Previous studies have proposed that oxygen radicals may play a role in the triggering of ischemic preconditioning. However, studies evaluating the effects of radical scavengers have yielded conflicting results, possibly because of differences in the number of preconditioning episodes used. The present study tested whether N-2-mercaptopropionylglycine (MPG) could block protection of both single and multiple episodes of preconditioning in in situ and in vitro rabbit hearts. All hearts were subjected to 30 min of regional ischemia followed by reperfusion for 2 (in vitro) or 3 (in situ) h. Infarct size was measured by tetrazolium. Infarction in control in situ hearts was 37.5+/-3.5% of the risk zone. A single cycle of preconditioning (PC1), with 5 min ischemia/10 min reperfusion, reduced infarct size to 12.3+/-2.0% (P<0.05). Four cycles of preconditioning (PC4) were equally protective. MPG (1 mg/kg/min i.v.) alone had no effect on infarction but abolished protection afforded by PC1 (35.4+/-3.9%). However, MPG failed to block protection in the PC4 group. In isolated control hearts, infarct size was 31.1+/-1.8% and was reduced to 10.2+/-2.2% (P<0.05) by preconditioning. MPG (300 microM) aborted protection. Infusion of hypoxanthine or xanthine oxidase separately in lieu of preconditioning had no effect on infarct size, but induced protection when combined (14.1+/-2.2%; P<0.05). Polymyxin B, an inhibitor of protein kinase C (PKC), abolished this protection (53.1+/-4.1%). In conclusion, oxygen radicals contribute to ischemic preconditioning in the rabbit and appear to do so via activation of PKC. The fact that MPG could not block protection by PC4 suggests that oxygen radicals act in concert with other triggers of preconditioning such as adenosine and bradykinin.