Mechanism of positive regulation by DsrA and RprA small noncoding RNAs: pairing increases translation and protects rpoS mRNA from degradation

Abstract

Small noncoding RNAs (sRNAs) regulate gene expression in Escherichia coli by base pairing with mRNAs and modulating translation and mRNA stability. The sRNAs DsrA and RprA stimulate the translation of the stress response transcription factor RpoS by base pairing with the 5' untranslated region of the rpoS mRNA. In the present study, we found that the rpoS mRNA was unstable in the absence of DsrA and RprA and that expression of these sRNAs increased both the accumulation and the half-life of the rpoS mRNA. Mutations in dsrA, rprA, or rpoS that disrupt the predicted pairing sequences and reduce translation of RpoS also destabilize the rpoS mRNA. We found that the rpoS mRNA accumulates in an RNase E mutant strain in the absence of sRNA expression and, therefore, is degraded by an RNase E-mediated mechanism. DsrA expression is required, however, for maximal translation even when rpoS mRNA is abundant. This suggests that DsrA protects rpoS mRNA from degradation by RNase E and that DsrA has a further activity in stimulating RpoS protein synthesis. rpoS mRNA is subject to degradation by an additional pathway, mediated by RNase III, which, in contrast to the RNase E-mediated pathway, occurs in the presence and absence of DsrA or RprA. rpoS mRNA and RpoS protein levels are increased in an RNase III mutant strain with or without the sRNAs, suggesting that the role of RNase III in this context is to reduce the translation of RpoS even when the sRNAs are acting to stimulate translation.

FIG. 1.

(A) Nucleotides in the 5′ UTR of the rpoS mRNA upstream of the start codon form a hairpin structure in the absence of binding to an sRNA. Bases are numbered backward from the start codon. The transcription start site was mapped at 584 bases upstream of the start codon (see text). The region that pairs with the sRNAs is highlighted, and the start codon is underlined. Mutated bases are boxed, and nucleotide changes in the rpoS* allele are indicated. (B) Predicted pairings are shown between the rpoS or rpoS* mRNA with DsrA or DsrA*. The plus sign and the arrow indicate 5′ ends of the cleaved forms of rpoS mRNA as mapped by 5′ RACE (+, high-temperature forms; arrow, DsrA*-rpoS* forms). Boxed nucleotides indicate the bases that were mutated in the dsrA* and rpoS* alleles. (C) Predicted pairings between the rpoS or rpoS* mRNA with RprA or RprA* are shown, and boxes indicate mutated nucleotides.

FIG. 2.

RpoS expression in strains with optimal or mismatched pairing between sRNAs and rpoS. CM1062 (ΔdsrA rprA::kan rpoS⁺ Δara714) and CM1063 (ΔdsrA rprA::kan rpoS* Δara714) with the pNM12 vector or its derivatives expressing dsrA or dsrA* (A) and CM1062 and CM1063 with the vector or its derivatives expressing rprA or rprA* (B) were grown in LB Ap at 30°C to mid-exponential phase, and sRNA expression was induced with 0.02% arabinose for 20 min. Total protein was precipitated and analyzed by Western blotting. RpoS accumulation in each strain is described as a percentage of the accumulation in the rpoS⁺/pBAD-dsrA⁺ (A) or rpoS⁺/pBAD-rprA⁺ (B) strain. Accumulation is presented as the mean percentage ± the standard deviation (n = 3).

FIG. 3.

Northern blot analysis of rpoS mRNA levels in strains with optimal or mismatched pairing between sRNAs and rpoS. CM1062 (ΔdsrA rprA::kan rpoS⁺ Δara714) and CM1063 (ΔdsrA rprA::kan rpoS* Δara714) with the pNM12 vector or its derivatives expressing dsrA or dsrA* (A) and CM1062 and CM1063 with the pNM12 vector or its derivatives expressing rprA or rprA* (B) were grown in LB Ap at 30°C to mid-exponential phase, and sRNA expression was induced with 0.02% arabinose for 20 min. RNA was collected by the hot phenol method and analyzed by Northern blotting with the probe RpoS-N3. rpoS mRNA accumulation is described as a percentage of the accumulation in the rpoS⁺/pBAD-dsrA⁺ (A) or rpoS⁺/pBAD-rprA⁺ (B) strains. Accumulation is presented as the mean percentage ± the standard deviation (n = 3). (C) rpoS mRNA was examined in the same manner as for panels A and B for the RNase III mutant strains (CM1082 and CM1083) containing the same plasmids. rpoS mRNA accumulation was measured relative to that of the rpoS⁺ rnc⁺-pBAD-dsrA⁺ strain.

FIG. 4.

rpoS mRNA half-life determination when it is paired with DsrA or RprA. CM1000 (ΔdsrA rpoS⁺)/pBAD-dsrA and CM1001 (ΔdsrA rpoS*)/pBAD-dsrA* (A) and JNB001 (ΔdsrA rprA::kan rpoS⁺) and JNB002 (ΔdsrA rprA::kan rpoS*) with pBAD-rprA or pBAD-rprA* (B) were grown in LB Ap at 30°C to mid-exponential phase, and sRNA expression was induced with 0.02% arabinose for 15 min. RNA samples were collected by the hot phenol method at the indicated times after the addition of 250 μg/ml rifampin (rif). rpoS mRNA levels were analyzed by Northern blotting. The experiment whose results are shown in panel A was performed three times, resulting in half-lives measured for rpoS and rpoS* of 2 to 3 min in each case. The experiment whose results are shown in panel B was performed four times, with some variability in half-lives being measured for rpoS and rpoS*. In each experiment, the mRNA half-lives of the mismatched pairs (rpoS⁺-rprA* and rpoS*-rprA⁺) are shorter than that of the optimally paired sets. However, the mRNA half-lives in the rpoS⁺-rprA⁺ and rpoS*-rprA* strains varied from 4 to 15 min, and the mRNA half-lives in the mismatched pairs varied from 1 to 10 min. (C and D) Graphical representation of rpoS mRNA decay.

FIG. 5.

rpoS mRNA and RpoS levels in a Δrnc mutant strain in the presence and absence of DsrA. CM1000 (rnc⁺ rne wild type), CM1010 [rnc⁺ rne(Ts)], CM1050 (Δrnc-1223::cat rne wild type), and CM1052 [Δrnc-1223::cat rne(Ts)] with the pNM12 vector or pBAD-dsrA were grown at 30°C in LB Ap to mid-exponential phase; sRNA expression was induced with 0.02% arabinose for 20 min. Samples were collected for protein and RNA isolation, as described in Materials and Methods (O.D.₆₀₀, optical density at 600 nm). (A) Northern blot analysis of RNA samples was performed using the oligonucleotide probe RpoS-N3 to detect rpoS mRNA. (B) Western blot analysis of protein samples was performed using the anti-RpoS antibody to detect RpoS. Graphical analyses show the mean accumulation of the full-length rpoS mRNA or RpoS protein ± the standard deviation (n = 3) relative to that of CM1000 with the vector control. Samples were also tested by Western blotting for EF-Tu, which was close to identical in each sample. Values over the bars indicate the mean fold change.

FIG. 6.

rpoS mRNA and RpoS levels in an rne(Ts) mutant strain in the presence and absence of DsrA. CM1000 (rnc⁺ rne wild type), CM1010 [rnc⁺ rne(Ts)], CM1050 (Δrnc rne wild type) and CM1052 [Δrnc rne(Ts)] with the pNM12 vector or pBAD-dsrA were grown at 30°C in LB Ap to mid-exponential phase; sRNA expression was induced with 0.02% arabinose for 20 min. Cultures were heat shocked for 10 min at 43.5°C, and RNA and protein samples were collected, as described in Materials and Methods (O.D.₆₀₀, optical density at 600 nm). (A) Northern blot analysis of RNA samples was performed using the oligonucleotide probe RpoS-N3 to detect rpoS mRNA. Graphical analysis shows the mean accumulation of full-length rpoS mRNA (dark bars) and the truncated form of the rpoS mRNA (light bars) ± the standard deviation (n = 3) relative to that of CM1000 with the vector control. rpoS mRNA accumulation in lane 1 is 3 times that for the same strain before heat shock (at 30°C). (B) Western blot analysis of protein samples was performed using the anti-RpoS antibody to detect RpoS. Protein samples were diluted 5-fold before electrophoresis. Graphical analysis shows mean accumulation of RpoS protein ± the standard deviation (n = 3) relative to that of CM1000 with the vector control. RpoS protein accumulation in lane 1 is 20 times that of the same strain before heat shock; in lane 5, RpoS accumulation is 10 times that of the same strain before heat shock. As with Fig. 5, equal protein input per lane was confirmed by Western blotting for EF-Tu. Values over the bars indicate the mean fold change.