Hemodynamic and electron spin resonance (ESR) analyses were performed on isolated ischemic and reperfused rat hearts to assess the cardioprotective and antioxidant effects of therapeutically relevant concentrations of Ginkgo biloba extract (EGb 761; 5, 50 or 200 microg/ml), its terpenoid constituents (ginkgolide A; 0.05 microg/ml and ginkgolide B; 0.05, 0.25 or 0.50 microg/ml), and a terpene-free fraction of EGb 761 (CP 205; 5 or 50 microg/ml). Hearts underwent 10 min of low-flow ischemia, 30 min of no-flow global ischemia, and 60 min of reperfusion. Test substances were added to the perfusion fluid during the last 10 min of control perfusion, low-flow ischemia and the first 10 min of reperfusion. A separate group of rats was treated with CP 205 (60 mg/kg/day; p.o.) for 15 days, after which the hearts were perfused with plain buffer. In ESR experiments, the spin-trap 5,5-dimethyl-1-pyrroline N-oxide (DMPO) was added to the perfusate to determine the effects of treatments on post-ischemic myocardial free radical generation. Results showed that in vitro exposure of hearts to EGb 761 (5 or 50 microg/ml) or to ginkgolides A and B (both at 0.05 microg/ml), or in vivo pretreatment of the rats with CP 205 delayed the onset of contracture during ischemia. The strong reperfusion-induced elevation of left ventricular end-diastolic pressure observed in untreated hearts was significantly reduced by in vitro exposure to the lowest concentrations of EGb 761, by ginkgolide A, and to a lesser extent by ginkgolide B, or by prior oral treatment with CP 205. Post-ischemic functional recovery.was significantly improved by in vivo administration of CP 205, by perfusion with 5 microg/ml of EGb 761 or with both terpenoids as compared to untreated group but in vitro CP 205 was not effective. ESR analyses revealed that DMPO-OH (the DMPO/hydroxyl radical spin-adduct) concentrations in coronary effluents were markedly decreased by all treatments, except for the lowest concentration of ginkgolide B. Perfusing 5 microg/ml EGb 761 resulted in a better inhibition of baseline DMPO-OH concentration than 5 microg/ml CP 205 (-70 % and -48 % vs. control, respectively), indicating that both terpenoid and flavonoid constituents of EGb 761 are required to produce this effect. CP 205 was significantly more efficient in reducing DMPO-OH concentration when administered in vivo than when applied in vitro, indicating that the antioxidant effect of flavonoid metabolites (formed in vivo) is superior to that of intact flavonol glycosides (present in vitro). Collectively, these findings provide the first evidence that part of the cardioprotection afforded by EGb 761 is due to a specific action of its terpenoid constituents and that this effect involves a mechanism independent of direct free radical-scavenging. Thus, the terpenoid constituents of EGb 761 and the flavonoid metabolites that are formed after in vivo administration of the extract act in a complementary manner to protect against myocardial ischemia-reperfusion injury.