The roles of different gene expression regulators in acoustic variation in the intermediate horseshoe bat revealed by long-read and short-read RNA sequencing data

Abstract

Gene expression changes contribute greatly to phenotypic variations in nature. Studying patterns of regulators of gene expression is important to fully understand the molecular mechanism underlying phenotypic variations. In horseshoe bats, the cochleae are finely tuned to echoes of call frequency. Here, using 2 recently diverged subspecies of the intermediate horseshoe bat (Rhinolophus affinis hainanus and R. a. himalayanus) with great acoustic variations as the system, we aim to explore relative roles of different regulators of gene expression (differential gene expression, alternative splicing (AS) and long non-coding RNAs (lncRNAs)) in phenotypic variation with a combination of Illumina short-read and Nanopore long-read RNA-seq data from the cochlea. Compared to R. a. hainanus, R. a. himalayanus exhibited much more upregulated differentially expressed genes (DEGs) and multiple of them may play important roles in the maintenance and damage repair of auditory hair cells. We identified 411 differentially expressed lncRNAs and their target DEGs upregulated in R. a. himalayanus were also mainly involved in a protective mechanism for auditory hair cells. Using 3 different methods of AS analysis, we identified several candidate alternatively spliced genes (ASGs) that expressed different isoforms which may be associated with acoustic divergence of the 2 subspecies. We observed significantly less overlap than expected between DEGs and ASGs, supporting complementary roles of differential gene expression and AS in generating phenotypic variations. Overall, our study highlights the importance of a combination of short-read and long-read RNA-seq data in examining the regulation of gene expression changes responsible for phenotypic variations.

Sex Differences in Cochlear Transcriptomes in Horseshoe Bats

“…This mainly resulted from the different methods used between studies (rMATs in [40]; DEXSeq in this study). Another recent study that used both rMATs and DEXSeq when performing AS analysis in the same tissue revealed that the former method detected much more ASGs than the latter [73] (1932 with rMATs and 1267 with DEXSeq). Nevertheless, our current study and [40] in diverse bat species and tissues support an important role of AS in encoding sexual differences [36,39,74,75].…”