Embed Size (px)
Citation preview
Progress in Neuro-Psychopharmacology & Biological Psychiatry 51 (2014) 119–125
Contents lists available at ScienceDirect
Progress in Neuro-Psychopharmacology & BiologicalPsychiatry
j ourna l homepage: www.e lsev ie r .com/ locate /pnp
Gene × gene × gender interaction of BDNF and COMT genotypesassociated with panic disorder
Yoshiaki Konishi a, Hisashi Tanii a,⁎, Takeshi Otowa b, Tsukasa Sasaki c, Mamoru Tochigi d, Tadashi Umekage e,Eishi Motomura a, Takashi Shiroyama a, Hisanobu Kaiya f, Yuji Okazaki g, Motohiro Okada a
a Department of Psychiatry, Division of Neuroscience, Graduate School of Medicine, Brain Science and Animal Model Research Center (BSAM), Mie University, Mie, Japanb Department of Neuropsychiatry, Graduate School of Medicine, The University of Tokyo, Tokyo, Japanc Laboratory of Health Education, Graduate School of Education, The University of Tokyo, Tokyo, Japand Department of Psychiatry, Teikyo University School of Medicine, Tokyo, Japane Division for Environment, Health and Safety, The University of Tokyo, Tokyo, Japanf Research Center for Panic Disorder, Nagoya Mental Clinic, Aichi, Japang Department of Psychiatry, Tokyo Metropolitan Matsuzawa Hospital, Tokyo, Japan
Abbreviations: BDNF, brain-derived neurotrophimethyltransferase; DSM-IV, Diagnostic and Statistical4th edition; HA, harm avoidance; MANCOVA, multivMet, methionine; MINI 24, Mini International NeurNEO-PI-R, NEO Inventory; PCR, polymerase chain rPTSD, posttraumatic stress disorder; SPSS, StatisticaSTAI, The State-Trait Anxiety Inventory; Val, valine.⁎ Corresponding author at: Department of Psychia
Graduate School of Medicine, Brain Science and Animal MMie University, 2-174, Edobashi, Tsu, Mie, Japan. Tel.: +231 5208.
E-mail address: [email protected] (H. Tan
0278-5846/$ – see front matter © 2014 Elsevier Inc. All rihttp://dx.doi.org/10.1016/j.pnpbp.2014.01.020
a b s t r a c t
a r t i c l e i n f oArticle history:Received 8 November 2013Received in revised form 29 January 2014Accepted 29 January 2014Available online 4 February 2014
Keywords:BDNFCOMTGenderGene × gene × gender interactionPanic disorder
Genetic and gender differences are among the factors that have a role in the etiology of panic disorder (PD). Itis thought that PD is related to neurotransmitter pathways, such as brain-derived neurotrophic factor (BDNF)and catechol-O-methyltransferase (COMT), both of which are involved in the regulation of the monoaminemechanism. We examined the interactions of BDNF, COMT and gender differences in terms of personalitycharacteristics in PD. The subjects were 470 patients (178 men, 292 women) with a DSM-IV diagnosis of PD,and 458 healthy controls (195 men, 263 women). The subjects were further clinically characterized using theRevised NEO Personality Inventory (NEO-PI-R) and State-Trait Anxiety Inventory (STAI). COMT Val158Met poly-morphisms (rs4680) and BDNF Val66Met (rs6265) polymorphisms were genotyped using allelic discriminationby a real-time PCR assay. A multivariate analysis of covariance (MANCOVA) was performed with STAI and NEO-PI-R scores as the dependent factor, gender and genotyping groups (BDNF and COMT) as fixed factors, and thecovariate of age in the PD and healthy control groups. Post hoc MANCOVA tests were conducted to evaluateCOMT × BDNF interactions. An interaction of BDNF × COMT × gender was confirmed in the PD group byMANCOVA on STAI scores and NEO-PI-R Neuroticism and Extraversion scores, whereas no association of suchinteractions was observed in the healthy controls. The anxiety sensitivity of the COMT Met + BDNF Val/Valcarriers was higher than that of the COMT Val/Val + BDNF Val/Val carriers by post hoc MANCOVA. A significantBDNF × COMT × gender interaction was observed in the PD patients but not in the controls. Our findings partlydemonstrated the involvement of a gene × gene × gender interaction in the pathogenesis of PD.
© 2014 Elsevier Inc. All rights reserved.
1. Introduction
Panic disorder (PD) is an anxiety disorder characterized by frequent,unexpected panic attacks and anticipatory anxiety. Epidemiologicalstudies have detected gender differences in the symptomatology ofPD. Women are affected with PD approximately twice as often as men,
c factor; COMT, catechol-O-Manual of Mental Disorders,ariate analysis of covariance;opsychiatric Interview 5.0.0;eaction; PD, panic disorder;l Package for Social Science;
try, Division of Neuroscience,odel Research Center (BSAM),81 59 231 5018; fax: +81 59
ii).
ghts reserved.
and experience more panic symptoms than men do (Crowe et al.,1983). Female patients with PD are more likely to have respiratorysymptoms (shortness of breath, feeling smothered, choking, difficultyswallowing) and faintness during panic attacks (Weissman, 1993).
It has been suggested that genetic factors may influence vulnerabil-ity to PD, as many current patients have a relative with the disorder(Crowe et al., 1983; Hettema et al., 2001; Weissman, 1993). Severalgenes related to neurotransmission and neurotrophic systemsmay con-tribute to the genetic variation of PD-related traits, and theymaymodifythe phenotypic expression of pathologic anxiety. One plausible geneticrisk factor involves brain-derived neurotrophic factor (BDNF), a proteinhypothesized to limit or repair the damage caused by stress, andcatechol-O-methyltransferase (COMT) (Domschke et al., 2007), bothofwhich has been investigated as the typical gene polymorphisms relatedto anxiety disorders such as panic disorder.
BDNF Met carriers may play a crucial role in increased sensitivity toanxiety, and persons with the Met/Met carriers polymorphism may be
120 Y. Konishi et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 51 (2014) 119–125
more sensitive to the stress-induced down-regulation of BDNF (Elzingaet al., 2011). The symptoms of Met/Met carriers tended to be moreserious compared to those of Val/Val or Val/Met carriers (Monteleoneet al., 2006). COMT Val158Met polymorphisms have been associatedwith “worrier” (Met allele) tendencies based on findings that theMet allele has been broadly associatedwith anxiety-related phenotypes(Domschke et al., 2004; McGrath et al., 2004). Studies of personalitytraits in anxiety disorders have clarified a relationship between neurot-icism (Ormel et al., 2004) and extraversion (Jylha et al., 2009). Anotherrecent study reported that resiliencewas influenced byCOMTand BDNFpolymorphisms in male — but not female — healthy college students(Kang et al., 2013). Resilience was reported to be negatively associatedwith neuroticism, and positively related to extraversion in young healthyadults (Campbell-Sills et al., 2006).
Regarding gene–gene interactionwith anxiety disorder, associationsbetweenBDNF and dopamine receptor 3 gene polymorphismhave beenshown in bipolar disorder comorbidwith anxiety disorder (Chang et al.,2013). Hemmings et al. (2013) reported that the interaction of BDNFVal66Met and DRD2 Taq1A polymorphisms influences posttraumaticstress disorder (PTSD) symptom severity. An interaction of promotervariants of the cannabinoid receptor 1 gene (CNR1) and 5-HTTLPRaffects the anxious phenotype (Lazary et al., 2009).
From the viewpoint of pharmacogenomics and gene–gene interac-tions, an association between polymorphisms in serotonin transporter(SLC6A4) and serotonin receptor 2A (HTR2A) genes was reported topredict treatment response to venlafaxine XR in generalized anxietydisorder (Lohoff et al., 2013). In addition Lee et al. (2013) reported a sig-nificant interaction effect for the Val/Val carriers of the BDNF Val66Metpolymorphism and the Met carriers of the COMT Val158Met polymor-phism in a comparative study between patients with bipolar II disorderwithout anxiety disorder and controls.
With regard to the gender difference observed in PD, several studieshave investigated BDNF and COMT as representative polymorphisms.Shalev et al. (2009) reported that in male subjects, BDNF Val/Val homo-zygotes showed a greater increase in salivary cortisol than Val/Metheterozygotes. In female subjects, the opposite trend was observed;that is, the Val/Val homozygotes had the lowest increase (Shalev et al.,2009). The female subjects also displayed significantly lower plateletBDNF levels compared to the males. Lommatzsch et al. (2005) foundthat platelet BDNF levels changed during the menstrual cycle. The196G/A (Val66Met) polymorphism of the BDNF gene was revealed tobe significantly associated with the severity of binge eating behaviorin women with bulimia nervosa or other binge eating disorders.
It has been suggested that the role of gender in the relationshipbetween the COMT gene and personality may be due to an interactionbetween estrogen and COMT activity (Harrison and Tunbridge, 2008).An earlier study showed that women with high estrogen levels hadlower COMT activity (Briggs and Briggs, 1973). Estrogen was reportedto inhibit COMT mRNA expression and reduce its activity (Jiang et al.,2003; Xie et al., 1999). COMT plays an important role in metabolizingcatechol estrogens, thereby lowering their levels (Creveling, 2003).Thus, gender differences may interact with the COMT gene to affectpersonality.
As noted above, gender differences in BDNF and COMThave been in-vestigated by many researchers, but the potential interaction of BDNF,COMT and gender is not clear, and no definitive results were obtainedin the studies of the relationship between PD and BDNF and COMT Wesuspected that it would be worthwhile to investigate the gender differ-ences between PD and healthy controls from the viewpoint of gender ×genotype interaction. Based on previous studies, we hypothesized thatthe gene–gene interaction of BDNF Met66 allele and COMT Met158allele would predispose PD individuals to higher anxiety sensitivity,and would show a gender difference.
The aim of our study was to explore the possible association be-tween the gender-specific characteristics of PD-related anxiety traitsand two functional polymorphisms of the BDNF and COMT genes. This
study focused particularly on a gene × gene interaction of BDNF andCOMT.
2. Methods
2.1. Subjects
The study subjects were 470 PD patients (178men and 292women)and 458 healthy controls (195 men and 263 women). The study wasexplained to all subjects prior to their participation, and they each pro-vided written informed consent to participate. In the healthy controlgroup, the inclusion criteria were as follows: drug-free, no previousdiagnosis of a psychiatric disorder, and no family history of psychiatricdisorder. We excluded individuals with a history of major physicalillness, neurological disorder, alcohol abuse, substance abuse, or loss ofconsciousness due to head injury. The healthy control subjects werescreened for the presence or absence of a Diagnostic and StatisticalManual of Mental Disorders, Fourth Edition (DSM-IV) axis I disorderusing the Japanese version of the Mini International NeuropsychiatricInterview 5.0.0 (MINI, 24) (Otsubo et al., 2005).
All of the patients with a DSM-IV diagnosis of PDwere outpatients atthe Nagoya Mental Clinic in Japan, and were diagnosed by at least twodoctors. The MINI is a structured interview used to assess psychiatricillness. Screenings were completed for psychiatric illnesses includingPD. Agoraphobia was found in 65.8% (n = 192) of the female patientsand 51.7% (n = 92) of the male patients. Depression was found in33.5% (n = 98) of the female patients and 27.5% (n = 49) of the malepatients. Bipolar disorder was found in 16.1% (n = 47) of the femalepatients and 11.2% (n = 20) of the male patients. Most of the 470patients were on medication and took antidepressants (n = 401),anxiolytics (n = 357), or a mood stabilizer including antipsychoticdrugs (n= 160). Most of the patients were being treated with an anti-depressant drug such as a serotonin reuptake inhibitor, and the averageimipramine equivalent was 74.4 ± 3.2 mg (standard error: SE) (Bolliniet al., 1999).
This study was approved by the institutional ethics committeesof the Mie University Graduate School of Medicine and the WarakukaiNagoya Mental Clinic.
2.2. Psychological tests
Psychological tests were administered to the participants by ques-tionnaire and by interview. The questionnaire was used to collectclinical information including basic data on family members with PDand genetic factors. The collected data included gender, age, height,weight, blood type, birthplace, growth history, body weight at birth,marital status, drinking habits, smoking habits, menstruation, medicalhealth history, family medical history, and disease under treatment.The questionnaire concerned the individual's experience with PD,including symptoms of first panic attack, frequency of attacks, andavoidance during the previous month of situations in which a panicattack might occur.
The State-Trait Anxiety Inventory (STAI) was administered to all ofthe subjects The STAI has 20 questions to assess state anxiety, and 20for trait anxiety (Nakazato and Mizuguchi, 1982; Spielberger et al.,1983). “Trait anxiety” represents stable individual differences in thepropensity for anxiety, and refers to a tendency to respond to life situa-tions with “general” anxiety. The Revised NEO Personality Inventory(NEO-PI-R) was also administered to all of the subjects The NEO-PI-Ris a standard instrument for measuring the personality traits of individ-uals over a wide range of ages, from the elderly to the young (Costa andMcCrae, 1992; Shimonaka, 1997). The inventory is based on a five-factor model of character: Neuroticism (N), Extroversion (E), Opennessto Experience (O), Agreeableness (A), and Conscientiousness (C). Thesubjects responded to the 240 items on a 5-point scale.
Table1
Compa
risonof
themea
nag
e,ge
nder,C
OMTVal15
8Met,B
DNFVal66
Met
geno
type
distribu
tion
,and
allele
freq
uenc
ybe
twee
nthePD
patien
tan
dco
ntrolg
roup
s.
COMTge
notype
s:n(%
)CO
MTallele:n
(%)
BDNFge
notype
s:n(%
)BD
NFallele:n
(%)
Group
nAge
Mea
n±
SDP⁎
Sex(n
)P⁎
Val/Val
Val/M
etMet/M
etP⁎
Val
Met
P⁎Val/Val
Val/M
etMet/M
etP⁎
Val
Met
P⁎
Controls
458
36.63±
13.96
Men
(195
)10
1(52%
)75
(38%
)19
(10%
)27
7(71%
)11
3(29%
)63
(32%
)94
(48%
)38
(20%
)25
0(60%
)17
0(40%
)W
omen
(263
)11
5(44%
)12
4(47%
)24
(9%)
354(67%
)17
2(33%
)96
(37%
)12
7(48%
)40
(15%
)31
9(61%
)20
7(39%
)To
tal(45
8)21
6(47%
)19
9(43%
)43
(9%)
631(69%
)28
5(31%
)15
9(35%
)22
1(48%
)78
(17%
)56
9(60%
)37
7(40%
)PD
patien
ts47
037
.93±
11.10
0.11
8Men
(178
)0.16
074
(42%
)87
(49%
)17
(9%)
0.11
423
5(66%
)12
1(34%
)0.16
361
(34%
)85
(48%
)32
(18%
)0.89
720
7(58%
)14
9(42%
)0.71
4W
omen
(292
)15
2(52%
)11
5(39%
)25
(9%)
0.13
741
9(72%
)16
5(28%
)0.12
397
(33%
)13
1(45%
)64
(22%
)0.12
932
5(56%
)25
9(44%
)0.10
0To
tal(47
0)22
6(48%
)20
2(43%
)42
(9%)
0.94
165
4(70%
)28
6(30%
)0.78
715
8(34%
)21
6(46%
)96
(20%
)0.41
853
2(57%
)40
8(43%
)0.12
3
Abb
reviations
:COMT=
catech
ol-O
-methy
ltrans
ferase;B
DNF=
brain-de
rive
dne
urotroph
icfactor;V
al=
valin
e;Met
=methion
ine.
PD=
panicdisorder;S
D=
stan
dard
deviation.
⁎Re
lative
tothecontrolg
roup
.
121Y. Konishi et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 51 (2014) 119–125
2.3. SNP genotyping
For each subject, a 7-mL blood sample was centrifuged at2000 rpm × 10 min in a collection tube and separated into the plasmalayer, buffy-coat layer (containing white blood corpuscles), andred blood corpuscle layer. A 150-μL sample of the buffy coat was usedfor automatic blood DNA extraction. DNA was extracted using aBioRobot EZ1 (Qiagen, Hilden, Germany). BDNF Val66Met (rs6265(assay ID C__11592758_10)) and COMT Val158Met (rs4680 (AssayID C__25746809_50)) polymorphisms were genotyped with allelicdiscrimination using the TaqMan real-time polymerase chain reac-tion (PCR) assay (Life Technologies Japan, Tokyo). The assay in-cludes a sequence-specific fluorogenic minor groove binder probefor each allele. Each probe is 5′-labeled with a different reporterfluorescent dye (4,7,2′-trichloro-7′-phenyl-6-carboxyfluoresceinand 6-carboxyfluorescein) to differentiate the amplification of eachallele. Primer–probe sets for the detection of the polymorphismswere purchased from Applied Biosystems (Foster City, CA, USA).PCR was carried out on an ABI StepOne+™ Real-Time PCR Systemaccording to the manufacturer's protocol (Applied Biosystems).
2.4. Statistical analyses
The Hardy–Weinberg equilibrium for genotyping frequencies wascalculated using χ2 tests. Statistical calculations were carried out usingSPSS for Windows (Release 17.0)®. A multivariate analysis of covari-ance (MANCOVA) was performed with STAI and NEO-PI-R scoresas the dependent factor, and health status (PD patients and control sub-jects), gender and genotyping groups (BDNF andCOMT) as fixed factors,and the covariate of age. As a sub-analysis, we performed a MANCOVAwith STAI and NEO-PI-R scores as the dependent factor, gender andgenotyping groups (BDNF and COMT) as fixed factors, and the covariateof age in the PD patients and healthy controls. Post hoc MANCOVA testswere conducted to evaluate COMT × BDNF interactions.
The statistical power of effect size was used for comparative effectsin the power analysis. Comparisons of gender-linked genotype distribu-tions and allele frequencies between PD and healthy controls were per-formedby theχ2 test. Student's t-testwas used for differences in age. Alltests were performed with a two-tailed type-I error rate of p b 0.05.
3. Results
The average age of the patients in the PD group was 37.93 ±11.10 years, and that of the healthy control group was 36.63 ±13.96 years (t = 1.566, p = 0.118). The gender ratios of the PD group(178 men, 292 women) and controls (195 men and 263 women)showed no significant difference (χ2 value = 1.942, p = 0.163)(Table 1).
The COMT Val158Met polymorphism followed Hardy–Weinbergequilibrium in each subgroup (p= 0.742 for PD, p= 0.770 for healthycontrol; p = 0.289 for male, PD, p = 0.360 for male healthy control;p = 0.356 for female PD, p = 0.248 for female healthy control).The Hardy–Weinberg equilibrium also applied to the BDNF Val66Metpolymorphism in each subgroup (p = 0.161 for PD, p = 0.935for healthy control; p = 0.800 for male PD, p = 0.782 for male healthycontrol; p = 0.119 for female, PD, p = 0.850 for female healthycontrol). No significant difference was observed in the genotypic orallele frequencies of COMT and BDNF polymorphisms between the PDand healthy control groups (Table 1). Equivalent doses to that of thestandard imipramine did not differ significantly among the COMTVal158Met or BDNF Val66Met polymorphism groups (all, male, orfemale PD patients; Supplementary Table 1).
There was a significant health status effect on STAI and NEO-PI-Rscores, and there were significant interactions between healthy status,gender, COMT and BDNF polymorphisms on the STAI, Neuroticism andExtraversion scores (Table 2).
Table 2Effects of health status, gender, COMT polymorphism and BDNF polymorphism on STAI and NEO-PI-R scores.
Category of groups Response variables
STAI state STAI trait Neuroticism Extraversion Openness Agreeableness Consciousness
Health status (HS) b0.001⁎⁎⁎ b0.001⁎⁎⁎ b0.001⁎⁎⁎ b0.001⁎⁎⁎ b0.001⁎⁎⁎ 0.024⁎ 0.001⁎⁎Gender (G) 0.989 0.180 0.001⁎⁎ 0.063 0.675 b0.001⁎⁎⁎ 0.617COMT polymorphism (CP) 0.546 0.362 0.940 0.960 0.811 0.738 0.370BDNF polymorphism (BP) 0.536 0.275 0.814 0.107 0.893 0.392 0.567HS × G 0.798 0.789 0.942 0.328 0.259 0.929 0.663HS × CP 0.001⁎⁎ 0.034⁎ 0.007⁎⁎ 0.008⁎⁎ 0.004⁎⁎ 0.999 0.109HS × BP 0.679 0.131 0.056 0.551 0.629 0.053 0.035⁎BP × CP 0.774 0.896 0.953 0.39 0.365 0.671 0.552HS × G × CP 0.051 0.098 0.216 0.006⁎⁎ 0.027⁎ 0.585 0.413HS × G × BP 0.465 0.640 0.938 0.742 0.716 0.976 0.396HS × CP × BP 0.021⁎ 0.051 0.045⁎ b0.001⁎⁎⁎ 0.045⁎ 0.666 0.141G × CP × BP 0.058 0.300 0.212 0.008⁎⁎ 0.777 0.890 0.603HS × G × CP × BP 0.042⁎ 0.022⁎ 0.030⁎ 0.001⁎⁎ 0.938 0.723 0.122
Abbreviations: COMT= catechol-O-methyltransferase; BDNF = brain-derived neurotrophic factor; STAI = State-Trait Anxiety Inventory; NEO-PI-R = Revised NEO Personality Inventory;MANCOVA was performed with STAI and NEO-PI-R scores as dependent factor, and health status, gender and genotyping groups as fixed factors and covariate of age.⁎ P b 0.05.
⁎⁎ P b 0.01.⁎⁎⁎ P b 0.001.
122 Y. Konishi et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 51 (2014) 119–125
Following the sub-analysis, Tables 3 and 4 summarize the inter-action of gene × gender revealed by the MANCOVA on STAI and NEO-PI-R scores. In the PD patients, BDNF × COMT× gender showed a signif-icant effect on STAI state (F = 5.70, p = 0.018, power = 0.662), STAItrait (F = 4.47, p = 0.035, power = 0.559), and the Neuroticism(F = 5.22, p = 0.023, power = 0.625), and Extraversion scores(F = 12.3, p = 0.001, power = 0.938). No interactive effects ofthe BDNF × COMT × gender interaction on STAI or NEO-PI-R scoreswere detected in the healthy controls, however.
Particularly in the male PD patients, our results indicated a signifi-cant gene × gene interaction by a post hoc MANCOVA on STAI stateand the Neuroticism and Extraversion scores (Fig. 1, Suppl. Table 2).The post hoc MANCOVA revealed that among the male PD patients,COMT Met + BDNF Val/Val carriers had significantly higher STAI stateand Neuroticism scores than the COMT Val/Val+ BDNF Val/Val carriers(STAI state; p = 0.009, Neuroticism; p = 0.014). Among the male PDpatients, the COMT Val/Val + BDNF Val/Val carriers and the COMTMet + BDNF Met carriers also had significantly higher Extraversionscores than the COMT Met + BDNF Val/Val carriers (p b 0.001, p =0.007). Finally, among the male PD patients, the COMT Val/Val +BDNF Val/Val carriers had significantly higher Extraversion scores thanthe COMT Val/Val + BDNF Met carriers (p = 0.048).
Table 3Effects of gender, BDNF polymorphism, and COMT polymorphism on STAI scores in the PD and
Panic disorder
F P
STAI state Gender (G) 0.05 0.822BDNF polymorphism (BP) 0.56 0.456COMT polymorphism (CP) 4.72 0.031G × BP 0.27 0.604G × CP 2.53 0.113BP × CP 1.15 0.283G × BP × CP 5.70 0.018
STAI trait Gender (G) 0.69 0.408BDNF polymorphism (BP) 2.71 0.101COMT polymorphism (CP) 0.56 0.456G × BP 0.00 0.972G × CP 3.55 0.061BP × CP 1.32 0.251G × BP × CP 4.47 0.035
Abbreviations: BDNF = brain-derived neurotrophic factor; COMT= catechol-O-methyltransfeMANCOVA was performed with STAI scores as dependent factor, gender and genotyping group⁎ P b 0.05.⁎⁎ P b 0.01.
4. Discussion
Our results showed that BDNF × COMT × gender had a significanteffect on STAI state, STAI trait, and the NEO-PI-R Neuroticism andExtraversion scores in PD patients. No interactive effects of theBDNF × COMT × gender interaction on STAI or NEO-PI-R scoreswere detected in the healthy controls. The gene effect was observedparticularly in the men with PD.
As shown in previous studies of personality traits in anxiety disor-ders, we found that higher anxiety scores are associated with higherNeuroticism (Ormel et al., 2004) and lower Extraversion scores (Jylhaet al., 2009). In the present study, the male PD patients who wereCOMT Met carriers had higher STAI state and Neuroticism scores andlower Extraversion scores than the COMT Val/Val carriers, especiallyin the BDNF Val/Val carrier group (Fig. 1), corresponding to oppositegender effects on BDNF Val/Val homozygotes (Shalev et al., 2009), andthese effects were enhanced in COMT Met carriers.
The results of our study suggest the possibility of BDNF–COMT–gender involvement in anxiety traits, personality, and panic disorder,without significant differences of genotype distribution related togene–gene interaction. Gender differences may interact with the BDNFand COMT genes' effects on anxiety sensitivity and related personality
healthy control groups.
Healthy control
Power F P Power
0.056 0.03 0.869 0.0530.115 0.27 0.601 0.082
⁎ 0.581 6.81 0.009⁎⁎ 0.7400.081 0.47 0.496 0.1040.354 1.86 0.173 0.2750.188 6.24 0.013⁎ 0.702
⁎ 0.662 0.01 0.942 0.0510.131 0.80 0.372 0.1450.375 0.01 0.921 0.0510.116 7.96 0.005⁎⁎ 0.8040.050 0.43 0.513 0.1000.467 0.09 0.763 0.0600.209 2.65 0.104 0.368
⁎ 0.559 1.02 0.312 0.172
rase; STAI = State-Trait Anxiety Inventory.s as fixed factors and covariate of age.
Table 4Effects of gender, BDNF polymorphism, and COMT polymorphism on NEO-PI-R (Neuroticism and Extraversion) scores in the PD and healthy control groups.
Panic disorder Healthy control
F P Power F P Power
Neuroticism Gender (G) 4.05 0.045⁎ 0.518 7.86 0.005⁎⁎ 0.799BDNF polymorphism (BP) 2.12 0.146 0.306 2.18 0.141 0.313COMT polymorphism (CP) 2.79 0.096 0.384 5.09 0.025⁎ 0.614G × BP 0.08 0.776 0.059 0.17 0.682 0.069G × CP 4.17 0.042⁎ 0.530 0.23 0.630 0.077BP × CP 1.59 0.208 0.242 2.36 0.125 0.335G × BP × CP 5.22 0.023⁎ 0.625 0.43 0.514 0.100
Extraversion Gender (G) 2.66 0.104 0.369 0.68 0.412 0.130BDNF polymorphism (BP) 0.42 0.515 0.100 3.60 0.059 0.473COMT polymorphism (CP) 2.47 0.117 0.347 5.23 0.023⁎ 0.626G × BP 0.15 0.695 0.068 0.00 0.976 0.050G × CP 7.20 0.008⁎⁎ 0.762 0.52 0.469 0.112BP × CP 7.13 0.008⁎⁎ 0.758 5.96 0.015⁎ 0.683G × BP × CP 12.3 0.001⁎⁎ 0.938 0.59 0.445 0.119
Abbreviations: BDNF=brain-derived neurotrophic factor; COMT= catechol-O-methyltransferase; NEO-PI-R= RevisedNEOPersonality Inventory;MANCOVAwas performedwith STAIscores as dependent factor, gender and genotyping groups as fixed factors and covariate of age.⁎ P b 0.05.⁎⁎ P b 0.01.
123Y. Konishi et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 51 (2014) 119–125
characteristics in the pathogenesis of this disorder through the mono-amine system comprised of serotonergic, norepinephrine and dopami-nergic systems.
Previous studies of the potential involvement of gene–gene interac-tions with PD examined COMT and serotonin receptor polymorphisms.Freitag et al. (2006) reported a nominally significant interaction be-tween the serotonin receptor 1A and COMT polymorphisms. Interactionof serotonin-2A receptor and COMT polymorphism (Karacetin et al.,2012) was reported with no significant interaction. Gene–gene interac-tions in PD have not been extensively studied.
*
STA
I-st
ate
Ext
rave
rsio
n
**
*****
60
120
100
80
60
40
20
0
BDNF Val/ValBDNF MetCarrier
BDNF Val/ValBDNF MetCarrier
50
40
30
20
10
0COMT Val/Val
m f
COMT MetCarrier
COMT Val/Val
m
COMT MetCarrier
COMT Val/Val
f
COMT MetCarrier
COMT Val/Val COMT MetCarrier
Fig. 1. Comparison between genotype groups with post hoc test in male and female PD pati(COMT Val/Val + BDNF Val/Val carriers = 40.2 ± 1.8, COMT Met + BDNF Val/Val carriers =carriers = 98.5 ± 4.5, COMT Met + BDNF Val/Val carriers = 117.2 ± 6.8), and the NEO-PI-RVal/Val + BDNF Met carriers = 88.1 ± 4.1, COMT Met + BDNF Val/Val carriers = 73.6 ± 5.1⁎p b 0.05, ⁎⁎p b 0.01, ⁎⁎⁎p b 0.001. PD, panic disorder; COMT, catechol-O-methyltransferase;Anxiety Inventory; m, males; f, females.
Most brain imaging studies of patients with PD have consistentlyshown abnormalities in the limbic system and ventromedial prefrontalcortex associated with emotional arousal (Gorman et al., 2000). In astudy of patients with PD, we observed hypoactivation in the lateralprefrontal cortex (PFC), which is responsible for executive functionthrough the dopaminergic system (Nishimura et al., 2007). A hypothet-ical invertedU-shaped curvewas proposed for the relationship betweenworking memory performance and prefrontal dopaminergic activity(Goldman-Rakic et al., 2000). In the present study, significant geneticeffects by COMT genotype were observed in healthy control subjects
Neu
rotic
ism
*140
120
100
80
60
40
20
0
BDNF Val/ValBDNF MetCarrier
COMT Val/Val
m
COMT MetCarrier
COMT Val/Val
f
COMT MetCarrier
ents. A gene effect was observed, particularly in the male PD patients, on the STAI state50.8 ± 2.6), the NEO-PI-R factor Neuroticism scores (COMT Val/Val + BDNF Val/Val
factor Extraversion scores (COMT Val/Val + BDNF Val/Val carriers = 93.5 ± 2.7, COMT, COMT Met + BDNF Met carriers = 93.5 ± 2.7). Error bars: standard error of the mean.BDNF, brain-derived neurotrophic factor; Val, valine; Met, methionine; STAI, State-Trait
124 Y. Konishi et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 51 (2014) 119–125
(Tables 3, 4; Suppl. Table 3), but the genotype effects of COMTwere theopposite between the PD patients and the healthy controls. The modu-lation of the COMT genotype on anxiety-related parameters seems to bedifferent in PD patients.
Our present data on the distinct effect of the COMT genotype couldbe explained by the application of an inverted U-shaped hypothesisregarding the prefrontal dopamine level. In accord with this invertedU-shaped hypothesis, the PFC dopamine levels of Val158 homozygotesand Met carriers fell into an inverted U-shaped curve in accord withthe increase/decrease of COMT activity and showed more complexresults than the simple ‘warrior/worrier’ hypothesis of COMT genotype(Mier et al., 2010). As prefrontal dopamine could not be measureddirectly, this speculation is only a working hypothesis, but it is thoughtthat anxiety sensitivity might be indirectly related to PD through thedopamine neurotransmission of the PFC.
There are few previous studies of gene–gene interactions relatedto gender differences. Niewold et al. (2010) reported a gene–gene–sexinteraction in cytokine gene polymorphisms in juvenile dermatomyositis;to the best of our knowledge, such an interaction has not yet been re-ported in psychiatric disease. We studied the gene × gene × genderinteraction of BDNF and COMT genotypes associated with panic dis-order, andwe suspect that the gender specificity of COMT Val158Metinfluences personality traits in male PD patients. Whether the inter-action of the BDNF and COMT genes leads to an elevation of anxietysensitivity and how such an interaction affects the etiology of PDrequires further clarification.
The present study has several limitations. First, only one representa-tive functional SNPwithin the BDNF and COMTwas examined. It wouldhave beendesirable to study all tagging SNPswithin BDNF andCOMT, aswell as other variants associatedwith dopaminergic neurotransmission.Second, a self-reported questionnaire assessment measure was usedto evaluate personality-related PD symptoms. Additional assessmentsby clinicians may be more accurate and/or reliable. The personalityinventory scores may reflect effects from the symptoms, and the STAIstate may correspond to the severity of the symptoms. Further datafor the severity of symptoms of patients with PD are necessary. Third,we included PD patients whose comorbidities included agoraphobia,depression or bipolar disorder as a composite group of PD. For replica-tion studies, it is desirable to collect more detailed information and ahomogeneous group with larger sample sizes that could lead to morespecific evidence about anxiety disorders such as PD.
5. Conclusion
Gene × gene × gender interactionswere observed in the associationof the BDNF and COMT polymorphisms with STAI scores and with theNEO-PI-R Neuroticism and Extraversion scores in the male PD group.Our results show the possibility of a gender effect in anxiety disorderssuch as PD, which has a higher prevalence rate in women. The BDNFandCOMT interaction appears to exhibit gender differences in personal-ity scores, with potential implications for gender differences in thepathogenesis of PD.
Contributors
Y.K. executed the experimental protocols and statistical analyses,and wrote the first draft of the manuscript. Y.K., Y.O., and H.T. were re-sponsible for the study concept and design. M.T., T.U., E.M., T.S., H.K.and Y.O. contributed to the acquisition of clinical data. Y.K., E.M. andM.O. conducted the data analysis. T.O., T.S. and M.O. interpreted thefindings and provided critical revisions of the manuscript.
Acknowledgments
This work was supported by KAKENHI (a Grant-in-Aid for ScientificResearch) in the Priority Area of Applied Genomics from the Ministry
of Education, Culture, Sports, Science and Technology of Japan,(No. 17019029; No. 21591483) awarded to H. Tanii and Y. Okazaki.We thank the staff of the Department of Psychiatry at theMie UniversityGraduate School of Medicine and the Nagoya Mental Clinic for theirenthusiastic cooperation. We are grateful to all of the PD patients andothers who answered the questionnaire and allowed their blood to besampled.
Appendix A. Supplementary data
Supplementary data to this article can be found online at http://dx.doi.org/10.1016/j.pnpbp.2014.01.020.
References
Bollini P, Pampallona S, Tibaldi G, Kupelnick B,Munizza C. Effectiveness of antidepressants.Meta-analysis of dose–effect relationships in randomised clinical trials. Br J Psychiatry1999;174:297–303.
Briggs MH, Briggs M. Hormonal influences on erythrocyte catechol-O-methyl transferaseactivity in humans. Experientia 1973;29:278–80.
Campbell-Sills L, Cohan SL, Stein MB. Relationship of resilience to personality, coping, andpsychiatric symptoms in young adults. Behav Res Ther 2006;44:585–99.
Chang YH, Lee SY, Chen SL, Tzeng NS, Wang TY, Hui Lee I, et al. Genetic variants of theBDNF and DRD3 genes in bipolar disorder comorbid with anxiety disorder. J AffectDisord 2013;151:967–72.
Costa Jr PT, McCrae RR. Revised NEO Personality Inventory (NEO PI-R) and NEOFive-Factor Inventory (NEO-FFI) professional manual. Odessa, Fl: PsychologicalAssessment Resources; 1992.
Creveling CR. The role of catechol-O-methyltransferase in the inactivation ofcatecholestrogen. Cell Mol Neurobiol 2003;23:289–91.
Crowe RR, Noyes R, Pauls DL, Slymen D. A family study of panic disorder. Arch GenPsychiatry 1983;40:1065–9.
Domschke K, Freitag CM, Kuhlenbaumer G, Schirmacher A, Sand P, Nyhuis P, et al.Association of the functional V158M catechol-O-methyl-transferase polymorphismwith panic disorder in women. Int J Neuropsychopharmacol 2004;7:183–8.
Domschke K, Deckert J, O'Donovan MC, Glatt SJ. Meta-analysis of COMT val158met inpanic disorder: ethnic heterogeneity and gender specificity. Am J Med Genet BNeuropsychiatr Genet 2007;144B:667–73.
Elzinga BM, Molendijk ML, Oude Voshaar RC, Bus BA, Prickaerts J, Spinhoven P, et al. Theimpact of childhood abuse and recent stress on serum brain-derived neurotrophicfactor and the moderating role of BDNF Val66Met. Psychopharmacology (Berl)2011;214:319–28.
Freitag CM, Domschke K, Rothe C, Lee YJ, Hohoff C, Gutknecht L, et al. Interaction ofserotonergic and noradrenergic gene variants in panic disorder. Psychiatr Genet2006;16:59–65.
Goldman-Rakic PS, Muly 3rd EC, Williams GV. D(1) receptors in prefrontal cells andcircuits. Brain Res Brain Res Rev 2000;31:295–301.
Gorman JM, Kent JM, Sullivan GM, Coplan JD. Neuroanatomical hypothesis of panicdisorder, revised. Am J Psychiatry 2000;157:493–505.
Harrison PJ, Tunbridge EM. Catechol-O-methyltransferase (COMT): a gene contributing tosex differences in brain function, and to sexual dimorphism in the predisposition topsychiatric disorders. Neuropsychopharmacology 2008;33:3037–45.
Hemmings SM, Martin LI, Klopper M, van der Merwe L, Aitken L, de Wit E, et al. BDNFVal66Met and DRD2 Taq1A polymorphisms interact to influence PTSD symptomseverity: a preliminary investigation in a South African population. ProgNeuropsychopharmacol Biol Psychiatry 2013;40:273–80.
Hettema JM, Neale MC, Kendler KS. A review and meta-analysis of the genetic epidemiol-ogy of anxiety disorders. Am J Psychiatry 2001;158:1568–78.
Jiang H, Xie T, Ramsden DB, Ho SL. Human catechol-O-methyltransferase down-regulation by estradiol. Neuropharmacology 2003;45:1011–8.
Jylha P, Melartin T, Isometsa E. Relationships of neuroticism and extraversion with axis Iand II comorbidity among patients with DSM-IV major depressive disorder. J AffectDisord 2009;114:110–21.
Kang JI, Kim SJ, Song YY, Namkoong K, An SK. Genetic influence of COMT and BDNF genepolymorphisms on resilience in healthy college students. Neuropsychobiology2013;68:174–80.
Karacetin G, Bayoglu B, Cengiz M, Demir T, Kocabasoglu N, Uysal O, et al. Serotonin-2Areceptor and catechol-O-methyltransferase polymorphisms in panic disorder. ProgNeuropsychopharmacol Biol Psychiatry 2012;36:5–10.
Lazary J, Lazary A, Gonda X, Benko A, Molnar E, Hunyady L, et al. Promoter variants of thecannabinoid receptor 1 gene (CNR1) in interaction with 5-HTTLPR affect the anxiousphenotype. Am J Med Genet B Neuropsychiatr Genet 2009;150B:1118–27.
Lee SY, Chen SL, Wang YS, Chang YH, Huang SY, Tzeng NS, et al. COMT and BDNFinteracted in bipolar II disorder not comorbid with anxiety disorder. Behav BrainRes 2013;237:243–8.
Lohoff FW, Narasimhan S, Rickels K. Interaction between polymorphisms in serotonintransporter (SLC6A4) and serotonin receptor 2A (HTR2A) genes predict treatmentresponse to venlafaxine XR in generalized anxiety disorder. Pharmacogenomics J2013;13:464–9.
Lommatzsch M, Zingler D, Schuhbaeck K, Schloetcke K, Zingler C, Schuff-Werner P, et al.The impact of age, weight and gender on BDNF levels in human platelets and plasma.Neurobiol Aging 2005;26:115–23.
125Y. Konishi et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 51 (2014) 119–125
McGrath M, Kawachi I, Ascherio A, Colditz GA, Hunter DJ, De Vivo I. Associationbetween catechol-O-methyltransferase and phobic anxiety. Am J Psychiatry2004;161:1703–5.
Mier D, Kirsch P, Meyer-Lindenberg A. Neural substrates of pleiotropic action of geneticvariation in COMT: a meta-analysis. Mol Psychiatry 2010;15:918–27.
Monteleone P, Zanardini R, Tortorella A, Gennarelli M, Castaldo E, Canestrelli B, et al. The196G/A (val66met) polymorphism of the BDNF gene is significantly associated withbinge eating behavior in women with bulimia nervosa or binge eating disorder.Neurosci Lett 2006;406:133–7.
Nakazato K, Mizuguchi T. How to use STAI. Kyoto: Sankyoubou Corp.; 1982.Niewold TB, Kariuki SN, Morgan GA, Shrestha S, Pachman LM. Gene–gene–sex interaction
in cytokine gene polymorphisms revealed by serum interferon alpha phenotype injuvenile dermatomyositis. J Pediatr 2010;157:653–7.
Nishimura Y, Tanii H, Fukuda M, Kajiki N, Inoue K, Kaiya H, et al. Frontal dysfunctionduring a cognitive task in drug-naive patients with panic disorder as investigatedby multi-channel near-infrared spectroscopy imaging. Neurosci Res 2007;59:107–12.
Ormel J, Rosmalen J, Farmer A. Neuroticism: a non-informative marker of vulnerability topsychopathology. Soc Psychiatry Psychiatr Epidemiol 2004;39:906–12.
Otsubo T, Tanaka K, Koda R, Shinoda J, Sano N, Tanaka S, et al. Reliability and validity ofJapanese version of the Mini-International Neuropsychiatric Interview. PsychiatryClin Neurosci 2005;59:517–26.
Shalev I, Lerer E, Israel S, Uzefovsky F, Gritsenko I, Mankuta D, et al. BDNF Val66Met poly-morphism is associated with HPA axis reactivity to psychological stress characterizedby genotype and gender interactions. Psychoneuroendocrinology 2009;34:382–8.
Shimonaka J. On the validity and reliability of NEO-PI: Japanese version. Annual Report ofScientific Investigation. Japanese Ministry of Education; 1997.
Spielberger CD, Gorsuch RL, Lushene R, Vagg PR, Jacobs GA. Manual for the State-TraitAnxiety Inventory. Palo Alto CA: Consulting Psychology Press; 1983.
WeissmanMM. Family genetic studies of panic disorder. J Psychiatr Res 1993;27(Suppl. 1):69–78.
Xie T, Ho SL, Ramsden D. Characterization and implications of estrogenic down-regulationof human catechol-O-methyltransferase gene transcription. Mol Pharmacol 1999;56:31–8.
