivation. Hum Mol Genet syndromes. Mol Cell Biol 2002; 22(24):8592-8600.
2002; 11(2):147-152. 169. Wieser R, Volz A, Vinatzer U et al. Transcription factor GATA-2
164. Yu C, Niakan KK, Matsushita M et al. X-linked thrombocytopenia gene is located near 3q21 breakpoints in myeloid leukemia. Biochem
with thalassemia from a mutation in the amino finger of GATA-1 Biophys Res Commun. 2000; 273(1):239-245.
affecting DNA binding rather than FOG-1 interaction. Blood 170. Wechsler J, Greene M, McDevitt M et al. Acquired mutations in
2002;100(6):2040-2045. GATA1 in the megakaryoblastic leukemia of Down syndrome. Nat
165. Van Esch H, Devriendt K. Transcription factor GATA3 and the Genet 2002; 32(1):148-152.
human HDR syndrome. Cell Mol Life Sci 2001; 58(9):1296-1300. 171. Groet J, McElwaine S, Spinelli M et al. Acquired mutations in
166. Momeni P, Glockner G, Schmidt O et al. Mutations in a new gene, GATA1 in neonates with Down’s syndrome with transient myeloid
encoding a zinc-finger protein, cause tricho-rhino-phalangeal disorder. Lancet 2003; 361(9369):1617-1620.
syndrome type I. Nat Genet 2000; 24(1):71-74.
CHAPTER 31
T
he androgen receptor belongs to the superfamily of nuclear
receptors, which are ligand-dependent transcription factors. where it is associated with heat shock proteins (HSPs), including
In the absence of the androgen, the receptor is localized HSP90, HSP70 and HSP56.8-10 Upon testosterone or DHT bind-
to the cytoplasm where it is associated with heat shock proteins. ing, the AR undergoes several sequential processes to interact with
Upon ligand binding, the receptor translocates into the nucleus specific DNA sequences, called androgen response elements
and interacts with specific DNA sequences, called androgen re- (AREs), located in the regulatory regions of androgen-dependent
sponse elements. The DNA-bound receptor interacts with the genes.
transcription initiation complex to regulate transcription. Regulation of steroid hormone receptor action occurs, in part,
The structural organization of the androgen receptor is very by post-translational modification, including phosphorylation,
similar to the other members of the steroid receptor family, with acetylation, and sumoylation (addition of the small
an N-terminal transcriptional regulatory domain, a centrally po- ubiquitin-related modifier SUMO).11 AR is rapidly phosphory-
sitioned C2C2 zinc finger DNA binding domain, and a C-terminal lated after synthesis, suggesting that this modification is impor-
tant for the hormone binding properties of the receptor.12 AR
ligand binding domain. The N-terminal domain contains a poly-
undergoes further phosphorylation in the presence of ligand.13
morphic polyglutamine tract encoded by a trinucleotide (CAG)
AR is acetylated in a ligand-dependent manner, and acetylation
repeat; the polyglutamine tract normally consists of 9-36 increases its transcriptional activity, implicating this
glutamines. Pathological expansion of the androgen receptor post-translational modification as a regulatory mechanism.14
polyglutamine tract to 40-62 glutamines causes spinal and bul- Aberrant acetylation and phosphorylation state lead to
bar muscular atrophy, a slowly progressive, X-linked motor neu- ligand-independent activation of mutant AR.15 AR is covalently
ron disease. modified by SUMO-1 in an androgen-enhanced fashion, and
sumoylation has been shown to correlate with downregulation of
Androgen Receptor transcriptional activity.16
The androgen receptor (AR) belongs to the superfamily of Binding of the inactive AR, complexed with HSP, by ligand
nuclear receptors, which are ligand dependent transcription fac- induces transformation of the receptor: the receptor changes to a
tors that specifically regulate the expression of target genes in- more compact conformation, a second conformation change likely
volved in metabolism, development and reproduction. Andro- results in dissociation of heat shock proteins and precedes nuclear
gen receptor mediates the transcriptional response to androgens uptake and the formation of a transcriptionally active complex;
such as testosterone and 5α-dihydrotestosterone (DHT) (for re- the receptor then dimerizes and binds to DNA. The DNA-bound
receptor interacts with the transcription initiation complex to
view see Zhou et al1). The AR has a higher affinity for DHT than
regulate transcription (Fig. 1).17 AR can interact with the basal
testosterone.2,3 Testosterone dissociates from AR three times faster transcription machinery by direct interaction with general tran-
than DHT, and degradation of AR is faster in the presence of scription factors7,18 or indirectly through coactivators such as ste-
testosterone compared to DHT, indicating that ligand occupancy roid receptor coactivator-1 (SRC-1) and CBP, which are proteins
of the receptor is a major factor in receptor stabilization.4 It is that interact with nuclear receptors and enhance their
possible that androgens slow AR degradation by prolonging transactivation.19 Transcriptional activation by AR requires the
nuclear retention and limiting recycling from the nucleus to the recruitment of RNA polymerase II. Transcription by RNA poly-
cytoplasm, preventing proteolysis of AR.1 Although testosterone merase II is a complex processes which requires assembly of a set
and DHT bind the same receptor, they may have different effects of basal transcription factors (TF) to initiate transcription at the
in sexual differentiation as a result of activation of different target promoters. First, TATA binding protein (TBP) (a component of
genes due to differences in activity.5-7 the transcription factor TFIID) recognizes the TATA element of
Zinc Finger Proteins: From Atomic Contact to Cellular Function, edited by Shiro Iuchi and Natalie Kuldell.
©2005 Landes Bioscience/Eurekah.com and Kluwer Academic/Plenum Publishers.
The Androgen Receptor and Spinal and Bulbar Muscular Atrophy 233
Figure 2. Structure of the AR gene and protein. Top) The gene contains
eight exons. Bottom) The protein structure shows how the exon orga-
nization translates into discrete functional regions of the receptor: the
N-terminal transactivation region, the DNA binding domain (DBD)
and the ligand binding domain (LBD).
normal. The mildness of symptoms and signs in these homozy- mechanisms of neuronal degeneration characteristic of each of
gous females suggests that females do not develop the full mani- the polyglutamine diseases. In the case of SBMA, recent work in
festations of SBMA because they have reduced levels of the AR transgenic fly85 and mouse models,86 has demonstrated that
ligand, testosterone, rather than because of X-chromosome inac- full-length AR requires the presence of ligand in order to cause
tivation. toxicity. Depletion of the ligand in male mice by castration pre-
Confirmation of SBMA diagnosis can be made with genetic vents motor neuron degeneration and administration of andro-
testing. CAG repeat number is determined by polymerase chain gens to female mice induces motor neuronal toxicity.86 This
reaction amplification of the CAG repeat region within the AR ligand-dependency explains the gender specificity of SBMA.
gene (chromosomal locus Xq11-q12).75 Accurate diagnosis is
important in order to provide information about prognosis and Inclusions and Aggregates
genetic counseling for the patient and his family members as well A common pathologic feature of the polyglutamine diseases is
as to avoid inappropriate treatment. Most patients with SBMA the presence of inclusion bodies in susceptible neurons. 87 In
have a normal lifespan and do not die of direct consequences of SBMA, inclusions have been identified in spinal cord motor neu-
their disease, although late in the course of illness some patients rons from patients88 and in animal and cell culture models. These
are at risk for aspiration pneumonia or respiratory failure. Cur- inclusions contain N-terminal epitopes of the AR protein that
rently, treatment is supportive. may arise from cleavage of mutant AR by caspases.89 The inclu-
sions also contain ubiquitin, components of the proteasome com-
Molecular Pathophysiology of SBMA plex, and several molecular chaperones, all of which are cellular
SBMA is caused by expansion of the trinucleotide CAG re- components involved in normal protein folding and degradation.
peat in the AR gene from a normal length of 9-36 to a Their presence in inclusions suggests that there has been accu-
disease-associated length of 40-62.59 SBMA is one of nine disor- mulation of nondegraded mutant AR protein within the cell.
ders that share this type of mutation. These disorders, known as Expanded polyglutamine tracts have been shown to cause pro-
the polyglutamine expansion diseases, include Huntington’s dis- tein misfolding and aggregation in a repeat length dependent
ease (HD), dentatorubropallidoluysian atrophy (DRPLA), and manner.90,91 It is postulated that aggregates result from the abil-
six forms of spinocerebellar ataxia (SCA).76 They are caused by ity of long polyglutamine stretches to form self-associating
CAG expansions in unrelated genes and are characterized by se- β-sheets.92,93 Aggregation of mutant AR protein into inclusions
lective neuronal degeneration in different nervous system regions. may be a central toxic event in polyglutamine-mediated
Except for SBMA, they all exhibit autosomal dominant inherit- neurodegeneration, however in some models, toxicity has been
ance with a similar age of onset and rate of progression. All occur dissociated from the presence of microscopically visible inclusions.
when the CAG tract achieves a critical threshold length, usually Recent work suggests that microscopically visible inclusions rep-
of approximately 40 repeats. Above this threshold, further in- resent structures formed by the cell in order to accelerate degra-
creased repeat length correlates with earlier disease onset and in- dation of excess misfolded protein, while monomeric or
creased disease severity.77-79 These shared features suggest that there micro-aggregated intermediates of polyglutamine protein are the
is a common pathogenic mechanism underlying these disorders. toxic species.94
References
1. Zhou ZX, Wong CI, Sar M et al. The androgen receptor: An
overview. Recent Prog Horm Res 1994; 49:249-274.
2. Wilson EM, French FS. Binding properties of androgen receptors.
Evidence for identical receptors in rat testis, epididymis, and prostate.
J Biol Chem 1976; 251:5620-5629.
3. Wilbert DM, Griffin JE, Wilson JD. Characterization of the cytosol
androgen receptor of the human prostate. J Clin Endocrinol Metab
1983; 56:113-120.
4. Zhou ZX, Lane MV, Kemppainen JA et al. Specificity of
ligand-dependent androgen receptor stabilization: Receptor domain
interactions influence ligand dissociation and receptor stability. Mol
Endocrinol 1995; 9:208-218.
5. Brinkmann AO. Molecular basis of androgen insensitivity. Mol Cell
Endocrinol 2001; 179:105-109.
Figure 3. Mutant AR containing a polyglutamine expansion (QQQQ) 6. Brinkmann AO, Trapman J. Androgen receptor mutants that affect
may cause neuronal degeneration through one or more mechanisms. normal growth and development. Cancer Surv 1992; 14:95-111.
After proteolytic cleavage (in the nucleus or cytoplasm), the polyglutamine 7. MacLean HE, Warne GL, Zajac JD. Localization of functional
tract assumes an altered conformation, resulting in aggregation and the domains in the androgen receptor. J Steroid Biochem Mol Biol 1997;
formation of inclusions. This abnormally folded protein may cause 62:233-242.
proteasome inhibition, mitochondrial toxicity, altered transport, or tran- 8. Pratt WB, Welsh MJ. Chaperone functions of the heat shock proteins
associated with steroid receptors. Semin Cell Biol 1994; 5:83-93.
scriptional dysregulation by sequestration of transcription factors. 9. Smith DF, Toft DO. Steroid receptors and their associated proteins.
DBD=DNA binding domain, NLS=nuclear localization signal, Mol Endocrinol 1993; 7:4-11.
LBD=ligand binding domain. (Adapted from Ross CA.)108 10. Veldscholte J, Berrevoets CA, Zegers ND et al. Hormone-induced
dissociation of the androgen receptor-heat-shock protein complex:
acetylation state of histones is an important determinant of tran- Use of a new monoclonal antibody to distinguish transformed from
nontransformed receptors. Biochemistry 1992; 31:7422-7430.
scriptionally active regions of chromatin; hyper-acetylation of
11. Moudgil VK. Phosphorylation of steroid hormone receptors. Biochim
histones marks transcriptionally active regions and Biophys Acta 1990; 1055:243-258.
hypo-acetylation marks transcriptionally silent regions. Seques- 12. Brinkmann AO, Blok LJ, de Ruiter PE et al. Mechanisms of
tration of CBP by mutant polyglutamine could have effects on androgen receptor activation and function. J Steroid Biochem Mol
gene expression that result in cellular toxicity. In a fly model, Biol 1999; 69:307-313.
polyglutamine-induced neurodegeneration has been shown to be 13. Jenster G, de Ruiter PE, van der Korput HA et al. Changes in the
associated with a histone acetylation defect and an alteration in abundance of androgen receptor isotypes: Effects of ligand treatment,
glutamine-stretch variation, and mutation of putative phosphorylation
the transcriptional profile. These abnormalities can be reversed sites. Biochemistry 1994; 33:14064-14072.
with over-expression of CBP.94 Restoration of the balance of his- 14. Fu M, Wang C, Reutens AT et al. p300 and p300/cAMP-response
tone acetylation might also be achieved by administering inhibi- element-binding protein-associated factor acetylate the androgen
tors of the histone deacetylase enzymes. Indeed, in cell, fly, and receptor at sites governing hormone-dependent transactivation. J Biol
mouse models of polyglutamine disease, histone deacetylase in- Chem 2000; 275:20853-20860.
hibitors have been shown to reverse the histone acetylation defect 15. Lieberman AP, Harmison G, Strand AD et al. Altered transcriptional
regulation in cells expressing the expanded polyglutamine androgen
and reduce polyglutamine–induced cell death.102-104 receptor. Hum Mol Genet 2002; 11:1967-1976.
With recent advances in the understanding of the basic patho- 16. Poukka H, Karvonen U, Janne OA et al. Covalent modification of
physiologic mechanisms of SBMA, potential therapeutic strate- the androgen receptor by small ubiquitin-like modifier 1 (SUMO-1).
gies can now be envisaged. One strategy might be to deplete the Proc Natl Acad Sci USA 2000; 97:14145-14150.
AR ligand, testosterone, thereby preventing mutant AR translo- 17. Kuil CW, Berrevoets CA, Mulder E. Ligand-induced conformational
cation to the nucleus. It has been recently shown that leuprolide, alterations of the androgen receptor analyzed by limited
trypsinization. Studies on the mechanism of antiandrogen action. J
a lutenizing hormone-releasing hormone agonist that reduces tes-
Biol Chem 1995; 270:27569-27576.
tosterone release from the testis, rescues the behavioral and patho- 18. Lee DK, Duan HO, Chang C. From androgen receptor to the
logic phenotypes in a transgenic mouse model of SBMA.105 Plans general transcription factor TFIIH. Identification of cdk activating
are underway to evaluate the efficacy of this compound in pa- kinase (CAK) as an androgen receptor NH(2)-terminal associated
tients. Another strategy might be to up-regulate the expression or coactivator. J Biol Chem 2000; 275:9308-9313.
enhance the function of molecular chaperones. This might pro- 19. McKenna NJ, Lanz RB, O’Malley BW. Nuclear receptor coregulators:
mote mutant AR refolding and/or degradation. A third strategy Cellular and molecular biology. Endocr Rev 1999; 20:321-344.
20. Nikolov DB, Burley SK. RNA polymerase II transcription initiation:
might be to restore the normal transcriptional state within cells. A structural view. Proc Natl Acad Sci USA 1997; 94:15-22.
If polyglutamine containing proteins alter histone acetylation levels 21. Chang CS, Kokontis J, Liao ST. Molecular cloning of human and
by sequestering histone acetyltransferases, normal histone acety- rat complementary DNA encoding androgen receptors. Science 1988;
lation might be restored with administration of histone deacetylase 240:324-326.
inhibitors. Screening of chemical compound libraries has been 22. Lubahn DB, Joseph DR, Sullivan PM et al. Cloning of human
done recently to identify agents that mitigate toxicity of mutant androgen receptor complementary DNA and localization to the X
chromosome. Science 1988; 240:327-330.
The Androgen Receptor and Spinal and Bulbar Muscular Atrophy 237
23. Tilley WD, Marcelli M, Wilson JD et al. Characterization and 47. Ikonen T, Palvimo JJ, Janne OA. Interaction between the amino-
expression of a cDNA encoding the human androgen receptor. Proc and carboxyl-terminal regions of the rat androgen receptor modulates
Natl Acad Sci USA 1989; 86:327-331. transcriptional activity and is influenced by nuclear receptor
24. Matias PM, Donner P, Coelho R et al. Structural evidence for ligand coactivators. J Biol Chem 1997; 272:29821-29828.
specificity in the binding domain of the human androgen receptor. 48. He B, Kemppainen JA, Voegel JJ et al. Activation function 2 in the
Implications for pathogenic gene mutations. J Biol Chem 2000; human androgen receptor ligand binding domain mediates
275:26164-26171. interdomain communication with the NH(2)-terminal domain. J Biol
25. Sack JS, Kish KF, Wang C et al. Crystallographic structures of the Chem 1999; 274:37219-37225.
ligand-binding domains of the androgen receptor and its T877A 49. He B, Wilson EM. The NH(2)-terminal and carboxyl-terminal
mutant complexed with the natural agonist dihydrotestosterone. Proc interaction in the human androgen receptor. Mol Genet Metab 2002;
Natl Acad Sci USA 2001; 98:4904-4909. 75:293-298.
26. Moras D, Gronemeyer H. The nuclear receptor ligand-binding 50. Gao T, Marcelli M, McPhaul MJ. Transcriptional activation and
domain: Structure and function. Curr Opin Cell Biol 1998; transient expression of the human androgen receptor. J Steroid
10:384-391. Biochem Mol Biol 1996; 59:9-20.
27. Pratt WB, Toft DO. Steroid receptor interactions with heat shock 51. Liu YZ, Chrivia JC, Latchman DS. Nerve growth factor up-regulates
protein and immunophilin chaperones. Endocr Rev 1997; the transcriptional activity of CBP through activation of the p42/
18:306-360. p44(MAPK) cascade. J Biol Chem 1998; 273:32400-32407.
28. Reynolds PD, Ruan Y, Smith DF et al. Glucocorticoid resistance in 52. Tanese N, Saluja D, Vassallo MF et al. Molecular cloning and analysis
the squirrel monkey is associated with overexpression of the of two subunits of the human TFIID complex: hTAFII130 and
immunophilin FKBP51. J Clin Endocrinol Metab 1999; 84:663-669. hTAFII100. Proc Natl Acad Sci USA 1996; 93:13611-13616.
29. Smith DF. Chaperones in progesterone receptor complexes. Semin 53. Chamberlain NL, Driver ED, Miesfeld RL. The length and location
Cell Dev Biol 2000; 11:45-52. of CAG trinucleotide repeats in the androgen receptor N-terminal
30. Jenster G, van der Korput HA, van Vroonhoven C et al. Domains domain affect transactivation function. Nucleic Acids Res 1994;
of the human androgen receptor involved in steroid binding, 22:3181-3186.
transcriptional activation, and subcellular localization. Mol Endocrinol 54. Kazemi-Esfarjani P, Trifiro MA, Pinsky L. Evidence for a repressive
1991; 5:1396-1404. function of the long polyglutamine tract in the human androgen
31. Evans RM. The steroid and thyroid hormone receptor superfamily. receptor: Possible pathogenetic relevance for the (CAG)n-expanded
Science 1988; 240:889-895. neuronopathies. Hum Mol Genet 1995; 4:523-527.
32. Claessens F, Verrijdt G, Schoenmakers E et al. Selective DNA binding 55. Tut TG, Ghadessy FJ, Trifiro MA et al. Long polyglutamine tracts
by the androgen receptor as a mechanism for hormone-specific gene in the androgen receptor are associated with reduced trans-activation,
regulation. J Steroid Biochem Mol Biol 2001; 76:23-30. impaired sperm production, and male infertility. J Clin Endocrinol
33. Kasper S, Rennie PS, Bruchovsky N et al. Cooperative binding of Metab 1997; 82:3777-3782.
androgen receptors to two DNA sequences is required for androgen 56. Schoenberg MP, Hakimi JM, Wang SP et al. Microsatellite Mutation
induction of the probasin gene. J Biol Chem 1994; 269:31763-769. (Cag24->18) in the Androgen Receptor Gene in Human Prostate
34. Schoenmakers E, Alen P, Verrijdt G et al. Differential DNA binding Cancer. Biochemical and Biophysical Research Communications
by the androgen and glucocorticoid receptors involves the second 1994; 198:74-80.
Zn-finger and a C-terminal extension of the DNA-binding domains. 57. Hsing AW, Gao Y-T, Wu G et al. Polymorphic CAG and GGN
Biochem J 1999; 341(Pt 3):515-521. Repeat Lengths in the Androgen Receptor Gene and Prostate Cancer
35. Beato M. Gene regulation by steroid hormones. Cell 1989; Risk: A Population-based Case-Control Study in China. Cancer Res
56:335-344. 2000; 60:5111-5116.
36. Umesono K, Evans RM. Determinants of target gene specificity for 58. Giovannucci E, Stampfer MJ, Krithivas K et al. The CAG repeat
steroid/thyroid hormone receptors. Cell 1989; 57:1139-1146. within the androgen receptor gene and its relationship to
37. Zoppi S, Marcelli M, Deslypere JP et al. Amino acid substitutions prostate cancer. Proc Natl Acad Sci USA 1997; 94:3320-3323.
in the DNA-binding domain of the human androgen receptor are a 59. La Spada AR, Wilson EM, Lubahn DB et al. Androgen receptor
frequent cause of receptor-binding positive androgen resistance. Mol gene mutations in X-linked spinal and bulbar muscular atrophy.
Endocrinol 1992; 6:409-415. Nature 1991; 352:77-79.
38. Quigley CA, De Bellis A, Marschke KB et al. Androgen receptor 60. Kennedy WR, Alter M, Sung JH. Progressive proximal spinal and
defects: Historical, clinical, and molecular perspectives. Endocr Rev bulbar muscular atrophy of late onset. A sex-linked recessive trait.
1995; 16:271-321. Neurology 1968; 18:671-680.
39. Lumbroso S, Lobaccaro JM, Belon C et al. A new mutation within 61. Harding AE, Thomas PK, Baraitser M et al. X-linked recessive
the deoxyribonucleic acid-binding domain of the androgen receptor bulbospinal neuronopathy: A report of ten cases. J Neurol Neurosurg
gene in a family with complete androgen insensitivity syndrome. Psychiatry 1982; 45:1012-1019.
Fertil Steril 1993; 60:814-819. 62. Hausmanowa-Petrusewicz I, Borkowska J, Janczewski Z. X-linked
40. Tsai MJ, O’Malley BW. Molecular mechanisms of action of steroid/ adult form of spinal muscular atrophy. J Neurol 1983; 229:175-188.
thyroid receptor superfamily members. Annu Rev Biochem 1994; 63. Sobue G, Hashizume Y, Mukai E et al. X-linked recessive bulbospinal
63:451-486. neuronopathy. A clinicopathological study. Brain 1989; 112:209-232.
41. Kallio PJ, Palvimo JJ, Mehto M et al. Analysis of androgen 64. Olney RK, Aminoff MJ, So YT. Clinical and electrodiagnostic
receptor-DNA interactions with receptor proteins produced in insect features of X-linked recessive bulbospinal neuronopathy. Neurology
cells. J Biol Chem 1994; 269:11514-11522. 1991; 41:823-828.
42. Jenster G, Trapman J, Brinkmann AO. Nuclear import of the human 65. Suzuki T, Endo K, Igarashi S et al. Isolated bilateral masseter atrophy
androgen receptor. Biochem J 1993; 293(Pt 3):761-768. in X-linked recessive bulbospinal neuronopathy. Neurology 1997;
43. Simental JA, Sar M, Lane MV et al. Transcriptional activation and 48:539-540.
nuclear targeting signals of the human androgen receptor. J Biol 66. Sumner CJ, Fischbeck KH. Jaw drop in Kennedy’s disease. Neurology
Chem 1991; 266:510-518. 2002; 59:1471-1472.
44. Picard D, Yamamoto KR. Two signals mediate hormone-dependent 67. Antonini G, Gragnani F, Romaniello A et al. Sensory involvement
nuclear localization of the glucocorticoid receptor. EMBO J 1987; in spinal-bulbar muscular atrophy (Kennedy’s disease). Muscle Nerve
6:3333-3340. 2000; 23:252-258.
45. Jenster G, van der Korput HA, Trapman J et al. Identification of 68. Polo A, Teatini F, D’Anna S et al. Sensory involvement in X-linked
two transcription activation units in the N-terminal domain of the spino-bulbar muscular atrophy (Kennedy’s syndrome): An
human androgen receptor. J Biol Chem 1995; 270:7341-7346. electrophysiological study. J Neurol 1996; 243:388-392.
46. McEwan IJ, Gustafsson J. Interaction of the human androgen 69. Arbizu T, Santamaria J, Gomez JM et al. A family with adult spinal
receptor transactivation function with the general transcription factor and bulbar muscular atrophy, X-linked inheritance and associated
TFIIF. Proc Natl Acad Sci USA1997; 94:8485-8490. testicular failure. J Neurol Sci 1983; 59:371-382.
238 Zinc Finger Proteins: From Atomic Contact to Cellular Function
70. Nagashima T, Seko K, Hirose K et al. Familial bulbo-spinal muscular 90. Scherzinger E, Lurz R, Turmaine M et al. Huntingtin-encoded
atrophy associated with testicular atrophy and sensory neuropathy polyglutamine expansions form amyloid-like protein aggregates in
(Kennedy-Alter-Sung syndrome). Autopsy case report of two brothers. vitro and in vivo. Cell 1997; 90:549-558.
J Neurol Sci 1988; 87:141-152. 91. Scherzinger E, Sittler A, Schweiger K et al. Self-assembly of
71. Warner CL, Griffin JE, Wilson JD et al. X-linked spinomuscular polyglutamine-containing huntingtin fragments into amyloid-like
atrophy: A kindred with associated abnormal androgen receptor fibrils: Implications for Huntington’s disease pathology. Proc Natl
binding. Neurology 1992; 42:2181-2184. Acad Sci USA 1999; 96:4604-4609.
72. Sobue G, Doyu M, Kachi T et al. Subclinical phenotypic expressions 92. Perutz MF, Johnson T, Suzuki M et al. Glutamine repeats as polar
in heterozygous females of X-linked recessive bulbospinal zippers: Their possible role in inherited neurodegenerative diseases.
neuronopathy. J Neurol Sci 1993; 117:74-78. Proc Natl Acad Sci USA 1994; 91:5355-5358.
73. Mariotti C, Castellotti B, Pareyson D et al. Phenotypic manifestations 93. Perutz MF, Pope BJ, Owen D et al. Aggregation of proteins with
associated with CAG-repeat expansion in the androgen receptor gene expanded glutamine and alanine repeats of the glutamine-rich and
in male patients and heterozygous females: A clinical and molecular asparagine-rich domains of Sup35 and of the amyloid beta-peptide
study of 30 families. Neuromuscul Disord 2000; 10:391-397. of amyloid plaques. Proc Natl Acad Sci USA 2002; 99:5596-5600.
74. Schmidt BJ, Greenberg CR, Allingham-Hawkins DJ et al. Expression 94. Taylor JP, Tanaka F, Robitschek J et al. Aggresomes protect cells by
of X-linked bulbospinal muscular atrophy (Kennedy disease) in two enhancing the degradation of toxic polyglutamine-containing protein.
homozygous women. Neurology 2002; 59:770-772. Hum Mol Genet 2003; 12:749-757.
75. Wang Z, Thibodeau SN. A polymerase chain reaction-based test for 95. Piccioni F, Pinton P, Simeoni S et al. Androgen receptor with
spinal and bulbar muscular atrophy. Mayo Clin Proc 1996; elongated polyglutamine tract forms aggregates that alter axonal
71:397-398. trafficking and mitochondrial distribution in motor neuronal
76. Taylor J, Lieberman A, Fischbeck K. Repeat expansion and processes. Faseb J 2002; 16:1418-1420.
neurological diseases. In: Asbury AM, GM McDonald, WI Goadsby, 96. Kobayashi Y, Kume A, Li M et al. Chaperones Hsp70 and Hsp40
PJ McArthur JC, eds. Diseases of the Nervous System. 3rd ed. suppress aggregate formation and apoptosis in cultured neuronal cells
Cambridge: Cambridge University Press, 2002. expressing truncated androgen receptor protein with expanded
77. La Spada AR, Roling DB, Harding AE et al. Meiotic stability and polyglutamine tract. J Biol Chem 2000; 275:8772-8778.
genotype-phenotype correlation of the trinucleotide repeat in X-linked 97. Bailey CK, Andriola IF, Kampinga HH et al. Molecular chaperones
spinal and bulbar muscular atrophy. Nat Genet 1992; 2:301-304. enhance the degradation of expanded polyglutamine repeat androgen
78. Doyu M, Sobue G, Mukai E et al. Severity of X-linked recessive receptor in a cellular model of spinal and bulbar muscular atrophy.
bulbospinal neuronopathy correlates with size of the tandem CAG Hum Mol Genet 2002; 11:515-523.
repeat in androgen receptor gene. Ann Neurol 1992; 32:707-710. 98. Adachi H, Katsuno M, Minamiyama M et al. Heat shock protein
79. Igarashi S, Tanno Y, Onodera O et al. Strong correlation between 70 chaperone overexpression ameliorates phenotypes of the spinal
the number of CAG repeats in androgen receptor genes and the and bulbar muscular atrophy transgenic mouse model by reducing
clinical onset of features of spinal and bulbar muscular atrophy. nuclear-localized mutant androgen receptor protein. J Neurosci 2003;
Neurology 1992; 42:2300-2302. 23:2203-2211.
80. Brooks BP, Paulson HL, Merry DE et al. Characterization of an 99. McCampbell A, Taylor JP, Taye AA et al. CREB-binding protein
expanded glutamine repeat androgen receptor in a neuronal cell sequestration by expanded polyglutamine. Hum Mol Genet 2000;
culture system. Neurobiol Dis 1997; 3:313-323. 9:2197-2202.
81. Mhatre AN, Trifiro MA, Kaufman M et al. Reduced transcriptional 100. Steffan JS, Kazantsev A, Spasic-Boskovic O et al. The Huntington’s
regulatory competence of the androgen receptor in X-linked spinal disease protein interacts with p53 and CREB-binding protein and
and bulbar muscular atrophy. Nat Genet 1993; 5:184-188. represses transcription. Proc Natl Acad Sci USA 2000; 97:6763-6768.
82. Lieberman AP, Harmison G, Strand AD et al. Altered transcriptional 101. Nucifora Jr FC, Sasaki M, Peters MF et al. Interference by huntingtin
regulation in cells expressing the expanded polyglutamine androgen and atrophin-1 with cbp-mediated transcription leading to cellular
receptor. Hum Mol Genet 2002; 11:1967-1976. toxicity. Science 2001; 291:2423-2428.
83. Kazemi-Esfarjani P, Benzer S. Genetic Suppression of Polyglutamine 102. McCampbell A, Taye AA, Whitty L et al. Histone deacetylase
Toxicity in Drosophila. Science 2000; 287:1837-1840. inhibitors reduce polyglutamine toxicity. Proc Natl Acad Sci USA
84. Ordway JM, Tallaksen-Greene S, Gutekunst CA et al. Ectopically 2001; 98:15179-15184.
expressed CAG repeats cause intranuclear inclusions and a progressive 103. Steffan JS, Bodai L, Pallos J et al. Histone deacetylase inhibitors
late onset neurological phenotype in the mouse. Cell 1997; arrest polyglutamine-dependent neurodegeneration in Drosophila.
91:753-763. Nature 2001; 413:739-743.
85. Takeyama K, Ito S, Yamamoto A et al. Androgen-dependent 104. Hockly E, Richon VM, Woodman B et al. Suberoylanilide
neurodegeneration by polyglutamine-expanded human androgen hydroxamic acid, a histone deacetylase inhibitor, ameliorates motor
receptor in Drosophila. Neuron 2002; 35:855-864. deficits in a mouse model of Huntington’s disease. Proc Natl Acad
86. Katsuno M, Adachi H, Kume A et al. Testosterone reduction prevents Sci USA 2003; 100:2041-2046.
phenotypic expression in a transgenic mouse model of spinal and 105. Katsuno M, Adachi H, Doyu M, et al. Leuprorelin rescues
bulbar muscular atrophy. Neuron 2002; 35:843-854. polyglutamine-dependent phenotypes in a transgenic mouse model
87. Davies SW, Beardsall K, Turmaine M et al. Are neuronal intranuclear of spinal and bulbar muscular atrophy. Nat Med 2003; 9:768-73.
inclusions the common neuropathology of triplet-repeat disorders 106. Piccioni F, Taylor J, Fischbeck K. Screen for compounds that inhibit
with polyglutamine-repeat expansions? Lancet 1998; 351:131-133. polyglutamine-induced Caspase-3 activation. (abstract) Neurology
88. Li M, Miwa S, Kobayashi Y et al. Nuclear inclusions of the androgen 2003; 60 (Suppl 1):A529.
receptor protein in spinal and bulbar muscular atrophy. Ann Neurol 107. Heemskerk J, Tobin AJ, Ravina B. From chemical to drug:
1998; 44:249-254. Neurodegeneration drug screening and the ethics of clinical trials.
89. Ellerby LM, Hackam AS, Propp SS et al. Kennedy’s disease: Caspase Nat Neurosci 2002; 5 (Suppl):1027-1029.
cleavage of the androgen receptor is a crucial event in cytotoxicity. J 108. Ross CA. Polyglutamine pathogenesis: Emergence of unifying
Neurochem 1999; 72:185-195. mechanisms for Huntington’s disease and related disorders. Neuron
2002; 35:819-822.
CHAPTER 32
Abstract
ties.4 Cells from XP patients have defects in NER, and therefore,
X
PA, a 273 amino acid protein, is involved in the early
stage of the nucleotide excision repair process, by which are hypersensitive to UV irradiation. Complementation analyses
a variety of DNA lesions are removed from the genome. have identified eight genetic complementation groups in XP:
NMR was used to analyze the structure of the central domain of (XP-A through XP-G) and a variant (XP-V).5
XPA, which encompasses residues 98 to 219, and contains a zinc The gene that complements XP group A cells encodes a
coordinating motif. Following chemical shift assignments of zinc-finger protein, XPA, which is composed of 273 amino acids.
the backbone and side-chain 1H, 15N, and 13C nuclei, the ter- XPA binds preferentially in vitro to a variety of DNA damages
tiary structure was determined by multi-dimensional and including (6-4) photoproducts and crosslinks caused by UV and
multi-resonance NMR methods. The structure shows that the chemicals. Therefore, it has been suggested that XPA is involved
central domain consists of two subdomains, a zinc-containing in the damage recognition step of NER.6-10 XPA also binds di-
subdomain and a C-terminal subdomain, which are connected rectly to many other repair factors: RPA (replication protein A),
by a short linker sequence. The fold adopted by the ERCC1 (excision repair cross-complementing rodent repair defi-
zinc-containing subdomain is similar to those of zinc fingers in ciency 1)/XPF(ERCC4) heterodimer, and TFIIH (transcription
transcription factors, which bind to DNA in a sequence specific factor IIH), which contains XPB and XPD as its subunits.11-18
manner. In contrast to these zinc fingers, the zinc-containing All of these factors are essential for the early steps of NER.1-19
subdomain of XPA is dominantly negatively charged, and thus These observations suggest that XPA also plays a role in loading
unlikely to directly bind to DNA. The interaction of the central the incision protein complex onto the damaged site as a multi-
domain of XPA with a damaged DNA was investigated by a functional protein that coordinates the early steps of NER. Of
chemical shift perturbation experiment, which suggests that the these proteins that function in the early stage of NER, the TFIIH
DNA interact with the positively charged cleft in the C-terminal components, XPB and XPD, have ATP-dependent helicase ac-
subdomain, and not with the zinc-containing subdomain. The tivities to the 3' to 5' and 5' to 3' directions, respectively, and
backbone dynamics were analyzed with 15N longitudinal (T1), form an open DNA complex in the damaged site. XPG may be
transverse (T2), and NOE (nuclear Overhauser effect) relaxation recruited to the lesion by its interaction with RPA. Likewise, the
data. The results show that the proposed DNA-binding surface ERCC1-XPF complex may be recruited to the lesion by the in-
exhibits a highly dynamical feature, which may be necessary for teractions between ERCC1 and XPA18 and between ERCC1-XPF
interacting structurally with various DNA damages. and RPA. An oligonucleotide of about 28 nucleotides including
the damage is excised by XPG and XPF, which are responsible for
Introduction nicking the 3'- and 5'-ends, respectively. As the final step of the
Ultraviolet (UV) irradiation and certain chemical agents can NER process, the DNA polymerase δε synthesizes a new DNA
damage DNA, which is potentially mutagenic or lethal to cells. strand. PCNA (proliferating cell nuclear antigen) and RFC (rep-
Usually in living cells, DNA repair machinery efficiently recog- lication factor C) may be required by RPA to displace the inci-
nizes and eliminates the legions. Nucleotide excision repair (NER) sion proteins from the post-incision complex, followed by liga-
is an important pathway, which removes various kinds of struc- tion of the DNA strand by DNA ligase I.3
turally unlike DNA damages caused by UV and chemical car- Recently, Tanaka et al have succeeded in making XPA
cinogens from the genome.1,2 The NER pathway first recognizes gene-knockout mice (XPA -/- mice).20 The mice were defective in
lesions, and then excises the oligomer carrying the damaged bases, NER and highly susceptible to UVB- or DMBA-induced skin
and finally synthesizes a repair patch using the opposite strand as tumorigenesis, and are therefore expected to be a good animal
the template.3 Defects in NER cause a human inherited (autoso- model for further study of XP-A patients.
mal recessive) disease, xeroderma pigmentosum (XP), which is char-
acterized by an extreme hyper-sensitivity to sunlight, a high fre- Separate Functional Domains of XPA
quency of skin cancers on sun-exposed areas as well as by many XPA consists of several distinct functional domains. The
cancers in other organs, and progressive neurological abnormali- amino-terminal part contains a region (residues 4 to 74) that binds
Zinc Finger Proteins: From Atomic Contact to Cellular Function, edited by Shiro Iuchi and Natalie Kuldell.
©2005 Landes Bioscience/Eurekah.com and Kluwer Academic/Plenum Publishers.
240 Zinc Finger Proteins: From Atomic Contact to Cellular Function
Figure 3. Structures of the XPA zinc-containing subdomain. a) A superposition of the backbone atoms of the 30 best structures of the zinc-containing
subdomain (residues 102 to 129). The side chains of the residues in the hydrophobic core (Val103, Phe112, and Met118) and those of the
zinc-binding cysteine residues (Cys105, Cys108, Cys126 and Cys129) are also shown. b) Stick representation of the zinc-containing subdomain.
Negatively charged residues (Asp101, Glu106, Glu107, Glu111, Asp114 and Asp122) in the vicinity of the β-hairpin, which form an acidic patch,
are indicated. Hydrophobic residues (Tyr116, Leu117, Phe121 and Leu123) on the helical turn, which interact with those from the C-terminal
subdomain, are also shown. (Reproduced with permission from: Ikegami et al. Nature Struct Biol 1998; 5:701-706. Nature Publishing Group.)
242 Zinc Finger Proteins: From Atomic Contact to Cellular Function
Figure 4. Structures of the XPA C-terminal subdomain. a) A superposition of the backbone atoms of the 30 best structures of the C-terminal
subdomain (residues 138 to 209). The backbone atoms of residues 138 to 155, 163 to 165, and 180 to 209 are superimposed. The side chains
of the residues in the hydrophobic core are also shown. They serve to fix the relative positions among the α-helices and β-sheet in the subdomain.
b) Stick representation of the C-terminal subdomain. The well-conserved arginine and lysine residues that contribute to the positive charges in
the basic cleft are shown. The six glutamic acids form an acidic patch on the outer surface of helix α3. The function of this acidic patch is still
unknown. It may be used for an electrostatic interaction with other parts of XPA, or with other NER factors. (Reproduced with permission from:
Ikegami et al. Nature Struct Biol 1998; 5:701-706. Nature Publishing Group.)
structural similarity with the DNA-binding zinc finger of Hydrophobic Contacts between the Two
GATA-1, it is unlikely that the zinc-containing subdomain of Subdomains
XPA directly contacts with DNA, due to electrostatic repulsive
The zinc-containing subdomain offers some hydrophobic resi-
forces between negative charges in the subdomain and DNA.
dues, Tyr116, Leu117, Phe121 and Leu123, on the outer surface
of the helical turn (Fig. 3b), which make contacts with hydro-
The Structure of the C-Terminal Subdomain phobic residues of the C-terminal subdomain, Leu138, Ile165,
The C-terminal subdomain consists of a sheet-helix-loop re- Leu182 and Leu184. These contacts fix the three-stranded β-sheet
gion (residues 138 to 182), and a helix-turn-helix region (resi- in the C-terminal subdomain with respect to the zinc-containing
dues 183 to 209) (Fig. 4). The sheet-helix-loop region is com- subdomain (Fig. 5), indicating that the zinc-containing subdomain
posed of an antiparallel β-sheet (strands β3-5), helix α1, and two is not a structurally independent domain. This feature is in con-
long loops, L1 and L2. Loop L1 (residues 148 to 163) connects trast to the zinc fingers of transcriptional factors, most of which
helix α1 and strand β4, and loop L2 (residues 166 to 179) con- are thought to exist as independent domains, separated from other
nects strands β4 and β5. A heteronuclear {1H}-15N steady-state domains by flexible linkers in modular proteins. This character
NOE experiment showed that a part of L1 (residues 157 to 162), probably makes finding a biochemical function of the isolated
L2, and the C-terminal sequence (residues 211 to 219) are highly zinc-containing subdomain difficult.10,13
mobile in solution. L2 apparently formed no definite structure
since few long-range NOEs were observed. The signals for resi- Exon/Intron Boundaries and Tertiary
dues 171 to 174 in L2 were not observed in the 15N-1H-HSQC Structure Elements
spectra, probably due to local exchange broadening and/or rapid It has been shown that the human xpa gene contains six ex-
exchange of the amide protons with solvent. The helix-turn-helix ons.29 Interestingly, the genomic structure of the xpa gene has a
region consists of two long helices, α2 (residues 183 to 194) and good correlation with the tertiary structure elements. Exon 3 en-
α3 (residues 197 to 209), connected by the conserved codes residues 96 to 130 of XPA, which correspond to the entire
Gly195-Ser196 sequence forming a short turn. This region is zinc-containing subdomain, and exon 4 (residues 131 to 185)
packed tightly against helix α1, loop L1, and the β-sheet of the encodes the region containing the linker (residues 130 to 137)
sheet-helix-loop region, stabilized by a hydrophobic core con- and the sheet-helix-loop region (residues 138 to 182). Finally,
taining Ala144, Tyr148, Leu149, Leu150, Leu155, Leu162, the region encoded by exon 5, residues 186 to 224, covers most
Tyr181 and Ile186 (Fig. 4a). The C-terminal subdomain has a of the helix-turn-helix region (residues 183 to 209) and the
large cleft between the sheet-helix-loop region and the C-terminal flanking sequence (residues 211 to 219) (Fig. 6). This
helix-turn-helix region. The surface electrostatic potentials show indicates that the exon/intron junctions are located at the sites
that many positively charged side chains, which are contributed corresponding to the inter-module junctions of the central do-
by well-conserved Lys141, Lys145, Lys151, Lys179, Lys204 and main of XPA. Such a correlation between the functional domains
Arg207, are present in the cleft (Fig. 4b). of XPA and its genomic structure does not seem to be limited to
The Role of XPA in DNA Repair 243
Buchko and Kennedy et al have recently proposed a model of RPA70181-422 nonspecifically interacts with the zinc-containing
the complex between the XPA central domain and a single-stranded subdomain, due primarily to the electrostatic force.
DNA nonamer, dCCAATAACC.34 They monitored chemical shift Through deletion analyses, Saijo et al have shown that residues
changes or peak broadenings in 31P resonances of the nonamer 98 to 187 of XPA, which include the zinc-containing subdomain
DNA as well as in amide 15N-1H resonances of the XPA central and the sheet-helix-loop region (residues 138 to 182) of the
domain in titration experiments, and calculated the model struc- C-terminal subdomain, are sufficient for binding to RPA70, but
ture using these chemical shift perturbation data with the assump- that further deletion of the sheet-helix-loop region causes the loss
tion that electrostatic interactions drive the complex formation. of RPA70 binding.13 This observation raises the possibility that
The model shows that the DNA covers a surface of the C-terminal the sheet-helix-loop region of XPA serves as another and indis-
subdomain, in particular α1 and the triple stranded β-sheets in the pensable binding surface for the N- or C-terminal region of RPA70,
subdomain, and is thus consistent with the result of our perturba- outside of RPA70181-422. Apparently, future experiments need to
tion experiment. be directed at identifying the major RPA-binding site of XPA.
Chazin and coworkers have shown that the C-terminal domain of
RPA32 interacts with the N-terminal 98 residues of XPA.39
Interaction with RPA70
RPA is a heterotrimeric single-stranded-DNA-binding pro-
tein, which is involved in DNA replication, homologous re-
Role of XPA in NER
combination, and NER. For NER, the interaction between Although various experiments conducted so far, including the
RPA and XPA has been shown to be essential for the opening ones described here, have shown that XPA interacts with dam-
of damaged DNA regions by TFIIH.11,12,35,36 Although XPA aged DNA, the role of XPA in damage recognition is still unclear.
Wakasugi et al have observed that the rate of damage removal was
binds to the two largest subunits, RPA70 and RPA34, a dele-
five times faster when the combination of XPA and RPA first
tion analysis of XPA showed that RPA70 contributes more
bound to a single (6-4) photoproduct in a 136-bp duplex than
predominately to the XPA-binding than RPA34 does, and the
when XPC first bound to the damaged DNA.40 They proposed a
central domain of XPA contains the XPA70 binding site.11-13 model for assembly of the NER-related proteins, in which
By means of limited proteolysis, mutational analysis, and crys- XPA+RPA first forms a complex with damaged DNA with a low
tal structure determination, Pfuetzner et al have shown that specificity and then becomes more discriminatory upon recruit-
residues 181 to 422 of RPA70, RPA70181-422, form a compact ing XPC and TFIIH. In contrast, recent findings have suggested
structural domain that maintains single-stranded DNA bind- a two-step mechanism for damage recognition during global ge-
ing activity.37,38 nome repair, in which various types of damages are detected by
In order to examine the interaction between RPA70181-422 and XPC-HR23B first,41 and then verified by XPA.42 You et al have
the central domain of XPA, we performed chemical shift pertur- suggested that damage recognition occurs in a multi-step pro-
bation experiments using the 15N-labeled central domain. Selec- cess, whereby XPC-HR23B initiates damage recognition and then
tive signal losses were observed in a 15N-1H-HSQC spectrum XPC-HR23B is replaced by a combined action of XPA and RPA.43
upon mixing with an equimolar amount of RPA70181-422 when As for the identification of cyclobutane pyrimidine dimers
the salt concentration was sufficiently low (see below). In gen- (CPDs), XPE may be involved.44
eral, signal losses occur when the chemical shifts are perturbed
considerably by interaction, when the exchange rate between the Relationships between XPA Mutations
free and bound states is comparable to the chemical shift differ-
ence between the two states, or when the ligand binding enhances and Disease
the amide proton exchange rate with a solvent. Almost all of The patients suffering from XP have mutations in the genes
the signals for which the peak intensities decreased to < 65% of encoding the proteins involved in nucleotide excision repair pro-
the original values were attributed to the residues in the cesses, XPA to XPG. The mutations lead the deficiency in effec-
zinc-containing subdomain (Fig. 7). Buchko and coworkers re- tive repairs of DNA lesions caused by UV and chemicals. Of these,
mutations in the xpa gene (shown in Fig. 8) tend to cause more
ported that not only the zinc-containing subdomain but also parts
severe symptoms than others, suggesting the functional signifi-
of the C-terminal subdomain, the β-sheet and loop L2, exhibited
cance of XPA in NER. The missense mutation of Cys108 (TGT)
perturbed chemical shifts upon addition of RPA70∆C327, which to Phe108 (TTT) most likely has a global structural consequence
ranges from M1 to Y326 and contains the N-terminal 146 resi- to the zinc-containing subdomain because this cysteine coordi-
dues (60%) out of RPA70181-422.22 They suggested that RPA may nates the zinc-ion. The nonsense mutation of Tyr116 (TAT) re-
interact only with the zinc-containing subdomain, but that the sults in deletion of a large part of the C-terminal subdomain of
interaction may generate a conformational change in the the central domain together with the following region of XPA,
C-terminal subdomain to enhance XPA binding to damaged and thus, probably results in a total loss of function of XPA. The
DNA. patient bearing the nonsence mutation of Arg207 (CGA) was
The interaction between the XPA central domain and reported to exhibit a severe symptom, whereas the mutation of
RPA70181-422 is highly sensitive to the salt concentration. When Arg228 (CGA) leads only to a mild symptom. This observation
the KCl concentration was increased from 28 to 77 mM, the suggests that the C-terminal part after the central domain, which
specific broadening observed for the zinc-containing subdomain contains the TFIIH binding site, is not essential for NER, but
of XPA became significantly smaller. Since the surface of the that the further deletion of a large part of the cleft region of the
zinc-containing subdomain is highly negatively charged, we can- C-terminal subdomain of the central domain of XPA leads to a
not exclude the possibility that a positively charged patch in severe deficiency in NER.
The Role of XPA in DNA Repair 245
Figure 8. Mutations in the xpa gene found in XP patients. a) The mutations are shown on the map of the xpa gene, which contains six exons
(I to VI). The central domain corresponds to the region composed of exons III, IV, and V. b) Some mutations in the central domain are mapped
on the tertiary structure. The terms “stop” and “splicing” represent nonsense and splicing mutations, respectively.
246 Zinc Finger Proteins: From Atomic Contact to Cellular Function
References 24. Lipton AS, Buchko GW, Sears JA et al. 67Zn solid-state NMR
1. Aboussekhra A, Wood RD. Repair of UV-damaged DNA by spectroscopy of the minimal DNA binding domain of human
mammalian cells and Saccharomyces cerevisiae. Curr Opin Genet nucleotide excision repair protein XPA. J Am Chem Soc 2001;
Dev 1994; 4:212-220. 123:992-993.
2. Cline SD, Hanawalt PC. Who’s on first in the cellular response to 25. Hess NJ, Buchko GW, Conradson SD et al. Human nucleotide
DNA damage? Nat Rev Mol Cell Biol 2003; 4:361-372. excision repair protein XPA: Extended X-ray absorption fine-structure
3. Thompson LH. Nucleotide excision repair in DNA damage and evidence for a metal-binding domain. Protein Sci 1998; 7:1970-1975.
repair. In: Nickoloff JA, Hoekstra MF, eds. DNA Repair in Higher 26. Morita EH, Ohkubo T, Kuraoka I et al. Implications of the
Eukaryotes. New Jersey: Human Press, 1998:II:335-393. zinc-finger motif found in the DNA-binding domain of the human
4. Cleaver JE, Kraemer KH. Xeroderma pigmentosum. In: Scriver CR, XPA protein. Genes Cells 1996; 1:437-442.
ed. The Metabolic Basis of Inherited Disease, 7th ed. New York: 27. Summers MF. Zinc fingers. In: Grant DM, Harris RK, eds.
McGraw-Hill, 1995:4393-4419. Encyclopedia of Nuclear Magnetic Resonance. West Sussex, England:
5. Hoeijmakers JH, Bootsma D. Molecular genetics of eukaryotic DNA John Wiley and Sons Ltd., 1996:8:5063-5071.
excision repair. Cancer Cells Mon Rev 1990; 2:311-320. 28. Omichinski JG, Clore GM, Schaad O et al. NMR structure of a
6. Tanaka K, Miura N, Satokata I et al. Analysis of a human DNA specific DNA complex of Zn-containing DNA binding domain of
excision repair gene involved in group A xeroderma pigmentosum GATA-1. Science 1993; 261:438-446.
and containing a zinc-finger domain. Nature 1990; 348:3-76. 29. Satokata I, Iwai K, Matsuda T et al. Genomic characterization of
7. Robins P, Jones CJ, Biggerstaff M et al. Complementation of DNA the human DNA excision repair-controlling gene XPAC. Gene 1993;
repair in xeroderma pigmentosum group A cell extracts by a protein 136:345-348.
with affinity for damaged DNA. EMBO J 1991; 10:3913-3921. 30. Go M. Protein structures and split genes. Adv Biophys 1985;
8. Jones CJ, Wood RD. Preferential binding of the xeroderma 19:91-131.
pigmentosum group A complementing protein to damaged DNA. 31. Nakashima T, Sekiguchi T, Sunamoto H et al. Structure of the
Biochemistry 1993; 32:12096-12104. human CCG1 gene: Relationship between the exons/introns and
9. Asahina H, Kuraoka I, Shirakawa M et al. The XPA protein is a functional domain/modules of the protein. Gene 1994; 141:193-200.
zinc metalloprotein with an ability to recognize various kinds of DNA 32. Moggs JG, Yarema KJ, Essigmann JM et al. Analysis of incision
damage. Mutat Res 1994; 315:229-237. sites produced by human cell extracts and purified proteins during
10. Kuraoka I, Morita EH, Saijo M et al. Identification of a nucleotide excision repair of a 1,3-intrastrand d(GpTpG)-cisplatin
damaged-DNA binding domain of the XPA protein. Mutat Res 1996; adduct. J Biol Chem 1996; 271:7177-7186.
362:87-95. 33. Ikegami T, Kuraoka I, Saijo M et al. Resonance assignments, solution
11. He Z, Henricksen LA, Wold MS et al. RPA involvement in the structure, and backbone dynamics of the DNA- and RPA-binding
damage-recognition and incision steps of nucleotide excision repair. domain of human repair factor XPA. J Biochemistry 1999;
Nature 1995; 374:566-569. 125:495-506.
12. Li L, Lu X, Peterson CA et al. An interaction between the DNA 34. Buchko GW, Tung CS, McAteer K et al. DNA-XPA interactions: A
31
repair factor XPA and replication protein A appears essential for P NMR and molecular modeling study of dCCAATAACC
nucleotide excision repair. Mol Cell Biol 1995; 15:5396-5402. association with the minimal DNA-binding domain (M98-F219) of
13. Saijo M, Kuraoka I, Masutani C et al. Sequential binding of DNA the nucleotide excision repair protein XPA. Nucleic Acids Res 2001;
repair proteins RPA and ERCC1 to XPA in vitro. Nucleic Acids 29:2635-2643.
Res 1996; 24:4719-4724. 35. Guzder SN, Habraken Y, Sung P et al. Reconstitution of yeast
14. Park CH, Mu D, Reardon JT et al. The general transcription-repair nucleotide excision repair with purified Rad proteins, replication
factor TFIIH is recruited to the excision repair complex by the XPA protein A, and transcription factor TFIIH. J Biol Chem 1995;
protein independent of the TFIIE transcription factor. J Biol Chem 270:12973-12976.
1995; 270:4896-4902. 36. Aboussekhra A, Biggerstaff M, Shivji MK et al. Mammalian DNA
15. Nocentini S, Coin F, Saijo M et al. DNA damage recognition by nucleotide excision repair reconstituted with purified protein
XPA protein promotes efficient recruitment of transcription factor components. Cell 1995; 80:859-868.
II H. J Biol Chem 1997; 272:22991-22994. 37. Bochkarev A, Pfuetzner RA, Edwards AM et al. Structure of the
16. Li L, Elledge SJ, Peterson CA et al. Specific association between the single-stranded-DNA-binding domain of replication protein A bound
human DNA repair proteins XPA and ERCC1. Proc Natl Acad Sci to DNA. Nature 1997; 385:176-181.
USA 1994; 91:5012-5016. 38. Pfuetzner RA, Bochkarev A, Frappier L et al. Replication protein A:
17. Li L, Peterson CA, Lu X et al. Mutations in XPA that prevent Characterization and crystallization of the DNA binding domain. J
association with ERCC1 are defective in nucleotide excision repair. Biol Chem 1997; 272:430-434.
Mol Cell Biol 1995; 15:1993-1998. 39. Mer G, Bochkarev A, Gupta R et al. Structural basis for the
18. Nagai A, Saijo M, Kuraoka I et al. Enhancement of damage-specific recognition of DNA repair proteins UNG2, XPA, and RAD52 by
DNA binding of XPA by interaction with the ERCC1 DNA repair replication factor RPA. Cell 2000; 103:449-456.
protein. Biochem Biophys Res Commun 1995; 211:960-966. 40. Wakasugi M, Sancar A. Order of assembly of human DNA repair
19. Mu D, Hsu DS, Sancar A. Reaction mechanism of human DNA excision nuclease. J Biol Chem 1999; 274:18759-18768.
repair excision nuclease. J Biol Chem 1996; 271:8285-8294. 41. Hey T, Lipps G, Sugasawa K et al. The XPC-HR23B complex
20. Tanaka K, Kamiuchi S, Ren Y et al. UV-induced skin carcinogenesis displays high affinity and specificity for damaged DNA in a
in xeroderma pigmentosum group A (XPA) gene-knockout mice with true-equilibrium fluorescence assay. Biochemistry 2002;
nucleotide excision repair-deficiency. Mutat Res 2001; 477:31-40. 41:6583-6587.
21. Ikegami T, Kuraoka I, Saijo M et al. Solution structure of the DNA- 42. Sugasawa K, Ng JM, Masutani C et al. Xeroderma pigmentosum
and RPA-binding domain of the human repair factor XPA. Nature group C protein complex is the initiator of global genome nucleotide
Struct Biol 1998; 5:701-706. excision repair. Mol Cell 1998; 2:223-232.
22. Buchko GW, Daughdrill GW, Lorimier R et al. Interactions of 43. You JS, Wang M, Lee SH. Biochemical analysis of the damage
human nucleotide excision repair protein XPA with DNA and recognition process in nucleotide excision repair. J Biol Chem 2003;
RPA70∆C327: Chemical shift mapping and 15N NMR relaxation 278:7476-7485.
studies. Biochemistry 1999; 38:15116-15128. 44. Tang JY, Hwang BJ, Ford JM et al. Xeroderma pigmentosum p48
23. Miyamoto I, Miura N, Niwa H et al. Mutational analysis of the gene enhances global genomic repair and suppresses UV-induced
structure and function of the xeroderma pigmentosum group A mutagenesis. Mol Cell 2000; 5:737-744.
complementing protein. Identification of essential domains for
nuclear localization and DNA excision repair. J Biol Chem 1992;
267:12182-12187.
CHAPTER 33
Abstract
T
he histone acetyltransferase MOF is central to the process histone N-terminal tails has emerged as a major principle under-
of dosage compensation complex in Drosophila, which lying dynamic transitions between permissive and repressed chro-
assures that the transcriptional activity of large parts of matin structure.
the male X chromosome is enhanced by about two-fold. Inacti- The acetylation status of chromosomal loci is negotiated
vation of the acetyl transferase activity of MOF is lethal to af- through a dynamic interplay between histone acetyl transferases
fected male flies. MOF belongs to the MYST family of acetylases (HATs) and corresponding deacetylases (HDACs), which may
that are involved in a wide variety of cellular processes. One char- vary in concentration and activity, and may be targeted to their
acteristic structural feature of this family is the presence of a sites of action by sequence-specific DNA binding factors.5 In gen-
CCHC zinc finger adjacent to their acetyl-CoA binding motif. eral, acetylation of core histones is associated with transcription-
Structure-function analyses of several family members suggest ally active genes, whereas deacetylation of histones often corre-
that this region is important for enzymatic activity. Our detailed lates with repression,6 but exceptions to these rules exist.7
characterization of MOF shows that the CCHC-type zinc fin- HDACs and HATs can be categorized according to their do-
ger is required for interaction with the histone substrate and hence main structure, function and substrate specificity.3,8,9 The cyto-
catalytic activity. plasmic HATs (of B-type) are involved in histone storage, trans-
port and assembly. The A-type HATs are nuclear regulators and
Introduction can be grouped into the GNAT family, the p300/CBP family
The organization of eukaryotic genomes as chromatin repre- and the MYST family.8,10 We will be focusing on MYST acetylases,
sents a considerable barrier for a variety of nuclear processes with which feature a CCHC zinc finger as an integral part of their
DNA substrate, such as transcription of genes, the replication of acetyl transferase domain.
the chromosome and the repair of damaged DNA. DNA is coiled
around histone octamers to form successions of nucleosomes with The MYST Family of Acetyl Transferases
beads-on-a-string appearance at low salt.1-3 The folding of the
nucleosomal fiber into compact structures under physiological The ‘MYST-signature’ emerged when similarities in the do-
conditions can render the wrapped DNA inaccessible to interact- main structure of the founding members hMOZ, ySas2, yYbf2/
ing regulators. The degree and type of folding and hence the func- Sas3 and hTip60 was noted. Later, additional members such as
tional state of a locus is defined by interactions between nucleo- yeast Esa1, Drosophila MOF, human MORF and HBO1 were
somes and interactions between nucleosomes and nonhistone added to the list.8,11,12 MYST proteins are mainly related by simi-
proteins. Main targets for these interactions are the histone larity around their acetyl transferase domains, which includes a
N-terminal domains, their ‘tails’ which protrude from the com- CxxCx12HxxxC zinc finger (all except Esa1, see below) (Fig. 1).
pact globular part of the nucleosome and are hence available for Apart from the ‘MYST signature’ the proteins differ in size and
interactions. Their potential for interactions can be modified by structure, although some recurring domains, such as
post-translational modifications. The histone N termini contain chromodomains and PHD fingers can be noted (Fig. 1). MYST
a number of conserved residues that can be acetylated, methy- histone acetyl transferases are evolutionarily conserved proteins
lated, phosophorylated, ubiquitinated and ribosylated. Distinct involved in a wide variety of regulatory functions. Although they
histone modification patterns serve as semaphore ‘flags’ which were originally identified as histone acetyl transferases (HATs) it
are decoded by nonhistone regulators of chromatin folding.4 The is quite likely that they may modify other, nonhistone substrates
reversible acetylation of the ε-amino group of lysines within the as well, as was recently shown for MOF (see below).
Zinc Finger Proteins: From Atomic Contact to Cellular Function, edited by Shiro Iuchi and Natalie Kuldell.
©2005 Landes Bioscience/Eurekah.com and Kluwer Academic/Plenum Publishers.
248 Zinc Finger Proteins: From Atomic Contact to Cellular Function
Since MYST acetylases were largely identified through sequence Drosophila MOF is essential to this process: inactivation of
similarity screens, our knowledge of their function is rudimen- MOF leads to male-specific lethality, a striking example of the
tary in many cases. However, they all are found in large drastic effect a two-fold difference in gene expression may have.
multisubunit complexes involved in gene regulation. Tip60, for MOF is part of a multisubunit dosage compensation complex
example, the first human MYST protein member, was identified (DCC), which contains at least four other proteins, the precise
in a yeast 2-hybrid screen in a search for Tat (HIV-1 function of which is not clear yet.28,29
transcactivator) interacting proteins.13 It associates with at least A hallmark of the hypertranscribed male X chromosome is
11 other proteins in a complex that may have other functions the acetylation of histone H4 at lysine 16 (H4K16ac).28-30 This
besides the nucleosomal acetylase and that can be recruited to epigenetic mark is set by MOF and failure to do so leads to lack
chromatin through interactions with the transcription factor of dosage compensation and male-specific lethality. Whether
MYC.14 It also interacts with the TEL/ETV6 repressor and has embedded in the DCC or as a recombinant protein, MOF spe-
been suggested to have corepressor function.15 Deregulation of cifically acetylates chromatin.31,32 This single modification suf-
MOZ (monocytic leukemia zinc finger protein) leads to onco- fices to relieve chromatin-mediated repression of transcription in
genic transformations.16,17 In acute myeloid leukemia, the MOZ vitro and in vivo.31 MOF binds chromatin nonspecifically in
gene is translocated to fuse with the acetylase CBP gene at vari- vitro31 and hence the selective association with the male X chro-
ous breakpoints.18 MOZ is itself an acetyl transferase able to in- mosome observed in vivo depends on other parts of the DCC.
teract with transcription factors, but additionally containing tran-
MOF interacts with other DCC subunits (Morales, Mengus,
scription repression and activation domains.19,20 The yeast Sas2
Akhtar and Becker, submitted),33-35 and most interestingly also
and Sas3/Ybf2 proteins are involved in silencing of mating type
requires a noncoding RNA, roX, for proper association with the
loci (Sas: ‘something about silencing’). Sas2 was discovered in a
screen for epigenetic silencing, while Sas3 was discovered because X chromosome.36-38 Besides histones, MOF is also able to acety-
of its homology with Sas2. Sas3 is part of the NuA3 complex, late other associated subunits within DCC, such as MSL1 and
which acetylates histone H3 in nucleosomes.21,22 Sas2, unlike Sas3, MSL3, suggesting a coordinated modification of the histone in-
does have not HAT activity as a recombinant protein but does frastructure and the nonhistone regulators themselves (Morales,
have HAT activity as part of a complex with Sas4 and Sas5.23 Mengus, Akhtar and Becker, submitted).35
Most is known about the regulatory role of MOF, a Drosophila Drosophila MOF is highly related to human MOF (hMOF)
HAT which is central to the process of dosage compensation, (52% identity and 69% similarity), which acetylates histones with
whereby transcription of many genes on the single male X chro- a specificity similar to the other members of the MYST family.39
mosome is increased by two-fold as compared the transcription The functional role of hMOF protein is currently unknown. The
from either of the two female X genes.24 As will be detailed, the closest relative of MOF in yeast is Esa1, which is part of the
role of the CCHC zinc finger in MOF function is particularly multisubunit NuA4 complex involved in transcriptional activa-
well documented. Thus, MOF serves as a paradigm for the activ- tion.40,41 Intriguingly, the NuA4 complex has other similarities
ity of MYST acetyl transferases. to the DCC in flies: the NuA4 subunit Eaf3 is homologous to
Drosophila MSL3.40,42
The Role of MOF in Dosage Compensation
Dosage compensation mechanisms ensure that, despite of The MOF CCHC Zinc Finger Is Involved in
the fact that male and female cells contain different numbers of Substrate Recognition
X chromosomes (one in males and two in females in humans Part of the sequence similarity that characterises MYST acetyl
and Drosophila) the expression of X-linked genes is equal in transferases is a CxxCx12HxxxC zinc finger close to their HAT
both sexes. Dosage compensation is achieved differently in dif- domains. Zinc fingers of this type are frequently involved in
ferent organisms. In mammals, one of the two female X chro- protein-protein interactions.43 Although they fold as indepen-
mosomes is inactivated by epigenetic means. 25 In Caenorhabditis dent units,44,45 they frequently perform their functions in com-
elegans, the activity of each of the two hermaphrodite X chro- bination with other elements, such as CCHH fingers in LIM
mosomes is halved.26 By contrast, in Drosophila, both female X domains,46 where they are again involved in protein-protein con-
chromosomes are transcribed, but transcription from the ma- tacts (see reference 47 and references therein). The CCHC fin-
jority of genes on the single male X chromosome is increased gers of the HIV1 nucleocapsid proteins bind single-stranded DNA
two-fold to make up for the reduced chromosome dosage.27 during the replication of the viral genome.48
MOF Uses a CCHC Finger for Substrate Recognition 249
Figure 2. Schematic drawing of the MOF CCHC zinc finger region and
position of the different point mutations. Each point mutation is shown
by the original residue, the position and the mutant residue. Residues
which, if mutated, affect chromatin binding and acetylase function, are
marked in black.
Acknowledgements
We thank Gerit Vriend for fruitful discussions and Mitzi
Kuroda for anti-MOF antibodies. Research on MOF in the lab
of P.B. is supported by the Deutsche Forschungsgemeinschaft.
References
1. Hayes JJ, Hansen JC. Nucleosomes and the chromatin fiber. Curr
Opin Genet Dev 2001; 11(2):124-129.
2. Woodcock CL, Dimitrov S. Higher-order structure of chromatin and
chromosomes. Curr Opin Genet Dev 2001; 11(2):130-135.
3. Gregory PD, Wagner K, Horz W. Histone acetylation and chromatin
remodeling. Exp Cell Res 2001; 265(2):195-202.
4. Jenuwein T, Allis CD. Translating the histone code. Science 2001;
293(5532):1074-1080.
5. Eberharter A, Becker PB. Histone acetylation: A switch between
repressive and permissive chromatin. Second in review series on
chromatin dynamics. EMBO Rep 2002; 3(3):224-229.
6. Allfrey VG, Pogo BG, Littau VC et al. Histone acetylation in insect
chromosomes. Science 1968; 159(812):314-316.
7. Suka N, Luo K, Grunstein M. Sir2p and Sas2p opposingly regulate
acetylation of yeast histone H4 lysine16 and spreading of
Figure 4. A) Nucleic acid binding of zinc finger mutants. DNA electro- heterochromatin. Nat Genet 2002; 32(3):378-383.
phoretic mobility shift assay (EMSA) were performed as described in 8. Sterner DE, Berger SL. Acetylation of histones and
reference 36, except 75 bp ss DNA oligo was used as a probe. 150 ng transcription-related factors. Microbiol Mol Biol Rev 2000;
64(2):435-459.
(lanes 2, 4, 6) or 300 ng (lanes 3, 5, 7) of WT MOF, C594H or H590C 9. de Ruijter AJ, van Gennip AH, Caron HN et al. Histone deacetylases
were used respectively. B) DNA EMSA as in (A), but including 5, 10 or (HDACs): Characterization of the classical HDAC family. Biochem
15 mM 1,10-phenanthroline (lanes 2-4, 6-8, 10-12, respectively). Lane J 2003; 370(Pt 3):737-749.
13, ss DNA probe alone. 10. Struhl K. Histone acetylation and transcriptional regulatory
mechanisms. Genes Dev 1998; 12(5):599-606.
11. Timmermann S, Lehrmann H, Polesskaya A et al. Histone acetylation
when Marmorstein and colleagues determined the structure of
and disease Cell Mol Life Sci. May 2001; 58(5-6):728-736.
the HAT domain of Esa1 in complex with coenzyme A and found 12. Utley RT, Cote J. The MYST family of histone acetyltransferases.
that the corresponding region forms a classical TFIIIA-type zinc Curr Top Microbiol Immunol 2003; 274:203-236.
finger fold entirely in the absence of zinc.50 This finding provides 13. Kamine J, Elangovan B, Subramanian T et al. Identification of a
unexpected support to the idea of the importance of the zinc cellular protein that specifically interacts with the essential cysteine
finger fold in MYST acetylase function. region of the HIV-1 Tat transactivator. Virology 1996;
216(2):357-366.
14. Frank SR, Parisi T, Taubert S et al. MYC recruits the TIP60 histone
Concluding Remarks acetyltransferase complex to chromatin. EMBO Rep 2003;
The detailed analysis of Esa1 structure revealed fundamen- 4(6):575-580.
tally different catalytic principles between MYST family mem- 15. Nordentoft I, Jorgensen P. The acetyltransferase Tip60 interacts with
bers and, for example, acetylases of the GCN5/PCAF family.51 TEL/ETV6 and functions as a transcriptional corepressor. Biochem
J 2003; Pt.
Although both classes of enzymes target similar histone substrates, 16. Deguchi K, Ayton PM, Carapeti M et al. MOZ-TIF2-induced acute
their mode of substrate recognition may also be very different. myeloid leukemia requires the MOZ nucleosome binding motif and
However, while MYST acetylases are specialized in using a TIF2-mediated recruitment of CBP. Cancer Cell 2003; 3(3):259-271.
CxxCx12HxxxC zinc finger for substrate recognition, other his- 17. Borrow J, Stanton Jr VP, Andresen JM et al. The translocation
tone modification enzymes require different types of zinc fingers t(8;16)(p11;p13) of acute myeloid leukaemia fuses a putative
for full function. For example, the acetyl transferases P300/CBP acetyltransferase to the CREB-binding protein. Nat Genet 1996;
14(1):33-41.
MOF Uses a CCHC Finger for Substrate Recognition 251
18. Panagopoulos I, Isaksson M, Lindvall C et al. Genomic 39. Neal KC, Pannuti A, Smith ER et al. A new human member of the
characterization of MOZ/CBP and CBP/MOZ chimeras in acute MYST family of histone acetyl transferases with high sequence
myeloid leukemia suggests the involvement of a damage-repair similarity to Drosophila MOF. Biochim Biophys Acta 2000;
mechanism in the origin of the t(8;16)(p11;p13). Genes 1490(1-2):170-174.
Chromosomes Cancer 2003; 36(1):90-98. 40. Eisen A, Utley RT, Nourani A et al. The yeast NuA4 and Drosophila
19. Champagne N, Pelletier N, Yang XJ. The monocytic leukemia zinc MSL complexes contain homologous subunits important for
finger protein MOZ is a histone acetyltransferase. Oncogene 2001; transcription regulation. J Biol Chem 2001; 276(5):3484-3491.
20(3):404-409. 41. Clarke AS, Lowell JE, Jacobson SJ et al. Esa1p is an essential histone
20. Pelletier N, Champagne N, Stifani S et al. MOZ and MORF histone acetyltransferase required for cell cycle progression. Mol Cell Biol
acetyltransferases interact with the Runt-domain transcription factor 1999; 19(4):2515-2526.
Runx2. Oncogene 2002; 21(17):2729-2740. 42. Gavin AC, Bosche M, Krause R et al. Functional organization of
21. Reifsnyder C, Lowell J, Clarke A et al. Yeast SAS silencing genes the yeast proteome by systematic analysis of protein complexes.
and human genes associated with AML and HIV-1 Tat interactions Nature 2002; 415(6868):141-147.
are homologous with acetyltransferases. Nat Genet 1996; 14(1):42-49. 43. Kowalski K, Liew CK, Matthews JM et al. Characterization of the
22. John S, Howe L, Tafrov ST. The something about silencing protein, conserved interaction between GATA and FOG family proteins. J
Sas3, is the catalytic subunit of NuA3, a yTAF(II)30-containing HAT Biol Chem 2002; 277(38):35720-35729.
complex that interacts with the Spt16 subunit of the yeast CP 44. Hammarstrom A, Berndt KD, Sillard R et al. Solution structure of
(Cdc68/Pob3)-FACT complex. Genes Dev 2000; 14(10):1196-1208. a naturally-occurring zinc-peptide complex demonstrates that the
23. Sutton A, Shia WJ, Band D et al. Sas4 and Sas5 are required for N-terminal zinc-binding module of the Lasp-1 LIM domain is an
the histone acetyltransferase activity of Sas2 in the SAS complex. J independent folding unit. Biochemistry 1996; 35(39):12723-12732.
Biol Chem 2003; 278(19):16887-16892. 45. Yao X, Perez-Alvarado GC, Louis HA et al. Solution structure of
24. Hilfiker A, Hilfiker-Kleiner D, Pannuti A et al. mof, a putative acetyl the chicken cysteine-rich protein, CRP1, a double-LIM protein
transferase gene related to the Tip60 and MOZ human genes and implicated in muscle differentiation. Biochemistry 1999;
to the SAS genes of yeast, is required for dosage compensation in 38(18):5701-5713.
Drosophila. EMBO J 1997; 16(8):2054-2060. 46. Schwabe JW, Klug A. Zinc mining for protein domains. Nat Struct
25. Lyon M. Gene action in the X-chromosome of the mouse (Mus Biol 1994; 1(6):345-349.
musculus). Nature 1961; 190:372-373. 47. Schmeichel KL, Beckerle MC. Molecular dissection of a LIM
26. Meyer BJ, Casson LP. Caenorhabditis elegans compensates for the domain. Mol Biol Cell 1997; 8(2):219-230.
difference in X chromosome dosage between the sexes by regulating 48. Guo J, Wu T, Kane BF et al. Subtle alterations of the native zinc
transcript levels. Cell 1986; 47(6):871-881. finger structures have dramatic effects on the nucleic acid chaperone
27. Mukherjee AS, Beermann W. Synthesis of ribonucleic acid by the activity of human immunodeficiency virus type 1 nucleocapsid
X-chromosomes of Drosophila melanogaster and the problem of protein. J Virol 2002; 76(9):4370-4378.
dosage compensation. Nature 1965; 207(998):785-786. 49. Akhtar A, Becker PB. The histone H4 acetyltransferase MOF uses a
28. Lucchesi JC. Dosage compensation in flies and worms: The ups and C2HC zinc finger for substrate recognition. EMBO Rep 2001;
downs of X-chromosome regulation. Curr Opin Genet Dev 2(2):113-118.
1998;8(2):179-184. 50. Yan Y, Barlev NA, Haley RH et al. Crystal structure of yeast Esa1
29. Akhtar A. Dosage compensation: An intertwined world of RNA and suggests a unified mechanism for catalysis and substrate binding by
chromatin remodelling. Curr Opin Genet Dev 2003; 13(2):161-169. histone acetyltransferases. Mol Cell 2000; 6(5):1195-1205.
30. Lucchesi JC. Dosage compensation in Drosophila and the “complex’ 51. Yan Y, Harper S, Speicher DW et al. The catalytic mechanism of
world of transcriptional regulation. Bioessays 1996; 18(7):541-547. the ESA1 histone acetyltransferase involves a self-acetylated
31. Akhtar A, Becker PB. Activation of transcription through histone intermediate. Nat Struct Biol 2002; 9(11):862-869.
H4 acetylation by MOF, an acetyltransferase essential for dosage 52. Bordoli L, Husser S, Luthi U et al. Functional analysis of the p300
compensation in Drosophila. Mol Cell 2000; 5(2):367-375. acetyltransferase domain: The PHD finger of p300 but not of CBP
32. Smith ER, Pannuti A, Gu W et al. The drosophila MSL complex is dispensable for enzymatic activity. Nucleic Acids Res 2001;
acetylates histone H4 at lysine 16, a chromatin modification linked 29(21):4462-4471.
to dosage compensation. Mol Cell Biol 2000; 20(1):312-318. 53. Kalkhoven E, Teunissen H, Houweling A et al. The PHD type zinc
33. Palmer MJ, Richman R, Richter L et al. Sex-specific regulation of finger is an integral part of the CBP acetyltransferase domain. Mol
the male-specific lethal-1 dosage compensation gene in Drosophila. Cell Biol 2002; 22(7):1961-1970.
Genes Dev 1994;8(6):698-706. 54. Kalkhoven E, Roelfsema JH, Teunissen H et al. Loss of CBP
34. Scott MJ, Pan LL, Cleland SB et al. MSL1 plays a central role in acetyltransferase activity by PHD finger mutations in
assembly of the MSL complex, essential for dosage compensation in Rubinstein-Taybi syndrome. Hum Mol Genet 2003; 12(4):441-450.
Drosophila. EMBO J 2000; 19(1):144-155. 55. Rea S, Eisenhaber F, O’Carroll D et al. Regulation of chromatin
35. Buscaino A, Kocher T, Kind JH et al. MOF-regulated acetylation of structure by site-specific histone H3 methyltransferases. Nature 2000;
MSL-3 in the Drosophila dosage compensation complex. Mol Cell 406(6796):593-599.
2003; 11(5):1265-1277. 56. Zhang X, Tamaru H, Khan SI et al. Structure of the Neurospora
36. Akhtar A, Zink D, Becker PB. Chromodomains are protein-RNA SET domain protein DIM-5, a histone H3 lysine methyltransferase.
interaction modules. Nature 2000; 407(6802):405-409. Cell 2002; 111(1):117-127.
37. Meller VH, Gordadze PR, Park Y et al. Ordered assembly of roX 57. Xiao B, Jing C, Wilson JR et al. Structure and catalytic mechanism
RNAs into MSL complexes on the dosage-compensated X of the human histone methyltransferase SET7/9. Nature 2003;
chromosome in Drosophila. Curr Biol 2000; 10(3):136-143. 421(6923):652-656.
38. Meller VH, Rattner BP. The roX genes encode redundant 58. Jacobs SA, Harp JM, Devarakonda S et al. The active site of the
male-specific lethal transcripts required for targeting of the MSL SET domain is constructed on a knot. Nat Struct Biol 2002;
complex. EMBO J 2002; 21(5):1084-1091. 9(11):833-838.
CHAPTER 34
M
Abstract
DM2 is an oncogene that is frequently over-expressed
in various human cancers, including sarcomas, glio-
sequence has revealed several structural and functional domains
(Fig. 1). These include the p53-binding domain located in the N
mas, melanomas, and breast cancers.1 The primary terminus between residues 19-102, the p300-binding domain
function of MDM2 is to inhibit the activity of the p53 tumor between residues 217-246, an acidic domain located between resi-
suppressor protein. P53 inhibits cell proliferation in response to dues 222-272, a zinc finger domain located between residues
DNA damage and other stresses by activating the transcription 289-331, and a RING finger domain located in the MDM2 C
of genes that mediate either cell-cycle arrest or apoptosis. MDM2 terminus between residues 438-478. RING fingers are zinc bind-
binds the transactivation domain of p53 and inhibits the ability ing domains with a defined octet of cysteine (C) and histidine
of p53 to activate transcription. Importantly, MDM2 binding (H) residues spaced at varying intervals in the coding sequence.11-14
also promotes the ubiquitin-mediated degradation of p53, as well In the RING H2 subtype, the cysteine and histidines are arranged
as the export of p53 from the nucleus to the cytoplasm. The in the order C3H2C3, while in the RING HC subtype the order
MDM2 gene itself is transcriptionally activated by p53, thus form- is C3HC4. The MDM2 RING finger is of the C3HC4 subtype
ing an autoregulatory feedback loop in which p53 promotes the and, like other RING fingers, is predicted to bind two zinc at-
expression of its own inhibitory factor. The ability of MDM2 to oms. RING finger domains are believed to function in the as-
promote the degradation and nuclear export of p53 depends on sembly and architecture of large protein complexes, including
an intact RING finger domain located in the MDM2 C termi- enzyme complexes of the ubiquitin system of protein degrada-
nus. This chapter will discuss the mechanisms by which MDM2 tion.15,16 As we will see below, the MDM2 protein can function
inhibits p53 function, with an emphasis on the requirement of as an ubiquitin-system E3 ligase that promotes the specific deg-
the MDM2 RING finger domain in p53 inhibition. radation of p53, and this activity requires the MDM2 RING
finger domain.
Introduction
The MDM2 gene was first discovered in 1987 as a gene that
was amplified in spontaneously transformed murine 3T3 fibro-
blasts.2 Subsequent studies revealed that the MDM2 protein in-
teracts strongly with the tumor suppressor protein p53 and in-
hibits p53 activity.3,4 It is now well-established that the primary
function of MDM2 is to inhibit p53.5,6 The importance of p53
inhibition by MDM2 is best illustrated through the attempted
generation of mice that lack MDM2. These MDM2 “knockout”
mice are not viable, dying at an early stage of embryonic develop-
ment.7,8 Importantly, however, the concomitant knockout of p53
allows survival of the mice. These results demonstrate that MDM2
normally inhibits p53 function, and that this inhibition is criti-
cally important during embryonic development.
Human MDM2 is a 491 amino acid protein that localizes
mostly in the nucleus, though it can also localize in the cyto-
plasm.6 The MDM2 gene has been mapped to chromosome Figure 1. The MDM2 protein. Full-length human MDM2 (also called
12q13-14 and includes several exons that encode the MDM2 HDM2) is a 491 amino acid protein. Positions of the nuclear localiza-
protein. Alternative splicing of the primary mRNA transcript can tion signal (NLS), nuclear export signal (NES), and nucleolar localiza-
result in the generation of different MDM2 isoforms, some of tion signal (NoLS) are indicated. Also indicated are the various func-
which inhibit the activity of the wild-type protein in a tional and structural domains of MDM2, including the RING-finger
dominant-negative fashion.9,10 Analysis of the MDM2 coding located in the C terminus.
Zinc Finger Proteins: From Atomic Contact to Cellular Function, edited by Shiro Iuchi and Natalie Kuldell.
©2005 Landes Bioscience/Eurekah.com and Kluwer Academic/Plenum Publishers.
MDM2: RING Finger Protein and Regulator of p53 253
MDM2 Promotes p53 Degradation zymes in turn may transfer the ubiquitin either directly to a sub-
In 1997, research groups led by Moshe Oren in Israel and strate or to the final class of enzymes known as ubiquitin protein
Karen Vousden in Frederick, Maryland were investigating the ligases (or E3s). The E3 enzymes catalyze the formation of an
binding interactions between MDM2 and p53.17,18 These stud- isopeptide bond between the carboxyl terminus of ubiquitin and
ies were complicated by the fact that p53 levels were consistently the ε-amino group of lysine residues on the target protein. A sub-
decreased with coexpression of MDM2 (see Fig. 2) as an example), strate may then be multiply ubiquitinated through the attach-
which made a quantitative analysis of the p53:MDM2 interac- ment of additional ubiquitin molecules to specific lysine residues
tion difficult. The researchers decided to investigate the molecu- of ubiquitin itself. In many cases the E1, E2, and E3 enzymes
lar basis for the decreased levels of p53. Small-molecule inhibi- involved form large, multi-protein complexes. This increases the
tors of the proteasome prevented the decrease in p53 levels efficiency of the process by allowing the rapid thiol-ester transfer
observed with MDM2 expression, suggesting that MDM2 was of ubiquitin molecules between proteins.
somehow promoting p53 degradation by the proteasome. More- In vitro studies with purified E1 enzyme, E2 enzyme, MDM2,
over, a high molecular weight ladder of p53 species was generated and p53 showed that MDM2 could form a direct thiol-ester link-
when p53 was coexpressed with MDM2 (see Fig. 2) as an ex- age with ubiquitin at Cys464 in the MDM2 RING finger do-
ample). The size of these p53 species was consistent with the ad- main, and could transfer the ubiquitin to p53.21 Removal of ei-
dition of multiple ubiquitin-moieties to p53, and analysis with ther the E1 or E2 enzymes prevented formation of the thiol-ester
anti-ubiquitin antibodies identified them as p53:ubiquitin con- bond between ubiquitin and MDM2, and also prevented the
jugates. These seminal studies provided the first evidence that transfer of ubiquitin to p53. Moreover, mutations that converted
MDM2 can promote the degradation of p53 through the Cys464 to Ala464 (C464A) also inhibited the transfer of ubiquitin
ubiquitin-proteasome pathway.17,18 to p53. Taken together, these data support a model in which
MDM2 functions as an E3 ubiquitin ligase that can accept acti-
vated ubiquitin from an E2 enzyme, and then transfer the
MDM2 Is an E3 Ubiquitin-Protein Ligase ubiquitin to p53 (Fig. 3).
The ubiquitin-proteasome system is the major pathway in the
cell for selective protein degradation.19,20 The covalent attach-
ment of multiple ubiquitin molecules to lysine residues of a tar-
get protein serves to signal its recognition and rapid degradation
by the 26S proteasome. Ubiquitination of a protein substrate re-
quires the concerted action of three classes of enzymes designated
E1, E2, and E3. The ubiquitin activating enzyme E1 initially
activates ubiquitin in an ATP-dependent reaction through the
formation of a thiol-ester bond between the carboxyl terminus of
ubiquitin and the thiol group of a specific cysteine residue of E1.
Ubiquitin is then transferred to a specific cysteine residue on one
of several ubiquitin-conjugating enzymes (Ubcs or E2s). E2 en-
The nature of p53 ubiquitination by MDM2 has not been P53 Is Stabilized and Its Levels Increase in
fully clarified. MDM2 can stimulate the formation of a “ladder”
of p53:ubiquitin conjugates in cells transiently expressing both
Response to Stress
proteins (see Fig. 2 as an example). This ladder could result from P53 levels increase in response to multiple genotoxic and
nongenotoxic stresses, including DNA damage, hypoxia, ribo-
the addition of single ubiquitin molecules to multiple lysine resi-
nucleotide depletion, and the abnormal activation of oncogene
dues in p53, or from the formation of poly-ubiquitin chains of
varying lengths on one or only a few lysines. Mutational analysis signaling pathways, among others.28-30 The effect of increasing
p53 is to inhibit cell proliferation through either cell cycle arrest
of p53, as well as in vitro studies using ubiquitin-aldehyde, sup-
or apoptosis (Fig. 5). These effects are mediated by p53-responsive
port the notion that MDM2 mono-ubiquitinates p53, transfer-
ring single ubiquitin moieties to multiple lysine residues located genes, such as p21 and bax. The importance of signaling either a
cell cycle arrest or apoptosis in response to these stresses is abun-
in the extreme C terminus of p53.22,23 However, the fact that
dantly clear. DNA damaging stresses have the potential to dam-
formation of poly-ubiquitin chains is necessary to target a pro-
tein to the proteasome24 indicates that a so called “E4” enzyme age chromosomes, alter genomic integrity, and cause mutations
that may activate oncogenes and/or inhibit tumor suppressors.
must exist that can promote p53 poly-ubiquitination. Recent stud-
By triggering a cell cycle arrest or apoptosis in response to DNA
ies suggest p300 may be such a factor. P300 is a large (approx.
300 kilodalton) protein that can bind the central region of MDM2 damaging stress, p53 prevents the propagation of cells that may
have an altered genome or acquired potentially cancer-promoting
(Fig. 1), as well as two separate regions of p53. While purified
mutations. Similarly, the mutational activation of oncogenes is
MDM2 can promote p53 mono-ubiquitination in an in vitro
system, addition of p300 stimulates the poly-ubiquitination of an early step in the development of most or all human cancers.
By triggering an arrest or apoptosis in response to the mutational
p53,25 probably through the formation of a p53:MDM2:p300
activation of oncogenes, p53 prevents the growth of cells that
trimeric complex. These and other studies indicate that p300 can
participate with MDM2 to promote p53 degradation by pro- would otherwise be prone to becoming cancerous. It is worth
noting that many chemotherapeutic drugs used in cancer treat-
moting p53 poly-ubiquitination25-27 (Fig. 4, left). However, it is
ment in some way modify DNA, resulting in the generation of
important to note that p300 can also function as an activator of
p53 by promoting p53 acetylation at one or more lysine residues single or double-stranded breaks. In cancers that retain wild-type
p53, these DNA breaks trigger an increase in p53 levels, and a
in the p53 C terminus (Fig 4, right). This acetylation increases
resulting activation of p53-induced cell death pathways.31,32 In
the DNA binding activity of p53, thus leading to an activation of
p53-responsive genes, and will be discussed in more detail below. contrast, the same thing does not occur in cancers that express
mutant p53 or in which the wild-type p53 gene has been de-
Regulatory interactions between p53, MDM2, and p300 remain
leted.31,32 This explains the increased sensitivity to chemothera-
both interesting and complex.
peutics observed in cancers with wild-type p53 compared with
cancers that lack wild-type p53 expression.
The increase in p53 levels following stress results from a stabi-
lization of the p53 protein.33,34 This is illustrated in Figure 6.
RKO cells (human colon cancer cells that express wild-type p53)
were untreated, or exposed to UV or ionizing (IR) radiation. Six
hours after radiation exposure, the cells were treated with
cyclohexamide (CHX) to inhibit de novo protein synthesis, and
p53 steady state levels were monitored at various time points af-
ter CHX addition. The rate at which p53 levels decrease under
these conditions is a measure of protein stability. In untreated
cells, p53 levels decrease rapidly following CHX treatment, with
a half-life (T1/2) of approximately 30 min. In contrast, p53 levels
Figure 4. Dual roles for p300 in regulating p53 levels and activity. Left)
MDM2 promotes the addition of single ubiquitin moieties to multiple
lysine residues in p53. p300 is believed to act in conjunction with MDM2
to promote the poly-ubiquitination of p53, and the subsequent degra-
dation of p53 by the proteasome. This process likely involves formation
of a trimeric complex between p53, MDM2 and p300 in which p300
binds simultaneously to MDM2 and a central region of p53. Right)
p300 can bind the N-teminus of p53 and promote the acetylation of
lysines in the p53 C terminus. This acetylation increases p53 DNA Figure 5. p53 levels increase in response to diverse stresses, resulting in
binding activity, leading to an activation of p53 responsive genes. cell cycle arrest or apoptosis.
MDM2: RING Finger Protein and Regulator of p53 255
not been demonstrated for the N-terminal NES. Despite the iden-
tification of these NESs, the molecular mechanisms that control
p53 nuclear export and the role of MDM2 in this process have
remained a matter of debate. Initial studies found that endog-
enous wild-type p53 accumulated in the nucleus and was stabi-
lized in cells exposed to leptomycin B (LMB), a specific inhibitor
of nuclear export.55,57 This suggested that nuclear export is re-
quired for p53 degradation. Further studies reported that MDM2
contains an NES within its sequence, and that blocking nuclear
export could inhibit the ability of MDM2 to promote p53 deg-
radation.57-59 Based on these findings a model was proposed in
which MDM2 transports p53 from the nucleus to the cytoplasm
for p53 to be degraded, and this transport depends on a func-
tional NES within MDM2. In conflict with this model were re-
ports that p53 contains its own NES, and that this NES is suffi-
cient to promote p53 nuclear export even in cells that lack MDM2
expression. 55 The NES identified in these studies was the
C-terminal NES located within the p53 tetramerization domain,
and structural analyses indicated that this NES would be inacces-
sible to the export machinery when p53 is a tetramer, but would
be exposed and accessible when p53 is in either a dimeric or
monomeric form. Taken together, these latter findings supported
a second model in which p53 is exported from the nucleus via its
own NES and independent of MDM2, and that this export may
be regulated by changes in p53 tetramerization.
In an attempt to clarify the role of MDM2 in p53 nuclear
export, cultured cells were transiently transfected with DNAs
Figure 8. Oncogene-mediated activation of p53. Activated oncogenes
stimulate expression of the tumor suppressor protein p14/Arf, which
stabilizes p53 by sequestering MDM2 in the nuclelous. P53 may then
be recruited to PML-nuclear bodies (PML-NBs) through its association
with PML. P300/CBP in the nuclear body can activate p53 by promot-
ing its acetylation.
MDM2 Interaction with Numb 13. Jackson PK, Eldridge AG, Freed E et al. The lore of the RINGs:
Substrate recognition and catalysis by ubiquitin ligases. Trends Cell
Numb is a regulator of neuronal differentiation that can inter- Biol 2000; 10:429-439.
act with the N terminus of MDM2. Interestingly, Numb levels 14. Borden KLB. RING domains: Master builders of molecular scaffolds.
are decreased in the presence of MDM2, suggesting that MDM2 J Mol Biol 2000; 295:1103-1112.
may target Numb for degradation.90 An effect of MDM2 on neu- 15. Lorick KL, Jensen JP, Fang S et al. RING fingers mediate
ronal differentiation has not been described. ubiquitin-conjugating enzyme (E2)-dependent ubiquitination. Proc
Natl Acad Sci USA 1999; 96:11364-11369.
16. Joazeiro CA, Weissman AM. RING finger proteins: Mediators of
MDM2 Interaction with PML ubiquitin ligase activity. Cell 2000; 102:549-552.
As mentioned earlier, PML is a tumor suppressor that can ac- 17. Haupt Y, Maya R, Kazaz A et al. MDM2 promotes the rapid
tivate p53 by recruiting it to multi-protein nuclear complexes degradation of p53. Nature 1997; 387:296-299.
termed PML-NBs.50 MDM2 colocalized with PML in PML-NBs 18. Kubbutat MH, Jones SN, Vousden KH. Regulation of p53 stability
by MDM2. Nature 1997; 387:299-303.
in cells that lacked p53 expression, and purified MDM2 protein 19. Hershko A, Ciechanover A. The ubiquitin system. Annu Rev
was shown to interact with a GST-PML fusion protein generated Biochem 1998; 67:425-479.
in bacteria. 91 These results indicate that MDM2 can interact 20. Ciechanover A, Schwartz AL. Ubiquitin-mediated degradation of
with PML in a p53-independent manner. Given that PML can cellular proteins in health and disease. Hepatology 2002; 35:3-6.
activate p53, it was speculated that the MDM2:PML interaction 21. Honda R, Tanaka H, and Yasuda H. Oncoprotein MDM2 is a
could indirectly affect the activity of p53. However, experiments ubiquitin ligase E3 for tumor suppressor p53. FEBS Lett 1997;
420:25-27.
to test this possibility have not yet been carried out. 22. Rodriguez MS, Desterro JM, Lain S et al. Multiple C-terminal lysine
residues target p53 for ubiquitin-proteasome-mediated degradation.
Conclusions Mol Cell Biol 2000; 20:8458-8467.
MDM2 is an oncoprotein that is frequently overexpressed in 23. Lai Z, Ferry KV, Diamond MA et al. Human mdm2 mediates
multiple mono-ubiquitination of p53 by a mechanism requiring
various human cancers. The primary function of MDM2 is to enzyme isomerization. J Biol Chem 2001; 27:357-367.
inhibit the tumor suppressor protein p53. MDM2 promotes the 24. Deveraux Q, Ustrell V, Pickart C et al. A 26S protease subunit that
ubiquitination of p53 and its degradation by the proteasome, as binds ubiquitin conjugates. J Biol Chem 1994; 269:7059-7061.
well as the export of p53 from the nucleus to the cytoplasm. Both 25. Grossman SR, Deato ME, Brignone C et al. Polyubiquitination of
of these activities require the RING finger domain located in the p53 by a ubiquitin ligase activity of p300. Science 2003;
MDM2 C terminus. DNA damage and other stresses stabilize 300:342-344.
26. Grossman SR, Perez M, Kung AL et al. p300/MDM2 complexes
p53 through an inhibition of MDM2-mediated degradation. participate in MDM2-mediated p53 degradation. Mol Cell 1998;
MDM2 interacts with a variety of other proteins in addition to 2:405-415.
p53, and these interactions may contribute to the oncogenic po- 27. Grossman SR. p300/CBP/p53 interaction and regulation of the p53
tential of MDM2. response. Eur J Biochem 2001; 268:2773-2778.
28. Prives C, Hall PA. The p53 pathway. J Pathol 1999; 187:112-126.
29. Ryan KM, Phillips AC, Vousden KH. Regulation and function of
References the p53 tumor suppressor protein. Curr Opin Cell Biol 2001;
1. Landers JE, Cassel SL, George DL. Translational enhancement of
13:332-337.
mdm2 oncogene expression in human tumor cells containing a
30. Woods DB, Vousden KH. Regulation of p53 function. Exp Cell
stabilized wild-type p53 protein. Cancer Res 1997; 57:3562-3568.
Res 2001; 264:56-66.
2. Cahilly-Snyder L, Yang-Feng T, Francke U et al. Molecular analysis
31. Lowe SW. Cancer therapy and p53. Curr Opin Oncol 1995;
and chromosomal mapping of amplified genes isolated from a
7:547-553.
transformed mouse 3T3 cell line. Somat Cell Molec Genet 1987;
32. Schmitt CA, Fridman JS, Yang M et al. A senescence program
13:235-244.
controlled by p53 and p16INK4a contributes to the outcome of
3. Momand J, Zambetti GP, Olson DC et al. The mdm-2 oncogene
cancer therapy. Cell 2002; 109:335-346.
product forms a complex with the p53 protein and inhibits
33. Maltzman W, Czyzyk L. UV irradiation stimulates levels of p53
p53-mediated transactivation Cell 1992; 69:1237-1245.
cellular tumor antigen in nontransformed mouse cells. Mol Cell Biol
4. Oliner JD, Pietenpol JA, Thiagalingam S et al. Oncoprotein MDM2
1984; 4:1689-1694.
conceals the activation domain of tumour suppressor p53. Nature
34. Maki CG, Howley PM. Ubiquitination of p53 and p21 is
1993; 362:857-860.
differentially affected by ionizing and UV radiation. Mol Cell Biol
5. Momand J, Wu H, Dasgupta G. MDM2-master regulator of the
1997; 17:355-363.
p53 tumor suppressor protein. Gene 2000; 242:15-29.
35. Giaccia AJ, Kastan MB. The complexity of p53 modulation:
6. Deb SP. Function and dysfunction of the human oncoprotein
Emerging patterns from divergent signals. Genes Dev 1998;
MDM2. Front Biosci 2002; 7:235-243.
12:2973-2983.
7. Jones SN, Roe AE, Donehower LA et al. Rescue of embryonic
36. Shieh SY, Ikeda M, Taya Y et al. DNA damage-induced
lethality in Mdm2-deficient mice by absence of p53. Nature 1995;
phosphorylation of p53 alleviates inhibition by MDM2. Cell 1997;
378:206-208.
91:325-334.
8. Montes de Oca Luna R, Wagner DS, Lozano G. Rescue of early
37. Chehab NH, Malikzay A, Stavridi ES et al. Phosphorylation of ser-20
embryonic lethality in mdm2-deficient mice by deletion of p53.
mediates stabilization of human p53 in response to DNA damage.
Nature 1995; 378:203-206.
Proc Natl Acad Sci USA 1999; 96:13777-13782.
9. Perry ME, Mendrysa SM, Saucedo LJ et al. p76(MDM2) inhibits
38. Unger T, Juven-Gershon T, Moallem E et al. Critical role for ser20
the ability of p90(MDM2) to destabilize p53. J Biol Chem 2000;
of p53 in the negative regulation of p53 by MDM2. EMBO J 1999;
275:5733-5738.
18:1805-1814.
10. Evans SC, Viswanathan M, Grier JD et al. An alternatively spliced
39. Chehab NH, Malikzay A, Appel M et al. Chk2/hCds1 functions as
HDM2 product increases p53 activity by inhibiting HDM2.
a DNA damage checkpoint in G(1) by stabilizing p53. Genes Dev
Oncogene 2001; 20:4041-4049.
2000; 14:278-288.
11. Deshaies RJ. SCF and cullin/RING H2-based ubiquitin ligases. Annu
40. Dumaz N, Meek DW. Serine 15 phosphorylation stimulates p53
Rev Cell Dev Biol 1999; 15:435-467.
transactivation but does not directly influence interaction with
12. Tyers M, Jorgensen P. Proteolysis and the cell cycle: With this RING
HDM2. EMBO J 1999; 18:7002-7010.
I do thee destroy. Curr Opin Genet Dev 2000; 10:54-64.
260 Zinc Finger Proteins: From Atomic Contact to Cellular Function
41. Ito A, Lai CH, Zhao X et al. p300/CBP-mediated p53 acetylation 66. Bosari S, Viale G, Roncalli M et al. p53 gene mutations, p53 protein
is commonly induced by p53-activating agents and inhibited by accumulation and compartmentalization in colorectal
MDM2. EMBO J 2001; 20:1331-1340. adenocarcinoma. Am J Pathol 1995; 147:790-798.
42. Khosravi R, Maya R, Gottlieb T et al. Rapid ATM-dependent 67. Ueda H, Ullrich SJ, Gangemi JD et al. Functional inactivation but
phosphorylation of MDM2 precedes p53 accumulation in response not structural mutation of p53 causes liver cancer. Nat Genet 1995;
to DNA damage. Proc Natl Acad Sci USA 1999; 96:14973-14977. 9:41-47.
43. Maya R, Balass M, Kim ST et al. ATM-dependent phosphorylation 68. Schlamp CL, Poulsen GL, Nork TM et al. Nuclear exclusion of
of Mdm2 on serine 395: Role in p53 activation by DNA damage. wild-type p53 in immortalized human retinoblastoma cells. J Natl
Genes Dev 2001; 15:1067-1077. Cancer Inst 1997; 89:1530-1536.
44. Lowe SW. Activation of p53 by oncogenes. Endocr Relat Cancer 69. Rodriguez-Lopez AM, Xenaki D, Eden TOB et al. MDM2 mediated
1999; 6:45-48. nuclear exclusion of p53 attenuates etoposide-induced apoptosis in
45. Lowe SW, Sherr CJ. Tumor suppression by Ink4a-Arf: Progress and neuroblastoma cells. Mol Pharmocol 2000; 59:135-143.
puzzles. Curr Opin Genet Dev 2003; 13:77-83. 70. Cordon-Cardo C, Latres E, Drobnjak M et al. Molecular
46. Weber JD, Taylor LJ, Roussel MF et al. Nucleolar Arf sequesters abnormalities of mdm2 and p53 genes in adult soft tissue sarcomas.
Mdm2 and activates p53. Nat Cell Biol 1999; 1:20-26. Cancer Res 1994; 54:794-799.
47. Zhang Y, Xiong Y. Mutations in human ARF exon 2 disrupt its 71. Dubs-Poterszman MC, Tocque B, Wasylyk B. MDM2 transformation
nucleolar localization and impair its ability to block nuclear export in the absence of p53 and abrogation of the p107 G1 cell-cycle
of MDM2 and p53. Mol Cell 1999; 3:579-591. arrest. Oncogene 1995; 11:2445-2449.
48. Borden KL. Pondering the promyelocytic leukemia protein (PML) 72. Lundgren K, Montes de Oca Luna R, McNeill YB et al. Targeted
puzzle: Possible functions for PML nuclear bodies. Mol Cell Biol expression of MDM2 uncouples S phase from mitosis and inhibits
2002; 22:5259-5269. mammary gland development independent of p53. Genes Dev 1997;
11:714-725.
49. Salomoni P, Pandolfi PP. The role of PML in tumor suppression.
73. Daujat S, Neel H, Piette J. MDM2: Life without p53. Trends Genet
Cell 2002; 108:165-170.
2001; 17:459-464.
50. Fogal V, Gostissa M, Sandy P et al. Regulation of p53 activity in
74. Davis PK, Dowdy SF. p73. Int J Biochem Cell Biol 2001;
nuclear bodies by a specific PML isoform. EMBO J 2000;
33:935-939.
19:6185-6195. 75. Jost CA, Marin MC, Kaelin WG Jr. p73 is a simian p53-related
51. Doucas V, Tini M, Egan DA et al. Modulation of CREB binding protein that can induce apoptosis. Nature 1997; 389:122-123.
protein function by the promyelocytic (PML) oncoprotein suggests 76. Yang A, Walker N, Bronson R et al. p73-deficient mice have
a role for nuclear bodies in hormone signaling. Proc Natl Acad Sci neurological, pheromonal and inflammatory defects but lack
USA 1999; 96:2627-2632. spontaneous tumours. Nature 2000; 404:99-103.
52. Ferbeyre G, de Stanchina E, Querido E et al. PML is induced by 77. Zeng X, Chen L, Jost CA et al. MDM2 suppresses p73 function
oncogenic ras and promotes premature senescence. Genes Dev 2000; without promoting p73 degradation. Mol Cell Biol 1999;
14:2015-2027. 19:3257-3266.
53. Pearson M, Carbone R, Sebastiani C et al. PML regulates p53 78. Gu J, Chen D, Rosenblum J et al. Identification of a sequence
acetylation and premature senescence induced by oncogenic ras. element from p53 that signals for Mdm2-targeted degradation. Mol
Nature 2000; 406:207-210. Cell Biol 2000; 20:1243-1253.
54. Pearson M, Pelicci PG. PML interaction with p53 and its role in 79. Dobbelstein M, Wienzek S, Konig C et al. Inactivation of the
apoptosis and replicative senescence. Oncogene 2001; 20:7250-7256. p53-homologue p73 by the mdm2-oncoprotein. Oncogene 1999;
55. Stommel JM, Marchenko ND, Jimenez GS et al. A leucine-rich 18:2101-2106.
nuclear export signal in the p53 tetramerization domain: Regulation 80. Flores ER, Tsai KY, Crowley D et al. p63 and p73 are required for
of subcellular localization and p53 activity by NES masking. EMBO p53-dependent apoptosis in response to DNA damage. Nature 2002;
J 1999; 18:1660-1672. 416:560-564.
56. Zhang Y, Xiong Y. A p53 amino-terminal nuclear export signal 81. Xiao ZX, Chen J, Levine AJ et al. Interaction between the
inhibited by DNA damage-induced phosphorylation. Science 2001; retinoblastoma protein and the oncoprotein MDM2. Nature 1995;
292:1910-1915. 375:694-698.
57. Freedman DA, Levine AJ. Nuclear export is required for degradation 82. Martin K, Trouche D, Hagemeier C et al. Stimulation of E2F1/
of endogenous p53 by MDM2 and human papillomavirus E6. Mol DP1 transcriptional activity by MDM2 oncoprotein. Nature 1995;
Cell Biol 1998; 18:7288-7293. 375:691-694.
58. Roth J, Dobbelstein M, Freedman DA et al. Nucleo-cytoplasmic 83. Hsieh JK, Chan FS, O’Connor DJ et al. RB regulates the stability
shuttling of the hdm2 oncoprotein regulates the levels of the p53 and the apoptotic function of p53 via MDM2. Mol Cell 1999;
protein via a pathway used by the human immunodeficiency virus 3:181-193.
rev protein. EMBO J 1998; 17:554-564. 84. Loughran O, Thangue NB. Apoptotic and growth-promoting activity
59. Tao W, Levine AJ. Nucleocytoplasmic shuttling of oncoprotein of E2F modulated by MDM2. Mol Cell Biol 2000; 20:2186-2197.
Hdm2 is required for Hdm2-mediated degradation of p53. Proc 85. Jackson MW, Berberich SJ. MdmX protects p53 from
Natl Acad Sci USA 1999; 96:3077-3080. Mdm2-mediated degradation. Mol Cell Biol 2000; 20:1001-1007.
60. Boyd SD, Tsai KY, and Jacks T. An intact HDM2 RING-finger 86. Parant J, Chavez-Reyes A, Little NA et al. Rescue of embryonic
domain is required for nuclear exclusion of p53. Nature Cell Biol lethality in Mdm4-null mice by loss of Trp53 suggests a
2000; 2:563-568. nonoverlapping pathway with MDM2 to regulate p53. Nat Genet
61. Geyer RK, Yu ZK, Maki CG. The MDM2 RING-finger domain is 2001; 29:92-95.
required to promote p53 nuclear export. Nature Cell Biol 2000; 87. Gu J, Kawai H, Nie L et al. Mutual dependence of MDM2 and
2:569-573. MDMX in their functional inactivation of p53. J Biol Chem 2002;
62. Gu J, Nie L, Wiederschain D et al. Identification of p53 sequence 277:19251-19254.
elements that are required for MDM2-mediated nuclear export. Mol 88. Marechal V, Ellenbaas B, Piette J et al. The ribosomal L5 protein is
Cell Biol 2001; 21:8533-8546. associated with mdm-2 and mdm-2-p53 complexes. Mol Cell Biol
63. Lohrum MA, Woods DB, Ludwig RL et al. C-terminal 1994; 14:7414-7420.
ubiquitination of p53 contributes to nuclear export. Mol Cell Biol 89. Elenbaas B, Dobbelstein M, Roth J et al. The MDM2 oncoprotein
2001; 21:8521-8532. binds specifically to RNA through its RING finger domain. Mol
64. Moll UM, Riou G, Levine AJ. Two distinct mechanisms alter p53 Med 1996; 2:439-451
90. Juven-Gershon T, Shifman O, Unger T et al. The Mdm2 oncoprotein
in breast cancer: Mutation and nuclear exclusion. Proc Natl Acad
interacts with the cell fate regulator Numb. Mol Cell Biol 1998;
Sci USA 1992; 89:7262-7266.
18:3974-3982.
65. Sun XF, Carstensen JM, Zhang H et al. Prognostic significance of
91. Wei X, Yu ZK, Ramalingam A et al. Physical and functional
cytoplasmic p53 oncoprotein in colorectal adenocarcinoma. Lancet interactions between PML and MDM2. J Biol Chem March 2003
1992; 340:1369-1373. (Epub ahead of publication).
CHAPTER 35
Abstract
plasm for subsequent utilization.13 Also in yeast, the Atx2 pro-
T
he ZIP family of transporters plays important roles in
supplying zinc to metalloproteins. These transporters are tein appears to transport manganese out of the Golgi.14
found in organisms at all phylogenetic levels including The biochemical mechanisms involved in these transport events
bacteria, fungi, plants, insects, and mammals. They have many have not been extensively investigated. In yeast, transport by Zrt1
conserved sequence elements and most have eight likely trans- and Zrt2 is known to be energy dependent.4,5 In contrast, zinc
membrane domains with similar predicted topologies. Intracel- uptake by the human hZip1 and hZip2 proteins was found to be
lular zinc homeostasis is facilitated by both transcriptional and energy independent.15,16 Zinc uptake by these proteins was not
post-transcriptional control mechanisms that regulate ZIP trans- dependent on K+ or Na+ gradients but hZip2 activity was stimu-
porter activity. These systems work in concert to provide a con- lated by HCO3- suggesting a Zn2+/HCO3- symport mechanism.15
stant supply of zinc in the face of changing extracellular levels Alternatively, zinc uptake by these proteins could be driven sim-
and limit overaccumulation of this potentially toxic metal. ply by the concentration gradient of labile zinc likely to exist across
the plasma membrane.17 It is also unclear if these proteins are
Introduction channels or carriers of Zn2+. Their high affinity for substrate (see
below) provides some support for the carrier model.
Charged ions like Zn2+ do not freely diffuse across lipid bi-
layer membranes. Transporter proteins are required to ferry Zn2+
into and out of cells and into and out of intracellular compart- Biochemical and Structural Aspects of Zip
ments. One group of proteins responsible for zinc uptake in many Transporters
different types of cells is the ZIP family of metal ion transport- ZIP transporters range in size from ~300 to over 600 amino
ers.1-3 ZIP transporters play an important role in supplying Zn2+ acids in length. The marked differences in lengths largely come
to metallate the various intracellular zinc binding sites that have from differences in the length of the N-terminal region before
been described throughout this book. the first transmembrane domain. Most ZIP proteins have eight
The “ZIP” designation stands for Zrt-, Irt-like Protein and transmembrane domains and similar predicted topologies (Fig.
reflects the first members of this transporter family to be identi- 1). Aspects of this topology model have been confirmed for
fied. Zrt1 and Zrt2 are the primary zinc uptake transporters in yeast18,19 and some mammalian ZIP transporters.15,16 While most
the yeast Saccharomyces cerevisiae and Irt1 is the major iron up- loops between transmembrane domains are quite short, a longer
take transporter in roots of Arabidopsis thaliana but has also been loop region is often found between transmembrane domains III
shown to transport zinc.4-7 Since the initial identification of these and IV. This region frequently contains a histidine-rich domain
proteins, the ZIP family has grown to over 90 members includ- with the sequence (HX)n where n generally ranges from 3-5. Al-
ing proteins in bacteria, nematodes, insects, and mammals.1 though conserved, the function of this domain is not yet clear.
Among these, ZupT is a zinc uptake transporter in Escherichia Mutation of this motif in the yeast Zrt1 protein resulted in slightly
coli 8 and several ZIPs have been implicated in zinc, iron, and/or reduced levels of the protein accumulating on the surface of the
manganese transport in plants (see below).9-12 Database analyses cell but did not alter zinc uptake affinity or activity when nor-
have indicated that there are 14 ZIP proteins encoded by the malized to surface protein levels.20
human genome. Human genes encoding ZIP proteins are referred Determinants of substrate specificity have been mapped to
to by their Human Genome Organization Nomenclature Com- the extracellular loop between TM II and III in Irt1.21 For some
ZIP proteins (for example the mammalian Liv-1 protein), this
mittee designation of “solute carrier family 39” (or SLC39).
region is also rich in histidine residues.22 Finally, some ZIP pro-
teins (for example hZip4) have long N-terminal domains that are
Functional Characteristics rich in potential metal-binding residues. These domains may play
With no exceptions yet known, ZIPs transport metal ion sub- a role in “harvesting” the substrate from the extracellular envi-
strates across cellular membranes into the cytoplasm. While many ronment prior to transport. Transmembrane domains IV and V
members are involved in the uptake of metal ions across the plasma are particularly amphipathic and contain conserved histidine resi-
membrane, some efflux metals from intracellular compartments. dues frequently with adjacent polar or charged amino acids. Given
For example, the yeast Zrt3 protein is responsible for transport- their sequence conservation and amphipathic nature, TM IV and
ing stored zinc out of the lysosome-like vacuole into the cyto- V are predicted to line a cavity in the transporter through which
Zinc Finger Proteins: From Atomic Contact to Cellular Function, edited by Shiro Iuchi and Natalie Kuldell.
©2005 Landes Bioscience/Eurekah.com and Kluwer Academic/Plenum Publishers.
262 Zinc Finger Proteins: From Atomic Contact to Cellular Function
25. Milon B, Dhermy D, Pountney D et al. Differential subcellular 31. Zhao H, Eide DJ. Zap1p, a metalloregulatory protein involved in
localization of hZip1 in adherent and nonadherent cells. FEBS Lett zinc-responsive transcriptional regulation in Saccharomyces cerevisiae.
2001; 507(3):241-246. Mol Cell Biol 1997; 17(9):5044-5052.
26. Yamaguchi S. Subtraction cloning of growth arrest inducible genes 32. Gomes DS, Fragoso LC, Riger CJ et al. Regulation of cadmium
in normal human epithelial cells. Kokubyo Gakkai Zasshi 1995; uptake by Saccharomyces cerevisiae. Biochim Biophys Acta 2002;
62:78-93. 1573(1):21-25.
27. Cao J, Bobo JA, Liuzzi JP et al. Effects of intracellular zinc depletion 33. Hicke L. Ubiquitin-dependent internalization and down-regulation
on metallothionein and ZIP2 transporter expression and apoptosis. of plasma membrane proteins. FASEB J 1997; 11:1215-1226.
J Leuk Biol 2001; 70(4):559-566. 34. Costello LC, Liu Y, Zou J et al. Evidence for a zinc uptake
28. Wang K, Pugh EW, Griffen S et al. Homozygosity mapping places transporter in human prostate cancer cells which is regulated by
the acrodermatitis enteropathica gene on chromosomal region 8q24.3. prolactin and testosterone. J Biol Chem 1999; 274:17499-17504.
Am J Hum Genet 2001; 68(4):1055-1060. 35. Cousins RJ, Blanchard RK, Popp MP et al. A global view of the
29. Wang K, Zhou B, Kuo YM et al. A novel member of a zinc selectivity of zinc deprivation and excess on genes expressed on genes
transporter family is defective in acrodermatitis enteropathica. Am J expressed in human THP-1 mononuclear cells. Proc Natl Acad Sci
Hum Genet 2002; 71(1):66-73. USA 2003; 100:6952-6957.
30. Kury S, Dreno B, Bezieau S et al. Identification of SLC39A4, a 36. Connolly EL, Fett JP, Guerinot ML. Expression of the IRT1 metal
gene involved in acrodermatitis enteropathica. Nature Genet 2002; transporter is controlled by metals at the levels of transcript and
31(3):239-240. protein accumulation. Plant Cell 2002; 14(6):1347-1357.
CHAPTER 36
Abstract
Z
inc is an essential trace element for all forms of life. Zinc highly adaptable to meeting the needs of proteins and enzymes
deficiency affects many systems because of the many roles that carry out diverse biological functions. These and many other
it encompasses, such as in metabolism (including the ac- chemical properties form the basis for the extensive participation
tivity of more than 300 enzymes), the structure of many pro- of zinc in protein, nucleic acid, carbohydrate and lipid metabo-
teins and control of genetic expression. Homeostatic regulation lism, as well as in the control of gene transcription and other
of this trace metal is extremely crucial, with zinc status affecting fundamental biological processes.
basic processes of cell division, growth, differentiation and The effects of zinc on cells is evidenced in the numerous roles
apoptosis. The role of zinc in the regulation of apoptosis is not this trace metal plays in basic cellular functions such as DNA
fully understood. The present review describes the current lit- replication, gene expression, cell division and cell activation. La-
erature surrounding zinc deprivation and the induction of bile pools of intracellular zinc are thought to be essential at sev-
apoptotic cell death, and attempts to dissect the various molecu- eral points in the cell cycle, since a number of enzymes involved
lar mechanisms involved in the process. in DNA synthesis such as thymidine kinase and DNA polymerase
are affected by availability of zinc.4 There is evidence that sug-
Introduction gests that zinc is required for a critical process in the mid G1
Two distinct forms of cell death, necrosis and apoptosis, are phase of the cell cycle where under conditions of adequate zinc
extremely well characterized by their different morphological and supply, cells can progress beyond this point to either replicate or
biological features. Apoptosis is regarded as the prominent form differentiate depending on other environmental factors. How-
of cell death during normal development, tissue homeostasis and ever, if zinc availability is limiting, the cells tend to undergo
in the regulation of the immune system. The phenomenon of apoptosis (Fig. 1). Therefore zinc may be involved in regulation
apoptosis in cells was first described by Kerr, Wyllie and Currie of cellular homeostasis by its dual role in both proliferation and
over 30 years ago, when they discovered morphological changes death by apoptosis. This chapter describes the current concepts
in cells undergoing cell death that lead to membrane blebbing of cell death and how zinc levels can affect the molecular mecha-
and digestion of the resulting vesicles by macrophages. Though it nisms of apoptosis.
was discovered in 1972, apoptosis took almost 20 years to come
into prominence in scientific research and only now can we ap-
preciate that apoptotic cells die in a highly coordinated manner
with a myriad of characteristic structural changes that include
chromatin condensation, nuclear DNA fragmentation and
externalization of phosphatidylserine on the surface of the cell
membrane. In the last few years there has been an exponential
growth in our understanding of the biochemical and molecular
processes underlying apoptotic pathways, a few of which will be
discussed in this chapter.
There is ample evidence that the induction and inhibition of
apoptosis in many cell types are thought to involve signal trans-
duction pathways in which the concentration of free intracellular
divalent metal ions seems to be important.1,2 The trace element
zinc appears to be particularly important in this regard and un-
like many other metals, zinc is virtually nontoxic.3 The homeo-
static mechanisms that regulate the entry into, distribution in,
and excretion from cells and tissues are so efficient that no disor-
ders are known to be associated with its excessive accumulation,
in contrast to iron and other metals. Furthermore, its physical Figure 1. Relationship of zinc to the cell cycle. Requirement for zinc
and chemical properties, including its general stable association during the mid-G1 phase of the cell cycle and its potential to cause cell
with macromolecules and its coordination flexibility, make it cycle exit by apoptosis.
Zinc Finger Proteins: From Atomic Contact to Cellular Function, edited by Shiro Iuchi and Natalie Kuldell.
©2005 Landes Bioscience/Eurekah.com and Kluwer Academic/Plenum Publishers.
266 Zinc Finger Proteins: From Atomic Contact to Cellular Function
Table 2.
the mitochondrial membrane potential (ψAm), which is required such as FADD/MORT1. Upon binding of their ligand (FasL and
to maintain asymmetric distribution of charges between the in- TNF respectively), a signal is transmitted to the cytoplasm that
ner and outer mitochondrial membrane.21,22 If disruption of the leads to recruitment of caspase-8 and possibly caspase-10, which
ψAm persists, the integrity of the outer membrane is compro- then becomes activated by proteolysis. Once activated, these proxi-
mised and results in the release of pro-apoptotic factors that in- mal caspases can also cleave and activate caspase-3 (Fig. 3).
clude cytochrome c23 and apoptosis inducing factor (AIF), which
can induce death independent of caspase activation.24 Release of Involvement of Zinc in Apoptotic Pathways
cytochrome c from the mitochondria into the cytosol has been The influence of zinc on apoptosis is well documented as both
demonstrated to be a critical initiator of caspase-dependent death in vitro and in vivo studies have shown that zinc influences
pathways. Once released, cytochrome c binds to Apaf-1 thereby apoptotic cell death.27-32 In particular, zinc supplementation in
triggering activation of caspase-9, which then cleaves and acti- thymocytes is known to inhibit apoptosis. In 1992, Cohen and
vates caspase-3 (Fig. 2). Activated caspase-3 then targets several colleagues observed that addition of zinc to dexamethasone-treated
proteins required for cellular integrity including Poly-ADP-ribose thymocytes (a potent inducer of the apoptotic program) in vitro
polymerase (PARP) and the inhibitor of calcium-activated blocked the transition from chromatin condensation to nuclear
DNAase (ICAD) which releases the endonuclease and enables it collapse. The authors attributed this finding to the specific inhi-
to cleave the DNA at internucleosomal sites.25 bition of the ICAD, which digests the DNA leading to the classic
One of the best-understood mechanisms for initiating the pro- DNA fragmentation hallmark of apoptosis.33 However a weak-
teolytic cascade of caspases involves the tumor necrosis factor ness of this study and many others using alternative cell types is
(TNF) family receptors.26 This form of apoptosis, triggered by that the concentrations of zinc required to prevent apoptosis are
external signals is the extrinsic or death receptor pathway. The extremely high, with most experimental works reporting preven-
cytosolic domains of some members of this family, such as Fas tion of DNA fragmentation with superfluous zinc concentrations
and the TNF receptor, are linked to caspases via adaptor proteins in the range of 250mM to 5mM.34 These are concentrations of
Figure 3. The Fas signal transduction pathway. When cell surface Fas
binds Fas ligand (FasL), several proteins become recruited to the intra-
cellular “death domain” of Fas to form a death-inducing signaling com-
Figure 2. Involvement of caspases in apoptosis. Activation of initiator plex, or DISC. Of these, the Fas-associated death domain (FADD)
caspase such as caspase-8, -9 and-10 can activate downstream caspase protein acts as an adaptor, linking pro-caspase-8 to the activated Fas.
members such as caspase-3, -6 and -7, which can then cleave down- This results in the proteolytic processing of pro-caspase-8 and activation
stream proteins associated with apoptosis. of the executioner caspases.
268 Zinc Finger Proteins: From Atomic Contact to Cellular Function
zinc that are 10-100 fold greater than the concentration found in intracellular apoptotic mediators other than the endonuclease
serum or tissues, albeit one that effectively blocks a wide array of came from revealing studies by Lazebnik and colleagues who
cell death inducers for a variety of cell types. Conversely, low con- showed that a more sensitive target of zinc was present in the
centrations of intracellular zinc have been shown to accelerate cytoplasm.46 In these studies, they demonstrated that cytosol from
apoptosis of lymphoid and myeloid cell lines35,36 as well as hu- cells primed to undergo apoptosis, induced chromatin condensa-
man leukemia lymphocytes, rat splenocytes and rat thy- tion and DNA fragmentation in isolated healthy nuclei and that
mocytes.37-39 Indeed, Telford et al suggested that a zinc concen- zinc suppressed these effects even when only added to the cyto-
tration below 300mM provides only limited protection against plasm. Subsequent studies by others demonstrated that this mys-
glucocorticoid-induced cell death. Concomitant with these stud- terious component was caspase-3. Although zinc did not block
ies numerous in vitro studies have now shown that depletion of the activity of caspase-3 to cleave its cellular substrates, it did
intracellular zinc by culture of cells in zinc-depleted medium or block the mechanism by which caspase 3 is activated from its
by treatment with the zinc chelator N’N’N’N-tetrakis-2 inactive precursor. Zinc deficiency induced apoptosis in vitro (that
pyridylmethyl-ethylenediamine (TPEN) results in apoptosis.37-39 is, using zinc chelators) appears to be dependent on caspase-3
Several in vivo models also provide evidence for adverse effects of activation since cytosolic caspase-3 activity is increased in
zinc deficiency in the whole animal.6,40 In particular, apoptosis zinc-deficient cells, while known inhibitors of caspase-3 (that is,
was shown to play an important role in the loss of precursor lym- z-DEVD-fmk) can partially suppress apoptosis, albeit at high
phocytes (lymphopenia) during zinc deficiency in mice.41 Con- concentrations. Interestingly, there is also evidence in vivo where
versely, there are also instances where addition of zinc can mark- zinc deficiency induced apoptosis in rat embryos was associated
edly increase apoptosis. Paramanantham et al used in vitro flow with increased caspase-3 activity.47 Many authors maintain that
cytometric studies to show that addition of zinc salts caused cell it is only the activation of caspase-3, which is the important de-
condensation and DNA fragmentation and suggested that zinc terminant as to whether a cell succumbs to apoptosis following
could enhance the generation of hydroxyl free radicals, results zinc deficiency. An interesting study by Chimienti et al demon-
which imply that zinc supplementation could induce features re- strated by western blot analysis that zinc depletion induced a strong
sembling apoptosis.42 Recently another group suggested that zinc activation of both caspase-3 and -8, but only a weak activation of
induces both necrosis and apoptosis in Molt-4 cells without caspase-9,48 which as suggested earlier may have resulted from
caspase-3 activation.43 Collectively however, all the data currently retroactivation by other caspases (Fig. 2). However, since the pro-
suggest that appropriate zinc homeostasis is essential to prevent teolytic processing and activation of effector caspases like caspase-3
apoptosis. However, we must appreciate that although generally is dependent on other caspases such as capase-8, -9 and caspase-6,
accepted to be a potent inhibitor of apoptosis, zinc can act as a these proteases should theoretically also be targeted during zinc
double-edged sword, where depending on its concentration and deficiency. Another study established that zinc depletion from
cell context, addition or depletion can induce apoptosis the use of chelators induced translocation of cytochrome c from
Several studies investigating zinc homeostasis imply that there the mitochondria with subsequent activation of caspases-3, -8
is a labile or exchangeable pool of intracellular zinc that is impor- and –9,49 thus it may be interesting for future experimentation if
tant in the regulation of apoptosis. Many have challenged this apical mediators of apoptosis are targeted by zinc depletion. One
hypothesis and suggest that this labile pool of zinc only consti- other possible target of zinc is caspase-6 which as stated earlier is
tutes a small part of the total cellular zinc, since most of the zinc also known to cleave and activate the proform of caspase-3. How-
in cells is tightly complexed to metalloenzymes and is not readily ever, studies have demonstrated that caspase-6 activation was typi-
accessible.44 Zinquin, a zinc-specific intracellular fluorophore has cally secondary to caspase-3 processing.50 Nevertheless, what ap-
been used to detect labile intracellular zinc. It is easily taken up pears clear from all these experiments is that the addition of zinc
by living cells and is retained intracellularly for several hours. It is salts following zinc deficiency can only effectively prevent late
essentially nonfluorescent until it is complexed with zinc. Zalewski stages of apoptosis such as DNA fragmentation, but appear rela-
et al investigated the subcellular localization of zinc using fluo- tively inefficient at inhibiting early events such as caspase activa-
rescence microscopy and suggested that zinc was largely restricted tion. Another question which warrants addressing is how impor-
to the extranuclear regions.45 However, zinquin probably only tant is the contribution of caspases to apoptosis potentiated by
detects the less tightly complexed zinc in cells, which would in- zinc deficiency? Given the limitation of most in vitro systems
clude free zinc and that which is loosely associated with cellular that are currently employed, several groups have made use of
proteins and lipids. The major pool of cellular zinc that is very gene-targeting techniques to generate mice that are deficient in
tightly bound to the active sites of enzymes and in zinc finger specific caspases to determine their physiological function.
transcription factors may not be available for interaction with Caspase-3-null mice have a perinatal lethality (though this is
zinquin. Therefore, interpretation as to which of the subcellular strictly strain dependent) and exhibit altered morphological and
pools of zinc are important in the suppression of apoptosis obvi- biochemical features when undergoing apoptosis.51 Various cell
ously requires more experimental evidence particularly in under- types such as mouse embryo fibroblasts (MEFs) from these mice
standing the interactions between zinc and known apoptotic regu- and other examples of caspase knockout mice challenged with
lators in these subcellular pools. zinc chelators such as TPEN may give us a better understanding
The precise step in the apoptosis signaling cascade that is sen- of the apoptotic pathways involved following zinc deficiency.
sitive to fluxes in zinc levels remains elusive, however there is a Important modulators of many apoptotic programs are the
growing body of evidence that points us in the direction of some members of the Bcl-2 family of anti-apoptotic proteins and dis-
of the putative targets of zinc in apoptosis. From a historical pro- turbances in the ratio of these proteins are indicative of whether a
spective, zinc was first shown to be an effective inhibitor of the cell is prone to die. Fukamachi and colleagues showed that zinc
endonuclease responsible for DNA fragmentation and indeed supplementation increased the Bcl-2/Bax ratio in the U937 mono-
following zinc deficiency it was this Ca/Mg-dependent enzyme cytic cell line, thereby increasing their resistance to cell death.34
that was targeted. The first evidence that zinc may affect other At this time, it has yet to be determined whether Bcl-2 levels are
Apoptosis by Zinc Deficiency 269
compromised following zinc deficiency, Ahn and colleagues have Conclusion and Perspective
speculated that Bcl-2 is not involved in apoptosis potentiated by This chapter has attempted to address the current research of
zinc deficiency, at least in their model of zinc-depleted induced how apoptosis can be potentiated by zinc deficiency. While re-
neuronal apoptosis.52 However, since these observations only in- search into the various mechanisms of apoptosis is at the fore-
corporated over-expression studies of Bcl-2, it is difficult to rec- front of scientific research, recognized with the 2002 Nobel Prize
oncile whether levels of Bcl-2 are truly an important determinant for Medicine awarded for contributions to the apoptosis field,
as to the outcome of zinc-depleted apoptosis. Nevertheless, oth- research into the cellular biology of zinc, its homeostasis and ef-
ers have speculated that the comparable phenotypes of Bcl-2 fects on apoptosis are still ambiguous. Clearly the best evidence
knockout mice to the zinc deficient mouse model (including, comes from the use of animal models and primary cells. Though
massive apoptotic involution of the thymus and spleen associ- nonetheless important to our understanding, in vitro studies are
ated with depletion of CD4+ T cells and growth retardation) are problematic in that it has been difficult to distinguish the physi-
not simply coincidental similarities.8 Another modulator of ological effects from the pharmacological and toxicological ef-
apoptosis is p53. A recent study by Fanzo et al indicated that fects of addition/suppression of zinc and the fact that immortal-
stress response genes, such as p53, are altered by zinc status. Here, izing events associated with cell line cultures usually involve
normal human bronchial epithelial cells cultured in zinc-deficient inactivation of the ARF-p53 tumor suppressor pathway, a path-
media resulted in elevated levels of both p53 mRNA and protein, way that has many ramifications on apoptosis. The utilization of
results which suggest that reductions in zinc may trigger the DNA cells from knock-out mouse models of apoptosis would help to
damage pathway to promote apoptosis.53 p53 is activated by a define and dissect out what the important mediators are in the
diverse number of signals that stress the cell, including apoptotic response following zinc deficiency. Furthermore, the
hyperproliferation, hypoxia, and DNA damage.54 Normally p53 development of better methods to assess intracellular zinc defi-
levels and function are harnessed by Mdm2,55 which programs ciency, coupled with the identification of which intracellular
p53 for destruction by initiating p53 ubiquitination and shut- pool(s) of zinc are important in mediating suppression of apoptosis
tling it to the cytosol for degradation by the proteasome. 56 Sig- is required.
nals that activate p53 disrupt this interaction, by inducing phos- One of the current challenges is to understand how regulators
phorylation of p5357,58 and/or Mdm2,59 or by inducing the ARF of the apoptosis pathway cooperate with metals such as zinc.
nucleolar tumor suppressor, which binds to and inhibits Mdm2 Adequate zinc levels are crucial for normal human development
functions.60,61 The net result is stabilization of p53 and activa- since clinical studies have revealed that the effects of even a mod-
tion of its transcription functions. In turn, p53 either induces G1 erate degree of zinc deficiency may be profound. Therefore, from
arrest by inducing transcription of the cyclin-dependent kinase a therapeutic perspective the ability to target a specific apoptotic
(cdk) inhibitor p21Cip1, allowing time for DNA repair, or elimi- pathway may ultimately prove to be a superior measure to mere
nates cells by triggering apoptosis. p53-induced apoptosis is zinc supplementation for individuals harboring clinical signs of
thought to occur through its ability to induce pro-apoptotic Bcl-2 zinc deficiency (such as those individuals with Acrodermatitis
family members, including Bax, Noxa and Puma.54 Several stud- Enterpathica).
ies have demonstrated that during apoptosis the cyclin-dependent
kinase (cdk) inhibitors p21Cip1 and p27Kip1 are cleaved by References
caspase-3.62 This observation was linked to zinc deficiency by 1. Golstein P, Ojcius DM, Young JD. Cell death mechanisms and the
Chai et al who showed that activation of caspase-3 in zinc-depleted immune system. Immunol Rev 1991; 121:29-65.
cells was rapidly followed by initial cleavage of p21Cip1 to a 15-kDa 2. Orrenius S, McCabe Jr MJ, Nicotera P. Ca(2+)-dependent
form, then to complete loss of p21Cip1 protein, perhaps mediated mechanisms of cytotoxicity and programmed cell death. Toxicol Lett
by proteosomal degradation.50 Moreover, one can infer that since 1992; 64-65(Spec No):357-364.
both p53 and Mdm2 require zinc fingers for their configuration, 3. Bertholf RL. Zinc. In: Seiler HG, Sigel H, eds. Handbook on toxicity
of inorganic compounds. New York: New York Dekker,
alterations in zinc homeostasis may cause the genes to become 1988:788-800.
mutated, resulting in their inactivation or suppression of their 4. Chesters JK. In: Mills CF, ed. Zinc in human biology. London:
function.63-65 Dysfunction/alterations in the p53 gene have par- Springer-Verlag, 1989:109-118.
ticularly been documented in head and neck (esophageal), and 5. Elmes ME. Apoptosis in the small intestine of zinc-deficient and
gynecological (breast) cancer patients.63,66 Interestingly, concomi- fasted rats. J Pathol 1977; 123(4):219-223.
tant dietary zinc deficiency has also frequently been found in 6. Yeiser EC, Vanlandingham JW, Levenson CW. Moderate zinc
deficiency increases cell death after brain injury in the rat. Nutr
esophageal cancer cohorts.67 Both correlate significantly with Neurosci 2002; 5(5):345-352.
poorer tumor prognosis and treatment outcome. Researchers have 7. Fraker PJ, Telford WG. A reappraisal of the role of zinc in life and
also discovered that not only do many esophageal cancer patients death decisions of cells. Proc Soc Exp Biol Med 1997;
harbor p53 mutations and acquire dietary zinc deficiency, but 215(3):229-236.
they also often have suppressed immune systems, which as indi- 8. Truong-Tran AQ, Carter J, Ruffin RE et al. The role of zinc in
cated earlier in the chapter, is demonstrated by a disturbance in caspase activation and apoptotic cell death. Biometals 2001;
14(3-4):315-330.
cytokine balance. There is a notable premature shift from Th-1 9. Kruse-Jarres JD. The significance of zinc for humoral and cellular
cytokines to Th-2 cytokines, with an accompanying poor T-cell immunity. J Trace Elem Electrolytes Health Dis 1989; 3(1):1-8.
function. The zinc-dependent hormone, thymulin, activates 10. Cunningham-Rundles S, Bockman RS, Lin A et al. Physiological
T-cell proliferation and function, and this same premature shift and pharmacological effects of zinc on immune response. Ann N Y
and T-cell dysfunction occurs with zinc deficiency. 68,69 Dissec- Acad Sci 1990; 587:113-122.
tion of this p53 damage pathway requires much more investiga- 11. Wellinghausen N, Kirchner H, Rink L. The immunobiology of zinc.
tion and consideration with regard to zinc deficiency. Immunol Today 1997; 18(11):519-521.
12. Dardenne M. Zinc and immune function. Eur J Clin Nutr 2002;
56(Suppl 3):S20-23.
270 Zinc Finger Proteins: From Atomic Contact to Cellular Function
13. Baum MK, Shor-Posner G, Campa A. Zinc status in human 36. Ning ZQ, Norton JD, Li J et al. Distinct mechanisms for rescue
immunodeficiency virus infection. J Nutr 2000; 130(5S from apoptosis in Ramos human B cells by signalling through CD40
Suppl):1421S-1423S. and interleukin-4 receptor: Role for inhibition of an early response
14. Neldner KH, Hambidge KM, Walravens PA. Acrodermatitis gene, Berg36. Biochem Soc Trans 1997; 25(2):306S.
enteropathica. Int J Dermatol 1978; 17(5):380-387. 37. Jiang S, Chow SC, McCabe MJ et al. Lack of Ca2+ involvement in
15. Hanna LA, Peters JM, Wiley LM et al. Enhancing effect of maternal thymocyte apoptosis induced by chelation of intracellular Zn2+. Lab
zinc deficiency and 137Cs gamma-irradiation on the frequency of Invest 1995; 73(1):111-117.
fetal malformations in mice. Teratog Carcinog Mutagen 1997; 38. Telford WG, Fraker PJ. Preferential induction of apoptosis in mouse
17(3):127-137. CD4+CD8+ alpha beta TCRloCD3 epsilon lo thymocytes by zinc.
16. Peters JM, Wiley LM, Zidenberg-Cherr S et al. Influence of J Cell Physiol 1995; 164(2):259-270.
short-term maternal zinc deficiency on the in vitro development of 39. Mathieu J, Ferlat S, Ballester B et al. Radiation-induced apoptosis
preimplantation mouse embryos. Proc Soc Exp Biol Med 1991; in thymocytes: Inhibition by diethyldithiocarbamate and zinc. Radiat
198(1):561-568. Res 1996; 146(6):652-659.
17. Hurley LS, Shrader RE. Abnormal development of preimplantation 40. Truong-Tran AQ, Ruffin RE, Foster PS et al. Altered zinc homeostasis
rat eggs after three days of maternal dietary zinc deficiency. Nature and caspase-3 activity in murine allergic airway inflammation. Am J
1975; 254(5499):427-429. Respir Cell Mol Biol 2002; 27(3):286-296.
18. Thornberry NA, Lazebnik Y. Caspases: Enemies within. Science 41. King LE, Fraker PJ. Zinc deficiency in mice alters myelopoiesis and
1998; 281(5381):1312-1316. hematopoiesis. J Nutr 2002; 132(11):3301-3307.
19. Fischer U, Janicke RU, Schulze-Osthoff K. Many cuts to ruin: A 42. Paramanantham R, Sit KH, Bay BH. Adding Zn2+ induces DNA
comprehensive update of caspase substrates. Cell Death Differ 2003; fragmentation and cell condensation in cultured human Chang liver
10(1):76-100. cells. Biol Trace Elem Res 1997; 58(1-2):135-147.
20. Coultas L, Strasser A. The role of the Bcl-2 protein family in cancer. 43. Hamatake M, Iguchi K, Hirano K et al. Zinc induces mixed types
Semin Cancer Biol 2003; 13(2):115-123. of cell death, necrosis, and apoptosis, in molt-4 cells. J Biochem
21. Vayssiere JL, Petit PX, Risler Y et al. Commitment to apoptosis is (Tokyo) 2000; 128(6):933-939.
associated with changes in mitochondrial biogenesis and activity in 44. Bettger WJ, O’Dell BL. A critical physiological role of zinc in the
cell lines conditionally immortalized with simian virus 40. Proc Natl structure and function of biomembranes. Life Sci 1981;
Acad Sci USA 1994; 91(24):11752-11756. 28(13):1425-1438.
22. Zamzami N, Marchetti P, Castedo M et al. Sequential reduction of 45. Zalewski PD, Forbes IJ, Betts WH. Correlation of apoptosis with
mitochondrial transmembrane potential and generation of reactive change in intracellular labile Zn(II) using zinquin
oxygen species in early programmed cell death. J Exp Med 1995; [(2-methyl-8-p-toluenesulphonamido-6-quinolyloxy)acetic acid], a
182(2):367-377. new specific fluorescent probe for Zn(II). Biochem J 1993; 296(Pt
23. Jiang X, Wang X. Cytochrome c promotes caspase-9 activation by 2):403-408.
inducing nucleotide binding to Apaf-1. J Biol Chem 2000; 46. Lazebnik YA, Cole S, Cooke CA et al. Nuclear events of apoptosis
275(40):31199-31203. in vitro in cell-free mitotic extracts: A model system for analysis of
24. Susin SA, Lorenzo HK, Zamzami N et al. Molecular characterization the active phase of apoptosis. J Cell Biol 1993; 123(1):7-22.
of mitochondrial apoptosis-inducing factor. Nature 1999; 47. Jankowski-Hennig MA, Clegg MS, Daston GP et al. Zinc-deficient
397(6718):441-446. rat embryos have increased caspase 3-like activity and apoptosis.
25. Janicke RU, Ng P, Sprengart ML et al. Caspase-3 is required for Biochem Biophys Res Commun 2000; 271(1):250-256.
alpha-fodrin cleavage but dispensable for cleavage of other death 48. Chimienti F, Seve M, Richard S et al. Role of cellular zinc in
substrates in apoptosis. J Biol Chem 1998; 273(25):15540-15545. programmed cell death: Temporal relationship between zinc depletion,
26. Peter ME, Krammer PH. The CD95(APO-1/Fas) DISC and beyond. activation of caspases, and cleavage of Sp family transcription factors.
Cell Death Differ 2003; 10(1):26-35. Biochem Pharmacol 2001; 62(1):51-62.
27. Fraker PJ, Haas SM, Luecke RW. Effect of zinc deficiency on the 49. Kolenko VM, Uzzo RG, Dulin N et al. Mechanism of apoptosis
immune response of the young adult A/J mouse. J Nutr 1977; induced by zinc deficiency in peripheral blood T lymphocytes.
107(10):1889-1895. Apoptosis 2001; 6(6):419-429.
28. Martin SJ, Mazdai G, Strain JJ et al. Programmed cell death 50. Chai F, Truong-Tran AQ, Evdokiou A et al. Intracellular zinc
(apoptosis) in lymphoid and myeloid cell lines during zinc deficiency. depletion induces caspase activation and p21 Waf1/Cip1 cleavage in
Clin Exp Immunol 1991; 83(2):338-343. human epithelial cell lines. J Infect Dis 2000; 182(Suppl 1):S85-92.
29. Migliorati G, Nicoletti I, Pagliacci MC et al. Interleukin-2 induces 51. Zheng TS, Hunot S, Kuida K et al. Caspase knockouts: Matters of
apoptosis in mouse thymocytes. Cell Immunol 1993; 146(1):52-61. life and death. Cell Death Differ 1999; 6(11):1043-1053.
30. Kuo IC, Seitz B, LaBree L et al. Can zinc prevent apoptosis of 52. Ahn YH, Koh JY, Hong SH. Protein synthesis-dependent but
anterior keratocytes after superficial keratectomy? Cornea 1997; Bcl-2-independent cytochrome C release in zinc depletion-induced
16(5):550-555. neuronal apoptosis. J Neurosci Res 2000; 61(5):508-514.
31. Marini M, Musiani D. Micromolar zinc affects endonucleolytic 53. Fanzo JC, Reaves SK, Cui L et al. Zinc status affects p53, gadd45,
activity in hydrogen peroxide-mediated apoptosis. Exp Cell Res 1998; and c-fos expression and caspase-3 activity in human bronchial
239(2):393-398. epithelial cells. Am J Physiol Cell Physiol 2001; 281(3):C751-757.
32. Maclean KH, Cleveland JL, Porter JB. Cellular zinc content is a 54. Vousden KH, Lu X. Live or let die: The cell’s response to p53. Nat
major determinant of iron chelator-induced apoptosis of thymocytes. Rev Cancer 2002; 2(8):594-604.
Blood 2001; 98(13):3831-3839. 55. Momand J, Zambetti GP, Olson DC et al. The mdm-2 oncogene
33. Cohen GM, Sun XM, Snowden RT et al. Key morphological features product forms a complex with the p53 protein and inhibits
of apoptosis may occur in the absence of internucleosomal DNA p53-mediated transactivation. Cell 1992; 69(7):1237-1245.
fragmentation. Biochem J 1992; 286(Pt 2):331-334. 56. Fuchs SY, Adler V, Buschmann T et al. JNK targets p53
34. Fukamachi Y, Karasaki Y, Sugiura T et al. Zinc suppresses apoptosis ubiquitination and degradation in nonstressed cells. Genes Dev 1998;
of U937 cells induced by hydrogen peroxide through an increase of 12(17):2658-2663.
the Bcl-2/Bax ratio. Biochem Biophys Res Commun 1998; 57. Banin S, Moyal L, Shieh S et al. Enhanced phosphorylation of p53
246(2):364-369. by ATM in response to DNA damage. Science 1998;
35. Trubiani O, Antonucci A, Palka G et al. Programmed cell death of 281(5383):1674-1677.
peripheral myeloid precursor cells in Down patients: Effect of zinc 58. Canman CE, Lim DS, Cimprich KA et al. Activation of the ATM
therapy. Ultrastruct Pathol 1996; 20(5):457-462. kinase by ionizing radiation and phosphorylation of p53. Science
1998; 281(5383):1677-1679.
Apoptosis by Zinc Deficiency 271
59. Maya R, Balass M, Kim ST et al. ATM-dependent phosphorylation 65. Ghosh M, Huang K, Berberich SJ. Overexpression of Mdm2 and
of Mdm2 on serine 395: Role in p53 activation by DNA damage. MdmX fusion proteins alters p53 mediated transactivation,
Genes Dev 2001; 15(9):1067-1077. ubiquitination, and degradation. Biochemistry 2003;
60. Pomerantz J, Schreiber-Agus N, Liegeois NJ et al. The Ink4a tumor 42(8):2291-2299.
suppressor gene product, p19Arf, interacts with MDM2 and 66. Zachos G, Spandidos DA. Transcriptional regulation of the c-H-ras1
neutralizes MDM2’s inhibition of p53. Cell 1998; 92(6):713-723. gene by the P53 protein is implicated in the development of human
61. Weber JD, Taylor LJ, Roussel MF et al. Nucleolar Arf sequesters endometrial and ovarian tumours. Oncogene 1998;
Mdm2 and activates p53. Nat Cell Biol 1999; 1(1):20-26. 16(23):3013-3017.
62. Levkau B, Koyama H, Raines EW et al. Cleavage of p21Cip1/Waf1 67. Prasad AS, Beck FW, Doerr TD et al. Nutritional and zinc status of
and p27Kip1 mediates apoptosis in endothelial cells through head and neck cancer patients: An interpretive review. J Am Coll
activation of Cdk2: Role of a caspase cascade. Mol Cell 1998; Nutr 1998; 17(5):409-418.
1(4):553-563. 68. Doerr TD, Marks SC, Shamsa FH et al. Effects of zinc and
63. Kihara C, Seki T, Furukawa Y et al. Mutations in zinc-binding nutritional status on clinical outcomes in head and neck cancer.
domains of p53 as a prognostic marker of esophageal-cancer patients. Nutrition 1998; 14(6):489-495.
Jpn J Cancer Res 2000; 91(2):190-198. 69. Prasad AS. Effects of zinc deficiency on Th1 and Th2 cytokine shifts.
64. Geyer RK, Yu ZK, Maki CG. The MDM2 RING-finger domain is J Infect Dis 2000; 182(Suppl 1):S62-68.
required to promote p53 nuclear export. Nat Cell Biol 2000;
2(9):569-573.
INDEX
A
Acetylation 11, 12, 29, 139-141, 145, 160, 169, 170, 176, Corepressor 134, 136, 139-142, 144, 145, 151, 154, 159,
183, 184, 187, 188, 201, 232, 236, 247-249, 254-256 169, 176, 185, 187, 222, 248
Aiolos 200-203 Crystal structure 3, 5, 15, 17, 22, 31, 35-37, 44, 47-49, 68,
Amino acid mutation 16, 81 91, 95, 96, 121, 122, 125, 129, 130, 134, 136, 141,
Androgen receptor 169, 171, 232-235 175, 176, 185-187, 197, 233, 244
Apoptosis 39, 44, 76-78, 106, 107, 110, 116-118, 134, Cysteine aspartate specific protease (Caspase) 235,
138, 141-144, 157, 171, 175, 178, 188, 217, 235, 252, 266-268
254, 258, 265-269
AreA 26-29 D
Arginine 9, 24, 33, 41, 52, 66, 69-71, 87, 93, 96, 129,
Damage recognition 239, 244
183, 216, 242
Deficiency 100, 144, 201-203, 239, 244, 262, 263, 265,
B 266, 268, 269
Development 7, 18, 29, 43, 56, 67, 78, 80, 88, 89, 99, 100,
Basonuclin 2 211, 212 101, 109, 116, 117, 121, 134, 138, 139, 141, 151,
Bcl-2 76, 266, 268, 269 152, 154, 160, 162, 171, 174, 177-179, 182, 183,
Bcl-6 134-136, 141-145 186-190, 195, 196, 200, 202-205, 208, 209, 212, 213,
BRCA1 107, 154 216-218, 221-227, 232, 250, 252, 254, 262, 265, 266,
Brg-1 203 269
BTB 134-142, 144, 145, 156, 157, 165 Differentiation 7, 27, 76, 78, 79, 99, 106, 107, 109, 110,
117, 134, 138-144, 156, 160-162, 171, 172, 177-179,
C 182, 187-190, 200-205, 207, 209, 210, 213, 219,
Cys2His2 (C2H2) zinc finger 1, 4, 5, 7-12, 14, 15, 20, 23, 221-227, 232, 258, 259, 265, 266
39-43, 47, 52, 66-71, 76-78, 80, 131, 134, 135, 137, Dimerization 5, 11, 28, 43, 51, 100, 109, 110, 122,
141, 142, 144, 151, 154, 156, 158-162, 164, 165, 168, 124-126, 129, 135, 136, 139, 141, 144, 157-159, 162,
174, 175, 177, 207, 208, 213, 214, 216, 218 165, 202, 233
CAG repeat 168, 235 Distance determinant 35, 37, 38
Cardiogenesis 223, 224 DNA binding 1-5, 7-12, 16-18, 20, 22-24, 26, 28, 29, 33,
Catalytic domain 35-38, 250 37-40, 42-44, 47, 48, 51, 56, 59, 60, 66-68, 72, 73,
Catalytic subunit 121, 122, 124, 125, 128 76, 78, 89, 91, 137, 139, 140, 142, 144, 152, 154,
Cell cycle 23, 43, 76, 107, 109, 110, 117, 118, 134, 138, 156, 159-163, 170, 174-177, 179, 182-189, 195, 196,
141, 142, 144, 145, 151, 188, 203-205, 254, 258, 265, 198, 200-204, 213, 216-218, 226, 232-234, 236, 241,
266 243, 244, 247, 249, 254, 255
Cell proliferation 39, 76, 78, 79, 109, 121, 171, 178, 188, DNA-binding domain (DBD) 3-5, 16, 20-24, 26-29, 35-37,
190, 212, 217, 218, 252, 254, 258, 269 47, 50, 51, 182, 183, 233, 234, 236
Cell survival 44, 121, 156 DNA glycosylase 31, 34
Chromatin remodeling 23, 29, 47, 109, 136, 169, 176, 183, DNA recognition 1-5, 16, 17, 20, 22, 26, 32, 36, 43, 44,
184, 200, 201, 205 48, 49, 51, 52, 136, 196
Chromatin structure 51, 52, 136, 153, 183, 186, 222, 247 DNA repair 43, 188, 198, 239, 269
Coactivator 114, 118, 161, 162, 169, 170, 176, 184, 185, DNase I footprint 16, 17, 59, 211
187, 222, 232, 234, 249 Double-stranded RNA (dsRNA) 66, 67, 76-78
Combinatorial regulation 185, 186 Drug design 39, 43, 44
Coordination chemistry 41, 43, 44 dsRBP-ZFa 67, 77, 78
274 Zinc Finger Proteins From Atomic Contact to Cellular Function
E I
E3 ubiquitin ligase 106, 107, 109, 110, 163, 253 Ikaros 11, 12, 200-205
Effector domain (ED) 47, 50-52, 116, 137, 267 Imprinting 12
Endocytosis 101, 131, 262, 263 Inhibition of cell growth 78
Endoderm 27, 221, 223-227
Endonuclease 32, 35-37, 267, 268 J
Endosome 128-132, 263 Just another zinc finger protein (JAZ) 67, 77, 78
Enigma 100-102
Equilibrium binding 56, 57, 59, 60, 62 K
Estrogen-responsive finger protein (Efp) 107, 109, 110
Evolutionary conservation 174, 223 κB-mediated transcription 218
Exonuclease III 16, 17, 32 Kaiso 12, 136, 139, 144, 145
KAP-1 151-154
F Kennedy’s disease 234
Keratinocyte 139, 144, 207-212
Folding 1-3, 5, 7, 9, 17, 40-42, 59, 62, 71, 80, 92, 94, 97, Kidney development 178
103, 122, 141, 153, 223, 233, 235, 247 KRAB domain 109, 151-154, 156, 159, 161, 163-165
Formamidopyrimidine DNA glycosylase (Fpg) 31
L
G
Leukemia 99, 106, 107, 110, 116, 117, 119, 134, 137-141,
GAGA 9, 20-24, 135, 136 177-179, 187, 189, 190, 203, 204, 218, 226, 248, 249,
GATA 20, 23, 26-29, 39, 103, 117, 138, 142, 143, 221-227, 268
241, 242, 249 LIM-homeodomain (LIM-HD) 99-104, 140, 176, 222, 248
Gene cluster 29, 161, 213 LIM-only (LMO) 99, 100, 104
Gene duplication 2, 8, 15, 29, 164 Linker 2, 3, 7, 8, 10-12, 14-18, 24, 35-38, 51, 52, 66, 68,
Gene silencing 142, 145, 151, 153, 154 70, 77, 78, 87, 93, 101, 102, 136, 151, 171, 175, 204,
Gene transcription 1, 11, 29, 44, 67, 81, 99, 109, 117, 216, 239, 240, 242, 255
136, 141, 168-171, 185, 187, 189, 190, 217, 218, 233, Lymphocyte development 189, 213
235, 265, 266 Lymphoma 76, 106, 119, 134, 141, 143, 144, 190, 203,
Glomerular nephropathy 179 204, 214, 218
Glomerular sclerosis 177, 178
Growth regulation 139 M
H Male-female sex reversal 177, 178
MDM2 76, 107, 118, 218, 252-259, 269
Hair follicle 207, 209, 210 Membrane trafficking 128, 131
HAT 170, 184, 248-250 Methylation 11, 12, 17, 144, 145, 153, 183, 184, 186, 187,
Heart 27, 109, 117, 143, 151, 154, 178, 221, 223, 224, 201
226, 227 Mi-2 201, 203
Hematopoiesis 139, 177, 179, 221-223 Miz-1 134, 145
Heterochromatin 23, 151, 153, 159, 203, 204 Modular design 1, 2
Heterochromatin protein 1 (HP1) 151, 153, 159 Motor neuron disease 232, 234
HIC-1 136, 144 Multiwavelength anomalous diffraction 31
Histone modification 183, 197, 201, 247, 250 Mutation 11, 16, 23, 27, 28, 31, 58, 59, 61, 69-71, 76, 84,
HIV 4, 41, 44, 50, 52, 67, 69-71, 109, 117, 184, 189, 213, 86, 87, 92-97, 100-102, 107, 109-111, 136, 139, 141,
214, 216, 217, 248, 249, 266 152, 168, 177, 178, 190, 202, 203, 217, 222, 224,
Holoenzyme 124, 125 226, 233-235, 240, 244, 249, 257, 261
Homeostasis 134, 141, 169, 178, 200, 203, 261, 262, 265, MutM 31-34
266, 268, 269 MYST family of acetylase 247
Homing 35-37 MZF1 158, 160-163
Human genome 1, 2, 4, 47, 51, 52, 66, 81, 128, 134, 151,
156, 157, 159, 164, 195, 261
Huntingtin 168-171
Hydrophobic interaction 7, 9, 11, 21, 114, 115, 121, 123,
124, 233
Hzf 78
Index 275
N R
N-terminal extension 20-23, 135 5S ribosomal RNA (5S rRNA) 14, 15, 56-63, 67-70, 177,
Nitric oxide (NO) 31, 39, 43, 44 195-198, 258
NMR 3, 4, 11, 17, 20, 22, 26, 29, 40, 43, 52, 58, 60, 72, 7S RNP 1, 5, 14, 15, 56, 57, 67, 197, 198
73, 82, 89, 100, 103, 115-118, 175, 239, 240 26S rRNA domain 94
Nuclear export 15, 18, 86, 87, 139, 168, 197, 252, 256, Rab GTPase 131
257 RBCC 108-111, 152
Nuclear localization 4, 14, 15, 51, 76-78, 81, 86, 87, 139, rDNA 94, 210, 211
145, 160, 168, 169, 175, 176, 183, 198, 200, 208, Receptor sorting 132
216, 233, 234, 236, 252, 256, 257 Recognition code 3, 22, 47-49, 51, 72
Nuclear Overhauser effect (NOE) 102, 239-243 Redox reaction 40, 42, 44
Nuclease footprinting 61, 62 Regulation 3, 4, 11, 12, 26, 28, 39, 40, 42, 44, 47, 51-53,
Nucleolar localization 252 56, 57, 66, 76-78, 81, 90, 99, 107-111, 116, 117, 121,
Nucleosome interaction 197 122, 124, 125, 130, 131, 134, 138-143, 145, 153, 156,
Nucleotide excision repair (NER) 239, 242, 244 160, 175, 176, 179, 183-188, 195-198, 200-202, 204,
NuRD 151, 153, 159, 201, 203 205, 209, 210, 212, 213, 216-219, 222, 223, 226, 232,
234, 248, 257, 262, 263, 265, 266, 268
O Regulatory subunit 109, 110, 121, 123-126, 128, 198
8-oxoguanine (GO) 31-33 Replication protein A (RPA) 239, 240, 244
Oncogenesis 99, 106, 107, 121, 189, 190, 217 Rev protein 69
Oxidative damage 31, 42 Ribosome 14, 56, 57, 66, 67, 91, 92, 94, 95, 97, 195, 197,
198, 258
P Ribosome biosynthesis 195, 197
RING finger 39, 106-111, 152, 252, 253, 256, 257, 259
p43 18, 56, 57, 60-63, 67-69, 72, 197 RNA binding 52, 56, 60-63, 66-73, 77, 78, 83-87, 89, 90,
p53 43, 44, 52, 67, 76-78, 107, 116-118, 144, 145, 169, 96, 176, 177, 195, 197, 249
176, 190, 218, 252-259, 269 RNA binding protein 14, 62, 66, 67, 69, 72, 73, 78, 89,
p53 target 52, 76 177
pag608 76, 78 RNA conformation 89
Particularly interesting new Cys-His (PINCH) 100, 102-104 RNA polymerase III 1, 14, 56, 195-197
Paxillin 100-102
Phage display 3, 11, 48, 49, 52, 66, 68-70, 72, 73 S
Phagosome 128-131
PHD 108, 109, 114, 128, 129, 153, 247, 250 7S storage particle 14
Phosphoinositide 128, 129 Saccharomyces cerevisiae 28, 52, 91-95, 121, 123, 128,
Phosphorylation 11, 12, 77, 90, 101, 121, 125, 128, 129, 134, 165, 168, 196-198, 261-263
131, 139, 140, 142, 160, 170, 174, 184, 187, 198, SCAN 151, 154, 156-165
201, 204, 205, 208, 216-218, 232, 255, 269 SCAN domain 151, 156-165
PI 3-kinase 128, 129 SCAND1 158, 161, 162
Plasma cell 142, 187, 203 Schnurri 213-219
PLZF 110, 134-141, 144, 145 Sclerosis 177, 178
Pol III 14, 18 Separated-paired C2H2 zinc finger 8
Polyglutamine 20, 24, 168-171, 232, 234-236 Separated-paired zinc finger 8, 11
Position effect variegation (PEV) 151, 204 Sequence repeat 15
POZ 20, 24, 134, 139, 156, 157, 165 Sequence specific DNA (ssDNA) binding 20, 139, 249
Proliferation 39, 76, 78, 79, 109, 121, 143, 161, 171, 178, Signal transduction 39, 43, 107, 116, 131, 134, 213, 217,
179, 188, 190, 203, 208, 210, 212, 213, 217-219, 221, 218, 255, 263, 265, 267
223, 227, 252, 254, 258, 265, 269 Site-directed mutagenesis 3, 5, 22, 31, 68, 69, 73
Protein kinase CK2 121 Solution NMR 20
Protein-DNA interaction 17, 37, 40, 66, 158, 195, 196 Splicing factor 67, 176
Protein-protein interaction 37, 38, 43, 99, 100, 109, Squamous epithelia 207, 209, 210
114-118, 123, 126, 132, 134, 135, 137, 141, 145, 153, Swi/Snf 201, 203
157, 159, 160, 169, 170, 175, 184-186, 188, 189, 196,
202, 216, 234, 248
276 Zinc Finger Proteins From Atomic Contact to Cellular Function
T X
TGF-β pathway 218 X chromosome 233, 247, 248
Thermus thermophilus (Tt) 31, 91 Xenobiotic 39, 41-43
Three-dimensional structure 1, 4, 5, 14, 15, 17, 20, 31, 36, Xenopus oocytes 1, 14, 67, 77, 187, 196, 198
42, 63, 86, 95, 97, 102, 106, 114, 115, 234 XPA 43, 239-244
Thymocyte differentiation 219 XPC 244
TIF1β 109, 151-154, 159
TNF-α pathway 213 Y
TPEN 263, 268 Yeast ribosomal protein YL37a 91
Transcription 1, 3, 4, 7-9, 11, 14-16, 18, 20, 23, 24, 26-29, Yin Yang 1 (YY1) 9, 11, 117, 182-190
39, 42-44, 47, 49-52, 56, 57, 66, 67, 71, 77, 80, 81,
94, 99, 100, 103, 106, 107, 109-111, 114-118, 123, Z
126, 131, 134, 136-139, 141-145, 151, 152, 154,
156-160, 162-165, 168-171, 175-179, 182-190, ZAS 11, 213-219
195-198, 201, 204, 207, 209-213, 216-218, 221-223, ZBRK1 154
226, 232-236, 239, 241, 247, 248, 252, 255, 258, 262, ZFR 78
265, 266, 268, 269 Zif268 3, 4, 7, 9, 10, 17, 21, 47-52, 71-73
Transcription activation 176, 186, 196, 198, 234 Zinc 1-5, 7-12, 14-18, 20-23, 26-29, 31-44, 47-52, 56, 57,
Transcription factor 1, 3, 4, 7-9, 14, 15, 18, 20, 23, 26-29, 59-63, 66-73, 76-78, 80, 81, 84-86, 89, 91-97,
39, 43, 44, 47, 49, 51, 52, 56, 57, 67, 66, 80, 81, 99, 99-104, 106, 107, 110, 114-118, 121-126, 128, 131,
103, 107, 109, 114-118, 123, 126, 134, 136, 138, 141, 132, 134-142, 144, 145, 151-154, 156, 158-165, 168,
156-160, 162, 164, 165, 168-171, 175-179, 182, 183, 171, 174-179, 182, 185-187, 195-198, 200, 202, 204,
185-187, 195, 196, 201, 207, 210-213, 217, 218, 205, 207-219, 221, 222, 226, 232-234, 239-244,
221-223, 226, 232, 234-236, 239, 241, 248, 258, 268 247-250, 252, 261-263, 265-269
Transcription factor IIIA (TFIIIA) 1-5, 7-10, 14-18, 20, 23, Zinc binding 5, 18, 20, 23, 27, 40-42, 44, 92, 93, 97, 100,
39-43, 52, 56-63, 67-70, 72, 177, 195-198, 250 103, 114, 115, 117, 121-124, 126, 240, 252, 261
Transcription initiation 14, 15, 18, 123, 183, 185, 195-197, Zinc coordination 39-44, 70, 78, 100, 240
232, 233, 262 Zinc finger 1-5, 7-12, 14-18, 20-23, 26, 28, 29, 31-44,
Transcriptional regulation 26, 47, 51, 76, 90, 109, 116, 47-52, 56, 57, 59-63, 66-73, 76-78, 80, 81, 84-86, 89,
117, 170, 175, 176, 179, 183-187, 201, 204, 205, 217, 91-97, 99-104, 106, 107, 110, 114, 122-124, 126, 128,
218 131, 132, 134-142, 144-149, 151-154, 156, 158-165,
Transcriptional repression 135-137, 139-142, 144, 145, 168, 171, 174-179, 182, 185-187, 195-198, 200, 202,
159, 162, 163, 169, 175, 176, 182, 184, 189 204, 205, 207-219, 221, 222, 226, 232-234, 239-242,
Transport 14, 161, 212, 235, 236, 247, 256 247-250, 252, 268, 269
Treble clef 36-39 Zinc finger motif 2, 3, 7, 15, 26, 31-34, 47, 52, 76-78, 80,
TRIM 108-110 84, 91-97, 99, 106, 114, 122, 124, 137, 139, 144, 145,
Tripartite motif 108 151, 156, 158, 159, 161, 162, 171, 241, 249
Zinc finger protein 1, 3, 5, 7-9, 11, 14, 15, 17, 18, 20,
U 39-44, 47, 56, 57, 60, 61, 66-72, 76-78, 80, 91, 134,
136, 137, 139, 144, 145, 151-154, 156, 159, 162-165,
Ubiquitin 106, 107, 109-111, 131, 136, 140, 160, 163, 168, 171, 174, 177, 179, 182, 197, 207, 212, 248
187, 188, 232, 235, 252-254, 257, 263 Zinc finger RNA binding 78
Uptake 42, 232, 261-263 Zinc ribbon motif 95, 121-126
V Zinc specificity 42
Zinquin 268
Vertebrate specific 164, 165 ZNF174 156-159, 161, 163
Viral infection 52, 121, 189, 190 ZNF197 158, 159, 161, 163
ZNF202 157, 158, 161-163
W Zyxin 99-102
Wig-1 67, 76-79
Wilms tumor 174, 177, 178