Staphylococcus aureus Shifts toward Commensalism in Response to Corynebacterium Species

Abstract

Staphylococcus aureus-human interactions result in a continuum of outcomes from commensalism to pathogenesis. S. aureus is a clinically important pathogen that asymptomatically colonizes ~25% of humans as a member of the nostril and skin microbiota, where it resides with other bacteria including commensal Corynebacterium species. Commensal Corynebacterium spp. are also positively correlated with S. aureus in chronic polymicrobial diabetic foot infections, distinct from acute monomicrobial S. aureus infections. Recent work by our lab and others indicates that microbe-microbe interactions between S. aureus and human skin/nasal commensals, including Corynebacterium species, affect S. aureus behavior and fitness. Thus, we hypothesized that S. aureus interactions with Corynebacterium spp. diminish S. aureus virulence. We tested this by assaying for changes in S. aureus gene expression during in vitro mono- versus coculture with Corynebacterium striatum, a common skin and nasal commensal. We observed a broad shift in S. aureus gene transcription during in vitro growth with C. striatum, including increased transcription of genes known to exhibit increased expression during human nasal colonization and decreased transcription of virulence genes. S. aureus uses several regulatory pathways to transition between commensal and pathogenic states. One of these, the quorum signal accessory gene regulator (agr) system, was strongly inhibited in response to Corynebacterium spp. Phenotypically, S. aureus exposed to C. striatum exhibited increased adhesion to epithelial cells, reflecting a commensal state, and decreased hemolysin activity, reflecting an attenuation of virulence. Consistent with this, S. aureus displayed diminished fitness in experimental in vivo coinfection with C. striatum when compared to monoinfection. These data support a model in which S. aureus shifts from virulence toward a commensal state when exposed to commensal Corynebacterium species.

Keywords: Corynebacterium; Staphylococcus aureus; agr system; commensal bacteria; microbiome; quorum sensing (QS).

FIGURE 1

Staphylococcus aureus agr quorum sensing decreases in response to Corynebacterium spp. Using a luminescent agr-I activity assay, we tested the ability of CFCM from the indicated species to inhibit activation of agrP3 after exposure to exogenous AIP-1. Dark gray bars indicate AIP-1 negative and positive controls. White bars indicate CFCM controls from S. aureus JE2 containing a Tn insertion in agrB and the plasmid-free derivative of the E. coli strain used to exogenously produce AIP-1. Light gray bars indicate CFCM from 48 h cultures of C. striatum, C. amycolatum, C. glutamicum, C. accolens, and C. pseudodiphtheriticum, respectively. Medium gray bars indicate C. striatum (Cst) CFCM treatments, ‘<3 kD’ indicates sub 3 kilodalton fractions, ‘HT’ is CFCM heat treated at 95°C for 15 m, ‘PK’ indicates Proteinase K digested CFCM. For each bar, n = 3 and error bars represent SEM. For each result between the indicated samples (horizontal line), *p < 0.05 by two-tailed Student’s t-test with Bonferroni correction for multiple testing.

FIGURE 2

Staphylococcus aureus attachment to human epithelial cells increases in response to C. striatum CFCM. When S. aureus WT (gray bars) and AgrA^- (white bar) cells were exposed to AIP-1 plus or minus C. striatum CFCM (Cst), we observed 4.5- and 3.6-fold increases in attachment respectively. These increases were statistically significant (*) when compared to the WT alone. S. aureus attachment to human A549 epithelial cells was quantified after 1 h of exposure. Attachment was measured as the percentage of attached cells divided by the total number of S. aureus planktonic cells added, as determined by CFU enumeration. Fold change was determined by dividing the percent attached for each condition by that of the % attached for WT exposed only to AIP. Error bars were omitted from the WT normalized data for clarity. Data were analyzed by two-tailed Student’s t-test with Bonferroni correction for multiple testing (*p < 0.03). Error bars represent SEM.

FIGURE 3

The proportion of S. aureus cells with active surface-associated SpA increases in coculture with C. striatum. S. aureus strains were grown in mono- versus coculture with C. striatum under conditions identical to those used in Table 1. SpA activity was detected by observing SpA-mediated capture of FITC-conjugated goat IgG (yellow in B,C). Cells were counterstained with DAPI (blue in B,C). Cell intensity was quantified for all S. aureus strains in three fields of view at 1000x magnification each from 3 biological replicates under identical exposure settings. (A) ImageJ software was used to count total cells versus FITC-positive cells. 10% of wild-type S. aureus (WT) were positive for SpA activity in monoculture compared to 60% in coculture with C. striatum (Cst). The spa-deletion mutant carrying the empty expression vector (Δspa) had no detectable FITC staining (ND). The spa-deletion mutant carrying an expression vector expressing spa from its native promoter (spa+) had FITC-positive cells in both conditions consistent with this construct producing SpA above wt levels. Error bars represent SEM. Data were analyzed by two-tailed Student’s t-test with Bonferroni correction for multiple testing (*p < 0.003). Representative micrographs of wild-type S. aureus cells (blue) in mono- (B) versus coculture (C) stained for SpA (yellow) at the cell surface. Scale bars represent 1.5 μm.

FIGURE 4

Staphylococcus aureus exhibits decreased hemolytic activity when grown with C. striatum CFCM. Hemolysis of rabbit erythrocytes was quantified 30 m after exposure to BHI, as the negative control (NC), or CFCM from S. aureus strains grown in the presence of AIP-1 alone or plus the addition of C. striatum CFCM (Cst) for the wild-type (WT) and agrA::Tn mutant (AgrA^-), n = 3 each. Decreased OD₆₃₀ is indicative of S. aureus hemolytic activity. Growing S. aureus in the presence of C. striatum CFCM significantly diminished the hemolytic activity produced by the WT. In contrast, the agrA::Tn mutant (AgrA^-) was incapable of significant hemolysis in either condition. Data were analyzed by two-tailed Student’s t-test with Bonferroni correction for multiple testing (*p < 0.005, **p < 0.00005). Error bars represent SEM.

FIGURE 5

Staphylococcus aureus abundance decreases in vivo when coinfected with C. striatum in a murine abscess model. (A) In a murine abscess infection model 4 days post-infection, wild-type S. aureus showed reduced numbers (CFU/g) during coinfection with C. striatum (light orange bar; Sa + Cst) compared to monoinfection (orange bar; Sa alone). (B) In the same model, C. striatum numbers increased significantly when coinfected with S. aureus (light blue bar; Cst + Sa) when compared to monoinfection (blue bar; Cst alone). For each bar, n = 9. Data were analyzed using the Mann Whitney U-test (*p < 0.03, **p < 0.02). Error bars represent SEM.

FIGURE 6

Corynebacterium striatum CFCM inhibits signaling of S. aureus agr types I, II, and III. For each respective agr type, post-exponential phase S. aureus CFCM from that same type was used to induce an agrP3:gfp reporter. CFCM containing AIP-1 was used to inhibit S. aureus agr types II and III strains and CFCM containing AIP-2 was used to inhibit S. aureus agr types I and IV strains (white bars). Cultures grown in the presence of inducing AIP-containing S. aureus CFCM and C. striatum CFCM showed pronounced inhibition of agr types I, II, and III and no significant inhibition of agr type IV (light gray bars). GFP was measured as fluorescence/OD₆₃₀ (n = 3). Error bars represent SEM. Data were analyzed by two-tailed Student’s t-test with Bonferroni correction for multiple testing (*p < 0.005 and **p < 0.01).

FIGURE 7

Staphylococcus aureus gene expression during in vivo colonization is similar to in vitro coculture with C. striatum. Gene expression data from S. aureus (orange spheres) in vivo colonization of humans (Burian et al., 2010b; Krismer et al., 2014), cotton rats (Burian et al., 2010a) or in vitro coculture with C. striatum (blue ovals) (Table 1/Supplementary Table S1) is depicted as upregulated (blue arrows) or downregulated (red arrows) in comparison to in vitro monoculture. Genes quantified by qRT-PCR in vivo share similar patterns of expression with those detected with RNAseq during in vitro coculture with C. striatum (“shared between conditions”). *In several cases, individual genes identified by qRT-PCR were part of operons whose members were also differentially regulated in our RNASeq results (Supplementary Table S1). †Type 5 capsule (CP5) production was quantified by ELISA in a mouse nasal colonization model (Kiser et al., 1999) and was overrepresented in vivo versus in vitro. We observed upregulation of several CP5 synthesis genes in coculture with C. striatum (Supplementary Table S1). It is unknown (“?”) whether or not Corynebacterium spp. were present in the referenced in vivo experiments. These data demonstrate the similarities in S. aureus gene expression during host commensal in vivo colonization and in vitro growth with Corynebacterium spp.