Significant outcomes
-
∙ There was no statistically significant difference in different allele and genotype frequencies of rs1635, rs11038167, rs10489202 between schizophrenia (SZ) cases and controls in Zhuang ethnic population.
-
∙ No significant difference was observed between NKAPL rs1635, TSPAN18 rs11038167, MPC2 rs10489202 polymorphism and the susceptibility to SZ in Han ethnic population.
-
∙ Further meta-analysis suggested that single-nucleotide polymorphism (SNP) rs10489202 was significantly associated with SZ in the Han Chinese population.
Limitations
-
∙ The limited sample size may not have been powerful enough to identify the tiny influence of some SNPs with susceptibility to SZ. Therefore, additional studies with larger sample sizes should be conducted.
-
∙ The recruited SZ patients of the study may have different levels of heterogeneity, which may confuse the association result. So, it could be feasible to narrow down the phenotype to a more homogeneous subgroup, enabling an accurate distinction of underlying genetic subtypes.
-
∙ We only investigated one SNP in each gene, which failed to cover all the genetic variants of NKAPL, TSPAN18, and MPC2 gene. Consequently, additional studies should take the combined effects of more polymorphisms and the potential gene–gene interactions into consideration.
Introduction
Schizophrenia (SZ) is a severe neuropsychiatric disorder characterised by a series of psychotic symptoms, such as delusions, hallucinations, and cognitive impairments (Reference Owen, Craddock and O’Donovan1), and epidemiological data have shown that the lifetime prevalence of SZ is up to about 1% worldwide (Reference Saha, Chant, Welham and McGrath2). SZ, as a complex disease, is influenced by both genetic and environmental factors (Reference Jia, Wang, Meltzer and Zhao3), whereas heritable factors of SZ account for 80% (Reference Sullivan, Daly and O’Donovan4). Thus, genetic research of SZ is popular worldwide. However, the definite genetic mechanism remains unclear.
Numerous genome-wide association studies (GWASs) in SZ have been conducted, and many susceptibility loci have been identified. However, these GWAS-identified genetic associations still require further validation. In 2011, a GWAS (768 cases and 1733 controls) found that two SNPs, rs1635 and rs11038167, have a significant association with SZ in a Han Chinese population, and the follow-up independent Chinese sample of 4027 cases and 5603 controls also successfully replicated this result (Reference Yue, Wang and Sun5). Another GWAS of Han Chinese descent (3750 cases and 6468 controls) in 2011 (Reference Shi, Li and Xu6) reported that the association between MPC2 variant rs10489202 and SZ reaches genome-wide statistical significance, thereby validating the significant association with an additional sample of 4383 cases and 4539 controls. SNP rs1635 is located in the NKAPL gene, which encodes for nuclear factor-κB-activating protein-like, whereas NKAPL is located at 6p21-p22.1, which belongs to the extended major histocompatibility complex (MHC) region. Accumulating evidence from GWASs (Reference Purcell, Wray and Stone7–Reference Stefansson, Ophoff and Steinberg9), pathway analysis studies (Reference Jia, Wang, Fanous, Chen, Kendler and Zhao10), and expression studies (Reference Mexal, Frank and Berger11,Reference Harrison12) demonstrated the significant role of MHC in SZ. The rs11038167 polymorphism is located at the TSPAN18 gene and it encodes tetraspanin 18, a member of a superfamily of tetraspanins that are involved in signalling, antigen presentation, and diverse cellular processes (Reference Berditchevski and Odintsova13). TSPAN18 was proposed as a susceptibility locus containing the rs11038167, rs11038172, and rs835784 polymorphisms for SZ in the GWAS by Yue et al. (Reference Yue, Wang and Sun5) of Han Chinese, and a similar study (Reference Yuan, Jin and Qin14) also revealed that TSPAN18 rs835784 is significantly associated with SZ in a Chinese population. Another SNP rs10489202 of the MPC2 gene was found on chromosome 1q24.2, which encodes mitochondrial pyruvate carrier (MPC) 2 protein. Previous studies (Reference Halestrap, Scott and Thomas15–Reference Olsen, Hansen and Jakobsen17) have shown that the MPC involved in some metabolic pathways is closely related to psychiatric disorders. Thus, these studies suggested that the NKAPL, TSPAN18, and MPC2 genes may be associated with susceptibility to SZ.
Since the 2011 GWASs (Reference Yue, Wang and Sun5,Reference Shi, Li and Xu6), subsequent validation research was carried out in succession in different regions of China. In 2012, Ma et al. (Reference Ma, Tang and Wang18) genotyped nine GWAS-identified risk loci in an independent case–control study of Han Chinese from Hunan Province, and they failed to replicate the associations of the SNP rs1635, rs11038167, and rs10489202 with SZ. In 2013, a replication study by Yuan et al. (Reference Yuan, Jin and Qin14) indicated no significant association in both the allele and genotype frequencies of rs11038167 in TSPAN18 with SZ among a Han Chinese population in Jiangsu. Another replication study (Reference Jin, Zhang and Wang19) for MPC2 did not support the previous GWAS findings, and results showed that rs10489202 is not associated with SZ in Han Chinese. In 2014, Chen et al. (Reference Chen, Chao, Shen, Chen and Weng20) used a sample of Han Chinese subjects from Taiwan, and validated the association of rs1635 in the NKAPL gene with SZ, which supported the view that NKAPL is a susceptible gene for SZ. In 2015, a recent study by Zhang et al. (Reference Zhang, Li, Lu, Fan, Li and Feng21) reported that TSPAN18 variant rs11038167 is not significantly associated with SZ risk in northwestern Han Chinese subjects in Shanxi Province. Another study by Wang et al. (Reference Wang, Yang and Liu22) detected a positive association of the NKAPL rs1635 polymorphism with SZ in Han Chinese from Jiangsu Province, but their meta-analysis failed to validate a significant association. Notably, these replication studies used samples from different provinces in China, which implied that differences in the genetic background of individuals in different regions may lead to conflicting results.
Hence, this study aimed to examine whether the three GWAS-identified positive SNPs (rs1635, rs11038167, and rs10489202) can influence susceptibility to SZ in a Chinese population. China is a united multi-ethnic state, where the Han is a major ethnic group, and the Zhuang nationality is the largest minority. First, we conducted an independent case–control study including subjects from Zhuang and Han Chinese populations in Guangxi. In addition, a meta-analysis of the SNPs combining our case–control study with previous replication studies was performed to accurately assess the relationship between these SNPs and SZ in the Chinese population.
Methods
Case–control study
Participants
The case group consisted of 700 SZ patients from the two ethnic groups of Mainland China: 400 Han subjects and 192 Zhuang subjects from Guangxi Brain Hospital, and 108 Zhuang subjects from a public health project funded by the government in Liujiang County, Guangxi. The diagnosis was confirmed by two independent, trained psychiatrists according to ICD-10 (The International Classification of Disease, tenth revision). All patients are of one ethnic origin and born in Guangxi within three generations based on their own reports or reports by their first-degree relatives. Exclusion criteria included mental disorders caused by various organic diseases of the nervous system or other system disorders; substance-induced psychotic disorders; mood or neurodevelopmental disorders or mental retardation; and stroke, epilepsy, and other neurological diseases or other serious physical illnesses. Healthy controls were selected by well-trained investigators using a simple non-structured interview; these controls included 154 Zhuang healthy volunteers and 400 Han healthy individuals recruited from two comprehensive hospitals in the same region, as well as 146 healthy Zhuang volunteers in rural communities of Liujiang County. The ICD-10 was used to assess the subjects to exclude individuals with psychiatric conditions from the control group. This group was free of past or present major psychiatric or neurological disorders; in addition, the control subjects had no family history of mental illnesses in first-degree relatives.
All participants were unrelated Han or Zhuang Chinese born and living in Guangxi, and all of their biological grandparents were of Han or Zhuang Chinese ancestry. All the research subjects or their legal guardians signed informed consents.
Genotyping assays
Blood samples were collected using K2EDTA tubes (Weihai Hongyu Medical Instrument Co. Ltd, Shangdong, China) from all participants. In 7 days, genomic DNA was extracted from peripheral blood leucocytes using a commercial kit (Tiangen Biotech, Beijing, China). DNA samples were then stored at −80°C for genotype analysis. Primers were designed using Sequenom Assay Designer 3.1 software (Sequenom, San Diego, CA, USA), and the primer sequences are shown in Table 1. The polymorphisms in TSPAN18 (rs11038167), NKAPL (rs1635), and MPC2 (rs10489202) were genotyped using the Sequenom MassARRAY method. The design and synthesis of primers for allele identification were completed by Bomiao Technologies Co. (Beijing, China). Moreover, a 5% random sample was repeatedly tested, and the concordance rate was 100%.
Table 1 Primers of the rs1635, rs11038167, and rs10489202 polymorphisms
SNP=single nucleotide polymorphism.
Statistical analysis
The PLINK program (http://pngu.mgh.harvard.edu/~purcell/plink/) was applied to evaluate the genetic association between SNP genotypes and SZ susceptibility. χ2 goodness of fit test was used to assess the genotypic distributions among the control groups of each SNP for Hardy–Weinberg equilibrium (HWE). The correlation between each SNP and SZ was performed using unconditional logistic regression. After adjusting for age and gender, we obtained odds ratios (ORs) and 95% confidence intervals (95% CIs) to evaluate the strength of genotypes or alleles in different genetic models. Three genetic models were used, namely, Additive (AA vs. Aa vs. aa), Dominant (AA+Aa vs. aa), and Recessive (AA vs. Aa+aa), in which A is the mutant genotype, and a is the wild type. Power analysis was undertaken by Quanto software (http://hydra.usc.edu/gxe). SPSS version 16.0 was used to complete the statistical analysis of general characteristics of subjects. The independent t-test was employed to compare ages between SZ patients and normal controls of the two ethnic groups, and Pearson’s χ2-test was used to compare the categorical variable gender.
Meta-analysis
Searches were done in PubMed, Embase, Chinese National Knowledge Infrastructure database, Chinese Wanfang, and Chongqing VIP database, using a combined retrieval strategy that includes the following key terms: ‘schizophrenia’, ‘rs1635’, ‘rs11038167’, ‘rs10489202’, ‘NKAPL’, ‘TSPAN18’, ‘MPC2’, ‘association study’, ‘case-control study’ and ‘China’. In addition, some potential studies which were closely related to the research subject were identified by examining reference lists. The available articles cover all English and Chinese publications from their commencement to 16 September 2015. The identified studies were required to comply with the following inclusion criteria: (1) case–control studies that examined the association of NKAPL rs1635 or TSPAN18 rs11038167 or MPC2 rs10489202 polymorphisms with SZ susceptibility in a Chinese population; and (2) studies provided with data on ‘ORs and 95% CIs’, or ‘genotype and allele frequencies’. Studies were unsuitable if they complied with any one of characteristics below, namely, exclusion criteria: (1) the articles were not original research, for instance, reviews, commentaries, editorials, conference papers, and so on; (2) the duplicate publications and incomplete articles. Data extraction was in line with the inclusion criteria, including name of first author, publication year, country, ethnicity, number of cases and controls, ORs, 95% CIs, genotype, and allele frequencies. STATA software (version 11.1) was used for meta-analysis. The associations between SNPs and suspectibility to SZ in all samples were tested by pooled ORs and 95% CIs. Here, the pooled ORs were calculated based on the original data including ORs and 95% CIs from each included study. Heterogeneity across the included studies was assessed using the Q-test and I 2 statistics. If I 2 was <50% and p>0.10, no significant heterogeneity existed, and the fixed-effect model (Mantel–Haenszel method) was used to merge the data; otherwise, the random-effect model (The DeSimonian and Laird method) was used. All of the tests were two-tailed, with statistical significance of p<0.05.
Results
General characteristics of participants
The Zhuang ethnic population comprised 300 SZ cases (with 207 males and 93 females aged 33.68±11.99 years) and 300 controls (with 198 males and 102 females aged 32.37±12.27 years). The Han ethnic sample included 400 patients (67.2% males, 32.29±11.56 of mean age) and 400 controls (62.0% males, 33.09±11.17 of mean age). In the total sample group, 700 SZ patients (68% males) and 700 controls (63.7% males) were included, with an average of 32.89±11.76 and 32.78±11.65 years, respectively. No significant differences in age or gender distributions were observed between the case and control subjects in the Zhuang, Han, and total sample population (all p>0.05).
Power analysis
The power test was calculated under the following parameters: the prevalence of SZ was 1%; the ORs were set as 1.3; the sample size of the Zhuang, Han, and total group was 300, 400, and 700, respectively; the allele frequencies of rs1635, rs11038167, and rs10489202 polymorphisms were 0.340, 0.375, and 0.128 in Zhuang controls, 0.364, 0.345, and 0.144 in Han controls, and 0.354, 0.358, and 0.137 in the total controls, respectively. According to the above parameters, the statistical power of rs1635, rs11038167, and rs10489202 polymorphisms was 59.05%, 60.51%, and 35.56% among Zhuang subjects, 72.54%, 71.72%, and 48.57% among Han subjects, and 92.05%, 92.15%, and 70.38% among the total subjects, respectively.
Association between the NKAPL rs1635, TSPAN18 rs11038167, and MPC2 rs10489202 polymorphisms and SZ susceptibility.
The genotype distributions of the SNPs rs1635, rs11038167, and rs10489202 in the healthy controls did not show any significant deviations from HWE in each group (Table 2). No significant difference in the genotypic and allelic frequency distribution of the three SNPs was observed between SZ patients and controls (Table 2). As presented in Table 3, we found no significant association between the NKAPL rs1635, TSPAN18 rs11038167, and MPC2 rs10489202 polymorphisms and SZ susceptibility in different genetic models in the Zhuang, Han, and total sample groups, respectively.
Table 2 Genotype and allele distribution of single-nucleotide polymorphisms (SNPs) and Hardy–Weinberg equilibrium (HWE) test
A1, minor allele; A2, major allele; P HWE, HWE for control subjects.
Table 3 Association between single-nucleotide polymorphisms (SNPs) and SZ in Zhuang and Han groups
SZ, schizophrenia; OR, odds ratio; CI, confidence interval.
Meta-analysis
Nine studies (Reference Yue, Wang and Sun5, Reference Shi, Li and Xu6, Reference Yuan, Jin and Qin14, Reference Ma, Tang and Wang18–Reference Wang, Yang and Liu22) (including our study) were finally included in this meta-analysis. This meta-analysis comprised five data sets containing 8070 SZ cases and 10 237 controls for the NKAPL rs1635 polymorphism, five data sets (7685 cases and 10 295 controls) for the TSPAN18 rs11038167 polymorphism, and seven data sets (10 602 cases and 13 472 controls) for the MPC2 rs10489202 polymorphism (Table 4). Results of the present meta-analysis indicated that the SNPs rs1635 and rs11038167 were not significantly associated with SZ susceptibility (both p OR>0.05), whereas the rs10489202 polymorphism showed a significant association with SZ susceptibility in the Han Chinese population (p OR=0.002), and the corresponding pooled OR and 95% CI are shown in Table 5. A random-effect model was applied to calculate the association between rs1635, rs11038167, rs10489202 polymorphisms and SZ risk with significant heterogeneity (p Heterogeneity=0.000, I 2=93.80%; p Heterogeneity=0.000, I 2=89.00%; p Heterogeneity=0.048, I 2=52.80%, respectively), showed in Table 5 and supplementary Fig. 1a–c. No publication bias of the meta-analysis upon the rs11038167 and rs10489202 polymorphisms were observed in the funnel plot by Egger’s test (both p>0.05, supplementary Fig. 2b–c), whereas a publication bias of the meta-analysis on the SNP rs1635 was found in the funnel plot by Egger’s test (p=0.018, supplementary Fig. 2a).
Table 4 Genotype distribution of the studied single-nucleotide polymorphisms (SNPs) in the meta-analysis
Table 5 Meta-analysis of single-nucleotide polymorphisms (SNPs) and the susceptibility of schizophrenia
OR, odds ratio; CI, confidence interval.Bold value showed a significant correlation, which was used for emphasis.
Discussion
In our study, the results suggested that the GWAS-identified SNPs rs1635, rs11038167, and rs10489202 may not be associated with SZ susceptibility in Zhuang and Han individuals of China. But, our meta-analysis showed that the SNP rs10489202 was significantly associated with SZ in Han Chinese. This study was conducted for the first time to examine the association of the SNPs rs1635, rs11038167, and rs10489202 with the susceptibility of SZ in the Zhuang Chinese population, and further evaluate the association of the three SNPs with SZ susceptibility in the Han Chinese population.
In 2011, a GWAS (Reference Yue, Wang and Sun5) reported that the association between the NKAPL rs1635 polymorphism, TSPAN18 rs11038167 polymorphism, and SZ reached the level of genome-wide significance in a Han Chinese population. In the same year, another GWAS (Reference Shi, Li and Xu6) also identified the MPC2 rs10489202 polymorphism as potential susceptibility loci for SZ in Han Chinese. Several replicated studies were conducted in different areas in China, including Hunan (Reference Ma, Tang and Wang18), Jiangsu (Reference Yuan, Jin and Qin14,Reference Jin, Zhang and Wang19,Reference Wang, Yang and Liu22), Taiwan (Reference Chen, Chao, Shen, Chen and Weng20), and Shanxi province (Reference Zhang, Li, Lu, Fan, Li and Feng21). In addition, our case–control study showed no significant association between the SNPs rs1635, rs11038167, and rs10489202 and SZ risk in both Zhuang and Han groups in Guangxi, whereas our meta-analysis validate the association between rs10489202 and SZ in Han Chinese. Thus, these subsequent validation studies presented inconsistent association results compared with previous GWAS studies. One possible reason for this inconsistency is that the limited sample size may not provide sufficient power. Our sample size was smaller than the two GWASs, with a sample size (case/control) of 4773/7207 (Reference Yue, Wang and Sun5) and 8133/11 007 (Reference Shi, Li and Xu6), respectively. Another important reason may be the genetic heterogeneity of these variants to SZ risk in different regions of China. Notably, these studies used samples from different geographic areas of China; the subjects of the 2011 GWAS (Reference Yue, Wang and Sun5) were recruited from northern China (including Beijing, Tianjin, Hebei, and Shandong Provinces), whereas the study subjects of Zhang et al. (Reference Zhang, Li, Lu, Fan, Li and Feng21) were from northwestern China (Shanxi Province). Subjects from Ma et al. (Reference Ma, Tang and Wang18), Yuan et al. (Reference Yuan, Jin and Qin14), and Wang et al. (Reference Wang, Yang and Liu22), as well as our present study, were enroled from southern China (Hunan, Jiangsu, and Guangxi Provinces). Meanwhile, another GWAS (Reference Shi, Li and Xu6) collected samples from northern, central, and southern Han Chinese, whereas subjects of the replicated study (Reference Jin, Zhang and Wang19) were only from Jiangsu Province (eastern China), and our subjects in the present study were from Guangxi Province (southern China). Previous studies by Chen et al. (Reference Chen, Zheng and Bei23) and Xu et al. (Reference Xu, Yin and Li24) showed a great genetic difference among the Han Chinese population, which implied that the population structure and geographic variation might generate conflicting association results.
Although our case–control study has failed to find the susceptibility of SZ, further meta-analysis of this study suggested that the SNP rs10489202 was significantly associated with SZ risk in Han Chinese. Rs10489202 polymorphism is located in intron 1 of MPC2 gene on chromosome 1q24.2. MPC2 gene encodes the MPC-2 protein that is the main component of MPC proteins (Reference Vigueira, McCommis and Schweitzer25). The main role of MPC proteins is to transport pyruvate by acting as a facilitative carrier. Pyruvate is a critical metabolite linking cytoplasmic and mitochondrial metabolism, which participates in the process of oxidative metabolism, glycolysis, amino acid catabolism, and anabolism (Reference Gray, Rauckhorst and Taylor26). Currently, research showed that defects in mitochondrial function are related to some progressive neurodegenerative diseases such as Alzheimer’s disease, Parkinson’s disease, and it is considered to be linked with perturbations in MPC activity (Reference Shetty, Galeffi and Turner27–Reference Ahmed, Santosh, Kumar and Christlet29). Though the mechanism of MPC gene disruption in neurons is currently not clear, mitochondrial pyruvate transport is believed to maintain neuronal function of the brain (Reference McCommis and Finck30). Therefore, further studies are warranted to explore the role of MPC2 gene variant in the occurrence of SZ.
Several possible limitations of this study should be noted. First, the limited sample size may not have been powerful enough to identify the minimal influence of some SNPs with susceptibility to SZ. Second, the recruited SZ patients of the study may have different levels of heterogeneity, which may confuse the results. SZ is recognised as a highly heterogeneous disease, and it possesses diverse clinical presentations and may have different risk genotypes. Thus, narrowing down the phenotype to a more homogeneous subgroup is feasible; for example, by using neuropsychological test scores, electrophysiological assessments, neuroimaging, or defined symptom subtypes with some symptom scales, an accurate distinction of underlying genetic subtypes may be created. Third, we only investigated one SNP in each gene, which failed to cover all the genetic variants of the NKAPL, TSPAN18, and MPC2 genes. An association study by Zhang et al.(Reference Zhang, Lu and Yan31) reported that the SNPs rs12214383 and rs12000 on the NKAPL gene are significantly associated with SZ in a Han Chinese population. Consequently, additional studies on the association of the three genes with SZ should consider the combined effects of more polymorphisms and the potential gene–gene interactions.
In conclusion, our case–control study suggested that the NKAPL rs1635, TSPAN18 rs11038167, and MPC2 rs10489202 polymorphisms might not be associated with susceptibility to SZ in the Zhuang or Han populations, whereas our meta-analysis validate a significant association between MPC2 rs10489202 polymorphism and SZ susceptibility. Further studies with larger samples should include more ethnicities to further confirm the association of the NKAPL, TSPAN18, and MPC2 genes with the risk of SZ.
Acknowledgements
All the authors would like to express their sincere thanks to all the subjects who have taken part in the study. Authors’ contributions: all authors have contributed towards different phases of the manuscript preparation (the study design, acquisition of data, interpretation of data, table layout and drafting the article, revising it critically for important intellectual content), and also approved the final submitted version.
Financial Support
This work was supported by grants from the National Natural Science Foundation of China (No. 81460518), the Guangxi Natural Science Foundation (No. 2013GXNSFAA019352), the Science and Technology Program of Guangxi Universities (No. 2013YB043), and the Youth Science Foundation of Guangxi Medical University (GXMUYSF201322).
Conflicts of Interest
None.
Ethical Standards
The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national and institutional committees on human experimentation and with the Helsinki Declaration of 1975, as revised in 2008.
Supplementary Material
For supplementary material/s referred to in this article, please visit http://dx.doi.org/doi:10.1017/neu.2016.36
