Decreased Catalytic Activity and Expression of Protein Kinase C Isozymesin Teenage Suicide Victims: Title and subTitle BreakA Postmortem Brain Study

Suicide is a major public health concern,¹ andapproximately 30 000 people die of suicide each year in the United Statesalone.² Teenage suicide is a special concernbecause it is the second leading cause of death among teenagers and becausethe suicide rate in the past 10 to 20 years has increased 4-fold in the whiteteenage male population.³ Whereas there issome understanding of the psychosocial and psychological factors associatedwith teenage suicide and of the neurobiologic abnormalities of adult suicide,little is known about the neurobiologic characteristics of teenage suicide.Studies^{4 - 9} ofthe neurobiologic factors in adult suicide have indicated abnormalities inneurotransmitter receptors, such as 5-hydroxytryptamine 2A (5-HT_2A)and α- or β-adrenergic receptors, and in the receptor-linked signalingsystems, namely, adenylate cyclase and the phosphoinositide signaling system,to which these receptors are linked.

The characteristics of teenage suicide may be similar to those of adultsuicide in some respects but may differ in others. Besides psychiatric disorders,other risk factors for suicide include psychosocial stressors, impulsivity,and aggression.¹⁰ In teenagers, the major drivingfactor leading to suicide is aggressive/impulsive behavior,^{10 - 11} whichhas been implicated in abnormalities in serotonin function.¹² Ina recent study, our group⁵ found that the numberof 5-HT_2A receptors was increased in the prefrontal cortex (PFC)and hippocampus of suicide victims, which was associated with an increasein protein and messenger RNA (mRNA) expression. Although these studies indicatedabnormalities in 5-HT_2A receptors, the biochemical consequencesand the physiologic significance of this observation remain unclear.

Protein kinase C (PKC) is an important component of the phosphoinositidesignaling system^{13 - 14} to which5-HT_2A and several other receptors, such as 5-HT_2C, α₁-adrenergic, and muscarinic M₁ receptors, are linked andmediate their functional response. In the phosphoinositide signaling system,stimulation of these receptors activates the effector phospholipase C, whichcauses hydrolysis of the substrate inositol-4,5-biphosphate and results inthe formation of 2 second messengers, diacylglycerol and inositol triphosphate.¹⁵ Diacylglycerol activates the phospholipid- and calcium-dependentPKC and increases the affinity of the enzyme for calcium.¹⁶ Inositoltriphosphate, on the other hand, mobilizes calcium from intracellular stores.¹⁴ Activation of PKC is associated with translocationof the enzyme from cytoplasm to membrane.¹⁷ OncePKC has been activated, it is involved in the phosphorylation of several membranal,cytosolic, and nuclear proteins.

Protein kinase C is a key regulatory enzyme present in various tissues,including the brain, and it is localized presynaptically and postsynaptically.^{13 ,18} The expression of PKC in the brainregion is isoform specific.¹⁹ Heterogeneityexists in PKC, and 12 different isozymes have now been described.¹⁶ Each individual isozyme is involved in specific cellularresponses, such as migration, proliferation, atrophy, apoptosis, and secretion,which suggests that these isozymes are important in clinical disorders.^{14 ,19 - 21} ThePKC family has been subgrouped into 3 classes: conventional, which includes α, βI, βII,and γ; novel, which includes δ, ϵ, and φ; and atypical.^{14 ,16} Each isozyme is encoded by a uniquegene, except for βI and βII isozymes, which are the splicing productsof the same transcript.²² The biochemical propertiesof each isozyme have been identified with respect to activation or to phosphorylation,proteolytic activation, degradation, and substrate specificity.²²

There is direct and indirect evidence^{20 ,23 - 24} suggestingthat PKC may play a crucial role in mental disorders. An earlier study²⁵ examined the role of PKC in suicide by determining[³H]phorbol 12,13-dibutyrate (PDBu) binding in the PFC and hippocampusof teenage suicide victims. [³H]phorbol 12,13-dibutyrate is a highlyspecific ligand that measures regulatory subunits of PKC.^{26 - 27} Theauthors²⁵ found that [³H]PDBu bindingwas significantly decreased in the PFC and hippocampus of teenage suicidevictims compared with nonpsychiatric control subjects. This study providedpreliminary evidence of abnormalities of PKC in teenage suicide; however,[³H]PDBu, as stated previously, measures only regulatory subunitsof PKC, and it was not clear from this study whether the decrease in PKC bindingsites was associated with either changes in catalytic activity or changesin the level of any specific isozyme. Because each isozyme is related to specificfunctions and is region specific, we determined the catalytic activity ofPKC and the protein levels of various isozymes in membrane and cytosol fractionsof the PFC and hippocampus obtained from postmortem brain samples of teenagesuicide victims and nonpsychiatric control subjects. To further examine whetherany changes in the isozymes are related to altered transcription, we determinedthe mRNA levels of these isozymes in total RNA.

The right hemispheres of the PFC (Brodmann area 9) and hippocampus from17 teenage suicide victims and 17 nonpsychiatric teenage control subjectsobtained from the Brain Collection Program of the Maryland Psychiatric ResearchCenter, in collaboration with the Medical Examiner's Office of the State ofMaryland, were used. Brain samples were free of neuropathologic abnormalitiesand human immunodeficiency virus antibodies. Toxicologic data were obtainedby analysis of urine and blood samples from the participants.

All patients in this study were diagnosed based on the Diagnostic Evaluation After Death²⁸ andthe Structured Clinical Interviews for the DSM-III-R asdescribed earlier.²⁹ Two senior psychiatristsprovided independent lifetime DSM-III-R diagnoses.Controls were verified as being free of mental illness and substance abuse.The protocols for tissue sampling and retrospective assessments were approvedby the institutional review board of the University of Maryland. This studywas also approved by the institutional review board of the University of Illinoisat Chicago. The demographic, clinical, and toxicologic characteristics ofthe participants are provided in Table 1.

Protein kinase C activity in membrane and cytosol fractions of bothbrain areas was measured by using the procedure described in a previous study.³⁰ Assay tubes contained 25 µL of a componentmixture (3mM Ca/[C₃H₃O₂]₂, L-α-phosphatidyl-L-serine [75 µg/mL], phorbol 12-myristate13-acetate [6 µg/mL], 225µM substrate peptide, and 7.5mM dithiothreitolin 50mM Tris hydrochloride containing 0.05% sodium azide, pH 75) and 25 µLof the membrane or cytosol fraction. The reaction was initiated by the additionof 25 µL of magnesium–adenosine triphosphate buffer ([γ-³²P]adenosine triphosphate [10 µCi/mL {370 000 Bq/mL}], 1.2mMadenosine triphosphate, 72mM magnesium chloride, and 30mM HEPES [4-(2-hydroxyethyl)poperazine-1-ethanesulfonicacid], pH 7.4) and incubated for 15 minutes at 37°C. The reaction wasterminated by the addition of 100 µL of 300mM orthophosphoric acid.An aliquot of the solution from each tube (35 µL) was blotted onto individualpeptide-binding papers, and the retained radioactivity was counted.

Immunolabeling of PKC α, βI, βII, and γ was determinedas described in a previous study.³⁰ Equal volumesof tissue samples (30 µg of protein in 20 µL) were loaded onto7.5% (weight per volume) polyacrylamide gel and electrophoresed. The blotswere initially developed using polyclonal anti–PKC α, βI, βII,or γ antibody (1:3000-1:5000 dilution) and subsequently using β-actinmonoclonal antibody (1:5000 dilution). The levels of PKC isozymes were calculatedas a ratio of the optical density of the PKC antibody of interest to the opticaldensity of β-actin antibody.

Messenger RNA levels of PKC isozymes and neuron-specific enolase (NSE)were determined in total RNA as described in detail in a previous study.³¹ Internal standards were used to determine the quantitationof PKC isozyme and NSE mRNAs.³¹ Primer pairswere designed to allow amplification of 700 to 1003 base pairs (bp) for PKC α(GenBank accession number X52479), 914 to 1399 bp for PKC β (GenBankaccession number X06318), 1506 to 1823 bp for PKC γ (GenBank accessionnumber AF345987), and 295 to 675 bp for NSE (GenBank accession number X14327).The internal primers for PKC isozymes and NSE were as follows: PKC α,843 to 866 bp; PKC β, 1131 to 1155 bp; PKC γ, 1645 to 1668 bp;and NSE, 403 to 423 bp. Internal standards contained Bg1II (PKC isozymes) or XhoI (NSE) restrictionendonuclease sites. Decreasing concentrations of PKC isozyme or NSE internalstandard complementary RNAs were added to 1 µg of total RNA. The polymerasechain reaction mixture was amplified for 30 cycles, digested with Bg1II or with XhoI in triplicate, and runon 1.5% agarose gel. The results were calculated as the counts incorporatedinto the amplified complementary RNA standard divided by the counts incorporatedinto the corresponding mRNA amplification product vs the known amount of PKCisozyme or NSE internal standard added to the test sample. The results areexpressed as attomoles per microgram of total RNA.

Data analyses were performed using a statistical software package (SPSS8.0; SPSS Inc, Chicago, Ill). All values are reported as mean ± SD.Differences in PKC catalytic activity, mRNA and protein levels of PKC isozymes,age, sex, postmortem interval (PMI), and pH of the brain between suicide victimsand control subjects were analyzed using the independent-sample t test. The relationships between PKC catalytic activity, mRNA andprotein levels of PKC isozymes, PMI, age, sex, and pH of the brain were determinedby using Pearson product moment correlation analysis. P values were 2-tailed. Statistical differences between subgroups ofsuicide victims (with and without mental disorders) and controls were evaluatedby using 1-way analysis of variance. During analysis of the data, we includedrace as a potential confounding variable.

The detailed demographic characteristics of the teenage suicide victims(n = 17) and the control subjects (n = 17) are given in Table 1. There were no significant differences in age (t₃₂ = 1.5; P = .12), PMI (t₃₁ = 0.84; P = .40),or pH of the brain (t₃₂ = 0.51; P = .61) between controls and suicide victims.

Protein kinase C activity was determined in membrane and cytosol fractionsof the PFC and hippocampus (Figure 1).There was a significant decrease in PKC activity in membrane and cytosol fractionsof the PFC and hippocampus of teenage suicide victims compared with controlsubjects (PFC: membrane, t₃₂ = 4.65; P<.001; cytosol, t₃₂ =3.59; P = .001; and hippocampus: membrane, t₃₂ = 3.2; P = .003;cytosol, t₃₂ = 2.77; P = .009).

Figure 2 shows representativeimmunoblots of the various PKC isozymes in membrane and cytosol fractionsobtained from the PFC of 2 teenage suicide victims and 2 control subjects.Protein kinase C α, βI, βII, and γ isozymes migratedto 80 kDa, whereas β-actin migrated to 46 kDa. Levels of PKC α, βI, βII,and γ decreased in membrane and cytosol fractions obtained from thePFC of teenage suicide victims. Mean levels of PKC α, βI, βII,and γ were significantly decreased in membrane and cytosol fractionsof the PFC (Figure 3) and hippocampus(Figure 4) of teenage suicide victimscompared with controls. The significance levels of these decreases in thePFC were as follows: membrane—PKC α, t₃₂ = 4.6; P<.001; PKC βI, t₃₂ = 4.1; P<.001; PKC βII, t₃₂ = 4.5; P<.001;and PKC γ, t₃₂ = 3.3; P = .002; cytosol—PKC α, t₃₂ = 3.5; P = .001; PKC βI, t₃₂ = 4.1; P<.001; PKC βII, t₃₂ = 3.8; P = .001;and PKC γ, t₃₂ = 6.3; P<.001. The significance levels of these decreases in the hippocampuswere as follows: membrane—PKC α, t₃₂ = 3.2; P = .003; PKC βI, t₃₂ = 3.8; P<.001; PKC βII, t₃₂ = 2.6; P = .02;and PKC γ, t₃₂ = 3.1; P = .004; cytosol—PKC α, t_2.032 = 2.8; P = .009; PKC βI, t₃₂ = 3.9; P<.001; PKC βII, t₃₂ = 2.0; P = .04;and PKC γ, t₃₂ = 2.4; P = .02.

To examine whether decreases in PKC isozyme levels were related to alteredtranscription of their respective mRNAs, we determined the mRNA levels ofthe various PKC isozymes by using quantitative reverse transcriptase–polymerasechain reaction (RT-PCR). Representative gel electrophoreses of the competitiveRT-PCR of PKC isozymes are shown in Figure5A-C for 1 control subject. The results of a competitive RT-PCRanalysis for PKC α, β, and γ are shown in Figure 5D-F, where the point of equivalence represents the amountof PKC isozyme mRNAs present. The mRNA levels of PKC α, β, and γwere significantly decreased in the PFC (PKC α, t₃₀ = 5.5; P<.001; PKC β, t₃₀ = 3.3; P = .002;and PKC γ, t₃₀ = 3.7; P = .001) and in the hippocampus (PKC α, t₃₀ = 5.5; P<.001; PKC β, t₃₀ = 3.3; P = .002;and PKC γ, t₃₀ = 3.7; P = .001) of teenage suicide victims (Figure 6). To establish whether neuronal RNA contributes equallyto the total RNA pool, we determined the mRNA level of NSE in the PFC of suicidevictims and control subjects. Levels of NSE mRNA in PFC in control subjectsvs suicide victims showed no significant differences (360 ± 89 vs 408± 98 attomoles/µg of total RNA; t₃₀ = 1.46; P = .15). However, the ratios ofPKC isozyme to NSE mRNA showed that PKC mRNA levels were significantly decreasedin the PFC of suicide victims when expressed as a function of NSE mRNA (PKC α:control, 0.26 ± 0.11; suicide, 0.11 ± 0.04; PKC β: control,0.46 ± 0.32; suicide, 0.15 ± 0.08; and PKC γ: control,1.2 ± 0.22; suicide, 0.69 ± 0.20).

We did not find any effect of age, PMI, and pH of the brain on any ofthe PKC measures in either the PFC or the hippocampus (data not shown). Inaddition, we examined whether antidepressant drug treatment affects any ofthe measures. Of 17 suicide victims, 3 had toxicologic evidence of antidepressantdrugs. Comparison of suicide victims with toxicologic findings positive forantidepressant drugs with those without such findings revealed no significantdifferences in PKC activity or in protein and mRNA levels of PKC isozymesin the PFC and hippocampus between these 2 groups (data not shown).

Some previous studies²⁴ have shown anassociation between abnormalities in PKC activity and PKC isozymes in theplatelets of patients with mood disorders. We, therefore, examined whetherthe decrease in PKC isozyme levels is associated with mental disorders orwhether it is associated with suicide independent of diagnosis. Of the 17suicide victims, 9 had a history of some kind of mental disorder, primarilymood disorders and adjustment disorders, and 8 had no history of mental disorders,although 2 had a history of alcohol or other drug abuse. When we comparedthe various PKC isozymes between patients who had a lifetime history of mentaldisorder and those who did not have such a history, there were no statisticallysignificant differences between the 2 groups in catalytic activity of PKCor in protein expression of any of the PKC isozymes in membrane and cytosolfractions. Similarly, there were no significant differences in mRNA expressionlevels in patients who had a lifetime history of mental disorders and thosewho did not have such a history; however, both subgroups were significantlylower than control subjects in terms of protein and mRNA expression of thevarious PKC isozymes (data not shown). These results thus suggest that thedecreases in protein and mRNA expression of PKC isozymes found in teenagesuicide victims were independent of diagnosis.

There were 16 males and 1 female in the control group and 10 males and7 females in the suicide group. To examine whether the uneven sex distributionhad any effects on PKC, we analyzed the data in males in the control and suicidegroups. We found significant differences in PKC activity (PFC: membrane, t₂₄ = 3.08; P = .01;cytosol, t₂₄ = 4.2 P<.001; and hippocampus: membrane, t₂₄ = 2.4; P = .02; cytosol, t₂₄ = 4.04; P = .001), proteinlevels (PFC—membrane: PKC α, t₂₄ = 3.9; P = .001; PKC β, t₂₄ = 4.2; P<.001; and PKC γ, t₂₄ = 2.7; P = .01;and cytosol: PKC α, t₂₄ = 2.6; P = .01; PKC β, t₂₄ = 4.7; P = .001; and PKC γ, t₂₄ = 5.4; P<.001; and hippocampus—membrane:PKC α, t₂₄ = 2.6; P = .01; PKC β, t₂₄ = 3.4; P = .005; and PKC γ, t₂₄ = 2.4; P = .02; and cytosol: PKC α, t₂₄ = 2.0; P = .05;PKC β, t₂₄ = 3.5; P = .002; and PKC γ, t₂₄ =2.2; P = .03), and mRNA levels (PFC: PKC α, t₂₂ = 3.5; P = .002;PKC β, t₂₂ = 2.4; P = .02; and PKC γ, t₂₂ =2.2; P = .03; and hippocampus: PKC α, t₂₂ = 3.5; P = .002;PKC β, t₂₂ = 2.4; P = .02; and PKC γ, t₂₂ =2.6; P = .01) between controls and suicide victimswith a similar magnitude as observed after inclusion of females. In the suicidegroup, we also analyzed the data comparing male and female suicide victims.We did not find any significant effects of sex on PKC activity or mRNA andprotein levels of PKC isozymes (data not shown). These data suggest that theeffects seen on PKC are not related to sex.

Because there was an uneven distribution of race between the controland suicide groups, we analyzed the effect of race by comparing PKC measuresbetween blacks and whites in both groups and found no significant differencesbetween them (data not shown). We also included race as a covariate in theanalysis.

The results of the present study reveal a statistically significantreduction in the catalytic activity of PKC in the PFC and hippocampus of teenagesuicide victims compared with control subjects. This reduction in PKC activitywas associated with decreases in protein and mRNA expression of PKC α, β,and γ isozymes in the PFC and hippocampus of teenage suicide victims.These changes were not related to age, PMI, or pH of the brain. Furthermore,these changes were not associated with mental disorders, as no statisticallysignificant differences were observed either in PKC activity or in proteinor mRNA expression of PKC α, β, and γ isozymes between suicidevictims with no history of mental disorders and those with a history of mentaldisorders in either the PFC or the hippocampus. However, PKC activity andprotein and mRNA expression of the PKC isozymes in suicide victims with orwithout mental disorders still were statistically significantly lower thanin control subjects. This observation has provided preliminary evidence suggestingthat alterations in PKC activity and isozymes in teenage suicide victims areindependent of diagnosis. However, these results need to be replicated ina larger number of suicide victims with or without mental disorders beforearriving at a definite conclusion on the effect of mental disorders on PKCisozymes.

There is direct and indirect evidence suggesting that PKC may play animportant role in the pathogenesis of mood disorders. Indirect evidence isderived from the observations that treatment with lithium and other mood-stabilizingdrugs causes a decrease in PKC activity and expression of certain PKC isozymesin rat brain.³² Because this effect of lithiumis also shared by other mood-stabilizing drugs, such as valproate,³² this implies that a PKC abnormality may be associatedwith bipolar illness and that treatment with lithium or other mood-stabilizingdrugs may normalize this abnormality. Direct evidence in support of the involvementof PKC in mood disorders is limited and is derived mainly from studies ofPKC in platelets obtained from patients with mood disorders. For example,a previous study²³ reported that PKC activityis decreased in platelets of bipolar patients compared with control subjects.This decrease in PKC activity was mainly due to a selective decrease in theprotein expression of PKC α, βI, βII, and δ in membraneand cytosol fractions of platelets from bipolar patients. Friedman et al²⁴ and Young et al³³ reportedthat PKC activity is increased in platelets of bipolar patients during themanic phase. Young et al³³ studied the expressionof PKC α in these patients and did not find any difference in cytosoland membrane fractions of platelets obtained from bipolar patients.

Protein kinase C has not been studied extensively in the postmortembrain. In a recent study, Pandey et al²⁵ determinedPKC binding sites using [³H]PDBu, which is a measure of the regulatorybinding site of PKC, in the PFC of teenage suicide victims and nonpsychiatriccontrol subjects and found that [³H]PDBu binding was significantlydecreased in teenage suicide victims compared with control subjects. Coullet al³⁴ also studied [³H]PBDu bindingin the PFC and hippocampus of antidepressant-treated and antidepressant-freeadult depressed suicide victims. They did not find any significant differencesin [³H]PBDu binding between antidepressant-treated suicide victimsand control subjects. On the other hand, they found a significant increasein the B_max of [³H]PDBu binding in the soluble fractionsof antidepressant-free suicide victims compared with control subjects. Theapparent differences between our study and that of the depressed suicide victimsreported by Coull et al³⁴ could be due to thedifference in age of the populations studied. It is possible that the changesin PKC may be different in teenage vs adult suicide victims because of developmentalfactors and factors associated with chronicity of illness or previous long-termtreatment with psychoactive drugs. Dean et al³⁵ reporteda significant decrease in the density of PKC in the parahippocampal gyrusof patients with schizophrenia compared with control subjects. The resultsof studies of PKC in postmortem brain thus seem to be mixed; however, almostall studies in the postmortem brain of suicide victims, patients with schizophrenia,or depressed suicide victims were carried out using [³H]PDBu forlabeling PKC binding sites, but none of these studies examined the proteinor mRNA expression of individual PKC isozymes. Although [³H]PDBuhas been used for studying PKC binding sites, it is believed that [³H]PDBu also labels other proteins, such as α-chimaerins and RAS-GRP,and it is possible that the failure to find any changes in PKC binding sitesmay not reflect abnormal expression of PKC protein. Studies of PKC activityand of expression of the various PKC isozymes may, therefore, be importantin examining the role of PKC in these disorders. To our knowledge, this isthe first study to examine protein and mRNA expression and PKC activity inpostmortem brains obtained from psychiatric patients and suicide victims.As mentioned earlier herein, we found that the expression of 3 different PKCisozymes—PKC α, β, and γ—was significantly decreasedin the PFC and hippocampus of teenage suicide victims.

The mechanism by which PKC is decreased in the postmortem brain of suicidevictims is not clear. The activation of PKC by an agonist such as diacylglycerolor phorbol myristate acetate causes the translocation of PKC from the cytosolto the membrane. This translocation also facilitates the proteolytic degradationof the enzyme. Another possibility could be related to a compensatory mechanismin response to the sustained activation of the receptors to which this signalingcascade is linked. For example, our group⁵ recentlyreported that 5-HT_2A receptors are increased in the postmortembrain of teenage suicide victims. It is possible that activation of theseincreased receptors may cause continued activation of the enzyme and the down-regulationof PKC and some of its specific isozymes. That this mechanism may be a reasonfor the decrease in PKC in the postmortem brain of suicide victims is supportedby the observation of some investigators that continued activation of theenzyme PKC by phorbol esters causes its translocation and subsequent degradationand decrease.³⁶ The other reason for a possibledecrease in PKC may be related to an overactive hypothalamic-pituitary-adrenal(HPA) axis in suicide. It has been reported that the HPA axis may be abnormalin suicide victims.³⁷ Earlier, our group³⁸ showed that activation of the HPA axis causes down-regulationof certain PKC isozymes in rat brain. Therefore, it is possible that the changesin PKC may be related to altered HPA function in suicide victims.

The pathophysiologic significance of the decrease in PKC in suicidevictims remains to be elucidated; however, the phosphorylation of proteins,mediated by PKC, is a key to many physiologic functions in the brain, includinggene transcription. Some examples of proteins phosphorylated by PKC are MARCKSand GPA-43. Both these proteins have been implicated in mood disorders.^{23 ,32} Furthermore, certain evidence indicatesthat besides phosphorylating proteins in cytosol, PKC α, β, and γtranslocate into the nucleus and phosphorylate many nuclear proteins,^{39 - 40} cell cycle–related proteins,⁴¹ and transcription factors, such as CREB and NFκB.^{42 - 43} Although specific functions of eachPKC isozyme are not clearly known, recent studies demonstrate that PKC αis closely associated with proliferation, cell differentiation, and metabolismand with some cell type–specific functions, such as β-cell developmentand activation.^{44 - 45} As far asPKC γ is concerned, it is the only PKC isozyme that is present solelyin the brain and spinal cord.^{16 ,46} Manyneuronal functions have been attributed to PKC γ, among them synapticformation, long-term potentiation, long-term depression, and modulation ofreceptor functions, particularly γ-aminobutyric acid A.⁴⁷ Proteinkinase C γ–deficient mice exhibit impaired motor coordinationand deficits in spatial and contextual learning.⁴⁸ Mutationin PKC γ also shows brain pathology resembling the parkinsonian syndrome.^{49 - 50} Whether a deficiency in PKC is associatedwith suicidal behavior is not clearly known; however, given the many substratesthat PKC phosphorylates, a deficiency in PKC isozymes may disrupt normal brainfunction.

Regardless of the mechanism by which PKC is down-regulated in the postmortembrain of suicide victims, our observation that PKC activity and some of itsisozymes are decreased in the postmortem brain of teenage suicide victimsmay be vitally important in understanding the neurobiologic abnormalitiesof suicide.

Correspondence: Ghanshyam N. Pandey, PhD, Psychiatric Institute,Department of Psychiatry, University of Illinois at Chicago, 1601 W TaylorSt, Chicago, IL 60612 (Gnpandey@psych.uic.edu)

Submitted for publication July 11, 2003; final revision received January27, 2004; accepted February 3, 2004.

This work was supported by grants RO1 MH 48153 (Dr Pandey) and KO1 MH01836 (Dr Dwivedi) from the National Institute of Mental Health, Rockville,Md, and by the American Foundation for Suicide Prevention, New York, NY (DrDwivedi).

We thank Dr Smialek, MD, chief medical examiner, and Dennis Chute, MD,assistant medical examiner, for their cooperation in the collection of brainsamples; Terri U'Prichard, MA, for performing the psychological autopsies;and Barbara Brown, BS, and Miljana Petkovic, BS, for their help in organizingthe brain tissue.