Igniting single-cell analysis of full-length RNA isoforms in brain development

Abstract

The expression of specific mRNA isoforms is central to many aspects of development, homeostasis, and disease. With long-read single-cell RNA-sequencing (LR scRNA-seq) we can now profile the expression of both known and novel RNA isoforms at the single cell level, elucidating how isoform expression varies both within and between cell types. However, accurate identification and quantification of isoforms in single cells remains a challenge, and the potential benefits of LR scRNA-seq compared to short-read (SR) scRNA-seq for many downstream analyses remains unclear. Here, we introduce an updated, faster, and more accurate version of our widely used FLAMES LR scRNA-seq analysis package and compared LR and SR scRNA-seq in a stem-cell model of cortical neurogenesis. We show long reads can provide more accurate cell-type identification, a key step in single-cell analysis, while short reads identified more false positive cell types such as mistaking radial glia for cancer cells. With LR scRNA-seq we achieved high-resolution identification of cell types, including outer radial glial progenitors and somatostatin interneuron subtypes. We identified >178,000 unique isoforms, including >10,000 that were previously unknown. We discovered thousands of differentially expressed genes and isoforms associated with synaptic transmission, neuronal projection, axonogenesis, and neuron development, confirming the validity of the expression quantification obtained from FLAMES. This included cell-type-specific isoform expression, such as in the PKM gene, which expressed different RNA isoforms in radial glia and neurons. Our work demonstrates the power of LR scRNA-seq for isoform-level quantification, particularly in unravelling the intricacies of brain development. Our refined methodology lays the groundwork for transformative insights in our understanding of complex developmental systems and diseases.

Bio

Dr. Mike Clark is a research fellow and group leader in the Centre for Stem Cell Systems at the University of Melbourne, Australia. His research sits at the intersection of genomics and neuroscience, developing and utilizing a number of genomic and transcriptomic approaches to investigate gene expression and function in the human brain and in neurological disorders. Mike is particularly interested in risk genes for mental health disorders and how their roles in brain development make some individuals more likely to develop mental health disorders.

Authors: Mike Clark