Profiling of adenosine to inosine editing in single neurons

Oregon Health & Science University, Portland OR

Investigators

Abstract

Adenosine to inosine (A-to-I) editing is a post-transcriptional modification which is prevalent in the developing nervous system. This chemical modification is catalyzed by adenosine deaminases (ADARs) acting on double- stranded RNA. Because inosine is structurally similar to guanine, A-to-I editing leads to inosine base pairing with cytosine and can change the encoded amino acid. A-to-I modification of many key neuronal mRNAs changes properties of their encoded proteins, including ion channels and mediators of synaptic transmission, signal transduction, and intracellular trafficking. Typically, editing increases as the brain matures during embryonic development, with less editing found in stem and neuronal progenitors and more editing in mature neurons. A- to-I editing is essential for normal brain development and abnormal editing has been associated with a number of diseases, including epilepsy, amyotrophic lateral sclerosis, psychiatric disorders, developmental disorders, and encephalopathy. The development of next generation sequencing approaches has enabled identification of A-to-I editing sites in different tissues. These approaches have identified specific neuronal target transcripts and the extent to which they are edited (editing efficiency varies widely and depends on transcript identity). Most mRNA target edit sites and their overall editing frequency are known. However, whether editing efficiency varies in different types of neurons within tissues and what fraction of a given transcript is edited in single neurons is not known. Recent developments in transcriptomic analysis of single cells allows assigning of individual neurons to various subtypes. However, common single cell RNA-sequencing (scRNA-seq) platforms, like 10x Chromium, use high-throughput short-read sequencing (e.g. Illumina Sequencing) to preferentially sequence the 3’UTR portion of transcripts to assign gene identity. As such, this approach does not allow comprehensive identification of editing events in coding regions of transcripts. On the other hand, approaches like Iso-Seq from Pacific Biosciences (PacBio), enable highly accurate, long-read sequencing (up to 15 kb), and do cover full transcripts in scRNA libraries. However, Iso-Seq read depth is too low (mostly due to sequencing costs) to sample enough transcripts per cell to assign detailed molecular identities. The goal of this proposal is to develop a methodology allowing for the simultaneous analysis of A-to-I editing sites in targeted RNAs (via long-read sequencing) in specific neuronal cell subtypes (characterized in detail via short-read sequencing). We will then benchmark our approach by performing a bulk RNA-seq in the same neuronal subtypes. Because multiple studies implicate A- to-I editing alterations in neurological and neuropsychiatric disorders, classification of A-to-I editing on a single cell level will enable a better understanding of disease pathogenesis and better targeting of potential treatments.