The phosphorelay BarA/SirA activates the non-cognate regulator RcsB in Salmonella enterica

Abstract

To survive an environmental stress, organisms must detect the stress and mount an appropriate response. One way that bacteria do so is by phosphorelay systems that respond to a stress by activating a regulator that modifies gene expression. To ensure an appropriate response, a given regulator is typically activated solely by its cognate phosphorelay protein(s). However, we now report that the regulator RcsB is activated by both cognate and non-cognate phosphorelay proteins, depending on the condition experienced by the bacterium Salmonella enterica serovar Typhimurium. The RcsC and RcsD proteins form a phosphorelay that activates their cognate regulator RcsB in response to outer membrane stress and cell wall perturbations, conditions Salmonella experiences during infection. Surprisingly, the non-cognate phosphorelay protein BarA activates RcsB during logarithmic growth in Luria-Bertani medium in three ways. That is, BarA's cognate regulator SirA promotes transcription of the rcsDB operon; the SirA-dependent regulatory RNAs CsrB and CsrC further increase RcsB-activated gene transcription; and BarA activates RcsB independently of the RcsC, RcsD, and SirA proteins. Activation of a regulator by multiple sensors broadens the spectrum of environments in which a set of genes is expressed without evolving binding sites for different regulators at each of these genes.

Fig 1. The phosphorelay BarA/SirA activates the non-cognate regulator RcsB.

Upon polymyxin B-induced outer membrane damage or mecillinam-induced cell wall damage (left panel), the sensor kinase RcsC and the phosphotransferase RcsD activate their cognate response regulator RcsB, which then alters transcription of its target genes (e.g., rprA). The single-headed arrows for the RcsC/RcsD/RcsB phosphotransfer cascade indicate that upon outer membrane and cell wall damage, the phosphate flow is directed toward RcsB. Under these conditions, BarA is dispensable for RcsB activation. Upon rapid growth in LB media (right panel), BarA uses multiple mechanisms to activate RcsB. That is, BarA’s cognate regulator SirA promotes transcription of the rcsDB operon, increasing the amounts of the RcsD and RcsB proteins; the SirA-dependent regulatory RNAs CsrB and CsrC reduce RcsC expression, probably through titration of the CsrA protein, thus limiting RcsB dephosphorylation by RcsC and RcsD. BarA also activates RcsB independently of the RcsC, RcsD, and SirA proteins. The latter mechanism may entail direct phosphotransfer from BarA to RcsB and is depicted by a dashed arrow and a question mark. The double-headed arrows for the RcsC/RcsD/RcsB phosphotransfer cascade indicate that upon rapid growth in LB media, the phosphate flow may be reversed, resulting in the phosphate being taken away from RcsB by RcsD and RcsC. RcsF is an outer membrane lipoprotein that senses envelope stress. IgaA is a negative regulator of the RcsC/RcsD/RcsB phosphorelay.

Fig 2. The response regulator RcsB can be activated independently of the RcsC, RcsD, and RcsF proteins.

Fluorescence from wild-type (14028s), rcsB (EG12925), rcsC (HS1350), rcsD (HS1382), rcsC rcsD (HS1383) and rcsF (HS1326) Salmonella harboring plasmid pRprA-GFP (rprA-gfp) following 24 h of growth on LB solid medium without (-NaCl) or with (+NaCl) NaCl. Data are representative of three independent experiments, which gave similar results. Quantification of the fluorescence is provided on the right panel of the figure. Values derived from three independent experiments (mean ± standard deviation) were statistically analyzed by Prism 8 using two-tailed unpaired t test. Statistical significance is indicated by * P<0.05, **P<0.01, *** P<0.001; ns, not significant. Error bars indicate standard deviation.

Fig 3. RcsB activation by the phosphorelay sensor BarA.

(A) BarA is required for full RcsB activation on LB agar plates. Fluorescence from wild-type (14028s), rcsB (EG12925), rcsC (HS1350), rcsD (HS1382), rcsC rcsD (HS1383), barA (HS1520), barA rcsC (HS1521), barA rcsD (HS1522) and barA rcsC rcsD (HS1523) Salmonella harboring plasmid pRprA-GFP (rprA-gfp) following 24 h of growth on LB solid medium without (-NaCl) or with (+NaCl) NaCl. Data are representative of two independent experiments, which gave similar results. (B) BarA activates RcsB in a time-dependent manner. Fluorescence from wild-type (14028s), rcsB (EG12925), rcsC (HS1350), rcsD (HS1382), rcsC rcsD (HS1383), barA (HS1520), barA rcsC (HS1521), barA rcsD (HS1522) and barA rcsC rcsD (HS1523) Salmonella harboring plasmid pRprA-GFP (rprA-gfp) or pVector (empty pFPV25 vector) following 9 h of growth in LB liquid medium without (-NaCl) or with (+NaCl) NaCl. For each strain, relative fluorescence was obtained by subtracting the fluorescence of the strain harboring pVector from the fluorescence of the strain harboring pRprA-GFP. The obtained value was then divided by OD₆₀₀. Error bars represent standard deviation from three independent experiments. Values derived from three independent experiments (mean ± standard deviation) were statistically analyzed by Prism 8 using two-tailed unpaired t test. Significance values (P) are reported in the text. Relative fluorescence values (left axis) are represented by solid lines and OD₆₀₀ values (right axis) by dotted lines. (C) Western blot analysis of extracts prepared from wild-type (14028s), barA (HS1520) and rcsB (EG12925) grown in LB liquid medium without (-NaCl) or with (+NaCl) NaCl in late exponential phase (OD₆₀₀ of ~1.0) following separation on Phos-tag SDS-PAGE to detect RcsB and RcsB-P. Samples were analyzed with antibodies directed to RcsB or AtpB proteins. Crude extracts from rcsBD56Q (HS1483) and rcsC11 (EG14873) strains overnight cultures were used as negative and positive controls for RcsB phosphorylation, respectively. The rcsBD56Q mutant encodes an RcsB protein that cannot be phosphorylated at the conserved aspartate located at position 56 [14] and the rcsC11 strain harbors the rcsC11 point mutation promoting constitutive Rcs system activation [67].

Fig 4. BarA activates RcsB via its cognate response regulator SirA.

Fluorescence from wild-type (14028s), rcsB (EG12925), rcsC (HS1350), rcsD (HS1382), rcsC rcsD (HS1383), sirA (HS1565), sirA rcsC (HS1566), sirA rcsD (HS1567) and sirA rcsC rcsD (HS1568) Salmonella harboring plasmid pRprA-GFP (rprA-gfp) following 24 h of growth on LB solid medium without (-NaCl) or with (+NaCl) NaCl. Data are representative of two independent experiments, which gave similar results.

Fig 5. BarA activation of RcsB via the SirA-activated regulatory RNAs CsrB and CsrC.

(A) The regulatory RNAs CsrB and CsrC are required for full RcsB activation on LB agar plates. Fluorescence from wild-type (14028s), rcsB (EG12925), rcsC (HS1350), rcsD (HS1382), rcsC rcsD (HS1383), csrB csrC (HS1651), csrB csrC rcsC (HS1654), csrB csrC rcsD (HS1655) and csrB csrC rcsC rcsD (HS1656) Salmonella harboring plasmid pRprA-GFP (rprA-gfp) following 24 h of growth on LB solid medium without (-NaCl) or with (+NaCl) NaCl. Data are representative of two independent experiments, which gave similar results. (B) Fluorescence from wild-type (14028s), barA (HS1520), sirA (HS1565) and csrB csrC (HS1651) Salmonella harboring plasmid pRprA-GFP (rprA-gfp) following 24 h of growth on LB solid medium without (-NaCl) or with (+NaCl) NaCl. Data are representative of three independent experiments, which gave similar results. Quantification of the fluorescence is provided next to the plate images. Values derived from three independent experiments (mean ± standard deviation) were statistically analyzed by Prism 8 using two-tailed unpaired t test. Statistical significance is indicated by * P<0.05, **P<0.01; ns, not significant. Error bars indicate standard deviation. (C) Western blot analysis of crude extracts from rcsC-3XFLAG (HS539) and rcsC-3XFLAG csrB csrC (HS2263) strains grown in LB NaCl-free broth. Samples were analyzed with antibodies directed to the FLAG epitope or the RpoB protein. Data are representative of three independent experiments, which gave similar results. RcsC levels for csrB csrC mutant strain relative to wild-type Salmonella are marked below (fold increase ± standard deviation). Values were statistically analyzed by Prism 8 using two-tailed unpaired t test. Statistical significance is indicated by * P<0.05, ** P<0.01; ns, not significant.

Fig 6. SirA promotes rcsDB transcription.

(A) SirA is required for full rcsDB transcription. Fluorescence from wild-type (14028s), sirA (HS1565), rcsC (HS1350) and rcsC sirA (HS1566) Salmonella harboring pRcsD_-293-GFP (rcsD-gfp) or pVector (empty pFPV25 vector) following 9 h of growth in LB liquid medium without (-NaCl) or with (+NaCl) NaCl. For each strain, relative fluorescence was obtained by subtracting the fluorescence of the strain harboring pVector from the fluorescence of the strain harboring pRcsD-GFP. The value obtained was then divided by OD₆₀₀. Error bars represent standard deviation from three independent experiments. Values derived from three independent experiments (mean ± standard deviation) were statistically analyzed by Prism 8 using two-tailed unpaired t test. Significance values (P) are reported in the text. Relative fluorescence values (left axis) are represented by solid lines and OD₆₀₀ values (right axis) by dotted lines. (B) SirA-mediated activation of rcsDB transcription requires BarA and occurs independently of CsrB and CsrC. Fluorescence from wild-type (14028s), barA (HS1520), sirA (HS1565) and csrB csrC Salmonella harboring pRcsD_-293-GFP (rcsD-gfp) with pSirA or pVector (empty pACYC184 vector) following 24 h of growth on LB solid medium without (-NaCl) or with (+NaCl) NaCl. Data are representative of two independent experiments, which gave similar results. (C) Primer extension analysis of rcsDB_-36 and rcsDB_-157 levels in wild-type (14028s) Salmonella harboring pSirA or pVector (empty pACYC184 vector) grown in LB broth. Primer extension reaction was carried out on total RNA samples using primer W4171. (D) Western blot analysis of crude extracts from rcsD-HA (HS1309) and rcsB-FLAG (HS717) Salmonella harboring pSirA or pVector (empty pACYC184 vector) grown in LB broth. Samples were analyzed with antibodies directed to the FLAG or HA epitopes or the RpoB protein. Data are representative of three independent experiments, which gave similar results. RcsB and RcsD levels for pSirA strain relative pVector strain are marked below. Values were statistically analyzed by Prism 8 using two-tailed unpaired t test. Statistical significance is indicated by * P<0.05, ** P<0.01; ns, not significant. (E) Identification of the region upstream of the rcsD coding region required for SirA-mediated activation of rcsDB. Fluorescence from wild-type (14028s) Salmonella harboring pRcsD_-293-GFP (rcsD-gfp), pRcsD_-270-GFP, pRcsD_-235-GFP, pRcsD_-220-GFP or pRcsD_-110-GFP with pSirA or pVector (empty pACYC184 vector) following 24 h of growth on LB solid medium without (-NaCl) or with (+NaCl) NaCl was monitored. SirA activation is defined by increased fluorescence in the pSirA strain as compared to the pVector strain (see S10 Fig for plates images). Plus sign (+) indicates fusion activation, plus/minus sign (+/-) indicates weak fusion activation, and minus sign (-) indicates no activation. The numbers refer to locations relative to the rcsD start codon. Data are representative of two independent experiments, which gave similar results.

Fig 7. BarA activates RcsB independently of the SirA, RcsC, and RcsD proteins.

(A) Effect of barA inactivation on RcsB activity in a rcsC rcsD sirA background. Fluorescence from rcsC rcsD (HS1383), barA rcsC rcsD (HS1523), sirA rcsC rcsD (HS1568), barA sirA rcsC rcsD (HS1796) Salmonella harboring plasmid pRprA-GFP (rprA-gfp) following 24 h of growth on LB solid medium without (-NaCl) or with (+NaCl) NaCl. Quantification of fluorescence is provided in the right panel of the figure. Values derived from three independent experiments (mean ± standard deviation) were statistically analyzed by Prism 8 using two-tailed unpaired t test. Statistical significance is indicated by **P<0.01, by ***P<0.001; ns, not significant. Error bars indicate standard deviation. (B) Heterologous expression of the barA gene recovers wild-type levels of fluorescence in a barA sirA rcsC rcsD strain. Fluorescence from wild-type (14028s) and HS1796 (barA sirA rcsC rcsD) Salmonella harboring plasmid pRprA-GFP (rprA-gfp) with pBarA or pVector (empty pACYC184 vector) following 24 h of growth on LB solid medium without (-NaCl) or with (+NaCl) NaCl. Data are representative of two independent experiments, which gave similar results. (C) The BarA and RcsB proteins interact. Anti-HA pull-down assay showing interactions between in vitro–synthesized BarA-HA, BarA_198-918-HA, RcsB-FLAG, SirA-FLAG and PhoP-FLAG proteins. Samples were analyzed by Western blotting using antibodies recognizing the HA or FLAG epitopes.

Fig 8. BarA is dispensable for RcsB activation upon damage of the outer membrane or cell wall.

(A) Fluorescence from wild-type (14028s), rcsB (EG12925), rcsC (HS1350), rcsD (HS1382), rcsC rcsD (HS1383) and barA (HS1520) Salmonella harboring plasmid pRprA-GFP (rprA-gfp) or pVector (empty pFPV25 vector) upon 2 h treatment with a sublethal concentration (0.5 μg/ml) of polymyxin B (+ PMB) or with water (-PMB) in LB liquid medium. Polymyxin B was added in early exponential phase (OD₆₀₀ of ~0.15). Relative fluorescence values (left axis) are represented by solid lines and OD₆₀₀ values (right axis) by dotted lines. Values derived from three independent experiments (mean ± standard deviation) were statistically analyzed by Prism 8 using two-tailed unpaired t test. Significance values (P) are reported in the text. (B) Fluorescence from wild-type (14028s), rcsB (EG12925), rcsC (HS1350), rcsD (HS1382), rcsC rcsD (HS1383) and barA (HS1520) Salmonella harboring plasmid pRprA-GFP (rprA-gfp) or pVector (empty pFPV25 vector) upon 2 h treatment with a sublethal concentration (10 μg/ml) of mecillinam (+ MPC) or with water (-MPC) in LB liquid medium. Mecillinam was added in early exponential phase (OD₆₀₀ of ~0.15). For each strain, relative fluorescence was obtained by subtracting the fluorescence of the strain harboring pVector from the fluorescence of the strain harboring plasmid pRprA-GFP. The obtained value was then divided by OD₆₀₀. Error bars represent standard deviation from three independent experiments. Relative fluorescence values (left axis) are represented by solid lines and OD₆₀₀ values (right axis) by dotted lines. Values derived from three independent experiments (mean ± standard deviation) were statistically analyzed by Prism 8 using two-tailed unpaired t test. Significance values (P) are reported in the text.