Journal of Psychiatric Research xxx (2014) 1e8

Contents lists available at ScienceDirect

Journal of Psychiatric Research journal homepage: www.elsevier.com/locate/psychires

Association of COMT and PRODH gene variants with intelligence quotient (IQ) and executive functions in 22q11.2DS subjectsq Miri Carmel a, b, *, Omer Zarchi a, c, f, Elena Michaelovsky a, b, Amos Frisch a, b, Miriam Patya a, b, Tamar Green a, c, d, Doron Gothelf a, c, Abraham Weizman a, b, e a

Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel Felsenstein Medical Research Center, Petah Tikva, Israel The Child Psychiatry Unit, Edmond and Lily Safra Children’s Hospital, Sheba MedicalCenter, Tel Hashomer, Ramat Gan, Israel d Nes-Ziyyona-Beer Yaakov Mental Health Center, Beer Yaakov, Israel e Geha Mental-Health Center, Petah Tikva, Israel f Rabin Medical Center, 49100 Petah Tikva, Israel b c

a r t i c l e i n f o

a b s t r a c t

Article history: Received 27 January 2014 Received in revised form 23 April 2014 Accepted 24 April 2014

The 22q11.2 deletion syndrome (22q11.2DS) carries the highest genetic risk factor for the development of schizophrenia. We investigated the association of genetic variants in two schizophrenia candidate genes with executive function (EF) and IQ in 22q11.2DS individuals. Ninety two individuals with 22q11.2 deletion were studied for the genetic association between COMT and PRODH variants and EF and IQ. Subjects were divided into children (under 12 years old), adolescents (between 12 and 18 years old) and adults (older than 18 years), and genotyped for the COMT Val158Met (rs4680) and PRODH Arg185Trp (rs4819756) polymorphisms. The participants underwent psychiatric evaluation and EF assessment. Our main finding is a significant influence of the COMT Val158Met polymorphism on both IQ and EF performance. Specifically, 22q11.2DS subjects with Met allele displayed higher IQ scores in all age groups compared to Val carriers, reaching significance in both adolescents and adults. The Met allele carriers performed better than Val carriers in EF tasks, being statistically significant in the adult group. PRODH Arg185Trp variant did not affect IQ or EF in our 22q11.2DS cohort. In conclusion, functional COMT variant, but not PRODH, affects IQ and EF in 22q11.2DS subjects during neurodevelopment with a maximal effect at adulthood. Future studies should monitor the cognitive performance of the same individuals from childhood to old age. Ó 2014 Elsevier Ltd. All rights reserved.

Keywords: 22q11.2DS Intelligence quotient (IQ) Executive function (EF) COMT

1. Introduction Common variants in several psychiatric risk genes were found to predict brain structure at birth which may interfere with normal neurodevelopment, affect cognitive functions and predispose to psychiatric pathologies (Knickmeyer et al., 2014). One of the major candidates among these genes is the catechol-O-methyltransferase (COMT) gene the variants of which were implicated in neurobehavioral phenotypes in healthy and unhealthy individuals. The

enzyme coded by the COMT gene participates in the inactivation of catecholamines such as dopamine, and its main pro-cognitive effect is in the prefrontal cortex. The most widely studied variation in the COMT gene is a functional single-nucleotide polymorphism (rs4680) coding for Val158Met. The Met variant has significantly lower COMT activity than the Val allele. It is thought that this COMT activity affects human prefrontal cortical functions such as cognition via dopamine neurotransmission (Rasetti and Weinberger, 2011; Gaysina et al., 2013). 1.1. COMT and cognition in healthy individuals

q Work was performed at the Felsenstein Medical Research Center, Petah Tikva, Israel and the Behavioral Neurogenetics Center, Sheba Medical Center. * Corresponding author. Felsenstein Medical Research Center (FMRC), Sackler Faculty of Medicine, Tel Aviv University, Rabin Medical Center, 49100 Petah Tikva, Israel. Tel.: þ972 3 9376793; fax: þ972 3 9211478. E-mail addresses: [email protected], [email protected], miri.carmel@ gmail.com (M. Carmel).

The effect of the COMT gene and especially the Val158Met polymorphism on IQ and executive function (EF) in healthy individuals has been extensively studied (Barnett et al., 2008; Squarcione et al., 2013). A range of studies have shown that the Val158Met polymorphism has a small but significant impact on

http://dx.doi.org/10.1016/j.jpsychires.2014.04.019 0022-3956/Ó 2014 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Carmel M, et al., Association of COMT and PRODH gene variants with intelligence quotient (IQ) and executive functions in 22q11.2DS subjects, Journal of Psychiatric Research (2014), http://dx.doi.org/10.1016/j.jpsychires.2014.04.019

2

M. Carmel et al. / Journal of Psychiatric Research xxx (2014) 1e8

prefrontal cognitive performance, with the Met allele-carrying individuals performing better than Val allele carriers (Barnett et al., 2007b, 2009; Farrell et al., 2012; Parasuraman and Jiang, 2012). On the other hand other studies did not replicate these findings (Blanchard et al., 2011; Diaz-Asper et al., 2008). 1.2. COMT and cognition in 22q11.2DS individuals The 22q11.2 deletion syndrome (22q11.2DS) which is characterized by a hemizygous deletion in chromosome 22 leaving only one copy of about 60 genes, among them COMT, is an interesting model to study the role of COMT variants and other genes that reside in the deleted region on cognition (Karayiorgou et al., 2010). These 22q11.2DS individuals are expected to have lower COMT activity than healthy individuals because they have only one allele (Val or Met) while normal individuals have three possible genotypes (Val/Val, Val/Met, Met/Met). Few studies have examined the effect of COMT genotype on cognitive functions and found that 22q11.2DS subjects carrying the Met allele performed significantly better than those carrying the Val allele in IQ test and EF tasks (Bearden et al., 2004; Shashi et al., 2006). However, other studies failed to demonstrate differences in these COMT allele groups (Baker et al., 2005; Glaser et al., 2006a; Barnett et al., 2008). 1.3. PRODH Another gene located on chromosome 22 and is hemizygously deleted in 22q11.2DS is PRODH. PRODH encodes for proline dehydrogenase, a mitochondrial rate-limiting enzyme in the proline degradation process. Homozygous mutations in the PRODH gene lead to hyperprolinemia type I, a rare neurologic disorder with variable manifestations such as seizures, mental retardation, psychiatric and behavioral disorders. Hemizygous deletions of PRODH in 22q11.2DS have been found to be associated with elevated proline levels (Drew et al., 2011). PRODH has a considerable number of functional SNPs. Multiple studies in human and mouse models of 22q11.2DS indicate that PRODH polymorphisms are associated with the risk of schizophrenia or schizophrenia-like behavior, although their role in the pathogenesis is not clear. Some of the common functional variants of PRODH have been characterized and it was found that the SNPs influence a wide range of enzymatic activity from a decrease to below 30% of the activity to an increase of 120%. We focused on rs4819756 (Arg185Trp) located on exon 5 because it was reported that this SNP alters the PRODH enzymatic activity by the reduction of 30e70% (Bender et al., 2005). Raux et al. (2007) reported that 22q11.2DS patients with severe hyperprolinemia performed significantly worse on a large number of cognitive tasks and exhibited a higher prevalence of psychosis compared to other 22q11.2DS subjects. 1.4. Interaction between COMT and PRODH Since both COMT and PRODH variants have been implicated in modulating cognitive functions and susceptibility to psychiatric manifestations in healthy and 22q11.2DS individuals, it was interesting to examine their combined effect on our 22q11.2DS subjects. The reduction of COMT activity may lead to increased dopamine availability (Bender et al., 2012; Witte and Flöel, 2012) while the reduction of PRODH activity may increase the presence of proline leading to the elevation in glutamatergic signaling in the hippocampus causing a release of dopamine in the prefrontal cortex (PFC) (Vorstman et al., 2009; Paterlini et al., 2005).

The effect of interactions between COMT and PRODH on schizophrenia has been investigated. Paterlini et al. (2005) reported that the reduced enzymatic activities of COMT and PRODH cause an increase in dopamine activity that may predispose to psychosis and schizophrenia. Raux et al. (2007) found that in the PRODH deficient mouse model possessing both hyperprolinemia and the Met-COMT allele the animals were at risk for a broad spectrum of psychotic disorders. It was shown in a mouse model, that alterations in COMT activity in the PFC may affect the GABA signaling-related genes, a system relevant to the pathophysiology of schizophrenia (Kimoto et al., 2012). The goal of this study was to assess the effect of the functional COMT and PRODH SNPs, rs4680 and rs4819756, jointly and separately, on cognitive capacity in 22q11.2DS subjects. We attempted to further clarify the possible role of these variations on cognitive functions, IQ and executive function, that are relevant to the phenotype of schizophrenia and other mental disorders associated with cognitive deficits in 22q11.2DS individuals. 2. Methods 2.1. Subjects Ninety two subjects with 22q11.2DS were recruited from the Behavioral Neurogenetics Center, a large tertiary referral center in Israel. The 92 participants were divided into three age groups (Table 1). The study protocol was approved by the Institutional Review Board. Written informed consent was obtained from all participants and/or their parents after the nature of this study was explained to the subjects and their parents or guardians. 2.2. Psychiatric assessment 22q11.2DS subjects and their parents were interviewed by a child psychiatrist (DG and TG) using the Hebrew version of the Schedule for Affective Disorders and Schizophrenia for School-Age Children, Present and Lifetime version (K-SADS-PL). The adults were interviewed with the Structured Clinical Interview for Axis I DSM-IV (SCID). All diagnoses were established according to DSMIV-TR. 2.3. IQ and EF assessments IQ was measured using the age-appropriate versions of WISCIII and WAIS-III (Caplan et al., 1997; Wechsler, 1991). EF evaluation was assessed by the Flanker Fish Tasks (FF) as previously described in details (Diamond et al., 2007; Zarchi et al., 2013b). 2.4. Genotyping Diagnosis of all subjects with 22q11.2DS was confirmed by the fluorescence in situ hybridization test and the multiplex ligationdependent probe amplification technique. COMT Val158Met polymorphism (rs4680) was genotyped by the C25746809-50 TaqMan kit (Applied Biosystems Incorporated, Foster City, CA) using the ABI 7000 instrument. Results were validated by RFLP using the NlaIII restriction enzyme (Daniels et al., 1996). PRODH-exon 5 Arg185Trp polymorphism (rs4819756) was genotyped by amplification of a 409 bp fragment (primers: F:50 caaggccactatgcttggag30 ; R:50 aacagtgagggacccaagtg30 ) followed by digestion by BseNI and analysis by gel electrophoresis. According to NCBI dbSNP (Build 37.5) the global minor allele frequency (MAF) for COMT: A ¼ 0.390 and PRODH: A ¼ 0.257. In our 22q11.2DS cohort 50 individuals had the COMT Met (A) allele (50/ 92 ¼ 0.543) and 23 had the PRODH Trp (A) allele (23/92 ¼ 0.25).

Please cite this article in press as: Carmel M, et al., Association of COMT and PRODH gene variants with intelligence quotient (IQ) and executive functions in 22q11.2DS subjects, Journal of Psychiatric Research (2014), http://dx.doi.org/10.1016/j.jpsychires.2014.04.019

M. Carmel et al. / Journal of Psychiatric Research xxx (2014) 1e8

3

Table 1 Demographic and clinical data of the study population stratified by age in the 22q11.2DS subjects.

Nb Age (mean  SD, yrs) Gender M/F (%/%) Psychotic disorder (%) OCD (%) MDD (%) ADHD (%)

Alla

Children (age < 12 yrs)

Adolescents (aged 12e18)

Adults (age > 18 yrs)

c2test

n ¼ 92 17.61  10.72 49/43 (53.3/46.7) 13/84 (15.5) 8/84 (9.5) 8/84 (9.5) 41/84 (48.8)

n ¼ 32 8.06  2.37 21/11 (65.6/34.4) 0/22 (0) 1/22 (4.5) 0/22 (0) 19/22 (86.4)

n ¼ 28 14.64  1.73 15/13 (53.6/46.4) 4/31 (12.9) 3/31 (9.7) 4/31 (12.9) 15/31 (48.4)

n ¼ 32 29.75  8.64 13/19 (40.6/59.4) 9/31 (29) 4/31 (12.9) 4/31 (12.9) 7/31 (22.6)

c2 ¼ 4.02, df ¼ 2, p ¼ 0.13 c2 ¼ 8.54, df ¼ 2, p ¼ 0.014c c2 ¼ 1.04, df ¼ 2, p ¼ 0.593 c2 ¼ 3.14, df ¼ 2, p ¼ 0.208 c2 ¼ 20.96, df ¼ 2, p ¼ 0.000d

ADHD ¼ attention deficit hyperactivity disorder, MDD ¼ major depressive disorder, OCD ¼ obsessive compulsive disorder. a Eight subjects did not undergo a psychiatric evaluation, thus 84 out of the 92 underwent a psychiatric assessment. b At IQ evaluation. c Children vs. adults: c2 ¼ 5.79, df ¼ 1, p ¼ 0.016. d Children vs. adults: c2 ¼ 6.96, df ¼ 1, p ¼ 0.008.

2.5. Statistical analysis Data were statistically analyzed using the SPSS software, Chicago, IL (ver. 17). The distribution of psychiatric morbidity and gender were calculated by c2 test. The effect of the SNP variants and age on IQ and FF scores was analyzed using multivariate General Linear Model (GLM). IQ score difference among all age groups was examined by analysis of variance (ANOVA) test. 3. Results Our study population consisted of 92 22q11.2DS individuals with the psychiatric morbidity distribution as follows: 13 individuals with psychotic disorders [schizophrenia (n ¼ 6), schizoaffective disorder (n ¼ 2), major depressive disorder with psychotic features (n ¼ 2) and psychotic disorder NOS (n ¼ 3)], 8 with obsessive compulsive disorder (OCD), 8 suffered from major depressive disorder (MDD) and 41 had attention deficit hyperactivity disorder (ADHD). Their age ranged from 4 to 55 years. We divided our 22q11.2DS individuals into 3 age groups according to their developmental stage: children, adolescents and adults. As shown in Table 1, as expected, significantly higher rates of psychotic disorders were found in the adult compared to the children’s group (p ¼ 0.016) whereas the ADHD was observed more in the children’s group than the adults (p ¼ 0.008).

3.1. Association of COMT Val158Met genetic variant with IQ and EF in the 22q11.2DS subjects 3.1.1. IQ The assessment of IQ in our 22q11.2DS population showed a steady decline with age in the IQ scores (for details see Fig. 1). In order to examine the factors that affect the IQ value, we performed a multivariate GLM with COMT genotype and age as fixed factors, gender as covariate and PIQ and VIQ as dependent variables (Table 2). We found that the COMT genotype (p ¼ 0.01) and age (p ¼ 0.0001) each one independently, affected PIQ with no significant interaction between them. Gender did not affect either PIQ or VIQ. Post-hoc analysis revealed that subjects with the COMT Met allele had higher IQ scores than the Val carriers in all age groups, reaching statistical significance in both adolescent (PIQ, p ¼ 0.017) and adult (PIQ, p ¼ 0.007; VIQ, p ¼ 0.012) groups (Fig. 2). 3.1.2. EF EF was assessed by the Flanker Fish Tasks (FF) (Diamond et al., 2007). We performed multivariate GLM with COMT alleles and age as fixed factors, gender as covariate and FF blue, FF pink and FF mixed as dependent variables. The results are shown in Table 3. As shown by the corrected GLM model both FF pink (p ¼ 0.009) and mixed (p ¼ 0.019) components of EF were affected by other examined factors, while the FF blue component approached significance (p ¼ 0.067). The main factor influencing the three FF tasks examined was age (FF blue, pink and mixed p ¼ 0.027, 0.008, 0.008, respectively). Adolescents outperformed children in all three FF performance tests (FF blue, pink and mixed p ¼ 0.043, 0.008 and 0.009, respectively). No differences were found between other age groups. COMT also affected FF performance, but a significant effect was expressed only in the mixed condition (p ¼ 0.041), the most Table 2 Multivariate General linear model (GLM) for the effect of the COMT Val158Met variant and age on PIQ and VIQ scores. Bold indicates values close to or less than 0.05. Source

Dependent variable

df

Mean square

F

Sig.

Corrected model

VIQ PIQ VIQ PIQ VIQ PIQ VIQ PIQ VIQ PIQ

6 6 1 1 1 1 2 2 2 2

486.090 582.333 12.390 204.742 481.028 983.366 975.977 1262.211 229.754 130.514

3.499 4.148 .089 1.459 3.463 7.005 7.026 8.992 1.654 .930

.004 .001 .766 .231 .066 .010 .001 .000 .197 .399

Gender Fig. 1. Full scale (FSIQ), verbal (VIQ) and performance (PIQ) IQ scores (Mean  SD) in the 22q11.2 deletion syndrome subjects at 3 developmental stages. Assessment of IQ in our 22q11.2DS population, showed a steady decline in the IQ scores from children through adolescents to adults: VIQ 83.72  12.31, 75.68  12.00, 73.22  11.67 respectively; F ¼ 6.66, df ¼ 2,89, p ¼ 0.002. PIQ 80.78  15.61, 75.14  9.46, 69.22  10.35 respectively; F ¼ 7.17, df ¼ 2,89, p ¼ 0.001. FSIQ 78.94  13.31, 73.73  9.65, 69.19  11.36 respectively; F ¼ 6.81, df ¼ 2,89, p ¼ 0.002. **p < 0.003 vs. adults; *p < 0.03 vs. adolescents.

COMT Age COMT  Age

PIQ ¼ performance IQ; VIQ ¼ verbal IQ.

Please cite this article in press as: Carmel M, et al., Association of COMT and PRODH gene variants with intelligence quotient (IQ) and executive functions in 22q11.2DS subjects, Journal of Psychiatric Research (2014), http://dx.doi.org/10.1016/j.jpsychires.2014.04.019

4

M. Carmel et al. / Journal of Psychiatric Research xxx (2014) 1e8 Table 3 Multivariate general linear model shows the effect of the COMT Val158Met variant and age on the Flanker Fish task. The Flanker Fish (FF) tasks examine working memory, inhibition and cognitive flexibility. There are three levels of difficulty: the congruent, incongruent and mixed conditions that correspond to the three sub-tests FF Blue, FF Pink and FF Mixed respectively (Diamond et al., 2007; Zarchi et al., 2013b). Bold indicates values close to or less than 0.05.

Corrected Model

Gender

COMT

Age

COMT  Age

a b c

Dependent variable

Type III sum of squares

Mean square

F

df

Sig.

FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF

.607a 1.062b .532c .045 .002 .024 .023 .194 .133 .370 .569 .331 .263 .369 .111

.101 .177 .089 .045 .002 .024 .023 .194 .133 .185 .284 .166 .131 .184 .056

2.22 3.54 3.03 0.99 0.04 0.82 0.50 3.88 4.55 4.06 5.69 5.66 2.88 3.69 1.90

6 6 6 1 1 1 1 1 1 2 2 2 2 2 2

.067 .009 .019 .328 .840 .371 .486 .058 .041 .027 .008 .008 .071 .037 .166

Blue Pink Mixed Blue Pink Mixed Blue Pink Mixed Blue Pink Mixed Blue Pink Mixed

R Squared ¼ .301 (Adjusted R Squared ¼ .165). R Squared ¼ .407 (Adjusted R Squared ¼ .292). R Squared = .369 (Adjusted R Squared = .247).

3.2. Association of PRODH Arg185Trp genetic variant with IQ and EF in the 22q11.2DS population In order to examine whether PRODH Arg185Trp affects IQ score, we performed multivariate GLM analysis with PRODH genotype and age as fixed factors, gender as covariate and PIQ and VIQ as dependent variables (data not shown). No significant effects were revealed. PRODH Arg185Trp genotype did not affect FSIQ, PIQ or VIQ in our 22q11.2DS subject sample. Namely, there was no significant difference between subjects with Arg allele and subjects with Trp allele in IQ performance. We further assessed the influence of PRODH on EF as evaluated by the FF tasks. We performed a multivariate GLM with PRODH and age as fixed factors, gender as covariate and FF blue, FF pink and FF mixed as dependent variables (results are not shown). PRODH Arg185Trp genotype had no significant effect on any of the FF tasks (FF blue, FF pink or FF mixed) in our sample. Namely, there was no significant difference between PRODH Arg carriers and Trp carriers Fig. 2. Effect of COMT Val158Met variant on performance (PIQ; A) and verbal (VIQ; B) IQ scores (Mean  SD) in the 22q11.2 deletion syndrome subjects. Subjects with the COMT Met allele had higher PIQ scores than the Val carriers in both adolescents (p ¼ 0.017) and adults (p ¼ 0.007). VIQ was significantly higher in adults carrying the Met allele (p ¼ 0.012).

difficult level of the task. No significant gender effect or COMT  age interaction was detected for any of the FF tasks. Both COMT and age, independently, had a significant impact on the performance in the mixed condition of the FF. The achievements of the carriers of the two COMT alleles were similar in the children’s group (Val carriers 0.55  0.2 vs. Met carriers 0.54  0.23, t ¼ 0.046, df ¼ 15, p ¼ 0.963, NS) (Fig. 3). A non-significant difference in the FF mixed condition performance was detected in the adolescent group: Met carriers performed better than Val carriers (0.81  0.04 vs. 0.69  0.14, respectively; t ¼ 2.12, df ¼ 11, p ¼ 0.06, NS). A significant difference between the Val and Met carriers in the FF mixed performance was revealed only in the adult group (Met carriers 0.80  0.11 vs. Val carriers 0.54  0.13; t ¼ 2.914, df ¼ 6, p ¼ 0.027) (Fig. 3).

Fig. 3. Effect of COMT Val158Met variant on accuracy in the Flanker Fish (FF) mixed condition (an executive function task) of the 22q11.2DS age groups. Accuracy was significantly higher (p ¼ 0.027) in the adults Met carriers than the Val carriers.

Please cite this article in press as: Carmel M, et al., Association of COMT and PRODH gene variants with intelligence quotient (IQ) and executive functions in 22q11.2DS subjects, Journal of Psychiatric Research (2014), http://dx.doi.org/10.1016/j.jpsychires.2014.04.019

M. Carmel et al. / Journal of Psychiatric Research xxx (2014) 1e8

5

In this study, we investigated the impact of COMT and PRODH polymorphisms on IQ and EF performance in 22q11.2DS subjects. Our main findings are a significant and selective influence of the COMT Val158Met polymorphism on both IQ and EF: (i) 22q11.2DS subjects with Met allele had a higher IQ score in all age groups compared to Val carriers, reaching statistical difference in both adolescent and adults (ii) 22q11.2DS Met carriers performed better than Val carriers in EF performance tasks, the statistical significance was revealed in the adult group, but the tendency started at younger age. (iii) PRODH Arg185Trp variant did not affect IQ or EF in our 22q11.2DS subject sample.

2005; Simon et al., 2005; Takarae et al., 2009). It was shown that children with 22q11.2DS had greater variation (as shown by larger standard deviations) in cognitive performance with relation to their healthy siblings or typically developed individuals (Niarchou et al., 2014; Shapiro et al., 2012; Sobin et al., 2004). EF performance in our 22q11.2DS subjects, as measured by the FF tasks, was significantly better in adolescents compared to children, with no further improvement in adults. A recent study that compared 22q11.2DS subjects to normal typically developed children aged 7e14 years, showed that the 22q11.2DS children performed EF tasks similarly to the matching age group of typically developed children. Significant differences between the two groups began to emerge only in the older children (Shapiro et al., 2013). The same researchers in an earlier study showed that young children with 22q11.2DS demonstrated reduced performance on tasks of attention compared to their older counterparts (Shapiro et al., 2012). Stoddard et al. (2011) revealed that children with 22q11.2DS demonstrated age-related impairment in the executive control of attention and children in the third age tertile (12.3e14.9 years) performed better in the incongruent flanker task compared to younger age tertile. Age of the individuals studied is the key to EF performance because major changes in brain structure and development over time (Philip and Bassett, 2011).

4.1. Effect of age on IQ

4.3. Effect of COMT variants on IQ

The majority of subjects with 22q11.2DS are known to have borderline IQ levels (the mean IQ in 22q11.2DS is typically around 75), but the IQ range can vary from as low as 40 to more than 100 (Gothelf et al., 2013; Squarcione et al., 2013). These data may influence and contribute to the large standard deviation obtained in the IQ test. In general, IQ assessment of our 22q11.2DS population, irrespective of genotype and gender, showed a steady decline with age in the IQ scores. The children’s group had the highest FSIQ score, the adolescent had intermediate, while the adults had the lowest IQ scores. The same observation was maintained in the PIQ and VIQ sub-tests. A similar observation has previously been reported in longitudinal studies. Gothelf et al. (2005) followed up for 5 years twenty four 22q11.2DS individuals from childhood (age 13.3  3.7) to adolescence (age 18.1  3.4) and showed a decline from baseline in VIQ in the 22q11.2DS population cohort. In a more recent study, Gothelf et al. (2013), enlarged the sample number and for four years followed up 125 subjects with 22q11.2DS. As in their previous study, they found a significant decline in IQ scores that was similarly significant for both VIQ and PIQ scores in children and adolescents. However, the decline was less robust for VIQ in adults. Another longitudinal study examining seventy 22q11.2DS individuals for 3 years from childhood (mean age 11.8 years) into mid-adolescence (mean age 15.0 years), demonstrated a general decrease in IQ (Antshel et al., 2010). Duijff et al. (2012) reported a similar observation in 22q11.2DS children at a younger age; they found a mean significant decline of 9.7 FSIQ points between 5.5 and 9.5 years of age. Similar results were also obtained from cross-sectional research. Niarchou et al. (2014) reported that older children with 22q11.2DS had lower IQs than younger children, ranging from 6.6 to 14.1 years of age; this age-related decline was restricted to 22q11.2DS subjects only and was not observed in control subjects.

COMT genotype affected IQ, with no significant interaction with age or gender. Namely, subjects with Met allele had higher IQ score in all age groups compared to Val carriers (significant in adolescents and adults). Our results are consistent with previous reports. Shashi et al. (2006) found that children (aged 7e16) with 22q11DS who have the Met allele have higher FSIQ and VIQ (more than 12 points), as compared with those with the Val allele. There are some works that are inconsistent with our findings: Baker et al. (2005) found in their 22q11.2DS population, that the COMT Met allele is associated with more than a 10-point decrement in IQ; however, the difference did not reach statistical significance. Bearden et al. (2004) screened 44 subjects with 22q11.2DS, mean age of 11.1 years, 28 of them were Val hemizygous and 16 were Met hemizygous and found that in this study group there was a tendency toward higher FSIQ in Val-hemizygous subjects (mean ¼ 77.6  10.5 versus 71.8  11.4, respectively). But when they controlled the IQ effect, they found that Met-hemizygous subjects performed better than Val-hemizygous individuals. There are several longitudinal studies that examined the effect of COMT genotype on IQ in healthy population. Barnett et al. (2007a) surveyed 8707 normal boys and girls at age 8 and 10 years and found that COMT genotype significantly affected VIQ. This effect was expressed primarily in pubertal boys; the children who were Met allele carriers had higher VIQ than Val carriers. Dumontheil et al. (2011) confirmed similar results in their longitudinal study of 260 typically developed individuals with benefit of the Met allele emerging after 10 years of age. A recent longitudinal study of Gaysina et al. (2013) which followed up more than 2000 healthy individuals from the British 1946 birth cohort and compared IQ outcomes tests between the same children aged eight and 15 did not corroborate the previous findings and found no association between COMT variation and IQ in childhood and adolescence. Concerning the effect of gender on 22q11.2DS subjects, as already mentioned in our 22q11.2DS cohort, gender did not affect either PIQ or VIQ. There are studies that demonstrated gender effect on IQ in 22q11.2DS individuals (Barnett et al., 2007a; Duijff et al., 2012), but recent reports have concluded that gender does not

in the FF tasks performance. No interaction was found between PRODH genotype and age. Despite the lack of effect of PRODH genotype on cognitive function we analyzed the interaction between the effects of COMT and PRODH variants on EF and IQ, and no significant interaction was detected (data not shown). No association was found between the various alleles and the mental diagnoses (data not shown). Due to the small sample size of each diagnostic group and the heterogeneity of age we were unable to analyze reliably the possible moderating role of diagnosis in our cohort. 4. Discussion

4.2. Effect of age on EF 22q11.2DS is characterized by variable decline in EF performance, developmental delay and learning disabilities (Gothelf et al., 2013; Zarchi et al., 2013a; Shashi et al., 2012; Bish et al.,

Please cite this article in press as: Carmel M, et al., Association of COMT and PRODH gene variants with intelligence quotient (IQ) and executive functions in 22q11.2DS subjects, Journal of Psychiatric Research (2014), http://dx.doi.org/10.1016/j.jpsychires.2014.04.019

6

M. Carmel et al. / Journal of Psychiatric Research xxx (2014) 1e8

affect IQ either in 22q11.2DS or in healthy individuals (Niarchou et al., 2014; Bearden et al., 2004; Gaysina et al., 2013). 4.4. Effect of COMT variants on EF A significant effect of COMT allele was detected only in the FF mixed condition, the most difficult level of the task. In the children’s group the EF performance of the two COMT allele carriers were similar. A non-significant difference in FF mixed condition performance was obtained in the adolescent group in which Met carriers performed better then Val carriers. A significant difference in FF mixed performance was revealed only in the adult group, adult Met carriers performed much better than Val carriers. Other studies in the healthy population reported on the effect of COMT genotype on EF, namely, Met allele carriers have advantage over the Val carriers (Barnett et al., 2007b; Farrell et al., 2012). A metaanalysis (Barnett et al., 2007b) of healthy individuals showed that the Met/Met homozygous individuals performed better than Val/ Val individuals in EF and gender did not affect this cognitive performance. Farrell et al. (2012) compared executive performance of 16 Met to Val COMT homozygous healthy men aged 18 to 50 and found that Met/Met-COMT subjects outperformed their Val/ValCOMT counterparts. Dumontheil et al. (2011) demonstrated in a population of typically developed 6e20 year-old subjects that the cognitive benefit of the Met allele emerges after 10 years of age. Similar findings were found in patient populations. Diaz-Asper et al. (2008), assessed the working memory/EF in schizophrenia probands (n ¼ 325), their non-psychotic siblings (n ¼ 359), and normal control subjects (n ¼ 330) in participants aged 18e60. Their findings support the contention that COMT Val158Met genotype influences EF, with Met homozygotes demonstrating the best performance. Similar results were reported by Shashi et al. (2006) and Bearden et al. (2004): Met-hemizygous 22q11.2DS children outperformed Val carriers on a composite measure of EF. Unfortunately in both studies there is no comparison of the performance of 22q11.2DS children and their adolescent counterparts. In contrast, there are some studies that are inconsistent with our findings. Gaysina et al. (2013) used data from the British 1946 birth cohort and investigated the effect of five COMT SNPs, among them Val158Met, on cognitive function in the same boys and girls at two time-points (age 8 and 15 years). They found no significant association between the Val158Met polymorphism and cognitive functioning. An earlier meta-analysis also failed to demonstrate a significant effect of the Val158Met SNP on frontal-related cognitive tasks (Barnett et al., 2008). Glaser et al. (2006a) found no evidence of an effect of COMT Val158Met genotype on EF in children and adults with 22q11.2DS. 4.5. Effect of PRODH variants on IQ and EF Although PRODH gene appeared to be a candidate gene with great potential to influence IQ and EF in the general population and especially in 22q11.2DS, we found no effect of PRODH Arg185Trp genotype on FSIQ, PIQ or VIQ and EF performance in our 22q11.2DS subjects. A 3 SNPs haplotype, rs4819756-rs2870983-rs450046, containing our PRODH Arg185Trp SNP, is significantly associated with quantitative changes in striatal activity and fronto-striatal connectivity during performance of cognitive tasks (Kempf et al., 2008). Roussos et al. (2009) showed significant associations between the haplotype comprising three PRODH SNPs (rs372055, rs450046, rs385440) and sensorimotor gating, memory, schizotypy and anxiety in healthy male subjects. Zinkstok et al. (2008) investigated the association between three PRODH SNPs (different from

ours) and brain morphometry in young patients with schizophrenia and schizoaffective disorder. They found that two of the three SNPs tested were associated with bilateral frontal white matter density reductions, a finding frequently reported in 22q11.2DS and schizophrenia subjects. Raux et al. (2007) characterized the PRODH genotype and the plasma proline level in 92 adult or adolescent 22q11.2DS subjects. They found an inverse correlation between the plasma proline level and IQ, attesting that mild-to-moderate hyperprolinemia is involved in cognitive impairment in humans. Interestingly, 27 of 92 subjects had the risk allele PRODH Arg185Trp genotype. Another haplotype of PRODH (rs385440-rs372055) was associated with impaired planning capacity in EF tasks (Li et al., 2008). A deficit in associative learning was detected in a mouse model of PRODH deficiency that mimics the status in 22q11.2DS (Paterlini et al., 2005). Several human studies failed to find an association between the various PRODH gene variants and schizophrenia (Glaser et al., 2006b; Abou Jamra et al., 2005; Ohtsuki et al., 2004; Williams et al., 2003a, 2003b; Fan et al., 2003). However, they did not assess the impact of the PRODH gene variants on EF and IQ. In our study PRODH gene variant did not affect EF in 22q11.2DS subjects, a population prone to develop schizophrenia. Thus, it seems that PRODH gene variants do not play a major role in EF performance and IQ in 22q11.2DS. 5. Conclusions Our findings show that the COMT genotype affects the performance of EF as assessed by the FF test. The beneficial effect of the COMT Met allele on EF performance is manifested in both adolescents and adults, but not in children. The benefit of the COMT low activity allele may be achieved by reduced inactivation of monoamines in the prefrontal cortex (Squarcione et al., 2013). In contrast, PRODH polymorphism did not affect EF performance or IQ score in our 22q11.2DS subjects. The limitations of our study include: a relatively small sample size, the cross-sectional nature of the study, the failure of correcting for multiple testing and the phenotypic/diagnostic and age heterogeneity of the sample. The strength of the study is the use of the appropriate cognitive tasks in three age groups that represent different neurodevelopmental stages, enabling the detection of the effect of age on IQ and EF performance in 22q11.2DS. We believe that our report contributes to a growing body of literature implicating genetic variation of the COMT locus in inefficient performance of cognitive functions (IQ and executive function) that are relevant to the phenotype of schizophrenia and other mental disorders associated with cognitive deficits. In conclusion, functional COMT, but not PRODH, variant affects IQ and EF in 22q11.2DS subjects during neurodevelopment with a maximal effect at adulthood. Future longitudinal studies should monitor the cognitive performance of the same 22q11.2DS individuals from childhood to old age in an attempt to determine the complex interaction between gene variants and cognition in the life cycle of this unique population. Role of funding source No role of funding source. Author’s contributions MC prepared the first draft of the manuscript. MC, AF, EM and MP managed the genotyping assays and analyses. DG, AW and AF designed the study and DG wrote the clinical protocol. TG participated in the clinical evaluations. MC, EM and OZ performed the

Please cite this article in press as: Carmel M, et al., Association of COMT and PRODH gene variants with intelligence quotient (IQ) and executive functions in 22q11.2DS subjects, Journal of Psychiatric Research (2014), http://dx.doi.org/10.1016/j.jpsychires.2014.04.019

M. Carmel et al. / Journal of Psychiatric Research xxx (2014) 1e8

statistical analysis. AW, EM, OZ and AF participated in designing the study, data analyses and participated in drafting the manuscript. DG recruited and followed up the patients, carried out the clinical evaluations and participated in the study design. All authors read and approved the final manuscript. Conflict of interest All authors declare that they have no conflict of interest. Acknowledgment No acknowledgment. References Abou Jamra R, Schumacher J, Becker T, Dahdouh F, Ohlraun S, Suliman H, et al. No evidence for an association between variants at the proline dehydrogenase locus and schizophrenia or bipolar affective disorder. Psychiatr Genet 2005;15(3):195e8. Antshel KM, Shprintzen R, Fremont W, Higgins AM, Faraone SV, Kates WR. Cognitive and psychiatric predictors to psychosis in velocardiofacial syndrome: a 3-year follow-up study. J Am Acad Child Adolesc Psychiatry 2010;49(4):333e44. Baker K, Baldeweg T, Sivagnanasundaram S, Scambler P, Skuse D. COMT Val108/158 Met modifies mismatch negativity and cognitive function in 22q11 deletion syndrome. Biol Psychiatry 2005;58(1):23e31. Barnett JH, Heron J, Goldman D, Jones PB, Xu K. Effects of catechol-Omethyltransferase on normal variation in the cognitive function of children. Am J Psychiatry 2009;166(8):909e16. Barnett JH, Heron J, Ring SM, Golding J, Goldman D, Xu K, et al. Gender-specific effects of the catechol-O-methyltransferase Val108/158Met polymorphism on cognitive function in children. Am J Psychiatry 2007a;164(1):142e9. Barnett JH, Jones PB, Robbins TW, Müller U. Effects of the catechol-Omethyltransferase Val158Met polymorphism on executive function: a metaanalysis of the Wisconsin Card Sort Test in schizophrenia and healthy controls. Mol Psychiatry 2007b;12(5):502e9. Barnett JH, Scoriels L, Munafò MR. Meta-analysis of the cognitive effects of the catechol-O-methyltransferase gene Val158/108Met polymorphism. Biol Psychiatry 2008;64(2):137e44. Bearden CE, Jawad AF, Lynch DR, Sokol S, Kanes SJ, McDonald-McGinn DM, et al. Effects of a functional COMT polymorphism on prefrontal cognitive function in patients with 22q11.2 deletion syndrome. Am J Psychiatry 2004;161(9):1700e2. Bender HU, Almashanu S, Steel G, Hu CA, Lin WW, Willis A, et al. Functional consequences of PRODH missense mutations. Am J Hum Genet 2005;76(3):409e 20. Bender S, Rellum T, Freitag C, Resch F, Rietschel M, Treutlein J, et al. Dopamine inactivation efficacy related to functional DAT1 and COMT variants influences motor response evaluation. PLoS One 2012;7(5):e37814. Bish JP, Ferrante SM, McDonald-McGinn D, Zackai E, Simon TJ. Maladaptive conflict monitoring as evidence for executive dysfunction in children with chromosome 22q11.2 deletion syndrome. Dev Sci 2005;8(1):36e43. Blanchard MM, Chamberlain SR, Roiser J, Robbins TW, Müller U. Effects of two dopamine-modulating genes (DAT1 9/10 and COMT Val/Met) on n-back working memory performance in healthy volunteers. Psychol Med 2011;41(3): 611e8. Caplan R, Arbelle S, Guthrie D, Komo S, Shields WD, Hansen R, et al. Formal thought disorder and psychopathology in pediatric primary generalized and complex partial epilepsy. J Am Acad Child Adolesc Psychiatry 1997;36(9):1286e94. Daniels JK, Williams NM, Williams J, Jones LA, Cardno AG, Murphy KC, et al. No evidence for allelic association between schizophrenia and a polymorphism determining high or low catechol O-methyltransferase activity. Am J Psychiatry 1996;153(2):268e70. Diamond A, Barnett WS, Thomas J, Munro S. Preschool program improves cognitive control. Science 2007;318(5855):1387e8. Diaz-Asper CM, Goldberg TE, Kolachana BS, Straub RE, Egan MF, Weinberger DR. Genetic variation in catechol-O-methyltransferase: effects on working memory in schizophrenic patients, their siblings, and healthy controls. Biol Psychiatry 2008;63(1):72e9. Drew LJ, Crabtree GW, Markx S, Stark KL, Chaverneff F, Xu B, et al. The 22q11.2 microdeletion: fifteen years of insights into the genetic and neural complexity of psychiatric disorders. Int J Dev Neurosci 2011;29(3):259e81. Duijff SN, Klaassen PW, de Veye HF, Beemer FA, Sinnema G, Vorstman JA. Cognitive development in children with 22q11.2 deletion syndrome. Br J Psychiatry 2012;200(6):462e8. Dumontheil I, Roggeman C, Ziermans T, Peyrard-Janvid M, Matsson H, Kere J, et al. Influence of the COMT genotype on working memory and brain activity changes during development. Biol Psychiatry 2011;70(3):222e9. Fan JB, Ma J, Zhang CS, Tang JX, Gu NF, Feng GY, et al. A family-based association study of T1945C polymorphism in the proline dehydrogenase gene and schizophrenia in the Chinese population. Neurosci Lett 2003;338(3):252e4.

7

Farrell SM, Tunbridge EM, Braeutigam S, Harrison PJ. COMT Val(158)Met genotype determines the direction of cognitive effects produced by catechol-Omethyltransferase inhibition. Biol Psychiatry 2012;71(6):538e44. Gaysina D, Xu MK, Barnett JH, Croudace TJ, Wong A, Richards M, et al. LHA genetics group. The catechol-O-methyltransferase gene (COMT) and cognitive function from childhood through adolescence. Biol Psychol 2013;92(2):359e64. Glaser B, Debbane M, Hinard C, Morris MA, Dahoun SP, Antonarakis SE, et al. No evidence for an effect of COMT Val158Met genotype on executive function in patients with 22q11 deletion syndrome. Am J Psychiatry 2006a;163(3):537e9. Glaser B, Moskvina V, Kirov G, Murphy KC, Williams H, Williams N, et al. Analysis of ProDH, COMT and ZDHHC8 risk variants does not support individual or interactive effects on schizophrenia susceptibility. Schizophrenia Res 2006b;87(1e 3):21e7. Gothelf D, Eliez S, Thompson T, Hinard C, Penniman L, Feinstein C, et al. COMT genotype predicts longitudinal cognitive decline and psychosis in 22q11.2 deletion syndrome. Nat Neurosci 2005;8(11):1500e2. Gothelf D, Schneider M, Green T, Debban M, Frisch A, Glaser B, et al. Risk factors and the evolution of psychosis in 22q11.2 deletion syndrome: a longitudinal 2-site study. J Am Acad Child Adolesc Psychiatry 2013;52(11):1192e203. Karayiorgou M, Simon TJ, Gogos JA. 22q11.2 microdeletions: linking DNA structural variation to brain dysfunction and schizophrenia. Nat Rev Neurosci 2010;11(6): 402e16. PMID: 20485365. Kempf L, Nicodemus KK, Kolachana B, Vakkalanka R, Verchinski BA, Egan MF, et al. Functional polymorphisms in PRODH are associated with risk and protection for schizophrenia and fronto-striatal structure and function. PLoS Genet 2008;4(11):e1000252. 1e9. Kimoto S, Muraki K, Toritsuka M, Mugikura S, Kajiwara K, Kishimoto T, et al. Selective overexpression of Comt in prefrontal cortex rescues schizophrenia-like phenotypes in a mouse model of 22q11 deletion syndrome. Transl Psychiatry 2012;2(8):e146. 1e16. Knickmeyer RC, Wang J, Zhu H, Geng X, Woolson S, Hamer RM, et al. Common variants in psychiatric risk genes predict brain structure at birth. Cereb Cortex 2014;24(5):1230e46. Li T, Ma X, Hu X, Wang Y, Yan C, Meng H, et al. PRODH gene is associated with executive function in schizophrenic families. Am J Med Genet Part B (Neuropsychiatr Genet) 2008;147B(5):654e7. Niarchou M, Zammit S, van Goozen SH, Thapar A, Tierling HM, Owen MJ, et al. Psychopathology and cognition in children with 22q11.2 deletion syndrome. Br J Psychiatry 2014;204(1):46e54. Ohtsuki T, Tanaka S, Ishiguro H, Noguchi E, Arinami T, Tanabe E, et al. Failure to find association between PRODH deletion and schizophrenia. Schizophrenia Res 2004;67(1):111e3. Parasuraman R, Jiang Y. Individual differences in cognition, affect, and performance: behavioral, neuroimaging, and molecular genetic approaches. NeuroImage 2012;59(1):70e82. Paterlini M, Zakharenko SS, Lai WS, Qin J, Zhang H, Mukai J, et al. Transcriptional and behavioral interaction between 22q11.2 orthologs modulates schizophrenia-related phenotypes in mice. Nat Neurosci 2005;8(11):1586e94. Philip N, Bassett A. Cognitive, behavioural and psychiatric phenotype in 22q11.2 deletion syndrome. Behav Genet 2011;41(3):403e12. Rasetti R, Weinberger DR. Intermediate phenotypes in psychiatric disorders. Curr Opin Genet Dev 2011;21(3):340e8. Raux G, Bumsel E, Hecketsweiler B, van Amelsvoort T, Zinkstok J, ManouvrierHanu S, et al. Involvement of hyperprolinemia in cognitive and psychiatric features of the 22q11 deletion syndrome. Hum Mol Genet 2007;16(1):83e 91. Roussos P, Giakoumaki SG, Bitsios P. A risk PRODH haplotype affects sensorimotor gating, memory, schizotypy, and anxiety in healthy male subjects. Biol Psychiatry 2009;65(12):1063e70. Shapiro HM, Takarae Y, Harvey DJ, Cabaral MH, Simon TJ. A cross-sectional study of the development of volitional control of spatial attention in children with chromosome 22q11.2 deletion syndrome. J Neurodev Disord 2012;4(1):5. 1e12. Shapiro HM, Wong LM, Simon TJ. A cross-sectional analysis of the development of response inhibition in children with chromosome 22q11.2 deletion syndrome. Front Psychiatry 2013;4(81):1e10. Shashi V, Francis A, Hooper SR, Kranz PG, Zapadka M, Schoch K, et al. Increased corpus callosum volume in children with chromosome 22q11.2 deletion syndrome is associated with neurocognitive deficits and genetic polymorphisms. Eur J Hum Genet 2012;20(10):1051e7. Shashi V, Keshavan MS, Howard TD, Berry MN, Basehore MJ, Lewandowski E, et al. Cognitive correlates of a functional COMT polymorphism in children with 22q11.2 deletion syndrome. Clin Genet 2006;69(3):234e8. Simon TJ, Bish JP, Bearden CE, Ding L, Ferrante S, Nguyen V, et al. A multilevel analysis of cognitive dysfunction and psychopathology associated with chromosome 22q11.2 deletion syndrome in children. Dev Psychopathol 2005;17(3): 753e84. Sobin C, Kiley-Brabeck K, Daniels S, Blundell M, Anyane-Yeboa K, Karayiorgou M. Networks of attention in children with the 22q11 deletion syndrome. Dev Neuropsychol 2004;26(2):611e26. Squarcione C, Torti MC, Di Fabio F, iondi M. 22q11 deletion syndrome: a review of the neuropsychiatric features and their neurobiological basis. Neuropsychiatr Dis Treat 2013;9:1873e84. Stoddard J, Beckett L, Simon TJ. Atypical development of the executive attention network in children with chromosome 22q11.2 deletion syndrome. J Neurodev Disord 2011;3(1):76e85.

Please cite this article in press as: Carmel M, et al., Association of COMT and PRODH gene variants with intelligence quotient (IQ) and executive functions in 22q11.2DS subjects, Journal of Psychiatric Research (2014), http://dx.doi.org/10.1016/j.jpsychires.2014.04.019

8

M. Carmel et al. / Journal of Psychiatric Research xxx (2014) 1e8

Takarae Y, Schmidt L, Tassone F, Simon TJ. Catechol-O-methyltransferase polymorphism modulates cognitive control in children with chromosome 22q11.2 deletion syndrome. Cognitive, Affect Behav Neurosci 2009;9(1):83e90. Vorstman JA, Turetsky BI, Sijmens-Morcus ME, de Sain MG, Dorland B, Sprong M, et al. Proline affects brain function in 22q11DS children with the low activity COMT 158 allele. Neuropsychopharmacology 2009;34(3):739e46. Wechsler D. Wechsler Intelligence Scale for children. 3rd ed. San Antonio, TX: Psychological Corporation; 1991. Williams HJ, Williams N, Spurlock G, Norton N, Ivanov D, McCreadie RG, et al. Association between PRODH and schizophrenia is not confirmed. Mol Psychiatry 2003a;8(7):644e5. Williams HJ, Williams N, Spurlock G, Norton N, Zammit S, Kirov G, et al. Detailed analysis of PRODH and PsPRODH reveals no association with schizophrenia. Am J Med Genet Part B, Neuropsychiatric Genet 2003b;120B(1):42e6.

Witte AV, Flöel A. Effects of COMT polymorphisms on brain function and behavior in health and disease. Brain Res Bull 2012;88(5):418e28. Zarchi O, Carmel M, Avni C, Attias J, Frisch A, Michaelovsky E, et al. Schizophrenialike neurophysiological abnormalities in 22q11.2 deletion syndrome and their association to COMT and PRODH genotypes. J Psychiatric Res 2013a;47(11): 1623e9. Zarchi O, Diamond A, Weinberger R, Abbott D, Carmel M, Frisch A, et al. A comparative study of the neuropsychiatric and neurocognitive phenotype in two microdeletion syndromes: Velocardiofacial (22q11.2 deletion) and Williams (7q11.23 deletion) syndromes. Eur Psychiatry; 2013. http://dx.doi.org/ 10.1016/j.eurpsy.2013.07.001. pii: S0924e9338(13)00097-7. Zinkstok J, Schmitz N, van Amelsvoort T, Moeton M, Baas F, Linszen D. Genetic variation in COMT and PRODH is associated with brain anatomy in patients with schizophrenia. Genes Brain, Behav 2008;7(1):61e9.

Please cite this article in press as: Carmel M, et al., Association of COMT and PRODH gene variants with intelligence quotient (IQ) and executive functions in 22q11.2DS subjects, Journal of Psychiatric Research (2014), http://dx.doi.org/10.1016/j.jpsychires.2014.04.019