Faculty Bibliography

Lithium inhibits glycogen synthase kinase-3 (GSK-3) at therapeutic concentrations; however, it is unclear if this inhibition and its downstream effects on specific signaling pathways are relevant to the treatment of bipolar disorder and depression. One of the targets of GSK-3 is the transcription factor beta-catenin. Normally active GSK-3 phosphorylates beta-catenin, leading to its degradation. Inhibition of GSK-3 therefore increases beta-catenin. We have utilized transgenic mice to investigate the behavioral consequences of CNS beta-catenin overexpression. Transgenic mice overexpressing beta-catenin demonstrated behavioral changes similar to those observed following the administration of lithium, including decreased immobility time in the forced swim test (FST). Further, we show that although acute administration of lithium and overexpression of the beta-catenin transgene inhibits d-amphetamine-induced hyperlocomotion, neither lithium nor the beta-catenin transgene prevents d-amphetamine-induced sensitization, as measured by locomotor activity. Both lithium-treated and beta-catenin mice had an elevated response to d-amphetamine following multiple administrations of the stimulant, though the difference in absolute locomotion was maintained throughout the sensitization time-course. Neither acute lithium nor beta-catenin overexpression had an effect on d-amphetamine-induced stereotyped behavior. The results of this study, in which beta-catenin transgenic mice exhibited behaviors identical to those observed in lithium-treated mice, are consistent with the hypothesis that the behavioral effects of lithium in these models are mediated through its direct inhibition of GSK-3 and the consequent increase in beta-catenin. By associating the behavioral effects of lithium with beta-catenin levels, these data suggest that increasing beta-catenin might be a novel therapeutic strategy for mood disorders.

Lithium attenuation of stimulant-induced hyperlocomotion is a rodent model that may be useful both to understand the mechanism of the therapeutic action of lithium and to develop novel lithium-mimetic compounds. To lay the foundation for future investigations into the neurobiology and genetics of lithium as a therapeutic agent, we studied the effect of lithium on d-amphetamine-induced hyperlocomotion in 12 (3 outbred) mouse strains. In our initial screening, mice received either (1) no drugs, (2) LiCl only, (3) d-amphetamine only, or (4) d-amphetamine and LiCl. Whereas there was no significant effect of LiCl alone on locomotion in any strain, there was a large degree of strain variation in the effects of LiCl combined with d-amphetamine. LiCl attenuated d-amphetamine-induced hyperlocomotion in C57BL/6J, C57BL/6Tac, Black Swiss, and CBA/J mice, whereas CD-1, FVB/NJ, SWR/J, and NIH Swiss mice, which were responsive to d-amphetamine, showed no significant effect of LiCl. d-Amphetamine-induced hyperlocomotion in the C3H/HeJ strain was increased by pretreatment with lithium. A subset of strains were treated for 4 weeks with lithium carbonate before the d-amphetamine challenge, and in each of these strains, lithium produced effects identical to those seen following acute administration. Strain responsiveness to lithium was not dependent upon the dose of either d-amphetamine or LiCl. Further, the results are not explained by brain lithium levels, which suggests that these behavioral responses to lithium are under the control of inherent genetic or other biological mechanisms specific to the effects of lithium on brain function.