FAMILY-BASED RNA SEQUENCING IN BIPOLAR DISORDER FOR CANDIDATE GENE AND PATHWAY IDENTIFICATION

Abstract

ImportanceGenetic risk variants associated with bipolar disorder (BD) are likely to result in transcriptomic changes. RNA sequencing (RNAseq) studies in multiplex BD families are limited. ObjectiveTo identify candidate genes and pathways involved in BD. Design, settings, and participantsThis cohort study includes Australian families with three or more members diagnosed with BD. RNAseq was performed on total mRNA from lymphoblastoid cell lines of eight multiplex BD families. Patients were enrolled between May 1999 and October 2004. Data were analysed from June 2022 to March 2024. ExposuresDifferential gene expression changes and co-expressed gene networks in affected and unaffected relatives. Main outcomes and measuresQuantified read counts were examined for differential gene expression and Weighted Gene Co-expression Network Analysis (WGCNA). Differentially Expressed Genes (DEGs) were validated through: i) gene-based association using BD Genome-Wide Association Study (GWAS) summary statistics; ii) Polygenic Priority Score (PoPS); and iii) replication using RNAseq data from brain tissues of 71 BD and 252 controls. DEGs and co-expressed modules were examined for enriched categories via Gene-Set Enrichment Analysis (GSEA) and Over-Representation Analysis (ORA), and were further validated through gene-set analysis using GWAS data. ResultsSixty significant DEGs were found after comparing 16 BD (37.5% female) and 15 unaffected relatives (46.7% female), with the long non-coding RNA (lncRNA) LINC01237 being the most significant. Gene expression patterns of DEGs were correlated between lymphoblastoid cell lines and brain tissues (r=0.83). After validation, five DEGs were prioritised, including ENSG00000279277, a lncRNA mapping in a GWAS locus for BD. Enrichment analyses of DEGs pointed to nervous system and gated channel activities. WGCNA identified five expression modules associated with BD. The most significant module showed enriched categories associated with ion transmembrane transport and hypoplasia on corpus callosum, together with a protein-protein interaction network related to solute carriers at plasma membrane. Conclusions and relevanceThis study implicates a role for lncRNAs in the pathophysiology of BD. Alterations in ion homeostasis, driven by the dysregulation of gated channels, appear to be a plausible impaired process in BD. In addition, the corpus callosum emerges as a key brain structure with a potential involvement in BD. KEY POINTSO_ST_ABSQuestionC_ST_ABSCan transcriptomic signatures in families with multiple members affected with bipolar disorder (BD), likely due to a higher genetic load, identify genes and pathways contributing to disease risk? FindingsIn this cohort study of eight multiplex BD families, transcriptomic data implicated a pathogenic role of long non-coding RNAs, highlighting dysregulation of ion homeostasis as a key mechanism in the pathophysiologic aetiology of this condition. MeaningNon-coding RNA genes may play a more significant role in BD pathogenesis than previously estimated. The dysfunction of ion channels in BD indicates an endophenotype likely associated with channelopathy.