Introduction
Schizophrenia has a substantial genetic basis, and a significant number of these genetic risk variants have recently been identified through genome-wide analyses (Sullivan et al. Reference Sullivan, Daly and O'Donovan2012). The mechanisms through which variation in risk genes, copy number variants, or rare mutations affect the developing brain and interact in its arising pathophysiology are still far from being clear. The ZNF804A gene (coding for zinc finger protein 804A) has been suggested to harbour common risk variants for schizophrenia (O'Donovan et al. Reference O'Donovan, Craddock, Norton, Williams, Peirce, Moskvina, Nikolov, Hamshere, Carroll, Georgieva, Dwyer, Holmans, Marchini, Spencer, Howie, Leung, Hartmann, Möller, Morris, Shi, Feng, Hoffmann, Propping, Vasilescu, Maier, Rietschel, Zammit, Schumacher, Quinn, Schulze, Williams, Giegling, Iwata, Ikeda, Darvasi, Shifman, He, Duan, Sanders, Levinson, Gejman, Cichon, Nothen, Gill, Corvin, Rujescu, Kirov, Owen, Buccola, Mowry, Freedman, Amin, Black, Silverman, Byerley and Cloninger2008), which has been replicated in several studies across different ethnic groups (Li et al. Reference Li, Shi, Shi, Luo, Zheng, Li, Liu, Chong, Lee, Wang, Liu, Yin, Pu, Diao, Xu and Su2012; Zhang et al. Reference Zhang, Yan, Valenzuela, Lu, Du, Zhong, Ren, Zhao, Gao, Wang, Guo and Ma2012; Schwab et al. Reference Schwab, Kusumawardhani, Dai, Qin, Wildenauer, Agiananda, Amir, Antoni, Arsianti, Asmarahadi, Diatri, Djatmiko, Irmansyah, Khalimah, Kusumadewi, Kusumaningrum, Lukman, Mustar, Nasrun, Naswati, Prasetiyawan, Semen, Siste, Tobing, Widiasih, Wiguna, Wulandari, Benyamin and Wildenauer2013). The rs1344706 single nucleotide polymorphism (SNP) has been of particular interest, as it has been linked to brain structural, functional and cognitive variation (Donohoe et al. Reference Donohoe, Morris and Corvin2010). Two recent studies have shown an impact of ZNF804A rs1344706 genetic variation on cognitive function and brain activation patterns in tasks related to cognitive control (Thurin et al. Reference Thurin, Rasetti, Sambataro, Safrin, Chen, Callicott, Mattay and Weinberger2013) and working memory for faces (Linden et al. Reference Linden, Lancaster, Wolf, Baird, Jackson, Johnston, Donev and Thome2013), respectively, as well as connectivity measures in healthy subjects (Paulus et al. Reference Paulus, Krach, Bedenbender, Pyka, Sommer, Krug, Knake, Nothen, Witt, Rietschel, Kircher and Jansen2013). As such functions are putative endophenotypes for schizophrenia, the link to ZNF804A might provide an understanding on how the gene or gene products act on the biological substrates of psychosis, including schizophrenia and other psychiatric disorders (Donohoe et al. Reference Donohoe, Morris and Corvin2010; Hess & Glatt, Reference Hess and Glatt2014).
ZNF804A might, however, also affect brain development; hence, brain structural changes related to variation in one or more SNPs may thus reflect an enduring impact of the gene's function (Donohoe et al. Reference Donohoe, Morris and Corvin2010). A first volumetric study of ZNF804A (rs1344706) in healthy subjects showed larger total white matter volumes in risk-allele carriers along with reduced grey matter in the angular gyrus, parahippocampal gyrus, posterior cingulate and medial orbitofrontal gyrus (Lencz et al. Reference Lencz, Szeszko, DeRosse, Burdick, Bromet, Bilder and Malhotra2010). A subsequent volumetric study in schizophrenia patients and healthy controls demonstrated an effect of the same marker on total and frontal white matter volume in schizophrenia patients (risk-allele carriers having larger volumes), but only total white matter volume in healthy controls (Wassink et al. Reference Wassink, Epping, Rudd, Axelsen, Ziebell, Fleming, Monson, Ho and Andreasen2012). A more detailed recent voxel-based morphometry (VBM) study also revealed effects on regional grey matter in schizophrenia patients and healthy controls (Donohoe et al. Reference Donohoe, Rose, Frodl, Morris, Spoletini, Adriano, Bernardini, Caltagirone, Bossu, Gill, Corvin and Spalletta2011), whereby homozygous risk-allele carriers for rs1344706 differed in regional grey matter of the dorsolateral prefrontal cortex, hippocampus and amygdala compared with heterozygotes and non-risk-allele homozygotes. This last study, in particular, also suggested that effects of this polymorphism might, in fact, diverge markedly between healthy controls and schizophrenia patients, i.e. that effects vary both regionally and in direction, such as findings of increased hippocampal grey matter in risk-allele homozygotes in the patient sample, but not within the control sample (Donohoe et al. Reference Donohoe, Rose, Frodl, Morris, Spoletini, Adriano, Bernardini, Caltagirone, Bossu, Gill, Corvin and Spalletta2011). Similarly, we found cortical thickness to show dissociated effects in patients and controls (Schultz et al. Reference Schultz, Nenadic, Riley, Vladimirov, Wagner, Koch, Schachtzabel, Muhleisen, Basmanav, Nothen, Deufel, Kiehntopf, Rietschel, Reichenbach, Cichon, Schlosser and Sauer2014), with risk-allele carriers within the patient group having thicker prefrontal and temporal cortices, while in the healthy control group risk-allele homozygotes had thinner cortices. Despite the existing evidence, the effects and regional distribution remain unclear, and as with many other imaging genetics findings in schizophrenia, there is a lack of replication.
In the present study, we aimed to extend the evidence for effects of rs1344706 and its risk allele (A) on brain structure in both healthy controls and schizophrenia subjects by using VBM to map regional brain structural differences on a voxel-by-voxel basis, with the particular emphasis to replicate and extend the previous findings on (a) the regional pattern of effects, and (b) the divergence of effects between schizophrenia patients and healthy controls.
Method
Subjects
We included 62 patients with a Diagnostic and Statistical Manual of Mental Disorders, fourth edition (DSM-IV) diagnosis of schizophrenia (44 male, 18 female; mean age 31.6 years, s.d. = 11.5 years; all on stable antipsychotic medication) and 54 healthy controls from the community (29 male, 25 female; mean age 29.5 years, s.d. = 9.9 years), who had no current or previous psychiatric treatment or diagnosis. All subjects provided written informed consent to a study protocol approved by the Jena Medical School Ethics Committee, in accordance with the Declaration of Helsinki. Groups did not differ in age (t test: p = 0.291, two-tailed) or gender (χ 2 test: p = 0.055). Groups differed in overall performance on the MWT-B (Mehrfachwahl-Wortschatz-Intelligenztest B), an estimate of pre-morbid intelligence quotient (IQ) [available for 58 schizophrenia patients and 37 controls; analysis of variance (ANOVA): F = 9.312, p = 0.003]. None of the subjects had any major medical or neurological condition. Part of this sample has been used in a recent analysis of ZNF804A effects on cortical thickness (Schultz et al. Reference Schultz, Nenadic, Riley, Vladimirov, Wagner, Koch, Schachtzabel, Muhleisen, Basmanav, Nothen, Deufel, Kiehntopf, Rietschel, Reichenbach, Cichon, Schlosser and Sauer2014), whereas for the present analysis, the sample has been expanded to allow analysis of a larger cohort with another methodological approach.
Genotyping using Sanger sequencing
Individual genotypes for the studied polymorphism were retrieved from sequence data generated by Sanger sequencing. Genomic DNA sequences covering rs1344706 and the flanking up- and downstream regions were retrieved from the UCSC Genome Browser based on the human genome build GRCh37/hg19 (http://genome.ucsc.edu/cgi-bin/hgGateway). Primer design was performed by keeping a minimum distance of about 150 base pairs (bp) up- and downstream from the target base. Primer sequences are obtainable from the authors upon request. Amplicons were generated under standard polymerase chain reaction (PCR) conditions. Resequencing was carried out with BigDye v3.1 (Applied Biosystems, USA) sequencing reagents employed according to the manufacturer's protocol. Data were generated on the Applied Biosystems 3130xl Genetic Analyzer, and SeqMan II (DNASTAR Inc., USA) software was used for data visualization. Electropherograms were evaluated independently by two raters.
Genotype distribution in schizophrenia patients was 23, 33 and six for AA, CA and CC, respectively, and 19, 28 and seven in healthy controls. Genotype distributions did not deviate significantly from Hardy–Weinberg equilibrium in patients and controls (p patients = 0.234, p controls = 0.504). The rs1344706 minor (C) allele frequency (MAF) observed in our samples (MAF = 0.375) was very similar to the MAF observed in 1000 Genomes (MAF = 0.393; http://www.1000genomes.org/). ANOVA with the factors diagnosis and genotype did not show effects of diagnosis (p = 0.758), genotype (p = 0.188) or their interaction (p = 0.543) on the variable age. For pre-morbid IQ estimates (see above, available for 58 patients and 37 controls), there was also no effect of genotype (p = 0.496), or diagnosis x genotype interaction (p = 0.114), but only for factor diagnosis (as shown above).
Magnetic resonance imaging (MRI) and VBM
We obtained T1-weighted high-resolution MRI scans on a 1.5 T Siemens Magnetom Vision plus system (Siemens, Germany) using a three-dimensional Fast Low Angle SHot (FLASH) sequence (repetition time = 15 ms, echo time = 5 ms, flip angle α = 30°, 192 sagittal slices, field of view = 256 × 256 mm, with voxel dimensions 1 × 1 × 1 mm3). Images were visually inspected for artefacts and underwent automated quality control using an algorithm provided in the VBM toolbox (http://dbm.neuro.uni-jena.de/vbm8/).
For post-processing and VBM analysis, we used the VBM8 toolbox, a freely available toolbox based on Statistical Parametric Mapping software SPM8 (Institute of Neurology, UK) and Matlab (Mathworks, USA). VBM8 is based on the general VBM approach proposed by Ashburner & Friston (Reference Ashburner and Friston2000, Reference Ashburner and Friston2005), implementing the Diffeomorphic Anatomical Registration Through Exponentiated Lie Algebra (DARTEL) algorithm for non-linear normalization (Ashburner, Reference Ashburner2007). Grey matter and white matter maps were extracted from MRI data for each subject and underwent non-linear normalization using DARTEL. We defined an internal grey matter threshold of 0.2 for both grey matter maps and white matter maps, thus choosing a more conservative value to limit edge effects. Smoothing was performed with a 12 mm full-width at half-maximum (FWHM) Gaussian kernel for grey matter maps and 20 mm FWHM for white matter maps. The Anatomical Automatic Labeling (AAL) toolbox was used for anatomical labelling (Tzourio-Mazoyer et al. Reference Tzourio-Mazoyer, Landeau, Papathanassiou, Crivello, Etard, Delcroix, Mazoyer and Joliot2002). In addition to regional analysis of grey and white matter, the VBM toolbox also estimates total brain grey matter and white matter volume by assigning the corresponding number of voxels to either tissue type multiplied by the voxel volumes, thus allowing analysis of effects of genotype on global parameters as well.
Statistical analysis
We first used multivariate ANOVA (with SPSS 20; IBM, USA) to test effects of genotype on total brain grey matter and total white matter volumes by defining the factors group (schizophrenia; controls) and genotype (AA; CA; CC) and the covariates age and gender (to remove variance related to these variables), and using Pillai's trace.
VBM statistics were carried out in SPM8 using a general linear model defining diagnostic group (schizophrenia; healthy control) and genotype (AA; CA; CC) while using age and gender as covariates (again, to remove their effect on variance) and applying a height threshold of p < 0.001 (uncorrected) throughout all analyses. This choice of threshold was based on the existence of previous VBM analyses, as mentioned above, and thus allowed us to either replicate (or refute) previous findings in these mentioned areas, while whole-brain testing allowed us to expand analysis to other areas as well. Our main hypothesis of divergent effects of rs1344706 was tested with an interaction of group x genotype; furthermore, we also tested effects of genotype in schizophrenia patients and effects of genotype in healthy controls separately; we performed each of these three tests for both grey and white matter, respectively.
Results
Total brain grey matter and white matter
Overall multivariate ANOVA (for grey matter and white matter) revealed no significant effect of factor diagnosis (F 2,107 = 0.223, p = 0.801) or genotype (F 4,216 = 1.831, p = 0.124), but a significant diagnosis x genotype interaction (F 4,216 = 3.109, p = 0.016). Tests of between-subject effects revealed significant group differences only for the interaction of diagnosis x genotype for total brain white matter (F 2,116 = 3.318, p = 0.04), but not for grey matter (F 2,216 = 1.267, p = 0.286) and neither for grey nor white matter for the diagnosis or genotype effects alone. Both age and gender had significant effects on total brain grey matter and white matter, respectively (both p < 0.001).
VBM analysis of grey matter
In schizophrenia patients, there were several significant regional effects of genotype on grey matter (see Fig. 1), including the left temporal pole and anterior part of the inferior temporal gyrus, bilateral lateral temporal cortices, right superior temporal and supramarginal gyri, bilateral inferior prefrontal gyri, left medial cerebellum, left orbitofrontal cortex and posterior right thalamus.
Voxel-based morphometry analysis of effects of ZNF804A genotype (rs1344706) on grey matter in schizophrenia patients (p < 0.001, uncorrected).
In healthy controls, we found several significant regional effects of genotype (see Fig. 2), including the left middle temporal gyrus (lateral temporal cortex), right insular cortex, left orbitofrontal cortex and right superior temporal cortex.
Voxel-based morphometry analysis of effects of ZNF804A genotype (rs1344706) on grey matter in healthy controls (p < 0.001, uncorrected).
The interaction analysis of diagnosis x genotype for regional grey matter also resulted in significant findings (see Fig. 3) in some of the above-mentioned clusters. Of note, parts of three of these clusters actually showed significance levels that also survived family-wise error (FWE) correction for multiple comparisons (as implemented in SPM, peak-level): one cluster in the left orbitofrontal cortex (coordinates −6, 62, −20; FWE-corrected p = 0.021), right middle/superior temporal gyrus (69, −39, 6; FWE-corrected p = 0.029) and left middle temporal gyrus (−68, −39, 3; FWE-corrected p = 0.034) (see Table 1 and Fig. 4).
Voxel-based morphometry analysis of interaction effects of ZNF804A genotype (rs1344706) x diagnostic group (schizophrenia; healthy controls) (p < 0.001, uncorrected).
Render view of voxel-based morphometry analysis of interaction effects of ZNF804A genotype (rs1344706) x diagnostic group (schizophrenia; healthy controls) (p < 0.05, family-wise error correction for multiple comparisons).
Anatomical overview of regions with significant effects (p < 0.001, uncorrected) in voxel-based morphometry analysis of ZNF804A (rs1344706) genotype effects (only clusters with at least k = 10 voxels listed) in schizophrenia patients, healthy controls, and interaction effects of diagnosis x genotype
MNI, Montreal Neurological Institute.
* Parts of these clusters are also significant at p < 0.05, family-wise error corrected.
Results of group comparison of grey matter differences between patients and healthy controls are provided in online Supplementary Fig. S1.
In addition to the above results, we inspected the plots for profiles of genotype-related changes, i.e. to identify the subgroup contributing most to the effects. Assessing the distribution of grey matter values across the groups, a rather consistent pattern emerged. For the interaction analysis, the effect was caused by smaller grey matter values for AA homozygotes (compared with AC heterozygotes) in the healthy control group, whereas in the schizophrenia group grey matter values for AA homozygotes were higher compared with AC heterozygotes. The CC homozygotes often took an intermediary position. However, these subgroups were very small, thus making direct comparison more difficult. For the genotype effects within each group, we found that AA homozygotes in the schizophrenia group had higher grey matter values than the AC heterozygotes across all significant larger clusters. Within the healthy control group analysis, inspection of plots also confirmed smaller grey matter values for AA homozygotes (in significant clusters) in contrast to AC heterozygotes (except for the anterior insula cluster, where the two groups were similar, but CC homozygotes showed larger effects), thus confirming the overall pattern of an inverted (or diverging) effect of AA versus AC carrier status in schizophrenia patients versus controls.
VBM analysis of white matter
There were no significant effects of genotype on white matter in either schizophrenia patients or healthy controls, and no significant effects of a diagnosis x genotype interaction for white matter (all at the predefined threshold of p < 0.001).
Discussion
As its main result, our study identified effects of ZNF804A on regional brain structure in the left orbitofrontal and left and right lateral temporal cortices with a significant diagnostic group x genotype interaction. In these regions, variation in rs1344706 led to diverging effects in patients and controls. Further major effects (at uncorrected levels of significance) were seen in multiple frontal, temporal and parietal cortical areas, where the gene's impact was different for the two diagnostic groups. While these findings add to the increasing identification of brain structural effects of this schizophrenia risk gene, they partially confirm a peculiar effect of this SNP. This effect is related to the markedly different impact of the SNP in patients and healthy subjects. Previous morphometry studies have already suggested such effects on grey matter density (Donohoe et al. Reference Donohoe, Rose, Frodl, Morris, Spoletini, Adriano, Bernardini, Caltagirone, Bossu, Gill, Corvin and Spalletta2011), as well as on cortical thickness in our own previous work (Schultz et al. Reference Schultz, Nenadic, Riley, Vladimirov, Wagner, Koch, Schachtzabel, Muhleisen, Basmanav, Nothen, Deufel, Kiehntopf, Rietschel, Reichenbach, Cichon, Schlosser and Sauer2014). There is also increasing evidence for ZNF804A effects on brain function. Two recent electroencephalography (EEG) studies revealed an impact of rs1344706 on P300 amplitude/responses (Del Re et al. Reference Del Re, Bergen, Mesholam-Gately, Niznikiewicz, Goldstein, Woo, Shenton, Seidman, McCarley and Petryshen2014; O'Donoghue et al. Reference O'Donoghue, Morris, Fahey, Da Costa, Moore, Cummings, Leicht, Karch, Hoerold, Tropea, Foxe, Gill, Corvin and Donohoe2014).
Imaging genetics studies have generally aimed to evaluate the effect of particular risk genes (or risk scores) on brain structures. Often, however, it is unclear how gene expression or gene products might influence brain structure in later life. For ZNF804A, there is growing evidence that effects start during early brain development. This has been shown in fetal brains, where rs1344706 variation had a strong impact on ZNF804A allelic expression during the second trimester of intra-uterine development, but not in adult brains (Hill & Bray, Reference Hill and Bray2012). Expression in postmortem brains seems to differ between patients and healthy controls, but this observation is only incompletely understood (Okada et al. Reference Okada, Hashimoto, Yamamori, Umeda-Yano, Yasuda, Ohi, Fukumoto, Ikemoto, Kunii, Tomita, Ito and Takeda2012; Schultz et al. Reference Schultz, Nenadic, Riley, Vladimirov, Wagner, Koch, Schachtzabel, Muhleisen, Basmanav, Nothen, Deufel, Kiehntopf, Rietschel, Reichenbach, Cichon, Schlosser and Sauer2014), as is the regional pattern of physiological expression (Buonocore et al. Reference Buonocore, Hill, Campbell, Oladimeji, Jeffries, Troakes, Hortobagyi, Williams, Cooper and Bray2010). Also, recent studies suggest an indirect action of ZNF804A by modulating the gene expression of other risk genes (Girgenti et al. Reference Girgenti, LoTurco and Maher2012).
While in many instances genetic risk factors have demonstrated stronger impact of a particular risk gene in patients compared with controls, this might not be the case (at least not generally) for rs1344706. While the mentioned VBM study of Donohoe et al. (Reference Donohoe, Rose, Frodl, Morris, Spoletini, Adriano, Bernardini, Caltagirone, Bossu, Gill, Corvin and Spalletta2011) reported preserved brain structure in patients, one volumetric study found larger regional white matter volumes in risk-allele carriers, together, however, with more severe psychotic symptoms (Wassink et al. Reference Wassink, Epping, Rudd, Axelsen, Ziebell, Fleming, Monson, Ho and Andreasen2012). In contrast, a large study of healthy subjects did not find statistically significant effects on brain structure (Cousijn et al. Reference Cousijn, Rijpkema, Harteveld, Harrison, Fernandez, Franke and Arias-Vasquez2012). Given that volumetry, VBM and analysis of cortical thickness assess three different aspects of brain morphometry, our findings are most comparable with the studies of Donohoe et al. (Reference Donohoe, Rose, Frodl, Morris, Spoletini, Adriano, Bernardini, Caltagirone, Bossu, Gill, Corvin and Spalletta2011) with respect to methods. In fact, these authors analysed schizophrenia and healthy control cohorts of roughly similar size to ours (using the older VBM5 toolbox). While their regional findings only partially overlap with ours (including anterior temporal pole and temporal cortical areas, although with somewhat different location), they also share similarities. The authors reported that AA risk homozygotes showed opposite effects in patient versus control groups, with higher grey matter density for AA homozygotes in patients, but not controls. Similarly, effects of ZNF804A on cortical thickness have received support from two recent studies (Voineskos et al. Reference Voineskos, Lerch, Felsky, Tiwari, Rajji, Miranda, Lobaugh, Pollock, Mulsant and Kennedy2011; Schultz et al. Reference Schultz, Nenadic, Riley, Vladimirov, Wagner, Koch, Schachtzabel, Muhleisen, Basmanav, Nothen, Deufel, Kiehntopf, Rietschel, Reichenbach, Cichon, Schlosser and Sauer2014), though there has been also one failure to replicate (Bergmann et al. Reference Bergmann, Haukvik, Brown, Rimol, Hartberg, Athanasiu, Melle, Djurovic, Andreassen, Dale and Agartz2013).
Functional studies, however, do suggest effects of ZNF804A genotypes in patients to exceed those seen in healthy controls (Rasetti et al. Reference Rasetti, Sambataro, Chen, Callicott, Mattay and Weinberger2011). Given that there are also a number of studies demonstrating effects of rs1344706 in healthy subjects (Esslinger et al. Reference Esslinger, Walter, Kirsch, Erk, Schnell, Arnold, Haddad, Mier, Opitz von Boberfeld, Raab, Witt, Rietschel, Cichon and Meyer-Lindenberg2009; Walter et al. Reference Walter, Schnell, Erk, Arnold, Kirsch, Esslinger, Mier, Schmitgen, Rietschel, Witt, Nothen, Cichon and Meyer-Lindenberg2011; Linden et al. Reference Linden, Lancaster, Wolf, Baird, Jackson, Johnston, Donev and Thome2013; Mohnke et al. Reference Mohnke, Erk, Schnell, Schutz, Romanczuk-Seiferth, Grimm, Haddad, Pohland, Garbusow, Schmitgen, Kirsch, Esslinger, Rietschel, Witt, Nothen, Cichon, Mattheisen, Muhleisen, Jensen, Schott, Maier, Heinz, Meyer-Lindenberg and Walter2014), further studies might elucidate (within the same cohort) the interaction of structural and functional effects.
Our study failed to identify any affects of ZNF804A on white matter density. Although VBM of white matter is inferior to diffusion tensor imaging (DTI) applications, which captures also quantitative measures of white matter damage (e.g. assessing fractional anisotropy), our results fit into the current literature on ZNF804A and white matter tracts. Previous DTI studies have mostly failed to confirm a clear impact of ZNF804A on white matter architecture or fractional anisotropy measures (Voineskos et al. Reference Voineskos, Lerch, Felsky, Tiwari, Rajji, Miranda, Lobaugh, Pollock, Mulsant and Kennedy2011; Sprooten et al. Reference Sprooten, McIntosh, Lawrie, Hall, Sussmann, Dahmen, Konrad, Bastin and Winterer2012; Wei et al. Reference Wei, Kang, Diao, Guidon, Wu, Zheng, Li, Guo, Hu, Zhang, Liu and Zhao2013). One previous white matter VBM study (of T1-weighted MRI) suggested an interaction of group and genotype (rs1344706) in left prefrontal white matter (Wei et al. Reference Wei, Kang, Diao, Shan, Li, Zheng, Guo, Liu, Zhang and Zhao2012), but our study did not replicate this finding. In two more recent studies, one found lower fractional anisotropy in risk-allele carriers in several regions including the corpus callosum and left forceps minor (Ikuta et al. Reference Ikuta, Peters, Guha, John, Karlsgodt, Lencz, Szeszko and Malhotra2014), while a second study failed to show any significant impact of ZNF804A on white matter microstructure in a large healthy subject sample (O'Donoghue et al. Reference O'Donoghue, Morris, Fahey, Da Costa, Moore, Cummings, Leicht, Karch, Hoerold, Tropea, Foxe, Gill, Corvin and Donohoe2014).
Conclusions
In conclusion, our study provides evidence for significant divergence of ZNF804A (rs1344706) effects on grey matter in prefrontal and temporal brain areas. Effects of prefrontal, temporal and parietal grey matter differ not only according to genotype, but the genotype-related effects are different between schizophrenia patients and controls. There was, however, no effect on white matter density. While our results confirm divergence of structural effects across different groups, further studies are needed to clarify the microstructural changes underlying these effects.
Supplementary material
For supplementary material accompanying this paper visit http://dx.doi.org/10.1017/S0033291714001159.
Acknowledgements
I.N. was supported by a Junior Scientist Grant from IZKF (Interdisiplinäres Zentrum für klinische Forschung, Jena). The work of M.M.N. and S.C. was supported by the German Federal Ministry of Education and Research (BMBF) through the Integrated Genome Research Network (IG) MooDS (Systematic Investigation of the Molecular Causes of Major Mood Disorders and Schizophrenia), under the auspices of the National Genome Research Network plus (NGFNplus) and through the Integrated Network IntegraMent (Integrated Understanding of Causes and Mechanisms in Mental Disorders), under the auspices of the e:Med Programme. M.M.N. is a member of the German Research Foundation (DFG)-funded Excellence-Cluster ImmunoSensation. He also received support from the Alfried Krupp von Bohlen und Halbach-Stiftung.
Declaration of Interest
None.
