Structure-seq2 probing of RNA structure upon amino acid starvation reveals both known and novel RNA switches inBacillus subtilis

Abstract

RNA structure influences numerous processes in all organisms. In bacteria, these processes include transcription termination and attenuation, small RNA and protein binding, translation initiation, and mRNA stability, and can be regulated via metabolite availability and other stresses. Here we use Structure-seq2 to probe the in vivo RNA structurome of Bacillus subtilis grown in the presence and absence of amino acids. Our results reveal that amino acid starvation results in lower overall dimethyl sulfate (DMS) reactivity of the transcriptome, indicating enhanced protection owing to protein binding or RNA structure. Starvation-induced changes in DMS reactivity correlated inversely with transcript abundance changes. This correlation was particularly pronounced in genes associated with the stringent response and CodY regulons, which are involved in adaptation to nutritional stress, suggesting that RNA structure contributes to transcript abundance change in regulons involved in amino acid metabolism.Structure-seq2 accurately reported on four known amino acid-responsive riboswitches: T-box, SAM, glycine, and lysine riboswitches. Additionally, we discovered a transcription attenuation mechanism that reduces yfmG expression when amino acids are added to the growth medium. We also found that translation of a leader peptide (YfmH) encoded just upstream of yfmG regulates yfmG expression. Our results are consistent with a model in which a slow rate of yfmH translation caused by limitation of the amino acids encoded in YfmH prevents transcription termination in the yfmG leader region by favoring formation of an overlapping antiterminator structure. This novel RNA switch offers a way to simultaneously monitor the levels of multiple amino acids. Ritchey -3

Small RNA profiling inMycobacteriuminsights into stress adapt ability

“…Glycine riboswitch was only reported in Bacillus subtilis , regarded as a unique cis-acting mRNA element that contains two tandem homologous glycine-binding domains, and act on a single expression platform to regulate gene expression in response to glycine (Babina, et al 2017). In fact, riboswitching-binding metabolites have been successfully used to inhibit growth of non-pathogenic Bacillus subtilis in vitro (Ritchey, et al 2020). 6C is a conserved sRNA in high GC Gram-positive bacteria including Mycobacterium ; it has many cellular functions via binding to multiple mRNA targets through the C-rich loops in Mycobacterium , and regulates DNA replication and protein secretion (Mai, et al 2019).…”

“…For all the annotated ncRNAs, around 20% were rRNA and tRNA, which are essential for ribosome and translation (Ravishankar, et al 2016) Bacillus subtilis, regarded as a unique cis-acting mRNA element that contains two tandem homologous glycine-binding domains, and act on a single expression platform to regulate gene expression in response to glycine (Babina, et al 2017). In fact, riboswitching-binding metabolites have been successfully used to inhibit growth of non-pathogenic Bacillus subtilis in vitro (Ritchey, et al 2020).…”

Perspective: a stirring role for metabolism in cells

“…Structural changes of biomolecules such as RNA or proteins, potentially induced by altered molecular motion caused by altered metabolic activity, could also affect regulation. Interestingly, changes in the secondary structure of RNAs were found in Bacillus subtilis upon the modification of the metabolic state (Ritchey et al , 2020). As another example, the rpoH transcript of E. coli undergoes temperature‐induced conformational changes that determine whether it is translated or not (Yuzawa et al , 1993).…”

NusG is an intrinsic transcription termination factor that stimulates motility and coordinates gene expression with NusA

“…All RNA-Seq reads posttrimming were pseudo-mapped using Kallisto in SE mode using the --rf-stranded option to a transcriptome built with Illumina generated RNA-seq data collected from strain PLBS338 (Bray et al, 2016;Ritchey et al, 2020). This method determined both TPM and raw count values for each annotated transcript for both merged and non-merged FASTQ files (Supplementary file 6 ).…”