CLIN.CHEM.40/2, 309-314 (1994)

CLINICALCHEMISTRY,Vol.40, No.2, 1994 309

Lithium and the Brain: A PsychopharmacologicalStrategy to a Molecular Basis forManic Depressive IllnessRobert H. Lenox1 and David G. Watson

Lithium, an effective treatment for mania and the preven-tion of recurrent episodes of both mania and depression inpatients with manic depressive illness, exerts multiplebiochemical effects. However, any clinically relevant siteof action of lithium must occur at therapeutic concentra-tions attained in the brain of patients and must account forthe lag period accompanying onset of action as well aseffects persisting beyond discontinuation of treatment.This monovalent cation acts as a potent uncompetitiveinhibitor in the receptor-coupled breakdown of inositolphospholipids, resulting in a relative depletion of inositoland an alteration in the generation of diacylglycerol, anendogenous activator of protein kinase C. In our labora-tory, we are examining the action of chronically adminis-tered lithium on posttranslational modification of specificphosphoproteins involved in regulating signal transduc-tion in the brain. We have found that chronic, but notacute, administrationof lithiumin rats markedly reduces amajor phosphoproteinsubstrate of protein kinase C in thehippocampus, an effect that persists beyond the cessa-tion of lithium treatment. This protein, myristoylated ala-nine-rich C kinase substrate (“MARCKS”), is implicated insynaptic neurotransmission,calcium regulation, and cyto-skeletal restructuring. These findings have relevance forthe long-term action of lithium in stabilizing an underlyingdysregulation in the brain and may move us closer toformulating a molecular basis of manic depressive illness.

IndexingTerms: protein kinase C/rat models/phosphoproteins/inositol phospholipids/mood disorders

HIstorIcal Perspective

Lithium was discovered as an element in 1817 andhas been used to treat a multitude of somatopsychicdisorders over the years (1). In particular, because lith-ium forms the most soluble salt known with uric acid, itwas proposed as early as the mid-l9th century as atreatment for gout or “uric acid diathesis,” which wasdescribed as including mood-related symptoms rangingfrom “gouty mania” to depression. By the late 19thcentury, physicians such as John Aulde in America andCarl Lange in Denmark reported clinical observationsrelated to the prophylactic efficacy of lithium in thetreatment of recurrent symptoms of depression (2, 3).The clinical use of lithium in conventional medicine fellinto disrepute in this country in the early and middle20th century, in light of exaggerated claims for its ther-apeutic efficacy as well as its potential for toxicity.

Division of Neuroscience, Department of Psychiatry, Universityof Vermont College of Medicine, Burlington, VT 05405.

‘Author for correspondence. Fax 802-656-0987.Received August 11, 1993; accepted September 22, 1993.

Widespread and unregulated marketing of “lithia salts”

in potions and mineral waters, often at negligible con-centrations, for a variety of ills contributed to the gen-eral disifiusionment of the traditional medical estab-lishment. The use of lithium salt as a substitute forsodium to treat patients with chronic cardiac and renaldiatheses resulted in an increased incidence of morbid-ity and mortality related directly to lithium toxicity.

Not until John Cade reported his observations in 1949was lithium considered once again a potential pharma-cotherapeutic agent for treating mania in patients withmanic depressive illness (4). In a series of studies of theuric acid diathesis of mania, Cade injected urine frommanic patients into guinea pigs. To solubilize the uricacid, he used lithium, and subsequently reported thatwithin 2 h of administration of lithium the animalsbecame lethargic and unresponsive but thereaftershowed full recovery. This seminal observation resultedin his administering lithium to a series of 10 manicpatients and reporting a rather startling reduction ofpsychotic excitement in all of them. Later, in a series ofnow-famous clinical investigations, Mogens Schou es-tablished lithium as an effective antimanic treatmentand a prophylactic therapy for recurrent affective epi-sodes in patients with manic depressive ifiness (5). Be-cause of its past history, however, the introduction oflithium into the clinical practice of psychiatry as a psy-chopharmacological treatment for manic depressive ill-ness in this country was delayed almost a decade, to themid-1960s.

ClInical Efficacy

Lithium, one of the most clinically efficacious phar-macological agents now available for treating a majorpsychiatric disorder, has been used clinically with avariety of behavioral disorders. Clinical indications forthe treatment of affective illness include:

1. Long-term treatment to reduce the frequency andseverity of recurrent episodes of mania and depressionin bipolar illness.

2. Acute treatment of mania in which the antimanic

effect is more specific than that of neuroleptics but has alonger latency of action (5-10 days).

3. Long-term prophylaxis for recurrent episodes ofdepression in umpolar illness.

4. Adjunctive treatment with antidepressants for re-fractory depressive episodes.Lithium is most effective and selective for the long-termprophylactic treatment of recurrent episodes of both ma-nia and depression in patients diagnosed with bipolardisorder (6). The unique clinical action of lithium in thetreatment of the profound mood cycling seen in this

310 CLINICAL CHEMISTRY, Vol. 40, No. 2, 1994

disorder requires a lag period for onset and is not im-mediately reversed upon discontinuation of treatment.In properly screened patient populations, published re-ports support its efficacy in >80% of patients studied (6,7). Lithium is also effective in the treatment of acutemania in this same patient population. Because its on-set of action is delayed for between 5 and 10 days,lithium is often initially combined with other medica-tions, e.g., benzodiazepines, to achieve adequate clinicalmanagement of the patient (8). To a lesser but clinicallyrelevant extent, lithium is also used to prevent recur-rent episodes of depression in patients with unipolar

disorder. This clinical treatment profile of lithiumclearly underscores its long-term properties as a moodstabilizer.

Several other characteristics of chronic treatmentwith lithium are also important in understanding itsmechanism of action. If lithium is abruptly withdrawn

from a patient with bipolar disorder after long-termtreatment, affective symptomatology does not immedi-ately reappear. However, there is now clear evidencethat as many as 50% of such patients will experience arecurrence of either a manic or depressive episode within5 months, with a clearly higher risk for a manic episodeearly in the withdrawal period (9, 10). Lithium also hasa rather narrow therapeutic index, clinical toxicity beingevident in some patients at plasma concentrations 50%

greater than the therapeutic range; thus, its therapeuticaction may be on a continuum with some of its toxicmanifestations. Our laboratory has been interested inaddressing the biological targets responsible for the long-term therapeutic action of lithium in the brain.

Mechanism of Action

The underlying biological processes responsible forthe episodic clinical manifestation of mania and depres-sion may be related to faulty homeostatic regulation inthe brain (6). Patients with this disorder may lack thenecessary system flexibility to adaptively respond to theperiodic fluctuations in the internal and external state,but instead experience sudden oscillations beyond im-mediate compensatory control (11, 12). This system fail-ure results in the clinical manifestations of disruption ofbehavior and profound changes in mood, circadianrhythm, and neurophysiology of sleep, as well as signif-icant alterations in neuroendocrine regulation, all con-sistent with a dysregulation within the limbic systemand associated regions of the brainstem and prefrontalcortex. A better understanding of the molecular mech-anisms by which lithium achieves its mood stabifization

properties in these regions of the brain might not onlylead to the development of even better psychopharma-cological treatment approaches but also bring us closerto a molecular basis of the disorder itself. This researchstrategy has the advantage of focussing the investigatorwithin the complex array of neurochemical systemswithin the brain, but has limitations in terms of extrap-olation to the pathogenesis of the disorder, as will bediscussed below.

Signal Transduction

Regulation of signal transduction within limbic andlimbic-related regions of the brain, altering the balanceof activity of the complex network of neurotransmitterpathways, remains an attractive site for the therapeuticaction of a drug like lithium. Over the past severalyears, emerging data have confirmed that the phospho-inositide second-messenger system is an important siteof action of lithium in the brain (see Fig. 1) (13, 14).Lithium, at therapeutically relevant concentrations, isa potent inhibitor of the intracellular enzyme inositolmonophosphatase (K1, 0.8 mmol/L), which results in anaccumulation of inositol 1-monophosphate and a reduc-tion in the generation of free inositol (15-17). Becausethe brain has limited access to inositol other than thatderived from recycling of the inositol polyphosphates,the ability of neurons to maintain sufficient supplies ofmyo-inositol can be crucial to the resynthesis of thephosphoinositides and thus maintenance and efficiencyof signaling (18). Furthermore, because the mode ofenzyme inhibition is uncompetitive, lithium’s effectshave been postulated to be most pronounced in systemsundergoing the highest rate of hydrolysis of phosphati-dylinositol 4,5-bisphosphate (PIP2), wherein the physi-ological consequence of lithium’s action in brain is de-rived through the relative depletion of free inositol.2Thus, the selective action of lithium has to be attributedto its preferential action in brain, which results in sup-pressing the hydrolysis of phosphatidylinositol (P1) inthe most overactive receptor-mediated neuronal path-ways, and thereby modulating their relative activity

2NonsdJ abbreviations: PKC, protein kinase C; MA.RCKS,

myristoylated alanine-rich C kinase substrate; P1, phosphatidyl-inositol; PIP2, phosphatidylinositol 4,5-bisphosphate; CMP-PA, cy-tidine monophosphate-phosphatidate; DAG, diacyiglycerol; andPA, phosphotidate.

Fig. 1. Agonist occupancy of the receptor binding site results in Gprotein-mediated coupling and activation of phospholipase C (PLC),which induces the hydrolysisof phosphatidylinositol4,5-bisphosphate(PIP2)and the generationof two second-messengers,i.e., diacylglyc-erol (DAG)and myo-inositol1,4,5-tiisphosphate(Ins 1,4,5 P2).Ins1,4.5P3bindsto receptorsin theendoplasmicreticulum(ER) and releasesIntracellularcalcium.DAGisanendogenousactivatorofproteinkinase C (PKC).Lithium inhibits the enzyme inositolmonophosphatase,which prevents the re-cyding of inositoland increasesaccumulationof metabolitesin the DAG path-way, potentiallyreducingcritical pools of PIP2requiredfor signaltransduction.PIP, phosphatidylinositol4-phosphate;G, a Gm-bindingprotein,e.g., Gp, Go;A, receptor;A, a receptor-specificagonist;Ins 1P. myo-inosrtol1-phosphate;Ins3P, myo-inositol3-phosphate;Ins 4P, myo.inositol4-phosphate.

150

125

No Inositol- + Inositol

75

0 25 10

lime (minutes)

of CMP-PA formation in lithium-exposed

CLINICALCHEMISTRY,Vol.40, No. 2, 1994 311

(13, 19). Because several subtypes of adrenergic (e.g.,a1), cholinergic (e.g., m1, m3, m5), serotonergic (e.g.,5HT2, 5HT1), and dopaminergic (e.g., D1) receptors arecoupled to PIP2 turnover in the central nervous systemand mediate both excitatory and inhibitory pathways(14,20-22), such a hypothesis offers a plausible expla-nation for the therapeutic efficacy of lithium in treatingboth poles of manic depressive illness by the compensa-tory stabilization of an inherent biogenic amine imbal-ance in critical regions of the brain (23).

PhosphoinositideSignalingThe evidence discussed above suggests that at least

some of lithium’s critical effects are mediated via long-term processes set in motion during exposure of thebrain to lithium. Although it is still difficult to demon-strate that the effect of chronic lithium on inositol can bea meaningful reduction of a phosphoinositide (e.g., PIP2)pool associated with signal transduction, there is evi-dence that it may mediate long-term events via post-translational changes in the phosphorylation of selec-tive protein substrates. As a consequence of lithium-induced inositol depletion in the presence of continuedreceptor-mediated hydrolysis of phosphoinositides (PIP2),diacyiglycerol (DAG) and its related metabolites, phos-phatidate (PA) and cytidine monophosphate-phosphati-date (CMP-PA), accumulate in a variety of cell types,including the brain, presumably because of a reductionin the recycling of inositol into critical PIP2 pools (seeFig. 1). These findings are further supported by datademonstrating that this increase can be reversed by theaddition of inositol (24-28). Studies in our own labora-tory have replicated these observations in a CHO-K1cell line transfected with the human m1 muscarinicreceptor subtype, which is known to couple to P1 re-sponse (29,30). As shown in Fig. 2, time-course analysisdetermined that, in the presence of lithium, theCMP-PA response to carbachol under these conditionswas very rapid (significant accumulation as early as 2

o 10000

‘C

a-0

50

25

Fig. 2. Time courseCHO-Ki (Hml) cells.Cells were grown in inositol-free Ham’s Fl 2 mediumsupplementedwithfetalbovine serum (25 milL). Cells were prelabeled with IHJcytidine for 60 mmbefore theadditionof UCI (5 mmol/L),with (U) orwithout(0) additional inositol(10 mmol/L).After 10 mm,carbachol (1mmol/L) was addedand the incubationcontinued for thetimesindicated.The reactionswere stopped by theadditionof chloroform/methanol(1/2 by vol). CMP-PA was extracted as described byGodfrey (26)and quantified by liquid scintillation spectrometry.

mm after the addition of carbachol) and could be signif-icantly reduced by the addition of inositol. We have alsoobserved that CMP-PA formation in cells grown andassayed in standard Ham’s F12 medium supplementedwith fetal bovine serum (100 milL) was markedly lessthan in cells grown in inositol-depleted media.

Protein Kinase C (PKC)Inasmuch as DAG is a known endogenous activator of

PKC, these findings are consistent with our earlier ob-servations that the action of chronic lithium may steminitially from its potent effects in inhibiting the recy-cling of inositol through the receptor-mediated hydroly-sis of PIP2 and its indirect action in accumulating DAG(23, 31) with subsequent changes in the activation ofPKC isoenzymes, altering the relative “constitutivephosphorylation” of key phosphoprotein substrates (32).Data from several laboratories, including our own, pro-vide convincing evidence for a role for PKC in mediatingthe effects of lithium exposure in various cell systemsincluding the brain (23, 33-38). PKC-mediated post-translational modification of proteins involved in re-ceptor-mediated signal transduction, neurotransmitterrelease, ion-channel activity, and transcriptional regu-lation have major physiological implications for neu-ronal function throughout the brain (39).

Because alterations in endogenous PKC substratesmight reveal both subtle and selective effects of lithiumon PKC activity, we examined the phosphoprotein sub-strates for PKC in the soluble and membrane fractionsfrom hippocampi of control animals and animals givenchronic lithium treatment (4 weeks), as well as from ratstreated acutely with lithium (2 h) and rats withdrawnfrom chronic lithium administration for 40 h. We found asignificant reduction (45%) in the in vitro phosphoryla-tion of a prominent protein with an apparent molecularmass of 83 kDa in the soluble fraction of hippocampi fromanimals subjected to chronic lithium treatment (37).Other phosphoproteins, including other PKC substrates,were not changed by the chronic lithium treatment. An-imals acutely achieving similRr therapeutic concentra-tions of lithium in the brain did not demonstrate a sig-nificant change in phosphorylation of these proteins.Animals withdrawn from lithium showed a partial rever-sal of the effects on the 83-kDa protein seen duringchronic lithium treatment. The phosphoprotein appearedto correspond to a major myristoylated alanine-rich Ckinase substrate (MARCKS) in brain. This was subse-quently confirmed by using an antibody raised to highlypurified MARCKS that recognizes both phosphorylatedand nonphosphorylated forms of the protein. Westernblot analyses provided evidence for a marked reductionin the amount of MARCKS in both the soluble (57%) andmembrane (7 1%) fraction from hippocampus of rats re-ceiving chronic lithium treatment (Table 1). However,there was no such evidence for a reduction of MARCKSin rats receiving acute doses of lithium, and the reductionobserved in the hippocampus of rats receiving repeateddoses of lithium was stifi present at 40 h after abruptdiscontinuation of the lithium diet. These findings ap-

Soluble

89.4 ± 6.443.3 ± 32b

499 ±

Pellet

96.4 ± 12.028.5 ± 48b

25.9 ± 68b

References1. Johnson FN. The history of lithium therapy. London/Basing-stoke: Macmillan, 1984:198 pp.2. Lange, C. Om periodiske depressionstilstande og deres patoge-nese. Copenhagen: Jacob Lunds Forlag, 1886:60 pp.3. Aulde J. The use of lithium bromide in combination withsolution of potassium citrate. Med Bull (Phila) 1887;9:35-9, 69-72,228.-33.4. Cede JFJ. Lithium salts in the treatment of psychotic excite-ment. Med J Aust 1949;36:349-52.5. Schou M. Lithium research at the Psychopharmacology Re-

312 CLINICAL CHEMISTRY, Vol.40, No.2, 1994

Table 1. ScannIng laser densitometric analysis ofWestern blots: relatIve density of MARCKS protein

Immunostalned bands from hippocampus.

Tr.atm.nr

Acute lithIum (3)Chroniclithium(7)Uthium withdrawal(3)

% of control (mean ± SEM)

#{149}Numberof separate expeilments is listed in parentheses.b SignifIcantlydifferent from control: P <0.001.

pear to parallel the clinical efficacy of lithium in thetreatment of bipolar disorder as noted above.

MARCKS is a well-known acidic protein that is one ofthe major substrates for PKC in brain; it belongs to ahighly homologous family of proteins that have wide-spread tissue and subcellular distribution (40, 41). Theprotein has been sequenced and cloned from animal (42,43) and, recently, human brain (44). MARCKS proteinsare phosphorylated reversibly by several pharmacolog-ical agents and growth factors, including muscarinicagonists, vasopressin, bradykinin, and phorbol ester, allof which are known to activate PKC (45-52). The pro-tein undergoes cotranslational myristoylation and pos-sesses a highly conserved lysine-rich region for PKCphosphorylation and the binding of calmodulin. Phos-phorylation appears to regulate calniodulin binding toMARCKS, making this protein, among others, poten-tially instrumental in coordinating PKC and calciumactivation through redistribution of calmodulin for ap-propriate regulation of intracellular signaling (53-55).MARCKS is found in both the soluble and membranefraction of the cell (56). It has been suggested that therelative state of fatty acylation vs phosphorylation maydetermine the subcellular distribution of the protein ata given time (42) and may be associated with the regu-lation of physiological events such as signal transduc-tion and neurotransmitter release in brain (40, 47).More recently, in other cell systems, Thelen et al. (57)suggested that MARCKS may represent a “critical ef-fector molecule in the transduction pathway” of agonist-dependent responses in cells, with additional data sup-porting a prominent role for this protein in plasticprocesses involving membrane-cytoskeletal elements ofthe cell (58, 59). MARCKS is a protein cross-linkingfilainentous actin, where regulation of its cross-bridgingbetween actin and the plasma membrane occurs by bothPKC-mediated phosphorylation and calcium-calmodu-lin binding (60). The abffity of MARCKS to bind bothcalinodulin and actin has implicated MARCKS in sev-eral cellular processes, including secretion and mem-brane trafficking, cell motility, cell cycle regulation, andcellular transformation (61).

Strategy for Drug Development and Pathogenesls ofManic Depressive Illness

The plasticity of synaptic processes responsible forregulating neurotransmission within the limbic systemmust play a significant role in mediating homeostasis of

behavioral-vegetative-emotional responses over time.Cytoskeletal restructuring and membrane traffickingare integral to the function of processes such as neuro-transmitter release and up/down regulation of receptor-response systems, so necessary for appropriate physio-logical adaptation of the organism to both internal andexternal events. In the presence of normal compensa-tory systems and fine-tuning, aberrant signals gener-ated by faulty neurotransmitter systems might be am-plified and degenerate into inappropriate oscillationsbeyond immediate physiological control. An interactionof chronic lithium with proteins critical to such long-term molecular events may represent an attractive tar-get for the therapeutic action of this psychotropic agentin achieving clinical stability in patients with manicdepressive illness.

However, although these proteins may represent anoptimal target for the clinical action of a mood stabilizersuch as lithium, they are not necessarily directly linkedto the pathogenesis of the disorder. The ultimate geneproduct(s) underlying the dysregulation in the brain

fundamental to the clinical manifestations of manic de-pressive illness may be at a level of neuronal expressionquite distant from the adaptive processes responsible forregulating the balance of neurotransmitter activity andsignal transduction. For example, in the case of diabe-tes, endocrinologists already have identified a moleculartarget for their pharmacotherapeutic strateT. Prevent-ing the clinical manifestations of diabetes by closelymonitoring and controlling insulin concentrationswould appear to be separate from understanding thegenetic predisposition associated with the various sub-types of type 1 and type 2 diabetes. Yet it is through ourunderstanding of the genetics of this disorder as well asthe evolving knowledge of the insulin receptor-responsecomplex and regulation of glucose transport that we willeventually discover the etiological processes underlyingthe disease. Similarly, with the identification of themolecular target(s) directly associated with the long-term clinical efficacy of lithium, we will be in a positionto better design future strategies for drug developmentand, in concert with molecular genetic approaches, toaddress the ultimate pathogenesis of manic depressiveillness.

We thank Jit Patel for providing antibody to the MARCKSprotein and John Ellis and Mark Brann for access to the trans-fected CHO-Ki (Hml) cells. In addition, we express our apprecia-tion for the fine technical assistance of Ann Wood and LorraineJean. This work was supported in part by grants R01-M1141571and PHS0705429-30 from the National Institute for MentalHealth.

CLINICALCHEMISTRY, Vol. 40, No. 2, 1994 313

search Unit, Risskov, Denmark: a historical account. In: Schou M,Stromgren E, eds. Origin, prevention and treatment of affectivedisorders. London: Academic Press, 1979:1-8.6. Goodwin FK, Jamison KR. Manic-depressive illness. New York:Oxford University Press, 1990:938 pp.7. Davis JM. Overview: maintenance therapy in psychiatry. II:Affective disorders. Am J Psychiatry 1976;133:1-13.8. Lenox RH, Modell JG, Weiner S. Acute treatment of manicagitation with lorazepam. Psychosomatics 1986;27:28-31.9. Klein E, Lavie P, Meiraz R, Sadeh A, Lenox RH. Increasedmotor activity and recurrent manic episodes: predictors of rapidrelapse in remitted bipolar disorder patients after lithium discon-tinuation. Biol Psychiatry 1992;31:279-84.10. Suppes T, Baldessarini RJ, Faedda (IL, Tohen M. Risk ofrecurrence following discontinuation of lithium treatment in bipo-lar disorder. Arch Gen Psychiatry 1991;48:1082-8.11. Mandell AJ, Knapp S, Ehlers C, Russo PV. The stabifity ofconstrained randomness: lithium prophylaxis at several neurobi-ological levels. In: Post EM, Ballenger JC, eds. Neurobiology ofmood disorders. Baltimore: Wffliams & Wilkins, 1984:744-76.12. Depue RA, Karuss SF, Spoont MR. A two-dimensional thresh-old model of seasonal bipolar affective disorder. In: Magnusson D,Ohman A, eds. Psychopathology: an interactional perspective.Orlando, FL: Academic Press, 1987:95-123.13. Berridge MJ. Inositol triaphosphate, calcium, lithium, and cellsignaling [Review]. JAm Med Assoc 1989;262:1834-41.14. Eana ES, Hokin LB. Role of phosphoinositides in transmem-brane signaling [Review]. Physiol Rev 1990;70:115-64.15. Allison JH, Stewart MA. Reduced brain inositol in lithium-treated rats. Nature New Biol 1971;233:267-8.16. Hallcher LM, Sherman WR. The effects of lithium ion andother agents on the activity of myo-inositol-1-phosphatase frombovine brain. J Biol Chem 1980;255:10896-901.17. Sherman WR, Gish BG, Honchar MP, Munsell LY. Effectsoflithium on phosphoinositide metabolism in vivo. Fed Proc 1986;45:2639-46.18. Sherman WE. Lithium and the phosphoinositide signallingsystem. In: Birch NJ, ed. Lithium and the cell. London: AcademicPress, 1991:121-57.19. Berridge MJ, Downes CP, Hanley MR. Lithium amplifiesagonist-dependent phoephatidylinositol responses in brain andsalivary glands. Biochem J 1982;206:587-95.20. Fisher SK, Heacock AM, Agranoff BW. Inositol lipids andsignal transduction in the nervous system: an update [Review]. JNeurochem 1992;58:18-38.21. Mahan LC, Burch EM, Monsma FJ Jr, Sibley DR. Expressionof striatal D1 dopamine receptors coupled to inositol phosphateproduction and Ca2 mobilization in Xenopus oocytes. Proc NatlAcad Sd USA 1990;87:2196-200.22. Vallar L, Muca C, Magni M, Albert P, Bunzow J, Meldolesi J,Civeffi 0. Differential coupling of dopaminergic D2 receptorsexpressed in different cell types. Stimulation of phosphatidylino-sitol 4,5-bisphosphate hydrolysis in LtK fibroblasta, hyperpolar-ization, and cytosolic-free Ca2 concentration decrease in GH4C1cells. J Biol Chem 1990;265:10320-6.23. Lenox RH. Role of receptor coupling to phosphoino8itide me-tabolism in the therapeutic action of lithium. In: Ehrlich Y, LenoxRH, Kornecki E, Berry WO, eds. Molecular mechanisms of neu-renal responsiveness (Mv Exp Biol Mod, VoL 221). New York:Plenum Press, 1987;515-30.24. Drummond All, Raeburn CA. The interaction of lithium withthyrotropin-releasing hormone-stimulated lipid metabolism inGil3 pituitary tumor cells. Biochem J 1984;224:129-36.25. Downes CP, Stone MA. Lithium-induced reduction in intra-cellular inositol supply in cholinergically stimulated parotidgland. Biochem J 1986;234:199-204.26. Godfrey PP. Potentiation by lithium of CMP-phosphatidateformation in carbachol-stimulated rat cerebral-cortical slices andits reversal by myo-inoeitoL Biochem J 1989;258:621-4.27. Watson SP, Shipman L, Godfrey PP. Lithium potentiateaagonist formation of [3H]CDP-diacylglycerol in human platelets.Eur J Pharmacol 1990;188:273-6.28. Brami BA, Leli U, Hauser G. Origin of the diacylglycerolproduced in excess of inositol phosphates by lithium in NG1OS-15cells. J Neurochem 1991;57(Suppl):S9.29. Buckley NJ, Banner TI, Buckley CM, Brann MR. Antagonist

binding properties of five cloned muscarinic receptors expressed inCHO-Kl cells. Mol Pharmacol 1989;35:469-76.30. Jones SVP, Levey Al, Weiner DM, El]is J, Novotny E, Yu S-H,et al. Muscarinic acetylcholine receptors. In: Brann MR, ed.Molecular biology of G-protein-coupled receptors. New York:Birkhauser, 1992:170-97.31. Lenox RH. Lithium interaction with the phosphoinositidesystem and protein kinase C in the nervous system. In: BelmakerRH, Sandier M, Dahlstrom A, eds. Progress in catecholamineresearch (Neurol Neurobiol, Vol. 42C). New York: Alan Lisa,1988:327-35.32. Blackshear PJ. Approaches to the study of protein kinase Cinvolvement in signal transduction. Am J Med Sci 1988;296:231-40.33. Zatz M, Reisine TD. Lithium induces corticotropin secretionand desensitization in cultured anterior pituitary cells. Proc NatlAced Sci USA 1985;82:1286-90.34. Reisine T, Zatz M. Interactions among lithium, calcium, din-cyiglycerides and phorbol esters in the regulation of adrenocorti-cotropin hormone release from AtT-20 cells. J Neurochem 1987;49:884-9.35. Evans MS, Zorumski CF, Clifford DB. Lithium enhancesneuronal muscarinic excitation by presynaptic facffitation. Neuro-science 1990;38:457-68.36. Bitran JA, Potter WZ, Manji HK, Gusovsky F. Chronic Liattenuates agonist- and phorbol ester-mediated Na/H anti-porter activity in HL-60 cells. EurJ Pharmacol 1990;188:193-202.37. Lenox RH, Watson DG, Patel J, Ellis J. Chronic lithiumadministration alters a prominent PKC substrate in rat hippocaxn-pus. Brain Res 1992;570:333-40.38. Manji H, Lenox R. Long term action of lithium: a role fortranscriptional and posttranscriptional factors regulated by pro-tein kinase C. Synapse (NY) (in press).39. Nishiz#{252}kaY.Intracellular signalingby hydrolysis of phospho-lipids and activation of protein kinase C [Review]. Science 1992;258:607-14.40. Wang JKT, Wa]aas SI, Greengard P. Protein phosphorylationin nerve terminals: comparison of calciumlcslmodulin-dependentand calcium/diacylglycerol-dependent systems. J Neurosci 1988;8:281-8.41. Wang JKT, Walaas SI, Sthra TS, Aderem A, Greengard P.Phosphorylation and associated translocation of the 87-kDa pro-tein, a major protein kinase C substrate, in isolated nerve termi-nals. Proc Natl Aced Sci USA 1989;86:2253-6.42. Stumpo DJ, Graff JM, Albert KA, Greengard P, BlackshearPJ. Molecular cloning, characterization, and expression of a cDNAencoding the “80- to 87-kDa” myristoylated alanine-rich C kinasesubstrate: a major cellular substrate for protein kinase C. ProcNatl Aced Sci USA 1989;86:4012-6.43. Erusalimsky JD, Brooks SF, Herget T, Morris C, Rozengurt B.Molecular cloning and characterization of the acidic 80-kDa pro-tein kinase C substrate from rat brain. J Biol Chem 1991;266:7073-80.44. Harlan DM, GraffJM, Stumpo DJ, Eddy RL, Shows TB, BoyleJM, Blackshear PJ. The human myristoylated a]anine-rich Ckinase substrate (MARCKS) gene (MACS).J Biol Chem 1991;266:14399-405.45. Isacke CM, Meisenhelder J, Brown lCD, Gould KL, Gould SJ,Hunter T. Early phoaphorylation events following the treatment ofSwiss 3T3 cells with bombesin and the mammalian bombesin-related peptide, gastrin-releasing peptide. EMBO J 1986;5:2889-98.46. Rodnight R, Perrett C. Protein phosphorylation and synaptictransmission: receptor mediated modulation of protein kinase C ina rat brain fraction enriched in synaptosomes. J Physiol (Paris)1986;81:340-8.47. Rodriguez-Pens A, Zachary I, Rozengurt E. Rapid dephosphor-ylation of a Mr 80000 protein, a specific substrate of protein kinaseC, upon removal of phorbol esters, bombesin and vasopressin.Biochem Biophys Res Commun 1986;140:379-85.48. Zachary I, Sinnett-Smith JW, Rozengurt E. Early eventselicited by bombesin and structurally related peptides in quiescentSwiss 3T3 cells. I. Activation of protein kinase C and inhibition ofepidermal growth factor binding. J Cell Biol 1986;102:2211-22.49. Blackshear PJ, Stumps DJ, Huang J-K, Nemenoff RA, SpachDH. Protein kinase C-dependent and -independent pathways of

314 CUNICAL CHEMISTRY, Vol. 40, No. 2, 1994

proto-oncogene induction in human astrocytoma cells. J Biol Chem1987;262:7774-81.50. Erusalimsky JD, Friedberg I, Rozengurt E. Bombesin, diacyl-glycerols, and phorbol esters rapidly stimulate the phosphoryla-lion of an Mr= 80,000 protein kinase C substrate in permeabilized3T3 cells. Effect of guanine nucleotides. J Biol Chem 1988;263:19188-94.51. ErusalimskyJD, Rozengurt E. Vasopressin rapidly stimulatesprotein kinase C in digitonin-permeabuized Swiss 3T3 cells: in-volvement of a pertussis toxin-insensitive guanine nucleotidebinding protein. J Cell Physiol 1989;141:253-61.52. Isaandou M, Rozengurt B. Bradykinin transiently activatesprotein kinase C in Swiss 3T3 cells. Distinction from activation bybombesin and vasopreesin. J Biol Chem 1990;265:11890-8.53. GraffJM, Young TN, Johnson JD, Blackshear PJ. Phosphory-lation-regulated ca]modu]in binding to a prominent cellular sub-strate for protein kinase C. J Biol Chem 1989;264:21818-23.54. Mangels LA, Gnegy ME. Muscarinic receptor-mediated trans-location of calinodulin in SK-N-SH human neuroblastoma cells.Mol Pharmacol 1990;37:820-6.55. Mdllroy BK, Walters JD, Blackahear PJ, Johnson JD. Phos-phorylation-dependent binding of a synthetic MARCKS peptide tocalinodulin. J Biol Chem 1991;266:4959-64.

56. Patel J, Kligman D. Purification and characterization of an Mr87,000 protein kinase C substrate from rat brain. J Biol Chem1987;262:16686-91.57. Thelen M, Rosen A, Nairn AC, Aderem A. Tumor necrosisfactor a modifies agonist-dependent responses in human neutro-phi]s by inducing the synthesis and myristoylation of a specificprotein kinase C substrate. Proc Natl Aced Sci USA 1990;87:5603-7.58. Ouiinet CC, Wang JKT, Walaas SI, Albert KA, Greengard P.Localization of the MARCKS (87 kDa) protein, a major specificsubstrate for protein kinase C, in rat brain. J Neurosci 1990;10:1683-98.59. Rosen A, Keenan KF, Thelen M, Nairn AC, Merem A.Activation of protein kinase C results in the displacement of itsmyristoylated, alanine-rich substrate from punctate structures inmacrophage filopodia. J Exp Med 1990;172:1211-5.60. Hartwig JH, Thelen M, Rosen A, Janmey PA, Nairn AC,Aderem A. MARCKS is an actin filament crosslinking proteinregulated by protein kinase C and calcium-calmodulin. Nature1992;356:618-22.61. Merem A. The MARCKS brothers: a family of protein kinaseC substrates. Cell 1992;71:713-6.