Several hypotheses have been proposed for the pathophysiology of the congenital long QT syndrome, however, the underlying mechanism has not yet been elucidated. This study evaluated G protein function in patients with congenital long QT syndrome (LQTS) and compared it with that of normal subjects. Platelet-rich plasma was collected and the cyclic AMP (cAMP) level of platelets was measured in three conditions utilizing radioimmunoassay: the basal state (Basal cAMP), after stimulation by PGE1 (PGE1-cAMP), and after stimulation by PGE1 followed by inhibition by adrenaline (Adr-cAMP), and the results were compared between 7 LQTS patients and 10 healthy volunteers (control). Gs function was defined as (PGE1-cAMP)/(Basal cAMP) and Gi function as ¿(PGE1-cAMP)-(Adr-cAMP)¿/(PGE1-cAMP). Basal cAMP was lower in patients than in the controls: 2.9 +/- 0.6 pmol/ 10(8) cells vs. 4.2 +/- 0.7 pmol/10(8) cells (p < 0.05). The increase in cAMP after PGE1 was similar in the two groups but the peak was lower in the patients: 16.8 +/- 6.2 pmol/10(8) cells vs. 24.8 +/- 7.4 pmol/10(8) cells (PGE1-cAMP). After addition of adrenaline, cAMP decreased to 14.2 +/- 5.8 pmol/10(8) cells vs. 16.2 +/- 7.6 pmol/10(8) cells and the change was significantly smaller in the patients than in the controls: 0.17 +/- 0.12 vs. 0.38 +/- 0.16 (p < 0.05). Basal cAMP was weakly correlated with sinus cycle length (r = -0.48, p > 0.3) and QTc was correlated with Gs function (r = 0.52, p > 0.3) but not with Gi function. Patients with associated Torsade de Pointes had a significantly lower Gi function compared to those without (p < 0.05). In LQTS patients, G protein function was abnormal and the abnormality was associated with clinical characteristics of long QT syndrome. The relationship between the abnormal G protein function and the regulation of the repolarization of the ventricular myocardium needs to be studied further.