We have investigated the role of poly(ADP-ribose) polymerase (PARP) activation in rat brain in a model of sublethal transient global ischemia. Adult male rats were subjected to 15 min of ischemia with brain temperature reduced to 34 degrees C, followed by 1, 2, 4, 8, 16, 24, and 72 h of reperfusion. PARP mRNA expression was examined in the hippocampus using quantitative RT-PCR. northern blot analysis, and in situ hybridization. Protein expression was assessed using western blot analysis. PARP enzymatic activity was investigated by measuring nuclear [H-3]NAD incorporation. The presence of poly(ADP-ribose) polymers was assessed immunocytochemically. Although PARP mRNA and protein expressions were not altered after ischemia, enzymatic activity was increased 4.37-fold at 1 h (p < 0.05 vs. sham) and 1.73-fold (p < 0.05 vs. sham) at 24 h of reperfusion. Immunostaining demonstrated the presence of poly-(ADP-ribose) polymers in CA1 neurons. Cellular NAD(+) levels were not significantly altered at any time point. Furthermore, systemic administration of 3-aminobenzamide (30 mg/kg), a PARP inhibitor, prevented the increase in PARP activity at 1 and 24 h of reperfusion, significantly decreased the number of surviving neurons in the hippocampal CA1 region 72 h after ischemia (p < 0.01 vs. sham), and increased DNA single-strand breaks assessed as DNA polymerase I-mediated biotin-dATP nick-translation (PANT)-positive cells (p < 0.01 vs. sham). Furthermore, using an in vitro DNA repair assay, 3-aminobenzamide (30 mg/kg) was shown to block DNA base excision repair activity. These data suggest that the activation of PARP, without subsequent NAD(+) depletion, following mild transient ischemia may be neuroprotective in the brain.