A stochastic RNA editing process targets a limited number of sites in individual Drosophila glutamatergic motoneurons

Abstract

RNA editing is a post-transcriptional source of protein diversity and occurs across the animal kingdom. Given the complete profile of mRNA targets and their editing rate in individual cells is unclear, we analyzed single cell RNA transcriptomes from Drosophila larval glutamatergic motoneuron subtypes to determine the most highly edited targets and identify single neuron editing rules. From [~]15,000 genes encoded in the genome, 316 high confidence A-to-I canonical RNA edit sites were identified, with 60 causing missense amino acid changes predicted to alter proteins regulating membrane excitability, synaptic transmission, or neuronal function. Twenty-seven canonical sites were edited at >90% frequency as observed for editing of mammalian AMPA receptors, including coding edits in the Rdl GABA receptor, the nAChRalpha5 and nAChRalpha6 acetylcholine receptors, the Shab K+ channel, the NCKX30C Na+/K+-dependent Ca2+ exchanger and the postsynaptic scaffold Shank. However, most sites were edited at lower levels and generated variable expression of edited and unedited mRNAs, suggesting stochastic editing that may provide a mechanism to fine-tune synaptic function similar to alternative splicing. Among these variably edited targets were proteins with well-known presynaptic functions, including the voltage-gated Ca2+ channel Cacophony, the synaptic vesicle fusion regulator Complexin, the active zone scaffolding proteins RBP and Rim, and the endocytosis regulators Lap/AP180 and Endophilin. Comparison of these editing targets across other publicly available Drosophila RNAseq datasets identified several sites present exclusively in larval motoneurons, indicating the presence of cell-type and/or developmental-specific editing. Further comparisons confirmed the co-transcriptional nature of canonical editing and revealed editing is largely resistant to changes in neuronal activity, with only a few sites displaying evidence of being activity-regulated. Noncanonical editing was also found to occur in these neurons, including a C-to-U edit that altered an amino acid in the capsid hinge domain of the synaptic plasticity regulator Arc1. Together, these data provide insights into how the RNA editing landscape may alter protein function to modulate the properties of two well-characterized neuronal populations in Drosophila.