Using stringent criteria, we followed-up a genome-wide association study (GWAS) finding1 to examine the involvement of the developmental regulatory gene NOTCH4 in schizophrenia. This is the first study in a Japanese population that satisfies genome-wide significance between a particular single-nucleotide polymorphism (SNP) and schizophrenia.
The past 5 years have seen significant genetic discoveries made through GWASs. Although in schizophrenia the effect sizes have been relatively small (odds ratio (OR) of <1.2), these studies collectively provide strong evidence for association of a number of susceptibility genes. The major loci identified for schizophrenia risk are located in the major histocompatibility complex region in 6p21.3–p22.1. The Psychiatric GWAS Consortium, a sample consisting mainly of Caucasian subjects, reported 136 SNPs with genome-wide significance (5 × 10−8), and among them 129 SNPs were located from 25 to 33 Mb at chromosome 6.2 In addition, a recent Asian GWAS from a Chinese population showed a stringent level of significance at ZKSCAN4 (chr6: 28.32 Mb),3 even though another larger Chinese population GWAS4 and our prior Japanese population-based GWAS (total sample size: 1108)1 did not detect any genome-wide significance in this region. In our previous study,1 however, only one signal within these loci maximized at a SNP (rs2071287: chr6: 32 170 433 bp on hg19) in NOTCH4 (chr6: 32.16 to 32.19 Mb). To provide a more complete analysis, in the current study, we expanded the sample size to verify whether this SNP (rs2071287: C>T) in NOTCH4 shows strong support for genetic association.
Six case–control data sets from Japanese population, including our screening GWAS (JPN_GWAS), replication set (Rep_JPN) and the sample of Tochigi et al.,1 were evaluated. The other three new data sets consist of REP1, REP2 and REP3, the sample sets being divided by the collection regions in Japan: REP1 (Mid-east Japan: case=3173, control=3540), REP2 (Mid-west Japan: case=672, control=5321), and REP3 (South island of Japan (Shikoku): case=569, control=1622) (Supplementary Method). The TaqMan assay (Applied Biosystems, Foster City, CA, USA) was used to determine the genotypic distribution of rs2071287. A meta-analysis was performed with a fixed model by ‘rmeta’, an R package, as we did not apply a principal component analysis, which is the most common approach to adjust the population structure. Instead, we conducted a stratified analysis according to the sample collection site in order to rule out possible population stratification. A test for heterogeneity did not show significant deviation for combined replication data sets (REP1–REP3, χ2=0.63, df=2, P=0.73) and all data sets, including samples from our prior study (χ2=4.2, df=5, P=0.52).
Previously, we reported an association based upon three data sets with Pmeta at 5.1 × 10−5 and OR of 0.83 (reference: T allele: minor allele, 95% confidence intervals: CIs: 0.77–0.91: ‘Summary1’ in Figure 1).1 In the current study, in two out of three replication sets, we detected statistical evidence (P<0.05, two-tailed) for a consistent association that was similar in magnitude and direction (T allele in control is more frequent than in case: Figure 1). Further meta-analysis of the combined samples for the new data sets (REP1–REP3: case=4414, control=10 483) suggests strong support for the association of this SNP (P=7.9 × 10−5: OR=0.89, 95% CI=0.84–0.94, ‘Summary2’ in Figure 1). Finally, genome-wide significance was detected in the meta-analysis combining all of the data sets (6668 case and 12 791 controls) in the Japanese population (P=3.4 × 10−8: OR=0.87, 95% CI=0.83–0.92, ‘ALL combined’ in Figure 1).
Our results provide stringent support for previous genetic findings reporting the association of an intronic SNP (rs3131296, chr6: 32 172 993 bp on hg19) in NOTCH4 in a European cohort.5 However, the pattern of linkage disequilibrium (LD) between the current (rs2071287) and previous (rs3131296) SNPs, in addition to the minor allele frequencies (MAFs) of these SNPs vary across populations (Supplementary Figures 1 and 2): (1) a greater level of LD is observed in the Caucasian (HapMap CEU, release28, D′=1, r2=0.13) compared with the Japanese population (HapMap JPT, release28, D′=0.49, r2=0); and (2) rs3131296, the SGENE-plus study5 reported significant association, shows substantially higher MAF in Caucasian (8.5% in HapMap CEU, 13% in SGENE-plus) than in Japanese (1.9% in HapMap JPT). Considering that the major histocompatibility complex region consists of large blocks with a very high LD,6 it would be difficult to ascribe the difference to particular SNP(s). Nevertheless, given the evidence for the association of NOTCH4 with schizophrenia based upon current and previous genetic and biological support,7 we suggest that NOTCH4 remains one of the strongest candidate susceptibility genes for schizophrenia, and we propose this SNP is a key to the identification of causal variant(s).
Ikeda M, Aleksic B, Kinoshita Y, Okochi T, Kawashima K, Kushima I et al. Biol Psychiatry 2011; 69: 472–478.
Ripke S, Sanders AR, Kendler KS, Levinson DF, Sklar P, Holmans PA et al. Nat Genet 2011; 43: 969–976.
Yue WH, Wang HF, Sun LD, Tang FL, Liu ZH, Zhang HX et al. Nat Genet 2011; 43: 1228–1231.
Shi Y, Li Z, Xu Q, Wang T, Li T, Shen J et al. Nat Genet 2011; 43: 1224–1227.
Stefansson H, Ophoff RA, Steinberg S, Andreassen OA, Cichon S, Rujescu D et al. Nature 2009; 460: 744–747.
Jeffreys AJ, Kauppi LNeumann R . Nat Genet 2001; 29: 217–222.
Imayoshi I, Kageyama R . Mol Neurobiol 2011; 44: 7–12.
This work was supported by research grants from the Ministry of Education, Culture, Sports, Science and Technology (MEXT) of Japan; Ministry of Health, Labor and Welfare of Japan; Academic Frontier Project for Private Universities, Comparative Cognitive Science Institutes; Core Research for Evolutional Science and Technology; Uehara Memorial Foundation; SENSHIN Medical Research Foundation; Takeda Science Foundation; Novaltis Foundation, Japan; Strategic Research Program for Brain Sciences of the MEXT of Japan.
The authors declare no conflict of interest.
Supplementary Information accompanies the paper on the Molecular Psychiatry website
About this article
Cite this article
Ikeda, M., Aleksic, B., Yamada, K. et al. Genetic evidence for association between NOTCH4 and schizophrenia supported by a GWAS follow-up study in a Japanese population. Mol Psychiatry 18, 636–638 (2013). https://doi.org/10.1038/mp.2012.74
Journal of Medical Genetics (2020)
A common variant of the NOTCH4 gene modulates functional connectivity of the occipital cortex and its relationship with schizotypal traits
BMC Psychiatry (2020)
Schizophrenia Research (2018)
Schizophrenia Research (2018)