Periodic Reporting for period 3 - RegRNA (Mechanistic principles of regulation by small RNAs)
Période du rapport: 2022-09-01 au 2024-02-29
Bacteria adapt to environmental changes by remodeling their gene expression programs. These changes in gene expression are tightly controlled by various regulators, among them small RNAs (sRNAs). sRNAs are 50-400 nucleotide-long RNA molecules that regulate gene expression post-transcriptionally by base pairing with the mRNA of their targets, while both the sRNA and mRNA are bound to a chaperon protein, Hfq. For many years sRNAs were considered as regulators of translation, but accumulating evidence suggest that they affect also the transcript levels, directly or indirectly, by affecting various molecular processes involving the mRNA transcript, such as pre-mature transcription termination or the cleavage of the transcript by endoribonucleases. In this project we aim to decipher the modes of regulation that the various sRNAs exerts on their different targets. Understanding the basic mechanistic principles of regulation by sRNAs and their interplay with other regulators in the cell is of utmost importance for understanding the integrated response of the bacteria to various stresses, among them antibiotic treatment. In a connected venue, we apply our methodologies to pathogenic bacteria, attempting to identify the involvement of sRNAs in pathogenicity pathways. Our discoveries are expected to provide new leads for the development of novel anti-bacterial or immunomodulatory agents. Both drug classes are urgently needed given the emerging health crisis caused by the increasing prevalence of antibiotic-resistant bacterial pathogens.
From the beginning of the project we accomplished the following:
1) Our results suggest that competition between targets over binding to Hfq plays a major role in the regulatory outcomes that a sRNA exerts on different targets (Faigenbaum-Romm et al, 2020).
2) We developed a machine learning-based algorithm that predicts novel sRNAs from Hfq-mediated RIL-seq data. Applying this new algorithm to our original RIL-seq data of Escherichia coli K-12 (Melamed et al, 2016), we identified and verified experimentally novel Hfq-dependent sRNAs encoded in various genomic regions (Bar et al., 2021).
3) We completed the study of inference of the relative contribution of a sRNA to the response of cells to an environmental change (Barsheshet et al., 2022). Applying a special design of differential gene expression analysis by RNA-seq, we extract the contribution of a sRNA to the total change in the transcript level of each gene in response to an environmental change (applied to RyhB and GadF in Escherichia coli grown under iron limitation and acidic stress, respectively, and to MgrR in enteropathogenic E. coli (EPEC) grown on DMEM).
4) We published the sRNA-target network of the pathogenic bacterium EPEC, studied in collaboration with Prof. Ilan Rosenshine from our department (Pearl Mizrahi et al., 2021). We discovered novel sRNAs encoded in non-core genomic regions and found that sRNAs mediate substantial information flow between the core genome and genomic islands. We discovered a few sRNAs that act as “regulatory hubs” of virulence genes, one of which, MgrR, we studied in detail.
5) We completed deciphering the networks of additional two pathogenic bacteria, Salmonella enertica and Klebsiella pneumonia, in collaboration with Joerg Vogel (HIRI, Germany) and Yunn-Hwen Gan (NUS, Singapore), respectively. In the Salmonella study (Matera et al, 2022), we discovered a new sponging mechanism by a sRNA (OmpX) of another sRNA (MicF). In Klebsiella (Goh et al., 2024, under review), we found that the capsule, a major virulence component, is a hub target of sRNAs. Having the three sRNA-target networks of pathogenic bacteria and the network of E. coli K-12, we identified sRNA-target interactions that are conserved in all four bacteria, implying the importance of this regulation.
6) We applied ribo-seq to E. coli strains grown in normal conditions and under iron limitation, where the sRNA RyhB is induced. We analyze these data to study the effect of sRNAs on translation initiation and elongation. For the latter, we have been developing algorithms to identify ribosome pausing sites that are dependent on sRNA binding, suggesting possible regulation of translation elongation by sRNAs.
7) We completed the development of an original and accurate algorithm to determine transcript termini using only gene expression data (Bar et al., 2023). To this end, we exploit a specific read pattern that we identified at 3’ termini of transcripts in gene expression data generated by the RNAtag-seq protocol, a protocol that is widely used in microbiology research. We now apply our new algorithm to the wealth of RNAtag-seq data available for 32 bacteria grown under 12 conditions, identifying condition-dependent pre-mature transcaription termination sites that are conserved in evolution.
8) We completed the study of the interplay between a sRNA and an endoribonuclease, using GcvB and RNAse E as our model system (Barsheshet et al., 2024, in revision). To infer quantitatively the functional interaction between GcvB and RNase E, we appplied the approach of the double mutant cycle. This approach involves a special design of single and double mutants to which RNA-seq is applied. We can extract the separate contributions and mutual contributions of the endoribonuclease and sRNA to changes in gene expression by a special analysis of the data. We carried out these experiments for GcvB and RNase E, using RNA from E. coli wild type, ΔgcvB, rne ts conditional mutant and the double mutant strains. In addition, we applied TIER-seq to the four strains, to identify cleavage sites of RNase E that are affected by GcvB. By intersecting these results with data of GcvB binding sites, we attempted to delineate the mechanisms by which GcvB interacts with RNase E, either in a translation-dependent or independent manner.
1) Our results suggest that competition between targets over binding to Hfq plays a major role in the regulatory outcomes that a sRNA exerts on different targets (Faigenbaum-Romm et al, 2020).
2) We developed a machine learning-based algorithm that predicts novel sRNAs from Hfq-mediated RIL-seq data. Applying this new algorithm to our original RIL-seq data of Escherichia coli K-12 (Melamed et al, 2016), we identified and verified experimentally novel Hfq-dependent sRNAs encoded in various genomic regions (Bar et al., 2021).
3) We completed the study of inference of the relative contribution of a sRNA to the response of cells to an environmental change (Barsheshet et al., 2022). Applying a special design of differential gene expression analysis by RNA-seq, we extract the contribution of a sRNA to the total change in the transcript level of each gene in response to an environmental change (applied to RyhB and GadF in Escherichia coli grown under iron limitation and acidic stress, respectively, and to MgrR in enteropathogenic E. coli (EPEC) grown on DMEM).
4) We published the sRNA-target network of the pathogenic bacterium EPEC, studied in collaboration with Prof. Ilan Rosenshine from our department (Pearl Mizrahi et al., 2021). We discovered novel sRNAs encoded in non-core genomic regions and found that sRNAs mediate substantial information flow between the core genome and genomic islands. We discovered a few sRNAs that act as “regulatory hubs” of virulence genes, one of which, MgrR, we studied in detail.
5) We completed deciphering the networks of additional two pathogenic bacteria, Salmonella enertica and Klebsiella pneumonia, in collaboration with Joerg Vogel (HIRI, Germany) and Yunn-Hwen Gan (NUS, Singapore), respectively. In the Salmonella study (Matera et al, 2022), we discovered a new sponging mechanism by a sRNA (OmpX) of another sRNA (MicF). In Klebsiella (Goh et al., 2024, under review), we found that the capsule, a major virulence component, is a hub target of sRNAs. Having the three sRNA-target networks of pathogenic bacteria and the network of E. coli K-12, we identified sRNA-target interactions that are conserved in all four bacteria, implying the importance of this regulation.
6) We applied ribo-seq to E. coli strains grown in normal conditions and under iron limitation, where the sRNA RyhB is induced. We analyze these data to study the effect of sRNAs on translation initiation and elongation. For the latter, we have been developing algorithms to identify ribosome pausing sites that are dependent on sRNA binding, suggesting possible regulation of translation elongation by sRNAs.
7) We completed the development of an original and accurate algorithm to determine transcript termini using only gene expression data (Bar et al., 2023). To this end, we exploit a specific read pattern that we identified at 3’ termini of transcripts in gene expression data generated by the RNAtag-seq protocol, a protocol that is widely used in microbiology research. We now apply our new algorithm to the wealth of RNAtag-seq data available for 32 bacteria grown under 12 conditions, identifying condition-dependent pre-mature transcaription termination sites that are conserved in evolution.
8) We completed the study of the interplay between a sRNA and an endoribonuclease, using GcvB and RNAse E as our model system (Barsheshet et al., 2024, in revision). To infer quantitatively the functional interaction between GcvB and RNase E, we appplied the approach of the double mutant cycle. This approach involves a special design of single and double mutants to which RNA-seq is applied. We can extract the separate contributions and mutual contributions of the endoribonuclease and sRNA to changes in gene expression by a special analysis of the data. We carried out these experiments for GcvB and RNase E, using RNA from E. coli wild type, ΔgcvB, rne ts conditional mutant and the double mutant strains. In addition, we applied TIER-seq to the four strains, to identify cleavage sites of RNase E that are affected by GcvB. By intersecting these results with data of GcvB binding sites, we attempted to delineate the mechanisms by which GcvB interacts with RNase E, either in a translation-dependent or independent manner.
We developed and made publicly accessible RILseqDB, a user-friendly database of sRNA-target interactions revealed by our RIL-seq technology. We applied RIL-seq to different bacteria, including pathogenic bacteria, generating precious resources of sRNA-target interactions. These are now presented in RILseqDB, which has sophisticated browsing and retrieval options(URL:https://rilseqdb.cs.huji.ac.il/).
We implemented the experimental/computational double mutant cycle approach in the sRNA field, for studying the interplay between sRNA and other regulators. We were approached by other researchers in the sRNA field who expressed interest in integrating the approach we developed in their research.
We discovered in RNA-seq data generated by RNAtag-seq protocol a distinct read pattern that classifies 3' transcript termini. Followingly, we developed an algorithm to determine 3' termini of transcripts, using only the read patterns in RNAtag-seq data. This opens the door to many large-scale studies of transcription termination exploiting the large body of RNAtag-seq sequencing data available. Especially, we now take advantage of this algorithm and the wealth of RNAtag-seq data available for many bacteria grown under different conditions to identify condition-dependent gene expression regulation by transcription termination and the effect of sRNAs on pre-mature transcription termination. Our results suggest that generation of sRNAs by premature transcription termination or processing by endoribonuclease is conserved across various bacterial species
.
We implemented the experimental/computational double mutant cycle approach in the sRNA field, for studying the interplay between sRNA and other regulators. We were approached by other researchers in the sRNA field who expressed interest in integrating the approach we developed in their research.
We discovered in RNA-seq data generated by RNAtag-seq protocol a distinct read pattern that classifies 3' transcript termini. Followingly, we developed an algorithm to determine 3' termini of transcripts, using only the read patterns in RNAtag-seq data. This opens the door to many large-scale studies of transcription termination exploiting the large body of RNAtag-seq sequencing data available. Especially, we now take advantage of this algorithm and the wealth of RNAtag-seq data available for many bacteria grown under different conditions to identify condition-dependent gene expression regulation by transcription termination and the effect of sRNAs on pre-mature transcription termination. Our results suggest that generation of sRNAs by premature transcription termination or processing by endoribonuclease is conserved across various bacterial species
.