We wished to determine if the previously observed cardioprotective effects of monophosphoryl lipid A (MLA, 65 micrograms/kg intravenously, i.v.), an endotoxin derivative, were time related and mediated by an enhancement of antioxidant defense mechanisms, i.e., myocardial catalase and superoxide dismutase (SOD) activities and neutrophil infiltration as assessed by myeloperoxidase (MPO) activity. We also wished to study the effect of pretreatment with MLA on vascular endothelial and smooth muscle function in vivo and in vitro. Barbital-anesthetized dogs were subjected to 60-min left circumflex coronary artery (LCX) occlusion followed by 5-h reperfusion. Myocardial catalase, SOD, and MPO activities were measured at the end of 5-h reperfusion. Pretreatment with MLA 24 h before ischemia produced a significant reduction in myocardial infarct size as measured by triphenyltetrazolium staining (15.3 +/- 4.4 vs. 30.9 +/- 5.2% in controls, p < 0.05), but 1-h pretreatment with MLA had no protective effect. MLA pretreatment for 24 h resulted in marked reduction (p < 0.05) in MPO activity in the border zone surrounding the infarct. Although a trend indicated an increase in catalase activity in the 24-h pretreatment group, no significant changes were observed in either catalase or SOD activities among the three groups. The cardioprotection produced by MLA was independent of differences in collateral blood flow to the ischemic region assessed by radioactive microsphere technique, systemic hemodynamics, myocardial oxygen demand, and ischemic bed size. Responses of the LCX bed to intracoronary acetylcholine (ACh) or nitroglycerin (NTG) in vivo and responses of isolated femoral artery rings to the endothelium-dependent vasodilators, ACh, A23187, bradykinin, or the nonendothelium-dependent vasodilator, sodium nitroprusside (SNP) in vitro were significantly decreased in the MLA 1-h pretreatment group but not in the 24-h pretreatment group. Incubation of the femoral artery rings from the MLA 1-h pretreatment group with 3 mM L-arginine for 1 h reversed the decreased endothelium-dependent responses to ACh and A23187, but not those to bradykinin. These results indicate that (a) the MLA-induced myocardial infarct size reduction was pronounced when MLA was administered for 24 h but was not evident at 1-h pretreatment; (b) a decrease in neutrophil infiltration into the site of ongoing tissue damage might be partially responsible for the protection; (c) vascular endothelial and/or smooth muscle function were transiently decreased by MLA administration and returned to nearly normal levels 24 h after treatment; and (d) the effect of MLA on endothelium-dependent responses might be mediated by the L-arginine/nitric oxide (NO) pathway.