The genomic basis of mood instability: identification of 46 loci in 363,705 UK Biobank participants, genetic correlation with psychiatric disorders, and association with gene expression and function

Abstract

Genome-wide association studies (GWAS) of psychiatric phenotypes have tended to focus on categorical diagnoses, but to understand the biology of mental illness it may be more useful to study traits which cut across traditional boundaries. Here, we report the results of a GWAS of mood instability as a trait in a large population cohort (UK Biobank, n = 363,705). We also assess the clinical and biological relevance of the findings, including whether genetic associations show enrichment for nervous system pathways. Forty six unique loci associated with mood instability were identified with a SNP heritability estimate of 9%. Linkage Disequilibrium Score Regression (LDSR) analyses identified genetic correlations with Major Depressive Disorder (MDD), Bipolar Disorder (BD), Schizophrenia, anxiety, and Post Traumatic Stress Disorder (PTSD). Gene-level and gene set analyses identified 244 significant genes and 6 enriched gene sets. Tissue expression analysis of the SNP-level data found enrichment in multiple brain regions, and eQTL analyses highlighted an inversion on chromosome 17 plus two brain-specific eQTLs. In addition, we used a Phenotype Linkage Network (PLN) analysis and community analysis to assess for enrichment of nervous system gene sets using mouse orthologue databases. The PLN analysis found enrichment in nervous system PLNs for a community containing serotonin and melatonin receptors. In summary, this work has identified novel loci, tissues and gene sets contributing to mood instability. These findings may be relevant for the identification of novel trans-diagnostic drug targets and could help to inform future stratified medicine innovations in mental health.

Figure 1. Manhattan and QQ plot of mood instability GWAS.

Figure 2. Different Mood associated genetic risk variants converge in a nervous specific gene network.

(A) Enrichments of gene functional communities from a generic PLN and from a Nervous-System (NS) PLN within Mood-GWA and subGWA loci (see Methods). The Community ID is given first in the descriptor followed by the number of genes within that community. Only communities formed from over 20 genes are shown. (B) Gene subnetwork of Community 26 from NS-PLN showing functional associations between genes residing in Mood-associated GWA (red squares) and subGWA (orange squares) intervals and genes whose unique mouse orthologues are annotated with abnormal synaptic transmission phenotype (cyan squares). To increase clarity, only genes with abnormal synaptic transmission phenotype annotation with at least three functional links to genes residing in GWA and subGWA regions are shown. The colour of the link connecting two genes indicates the strongest information source supporting the functional association.