Significance of risk polymorphisms for depression depends on stress exposure

Xenia Gonda,John Francis William Deakin,Tomas Gm Hökfelt,Gabriella Juhasz,Ian Muir Anderson,Nora Eszlari,Gyorgy Bagdy,Peter Antal,Gabor Hullam,Peter Petschner

doi:10.1038/s41598-018-22221-z

Abstract

Depression is a polygenic and multifactorial disorder where environmental effects exert a significant impact, yet most genetic studies do not consider the effect of stressors which may be one reason for the lack of replicable results in candidate gene studies, GWAS and between human studies and animal models. Relevance of functional polymorphisms in seven candidate genes previously implicated in animal and human studies on a depression-related phenotype given various recent stress exposure levels was assessed with Bayesian relevance analysis in 1682 subjects. This Bayesian analysis indicated a gene-environment interaction whose significance was also tested with a traditional multivariate analysis using general linear models. The investigated genetic factors were only relevant in the moderate and/or high stress exposure groups. Rank order of genes was GALR2 > BDNF > P2RX7 > HTR1A > SLC6A4 > CB1 > HTR2A, with strong relevance for the first four. Robust gene-gene-environment interaction was found between BDNF and HTR1A. Gene-environment interaction effect was confirmed, namely no main effect of genes, but a significant modulatory effect on environment-induced development of depression were found. Our data support the strong causative role of the environment modified by genetic factors, similar to animal models. Gene-environment interactions point to epigenetic factors associated with risk SNPs. Galanin-2 receptor, BDNF and X-type purin-7 receptor could be drug targets for new antidepressants.

Highlights

Animal models of depression usually imply environmental factors, such as chronic unpredictable stress or learned helplessness
Bayesian relevance analysis revealed that the investigated genetic factors were only relevant to the multiple depression-related phenotype in case of moderate or high recent negative life events (RLE) exposure (Fig. 1)
In case of the low RLE exposure group all genetic factors were non-relevant with low posterior probabilities

Summary

Introduction

Animal models of depression usually imply environmental factors, such as chronic unpredictable stress or learned helplessness. Candidate genes were usually selected from data of stressed animals[1,2], and evidence has accumulated that depressogenic effects of different stressors are mediated by different biological pathways[3] This approach sharply contrasts genome-wide association studies (GWAS), where several thousands of patients (cases) and controls are included without any knowledge of presence or type of previous stressors[4]. The applied Bayesian multivariate methods allow these 3 approaches to be combined; they blur the distinction between dependent and independent variables and instead evaluate the strength of all, or a predefined set of possible relationships across variables Using these methods, we have reported convergence of functionally related genes within the signaling pathway of the neuropeptide galanin on a multivariate phenotype, including quantitative measure of anxiety and depression, which operated only in those exposed to environmental stress. The use of the broader multivariate phenotype including symptoms of depression and anxiety after stressful events parallels animal models like chronic unpredictable stress

Objectives

Methods

Results

Conclusion