σ²⁸-dependent small RNA regulation of flagella biosynthesis

Reviewer #1 (Public Review):

Bacteria can adapt to extremely diverse environments via extensive gene reprogramming at transcriptional and post-transcriptional levels. Small RNAs are key regulators of gene expression that participate in this adaptive response in bacteria, and often act as post-transcriptional regulators via pairing to multiple mRNA-targets.

In this study, Melamed et al. identify four E. coli small RNAs whose expression is dependent on sigma 28 (FliA), involved in the regulation of flagellar gene expression. Even though they are all under the control of FliA, expression of these 4 sRNAs peaks under slightly different growth conditions and each has different effects on flagella synthesis/number and motility. Combining RILseq data, structural probing, northern-blots and reporter assays, the authors show that 3 of these sRNAs control fliC expression (negatively for FliX, positively for MotR and UhpU) and two of them regulate r-protein genes from the S10 operon (again positively for MotR, and negatively for FliX). UhpU also directly represses synthesis of the LrhA transcriptional regulator, that in turn regulates flhDC (at the top of flagella regulation cascade). Based on RILseq data, the fourth sRNA (FlgO) has very few targets and may act via a mechanism other than base-pairing.

As r-protein S10 is also implicated in anti-termination via the NusB-S10 complex, the authors further hypothesize that the up-regulation of S10 gene expression by MotR promotes expression of the long flagellar operons through anti-termination. Consistent with this possible connection between ribosome and flagella synthesis, they show that MotR overexpression leads to an increase in flagella number and in the mRNA levels of two long flagellar operons, and that both effects are dependent on the NusB protein. Lastly, they provide data supporting a more general activating and repressing role for MotR and FliX, respectively, in flagellar genes expression and motility.

This study brings a lot of new information on the regulation of flagellar genes, from the identification of novel sigma 28-dependent sRNAs to their effects on flagella production and motility. It represents a considerable amount of work; the experimental data are clear and solid and support the conclusions of the paper. Even though mechanistic details underlying the observed regulations by MotR or FliX sRNAs are lacking, the effect of these sRNAs on fliC, several rps/rpl genes, and flagellar genes and motility is convincing.
The connection between r-protein genes regulation and flagellar operons is exciting and raises a few questions. First, from the RILseq data, chimeric reads with mRNA for r-proteins (including rpsJ) are not restricted to the sigma 28-dependent sRNAs (e.g. rpsJ-sucD3'UTR, rpsF-DicF, rplN-DicF, rplK-ChiX, rplU-CyaR, rpsT-CyaR, rpsK-CyaR, rpsF-MicA...), suggesting that regulation of r-protein synthesis by sRNAs is not necessarily related to flagella/motility. Second, it would be interesting to know if the flagellar operons are more sensitive than other long operons to antitermination following MotR overexpression? In other words, does pMotR similarly affect antitermination in rrn or other long operons?

The general effect of pMotR or pFliX on the expression of multiple middle and late flagellar genes is also interesting even though the mechanism is not clear. While it may be difficult to fully address it, testing whether some of these regulatory events depend on the control of fliC and/or the S10 operon could be relevant (by analyzing the effects in strains deleted for fliC or nusB for instance).

Reviewer #2 (Public Review):

This manuscript discusses the posttranscriptional regulation of flagella synthesis in Escherichia coli. The bacterial flagellum is a complex structure that consists of three major domains, and its synthesis is an energy-intensive process that requires extensive use of ribosomes. The flagellar regulon encompasses more than 50 genes, and the genes are activated in a sequential manner to ensure that flagellar components are made in the order in which they are needed. Transcription of the genes is regulated by various factors in response to environmental signals. However, little is known about the posttranscriptional regulation of flagella synthesis. The manuscript describes four UTR-derived sRNAs (UhpU, MotR, FliX, and FlgO) that are controlled by the flagella sigma factor σ28 (fliA) in Escherichia coli. The sRNAs have varied effects on flagellin protein levels, flagella number, and cell motility, and they regulate different aspects of flagella synthesis.
UhpU corresponds to the 3´ UTR of uhpT.

UhpU is transcribed from its own promoter inside the coding sequence of uhpT.

MotR originates from the 5´ UTR of motA. The promoter for motR is within the flhC CDS and is also the promoter of the downstream motAB-cheAW operon.

FliX originates from the 3´ UTR of fliC. Probably processed from parental mRNA.

FlgO originates from the 3´ UTR of flgL. Probably processed from parental mRNA.

This is a very interesting study that shows how sRNA-mediated regulation can create a complex network regulating flagella synthesis. The information is new and gives a fresh outlook at cellular mechanisms of flagellar synthesis. The presented work could benefit from additional experiments to confirm the effect of endogenous sRNAs expressed at natural level.

Reviewer #3 (Public Review):

Flagella are crucial for bacterial motility and virulence of pathogens. They represent large molecular machines that require strict hierarchical expression control of their components. So far, mainly transcriptional control mechanisms have been described to control flagella biogenesis. While several sRNAs have been reported that are environmentally controlled and regulate motility mainly via control of flagella master regulators, less is known about sRNAs that are co-regulated with flagella genes and control later steps of flagella biogenesis.

In this carefully designed and well-written study, the authors explore the role of four E. coli σ28-dependent 3' or 5' sUTR-derived sRNA in regulating flagella biogenesis. UhpU and MotR sRNAs are generated from their own σ28(FliA)-dependent promoter, while FliX and FlgO sRNAs are processed from the 3'UTRs of flagella genes under control of FliA. The authors provide an impressive amount of data and different experiments, including phenotypic analyses, genomics approaches as well as in-vitro and in-vivo target identification and validation methods, to demonstrate varied effects of three of these sRNAs (UhpU, FliX and MotR) on flagella biogenesis and motility. For example, they show different and for some sRNAs opposing phenotypes upon overexpression: While UhpU sRNA increases flagella number and motility, FliX has the opposite effect. MotR sRNA also increases the number of flagella, with minor effects on motility.

While the mechanisms and functions of the fourth sRNA, FlgO, remain elusive, the authors provide convincing experiments demonstrating that the three sRNAs directly act on different targets (identified through the analysis of previous RIL-seq datasets), with a variety of mechanisms. The authors demonstrate, UhpU sRNA, which derives from the 3´UTR of a metabolic gene, downregulates LrhA, a transcriptional repressor of the flhDC operon encoding the early genes that activate the flagellar cascade. According to their RIL-seq data analyses, UhpU has hundreds of additional potential targets, including multiple genes involved in carbon metabolism. Due to the focus on flagellar biogenesis, these are not further investigated in this study and the authors further characterize the two other flagella-associated sRNAs, FliX and MotR. Interestingly, they found that these sRNAs seem to target coding sequences rather than acting via canonical targeting of ribosome binding sites. The authors show FliX sRNA represses flagellin expression by interacting with the CDS of the fliC mRNA. Both FliX and MotR sRNA turn out to modulate the levels of ribosomal proteins of the S10 operon with opposite effects. MotR, which is expressed earlier, interacts with the leader and the CDS of rpsJ mRNA, leading to increased S10 protein levels and S10-NusB complex mediated anti-termination, promoting readthrough of long flagellar operons. FliX interacts with the CDSs of rplC, rpsQ, rpsS-rplV, repressing the production of the encoded ribosomal proteins. The authors also uncover MotR and FliX affect transcription selected representative flagellar genes, with an unknown mechanism.

Overall, this comprehensive study expands the repertoire of characterized UTR derived sRNAs and integrate new layers of post-transcriptional regulation into the highly complex flagellar regulatory cascade. Moreover, these new flagella regulators (MotR, FliX) act non-canonically, and impact protein expression of their target genes by base-pairing with the CDS of the transcripts. Their findings directly connect flagella biosynthesis and motility, highly energy consuming processes, to ribosome production (MotR and FliX) and possibly to carbon metabolism (UhpU).

Specific points to be considered:

- The authors use a crl- hyper-motile strain as WT strain for the study and sometimes also a crl+ strain is used. Can the authors comment on potential reasons why some phenotypes (e.g., UhpU and MotR effects on motility) are only detectable in the crl+ strain or vice versa? Is σS regulation important for the function of these sRNAs?

- In several experiments, a variant of MotR sRNA, MotR* that harbors a 3 nt mutation upstream of the seed sequence is used and seems to mediate stronger phenotypes (impact on flagellar number) upon overexpression compared to WT or phenotypes not retrieved for WT MotR (increased flagellin expression). It would be helpful to have some more clarification throughout the text, why this variant was used, even when OE of WT MotR already has impact on the target and how these three mutated nucleotides impact target regulation. For example, does MotR* show increased RNA stability or Hfq binding compared to MotR? Does the mutation in MotR* impact MotR structure (e.g., based on secondary structure predictions) or increase the complementarity with selected targets at potential secondary binding sites (e.g., based on target predictions)? For example, Fig. S7 shows additional regions of interaction between MotR and fliC mRNA beside the seed sequence. It is also suggested that MotR might have multiple interaction sites on rpsJ mRNA. Additional structure probing or biocomputational predictions could clarify these points.

- It is suggested that UphU impacts on motility via regulation of LrhA, which represses transcription of flhDC, and therefore the flagellar cascade. While LhrA-mediated regulation by UphU is validated based on reporter genes, the effect of UhpU OE on FlhDC levels is not directly examined (Fig. 3). Furthermore, as deletion of LrhA de-represses the flagellar cascade and UhpU was also shown to increase motility, the conclusions could be further strengthened by examining flhDC levels and/or the effect of ∆UhpU (if the sRNA part can be deleted) on motility (reduction) due to relieved down-regulation of LrhA.

-This study provides many opportunities for future follow-work. Now that the four sRNAs and some of their targets and opposing effects on flagella biogenesis have been identified, it will be interesting to see how the sRNAs themselves are temporally regulated throughout the flagella biogenesis cascade and which other targets are regulated by them. Future studies could also provide insights into the mechanism and function of FlgO sRNA, which seems to act via a different mechanism than base-pairing to target RNAs, as well as the global effects of regulation of ribosomal genes via FliX and MotR.

Author Response

Reviewer #1 (Public Review):

[...] This study brings a lot of new information on the regulation of flagellar genes, from the identification of novel sigma 28-dependent sRNAs to their effects on flagella production and motility. It represents a considerable amount of work; the experimental data are clear and solid and support the conclusions of the paper. Even though mechanistic details underlying the observed regulations by MotR or FliX sRNAs are lacking, the effect of these sRNAs on fliC, several rps/rpl genes, and flagellar genes and motility is convincing.

The connection between r-protein genes regulation and flagellar operons is exciting and raises a few questions. First, from the RILseq data, chimeric reads with mRNA for r-proteins (including rpsJ) are not restricted to the sigma 28-dependent sRNAs (e.g. rpsJ-sucD3'UTR, rpsF-DicF, rplN-DicF, rplK-ChiX, rplU-CyaR, rpsT-CyaR, rpsK-CyaR, rpsF-MicA...), suggesting that regulation of r-protein synthesis by sRNAs is not necessarily related to flagella/motility. Second, it would be interesting to know if the flagellar operons are more sensitive than other long operons to antitermination following MotR overexpression? In other words, does pMotR similarly affect antitermination in rrn or other long operons?

The general effect of pMotR or pFliX on the expression of multiple middle and late flagellar genes is also interesting even though the mechanism is not clear. While it may be difficult to fully address it, testing whether some of these regulatory events depend on the control of fliC and/or the S10 operon could be relevant (by analyzing the effects in strains deleted for fliC or nusB for instance).

We also think the connection between r-protein genes regulation and flagellar operons is exciting and raises some intriguing questions. While there are other RIL-seq chimeras for r-protein genes, the highest numbers are found for MotR and FliX. Nevertheless, understanding the impact of these other sRNAs on the r-protein operons and elucidating which long operons are most sensitive to antitermination following MotR overexpression are important directions for further studies.

Reviewer #2 (Public Review):

[...] This is a very interesting study that shows how sRNA-mediated regulation can create a complex network regulating flagella synthesis. The information is new and gives a fresh outlook at cellular mechanisms of flagellar synthesis. The presented work could benefit from additional experiments to confirm the effect of endogenous sRNAs expressed at natural level.

We agree that experiments regarding the endogenous effects of endogenous sRNAs are important. We provide such data in Figures 8 and S14 for MotR and FliX in a variety of assays: flagella numbers by electron microscopy, motility and competition assays, expression of flagellar genes by RT-qPCR and western analysis. We went to the trouble of constructing strains carrying point mutations in the chromosomal copies of these genes rather than deletions to avoid interfering with expression of motA and fliC given that MotR and FliX encompass the 5’ and 3’ UTRs respectively.

Reviewer #3 (Public Review):

[...] Overall, this comprehensive study expands the repertoire of characterized UTR derived sRNAs and integrate new layers of post-transcriptional regulation into the highly complex flagellar regulatory cascade. Moreover, these new flagella regulators (MotR, FliX) act non-canonically, and impact protein expression of their target genes by base-pairing with the CDS of the transcripts. Their findings directly connect flagella biosynthesis and motility, highly energy consuming processes, to ribosome production (MotR and FliX) and possibly to carbon metabolism (UhpU).

Specific points to be considered:

• The authors use a crl- hyper-motile strain as WT strain for the study and sometimes also a crl+ strain is used. Can the authors comment on potential reasons why some phenotypes (e.g., UhpU and MotR effects on motility) are only detectable in the crl+ strain or vice versa? Is σS regulation important for the function of these sRNAs?

• In several experiments, a variant of MotR sRNA, MotR* that harbors a 3 nt mutation upstream of the seed sequence is used and seems to mediate stronger phenotypes (impact on flagellar number) upon overexpression compared to WT or phenotypes not retrieved for WT MotR (increased flagellin expression). It would be helpful to have some more clarification throughout the text, why this variant was used, even when OE of WT MotR already has impact on the target and how these three mutated nucleotides impact target regulation. For example, does MotR* show increased RNA stability or Hfq binding compared to MotR? Does the mutation in MotR* impact MotR structure (e.g., based on secondary structure predictions) or increase the complementarity with selected targets at potential secondary binding sites (e.g., based on target predictions)? For example, Fig. S7 shows additional regions of interaction between MotR and fliC mRNA beside the seed sequence. It is also suggested that MotR might have multiple interaction sites on rpsJ mRNA. Additional structure probing or biocomputational predictions could clarify these points.

• It is suggested that UphU impacts on motility via regulation of LrhA, which represses transcription of flhDC, and therefore the flagellar cascade. While LhrA-mediated regulation by UphU is validated based on reporter genes, the effect of UhpU OE on FlhDC levels is not directly examined (Fig. 3). Furthermore, as deletion of LrhA de-represses the flagellar cascade and UhpU was also shown to increase motility, the conclusions could be further strengthened by examining flhDC levels and/or the effect of ∆UhpU (if the sRNA part can be deleted) on motility (reduction) due to relieved down-regulation of LrhA.

• This study provides many opportunities for future follow-work. Now that the four sRNAs and some of their targets and opposing effects on flagella biogenesis have been identified, it will be interesting to see how the sRNAs themselves are temporally regulated throughout the flagella biogenesis cascade and which other targets are regulated by them. Future studies could also provide insights into the mechanism and function of FlgO sRNA, which seems to act via a different mechanism than base-pairing to target RNAs, as well as the global effects of regulation of ribosomal genes via FliX and MotR.

We thank the reviewer for the constructive comments about the variation between the crl- and crl+ strains, and about the use of MotR versus MotR*, and will address these points in a revised version of the manuscript. Regarding the UhpU-mediated regulation, we agree that assays of flhDC expression will strengthen our conclusions. We share the reviewer opinion regarding many opportunities for future follow-up work.