RESEARCH ARTICLE Neuropsychiatric Genetics
Genetic Association of ACSM1 Variation with Schizophrenia and Major Depressive Disorder in the Han Chinese Population Wenjin Li,1,2,3a Weidong Ji,5a Zhiqiang Li,1,2,3 Kuanjun He,1,2,3 Qingzhong Wang,1,2,3 Jianhua Chen,1,2,3,4 Yu Qiang,1,2,3 Guoyin Feng,4 Xingwang Li,1 Jiawei Shen,1,2,3 Zujia Wen,1,2,3 Jue Ji,1,2,3 and Yongyong Shi1,2,3* 1
Bio-X Institutes, Key Laboratory for the Genetics of Developmental and Neuropsychiatric Disorders, Ministry of Education, Shanghai Jiao Tong University, Shanghai, P.R. China 2
Bio-X Institutes, Key Laboratory of Social Cognitive and Behavioral Sciences, Shanghai Jiao Tong University, Shanghai, P.R. China Institute of Neuropsychiatric Science and Systems Biological Medicine, Shanghai Jiao Tong University, Shanghai, P.R. China
3 4
Shanghai Mental Health Center, Shanghai, P.R. China
5
Shanghai Changning Mental Health Center, Shanghai, P.R. China
Manuscript Received: 20 August 2014; Manuscript Accepted: 25 November 2014
Schizophrenia (SCZ) and major depressive disorder (MDD) are two of the most common and severe mental disorders, the etiologies of which are not yet clearly elucidated. The ACSM1 gene has been identified as a susceptibility gene for SCZ in two previous genomewide association studies (GWAS). ACSM1 catalyzes the activation of fatty acids and plays an important role in the metabolic system. Some evidence has suggested that ACSM1 contributes to a genetic risk for MDD. The present study aimed to evaluate the common genetic risk of the ACSM1 gene in these two disorders in the Han Chinese population. In total, 1235 patients with SCZ, 1045 patients with MDD and 1235 control subjects of Chinese origin were recruited. Six single nuclear polymorphisms (SNPs) in ACSM1 were genotyped to test their associations with SCZ and MDD. SNP rs163234 was found to be significantly associated with both SCZ (permutated Pallele ¼ 1.700 103, OR ¼ 1.350 [95% CI ¼ 1.152–1.581]) and MDD (permutated Pallele ¼ 4.800 103, OR ¼ 1.329 [95% CI ¼ 1.127– 1.567]). SNP rs433598 showed a strong association with SCZ (permutated Pallele ¼ 4.300 103, OR ¼ 1.303 [95% CI ¼ 1.117– 1.520]). Haplotype analysis of the blocks containing the two positive markers also revealed a significant association. This is the first study to assess the possible association of the ACSM1 gene with a genetic susceptibility for MDD. Our data are the first to suggest a positive association of the ACSM1 gene with a genetic susceptibility for SCZ and MDD in the Han Chinese population. Ó 2015 Wiley Periodicals, Inc.
Key words: Schizophrenia; major depressive disorder; ACSM1; association; SNP
INTRODUCTION Psychiatric disorders place a large burden not only on affected individuals and their families but also on society and health services.
Ó 2015 Wiley Periodicals, Inc.
How to Cite this Article: Li W,Ji W, Li Z, He K, Wang Q, Chen J, Qiang Y, Feng G, Li X, Shen J, Wen Z, Ji J, Shi Y. 2015. Genetic Association of ACSM1 Variation with Schizophrenia and Major Depressive Disorder in the Han Chinese Population. Am J Med Genet Part B 168B:144–149.
Schizophrenia (SCZ) and major depressive disorder (MDD) are two of the most common and most severe mental disorders and are consistently related to increased disability. The prevalence of SCZ is up to 1% worldwide, and the lifetime risk of MDD ranges from 5% to 25% [Demyttenaere et al., 2004]. The highest lifetime rates of suicidal behavior were also found in patients with schizoaffective disorder and major depression with psychotic features [Radomsky et al., 1999]. In addition, numerous studies have suggested that SCZ and MDD, similar to other mental illnesses, may be typically complex diseases involving both genetic and environmental factors [Vaswani and Kapur, 2001; Kendler Ks, 2005] with substantial a These authors contributed equally to this paper. Conflict of interest: None. Correspondence, Address for correspondence: Dr. Yongyong Shi, Bio-X Institutes, Key Laboratory for the Genetics of Developmental and Neuropsychiatric Disorders (Ministry of Education), Shanghai Jiao Tong University, Shanghai 200030, China. E-mail: [email protected] Article first published online in Wiley Online Library (wileyonlinelibrary.com) DOI 10.1002/ajmg.b.32291
144
LI ET AL. heritabilities of 80% and 37%, respectively [Baron, 2001; Kendler, 2001; Levinson, 2006]. Despite the presence of so significant a genetic component, the etiology underlying psychiatric disorders remains to be elucidated. Moreover, research from different areas has provided convergent lines of evidence for a common genetic risk background in mental illnesses, such as SCZ, MDD and bipolar disorder [Baum et al., 2008; Schulze et al., 2012]. The psychopathological dimensions and psychiatric symptoms shared by patients further support this overlap. Acyl-CoA synthetase medium-chain family member 1 (ACSM1), located on chromosome 16p12.3, has medium-chain fatty acid: CoA ligase activity with broad substrate specificity. Its major role lies in the degradation of medium-chain fatty acids for the production of energy [Fujino et al., 2001]. Previous research has shown that genetic polymorphisms in the acyl-CoA synthetase gene family have contributed to multiple risk factors, especially hypertriglyceridemia [Iwai et al., 2003]. Notably, accumulating evidence has suggested that individuals with schizophrenia have high levels of metabolic disorders and cardiovascular risk factors [Mitchell et al., 2013]. Recently, Sullivan et al. identified certain alleles of ACSM1 as among the top 25 susceptibility markers of schizophrenia without attaining genome-wide significance [Sullivan et al., 2008]. Next, the significant association between ACSM1 and schizophrenia was replicated in Europeans [Athanasiu et al., 2010]. However, up to now, there has been no replication of the association between ACSM1 and neuropsychiatric diseases in the Han Chinese population. In this study, we not only undertook analysis to validate the associations of the common variants of ACSM1 with SCZ previously reported in European population but also genotyped those polymorphisms in MDD patients to reveal potential cross-disorder associations.
EXPERIMENTAL PROCEDURES Participants In total, 1,235 unrelated SCZ cases (mean age SD, 36.4 9.0; 805 males and 430 females) and 1,045 unrelated MDD cases (mean age SD, 34.4 12.1; 729 males and 316 females) were studied, who were diagnosed by at least two independent psychiatrists according to Diagnostic and Statistical Manual of Mental Disorders Fourth Edition (DSM-IV). Patients with two or more types of mental illness or substance dependence were excluded. A group of 1,235 normal controls (mean age SD, 30.6 11.4; 665 males and 570 females) were healthy volunteers randomly selected with no family history of psychiatric diseases in their first degree relatives. All subjects were recruited from the Han Chinese population and originated from Shanghai. The study was reviewed and approved by the local Ethical Committee of Human Genetics Resources. Informed consent was obtained from all participants before data collection.
Genotyping Rs234993 and rs433598 were given preferential selection because of their significant associations reported in previous GWAS studies [Sullivan et al., 2008; Athanasiu et al., 2010]. A total of 6 SNPs, including rs163234, rs2301672, rs234993, rs3087812, rs433598 and rs9930187 in ACSM1, were selected using the Haploview software
145 version 4.2 with the r2 threshold set at 0.8 and with the HapMap 3 release 27 CHB database as reference[Barrett et al., 2005]. Except for rs3087812, which was located in the 30 UTR, and rs2301672 in exon11, the rest of the markers were all intronic SNPs (supplementary Figure S1). All probes were designed and synthesized by Life Technology. Genomic DNA was extracted from the peripheral blood samples of the subjects using the QuickGene DNA whole blood kit L (FUJIFILM) protocol. SNP genotyping was performed by fluores1 cence-based TaqMan SNP discrimination assays on high-throughput Fludigm EP1TM platforms. Genotype calls were further reviewed manually to correct any uncertain calls due to clustering artifacts.
Statistical Analysis Pearson chi-square tests were used to determine the differences in the distribution of allele and genotype frequencies between cases and controls, as performed on the SHEsis software platform (http:// analysis.bio-x.cn/) [Shi and He, 2005; Li et al., 2009]. All genotype frequencies of the control population were tested against the Hardy–Weinberg equilibrium (HWE) using the chi-square test. The HWE test, linkage disequilibrium analysis and haplotype selection were performed using Haploview software. The marker-based 10,000 permutation tests for multiple test adjustments and the haplotype association analysis were carried out by the PLINK software (http://pngu.mgh.harvard.edu/purcell/plink/) [Purcell et al., 2007]. All of the P values presented were two-sided, and only significant results with a permuted P value < 0.05 were considered noteworthy associations.
Population Stratification Analysis STRUCTURE version 2.3.4 [Pritchard et al., 2000] was used to correct for ethnic stratification. Overall, 79 random SNPs from 174 Hapmap CEU samples, 209 Hapmap YRI samples and 139 samples randomly selected from our samples were included in the admixture model and correlated frequencies model, with a burn-in length of 10,000 and MCMC (Markov chain Monte Carlo) repeats of 10,000. We made several runs for each K (number of assumed populations) from 2 to 7 to make sure the results were consistent.
RESULTS Single Site Associations The allelic and genotype frequencies of the 6 SNPs in ACSM1 were determined in all of the subjects (Table I). The average genotyping call rate for all markers was 99.7%. No significant deviation from HWE was found for all of the SNPs in either the case or control samples. The analysis with the highest likelihood was considered to provide the best-fit result for each SNP. In SCZ case-control samples, a significant association was observed for rs163234 (permutated Pallele ¼ 1.700 103, OR ¼ 1.350 [95% CI ¼ 1.152– 1.581]) and rs433598 (permutated Pallele ¼ 4.300 103, OR ¼ 1.303 [95% CI ¼ 1.117–1.520]). Additionally, a comparison between patients with MDD and the controls revealed positive results in rs163234 (permutated Pallele ¼ 4.800 103, OR ¼ 1.329 [95% CI ¼ 1.127–1.567]). The minor allele A of rs163234 was more frequent in MDD patients than in controls.
Allele counts(frequencies) A G 412 (0.167) 2058 (0.833) 344 (0.165) 1746 (0.835) 319 (0.129) 2151 (0.871) A G 1203 (0.490) 1253 (0.510) 981 (0.482) 1055 (0.518) 1230 (0.499) 1236 (0.501) C T 578 (0.236) 1872 (0.764) 481 (0.232) 1595 (0.768) 549 (0.223) 1915 (0.777) C T 901 (0.366) 1560 (0.634) 817 (0.392) 1267 (0.608) 947 (0.384) 1519 (0.616) C T 430 (0.174) 2036 (0.826) 277 (0.139) 1719 (0.861) 344 (0.139) 2122 (0.861) A G 2142 (0.871) 318 (0.129) 1796 (0.876) 254 (0.124) 2168 (0.880) 296 (0.120)
OR (95%CI) 1.350 (1.152–1.581) 1.329 (1.127–1.567) 0.965 (0.863–1.079) 0.934 (0.831–1.051) 1.077 (0.943–1.230) 1.052 (0.915–1.209) 0.926 (0.825–1.040) 1.034 (0.918–1.166) 1.303 (1.117–1.520) 0.994 (0.838–1.179) 0.920 (0.777–1.089) 0.965 (0.807–1.155)
Pallele (permP) 1.960E-04 (1.700E-03) 7.220E-04 (4.800E-03) 5.295E-01 (9.725E-01) 2.574E-01 (7.482E-01) 2.745E-01 (7.754E-01) 4.763E-01 (9.538E-01) 1.942E-01 (1.700E-03) 5.805E-01 (9.873E-01) 7.670E-04 (4.300E-03) 9.450E-01 (1.000E þ 00) 3.318E-01 (8.484E-01) 6.997E-01 (9.974E-01)
Genotype counts(frequency) AA AG GG 33 (0.027) 346 (0.280) 856 (0.693) 33 (0.032) 278 (0.266) 734 (0.702) 17 (0.014) 285 (0.231) 933 (0.755) AA AG GG 284 (0.231) 635 (0.517) 309 (0.252) 223 (0.219) 535 (0.526) 260 (0.255) 304 (0.247) 622 (0.504) 307 (0.249) CC CT TT 78 (0.064) 422 (0.344) 725 (0.592) 49 (0.047) 383 (0.369) 606 (0.584) 66 (0.054) 417 (0.338) 749 (0.608) CC CT TT 173 (0.141) 555 (0.451) 502 (0.408) 159 (0.153) 499 (0.479) 384 (0.369) 199 (0.161) 549 (0.445) 485 (0.393) CC CT TT 45 (0.036) 340 (0.276) 848 (0.688) 19 (0.019) 239 (0.239) 740 (0.741) 24 (0.019) 296 (0.240) 913 (0.740) AA AG GG 928 (0.754) 286 (0.233) 16 (0.013) 799 (0.780) 198 (0.193) 28 (0.027) 958 (0.778) 252 (0.205) 22 (0.018)
1.678E-01 (6.025E-01) 2.691E-01 (7.895E-01)
2.711E-03 (1.350E-02) 9.966E-01 (1.000E þ 00)
3.433E-01 (8.705E-01) 2.759E-01 (7.975E-01)
4.965E-01 (9.650E-01) 2.883E-01 (8.113E-01)
6.666E-01 (9.968E-01) 3.052E-01 (8.307E-01)
7.810E-04 (4.100E-03) 1.358E-03 (6.799E-03)
Pgeno (permP)
Abbreviations: SCZ, schizophrenia patients; MDD, major depressive disorder patients; C, controls; SNP, single-nucleotide polymorphism; P (permP), uncorrected P-value (corrected P-value based on 10,000 permutations); OR (odds ratio). Significant P values (permP < 0.05) are in bold.
SNP rs163234 SZ MDD C rs2301672 SZ MDD C rs234993 SZ MDD C rs3087812 SZ MDD C rs433598 SZ MDD C rs9930187 SZ MDD C
TABLE I. Allele and Genotype Analysis for ACSM1 in the SCZ and MDD Case-Control Samples
146 AMERICAN JOURNAL OF MEDICAL GENETICS PART B
LI ET AL.
147 P value ¼ 1.840 102, OR ¼ 0.826) in block 1, consisting of rs3087812, rs2301672, rs234993, and rs163234 (Table II). In block 2, a haplotype (C-A) including the minor allele (C) of rs433598 was over-represented in SCZ patients compared with controls (permutated P value ¼ 8.000 104, OR ¼ 1.550). Meanwhile, a notable association with SCZ was also found in a protective haplotype (T-A) carrying the major allele of rs433598 (permutated P value ¼ 5.299 103, OR ¼ 0.797).
Population Stratification Analysis The potential population stratification analysis is illustrated by the triangle charts produced by STRUCTURE software (supplementary Figure S2). The combined population of CEU, YRI and our samples displayed a stratified pattern, whereas the two disorders and controls were distributed evenly in the triangle. Thus, there was no significant stratification detected in our samples, indicating that our results should not be caused by population stratification.
DISCUSSION Fig. 1. Linkage disequilibrium analysis of the six SNPs investigated in healthy controls. r2 value between marker pairs is indicated by the shaded matrices. Two blocks were identified using Haploview software: block 1 (rs3087812-rs2301672rs234993-rs163234); block 2 (rs433598-rs9930187).
Haplotype Analysis We identified two haplotype blocks according to the pairwise linkage disequilibrium (LD) evaluation (Fig. 1 ). The structure of the haplotype blocks and pairwise LDs for the SCZ and MDD patient groups were roughly similar to those for controls (data not shown). One protective haplotype, T-A-T-G, was found to be nominally associated with SCZ (P value ¼ 4.970 102, OR ¼ 0.894) and significantly associated with MDD (permutated
In the present study, we carried out a genetic association analysis to evaluate the role of the ACSM1 gene in psychiatric diseases. Although the pathogenic mechanisms underlying psychiatric diseases are largely unknown, a growing amount of evidence from clinical, epidemiological, and molecular genetic studies, suggests that some genetic risk factors are shared among neuropsychiatric disorders [Cross-Disorder Group of the Psychiatric Genomics, 2013]. We therefore postulated that the genes associated with schizophrenia might also play a role in other mental disorders. To our knowledge, this is the first exploratory analysis on the possible association between polymorphisms of ACSM1 and MDD patients. With its relatively large sample size, this study provided initial evidence that common variants in ACSM1 affect the susceptibility to SCZ and MDD of the Han Chinese population. Both rs163234 and rs433598 were strongly associated with SCZ, and minor allele A and minor allele C were over-represented in the SCZ patients. Moreover, a significant association was found between
TABLE II. Haplotype Association Analysis Controls
SCZ
Frequency Frequency P (permP) Block1: rs3087812-rs2301672-rs234993-rs163234 OMNIBUS NA NA 3.698E-05 (9.999E-05) TATG 0.477 0.461 4.970E-02 (4.004E-01) Block2: rs433598-rs9930187 OMNIBUS NA NA 1.170E-03 (2.200E-03) CA 0.045 0.070 3.570E-04 (8.000E-04) TA 0.836 0.801 1.960E-03 (5.299E-03)
MDD OR
Frequency
P (permP)
OR
NA 0.894
NA 0.461
9.999E-06 (9.999E-05) 1.890E-03 (1.840E-02)
NA 0.826
NA 1.550 0.797
NA 0.04 0.835
4.500E-01 (4.770E-01) 3.928E-01 (1.000) 4.500E-01 (4.708E-01)
NA 0.797 0.835
Haplotypes observed in
Genetic Association of ACSM1 Variation with Schizophrenia and Major Depressive Disorder in the Han Chinese Population Wenjin Li,1,2,3a Weidong Ji,5a Zhiqiang Li,1,2,3 Kuanjun He,1,2,3 Qingzhong Wang,1,2,3 Jianhua Chen,1,2,3,4 Yu Qiang,1,2,3 Guoyin Feng,4 Xingwang Li,1 Jiawei Shen,1,2,3 Zujia Wen,1,2,3 Jue Ji,1,2,3 and Yongyong Shi1,2,3* 1
Bio-X Institutes, Key Laboratory for the Genetics of Developmental and Neuropsychiatric Disorders, Ministry of Education, Shanghai Jiao Tong University, Shanghai, P.R. China 2
Bio-X Institutes, Key Laboratory of Social Cognitive and Behavioral Sciences, Shanghai Jiao Tong University, Shanghai, P.R. China Institute of Neuropsychiatric Science and Systems Biological Medicine, Shanghai Jiao Tong University, Shanghai, P.R. China
3 4
Shanghai Mental Health Center, Shanghai, P.R. China
5
Shanghai Changning Mental Health Center, Shanghai, P.R. China
Manuscript Received: 20 August 2014; Manuscript Accepted: 25 November 2014
Schizophrenia (SCZ) and major depressive disorder (MDD) are two of the most common and severe mental disorders, the etiologies of which are not yet clearly elucidated. The ACSM1 gene has been identified as a susceptibility gene for SCZ in two previous genomewide association studies (GWAS). ACSM1 catalyzes the activation of fatty acids and plays an important role in the metabolic system. Some evidence has suggested that ACSM1 contributes to a genetic risk for MDD. The present study aimed to evaluate the common genetic risk of the ACSM1 gene in these two disorders in the Han Chinese population. In total, 1235 patients with SCZ, 1045 patients with MDD and 1235 control subjects of Chinese origin were recruited. Six single nuclear polymorphisms (SNPs) in ACSM1 were genotyped to test their associations with SCZ and MDD. SNP rs163234 was found to be significantly associated with both SCZ (permutated Pallele ¼ 1.700 103, OR ¼ 1.350 [95% CI ¼ 1.152–1.581]) and MDD (permutated Pallele ¼ 4.800 103, OR ¼ 1.329 [95% CI ¼ 1.127– 1.567]). SNP rs433598 showed a strong association with SCZ (permutated Pallele ¼ 4.300 103, OR ¼ 1.303 [95% CI ¼ 1.117– 1.520]). Haplotype analysis of the blocks containing the two positive markers also revealed a significant association. This is the first study to assess the possible association of the ACSM1 gene with a genetic susceptibility for MDD. Our data are the first to suggest a positive association of the ACSM1 gene with a genetic susceptibility for SCZ and MDD in the Han Chinese population. Ó 2015 Wiley Periodicals, Inc.
Key words: Schizophrenia; major depressive disorder; ACSM1; association; SNP
INTRODUCTION Psychiatric disorders place a large burden not only on affected individuals and their families but also on society and health services.
Ó 2015 Wiley Periodicals, Inc.
How to Cite this Article: Li W,Ji W, Li Z, He K, Wang Q, Chen J, Qiang Y, Feng G, Li X, Shen J, Wen Z, Ji J, Shi Y. 2015. Genetic Association of ACSM1 Variation with Schizophrenia and Major Depressive Disorder in the Han Chinese Population. Am J Med Genet Part B 168B:144–149.
Schizophrenia (SCZ) and major depressive disorder (MDD) are two of the most common and most severe mental disorders and are consistently related to increased disability. The prevalence of SCZ is up to 1% worldwide, and the lifetime risk of MDD ranges from 5% to 25% [Demyttenaere et al., 2004]. The highest lifetime rates of suicidal behavior were also found in patients with schizoaffective disorder and major depression with psychotic features [Radomsky et al., 1999]. In addition, numerous studies have suggested that SCZ and MDD, similar to other mental illnesses, may be typically complex diseases involving both genetic and environmental factors [Vaswani and Kapur, 2001; Kendler Ks, 2005] with substantial a These authors contributed equally to this paper. Conflict of interest: None. Correspondence, Address for correspondence: Dr. Yongyong Shi, Bio-X Institutes, Key Laboratory for the Genetics of Developmental and Neuropsychiatric Disorders (Ministry of Education), Shanghai Jiao Tong University, Shanghai 200030, China. E-mail: [email protected] Article first published online in Wiley Online Library (wileyonlinelibrary.com) DOI 10.1002/ajmg.b.32291
144
LI ET AL. heritabilities of 80% and 37%, respectively [Baron, 2001; Kendler, 2001; Levinson, 2006]. Despite the presence of so significant a genetic component, the etiology underlying psychiatric disorders remains to be elucidated. Moreover, research from different areas has provided convergent lines of evidence for a common genetic risk background in mental illnesses, such as SCZ, MDD and bipolar disorder [Baum et al., 2008; Schulze et al., 2012]. The psychopathological dimensions and psychiatric symptoms shared by patients further support this overlap. Acyl-CoA synthetase medium-chain family member 1 (ACSM1), located on chromosome 16p12.3, has medium-chain fatty acid: CoA ligase activity with broad substrate specificity. Its major role lies in the degradation of medium-chain fatty acids for the production of energy [Fujino et al., 2001]. Previous research has shown that genetic polymorphisms in the acyl-CoA synthetase gene family have contributed to multiple risk factors, especially hypertriglyceridemia [Iwai et al., 2003]. Notably, accumulating evidence has suggested that individuals with schizophrenia have high levels of metabolic disorders and cardiovascular risk factors [Mitchell et al., 2013]. Recently, Sullivan et al. identified certain alleles of ACSM1 as among the top 25 susceptibility markers of schizophrenia without attaining genome-wide significance [Sullivan et al., 2008]. Next, the significant association between ACSM1 and schizophrenia was replicated in Europeans [Athanasiu et al., 2010]. However, up to now, there has been no replication of the association between ACSM1 and neuropsychiatric diseases in the Han Chinese population. In this study, we not only undertook analysis to validate the associations of the common variants of ACSM1 with SCZ previously reported in European population but also genotyped those polymorphisms in MDD patients to reveal potential cross-disorder associations.
EXPERIMENTAL PROCEDURES Participants In total, 1,235 unrelated SCZ cases (mean age SD, 36.4 9.0; 805 males and 430 females) and 1,045 unrelated MDD cases (mean age SD, 34.4 12.1; 729 males and 316 females) were studied, who were diagnosed by at least two independent psychiatrists according to Diagnostic and Statistical Manual of Mental Disorders Fourth Edition (DSM-IV). Patients with two or more types of mental illness or substance dependence were excluded. A group of 1,235 normal controls (mean age SD, 30.6 11.4; 665 males and 570 females) were healthy volunteers randomly selected with no family history of psychiatric diseases in their first degree relatives. All subjects were recruited from the Han Chinese population and originated from Shanghai. The study was reviewed and approved by the local Ethical Committee of Human Genetics Resources. Informed consent was obtained from all participants before data collection.
Genotyping Rs234993 and rs433598 were given preferential selection because of their significant associations reported in previous GWAS studies [Sullivan et al., 2008; Athanasiu et al., 2010]. A total of 6 SNPs, including rs163234, rs2301672, rs234993, rs3087812, rs433598 and rs9930187 in ACSM1, were selected using the Haploview software
145 version 4.2 with the r2 threshold set at 0.8 and with the HapMap 3 release 27 CHB database as reference[Barrett et al., 2005]. Except for rs3087812, which was located in the 30 UTR, and rs2301672 in exon11, the rest of the markers were all intronic SNPs (supplementary Figure S1). All probes were designed and synthesized by Life Technology. Genomic DNA was extracted from the peripheral blood samples of the subjects using the QuickGene DNA whole blood kit L (FUJIFILM) protocol. SNP genotyping was performed by fluores1 cence-based TaqMan SNP discrimination assays on high-throughput Fludigm EP1TM platforms. Genotype calls were further reviewed manually to correct any uncertain calls due to clustering artifacts.
Statistical Analysis Pearson chi-square tests were used to determine the differences in the distribution of allele and genotype frequencies between cases and controls, as performed on the SHEsis software platform (http:// analysis.bio-x.cn/) [Shi and He, 2005; Li et al., 2009]. All genotype frequencies of the control population were tested against the Hardy–Weinberg equilibrium (HWE) using the chi-square test. The HWE test, linkage disequilibrium analysis and haplotype selection were performed using Haploview software. The marker-based 10,000 permutation tests for multiple test adjustments and the haplotype association analysis were carried out by the PLINK software (http://pngu.mgh.harvard.edu/purcell/plink/) [Purcell et al., 2007]. All of the P values presented were two-sided, and only significant results with a permuted P value < 0.05 were considered noteworthy associations.
Population Stratification Analysis STRUCTURE version 2.3.4 [Pritchard et al., 2000] was used to correct for ethnic stratification. Overall, 79 random SNPs from 174 Hapmap CEU samples, 209 Hapmap YRI samples and 139 samples randomly selected from our samples were included in the admixture model and correlated frequencies model, with a burn-in length of 10,000 and MCMC (Markov chain Monte Carlo) repeats of 10,000. We made several runs for each K (number of assumed populations) from 2 to 7 to make sure the results were consistent.
RESULTS Single Site Associations The allelic and genotype frequencies of the 6 SNPs in ACSM1 were determined in all of the subjects (Table I). The average genotyping call rate for all markers was 99.7%. No significant deviation from HWE was found for all of the SNPs in either the case or control samples. The analysis with the highest likelihood was considered to provide the best-fit result for each SNP. In SCZ case-control samples, a significant association was observed for rs163234 (permutated Pallele ¼ 1.700 103, OR ¼ 1.350 [95% CI ¼ 1.152– 1.581]) and rs433598 (permutated Pallele ¼ 4.300 103, OR ¼ 1.303 [95% CI ¼ 1.117–1.520]). Additionally, a comparison between patients with MDD and the controls revealed positive results in rs163234 (permutated Pallele ¼ 4.800 103, OR ¼ 1.329 [95% CI ¼ 1.127–1.567]). The minor allele A of rs163234 was more frequent in MDD patients than in controls.
Allele counts(frequencies) A G 412 (0.167) 2058 (0.833) 344 (0.165) 1746 (0.835) 319 (0.129) 2151 (0.871) A G 1203 (0.490) 1253 (0.510) 981 (0.482) 1055 (0.518) 1230 (0.499) 1236 (0.501) C T 578 (0.236) 1872 (0.764) 481 (0.232) 1595 (0.768) 549 (0.223) 1915 (0.777) C T 901 (0.366) 1560 (0.634) 817 (0.392) 1267 (0.608) 947 (0.384) 1519 (0.616) C T 430 (0.174) 2036 (0.826) 277 (0.139) 1719 (0.861) 344 (0.139) 2122 (0.861) A G 2142 (0.871) 318 (0.129) 1796 (0.876) 254 (0.124) 2168 (0.880) 296 (0.120)
OR (95%CI) 1.350 (1.152–1.581) 1.329 (1.127–1.567) 0.965 (0.863–1.079) 0.934 (0.831–1.051) 1.077 (0.943–1.230) 1.052 (0.915–1.209) 0.926 (0.825–1.040) 1.034 (0.918–1.166) 1.303 (1.117–1.520) 0.994 (0.838–1.179) 0.920 (0.777–1.089) 0.965 (0.807–1.155)
Pallele (permP) 1.960E-04 (1.700E-03) 7.220E-04 (4.800E-03) 5.295E-01 (9.725E-01) 2.574E-01 (7.482E-01) 2.745E-01 (7.754E-01) 4.763E-01 (9.538E-01) 1.942E-01 (1.700E-03) 5.805E-01 (9.873E-01) 7.670E-04 (4.300E-03) 9.450E-01 (1.000E þ 00) 3.318E-01 (8.484E-01) 6.997E-01 (9.974E-01)
Genotype counts(frequency) AA AG GG 33 (0.027) 346 (0.280) 856 (0.693) 33 (0.032) 278 (0.266) 734 (0.702) 17 (0.014) 285 (0.231) 933 (0.755) AA AG GG 284 (0.231) 635 (0.517) 309 (0.252) 223 (0.219) 535 (0.526) 260 (0.255) 304 (0.247) 622 (0.504) 307 (0.249) CC CT TT 78 (0.064) 422 (0.344) 725 (0.592) 49 (0.047) 383 (0.369) 606 (0.584) 66 (0.054) 417 (0.338) 749 (0.608) CC CT TT 173 (0.141) 555 (0.451) 502 (0.408) 159 (0.153) 499 (0.479) 384 (0.369) 199 (0.161) 549 (0.445) 485 (0.393) CC CT TT 45 (0.036) 340 (0.276) 848 (0.688) 19 (0.019) 239 (0.239) 740 (0.741) 24 (0.019) 296 (0.240) 913 (0.740) AA AG GG 928 (0.754) 286 (0.233) 16 (0.013) 799 (0.780) 198 (0.193) 28 (0.027) 958 (0.778) 252 (0.205) 22 (0.018)
1.678E-01 (6.025E-01) 2.691E-01 (7.895E-01)
2.711E-03 (1.350E-02) 9.966E-01 (1.000E þ 00)
3.433E-01 (8.705E-01) 2.759E-01 (7.975E-01)
4.965E-01 (9.650E-01) 2.883E-01 (8.113E-01)
6.666E-01 (9.968E-01) 3.052E-01 (8.307E-01)
7.810E-04 (4.100E-03) 1.358E-03 (6.799E-03)
Pgeno (permP)
Abbreviations: SCZ, schizophrenia patients; MDD, major depressive disorder patients; C, controls; SNP, single-nucleotide polymorphism; P (permP), uncorrected P-value (corrected P-value based on 10,000 permutations); OR (odds ratio). Significant P values (permP < 0.05) are in bold.
SNP rs163234 SZ MDD C rs2301672 SZ MDD C rs234993 SZ MDD C rs3087812 SZ MDD C rs433598 SZ MDD C rs9930187 SZ MDD C
TABLE I. Allele and Genotype Analysis for ACSM1 in the SCZ and MDD Case-Control Samples
146 AMERICAN JOURNAL OF MEDICAL GENETICS PART B
LI ET AL.
147 P value ¼ 1.840 102, OR ¼ 0.826) in block 1, consisting of rs3087812, rs2301672, rs234993, and rs163234 (Table II). In block 2, a haplotype (C-A) including the minor allele (C) of rs433598 was over-represented in SCZ patients compared with controls (permutated P value ¼ 8.000 104, OR ¼ 1.550). Meanwhile, a notable association with SCZ was also found in a protective haplotype (T-A) carrying the major allele of rs433598 (permutated P value ¼ 5.299 103, OR ¼ 0.797).
Population Stratification Analysis The potential population stratification analysis is illustrated by the triangle charts produced by STRUCTURE software (supplementary Figure S2). The combined population of CEU, YRI and our samples displayed a stratified pattern, whereas the two disorders and controls were distributed evenly in the triangle. Thus, there was no significant stratification detected in our samples, indicating that our results should not be caused by population stratification.
DISCUSSION Fig. 1. Linkage disequilibrium analysis of the six SNPs investigated in healthy controls. r2 value between marker pairs is indicated by the shaded matrices. Two blocks were identified using Haploview software: block 1 (rs3087812-rs2301672rs234993-rs163234); block 2 (rs433598-rs9930187).
Haplotype Analysis We identified two haplotype blocks according to the pairwise linkage disequilibrium (LD) evaluation (Fig. 1 ). The structure of the haplotype blocks and pairwise LDs for the SCZ and MDD patient groups were roughly similar to those for controls (data not shown). One protective haplotype, T-A-T-G, was found to be nominally associated with SCZ (P value ¼ 4.970 102, OR ¼ 0.894) and significantly associated with MDD (permutated
In the present study, we carried out a genetic association analysis to evaluate the role of the ACSM1 gene in psychiatric diseases. Although the pathogenic mechanisms underlying psychiatric diseases are largely unknown, a growing amount of evidence from clinical, epidemiological, and molecular genetic studies, suggests that some genetic risk factors are shared among neuropsychiatric disorders [Cross-Disorder Group of the Psychiatric Genomics, 2013]. We therefore postulated that the genes associated with schizophrenia might also play a role in other mental disorders. To our knowledge, this is the first exploratory analysis on the possible association between polymorphisms of ACSM1 and MDD patients. With its relatively large sample size, this study provided initial evidence that common variants in ACSM1 affect the susceptibility to SCZ and MDD of the Han Chinese population. Both rs163234 and rs433598 were strongly associated with SCZ, and minor allele A and minor allele C were over-represented in the SCZ patients. Moreover, a significant association was found between
TABLE II. Haplotype Association Analysis Controls
SCZ
Frequency Frequency P (permP) Block1: rs3087812-rs2301672-rs234993-rs163234 OMNIBUS NA NA 3.698E-05 (9.999E-05) TATG 0.477 0.461 4.970E-02 (4.004E-01) Block2: rs433598-rs9930187 OMNIBUS NA NA 1.170E-03 (2.200E-03) CA 0.045 0.070 3.570E-04 (8.000E-04) TA 0.836 0.801 1.960E-03 (5.299E-03)
MDD OR
Frequency
P (permP)
OR
NA 0.894
NA 0.461
9.999E-06 (9.999E-05) 1.890E-03 (1.840E-02)
NA 0.826
NA 1.550 0.797
NA 0.04 0.835
4.500E-01 (4.770E-01) 3.928E-01 (1.000) 4.500E-01 (4.708E-01)
NA 0.797 0.835
Haplotypes observed in
