An RNase III-processed sRNA coordinates sialic acid metabolism of Salmonella enterica during gut colonization

Abstract

Sialic acids derived from colonic mucin glycans are crucial nutrients for enteric bacterial pathogens like Salmonella. The uptake and utilization of sialic acid in Salmonella depend on coordinated regulons, each activated by specific metabolites at the transcriptional level. However, the mechanisms enabling crosstalk among these regulatory circuits to synchronize gene expression remain poorly understood. Here, we identify ManS, a small noncoding RNA derived from the 3' UTR of STM1128 mRNA transcribed from a Salmonella enterica-specific genetic locus, as an important posttranscriptional regulator coordinating sialic acid metabolism regulons. ManS is primarily processed by RNase III and, along with its parental transcripts, is specifically activated by N-acetylmannosamine (ManNAc), the initial degradation product of sialic acid. We found that the imperfect stem-loop structure at the 5' end of ManS allows RNase III to cleave in a noncanonical manner, generating two functional types of ManS with the assistance of RNase E and other RNases: short isoforms with a single seed region that regulate the uptake of N-acetylglucosamine, an essential intermediate in sialic acid metabolism; and long isoforms with an additional seed region that regulate multiple genes involved in central and secondary metabolism. This sophisticated regulation by ManS significantly impacts ManNAc metabolism and S. enterica's competitive behavior during infection. Our findings highlight the role of sRNA in coordinating transcriptional circuits and advance our understanding of RNase III-mediated processing of 3' UTR-derived sRNAs, underscoring the important role of ManNAc in Salmonella adaptation within host environments.

Keywords: 3’ UTR-derived sRNAs; N-Acetylmannosamine; RNase III; Salmonella enterica; sialic acid metabolism.

Fig. 1.

STM1127-STM1131 is an N-acetylmannosamine utilization cluster. (A) Genomic context schematic of the STM1127-STM1131 genetic cluster in S. enterica. Transcription start sites (TSS) identified by dRNA-seq are indicated with arrows. The positions of the stop codon of STM1127 (Left) and STM1131 (Right) as well as the start and stop nucleotides of STnc500 in the S. Typhimurium LT2 genome (NC_003197.2) are marked in brackets. Predicted structures of five ORFs are displayed, generated using AlphaFold, with their functions determined by the HHpred server: STM1127, MurR family transcriptional regulator (TF); STM1128, Neu5Ac sodium-solute symporter (SSS); STM1129, ManNAc-6P epimerase; STM1130, Neu5Ac epimerase; STM1131, Outer membrane protein (OMP). (B) RNAfold-predicted secondary structure of STnc500. The numbers indicate the distance from the 5’ end. (C) Expression analysis of STM1128 and nanA in S. Typhimurium SL1344 wild-type cells grown in M9CA medium supplemented with various carbon sources (0.1% w/v). Total RNA samples were collected at OD₆₀₀ ~ 1.0 and analyzed by qRT-PCR, with rfaH used as the reference gene for data normalization. Abbreviations: Neu5Ac, N-acetylneuraminic acid; ManNAc, N-acetylmannosamine; GlcNAc-6P, N-acetylglucosamine-6-phosphate; GlcNAc, N-acetylglucosamine; UDP-GlcNAc, uridine diphosphate N-acetylglucosamine; Glc-6P, glucosamine-6-phosphate; Pyruvic, pyruvic acid; Fru, fructose; Man, mannose; Gly, glycerol; Glu, glucose. Error bars indicate SD (n = 3). ***P < 0.001 when comparing the ManNAc group with any other group, except Neu5Ac, in a two-tailed t test. (D) qRT-PCR analysis of STM1128 and nanA expression in S. Typhimurium SL1344 wild-type and indicated repressor-deleted strains grown in LB medium (Lennox). Total RNA samples were collected at OD₆₀₀ ~ 1.0. Error bars indicate SD (n = 3). ***P < 0.001 and ns, no significant difference in a two-tailed t test. (E) β-galactosidase activities of chromosomally encoded lacZ-transcriptional fusions to the STM1129-STM1128 promoter. Error bars indicate SD (n = 3). ***P < 0.001 in a one-way ANOVA test. (F) Growth curves of S. Typhimurium SL1344 wild-type, ΔnanKEAT-yhcH (Δnan), ΔSTM1127-STM1131 and double clusters deleted Δnan-ΔSTM1127-STM1131 strains in M9CA medium supplemented with indicated sugars. Shaded color represents SD of six replicates.

Fig. 2.

ManS is an RNase-processed sRNA with multiple length variants. (A) Northern blot analysis of ManS expression in S. Typhimurium SL1344 wild-type cells grown in M9CA medium supplemented with various carbon sources. Total RNA samples were probed for ManS, with 5S rRNA used as a loading control. Sugar abbreviations are consistent with those in Fig. 1. (B) Time-course analysis of ManS expression in response to ManNAc. S. Typhimurium SL1344 wild-type cells were initially grown in M9CA medium supplemented with 0.1% glucose to OD₆₀₀~1.0, then shifted to M9CA medium supplemented with 0.1% ManNAc. Total RNA samples were collected at indicated time points and analyzed for ManS expression. (C) Comparative analysis of ManS expression in Salmonella strains with deletions of indicated genes, grown in M9CA medium supplemented with 0.1% ManNAc.

Fig. 3.

RNase III is the primary RNase responsible for processing ManS. (A) Analysis of ManS expression in Salmonella strains with deletions of the indicated RNases, cultured in M9CA medium supplemented with 0.1% ManNAc. rne^TS and rne^Ctr refer to RNase E temperature-sensitive and control strains, respectively. The RNases encoded by the deleted genes are RNase I (rna), RNase II (rnb), RNase III (rnc), RNase G (rng), RNase H (rnhA and rnhB), RNase BN (rbn), Poly(A) Polymerase I (pcnB), PNPase (pnp), and RNase R (rnr). Representative of four independent experiments. (B) Predicted secondary structure of the 81 nt ManS (black) with its upstream region (blue). Red nucleotides represent 5’ ends detected by RACE in vivo, while blue and green triangles denote 5’ ends detected by in vitro RNase III cleavage assay, corresponding to cleavage sites shown in panel D. The numbers indicate the distance from the 5’ end of the 81 nt ManS. Proposed RNase III and RNase E cleavage sites are shown with black and red arrows, respectively, and product lengths are indicated in brackets. The gray line marks the 5’ end region of the 65 to 70 nt ManS isoforms processed by an unidentified RNase. (C) Biogenesis of ManS from various RNA precursors. A ManS-deleted strain was transformed with a pBAD plasmid carrying either the wild-type 3’ UTR of STM1128 or a version with the bulge deleted. Expression was induced with 0.2% L-arabinose for the indicated durations. Total RNA isolated from wild-type cells (ManS ^native) was included on the same blot for size comparison. (D) RNase III cleavage assay using in vitro-transcribed ManS with wild-type or bulge-deleted preceding sequences. RNAs were incubated with E. coli RNase III for the indicated times, collected, and resolved on a 12% polyacrylamide gel. The ManS^native lane was intentionally overexposed for size comparison; colored triangles indicate cleavage sites as depicted in panel B, with product sizes verified by RACE as indicated.

Fig. 4.

ManS acts posttranscriptionally to repress target gene expression. (A) Localization of base-pairing regions R1 and R2 within ManS, predicted by the IntaRNA program, with identified targets from RNA-sequencing data. Nucleotides mutated for fluorescence assays are highlighted in red. (B–D) Fluorescence measurements in the ΔmanS strain carrying either the control plasmid pXG-1 (B) or pXG10-sfGFP fused in-frame with wild-type R1 (C) or R2 (D) target sequences, or compensatory mutant versions (Mu). These were coexpressed with pZE12-based plasmids encoding either an empty vector (pJV300), wild-type (WT) ManS, or the indicated ManS mutants. Predicted RNA duplex formations between selected target mRNAs and ManS are shown below. Distances from the +1 site of the 81 nt ManS, or to the start codons (underlined) of the target mRNA sequences are indicated. Blue nucleotides represent annotated ribosome-binding sites on the mRNA. Red lines denote the positions of tested mutations. Error bars represent SD (n = 3). Statistical significance was assessed using a two-tailed t test. *P < 0.05, **P < 0.01, ***P < 0.001, with “ns” indicating no significant difference.

Fig. 5.

Selective ManS-mediated regulation in different conditions. (A–C) Western blot (A) and northern blot (B) analyses of FLAG-tagged target proteins and ManS in wild-type (WT + pCtr) and ManS-deleted strains with the indicated plasmids. “ManS M” in panel A represents ManS mutants: R1-M for DppA and LdhA, and R2-M1 for NagE. Samples were collected at OD₆₀₀ 0.4 to 0.5 during growth in M9CA medium with 0.1% ManNAc. RpoB and 5S rRNA were used as loading controls for protein and RNA analyses, respectively. Northern blot images were overexposed to make ManS visible in the WT+pJV300 strain. Protein levels were quantified using ImageJ, normalized to RpoB levels, and expressed relative to WT + pJV300 (set to 1). Data from three biological replicates are shown in panel C. (D–E) Expression levels of NagE::3xFLAG, DppA::3xFLAG, and ManS in wild-type and ManS-deleted strains during growth in ManNAc or Neu5Ac medium. Relative protein levels compared to the WT strain are shown in (E), calculated from three biological replicates. Average fold-changes calculated from three biological replicates are indicated above each bar. Error bars represent SD (n = 3). Statistical significance was calculated using a two-tailed t test. *P < 0.05, **P < 0.01, ***P < 0.001, “ns” indicates no significant difference.

Fig. 6.

Global effect of ManS-mediated regulations during ManNAc metabolism and infection. (A) Scatter plot displaying log2 fold changes (ΔmanS /WT) of genes with significant differences (P < 0.05, n = 3) in both RNA sequencing (x-axis) and mass spectrometry (y-axis) data. Genes showing significant changes in both RNA and protein levels are highlighted in red. Genes with significant changes at the RNA (yellow) or protein (cyan) level are shown in respective colors. (B) Metabolite abundance of amino-sugars from the sialic acid metabolic pathway (Left) and proline-related dipeptides (Right) with significant changes between wild-type and ΔmanS strains. Error bars represent SD (n = 5). Statistical significance is indicated as *P < 0.05, ***P < 0.001, using a two-tailed t test. (C) Competitive index (CI) values comparing S. Typhimurium SL1344 wild-type, ΔmanS, and complemented (ΔmanS::P_STM1129-manS) strains in the cecum and ileum of C57BL/6 mice at day 4 postinoculation (n = 15). Two inoculum groups with opposite antibiotic resistance markers were tested: Green dots indicate the wild-type background strain marked with kanamycin versus the competed strain with chloramphenicol; blue dots indicate the reverse. The CI was calculated as the ratio of recovered CFU of WT/competed strain. The dashed line represents CI = 1. Horizontal bars represent median CI values with 25 to 75% ranges. Statistical significance was determined using the Wilcoxon matched-pair test. **P < 0.01, ***P < 0.001, “ns” indicates no significant difference.

Fig. 7.

Model depicting the function of ManS in balancing ManNAc metabolism. The enrichment of ManNAc generated by either Neu5Ac degradation intracellularly or transportation from the extracellular environment activates the STM1129-STM1128 operon at the transcriptional level by releasing the repressor STM1127. ManNAc is then converted to GlcNAc-6P by STM1129 and other enzymes, which triggers the transcriptional activation of GlcNAc PTS permease NagE by releasing the repressor NagC. To avoid the accumulation of excess GlcNAc-6P, ManS is generated through noncanonical RNase III processing, repressing the translation of nagE mRNA via its seed regions in the loop of the Rho-independent terminator (R2) to balance the metabolism of ManNAc and its isomer GlcNAc. Additionally, the dipeptide ABC transport system DppABCDF and other metabolic genes are repressed by the long isoform to ensure homeostasis of related peptides.