Lithium has long been one of the primary drugs used to treat bipolar mood disorder. However, neither the etiology of this disease nor the therapeutic mechanism(s) of this drug is well understood. Several lines of clinical evidence suggest that lithium has neurotrophic actions. For example chronic lithium treatment increases the volume of gray matter and the content of N-acetyl-aspartate, a cell survival marker, in bipolar mood disorder patients (Moore et al., 2000). Moreover, treatment with this mood-stabilizer suppresses the decrease in the volume of the subgenual pre-frontal cortex found in bipolar patients (Drevets, 2001). To elucidate molecular mechanisms underlying the neuroprotective and neurotrophic actions of lithium, we employed a preparation of cultured cortical neurons prepared form embryonic rats. We found that treatment with therapeutic doses (0.2-1.2 mM) of lithium robustly protects cortical neurons from multiple insults, notably glutamate-induced excitotoxicity. The neuroprotection against glutamate excitotoxicity is time-dependent, requiring treatment for 5-6 days for maximal effect, and is associated with a reduction in NMDA receptor-mediated Ca2+ influx. The latter is correlated with a decrease in Tyrosine 1472 phosphorylation levels in the NR2B subunit of NMDA receptors and a loss of Src kinase activity which is involved in NR2B tyrosine phosphorylation. Neither the activity of total tyrosine protein kinase nor that of tyrosine protein phosphatase is affected by this drug, indicating the selectivity of the modulation. Lithium neuroprotection against excitotoxicity is inhibited by a BDNF-neutralizing antibody and K252a, a Trk antagonist. Lithium treatment time-dependently increases the intracellular level of BDNF in cortical neurons and activates its receptor, TrkB. The neuroprotection can be completely blocked by either heterozygous or homozygous knockout of the BDNF gene. These results suggest a central role of BDNF and TrkB in mediating the neuroprotective effects of this mood-stabilizer. Finally, long-term lithium treatment of cortical neurons stimulates the proliferation of their progenitor cells detected by co-labeling with BrdU and nestin. Lithium pretreatment also blocks the decrease in progenitor proliferation induced by glutamate, glucocorticoids and haloperidol, suggesting a role in CNS neuroplasticity. We used animal models to investigate further therapeutic potentials for lithium. In the MCAO/reperfusion model of stroke, we found that post-insult treatment with lithium robustly reduced infarct volume and neurological deficits. These beneficial effects were evident when therapeutic concentrations of lithium were injected at least up to 3 h after ischemic onset. The neuroprotection was associated with activation of heat-shock factor-1 and induction of heat-shock protein-70, a cytoprotective protein. In a rat excitotoxic model of Huntington's disease, the excitotoxin-induced loss of striatal medium-sized neurons was markedly reduced by lithium. This lithium protection was correlated with up-regulation of cytoprotective Bcl-2 and down-regulation of apoptotic proteins p53 and Bax, and neurons showing DNA damage and caspase-3 activation. Taken together, our results provide a new insight into the molecular mechanisms involved in lithium neuroprotection against glutamate excitotoxicity. Moreover, these novel molecular and cellular actions might contribute to the neurotrophic and neuroprotective actions of this mood-stabilizer in patients, and could be related to its clinical efficacy for treating mood disorder patients. Clearly, mood-stabilizers may have expanded use for treating excitotoxin-related neurodegenerative diseases.