Dissection of phenotype reveals possible association between schizophrenia and Glutamate Receptor Delta 1 (GRID1) gene promoter

Schizophrenia Research 111 (2009) 123–130

Contents lists available at ScienceDirect

Schizophrenia Research j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / s c h r e s

Dissection of phenotype reveals possible association between schizophrenia and Glutamate Receptor Delta 1 (GRID1) gene promoter Jens Treutlein a,1, Thomas W. Mühleisen b,1, Josef Frank a, Manuel Mattheisen b, Stefan Herms b, Kerstin U. Ludwig b, Tsendsesmee Treutlein a, Christine Schmael a, Jana Strohmaier a, Katja Veronika Böβhenz a, René Breuer a, Torsten Paul a, Stephanie H. Witt a, Thomas G. Schulze a, Ralf G.M. Schlösser d, Igor Nenadic d, Heinrich Sauer d, Tim Becker e, Wolfgang Maier f, Sven Cichon b,c, Markus M. Nöthen b,c, Marcella Rietschel a,⁎ a

Department of Genetic Epidemiology, Central Institute of Mental Health, J5, D-68159 Mannheim, Germany Department of Genomics, Life & Brain Center, University of Bonn, Sigmund-Freud-Str. 25, D-53127 Bonn, Germany Institute of Human Genetics, University of Bonn, Wilhelmstrasse 31, D-53111 Bonn, Germany d Department of Psychiatry, University of Jena, Philosophenweg 3, D-07743 Jena, Germany e Institute for Medical Biometry, Informatics and Epidemiology, University of Bonn, Sigmund-Freud-Str. 25, D-53105 Bonn, Germany f Department of Psychiatry, University of Bonn, Sigmund-Freud-Str. 25, D-53127 Bonn, Germany b c

a r t i c l e

i n f o

Article history: Received 17 December 2008 Received in revised form 25 February 2009 Accepted 4 March 2009 Available online 5 April 2009 Keywords: Schizophrenia GRID1 Association Case–control Haplotype block-wise tagging Lifetime history of major depression

a b s t r a c t Recent linkage and association data have implicated the Glutamate Receptor Delta 1 (GRID1) locus in the etiology of schizophrenia. In this study, we sought to test whether variants in the promoter region are associated with this disorder. The distribution of CpG islands, which are known to be relevant for transcriptional regulation, was computationally determined at the GRID1 locus, and the putative transcriptional regulatory region at the 5′-terminus was systematically tagged using HapMap data. Genotype analyses were performed with 22 haplotype-tagging single nucleotide polymorphisms (htSNPs) in a German sample of 919 schizophrenia patients and 773 controls. The study also included two SNPs in intron 2 and one in intron 3 which have been found to be significantly associated with schizophrenia in previous studies. For the transcriptional regulatory region, association was obtained with rs3814614 (p = 0.0193), rs10749535 (p = 0.0245), and rs11201985 (p = 0.0222). For all further analyses, the patient samples were divided into more homogeneous subgroups according to sex, age at onset, positive family history of schizophrenia and lifetime history of major depression. The pvalue of the schizophrenia association finding for the three markers decreased by approximately one order of magnitude, despite the reduction in the total sample size. Marker rs3814614 (unadjusted p = 0.0005), located ~ 2.0 kb from the transcriptional start point, also withstood a two-step correction for multiple testing (p = 0.030). No support was obtained for previously reported associations with the intronic markers. Our results suggest that genetic variants in the GRID1 transcriptional regulatory region may play a role in the etiology of schizophrenia, and that future association studies of schizophrenia may require stratification to ensure more homogeneous patient subgroups. © 2009 Elsevier B.V. All rights reserved.

1. Introduction ⁎ Corresponding author. Tel.: +49 621 1703 6051. E-mail address: [email protected] (M. Rietschel). 1 These authors contributed equally to this work. 0920-9964/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.schres.2009.03.011

The “glutamate hypothesis” aims to explain the symptoms of schizophrenia in terms of changes occurring in the glutamatergic neurotransmitter system (Kim et al., 1980;

124

J. Treutlein et al. / Schizophrenia Research 111 (2009) 123–130

Pilowsky et al., 2006). Recent genome-wide linkage scans for schizophrenia susceptibility loci (Fallin et al., 2003; Lerer et al., 2003, Faraone et al., 2006) have shown linkage to schizophrenia on human chromosome 10q in various ethnicities. A follow-up association analysis to the Fallin et al. (2003) study implicated the Glutamate Receptor Delta 1 (GRID1), which is located 1.35 Mb from the peak chromosome 10q22–q23 linkage signal (Fallin et al., 2005), as being the causative gene in Askenazi Jewish families, and added GRID1 to the list of glutamatergic candidate genes for schizophrenia. Guo et al. (2007) recently confirmed the association of GRID1 with schizophrenia in a Chinese Northern Han population and hypothesise that polymorphisms, or variants which are in linkage disequilibrium (LD) to them, may influence either the splicing or the level of gene expression of GRID1. The transcriptional regulatory region influences the abundance of messenger ribonucleic acid (mRNA) transcript and consequently the level of gene expression. The occurrence of CpG islands around transcriptional start sites is important for transcriptional efficiency (Antequera, 2003), and has been described as a characteristic of some glutamate receptor subunit genes (Myers et al., 1999). We therefore computationally determined the distribution of CpG islands at the GRID1 locus and applied haplotype block-wise tagging to this putative transcriptional regulatory region. In addition to SNPs in the transcriptional regulatory region, the three most significantly associated SNPs from the study of Guo et al. (2007) were included for replication, i.e. rs999383, rs11591408, and rs1902666. 2. Materials and methods 2.1. Subjects and psychiatric assessment Psychiatric evaluation was approved by the ethics committees of the Faculties of Medicine at the Universities of Heidelberg, Bonn and Jena. All study participants were recruited from consecutive hospital admissions and gave written informed consent. For the analyses, we included 919 German schizophrenia patients (403 females and 516 males; aged between 17 and 85 years, mean age = 36.3 ± 11.7) and 773 controls (307 females and 466 males; aged between 19 and 81 years, mean age = 44.8 ± 13.9). All patients were of German origin (defined as: parents had to be German. Patients were excluded if a grandparent was reported to be of non European origin or more then one grandparent not to be German). Diagnostic assessments using a best estimate approach were performed using a comprehensive inventory for phenotype characterisation, the Interviews for Psychiatric Genetic Studies (IPGS; Fangerau et al., 2004). This included the Structured Clinical Interview for the DSM-IV Axis I Disorders (SCID-I; First et al., 1994) for assessment according to the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV) and the Operational Criteria Checklist for Psychotic Illness (OPCRIT; McGuffin et al., 1991), as well as a review of medical records. Lifetime ratings were made for all items. All interviews were conducted by a trained psychologist/psychiatrist. Consensus diagnoses were performed by two psychiatrists, and additional psychiatrists were included in the decision making process when necessary. The population-based sample of controls was established with the help of the local Census Bureau of the city of Bonn (North Rhine-Westphalia, Germany).

In addition to the DSM-IV diagnosis of schizophrenia we also tested for association with the following phenotypic subgroups: female and male sex, positive family history (at least one first or second degree relative affected by schizophrenia), early age at onset (b20 years) and lifetime history of major depression. The latter three phenotypes were selected because of a presumed higher familial/genetic load for schizophrenia, such as it is likely to be present in patients with a positive family history of schizophrenia, patients with an earlier age at onset (Kendler and MacLean, 1990; Pulver et al., 1990; Suvisaari et al., 1998; Byrne et al., 2002), and in patients with a history of major depressive episodes (DeLisi et al., 1987; Kendler et al., 1997). As concerns the latter, latent class analyses have additionally shown mood symptoms to be the best discriminators of subgroups of psychosis (Boks et al., 2007). Furthermore many studies—among these a genomewide linkage study specifically stratifying for the occurrence of depressive episodes (Hamshere et al., 2006)—have shown different, or stronger, association with genetic markers than the non-stratified group of schizophrenia patients (Schumacher et al., 2005, Corvin et al., 2007, Georgi et al., 2007, Boks et al., 2008). In addition there is evidence that depressive symptoms have an impact on the course and outcome of schizophrenia (an der Heiden et al., 2005). 2.2. CpG islands and selection of tagging polymorphisms Positions of CpG islands across the GRID1 locus were retrieved from a data track (‘Regulation → CpG islands’; http:// genome.ucsc.edu/cgi-bin/hgTrackUi?hgsid=119874356&c= chr2&g=cpgIslandExt) of the University of California Santa Cruz (UCSC) genome browser (Karolchik et al., 2003). The track displays islands which were predicted according to the formula cited in Gardiner-Garden and Frommer (1987). For a 82.2 kb region of GRID1 (rs4454627–rs7907234, chr10:88082991–88165186, UCSC Genome Browser on Human March 2006 Assembly; dbSNP build 125), which includes the transcriptional start site and which contains the majority of the CpG islands, we selected 22 tagging SNPs for haplotype blocks representing haplotypes of frequencies N1% in the Centre d'Etude du Polymorphisme Humain (CEPH) Project's Utah (CEU) population using HapMap phase I + II data (HapMap release 20, dbSNP build 125). The blocks were defined by the Gabriel et al. (2002) algorithm, as implemented in Haploview 3.32, which was run under default settings (Barrett et al., 2005). At least two tagging SNPs were selected for each haplotype block. In addition, the three SNPs most significantly associated in the 2007 study of Guo et al. (rs999383, rs11591408, and rs1902666) were genotyped in order to replicate previous findings. 2.3. Genotyping and quality control The conventional salting-out protocol of Miller et al. (1988) was applied in extracting genomic DNA from ethylenediaminetetraacetic acid (EDTA) anticoagulated whole blood samples. Samples were genotyped using the GoldenGate assay (Illumina, San Diego). Microtitre plates contained mixtures of patients and controls; two individuals (2.1%) were genotyped as technical controls on each plate, and genotype replicate consistency was 100%. We applied a standardized quality

J. Treutlein et al. / Schizophrenia Research 111 (2009) 123–130

control to the raw data, i.e. the genotype scatterplot as a graphical representation of normalized probe intensities of each marker was visually assessed by two investigators using Illumina's BeadStudio software. We applied a DNA sample call rate of ≥95% and a marker call rate of ≥96%. 2.4. Power calculation For power calculation, we assumed a) a multiplicative model of inheritance, b) a disease prevalence of 0.01 c) a disease allele frequency of 0.1 (0.25, 0.4) and d) an odds ratio of 1.5 (average odds ratio for the GRID1 SNPs reported by Guo et al., 2007). Calculations were performed using Quanto (Gauderman and Morrison, 2006). Analysis revealed that the power of our study to detect an effect of the abovementioned size was 0.966 (0.9996, 0.9999) without correction for multiple testing. With Bonferroni correction for the number of SNPs assessed, the power was 0.77 (0.987, 0.997).

125

2.5. Association analysis For differences between actual and expected frequencies of the haplotype-tagging SNPs, we employed the Hardy– Weinberg equilibrium (HWE) equation, as implemented in the PLINK program (Purcell et al., 2007). Single SNP association tests (two-tailed, significance level set at 0.05) were carried out using the Cochran–Armitage test for linear trend. To correct for multiple testing, we used a two-step approach: (1) To take into account the number of singlemarkers and their LD, we applied a Monte Carlo simulation method using the case–control analysis functionality of FAMHAP (Becker and Knapp, 2004) with the option “hapcc allcombi maxmarker 1”. The validity of this procedure has been demonstrated by Becker et al. (2005). (2) To additionally correct p-values for the number of all phenotypic groups, we applied a Bonferroni factor of 6, accounting for all phenotypic traits assessed, including the initial test and the number of

Fig. 1. Localization of the analysed polymorphisms and distribution of CpG islands at the GRID1 locus (UCSC Genome Browser on Human March 2006 Assembly). Numbers on the left indicate CpG counts. The CpG count is the number of CG dinucleotides in the island. The orientation of the human sequence is reversed since GRID1 is coded by the opposite strand. CpG islands b 300 bases are marked in light grey. The coordinates refer to the entire length of the displayed browser window.

126

J. Treutlein et al. / Schizophrenia Research 111 (2009) 123–130

Table 1 Single marker associations: Hardy–Weinberg equilibrium (HWE); HapMap haplotype blocks (shaded dark grey/light grey); block marker coverage at a haplotype frequency N=1%; p-values of single-markers; p-values below 0.05 are highlighted in bold; allelic odds ratios (OR).

The three markers which had been most significantly associated in the study of Guo et al. (2007) are boxed in black. a Cochran–Armitage-test for linear trend in proportions. b Frequency of allele “A” in cases. c Frequency of allele “A” in controls. d Phenotype information available for 758 patients. e Phenotype information available for 719 patients. f Phenotype information available for 817 patients.

J. Treutlein et al. / Schizophrenia Research 111 (2009) 123–130

subgroups which we analysed (i.e. the initial analysis of the DSM diagnosis of schizophrenia, schizophrenia patients with a lifetime history of major depression, a positive family history of schizophrenia, early age at onset, male and female gender). 3. Results 3.1. Putative transcriptional regulatory region of the GRID1 locus The distribution of CpG islands at the GRID1 locus is nonrandom. Across three species, CpG islands are clustered within the proximal 5′-flanking region and the first and second exons/introns of GRID1, but not in the rest of the gene (Fig. 1). The position of the largest CpG island is conserved around the start of transcription (Fig. 1). 3.2. Association analysis 20 of 22 SNPs of the GRID1 putative transcriptional regulatory region passed our quality control procedures. The final SNP set of 20 SNPs tagged common haplotypes within 5 blocks, distributed across 82.2 kb transcriptional regulatory region. Three SNPs (rs3814614, rs10749535, rs11201985) showed nominally significant associations (pb 0.05) with the overall diagnosis of schizophrenia (Table 1). Following correction for multiple testing for 20 markers, these results were no longer significant. The three SNPs reported as being most significant in the study of Guo et al. (2007) were not found to be associated. A further five phenotypic groups were tested for genetic association: female gender, male gender, positive family history, early age at onset and lifetime history of major depression. Marker rs3814614 had an unadjusted p-value of p = 0.0005 for schizophrenia patients with a lifetime history of major depression, which is the strongest association signal of our

127

study (Table 1). In the other subgroup analyses, p-values did not survive correction for multiple testing. To account for the burden of multiple testing, we corrected the p-value of each SNP by a two-step approach: (1) To consider the number of single-markers and their LD, we applied a Monte Carlo simulation method to all analyses of the 6 phenotypes. The best marker-corrected p-value was found for rs3814614 in the subgroup of schizophrenia patients with a lifetime history of major depression (p = 0.005). (2) We then applied a Bonferroni factor of 6, accounting for all phenotypic tests. Following this step, rs3814614 (investigated in the subgroup with a lifetime history of major depression) was the only SNP amongst all comparisons to remain significant after two-step correction for multiple testing (p = 0.030). To illustrate our approach we have also carried out a Monte Carlo simulation using the PLINK program (Purcell et al., 2007) which might be more familiar to the readers. Using this method to correct for the number of markers, with simultaneous consideration of their LD, yielded essentially the same results. The nominal and adjusted p-values resulting from this analysis of the patients with a lifetime history of major depression and controls are given in Table 2. The only significant marker after two-step correction for multiple testing, rs3814614, is located ~2.0 kb from the transcriptional start point and ~0.9 kb from the island with 356 CpG counts. This island has a length of ~ 4.4 kb and spans a part of the proximal promoter, the entire first and second exon as well as parts of intron 2. The results therefore suggest the relevance of GRID1 for those schizophrenia patients with a lifetime history of major depression. 3.3. Assessment of the LD structure The pairwise D′ and r2 values of the SNPs in the control individuals are illustrated in Fig. 2. For all 23 SNPs at the

Table 2 Results of Monte Carlo simulation analysis using PLINK: patients with lifetime history of major depression vs. all controls; p-values below 5% are highlighted in bold; rows are ordered by p-values. SNP

Asymptotic single marker p-value

Simulation based single marker p-value

p-value after correction for number of markers using MC-simulation

p-value after further Bonferroni correction for number of assessed phenotypes (n = 6)

rs3814614 rs10788494 rs10887578 rs10749535 rs11201985 rs17106600 rs2018507 rs4933391 rs10887579 rs11202014 rs2217223 rs10887583 rs4934199 rs4454627 rs10509530 rs1863823 rs1902666 rs1863826 rs11591408 rs11201988 rs999383 rs7907234 rs733013

0.0005 0.0015 0.0015 0.0024 0.0026 0.0143 0.0153 0.0154 0.0208 0.0241 0.0243 0.0397 0.0463 0.0604 0.0630 0.0682 0.0938 0.1199 0.2444 0.2479 0.3113 0.3962 0.8808

0.0005 0.0017 0.0019 0.0027 0.0033 0.0153 0.0159 0.0166 0.0213 0.0255 0.0257 0.0410 0.0489 0.0670 0.0587 0.0754 0.0961 0.1220 0.2448 0.2618 0.3099 0.4182 0.9162

0.0073 0.0222 0.0225 0.0353 0.0380 0.1739 0.1813 0.1824 0.2350 0.2601 0.2629 0.3841 0.4314 0.5211 0.5355 0.5591 0.6739 0.7642 0.9534 0.9569 0.9841 0.9968 1.0000

0.0436 0.1331 0.1348 0.2117 0.2278 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000

128

J. Treutlein et al. / Schizophrenia Research 111 (2009) 123–130

Fig. 2. LD structure for 23 SNPs at the GRID1 locus, based on the genotypes from all German controls (genome build 36 for coordinates and gene structure). Pairwise D′ values are reflected by the GOLD heatmap (Abecasis and Cookson, 2000), while the corresponding r2 values are given in the diamonds.

GRID1 locus, the observed LD values ranged from 0 to 1 (median ± standard deviation, D′ = 0.78 ± 0.36, r2 = 0.09 ± 0.24). For association analysis, the LD between SNPs was taken into account by the Monte Carlo simulation method of FAMHAP, as described in Materials and methods section. 4. Discussion An increase or decrease in the expression of any gene involved in glutamatergic neurotransmission can provide a plausible explanation of the symptoms of schizophrenia according to a genetic interpretation of the “glutamate hypothesis”. Early research into the role of glutamate in schizophrenia mainly focussed on the N-methyl-D-aspartic acid (NMDA) receptor system (Kornhuber et al., 1989; Paus et al., 2004; Clinton and Meador-Woodruff, 2004). Most of the glutamatergic genes reported to confer susceptibility to schizophrenia were therefore either NMDA receptor subunits, or proteins with a modulatory effect upon them (e.g., Schumacher et al., 2004). Recent genetic research has also implicated non-NMDA glutamatergic receptors (namely alpha-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid (AMPA) and metabotropic glutamate receptors) as being susceptibility genes for schizophrenia (Egan et al., 2004; Beneyto and Meador-Woodruff, 2006; O'Connor et al., 2007). The

association findings of the GRID1 gene with schizophrenia in two medium-sized studies (Fallin et al., 2005: 274 trios; Guo et al., 2007: 260 cases and 307 controls) indicated that variation at delta glutamate receptors may also play a role. In our study, no association was observed for the three SNPs which had previously been reported to be associated with schizophrenia in Chinese Northern Han individuals (Guo et al., 2007). It seems unlikely that the lack of association can be explained by an underpowered study design, given the size of our sample, which included 919 cases and 773 controls and which had a resulting power of 0.966 (without correction for multiple testing) to detect an effect of 1.5 (average OR of the GRID1 SNPs reported by Guo et al., 2007). Besides the possibility that the first reports were false positives, a plausible explanation for non-replication may be the different LD structure between individuals of German and Chinese ethnicity. The assumption of differences in LD at the GRID1 locus is supported by the observation that heterozygosity differed greatly between Askenazi Jewish (Fallin et al., 2005) and Chinese Northern Han samples (Guo et al., 2007). Marker rs2607857, e.g., was monomorphic in the Chinese Northern Han in contrast to the Askenazi Jewish population (Guo et al., 2007). Another reason for non-replication of the markers of Guo et al. (2007) in our German sample may be heterogeneity

J. Treutlein et al. / Schizophrenia Research 111 (2009) 123–130

between the Chinese and German study populations: recruitment bias, phenotypic and genetic heterogeneity are common reasons that may cause conflicting results between studies (e.g., Massat et al., 2005; Jablensky, 2006). In our systematic investigation of the putative transcriptional regulatory region, we obtained association between the markers rs3814614, rs10749535 and rs11201985 and a diagnosis of schizophrenia, a finding which did not remain significant after correction for multiple testing. However, the clinical picture of schizophrenia is heterogeneous, and previous research on modelling genetic heterogeneity has demonstrated the importance of using clinically defined subgroups for association studies (Schulze and McMahon, 2004; Schulze et al., 2005). We therefore stratified our patient group for sex, family history, early age at onset and lifetime history of major depression. In a high proportion of schizophrenia patients, depressive episodes occur before or after the onset of schizophrenia symptoms and constitute an important clinical feature (Siris and Bench, 2003; Häfner et al., 2005). Schizophrenia patients (n = 277; 30.1%) from our sample had suffered from at least one major depressive episode. Formal genetic studies suggest that this comorbidity may have a common genetic background (DeLisi et al., 1987; Maier et al., 1993; Kendler et al., 1997). Schizophrenia patients with depressive episodes may thus represent a distinct and more genetically homogeneous subgroup. This hypothesis is further supported by previous genetic association studies which found associations with the subtype of schizophrenia patients with depressive symptoms (Schumacher et al., 2005; Hamshere et al., 2006; Corvin et al., 2007; Georgi et al., 2007; Boks et al., 2008). For example, genetic variation in the brain derived neurotrophic factor (BDNF) gene (Schumacher et al., 2005), which regulates glutamatergic synapses (Carvalho et al., 2008), is a frequently reported molecular risk factor for both schizophrenia and major depression (Angelucci et al., 2005). Members of the glutamate receptor gene family, which includes GRID1, have also been reported as contributing to the development of these disorders (Schiffer, 2002; McNally et al., 2008). The association finding for GRID1 did not survive correction for multiple testing in the sample with the overall diagnosis of schizophrenia, but showed stronger association with the subgroup of patients with lifetime symptoms of major depression. The p-values of the schizophrenia association finding for the three markers decreased by approximately one order of magnitude (i.e. rs3814614: p = 0.0193 to p = 0.0 0 05; rs10749535: p = 0.0245 to p = 0.0 024; rs11201985: p = 0.0222 to p = 0.0026), despite a substantial reduction in the total sample size, with marker rs3814614 (p = 0.0005) withstanding correction for multiple testing. However, further true association findings may have disappeared through correction for multiple testing. The finding for rs3814614 suggests that this subgroup is genetically more homogeneous. A limitation of the present study is that it is not possible to differentiate whether our finding is due to an association of GRID1 with depressive symptoms in schizophrenia, or with depression itself. Future replication studies are warranted in order to clarify whether GRID1 is implicated in the etiology of schizophrenia and/or whether it modifies the risk for major depression in schizophrenia. These studies should specifically include subgroups of schizophrenia

129

patients with a lifetime history of major depression, and populations of depressed patients. Role of funding source The funding sources, which are all mentioned in the acknowledgements, had no further role in study design, in the collection, analysis and interpretation of data, in the writing of the report, or in the decision to submit the paper for publication. Contributors Authors Marcella Rietschel, Sven Cichon and Markus M. Nöthen designed the study. Authors Jana Strohmaier, Katja Veronika Bößhenz, Torsten Paul, Thomas G. Schulze, Ralf G. M. Schlösser, Wolfgang Maier, Igor Nenadic and Heinrich Sauer phenotyped the individuals and collected the samples. Authors Jens Treutlein, Thomas W. Mühleisen, Christine Schmael, Tsendsesmee Treutlein and Stephanie H. Witt conducted the literature searches and analyses, bioinformatic analyses and the writing of the first draft of the manuscript. Authors Thomas W. Mühleisen, Stefan Herms and Kerstin U. Ludwig selected the haplotype-tagging SNPs and performed the genotyping. Authors René Breuer, Josef Frank, Manuel Mattheisen and Tim Becker collected the data and undertook the statistical analyses. All authors contributed to and have approved the final manuscript. Conflict of interest All authors declare that they have no conflicts of interest. Acknowledgements We thank Margrieta Alblas and Axel M. Hillmer for their expert technical assistance. This work was supported by the National Genome Research Network (NGFN) of the German Federal Ministry of Education and Research (BMBF) to MR, MMN and SC. MMN and SC received support from the Alfried Krupp von Bohlen und Halbach-Stiftung. The study was also supported by grants from the IZKF Jena (BMBF-FKZ 01 ZZ 0405 to IN), NARSAD (Young Investigator Award to TGS) and the German Research Foundation (BE 3828/3-1 to TB).

References Abecasis, G.R., Cookson, W.O., 2000. GOLD—graphical overview of linkage disequilibrium. Bioinformatics 16, 182–183. an der Heiden, W., Könnecke, R., Maurer, K., Ropeter, D., Häfner, H., 2005. Depression in the long-term course of schizophrenia. Eur. Arch. Psychiatry Clin. Neurosci. 255 (3), 174–184. Angelucci, F., Brenè, S., Mathé, A.A., 2005. BDNF in schizophrenia, depression and corresponding animal models. Mol. Psychiatry 10 (4), 345–352. Antequera, F., 2003. Structure, function and evolution of CpG island promoters. Cell Mol. Life Sci. 60 (8), 1647–1658. Barrett, J.C., Fry, B., Maller, J., Daly, M.J., 2005. Haploview: analysis and visualization of LD and haplotype maps. Bioinformatics 21 (2), 263–265. Becker, T., Knapp, M., 2004. Maximum-likelihood estimation of haplotype frequencies in nuclear families. Genet. Epidemiol. 27 (1), 21–32. Becker, T., Cichon, S., Jönson, E., Knapp, M., 2005. Multiple testing in the context of haplotype analysis revisited: application to case–control data. Ann. Hum. Genet. 69 (Pt 6), 747–756. Beneyto, M., Meador-Woodruff, J.H., 2006. Lamina-specific abnormalities of AMPA receptor trafficking and signaling molecule transcripts in the prefrontal cortex in schizophrenia. Synapse 60 (8), 585–598. Boks, M.P., Leask, S., Vermunt, J.K., Kahn, R.S., 2007. The structure of psychosis revisited: the role of mood symptoms. Schizophr. Res. 93 (1–3), 178–185. Boks, M.P., Hoogendoorn, M., Jungerius, B.J., Bakker, S.C., Sommer, I.E., Sinke, R.J., Ophoff, R.A., Kahn, R.S., 2008. Do mood symptoms subdivide the schizophrenia phenotype? Association of the GMP6A gene with a depression subgroup. Am. J. Med. Genet. B Neuropsychiatr. Genet. 147B (6), 707–711. Byrne, M., Agerbo, E., Mortensen, P.B., 2002. Family history of psychiatric disorders and age at first contact in schizophrenia: an epidemiological study. Br. J. Psychiatr. Suppl. 43, s19–s25. Carvalho, A.L., Caldeira, M.V., Santos, S.D., Duarte, C.B., 2008. Role of the brain-derived neurotrophic factor at glutamatergic synapses. Br. J. Pharmacol. 153 (Suppl 1), S310–S324. Clinton, S.M., Meador-Woodruff, J.H., 2004. Abnormalities of the NMDA receptor and associated intracellular molecules in the thalamus in schizophrenia and bipolar disorder. Neuropsychopharmacology 29 (7), 1353–1362. Corvin, A., Donohoe, G., McGhee, K., Murphy, K., Kenny, N., Schwaiger, S., Nangle, J.M., Morris, D., Gill, M., 2007. D-amino acid oxidase (DAO)

130

J. Treutlein et al. / Schizophrenia Research 111 (2009) 123–130

genotype and mood symptomatology in schizophrenia. Neurosci. Lett. 426 (2), 97–100. DeLisi, L.E., Goldin, L.R., Maxwell, M.E., Kazuba, D.M., Gershon, E.S., 1987. Clinical features of illness in siblings with schizophrenia or schizoaffective disorder. Arch. Gen. Psychiatry 44 (10), 891–896. Egan, M.F., Straub, R.E., Goldberg, T.E., Yakub, I., Callicott, J.H., Hariri, A.R., Mattay, V.S., Bertolino, A., Hyde, T.M., Shannon-Weickert, C., Akil, M., Crook, J., Vakkalanka, R.K., Balkissoon, R., Gibbs, R.A., Kleinman, J.E., Weinberger, D.R., 2004. Variation in GRM3 affects cognition, prefrontal glutamate, and risk for schizophrenia. Proc. Natl. Acad. Sci. U.S.A. 101 (34), 12604–12609. Fallin, M.D., Lasseter, V.K., Wolyniec, P.S., McGrath, J.A., Nestadt, G., Valle, D., Liang, K.Y., Pulver, A.E., 2003. Genomewide linkage scan for schizophrenia susceptibility loci among Ashkenazi Jewish families shows evidence of linkage on chromosome 10q22. Am. J. Hum. Genet. 73 (3), 601–611. Fallin, M.D., Lasseter, V.K., Avramopoulos, D., Nicodemus, K.K., Wolyniec, P.S., McGrath, J.A., Steel, G., Nestadt, G., Liang, K.Y., Huganir, R.L., Valle, D., Pulver, A.E., 2005. Bipolar I disorder and schizophrenia: a 440-singlenucleotide polymorphism screen of 64 candidate genes among Ashkenazi Jewish case-parent trios. Am. J. Hum. Genet. 77 (6), 918–936. Fangerau, H., Ohlraun, S., Granath, R.O., Nöthen, M.M., Rietschel, M., Schulze, T.G., 2004. Computer-assisted phenotype characterization for genetic research in psychiatry. Hum. Hered. 58 (3–4), 122–130. Faraone, S.V., Hwu, H.G., Liu, C.M., Chen, W.J., Tsuang, M.M., Liu, S.K., Shieh, M.H., Hwang, T.J., Ou-Yang, W.C., Chen, C.Y., Chen, C.C., Lin, J.J., Chou, F. H., Chueh, C.M., Liu, W.M., Hall, M.H., Su, J., Van Eerdewegh, P., Tsuang, M.T., 2006. Genome scan of Han Chinese schizophrenia families from Taiwan: confirmation of linkage to 10q22.3. Am. J. Psychiatry 163 (10), 1760–1766. First, M.B., Spitzer, R.L., Gibbon, M., Williams, J., 1994. Structured Clinical Interview for Axis I DSM-IV Disorders. Biometrics Research, New York. Gabriel, S.B., Schaffner, S.F., Nguyen, H., Moore, J.M., Roy, J., Blumenstiel, B., Higgins, J., DeFelice, M., Lochner, A., Faggart, M., Liu-Cordero, S.N., Rotimi, C., Adeyemo, A., Cooper, R., Ward, R., Lander, E.S., Daly, M.J., Altshuler, D., 2002. The structure of haplotype blocks in the human genome. Science 296 (5576), 2225–2229. Gardiner-Garden, M., Frommer, M., 1987. CpG islands in vertebrate genomes. J. Mol. Biol. 196, 261–282. Gauderman, W.J. and Morrison, J.M., 2006. QUANTO 1.1: A computer program for power and sample size calculations for genetic-epidemiology studies. http://hydra.usc.edu/gxe. Georgi, A., Jamra, R.A., Klein, K., Villela, A.W., Schumacher, J., Becker, T., Paul, T., Schmael, C., Höfels, S., Klopp, N., Illig, T., Propping, P., Cichon, S., Nöthen, M.M., Schulze, T.G., Rietschel, M., 2007. Possible association between genetic variants at the GRIN1 gene and schizophrenia with lifetime history of depressive symptoms in a German sample. Psychiatr. Genet. 17 (5), 308–310. Guo, S.Z., Huang, K., Shi, Y.Y., Tang, W., Zhou, J., Feng, G.Y., Zhu, S.M., Liu, H.J., Chen, Y., Sun, X.D., He, L., 2007. A case–control association study between the GRID1 gene and schizophrenia in the Chinese Northern Han population. Schizophr. Res. 93 (1–3), 385–390. Häfner, H., Maurer, K., Trendler, G., an der Heiden, W., Schmidt, M., Könnecke, R., 2005. Schizophrenia and depression: challenging the paradigm of two separate diseases—a controlled study of schizophrenia, depression and healthy controls. Schizophr. Res. 77 (1), 11–24. Hamshere, M.L., Williams, N.M., Norton, N., Williams, H., Cardno, A.G., Zammit, S., et al., 2006. Genome-wide significant linkage in schizophrenia conditioning on occurrence of depressive episodes. J. Med. Genet. 43 (7), 563–567. Jablensky, A., 2006. Subtyping schizophrenia: implications for genetic research. Mol. Psychiatry 11 (9), 815–836. Karolchik, D., Baertsch, R., Diekhans, M., Furey, T.S., Hinrichs, A., Lu, Y.T., Roskin, K.M., Schwartz, M., Sugnet, C.W., Thomas, D.J., Weber, R.J., Haussler, D., Kent, W.J., 2003. The UCSC genome browser database. Nucl. Acids Res. 31 (1), 51–54. Kendler, K.S., MacLean, C.J., 1990. Estimating familial effects on age at onset and liability to schizophrenia. I. Results of a large sample family study. Genet. Epidemiol. 7 (6), 409–417. Kendler, K.S., Karkowski-Shuman, L., O'Neill, F.A., Straub, R.E., MacLean, C.J., Walsh, D., 1997. Resemblance of psychotic symptoms and syndromes in affected sibling pairs from the Irish Study of High-Density Schizophrenia Families: evidence for possible etiologic heterogeneity. Am. J. Psychiatry 154 (2), 191–198. Kim, J.S., Kornhuber, H.H., Schmid-Burgk, W., Holzmuller, B., 1980. Low cerebrospinal fluid glutamate in schizophrenic patients and a new hypothesis on schizophrenia. Neurosci. Lett. 20 (3), 379–382.

Kornhuber, J., Mack-Burkhardt, F., Riederer, P., Hebenstreit, G.F., Reynolds, G.P., Andrews, H.B., Beckmann, H., 1989. [3H]MK-801 binding sites in postmortem brain regions of schizophrenic patients. J. Neural Transm. 77 (2–3), 231–236. Lerer, B., Segman, R.H., Hamdan, A., Kanyas, K., Karni, O., Kohn, Y., Korner, M., Lanktree, M., Kaadan, M., Turetsky, N., Yakir, A., Kerem, B., Macciardi, F., 2003. Genome scan of Arab Israeli families maps a schizophrenia susceptibility gene to chromosome 6q23 and supports a locus at chromosome 10q24. Mol. Psychiatry 8 (5), 488–498. Maier, W., Lichtermann, D., Minges, J., Hallmayer, J., Heun, R., Benkert, O., Levinson, D.F., 1993. Continuity and discontinuity of affective disorders and schizophrenia. Results of a controlled family study. Arch. Gen. Psychiatry 50 (11), 871–883. Massat, I., Souery, D., Del-Favero, J., Nothen, M., Blackwood, D., Muir, W., Kaneva, R., Serretti, A., Lorenzi, C., Rietschel, M., Milanova, V., Papadimitriou, G.N., Dikeos, D., Van Broekhoven, C., Mendlewicz, J., 2005. Association between COMT (Val158Met) functional polymorphism and early onset in patients with major depressive disorder in a European multicenter genetic association study. Mol. Psychiatry 10 (6), 598–605. McGuffin, P., Farmer, A.E., Harvey, I., 1991. A polydiagnostic application of the operational criteria in studies of psychotic illness. Arch. Gen. Psychiatry 48 (8), 764–770. McNally, L., Bhagwagar, Z., Hannestad, J., 2008. Inflammation, glutamate, and glia in depression: a literature review. CNS Spectr. 13 (6), 501–510. Miller, S.A., Dykes, D.D., Polesky, H.F., 1988. A simple salting out procedure for extracting DNA from human nucleated cells. Nucleic Acids Res. 16 (3), 1215. Myers, S.J., Dingledine, R., Borges, K., 1999. Genetic regulation of glutamate receptor ion channels. Annu. Rev. Pharmacol. Toxicol. 39, 221–241. O'Connor, J.A., Muly, E.C., Arnold, S.E., Hemby, S.E., 2007. AMPA receptor subunit and splice variant expression in the DLPFC of schizophrenic subjects and rhesus monkeys chronically administered antipsychotic drugs. Schizophr. Res. 90 (1–3), 28–40. Paus, S., Rietschel, M., Schulze, T.G., Ohlraun, S., Diaconu, C.C., Van Den Bogaert, A., Maier, W., Propping, P., Cichon, S., Nothen, M.M., 2004. Systematic screening for mutations in the human N-methyl-D-aspartate receptor 1 gene in schizophrenic patients from the German population. Psychiatr. Genet. 14 (4), 233–234. Pilowsky, L.S., Bressan, R.A., Stone, J.M., Erlandsson, K., Mulligan, R.S., Krystal, J.H., Ell, P.J., 2006. First in vivo evidence of an NMDA receptor deficit in medicationfree schizophrenic patients. Mol. Psychiatry 11 (2), 118–119. Pulver, A.E., Brown, C.H., Wolyniec, P., McGrath, J., Tam, D., Adler, L., Carpenter, W.T., Childs, B., 1990. Schizophrenia: age at onset, gender and familial risk. Acta Psychiatr. Scand. 82 (5), 344–351. Purcell, S., Neale, B., Todd-Brown, K., Thomas, L., Ferreira, M.A.R., Bender, D., Maller, J., Sklar, P., de Bakker, P.I.W., Daly, M.J., Sham, P.C., 2007. PLINK: a toolset for whole-genome association and population-based linkage analysis. Am. J. Hum. Genet. 81 (3), 559–575. Schiffer, H.H., 2002. Glutamate receptor genes: susceptibility factors in schizophrenia and depressive disorders? Mol. Neurobiol. 25 (2), 191–212. Schulze, T.G., McMahon, F.J., 2004. Defining the phenotype in human genetic studies: forward genetics and reverse phenotyping. Hum. Hered. 58 (3–4), 131–138. Schulze, T.G., Ohlraun, S., Czerski, P.M., Schumacher, J., Kassem, L., Deschner, M., et al., 2005. Genotype–phenotype studies in bipolar disorder showing association between the DAOA/G30 locus and persecutory delusions: a first step toward a molecular genetic classification of psychiatric phenotypes. Am. J. Psychiatry 162 (11), 2101–2108. Schumacher, J., Jamra, R.A., Freudenberg, J., Becker, T., Ohlraun, S., Otte, A.C., Tullius, M., Kovalenko, S., Bogaert, A.V., Maier, W., Rietschel, M., Propping, P., Nothen, M.M., Cichon, S., 2004. Examination of G72 and D-amino-acid oxidase as genetic risk factors for schizophrenia and bipolar affective disorder. Mol. Psychiatry 9 (2), 203–207. Schumacher, J., Jamra, R.A., Becker, T., Ohlraun, S., Klopp, N., Binder, E.B., Schulze, T.G., Deschner, M., Schmäl, C., Höfels, S., Zobel, A., Illig, T., Propping, P., Holsboer, F., Rietschel, M., Nöthen, M.M., Cichon, S., 2005. Evidence for a relationship between genetic variants at the brain-derived neurotrophic factor (BDNF) locus and major depression. Biol. Psychiatry 58 (4), 307–314. Siris, S.G., Bench, C., 2003. Depression and schizophrenia. In: Hirsch, S.R., Weinberger, D.R. (Eds.), Schizophrenia. Blackwell Publishing, Oxford, pp. 142–167. Suvisaari, J.M., Haukka, J., Tanskanen, A., Lönnqvist, J.K., 1998. Age at onset and outcome in schizophrenia are related to the degree of familial loading. Br. J. Psychiatry 173, 494–500.