REVIEW

C foSJ. avidPa , FaJo e KGa y, J

Tw f suirel rs inph relatsui pproint rkerpsy n FoCo ht tosui ggresfun enottra disor

Ke

SarestuheRethereccidcognepostagenshoAmment of Psychiatry at Columbia University that addressed thebioofme

Su

theouletatt11of3.3

unCobesuiachPo

Su

Fro

Ad

Rec

000dological and behavioral diathesis for suicidality for the purposeidentifying promising clinical, cognitive, and biological inter-diate phenotypes for finding candidate gene studies.

HeritabilityThe contribution of additive genetic factors is estimated to be

30% to 50% for a broad phenotype of suicidality that includesideation, plans, and attempts and is largely independent of theinheritance of psychiatric disorder (4). There is a higher concor-dance rate for suicide in monozygotic than dizygotic twins(24.1% vs. 2.8%) (4) and a 2.0 to 4.8 times greater prevalence ofsuicide in the relatives of individuals who die by suicide, evenafter adjusting for psychiatric disorder. Adoption studies docu-ment fourfold to sixfold higher rates of suicide in the biologicalrelatives of adoptees who die by suicide compared with adoptiverelatives (see Brent and Melhem [5] for a review).

Nonfatal suicide attempts have heritability estimates of 17% to45%, even after controlling for psychiatric disorder, and familystudies consistently report higher rates of suicide attempt inrelatives of suicide attempters compared with relatives of nonat-tempters. However, an adoption study did not find higher ratesof suicide attempt in the biological parents than in the adoptiveparents of suicide attempters (5). Data on suicidal ideation aresparser. Twin studies report 36% to 43% heritability, and family

m Department of Psychiatry (JJM, VAA, FAC, DC, FH, SEH, MAO, BS, JT),Columbia University, New York, New York; National Institutes of Health/National Institute of Mental Health (SA, JK, TL, FM, EKM, JP), Bethesda,Maryland; Western Psychiatric Institute and Clinic (DAB); AmericanFoundation for Suicide Prevention (PC), New York, New York; Universityof Texas Health Science Center at San Antonio (DMD), San Antonio,Texas; Department of Psychiatry (GNP), University of Illinois at Chicago,Chicago, Illinois; andDepartment of Psychiatry (AW), University of Penn-sylvania, Philadelphia, Pennsylvaniadress reprint requests to J. John Mann, M.D., Department of MolecularImaging&Neuropathology, NewYork State Psychiatric Institute/Colum-bia University, 1051 Riverside Drive, Box 42, New York, NY 10032; E-mail:jjm@columbia.edu.eived July 11, 2008; revised October 21, 2008; accepted November 12,2008.

BIOL PSYCHIATRY 2009;65:5565636-3223/09/$36.00i:10.1016/j.biopsych.2008.11.021 2009 Society of Biological Psychiatryandidate Endophenotypesuicidal BehaviorJohn Mann, Victoria A. Arango, Shelli Avenevoli, Dula Clayton, Dianne Currier, Donald M. Doughertyel Kleinman, Thomas Lehner, Francis McMahon, Evnshayam N. Pandey, Jane Pearson, Barbara Stanle

in, adoption, and family studies have established the heritability oated genes has proven more difficult. As with psychiatric disordeenotype for suicidal behavior and distinguishing suicide diathesis-cide-associated psychiatric illness. Adopting an endophenotype aermediate phenotypes, including biological and/or behavioral machiatric disorders. Therefore, aworkshop convenedby theAmericalumbia University, and the National Institute of Mental Health sougcidal behavior. The most promising endophenotypes were trait action, and cortisol social stress response. Other candidate endophnsmission, second messenger systems, and borderline personality

yWords: Endophenotype, genetics, suicide

uicidal behavior is an important preventable cause ofinjury, disability, and death worldwide. In most countries,the vast majority of suicides and nonfatal suicide attempts

associated with psychiatric disorders. Twin and adoptiondies find that the predisposition to suicidal behavior is partlyritable; however, the genes responsible are mostly unknown.cently, there have been substantial advances in knowledge ofpathophysiology of suicide and nonfatal suicide attempts;ognition of behavioral components of the diathesis for sui-al behavior including aggressive/impulsive traits and neuro-nitive deficits; development of methods for imaging relevantural circuitry of the brain; finer grained single nucleotidelymorphism (SNP) maps of the human genome; improvedtistical approaches; and greater appreciation of the role ofe-environment interactions. This article is based on a work-p convened by the National Institute of Mental Health, theerican Foundation for Suicide Prevention, and the Depart-r Genetic Studies of

A. Brent, Frances A. Champagne,temah Haghighi, Susan E. Hodge,. Moscicki, Maria A. Oquendo,oseph Terwilliger, and AmyWenzel

cide attempts and suicide. Identifying specific suicide diathesis-general, methodological difficulties include complexity of theed genes from genes associated with mood disorders and otherach involving identification of genes associated with heritables more proximal to genes, is an approach being used for otherundation for Suicide Prevention, theDepartment of Psychiatry atidentify potential target endophenotypes for genetic studies ofsion/impulsivity, early-onset major depression, neurocognitiveypes requiring further investigation include serotonergic neuro-der traits.

icidal Behavior Phenotypes

There are multiple levels of severity and injury burden withinconstruct of suicidality, ranging from suicidal ideation with-

t a specific plan, suicidal ideation with a specific plan, low-hality suicide attempts, high-lethality but nonfatal suicideempts, and death by suicide. In 2005, the U.S. suicide rate wasper 100,000 persons (1). In the general population, prevalencesuicide attempt is .4 % to .6%, and suicidal ideation 2.8% to% (2).Comparison of findings between studies is limited by lack ofiform definitions of suicidal ideation and behaviors. Thelumbia Classification Algorithm of Suicidal Assessment ising widely adopted and provides operational definitions ofcide and suicide attempt that set two minimal requirements: 1)suicidal act must be a self-directed act; and 2) it must bearacterized by some intent to die (for detailed definitions, seesner, et al.) (3).

icidal Behavior and Genetics

studies have documented transmission of suicidal ideation (5).However, familial transmission of ideation, in contrast to trans-mipsy

Ge

antypchrvarde6q

sanideuncomofdespefropre

sellattheserrecsibdrodroensui[6]TPdemoMetrafoutherecmibracan(6)

varmuatiorec(en

vusuiapcidfaichivaransui

No single gene will explain complex multidetermined behav-iors such as suicide and nonfatal suicide attempt. Thus, givenlikanproph

De

typstipillnstawinopo

pucostr

Im

beulaincstudieGrattulaoftw47dephhage5-Hantraanfrewiofin

proTwassratbetradifratge(32tiomoanchonSin

J.J. Mann et al. BIOL PSYCHIATRY 2009;65:556563 557ssion of behavior, appears to be related to the transmission ofchiatric disorder (5).

nes and Suicidal BehaviorWhole-genome linkage studies seek to identify genetic locid ultimately candidate genes associated with disease pheno-es. Such studies have reported linkage of suicide attempt andomosome 2p11, 2p12, and 2p, 5q, 6q, 8p, 11q, and Xq inied psychiatric disorder pedigrees (6). A study of suicideaths in bipolar pedigrees reported linkage at chromosome25.2 (6).Microarray technology enables expression profiling of thou-ds of genes in the brains of suicides and may also aid inntification of candidate genes, as well as extending thederstanding of neurobiological pathways in suicide. Studiesparing suicides and control subjects report higher expression

serotonin 2A receptor (5-HT2A) genes in Brodmann area 11 inpressed suicides (7) and lower expression of the spermine/rminidine N1-acetyltransference gene in the dorsolateral pre-ntal cortex and motor cortex in both depressed and nonde-ssed suicides (8).To date, candidate genes for suicidal behavior have beenected largely on the basis of established biological corre-es of suicidal behavior and thus have focused primarily onserotonergic system. Genes for the serotonin transporter,otonin 1A (5-HT1A), serotonin 1B (5-HT1B), and 5-HT2Aeptors; monoamine oxidase A (MAOA) (an enzyme respon-le for the degradation of serotonin); and tryptophan hy-xylase (tryptophan hydroxylase 1 [TPH1], tryptophan hy-xylase 2 [TPH2] isoforms), the rate-limiting biosyntheticzyme for serotonin, have been investigated with respect tocidal behavior with suggestive but inconclusive results (seefor a review). Preliminary studies of the recently identifiedH2 gene link an intronic polymorphism to suicide andpression (9), although others find an association only withod disorder and not independently with suicide (10).ta-analysis of 14 5-HT2A receptor gene and 12 serotoninnsporter promoter (5-HTTLPR) gene association studiesnd an association between the low-expressing alleles of5-HTTLPR and suicide but no association for the 5-HT2A

eptor 102T/C polymorphism (11). The noradrenergic and dopa-nergic systems, hypothalamic-pituitary-adrenal (HPA) axis, andin-derived neurotrophic factor have also been examined fordidate genes, with no consistent associations identified as yet.Pharmacogenetic studies have examined the effect of geneticiants on treatment-emergent suicidal ideation, and a largelticenter treatment trial of major depression reported associ-ns with GRIK2 and GRIA3 (encode ionotropic glutamateeptors), IL28R (encodes an interleukin receptor), and PAPLNcodes a protoglycan-like sulfated glycoprotein) (12).Association studies suggest that interaction between geneticlnerability and environmental conditions increases risk forcidal behavior. The lower-expressing allele of the 5-HTTLPRpears associated with increased risk for depression and sui-ality in response to stressful life events, though others haveled to replicate those findings (see [6] for a review). Adverseldhood experiences in conjunction with a lower-expressingiant of the MAOA gene contributed to the development oftisocial behavior and greater impulsivity (risk factors forcidal behavior) in male subjects (13).ely gene-gene relationships, gene-environment interactions,d the multiple pathways resulting in suicidal acts, a moreductive approach is to identify biological and clinical endo-enotypes.

fining Endophenotypes for Suicidal Behavior

Gottesman and Gould (14) have described an endopheno-e as an internal phenotype between gene and disease. Theyulate five criteria for an endophenotype: 1) association withess in population; 2) heritable (20% or greater); 3) primarilyte-independent; 4) illness and endophenotype co-segregatethin families (linkage of trait to gene variant); and 5) found innaffected family members more frequently than in the generalpulation (14).Proposed endophenotypes for suicidal behavior include im-lsive-aggressive traits, early onset of major depression, neuro-gnitive function, and heightened cortisol response to socialess.

pulsive and Aggressive TraitsBehavioral dysregulation is thought to characterize suicidalhavior, with traits of impulsivity and aggression being partic-rly salient because impulsiveness favors acting on emotionsluding anger and suicidal ideation. Psychological autopsydies have found higher levels of aggression in individuals whoby suicide than in living psychiatric control subjects (15,16).

eater lifetime aggression is associated with nonfatal suicideempts in prospective and cross-sectional studies (17), partic-rly attempts with higher medical lethality (18). The heritabilityaggressive traits has been demonstrated, with studies of adultins, using a variety of measures, reporting heritability of 40% to% (19,20). Aggression has a trait dimension (21). Family studiesmonstrate co-segregation of the aggressive/impulsive endo-enotype and suicidal behavior in families (22,23). Associationsve been reported between aggressive/violent behaviors andnes related to the serotonergic system, including 5-HT1B andT2A receptors and MAOA (2426). Violent suicidal behavior

d/or medical seriousness of suicide attempts are correlates ofit aggression and possibly associated with 5-HTTLPR genotyped TPH1 (6). There are no direct data available regardingquency of trait aggression in nonsuicidal relatives comparedth the general population, although the finding of higher levelsaggression in first-degree relatives of suicide completers thanfirst-degree relatives of control subjects (27) is suggestive.Impulsivity has been associated with suicidal behavior inspective and retrospective studies (see [16] for a review).in and family studies support the heritability of impulsivity,essed using various personality trait scales (28,29) and labo-ory behavior measures (30), with estimates of heritabilitytween 30% and 45% (28,29). Impulsivity is understood to be ait but is a multifaceted construct and likely the outcome offerent underlying deficits. Impulsivity assessed using the Bar-t Impulsivity Scale (BIS) has been associated with the 5-HT2Ane in alcohol-dependent subjects (31) but not with 5-HTTLPR). Behavioral indices of impulsivity, such as response initia-n, response inhibition, and consequence sensitivity, may bere suitable endophenotypes, as they are more basic constructsd more reliably quantified through laboratory-based neuropsy-ological testing than self-report impulsivity scales that dependrecall and the nature of stressful events in the patients life.gle nucleotide polymorphism studies have detected associa-www.sobp.org/journal

tions with measures of neuropsychological performance but notself-report scale scores (33) or with the hyperactive/impulsiveclin(34hasevwibeassexativtiattie

forbeof5-H5-Hmoasshe(40resageStoadtivrecmetivundeobsamproOCanneessba

Ea

be(47DeonthrwiagemaestEacouporioorideenrel

Neurocognitive FunctionSuicidal behavior has been associated with altered neurocog-

nittioa rtiostimtheuawhdelettrorelmoinfingwistuhesub(CintOtCOassinganalsonsubphunThassfre

incremaltidenitsolreg

Co

wicodeobcer(CRnoanstuStrrepHPadstrmi5-H

558 BIOL PSYCHIATRY 2009;65:556563 J.J. Mann et al.

wwical subtype in attention-deficit/hyperactivity disorder (ADHD)). Response initiation can be assessed by commission errors ands been shown to be related to suicide attempt history (35) anderity of attempt (36). Response inhibition can be measuredth go-stop tests and has not yet been linked to suicidalhavior, although deficits in response inhibition appear to beociated with particular subtypes of impulsive aggression, formple, fighting in conduct disorder (37). Consequence sensi-ity can be assessed with the delayed reward test and differen-es suicide attempters from nonattempters in self-injury pa-nts (38).There is some evidence of heritability and genetic associationbehavioral measures of impulsivity. Commission errors haveen shown to be heritable in family studies (30,39). Higher ratescommission errors were associated with the T allele in theT2A T102C SNP (39) and the lower-expressing allele of theTTLPR (M.A. Dawes, et al., unpublished data, 2008). Poly-rphisms in the dopamine D4 receptor gene have beenociated with response inhibition, assessed by the stop task inalthy adults (33) and stop task and go/no-go in ADHD children). Dopamine transporter genotype has been associated withponse inhibition assessed by the Opposite World test in boysd 6 to 11 (41) and the stop-signal task in healthy adults (33).p task scores varied by TPH2 genotype in healthy college-ageults (42). Delayed reward, a measure of consequence sensi-ity, is associated with a polymorphism in the dopamineeptor D2 subtype (DRD2) gene but not with clinical scaleasures of impulsivity in healthy adults (43). Unaffected rela-es, compared with control subjects, affected siblings, andaffected siblings of ADHD patients, have response inhibitionficits in the stop-signal task (44). Unaffected relatives ofsessive-compulsive disorder (OCD) probands showed thee response inhibition deficits in the stop-signal task asbands compared with individuals with no family history ofD (45). Thus, impulsivity appears to meet all five criteria forendophenotype, assessed either using clinical instruments oruropsychological tasks; however, additional studies are nec-ary to reproduce these findings in study samples ascertainedsed on suicidal behavior.

rly-Onset Major DepressionThe association of early-onset major depression and suicidalhavior has been documented in clinical (46) and community) samples. In the Sequenced Treatment Alternatives to Relievepression (STAR*D) treatment trial cohort, individuals withset of major depressive disorder (MDD) before age 18 wereee times more likely to report a suicide attempt than thoseth an onset after age 18, adjusted for duration of illness, current, and gender (46). Family studies have shown early-onsetjor depression to be heritable (46), and a study of male twinsimated heritability of early-onset MDD (age 30) at 47% (48).rly-onset MDD is a subtype characterized by a more severerse with greater chronicity and psychiatric comorbidity andorer psychosocial function (49), thus meeting the trait crite-n. Genome-wide linkage studies in affected family pedigreesrelative pairs with early-onset recurrent major depressionntified linkage to chromosomes 15q, 17p, 8p, and 6p (50). Thedophenotype criterion of greater frequency in unaffectedatives has not been investigated.w.sobp.org/journalive function in multiple domains, including executive func-n, attention, verbal fluency, and decision making (see [50] foreview). Attentional deficits, such as impaired selective atten-n leading to difficulty in shifting attention from inappropriateuli and the inability to generate new solutions, may underlieclinical observations of cognitive rigidity in suicidal individ-

ls (51,52). Poorer performance in the Stroop Interference Test,ich assesses selective attention, is found in high-lethalitypressed suicide attempters compared with depressed low-hality attempters, depressed nonattempters, and healthy con-l subjects (51). Moreover, selective attention toward suicide-evant cues is demonstrated in suicidal individuals in adified Stroop task (53,54). Attentional fixation is a relatedormation-processing bias, characterized by agitation, narrow-of attention on suicide-specific cues, and a preoccupation

th suicide as the only solution to ones problems (55). Twindies in children and adults (56,57) indicate 39% to 50%ritability in Stroop Interference Test performance. In healthyjects, the DRD2 gene and the catechol-O-methyltransferase

OMT) valine (val)/methionine (met) polymorphism had aneraction effect on Stroop Interference Test performance (58).her gene association studies of dopamine receptor genes andMT in schizophrenia and/or ADHD subjects report geneticociations with various domains of cognitive function, includ-executive function, verbal and working memory, attention,

d performance monitoring (59,60). Bipolar disorder individu-and their unaffected relatives exhibit impaired performancestests of cognitive flexibility compared with healthy controljects (61), and attention deficits (Stroop) are seen in schizo-renia (62) and bipolar disorder (63) individuals and theiraffected relatives compared with healthy control subjects.us, attentional deficits meet the endophenotype criteria ofociation with illness, heritability, gene association, and greaterquency in unaffected relatives.There is evidence of the persistence of cognitive deficits,luding attention, verbal fluency, memory, and learning, afterission of depressive symptoms, indicative of trait status (64),

hough one study found executive dysfunction in suicideators to be state-related (65). For other domains of neurocog-ive function, including memory, verbal fluency, and problemving, association with suicidal behavior is suggested, but dataarding heritability and familial co-segregation are lacking.

rtisol Response to Psychosocial StressHypothalamic-pituitary-adrenal axis dysfunction is associatedth suicide death, with dexamethasone nonsuppression ofrtisol increasing risk for suicide more than fourfold in majorpression (66). Disturbances in HPA axis function have beenserved in suicide attempters using various indices, includingebrospinal fluid (CSF) corticotrophin-releasing hormoneH), postdexamethasone cortisol, and urinary cortisol, thought all agree (67). Twin studies of cortisol levels in blood, urine,d saliva indicate approximately 60% heritability (68). A twindy of cortisol response to psychosocial stress (Trier Socialess Test [TSST]) estimated heritability of cortisol response withetition of the stressor between 56% to 97% (69). AlteredA axis stress responsivity is a trait influenced by genes andversity in early life (70). Cortisol response to psychosocialess (TSST) appears associated with polymorphisms in theneralocorticoid and glucocorticoid receptor genes (71), theTTLPR (72), and the gamma-aminobutyric acid (GABA) A

alpha 6 receptor gene (73). Studies are necessary to establish ifthe same deficits in cortisol response are more frequent inungen

Ca

beofestgen

SeSu

beMayiedehatracindosteexpbraaltecluatioind

powidetranostuhubosarexh

gictraim5-Hscotraobcoringcorstugenneto

Bra

gluanlettra

with administration of the serotonergic agonist fenfluramine (86).Impulsivity correlates with bilateral hypometabolism in the me-diafroanredtivinlowsiocobrapalie

Se

pomaadnapaproanmobratyrbe(94theavaseg

Bo

higiorranofThphchovdycidaffcluwhfuningcridyentioassbiotiotheMDtemstuafffam

J.J. Mann et al. BIOL PSYCHIATRY 2009;65:556563 559affected relatives of suicidal individuals compared with theeral population.

ndidate Endophenotypes

Other biological and clinical factors associated with suicidalhavior are candidate endophenotypes. These meet the criteriaassociation with the disease but more work is needed toablish heritability, trait status, co-segregation in families, ande variant associations.

rotonergic System Alterations in Suicide and Nonfatalicide AttemptAlterations in various indices of serotonergic function haveen associated with suicide attempt and suicide (reviewed innn [74]). Studies of postmortem brain tissue of suicides mayld molecular endophenotypes for genetic studies of suicideath. Postmortem studies comparing suicides with nonsuicidesve found death by suicide associated with low serotoninnsporter binding in orbital prefrontal cortex and anteriorgulate, higher 5-HT1A and 5-HT2A receptor binding in thersolateral prefrontal cortex and 5-HT1A binding in the brain-m dorsal raphe nucleus, and higher transcript level, proteinression, and number of TPH2 expressing neurons in theinstem (74). Evidence of genetic association with serotonergicrations observed postmortem in suicides is sparse and incon-sive. Studies are necessary to establish if serotonergic alter-ns observed in suicides meet criteria of heritability, state-ependence, and presence in nonaffected family members.The level of CSF 5-hydroxyindoleacetic acid (5-HIAA) is atential endophenotype, as lower levels have been associatedth suicide attempt and suicide death across psychiatric disor-rs (75). In nonhuman primates, CSF 5-HIAA levels have shownit characteristics (76). Heritability estimates in humans andnhuman primates range from 25% to 50% (77,78). Animaldies show CSF 5-HIAA is under genetic regulation (77), and inmans, studies of gene association and CSF 5-HIAA have beenth positive (79) and negative (80). Further studies are neces-y in humans to establish if nonaffected family membersibit this anomaly more than the general population.Imaging studies support the association of altered serotoner-function and suicidal behavior, reporting lower serotonin

nsporter binding in the frontal and midbrain regions inpulsive violent subjects (81). An inverse correlation betweenT1A binding in the orbital frontal cortex and aggression scaleres has been reported (82). Lower C--methyl-L-tryptophanpping in the orbital and ventromedial prefrontal cortex wasserved in high-lethality suicide attempters with a negativerelation with suicide intent (83). Serotonin 2A receptor bind-also correlated negatively with levels of hopelessness, a

relate of suicide and suicide attempt (84). There have been nodies of heritability in imaging studies of serotonin (5-HT) andetic association studies involving brain imaging are few and

gative (85). The state-independent status of these indices is yetbe established, as is the frequency in nonaffected relatives.

in Structure and Function In VivoAlterations in resting condition brain blood flow and/orcose metabolism have been associated with suicidal behaviord related clinical traits. Relative hypometabolism in high-hality compared with low-lethality suicide attempters in ven-l, medial, and lateral prefrontal cortex become more markedl frontal cortex (87) and reduced perfusion in the right sidentotemporal cortex (88). Likewise, impaired frontal perfusiond metabolism are consistent with cognitive deficits, includinguced executive function, and verbal fluency correlates posi-ely with regional cerebral glucose metabolic rates (rCMRGlu)the same brain regions that differ between high-lethality and-lethality suicide attempters (86). Blunted prefrontal perfu-n is related to poorer verbal fluency in suicide attemptersmpared with control subjects (89). The heritability of alteredin structure and function has not been investigated in imagingradigms nor has state-independence, co-segregation in fami-s, or frequency in nonaffected family members.

condMessengersAlteration in markers of second messenger function is anothertential biological endophenotype for suicide death. Severalrkers have been shown to be altered in teenagers and/orults who died by suicide, including glycogen synthase ki-se-3 (GSK-3), an important component of the Wnt signalingthway (90); protein and messenger RNA (mRNA) expression oftein kinase C (PKC) isozymes, protein kinase PKC, PKC,d PKC (91); transcription factors, including cyclic adenosinenophosphate (cAMP) response element binding (CREB) andin-derived neurotrophic factor (BDNF) (92,93); and lowerosine receptor kinase B (TrkB) mRNA and protein levels haveen found in prefrontal cortex and hippocampus of suicides). It remains to be determined whether these alterations meetother criteria for endophenotype, as there are no datailable as yet regarding their heritability, trait status, co-regation in families, or frequency in nonaffected relatives.

rderline Personality DisorderBorderline personality disorder (BPD) is characterized by ah rate of suicide, suicide attempt, and other self-harm behav-. In twin studies, the heritability of BPD is reported in thege of 35% to 60% (95) and family studies also provide supportheritability of both the diagnosis and of symptom clusters (96).ree main BPD symptom clusters that may yield useful endo-enotypes are 1) disturbances in interpersonal relatedness,aracterized by interpersonal deficits and dysfunction and anerreliance on others to maintain a sense of self; 2) behavioralsregulation, characterized by impulsivity, self-injury, and sui-al behavior; and 3) affective dysregulation, characterized byect lability and intense emotional reactions. For the firstster, a potential endophenotype is interpersonal reactivity,ich can be assessed by measures of rejection sensitivity or inctional magnetic resonance imaging (fMRI) paradigms assess-trust but has not yet been investigated with respect to the five

teria for an endophenotype. The second cluster is behavioralsregulation, described earlier as the impulsive aggressiondophenotype. For the third cluster, affect dysregulation, emo-nal or pain sensitivity is a potential endophenotype that can beessed though social stress paradigms, including examininglogical indices such as HPA axis dysfunction, opioid dysfunc-n, and/or autonomic dysfunction. Higher cortisol response toTrier Social Stress Test was observed in comorbid BPD andD suicide attempters compared with MDD-only suicide at-pters (Barbara Stanley, Ph.D., unpublished data, 2008). Twindies of borderline personality traits report a 45% heritability ofective lability (96), but there are no data on co-segregation inilies or frequency in nonaffected relatives.www.sobp.org/journal

sui

Ge

Ep

dirgenearingfacinillnatioofDNtheThceloccwhincmerat

beeffexaingcanmereg

Me

dretheicacau

erybeasssafTywhvalbesizhoexfouancofor

Co

Table 1. Proposed and Candidate Endophenotypes for Suicidal Behavior

State

EndAENC

Can5C5BSB

ydroxme

560 BIOL PSYCHIATRY 2009;65:556563 J.J. Mann et al.

wwA summary of proposed and candidate endophenotypes forcidal behavior appears in Table 1.

netic StudiesOther Considerations

igeneticsGenetic impact on suicidality may not just be a matter of aect generation-to-generation transmission of vulnerabilityes. Familial transmission may also occur in the context ofly-life environment effects on epigenetic mechanisms result-in altered neurobiological function. Evidence that epigenetictors play a role in psychiatric disorders includes discordancemonozygotic (MZ) twins and discordance for psychiatricess in twins being associated with differential DNA methyl-n (97). Epigenetic events alter expression for different copiesthe same gene in a given cell nucleus. Methylation of parts ofA mostly blocks transcriptional factors from gaining access togene and thus effectively silences expression of the gene.

is is a stable epigenetic modification that is maintained afterl division and is the means by which cellular differentiationurs during development. Microarray SNP chips allow forole-genome DNA methylation profiling in human tissues,luding brain, and can provide data on total and allele-specificthylation that can then be examined for disease-related aber-ions.Animal studies can examine environmental, genetic, andhavioral circuitry, and in rodents there is clear evidence for theects of early environment on gene expression (98). Formple, differences in maternal behavior (licking and groom-) result in alterations in HPA axis response to stress, one of thedidate endophenotypes, in offspring (99) and in differentialthylation of the noncoding glucocorticoid receptor promotorion 17 in hippocampal tissues in adult offspring (100).

thodological IssuesIdentifying endophenotypes for suicidal behavior may ad-ss some difficulties related to heterogeneity and complexity ofphenotype; however, there remain substantial methodolog-l challenges in identifying relevant genes and elucidating theirsal pathways.

Association withSuicidal Behavior

Heritability(20%)

ophenotypesggression/impulsivity YES YESarly-onset major depression YES YESeurocognitive function YES YESortisol stress response YES YESdidate Endophenotypes-HT in postmortem brain YES No dataSF 5-HIAA YES YES-HT in vivo (imaging) YES No datarain imaging (rCMRGlu) YES No dataecond messengers YES No dataPDInterpersonal reactivity No data YESAffect dysregulation YES YES

BPD, borderline personality disorder; CSF, cerebrospinal fluid; 5-HIAA, 5-htabolic rates.aAnimal studies.w.sobp.org/journalGene expression array data sets are enormous. False discov-rate corrections are often too stringent, and real findings canmissed (type 2 error). Verifying positive findings by a seconday method such as in situ hybridization histochemistry is oneeguard, and better statistical methods offer another approach.pe 2 error is also an issue in SNP association studies, evenere polymorphisms have established functional effects, i.e.,/met in COMT, where the small difference in frequenciestween affected and nonaffected groups means large samplees are required to detect any effect. Even in relatively largemogenous samples, it can be difficult to detect an effect; forample, a Finnish study of 2000 families with schizophreniand heritability of 54%, but despite examining 317,000 SNPsd characterizing the genome so precisely that individualsuld be identified by village of origin, no locus was significantschizophrenia (101).

nclusions

Given the heritability of suicidal behavior, more knowledgethe risk and/or protective genes would be valuable in thentification of high-risk individuals and the development ofatment interventions. Discovering the relevant genes and theirlogical role in complex and multidetermined phenotypesh as suicide and attempted suicide is a challenge. However,vances in clinical and biological understanding of suicide andempted suicide and methodological and technical develop-nts in genetic studies assist in this task. Moving from broadd point phenotypes to more specific narrower endopheno-es is one approach. Several promising endophenotypes thatgely meet the criteria of Gottesman and Gould (14) includepulsive-aggressive traits, early-onset major depression, neuro-gnitive function, and increased cortisol response to socialess. Systematic genetic studies of these endophenotypes areposed using genome-wide structural and expression arraysd superimposed mapping of methylation sites. Other potentialdidate biological and clinical endophenotypes have beenntified, based largely on association with the phenotype. Forse, studies are needed to establish that they can fulfill theteria for endophenotype. The heritability of suicidal behavior

Endophenotype Criteria

-IndependentCosegregate in Families

(Gene Association)More Frequent in

Nonaffected Relatives

YES YES YESYES YES No dataYES YES YESYES YES No data

No data Insufficient data No dataYESa Insufficient data No dataNo data No data No dataNo data No data No dataNo data No data No data

YES No data No dataYES No data No data

yindoleacetic acid; 5-HT, serotonin; rCMRGlu, regional cerebral glucoseofidetrebiosucadattmeentyplarimcostrproancanidethecri

is comparable with major psychiatric disorders and warrants acomparable effort to identify the responsible genes. Such studieswiMean

neexc

maJoean

ertPabioreptisFor

NasomBodevinvtioroyan

1

2

3

4

5

6

7

8

9

10

11

12

13. HuangYY, Cate SP, Battistuzzi C,OquendoMA, BrentD,Mann JJ (2004):An association between a functional polymorphism in the mono-

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

J.J. Mann et al. BIOL PSYCHIATRY 2009;65:556563 561ll require large sample sizes, and the National Institute ofntal Health (NIMH) Center for Collaborative Genetic Studies isimportant resource (http://nimhgenetics.org/).

The views expressed here are those of the authors and do notcessarily reflect the views of the authors institutions. With theeption of the first author, order of authorship is alphabetical.We thank the following for input and discussion of theterial presented in this manuscript: Yates Conwell, M.D.; Sean, Ph.D., L.M.S.W.; Gonzalo Laje, M.D.; Roy Perlis, M.D., M.Sc.;d Steven Zalcman, M.D.Drs. Arango, Avenevoli, Brent, Champagne, Currier, Dough-y, Haghighi, Hodge, Kleinman, Lehner, Moscicki, Oquendo,ndey, Pearson, Stanley, Terwilliger, and Wenzel have nomedical financial interests or potential conflicts of interest toort. Dr. Mann receives research funding support from Novar-and GlaxoSmithKlein. Dr. Clayton has been a speaker forest laboratories.Dr. McMahon makes the following statement of interest: thetional Institutes of Health (NIH) has filed a patent based one the diagnostic technology discussed herein. NeuroMark of

ulder, Colorado, negotiated a nonexclusive license with NIH toelop this technology commercially. Federal law prohibits theentors from any involvement in the negotiation and execu-n of this license but requires NIH to pay them a portion of thealties received. The inventors may not and have not endorsedy commercial use of the patent.

. U.S. Department of Health and Human Services, Centers for DiseaseControl and Prevention (2008): Injury center. Welcome to WISQARS.Available at: http://www.cdc.gov/ncipc/wisqars/. Accessed July 15,2008.

. Kessler RC, Berglund P, Borges G, Nock M, Wang PS (2005): Trends insuicide ideation, plans, gestures, and attempts in the United States,19901992 to 20012003. JAMA 293:24872495.

. Posner K, OquendoMA,GouldM, Stanley B, DaviesM (2007): Columbiaclassification algorithm of suicide assessment (C-CASA): Classificationof suicidal events in the FDAs pediatric suicidal risk analysis of antide-pressants. Am J Psychiatry 164:10351043.

. VoracekM, Loibl LM (2007): Genetics of suicide: A systematic review oftwin studies.Wien Klin Wochenschr 119:463475.

. Brent DA, MelhemN (2008): Familial transmission of suicidal behavior.Psychiatr Clin North Am 31:157177.

. Brezo J, Klempan T, Turecki G (2008): The genetics of suicide: A criticalreview of molecular studies. Psychiatr Clin North Am 31:179203.

. Gwadry FG, Sequeira A, Hoke G, Ffrench-Mullen JM, Turecki G (2005):Molecular characterization of suicide bymicroarray analysis. Am JMedGenet C SeminMed Genet 133:4856.

. SequeiraA,GwadryFG,Ffrench-MullenJM,Canetti L,GingrasY,CaseroRAJr, et al. (2006): Implication of SSATby gene expression andgenetic varia-tion in suicide andmajor depression.ArchGen Psychiatry 63:3548.

. Zill P, Buttner A, Eisenmenger W, Moller HJ, Bondy B, Ackenheil M(2004): Single nucleotide polymorphism and haplotype analysis of anovel tryptophan hydroxylase isoform (TPH2) gene in suicide victims.Biol Psychiatry 56:581586.

. Haghighi F, Bach-MizrachiH,HuangYY, ArangoV, Shi S, DworkAJ, et al.(2008): Genetic architecture of the human tryptophan hydroxylase 2Gene: Existenceof neural isoformsand relevance formajor depression.Mol Psychiatry 13:813820.

. Anguelova M, Benkelfat C, Turecki G (2003): A systematic review ofassociation studies investigating genes coding for serotonin receptorsand the serotonin transporter: II. Suicidal behavior. Mol Psychiatry8:646653.

. LajeG, Paddock S,Manji H, RushAJ,WilsonAF, CharneyD,McMahonFJ(2007): Genetic markers of suicidal ideation emerging during citalo-pram treatment of major depression. Am J Psychiatry 164:15301538.amine oxidase a gene promoter, impulsive traits and early abuse ex-periences. Neuropsychopharmacology 29:14981505.

. Gottesman II, Gould TD (2003): The endophenotype concept in psychi-atry: Etymologyand strategic intentions.AmJPsychiatry160:636645.

. Dumais A, Lesage AD, Alda M, Rouleau G, Dumont M, Chawky N, et al.(2005): Risk factors for suicide completion inmajor depression: A case-control study of impulsive and aggressive behaviors in men. Am JPsychiatry 162:21162124.

. McGirr A, Renaud J, Bureau A, Seguin M, Lesage A, Turecki G (2008):Impulsive-aggressivebehaviours andcompleted suicide across the lifecycle: A predisposition for younger ageof suicide. PsycholMed 38:407417.

. Oquendo MA, Galfalvy H, Russo S, Ellis SP, Grunebaum MF, Burke A,Mann JJ (2004): Prospective study of clinical predictors of suicidal actsafter a major depressive episode in patients with major depressivedisorder or bipolar disorder. Am J Psychiatry 161:14331441.

. Placidi GP, Oquendo MA, Malone KM, Huang YY, Ellis SP, Mann JJ(2001): Aggressivity, suicide attempts, and depression: Relationship tocerebrospinal fluid monoamine metabolite levels. Biol Psychiatry 50:783791.

. Coccaro EF, Bergeman CS, Kavoussi RJ, Seroczynski AD (1997): Herita-bility of aggression and irritability: A twin study of the Buss-Durkeeaggression scales in adult male subjects. Biol Psychiatry 41:273284.

. Rushton JP, Fulker DW, NealeMC, Nias DK, Eysenck HJ (1986): Altruismand aggression: The heritability of individual differences. J Pers SocPsychol 50:11921198.

. Gothelf D, Apter A, van Praag HM (1997): Measurement of aggressionin psychiatric patients. Psychiatry Res 71:8395.

. Brent DA, Oquendo MA, Birmaher B, Greenhill L, Kolko D, Stanley B, etal. (2003): Peripubertal suicide attempts in offspring of suicide at-tempters with siblings concordant for suicidal behavior. Am J Psychia-try 160:14861493.

. MelhemNM, Brent DA, ZieglerM, Iyengar S, KolkoD, OquendoM, et al.(2007): Familial pathways to early-onset suicidal behavior: Familial andindividual antecedents of suicidal behavior.Am J Psychiatry 164:13641370.

. ZoukH,McGirr A, Lebel V, Benkelfat C, RouleauG, Turecki G (2007): Theeffect of genetic variation of the serotonin 1B receptor gene on impul-sive aggressive behavior and suicide. Am JMedGenet B NeuropsychiatrGenet 144B:9961002.

. Giegling I, Hartmann AM, Moller HJ, Rujescu D (2006): Anger- andaggression-related traits are associated with polymorphisms in the5-HT-2A gene. J Affect Disord 96:7581.

. Manuck SB, Flory JD, Ferrell RE, Mann JJ, MuldoonMF (2000): A regula-tory polymorphism of themonoamine oxidase-A genemay be associ-ated with variability in aggression, impulsivity, and central nervoussystem serotonergic responsivity. Psychiatry Res 95:923.

. Kim CD, Seguin M, Therrien N, Riopel G, Chawky N, Lesage AD, TureckiG (2005): Familial aggregation of suicidal behavior: A family study ofmale suicide completers from the general population. Am J Psychiatry162:10171019.

. Scarr S,Webber PL,Weinberg RA,WittigMA (1981): Personality resem-blance among adolescents and their parents in biologically relatedand adoptive families. Prog Clin Biol Res 69(Pt B):99120.

. Eaves L, Eysenck H (1975): The nature of extraversion: A geneticalanalysis. J Pers Soc Psychol 32:102112.

. Dougherty DM, Bjork JM, Moeller FG, Harper RA, Marsh DM, MathiasCW, Swann AC (2003): Familial transmission of Continuous Perfor-mance Test behavior: Attentional and impulsive response characteris-tics. J Gen Psychol 130:521.

. Preuss UW, Koller G, Bondy B, Bahlmann M, Soyka M (2001): Impulsivetraits and 5-HT2A receptor promoter polymorphism in alcohol depen-dents: Possible association but no influence of personality disorders.Neuropsychobiology 43:186191.

. Baca-Garcia E, Vaquero C, Diaz-Sastre C, Garcia-Resa E, Saiz-Ruiz J,Fernandez-Piqueras J, de Leon J (2004): Lack of association betweenthe serotonin transporterpromotergenepolymorphismand impulsiv-ity or aggressive behavior among suicide attempters and healthy vol-unteers. Psychiatry Res 126:99106.www.sobp.org/journal

33. Congdon E, Lesch KP, Canli T (2008): Analysis of DRD4 and DAT poly-morphisms and behavioral inhibition in healthy adults: Implications

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54. Williams JM, Broadbent K (1986): Distraction by emotional stimuli: Useof a Stroop task with suicide attempters. Br J Clin Psychol 25(Pt 2):101

55

56

57

58

59

60

61

62

63

64

65

66

67

68

69

70

71

72

73

74

75

76

562 BIOL PSYCHIATRY 2009;65:556563 J.J. Mann et al.

wwfor impulsivity. Am JMed Genet B Neuropsychiatr Genet 147B:2732.. Langley K, Marshall L, van den BM, Thomas H, OwenM, ODonovan M,Thapar A (2004): Association of the dopamine D4 receptor gene 7-re-peat allele with neuropsychological test performance of children withADHD. Am J Psychiatry 161:133138.

. Dougherty DM, Mathias CW, Marsh DM, Papageorgiou TD, Swann AC,Moeller FG (2004): Laboratorymeasured behavioral impulsivity relatesto suicide attempt history. Suicide Life Threat Behav 34:374385.

. SwannAC, Dougherty DM, Pazzaglia PJ, PhamM, Steinberg JL,MoellerFG (2005): Increased impulsivity associated with severity of suicideattempt history in patients with bipolar disorder. Am J Psychiatry 162:16801687.

. DoughertyDM,Mathias CW,MarshDM (2004): Impulsivity and aggres-sion in adolescents with conduct disorder. Presented at the WorldPsychiatric Association, Clinical Research on Impulsivity: New Devel-opments and Directions for Possible Treatments, November 12, 2004,Florence, Italy.

. Dougherty DM, Mathias CW, Marsh-Richard DM, Prevette KN, DawesMA, Shannon EE, et al. (in press): Impulsivity and clinical symptomsamong adolescents with non-suicidal self-injury with or without at-tempted suicide. Psychiatry Res.

. Bjork JM, Moeller FG, Dougherty DM, Swann AC, Machado MA, HanisCL (2002): Serotonin 2a receptor T102C polymorphism and impairedimpulse control. Am J Med Genet 114:336339.

. Langley K, Marshall L, van den BM, Thomas H, OwenM, ODonovan M,Thapar A (2004): Association of the dopamine D4 receptor gene 7-re-peat allele with neuropsychological test performance of children withADHD. Am J Psychiatry 161:133138.

. Cornish KM, Manly T, Savage R, Swanson J, Morisano D, Butler N, et al.(2005): Association of the dopamine transporter (DAT1) 10/10-repeatgenotype with ADHD symptoms and response inhibition in a generalpopulation sample.Mol Psychiatry 10:686698.

. Stoltenberg SF, Glass JM, Chermack ST, Flynn HA, Li S, Weston ME,Burmeister M (2006): Possible association between response inhibi-tion and a variant in the brain-expressed tryptophan hydroxylase-2gene. Psychiatr Genet 16:3538.

. Eisenberg DT, Mackillop J, Modi M, Beauchemin J, Dang D, Lisman SA,et al. (2007): Examining impulsivity as an endophenotype using a be-havioral approach: A DRD2 TaqI A and DRD4 48-bp VNTR associationstudy. Behav Brain Funct 3:2.

. Schachar RJ, Crosbie J, Barr CL, Ornstein TJ, Kennedy J, MaloneM, et al.(2005): Inhibition of motor responses in siblings concordant and dis-cordant for attention deficit hyperactivity disorder. Am J Psychiatry162:10761082.

. Chamberlain SR, Fineberg NA, Menzies LA, Blackwell AD, Bullmore ET,Robbins TW, Sahakian BJ (2007): Impaired cognitive flexibility andmotor inhibition in unaffected first-degree relatives of patients withobsessive-compulsive disorder. Am J Psychiatry 164:335338.

. Zisook S, Rush AJ, Lesser I, Wisniewski SR, Trivedi M, Husain MM, et al.(2007): Preadult onset vs. adult onset of major depressive disorder: Areplication study. Acta Psychiatr Scand 115:196205.

. Thompson AH (2008): Younger onset of depression is associated withgreater suicidal intent. Soc Psychiatry Psychiatr Epidemiol 43:538544.

. LyonsMJ, Eisen SA, Goldberg J, TrueW, Lin N, Meyer JM, et al. (1998): Aregistry-based twin study of depression in men. Arch Gen Psychiatry55:468472.

. Pettit JW, Lewinsohn PM, Roberts RE, Seeley JR, Monteith L (2008): Thelong-term course of depression: Development of an empirical indexand identification of early adult outcomes [published online ahead ofprint July 8]. Psychol Med.

. Holmans P, Weissman MM, Zubenko GS, Scheftner WA, Crowe RR,DePaulo JR Jr, et al. (2007): Genetics of recurrent early-onset majordepression (GenRED): Final genome scan report. Am J Psychiatry 164:248258.

. Keilp JG, Gorlyn M, Oquendo MA, Burke AK, Mann JJ (2008): Attentiondeficit in depressed suicide attempters. Psychiatry Res 159:717.

. Marzuk PM, Hartwell N, Leon AC, Portera L (2005): Executive function-ing in depressed patients with suicidal ideation. Acta Psychiatr Scand112:294301.

. Becker ES, Strohbach D, Rinck M (1999): A specific attentional bias insuicide attempters. J Nerv Ment Dis 187:730735.w.sobp.org/journal110.. Wenzel A, Brown GK, Beck AT (2008): Cognitive Therapy for SuicidalPatients: Scientific and Clinical Applications. Washington, DC: APABooks.

. Stins JF, van Baal GC, Polderman TJ, Verhulst FC, Boomsma DI (2004):Heritability of Stroop and flanker performance in 12-year old children.BMC Neurosci 5:49.

. Swan GE, Carmelli D (2002): Evidence for genetic mediation of execu-tive control: A study of agingmale twins. J Gerontol B Psychol Sci Soc Sci57:133143.

. Reuter M, Peters K, Schroeter K, Koebke W, Lenardon D, Bloch B, Hen-nig J (2005): The influence of the dopaminergic system on cognitivefunctioning:Amolecular genetic approach.BehavBrainRes164:9399.

. Kramer UM, Cunillera T, Camara E, Marco-Pallares J, Cucurell D, NagerW, et al. (2007): The impact of catechol-O-methyltransferase anddopa-mineD4 receptor genotypes onneurophysiologicalmarkers of perfor-mance monitoring. J Neurosci 27:1419014198.

. Alfimova MV, Golimbet VE, Gritsenko IK, Lezheiko TV, Abramova LI,StreltsovaMA, et al. (2007): Interactionof dopamine systemgenes andcognitive functions in patients with schizophrenia and their relativesand in healthy subjects from the general population. Neurosci BehavPhysiol 37:643650.

. Szoke A, Schurhoff F, Golmard JL, Alter C, Roy I, Meary A, et al. (2006):Familial resemblance for executive functions in families of schizo-phrenic and bipolar patients. Psychiatry Res 144:131138.

. Szoke A, Schurhoff F, Mathieu F, Meary A, Ionescu S, Leboyer M (2005):Tests of executive functions in first-degree relatives of schizophrenicpatients: A meta-analysis. Psychol Med 35:771782.

. Zalla T, Joyce C, Szoke A, Schurhoff F, Pillon B, Komano O, et al. (2004):Executive dysfunctions as potential markers of familial vulnerability tobipolar disorder and schizophrenia. Psychiatry Res 121:207217.

. Reppermund S, IsingM, Lucae S, Zihl J (2008): Cognitive impairment inunipolar depression is persistent and non-specific: Further evidencefor the final common pathway disorder hypothesis [published onlineahead of print July 30]. Psychol Med.

. Westheide J, Quednow BB, Kuhn KU, Hoppe C, Cooper-Mahkorn D,Hawellek B, et al. (2008): Executive performance of depressed suicideattempters: The role of suicidal ideation. Eur Arch Psychiatry Clin Neu-rosci 258:414421.

. Mann JJ, Currier D, Stanley B, Oquendo MA, Amsel LV, Ellis SP (2006):Can biological tests assist prediction of suicide in mood disorders? IntJ Neuropsychopharmacol 9:465474.

. Lindqvist D, Isaksson A, Lil TB, Brundin L (2008): Salivary cortisol andsuicidal behaviora follow-up study. Psychoneuroendocrinology 33:10611068.

. Bartels M, Van den BM, Sluyter F, Boomsma DI, de Geus EJ (2003):Heritability of cortisol levels: Review and simultaneous analysis of twinstudies. Psychoneuroendocrinology 28:121137.

. Federenko IS, Nagamine M, Hellhammer DH, Wadhwa PD, Wust S(2004): The heritability of hypothalamus pituitary adrenal axis re-sponses to psychosocial stress is context dependent. J Clin EndocrinolMetab 89:62446250.

. Heim C, Nemeroff CB (2001): The role of childhood trauma in theneurobiology of mood and anxiety disorders: Preclinical and clinicalstudies. Biol Psychiatry 49:10231039.

. DeRijk RH, van Leeuwen N, Klok MD, Zitman FG (2008): Corticosteroidreceptor-gene variants: Modulators of the stress-response and impli-cations for mental health. Eur J Pharmacol 585:492501.

. Gotlib IH, Joormann J,Minor KL, Hallmayer J (2008): HPA axis reactivity:A mechanism underlying the associations among 5-HTTLPR, stress,and depression. Biol Psychiatry 63:847851.

. Uhart M, McCaul ME, Oswald LM, Choi L, Wand GS (2004): GABRA6genepolymorphismand an attenuated stress response.Mol Psychiatry9:9981006.

. Mann JJ (2003): Neurobiology of suicidal behaviour. Nat Rev Neurosci4:819828.

. Asberg M (1997): Neurotransmitters and suicidal behavior. The evi-dence from cerebrospinal fluid studies. Ann N Y Acad Sci 836:158181.

. Higley JD, Linnoila M (1997): Low central nervous system serotonergicactivity is traitlike and correlates with impulsive behavior. A nonhu-

man primate model investigating genetic and environmentalinfluences on neurotransmission. Ann N Y Acad Sci 836:3956.

77. Higley JD, ThompsonWW, ChampouxM, GoldmanD, Hasert MF, Krae-mer GW, et al. (1993): Paternal andmaternal genetic and environmen-tal contributions to cerebrospinal fluid monoamine metabolites inRhesus monkeys (Macaca mulatta). Arch Gen Psychiatry 50:615623.

78. Rogers J, Martin LJ, Comuzzie AG, Mann JJ, Manuck SB, Leland M,Kaplan JR (2004): Genetics of monoamine metabolites in baboons:Overlapping sets of genes influence levels of 5-hydroxyindolaceticacid, 3-hydroxy-4-methoxyphenylglycol, and homovanillic acid. BiolPsychiatry 55:739744.

79. Williams RB, Marchuk DA, Gadde KM, Barefoot JC, Grichnik K, HelmsMJ, et al. (2003): Serotonin-related gene polymorphisms and cen-tral nervous system serotonin function. Neuropsychopharmacology28:533541.

80. Mann JJ, Currier D, Murphy L, Huang YY, Galfalvy H, Brent D, et al.(2008): No association between a TPH2 promoter polymorphism andmood disorders or monoamine turnover. J Affect Disord 106:117121.

81. Tiihonen J, Kuikka JT, Bergstrm KA, Karhu J, Lehtonen J, Hallikainen T,et al. (1997): Single-photon emission tomography imaging of mono-amine transporters in impulsive violent behaviour. Eur J Nucl Med 24:12531260.

82. Parsey RV, Oquendo MA, Simpson NR, Ogden RT, Van Heertum R,Arango V, Mann JJ (2002): Effects of sex, age, and aggressive traits inman on brain serotonin 5-HT1A receptor binding potential measuredby PET using [11C]WAY-100635. Brain Res 954:173182.

83

84

85

86

87

88

89. Audenaert K, Goethals I, Van LK, Lahorte P, Brans B, Versijpt J, et al.(2002): SPECT neuropsychological activation procedure with the Ver-bal Fluency Test in attempted suicide patients. Nucl Med Commun23:907916.

90. Pandey GN, Dwivedi Y, Rizavi HS, Teppen T, Gaszner GL, Roberts RC,Conley RR (2008): GSK-3beta gene expression in human postmortembrain: Regional distribution, effects of age and suicide [published on-line ahead of print June 28]. Neurochem Res.

91. Pandey GN, Dwivedi Y, Rizavi HS, Ren X, Conley RR (2004): Decreasedcatalytic activity and expression of protein kinase C isozymes in teen-age suicide victims: A postmortem brain study. Arch Gen Psychiatry61:685693.

92. Dwivedi Y, Rao JS, Rizavi HS, Kotowski J, Conley RR, Roberts RC, et al.(2003): Abnormal expression and functional characteristics of cyclicadenosine monophosphate response element binding protein inpostmortembrain of suicide subjects. Arch Gen Psychiatry 60:273282.

93. Dwivedi Y, Rizavi HS, Conley RR, Roberts RC, Tamminga CA, PandeyGN(2003): Altered gene expression of brain-derived neurotrophic factorand receptor tyrosine kinaseB inpostmortembrainof suicide subjects.Arch Gen Psychiatry 60:804815.

94. Pandey GN, Ren X, Rizavi HS, Conley RR, Roberts RC, Dwivedi Y (2008):Brain-derived neurotrophic factor and tyrosine kinase B receptor sig-nalling inpost-mortembrainof teenage suicide victims. Int J Neuropsy-chopharmacol 11:10471061.

95. Torgersen S, Czajkowski N, Jacobson K, Reichborn-Kjennerud T,Roysamb E, NealeMC, Kendler KS (2008): Dimensional representationsofDSM-IV cluster Bpersonalitydisorders in apopulation-based sample

96

97

98

99

100

101

J.J. Mann et al. BIOL PSYCHIATRY 2009;65:556563 563. Leyton M, Paquette V, Gravel P, Rosa-Neto P, Weston F, Diksic M,Benkelfat C (2006): alpha-[11C]Methyl-L-tryptophan trapping in theorbital and ventral medial prefrontal cortex of suicide attempters. EurNeuropsychopharmacol 16:220223.

. van Heeringen C, Audenaert K, Van Laere K, Dumont F, Slegers G,Mertens J, Dierckx RA (2003): Prefrontal 5-HT2a receptor binding in-dex, hopelessness and personality characteristics in attempted sui-cide. J Affect Disord 74:149158.

. OquendoMA, Hastings RS, Huang YY, Simpson N, Ogden RT, Hu XZ, etal. (2007): Brain serotonin transporter binding in depressed patientswith bipolar disorder using positron emission tomography. Arch GenPsychiatry 64:201208.

. OquendoMA, Placidi GP,Malone KM, Campbell C, Keilp J, Brodsky B, etal. (2003): Positron emission tomography of regional brain metabolicresponses to a serotonergic challengeand lethality of suicide attemptsin major depression. Arch Gen Psychiatry 60:1422.

. Soloff PH, Meltzer CC, Greer PJ, Constantine D, Kelly TM (2000): Afenfluramine-activated FDG-PET study of borderline personality disor-der. Biol Psychiatry 47:540547.

. Goethals I, Audenaert K, Jacobs F, Van den EF, Bernagie K, Kolindou A,et al. (2005): Brain perfusion SPECT in impulsivity-related personalitydisorders. Behav Brain Res 157:187192.of Norwegian twins: A multivariate study. Psychol Med 38:16171625.. Skodol AE, Siever LJ, Livesley WJ, Gunderson JG, Pfohl B, Widiger TA(2002): The borderline diagnosis II: Biology, genetics, and clinicalcourse. Biol Psychiatry 51:951963.

. Mill J, Dempster E, Caspi A,WilliamsB,Moffitt T, Craig I (2006): Evidencefor monozygotic twin (MZ) discordance in methylation level at twoCpG sites in the promoter region of the catechol-O-methyltransferase(COMT) gene. Am JMed Genet B Neuropsychiatr Genet 141:421425.

. Meaney MJ (2001): Maternal care, gene expression, and the transmis-sion of individual differences in stress reactivity across generations.Annu Rev Neurosci 24:11611192.

. Liu D, Diorio J, Tannenbaum B, Caldji C, Francis D, Freedman A, et al.(1997): Maternal care, hippocampal glucocorticoid receptors, and hy-pothalamic-pituitary-adrenal responses to stress. Science 277:16591662.

. Weaver IC, Champagne FA, Brown SE, Dymov S, Sharma S, MeaneyMJ,Szyf M (2005): Reversal of maternal programming of stress responsesin adult offspring through methyl supplementation: Altering epige-netic marking later in life. J Neurosci 25:1104511054.

. Marchini J, Donnelly P, Cardon LR (2005): Genome-wide strategies fordetecting multiple loci that influence complex diseases. Nat Genet37:413417.www.sobp.org/journal

Candidate Endophenotypes for Genetic Studies of Suicidal BehaviorSuicidal Behavior PhenotypesSuicidal Behavior and GeneticsHeritabilityGenes and Suicidal Behavior

Defining Endophenotypes for Suicidal BehaviorImpulsive and Aggressive TraitsEarly-Onset Major DepressionNeurocognitive FunctionCortisol Response to Psychosocial Stress

Candidate EndophenotypesSerotonergic System Alterations in Suicide and Nonfatal Suicide AttemptBrain Structure and Function In VivoSecond MessengersBorderline Personality Disorder

Genetic StudiesOther ConsiderationsEpigeneticsMethodological Issues

ConclusionsAcknowledgmentReferences