We tested the direct effects of thromboxane A2/prostaglandin endoperoxide (TP) receptor agonists and antagonists on ischemic rat hearts to determine if any significant actions of TP may be occurring in a buffer-perfused system (without blood). Buffer-perfused rat hearts were treated with the TP antagonist SQ 30,741 (0.5-1.0 microM) during 15 min of ischemia and 30 min reperfusion. SQ 30,741 had no effect on severity of ischemia. In the same model, the TP receptor agonists U-46619 (0.01-1.0 microM) and SQ 26,655 (0.1 microM) reduced coronary flow and cardiac function both before and after ischemia. The decrease in contractile function appeared to be secondary to flow decrement. Despite the flow effects, U-46619 reduced ischemia-induced lactate dehydrogenase (LDH) release and contracture, indicating some beneficial effects. Measurement of prostacyclin release during reperfusion with and without U-46619 treatment showed that U-46619 significantly increased prostacyclin production. Meclofenamate (5.0 microM) did not reverse the vasoconstrictor and cardiodepressant effects of U-46619 but completely reversed its beneficial effect on LDH release. TP receptor blockade with 1.0 microM SQ 30,741 completely reversed the flow and cardiodepressant effects of SQ 26,655 but did not reverse the beneficial effects of this compound on LDH release. Receptor binding studies using [3H]-SQ 29,548 and [3H]-U-46619 indicated that few if any TP receptors exist in myocytes. In conclusion, TP antagonists are not cardioprotective in this model, but exogenous TP receptor agonists have complex actions in buffer-perfused hearts, some of which are mediated by vascular TP receptors and others which are not.