Contribution of common and rare damaging variants in familial forms of bipolar disorder and phenotypic outcome

Elisa Courtois,Jean-Pierre Kahn,Philippe Le Corvoisier,Mircea Polosan,Christine Passerieux,Emilie Olie,Bruno Etain, ,StéPhane Jamain,Mark Schmid,Marion Leboyer,Frank Bellivier,Bruno Aouizerate,Orly Wajsbrot,Caroline Dubertret,Caroline Barau,Raoul Belzeaux

doi:10.1038/s41398-020-0783-0

Abstract

Genome-wide association studies on bipolar disorders (BD) have revealed an additive polygenic contribution of common single-nucleotide polymorphisms (SNPs). However, these SNPs explain only 25% of the overall genetic variance and suggest a role of rare variants in BD vulnerability. Here, we combined high-throughput genotyping data and whole-exome sequencing in cohorts of individuals with BD as well as in multiplex families with a high density of affected individuals in order to determine the contribution of both common and rare variants to BD genetic vulnerability. Using polygenic risk scores (PRS), we showed a strong contribution of common polymorphisms previously associated with BD and schizophrenia (SZ) and noticed that those specifically associated with SZ contributed more in familial forms of BD than in non-familial ones. The analysis of rare damaging variants shared by affected individuals in multiplex families with BD revealed a single interaction network enriched in neuronal and developmental biological pathways, as well as in the regulation of gene expression. We identified four genes with a higher mutation rate in individuals with BD than in the general population and showed that mutations in two of them were associated with specific clinical manifestations. In addition, we showed a significant negative correlation between PRS and the number of rare damaging variants specifically in unaffected individuals of multiplex families. Altogether, our results suggest that common and rare genetic variants both contribute to the familial aggregation of BD and this genetic architecture may explain the heterogeneity of clinical manifestations in multiplex families.

Highlights

Bipolar disorders (BD) are chronic, heterogeneous, and complex mental disorders with a worldwide prevalence of ~1%1
In order to determine whether the polygenic contribution was similar in familial and non-familial forms of BD, we used imputed genotyping data of 1,709,567 single-nucleotide polymorphisms (SNPs) for 445 patients with BD, 1636 controls, and 37 individuals in multiplex families to calculate two polygenic risk scores (PRS) based on Psychiatric Genomics Consortium (PGC) BD4 and SZ29 summary statistics
There was no difference in based PRS (BD-PRS) between individuals with and without a family history of BD in our cohort (Wilcoxon rank-sum test, W = 16,212; p = 0.60; padj = 1.00; Fig. 1a), a higher SZPRS was observed for familial forms when compared to non-familial ones (Wilcoxon rank-sum test, W = 13,225; p = 0.01; padj = 0.12; Fig. 1b)

Summary

Introduction

Bipolar disorders (BD) are chronic, heterogeneous, and complex mental disorders with a worldwide prevalence of ~1%1. Their etiopathology is poorly understood but genetic contribution arises has an important factor, with an estimated heritability ranging between 60% and 80%2,3. Associated variants had a low effect size, but when combined in an additive model, they were able to significantly distinguish individuals with BD from controls[5]. These overall results explain only 25% of the genetic variance[6], revealing a “missing heritability” and increasing the potential interest for rare penetrant variants. E.g. autism spectrum disorder (ASD) and schizophrenia (SZ), whole-exome sequencing (WES) has already shown a higher frequency of de novo mutations in affected

Methods

Results

Conclusion