5-HT2A SNPs and the Temperament and Character Inventory

Progress in Neuro-Psychopharmacology & Biological Psychiatry 31 (2007) 1275 – 1281 www.elsevier.com/locate/pnpbp

5-HT2A SNPs and the Temperament and Character Inventory Alessandro Serretti a,⁎, Raffaella Calati a , Ina Giegling b , Annette M. Hartmann b , Hans-Jürgen Möller b , Cristina Colombo c , Dan Rujescu b a

Institute of Psychiatry, University of Bologna, Viale Carlo Pepoli 5, 40123 Bologna, Italy b Department of Psychiatry, Ludwig Maximilians University, Munich, Germany c Psychiatry Department, San Raffaele Scientific Institute, Milan, Italy Received 29 January 2007; received in revised form 7 May 2007; accepted 18 May 2007 Available online 26 May 2007

Abstract Temperamental traits, the most basic part of personality, have been largely correlated with neurotransmitter systems and are under genetic control. Among serotonin candidates, the 2A receptor (5-HT2A) received considerable attention. We analyzed four SNPs (rs643627, rs594242, rs6311 and rs6313) in the 5-HT2A gene and their association with personality traits, as measured with the Temperament and Character Inventory (TCI). The sample was composed of three sub-groups: two German sub-samples, consisting of a healthy group of 289 subjects (42.6% males, mean age: 45.2 ± 14.9) and a psychiatric patient group of 111 suicide attempters (38.7% males, mean age: 39.2 ± 13.6), and an Italian sub-sample, composed of 60 mood disorder patients (35.0% males, mean age: 44.0 ± 14.8). Controlling for sex, age and educational level, the SNPs were not strongly associated with personality dimensions. Only the rs594242 showed an association with Self-Directedness (p = 0.003) in the German sample, while rs6313 was marginally associated with Novelty Seeking (p = 0.01) in the Italian sample. We conclude that 5-HT2A SNPs may marginally modulate personality traits but further studies are required. © 2007 Elsevier Inc. All rights reserved. Keywords: 5-HT2A; Character; Personality; TCI; Temperament

1. Introduction Personality traits influence many aspects of normal and pathologic behaviours (Bienvenu and Stein, 2003; Cloninger et al., 1994; Grucza et al., 2005; Klein et al., 1993; Velting, 1999). Temperamental traits, the most basic part of personality, have been correlated with neurotransmitter systems and are under genetic control (Cloninger et al., 1994). In the last decade, a large number of studies focused on the detection of gene variants associated with specific temperament traits and numerous finding have been reported, though with conflicting

Abbreviations: 5-HT, serotonin; 5-HT2A, serotonin 2A receptor; SNPs, single-nucleotide polymorphisms; LD, linkage disequilibrium; TCI, Temperament and Character Inventory; HA, harm avoidance; NS, novelty seeking; RD, reward dependence; P, persistence; SD, self-directedness; C, cooperativeness; ST, self-transcendence; SD, standard deviation; df, degree of freedom; HWE, Hardy–Weinberg equilibrium. ⁎ Corresponding author. Tel.: +39 051 6584233; fax: +39 051 521030. E-mail address: [email protected] (A. Serretti). 0278-5846/$ - see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.pnpbp.2007.05.008

results (Noblett and Coccaro, 2005; Van Gestel and Van Broeckhoven, 2003). Serotonergic genes and particularly the serotonin transporter have been extensively studied, as serotonin (5-HT) seems involved in specific temperamental traits (Carver and Miller, 2006; Ebstein, 2006; Ebstein et al., 2000; Serretti et al., 2006). Among 5-HT candidates, the 2A receptor (5-HT2A) received considerable attention. It plays an important role in cognitive functions, like learning (Harvey, 2003) and memory (Buhot et al., 2000). It consists of 4 exons separated by 3 introns spanning over 62 kb (Chen et al., 1992) and it is located on 13q14–q21 (Sparkes et al., 1991). In the mouse, homologous recombination knockout genes encoding various proteins involved in 5-HT neurotransmission has recently allowed further assessment of the role of the 5-HT2A receptor in the regulation of sleep (Adrien, 2004). 5-HT2A−/− knockout mice exhibited more wakefulness and less slow wave sleep (SWS) than wild-types. In humans, a large number of studies investigated the association of single-nucleotide polymorphisms (SNPs) in this

1276

A. Serretti et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 31 (2007) 1275–1281

gene with a wide range of phenotypes such as psychiatric disorders, drug response and personality traits. For what concerns psychiatric phenotypes and 5-HT2A polymorphisms, in a recent review, association studies on 5-HT2A gene and its relation with schizophrenia, mood, eating and anxiety disorders, attention-deficit hyperactivity disorders (ADHD), suicide and Alzheimer's disease were considered, with mainly negative or conflicting results plausibly due to the small sample size (Norton and Owen, 2005). The main SNPs analysed were rs6311 (−1438 A/G) and rs6313 (102 T/C). After the publication of this review, other studies were published for each of these associations. A recent meta-analysis, conducted on 73 studies, failed to find significant association of the 5-HT2A rs6313 polymorphism with either schizophrenia or suicidal behavior. Evidence of significant association was only detected between rs6311 and suicidal behavior (Li et al., 2006). For what concerns mood disorders, a study has examined their possible association with the 5-HT2A rs6313 polymorphism and no differences were found in genotype and allele distribution between the mood disordered subjects, with and without suicide attempt history, and controls (Khait et al., 2005). Nevertheless, the relation between 5HT2A and mood disorders is strengthened by the publication of a large number of studies on 5-HT2A and its role such as target of antidepressants (Choi et al., 2005; McMahon et al., 2006). Nevertheless, negative studies were also published (Cusin et al., 2002; McMahon et al., 2006; Sato et al., 2002). For what concerns drug response, the 5-HT2A receptors are also targets for modulation of psychostimulant use (Bubar and Cunningham, 2006), in particular for hallucinogens (Nichols, 2004). The SNPs predominantly analysed in association with personality traits to date were rs6311 and rs6313. These SNPs are in strong linkage disequilibrium (LD) being in the promoter and at the beginning of the gene, respectively. A large quantity of studies was published. In particular, platelet serotonin 2A receptor sites have been found associated with impulsivity and aggression in a Positron Emission Tomography (PET) study (Coccaro et al., 1997). Moreover, rs6311 was found to be associated with impulsive traits in alcohol dependents, measured with the Baratt Impulsiveness Scale (BIS) (Preuss et al., 2001). This was the first report on an association of a 5-HT2A promoter SNP with personality dimensions, which is partially confirmed by more recent reports on the same variant (Ni et al., 2006; Nomura, 2006). Subsequently, 5-HT2A SNPs have been found to be associated with low anxiety-related traits, which are negatively related to impulsive traits (Golimbet et al., 2004; Rybakowski et al., 2006). In addition, TCI Self-Transcendence was related to the rs6311 SNP (Ham et al., 2004). On the other hand, a considerable number of negative results have also been reported regarding the involvement of 5-HT2A in personality traits (Berggard et al., 2003; Blairy et al., 2000; Jonsson et al., 2001; Kusumi et al., 2002; Lochner et al., 2006; Tochigi et al., 2005). The lack of unequivocal results may be due to several factors, such as sample selection for its heterogeneous composition in terms of gender, age and ethnicity, the small number of subjects in each study or the use of different per-

Table 1 5-HT2A SNPs considered in this study SNP ID

Position a

rs643627

46326612 (41430)

rs594242

46356053 (11989)

Distance (bp)

Alleles

Role

A/G

Intron

C/G

Intron

C/T

Coding sequence

C/T

Promoter

29,441 13,426 rs6313

46367941 (101) 1538

rs6311

46369479 (− 1437)

a

Absolute chromosomal position. The relative position to the start codon is given in parenthesis.

sonality scales. Conflicting findings may also be due to the poor gene coverage in the majority of the studies to date. With the aim to better cover the gene, we investigated several 5-HT2A SNPs (rs643627, rs594242, rs6311 and rs6313; Table 1) and their association with personality traits as measured with the TCI. We studied the first three SNPs in two sub-samples of German subjects, composed respectively of normal controls and suicide attempters with a psychiatric diagnosis. Moreover, we also investigated an independent, smaller Italian mood disorder sample for association with the rs6313 SNP, which is in strong LD with rs6311. 2. Materials and methods 2.1. Samples 2.1.1. German sample Healthy volunteers were randomly selected from the city registry (Einwohnermeldeamt) of Munich, Germany, and contacted by mail. To exclude subjects with neuropsychiatric disorders, we conducted further screenings before the volunteers were enrolled in the study. First, subjects who responded were initially screened by phone. Detailed medical and psychiatric histories were assessed by using systematic forms. Second, they were invited to a comprehensive interview including the Structured Clinical Interview for DSM-IV (SCID I and SCID II) (First et al., 1990; First et al., 1995; Wittchen et al., 1997) to evaluate their lifetime Axis I and II disorders. Subjects with relevant somatic diseases or a lifetime history of any Axis I or II disorders or suicidal behavior were excluded. Finally, 289 healthy subjects (males/females: 123/ 166; 42.6% males) were included in the study. Their mean age was 45.2 ± 14.9 years. 26.3% of the subjects had low education level, 30.1% had middle level and 43.6% had high level. The education level was rated as low (secondary school), middle (junior high school) and high (general qualification for university entrance). The patient group consisted of suicide attempters. The attempters were consecutively referred to general psychiatric wards of the Department of Psychiatry, Ludwig Maximilians University of Munich. 111 suicide attempters (males/females: 43/ 68; 38.7% males) were included in the study. Their mean age was 39.2± 13.6 years. 30.6% of the patients had low education level, 36.9% had middle level and 32.4% had high level.


The current and lifetime diagnoses of mental disorders were assessed close to discharge by applying SCID I and II. Patients with mental disorders due to a general medical condition or with dementia were excluded. DSM-IV lifetime diagnoses of mental disorders among the patients were affective spectrum (n = 76; 68.5%), schizophrenia spectrum (n = 17; 15.3%) and borderline personality disorder (n = 18; 16.2%). Written informed consent was obtained from all subjects after a detailed and extensive description of the study. The study was approved by the local ethics committee and carried out in accordance to the ethical standards laid down in the 1964 Declaration of Helsinki. 2.1.2. Italian sample The sample was composed of 60 patients affected by bipolar disorder (n = 46) and recurrent major depression disorder (n = 14) and consecutively admitted to the Department of Neuropsychiatry at the Institute H. San Raffaele. 35.0% were males (males/females: 21/39). The mean age was 44.0 ± 14.8 years. The mean educational level, in terms of years of education, was 11.0 ± 4.2 years. The mean age at onset was 28.5 ± 10.4 years. Among patients, suicide attempters were 16 (28.6%) and subjects with a diagnosis of a personality disorder were 29 (52.7%). The present sample is part of a larger sample that has been published for analyses of other clinical features, such as symptomatology and drug response (Serretti et al., 2001a; Serretti et al., 2002; Serretti et al., 2001b). Lifetime diagnoses were assigned by two independent psychiatrists on the basis of unstructured clinical interviews and medical records, according to DSM-IV criteria following a best estimate procedure (Leckman et al., 1982). The presence of concomitant diagnoses of mental retardation or drug dependence, together with somatic

1277

Table 2 Temperament and character dimensions stratified for 5-HT2A genotypes – German healthy sample rs643627 A/A, mean ± S.D. (n = 138)

A/G, mean ± G/G, mean ± S.D. (n = 119) S.D. (n = 32)

F

df

HA NS RD P SD C ST

46.68 ± 5.08 60.10 ± 5.24 39.01 ± 3.16 11.13 ± 1.93 80.55 ± 4.41 75.11 ± 4.46 42.97 ± 5.76

0.16 2.31 0.29 0.76 1.18 0.008 0.40

2 2 2 2 2 2 2

rs594242 C/C, mean ± C/G, mean ± G/G, mean ± S.D. (n = 173) S.D. (n = 102) S.D. (n = 14)

F

df

HA NS RD P SD C ST

1.69 0.61 0.18 1.00 6.09 1.32 0.54

2 2 2 2 2 2 2

46.57 ± 5.40 60.55 ± 5.31 39.37 ± 3.52 11.40 ± 2.06 79.73 ± 5.53 75.17 ± 3.67 43.47 ± 6.47

46.44 ± 5.61 59.12 ± 5.33 39.03 ± 4.22 11.62 ± 2.01 81.28 ± 4.23 75.19 ± 5.04 44.19 ± 6.66

46.34 ± 5.28 60.65 ± 4.97 39.32 ± 3.42 11.36 ± 1.99 80.50 ± 5.15 74.99 ± 3.95 42.99 ± 5.84

46.72 ± 5.28 59.57 ± 5.36 39.01 ± 3.51 11.17 ± 1.94 80.40 ± 4.14 75.64 ± 4.49 43.77 ± 6.79

48.93 ± 4.04 59.36 ± 7.83 38.79 ± 3.64 11.79 ± 2.49 75.79 ± 6.45 73.57 ± 3.88 44.64 ± 6.03

rs6311 T/T, mean ± S.D. (n = 48)

T/C, mean ± S.D. (n = 82)

C/C, mean ± F S.D. (n = 131)

df

HA NS RD P SD C ST

47.27 ± 4.94 60.10 ± 5.26 39.33 ± 3.70 11.37 ± 2.02 80.16 ± 4.57 74.95 ± 4.68 43.76 ± 6.78

46.40 ± 5.55 59.97 ± 5.26 39.12 ± 3.37 11.25 ± 1.97 79.96 ± 4.90 74.85 ± 4.22 42.82 ± 5.82

2 2 2 2 2 2 2

46.87 ± 5.14 60.92 ± 5.31 39.27 ± 3.50 11.06 ± 1.86 80.94 ± 5.88 75.48 ± 3.68 43.98 ± 5.76

1.25 0.40 0.43 0.28 0.08 0.06 0.45

p 281 281 281 281 281 281 281

0.85 0.10 0.75 0.47 0.31 0.99 0.67

p 281 281 281 281 281 281 281

0.19 0.54 0.84 0.37 0.003 0.27 0.58

p 253 253 253 253 253 253 253

0.29 0.67 0.65 0.76 0.93 0.94 0.64

HA = Harm Avoidance, NS = Novelty Seeking, RD = Reward Dependence, P = Persistence, SD = Self-Directedness, C = Cooperativeness, ST = SelfTranscendence.

or neurological illnesses that impaired psychiatric evaluation represented exclusion criteria. Informed consent was obtained from all probands after the procedure had been fully explained to them; the probands were unrelated and of Italian descent with antecedents from all parts of the country. 2.2. Temperament and Character Inventory (TCI)

Fig. 1. Linkage disequilibrium between the three 5-HT2A rs643627, rs594242 and rs6311 loci. Rs594242 and rs6311 were in significant strong linkage disequilibrium (D′ = 0.84), while D′ scores were very low between rs643627 and both rs594242 (D′ = 0.09) and rs6311 (D′ = 0.02).

The Temperament and Character Inventory (TCI) is the tool to assess individual differences in the basic dimensions of the Cloninger biosocial model of personality (Cloninger et al., 1994). This model was based on the assumption that a part of the individual's personality is heritable. In particular, Cloninger hypothesized that personality is composed both by traits that are heritable and stable throughout life and traits that are influenced by socio-cultural learning and mature throughout life. He defined Temperament the totality of heritable dimensions and Character the totality of non-heritable dimensions. Temperament consists of four traits, so-called Harm Avoidance (HA), Novelty Seeking (NS), Reward Dependence (RD) and Persistence (P). HA denotes the individual's inclination to behavioral inhibition in front of potentially dangerous stimuli

1278


and to anticipate negative effects; NS relates to exploratory behaviours and activation in response to novel stimuli; RD concerns relational and affective skills but also dependent attitudes; finally, P characterizes industrious, hard working and stable despite frustration and fatigue individuals. Character consists of three dimensions: Self-Directedness (SD), Cooperativeness (C) and Self-Transcendence (ST). SD expresses the individual's competence towards autonomy, reliability and maturity; C is related to social skills, like support, collaboration, and partnership; finally, ST denotes the aptitude towards mysticism, religion and idealism. The TCI is a battery of 240 items that has been developed to account for individual differences in both normal and abnormal behavioural patterns. In both samples, the same version of the TCI was administered, with a true–false Likert scale. The Italian version was previously validated (Fossati et al., 2001) as well as the German version (Brandstrom et al., 2003). 2.3. DNA analysis 5-HT2A rs643627, rs594242 and rs6311 were genotyped by Illumina (Illumina, Inc., San Diego, USA) through use of their Integrated BeadArray System. Additionally, 23 SNPs in genes spanning all chromosomes were genotyped as genomic controls (rs2006727; rs586726; rs724529; rs206847; rs1868155; rs2009602; rs1383628; rs876635; rs1367828; rs2076940; rs2025557; rs1993181; rs2168213; rs948184; rs2227973; rs1713449; rs2295152; rs1800404; rs1469122; rs731835; rs418251; rs725493; rs1555048). These genes have been selected based on the low a priori probability of involvement in behavioral traits and phenotypes. No association was observed for personality or diagnostic status. We supplied Illumina with barcoded DNA microliter plates containing the DNA quantified with Pico Green to be at 100 ng/μl and Illumina delivered genotypes with a quality score calculated by proprietary Illumina algorithms. Rs6313 in the Italian sample was genotyped as previously reported (Cusin et al., 2002). 2.4. Statistical analysis Multivariate analysis of covariance (MANCOVA) was used to test possible influence of single SNPs on TCI scores. With the aim of reducing possible sources of variance, we included in all analyses genotype and sex as main factors and education and age as covariates in the MANCOVA model. This was performed because of the influence of such variables on TCI (Cloninger et al., 1994) and unequal distribution in our study sample (Giegling et al., 2006; Rujescu et al., 2003). Haploview 3.2 was used to generate a LD map and to test for Hardy–Weinberg equilibrium (HWE) (Barrett et al., 2005). Tests for associations using multi-marker haplotypes were performed using the statistics software “R” (http://www.Rproject.org), package “haplo.score” to compare TCI scores between haplotypes. Sex, age and education were added as covariates. Permutations (n = 10,000) were performed to estimate the global significance of the results for all haplotype analyses and to validate the expectation-maximization values.

All p values were 2-tailed, and statistical significance was conservatively set at the 0.007 level (0.05 divided by 7 TCI factors). Traditional statistical analyses were performed using “Statistica” package (StatSoft, 1995). With these parameters we had a sufficient power on the German control sample (0.80) to detect a medium effect size (d = 0.45) between two main genotypes, that, as an example, corresponds to 2.3 points on NS scores (TCI) (Cohen, 1988). The other samples had lower power given the reduced sample size (data not shown). 3. Results 3.1. German sample 3.1.1. Hardy–Weinberg equilibrium and linkage disequilibrium for 5-HT2A rs643627, rs594242 and rs6311 SNPs 5-HT2A SNPs were in HWE in the whole sample (rs643627: p = 0.81; rs594242: p = 1.00; rs6311: p = 1.00), as well as separately in controls (rs643627: p = 0.43; rs594242: p = 1.00; rs6311: p = 0.88) and in suicide attempters (rs643627: p = 0.99; rs594242: p = 0.94; rs6311: p = 0.85).

Table 3 Temperament and character dimensions stratified for 5-HT2A genotypes – German suicide sample rs643627 A/A, mean ± S.D. (n = 50)

A/G, mean ± S.D. (n = 50)

G/G, mean ± S.D. (n = 11)

F

df

HA NS RD P SD C ST

58.50 ± 7.59 58.50 ± 5.68 39.60 ± 3.79 11.63 ± 1.92 68.98 ± 9.51 72.46 ± 5.93 47.40 ± 7.94

56.54 ± 9.34 56.82 ± 6.18 40.18 ± 4.40 11.90 ± 2.47 70.60 ± 9.84 70.82 ± 6.72 47.64 ± 8.91

1.17 0.90 0.22 0.41 0.31 0.49 1.99

2 2 2 2 2 2 2

rs594242 C/C, mean ± S.D. (n = 67)

C/G, mean ± S.D. (n = 37)

G/G, mean ± S.D. (n = 7)

F

df

HA NS RD P SD C ST

57.72 ± 8.06 58.12 ± 5.63 39.60 ± 4.06 11.71 ± 2.11 68.65 ± 8.99 71.39 ± 7.05 46.78 ± 7.78

57.05 ± 7.50 59.19 ± 6.36 39.49 ± 3.33 11.84 ± 2.03 70.44 ± 8.54 73.59 ± 4.56 45.24 ± 6.32

59.43 ± 8.83 58.57 ± 6.12 39.00 ± 2.31 11.14 ± 1.46 67.83 ± 10.17 72.14 ± 4.56 43.57 ± 6.55

0.49 0.29 0.06 0.39 0.52 2.43 0.70

2 2 2 2 2 2 2

T/T, mean ± S.D. (n = 14)

T/C, mean ± S.D. (n = 36)

C/C, mean ± S.D. (n = 43)

F

df

57.86 ± 8.45 59.57 ± 5.23 39.50 ± 4.15 12.00 ± 2.52 66.92 ± 10.18 68.86 ± 10.22 47.86 ± 9.09

58.03 ± 7.27 58.89 ± 6.68 38.78 ± 3.66 11.54 ± 2.00 67.68 ± 8.22 71.22 ± 5.15 44.31 ± 6.60

58.65 ± 7.99 57.67 ± 5.66 39.95 ± 3.64 11.72 ± 1.97 68.14 ± 8.08 73.23 ± 5.19 46.84 ± 6.82

0.04 0.93 1.17 2.29 0.04 2.71 1.58

2 2 2 2 2 2 2

56.94 ± 7.88 58.88 ± 6.06 39.30 ± 3.55 11.76 ± 2.10 69.14 ± 8.18 72.18 ± 6.50 44.38 ± 5.80

p 103 103 103 101 99 103 103

0.31 0.41 0.80 0.66 0.73 0.61 0.14

p 103 103 103 101 99 103 103

0.61 0.74 0.94 0.68 0.59 0.09 0.50

rs6311

HA NS RD P SD C ST

p 85 85 85 83 81 85 85

0.96 0.40 0.31 0.11 0.96 0.07 0.21


A. Serretti et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 31 (2007) 1275–1281 Table 4 Temperament and character dimensions stratified for rs6313 genotypes – Italian patient sample rs6313

HA NS RD P SD C ST

T/T, mean ± S.D. (N = 15)

T/C, mean ± S.D. (N = 34)

C/C, mean ± S.D. (N = 11)

F

df

53.80 ± 5.92 62.40 ± 5.82 38.67 ± 2.82 12.00 ± 1.60 71.40 ± 7.31 69.00 ± 6.91 46.36 ± 5.50

55.18 ± 7.74 57.88 ± 4.15 38.79 ± 3.46 11.91 ± 1.78 72.00 ± 9.30 72.97 ± 4.43 46.09 ± 5.95

56.50 ± 8.21 59.45 ± 4.95 38.36 ± 3.07 11.64 ± 2.42 67.09 ± 8.20 70.36 ± 4.63 47.54 ± 8.85

0.41 4.71 0.07 0.13 1.36 3.38 0.21

2 2 2 2 2 2 2

p 56 57 57 57 56 57 56

0.66 0.01 0.93 0.88 0.26 0.04 0.81


In the whole sample, rs594242 and rs6311 were in strong LD (Fig. 1), also considering separately controls and suicide attempters (data not shown), while D′ scores were lower between rs643627 and both rs594242 and rs6311. 3.1.2. 5-HT2A rs643627, rs594242 and rs6311 polymorphisms and personality traits in healthy sample Single SNP MANCOVA results are reported in Table 2. Rs594242 showed an association with SD (main effect: F = 6.09; df = 2,281; p = 0.003). No other association was observed with other TCI factors. Rs643627 and rs6311 were not associated with TCI personality traits. 3.1.3. 5-HT2A rs643627, rs594242 and rs6311 polymorphisms and personality traits in suicide attempters Single SNP MANCOVA results are reported in Table 3. No associations were observed between any of the three polymorphisms and TCI factors. Rs594242 and Rs6311 only showed a trend for association with C (respectively, main effect: F = 2.43; df = 2,103; p = 0.09 and F = 2.71; df = 2,85; p = 0.07). 3.2. Italian sample 3.2.1. 5-HT2A rs6313 SNP and personality dimensions Rs6313 was marginally departing from HWE in the Italian sample (p = 0.042). The SNP was marginally associated with NS (F = 3.42; df = 2,39; p = 0.04; Table 4), with T/T subjects reporting higher scores for NS. None of the other personality dimensions were associated with the investigated SNP. 4. Discussion We investigated the association of 5-HT2A SNPs (5-HT2A rs643627, rs594242, rs6311 and rs6313) with personality traits as measured with the TCI in three independent samples including healthy subjects and patients. We observed mainly negative findings, nevertheless, a marginal association between the rs594242 polymorphism and SD was found in the healthy subject German sub-sample. Interestingly a non-significant similar direction of scores was observed in the Italian sample. From a purely probabilistic calculation, the chance of the

1279

observed finding is about 3 out of 1000, however much caution should be applied given the high number of genetic investigations published in literature. We selected several SNPs to better cover the gene; however, some genetic variance has not been captured as also demonstrated by the low LD between rs643627 and the other markers. In fact, rs643627 is located almost at the 3′ end of the gene in intron 2, and about 29 kbp downstream from rs594242, which is also located in the same intron. Rs6311 is in the promoter region further 13 kbp upstream from rs594242, an area that has been much investigated in previous studies (Blairy et al., 2000; Ni et al., 2006; Tochigi et al., 2005). Finally, rs6313, investigated in the Italian sample only, is located in the coding sequence and it is about 1.5 kbp distant from rs6311. This SNP is in almost complete LD with rs6311 in Caucasians (Ni et al., 2006). Since rs6313 was not in HWE in the Italian sample, it is important to note that it has been observed that approximately 10% of the published studies report HWE departures; reasons may be linked to ethnic stratification of the sample, lack of random matching, spurious associations with undetected conditions, genetic drift, genotype errors, small sample size and chance (Trikalinos et al., 2006; Vogel and Motulsky, 1997). Moreover, HWE departure may bias gene–disease associations; in our sample, we carefully checked for laboratory errors with a double operator reading. As for other findings, rs594242 and rs6311 showed a trend for association with C (p = 0.09 and p = 0.07, respectively). These are novel and marginal results, so replication studies are clearly necessary. Interestingly, rs6313 was marginally associated with NS (p = 0.01) in the Italian mood disorder sample. T/T subjects reported higher scores in NS than the others. However, this result is not in line with findings of two previously published studies. In a healthy subject sample, rs6313 heterozygotes scored lower on the State–Trait Anxiety Expression Inventory (STAI) scale of Anxiety and had a lower score on the Minnesota Multiphasic Personality Inventory (MMPI) scale of Social Introversion, which is correlated with the TCI HA dimension. For what concerns TCI, rs6313 heterozygotes had significantly lower scores in HA than homozygous schizophrenics (Golimbet et al., 2004). In a second study in borderline personality disorder patients and controls, there was no association between the C allele of rs6313 and higher extraversion scores (Ni et al., in press). No other studies on the association of rs6313 with personality traits are published to date, consequently further studies are required. The results of the present investigation are mainly negative, considering the weakness of the p values, nevertheless, it is however possible to hypothesize a link between impulsivity and 5HT2A SNPs, that should be more deeply dissected. In a recent study carried out on the same German sample, C/C homozygotes for the functional rs6311 SNP reported more anger- and aggression-related behaviours. In particular, regarding the State– Trait Anger Expression Inventory (STAXI), C/C homozygotes had higher scores on Trait Anger and Angry Reaction and lower scores on Anger Out subscales and regarding the Questionnaire for Measuring Factors of Aggression (FAF) C-allele carriers

1280


showed less Aggression Inhibition (Giegling et al., 2006). Since we have previously observed (data not published) a positive correlation between NS and the Anger Out subscale and a negative one between NS and Aggression Inhibition, also the rs6311 could be related to NS. Consequently all SNPs analyzed in this study should be considered in future research. The present study has relevant strengths: the number of the polymorphisms considered, two of which scarcely studied in literature; the sample, large and composed of control subjects in the German group. Also the fact that the two samples are made up of subjects from two different ethnic groups is relevant. Actually, ethnic origin is a frequent cause of stratification bias, but our sample was composed of subjects mainly from Germany and the North of Italy with local antecedents for at least two generations. The German population is considered genetically homogeneous (Cavalli Sforza, 1994). North Italy is also characterized by a substantial genetic homogeneity (Barbujani and Sokal, 1991a; Barbujani and Sokal, 1991b; Gasparini et al., 1997). Finally, a collaborative paper presenting 5-HT2C genotypes from different countries reported similar frequencies for Italy and Germany (Lerer et al., 2001). Nevertheless, limitations are also present: the difference in the polymorphisms investigated in Italians and Germans; the dissimilarity of the psychiatric diagnoses between the two samples; the lack of the control sample among Italian subjects; and the difference in terms of gender ratio and mean age among controls and the clinical samples. Moreover, candidate gene studies like the present are highly likely to produce false positive findings (Sullivan, in press). A number of clinical factors influence TCI scores, among them sex and age in particular. In a recent meta-analysis on gender differences in the TCI temperament dimensions, women scored higher in RD and HA. There were no differences in NS or in P (Miettunen et al., 2007). In another work distribution by age and sex of the dimensions of the TCI were assessed cross-culturally for samples in Sweden, Germany and the USA (Brandstrom et al., 2001). Significant effects of age, sex and culture in univariate and multivariate comparisons on the personality dimensions were found. Also genetic analyses revealed gender specific differences (Samochowiec et al., 2004). Consequently, we put gender, age and education as covariants. Another question could be the validity problems arising from the administration of the TCI to schizophrenia spectrum patients. A few studies were published in literature on the temperamental profile of schizophrenic patients (Eklund et al., 2004; Guillem et al., 2002), consequently future investigations are required. In conclusion, we observed a minor effect of 5-HT2A variants on personality scores in both patients and controls; a false-positive finding is possible but further analyses may be worthwhile. References Adrien J. Implication of serotonin in the control of vigilance states as revealed by knockout-mouse studies. J Soc Biol 2004;198:30–6. Barbujani G, Sokal RR. Genetic population structure of Italy. I. Geographic patterns of gene frequencies. Hum Biol 1991a;63:253–72. Barbujani G, Sokal RR. Genetic population structure of Italy: II. Physical and cultural barriers to gene flow. Am J Hum Genet 1991b;48:398–411.

Barrett JC, Fry B, Maller J, Daly MJ. Haploview: analysis and visualization of LD and haplotype maps. Bioinformatics 2005;21:263–5. Berggard C, Damberg M, Longato-Stadler E, Hallman J, Oreland L, Garpenstrand H. The serotonin 2A − 1438 G/A receptor polymorphism in a group of Swedish male criminals. Neurosci Lett 2003;347:196–8. Bienvenu OJ, Stein MB. Personality and anxiety disorders: a review. J Personal Disord 2003;17:139–51. Blairy S, Massat I, Staner L, Le Bon O, Van Gestel S, Van Broeckhoven C, et al. 5-HT2a receptor polymorphism gene in bipolar disorder and harm avoidance personality trait. Am J Med Genet 2000;96:360–4. Brandstrom S, Richter J, Przybeck T. Distributions by age and sex of the dimensions of Temperament and Character Inventory in a cross-cultural perspective among Sweden, Germany, and the USA. Psychol Rep 2001;89:747–58. Brandstrom S, Richter J, Nylander PO. Further development of the temperament and character inventory. Psychol Rep 2003;93:995–1002. Bubar MJ, Cunningham KA. Serotonin 5-HT2A and 5-HT2C receptors as potential targets for modulation of psychostimulant use and dependence. Curr Top Med Chem 2006;6:1971–85. Buhot MC, Martin S, Segu L. Role of serotonin in memory impairment. Ann Med 2000;32:210–21. Carver CS, Miller CJ. Relations of serotonin function to personality: current views and a key methodological issue. Psychiatry Res 2006;144:1–15 [Epub 2006 Aug 17]. Cavalli Sforza L. The history and geography of human genes. Princeton, New Jersey, USA: Princeton University Press; 1994. Chen K, Yang W, Grimsby J, Shih JC. The human 5-HT2 receptor is encoded by a multiple intron–exon gene. Brain Res Mol Brain Res 1992;14:20–6. Choi MJ, Kang RH, Ham BJ, Jeong HY, Lee MS. Serotonin receptor 2A gene polymorphism (− 1438A/G) and short-term treatment response to citalopram. Neuropsychobiology 2005;52:155–62. Cloninger CR, Przybeck TR, Svrakic DM, Wetzel RD. The Temperament and Character Inventory (TCI): a guide to its development and use. St.Louis, Missouri: Center for Psychobiology of Personality Washington University; 1994. Coccaro EF, Kavoussi RJ, Sheline YI, Berman ME, Csernansky JG. Impulsive aggression in personality disorder correlates with platelet 5-HT2A receptor binding. Neuropsychopharmacology 1997;16:211–6. Cohen J. Statistical power analysis for the behavioral sciences. Hillsdale, New Jersey: Lawrence Erlbaum Associates; 1988. p. 8–14. Cusin C, Serretti A, Zanardi R, Lattuada E, Rossini D, Lilli R, et al. Influence of monoamine oxydase A and serotonin receptor 2A polymorphisms in SSRIs antidepressant activity. Int J Neuropsychopharmacol 2002;5:27–35. Ebstein RP. The molecular genetic architecture of human personality: beyond self-report questionnaires. Mol Psychiatry 2006;11:427–45. Ebstein RP, Benjamin J, Belmaker RH. Personality and polymorphisms of genes involved in aminergic neurotransmission. Eur J Pharmacol 2000;410:205–14. Eklund M, Hansson L, Bengtsson-Tops A. The influence of temperament and character on functioning and aspects of psychological health among people with schizophrenia. Eur Psychiatry 2004;19:34–41. First MB, Spitzer RL, Gibbon M, Williams BW, Benjamin L. Structured Clinical Interview for DSM-IV Axis II Personality Disorders (SCID-II). New York State Psychiatric Institute, New York: Biometrics Research Department; 1990. First MB, Spitzer RL, Gibbon M, Williams JB. Structured Clinical Interview for DSM-IVAxis I Disorders – Patient Edition (SCID – I/P, Version 2.0). New York State Psychiatric Institute, New York: Biometrics Research Department; 1995. Fossati A, Donati D, Donini M, Novella L, Bagnato M, Maffei C. Temperament, character, and attachment patterns in borderline personality disorder. J Personal Disord 2001;15:390–402. Gasparini P, Estivill X, Volpini V, Totaro A, Castellvi-Bel S, Govea N, et al. Linkage of DFNB1 to non-syndromic neurosensory autosomal-recessive deafness in Mediterranean families. Eur J Hum Genet 1997;5:83–8. Giegling I, Hartmann AM, Moller HJ, Rujescu D. Anger- and aggression-related traits are associated with polymorphisms in the 5-HT-2A gene. J Affect Disord 2006;96:75–81. Golimbet VE, Alfimova MV, Mitiushina NG. Polymorphism of the serotonin 2A receptor gene (5HTR2A) and personality traits. Mol Biol (Mosk) 2004;38:404–12.

A. Serretti et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 31 (2007) 1275–1281 Grucza RA, Przybeck TR, Cloninger CR. Personality as a mediator of demographic risk factors for suicide attempts in a community sample. Compr Psychiatry 2005;46:214–22. Guillem F, Bicu M, Semkovska M, Debruille JB. The dimensional symptom structure of schizophrenia and its association with temperament and character. Schizophr Res 2002;56:137–47. Ham BJ, Kim YH, Choi MJ, Cha JH, Choi YK, Lee MS. Serotonergic genes and personality traits in the Korean population. Neurosci Lett 2004;354:2–5. Harvey JA. Role of the serotonin 5-HT(2A) receptor in learning. Learn Mem 2003;10:355–62. Jonsson EG, Nothen MM, Gustavsson JP, Berggard C, Bunzel R, Forslund K, et al. No association between serotonin 2A receptor gene variants and personality traits. Psychiatr Genet 2001;11:11–7. Khait VD, Huang YY, Zalsman G, Oquendo MA, Brent DA, Harkavy-Friedman JM, et al. Association of serotonin 5-HT2A receptor binding and the T102C polymorphism in depressed and healthy Caucasian subjects. Neuropsychopharmacology 2005;30:166–72. Klein, M., Kupfer, D., Shea, T. 1993. Personality and depression. In: The Guilford Press NY (ed). Kusumi I, Suzuki K, Sasaki Y, Kameda K, Sasaki T, Koyama T. Serotonin 5-HT (2A) receptor gene polymorphism, 5-HT(2A) receptor function and personality traits in healthy subjects: a negative study. J Affect Disord 2002;68:235–41. Leckman JF, Sholomskas D, Thompson WD, Belanger A, Weissman MM. Best estimate of lifetime psychiatric diagnosis: a methodological study. Arch Gen Psychiatry 1982;39:879–83. Lerer B, Macciardi F, Segman RH, Adolfsson R, Blackwood D, Blairy S, et al. Variability of 5-HT2C receptor Cys23Ser polymorphism among European populations and vulnerability to affective disorder. Mol Psychiatry 2001;6:579–85. Li D, Duan Y, He L. Association study of serotonin 2A receptor (5-HT2A) gene with schizophrenia and suicidal behavior using systematic meta-analysis. Biochem Biophys Res Commun 2006;340:1006–15 [Epub 2006 Jan 4]. Lochner C, Hemmings S, Seedat S, Kinnear C, Schoeman R, Annerbrink K, et al. Genetics and personality traits in patients with social anxiety disorder: a case–control study in South Africa. Eur Neuropsychopharmacol 2006. McMahon FJ, Buervenich S, Charney D, Lipsky R, Rush AJ, Wilson AF, et al. Variation in the gene encoding the serotonin 2A receptor is associated with outcome of antidepressant treatment. Am J Hum Genet 2006;78:804–14 [Epub 2006 Mar 20]. Miettunen J, Veijola J, Lauronen E, Kantojarvi L, Joukamaa M. Sex differences in Cloninger's temperament dimensions—a meta-analysis. Compr Psychiatry 2007;48:161–9 [Epub 2007 Jan 2]. Ni X, Bismil R, Chan K, Sicard T, Bulgin N, McMain S, et al. Serotonin 2A receptor gene is associated with personality traits, but not to disorder, in patients with borderline personality disorder. Neurosci Lett 2006;408:214–9. Ni, X., Bismil, R., Chan, K., Sicard, T., Bulgin, N., McMain, S., Kennedy, J.L., In press. Serotonin 2A receptor gene is associated with personality traits, but not to disorder, in patients with borderline personality disorder. Nichols DE. Hallucinogens. Pharmacol Ther 2004;101:131–81. Noblett KL, Coccaro EF. Molecular genetics of personality. Curr Psychiatry Rep 2005;7:73–80. Nomura M. Involvement of a polymorphism in the 5-HT2A receptor gene in impulsive behavior. Nippon Yakurigaku Zasshi 2006;127:9–13.

1281

Norton N, Owen MJ. HTR2A: association and expression studies in neuropsychiatric genetics. Ann Med 2005;37:121–9. Preuss UW, Koller G, Bondy B, Bahlmann M, Soyka M. Impulsive traits and 5-HT2A receptor promoter polymorphism in alcohol dependents: possible association but no influence of personality disorders. Neuropsychobiology 2001;43:186–91. Rujescu D, Giegling I, Gietl A, Hartmann AM, Moller HJ. A functional single nucleotide polymorphism (V158M) in the COMT gene is associated with aggressive personality traits. Biol Psychiatry 2003;54:34–9. Rybakowski F, Slopien A, Dmitrzak-Weglarz M, Czerski P, Rajewski A, Hauser J. The 5-HT2A−1438 A/G and 5-HTTLPR polymorphisms and personality dimensions in adolescent anorexia nervosa: association study. Neuropsychobiology 2006;53:33–9 [Epub 2006 Jan 4]. Samochowiec J, Syrek S, Michal P, Ryzewska-Wodecka A, Samochowiec A, Horodnicki J, et al. Polymorphisms in the serotonin transporter and monoamine oxidase A genes and their relationship to personality traits measured by the Temperament and Character Inventory and NEO Five-Factor Inventory in healthy volunteers. Neuropsychobiology 2004;50:174–81. Sato K, Yoshida K, Takahashi H, Ito K, Kamata M, Higuchi H, et al. Association between − 1438G/A promoter polymorphism in the 5-HT(2A) receptor gene and fluvoxamine response in Japanese patients with major depressive disorder. Neuropsychobiology 2002;46:136–40. Serretti A, Lilli R, Lorenzi C, Lattuada E, Cusin C, Smeraldi E. Tryptophan hydroxylase gene and major psychoses. Psychiatric Research 2001a;103:79–86. Serretti A, Lilli R, Lorenzi C, Lattuada E, Smeraldi E. DRD4 exon 3 variants associated with delusional symptomatology in major psychoses: a study on 2011 affected subjects. Am J Med Genet 2001b;105:283–90. Serretti A, Lilli R, Lorenzi C, Lattuada E, Cusin C, Smeraldi E. Serotonin transporter gene (5-httlpr) and major psychoses. Mol Psychiatry 2002;7:95–9. Serretti A, Mandelli L, Lorenzi C, Landoni S, Calati R, Insacco C, et al. Temperament and character in mood disorders: influence of DRD4, SERTPR, TPH and MAO-A polymorphisms. Neuropsychobiology 2006;53:9–16. Sparkes RS, Lan N, Klisak I, Mohandas T, Diep A, Kojis T, et al. Assignment of a serotonin 5HT-2 receptor gene (HTR2) to human chromosome 13q14–q21 and mouse chromosome 14. Genomics 1991;9:461–5. StatSoft, I. 1995. STATISTICA per Windows. StatSoft Italia srl. Sullivan, P.F. in press. Spurious genetic associations. Biol Psychiatry – Epub ahead of print. Tochigi M, Umekage T, Kato C, Marui T, Otowa T, Hibino H, et al. Serotonin 2A receptor gene polymorphism and personality traits: no evidence for significant association. Psychiatr Genet 2005;15:67–9. Trikalinos TA, Salanti G, Khoury MJ, Ioannidis JPA. Impact of violations and deviations in Hardy–Weinberg Equilibrium on postulated gene–disease associations. Am J Epidemiol 2006;163:300–9. Van Gestel S, Van Broeckhoven C. Genetics of personality: are we making progress? Mol Psychiatry 2003;8:840–52. Velting DM. Suicidal ideation and the five-factor model of personality. Pers Individ Differ 1999;27:943–52. Vogel F, Motulsky AG. Human genetics: problems and approaches. Berlin: Springer Verlag; 1997. Wittchen H, Zaudig M, Fydrich T. Strukturiertes Klinisches Interview für DSMIV. Achse I und II. Göttingen: Hogrefe; 1997.