Uncovering Transcriptional Regulators and Targets of sRNAs Using an Integrative Data-Mining Approach: H-NS-Regulated RseX as a Case Study.

Mia K Mihailovic,Javier González Martínez,Éric Massé,Lydia M Contreras,Bridget Li,Angela Chen,Alyssa M Ekdahl,Peter L Freddolino,Abigail N Leistra,Matthew Law,Cindy Ejindu

doi:10.3389/fcimb.2021.696533

Abstract

Bacterial small RNAs (sRNAs) play a vital role in pathogenesis by enabling rapid, efficient networks of gene attenuation during infection. In recent decades, there has been a surge in the number of proposed and biochemically-confirmed sRNAs in both Gram-positive and Gram-negative pathogens. However, limited homology, network complexity, and condition specificity of sRNA has stunted complete characterization of the activity and regulation of these RNA regulators. To streamline the discovery of the expression of sRNAs, and their post-transcriptional activities, we propose an integrative in vivo data-mining approach that couples DNA protein occupancy, RNA-seq, and RNA accessibility data with motif identification and target prediction algorithms. We benchmark the approach against a subset of well-characterized E. coli sRNAs for which a degree of in vivo transcriptional regulation and post-transcriptional activity has been previously reported, finding support for known regulation in a large proportion of this sRNA set. We showcase the abilities of our method to expand understanding of sRNA RseX, a known envelope stress-linked sRNA for which a cellular role has been elusive due to a lack of native expression detection. Using the presented approach, we identify a small set of putative RseX regulators and targets for experimental investigation. These findings have allowed us to confirm native RseX expression under conditions that eliminate H-NS repression as well as uncover a post-transcriptional role of RseX in fimbrial regulation. Beyond RseX, we uncover 163 putative regulatory DNA-binding protein sites, corresponding to regulation of 62 sRNAs, that could lead to new understanding of sRNA transcription regulation. For 32 sRNAs, we also propose a subset of top targets filtered by engagement of regions that exhibit binding site accessibility behavior in vivo. We broadly anticipate that the proposed approach will be useful for sRNA-reliant network characterization in bacteria. Such investigations under pathogenesis-relevant environmental conditions will enable us to deduce complex rapid-regulation schemes that support infection.

Highlights

Bacterial small RNAs enable rapid post-transcriptional regulatory responses to external stressors that are often present within host environments (Villa et al, 2018), including envelope stress and carbon and metal ion limitation (Holmqvist and Wagner, 2017)
We have developed a computational approach (ID-sRnA) for identifying experimentally-supported regulators and targets of bacterial sRNAs by coupling multiple large and distinct omics datasets as well as bioinformatic prediction tools
The outputs of these two nodes, respectively, are (i) identities and putative binding positions of IPOD-HR- and Generation Sequencing (NGS)-supported DNA binding proteins (DBPs), namely, transcription factors (TFs), sigma factors, or nucleoidassociating proteins (NAPs), that may influence sRNA-specific expression, and (ii) computational sRNA target predictions informed by sRNA regional hybridization patterns in vivo, many of which are further supported by Gene Ontology analysis

Summary

Introduction

Bacterial small RNAs (sRNAs) enable rapid post-transcriptional regulatory responses to external stressors that are often present within host environments (Villa et al, 2018), including envelope stress and carbon and metal ion limitation (Holmqvist and Wagner, 2017). Most commonly, these 50-500 nucleotide transcripts (Villa et al, 2018) are induced under distinct environmental conditions and do not encode proteins, with a few exceptions (Gimpel and Brantl, 2017). The interest in understanding sRNA roles within larger stress-response networks has increased in recent years due to recognized links to pathogenicity (Chakravarty and Massé, 2019) and antibiotic resistance (Mediati et al, 2020)

Methods

Results

Conclusion