Quantitative DNA Methylation Analysis of DLGAP2 Gene using Pyrosequencing in Schizophrenia with Tardive Dyskinesia: A Linear Mixed Model Approach

Yanli Li,Ping Zhang,Shuping Tan,Kesheng Wang,Huimei An,Fu De Yang,Song Chen,Junchao Huang,Yunlong Tan,Nianyang Wang,Zhiren Wang

doi:10.1038/s41598-018-35718-4

Abstract

Tardive dyskinesia (TD) is a side effect of antipsychotic medications used to treat schizophrenia (SCZ) and other mental health disorders. No study has previously used pyrosequencing to quantify DNA methylation levels of the DLGAP2 gene; while the quantitative methylation levels among CpG sites within a gene may be correlated. To deal with the correlated measures among three CpG sites within the DLGAP2 gene, this study analyzed DNA methylation levels of the DLGAP2 gene using a linear mixed model (LMM) in a Chinese sample consisting of 35 SCZ patients with TD, 35 SCZ without TD (NTD) and 34 healthy controls (HCs) collected in Beijing, China. The initial analysis using the non-parametric Kruskal-Wallis test revealed that three groups (TD, NTD and HC) had significant differences in DNA methylation level for CpG site 2 (p = 0.0119). Furthermore, the average methylation levels among the three CpG sites showed strong correlations (all p values < 0.0001). In addition, using the LMM, three groups had significant differences in methylation level (p = 0.0027); while TD, NTD and TD + NTD groups showed higher average methylation levels than the HC group (p = 0.0024, 0.0151, and 0.0007, respectively). In conclusion, the LMM can accommodate a covariance structure. The findings of this study provide first evidence of DNA methylation levels in DLGAP2 associated with SCZ with TD in Chinese population. However, TD just showed borderline significant differences to NTD in this study.

Highlights

Epigenetic changes affect gene expression and function by mechanisms other than those from changes in the DNA sequence; whereas DNA methylation is an important epigenetic modification and involves the addition of a methyl group at the 5th carbon of cytosines preceding guanines (CpG dinucleotides)
There were no statistical significances in age (p = 0.993) and sex composition (p = 0.983) among the three groups; whereas there were significant differences in the DNA methylation levels in site 2 and the average of three CpG sites among three groups using generalized linear model (GLM) (p = 0.0307 and 0.0465, respectively)
The Kruskal-Wallis test revealed that the three groups had significant differences in DNA methylation level for CpG site 2 (p = 0.0119) and borderline differences for the average DNA methylation level of 3 CpG sites (p = 0.0945) (Table 2 and Fig. 1)

Summary

Introduction

Epigenetic changes affect gene expression and function by mechanisms other than those from changes in the DNA sequence; whereas DNA methylation is an important epigenetic modification and involves the addition of a methyl group at the 5th carbon of cytosines preceding guanines (CpG dinucleotides). Pyrosequencing offers a robust, versatile platform yielding rapid quantitative analysis of DNA methylation levels and providing information on the methylation status of single CpG sites[18,19,20]. No study has used pyrosequencing to quantify DNA methylation levels of the DLGAP2 gene; in addition, no study has examined the methylation of DLGAP2 gene in TD. The DNA methylation levels using pyrosequencing among CpG sites within a gene may be correlated[28]. This study sought to quantify DNA methylation levels of DLGAP2 gene in SCZ with or without TD using pyrosequencing and to deal with the possible correlations among 3 CpG sites within the DLGAP2 gene using a linear mixed model (LMM) in a Chinese population

Methods

Results

Conclusion