Lithium and certain anticonvulsants, including carbamazepine and valproic acid, are effective antimanic drugs for treating bipolar disorder, but their mechanisms of action remain uncertain.
Feeding rats lithium chloride for 6 weeks, to produce a brain lithium concentration of 0.7mM, reduced arachidonic acid turnover within brain phospholipids by 75%. The effect was highly specific, as turnover rates of docosahexaenoic acid and palmitic acid were unaffected. Arachidonate turnover in rat brain also was reduced by long-term valproic acid administration. Lithium's reduction of arachidonate turnover corresponded to its down-regulating gene expression and enzyme activity of cytosolic phospholipase A2, an enzyme that selectively liberates arachidonic but not docosahexaenoic acid from phospholipids. Lithium also reduced the brain protein level and activity of cyclooxygenase2, as well as the brain concentration of prostaglandin E2, an arachidonate metabolite produced via cyclooxygenase 2.
These results give rise to the hypothesis that lithium and antimanic anticonvulsants act by targeting parts of the "arachidonic acid cascade," which may be functionally hyperactive in mania. Thus, drugs that target enzymes in the cascade, such as cyclooxygenase 2 inhibitors, might be candidate treatments for mania. Also, in view of competition between arachidonic and docosahexaenoic acids in a number of functional processes, docosahexaenoic acid or its precursors would be expected to be therapeutic. Neither of these predictions is evident from other current hypotheses for the antimanic action of lithium and anticonvulsant drugs.