The Small RNA Teg41 Is a Pleiotropic Regulator of Virulence in Staphylococcus aureus

Abstract

Previously, our group demonstrated a role for the small RNA (sRNA) Teg41 in regulating production of the alpha phenol-soluble modulin toxins (αPSMs) in Staphylococcus aureus. Overexpressing Teg41 increased αPSM production while deleting the 3' end of Teg41 (Teg41Δ3' strain) resulted in a decrease in αPSM production, reduced hemolytic activity of S. aureus culture supernatants, and attenuated virulence in a murine abscess model of infection. In this study, we further explore the attenuation of virulence in the Teg41Δ3' strain. Using both localized and systemic models of infection, we demonstrate that the Teg41Δ3' strain is more severely attenuated than an ΔαPSM mutant, strongly suggesting that Teg41 influences more than the αPSMs. Proteomic and transcriptomic analysis of the wild-type and Teg41Δ3' strains reveals widespread alterations in transcript abundance and protein production in the absence of Teg41, confirming that Teg41 has pleiotropic effects in the cell. We go on to investigate the molecular mechanism underlying Teg41-mediated gene regulation. Surprisingly, results demonstrate that certain Teg41 target genes, including the αPSMs and βPSMs, are transcriptionally altered in the Teg41Δ3' strain, while other targets, specifically spa (encoding surface protein A), are regulated at the level of transcript stability. Collectively, these data demonstrate that Teg41 is a pleiotropic RNA regulator in S. aureus that influences expression of a variety of genes using multiple different mechanisms.

Keywords: AgrA; MgrA; PSM; Staphylococcus aureus; Teg41; protein A; regulation; virulence.

FIG 1

The Teg41Δ3′ strain is more attenuated in vivo than an ΔαPSM strain. (A and B) Female C57BL/6J mice were infected with ~3.6 × 10⁷ CFU of the indicated strains, and infection was allowed to proceed for 9 days. (A) Surface lesion size was quantified over 9 days. Data represent the mean (n = 13 to 23) with standard error of the mean (SEM). Statistical significance was determined by Mann-Whitney test; **** indicates P values of <0.0001 for the Teg41Δ3′ strain versus either WT or ΔαPSM mutant. Data are combined from multiple independent experiments. (B) At 0.3 days (8 h), 5 days, and 9 days postinfection (p.i.), mice were euthanized, and tissue was homogenized to determine bacterial titers. Each symbol represents a single animal (n = 10 to 15), and the line represents the median. Data are combined from multiple independent experiments. Statistical significance was determined by Mann-Whitney test (ns, not significant; *, P < 0.05; ***, P < 0.005; ****, P < 0.001). (C to E) Groups of 16 mice were retro-orbitally injected with the indicated strains. After 7 days, mice were euthanized, and organs were harvested. Bacteria were plated, and CFU were normalized to organ mass. Bacterial loads in kidneys (C), lungs (D), and liver (E) are displayed. Results are shown as median with 95% confidence interval (CI). Statistical tests were performed as follows: for panel A, two-way analysis of variance (ANOVA) with Šídák's multiple-comparison test (****, adjusted P [P_adj] < 0.0001; ***, P_adj < 0.001; **, P_adj < 0.01; *, P_adj < 0.05); for panels B to E, Kruskal-Wallis test with Dunn’s multiple-comparison test (****, P_adj < 0.0001; ***, P_adj < 0.001; **, P_adj < 0.01; *, P_adj < 0.05).

FIG 2

Comparison of the transcriptomic profiles of an ΔαPSM mutant with the Teg41Δ3′ strain. Both strains were inoculated in triplicate in TSB and grown for 6 h at 37°C. Cells were harvested, and RNAs were extracted and sequenced. (A) Principal-component analysis (PCA) comparing transcriptomes of the Teg41Δ3′ and ΔαPSM strains. The x axis displays principal component 1, explaining 43.7% of variance, and the y axis shows principal component 2, explaining 19.7% of variance. (B) Volcano plot of differential expression analysis (DEA) between ΔαPSM and Teg41Δ3′ strains. Red dots depict significant changes with a log₂ fold change of >1 and a −log₁₀ P value of >1.3 (P < 0.05) as cutoffs. Black dots represent transcripts not differentially impacted. The 5 most highly up- and downregulated genes are indicated.

FIG 3

Removal of Teg41 causes widespread gene expression changes in S. aureus. Strains were inoculated in triplicate in TSB and grown for 3 h, 6 h, and 9 h at 37°C. Cells were harvested, and RNA was extracted and sequenced. (A) Cluster of Orthologous Group (COG) analysis comparing dysregulated genes between Teg41Δ3′ strain and WT at 3 h (blue), 6 h (red), and 9 h (yellow). The number of genes in each category is represented by concentric circles moving outward from the center (each circle representing 10 genes). (B to D) Volcano plot of DEA between Teg41Δ3′ strain and WT at 3 h (B), 6 h (C), and 9 h (D). Red dots depict significant changes with a log₂ fold change of >1 and a −log₁₀ P value of >1.3 (P < 0.05) as cutoffs. Black dots represent transcripts not differentially impacted. The 5 most highly up- and downregulated genes are indicated. (E) Comparison of genes dysregulated in the Teg41Δ3′ strain at each growth phase. (F) List of the 7 genes that were dysregulated in all three RNA-seq data sets, including linear fold change values obtained in each experiment (i.e., 3 h, 6 h, and 9 h).

FIG 4

A small subset of protein targets can be explained by transcriptional changes in the Teg41Δ3′ strain. RNA-seq and proteomic data from the WT and Teg41Δ3′ strain grown for 6 h in TSB at 37°C were compared, and a list of common impacted targets upon Teg41 removal was made. Proteins found in both cytosolic and secreted fractions were combined and treated as one protein. FCs for secreted fractions were indicated for proteomic values. RGD, Arg-Gly-Asp.

FIG 5

Teg41 regulates expression of the αPSMs and βPSMs at the transcriptional level. (A and B) Teg41 influences promoter activity of the αPSMs (A) and βPSMs (B). Bacterial strains were grown in TSB for 6 h at 37°C, and cells were harvested and mechanically lysed. β-Galactosidase assays were performed, values were normalized to total protein concentration, and modified Miller units were calculated. Results are shown as mean ± standard deviation (SD). Significances were computed with one-way ANOVA corrected for multiple comparisons using Tukey’s test. ***, P_adj < 0.001; **, P_adj < 0.01; *, P_adj < 0.05. (C) AgrA and MgrA are required for Teg41-induced hemolytic activity. Bacterial strains were grown overnight in TSB at 37°C. Cell-free supernatants were incubated with human blood, and hemolysis was measured at OD₅₄₃. Results are shown as mean ± SD with one-way ANOVA corrected for multiple comparisons using Tukey’s test. ****, P_adj < 0.0001; ***, P_adj < 0.001. (D) MgrA is a transcriptional activator of the αPSM promoter. Bacterial strains were grown in TSB for 8 h at 37°C, and at each hour, cells were harvested and mechanically lysed. β-Galactosidase activity was recorded as stated for panels A and B. Results are shown as mean ± SD. Significances were computed with one-way ANOVA corrected for multiple comparisons using Tukey’s test at each hour. ****, P_adj < 0.0001; ***, P_adj < 0.001; **, P_adj < 0.01; *, P_adj < 0.05. Significance for Teg41Δ3′ versus WT is shown in red, for agrA versus WT in green, and for mgrA versus WT in purple.

FIG 6

Teg41 represses protein A production posttranscriptionally. (A) Teg41 does not influence protein A promoter activity. Bacteria were grown for 3 h, and the β-galactosidase activity was assessed. Results are shown as mean ± SD. (B) spa transcript levels and stability increase upon Teg41 removal. Bacteria were grown for 3 h in TSB prior to rifampicin treatment. Cells were pelleted at T₀ (before rifampicin addition), T_2.5, T₅, and T₁₀ (i.e., 2.5, 5, and 10 min after rifampicin treatment). RNA was extracted at each time point, and spa mRNA levels were monitored by Northern blotting. hup levels were also detected as a loading control. (C) Quantification of spa levels from Northern blots by densitometry. Times indicated represent the half-life of spa mRNA in each strain. Rif, rifampicin; AU, absorbance units. (D) Predicted interaction region between Teg41 (red) and spa (blue). The start codon of spa is underlined, and the predicted RBS sequence is in italic. Prediction was performed using IntaRNA (31). (E) spa transcript levels were detected by Northern blotting in WT S. aureus containing the pCN51 empty vector (1), the Teg41Δ3′ strain containing pCN51 (2), the Teg41Δ3′ strain overexpressing full-length Teg41 (3), and the Teg41Δ3′ strain overexpressing the 3′ end of Teg41 (4). For panels B and E the ladders shown are RNA molecular weight markers. Numbers shown indicate size of marker in nucleotides.