This is an ongoing assignment and will be worked on each week. This is what is required this week. I also picked to focus on Borderline Personality for my topic.Select the topic for your Critical Revi

The world journal of biological psychiatry , 2010 ; 11: 45 –58

Correspondence: André Tadic´, MD, Department of Psychiatry, University of Mainz, Untere Zahlbacher Str. 8, 55131 Mainz, Germany. Tel: +49 6131 17 3950. Fax: +49 6131 17 3459. E-mail: [email protected] e

(Received 19 june 2009 ; accepted 01 october 2009)

Association analysis between gene variants of the tyrosine

hydroxylase and the serotonin transporter in borderline personality

disorder

ANDRÉ TADIC ´1*, AMELIE ELSÄSSER 2, NIELS STORM 3, ULRIKE BAADE 3, STEFANIE

WAGNER 1, ÖMÜR BAS ¸KAYA 1, KLAUS LIEB 1 & NORBERT DAHMEN 1

1Department of psychiatr y and psychotherapy, University Medical Center Mainz, 2Institute for Medical biometr y, Epidemiology

and Informatics, University Medical Center Mainz, Mainz, Germany, and 3NS and Ub: bioglobe GmpH, Hamburg, Germany

Abstractobjectives. For patients with borderline personality disorder (BPD), we previously reported an independent effect of the catechol- o-methyl-transferase (COMT) low-activity (Met 158 ) allele and an interaction with the low-expression allele of the deletion/insertion (short/long or S/L, resp.) polymorphism in the serotonin transporter-linked promoter region (5-HT - TLPR). The purpose of the present study was to extend these findings to the tyrosine hydroxylase (TH) Val 81Met single nucleotide polymorphism (SNP), the 5-HTTLPR S/L polymorphism incorporating the recently described functional A/G SNP within the long allele of the 5-HTTLPR (rs25531) as well as the variable number of tandem repeat (VNTR) poly - morphism within intron 2 of the serotonin transporter gene (STin2). Methods. In 156 Caucasian BPD patients and 152 healthy controls, we tested for association between BPD and the TH Val 81Met SNP, the 5-HTTLPR/rs25531 polymor - phism, the STin2, the interaction of the TH Val 81Met SNP with the tri-allelic 5-HTTLPR/rs25531, the interaction of the TH Val 81Met SNP with STin2. Results. Between BPD patients and controls, we observed a slight over-representation of the TH Met 81Met genotype in BPD patients compared to controls, but no statistically significant differences in genotype distribution of the individual markers after adjusting for multiple testing. Logistic regression analysis showed a lack of interaction between the TH Val 81Met and the 5-HTTLPR/rs25531 as well as between the TH Val 81Met and the STin2 polymorphism. Conclusions. These data do not suggest independent or interactive effects of the TH Val 81Met, the 5-HT - TLPR/rs25531, or the STin2 polymorphisms in BPD.

Key words: borderline personality disorder, genetics, 5-HTT, tyrosine-hydroxylase, case-control

However, reported associations between BPD and

gene variants of the monoamino-oxidase A (Ni et al.

2007), the serotonin receptor 2A (Ni et al. 2006a),

the tryptophane hydroxylase 1 (Wilson et al. 2009;

Zaboli et al. 2006) have not yet been replicated. Ni

et al. (2006b) further reported on an association

analysis between the serotonin transporter promoter

and BPD. No significant differences in allele or gen -

otype distributions of 5-HTTLPR S/L and rs25531

were detected between BPD patients and controls,

but BPD patients showed a higher frequency of the

low expression 10-repeat allele of the VNTR in intron

2 and a higher frequency of the S-10 haplotype.

However, this latter finding could not be replicated

in a small sample of Spanish BPD patients (Pascual

et al. 2008). Analyses from our laboratory in a

German sample of 161 patients and 157 healthy con -

trols revealed an independent effect of the

Introduction

Borderline personality disorder (BPD) is a severe

mental disorder with a high prevalence affecting

about 1–6% of the general population (Grant et al.

2008; Lenzenweger et al. 2007; Torgersen et al.

2001). The disease has a high mortality rate as a

result of suicide and impulsive behaviour; up to 10%

of patients commit suicide (American Psychiatric

Association 2001). Its polymorphic symptomatology

with affective instability, impulsive aggression, self-

injurious behaviour, disturbances in identity and

interpersonal relations has been related to dysfunc -

tions of neurotransmitter systems including sero -

tonin (5-HT), dopamine (DA), and norepinephrine

(NE) (Gurvits et al. 2000; Skodol et al. 2002; Friedel

2004; Lieb et al. 2004).

Genetic studies in BPD patients have focused on

genes involved in these neurotransmitter systems. 46 A. Tadi´c et al.

transporter (5-HTT) expression levels, resulting in

lower serotonin uptake activity when compared with

the L variant (Lesch et al. 1996). Recently, a common

A/G SNP within the 'L G' allele of the 5-HTTLPR

polymorphism has been described (Hu et al. 2006).

The L G allele reduces 5-HTT messenger RNA

expression to levels nearly equivalent to those of the

S allele, whereas the L A allele confers higher 5-HTT

expression (Hu et al. 2006). The variable number of

17 bp tandem repeat polymorphism in intron 2 of the

5-HTT gene (STin2) was found to be in linkage dis -

equilibrium with 5-HTTLPR in different popula -

tions (Ebstein 2006). This gene variant may function

as a transcriptional enhancer (Lesch et al. 1994).

Transgenic mice with the VNTR enhancer region

fused to a reporter gene showed higher expression of

the 12-repeat in the dorsal raphe nucleus during

embryonic development when compared to the

10-repeat (MacKenzie and Quinn 1999). This VNTR

was also associated with different levels of reporter

gene expression in differentiating embryonic stem

cells. The 12 repeat acted as a stronger enhancer than

the 10 repeat (Fiskerstrand et al. 1999). Individual

repeat elements within the VNTR domain also dif -

fered in their enhancer activity in an embryonic stem

cell model (Lovejoy et al. 1998). These observations

suggest a role for the VNTR during brain develop -

ment and adult synaptic plasticity.

Based on the above-mentioned evidences, gene

variants of TH and 5-HTT are strong candidates for

genetic association studies on BPD. To the best of

our knowledge, there are no reports on association

studies between TH gene variants and BPD. The

purpose of the present study was to extent our previ -

ous results in multiple ways. We tested for association

between BPD and (i) the tyrosine hydroxylase Val - 81Met single nucleotide polymorphism (SNP), (ii)

the 5-HTTLPR S/L polymorphism incorporating the

recently described functional A/G SNP within

the long allele of the 5-HTTLPR (rs25531), (iii) the

variable number of tandem repeat (VNTR) poly -

morphism within intron 2 of the serotonin trans -

porter gene (STin2), (iv) the interaction of the TH

Val 81Met SNP with the complex 5-HTTLPR/

rs25531 polymorphism, and (v) the interaction of

the TH Val 81Met SNP with STin2.

Materials and methods

Subjects

All participants gave their written informed consent

to participate in the study after a complete and

extensive description. The study was approved by the

local ethical committee of the Landesärztekammer

Rheinland-Pfalz and was compliant to the Code of

catechol- o-methyl-transferase (COMT) low-activity

(Met 158 ) allele and an interaction with the low-

expression S allele of the 5-HTTLPR polymorphism

in BPD: 5-HTTLPR S carriers had an increasing

risk for BPD with the number of COMT Met 158

alleles (Tadic´ et al. 2009). These data suggested (i)

an interaction between the noradrenergic/dopamin -

ergic and the serotonergic neurotransmission in the

aetiology of BPD and (ii) the usefulness of analyses

of gene–gene interactions in BPD, given the extraor -

dinary heterogeneous symptomatology of BPD,

which suggests the involvement of multiple genes

with independent and interacting effects.

In NE and DA biosynthesis, tyrosine hydroxylase

(TH) catalyzes the rate limiting step, converting

l-tyrosine to l-DOPA (3,4 dihydroxy- l-phenylalanine).

TH gene is located on chromosome 11p15, contains

13 exons and spans approximately 8 kb. It encodes

for four alternatively spliced transcripts, two of which

(both lacking exon 2) are expressed in the brain

(O’Malley et al. 1987; Persson et al. 1997). Immu -

nohistochemical evidence for differences in TH

expression in the locus coeruleus of suicide victims

compared to controls has been presented, and genetic

differences have been postulated as an explanation

(Souery et al. 2003). However, the reported findings

are conflicting, showing decreased (Biegon and Fiel -

dust 1992) and increased (Ordway et al. 1994)

expression levels. The TH Val 81Met polymorphism

(Lüdecke and Bartholomé 1995) is the most fre -

quently investigated coding sequence variant. Codon

81 in exon 3 is altered from GTG to ATG, resulting

in an amino acid exchange from valine to methion -

ine. The polymorphism is located in the regulatory

domain of the tetrameric enzyme. Exon 3 is not pres -

ent in all known splicing variants of the gene (Dumas

et al. 1996). The catalytic domain does preserve its

function when expressed alone, and the regulatory

domain seems to have an inhibiting effect on enzyme

function (Kumer and Vrana 1996).

Serotonergic neurotransmission is regulated by

the 5-HT transporter (5-HTT). 5-HTT is believed

to play an important role in suicide (Anguelova et al.

2003; Li and He 2007), aggression (Cadoret et al.

2003; Retz et al. 2004; Beitchman et al. 2006), impul -

sive behaviour (Coccaro et al. 1996; Frankle

et al. 2005), and emotional lability (Collier et al.

1996; Hoefgen et al. 2005), i.e. clinical features that

occur simultaneously in BPD. Regulation of 5-HTT

expression is under complex control involving mul -

tiple genes. A functional 22–23-bp imperfect repeat

polymorphism in the promoter region (5-HTTLPR)

of the serotonin transporter gene (SLC6A4) creates

a short (S) and a long (L) allele (14- and 16-repeat

alleles) and alters promoter activity. The S variant has

been reported to be associated with lower serotonin TH Val 81Met, 5-HTTLpR and STin2 polymorphisms in bpD 47

and MALDI-TOF mass spectrometry. This method

allows the simultaneous determination of the 43-bp

insertion/deletion variant together with the typing of

SNPs rs25531. PCR primers were generated with

MassARRAY Assay Design software version 3.1. The

region of interest was PCR amplified using a primer

pair, where the forward primer is tagged with T7

RNA polymerase promoter sequence. After removal

of remaining dNTPs with SAP and heat inactivation

of the enzyme, the PCR product was divided in two

halves and MassCLEAVE cocktail (Sequenom) was

directly added for T- or C-specific cleavage. Reaction

products were conditioned with CLEAN resin

(Sequenom) and dispensed onto a SpectroChip with

a MassARRAY nanodispenser. The chips were

scanned using a mass spectrometry workstation

(MassARRAY compact analyzer, Sequenom). Anal -

ysis of resulting spectra was aided by SNP Discovery

Analyzer software version 1.2. SNP-genotypes were

operator determined by identification of

allele-identifying signals and correlating their signal

heights to constant reference signals for determina -

tion of the S/L genotype.

The determination of VNTRs was achieved by

electrophoretic size analysis of PCR products (Wang

et al. 1955; Saiki et al. 1988; Edwards et al. 1989).

For the analysis of STin2 repeat variants, samples

were PCR amplified. The forward primer was fluo -

rescently labelled with 5'-FAM. Fragment analysis

was done with MegaBACE ET550-R size standard

by capillary-electrophoresis on a MegaBACE 1000

DNA Analyzer. Data were analyzed with Softgenet -

ics Genemarker 1.5.

The detailed description of applied protocols can

be reviewed in the on-line supplemental.

Statistics

In the whole study sample and for cases and controls

separately, deviations from HWE were tested by com -

paring the observed and expected genotype frequen -

cies by exact test for HWE; alleles of the 5-HTTLPR

S/L polymorphism and the SNP rs25531 were treated

as two independent bi-allelic loci; for STin2, geno -

types including the 10- or the 12-repeat allele, but

not the 9-repeat allele were used for HWE testing

(Wendland et al. 2007). Regarding the TH Val 81Met

SNP, differences of genotype distribution between

cases and controls were analyzed with the Armitage-

Trend-test, allele distribution differences were ana -

lyzed by 2-tests. 2-Tests were also used to test for

differences between cases and controls in genotype

and allele distributions of the complex 5-HTTLPR/

rs25531 and the tri-allelic VNTR in Intron 2 of the

5-HTT gene (STin2). For interaction analysis logis -

tic regression models with the outcome variable

Ethics of the World Medical Association (Declaration

of Helsinki). One hundred and sixty-one patients of

Caucasian descent with DSM-IV BPD (American

Psychiatric Association 1994) were recruited in the

region of Mainz, Germany. The German version of

the Structured Clinical Interview for DSM-IV axis

II disorders (SCID-II) (Wittchen et al. 1996) was

applied by trained research assistants to each

patient. For the evaluation of DSM-IV axis I disor -

ders, the computerized DSM-IV version of the

Munich-Composite International Diagnostic Inter -

view (M-CIDI) (Wittchen et al. 1998) was applied.

Patients were not included if one or more of the fol -

lowing criteria were present: (i) dementia or any

other organic brain syndrome with obvious cognitive

impairment, (ii) doubts in the patients’ ability to give

informed consent (e.g., patients suffering from

depression with psychotic symptoms), (iii) a current

diagnosis of a psychotic disorder, (iv) the patient was

involved in a legal procedure at the time of study

participation, and (v) the patient was not of Caucasian

descent. The control group ( n156) consisted of

healthy volunteers. They were recruited by board

notices at the University of Mainz and were thoroughly

assessed by means of detailed medical and psychiatric

history forms. Moreover, both SCID-II and the com -

puterized M-CIDI interview were performed. Volun -

teers with a positive history or a current major

psychiatric disorder were not included in the study.

Genotyping

DNA was obtained from venous blood using stan -

dard techniques. The TH Val 81Met SNP (rs6356) was

genotyped by the analysis of primer extension prod -

ucts generated from amplified genomic DNA using a

Sequenom (San Diego, CA) chip-based MALDI-

TOF mass spectrometry platform (Storm et al. 2003;

Jurinke et al. 2004; Ragoussis et al. 2006) . In brief,

MassARRAY Assay Design software version 3.1 was

applied for system specific assay design. After the

PCR reaction, unincorporated nucleotides were

digested with shrimp alkaline phosphatase (SAP,

Sequenom) to prevent interference in the following

MassEXTEND reaction. The primer extension prod -

ucts were conditioned with CLEAN resin (Seque -

nom) and dispensed onto a SpectroChip with a

MassARRAY nanodispenser. The chips were scanned

using a mass spectrometry workstation (MassARRAY

compact analyzer, Sequenom) and the resulting spec -

tra were analyzed and genotypes were determined

using the MassARRAY Typer software version 3.4.

The 5-HTTLPR-S/L variant (rs4795541), in

which the L-allele additionally harbours SNP site

rs25531, was analyzed with MassCLEAVE biochem -

istry (Stanssens et al. 2004) (Sequenom, San Diego) 48 A. Tadi´c et al.

had to be excluded, resulting in 156 patients (48

males and 108 females, mean age  SD  33.5 

9.4 years; median 34) and 152 controls (46 males

and 106 females, mean age  SD  33.6  9.6

years, median 31) eligible for this analysis. For cases

and controls the distributions of the genotype fre -

quencies of the analyzed polymorphisms did not

show any divergence from Hardy–Weinberg equilib -

rium ( p  0.4 for each analysis).

Genotype distribution of the TH Val 81Met poly -

morphism is shown in Table I. There was a slight

over-representation of the TH Met 158 Met genotype

in BPD patients compared to healthy controls. How -

ever, after adjusting for multiple testing we could not

observe a significant association between BPD and

the TH Val 81Met polymorphism ( p  0.048, adjusted

p  0.144).

Distribution of the combined genotype of the

5-HTTLPR/rs25531 polymorphism is shown in

Table II. We detected no association with BPD for

the genotype distribution ( p  0.822, adjusted

p  1.000).

Genotype and allele distribution of the STin2

VNTR is shown in Table III. We detected no asso -

ciation with BPD for the genotype distribution

(p  0.475, adjusted p  1.000).

The logistic regression analysis with “BPD patient”

as outcome variable revealed no significant interaction

between the TH Val 81Met polymorphism and the tri-

allelic 5-HTTLPR/rs25531 genotype ( p  0.698)

(model not presented in detail). Because of small

sample sizes in some genotype groups of the

5-HTTLPR/rs25531 an additional logistic regres -

sion was calculated using binary variables (presence

vs. absence of the 5-HTTLPR/rs25531 ‘S’ or ‘L G’

allele instead of genotypes. No significant interaction

effect of the 5-HTTLPR/rs25531 ‘L G’ or ‘S’ allele

with TH Val 81Met polymorphism could be found

(p  0.428; see Table IV). For the evaluation

of potential allele-dose effects a regression model

including the TH genotype as a continuous covariate,

‘‘BPD patient’’ (yes vs. no) were calculated including

genotype variables and two-way interaction terms.

Dummy-coding was used for categorical variables.

For the interaction analysis of TH Val 81Met and

5-HTTLPR/rs25531 polymorphisms a logistic

regression model for the outcome ‘‘BPD patient’’

(yes vs. no) genotypes was performed; both polymor -

phisms were included as categorical variables as well

as their two-way interaction term. In a further mul -

tiple logistic regression model for the outcome ‘‘BPD

patient’’ the genotypes of the TH Val 81Met SNP as

categorical variable and a binary coded variable for

the tri-allelic 5-HTTLPR/rs25531 ‘S’ or ‘L G’ carri -

ers (present vs. absent) as well as the two-way inter -

action term of these two variables were included. The

binary coding of the tri-allelic 5-HTTLPR/rs25531

and the STin2 VNTR polymorphisms in the logistic

regression analyses was done because of small sam -

ple sizes in some of the genotype groups. For the

evaluation of allele-dose effects, a modification of the

last model was calculated: in this model the TH

genotype was included as a continuous covariate by

coding the number of Met alleles with 0 (for ValVal),

1 (for ValMet), and 2 (for MetMet).

For the interaction analysis of TH and STin2 again

a multiple logistic regression model for the outcome

‘‘BPD patient’’, including both polymorphisms as cat -

egorical variables plus their two-way interaction term,

was performed. In a further analysis a multiple logis -

tic regression model for the outcome ‘BPD patient’

including the genotypes of the TH Val 81Met SNP as

a categorical variable and the STin2 polymorphism as

binary coded variable for VNTR ‘9 or 10-repeat’carriers

(present vs. absent) as well as their two-way interac -

tion term was computed. Furthermore, a modified

multiple logistic regression model including the TH

genotype as a continuous covariate by coding the

number of Met alleles with 0 (for ValVal), 1 (for Val -

Met), and 2 (for MetMet) was calculated.

For the primary hypotheses of association between

genotypes of individual polymorphisms and BPD,

the global significance level was set to 5% and a Bon -

ferroni correction was used to adjust for the three

tests. Because of their exploratory character, the

results of all other comparisons cannot be considered

as significant at any level and p values were only

calculated for the purpose of description. The analy -

ses were performed using SAS V9.1 and SPSS 15, as

well as R version 2.6.1 (R Foundation for Statistical

Computing, Vienna, Austria).

Results

A total of 161 patients meeting DSM-IV criteria for

BPD and 156 healthy volunteers were recruited. Due

to genotyping failures, five patients and four controls

Table I. Genotype and allele distribution of the TH Val 81Met polymorphism in BPD patients and healthy controls.

BPDN (%) ControlsN (%) P value1 (adj. p value) 2

GenotypesA–A (Met 81Met) 21 (14) 15 (10) 0.048 1 (0.144) 2

A–G (Met 81Val) 80 (51) 66 (43) G–G (Val 81Val) 55 (35) 71 (47)

AllelesA (Met 81) 122 (39) 96 (32) 0.062 3

G (Val 81) 190 (61) 208 (68)

1p value is given for case–control comparison (Armitage trend-test).2Adjusted p value following Bonferroni-correction. 3p values are given for case–control comparison ( χ2-test). This p value is descriptive and cannot be considered significant at any level. TH Val 81Met, 5-HTTLpR and STin2 polymorphisms in bpD 49

behavioural dimensions of BPD is based on several

lines of evidence. These include the efficacy of tradi -

tional and atypical antipsychotics in the treatment of

BPD (American Psychiatric Association 2001;

Herpertz et al. 2007) because of their DA receptor

inhibiting capacity as well as human and animal

studies implicating DA activity in emotion informa -

tion processing, impulse control, and cognition

(Friedel 2004). Furthermore, suicidal behaviour,

which is a clinical feature of BPD, has repeatedly

been associated to components of the dopaminergic

system (Oquendo and Mann 2000; Rujescu et al.

2003; Ryding et al. 2006; Giegling et al. 2008). Addi -

tionally, we (Tadic´ et al. 2009) and others (Ni et al.

2007) have reported on associations between gene

variants involved in the metabolism of NE and DA

and BPD. In the herein presented analysis, the TH

representing the number of Met alleles, was calcu -

lated. There was no interaction effect of the TH Val - 81Met polymorphism and the 5-HTTLPR/rs25531

‘LG’ or ‘S’ allele ( p  0.234; see Table V).

There was also no significant interaction between

the TH Val 81Met polymorphism and the STin2

VNTR ( p  0.618) (model not presented in detail).

The additional logistic regression using binary vari -

ables (presence vs. absence of the ‘9’ or ‘10’-repeat

allele) revealed no significant interaction effect ( p 

0.135; see Table VI). The additional regression model

for the evaluation of potential allele-dose effects

revealed no interaction effect of the TH Val 81Met

polymorphism and the STin2 VNTR ‘9’ or

‘10’-repeat allele ( p  0.401; see Table VII).

Discussion

The involvement of an altered dopaminergic and/or

noradrenergic neurotransmission in the aetiology of

Table II. Genotype and allele frequencies of the complex 5-HTTLPR/rs25531 polymorphism in BPD patients and controls.

BPDN (%) ControlsN (%) p value 1 (adj. p value) 2

Genotypes

SS 25 (16) 26 (17) 0.822 1 (1.000) 2

S–L G 8 (5) 10 (6) S–L A 67 (43) 71 (47) LA–LG 13 (8) 11 (7) LA–LA 43 (28) 34 (22)

AllelesS 125 (40) 133 (44) 0.621 3

LG 21 (7) 21 (7) LA 166 (53) 150 (49)

1,3p value is given for case–control comparison ( χ2-test). 2Adjusted p value following Bonferroni correction. 3This p value is descriptive and cannot be considered significant at any level.

Table III. Genotype and allele frequencies of the VNTR in intron 2 of the HTT gene (STin2) in BPD patients and controls.

BPDN (%) ControlsN (%) p value 1 (adj. p value) 2

Genotypes9–10 6 (4) 3 (2) 0.475 1 (1.000) 2

9–12 2 (1) 3 (2) 10–10 25 (16) 17 (11) 10–12 76 (49) 73 (48) 12–12 47 (30) 56 (37)

Alleles9 8 (3) 6 (2) 0.235 3

10 132 (42) 110 (36) 12 172 (55) 188 (62)

1,3p value is given for case–control comparison ( χ2-test). 2Adjusted p value following Bonferroni-correction. 3This p value is descriptive only and cannot be considered significant at any level.

Table IV . Results of the multiple logistic regression analysis with disease status (BPD patient yes vs. no) as outcome using the TH Val 81Met genotype and a binary coded variable for 5-HTTLPR/ rs25531 ‘L G’ or ‘S’ carriers, as well as the interaction of these two variables.

Regression coefficient p value 1Exp(B) OR

5-HTTLPR/rs25531 ‘L G’ or ‘S’ –0.691 0.103 0.501

TH genotype 0.969 TH genotype Val 81Met –0.035 0.942 0.966 TH genotype Met 81Met –0.268 0.801 0.765 5-HTTLPR/rs25531 ‘L G’ or ‘S’ × TH genotype 0.428

5-HTTLPR/rs25531 ‘L G’ or ‘S’ × TH genotype Val 81Met

0.633 0.258 1.883

5-HTTLPR/rs25531 ‘L G’ or ‘S’ × TH genotype Met 81Met

1.071 0.349 2.917

Intercept 0.268 0.467 1.308

For TH genotype, the genotype Val 81Val was used as reference group.1Wald test used. All p values are descriptive and cannot be considered significant at any level.

Table V. Results of the multiple logistic regression analysis with disease status (BPD patient yes vs. no) as outcome using the TH Val 81Met genotype as a continuous covariate with values 0, 1, 2 representing the number of Met alleles and a binary coded variable for 5-HTTLPR/rs25531 ‘L G’ or ‘S’ carriers, as well as the interaction of these two variables.

Regression coefficient p value 1 Exp(B) OR

5-HTTLPR/rs25531 ‘L G’ or ‘S’ –0.657 0.102 0.518

TH genotype –0.077 0.847 0.926 5-HTTLPR/rs25531 ‘L G’ or ‘S’ × TH genotype (number of Met-alleles)

0.531 0.234 1.701

Intercept 0.286 0.416 1.331

1Wald test used. All p values are descriptive and cannot be

considered significant at any level. 50 A. Tadi´c et al.

the same result has been reported for a Canadian

sample of BPD patients (Ni et al. 2006b). Regarding

the STin2, the same results has been reported for a

Canadian (Ni et al. 2006b) and a Spanish sample of

BPD patients (Pascual et al. 2008). Taken together,

these data suggest that the 5-HTTLPR S/L polymor -

phism has no independent effects on BPD. Previous

studies in BPD (Ni et al. 2006b; Tadic´ et al. 2009),

on the intermediate endophenotype “central process -

ing of aversive stimuli” (Smolka et al. 2007) as well as

the herein reported results support the idea of modu -

lating effects of the 5-HTTLPR on the effects of other

genetic variants on BPD. Furthermore, these modu -

lating effects are likely not restricted to gene variants

but also affect the consequences of environmental

influences like serious life events on particular behav -

ioural dimensions of BPD (Wagner et al. 2009). This

principle has been already documented for depression

(Caspi et al. 2003), suicide (Roy et al. 2007), sensa -

tion seeking and insecure attachment (Steiger et al.

2007), i.e. clinical features often occurring in BPD.

Based on our previous finding of a significant inter -

action between the COMT Met 158 allele and the

5-HTTLPR S allele, we investigated the interaction

between the 5-HTTLPR (extended and refined by the

incorporation of the SNP rs25531), the STin2 and the

TH Val 81Met polymorphism. The interaction analysis

of the TH Val 81Met and the 5-HTTLPR/rs25531poly -

morphisms was completely negative, which suggests

that there is no major interaction between these poly -

morphisms substantiating a genetic predisposition to

BPD. However, we cannot completely rule out the

possibility that we might have missed small effects due

to sample size. The regression analysis for the interac -

tion between the TH Val 81Met and the STin2 poly -

morphism revealed a low p value for TH polymorphism

(0.033) as well as for TH Val81Met genotype (0.010).

These results cannot be regarded as statistically sig -

nificant because the analysis was exploratory and sig -

nificance would not withstand a correction for multiple

testing considering all tests done in the present analy -

sis. Nevertheless, the results suggest a small effect of

the TH Val81Met polymorphism which has to be rep -

licated in independent and larger populations.

In conclusion, this is the first study on an asso -

ciation between variants of the TH Val 81Met poly -

morphism with BPD as well as on an interaction

between the TH Val 81Met, the 5-HTTLPR/rs25531

and the STin2 polymorphism in BPD. The strength

of the study is the hypothesis driven, rigorous statisti -

cal approach to identify potential gene–gene interac -

tions (Tadic´ et al. 2009) that helps lessen the generally

high risk of spurious false positive results. Moreover,

the same method can also be used to examine gene–

environment effects in BPD (Wagner et al. 2009).

The major limitation is the limited size of the study.

Met 81Met genotype was slightly over-represented in

BPD patients; however, this association did not with -

stand correction for multiple testing. A similar result

has been previously reported by Giegling et al.

(2008), who found an association of suicidal behav -

iour with gene variants of the DOPA decarboxylase,

which catalyzes the last step in dopamine biosynthe -

sis, but not with the TH Val 81Met polymorphism.

Taken together, these results do not support the idea

that the TH Val 81Met polymorphism plays a relevant

role for the alterations of the DA/NE system involved

in BPD and its behavioural dimensions.

There was no significant difference in genotype dis -

tribution of the tri-allelic 5-HTTLPR/rs25531 and

the STin2 polymorphisms between BPD patients and

healthy controls. Regarding the 5-HTTLPR/rs25531,

Table VI. Results of the multiple logistic regression analysis with disease status (BPD Patient yes vs. no) as outcome using the TH Val 81Met genotype and a binary coded variable for STin2 VNTR 9-/10-repeat allele carriers, as well as their interaction.

Regression coefficient P value 1 Exp(B) OR

HTT VNTR 9/10 0.898 0.031 2.455 TH genotype 0.033 TH genotype Val 81Met 1.198 0.010 3.314

TH genotype Met 81Met 0.898 0.129 2.455

HTT VNTR 9/10 × TH genotype 0.135

HTT VNTR 9/10 × TH genotype Val 81Met

–1.056 0.054 0.348

HTT VNTR 9/10 × TH genotype Met 81Met

–0.205 0.799 0.815

Intercept –0.898 0.012 0.407

For TH genotype, the genotype Val 81Val was used as reference group.1Wald test used. All p values are descriptive and cannot be considered significant at any level.

Table VII. Results of the multiple logistic regression analysis with disease status (BPD patient yes vs. no) as outcome using the TH Val 81Met genotype as a continuous covariate with values 0, 1, 2 representing the number of Met alleles and a binary coded variable for STin2 VNTR 9-/10-repeat allele carriers, as well as their interaction.

Regression coefficient p value 1 Exp(B) OR

HTT VNTR 9/10 0.593 0.110 1.809 TH genotype (number

of Met-alleles)

0.563 0.051 1.756

HTT VNTR 9/10 × TH genotype (number of Met-alleles)

–0.306 0.401 0.737

Intercept –0.634 0.041 0.530

1Wald test used. All p values are descriptive and cannot be considered significant at any level. TH Val 81Met, 5-HTTLpR and STin2 polymorphisms in bpD 51

American Psychiatric Association Practice Guidelines 2001. Practice guideline for the treatment of patients with borderline personality disorder. Am J Psychiatry 158(10 Suppl):1–52.Anguelova M, Benkelfat C, Turecki G 2003. A systematic review of association studies investigating genes coding for serotonin receptors and the serotonin transporter: II. Suicidal behavior. Mol Psychiatry 8:646–653.Beitchman JH, Baldassarra L, Mik H, De Luca V, King N, Bender D, et al. 2006. Serotonin transporter polymorphisms and persistent, pervasive childhood aggression. Am J Psychiatry 163:1103–1105.Biegon A, Fieldust S 1992. Reduced tyrosine hydroxylase immu - noreactivity in locus coeruleus of suicide victims. Synapse 10:79–82.Cadoret RJ, Langbehn D, Caspers K, Troughton EP, Yucuis R, Sandhu HK, et al. 2003. Associations of the serotonin trans - porter promoter polymorphism with aggressivity, attention deficit, and conduct disorder in an adoptee population. Compr Psychiatry 44:88–101.Caspi A, Sugden K, Moffitt TE, Taylor A, Craig IW, Harrington H, et al. 2003. Influence of life stress on depression: moderation by a polymorphism in the 5-HTT gene. Science 301:386–389.Coccaro EF, Kavoussi RJ, Sheline YI, Lish JD, Csernansky JG 1996. Impulsive aggression in personality disorder correlates with tritiated paroxetine binding in the platelet. Arch Gen Psy - chiatry 53:531–536.Collier DA, Stöber G, Li T, Heils A, Catalano M, Di Bella D, et al. 1996. A novel functional polymorphism within the pro - moter of the serotonin transporter gene: possible role in sus - ceptibility to affective disorders. Mol Psychiatry 1:453–460.Dumas B, Le Hir H, Bodeau-Péan S, Hirsch E, Thermes C, Mallet J 1996. New species of human tyrosine hydroxylase mRNA are produced in variable amounts in adrenal medulla and are overexpressed in progressive supranuclear palsy. J Neurochem 67:19–25.Ebstein RP 2006. The molecular genetic architecture of human personality: beyond self-report questionnaires. Mol Psychiatry 11:427–445.

Edwards A, Gibbs RA, Nguyen PN, Ansorge W, Caskey CT 1989.

Automated DNA sequencing methods for detection and anal -

ysis of mutations: applications to the Lesch-Nyhan syndrome.

Trans Assoc Am Phys 102:185–194.

Fiskerstrand CE, Lovejoy EA, Quinn JP 1999. An intronic poly -

morphic domain often associated with susceptibility to affective

disorders has allele dependent differential enhancer activity in

embryonic stem cells. FEBS Lett 458:171–174.

Frankle WG, Lombardo I, New AS, Goodman M, Talbot PS,

Huang Y, et al. 2005. Brain serotonin transporter distribution

in subjects with impulsive aggressivity: a positron emission

study with [11C] McN 5652. Am J Psychiatry 162:915–923.

Friedel RO 2004. Dopamine dysfunction in borderline personal -

ity disorder: a hypothesis. Neuropsychopharmacology 29:

1029–1039.

Giegling I, Moreno-De-Luca D, Rujescu D, Schneider B,

Hartmann AM, Schnabel A, et al. 2008. Dopa decarboxylase

and tyrosine hydroxylase gene variants in suicidal behavior.

Am J Med Genet B Neuropsychiatr Genet 147:308–315.

Grant BF, Chou SP, Goldstein RB, Huang B, Stinson FS, Saha

TD, et al. 2008. Prevalence, correlates, disability, and comor -

bidity of DSM-IV borderline personality disorder: results from

the Wave 2 National Epideiologic Survey on Alcohol and

Related Conditions. J Clin Psychiatry 69:533–545.

Gurvits IG, Koenigsberg HW, Siever LJ 2000. Neurotransmitter

dysfunction in patients with borderline personality disorder.

Psychiatr Clin North Am 23:27–40.Herpertz SC, Zanarini M, Schulz CS, Siever L, Lieb K, Möller HJ; WFSBP Task Force on Personality Disorders; World

Although our patient group is – to our knowledge –

the largest independent sample currently used in

BPD genetics, it is small compared to large scale

international case-control consortia in other areas of

medicine. Thus, establishing suitable co-operations to

increase sample sizes will be one future task. In any

case, in this study we neither identified a significant

association between individual polymorphisms and

BPD nor a significant gene–gene interaction increas -

ing the susceptibility for BPD. Therefore, we cannot

provide further evidence for the involvement of the

NE/DA system in BPD. A slight over-representation

of the individual TH Met81Met genotype in BPD

patients as well as a low p value for the factor TH

genotype in a regression analysis including the STin2

polymorphism need to be replicated in independent

and larger populations. These largely negative results

do not rule out the importance of alterations in DA

and NE neurotransmission, because these are regu -

lated by many enzymes and receptors, which have not

yet been investigated for their role in BPD. Future

studies in this field should extent our findings to vari -

ants of the whole TH gene for fine-mapping and the

identification of possible risk haplotypes as well as to

further genes coding for proteins involved in DA/NE

pathways. Furthermore, it should be of particular

interest, whether the TH Val 81Met polymorphism

modulates the risk of developing BPD symptoms as

a function of life events (e.g., childhood trauma),

which might lead to better understanding of the neu -

robiological correlates of the G  E model in the

aetiology of BPD (Lieb et al. 2004).

Acknowledgements

We are grateful to all participants as well as to our

cooperating clinical partners (Drs Guth and Krafc -

zyk at the Clinic for Psychiatry in Alzey; Dr Lehr at

the Center for Social Psychiatry in Eltville) for study

support. We wish to thank Dr Anja Victor from the

Institute for Medical Biometry, Epidemiology and

Informatics, University of Mainz, for statistical

advice. Funding source: University of Mainz, Faculty

of Medicine (MAIFOR 2007) (AT).

Statement of Interest

The authors declare that they have no conflicts of

interest.

References

American Psychiatric Association 1994. Diagnostic and statistical manual of mental disorders, 4th ed. Washington, DC: American Psychiatric Press. 52 A. Tadi´c et al.

serotonin transporter gene and borderline personality disorder. J Psychiatr Res 42:87–88.Persson ML, Wasserman D, Geijer T, Jonsson EG, Terenius L 1997. Tyrosine hydroxylase allelic distribution in suicide attempters. Psychiatry Res 72:73–80.Ragoussis J, Elvidge GP, Kaur K, Colella S 2006. Matrix-assisted laser desorption/ionisation, time-of-flight mass spectrometry in genomics research. PLoS Genet (2)7:e100.Retz W, Retz-Junginger P, Supprian T, Thome J, Rösler M 2004. Association of serotonin transporter promoter gene polymor - phism with violence: relation with personality disorders, impulsivity, and childhood ADHD psychopathology. Behav Sci Law 22:415–425.Roy A, Hu XZ, Janal MN, Goldman D 2007. Interaction between childhood trauma and serotonin transporter gene variation in suicide. Neuropsychopharmacology 32:2046–2052.Rujescu D, Giegling I, Gietl A, Hartmann AM, Möller HJ 2003. A functional single nucleotide polymorphism (V158M) in the COMT gene is associated with aggressive personality traits. Biol Psychiatry 54:34–39.Ryding E, Ahnlide JA, Lindstrom M, Rosen I, Traskman-Bendz L

2006. Regional brain serotonin and dopamine transporter binding capacity in suicide attempters relate to impulsiveness and mental energy. Psychiatry Res 148:195–203.Saiki RK, Gelfand DH, Stoffel S, Scharf SJ, Higuchi R, Horn GT, et al. 1988. Primer-directed enzymatic amplification of DNA with thermostable polymerase. Science 239:487–491.Skodol AE, Siever LJ, Livesley WJ, Gunderson JG, Pfohl B, Widiger TA 2002. The borderline diagnosis II: biology, genet - ics, and clinical course. Biol Psychiatry 51:951–963.Smolka MN, Bühler M, Schumann G, Klein S, Hu XZ, Moayer M, et al. 2007. Gene-gene effects on central processing of aver - sive stimuli. Mol Psychiatry 12:307–317.Souery D, Oswald P, Linkowski P, Mendlewicz J 2003. Molecular genetics in the analysis of suicide. Ann Med 35:191–196.Stanssens P, Zabeau M, Meersseman G, Remes G, Gansemans Y, Storm N, et al. 2004. High-throughput MALDI-TOF discov - ery of genomic sequence polymorphisms. Genome Res 14: 126–133.Steiger H, Richardson J, Joober R, Gauvin L, Israel M, Bruce KR, et al. 2007. The 5-HTTLPR polymorphism, prior maltreat - ment and dramatic-erratic personality manifestations in women with bulimic syndromes. J Psychiatry Neurosci 32:354–362.Storm N, Darnhofer-Patel B, van den Boom D, Rodi CP 2003. MALDI-TOF mass spectrometry based SNP genotyping. Methods Mol Biol 212:241–262.Tadic´ A, Victor A, Bas¸ kaya O, von Cube R, Hoch J, Kouti I, et al. 2008. Interaction between gene variants of the serotonin transporter promoter region (5-HTTLPR) and catechol O-methyltransferase (COMT) in borderline personality disorder. Am J Med Genet B Neuropsychiatr Genet. 150: 487–495.Torgersen S, Kringlen E, Cramer V 2001. The prevalence of personality disorders in a community sample. Arch Gen Psychiatry 58:590–596.Wagner S, Baskaya Ö, Lieb K, Dahmen N, Tadic´ A 2009. The 5-HTTLPR polymorphism modulates the association of seri - ous life events (SLE) and impulsivity in patients with Border - line Personality Disorder. J Psychiatr Res 43:1067–1072.Wang Y, Ju J, Carpenter BA, Atherton JM, Sensabaugh GF, Mathies RA 1995. Rapid sizing of short tandem repeat alleles using capillary array electrophoresis and energy-transfer fluorescent primers. Anal Chem 67:1197–1203.Wendland JR, Kruse MR, Cromer KR, Murphy DL 2007. A large case-control study of common functional SLC6A4 and BDNF variants in obsessive-compulsive disorder. Neuropsychophar - macology 32:2543–2551.

Federation of Societies of Biological Psychiatry (WFSBP) 2007. World Federation of Societies of Biological Psychiatry (WFSBP) guidelines for biological treatment of personality disorders. World J Biol Psychiatry 8:212–244.Hoefgen B, Schulze TG, Ohlraun S, von Widdern O, Höfels S, Gross M, et al. 2005. The power of sample size and homoge - nous sampling: association between the 5-HTTLPR serotonin transporter polymorphism and major depressive disorder. Biol Psychiatry 57:247–251.Hu XZ, Lipsky RH, Zhu G, Akhtar LA, Taubman J, Greenberg BD, et al. 2006. Serotonin transporter promoter gain-of-func - tion genotypes are linked to obsessive-compulsive disorder. Am J Hum Genet 78:815–826.Jurinke C, Oeth P, van den Boom D 2004. MALDI-TOF mass spectrometry: a versatile tool for high-performance DNA anal - ysis. Mol Biotechnol 26:147–164.Kumer SC, Vrana KE 1996. Intricate regulation of tyrosine hydroxylase activity and gene expression. J Neurochem 67: 443–462.Lenzenweger MF, Lane MC, Loranger AW, Kessler RC 2007. DSM-IV personality disorders in the National Comorbidity

Survey Replication. Biol Psychiatry 62:553–564.Lesch KP, Balling U, Gross J, Strauss K, Wolozin BL, Murphy DL, et al. 1994. Organization of the human serotonin trans - porter gene. J Neural Transm Gen Sect 95:157–162.Lesch KP, Bengel D, Heils A, Sabol SZ, Greenberg BD, Petri S, et al. 1996. Association of anxiety-related traits with a polymor - phism in the serotonin transporter gene regulatory region. Sci - ence 274:1527–1531.Li D, He L 2007. Meta-analysis supports association between serotonin transporter (5-HTT) and suicidal behavior. Mol Psy - chiatry 12:47–54.Lieb K, Zanarini MC, Schmahl C, Linehan MM, Bohus M 2004. Borderline personality disorder. Lancet 364:453–461.Lovejoy EA, Scott AC, Fiskerstrand CE, Bubb VJ, Quinn JP 1998. The serotonin transporter intronic VNTR enhancer correlated with a predisposition to affective disorders has distinct regula - tory elements within the domain based on the primary DNA sequence of the repeat unit. Eur J Neurosci 17:417–420.Lüdecke B, Bartholomé K 1995. Frequent sequence variant in the human tyrosine hydroxylase gene. Hum Genet 95:716.MacKenzie A, Quinn J 1999. A serotonin transporter gene intron 2 polymorphic region, correlated with affective disorders, has allele-dependent differential enhancer-like properties in the mouse embryo. Proc Natl Assoc Sci USA 96:15251–15255.Ni X, Bismil R, Chan K, Sicard T, Bulgin N, McMain S, et al. 2006a. Serotonin 2A receptor gene is associated with personal - ity traits, but not to disorder, in patients with borderline per - sonality disorder. Neurosci Lett 408:214–219.Ni X, Chan K, Bulgin N, Sicard T, Bismil R, McMain S, et al. 2006b. Association between serotonin transporter gene and borderline personality disorder. J Psychiatr Res 40:448–453.Ni X, Sicard T, Bulgin N, Bismil R, Chan K, McMain S, et al. 2007. Monoamine oxidase A gene is associated with borderline personality disorder. Psychiatr Genet 17:153–157.O’Malley KL, Anhalt MJ, Martin BM, Kelsoe JR, Winfield SL, Ginns EI 1987. Isolation and characterization of the human tyrosine hydroxylase gene: identification of 50 alternative splice sites responsible for multiple mRNAs. Biochemistry 26: 6910–6914.Oquendo MA, Mann JJ 2000. The biology of impulsivity and suicidality. Psychiatr Clin North Am 23:11–25.Ordway GA, Smith KS, Haycock JW 1994. Elevated tyrosine hydroxylase in the locus coeruleus of suicide victims. J Neuro - chem 62:680–685.Pascual JC, Soler J, Barrachina J, Campins MJ, Alvarez E, Pérez V, et al. 2008. Failure to detect an association between the TH Val 81Met, 5-HTTLpR and STin2 polymorphisms in bpD 53

Wilson ST, Stanley B, Brent DA, Oquendo MA, Huang YY, Mann JJ 2009. The tryptophan hydroxylase-1 A218C polymor - phism is associated with diagnosis, but not suicidal behavior, in borderline personality disorder. Am J Med Genet B Neuro-psychiatr Genet 150B:202–208.Wittchen HU, Zaudig M, Fydrich T 1996. SKID-I/-II: Strukturiertes klinisches Interview für DSM-IV. Göttingen: Hogrefe.

Wittchen H, Lachner G, Wunderlich U, Pfister H 1998. Test-retest reliability of the computerized DSM-IV version of the Munich-Composite International Diagnostic Interview (M-CIDI). Soc Psychiatry Psychiatr Epidemiol 33:568–578.Zaboli G, Gizatullin R, Nilsonne A, Wilczek A, Jönsson EG, Ahnemark E, et al. 2006. Tryptophan hydroxylase-1 gene variants associate with a group of suicidal borderline women. Neuropsychopharmacology 31:1982–1990. 54 A. Tadi´c et al.

Supplemental material: Methods

All primers (Table I) were purchased from Metabion,

Martinsried Germany. If not otherwise stated, all

reagents, instruments and software tools were from

Sequenom, San Diego USA.

Genotyping of TH Val 81Met (rs6356, dbSNp build

52/129; G/A )

pCR. PCR was performed in a 5-µl reaction volume

and composed of approximately 5 ng human genomic

DNA, 1,25× Hot-Star Taq PCR buffer (Qiagen), 3.5

mM MgCl 2, 500 µM each dNTP (Roche), 100 nM

each PCR primer (PCR1_rs6356 and PCR2_rs6356,

respectively; see Table I) , 0.15 U HotStarTaq (Qia -

gen). The reaction mixture was initially denaturated

for 15 min at 95°C followed by 45 cycles of 95°C

for 20 s, 56°C for 30 s, 72°C for 1 min. The final

elongation was carried out for 3 min at 72°C.

Removal of not incorporated dNTps . Nucleotide

triphosphates not incorporated into the PCR prod -

uct during PCR were digested with shrimp alkaline

phosphatase (SAP) to prevent interference in the fol -

lowing MassEXTEND reaction. The enzyme cock -

tail consisting of 0.3 U SAP in 2 µl of 1× SAP buffer

was directly added to the crude PCR product. The

reaction mixture was incubated for 20 min at 37°C

followed by 5 min at 85°C for SAP inactivation.

hME reaction. After SAP-enzyme heat inactivation,

the MassEXTEND cocktail was directly added to

the crude product. Final reaction was composed of

625 nM MassEXTEND primer (ME_rs6356, see

Table I), 50 µM of each d/ddNTP (ACG stop mix)

and 0.063 U Thermosequenase. The analytical

Table I. Primers

FAM = 5  FAM label for fluorescent detection of VNTR analysis

Bases in lower case letters indicate method-specific tags, which are not target specific. These tags are attached at 5  end of the primer.

Gene Target polymorphism Primer name/Function Sequence 5 3

Tyrosine hydroxylase (TH) Val 81Met; rs6356; GA

PCR1_rs6356 acgttggatgTGGCCTTTGAGGAGAAGGAG

PCR2_rs6356 acgttggatgTCAAACACCTTCACAGCTCG

ME_rs6356 GAAGAGCAGGTTTAGCA

Serotonin transporter (5HTT)

LPR and rs25531 L/S and A G PCR1_5HTT_T7 cagtaatacgactcactatagggagaaggctATGCTGGAAGGGCTGCAGG

PCR2_5HTT acgttggatgCTCCTGCATCCCCCATTATC

intron 2 (STin2)VNTR 9, 10, 12 repeats

5HTT_VNTR_INT2_F FAM-GTGATTGGCTATGCTGTGGACC

5HTT_VNTR_INT2_F CATAATCTGTCTTCTGGCCTCTCAA

primer extension reaction was incubated for 2 min

at 94°C and then 75× cycled at 94°C for 5 s, 52°C

for 5 s, 72°C for 5 s.

Sample conditioning and MALDI-ToF MS; genotype

determination . The primer extension products were

diluted with 16 µl water, conditioned with 3 mg

CLEAN Resin, and then dispensed onto a Spectro -

Chip with a MassARRAY nanodispenser using the

supplier’s recommended standard settings. The chips

were scanned using a MassARRAY compact ana -

lyzer workstation with standard parameter settings.

The obtained raw data spectra were analyzed and

genotypes were determined using the MassARRAY

Typer software version 3.4. Genotype calls were

inspected and where necessary operator revised and/

or operator calls applied.

Simultaneous Genotyping the 5-HTTLpR L/S-variant

and SNp rs25531 (dbSNp build 72/129; A/G),

located within the insertion

This study aimed the analysis of the L/S-variant, 43

bp insertion/deletion in correlation to rs25531, A/G.

However, the applied MassCLEAVE method sup -

ports and requires the simultaneous determination

of these complex variations together with the deter -

mination of further SNPs (rs25530, C/T and

rs4795542, A/C) located in close proximity to the

targeted SNP rs25531. The applied T-forward and

C-forward cleavage reactions assist the non-ambigu -

ous detection of L/S variants and the haplotype

determination of SNPs rs25530 and rs25531. The

third SNP, rs4795542, was not of interest in this

study. Its accurate genotyping would require the

analysis of T-reverse cleavage reaction. However pres -

ence or absence of the A-allele can be determined

with the performed T- and C-forward reaction. TH Val 81Met, 5-HTTLpR and STin2 polymorphisms in bpD 55

Figure 1. Simulation results to determine allele identifiers for the 5HTT S/L and rs25531 genotyping.The presented data show the expected signal pattern of T-specific and C-specific cleavage reaction in forward direction of the generated PCR product. These two reactions allow non-ambiguous genotyping of the L/S alleles together with simultaneous genotype determination of rs25531. C-specific cleavage of the reverse strand is not informative for any of the considered sequence variants (data not shown). The pattern changes in the considered T- and C-specific cleavage analysis of the forward strand allow only the determination if the A-allele of SNP rs4795542 (A/C, position 1) is present or absent. The C-allele is not represented by an unique identifier. T- specific cleavage of the reverse strand would be required for non-ambiguous genotyping of this SNP (data not shown). SNP rs4795542 was not subject of this study. Therefore both reverse cleavage reactions were not performed. Illustration of possible and considered haplotype sequences. SNP definitions are extracted from dbSNP, build 129. Positions 1, 2, and 3 define the haplotype bases indicated in the sequence name of the insertion variant. The insert region itself is underlined. Position 1 = rs4795542 (A/C), position 2 = rs25531 (T/C), and position 3 = rs25530 (A/G). Two additional dbSNP annotated variants are the single base insertions of "A" and "G", respectively. These variants refer to rs4795541 and include the insertion/deletion polymorphism.

The tables present the expected signal pattern and the proposed masses for each of the considered alleles. The allele identifying signals are highlighted and indicated. In contrast to the standard MassCLEAVE analysis where the presence of each allele is determined by its own unique identifier, the highly repetitive sequence in the region of interest requires semi-quantitative considerations. The 43bp segment ("L"-allele or insertion) is identified by the yellow highlighted signals (x). The absence of the 43bp segment ("s"-allele or deletion) can be determined only by "missing" these signals. In case of a heterozygous sample these signals are still present, but with reduced intensity. Therefore two so called constant signals (labeled with an yellow boxed arrow, ↓) were chosen as reference signals. These signals are independently generated from the insertion / deletion genotype of a sample and appear with a conserved relative intensity of "1". That way the relative intensity of the "L"-identifiers indicates also the zygosity respectively the L/S variant of a sample. Signals labeled in light blue identify the insertion of a single "A" and "G", respectively instead of the entire 43bp insert. Signals in dark blue indicate the identifier for the "A"-allele at position 1. This position can be always independently analyzed. Fragments are not interfered by the other SNPs at position 2 and 3. SNPs at position 2 and 3 do influence each other with respect to the observed identifying signals. The haplotypes "CA", "CG", and "TA" can be identified by the presence of one of the dark green labeled signals (x). The "TG"-haplotype does not generate a new signal. Here a mass overlap with a constant signal will be observed. The signal is light green labeled (x). This already present signal will be observed with increased relative intensity and is therefore indirectly suitable for the determination of this haplotype. The intensity comparison of the "L/S" genotype identifier with intensity of the haplotype signals will support / confirm the presence or absence of this

haplotype.

C Fo r 1607.0

19 52.2

20 18.3

26 83.7

26 99.7

29 87.9

30 03.9

30 12.9

30 44.9

36 23.0

36 87.3

37 03.3

40 32.5

40 48.5

46 81.9

10 404.4 5H TT_ Ax xx xxxxx

5H TT_ Delx xx xxxx

5H TT_ Gx xx xxxx x

5H TT_ ins-AC Ax xx xxxxx xxx

5H TT_ ins-AC Gx xx xxxxx xxx

5H TT_ ins-AT Ax xx xxxxx xxx

5H TT_ ins-AT Gx xx xxxxxxx

5H TT_ ins-CC Ax xx xxxxx xx

5H TT_ ins-CC Gx xx xxxxx xx

5H TT_ ins-CT Ax xx xxxxx xx

5H TT_ ins-CT Gx xx xxxxxx

T Fo r 2035. 2

20 51. 2

21 16. 4

23 80. 5

29 58. 8

29 74. 8

29 98. 8

30 30. 8

32 72. 0

36 73. 3

36 89. 3

37 29. 3

39 13. 2

39 38. 4

40 18. 5

40 34. 5

40 74. 5 5HTT _A xxx xxxxxx x

5H TT _Del xxx xxxxxx

5H TT _G xxx xxxxxxx

5H TT _ins -A CA xxx xxxxxxx xxx

5H TT _ins -A CG xxx xxxxxxx xxx

5H TT _ins -A TA xxxx xxxxxx xxx

5H TT _ins -A TG xxx xxxxxx xxx

5H TT _ins -C CA xxx xxxxxx xxx

5H TT _ins -C CG xxx xxxxxx xxx

5H TT _ins -C TA xxxx xxxxx xxx

5H TT _ins -C TG xxx xxxxx xxx

Table 1. Simulation results for T-forward cleavage reaction (a) and for C-forward cleavage reaction (b). 56 A. Tadi´c et al.

2700 2800 2900 3000 3100 3200 3300 3400 3500 3600 3700

2683.659

3003.85

3012.86

3044.86

3623.03

3687.28

3703.28

B59 B63

B152 B26 4 5 1 2 3

Figure 2. Examples of genotype determination of SNPs located within the "L"-allele of 5HTT gene; C-forward reaction.As outlined in figure 1 legend, genotype determination of SNP rs25531 has to consider always the insertion/deletion as well as SNP rs25530. All shown example spectra are obtained from C-specific cleavage reaction of the forward.The applied reference sequence for data acquisition of all samples was the haplotype ins-ACA (see figure 1). Mass signals representing this haplotype are indicated by green lines and the mass label displayed by the software. The general identifiers of the "L"-allele are the mass signals at 3044.9 Da (4) and 3867.3 Da (5).The yellow boxed arrows ( ↓) indicate the reference signals used to determine the relative intensity of signals indicative for the presence of the insertion ("L"-allele, signals 4 and 5) to conclude sample's zygosity and to determine the SNP-haplotype of the "L"-alleles. The reference signals are set to an arbitrary intensity of "1". The intensity of signals indicative for the insertion is compared with the reference signals.In case of sample B59 (black spectrum) none of "L"-allele identifying signals is present. In conclusion this sample is homozygous for the 43bp deletion (genotype "SS"). Sample B152 (ochre spectrum) displays both identifiers 4 and 5. The observed intensity is approximately ½ of the reference signals with closest mass. Therefore this sample is genotyped "SL" for the deletion/insertion variant. Samples B63 (green spectrum) and B26 (brown spectrum) also show these signals, however with about equal intensity as the corresponding reference signals. In conclusion B26 and B63 are genotyped "LL" homozygous.Regarding rs25531, sample B59 does not need further consideration, since its homozygous "SS" genotype. In concordance none of the haplotype identifying signals 1, 2, nor 3 is observed.Sample B63 does not display additional signals as well. However the relative intensity of the mass signal at 3003.8 Da (signal x) is about 2-fold increased compared to the reference signal at 3012.8 ( ↓). Since all other haplotype identifying signals are absent this observation leads to the conclusion of a homozygous "TG"-haplotype (for SNPs rs25531-rs25530; positions 2 and 3 in figure 1a).Sample B26 displays also an increased intensity of the signal at 3003.8 Da (x), but relatively lower than in sample B63. The intensity increase is about 1.5-fold when compared to the reference signal at 3012.8 Da. That suggests only one of the alleles carries the "TG"-haplotype. The other allele is described by the appearance of the "TA"-haplotype identifying signal at 2987.9 Da (3). Sample B152 was characterized as "SL". In addition to the insertion identifying signals the signal 2 is observed, which indicates the presence of one "CG"-haplotype allele. The intensity of signal x is not significantly increased compared to signal ↓ supporting the concluded result.

There was no sample displaying the signal 1 indicative for the "CA"-haplotype for SNPs rs25531-rs25530.All determined genotypes/haplotypes in the discussed C-forward cleavage reactions were confirmed in analysis of the corresponding T-forward cleavage reaction (see figure 3).

product during PCR were digested with shrimp alka -

line phosphatase (SAP) to prevent interference in the

following MassCLEAVE reaction. The enzyme cock -

tail consisting of 0.3 U SAP in 2 µl of 1× SAP buffer

was directly added to the crude PCR product. The

reaction mixture was incubated for 20 min at 37°C

followed by 5 min at 85°C for SAP inactivation.

Transcription and base specific enzymatic cleavage.

After SAP treatment the crude reaction product

was divided in two halves and MassCLEAVE

cocktails for T-specific and C-specific cleavage,

respectively were directly added to following final

pCR. The reaction was performed in a 5 µl volume

containing 5 ng human genomic DNA, 1× HiFi-

PCR buffer (Invitrogen), 2 mM MgSO 4, 200 µM

each dNTP (Roche), 10% DMSO, 1× Q-Solution

(Qiagen), 0.3 U HiFi-Taq DNA polymerase (Invit -

rogen), 400 nM of each primer (PCR1_5HTT-

rs25531 and PCR2_5HTT-rs25531, respectively;

see Table I). The reaction mixture was initially dena -

turated for 2 min at 94°C followed by 35 cycles of

94°C for 30 s, 58°C for 30 s, 68°C for 60 s.

Removal of not incorporated dNTps . Nucleotide

triphosphates not incorporated into the PCR TH Val 81Met, 5-HTTLpR and STin2 polymorphisms in bpD 57

concentrations 1× transcription buffer, 5 mM DTT,

20 U T7 R&DNA polymerase, 0.02 µg/µl RNase A,

and 1× C- and T-cleavage nucleotide mix, respec -

tively. The reaction mixture was incubated for 3 h

at 37°C.

Sample conditioning and MALDI-T oF MS . The

reaction products were diluted with 25 µl water

conditioned with 6 mg CLEAN resin, and then dis -

pensed onto a SpectroChip with a MassARRAY

nanodispenser using the suppliers recommended

standard settings. The chips were scanned using a

MassARRAY compact analyzer workstation with

standard parameter settings. The obtained raw data

were visualized with SNP Discovery Analyzer soft -

ware version 1.2.

2000 2250 2500 2750 3000 3250 3500 3750 4000

2051.238 2380.447 2942.818 2958.818 2998.842 3030.841 3673.261 3689.260 3729.285 3913.200 3938.421

B59

B152

5 2 3272.029 6 2116.420

1 x 4 x

Figure 3: Examples of genotype determination of SNPs located within the "L"-allele of 5HTT gene; T-forward reaction.As outlined in figure 1 legend, genotype determination of SNP rs25531 has to consider always the insertion/deletion as well as SNP rs25530. All shown example spectra are obtained from T-specific cleavage reaction of the forward strand.The applied reference sequence for data acquisition of all samples was the haplotype ins-ACA (see figure 1). Mass signals representing this haplotype are indicated by green lines and the mass label displayed by the software. The general identifiers of the "L"-allele are the mass signals at 3030.8 Da (5) and 3673.3 Da (6).The yellow boxed arrows ( ↓) indicate the reference signals used to determine the relative intensity of signals indicative for the presence of the insertion ("L"-allele, signals 4 and 5) to conclude sample's zygosity and to determine the SNP-haplotype of the "L"-alleles. The reference signals are set to an arbitrary intensity of "1". The intensity of signals indicative for the insertion is compared with the reference signals.In case of sample B59 (black spectrum) none of "L"-allele identifying signals is present. In conclusion this sample is homozygous for the 43bp deletion (genotype "SS"). Sample B152 (ochre spectrum) displays both identifiers 4 and 5. The observed intensity is approximately ½ of the reference signals with closest mass. Therefore this sample is genotyped "SL" for the deletion/insertion variant. Samples B63 (green spectrum) and B26 (brown spectrum) also show these signals, however with about equal intensity as the corresponding reference signals. In conclusion B26 and B63 are genotyped "LL" homozygous.Regarding rs25531, sample B59 does not need further consideration, since its homozygous "SS" genotype. In concordance none of the haplotype identifying signals 1, 2, nor 3 is observed.Sample B63 does not display any additional signals as well. However the relative intensity of the mass signal at 2051.2 Da (signal x) is about 2-fold increased compared to the reference signal at 2116.4 Da ( ↓). Since all other haplotype identifying signals are absent this observation leads to the conclusion of a homozygous "TG"-haplotype (for SNPs rs25531-rs25530; positions 2 and 3 in figure 1a).Sample B26 displays also an increased intensity of the signal at 2051.2 Da (x), but relatively lower than in sample B63. The intensity increase is about 1.5-fold when compared to the reference signal at 2116.4 Da. That suggests only one of the alleles carries the "TG"-haplotype. The other allele is described by the appearance of the "TA"-haplotype identifying signal at 2035.2 Da (1). Sample B152 was characterized as "SL". In addition to the insertion identifying signals (4 and 5) the signal at 2974.8 Da (3) is observed,

which indicates the presence of one "CG"-haplotype allele. The intensity of signal 3 is about equal to signals 4 and 5 supporting the concluded result.There was no sample displaying the signal 2 indicative for the "CA"-haplotype for SNPs rs25531-rs25530.All determined genotypes/haplotypes in the discussed T-forward cleavage reactions were confirmed in analysis of the corresponding C-forward cleavage reaction (see figure 2).The signal at 3729.3 Da (6) was never observed indicating that the A-allele of rs4795542 (position 1 figure 1a) is not present in this sample set. Data were confirmed in C-forward reaction, where the A-allele identifying signal at 4681.9 Da was never detected (signal is outside the displayed mass range of figure 2). Exact genotyping of this SNP would require the analysis of T-reverse reaction. This analysis was not performed because rs4795542 was not of interest in this study. 58 A. Tadi´c et al.

1.5 µl Q-Solution, 20 ng human genomic DNA

template, and 0.75 µM of each primer (TH01_

VNTR_F and TH01_VNTR_R, respectively; see

Table I). Cycling was performed to heat once at

95°C for 10 min followed by 30 cycles at 94°C for

30 s, 58°C for 90 s and 72°C for 60 s, final extension

was at 60°C for 30 min.

Capillar y electrophoresis, analyte detection and data anal -

ysis. For the fragment analysis, PCR products were

diluted 1:4 with MegaBACE ET550-R size standard

working solution which results from MegaBACE

ET550-R stock solution diluted 1:32 in water. Sam -

ples were then denatured at 98°C for 3 min, quickly

chilled on ice for 4 min and subjected to capillary-

electrophoresis on MegaBACE 1000 DNA Analyzer.

Electrokinetic injection was at 3 kV for 55 s and elec -

trophoresis was continued at 3 kV for 80 min. The

data were semi-automated analyzed with Softgenetics

Genemarker 1.5 and operator inspected. Where nec -

essary, operator calls were applied.

Genotype determination . SNP genotypes were operator

determined by identification of allele-identifying signals

and correlating their relative signal heights to constant

reference signals for determination of the S/L genotype.

First, a simulation was performed to calculate the allele

identifying mass signals. The sequences were in-silico

cleaved at each C and T nucleotide, respectively. The

molecular mass of obtained fragments was calculated

according to the composition of fragments. In result,

each of the considered alleles and their combinations

delivered at least one identifying signal, which allows

genotyping. The sequences are presented in Figure 1 and

the simulation results are presented in Tables 1a and 1b. The

Figures 2 and 3 present examples of genotype determination

of SNps located within the "L" allele of 5HTT gene by

C-forward and T-forward reaction, respectively.

Analysis of VNTR variants in intron 2 of the serotonin

transporter gene (5HTT STin2)

PCR: The target region was PCR amplified contain -

ing 7.5 µl 2 × Multiplex PCR Master Mix (Qiagen),

may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express

written permission. However, users may print, download, or email articles for individual use.

may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express

written permission. However, users may print, download, or email articles for individual use.