Cross‐disorder Association Studies Of Gene MYT1L And FGFR2with Three Common Psychiatric Disorders

Schizophrenia (SCZ), bipolar disorder (BPD), and major depressive disorder (MDD) are three major mental disorders. Both bipolar disorder and schizophrenia have a population prevalence of1‐2%, while major depressive disorder affects an even higher proportion (8‐12%in most countries). All the three disorders show high heritability:60‐85%for bipolar disorder and schizophrenia, and40%for major depressive disorder. Previous studies have revealed an overlap in clinical spectrum and genetic risk for the three illnesses. BPD and MDD are both major mood disorders, sharing the characteristic of periods of depression. Since different mental disorders may have similar pathogenesis, we conducted two large scale population‐based cross‐disorder association studies to investigate the possible association between MYT1L and FGFR2within three disorders in the Chinese Han population.Myelin transcription factor1‐like (MYT1L) is a member of the myelin transcription factor1(MYT1) gene family, and the neural specific, zinc‐finger‐containing, DNA‐binding protein that it encodes plays a role in the development of the nervous system. On the basis of a recent copy number variation (CNV) study showing that this gene is disrupted in mental disorder patients; we investigated whether MYT1L also plays a role in SCZ, BPD and MDD. Our sample sets consisted of1139unrelated bipolar disorder patients (621males and518females, with a mean age of36.6years, SD=13.8),1112unrelated schizophrenia patients (618males and494females, with a mean age of35.4years, SD=7.2),1119unrelated major depressive disorder patients (476males and643females, with a mean age of 35.1years, SD=11.6) and1135shared healthy controls (369males and766females, with a mean age of58.7years, SD=9.9). All subjects were of Chinese Han origin. We genotyped8SNPs on all the samples, and found that MDD patients and controls showed statistically significant differences for3SNPs in the allelic and genotypic distribution (rs3748989, p=0.0008for allele, p=0.0034for genotype; rs3748988, p=0.024for allele, p=0.0051for genotype; rs7592630, p=0.0369for genotype). Even after permutation on Haploview and correction by SNPSpD (the effective number of tests is6), rs3748989continued to be significantly associated with MDD (for allele, permutated p=0.0079, corrected p=0.0048; for genotype, corrected p=0.0204), and the genotypic frequency of rs3748988still remained significant (for genotype, corrected p=0.0306). Our results indicate that rs3748989is probably associated with MDD and that MYT1L may be a potential risk gene for MDD in the Chinese Han population.The fibroblast growth factor receptor2(FGFR2) gene is located on chromosome10q25‐q26, which is a hot linkage region for SCZ and BPD. O’Donovan et al. recently proposed the first susceptibility gene for SCZ in this region by a GWAS and a series of follow‐up studies. From3606single‐nucleotide polymorphisms (SNPs) mapping to this region genotyped in their GWAS, they identified a SNP, rs17101921, as the only one that survived all replication studies and showed nominal association with SCZ (OR=1.17,95%CI=1.06‐1.29; p=0.0009) in the meta‐analysis of9additional independent case–control series. Rs17101921maps85kb downstream from the nearest gene encoding fibroblast growth factor receptor2(FGFR2). Given the possible role of FGFR2in the pathogenesis of SCZ and the genetic etiology overlap of SCZ, BPD and MDD, we conducted a large scale population‐based study to investigate the possible association between FGFR2and susceptibility to the three disorders in the Chinese Han population. We genotyped8SNPs in the same sample sets. The pair wise LD between the8investigated SNPs was quite similar in different disorder sample sets. Only two adjacent SNPs (rs7090018and rs11199993) were in relatively high LD (D’>0.9) and were classified as in one haplotype block. No SNPs were associated with SCZ or MDD in allelic analysis after permutation. One SNP, rs11199993, showed significant association with BPD (OR=1.291,95%CI=1.115‐1.496; p=0.0006, p=0.007after permutation). The haplotype block was also found to be associated with BPD, after permutation. T‐C was found to be a risk haplotype (OR=1.298,95%CI=1.120~1.504; p=0.0005, p=0.006after permutation). Our findings support the association between FGFR2and BPD in the Chinese Han population.