Tryptophan hydroxylase polymorphism and suicidality in unipolar and bipolar affective disorders: a multicenter association study

ORIGINAL ARTICLES Tryptophan Hydroxylase Polymorphism and Suicidality in Unipolar and Bipolar Affective Disorders: A Multicenter Association Study Daniel Souery, Sophie Van Gestel, Isabelle Massat, Sylvie Blairy, Rolf Adolfsson, Douglas Blackwood, Jurgen Del-Favero, Dimitris Dikeos, Miro Jakovljevic, Radka Kaneva, Enrico Lattuada, Bernard Lerer, Roberta Lilli, Vihbra Milanova, Walter Muir, Markus No¨then, Lilijana Oruc, George Papadimitriou, Peter Propping, Thomas Schulze, Alessandro Serretti, Baruch Shapira, Enrico Smeraldi, Costas Stefanis, Marian Thomson, Christine Van Broeckhoven, and Julien Mendlewicz Background: Being the rate-limiting enzyme in the biosynthesis of serotonin, the tryptophan hydroxylase gene (TPH) has been considered a possible candidate gene in bipolar and unipolar affective disorders (BPAD and UPAD). Several studies have investigated the possible role of TPH polymorphisms in affective disorders and suicidal behavior. Methods: The TPH A218C polymorphism has been investigated in 927 patients (527 BPAD and 400 UPAD) and their matched healthy control subjects collected within the European Collaborative Project on Affective Disorders. Results: No difference of genotype distribution or allele distribution was found in BPAD or UPAD. No statistically significant difference was observed for allele frequency and genotypes counts. In a genotype per genotype analysis in UPAD patients with a personal history of suicide attempt, the frequency of the C–C genotype (homozygosity for the short allele) was lower in UPAD patients (24%) than in control subjects (43%) (␹2 ⫽ 4.67, p ⫽ .03). There was no difference in allele or genotype frequency between patients presenting violent suicidal behavior (n ⫽ 48) and their matched control subjects. From the Department of Psychiatry, University Clinics of Brussels, Erasme Hospital, Free University of Brussels, Brussels (DS, IM, SB, JM), and Flanders Interuniversity Institute for Biotechnology, Department of Biochemistry, University of Antwerpen, Antwerp (SVG, JD-F, CVB), Belgium; the Department of Psychiatry, University of Umea¨, Umea¨, Sweden (RA); Istituto Scientifico Ospedale San Raffaele, Department of Neuropsychiatric Sciences, University of Milan School of Medicine, Milan, Italy (EL, RL, AS, ES); University Department of Psychiatry and MRC Human Genetic Unit, Edinburgh, United Kingdom (DB, WM, MT); the Department of Psychiatry, Athens University Medical School and University Mental Health Research Institute, Athens, Greece (DD, GP, CS); the Department of Psychiatry, University Hospital “REBRO”, University of Zagreb, Zagreb, Croatia (MJ); First Psychiatry Clinic, Department of Psychiatry, Alexander University Hospital and Genetics Laboratory, Maternity Hospital, Medical University, Sofia, Bulgaria (RK, VM); Biological Psychiatry Laboratory, Hadassah Medical Organization, Jerusalem, Israel (BL, BS); Institute of Human Genetics and Department of Psychiatry, University of Bonn, Bonn, Germany (MN, PP, TS); University Hospital Sarajevo, Sarajevo, Bosnia and Herzegovina (LO). Address reprint requests to Daniel Souery, Free University of Brussels, University Clinics of Brussels, Erasme Hospital, 808 route de Lennik, B-1070, Brussels, Belgium. Received March 20, 2000; revised July 6, 2000; accepted July 12, 2000.

© 2001 Society of Biological Psychiatry

Conclusions: We failed to detect an association between the A218C polymorphism of the TPH gene and BPAD and UPAD in a large European sample. Homozygosity for the short allele is significantly less frequent in a subgroup of UPAD patients with a history of suicide attempt than in control subjects. Biol Psychiatry 2001;49:405– 409 © 2001 Society of Biological Psychiatry Key Words: Bipolar affective disorders, suicide, tryptophan hydroxylase polymorphism, association study, linkage disequilibrium

Introduction

D

uring the past two decades, the search for genes for mood disorders has highlighted the genetic complexity of these disorders. Studies screening parts of or the total genome have identified promising candidate regions on chromosomes 4, 5, 11, 12, 18, 21, and X (Souery et al 1996). A number of candidate genes have also been investigated, and some of these are directly linked to neurobiological hypotheses of the etiology of affective disorders. Since gene(s) of minor effect may be involved in affective disorders, the association method testing candidate genes is a powerful alternative approach to linkage (No¨then et al 1993). Given that the affective disorders are genetically heterogeneous, one of the principal limitations of association studies is the large sample size required to have power to detect gene(s) of minor effect. This has been taken into account within the European Collaborative Project on Affective Disorders (Souery et al 1998). This project (BIOMED 1 and 2 projects in the area of brain research; European Community Grants Nos. CT 92-1217 and BMH4-CT-97-2307; project leader, JM) has as its major objective the use of both linkage and association strategies, in very large samples of patients and families from several populations, to locate genes contributing to these complex traits. 0006-3223/01/$20.00 PII S0006-3223(00)01043-X

406

D. Souery et al

BIOL PSYCHIATRY 2001;49:405– 409

Being the rate-limiting enzyme in the biosynthesis of serotonin, the tryptophan hydroxylase gene (TPH) has been considered a possible candidate gene in bipolar and unipolar affective disorders (BPAD and UPAD). The TPH gene is located on 11p15.3-p14 (Craig et al 1991). Two biallelic polymorphisms have been identified in intron 7 of TPH, on positions 218 and 779 (A218C and A779C; Bellivier et al 1998; Nielsen et al 1997). Several studies have investigated the possible role of these polymorphisms in affective disorders, suicidal behavior, and lithium prophylaxis. The A779C (allele 779C) polymorphism was first reported to be associated with a history of suicide attempts in violent offenders in a sample of Finnish alcoholics (Nielsen et al 1994), whereas an association with allele 779A was observed in a subsequent study (Mann et al 1997). The first finding was replicated in a larger sample including siblings (Nielsen et al 1998), but Abbar et al (1995) found no association between suicidal behavior and TPH polymorphism. Other studies have included patients with major affective disorders (BPAD and/or UPAD). The A218C polymorphism was associated with a small increase in susceptibility to BPAD in a European sample (Bellivier et al 1998), but this finding was not replicated in several other studies (Furlong et al 1998; Kirov et al 1999; Kunugi et al 1999; McQuillin et al 1999; Vincent et al 1999). Serretti et al (1999) have recently investigated the possible association between TPH and the prophylactic efficacy of lithium in mood disorders and found an association between TPH variants (patients with genotype A/A of the A218C polymorphism had a trend toward a worse response) and lithium outcome. In this article we present the results of a large multicenter association study between the TPH A218C polymorphism and UPAD and BPAD. Subgroups of patients with suicidal behavior have also been analyzed.

Methods and Materials Sample A total of 927 patients (527 BPAD and 400 UPAD) (Table 1) and their matched healthy control subjects have been included with DNA available. Each patient is matched to a control subject of the same ethnogeographic origin, the same gender, and a maximal age difference of 5 years. All these subjects have been diagnosed using the Schedule for Affective Disorders and Schizophrenia—Lifetime Versions (SADS-LA) or the Schedules for Clinical Assessment of Neuropsychiatry (SCAN). All the patients met the diagnosis of BPAD or UPAD according to Research Diagnostic Criteria, DSM-III-R, and DSM-IV classification systems. One of the two diagnostic interviews was used for all patients and control subjects recruited for the project. The decision to adopt two instruments arose out of different research experience within individual research teams having their own preferences and expertise with the two interviews. Comparability

Table 1. Number of Patients Selected for Each Country Country Belgium Bulgaria Croatia Germany Greece Israel Italy Sweden United Kingdom Total

BPAD

UPAD

72 79 67 36 34 52 73 — 114 527

48 34 35 25 12 36 57 83 70 400

Each patient is matched to a control subject of the same ethnogeographic origin, the same gender, and a maximal age difference of 5 years (527 control subjects for BPAD patients and 400 control subjects for UPAD patients). BPAD, bipolar affective disorder; UPAD, unipolar affective disorder.

between SADS and SCAN instruments is inferred from data published by the European Science Foundation (Farmer et al 1993) showing good concordance between the two instruments. Within the control group, subjects with a positive personal or family history (assessed by the Family History Research Diagnostic Criteria) of major affective disorder were not included. Specific clinical characteristics recorded in the central data base include age at onset and suicide attempts (both violent and nonviolent). Suicide attempt is defined as having at least one suicide attempt in the lifetime. After description of the study written informed consent was obtained from each subject, and the research conformed to the standards set by local ethics of medical research committees for each of the participating centers.

DNA Analysis Standard polymerase chain reaction (PCR) was performed on genomic DNA using the primers TPH-F (5⬘-TTCAGATCCCTTCTATACCCCAGAG-3⬘) and TPH-R (5⬘-GGACATGACCTAAGAGTTCATGGCA-3⬘). Polymerase chain reaction was performed in a 20-␮L volume containing 100 ng genomic DNA, 20 mmol/L of each dinucleotide triphosphate (dNTP) (20 mmol/L of a 7-deaza-deoxyguanosine triphosphate [dGTP]/ dGTP mixture was used), 1.5 mmol/L MgCl2, 10 pmol of each primer, and 1 U Taq DNA polymerase. After an initial denaturation step at 94° for 4 min, 30 cycles were performed at 94° for 1 min, annealing at 64° for 0.5 min, and extension at 72° for 1 min. Finally, an additional elongation step was performed at 72° for 6 min. The PCR products were digested using the BfaI restriction enzyme at 37° for 1 hour. Unrestricted fragments are 860 bp long and, after digestion, 615 and 245 bp (alleles A and C, respectively). The fragments were electrophoresed on a 1.5% agarose gel and were visualized by ethidium bromide staining.

Statistical Analysis The analysis of allelic association consisted of comparison of marker allele and genotype frequencies between patients and control subjects using ␹2 statistics. Special effort was made in the final analysis to carefully match

Affective Disorders and TPH Polymorphism

BIOL PSYCHIATRY 2001;49:405– 409

407

Table 2. Allele and Genotype Counts and Frequencies for the TPH Polymorphism in BPAD and UPAD Patients and Their Matched Control Subjects

Allelea A C Genotypesb A–A A–C C–C

BPAD [n ⫽ 527 (%)]

Control subjects, [n ⫽ 527 (%)]

UPAD [n ⫽ 400 (%)]

Control subjects, [n ⫽ 400 (%)]

438 (42) 616 (58)

434 (41) 620 (59)

349 (44) 451 (56)

358 (45) 442 (55)

90 (17) 258 (49) 179 (34)

82 (16) 270 (51) 175 (33)

70 (18) 209 (52) 121 (30)

78 (19) 202 (51) 120 (30)

Number for each allele represents a total number of times the allele was observed (n ⫻ 2). Frequencies are given in parentheses. TPH, tryptophan hydroxylase gene; BPAD, bipolar affective disorder; UPAD, unipolar affective disorder. a BPAD vs. control subjects, ␹2 ⫽ 0.03, p ⫽ .86; UPAD vs. control subjects, ␹2 ⫽ 0.205, p ⫽ .65. b BPAD vs. control subjects, ␹2 ⫽ 0.69, p ⫽ .708; UPAD vs. control subjects, ␹2 ⫽ 0.55, p ⫽ .757.

patients and control subjects for ethnogeographic origin to reduce the possibility of stratification bias. Analyses were performed on the pairwise sample of patients and control subjects according to ethnogeographic origin. Hardy–Weinberg equilibrium is calculated for the overall sample with the updated version of GENEPOP software (version 3.1d, March 1999, Raymond and Rousset 1995).

Results The genotypic distributions of patients and controls were in Hardy–Weinberg equilibrium in the overall sample (␹2 ⫽ 20.6, p ⫽ .298 for UPAD patients; ␹2 ⫽ 17.4, p ⫽ .492 for control subjects matched to UPAD patients; ␹2 ⫽ 16.9, p ⫽ .393 for BPAD patients; and ␹2 ⫽ 21.5, p ⫽ .160 for control subjects matched to BPAD patients). We tested the power of our sample considering an alpha value of .05, two-tailed. For the BPAD sample we had a good power (.80) to detect an odds ratio (OR) of 1.43 (considering a frequency of the risk allele of .50 among patients and of .41 among control subjects, effect size w ⫽ 0.09) (G POWER computer program, Bonn University, Department of Psychiatry, Germany). For the UPAD sample we had a power (.80) to detect an OR of 1.49 (considering a frequency of the risk allele of .55 among patients and of .45 among control subjects, effect size w ⫽ 0.1). The power was 1 for both samples, considering a moderate effect size of 0.2. The allele frequencies used in these estimations differ from those described by Nielsen et al (1997), but we investigated a different polymorphism in a different population. The frequencies observed are also different from the Genome Data Bank, in which the shorter (A) allele is found in 60% of the population and the longer (C) allele in 40% of the population. Table 2 shows the genotype and allele distributions for the TPH polymorphism in the overall sample of 527 BPAD patients and 400 UPAD patients and their matched normal control subjects. No difference of genotype distribution (BPAD vs. control subjects, ␹2 ⫽ 0.69, p ⫽ .71;

UPAD vs. control subjects, ␹2 ⫽ 0.55, p ⫽ .76) or allele distribution (BPAD vs. control subjects, ␹2 ⫽ 0.031, p ⫽ .86; UPAD vs. control subjects, ␹2 ⫽ 0.205, p ⫽ .65) was found in BPAD or UPAD. In a next step we stratified the patient samples with regard to presence or absence of suicide attempt. Finally, we looked at the patients who presented violent suicidal behavior such as hanging, drowning, jumping, and wrist cutting. For this last subgroup, BPAD and UPAD patients were analyzed together due to the small number available (48 patients). Among the 927 patients, 167 patients had at least one suicide attempt (58 UPAD and 109 BPAD). As shown in Table 3, no statistically significant difference was observed for allele frequency (BPAD vs. control subjects, ␹2 ⫽ 0.038, p ⫽ .85; UPAD vs. control subjects, ␹2 ⫽ 2.97, p ⫽ .084) and genotypes counts (BPAD vs. control subjects, ␹2 ⫽ 0.088, p ⫽ .96; UPAD vs. control subjects, ␹2 ⫽ 4.75, p ⫽ .09). However, when a genotype per genotype analysis was done, the frequency of the C–C genotype was lower in UPAD patients (24%) with suicide attempts than in control subjects (43%) (␹2 ⫽ 4.67, p ⫽ .03). When corrected for multiple testing (Bonferoni correction, three tests), this observation did not remain significant (p ⫽ .09). There was no difference in allele (p ⫽ .31) or genotype frequency (p ⫽ .30) between patients presenting violent suicidal behavior (n ⫽ 48) and their matched control subjects (data not shown). A logistic regression (stepwise method) was conducted with diagnosis as the dependent variable and genotypes (A–A, A–C, C–C) as predictors. Genotype C–C entered in the equation ␹2(1) ⫽ 4.67, p ⫽ .03. However, R2 ⬍ .16 indicated a weak relationship between the two variables. Stratification of the sample according to age at onset did not reveal any significant difference between BPAD (n ⫽ 236) and UPAD (n ⫽ 93) patients having an age at onset inferior to 25 and normal control subjects for genotype frequency (p ⫽ .451 and p ⫽ .423).

408

D. Souery et al

BIOL PSYCHIATRY 2001;49:405– 409

Table 3. Allele and Genotype Counts and Frequencies for TPH in BPAD and UPAD Patients with Suicide Attempts and Their Matched Control Subjects

Allelea A C Genotypesb A–A A–C C–C

BPAD [n ⫽ 109 (%)]

Control subjects, [n ⫽ 109 (%)]

UPAD [n ⫽ 58 (%)]

Control subjects, [n ⫽ 58 (%)]

87 (39) 131 (61)

85 (38) 133 (62)

56 (48) 60 (52)

43 (37) 73 (63)

17 (16) 53 (48) 39 (36)

17 (16) 51 (47) 41 (37)

12 (21) 32 (55) 14 (24)

10 (17) 23 (40) 25 (43)

TPH, tryptophan hydroxylase gene; BPAD, bipolar affective disorder; UPAD, unipolar affective disorder. a BPAD vs. control subjects, ␹2 ⫽ 0.04, p ⫽ .85; UPAD vs. control subjects, ␹2 ⫽ 2.97, p ⫽ .08. b BPAD vs. control subjects, ␹2 ⫽ 0.09, p ⫽ .96; UPAD vs. control subjects, ␹2 ⫽ 4.75, p ⫽ .09.

Discussion Some association studies have implicated TPH gene variants in the etiology of mood disorders or in suicidal behavior, supporting a role of serotonin both in mood disorders and suicidality. However, these findings have not been replicated in this study nor by several other groups. The absence of a significant difference of TPH variants between patients and control subjects should, however, be interpreted with caution because sample sizes used in previous studies may not be large enough to detect association, making it difficult to conclude on the exact role of this marker in the phenotypes investigated. Two different polymorphisms (A779C and A218C), both in intron 7 of the TPH gene and in tight linkage disequilibrium, were investigated in these studies (Bellivier et al 1998; Nielsen et al 1994). The A218C polymorphism was not associated with BPAD or UPAD in our large European sample. Previously we also reported the lack of association between serotonin 2A (5-HT2A) receptor gene T102C variants and BPAD and between a tyrosine hydroxylase polymorpism and BPAD in the same population (Massat et al 2000; Souery et al 1999). The large sample size provided by the multicenter approach in our study (527 BPAD and 400 UPAD patients and their matched control subjects) allows reaching a high statistical power (see Results). Ethnic and geographic origin are frequently a cause of stratification bias in case-control association studies. This observation stresses the importance of carefully taking into account population stratification. To circumvent this problem, each patient was matched to an unrelated control subject of the same ethnogeographic origin. In addition, the overall sample in the analysis was in Hardy–Weinberg equilibrium. We also investigated the possible role of TPH polymorphism in suicidal behavior in mood-disordered patients. Only in UPAD patients with prior personal histories of suicide attempts was an association found with TPH. The

frequency of the genotype C–C, which means homozygosity for the short allele, was lower in UPAD than in control subjects. No difference was found for BPAD patients, nor for patients with violent suicidal behavior. However, for this last subgroup results should be interpreted with caution because BPAD and UPAD patients were analyzed together to reach a reasonable sample size of 48 patients. The negative findings of Abbar et al (1995) and Kunugi et al (1999) are not consistent with our finding for suicidality. Our population is very different from the population of Finnish alcoholics studied by Nielsen et al (1998), and the TPH polymorphism investigated is not the same. However, both polymorphisms are in tight linkage disequilibrium (Kunugi et al 1999; Nielsen et al 1997), and in both studies an association with suicidality was found. This suggests the implication of serotonin disturbances in suicidal behavior. In conclusion, we failed to detect an association between the A218C polymorphism of the TPH gene and BPAD and UPAD in a large European sample. Homozygosity for the short allele is significantly less frequent in a subgroup of UPAD patients with a history of suicide attempts than in control subjects. In the small sample of patients with violent suicide attempts, no association with TPH polymorphism was found.

Dr. Souery and Dr. Van Gastel contributed equally to this project, which was supported by the National Foundation for Scientific Research (Grant No. 3.4504.97), Belgium; the Fund for Scientific Research (FWO), Flanders, Belgium (SVG is a Ph.D. fellow of the FWO); the European Community Biomed 1 and Biomed 2 Grants (Nos. CT 92-1217 and BMH4-97-2307); and the Association for Mental Health Research, Belgium.

References Abbar M, Courtet P, Amadeo S, Caer Y, Mallet J, BaldyMoulinier M, et al (1995): Suicidal behaviours and the

Affective Disorders and TPH Polymorphism

tryptophan hydroxylase gene. Arch Gen Psychiatry 52:846 – 849. Bellivier F, Leboyer M, Courtet P, Buresi C, Beaufils B, Samolyk D, et al (1998): Association between the tryptophan hydroxylase gene and manic-depressive illness. Arch Gen Psychiatry 55:33–37. Craig SP, Boularand S, Darmon MC, Mallet J, Craig IW (1991): Localization of human tryptophan hydroxylase (TPH) to chromosome 11p15.3-p14 by in situ hybridization. Cytogenet Cell Genet 56:157–159. Farmer AE, Cosyns P, Leboyer M, Maier W, Mors O, Sargeant M, et al (1993): A SCAN-SADS comparison study of psychotic subjects and their first degree relatives. Eur Arch Psychiatry Clin Neurosci 242:352–357. Furlong RA, Ho L, Rubinsztein JS, Walsh C, Paykel ES, Rubinsztein DC (1998): No association of the tryptophan hydroxylase gene with bipolar affective disorder, unipolar affective disorder, or suicidal behaviour in major affective disorder. Am J Med Genet 81:245–247. Kirov G, Owen MJ, Jones I, McCandless F, Craddock N (1999): Tryptophan hydroxylase gene and manic-depressive illness. Arch Gen Psychiatry 56:98 –99. Kunugi H, Ishida S, Kato T, Sakai T, Tatsumi M, Hirose T, Nanko S (1999): No evidence for an association of polymorphisms of the tryptophan hydroxylase gene with affective disorders or attempted suicide among japonese patients. Am J Psychiatry 156:774 –776. Mann JJ, Malone KM, Nielsen DA, Goldman D, Erdos J, Gelernter J (1997): Possible association of a polymorphism of the tryptophan hydroxylase gene with suicidal behaviour in depressed patients. Am J Psychiatry 154:1451–1453. Massat I, Souery D, Lipp O, Blairy S, Papadimitriou G, Dikeos D, et al (2000): A European multicenter association study of HTR2A receptor polymorphism in bipolar affective disorder. Am J Med Genet 96:136 –140. McQuillin A, Lawrence J, Kalsi G, Gurling H, Curtis D (1999): No allelic association between bipolar affective disorder and

BIOL PSYCHIATRY 2001;49:405– 409

409

the tryptophan hydroxylase gene. Arch Gen Psychiatry 56: 99 –100. Nielsen DA, Goldman D, Virkkunen M, Tokola R, Rawlings R, Linnoila M (1994): Suicidality and 5-hydroxyindoleacetic acid concentration associated with a tryptophan hydroxylase polymorphism. Arch Gen Psychiatry 51:34 –38. Nielsen DA, Jenkins GL, Stefanisko KM, Jefferson KK, Goldman D (1997): Sequence, splice site and population frequency distribution analyses of the polymorphic human tryptophan hydroxylase intron 7. Brain Res Mol Brain Res 45:145–148. Nielsen DA, Virkkunen M, Lappalainen J, Eggert M, Brown GL, Long JC, et al (1998): A tryptophan hydroxylase gene marker for suicidality and alcoholism. Arch Gen Psychiatry 55:593– 602. No¨then MM, Fimmers R, Propping P (1993): Association versus linkage studies in psychosis genetics. J Med Genet 30:634 – 637. Raymond M, Rousset F (1995): Population genetics software for exact tests and eucumenism. J Hered 86:248 –249. Serretti A, Lilli R, Lorenzi C, Gasperini M, Smeraldi E (1999): Tryptophan hydroxylase and response to lithium in in mood disorders. J Psychiatr Res 33:371–377. Souery D, Lipp O, Mahieu B, Serretti A, Cavallini C, Macciardi F, et al (1998): European Collaborative Project on Affective Disorders. Interactions between genetic and psychosocial vulnerability factors. Psychiatr Genet 8:197–206. Souery D, Lipp O, Rivelli SK, Massat I, Serretti A, Cavallini C, et al (1999): Tyrosine hydroxylase gene and phenotypic heterogeneity in bipolar affective disorder: A multicenter association study. Am J Med Genet 88:527–532. Souery D, Papadimitriou GN, Mendlewicz J (1996): New genetic approaches in affective disorders. Baillieres Clin Psychiatry 2(1):1–13. Vincent JB, Masellis M, Lawrence J, Choi V, Gurling HM, Parikh SV, Kennedy JL (1999): Genetic association analysis of serotonin system genes in bipolar affective disorder. Am J Psychiatry 156:136 –138.