Prediction of bacterial small RNAs in the RsmA (CsrA) and ToxT pathways: a machine learning approach

Abstract

BackgroundSmall RNAs (sRNAs) constitute an important class of post-transcriptional regulators that control critical cellular processes in bacteria. Recent research using high-throughput transcriptomic approaches has led to a dramatic increase in the discovery of bacterial sRNAs. However, it is generally believed that the currently identified sRNAs constitute a limited subset of the bacterial sRNA repertoire. In several cases, sRNAs belonging to a specific class are already known and the challenge is to identify additional sRNAs belonging to the same class. In such cases, machine-learning approaches can be used to predict novel sRNAs in a given class.MethodsIn this work, we develop novel bioinformatics approaches that integrate sequence and structure-based features to train machine-learning models for the discovery of bacterial sRNAs. We show that features derived from recurrent structural motifs in the ensemble of low energy secondary structures can distinguish the RNA classes with high accuracy.ResultsWe apply this approach to predict new members in two broad classes of bacterial small RNAs: 1) sRNAs that bind to the RNA-binding protein RsmA/CsrA in diverse bacterial species and 2) sRNAs regulated by the master regulator of virulence, ToxT, in Vibrio cholerae.ConclusionThe involvement of sRNAs in bacterial adaptation to changing environments is an increasingly recurring theme in current research in microbiology. It is likely that future research, combining experimental and computational approaches, will discover many more examples of sRNAs as components of critical regulatory pathways in bacteria. We have developed a novel approach for prediction of small RNA regulators in important bacterial pathways. This approach can be applied to specific classes of sRNAs for which several members have been identified and the challenge is to identify additional sRNAs.

High-resolution small RNAs landscape provides insights into alkane adaptation in the marine alkane-degraderAlcanivorax dieseloleiB-5

“…Importantly, combined the InvenireSRNA prediction (Fakhry et al, 2017), Rfam alignment and secondary structure analysis, we identified two putative Csr A-related s R NAs (named CsrR1 and CsrR2) from the 549 sRNAs of A. dieselolei (Figures 4A, 7A to 7C, Tables S10 and S18), which may sequester the CsrA regulator and therefore indirectly involve in the alkane metabolism. CsrR1 originated from the 5’UTR of hfq (B5T_00774).…”

“…CsrA-related sRNAs were identified with the assistance of the R package “InvenireSRNA”, which integrated sequence- and structure-based features to train machine-learning models for detecting the bacterial sRNAs in the CsrA pathway (Fakhry et al, 2017). All the sRNA candidates and the extracted intergenic sequences of the whole genome were scanned using the InvenireSRNA to discover all the possible CsrA-related sRNAs in A. dieselolei .…”

Binding of Campylobacter jejuni FliW Adjacent to the CsrA RNA-Binding Pockets Modulates CsrA Regulatory Activity

“…CsrA from C. jejuni is longer (75 amino acids), and its activity is modulated by FliW ( Dugar et al, 2016 ; Radomska et al, 2016 ; Li et al, 2018 ). In contrast, E. coli CsrA is shorter (61 amino acids) ( Figure 1A ) and its activity is regulated by binding to sRNAs ( Jin et al, 2002 ; Thomson et al, 2006 ; Babitzke and Romeo, 2007 ; Fakhry et al, 2017 ; Janssen et al, 2018 ; Potts et al, 2019 ). As the C-terminus of C. jejuni CsrA is longer than the analogous region of CsrA proteins of E. coli and other bacteria, whose activities are regulated by sRNAs rather than by FliW ( Fields and Thompson, 2012 ), and adjacent to the most C-terminal β-strand (β 5 ) that is involved in RNA binding in both E. coli and C. jejuni ( Figure 1A ; Mercante et al, 2006 ; El Abbar et al, 2019 ), we hypothesized that FliW would bind to this region.…”

Three Ribosomal Operons of Escherichia coli Contain Genes Encoding Small RNAs That Interact With Hfq and CsrA in vitro

“…The machine-learning-based algorithm InvenireSRNA ( Fakhry et al, 2017 ) is designed to predict sRNAs of the CsrB/C family. To gauge how RrA, B, F rank in this algorithm compared to other E. coli non-coding RNAs, we provided the algorithm with a total of 2902 sequences, including all annotated E. coli sRNA sequences, the 22 TLR sequences, all sequences identified as CsrA-binding peaks from CLIP-seq data ( Potts et al, 2017 ), and finally all intergenic regions of E. coli shorter than 1,000 bp.…”