Mapping cell type specific isoform diversity in the human brain: dissecting mechanisms of alternative splicing in ASD

PROJECT SUMMARY/ABSTRACT Large-scale RNA sequencing studies have provided remarkable insight into the brain molecular processes that underlie complex neurodevelopmental disorders like autism spectrum disorder (ASD). Bulk processing of brain tissue has linked the disruption of alternative splicing, a mechanism by which a gene’s exons and introns are differentially processed into mRNA isoforms, to the transcriptomic changes seen in ASD. Furthermore, single- cell and single-nucleus analyses have localized these alterations to pro-inflammatory microglia, astrocytes, and deep excitatory neuron populations. However, prior splicing analyses were limited to well-annotated isoforms from short-read data, which account for only a small subset of all isoforms present in the brain transcriptome. Now, emerging third generation “long-read” sequencing technologies allow for the processing of full-length transcripts, revealing the full profile of introns and exons for both known and novel gene isoforms. Here, we propose single nucleus long-read sequencing of over 60 human prefrontal cortex samples, including 33 individuals with a diagnosis of ASD and 30 without such a diagnosis, to fully interrogate the biological function of isoforms in neuropsychiatric disease. Given the existing evidence of splicing alterations in ASD, and the growing body of evidence that isoform expression captures the brain’s transcriptomic diversity better than gene expression, we hypothesize that long-read sequencing will reveal broader transcriptional dysregulation than previously captured, and that we will localize these changes to highly refined cell subpopulations. Leveraging long-read technology on our dataset, we will assess the cellular distribution of isoforms in the postnatal human brain (Aim 1). Next, we will perform case-control differential expression analysis, paired with genomic enrichment, to identify isoform-level drivers of ASD pathophysiology (Aim 2). Finally, we will examine correlations between isoforms and identify systems-level regulators of expression, across both control samples and ASD samples (Aim 3). Altogether, these aims serve to systematically characterize the role of isoform diversity in the postnatal human brain across both control and ASD populations, aligning with the NIMH’s mission to uncover the neurobiological basis of brain-related disorders. This proposal will be carried out by Michael Margolis, an MD/PhD student at UCLA, who will receive comprehensive training in human genetics and genomics. Mentorship will be provided by Dr. Daniel H. Geschwind as the primary mentor, with additional mentorship from Dr. Michael J. Gandal, both of whom are experts in the fields of functional genomics and neurobehavioral genetics.