CsrA selectively modulates sRNA-mRNA regulator outcomes

Abstract

Post-transcriptional regulation, by small RNAs (sRNAs) as well as the global Carbon Storage Regulator A (CsrA) protein, play critical roles in bacterial metabolic control and stress responses. The CsrA protein affects selective sRNA-mRNA networks, in addition to regulating transcription factors and sigma factors, providing additional avenues of cross talk between other stress-response regulators. Here, we expand the known set of sRNA-CsrA interactions and study their regulatory effects. In vitro binding assays confirm novel CsrA interactions with ten sRNAs, many of which are previously recognized as key regulatory nodes. Of those 10 sRNA, we identify that McaS, FnrS, SgrS, MicL, and Spot42 interact directly with CsrA in vivo. We find that the presence of CsrA impacts the downstream regulation of mRNA targets of the respective sRNA. In vivo evidence supports enhanced CsrA McaS-csgD mRNA repression and showcases CsrA-dependent repression of the fucP mRNA via the Spot42 sRNA. We additionally identify SgrS and FnrS as potential new sRNA sponges of CsrA. Overall, our results further support the expanding impact of the Csr system on cellular physiology via CsrA impact on the regulatory roles of these sRNAs.

Keywords: CsrA/RsmA; Posttranscriptional regulation; RNA-protein interactions; sRNA (small RNA); sRNA regulatory network.

FIGURE 1

Overlap of sRNA-mRNA and CsrA-RNA post-transcriptional regulatory networks. (A) The canonical CsrA regulatory network: CsrA binds mRNA targets to either repress or activate translation. CsrA is then regulated by sRNAs CsrB and CsrB, which can sequester up to 9 and 5 copies of CsrA, respectively. (B) Concurrently, sRNAs regulate mRNAs targets in response to external stimuli, these interactions are assisted by RNA Binding Proteins (RBPs) such as Hfq and ProQ. (C) sRNAs such as McaS are overlapping sRNAs that both regulate mRNA targets, as well as CsrA. There is potentially many other sRNAs that can fall into both regulatory modes.

FIGURE 2

sRNAs bind CsrA in vitro. In vitro evaluation of 14 sRNAs with CsrA using Electrophoretic Mobility Shift Assays (EMSAs). These assays were performed by titrating CsrA concentrations (purple), as well as in the presence and absence of Hfq (black). The sRNAs tested are as follows: (A) SgrS, (B) ChiX, (C) FnrS, (D) RprA, (E) DsrA, (F) CyaR, (G) Spot42, (H) GadY, (I) McaS. Arrows indicate bound complexes: CsrA-RNA complexes (purple), Hfq-RNA complexes (black), terniary CsrA-Hfq-RNA complexes (gray).

FIGURE 3

FnrS, SgrS, McaS, and Spot42 alter fluorescence of glgC-gfp in a CsrA-dependent manner. (A) Plasmid schematics of in vivo reporter assays to evaluate the effects of an sRNA on CsrA regulation of a known target, the 5′ UTR of the glgC mRNA. The 5′ UTR of the glgC + 100 nts of the CDS were fused to gfp and constitutively expressed from a plasmid. CsrA and each sRNA were expressed from a second plasmid under aTc and IPTG-inducible control, respectively. (B) Each sRNA may interact with CsrA and affect CsrA repression of the glgC-gfp mRNA fusion. (C) Additionally, each sRNA may interact with the mRNA fusion directly. (D) Fluorescence ratio of the glgC-gfp mRNA fusion between the presence and absence of each sRNA, as well as the presence of CsrA. (E) Fluorescence ratio of the glgC-gfp mRNA fusion between the presence and absence of each sRNA, in the absence of CsrA.

FIGURE 4

In vivo mutational assays confirm that direct sRNA-CsrA interaction enables CsrA-sponging activity of McaS, SgrS, and FnrS. (A) Fluorescence ratio of the glgC-gfp mRNA fusion reporter between the presence and absence of the FnrS, McaS, SgrS, CyaR, and Spot42 sRNAs. Both the wild type (blue) and mutant (grey) sRNAs were tested. For descriptions of how the sRNAs mutants were generated to abrogate CsrA interactions see Materials and Methods. (B–E) Northern Blots of each wild and mutant sRNA to ensure consistent stability between the wild type and mutant sRNAs.

FIGURE 5

McaS-csgD repression is enhanced by direct CsrA interactions under native expression conditions. (A) Fluorescence of a csgD-gfp mRNA reporter in response to both the McaS sRNA and CsrA expression tested combinatorially. The fluorescence ratios of the csgD-gfp reporter between the presence and absence of McaS are calculated with and without CsrA expression. The csgD-gfp mRNA reporter was constructed similarly to the glgC-gfp mRNA reporter. (B) Effect of CsrA on McaS-csgD regulation evaluated by the Congo Red plate to measure endogenous curli expression. (Quadrant I) Wild type E. coli transformed with -an empty control version of a low expression plasmid (pEmpty), (II) ΔmcaS E. coli transformed with pEmpty (III) pMcaS wild type, and (IV) pMcaS mutant (5′-most GGA:CCA) were grown for 48 h with limited NaCl as to induce native curli expression. Plasmid-based induction of wild type McaS (1 mM IPTG) effectively repressed the csgD mRNA (white colonies) compared to wild type E. coli (dark red colonies), while mutant McaS (5′-most GGA:CCA) showed minor repression (light red colonies). (C) Predicted secondary structure of McaS (Vienna RNA) (Gruber et al., 2008; Lorenz et al., 2011), with GGA motifs (red-circled nts), mutations (blue arrows and nucleotides), and known csgD and flhD mRNA binding sites indicated (purple and yellow outlines, respectively). (D) Proposed effect of CsrA on McaS-csgD regulation: CsrA enhances repression by increasing mRNA-sRNA binding site accessibility.

FIGURE 6

CsrA is necessary for Spot42 regulation of fucP target under native conditions. (A) Fluorescence of a fucP-gfp mRNA reporter in response to both the Spot42 sRNA and CsrA expression tested combinatorically. The fluorescence ratios of the fucP-gfp reporter between the presence and absence of Spot42 are calculated with and without CsrA expression. The fucP-gfp mRNA reporter was constructed similarly to the glgC-gfp mRNA reporter. (B) Expression of the fucP-gfp mRNA reporter in WT, a mutant Spot42, and a Δspf strain. The mutations made to Spot42 reduced but did not fully eliminate CsrA-Spot42 interactions. In this assay, cultures containing the fucP-gfp reporter plasmid were grown in minimal media and Spot42, when present, was expressed from the genome using a bolus addition of glucose. (C) Predicted secondary structure of Spot42 (Vienna RNA), with GGA or GGN motifs (red-circled nts) and known fucP mRNA (maroon and yellow) and Hfq (green) binding sites indicated. (D) General workflow of low copy plasmid reporter assay and results. Fluorescence a fucP-gfp mRNA reporter in response to Spot42 and CsrA expression in a genomic context. The fucP-mRNA reporter was constitutively expressed from a low copy plasmid, while Spot42 is induced from the same plasmid. This was tested in the Δspf and ΔspfΔcsrA::kan strains of E. coli, to evaluate Spot42-fucP regulation under native CsrA expression. Ratios of the fucP-gfp reporter we calculated in the presence and absence of CsrA for the WT and csrA::kan strains.