BACKGROUND: Recent studies have demonstrated that phenolic alkaloids from Menispermum dauricum (PAMD) can protect the heart and brain from ischemia/reperfusion injury, and promote neuron survival by inhibiting neuronal Bax and upregulating Bcl-2 expression following ischemia/reperfusion. OBJECTIVE: To investigate the neuroprotective effects of PAMD versus exogenous melatonin against ischemia/reperfusion injury. DESIGN, TIME AND SETTING: Observation and comparison experiments at a cellular level were performed at the Department of Biochemistry and Molecular Biology, Tongji Medical College of Huazhong University of Science and Technology between February 2007 and February 2008.MATERIALS: PAMD (95% purity) was provided by Kunming Institute of Botany, Chinese Academy of Sciences; melatonin was provided by Sigma, USA.METHODS: N2a mouse neuroblastoma cells were cultured in vitro deprived of glucose, serum and oxygen for 90 minutes, then cultured in normal medium containing different concentrations of PAMD (0.1, 1.0, 10 mg/L) or melatonin (1, 10, and 100 μmol/L). Cells cultured in normal conditions served as a control. MAIN OUTCOME MEASURES: The culture solution was collected to determine the content of excitatory neurotransmitters such as glutamic acid and aspartic acid; cell viability was detected by MTT methods; reactive oxygen species production was determined by fluorescence spectroscopy; mitochondrial transmembrane potential (?Ψm) was detected by laser confocal scanning; cytochrome C was measured by western blotting; and caspase-3 activity was determined by visible spectrophotometry. RESULTS: Melatonin and PAMD both promoted oxygen-glucose-serum deprivation-mediated N2a cell survival (P < 0.01) and inhibited glutamic acid release (P < 0.01), but melatonin did not inhibit aspartic acid production. The protective effects were the strongest using melatonin 100 μmol/L and PAMD 10 mg/L, so subsequent experiments were the performed at those doses. Although PAMD could no longer maintain mitochondrial transmembrane potential 6 hours after reperfusion, its inhibitory effects on cytochrome C release from mitochondria and scavengers of reactive oxygen species were stronger than those of melatonin (P < 0.01). However, its inhibitory effect on caspase-3 activity was weaker than that of melatonin: PAMD could inhibit caspase-3 activity 12 hours after reperfusion (P < 0.01), but melatonin inhibited caspase-3 activity 28 hours after reperfusion (P < 0.01). CONCLUSION: The results show that melatonin and PAMD have neuroprotective effects, but that the mechanisms are varied. Melatonin can maintain mitochondrial transmembrane potential, but its inhibitory effects on cytochrome C release, caspase-3 activity, and reactive oxygen species scavenging are different from those of PAMD.