PKC, MAP kinases and the bcl-2 family of proteins as long-term targets for mood stabilizers

Transcription

1 (2002) 7, S46 S Nature Publishing Group All rights reserved /02 $ ORIGINAL RESEARCH ARTICLE PKC, MAP kinases and the bcl-2 family of proteins as long-term targets for mood stabilizers Laboratory of Molecular Pathophysiology, National Institute of Mental Health, Bethesda, MD, USA The complexity of the unique biology of bipolar disorder which includes the predisposition to episodic, and often progressive, mood disturbance and the dynamic nature of compensatory processes in the brain, coupled with limitations in experimental design, have hindered our ability to identify the underlying pathophysiology of this fascinating neuropsychiatric disorder. Although we have yet to identify the specific abnormal genes in mood disorders, recent studies have implicated critical signal transduction pathways as being integral to the pathophysiology and treatment of bipolar disorder. In particular, a converging body of preclinical data has shown that chronic lithium and valproate, at therapeutically relevant concentrations, regulate the protein kinase C signaling cascade. This has led to the investigation of the antimanic efficacy of tamoxifen (at doses sufficient to inhibit protein kinase C), with very encouraging preliminary results. A growing body of data also suggests that impairments of neuroplasticity and cellular resilience may also underlie the pathophysiology of bipolar disorder. It is thus noteworthy that mood stabilizers, such as lithium and valproate, indirectly regulate a number of factors involved in cell survival pathways including camp response element binding protein, brain derived neurotrophic factor, bcl-2 and mitogen-activated protein kinases and may thus bring about some of their delayed long-term beneficial effects via under-appreciated neurotrophic effects. The development of novel treatments, which more directly target molecules involved in critical central nervous system cell survival and cell death pathways, has the potential to enhance neuroplasticity and cellular resilience, thereby modulating the long-term course and trajectory of these devastating illnesses. (2002) 7, S46 S56. DOI: /sj/mp/ Keywords: neuroplasticity; cell survival; bipolar; manic; lithium; antidepressant; MAP kinase Introduction Although genetic factors play a major, unquestionable role in the etiology of bipolar disorder (BD), the biochemical abnormalities underlying the predisposition to, and the pathophysiology of, BD has yet to be elucidated fully. Early biological theories regarding the pathophysiology of BD have focused on various neurotransmitters, in particular, the biogenic amines. In recent years, however, advances in our understanding of the fundamental mechanisms underlying cell to cell communication have focused research into the role of post-receptor sites. Indeed, this general focus on postreceptor sites has been credited with playing a major role in the molecular medicine revolution, which has resulted in a more complete understanding of the etiology and pathophysiology of a variety of medical disorders. However, in contrast to the progress that has been made in elucidating the etiology and/or pathophysiology of a variety of medical conditions, we have Correspondence: HK Manji, Chief, Laboratory of Molecular Pathophysiology, National Institute of Mental Health, Building 49, Room B1EE16, 49 Convent Dr MSC 4405, Bethesda, MD 20892, USA. manjih intra.nimh.nih.gov yet to identify the specific abnormal genes or proteins in BD. The behavioral and physiological manifestations of BD are complex and must account not only for the profound changes in mood, but also for the constellation of neurovegetative and psychomotoric features. The pathophysiology is undoubtedly mediated by a network of interconnected limbic, striatal and frontocortical neurotransmitter neuronal circuits, 1 and it is thus not surprising that the brain systems which have heretofore received the greatest attention in neurobiological studies of BD have been the monoaminergic neurotransmitter systems. These systems were implicated by discoveries that effective antidepressant drugs exerted their primary biochemical effects by regulating intrasynaptic concentrations of serotonin and norepinephrine, and that antihypertensives which depleted these monoamines sometimes precipitated depressive episodes in susceptible individuals. 2 Furthermore, the monoaminergic systems are extensively distributed throughout the network of limbic, striatal and prefrontal cortical neuronal circuits thought to support the behavioral and visceral manifestations of mood disorders. 2,3 Thus, clinical studies over the past 40 years have attempted to uncover the specific defects in these neurotransmitter systems in BD by utilizing a variety

2 of biochemical and neuroendocrine strategies. Indeed, assessments of cerebrospinal fluid chemistry, neuroendocrine responses to pharmacological challenge and neuroreceptor binding have, in fact, demonstrated a number of abnormalities of the serotonergic, noradrenergic and other neurotransmitter and neuropeptide systems in major depressive disorder. 2,4 Table 1 summarizes the most salient findings in mood disorders. While such investigations have been heuristic over the years, they have been of limited value in elucidating the unique neurobiology of BD. Thus, BD arises from the complex interaction of multiple susceptibility (and likely protective) genes and environmental factors, and the phenotypic expression of the disease includes not only episodic and often profound mood disturbance, but also a constellation of cognitive, motoric, autonomic, endocrine and sleep/wake abnormalities. Furthermore, while most antidepressants exert their initial effects by increasing the intrasynaptic levels of serotonin and/or norepinephrine, their clinical antidepressant effects are only observed after chronic (days to weeks) administration, suggesting that S47 Table 1 Direct and indirect evidence implicating multiple systems in mood disorders: a role for more proximal abnormalities in signaling pathways? Serotonergic system Cholinergic system Reduced CSF 5-HIAA Depressogenic effect of cholinomimetics Blunted neuroendocrine and temperature responses to Enhanced cholinergic sensitivity 5-HT agonists Role in sleep EEG abnormalities (commonly observed in Reduced [ 3 H]IMI binding in platelets and postmortem MDD) brain Antimanic effects of cholinomimetics Reduced 5HT 1A receptor binding in living brain and postmortem brain tissue Glutamatergic system Antidepressant efficacy of agents that increase Stress increases Glu signaling intrasynaptic 5-HT Lithium facilitates Glu reuptake Depressogenic effects of Trp depletion in AD-treated Lamotrigine, which has antidepressant efficacy, patients decreases Glu release ADs decrease 5HT2 density, ECS increases Ketamine may have antidepressant effects Antidepressants chronically reduce NMDA receptor subunit expression Noradrenergic system Neuronal atrophy and dendritic reshaping (very indirect) Reduced CSF and urinary MHPG GABAergic system Elevated plasma NE Reduced CSF and plasma GABA Blunted neuroendocrine responses to clonidine? reduced Occ. cortex GABA Altered 2 AR and AR density and responsivity in Lithium, VPA ± antidepressants may increase GABA peripheral circulating cells signaling Altered densities of 2 ARs and ARs in areas of postmortem brain CRF and HPA axis Antidepressant efficacy of agents whose biochemical Hypercortisolemia and resistance to feedback inhibition effects include increasing NE Adrenal and pituitary hypertrophy Reduced internal jugular venoarterial NE metabolite Increased CSF CRF, and reduced CRF receptors in concentration gradients postmortem brain Depressogenic/anxiogenic effects of CRF agonists in preclinical models Dopaminergic system Hypercortisolemia normalized by successful Reduced CSF and HVA antidepressant treatment Blunted neuroendocrine and temperature responses to DA agonists Neurophysiology Antidepressant efficacy of agents whose biochemical Reduced CBF and metabolism in the dorsomedial and effects include increasing DA dorsal anterolateral PFCx Depressogenic effects of AMPT and reserpine in Elevated CBF and metabolism in the lateral Orb Cx and susceptible individuals anterior insula in MDD Reduced internal jugular venoarterial (DA metabolite Elevated amygdala CBF and metabolism in some concentration gradients) subtypes Depression in Parkinson disease Abnormal CBF and metabolism in subgenual and Prominent anhedonia and amotivation; role of DA in pregenual portions of the anterior cingulate gyrus and in reward and motivation circuits the posterior cingulate gyrus CSF, cerebrospinal fluid; 5-HIAA, 5-hydroxyindole acetic acid (a major serotonin metabolite); IMI, impramine; 5-HT, 5-hydroxytryptamine (serotonin); Trp, tryptophan, NE, norepinephrine, MHPG, 3-methoxy 4-hydroxyphenylglycol (a major NE metabolite); DA, dopamine; HVA, homovanillic acid (a major DA metabolite); AR, adrenergic receptor; AMPT, a-methyl-paratyrosine (an inhibitor of catecholamine biosynthesis); Glu, glutamate; NMDA, N-methyl-D-aspartate; VPA, valproate; CRF, corticotrophin releasing factor; PFCx, prefrontal cortex; CBF, cerebral blood flow; Orb Cx, orbital cortex. Adapted from Manji et al. 2

3 S48 a cascade of downstream effects are ultimately responsible for their therapeutic effects. These observations have led to the appreciation that while dysfunction within the monoaminergic neurotransmitter systems is likely to play important roles in mediating some facets of the pathophysiology of mood disorders, it likely represents the downstream effects of other, more primary abnormalities. 2,5,6 Most recently, research into the pathophysiology and treatment of BD has focused on intracellular signaling pathways. Multicomponent, cellular signaling pathways interact at various levels, thereby forming complex signaling networks that allow neurons to receive, process and respond to information, and to modulate the signal generated by multiple distinct, but interacting, neurotransmitter and neuropeptide systems. 7 These signaling pathways are undoubtedly involved in neuroplastic events, which regulate complex psychological and cognitive processes, as well as diverse vegetative functions, such as appetite and wakefulness. Consequently, recent evidence demonstrating that mood stabilizers exert major effects on selected intracellular signaling cascades, which regulate neuroplasticity and cellular resilience, have generated considerable excitement among the clinical neuroscience community, and are reshaping views about the neurobiological underpinnings of BD. In this article, we review these data and discuss their implications, not only for changing existing conceptualizations regarding the pathophysiology of BD, but also for delineating the biochemical properties of primary mood stabilizers and for strategic development of novel improved therapeutics. The protein kinase C (PKC) signaling cascade in the treatment of BD The inositol-depletion hypothesis posited that lithium, as an uncompetitive inhibitor of inositol-1-phosphatase, produced its therapeutic effects via a depletion of neuronal myoinositol levels. Although this hypothesis has been of great heuristic value, numerous studies have examined the effects of lithium on receptor-mediated phosphoinositide (PI) responses, and although some report a reduction in agonist-stimulated PIP 2 hydrolysis in rat brain slices following acute or chronic lithium, these findings have often been small and inconsistent, and subject to numerous methodological differences. 8 Most recently, a magnetic resonance spectroscopy study has demonstrated that lithium-induced myoinositol reductions are observed in the frontal cortex (FCx) of BD patients after only 5 days of lithium administration, at a time when the patients clinical state is completely unchanged. 9 Consequently, these and other studies suggest that, while inhibition of inositol-1-phosphatase may represent an initiating lithium effect, reducing myo-inositol levels per se is not associated with therapeutic response. This has led to the working hypothesis that some of the initial actions of lithium may occur with a relative reduction of myo-inositol. This reduction of myo-inositol initiates a cascade of secondary changes in the PKC signaling pathway and gene expression in the central nervous system (CNS), effects which are ultimately responsible for lithium s therapeutic efficacy. Indeed, evidence accumulating from various laboratories has clearly demonstrated that lithium, at therapeutically relevant concentrations, exerts major effects on the PKC signaling cascade Table 2 summarizes the most salient data implicating the PKC signaling cascade in the pathophysiology and treatment of BD. 11 The family of PKC isozymes is highly enriched in brain, and plays a major role in regulating both preand postsynaptic aspects of neurotransmission PKC isozymes are major intracellular mediators of signals generated on external stimulation of cells via a variety of neurotransmitter receptor subtypes, which induce the hydrolysis of membrane phospholipids PKC is now known to exist as a family of closely related subspecies, has a heterogeneous distribution in brain (with particularly high levels in presynaptic nerve terminals), and plays a major role in the regulation of neuronal excitability, neurotransmitter release and long-term alterations in gene expression and plasticity. The preponderance of the currently available data suggests that acute lithium may transiently activate PKC, whereas chronic lithium exposure results in an attenuation of phorbol ester mediated responses, which is accompanied by a down-regulation of PKC isozymes in brain. Using quantitative autoradiographic techniques, it was demonstrated that chronic lithium resulted in a significant decrease in membrane-associated PKC in several hippocampal structures, most notably the subiculum and CA1 region, in the absence of any significant changes in the various other cortical and subcortical structures examined. 17,18 Furthermore, immunoblotting using monoclonal anti-pkc antibodies revealed an isozyme-specific decrease in PKC and PKC (which have been particularly implicated in facilitating neurotransmitter release), in the absence of significant alterations in PKC, PKC, PKC or PKC. Similarly, exposure of immortalized hippocampal cells, neuroblastoma cells or PC12 cells to lithium (1.0 mm) in vitro also produced isozyme-selective decreases in PKC and/or PKC. 11 A strategy that has been utilized to investigate the downstream consequences of lithium-induced alteration in PKC isozymes is the examination of the effects of chronic lithium on Table 2 Effects of lithium and VPA on PKC signaling Lithium VPA PKC activity PKC PKC MARCKS levels Inositol responsive + Reprinted with permission from Manji and Lenox. 11

4 endogenous PKC substrates in brain. The most prominent substrate for PKC in brain is an acidic protein, myristoylated alanine-rich C-kinase substrate (MARCKS), which has been implicated in regulating long-term neuroplastic events. Lenox et al 19 demonstrated that chronic lithium administration dramatically reduced MARCKS expression in hippocampus effects that were not immediately reversed following lithium discontinuation. In the absence of suitable animal models for BD, a major problem inherent in neuropharmacological research has been the difficulty in precisely ascribing therapeutic relevance to any observed biochemical finding. One approach that has been utilized is the identification of common biochemical targets, which are modified by drugs belonging to the same therapeutic class (eg, antimanic agents), but possessing distinct chemical structures (eg, lithium and valproate (VPA, a substituted pentanoic acid)) when administered in a therapeutically relevant paradigm. Although they likely do not work by precisely the same mechanisms, identifying the biochemical targets, which are regulated in concert by these two chemically distinct agents, may provide important clues about molecular and cellular mechanisms underlying mood stabilization in the brain. In view of lithium s significant effects on PKC outlined above, the effects of VPA on various aspects of PKC functioning have also been investigated. It was found that the structurally highly dissimilar agent, VPA, produces strikingly similar effects on PKC and PKC isozymes, and on MARCKS, as does lithium (Table 3). 11,20 Interestingly, chronic lithium and VPA appear to regulate PKC isozymes and MARCKS by distinct mechanisms, with VPA s effects Table 3 PKC: pathophysiology and treatment of BD Kindling produces dramatic increases in membraneassociated PKC in hippocampus and amygdala Amphetamine produces increases in PKC activity and GAP-43 phosphorylation (implicated in neurotransmitter release) PKC inhibitors block the biochemical and behavioral responses to amphetamine and cocaine, and also block cocaine-induced sensitization Dexamethasone administration increases PKC activity and increases the levels of PKC and PKC in rat FCx and hippocampus Increased membrane/cytosol PKC partitioning in platelets from manic subjects; normalized with lithium treatment Increased PKC activity and translocation in BD brains compared with controls Lithium and VPA regulate PKC activity, PKC, PKC and MARCKS Preliminary data suggest that PKC inhibitors may have efficacy in the treatment of acute mania PKC, protein kinase C; GAP, growth cone associated protein; FCx, frontal cortex; MARCKS, myristoylated alanine rich C kinase substrate. Adapted from Manji and Lenox. 11 appearing to be largely independent of myo-inositol. 11 This biochemical observation is consistent with the clinical observations that some patients show preferential response to one or other of the agents, and that one often observes additive therapeutic effects in patients when the two agents are co-administered. In view of the pivotal role of the PKC signaling pathway in the regulation of neuronal excitability, neurotransmitter release, and long-term synaptic events, 12,21 it was postulated that the attenuation of PKC activity may play a role in the antimanic effects of lithium and VPA. In a pilot study, it was found that tamoxifen (a non-steroidal antiestrogen known to be a PKC inhibitor at higher concentrations) 22 may indeed possess antimanic efficacy. 23 Clearly, these results have to be considered preliminary due to the small sample size thus far. In view of the preliminary data suggesting the involvement of the PKC signaling system in the pathophysiology of BD (Table 2), these results suggest that PKC inhibitors may be agents very useful in the treatment of mania. In the absence of more selective, clearly CNS-penetrant PKC inhibitors currently available for human use, larger, double-blind, placebo-controlled studies of tamoxifen in the treatment of mania are currently underway. Long-term molecular mechanisms underlying mood stabilization in BD: the identification of the cytoprotective protein bcl-2 as a novel target The therapeutic effects of mood stabilizers in the treatment of BD are only seen after chronic administration, thereby precluding any simple mechanistic interpretation based on their acute biochemical effects. Patterns of effects requiring such prolonged administration of the drug suggest that the therapeutic effects involve the strategic regulation of gene expression in critical neuronal circuits. 10,13,18,24 In this context, it is noteworthy that substantial progress has been made in recent years, both in identifying the genes responsive to transsynaptic stimulation, and in elucidating the processes that convert sometimes ephemeral second messengermediated events into long-term cellular phenotypic alterations. This has been particularly important for neurobiology, wherein we attempt to understand the mechanisms by which short-lived events (eg, stressors) can have profound, long-term (perhaps life-long), behavioral consequences, 25 and more importantly for the present discussion, to help unravel the processes by which seemingly simple molecules, including monovalent cations (eg, lithium) and fatty acids (eg, valproic acid) may produce a long-term stabilization of mood in individuals vulnerable to BD. There have, however, been several impediments in our attempts to fully understand the molecular and cellular mechanisms of action of mood stabilizers. 18 Firstly, a suitable experimental model of BD is not currently available, and many studies are of necessity thus conducted in normal rodents, with the view that the targets identified may have evolutionary-conserved functions and, therefore, therapeutic relevance in the S49

5 S50 treatment of this very complex human illness. 26 In this context, it is likely that the animal models of drug dependence have been instrumental in accelerating the pace of research on their molecular mechanisms. 24 Secondly, another problem inherent in the identification of therapeutically relevant target genes for the actions of mood stabilizers is the relative paucity of easily detectable phenotypic changes induced by these agents. 26 This makes the task of ascribing functional significance to the multiple treatment-induced changes at the genomic level quite daunting. Moreover, the genetic basis of mood as a quantitative trait is still at its inception 27 and we therefore cannot focus on a group of already-known genes. Finally, as previously alluded to, there is a real dearth of knowledge concerning the underlying etiology and pathophysiology of what is likely a group of complex, heterogeneous disorders that shows overlap of symptom clusters, and is subsumed under the rubric of manic-depressive illness or bipolar disorder. 28 Despite these many formidable obstacles, there is currently considerable excitement about the progress that is being made, using two fundamental strategies to identify changes in gene expression that may have therapeutic relevance in the long-term treatment of mood disorders. 18 Firstly, investigators have been focusing upon the known primary biochemical targets for the actions of mood stabilizers (eg, inositol monophosphatases), and have been subsequently identifying alterations in downstream signaling cascades, transcription factors and, ultimately, the expression of genes known to be regulated by these primary biochemical targets. Secondly, several technological advances are allowing more black-box screening approaches to be increasingly utilized; these approaches attempt to focus directly on changes in gene expression produced by the administration of mood stabilizers in therapeutically meaningful paradigms, without necessarily focusing on the initiating biochemical events, ie, the medications primary biochemical target. Using screening methods, such as subtractive hybridization, microarrays and mrna differential display, this strategy usually attempts to simultaneously identify treatment-induced changes in multiple, often thousands, of genes, without any a priori focus on specific candidate genes. Both these strategies require an initial reductive step which attempts to isolate the specific genes and proteins that are the targets of mood-stabilizing agents and, ideally, a subsequent integrative step that attempts to establish the relationship between the molecular/cellular changes and certain facets of the therapeutic response. 26 Using such a strategy, recent mrna reverse transcriptase-polymerase chain reaction differential display studies have led to the identification of a completely unexpected target for the actions of chronic lithium and VPA in the FCx the cytoprotective protein bcl-2. 13,29,30 Chronic treatment of rodents with therapeutic doses of lithium and VPA was found to produce a doubling of bcl-2 levels in FCx, effects primarily due to a marked increase in the number of bcl-2 immunoreactive cells in layers II and III of the FCx. Interestingly, the importance of neurons in layers II IV of the FCx in mood disorders has recently been emphasized, since primate studies have indicated that these are important sites for connections with other cortical regions, and major targets for subcortical input. 31 Furthermore, as discussed earlier, these are the very same brain areas where the greatest neuronal diminution has been observed in postmortem studies of subjects with major depression (Table 4). 2,31 Chronic lithium also markedly increased the number of bcl-2 immunoreactive cells in the dentate gyrus and striatum; 30 detailed immunohistochemical studies following chronic VPA treatment are currently underway. It has subsequently been demonstrated that lithium also increases bcl-2 levels in C57BL/6 mice, 32 in human neuroblastoma SH-SY5Y cells in vitro 13 and in rat cerebellar granule cells in vitro. 33 This latter study was undertaken to investigate the molecular and cellular mechanisms underlying the neuroprotective actions of lithium against glutamate excitotoxicity (see below). These investigators found that lithium produced a remarkable increase in bcl-2 protein and mrna levels. Moreover, lithium has recently been demonstrated to reduce the levels of the pro-apoptotic protein p53 both in cerebellar granule cells 33 and in SH-SY5Y cells. 34 Thus, overall the data clearly show that chronic lithium robustly increases the levels of the neuroprotective protein bcl-2 in areas of rodent FCx, hippocampus and Table 4 Postmortem morphometric brain studies in mood disorders demonstrating cellular atrophy and/or loss Volume/cortical thickness Volumes of NAcc (L), basal ganglia (bilateral) in MDD and BD Cortical thickness, rostral oribitofcx, MDD Parahippocampal cortex size in suicide Volume of subgenual PFCx in familial MDD and BD Neurons Non-pyramidal neurons density in the CA2 region in BD Layer-specific interneurons in anterior cingulate cortex in BD Layer-specific interneurons in anterior cingulate cortex in MDD Pyramidal neurons density (layers III, V) in dorsolateral PFCx in BD Neuronal density and size in rostral oribitofcx, layer II/III in MDD Glia Density/size of glia in dorsolateral PFCx and caudal oribitofcx, in MDD and BD layer specific Glial (but not neurons) number in subgenual PFCx in familial MDD ( 24%) and BD ( 41%) Glial cell counts, glial density and glia-to-neuron ratios in amygdala NAcc, nucleus accumbens. Adapted from Manji et al. 2

6 striatum in vivo, and in cultured cells of both rodent and human neuronal origin in vitro. Furthermore, at least in cultured cell systems, lithium has also been demonstrated to reduce the levels of the pro-apoptotic protein p53. Consistent with bcl-2 s known cytoprotective effects, at therapeutically relevant concentrations, lithium has been shown to exert neuroprotective effects in a variety of preclinical paradigms. Thus, lithium has been demonstrated to protect against the deleterious effects of glutamate, NMDA receptor activation, aging, serum/nerve growth factor deprivation, ouabain, thapsigargin (which mobilizes intracellular MPP +,Ca 2+ ) and beta amyloid in vitro (Table 5). 13,30 More importantly, lithium s neurotrophic and cytoprotective effects have also been demonstrated in rodent brain in vivo. Thus, lithium treatment has been shown to attenuate the biochemical deficits produced by kainic acid infusion, ibotenic acid infusion and forebrain cholinergic system lesions, 13,30 exert dramatic protective effects against middle cerebral artery occlusion 35 and enhance hippocampal neurogenesis in the adult rodent hippocampus (Table 5). 32 Endogenous neurotrophic factors utilize bcl-2 to mediate many of their neurotrophic and neuroprotective effects The demonstration that the cytoprotective protein bcl- 2 is a target for the actions of mood stabilizers is particularly noteworthy since recent research has shown that bcl-2 mediates many of the beneficial effects of endogenous neurotrophic factors. Thus, neurotrophic factors (eg, nerve growth factor (NGF) and brain derived neurotrophic factor (BDNF), as well as cytokines, insulin-like growth factor-1 and glial derived neurotrophic factor, have been known to increase cell survival for several years However, these factors are now known to promote cell survival by suppressing intrinsic, cellular apoptotic machinery, not by Table 5 Neuroprotective effects of lithium Protects rat and human cultured neurons in vitro from: Glutamate, NMDA Calcium MPP + -amyloid Aging Growth factor deprivation Protects rodent brain in vivo from: Cholinergic system lesions Radiation injury Middle cerebral artery occlusion Quinolinic acid Human effects: No subgenual PFCx atrophy Increases NAA levels Increases gray matter volumes inducing cell survival pathways. 38,39 This occurs via binding of these factors to membrane receptors and regulation of intracellular signal transduction pathways that can control apoptosis, including regulation of bcl-2 family members (Figure 1). The signal transduction cascades that are currently believed to mediate many of the effects of neurotrophic factors are the mitogen-activated protein (MAP) kinase cascade and the phosphotidylinositol-3 kinase/akt pathway. 40,41 Neurotrophic factor signaling is mediated by a family of receptors known as Trks, which contain an intrinsic tyrosine kinase domain. NGF binds to the TrkA receptor and BDNF to TrkB. Receptor activation results in phosphorylation and activation of effectors, including phosphotidylinositol-3 kinase, as well as coupling with a series of proteins leading to activation of the MAP kinase cascade. 40,41 Recent studies have demonstrated that the activation of the MAP kinase pathway can inhibit apoptosis by inducing the phosphorylation of Bad (a major pro-apoptotic protein) and increasing the expression of bcl-2 (a major anti-apoptotic protein). The latter effect likely involves the camp response element binding protein (CREB) (Figure 1). 42,43 Phosphorylation of Bad occurs via activation of a downstream target of the MAP kinase cascade, ribosomal S- 6 kinase (Rsk). Rsk phosphorylates Bad and thereby promotes its inactivation. Activation of Rsk also mediates the actions of the MAP kinase cascade and neurotrophic factors on the expression of Bcl-2. Rsk can phosphorylate the CREB, and this leads to induction of bcl-2 gene expression (Figure 1). Valproate exerts major effects on a signaling cascade utilized by endogenous neurotrophic factors the MAP kinase signaling cascade As discussed previously, several endogenous growth factors including NGF and BDNF exert many of their neurotrophic effects via the MAP kinase signaling cascade. MAP kinases transmit extracellular signals to the nucleus, where the transcription of specific genes is induced by the synthesis, phosphorylation and activation of transcription factors. Three distinct MAP kinase signal transduction pathways have been identified in mammalian cells, leading to activation of the MAP kinases: extracellular signal regulated kinases (ERKs), c-jun NH2-terminal kinases, and p MAP kinases are abundantly present in brain, and in recent years a broad role for the MAP kinase cascade in regulating gene expression in long-term forms of synaptic plasticity has been demonstrated Thus, MAP kinases play important physiological roles in the mature CNS, and have been postulated to represent important targets for the actions of CNS-active agents In view of the important role of MAP kinases in mediating long-term neuroplastic events, it is noteworthy that VPA has recently been demonstrated to robustly activate the ERK MAP kinase cascade. 50 Since the ERK MAP kinases are known to mediate many of the effects of various neurotrophic factors and to promote neurite outgrowth, 40,51 VPA s effects on the mor- S51

7 S52 Figure 1 Model of the neurotrophic factor mitogen-activated protein (MAP) kinase cascade and regulation of cell survival. Cell survival is dependent on neurotrophic factors, such as BDNF and NGF, and the expression of these factors can be induced by synaptic activity. The influence of neurotrophic factors on cell survival is mediated by activation of the MAP kinase cascade. Activation of neurotrophic factor receptors, also referred to as Trks, results in activation of the MAP kinase cascade via several intermediate steps, including phosphorylation of the adaptor protein SHC and recruitment of the guanine nucleotide exchange factor Sos. This results in activation of the small guanosine triphosphate-binding protein Ras, which leads to activation of a cascade of serine/threonine kinases. This includes Raf, MAP kinase kinase (MEK) and MAP kinase (also referred to as extracellular response kinase, or Erk). One target of the MAP kinase cascade is Rsk, which influences cell survival in at least two ways. Rsk phosphorylates and inactivates the pro-apoptotic factor BAD. Rsk also phosphorylates camp response element binding protein (CREB), and thereby increases the expression of the anti-apoptotic factor bcl-2. This diagram also shows that G proteincoupled receptors (R), which couple to second messenger effector (E) systems, can lead to activation of the MAP kinase cascade. From Manji and Duman, 52 with permission. phology of human neuroblastoma cells have been examined in detail. Human neuroblastoma SH-SY5Y cells exposed to VPA (1.0 mm) in serum-free media for 5 days exhibited prominent growth cones and long neurites. Growth cone associated protein-43 is a protein expressed at elevated levels during neurite growth during development or regeneration, and a greater than three-fold increase in growth cone associated protein- 43 levels was observed after 5-day VPA exposure. 50 To further analyze the VPA-induced morphological changes, SH-SY5Y cells were seeded at a very low density and treated with VPA in serum-free media for 1 9 days. VPA (0.5 mm) dramatically increased neurite length 9-fold after 4-day incubation, and 14-fold after 9-day incubation. In view of VPA s apparent trophic effects, SH-SY5Y cells were grown in the presence of therapeutic concentration of VPA without any additional neurotrophic factors. Remarkably, cells grown in the presence of VPA, but without other neurotrophic factors, continued to grow well for 40 days. Are the effects of mood stabilizers on neurotrophic signaling cascades relevant in the treatment of BD? Although severe, recurrent mood disorders have traditionally been conceptualized as neurochemical disorders, there is now evidence from a variety of sources demonstrating regional reductions in CNS volume, as well as reductions in the numbers and/or sizes of glia and neurons in discrete brain areas. One line of evidence comes from structural imaging studies, which have recently begun to provide important clues about the neuroanatomical basis of mood disorders. In toto, the volumetric neuroimaging studies have demonstrated an enlargement of third and lateral ventricles,

8 as well as reduced gray matter volumes in parts of the orbital and medial prefrontal cortex (PFCx), the ventral striatum and the mesiotemporal cortex. 3,52 In addition to the accumulating neuroimaging evidence, several postmortem brain studies are now providing direct evidence for reductions in regional CNS volume and cell number (Table 4). 2,13,31 These recent data are leading to a reconceptualization about the cellular underpinnings of recurrent mood disorders; while they are clearly associated with a variety of neurochemical abnormalities, recurrent mood disorders are also associated with an impairment of neuroplasticity and cellular resilience Neuroplasticity subsumes diverse processes of vital importance by which the brain perceives, adapts to and responds to a variety of internal and external stimuli. The manifestations of neuroplasticity in the adult CNS have been characterized as including alterations of dendritic function, synaptic remodeling, long-term potentiation, axonal sprouting, neurite extension, synaptogenesis and even neurogenesis. Although the potential relevance of neuroplastic events for the pathophysiology of psychiatric disorders has been articulated for some time, recent morphometric studies of the brain (both in vivo and postmortem) are beginning to lead to a fuller appreciation of the magnitude and nature of the neuroplastic events involved in the pathophysiology of mood disorders. 13,54 While the body of preclinical data demonstrating neurotrophic and neuroprotective effects of lithium is striking, it is clear that considerable caution must be exercised in extrapolating to the clinical situation with humans. In view of lithium and VPA s robust effects on the levels of the cytoprotective protein bcl-2 in the FCx, Drevets et al 55 have re-analyzed their data demonstrating 40% reductions in subgenual PFCx volumes in familial mood disorder subjects. Consistent with neurotrophic/neuroprotective effects of lithium and VPA, they found that the patients treated with chronic lithium or VPA exhibited subgenual PFCx volumes, which were significantly higher than the volumes in non-lithium- or non-vpa-treated patients, and not significantly different from controls. 3 Although the results of the study by Drevets 3 suggest that mood stabilizers may have provided neuroprotective effects during naturalistic use, considerable caution is warranted in view of the small sample size and cross-sectional nature of the study. To investigate the potential neurotrophic effects of lithium in humans more definitively, a longitudinal clinical study was recently undertaken using proton magnetic resonance spectroscopy to quantitate N-acetyl-aspartate (NAA) levels. NAA is a putative neuronal marker, localized to mature neurons and not found in mature glial cells, cerebrospinal fluid or blood. 56 A number of studies have now shown that initial abnormally low brain NAA measures may increase, and even normalize, with remission of CNS symptoms in disorders such as demyelinating disease, amyotrophic lateral sclerosis, mitochondrial encephalopathies and human immunodeficiency virus (HIV) dementia. 56 NAA is synthesized within mitochondria, and inhibitors of the mitochondrial respiratory chain decrease NAA concentrations, effects which correlate with reductions in adenosine triphosphate and oxygen consumption. 13 Thus, NAA is now generally regarded as a measure of neuronal viability and function, rather than strictly as a marker for neuronal loss, per se. 56 It was found that chronic Li administration at therapeutic doses increased NAA concentration in the human brain in vivo. 57 These findings provide intriguing indirect support for the contention that, similar to the findings observed in the rodent brain and in human neuronal cells in culture, chronic lithium increases neuronal viability/function in the human brain. Furthermore, a striking 0.97 correlation between lithiuminduced NAA increases and regional voxel gray matter content was observed, 57 thereby providing evidence for co-localization with the regional-specific bcl-2 increases observed (eg, gray vs white matter) in the rodent brain cortices. These results suggest that chronic lithium may not only exert robust neuroprotective effects (as has been demonstrated in a variety of preclinical paradigms), but also exerts neurotrophic effects in humans. In follow-up studies to the NAA findings, it was hypothesized that in addition to increasing functional neurochemical markers of neuronal viability lithium-induced increases in bcl-2 would also lead to neuropil increases, and thus to increased brain gray matter volume in BD patients. In this clinical research investigation, brain tissue volumes were examined using high-resolution three-dimensional magnetic resonance imaging and validated quantitative brain tissue segmentation methodology to identify and quantify the various components by volume, including total brain white and gray matter content. Measurements were made at baseline (medication free, after a minimum 14- day washout) and then repeated after 4 weeks of lithium at therapeutic doses. This study revealed the extraordinary finding that chronic lithium significantly increases total gray matter content in the human brain of patients with BD. 58 No significant changes were observed in brain white matter volume, or in quantitative measures of regional cerebral water content, thereby providing strong evidence that the observed increases in gray matter content are likely due to neurotrophic effects, as opposed to any possible cell swelling and/or osmotic effects associated with lithium treatment. A finer-grained subregional analysis of these brain imaging data is ongoing. Since it is believed that the majority of neuron-specific NAA is localized to the neurites rather than the cell body, the observed increase in NAA is likely due to expansion of neuropil content. Taken together, these exciting new results support the contention that lithium does indeed exert neurotrophic/neuroprotective events in the human brain in vivo (Table 5). Concluding remarks As discussed, there is a considerable body of evidence, both conceptually and experimentally, in support of S53

9 S54 the regulation of the PKC and neurotrophic signaling cascades in the treatment, and potentially the pathophysiology, of BD. Regulation of signal transduction within critical regions of the brain affects the intracellular signal generated by multiple neurotransmitter systems. These effects thus represent attractive putative mediators of the pathophysiology of BD and the therapeutic actions of mood stabilizers. The biological processes in the brain responsible for the episodic clinical manifestation of mania and depression may be due to an inability to mount the appropriate compensatory responses necessary to maintain homeostatic regulation, with the resultant clinical picture being reflected in disruption of behavior, circadian rhythms, neurophysiology of sleep and neuroendocrine and biochemical regulation within the brain. 28,59 The molecular and cellular targets underlying lithium s ability to stabilize an underlying dysregulation of limbic and limbic-associated function is thus critical to our understanding of its mechanism of action. The complex effects of lithium on PKC isozymes represent both an attractive and heuristic mechanism by which the expression of various proteins involved in long-term neuronal plasticity and cellular response is modulated, thereby compensating for as yet genetically undefined physiological abnormalities in critical regions of the brain. Current studies of the long-term lithium-induced changes in the PKC signaling pathway including PKC isozyme regulation, posttranslational modification of key phosphoproteins and PKCmediated alterations in gene and protein expression are promising avenues for future investigation. The evidence demonstrating the neurotrophic effects of lithium and VPA, as well as the growing appreciation that mood disorders are associated with cell loss and atrophy, suggest that these effects may be very relevant to the long-term treatment of mood disorders. Does the long-term administration of these agents actually retard disease-induced or affective episodeinduced cell loss or atrophy? At present, there are no longitudinal studies that can fully address this question, but this is clearly an important and fundamental issue worthy of investigation. The findings that lithium administration increases brain NAA levels and gray matter volumes, as well as the cross-sectional study demonstrating normalized subgenual PFCx volumes in lithium- and VPA-treated patients provides indirect support for such a contention. The evidence also suggests that, somewhat akin to the treatment of conditions like hypertension and diabetes, early and potentially sustained treatment may be necessary to adequately prevent many of the deleterious long-term sequelae associated with BD. While data suggest that hippocampal atrophy in depression may be related to illness duration, 60 it is presently not clear if the volumetric and cellular changes that have been observed in other brain areas (most notably, the FCx) are related to affective episodes per se. Indeed, some studies have observed reduced gray matter volumes and enlarged ventricles at first onset in patients with mood disorders. 61,62 It is thus perhaps useful to conceptualize the cell atrophy that occurs in BD as arising from an endogenous impairment of cellular resiliency, rather than simply representing the toxic sequelae of affective episodes per se. In conclusion, emerging results from a variety of clinical and preclinical experimental and naturalistic paradigms suggest that a reconceptualization about the pathophysiology, course and optimal long-term treatment of BD may be warranted. Optimal long-term treatment for these severe illnesses may require the early use of agents with neurotrophic/neuroprotective effects, irrespective of the primary, symptomatic treatment. Such treatment modalities, via their effects on molecules involved in cell survival and cell death pathways, such as CREB, BDNF, Bcl-2 and MAP kinases, would be envisioned as enhancing neuroplasticity and cellular resilience (Figure 2). It is also becoming increasingly clear that for many refractory BD patients, new drugs simply mimicking many traditional drugs, which directly or indirectly alter neurotransmitter levels, and those that bind to cell surface receptors may be of limited benefit. 48 This is because such strategies implicitly assume that the target receptor(s) is functionally intact, and that altered synaptic activity will thus be transduced to modify the postsynaptic throughput of the system. However, the possible existence of abnormalities in signal transduction pathways 13,63 suggests that, for patients refractory to conventional medications, improved therapeutics Figure 2 Schematic representation of the continuum between neuroplasticity and cell survival/cell death. Recent preclinical studies have shown that signaling pathways involved in regulating more acute neuroplastic events also play a major role in regulating cell survival and cell death. Interestingly, the antidepressants lithium and valproate indirectly regulate a number of factors involved in cell survival pathways, including CREB, BDNF, bcl-2 and MAP kinases, and may thus bring about some of their delayed longterm beneficial effects via under-appreciated neurotrophic effects. The future development of treatments that more directly target molecules involved in critical CNS cell survival and cell death pathways thus hold promise as novel, improved, long-term treatments for mood disorders.

10 may only be obtained by the direct targeting of postreceptor sites. Recent discoveries concerning a variety of mechanisms involved in the formation and inactivation of second messengers offer promise for the development of novel pharmacological agents designed to target signal transduction pathways site specifically. 64 Although clearly more complex than the development of receptor-specific drugs, it may be possible to design novel agents to selectively affect second messenger systems because they are quite heterogeneous at the molecular and cellular level, are linked to receptors in a variety of ways and are expressed in different stoichiometries in different cell types. 52 Additionally, since signal transduction pathways display certain unique characteristics depending on their activity state, they offer built-in targets for relative specificity of action, depending on the set-point of the substrate. It is interesting to speculate that mood disorder patients who exhibit cell loss and atrophy, despite adequate treatment with psychotropic medications known to exert neurotrophic effects, may do so because of potential impairments of the intricate cellular machinery involved in mediating neurotrophic effects (eg, CREB/BDNF/trkB/MAP kinase/bcl-2) at distinct levels. 52 For such patients with putative abnormalities in neurotrophic pathways, improved therapeutics may only be obtained by the more direct targeting of downstream sites. It is thus noteworthy that a variety of strategies to enhance neurotrophic factor signaling are currently under investigation. An increasing number of strategies are also being investigated to develop small molecule switches for protein protein interactions, which have the potential to regulate the activity of growth factors, MAP kinases cascades and interactions between homo- and heterodimers of the bcl-2 family of proteins. 64 These developments hold much promise for the development of novel therapeutics for the long-term treatment of severe mood disorders, and for improving the lives of millions of people. Acknowledgements The author s research was supported by the NIMH, the Theodore and Vada Stanley Foundation and NARSAD. Ms Kerri R Gibala provided outstanding editorial assistance. References 1 Lenox RH. Role of receptor coupling to phosphoinositide metabolism in the therapeutic action of lithium. Adv Exp Med Biol 1987; 221: Manji HK, Drevets WC, Charney DS. The cellular neurobiology of depression. Nature Med 2001; 7: Drevets WC. Neuroimaging studies of mood disorders. Biol Psychiatry 2000; 48: Garlow SJ, Musselman DL, Nemeroff CB. The neurochemistry of mood disorders clinical studies. In: Charney DS, Nester EJ, Bunney BS (eds). Neurobiology of Mental Illness. Oxford University Press: New York, 1999, pp Wang JF, Young LT, Li PP et al. Signal transduction abnormalities in bipolar disorder. In: Joffe RT, Young LT (eds). Bipolar Disorder: Biological Models and Their Clinical Application. Marcel Dekker: New York, 1996, pp Bowden CL. Toward an integrated biological model of bipolar disorder. In: Young LT, Joffe RT (eds). Bipolar Disorder: Biological Models and Their Clinical Application. Marcel Dekker: New York, 1997, pp Bhalla US, Iyengar R. Emergent properties of networks of biological signaling pathways. Science 1999; 283: Jope RS, Williams MB. Lithium and brain signal transduction systems. Biochem Pharmacol 1994; 77: Moore GJ, Bebchuk JM, Parrish JK et al. Temporal dissociation between lithium-induced CNS myo-inositol changes and clinical response in manic-depressive illness. Am J Psychiatry 1999; 156: Jope RS. Anti-bipolar therapy: mechanism of action of lithium. Mol Psychiatry 1999; 4: Manji HK, Lenox RH. Protein kinase C signaling in the brain: molecular transduction of mood stabilization in the treatment of bipolar disorder. Biol Psychiatry 1999; 46: Hahn CG, Friedman E. Abnormalities in protein kinase C signaling and the pathophysiology of bipolar disorder. Bipolar Disord 1999; 1: Manji HK, Moore GJ, Rajkowska G, Chen G. Neuroplasticity and cellular resilience in mood disorders. Millennium article. Mol Psychiatry 2000; 5: Stabel S, Parker PJ. Protein kinase C. Pharmacol Ther 1991; 51: Newton AC. Protein kinase C: structure, function and regulation. J Biol Chem 1995; 270: Nishizuka Y. Protein kinase C and lipid signaling for sustained cellular responses. FASEB J 1995; 9: Manji HK, Etcheberrigaray R, Chen G, Olds JL. Lithium decreases membrane-associated protein kinase C in hippocampus: selectivity for the isozyme. J Neurochem 1993; 61: Chen G, Huang LD, Manji HK. Mood stabilizers regulate cytoprotective and mrna binding proteins in the brain: long term effects on cell survival and transcript stability. Int J Neuropsychopharmacol 2001; 4: Lenox RH, Watson DG, Ellis J. Chronic lithium administration alters a prominent PKC substrate in rat hippocampus. Brain Res 1992; 570: Chen G, Manji HK, Hawver DB, Wright CB, Potter WZ. Chronic sodium valproate selectively decreases protein kinase C alpha and epsilon in vitro. J Neurochem 1994; 63: Conn PJ, Sweatt JD. Protein kinase C in the nervous system. In: Kuo JF (ed). Protein Kinase C. Oxford University Press: New York, 1994, pp Couldwell WT, Weiss MH, DeGiorgio CM et al. Clinical and radiographic response in a minority of patients with recurrent malignant gliomas treated with high dose tamoxifen. Neurosurgery 1993; 32: Bebchuk JM, Arfken CL, Dolan-Manji S et al. A preliminary investigation of a protein kinase C inhibitor (tamoxifen) in the treatment of acute mania. Arch Gen Psychiatry 2000; 57: Hyman SE, Nestler EJ. Initiation and adaptation: a paradigm for understanding psychotropic drug action. Am J Psychiatry 1996; 153: Kandel ER. A new intellectual framework for psychiatry. Am J Psychiatry 1998; 155: Ikonomov O, Manji HK. Molecular mechanisms underlying moodstabilization in manic-depressive illness: the phenotype challenge. Am J Psychiatry 1999; 156: Flint J, Corley R. Do animal models have a place in the genetic analysis of quantitative human behavioral traits? J Mol Med 1996; 74: Manji HK, Lenox RH. Signaling: cellular insights into the pathophysiology of bipolar disorder. Biol Psychiatry 2000; 48: Chen G, Zeng WZ, Jiang L et al. The mood stabilizing agents lithium and valproate robustly increase the expression of the neuroprotective protein bcl-2 in the CNS. J Neurochem 1999; 72: Manji HK, Moore GJ, Chen G. Lithium at 50: have the neuroprotective effects of this unique medication been overlooked? Biol Psychiatry 1999; 46: Rajkowska G, Halaris A, Selemon LD. Reductions in neuronal and S55