Signaling: Cellular Insights into the Pathophysiology of

Bipolar Disorder
Husseini K. Manji and Robert H. Lenox

Clinical studies over the years have provided evidence have for the most part rested upon the conceptual founda-
that monoamine signaling and hypothalamicpituitary tion that monoamine signaling and hypothalamicpitu-
adrenal axis disruption are integral to the pathophysiol- itaryadrenal axis disruption are integral to the pathophys-
ogy of bipolar disorder. A full understanding of the iology of both depression and mania (Bowden 1997;
pathophysiology from a molecular to a systems level must Goodwin and Jamison 1990). Although such investiga-
await the identification of the susceptibility and protective
tions have been heuristic over the years, they have been of
genes driving the underlying neurobiology of bipolar
disorder. Furthermore, the complexity of the unique biol- limited value in elucidating the unique biology of this
ogy of this affective disorder, which includes the predis- affective disorder, which must include an understanding of
position to episodic and often progressive mood distur- the underlying basis for the predilection to episodic and
bance, and the dynamic nature of compensatory processes often profound mood disturbance that can become pro-
in the brain, coupled with limitations in experimental gressive over time. Kindling and progressive sensitization
design, have hindered our progress to date. Imaging to seizures, taken from the epilepsy literature, has been
studies in patient populations have provided evidence of a very useful as a model for focusing attention to the
role for anterior cingulate, amygdala, and prefrontal long-term course of the illness, but thus far has been less
cortex in the pathophysiology of bipolar disorder. More useful in explicating the biological processes leading to
recent research strategies designed to uncover the molec- instability in the regulation of mood; however, recent
ular mechanisms underlying our pharmacologic treat- research characterizing the contributory roles of signaling
ments and their interaction in the regulation of signal
pathways in various facets of kindling and sensitization
transduction as well as more advanced brain imaging
studies remain promising approaches. This experimental holds promise in this regard. More recent research strate-
strategy provides data derived from the physiologic re- gies designed to uncover the molecular mechanisms un-
sponse of the system in affected individuals and addresses derlying our pharmacologic treatments coupled with more
the critical dynamic interaction with pharmacologic advanced brain imaging studies remain promising ap-
agents that effectively modify the clinical expression of the proaches as we enter the next millennium.
pathophysiology. Biol Psychiatry 2000;48:518 530 We should also keep in mind that a true understanding
2000 Society of Biological Psychiatry of the pathophysiology of BD must address its neurobiol-
ogy at different physiologic levels (i.e., molecular, cellu-
Key Words: Bipolar disorder, lithium, signal transduc- lar, systems, and behavioral; Figure 1). Abnormalities in
tion, protein kinase C, G proteins, protein kinase A gene expression undoubtedly underlie the neurobiology of
the disorder at the molecular level, and this will become
evident as we identify the susceptibility and protective
Introduction genes for BD in the coming years. Once this has been
accomplished, however, the even more difficult work of
C linical studies over the past 40 years have attempted
to uncover the biological factors mediating the patho-
physiology of bipolar disorder (BD) utilizing a variety of
examining the impact of the faulty expression of these
gene products (proteins) on integrated cell function must
biochemical and neuroendocrine strategies. These studies begin. It is at these levels that we have identified some
protein candidates using the psychopharmacologic strate-
gies noted above and that will be more fully elucidated
From the Laboratory of Molecular Pathophysiology, Departments of Psychiatry and
Behavioral Neurosciences, and Cellular and Clinical Neurobiology Program,
below. The precise manner in which these candidate
Wayne State University School of Medicine, Detroit, Michigan (HKM) and the molecular and cellular targets may or may not relate to the
Molecular Neuropsychopharmacology Program, Departments of Psychiatry,
Pharmacology, and Neuroscience, University of Pennsylvania School of
faulty expression of susceptibility gene products is yet to
Medicine, Philadelphia (RHL). be determined. The task becomes even more daunting
Address reprint requests to Husseini K. Manji, M.D., Wayne State University
School of Medicine, 4201 St. Antoine, UHC 9B-29, Detroit MI 48201.
when one considers the possibility that a major component
Received February 4, 2000; revised May 11, 2000; accepted May 17, 2000. of the pathophysiology of BD may stem from discordant

2000 Society of Biological Psychiatry 0006-3223/00/$20.00

PII S0006-3223(00)00929-X
Signaling and Bipolar Disorder BIOL PSYCHIATRY 519
2000;48:518 530

Figure 1. The pathophysiology of bipolar disorder (BD). This figure highlights the fact that a complete understanding of the
pathophysiology of BD must address its neurobiology at different physiologic levels (i.e., molecular, cellular, systems, and behavioral).
PKC, protein kinase C; MARCKS, myristoylated alanine-rich C kinase substrate; GSK-3, glycogen synthase kinase-3; MAP kinase,
mitogen-activated protein kinase; Bcl-2, B-cell leukemia/lymphoma; proteome, the population of cellular protein species and their
expression level; transcriptome, the population of cellular messenger RNA species and their expression level.

biological rhythms ranging from ultradian to infradian that processes predisposing to recurrent mood disturbance may
ultimately drive the periodic recurrent nature of the disor- share common features. Furthermore, the clinical picture
der (Bunney and Bunney 2000; Ehlers et al 1988; Ikono- and system response are the result of a complex dynamic
mov and Manji 1999; Lenox and Manji 1998; Mandell et interaction between the dysregulated signaling systems
al 1984; Wehr and Goodwin 1983). The subsequent and activation of existing physiologic feedback mecha-
challenge for the basic and clinical neuroscientist will be nisms designed to compensate for extreme changes (Post
the integration of these molecular/cellular changes to the and Weiss 1999). In this way, the constellation of symp-
systems and ultimately to the behavioral level, wherein the toms including not only mood but also autonomic, endo-
clinical expression of BD becomes fully elaborated.
Before critically reviewing the data relevant to the
Table 1. The Pathophysiology of Bipolar Disorder: Constraints
pathophysiology of BD, it is necessary to consider the for Experimental Design
complexity of our task (Table 1). To begin with, we would
suggest that altered expression of critical proteins resulting Complex disease with diagnostic heterogeneity
Episodic nature of symptoms and distinct symptom clusters
from a series of susceptibility genes predisposes to a
The biology underlying recurrence/cyclicity may be distinct from
dysregulation of signaling in regions of the brain predis- that responsible for specific symptom clusters
posing to periodic loss of homeostasis and clinical mani- Disease progression dictates changes over the course of the illness
festation of affective symptomatology (i.e., mania and/or Dynamic interaction between compensatory/adaptive changes in the
depression; Depue et al 1987; Goodwin and Jamison brain and primary neurobiology of the disorder
Effect of treatment and treatment withdrawal on measures
1990). Thus the biological processes underlying the risk
Potential circadian rhythm abnormalities suggest that single time-
for mood cycling may be very distinct from the biology point studies may be inadequate
driving the clinical symptoms of mania or depression per Relative inaccessibility of target organ and dependence on peripheral
se (Goodwin and Jamison 1990). In this regard it should be models
noted that there is increasing evidence for a shared genetic Lack of suitable animal models
The characterization of mood as a quantitative trait (quantitative
risk for both bipolar and unipolar disorders (Berrettini
trait locus analysis) has not yet been accomplished
2000), suggesting that the underlying pathophysiologic
520 BIOL PSYCHIATRY H.K. Manji and R.H. Lenox
2000;48:518 530

crine, sleep/wake, and circadian activity determinants will Table 2. Bipolar Disorder: A Putative Role for Signaling
reflect not only the stage and progression of illness, but Pathways
also the unique individual characteristics conferring het- Regulate the functional balance and multiple neurotransmitter
erogeneity in clinical presentation and diagnosis. In light systems
of this complexity and the dynamic properties of the Dynamic regulation of complex signaling networks forms the basis
system, research strategies examining biochemical and for higher order brain function mediating mood and cognition
Critical role in fine-tuning of signals (amplifiers, attenuators, and
endocrine variables would be expected to be subject to a integrators of multiple signals)
high degree of inherent variability not only when using Critical role in maintaining cellular memory and long-term
cross-sectional analyses between patients, but also when neuroplastic events
utilizing longitudinal designs over time within individual Major targets for the actions of hormones, including gonadal
patients. Furthermore, the use of peripheral sources of steroids, thyroid hormones, and glucocorticoids
Abnormalities are compatible with life (e.g., a variety of human
tissue and postmortem brains to address biochemical/ diseases are from defects in G protein coupled signal
neuroendocrine activity within patients brains introduces transduction pathways)
another set of variables inherent in the experimental Brain regional dysregulation and circumscribed symptomatology are
design and most often conferring significant constraints on indeed possible despite the ubiquitous expression of signaling
data interpretation. proteins
Signaling proteins have been identified as targets for medications
that are most effective in the treatment of mood disorders

Classical Monoaminergic Neurotransmitter

and Neuroendocrine Systems
crine systems, however, make it difficult to conceptualize,
The stimulus for the study of the biogenic amines in let alone demonstrate, true specificity of detected abnor-
patients with BD was provided by the discovery of malities. Since these earlier studies have been amply
effective pharmacologic treatments for depression and reviewed elsewhere, in this perspectives article we focus
mania (Goodwin and Jamison 1990). Although these upon more recent data linking signaling abnormalities
pharmacologic agents have prominent effects on single with the underlying neurobiology of BD.
episodes of depression and mania, even more pertinent to
our discussion are the major effects that these agents exert
Signaling Networks: The Cellular
on the long-term course of the illness (Goodwin and
Machinery Underlying Information
Jamison 1990; Manji and Potter 1997). Antidepressants in
Processing and Long-Term Neuroplastic
general, and tricyclics in particular, may increase the
Events
frequency of cycles and worsen long-term outcome (Wehr
and Goodwin 1987). In addition to this compelling phar- It is hardly surprising that abnormalities in multiple
macologic data, the biogenic amine neurotransmitter sys- neurotransmitter systems and physiologic processes have
tems are distributed extensively in the limbic system, been found in a disorder as complex as BD. Signal
which is implicated in the regulation of sleep, appetite, transduction pathways are in a pivotal position in the
arousal, sexual function, endocrine function, and emo- central nervous system (CNS), able to affect the functional
tional states such as fear and rage. The clinical picture of balance between multiple neurotransmitter systems, and
BD involves disruption of behavior, circadian rhythms, may therefore play a role in mediating the more down-
neurophysiology of sleep, and neuroendocrine and bio- stream abnormalities in multiple neurotransmitter sys-
chemical regulation within the brain (for reviews, see tems and physiologic processes. Moreover, recent research
Goodwin and Jamison 1990; Holsboer 1995). Thus it is has clearly identified signaling pathways as therapeuti-
not surprising that clinical research strategies over the cally relevant targets for our most effective pharmacologic
years have identified dysregulation of noradrenergic, do- treatments (Table 2). Indeed, the molecular and cellular
paminergic, serotonergic, and cholinergic systems as well targets underlying lithiums ability to stabilize an under-
as the hypothalamicpituitary axis in patients with BD. lying dysregulation of limbic and limbic-associated func-
The behavioral and physiologic manifestations of the tion strongly suggest that abnormalities in signaling path-
illness are complex and undoubtedly mediated by a net- ways may also play a critical role in the pathophysiology
work of interconnected neurotransmitter pathways (Lenox of BD.
1987; Manji 1992); the monoamine neurotransmitter sys- Multicomponent, cellular signaling pathways interact at
tems are ideally placed to mediate such complex behav- various levels, thereby forming complex signaling net-
ioral effects, and thus have represented attractive candi- works that allow the cell to receive, process, and respond
date systems underlying the pathophysiology of BD. The to information (Bhalla and Iyengar 1999; Bourne and
multiple functions of central neurotransmitter/neuroendo- Nicoll 1993). These networks facilitate the integration of
Signaling and Bipolar Disorder BIOL PSYCHIATRY 521
2000;48:518 530

signals across multiple time scales and the generation of Table 3. Protein Kinase C: Pathophysiology and Treatment of
distinct outputs depending on input strength and duration, Bipolar Disorder
and regulate intricate feed-forward and feedback loops Kindling produces dramatic increases in membrane-associated PKC
(Weng et al 1999). These properties of signaling networks in the hippocampus and amygdala
suggest that they play critical roles in cellular memory; Amphetamine produces increases in PKC activity, and GAP-43
thus, cells with different histories, and therefore express- phosphorylation (implicated in neurotransmitter release)
PKC inhibitors block the biochemical and behavioral responses to
ing different repertoires of signaling molecules and tran- amphetamine and cocaine and also block cocaine-induced
scription factors, interacting at different levels, may re- sensitization
spond quite differently to the same signal over time. Given Dexamethasone administration increases PKC activity, and increases
their widespread and crucial role in the integration, regu- the levels of PKC and PKC in the rat frontal cortex and
lation, amplification, and fine-tuning of physiologic pro- hippocampus
Increased membrane/cytosol PKC partitioning in platelets from
cesses, it is not surprising that abnormalities in signaling manic subjects; normalized with lithium treatment
pathways have now been identified in a variety of human Increased PKC activity and translocation in bipolar disorder brains,
diseases (Milligan and Wakelam 1992; Spiegel 1998; as compared with control subjects
Weintraub 1995). Pertinent to the present discussion is the Lithium and valproate regulate PKC activity, PKC , PKC , and
observation that a variety of diseases manifest relatively MARCKS
Preliminary data suggest that PKC inhibitors may have efficacy in
circumscribed symptomatology, despite the widespread, the treatment of acute mania
often ubiquitous expression of the affected signaling Typical and atypical antipsychotics regulate the levels of PKC and
proteins. PKC in areas of the rat brain
Although complex signaling networks are likely present PKC, protein kinase C; GAP, growth coneassociated protein; MARCKS,
in all eukaryotic cells and control various metabolic, myristoylated alanine-rich C kinase substrate.
humoral, and developmental functions, they may be espe-
cially important in the CNS, where they serve the critical
(either behavioral or electrophysiologic) to repeated low
roles of first amplifying and weighting numerous extra-
dose stimulation (pharmacologic or electric) over time.
cellularly generated neuronal signals and then transmitting
Indeed, the sensitized behavioral response has been re-
these integrated signals to effectors, thereby forming the
ported to persist for months and years after discontinuation
basis for a complex information processing network
of drug administration. Considerable evidence implicates
(Bhalla and Iyengar 1999; Bourne and Nicoll 1993; Manji
long-term alterations in midbrain dopaminergic transmis-
1992). The high degree of complexity generated by these sion in the development of behavioral sensitization, but the
signaling networks may be one mechanism by which cellular mechanism(s) underlying the long-term changes
neurons acquire the flexibility for generating the wide in excitability observed in kindled or stimulant-sensitized
range of responses observed in the nervous system (Table animals have not been fully elucidated. In recent years, an
2). These pathways are thus undoubtedly involved in appreciation of the major role of signal transduction
regulating such diverse vegetative functions as mood, pathways in the regulation of neuronal excitability has led
appetite, and wakefulness and are therefore likely to be to extensive research into their involvement in kindling
involved in the pathophysiology of BD. We now turn to a and behavioral sensitization, and a growing body of
discussion of the direct and indirect evidence supporting a evidence implicates alterations in both protein kinase C
role for abnormalities in signaling pathways in the patho- (PKC; Table 3) and certain G proteins (especially Gi and
physiology of BD. Go). Studies have demonstrated that pertussis toxin (which
inactivates Gi and Go) produces a significant augmentation
of psychostimulant-induced motor activity, and dopamine
Alterations of Transmembrane Cellular release in the nucleus accumbens (Chen et al 1999;
Signaling Pathways in Animal Models of BD Steketee and Kalivas 1991). In keeping with these studies,
The need to use caution in the appropriate application of specific decreases in Gi and Gi levels have been
animal models to an understanding of complex neuropsy- observed in response to chronic cocaine in the ventral
chiatric disorders has been well articulated (Einat et al, in tegmental area and nucleus accumbens (Nestler et al
press; Post and Weiss 1999), and in fact it is unlikely we 1990), once again consistent with the hypothesis that
will ever develop animal models that display the full range regulation of G proteins represents part of the biochemical
of symptomatology as clinically expressed in man. Two changes that underlie the effects of chronic psychostimu-
current models that have had reasonable heuristic value in lants. Similarly, injection of pertussis toxin into the amyg-
the study of mood disorders are kindling and behavioral dala of kindled rats markedly modifies kindled seizures,
sensitization (Einat et al, in press; Post and Weiss 1999). effects that appear to be mediated via an alteration in the
In both of these paradigms, there is increased responsivity after-discharge threshold.
522 BIOL PSYCHIATRY H.K. Manji and R.H. Lenox
2000;48:518 530

With respect to the PKC signaling system, dramatic of AC both by Gpp(NH)p (a poorly hydrolyzable analog
increases in membrane-associated PKC have been ob- of guanosine triphosphate [GTP]) and by Ca/calmodu-
served in the bilateral hippocampus up to 4 weeks after the lin, suggesting that lithium in vitro is directly able to
last kindled seizure and in the amygdala/pyriform cortex at inhibit the catalytic unit of AC (Mork et al 1992; Newman
4 weeks after (Daigen et al 1991). Studies have also and Belmaker 1987). Since these inhibitory effects of
implicated alterations in PKC activity as mediators of lithium in vitro can be overcome by Mg, it has been
long-term alterations in neuronal excitability in the brain postulated that they are mediated (at least in part) by a
following chronic stimulant use. Thus, both acute and direct competition with magnesium (whose hydrated ionic
chronic amphetamine produce an alteration in PKC activ- radius is similar to that of lithium) for a binding site on the
ity, its cytosol relative to membrane distribution, and the catalytic unit of AC (Mork and Geisler 1989; Mork et al
phosphorylation of a major PKC substrate, growth cone 1992; Newman and Belmaker 1987); however, the inhib-
associated protein 43 (GAP-43), which has been impli- itory effects of chronic lithium treatment on rat brain AC
cated in long-term alterations of neurotransmitter release are not reversed by Mg and still persist after washing of
(Kantor and Gnegy 1998). Furthermore, PKC inhibitors the membranes, but are reversed by increasing concentra-
have been shown to block the acute responses to both tions of GTP (Mork and Geisler 1989; Mork et al 1992;
amphetamine and cocaine (as assessed by both behavioral Newman and Belmaker 1987). These results suggest that
and in vivo microdialysis studies) as well as cocaine- the effects of chronic lithium (those that are more likely to
induced sensitization (Kantor and Gnegy 1998). be therapeutically relevant) may be exerted at the level of
It is indeed striking that behavioral sensitization and signal-transducing G proteins at a GTP-responsive step
kindling produce robust alterations in the PKC signaling (discussed below). These two distinct actions of lithium on
pathway in critical limbic structures, since lithium and the AC system may explain the differences in results that
valproate (VPA) also target the very same biochemical have been obtained by investigators using rat membrane
targets (see below). Thus, although considerable caution preparations and those obtained with the use of rat slice
clearly needs to be employed when extrapolating from the preparations (Lenox and Manji 1998). This has led to an
findings observed in these two models, the fact that they investigation of lithiums effects on the AC system in
are associated with effects on PKC signaling that are the vivo, using microdialysis. These studies found that chronic
opposite of those observed with chronic lithium or VPA is lithium treatment produced a significant increase in basal
compelling indeed. and postreceptor stimulated (cholera toxin or forskolin)
AC activity, while attenuating the -adrenergicmediated
effect (Manji et al 2000; Masana et al 1992). Interestingly,
Are G ProteinCoupled Signal Transduction chronic lithium treatment resulted in an almost absent
Pathways Targets for Mood-Stabilizing cyclic adenosine monophosphate (cAMP) response to
Agents? pertussis toxin, suggesting a lithium-induced attenuation
The cAMP/Protein Kinase A Signaling Pathway as of Gi function. It should be noted, however, that chronic
a Target for the Actions of Mood Stabilizers and lithium has been found to increase not only cAMP levels
Antidepressants (Wiborg et al 1999), but also the levels of AC type I and
type II messenger RNA (mRNA) and protein levels in the
As discussed above, dysregulation of cellular signaling
frontal cortex (Colin et al 1991; Jensen et al 2000),
networks in the brain likely underlies the oscillations in
suggesting that lithiums complex effects on the system
behavioral states observed in mood disorders. Thus, it is
may represent the net effects of direct inhibition of AC,
not surprising that increasingly recent research on the
upregulation of AC subtypes, and effects on the stimula-
cellular mechanisms underlying lithiums therapeutic ef-
fects has focused on the transmembrane signaling path- tory and inhibitory G proteins. Most recently, lithiums
ways in the brain (for reviews, see Lenox and Manji 1998; effects on the phosphorylation and activity of cyclic AMP
Mork et al 1992; Post et al 2000). response element binding protein (CREB) have been
examined in the rodent brain and in cultured human
neuroblastoma cells, with somewhat conflicting results
Lithium and the AC System (Ozaki and Chuang 1997; Wang et al 1999).
It has been demonstrated that lithium exerts complex A series of studies have also examined lithiums effects
effects on the activity of adenylyl cyclase (AC), with the on AC in humans. In a longitudinal study of healthy
preponderance of the data demonstrating an elevation of volunteers (utilized to overcome the potentially confound-
basal AC activity but an attenuation of variety of receptor- ing, significant effects of alterations in mood state
mediated responses (Jope 1999a, 1999b; Manji et al 2000; dependent biochemical and neuroendocrine parameters), 2
Wang et al 1997). Lithium in vitro inhibits the stimulation weeks of lithium administration was found to significantly
Signaling and Bipolar Disorder BIOL PSYCHIATRY 523
2000;48:518 530

increase platelet basal and postreceptor stimulated AC ulin-dependent protein kinases, effects that are accompa-
activity (Risby et al 1991), effects that are strikingly nied by increases in the endogenous phosphorylation of
similar to those observed in the rodent brain. Consistent selected substrates (microtubule-associated protein 2 and
with a lithium-induced increase in basal cAMP and AC synaptotagmin; Popoli et al 2000). Furthermore, recent
levels, a more recent study found that platelets obtained data from Duman and associates (1997) have demon-
from lithium-treated euthymic bipolar patients exhibited strated that the chronic treatment of rats with a variety of
enhanced basal and cAMP-stimulated phosphorylation of ADs increases the levels of CREB mRNA, CREB, and
Rap1 (a protein kinase A [PKA] substrate), as well as CRE DNA binding activity in the hippocampus. The same
another unidentified 38-kd phosphoprotein (Perez et al laboratory has demonstrated that chronic AD treatment
2000a). Somewhat surprisingly, these investigators did not also increases the expression of two important genes
find similar effects of lithium in healthy subjects. known to be regulated by CREBnamely, brain-derived
neuotrophic factor and its receptor trkB (Duman et al
Carbamazepine and the AC System 1997). Preliminary postmortem human brain studies have
also revealed increased CREB levels in patients treated
The effects of carbamazepine (CBZ) on the AC system with ADs, providing indirect support for the rodent and
have been less extensively studied. It is noteworthy, cell culture studies (Dowlatshahi et al 1998).
however, that CBZ has been reported to decrease not only
basal and stimulated cAMP production in rodents, but also
cerebrospinal fluid levels of cAMP in manic patients (Post Lithium and G Proteins
et al 2000). Recent studies have found that CBZ, at
therapeutically relevant concentrations, inhibits both basal Abundant experimental evidence has shown that lithium
AC and FSK-stimulated cAMP accumulation in C6 gli- attenuates receptor-mediated second messenger generation
oma cells (Chen et al 1996). Furthermore, CBZ also in the absence of consistent changes in the density of the
inhibits both basal and FSK-stimulated activity of AC receptors themselves (Lenox and Manji 1998; Mork et al
purified from the rat cortex, suggesting that CBZ inhibits 1992). There is now considerable evidence that chronic
cAMP production by acting directly on AC and/or through lithium administration affects G protein function (Jope
factor(s) that are tightly associated with and copurify with 1999a, 1999b; Lenox and Manji 1998; Wang and Fried-
AC (Chen et al 1996). Carbamazepines effects on cAMP- man 1999); however, the preponderance of data suggests
mediated transcriptional events have also been investi- that lithium, at therapeutically relevant concentrations,
gated, and CBZ has been found to inhibit forskolin- does not have any direct effects on G proteins and does not
stimulated phosphorylation of CREB (Chen et al 1996) consistently alter G protein subunit levels (Ellis and Lenox
and c-fos expression (Divish et al 1991). Together, the 1991; Lenox and Manji 1998). Although some studies
results suggest that CBZ brings about a dampening of the have reported modest changes in the levels of G protein
AC system; these effects may play a role in both its subunits, the preponderance of the data suggests that the
antiepileptic and its antimanic effects, but such a conten- effects of chronic lithium on signaling pathways occur in
tion requires extensive further validation. the absence of changes in the levels of G protein subunits
per se (Lenox and Manji 1998); however, chronic in vivo
lithium treatment has been shown to produce a significant
Antidepressants and the AC System increase in pertussis toxin catalyzed [32P]adenosine
The chronic administration of antidepressants (ADs) is diphosphate ([32P]ADP) ribosylation in the rat frontal
known to bring about a variety of biochemical effects, cortex. Since pertussis toxin selectively ADP-ribosylates
including an enhanced coupling between Gs and the the undissociated, inactive heterotrimeric form of Gi,
catalytic unit of AC (Rasenick et al 2000). Several studies these results suggest that lithium attenuates Gi function via
have also demonstrated that the postreceptor components a stabilization of the inactive conformation. These obser-
of the cAMP system are regulated by long-term AD vations suggest that the removal of the inhibitory tone
treatments, including cAMP-dependent protein kinase en- by lithium may be responsible for the elevations in basal
zyme activity (Nestler et al 1989; Popoli et al 2000). AC and the responses to agents activating the stimulatory
Taken together, these results suggest that ADs, via their pathway distal to the receptor (Jope 1999b). Consistent
complex effects on G proteins and the AC signaling with this hypothesis, lithium has been shown to potentiate
system, may attenuate -adrenergicmediated activation the behavioral hyperactivity induced by intra-accumbens
of Gs, while enhancing the effects of agents operating by cholera toxin administration (which activates the stimula-
pathways independent of the -adrenergic receptor (AR). tory G proteins, Gs and Golf; Kofman et al 1998).
Consistent with these results, ADs have been demon- Recent studies have also examined the effects of
strated to activate cAMP-dependent and calcium/calmod- chronic lithium on G protein function in humans and have
524 BIOL PSYCHIATRY H.K. Manji and R.H. Lenox
2000;48:518 530

generally observed reduced receptor/G protein coupling Indeed, evidence accumulating from various laborato-
(Lenox and Manji 1998; Risby et al 1991). The effects of ries has clearly demonstrated that lithium, at therapeuti-
2 weeks of lithium administration on G protein measures cally relevant concentrations, exerts major effects on the
has also been examined in healthy volunteers. Similar to PKC signaling cascade (Table 3). Currently available data
the findings in the rat brain, lithium did not affect the suggest that acute lithium exposure facilitates a number of
levels of platelet G protein subunits, but produced a PKC-mediated responses, whereas longer exposure results
significant increase in pertussis toxin catalyzed [32P]ADP in an attenuation of phorbol estermediated responses,
ribosylation, once again suggesting a stabilization of the which is accompanied by a downregulation of specific
inactive undissociated heterotrimeric form of Gi. PKC isozymes (discussed in Manji and Lenox 1999).
These long-term effects of chronic lithium on G protein Studies in rodents have demonstrated that chronic (but not
are likely attributable to an indirect posttranslational mod- acute) lithium produces an isozyme-selective reduction in
ification of the G protein(s) and a relative change in the PKC and in the frontal cortex and hippocampus, in the
dynamic equilibrium of the active/inactive states of pro- absence of significant alterations in the , , , or
tein conformation. In this context, it is noteworthy that isozymes. Concomitant studies carried out in immortalized
investigators have demonstrated that lithium alters the hippocampal cells in culture exposed to chronic lithium
levels of endogenous ADP-ribosylation in C6 glioma cells show a similar reduction in the expression of both the PKC
(Young and Woods 1996) and in the rat brain (Nestler et and isozymes in the cell as determined by immunoblot
al 1995), suggesting another novel mechanism by which (Chen et al, in press; Manji and Lenox 1999).
chronic lithium may indirectly regulate the activity of A major strategy that has been utilized to investigate the
these critical signaling proteins. downstream consequences of lithium-induced alteration in
PKC isozymes is the examination of the effects of chronic
lithium on endogenous PKC substrates in the brain. The
The Protein Kinase C Signaling Cascade most prominent substrate for PKC in the brain is an acidic
Over the last decade there have been major advances in our protein, myristoylated alanine-rich C kinase substrate
understanding of the critical role of the PKC signaling (MARCKS), which has been implicated in regulating
pathway as a therapeutically relevant target for the long-term long-term neuroplastic events. Lenox and associates
(1992) provided the first evidence for the effects of
actions of mood stabilizers. The inositol depletion hypoth-
chronic administration of lithium on the regulation of
esis posited that lithium, as an uncompetitive inhibitor of
MARCKS in the rat brain. They demonstrated that chronic
inositol-1-phosphatase, produced its therapeutic effects via a
lithium administration dramatically reduced MARCKS
depletion of neuronal myo-inositol (mI) levels. Although this
hypothesis has been of great heuristic value, numerous expression in the hippocampus, effects that were not
studies have examined the effects of lithium on receptor- immediately reversed following lithium discontinuation.
mediated phosphoinositide (PI) responses, and although Subsequent studies carried out in immortalized hippocam-
some report a reduction in agonist-stimulated phosphatidyl- pal cells have demonstrated that this action of chronic
inositol 4,5-bisphosphate (PIP2) hydrolysis in rat brain slices lithium on MARCKS regulation is dependent upon both
the inositol concentration and the level of receptor-medi-
following acute or chronic lithium, these findings have often
ated activation of PI hydrolysis (Watson and Lenox 1996).
been small and inconsistent, and subject to numerous meth-
Although these effects of lithium on PKC isozymes and
odological differences (Jope and Williams 1994; Lenox and
MARCKS are striking, a major problem inherent in
Manji 1998). Most recently, a magnetic resonance spectros-
neuropharmacologic research is the difficulty in attribut-
copy study has demonstrated that lithium-induced mI reduc-
ing therapeutic relevance to any observed biochemical
tions are observed in the frontal cortex of BD patients after
finding. It is thus noteworthy that the structurally dissim-
only 5 days of lithium administration, at a time when the
ilar antimanic agent VPA produces effects strikingly
patients clinical state is completely unchanged (Moore et al
similar to those of lithium on PKC isozymes and on the
1999). Consequently, these and other studies suggest that
expression of the MARCKS protein (Manji and Lenox
whereas inhibition of inositol monophosphatase may repre-
1999; Watson et al 1998).
sent an initiating lithium effect, reducing mI levels per se is
not associated with therapeutic response. This led to our
working hypothesis that some of the initial actions of lithium Are There Abnormalities in Signaling
may occur with a relative reduction of mIthis reduction of Pathways in BD?
mI initiates a cascade of secondary changes in the PKC
signaling pathway and gene expression in the CNS, effects The cAMP-Generating System
that are ultimately responsible for lithiums therapeutic The data reviewed above strongly suggest that regulation
efficacy. of cellular signaling transduction pathways play a major
Signaling and Bipolar Disorder BIOL PSYCHIATRY 525
2000;48:518 530

role in the therapeutic effects of lithium and VPA. Do Whether such changes result in altered endogenous
these clinically effective agents, in fact, correct an under- cAMP-stimulated phosphorylation in the brain, similar to
lying abnormality in signaling pathways in BD? The what has been reported in platelets from euthymic BD
future development of selective receptor and second mes- patients (Perez et al 2000a, 2000b), remains to be demon-
senger ligands for positron emission tomography studies strated. Nevertheless, these observed changes in PKA
may permit the direct assessment of CNS receptor and provide additional important evidence for a potential
postreceptor sensitivity in humans. To date, however, dysregulation of the Gs-mediated cAMP signaling path-
studies of receptor and postreceptor function in mood way in BD.
disorders have been limited to indirect research strategies.
The most commonly utilized strategy has been to charac- G Proteins in Mood Disorders
terize receptor function in readily accessible blood ele-
ments, and much clinical research has focused on the In view of the indirect evidence for abnormalities at
activity of the cAMP-generating system in mood disor- postreceptor sites described above, it is not surprising that
ders. Overall, the preponderance of the evidence suggests several independent laboratories have examined G pro-
altered receptor and/or postreceptor sensitivity of the teins in patients with mood disorders (Chen et al 1999;
cAMP-generating system in the absence of consistent Mathews et al 1997; Wang et al 1997; Warsh et al 2000).
alterations in the number of receptors themselves (for a Young and associates (1993) were the first to report
review, see Wang et al 1997). A consistently observed increased levels of Gs in BD. Compared with control
decrease in leukocyte AR function in depression could subjects matched with respect to age, postmortem interval,
reflect an inherited abnormality of the AR/Gs/AC com- and brain pH, they found increased levels of Gs in the
plex, as suggested by the findings utilizing immortalized frontal, temporal, and occipital cortices (but not in the
(EpsteinBarr virustransformed) lymphocytes (Wright et hippocampus, thalamus, or cerebellum) in postmortem
al 1984). These findings need to be replicated, however, brain tissue from patients with BD. This group also found
and a number of additional confounding factors need to be increases in forskolin-stimulated AC activity in postmor-
considered. In this context, twin studies suggest that tem brains of BD patients, compatible with enhanced
variations in isoproterenol-stimulated cAMP production Gs/AC signaling in BD (Warsh et al 2000). The findings of
are most likely due to environmental effects on the elevated Gs levels and/or function are also supported by
number or sensitivity of ARs (Ebstein et al 1986). A the recent observations of Friedman and Wang (1996),
recent study found higher levels of cAMP-stimulated who found increased agonist-activated [35S]GTPS bind-
phosphorylation of a 22-kd protein in platelets obtained ing to G protein subunits in frontal cortical membrane
from 10 treated euthymic BD patients, as compared with preparations from BD patients. Several studies have also
healthy subjects; by contrast, there was no significant examined G proteins in peripheral circulating cells and
difference in the basal phosphorylation between the have found elevated levels of Gs protein (Chen et al
groups (Perez et al 2000a, 2000b). Follow-up studies 1999; Wang et al 1997; Warsh et al 2000) and mRNA
identified the 22-kd protein as Rap1, and once again levels (Spleiss et al 1998), although the dependency on
found higher cAMP-stimulated phosphorylation in the BD clinical state remains unclear. Similar to what has been
patients (Perez et al 2000a, 2000b). observed in the CNS, one recent study has evaluated the
Warsh and associates have undertaken the most thor- role of platelet G proteins as signal coincidence detec-
ough series of studies investigating the cAMP/PKA sys- tors, and has found this function to be impaired in
tem in postmortem brains of BD patients. Due to the depressed patients (Mooney et al 1998). Overall, the most
instability of cAMP in postmortem brain tissue, this consistent finding to emerge is that in both peripheral cells
research group has investigated downstream targets of and postmortem brain tissue from BD patients elevations
cAMP. They found that the levels of PKA regulatory are observed in the predominant subspecies of Gs present
subunits (as assessed by [3H]cAMP binding) were signif- in the tissues examined. It should be emphasized, how-
icantly lower in cytosolic fractions of BD frontal, tempo- ever, that there is at present no evidence to suggest that the
ral, occipital, and parietal cortices; cerebellum; and thala- alterations in the levels of Gs are due to a mutation in the
mus, as compared with matched control subjects (Rahman Gs gene itself. Indeed, there are numerous transcriptional
et al 1997). Furthermore, preliminary findings show that and posttranscriptional mechanisms that regulate the lev-
the reduction of regulatory subunits of PKA in the cyto- els of G protein subunits (Warsh et al 2000), and the
solic fractions of the BD temporal cortex is accompanied elevated levels of Gs could potentially represent the
by a higher basal kinase activity and significantly lower indirect sequelae of alterations in any one of these other
apparent activation constant for cAMP in the cytosolic biochemical pathways. Thus, at this point considerable
fractions of the BD temporal cortex (Fields et al 1999). caution is required in the interpretation of the data, since
526 BIOL PSYCHIATRY H.K. Manji and R.H. Lenox
2000;48:518 530

they derive primarily from peripheral cell models and tained from BD subjects before and during lithium
postmortem studies involving a small number of patients. treatment. They reported that the ratios of platelets
membrane-bound to cytosolic PKC activities were ele-
Abnormalities of Calcium Signaling in BD vated in the manic subjects. In addition, serotonin-
elicited platelet PKC translocation was found to be
Acting via intracellular proteins such as MARCKS and enhanced in those subjects. With respect to brain tissue,
calmodulin, and enzymes such as PKC, AC, and Ca/ Wang and Friedman (1996) measured PKC isozyme
calmodulin-dependent kinase, calcium ions have been levels, activity, and translocation in postmortem brain
shown to regulate the synthesis and release of neuro- tissue from BD patients; they reported increased PKC
transmitters, neuronal excitability, cytoskeletal remod- activity and translocation in BD brains, as compared
eling, and long-term neuroplastic events. Thus, it is not with control subjects, effects that were accompanied by
surprising that a large number of studies have investi- elevated levels of selected PKC isozymes in cortices of
gated intracellular Ca2 in peripheral cells in BD BD subjects.
(discussed in Dubovsky et al 1992; Emamghoreishi et al In view of the pivotal role of the PKC signaling
1997; Wang et al 1997). In view of the caveats pathway in the regulation of neuronal excitability,
associated with studies of peripheral circulating cells, neurotransmitter release, and long-term synaptic events,
the remarkable consistency of the findings is surprising it was postulated that the attenuation of PKC activity
indeed. Studies have consistently revealed elevations in may play a role in the antimanic effects of lithium and
both resting and stimulated intracellular Ca2 levels in VPA. In a pilot study it was found that tamoxifen (a
platelets, lymphocytes, and neutrophils of patients with nonsteroidal antiestrogen known to be a PKC inhibitor
BD. The calcium abnormalities have been postulated to at higher concentrations) may, indeed, possess anti-
represent state-dependent findings (Dubovsky et al manic efficacy (Bebchuk et al 2000). These data further
1992), but recent studies using transformed lympho- support the potential role of the PKC pathway in
blasts from BD patients have revealed similar abnor- mediating the pathophysiology of BD. The data impli-
malities, suggesting that they may be trait dependent cating the PKC signaling system in the pathophysiology
(Emamghoreishi et al 1997). The regulation of free of BD are particularly compelling in view of the
intracellular Ca2 is a complex, multifaceted process growing body of data suggesting that this signaling
involving extracellular entry, release from intracellular pathway represents a therapeutically relevant target for
stores following receptor-stimulated PI hydrolysis, up- both lithium and VPA (see above). The complex effects
take into specific organelles, and binding to specific of lithium on the PKC signaling pathway represent an
proteins. Thus, the abnormalities observed in BD could attractive and heuristic mechanism by which the expres-
arise from abnormalities at a variety of levels, and sion of various proteins involved in long-term neuronal
recent studies suggest that the abnormality lies beyond plasticity and cellular response is modulated, thereby
the receptor (Hough et al 1999). Interestingly, recent compensating for as yet genetically undefined physio-
studies have demonstrated that Rap1 phosphorylation logic abnormalities in critical regions of the brain. It is
state is related to platelet intracellular calcium signaling our strong conviction that it is at the cellular and
(Magnier et al 1995), suggesting a possible relationship molecular level that some of the most exciting advances
between these two documented abnormalities. Since in our understanding of the pathophysiology of BD will
PKC is also known to regulate calcium signaling at take place in the coming years. Current studies of the
multiple levels (Ozaki et al 1997; Shibata et al 1996; long-term lithium-induced changes of the PKC signal-
Si-Tahar et al 1996), more recent studies have investi- ing pathway (including PKC isozyme regulation, post-
gated the putative role of PKC in mediating the calcium translational modification of key phosphoproteins, and
abnormalities in BD; preliminary analysis suggests that PKC-mediated alterations in gene and protein expres-
alterations in tonic PKC activity may play an important sion) are a most promising avenue for future
role in mediating the abnormal intracellular calcium investigation.
responses observed in BD (H.K. Manji and R.M. Post,
unpublished observations, June 1998).
Concluding Remarks
Are There Abnormalities of PKC Signaling in BD? As we have demonstrated, there is a considerable body
To date, there have only been a limited number of of evidence both conceptually and experimentally in
studies directly examining PKC in BD. Friedman and support of abnormalities in the regulation of signaling
associates (1993) investigated PKC activity and PKC as integral to the underlying neurobiology of BD. The
translocation in response to serotonin in platelets ob- pathophysiology of this illness must account for not
Signaling and Bipolar Disorder BIOL PSYCHIATRY 527
2000;48:518 530

only the profound changes in mood but also a constel- The authors research is supported by the National Institute of Mental
lation of neurovegetative features derived from dys- Health (Grants Nos. RO1-MH57743 and RO1-MH59107 [HKM] and
function in limbic-related regions such as the hip- RO1-MH6247 [RHL]), Theodore and Vada Stanley Foundation (HKM,
pocampus, hypothalamus, and brain stem. The highly RHL), National Alliance for Research on Schizophrenia and Depression
integrated monoamine and prominent neuropeptide (HKM, RHL), and Joseph Young Sr. Research grants (HKM). Outstand-
ing editorial assistance was provided by Celia Knobelsdorf. Space
pathways are known to originate and project heavily limitations necessitate our citing of review articles in many cases. A full
within these regions of the brain, and it is thus not reference list upon which this article was based is available upon request.
surprising that abnormalities have been noted in their Aspects of this work were presented at the conference Bipolar
function across clinical studies. In fact, the contribution Disorder: From Pre-Clinical to Clinical, Facing the New Millennium,
of these pathways to the pathophysiology of BD must January 19 21, 2000, Scottsdale, Arizona. The conference was spon-
sored by the Society of Biological Psychiatry through an unrestricted
be reasonably robust, given the variability that might be educational grant provided by Eli Lilly and Company.
expected in assessing such dynamic systems under the
constraints in experimental design imposed upon such
research. Whereas in the past much of the research References
effort has been focused upon identifying which of these Bebchuk JM, Arfken CL, Dolan-Manji S, Murphy J, Manji HK
systems might be etiologic in nature, we suggest that we (2000): A preliminary investigation of a protein kinase C
might be better served by addressing their relative inhibitor (tamoxifen) in the treatment of acute mania. Arch
contributions in the system response to the underlying Gen Psychiatry 57:9597.
neurobiology of the disease process. This will become Berrettini WH (2000): Are schizophrenic and bipolar disorders
particularly important as we begin to identify the related? A review of family and molecular studies. Biol
Psychiatry 48:531538.
susceptibility genes for BD in the coming years. More
Bhalla US, Iyengar R (1999): Emergent properties of networks of
recently, investigators have begun to identify compo- biological signaling pathways. Science 283(5400):381387.
nents of signal transduction pathways in the brain such
Bourne HR, Nicoll R (1993): Molecular machines integrate
as PKC that may not only play a role in the pathophys- coincident synaptic signals. Cell 72(suppl):6575.
iology of the disease but also represent targets for the Bowden CL (1997): Toward an integrated biological model of
action of our most efficacious treatments (i.e., lithium bipolar disorder. In: Young LT, Joffe RT, editors. Bipolar
and VPA). This experimental strategy may prove to be Disorder: Biological Models and Their Clinical Application.
most promising, since it provides data derived from the New York: Marcel Dekker, 235254.
physiologic response of the system in affected individ- Bunney WE, Bunney BG (2000): Molecular clock genes in man
uals and addresses the critical dynamic intereraction and lower animals. Possible implications for circadian abnor-
malities in depression. Neuropsychopharmacology 22:335
with pharmacologic agents that effectively modify the 345.
clinical expression of the pathophysiology. It is also
Chen G, Hasanat KA, Bebchuk JM, Moore GJ, Glitz D, Manji
noteworthy that a growing body of data is demonstrat- HK (1999): Regulation of signal transduction pathways and
ing significant reductions in regional CNS volume and gene expression by mood stabilizers and antidepressants.
cell numbers (both neurons and glia) in BD (Drevets et Psychosom Med 61:599 617.
al 1997; Manji et al, in press; Rajkowska et al 1999). Chen G, Hawver D, Wright C, Potter WZ, Manji HK (1996):
Furthermore, it has been increasingly recognized that Attenuation of adenylyl cyclases by carbamazepine in vitro.
for many patients there is a progressive decline in J Neurochem 67:2079 2086.
overall functioning; the potential role of cell death/ Chen G, Masana M, Manji HK (in press): Lithium regulates
PKC-mediated intracellular cross-talk and gene expression in
atrophy in determining long-term outcome remains to
the CNS in vivo. Bipolar Disord.
be elucidated. Nevertheless, it is noteworthy that psy-
Chen G, Rajkowska G, Du F, Seraji-Bozorgzad N, Manji HK (in
chopharmacologic research strategies have most re- press): Enhancement of hippocampal neurogenesis by lith-
cently identified long-term actions of these mood sta- ium. J Neurochem.
bilizers in the regulation of expression of genes Colin SF, Chang HC, Mollner S, Pfeuffer T, Reed RR, Duman
implicated in processes involved in neuroplasticity, RS, Nestler EJ (1991): Chronic lithium regulates the expres-
neuroprotection, and even neurogenesis (Chen et al, in sion of adenylate cyclase and Gi-protein alpha subunit in rat
press; Lenox and Watson 1994; Manji et al 1999, in cerebral cortex. Proc Natl Acad Sci U S A 88:10634 10637.
press; Moore et al, in press). To what extent these Daigen A, Akiyama K, Otsuki S (1991): Long-lasting change in
the membrane-associated protein kinase C activity in the
long-term processes in the brain are critical to the
hippocampal kindled rat. Brain Res 545:131136.
ability of these drugs to compensate for the pathophys-
Depue RA, Karuss SP, Spoont MR (1987): A two-dimensional
iology of BD as well as stabilize the clinical course and threshold model of seasonal bipolar affective disorder. In:
progression of BD has yet to be defined (Manji et al Magnusson D, Ohman A, editors. Psychopathology: An Inter-
1999). national Perspective. Orlando, FL: Academic Press, 95123.
528 BIOL PSYCHIATRY H.K. Manji and R.H. Lenox
2000;48:518 530

Divish MM, Sheftel G, Boyle A, Kalasapudi VD, Papolos DF, lying mood-stabilization in manic-depressive illness: The
Lachman HM (1991): Differential effect of lithium on fos phenotype challenge. Am J Psychiatry 156:1506 1514.
protooncogene expression mediated by receptor and postre- Jensen JB, Mikkelsen JD, Mork A (2000): Increased adenylyl
ceptor activators of protein kinase C and cyclic adenosine cyclase type 1 mRNA, but not adenylyl cyclase type 2 in the
monophosphate: model for its antimanic action. J Neurosci rat hippocampus following antidepressant treatment. Eur
Res 28:40 48. Neuropsychopharmacol 10:105111.
Dowlatshahi D, MacQueen GM, Wang JF, Young LT (1998): Jope RS (1999a): Anti-bipolar therapy: Mechanism of action of
Increased temporal cortex CREB concentrations and antide- lithium. Mol Psychiatry 4:117128.
pressant treatment in major depression. Lancet 2352(9142):
1754 1755. Jope RS (1999b): A bimodal model of the mechanism of action
of lithium. Mol Psychiatry 4:2125.
Drevets WC, Price JL, Simpson JR Jr, Todd RD, Reich T,
Vannier M, Raichle ME (1997): Subgenual prefrontal cortex Jope RS, Williams MB (1994): Lithium and brain signal trans-
abnormalities in mood disorders. Nature 386:824 827. duction systems. Biochem Pharmacol 77:429 441.
Kantor L, Gnegy ME (1998): Protein kinase C inhibitors block
Dubovsky S, Murphy J, Christiano J, Lee C (1992): The calcium
amphetamine-mediated dopamine release in rat striatal slices.
second messenger system in bipolar disorders: Data supporting
J Pharmacol Exp Ther 284:592598.
new research directions. J Neuropsychiatry Clin Neurosci 4:3
14. Kofman O, Li PP, Warsh JJ (1998): Lithium, but not carbamaz-
epine, potentiates hyperactivity induced by intra-accumbens
Duman RS, Heninger GR, Nestler EJ (1997): A molecular and
cholera toxin. Pharmacol Biochem Behav 59:191200.
cellular theory of depression. Arch Gen Psychiatry 54:597 606.
Lenox RH (1987): Role of receptor coupling to phosphoinositide
Ebstein RP, Oppenheim G, Ebstein BS, Amiri Z, Stessman J
metabolism in the therapeutic action of lithium. Adv Exp Med
(1986): The cyclic AMP second messenger system in man:
Biol 221:515530.
The effects of heredity, hormones, drugs, aluminum, age and
disease on signal amplification. Prog Neuropsychopharmacol Lenox RH, Manji HK (1998): Lithium. In: Nemeroff CB,
Biol Psychiatry 10:323353. Schatzberg AF, editors. American Psychiatric Press Textbook
of Psychopharmacology, 2nd ed. Washington, DC: American
Ehlers CL, Frank E, Kupfer DJ (1988): Social zeitgebers and
Psychiatric Press, 379 430.
biological rhythms. A unified approach to understanding the
etiology of depression. Arch Gen Psychiatry 45:948 952. Lenox RH, Watson DG (1994): Lithium and the brain: A
psychopharmacological strategy to a molecular basis for
Einat H, Kofman O, Belmaker RH (in press): Animal models of
manic depressive illness. Clin Chem 40:309 314.
bipolar disorder: From a single episode to progressive cycling
models. In: Myslobodsky MS, Weiner I, editors. Contempo- Lenox RH, Watson DG, Patel J, Ellis J (1992): Chronic lithium
rary Issues in Modeling Psychopathology. Boston: Kluwer administration alters a prominent PKC substrate in rat hip-
Academic. pocampus. Brain Res 570:333340.
Ellis J, Lenox RH (1991): Receptor coupling to G proteins: Magnier C, Corvazier E, Aumont MC, Le Jemtel TH, Enouf J
Interactions not affected by lithium. Lithium 2:141147. (1995): Relationship between Rap1 protein phosphorylation
and regulation of Ca2 transport in platelets: A new ap-
Emamghoreishi M, Schlichter L, Li PP, Parikh S, Sen J, Kamble proach. Biochem J 310:469 475.
A, Warsh JJ (1997): High intracellular calcium concentrations
in transformed lymphoblasts from subjects with bipolar I Mandell AJ, Knapp S, Ehlers C, Russo PV (1984): The stability
disorder. Am J Psychiatry 154:976 982. of constrained randomness: Lithium prophylaxis at several
neurobiological levels. In: Post RM, Ballenger JC, editors.
Fields A, Li PP, Kish SJ, Warsh JJ (1999): Increased cyclic Neurobiology of Mood Disorders. Baltimore: Williams &
AMP-dependent protein kinase activity in postmortem brain Wilkins, 744 776.
from patients with bipolar affective disorder. J Neurochem
73:1704 1710. Manji HK (1992): G proteins: Implications for psychiatry. Am J
Psychiatry 149:746 760.
Friedman E, Hoau YW, Levinson D, Connell TA, Singh H (1993):
Altered platelet protein kinase C activity in bipolar affective Manji HK, Chen G, Hsiao JK, Masana MI, Moore GJ, Potter WZ
disorder, manic episode. Biol Psychiatry 33:520 525. (2000): Regulation of signal transduction pathways by mood
stabilizing agents: Implications for the pathophysiology and
Friedman E, Wang HY (1996): Receptor-mediated activation of treatment of bipolar affective disorder. In: Manji HK, Bow-
G proteins is increased in postmortem brains of bipolar den CL, Belmaker RH, editors. Bipolar Medications: Mech-
affective disorder subjects. J Neurochem 67:11451152. anisms of Action. Washington, DC: American Psychiatric
Goodwin FK, Jamison KR (1990): Manic-Depressive Illness. Press, 129 177.
New York: Oxford University Press. Manji HK, Lenox RH (1999): Protein kinase C signaling in the
Holsboer F (1995): Neuroendocrinology of mood disorders. In: brain: Molecular transduction of mood stabilization in the
Bloom FE, Kupfer DJ, editors. Psychopharmacology: The treatment of bipolar disorder. Biol Psychiatry 46:1328 1351.
Fourth Generation of Progress. New York: Raven, 957969. Manji HK, Moore GJ, Chen G (1999): Lithium at 50: Have the
Hough C, Lu SJ, Davis CL, Chuang DM, Post RM (1999): neuroprotective effects of this unique cation been over-
Elevated basal and thapsigargin stimulated intracellular cal- looked? Biol Psychiatry 46:929 940.
cium of platelets and lymphocytes from bipolar affective Manji HK, Moore GJ, Chen G (in press): Clinical and preclinical
disorder patients measured by a fluorometric microassay. Biol evidence for the neurotrophic effects of mood-stabilizing
Psychiatry 46:247255. agents: Implications for the pathophysiology and treatment of
Ikonomov O, Manji HK (1999): Molecular mechanisms under- manic-depressive illness. Biol Psychiatry.
Signaling and Bipolar Disorder BIOL PSYCHIATRY 529
2000;48:518 530

Manji HK, Potter WZ (1997): Monoaminergic mechanisms. In: Perez J, Tardito D, Mori S, Racagni G, Smeraldi E, Zanardi R
Joffe RT, Young LT, editors. Bipolar Disorder: Biological (2000b): Altered Rap1 endogenous phosphorylation and lev-
Models and Their Clinical Application. New York: Marcel els in platelets from patients with bipolar disorder. J Psychiatr
Dekker, 1 40. Res 34:99 104.
Masana MI, Bitran JA, Hsiao JK, Potter WZ (1992): In vivo Popoli M, Brunello N, Perez J, Racagni G (2000): Second
evidence that lithium inactivates Gi modulation of adenylate messenger-regulated protein kinases in the brain: Their func-
cyclase in brain. J Neurochem 59:200 205. tional role and the action of antidepressant drugs. J Neuro-
Mathews R, Li PP, Young LT, Kish SJ, Warsh JJ (1997): chem 74:2133.
Increased G alpha q/11 immunoreactivity in postmortem Post RM, Chuang D-M (1991): Mechanism of action of lithium:
occipital cortex from patients with bipolar affective disorder. Comparison and contrast with carbamazepine. In: Birch NJ,
Biol Psychiatry 41:649 656. editor. Lithium and the Cell: Pharmacology and Biochemis-
try. London: Academic Press, 199 241.
Milligan G, Wakelam M (1992): G Proteins: Signal Transduc-
tion & Disease. San Diego: Academic Press. Post RM, Weiss SRB (1999): Neurobiological models of recur-
rence in mood disorder. In: Charney DS, Nester EJ, Bunney
Mooney JJ, Samson JA, McHale NL, Colodzin R, Alpert J,
BS, editors. Neurobiology of Mental Illness. New York:
Koutsos M, Schildkraut JJ (1998): Signal transduction by
Oxford University Press, 365384.
platelet adenylate cyclase: Alterations in depressed patients
may reflect impairment in the coordinated integration of Post RM, Weiss SRB, Clark M, Chuang DM, Hough C, He L
cellular signals (coincidence detection). Biol Psychiatry 43: (2000): Lithium, carbamazepine and valproate in affective
574 583. illness: Biochemical and neurobiological mechanisms. In:
Manji HK, Bowden CL, Belmaker RH, editors. Bipolar
Moore GJ, Bebchuk JM, Parrish JK, Faulk MW, Arfken CL, Medications: Mechanisms of Action. Washington, DC: Amer-
Strahl-Bevacqua J, Manji HK (1999): Temporal dissociation ican Psychiatric Press, 219 248.
between lithium-induced CNS myo-inositol changes and
clinical response in manic-depressive illness. Am J Psychiatry Rahman S, Li PP, Young LT, Kofman O, Kish SJ, Warsh JJ
156:19021908. (1997): Reduced [3H] cyclic AMP binding in postmortem
brain from subjects with bipolar affective disorder. J Neuro-
Moore GJ, Bebchuk JM, Wilds IB, Chen G, Manji HK (in press): chem 68:297304.
Pharmacologic increase in human gray matter. Lancet.
Rajkowska G, Miguel-Hidalgo JJ, Wei J, Dilley G, Pittman SD,
Mork A, Geisler A (1989): The effects of lithium in vitro and ex Meltzer HY, et al (1999): Morphometric evidence for neuro-
vivo on adenylate cyclase in brain are exerted by distinct nal and glial prefrontal cell pathology in major depression.
mechanisms. Neuropharmacology 28:307311. Biol Psychiatry 45:10851098.
Mork A, Geisler A, Hollund P (1992): Effects of lithium on Rasenick MM, Chen J, Ozawa H (2000): Effects of antidepres-
second messenger systems in the brain. Pharmacol Toxicol sant treatments on the G protein-adenylyl cyclase axis as the
71(suppl 1):4 17. possible basis of therapeutic action. In: Manji HK, Bowden
Nestler EJ, Terwilliger RZ, Duman RS (1989): Chronic antide- CL, Belmaker RH, editors. Bipolar Medications: Mecha-
pressant administration alters the subcellular distribution of nisms of Action. Washington, DC: American Psychiatric
cyclic AMP-dependent protein kinase in rat frontal cortex. Press, 87108.
J Neurochem 53:1644 1647. Risby ED, Hsiao JK, Manji HK, Bitran J, Moses F, Zhou DF,
Nestler EJ, Terwilliger RZ, Duman RS (1995): Regulation of Potter WZ (1991): The mechanisms of action of lithium. Arch
endogenous ADP-ribosylation by acute and chronic lithium in Gen Psychiatry 48:513524.
rat brain. J Neurochem 64:2319 2324. Shibata K, Morita K, Kitayama S, Okamoto H, Dohi T (1996):
Nestler EJ, Terwilliger RZ, Walker JR, Sevarino KA, Duman RS Ca2 entry induced by calcium influx factor and its regula-
(1990): Chronic cocaine treatment decreases levels of the G tion by protein kinase C in rabbit neutrophils. Biochem
protein subunits Gi alpha and Go alpha in discrete regions of Pharmacol 52:167171.
rat brain. J Neurochem 55:1079 1082. Si-Tahar M, Renesto P, Falet H, Rendu F, Chignard M (1996):
Newman ME, Belmaker RH (1987): Effects of lithium in vitro The phospholipase C/protein kinase C pathway is involved in
and ex vivo on components of the adenylate cyclase system in cathepsin G-induced human platelet activation: comparison
membranes from the cerebral cortex of the rat. Neurophar- with thrombin. Biochem J 313:401 408.
macology 26:211217. Spiegel AM (1998): G Proteins, Receptors, and Disease. To-
Ozaki N, Chuang D-M (1997): Effect of lithium on transcription towa, NJ: Humana Press.
factor binding in cultured cerebellar granule cells and the Spleiss O, van Calker D, Scharer L, Adamovic K, Berger M,
brain of rats. J Neurochem 69:2336 2344. Gebicke-Haerter PJ (1998): Abnormal G protein alpha(s)-
and alpha(i2)-subunit mRNA expression in bipolar affective
Ozaki N, Manji HK, Luiberman V, Lu S, Rosenthal N, Goldman
disorder. Mol Psychiatry 3:512520.
D (1997): A naturally occurring amino acid substitution of the
human serotonin (5-HT) 2A receptor influences amplitude Steketee JD, Kalivas PW (1991): Sensitization to psychostimu-
and timing of intracellular calcium mobilization. J Neuro- lants and stress after injection of pertussis toxin into the A10
chem 68:2186 2193. dopamine region. J Pharmacol Exp Ther 259:916 924.
Perez J, Tardito D, Mori S, Racagni G, Smeraldi E, Zanardi R Wang HY, Friedman E (1996): Enhanced protein kinase C
(2000a): Abnormalities of cAMP signaling in affective dis- activity and translocation in bipolar affective disorder brains.
orders: Implications for pathophysiology and treatment. Bi- Biol Psychiatry 40:568 575.
polar Disord 2:2736. Wang HY, Friedman E (1999): Effects of lithium on receptor-
530 BIOL PSYCHIATRY H.K. Manji and R.H. Lenox
2000;48:518 530

mediated activation of G proteins in rat brain cortical mem- depressive illness. In: Wehr TA, Goodwin FK editors. Cir-
branes. Neuropharmacology 38:403 414. cadian Rhythms in Psychiatry. Pacific Grove, CA: Boxwood
Wang JF, Asghari V, Rockel C, Young LT (1999): Cyclic AMP Press, 129 184.
responsive element binding protein phosphorylation and Wehr TA, Goodwin FK (1987): Can antidepressants cause mania
DNA binding is decreased by chronic lithium but not val- and worsen the course of affective illness? Am J Psychiatry
proate treatment of SH-SY5Y neuroblastoma cells. Neuro- 144:14031411.
science 91:771776.
Weintraub BD (1995): Molecular Endocrinology: Basic Con-
Wang JF, Young LT, Li PP, Warsh JJ (1997): Signal transduc- cepts and Clinical Correlations. New York: Raven.
tion abnormalities in bipolar disorder. In: Joffe RT, Young
LT, editors. Bipolar Disorder: Biological Models and Their Weng G, Bhalla US, Iyengar R (1999): Complexity in biological
Clinical Application. New York: Marcel Dekker, 4179. signaling systems. Science 284(5411):9296.
Warsh JJ, Young LT, Li PP (2000): Guanine nucleotide binding Wiborg O, Kruger T, Jakobsen SN (1999): Region-selective
(G) protein disturbances in bipolar affective disorder. In: effects of long-term lithium and carbamazepine administra-
Manji HK, Bowden CL, Belmaker RH, editors. Bipolar tion on cyclic AMP levels in rat brain. Pharmacol Toxicol
Medications: Mechanisms of Action. Washington, DC: Amer- 84:88 93.
ican Psychiatric Press, 299 329. Wright AF, Crichton DN, Loudon JB, Morten JE, Steel CM
Watson DG, Lenox RH (1996): Chronic lithium-induced down- (1984): Beta-adrenoceptor binding defects in cell lines from
regulation of MARCKS in immortalized hippocampal cells: families with manic-depressive disorder. Ann Hum Genet
Potentiation by muscarinic receptor activation. J Neurochem 48:201214.
67:767777. Young LT, Li PP, Kish SJ, Siu KP, Kamble A, Hornykiewicz O,
Watson DG, Watterson JM, Lenox RH (1998): Sodium valproate Warsh JJ (1993): Cerebral cortex Gs protein levels and
down-regulates the myristoylated alanine-rich C kinase sub- forskolin-stimulated cyclic AMP formation are increased in
strate (MARCKS) in immortalized hippocampal cells. A bipolar affective disorder. J Neurochem 61:890 898.
property of protein kinase C-mediated mood stabilizers. Young LT, Woods CM (1996): Mood stabilizers have differen-
J Pharmacol Exp Ther 285:307316. tial effects on endogenous ADP ribosylation in C6 glioma
Wehr TA, Goodwin FK (1983): Biological rhythms in manic- cells. Eur J Pharmacol 309:215218.