Limited role for the DsrA and RprA regulatory RNAs in rpoS regulation in Salmonella enterica

Abstract

RpoS, the sigma factor of enteric bacteria that responds to stress and stationary phase, is subject to complex regulation acting at multiple levels, including transcription, translation, and proteolysis. Increased translation of rpoS mRNA during growth at low temperature, after osmotic challenge, or with a constitutively activated Rcs phosphorelay depends on two trans-acting small regulatory RNAs (sRNAs) in Escherichia coli. The DsrA and RprA sRNAs are both highly conserved in Salmonella enterica, as is their target, an inhibitory antisense element within the rpoS untranslated leader. Analysis of dsrA and rprA deletion mutants indicates that while the increased translation of RpoS in response to osmotic challenge is conserved in S. enterica, dependence on these two sRNA regulators is much reduced. Furthermore, low-temperature growth or constitutive RcsC activation had only modest effects on RpoS expression, and these increases were, respectively, independent of dsrA or rprA function. This lack of conservation of sRNA function suggests surprising flexibility in RpoS regulation.

FIG. 1.

Comparison of the regulation of lac operon fusions formed by MudJ insertion or plasmid integration. A. The rpoS coding sequence (bold line) and upstream leader are indicated, with labels indicating four sites at which fusions were isolated or constructed as described in the text. These positions are: 1 (+9 of the leader), 2 (+276 of the leader), 3 (codon 36 of rpoS), and 4 (codon 222 of rpoS). B. Analysis of lac expression from MudJ insertions at sites 1 to 4. Open bars indicate exponential-phase cultures, and filled bars indicate SP cultures, as defined in the text. SP induction of rpoS transcription is measured as the ratio of the value from the filled bars to open bars. C. The same experiment as in panel B, except that each fusion was made by integration of plasmid pKG137, as described in Materials and Methods. Strains for panel B were TE8804, TE8935, TE8936, and TE8794; strains for panel C were TE9052, TE9049, TE9050, and TE9051. Each bar represents the average of at least three experiments. Standard deviations were within 15% of the mean.

FIG. 2.

A. RNA sequence of a segment of the 565-nt S. enterica rpoS leader RNA, starting at nt 456 (110 nt upstream from the start codon), and folded to show pairing between the antisense element and the rpoS RBS region (stems I, II, and III). Paired regions flank the Shine-Dalgarno (S.D.) complementarity to 16S rRNA and extend to the start codon (UUG in S. enterica, AUG in E. coli). Nucleotides that differ between S. enterica and E. coli are marked by filled circles. The antisense element and the RBS region are connected by 63 nt, which are not shown but are indicated by the oval. B. Pairing is shown between the antisense element (extended by an additional 18 nt on the upstream side) and two different sRNA regulators of RpoS: DsrA and RprA. Paired nucleotides are indicated by vertical lines, and spaces have been introduced where needed to facilitate the alignment. The stems of the antisense element are overlined for reference. The gray boxes indicate a hexameric sequence whose complementarity with the target RpoS RNA is essential for sRNA function, as shown for E. coli (28, 29). The DsrA sequence shown starts at +1, and the RprA sequence starts at +28.

FIG. 3.

Effect of low temperature on RpoS. Cells carrying the RpoS-dependent reporter katE-lac [op] were grown in LB medium at either 37°C (A) or 18°C (B) to early exponential phase (OD₆₀₀ of 0.18) and assayed for β-galactosidase activity. The S. enterica strains used were wild type (TE6153), ΔdsrA (TE8608), ΔrprA (TE8610), and ΔdsrA ΔrprA (TE8613). The E. coli strains were wild type (TE6897) and dsrA1::cat (TE6913). The indicated strains were also analyzed for RpoS protein by Western blotting as described in Materials and Methods. Cells were grown to exponential phase (C) or stationary phase (D) in LB medium at 18°C. (E) A lac operon fusion to +11 of the S. enterica dsrA gene was constructed as described in Materials and Methods, and its expression was assayed by measuring β-galactosidase activity after growth to exponential phase in LB medium at the indicated temperatures.

FIG. 4.

Effect of osmotic challenge on rpoS-lac. The lac protein fusion employed was formed by insertion of the transposon MudK at codon 250 of the S. enterica rpoS gene (see Materials and Methods for details). Cultures were grown in LB medium at 30°C to early exponential phase, split into duplicate cultures, and grown to an OD₆₀₀ of 0.12. Cultures were challenged with prewarmed aliquots of either sucrose dissolved in LB medium (addition of ≈1/5 volume of 2 M sucrose to give 0.464 M final concentration; ≈16%) or LB alone. Samples were removed at time zero and at 15, 30, and 45 min after challenge. Cells were concentrated and assayed for total protein and for β-galactosidase activity. (A) Comparison of wild-type cells (squares) and a dsrA mutant (circles); sucrose-challenged cells are represented by filled symbols. (B) Comparison of the wild type to the ΔdsrA ΔrprA double mutant under the same conditions. (C) Data from a set of such experiments. Each bar represents the ratio of enzyme activities from the 45-min and time-zero samples, where both values have been normalized to the corresponding protein concentration. Dark bars represent sucrose-challenged cultures, and light bars are untreated controls. The strains used were wild type (TE9160), ΔdsrA (TE9213), and ΔdsrA ΔrprA (TE9219).

FIG. 5.

Effect of activated rcsC on rpoS-lac. In the left half of the figure, the lac protein fusion was formed by insertion of the transposon MudK at codon 22 of the S. enterica rpoS gene. In the right half, the lac fusion was the gmm-21::MudJ insertion (also called wcaH) (15). Two activated alleles of rcsC were compared to wild type, and each set consisted of strains either wild type or mutant for rprA. Cells were grown overnight to stationary phase in LB medium at room temperature (23 to 25°C) and assayed for β-galactosidase activity, normalized to the OD₆₀₀ as described in Materials and Methods. Results shown are the averages and standard deviations for at least seven independent experiments. Strains for rpoS-lac were rcsC⁺ rprA⁺ (TE9317), rcsC⁺ ΔrprA (TE9353), rcsC55 rprA⁺ (TE9316), rcsC55 ΔrprA (TE9352), rcsC64 rprA⁺ (TE9318), and rcsC64 ΔrprA (TE9354). The corresponding strains with the gmm-21::MudJ insertion were TE9334, TE9394, TE9333, TE9395, TE9368, and TE9396.

FIG. 6.

Expression of dsrA from P_BAD. A. Strains of E. coli or S. enterica with an rpoS-lac [pr] fusion in the bacterial chromosome, deleted for dsrA, and also carrying a plasmid P_BAD-dsrA construct from the species indicated, were induced with 0.02% arabinose and grown overnight to stationary phase in LB-Amp medium at 32°C. Activity of β-galactosidase is plotted for each strain, normalized to the expression seen with a vector control. B and C. Northern analysis of DsrA accumulation in E. coli and S. enterica strains carrying plasmids expressing either S. enterica dsrA (B) or E. coli dsrA (C) under the control of the P_BAD promoter. Strains carrying the appropriate empty vector were used as controls. Culture conditions, RNA purification, and Northern blotting were carried out as described in Materials and Methods. Strains were TE9416, TE9418, TE9419, TE9422, TE9424, TE9425, TE9426, TE9427, TE9428, TE9429, TE9430, and TE9431.

FIG. 7.

Effects of point mutations in the antisense element and RBS region on expression of rpoS in E. coli and S. enterica. S. enterica strains with mutations in the rpoS leader on the bacterial chromosome were constructed as described in Materials and Methods. The rpoS::MudK (codon 216) fusion and an hfq::Mud-Cam insertion were introduced by P22 transduction. E. coli fusion strains have been described elsewhere or were made in the same way (7). Cultures were grown overnight in LB medium at 23 to 25°C to SP and assayed for β-galactosidase activity. A. Locations of point mutations within the antisense element. B. Results with strains TE6266, TE6557, TE6558, TE6590, TE6369-2, TE6369-3, and TE6382. C. Results with strains TE8808, TE8815, and TE8852 to TE8860. D. Results for strains TE8808, TE8815, and TE9236 to TE9247.