Normal brain development is contingent on precise spatial and temporal control over several layers of regulation. Alternative splicing of pre-mRNA is enriched in brain and is a key component of this regulation, contributes to expansive protein diversity and its disruption has been associated with human disease including neuropsychiatric disorders [1, 2]. Recent genome-wide association studies have identified hundreds of risk loci for neuropsychiatric disorders [3]. Often located in noncoding areas, how these loci contribute to disease risk is not well understood. Recent evidence suggests that risk loci may contribute to disorder risk by disrupting alternative pre-mRNA splicing of risk genes, however most risk gene isoform profiles remain unknown [4].
Using Nanopore long-read sequencing we mapped the RNA isoform profiles of 31 neuropsychiatric risk genes. High-confidence risk gene mRNAs were sequenced on seven post-mortem human brain regions from five control individuals. The bioinformatics pipelines IsoLamp and IsoVis were used to identify and annotate known and novel isoforms. In total, we identified 747 novel isoforms and 21 novel exons. We also find highly expressed novel isoforms in the schizophrenia (SCZ) risk genes ATG13 and CSMD1 for which novel isoforms accounted for >50% of expression. Tissue specific expression of novel isoforms was identified including within the caudate for the SCZ risk gene DOC2A and cerebellum for the depression risk gene DCC. Mass spectrophotometry of brain protein isolates also confirmed translation of a novel exon skipping event in the SCZ risk gene ITIH4, suggesting a new regulatory mechanism for this gene.
Our results show current gene transcript annotations are incomplete and support the use of long-read sequencing to identify novel isoforms and exons in human brain. Uncovering the splicing repertoire for neuropsychiatric risk genes will help identify pathogenic isoforms, predict their function and potentially reveal new therapeutic pathways.