Transcriptome-wide alternative splicing and transcript-level differential expression analysis of post-mortem Lewy body dementia brains.

Abstract

Lewy body dementias (LBD) are the second most common dementia. Several genes have been associated with LBD, but little is known about their contributions to LBD pathophysiology. Each gene may transcribe multiple RNA, and LBD brains have extensive RNA splicing dysregulation. Hence, we completed the first transcriptome-wide transcript-level differential expression analysis of post-mortem LBD brains for gaining more insights into LBD molecular pathology that are essential for facilitating discovery of novel therapeutic targets and biomarkers for LBD. We completed transcript-level quantification of next-generation RNA-sequencing data from post-mortem anterior cingulate (ACC) and dorsolateral prefrontal cortices (DLPFC) of people with pathology-verified LBD (LBD = 14; Controls = 7) using Salmon. We identified differentially expressed transcripts (DET) using edgeR and investigated their functional implications using DAVID. We performed transcriptome-wide alternative splicing analysis using DRIMseq. We identified 74 DET in ACC and 96 DET in DLPFC after Benjamini-Hochberg false discovery rate (FDR) correction (5%). There were 135 and 98 FDR-corrected alternatively spliced genes in ACC and DLPFC of LBD brains, respectively. Identified DET may contribute to LBD pathology by altering DNA repair, apoptosis, neuroplasticity, protein phosphorylation, and regulation of RNA transcription. We confirm widespread alternative splicing and absence of chronic neuroinflammation in LBD brains. Transcript-level differential expression analysis can reveal specific DET that cannot be detected by gene-level expression analyses. Therapeutic and diagnostic biomarker potential of identified DET, especially those from TMEM18, MICB, MPO, and GABRB3, warrant further investigation. Future LBD blood-based biomarker studies should prioritise measuring the identified DET in small extracellular vesicles.