Genome-wide mutagenesis identifies factors involved in MRSA vaginal colonization

Abstract

Methicillin-resistant Staphylococcus aureus (MRSA) is an opportunistic pathogen that colonizes various body sites, including the nares, skin, and vagina. During pregnancy,colonization can lead to dysbiosis, adverse pregnancy outcomes, and invasive disease. To identify genes contributing to MRSA vaginal fitness, we performed transposon sequencing (Tn-seq) using a murine model of vaginal colonization, identifying over 250 conditionally essential genes. Five genes were validated in our murine model, including those encoding the aerobic respiration protein QoxB, bacillithiol biosynthesis component BshB2, sialic acid catabolism enzyme NanE, and staphylococcal regulator of respiration SrrAB. RNA sequencing and comparative analysis identified over 30 SrrAB-regulated genes potentially important for fitness in vaginal-like conditions, particularly under oxygen stress. These findings highlight pathways such as aerobic respiration, bacillithiol biosynthesis, sialic acid catabolism, and transcriptional regulation that support MRSA's competitive fitness in the vaginal tract.

Keywords: CP: Microbiology; MRSA; RNA-seq; SrrAB regulator; Staphylococcus aureus; Tn-seq; bacillithiol biosynthesis; host-pathogen interactions; murine vaginal model; sialic acid catabolism; transposon sequencing; vaginal colonization.

Figure 1.. Tn-seq with a MRSA Tn library assesses genetic determinants of fitness in the murine female genital tract

(A) Schematic of MRSA in vivo Tn-seq in the murine female reproductive tract. (B) Number of unique TA insertion sites covered for each condition. (C) Average number of reads mapped to each Tn insertion site. (D) CIRCOS plot of insertion counts/gene across each sample aligned to kilobases of genome (gray ring) with an insertion number cutoff value of 100 reads based on the average number of reads per insertion site for the input sample (C). (B and C) Statistics performed by Kruskal-Wallis test with multiple comparisons, with p < 0.05 indicating statistical significance. See also Figure S1.

Figure 2.. Analysis of genes with conditional essentiality in the murine vaginal tract

(A–D) Volcano plots of log2 fold change (log2FC) vs. −log10 q value for (A) input vs. day 1 vaginal lavage, (B) input vs. day 3 vaginal lavage, (C) input vs. vaginal tissue, and (D) input vs. cervical tissue. Data points under the lower dotted line have a q > 0.05 (gray) and are not statistically significant. Values with negative log2FC above the q value cutoff are significantly underrepresented (conditionally essential), and values with positive log2FC above the q value cutoff are significantly overrepresented. Values at the second dotted line are shown at y = 3 because TRANSIT software does not calculate q values of %0.001. (E) Venn diagram of underrepresented genes at day 3. (F) Clusters of orthologous groups (COG) determined by eggNOG 5.0 for genes deemed conditionally essential. (G) Heatmap of fold changes for selected significantly underrepresented genes grouped by challenge condition (columns). Genes are ordered according to log2FC within functional groups. Generalized functional groups are labeled to the left, and groups include transcriptional regulators associated with each function. Asterisks indicate genes selected for further validation. See also Figure S2–S3 and Tables S1 and S2.

Figure 3.. Validating reduced fitness of qoxB, bshB2, and nanE mutants when challenged against WT in the murine vaginal tract

(A, C, and E) CFU/lavage of (A) WT and qoxB::Tn (n = 21), (C) WT tet-r and bshB2::Tn (n = 10), and (E) WT MRSA tet-r and DnanE (n = 20) lavaged daily for 7–8 days and plated on differential media. Statistics were calculated by two-way ANOVA with Geisser-Greenhouse correction and multiple comparisons of each cell mean with the other cell means in that row. (B, D, and F) Competitive index of (B) WT and qoxB::Tn, (D) WT tet-r and bshB2::Tn, and (F) WT MRSA tet-r and DnanE. Statistics were calculated by one-sample t test compared to a theoretical value of 1 (indicative of equal competition). (A–F) All data shown are a combination of two independent experimental replicates. The limit of detection (LOD) for bacterial CFUs in these experiments is 100; therefore, any strain under the LOD in a lavage is marked as 99. Any mouse with a competitive index (CI) of 99 /99 (both mutant and WT under the LOD) on a given day was excluded from the CI. See also Figure S4.

Figure 4.. A mutant lacking the SrrAB two-component system is less able to colonize and persist in the murine vaginal tract

(A) CFU/vaginal lavage of WT tet-r and ΔsrrAB challenged at 1:1 for a total of 10⁷ CFU, lavaged daily for 7 days, and plated on differential medium (n = 19). (B) Competitive index of WT vs. ΔsrrAB (mutant/WT) in colonized mice shows that WT outcompetes ΔsrrAB. (C) CFU/vaginal lavage of WT and ΔsrrAB LL29::srrAB challenged at 1:1 for a total of 10⁷ CFU, lavaged daily for 7 days, and plated on differential medium (n = 19). (D) Competitive index of WT vs. ΔsrrAB LL29::srrAB in colonized mice shows that WT does not outcompete the ΔsrrAB LL29::srrAB complement strain. (E and F) CFU/vaginal lavage of WT and ΔsrrAB challenged at 10⁷ CFU, lavaged daily for 12 days, and plated on differential medium (n = 16/strain) represented as (E) individual mice and (F) mean and SD with the 2-day period of mutant clearance highlighted in blue. (A, C, and E) Statistics were calculated by two-way ANOVA with column p value (strain) reported and (B and D) one-sample t test compared to a theoretical value of 1 (indicative of equal competition). (A–E) As with Figure 3, all data shown are a combination of two independent experimental replicates. Limit of detection (LOD) for bacterial CFUs is 100—any strain under the LOD in a lavage is marked as 99. Any mouse with both mutant and WT under the LOD on a given day was excluded from the competitive index (CI). See also Figure S4.

Figure 5.. In vitro WT vs. ΔsrrAB RNA-seq in differing growth phases and oxygen levels reveals SrrAB-regulated factors in vaginal-like atmospheric conditions

(A–C) Volcano plots of log2 fold change (log2FC) vs. −log10 q value for all transcripts with adj. p < 0.05 for (A) WT vs. ΔsrrAB during mid-log aerobic growth in TSB (OD₆₀₀ = 0.5), (B) WT vs. ΔsrrAB during mid-log reduced oxygen growth in TSB (OD₆₀₀ = 0.5), and (C) WT vs. ΔsrrAB during late-log aerobic growth in TSB (OD₆₀₀ = 1.5). Dots in gray do not make the significance cutoff of −1 > log2FC > 1. Upregulation in WT vs. ΔsrrAB indicates positive regulation by SrrAB, while downregulation in WT vs. ΔsrrAB indicates negative regulation by SrrAB. Labeled genes of interest are highlighted in yellow. (D and E) Venn diagrams (D) and clusters of orthologous groups (COG) determined by eggNOG 5.0 (E) for significantly up- and downregulated genes in each condition. See also Figure S5 and Tables S3 and S4.

Figure 6.. Cross-comparison identifies SrrAB-regulated loci of interest for vaginal colonization fitness

(A) Heatmap showing log2FC of all loci both significantly underrepresented in the Tn-seq and significantly upregulated in the RNA-seq in at least one condition. The genes above the horizontal line either are known to be SrrAB regulated (hemX, pflA, and qoxABC) or were identified as significantly upregulated (or members of a significantly upregulated operon) in a previously published murine vaginal RNA-seq (see Figure S6). Bolded genes were selected based on their comparative expression profiles (and/or operon position if applicable) for qPCR validation. (B–D) qPCR was performed on genes of interest in both the WT and the ΔsrrAB strains during (B) mid-log aerobic growth in TSB (OD₆₀₀ = 0.5), (C) mid-log reduced-oxygen growth in TSB (OD₆₀₀ = 0.5), and (D) late-log aerobic growth in TSB (OD₆₀₀ = 1.5). All quantitative expression values are shown as the mean values ±SD of mutant over WT transcription levels. Each datapoint is a biological replicate representing the average of two technical replicates. A one-sample t test was used to compare each gene’s expression to a hypothetical value of 1 (indicative of equal gene expression in ΔsrrAB and WT). See also Figure S7.

Figure 7.. Overview of genetic factors contributing to MRSA murine vaginal colonization investigated in this study

The four MRSA genetic factors confirmed to have fitness defects in the murine vaginal tract are shown (bolded, red text) in the context of their primary functional roles. QoxB is a known component of the respiratory chain and was shown to be important for competitive and single-challenge fitness in the murine vaginal tract. The SrrAB two-component system is known to sense redox stress within the membrane (SrrB) and respond by regulating transcription (SrrA). This regulon, which includes qoxB, and >30 members that may be important in the vaginal tract are described in Figures 5 and 6 to have differential expression in the srrAB mutant and/or differential expression in murine vaginal lavage. NanE, the third protein in the sialic acid catabolism pathway, and BshB2, which contributes to biosynthesis and recycling of the protective thiol bacillithiol (BSH), were both found to contribute to competitive fitness in the murine vaginal lumen. Created in https://www.biorender.com.