Perspectives on Neuroscience and Behavior.

The serendipitous discovery that lithium could treat bipolar disorder (BD) was published in 1949. In 1967, a published diagrammatic display of the clinical course of 88 patients with BP treated with lithium for one to six years depicted lithium’s extremely robust efficacy in preventing BP episodes, but there was also considerable variability across patients. In 1970, a five-month double-blind withdrawal study of lithium was published in which half of stable 50 patients with BD and 34 patients with recurrent depression were switched from lithium to placebo. Relapse occurred in 21 on placebo and none on lithium, which unequivocally demonstrated the robust efficacy of lithium in preventing relapse in BD and recurrent depression. Over the past 53 years, there have been an extensive number of studies attempting to discover the mechanism by which lithium produces such an important therapeutic effect. Now, in an outstanding and penetrating mechanistic study, it has been found that in two mouse models of ankyrin-G (AnkG) deficiency that displayed decreased dendritic complexity and decreased dendritic spine numbers in cortical neurons, lithium treatment corrected both abnormalities in both models. In the cortical neuron culture model with AnkG knockdown, a selective glycogen synthase kinase 3β (GSK3β) inhibitor rescued the spine morphology defects but not the dendritic complexity, and forskolin, which increases cAMP, rescued the dendritic complexity but not the spine morphology. A synergistic effect of both drugs was required to correct both the spine morphology and dendritic complexity (Piguel and others 2023). These findings are an important advance, since the ANK3 gene is linked to BD, single-nucleotide polymorphisms within the ANK3 regulatory domains have been found to be associated with lithium response in patients with BP, and mouse ANK3 knockout models have behavioral features like BD that respond to lithium treatment. Lithium directly or indirectly, through autoinhibition, acts to inhibit GSK3β, and it rescues several behavioral deficits like BD in ANKG knockout mice. In addition, lithium increases cAMP levels in frontal cortex. It should now be possible to assess the effects of lithium in individuals with BP who have genetic AnkG abnormalities to see if they have a more beneficial therapeutic response. Most important, the discovery that lithium’s mechanism of action involves both GSK3β inhibition and increased cAMP can help guide new research to discover alternatives to lithium, because lithium has so many toxic side effects.