Lithium and valproic acid: parallels and contrasts in diverse signaling contexts

Abstract

The introduction of lithium salts to treat bipolar disorder (BPD) revolutionized the therapy of psychiatric illnesses, but the pathogenesis of the disease and the mechanism of lithium action remain unknown. While several direct molecular targets of lithium have been identified, it is unknown which, if any, of these targets plays a role in the therapeutic response to lithium. Exposure to lithium evokes a wide spectrum of behavioral, physiological, and developmental responses in diverse organisms, and these effects have been exploited to explore the mechanisms of lithium action. Valproic acid (VPA), a widely used anticonvulsant, is also an effective therapy for BPD, and again its mechanism of action is not known, although new in vitro targets have been identified recently. In this review, the clinical, physiological, and developmental effects of lithium and VPA are summarized and recent work on direct targets for lithium and VPA is discussed in the context of these effects. We then describe some of the physiological effects common to the two drugs, in addition to treatment of BPD, and address signaling pathways that could be regulated by both lithium and VPA.

Introduction

The introduction of lithium salts for the treatment of bipolar disorder (BPD) by John Cade in 1949 revolutionized the therapy of this and other severe psychiatric illnesses. This landmark discovery not only brought an effective pharmacotherapy to this disorder for the first time, it also changed the public and scientific perception of psychiatric illnesses and triggered still ongoing attempts to define a pathophysiological mechanism for the origin and treatment of this common disorder (Manji et al., 1999a).

BPD is a prevalent and debilitating illness, affecting ∼1% of the population worldwide (Weissman et al., 1996). BPD is associated with significant morbidity and mortality, with suicide as a cause of death in as many as 10% of patients. Little is known about the pathogenesis of this or other affective disorders, but it responds remarkably well to mood-stabilizing drugs, such as lithium and the anticonvulsant valproic acid (VPA). More recently, other anticonvulsants, including carbamazepine (CBZ) and lamotrigine, have also been used in bipolar patients, and the effectiveness of these alternative regimens is reviewed in depth elsewhere (De Leon, 2001). In this review, we will focus on the physiological effects and potential molecular targets of lithium and VPA.

Despite their wide use in the pharmacotherapy of BPD, the mechanism of action of these drugs remains unknown. Both lithium and VPA have multiple actions in humans, model organisms, and cell culture systems, and these actions only partially overlap. Furthermore, a large number of indirect targets, as well as a more limited number of direct in vitro targets, for these two drugs has been described. The direct targets described in vitro offer plausible, but unproven, candidates to explain the in vivo effects of these drugs in various settings. In some cases, a strong argument can be made for inhibition of a given target molecule to explain in vivo drug effects. For example, inhibition of glycogen synthase kinase-3 (GSK-3) likely explains the effects of lithium on the developmental program of Xenopus embryos and Dictyostelium discoideum, inhibition of inositol polyphosphate 1-phosphatase (IPPase) provides a compelling explanation for the effect of lithium on the neuromuscular junction in Drosophila, and inhibition of phosphoglucomutase (PGM) is a likely explanation for the toxicity of lithium in yeast grown on galactose. However, most of the known, direct targets of lithium and VPA are ubiquitously expressed and are implicated as regulatory factors in diverse settings. Therefore, it is important to keep in mind that perturbation of a single molecular target is unlikely to explain all of the effects of either drug, and also that modulation of more than one direct target may be responsible for their efficacy in a complex disease process such as BPD Jope, 1999, Manji et al., 1996.

Several recent reviews have addressed the possible mechanisms of action for either lithium or VPA Baraban, 1994, Johannessen, 2000, Jope & Williams, 1994, Lenox et al., 1998, Lenox et al., 2002, Manji & Lenox, 1999, Stoll & Severus, 1996, Phiel & Klein, 2001, Tunnicliff, 1999, Williams & Harwood, 2000 and, in addition, numerous reviews and an excellent monograph are available concerning the indirect, as well as clinical, effects of these drugs (Goodwin & Jamison, 1990). Thus, we will only briefly discuss the clinical, physiological, and developmental effects of lithium and VPA to provide a context for the molecular discussion to follow. We will also summarize some of the recent work on direct, in vitro targets for lithium and VPA. Finally, physiological effects and potential signaling pathways that may be regulated by both drugs will be discussed, as these may reflect a common molecular pathway for the action of lithium and VPA in the treatment of BPD, and this would provide insight into the pathogenesis and pharmacotherapy of BPD.