Coagulase-Negative Staphylococcal Strain Prevents Staphylococcus aureus Colonization and Skin Infection by Blocking Quorum Sensing

Summary

Coagulase-negative staphylococci (CoNS) and Staphylococcus aureus are part of the natural flora of humans and other mammals. We found that spent media from the CoNS species Staphylococcus caprae can inhibit agr-mediated quorum sensing by all classes of S. aureus. A biochemical assessment of the inhibitory activity suggested that the S. caprae autoinducing peptide (AIP) was responsible, and mass spectrometric analysis identified the S. caprae AIP as an eight-residue peptide (YSTCSYYF). Using a murine model of intradermal MRSA infection, the therapeutic efficacy of synthetic S. caprae AIP was evident by a dramatic reduction in both dermonecrotic injury and cutaneous bacterial burden relative to controls. Competition experiments between S. caprae and MRSA demonstrated a significant reduction in MRSA burden using murine models of both skin colonization and intradermal infection. Our findings indicate that important interactions occur between commensals that can impact disease outcomes and potentially shape the composition of the natural flora.

eTOC blurb

Paharik, Parlet, et al. demonstrate that the human commensal Staphylococcus caprae competes with Staphylococcus aureus by inhibiting quorum-sensing. Through signal interference, S. caprae reduces Methicillin-resistant S. aureus burden in both skin colonization and infection, highlighting the benefits of healthy skin flora and suggesting a new avenue for probiotic therapy.

INTRODUCTION

Coagulase-negative staphylococci (CoNS) are a heterogeneous group of nearly 40 species that compose the majority of the Staphylococcus genus (Becker et al., 2014). CoNS are generally non-pathogenic commensals of humans and other animals, and have received less attention in the literature than the more virulent Staphylococcus aureus. However, recent studies have described a number of interactions between CoNS and S. aureus that share similar host niches (Iwase et al., 2010; Janek et al., 2016; Nakatsuji et al., 2017; Sugimoto et al., 2013; Zipperer et al., 2016), demonstrating that these species can compete with S. aureus during colonization and potentially alter its pathogenic behaviour. CoNS - S. aureus interactions are only beginning to be appreciated, and are important for a complete understanding of polymicrobial colonization dynamics. Investigating these interactions could shed light on the influence of the host microbiota on the ability of S. aureus to colonize different environments and cause disease.

S. aureus is one of the most problematic and common causes of bacterial infections (Dantes et al., 2013; Lowy, 1998; Tong et al., 2015). In particular, S. aureus is responsible for 76% of all skin and soft tissue infections (Moran et al., 2006), leading to 500,000 hospital visits and 10 million outpatient visits per year (Hersh et al., 2008). Infections caused by methicillin-resistant S. aureus (MRSA) are more difficult to treat and costly for healthcare systems (Filice et al., 2010), and the level of MRSA infections has remained high, with over 80,000 invasive infections occurring each year (Dantes et al., 2013). Community-associated MRSA (CA-MRSA) of the USA300 group have emerged as the most common isolates from skin and invasive infections (Chambers and Deleo, 2009; King et al., 2006) and are a major societal economic burden (Lee et al., 2013). In the US alone, antimicrobial-resistant pathogens cause over 2 million infections per year, resulting in over 23,000 annual mortalities and leading to over $20 billion in excess medical costs (Marston et al., 2016), highlighting the need to develop innovative approaches for treatment.

Due to the major healthcare impact of S. aureus infections, this pathogen is the subject of ongoing discovery efforts aimed at identifying novel antimicrobial and anti-virulence treatments (Cech and Horswill, 2013; Daly et al., 2015; Munguia and Nizet, 2017; Quave and Horswill, 2014; Spellberg et al., 2013; Sully et al., 2014). S. aureus has several global regulators that control the production of its virulence factors, and one such regulator is the agr quorum sensing system, which the staphylococci use to coordinate behaviour in response to an autoinducing peptide (AIP) signal (Kavanaugh and Horswill, 2016; Thoendel et al., 2011). The agr system has been extensively characterized in terms of the genes regulated, functions in virulence, interactions with other global regulatory systems, and variability across the genus (Novick and Geisinger, 2008; Thoendel et al., 2011). Briefly, the agrBDCA operon encodes the proteins (AgrB & AgrD) needed to produce the AIP signal, as well as a two-component system (AgrC & AgrA) that senses the AIP (Fig. 1A). When the AIP reaches a sufficient local concentration, it binds to the extracellular face of the AgrC histidine kinase, activating this kinase, which in turn phosphorylates the response regulator AgrA. AgrA then binds to the chromosomal P2 and P3 promoter regions to upregulate transcription of the agrBDCA operon and the RNAIII effector molecule, respectively, and in turn RNAIII induces global changes in gene expression (reviewed in (Novick and Geisinger, 2008; Thoendel et al., 2011)).

Figure 1. S. caprae inhibits all classes of S. aureus quorum-sensing.

A. Schematic of agr system operon. Staphylococcal autoinducing peptides (AIP) are processed and secreted in an AgrB-dependent process. Binding of the cognate AIP to AgrC results in phosphotransfer to AgrA, which induces transcription at the P2 and P3 promoters of the agr operon and RNAIII, respectively. Non-cognate staphylococcal AIPs can bind to AgrC and prevent signal transduction. B. Screen of clinical coagulase-negative staphylococcal (CoNS) isolates for inhibition of S. aureus agr type I–IV P3-GFP reporter. CoNS spent media was added to 10% final, and the 24 hr fluorescence time point is shown. Results are pooled from three experiments, each with N=4 replicates per condition. Two-way ANOVA with multiple comparisons (Dunnett’s correction) was performed. **** indicates p < 0.0001; *, p < 0.05.

Every staphylococcal species contains an agr locus and produces a unique AIP molecule that varies in sequence and length. All known staphylococcal AIPs contain a 5-membered thiolactone or lactone ring with an N-terminal extension, and they range in full length from 7 to 12 amino acids (Olson et al., 2014; Thoendel et al., 2011). Non-cognate AIPs have been shown to inhibit AgrC activation of AgrA, a phenomenon initially referred to as “bacterial interference” (Ji et al., 1997). Intraspecies cross-talk has been examined extensively in different agr types of S. aureus and S. epidermidis (Ji et al., 1997; Olson et al., 2014). However, interspecies cross-talk has received less attention, especially considering the depth and diversity of CoNS species (Becker et al., 2014). Otto and colleagues first observed that S. epidermidis AIP type I inhibits S. aureus quorum-sensing, although this was only potent in agr type III strains (Otto et al., 2001). Since this initial observation, there have been several reports that interspecies agr interactions occur (Canovas et al., 2016; Geisinger et al., 2009; Ji et al., 2005; Sung et al., 2006), but in many cases, the CoNS AIP structure has not been experimentally determined, but rather predicted based on sequence.

Recently, a screen with a number of animal CoNS isolates identified several that affected S. aureus agr function (Canovas et al., 2016). We performed our own initial screen, and identified species of CoNS whose spent media repressed S. aureus agr. One of the intriguing CoNS isolates that inhibited S. aureus agr function was Staphylococcus caprae, which was originally isolated from goat milk and is considered to be a commensal of goats and sheep (Devriese et al., 1983). However, recent studies have identified S. caprae on the skin of healthy humans (Cosseau et al., 2016; Gao et al., 2007; Kwaszewska et al., 2014), indicating that it may reside in a similar host environment as S. aureus. Little is known about the S. caprae agr system, although predictions based on AgrD gene sequence suggest that it has two agr types that result in different AIP structures (Thoendel et al., 2011; Zheng et al., 2015).

In this work, we characterized the S. caprae inhibitory activity and further investigated its impact on S. aureus colonization and pathogenesis. Genetic and biochemical characterization indicated that the inhibitor is the native S. caprae AIP, and mass spectrometry analysis of spent media revealed its structure. A synthetic version of the AIP was a potent inhibitor of S. aureus quorum sensing, and it prevented the progression of MRSA skin infection in a murine model. S. caprae successfully competed against S. aureus in colonization and infection models, suggesting S. caprae may have beneficial protective properties as a human commensal.

RESULTS

Commensal staphylococcal strains secrete inhibitors of S. aureus quorum-sensing

To test for cross-talk between staphylococcal strains, we selected a small group of CoNS clinical isolates and assessed whether their spent media impacted S. aureus quorum-sensing function. Staphylococcus epidermidis has already been tested in depth (Otto et al., 2001; Otto et al., 1999), and thus we focused on other staphylococcal species. For reporters, we used S. aureus strains representing all four agr types containing P3 promoter YFP-fusions that have been previously used to assess quorum-sensing activity (Kirchdoerfer et al., 2011; Quave et al., 2015). In our assays, Staphylococcus capitis and multiple clinical isolates of Staphylococcus lugdunensis did not affect S. aureus agr activation, while agr P3 inhibition was observed in spent media from Staphylococcus intermedius and Staphylococcus caprae (Fig. 1B). S. intermedius inhibited agr P3 activation in S. aureus agr types I, II, and III. The AIP from S. intermedius has been identified as an unusual lactone-based ring structure, and evidence that this AIP can inhibit different S. aureus agr types was previously noted (Ji et al., 2005). S. caprae spent media strongly inhibited agr type I, type II, and type III function in S. aureus. Interestingly, S. intermedius had no impact on S. aureus agr type IV (Fig. 1B), while S. caprae was slightly inhibitory. However, type IV strains are extremely rare and often absent from culture collections (Shopsin et al., 2003). Geisinger et al. previously noted that S. caprae spent media had inhibitory activity against S. aureus agr types I and II (Geisinger et al., 2009), although the S. caprae AIP structure was not identified. We chose to focus our efforts on S. caprae, given that it is found on human skin (Cosseau et al., 2016; Gao et al., 2007; Kwaszewska et al., 2014), and may interact with colonizing S. aureus to provide a protective benefit to the host. Additionally, S. caprae has not been extensively studied, and little is known about its AIP signals or agr system.

S. caprae spent media characterization suggests the inhibitory activity is AIP

We performed biochemical tests on the S. caprae spent media to gain insight on the nature of the inhibitor. The spent media did not inhibit S. aureus growth (Fig. S1), indicating that S. caprae does not produce an antimicrobial agent such as those that have been recently reported in other commensal staphylococci (Janek et al., 2016; Nakatsuji et al., 2017; Zipperer et al., 2016). Next, the S. caprae spent media was subjected to various treatments and tested for impact on an S. aureus agr type I reporter. The media was exposed to heat (65 °C for 1 hr), proteinase K to degrade peptides and proteins (1 mg/mL for 1 hr), chelation with EGTA (5 mM), and NaOH (to pH 11). With the exception of NaOH-treated supernatant, all the conditions retained some inhibitory activity against S. aureus agr type I (Fig. 2A). The thiolactone ring is critical for AIP functionality (Ji et al., 1997; Mayville et al., 1999; P et al., 2001), and labile in the presence of high pH; therefore, NaOH treatment of the spent media should destroy the S. caprae AIP structure. Based on these preliminary observations, we hypothesized the S. caprae inhibitory agent might be the AIP signal.

Figure 2. Initial characterization and identification of the S. caprae quorum-sensing inhibitor.

A. The S. caprae spent media was treated with 65 °C heat, proteinase K (1 mg/mL final), EGTA (5 mM final), or NaOH (to a pH of 11, then neutralized). Each sample was added to MRSA agr type I reporter to 10% final and compared to no treatment control. Florescence was measure at 24 hr, and the results are pooled from two experiments, each with N=3 replicates per condition and normalized to the media control. One-way ANOVA with multiple comparisons (Dunnett’s correction) was performed. **** indicates p < 0.0001. B. Hemolysis activity assay of MRSA wildtype and constitutive AgrC (R238H) treated with S. caprae spent media. Activity is expressed as the percent of MRSA spent media needed to achieve 50% rabbit blood cell lysis as calculated by a four-parameter logistic curve. Results are pooled from three experiments, each with N=3 replicates per condition. The pooled EC₅₀s were then analyzed by two-way ANOVA with each untreated vs. treated pair compared (Sidak multiple comparison test). ** indicates p < 0.01. C. The S. caprae agrBD genes were cloned into pEPSA5, and transformed into an S. aureus agr deletion mutant. Spent media from this strain, and additionally an empty vector control, were tested against a MRSA agr type I P3-YFP reporter. The six-hour time point is shown, and the results are pooled from two experiments, each with N=4 replicates per condition. One-way ANOVA with multiple comparisons (Dunnett’s correction) was performed. **** indicates p < 0.0001. D. Structure of S. caprae AIP with predicted y₇, y₆ and y₅ fragments. The AIP is an eight-residue peptide with the last five residues constrained as a thiolactone ring. E. Selected ion chromatogram within ± 5 ppm of m/z 1015.3871 for synthetic AIP standard (retention time 3.91, measured m/z 1015.3843). F. Selected ion chromatogram within ± 5 ppm of m/z 1015. 3871 for 30% acetonitrile/70% water fraction of S. caprae spent media (retention time 3.92, measured m/z 1015.3836).

To further characterize the S. caprae inhibitory properties, we used a constitutively active AgrC (R238H) mutant of S. aureus. AgrC (R238H) contains a point mutation in the dimerization histidine phosphotransfer subdomain of the cytoplasmic region, resulting in an “irreversibly constitutive” histidine kinase that is not inhibited by AgrC receptor antagonists (Geisinger et al., 2009). We tracked quorum-sensing induction of S. aureus using a quantitative assay for alpha toxin (Hla) function (Daly et al., 2015), which is induced when the agr system is activated. Upon treatment with S. caprae spent media, Hla activity decreased in S. aureus wildtype, but the AgrC R238H mutant retained Hla activity (Fig. 2B), similar to the expected response for treatment with a competitive antagonist like AIP-II (Daly et al., 2015). Collectively, these findings indicated that the S. caprae inhibitory activity was likely the native AIP, and its inhibitory function was upstream of phosphorylation of AgrA.

S. caprae agrBD are sufficient to inhibit quorum sensing in S. aureus

We took a genetic approach to further evaluate the hypothesis that the S. caprae inhibitor is an AIP. The S. caprae DSM 20608 type strain is not fully sequenced, but two S. caprae agr types have been annotated in the available sequences of S. caprae strains (Thoendel et al., 2011). We sequenced the agrBD locus of the 20608 strain and found that it encodes the sequence matching the proposed agr type I. To produce this peptide, we cloned agrBD from the S. caprae 20608 strain into an expression vector containing a xylose-inducible promoter (pEPSA5), resulting in the pAgrBD plasmid. Previous work has found that expression of the agrBD genes from a staphylococcal strain in a heterologous host is sufficient to produce an AIP (Thoendel and Horswill, 2013). We therefore expressed the pAgrBD plasmid in an agr deletion mutant of the strain S. aureus 4220, which does not make endogenous AIP. Background expression and xylose induction of this engineered strain conferred inhibitory activity against a S. aureus agr type I reporter strain, while induction of an empty vector control strain had no activity (Fig. 2C), indicating that the S. caprae agrBD genes are sufficient for production of the quorum-sensing inhibitor.

Identification of the S. caprae AIP structure

To identify the structure of the S. caprae AIP, mass spectrometry analysis of S. caprae spent media was performed. This experiment revealed that the native S. caprae type I AIP structure is 8 amino acids (YSTCSYYF, calculated m/z 1015.3871), with the expected C-terminal 5-membered thiolactone ring characteristic of AIPs (Fig. 2D). In a partially purified fraction of S. caprae media (eluted with 70% water/30% acetonitrile), an ion at m/z 1015.3836 was identified as being within 5 ppm of the calculated m/z of peptide YSTCSYYF (Fig. 2F). Fragmentation of the 1015.3836 ion produced ions matching the mass of the predicted y₇, y₆, and y₅ fragments (Fig. 2D) within 5 ppm.

Synthetic versions of the AIP and two derivative structures (+1 and −1 in N-terminal length) were obtained. LC-MS analysis of the synthetic eight-residue AIP matched retention time (Fig. 2E), m/z, and fragmentation of the putative AIP from the S. caprae spent media (Fig. 2F), confirming the assignment. Next, we tested all three of these synthetic AIPs against the S. aureus agr Type I–IV reporters, and found that each was able to inhibit agr function with varying efficacy (Fig. 3A–D, Table S2). The structure corresponding to the native AIP was the strongest inhibitor, with an IC₅₀ of 0.6 nM against type I S. aureus, while the 9 amino acid and 7 amino acid structures inhibited with IC₅₀’s of 0.83 nM and 3.0 nM, respectively. For all S. aureus agr types, the native AIP and the 9 amino acid structure had nearly identical activity, while the truncated 7 amino acid structure was significantly less inhibitory, especially against type IV (Fig. 3D, Table S2).

Figure 3. Inhibition of S. aureus agr types I–IV with synthetic S. caprae AIPs.

The agr P3-YFP reporters of S. aureus types I–IV (shown in panels A, B, C, and D) were treated with a dose-response of synthetic S. caprae AIP, and versions with one additional residue (AIP+1) and one less residue (AIP-1) than the native structure. Fluorescence was monitored throughout growth and fitted to a four-parameter logistic curve (GraphPad PRISM 7). One representative result is shown for each reporter (N=4 replicates per condition), and the entire experiment was repeated with similar results. Specific IC₅₀ values are reported in Table S2.

S. caprae AIP inhibits the progression of a murine MRSA skin infection

To investigate the potential of the S. caprae AIP as an anti-MRSA intervention, we evaluated its capacity to influence the infectious outcome of cutaneous challenge with the USA300 CA-MRSA strain LAC (agr type I, hereafter called MRSA) (Muhs et al., 2017; Quave et al., 2015; Todd et al., 2017). The USA300 MRSA strains are the dominant cause of skin infection and thus serve as an appropriate output for testing (King et al., 2006; Moran et al., 2006). To this end, 5 or 10 μg of AIP was intradermally administered as part of an inoculum suspension containing MRSA. When compared to controls, AIP-treated animals exhibited significant protection from both MRSA-induced local and systemic disease as measured by dermonecrosis and weigh loss respectively (Fig. 4A–C). Highlighting the potency of AIP-mediated quorum sensing inhibition, the cutaneous injuries sustained by AIP treated and agr null challenged animals were equivalently minor in severity.

Figure 4. S. caprae AIP attenuates MRSA pathogenesis.

A. Dermonecrotic lesion size following infection with MRSA WT +/− S. caprae AIP (n=4–8). Error bars represent SEM. Post test p value for every indicated time point =<0.0005 for comparisons of lesion size reduction resulting from both 5 μg and 10 μg AIP treatments. B. Weight loss and measurements following infection for the indicated groups (n=4–8). Error bars represent SEM. Post test p value (*)=<0.05, (**)=<0.01, (**)=<0.005, (***)=<0.001. C. Representative images of tissue injury following infection with MRSA WT +/− the indicated amounts of S. caprae AIP or agr null strain five days after infection. D. Representative images of in vivo bioluminescence induction one day after challenge with MRSA constitutive Lux +/− 10 μg of S. caprae-AIP or an Δagr MRSA Lux. E. Time course comparison of in vivo bioluminescence after intradermal challenge with MRSA Lux +/− 10 μg of S. caprae-AIP or an Δagr MRSA Lux (n=8–12). Error bars represent SEM. Post test p value (*)=<0.05, (**)=<0.01, (***)=<0.001.

To determine if the AIP-mediated attenuation of infectious injury occurred alongside enhanced MRSA clearance, analogous challenge experiments were performed with a MRSA Lux strain for live imaging (IVIS). This approach enabled the measurement of cutaneous bacterial loads in a non-invasive and longitudinal manner. When compared to controls, AIP-treated animals exhibited a significantly lower bacterial burden throughout the course of infection (Fig. 4D, E), and achieved levels of MRSA clearance that closely approached those observed after challenge with an agr null strain (Δagr Lux). Representative live imaging at day 1 demonstrates the dramatic difference in MRSA burden when the agr communication system is intact (MRSA Lux control), completely absent (agr null) or profoundly disrupted (AIP-treated) (Fig. 4E). Together, these data demonstrate that a single treatment of S. caprae AIP given at the time of infection affords significant protection against multiple disease outcomes, including tissue destruction, bacterial burden, and weight loss.

Given that direct AIP exposure does not inhibit MRSA growth in vitro (Fig S2), the rapid drop in MRSA burden in vivo suggests that within the context of infection, AIP promotes the bactericidal activity of host effector cells. Congruent with this, we observed a profound increase in the number of neutrophils (PMNs) accumulating at the cutaneous challenge sites of AIP-treated animals relative to controls (Fig. S3). Furthermore, parallel assessment of PMN phagocytosis within infected skin revealed that the AIP-induced reduction in bacterial burden corresponded with enhanced MRSA uptake by host PMNs (Fig S3). Considering that the magnitude of PMN infiltration elicited by sterile injection of AIP or vehicle were equivalent (Fig S4), it appears that the capacity of AIP to potentiate host defense is not a function of its inherent immunogenicity, but rather an indirect effect of sensitizing MRSA organisms for phagocytic clearance.

Finally, we assessed the efficacy of AIP as an intervention against established infections. To this end, at both 24 and 48 hours after MRSA Lux infection, mice were intradermally administered AIP (10 μg or 50 μg) at skin challenge sites. Similar to previous MRSA challenge experiments, grossly evident dermonecrosis peaked within the first 5 days following infection. Although the severity of acute injury did not differ between treatment groups, mice receiving a high dose of AIP treatment (50 μg) resolved their ulcers at an accelerated rate relative to controls. The AIP-induced promotion of cutaneous wound healing became significant late in the course of infection (>day 9 post infection) and corresponded with reduced MRSA burden at multiple time points (Fig S5). Altogether these results show that in contrast to the overwhelming protection afforded by prophylactic AIP administration, the therapeutic effects of late-stage delivery are more subtle and manifest later during the course of infection.

S. caprae AIP inhibits agr activation during a MRSA skin infection

The efficacy of AIP encouraged us to better define the relationship between MRSA hypo-virulence and quorum sensing inhibition in vivo. For this purpose, we carefully monitored agr activation kinetics within the infectious environment by intradermally challenging animals with MRSA carrying an agr P3-Lux reporter plasmid. Over the course of the first five hours of infection, exposure to S. caprae-derived AIP effectively blunted MRSA agr activation (Fig. 5A & B). By measuring infection progression in these animals (Fig. 5C), we showed that the level of AIP-mediated agr interference occurring during this early period after challenge corresponded with a marked attenuation of MRSA-induced dermatopathology. Histological assessment of AIP-treated and control animals (day 5 post infection), revealed characteristic ulcerative cratering with extensive dermal necrosis and complete epidermal destruction in the latter group (Fig. 5D). In contrast, the challenge sites of their AIP-treated counterparts were typified by narrow bands of necrosis that were restricted to the boundary of a highly consolidated dermal mass of inflammatory cells, all which was overlaid by an intact epidermis. Altogether, these data show that AIP-mediated quorum quenching in vivo profoundly attenuated the pathologic outcome of MRSA infection.

Figure 5. S. caprae AIP-mediated quorum quenching in vivo corresponds with protection against MRSA-induced dermatopathology.

A. Representative images of in vivo bioluminescence induction 2.5 hours after challenge with MRSA agr P3-Lux +/− 10 μg of S. caprae AIP. B. Time course comparison of in vivo bioluminescence after intradermal challenge with agr P3-Lux +/− 10 μg of S. caprae AIP (n=5). Error bars represent SEM. Post test p value (***)=<0.005. C. Dermonecrotic lesion size following infection with agr P3-Lux +/− 10 μg of S. caprae AIP (n=5). Error bars represent SEM. Post test p value (*)=<0.05, (**)=<0.01, (***)=<0.001. D. Left: representative images showing gross appearance of gross tissue injury five days after infection with agr P3-Lux +/− 10 μg of S. caprae AIP. Right: Representative H&E stained tissue sections prepared 5 days after infection agr P3-Lux +/− 10 μg of S. caprae AIP. Arrows denote distinct areas of tissue necrosis.

S. caprae provides protection during skin colonization and infection competition

When S. caprae and MRSA occupy the same host environment, signaling cross-talk and competition for resources could potentially influence colonization and infection progression. To address this possibility, we performed a series of co-inoculation experiments, mixing S. caprae and MRSA Lux in a 1:1 ratio, within settings of cutaneous carriage or invasive infection. Following epicutaneous application of both strains, a significant enhancement of early MRSA clearance (day 1) was noted using live imaging (Fig. 6A & 6C), without marked injury (Fig. 6B). Consistent with the known transient nature of S. aureus colonization in immune competent mice (Hashimoto et al., 2004), we observed that by day 4 the bioluminescence signals were quite low (Fig. 6C), and by day 5 had reached baseline, indicating that bacterial clearance had occurred.

Figure 6. S. caprae reduces early MRSA burden following epicutaneous inoculation.

A. Representative images of in vivo bioluminescence induction one day after epicutaneous administration of MRSA constitutive Lux alone or together with S. caprae. B. Representative images of tissue injury following epicutaneous application of the indicated groups 1 and 3 days after challenge. C. Time course comparison of in vivo bioluminescence after MRSA mono (MRSA alone) or co (MRSA/S. caprae) application (n=6). Error bars represent SEM. Post test p value (*)=<0.05.

To assess the capacity of S. caprae to impact the outcome of an outright MRSA infection, equal CFU of both strains (S. caprae and MRSA agr P3-Lux) were administered as part of the same inoculum suspension. Despite effectively doubling the incoming bacterial load, when compared to single culture infections, the presence of S. caprae afforded striking protection against MRSA pathogenesis (Fig. 7A & B). By using the MRSA agr P3-Lux reporter for these challenges, we were able to show that the equivalent efficacy of both the synthetic AIP and S. caprae bacteria during the first 5 hr of infection (Fig. 7C). Monitoring of the infection for 12 days showed the continued effectiveness of synthetic AIP and S. caprae bacteria (Fig. 7D).

Figure 7. S. caprae reduces early MRSA burden following intradermal inoculation.

A. Representative images of in vivo bioluminescence induction 3.5 hours after challenge with MRSA agr P3-Lux +/− 10 μg of S. caprae AIP, or equal CFU of the CoNS S. caprae or S. capitis. B. Representative images show dermonecrosis 5 days after the bacterial challenge. C. Time course comparison of in vivo bioluminescence after intradermal challenge with within the indicated conditions (n=5–10). Error bars represent SEM. Post test p value (*)=<0.05, (**)=<0.01, (***)=<0.005. D. Time course of dermonecrotic lesion size in the indicated challenge conditions (n=5–10). Error bars represent SEM. Post test p value (*)=<0.05, (**)=<0.01.

As a control, analogous challenge experiments were performed with a closely related CoNS species, Staphylococcus capitis, which did not inhibit S. aureus agr function during in vitro testing (Fig. 1B). Compared to S. caprae, the attenuation of MRSA virulence and agr activity with S. capitis organisms was minimal compared to vehicle control (Fig. 7A–D), indicating these properties are specific to S. caprae and cannot be broadly generalized throughout the genus. Taken together, these data demonstrate the ability of the S. caprae to employ agr antagonism as means to oppose both the commensal carriage and pathogenic invasion of MRSA in a cutaneous environment.

DISCUSSION

Infections from the bacterial pathogen S. aureus place a tremendous burden on our healthcare system (Dantes et al., 2013; Lowy, 1998; Tong et al., 2015). Often termed a pathobiont, S. aureus is both a potent bacterial pathogen and a common constituent of human cutaneous and mucosal flora that persistently colonizes 20% of the healthy adult population (Kluytmans et al., 1997). Given that colonization is a known risk factor for invasive autoinfection (Dantes et al., 2013; Lowy, 1998; Tong et al., 2015), there is great interest in elucidating the ecological factors that either prevent carriage of S. aureus, result in benign carriage, or induce a transition to invasive infection. There is growing appreciation that the quality and composition of the resident microbiota impacts these different states, and that the skin microbiome can function as a natural protective barrier for the host (van Rensburg et al., 2015). As commensal bacteria colonize a specific niche and compete for resources, extensive interactions occur (Ghoul and Mitri, 2016), and competing bacteria can produce factors that limit S. aureus growth. For commensal CoNS species, it has been observed that S. epidermidis secretes a protease, and S. lugdunensis and other species release antimicrobials, each of which restricts S. aureus colonization (Iwase et al., 2010; Janek et al., 2016; Nakatsuji et al., 2017; Zipperer et al., 2016). These examples illustrate how resident commensal species directly target S. aureus to gain a competitive advantage for resources, thereby limiting pathogenic outgrowth and invasion within shared ecological niches.

There are some reports that targeting quorum sensing could be an effective strategy enabling commensals to reduce the fitness of their competitors (Dong et al., 2002; Fleming et al., 2006; Ghoul and Mitri, 2016). In this work, we discovered that S. caprae secretes an AIP signal that strongly inhibits S. aureus quorum sensing function. Based on sequence analysis, Geisinger et al. predicted the S. caprae AIP could inhibit the AgrC receptor of S. aureus types I and II, and demonstrated its efficacy by in vitro spent media testing (Geisinger et al., 2009). Our findings confirm preliminary findings by Geisinger and coworkers, and extend them to all classes of S. aureus agr. Using high resolving power mass spectrometry, we directly identified the correct S. caprae AIP structure from the spent media, which was previously unknown. Considering that AIP tail length can be highly variable (Olson et al., 2014), and tail length strongly influences activity (Gordon et al., 2013), it is critical to directly measure the native AIP released by the bacteria to confirm its structure. Although S. caprae was historically regarded as a primarily animal-associates species, recent studies reveal that the presence of this microbe on human skin has been underappreciated. Knowing that S. caprae is present at transiently on human skin (Cosseau et al., 2016; Gao et al., 2007; Kwaszewska et al., 2014), it seems likely that S. caprae and S. aureus interact within the cutaneous environment, and the production of the inhibitory AIP could give S. caprae a competitive advantage. In the context of these findings, our work suggests that further studies should be carried out to characterize the human skin microbiota to the species level to better appreciate the contribution of S. caprae. Recently, another human skin commensal, Corynebacterium striatum, was found to suppress S. aureus quorum-sensing function and showed promise in a co-infection model (Ramsey et al., 2016). These previous findings and those presented herein suggest that antagonistic cross-talk could be an effective strategy for a commensal to gain advantage in the complex skin environment.

To test the therapeutic efficacy of interspecies cross-talk, the impact of S. caprae AIP on MRSA skin infection progression was examined using a similar approach that was pioneered for assessment of intraspecies bacterial interference (Mayville et al., 1999). We observed profound attenuation of MRSA virulence when it was inoculated along with synthetic S. caprae AIP, which essentially phenocopied analogous challenges with agr null mutants. This success encouraged us to investigate whether the magnitude of S. caprae AIP production was sufficient to confer protection against MRSA within a shared host environment. The abatement of MRSA quorum sensing within the polymicrobial challenge setting suggests that S. caprae-derived AIP does, indeed, achieve the potency needed to blunt MRSA virulence factor induction. To explore S. caprae/MRSA interactions in a more natural context, we developed a model of epicutaneous administration/colonization that elicited little tissue injury and permitted longitudinal tracking of MRSA burden. Consistent with other models of colonization with immune competent mice (Hashimoto et al., 2004), we found that MRSA carriage was relatively transient. Importantly, the occupancy of MRSA upon the skin surface was substantial enough to produce a robust signal for bioluminescent imaging over a three-day period (Fig. 6). Using this approach, we observed that the presence of S. caprae upon the skin significantly restricted the MRSA growth and accelerated its clearance, indicating that competing commensals may serve as a means by which the microbiota restrict MRSA outgrowth and invasion.

An ongoing challenge to understanding the contribution of commensal cross-talk is that the role of agr during staphylococcal colonization has not received much attention. We and others have observed that the S. aureus agr genes are repressed during nasal colonization (Burian et al., 2010; Kiedrowski et al., 2016; Tulinski et al., 2014), but staphylococcal skin colonization has comparatively received less attention at the mechanistic level. Using a porcine skin explant model, we demonstrated that S. epidermidis requires quorum sensing to successfully colonize (Olson et al., 2014), suggesting the regulatory system is important for survival in this environment. Considering that every staphylococcal strain has an agr system (Wuster and Babu, 2008) (with varying AIP structures and receptors), it seems probable that both S. caprae and S. aureus require a functional agr system for skin colonization. Across all CoNS strains, the agr system controls production of many secreted enzymes necessary for growth on host substrates (Kolar et al., 2013; Olson et al., 2014), and controls the production of PSM peptides, which have broad-ranging properties that impact motility, host cell lysis, and biofilm restructuring (Peschel and Otto, 2013). However, our knowledge of exo-enzyme and PSM function in most CoNS species is relatively limited. Despite these knowledge gaps, our findings with interspecies quorum sensing antagonism suggest it could be an effective strategy for survival within a competitive host environment.

Although S. caprae has received little attention, its appearance in species-level analysis of skin bacterial communities suggests that its contribution as a human colonizer is underestimated (Cosseau et al., 2016; Gao et al., 2007; Kwaszewska et al., 2014). Studies to date have relied mostly on 16s rRNA sequencing, and there are inherent challenges in separating closely-related species, such as each individual CoNS, and analyzing the temporal population dynamics with a single marker gene (Poretsky et al., 2014). As microbiome studies advance, our understanding of the prevalence of S. caprae as part of the healthy skin flora will improve, and the frequency of association or disassociation with S. aureus colonization will be revealed. Given that other CoNS make competing AIPs (Fig. 1B and (Canovas et al., 2016)), and the observation that many isolates produce antimicrobials (Janek et al., 2016; Nakatsuji et al., 2017; Zipperer et al., 2016), there is clearly a network of complex, competitive interactions among commensals in the skin environment. Nakatsuji et. al recently demonstrated the importance of understanding these interactions by showing that a Staphylococcus hominus strain can protect atopic dermatitis patients from S. aureus colonization (Nakatsuji et al., 2017). This exciting advancement with S. hominus demonstrates the successful translation of a commensal’s protective properties into an effective treatment, and it highlights the significance of investigating the contribution of normal skin flora to human health. With the post-antibiotic era looming ever-closer, there is a clear need for these types of innovative scientific solutions to control antibiotic resistant infections (Spellberg et al., 2013). It is becoming increasingly clear that breakthroughs in this area may arise from the pursuit of pathogen-specific interventions that exploit specific microbial pathways, or strategies to take advantage of protective environmental conditions, such as the beneficial properties of normal flora. For S. aureus, a ubiquitous human pathogen with a remarkable propensity for acquiring drug resistance, there are mounting efforts to develop both of these alternative approaches into effective treatments for limiting infections.

STAR METHODS

CONTACT FOR REAGENT AND RESOURCE SHARING

Further information and request for resources should be forwarded to and will be fulfilled by the Lead Contact Dr. Alexander R. Horswill at the University of Colorado, Department of Immunology & Microbiology at the Anschutz Medical Campus in Aurora, CO, 80045. Phone: (303) 724 3534. E-mail: alexander.horswill@ucdenver.edu

EXPERIMENTAL MODEL AND SUBJECT DETAILS

Bacterial strains and plasmids

Strains, plasmids, and bacteriophages used in this chapter are listed in Supplementary Table 1. Unless otherwise noted, strains were cultured at 37°C with 200 RPM shaking. E. coli was grown in Luria broth (LB) or on LB agar plates containing 100 μg/mL ampicillin to maintain plasmids when necessary. S. aureus and S. caprae were grown in tryptic soy broth (TSB) or on TSB agar plates. Staphylococcal plasmids were maintained with growth in 10 μg/mL chloramphenicol.

Mice

Male BALB/c mice were purchased from the Charles River and housed in specific pathogen free facilities at the University of Iowa. For in vivo studies, 8 to 20-week old age-matched mice were randomly assigned to treatment groups and at experimental end points, mice were humanely euthanized using carbon dioxide inhalation. Prior to their inclusion in the study, mice were allowed to acclimate to the ABSL-2 animal housing facility at the University of Iowa for at least seven days. The animal studies were reviewed and protocol approved by the University of Iowa Institutional Animal Care and Use Committee. The University of Iowa is AAALAC accredited, and the centralized facilities meet and adhere to the standards in the “Guide and Care of Laboratory Animals.”

METHOD DETAILS

agr reporter assay

Overnight cultures of S. aureus reporter strains were grown in TSB with 10 μg/mL chloramphenicol, then subcultured at 1:200 in fresh TSB with 10 μg/mL chloramphenicol. Overnight culture of S. caprae was centrifuged and the supernatant was passed through a 0.22 μM filter to generate spent media. Spent media, or fresh TSB for the negative control, were added at the indicated concentrations to the reporter cultures. The final cultures were then plated at 200 μL/well into a 96-well plate and grown in a Stuart humidified incubator at 1000 RPM and 37°C. At hourly time points, plates were measured on a TECAN plate reader to quantify OD₆₀₀ and YFP fluorescence (Ex 480/Em 515, Gain 60). For synthetic peptide addition, peptides or DMSO control were added from stocks of 20 μM in DMSO. Data from each time point were analyzed by a four-parameter logistic curve in Graphpad PRISM, where the IC₅₀ is the midpoint of the curve. The IC₅₀ for the time point with maximal agr activation is reported in Table 2, and error is reported as standard deviation.

S. caprae supernatant treatments

For heat treatment, S. caprae spent media from an overnight culture was incubated in a 65 °C heat block for one hour. For proteinase K treatment, 5 μL of a 20 mg/mL proteinase K solution (Fermentas) were added to 100 μL of spent media and incubated in a 37 °C heat block for one hour. For EGTA treatment, 0.5 μL of a 1M stock were added to 100 μL of spent media and incubated at room temperature for one hour. For NaOH treatment, the initial pH of the spent media was measured using 4–7 pH strips. Then, 20 μL of 10 N NaOH were added to 3 mL of spent media, achieving a pH of 11 according to a 0–14 pH strip. This spent media incubated for one hour at room temperature, then was returned to the initial pH of approximately 6.5 by adding approximately 15 μL of 37% HCl. Untreated S. caprae spent media also incubated at room temperature for one hour as a control. All treated supernatants were added to S. aureus agr type I reporter cultures at 10% vol/vol in 96-well plates as above.

Hemolysis assay

Hla activity assay was performed as in (Daly et al., 2015). Overnight cultures of LAC and the constitutive agrC mutant were subcultured at 1:100 in 5 mL TSB and grown for 18 hours with 10% fresh S. caprae spent media. Spent media from these cultures was serally diluted 2-fold across a 96-well plate. Rabbit defibrinated blood (HemoStat Laboratories) was centrifuged for ten minutes at 500 × g, 4°C to pellet rabbit erythrocytes, which were washed three times in cold PBS and resuspended to a final concentration of 3% in PBS. 30 μL of each spent media dilution was mixed with 70 μL of the erythrocyte solution in triplicate in 96-well plates, and then incubated statically at room temperature for 1 hour, after which OD₆₃₀ was measured using a TECAN plate reader. OD₆₃₀ versus percent spent media was plotted on a four-parameter logistic curve using Graphpad PRISM 7 and the midpoint of the curve (EC₅₀) was used as an indicator of activity.

S. caprae agrD sequencing

Genomic DNA was isolated from the S. caprae strain DSMZ 20608 using the Gentra PureGene Yeast/Bact Kit (Qiagen) according to the following modified protocol: after cells were pelleted, they were suspended in 300 μL 50 mM Tris-HCl, pH 8.0 with 100 ug/mL RNAseA (Qiagen), and lysozyme (Sigma-Aldrich) was added to a final concentration of 50 ug/mL (instead of the Qiagen Lytic Enzyme solution). After one hour incubation at 37°C, the manufacturer’s protocol was continued, with the omission of the RNAseA step and the addition of ten-minute incubation on ice after adding the Protein Precipitation Solution. DNA was submitted for Sanger sequencing to the Iowa Institute of Human Genetics (Univ. of Iowa) using primers AP53A, sequence 5′ CGTTAATGTGTCATATTGTG 3′ and AP54A, sequence 5′ CACACTTTCTGTAAATGACT 3′. The primers were chosen from a partial sequence of Staphylococcus caprae strain N900362, GenBank: AF346717.1.

pAgrBD induction experiment

For construction of the pAgrBD plasmid, the agrBD genes from the S. caprae strain DSMZ 20608 were amplified from genomic DNA (isolated as above) using Phusion High-Fidelity DNA Polymerase (NEB) and primers AP52, sequence 5′ GAACGAATTCTAATGAATTATGAGGAGAGTAGTAGATAAGTG 3′ and AP53, sequence 5′ GCGTGGATCCGTTATCCAATCATTCAAGCCTTCC 3′, followed by cleanup with the QIAquick PCR purification kit (Qiagen). The resulting product and pEPSA5 were digested with EcoRI (NEB) and BamHI (NEB), gel purified using the QIAquick gel extraction kit (Qiagen), ligated using T4 DNA ligase (NEB), and electroporated into E. coli BW2151. The resulting plasmid was then isolated using the HiSpeed Mini Plasmid kit (IBI) and transformed by electroporation into the S. aureus 4220 agr mutant (AH2492). To test the plasmid, S. aureus AH2492 containing the empty vector pEPSA5 or pAgrBD plasmid was grown in TSB with 10 μg/mL chloramphenicol. The strains were then subcultured at 1:500 in TSB with chloramphenicol and induced with xylose (from a 20% stock) at various concentrations for 14–15 hours. Spent media was obtained from these cultures and used in the reporter assay at 10% vol/vol.

Flow cytometry

Approximately 24 hr following intradermal challenge with 1×10⁸ MRSA-GFP (+/− 10μg AIP), or its agr deficient counterpart, the affected abdominal regions were carefully excised with a surgical scissors. For sterile injections of AIP or vehicle, an 8mm biopsy punch was used to remove the tissue surrounding the challenge site. In all cases, recovered tissue was incubated in trypsin (0.6% in PBS) for 75 min at 37°C; then cut into small pieces and incubated in Collagenase type II (1 mg/mL RPMI) for 90 min at 37°C. Cell suspensions were generated by serial passage of skin fragments through 18 and 20 gauge syringes. Single cell suspensions from skin preparations were stained with the following antibodies: Anti-Ly6G (AI8), anti-CD11b (M1/70), CD45 (30-F11), which were purchased from BioLegend. To block nonspecific binding, cells were incubated with rat anti-mouse CD16/32 FcγRIII/II (2.4G2) and vortexed prior to surface staining. In all experiments, cells were collected on a FACS LSR using Diva software, and analyzed using FlowJo software. Dead cells were excluded by low forward-scatter and side-light scatter. Spectral overlaps between fluorochrome channels were corrected by automated compensation on singly stained, positive controls for each fluorochrome. In general, 50,000 cells were collected/tube.

LC-MS identification of S. caprae AIP structure

Liquid chromatography-mass spectrometry analysis of S. caprae spent media was performed using a similar approach to that previously described (Olson et al., 2014). This media was fractionated using a Teledyne-Isco CombiFlash RF system. Spent media (9 mL) was combined with acetonitrile (1 mL), and injected onto a 130g Redisep RF C18 column. Sample was eluted from column using the following binary solvent gradient consisting of acetonitrile (solvent A) and water (solvent B) at a flow rate of 80 mL/min. Gradient initiated at 10% A isocratic hold for 1 column volume, then increased to 20% A and held for 1 column volume, increased to 30% and held for 1 column volume, these steps continued until solvent composition consisted of 100% A. The final step was held for 3 column volumes. Fractions were collected in 19 mL increments, and pooled based on percent B for a total of 10 fractions. Fractions were analyzed using a Waters Acquity UPLC coupled to a Thermo Fisher Q Exactive Plus orbitrap mass spectrometer. A 7 μL injection of each fraction was eluted from an Acquity BEH C18 UPLC column with the following binary solvent system containing water with 0.1% formic acid (solvent A) and acetonitrile with 0.1% formic acid (solvent B) at a 0.3 mL/min flow rate. The gradient initiated with a 0.5 min isocratic hold at 10% B. From 0.5 to 7.0 the gradient increased linearly to 60% B, followed by an isocratic hold for 0.5 min. The column was washed with 100% B and equilibrated to starting conditions from 7.5 min to 10 min. The Q-Exactive Plus was operated in the positive ion mode using the following settings: Ion source, HESI; spray voltage set, 4.0 kV; probe heater temperature, 412°C; sheath gas flow, 50; auxiliary gas, 15; sweep gas, 2; capillary temperature, 300°C; S-lens RF level, 80. A full scan experiment, scan range 500–2000, was used to identify ions matching within 5 ppm of possible AIP m/z values. A targeted MS² experiment was performed using the detected AIP m/z as the precursor ion, with an isolation window of 4.0 m/z and a normalized collision energy of 23 (arbitrary units).

Mouse dermonecrosis model and colonization models

Mice and S. aureus skin challenge model

For intradermal challenges: Overnight cultures (grown in TSB) of USA300 MRSA strain (AH1263) constructs derived from this parental strain (e.g., Lux+ MRSA) or S. caprae were subcultured (1:100) in TSB and to an OD₆₀₀ of 0.5. Bacterial cells were then pelleted and resuspended in sterile saline. For all challenge experiments the administration of bacterial cells was preceded by shaving, ethanol cleansing and performed with isoflurane anesthesia.

For AIP efficacy testing

50 μL inoculum suspensions containing 1×10⁸ CFUs and either S. caprae AIP (5 μg or 10 μg at a concentration of 10 μg/μL in neat DMSO) or DMSO alone were injected intradermally into abdominal skin using 0.3 mL/31 gauge insulin syringe (BD, Franklin Lakes, NJ). Baseline body weights of mice were measured before infection and every day thereafter for a period of 14 days. For determination of lesion size, digital photos of skin lesions were taken daily with a Canon Rebel Powershot (ELPH 330 HS) and analyzed via ImageJ software (National Institutes of Health Research Services Branch, Bethesda, MD, USA). For AIP treatment experiments, 50 μL inoculum suspensions containing 1×10⁸ CFUs MRSA Lux were injected intradermally into abdominal skin in the same fashion. At both 24 and 48 hours following infection mice received intradermal 50 μL injections of saline containing AIP (10 or 50 μg) or neat DMSO.

Colonization competition experiments: an established method of epicutaneous challenge was employed with modifications (Parlet et al., 2015). Briefly, 20 μL inoculum suspensions containing 1×10⁸ CFUs of both MRSA and S. caprae and MRSA (2×10⁸ total organisms) or 1×10⁸ CFUs of MRSA alone was pipetted directly onto the abdomen immediately after the skin surface received 5 gentle strokes with UV irradiated 220 grit sandpaper. A gentle stream of air was then focused over the challenge site until the skin appeared moist but absent of any standing liquid. Finally, challenge sites were covered with an adhesive bandage for ≈30 minutes. For all infections, challenge dose was confirmed by plating serial dilutions of inoculum on TSA and counting ensuing colonies after overnight culture.

Inoculum preparation for assessing quorum quenching in vivo

AH2759, a USA300 MRSA (LAC) strain containing the agr P3::lux reporter plasmid was tested in a manner that was similar to previously described (Muhs et al., 2017; Quave et al., 2015). The strain was grown overnight in TSB with 10 μg/mL chloramphenicol. Overnight cultures were subcultured at 1:100 in TSB with 10 μg/mL chloramphenicol and grown to an optical density of ≈ 0.1 at 600nm. Bacterial cells were then pelleted and resuspended in sterile saline. Inoculum suspensions containing 1×10⁷ CFUs and either 10 μg of S. caprae AIP, DMSO alone or an equal number S. caprae or S. capitis organisms (prepared in the prepared in the same manner) were injected intradermally into abdominal skin using 0.3 mL/31 gauge insulin syringe. As a technical control, several mice were injected in the same manner with 50μL of sterile saline only. Beginning immediately after infection, mice were imaged under isoflurane inhalation anesthesia (2%). Photons emitted from luminescent bacteria were collected during a 2 min exposure using the Xenogen IVIS Imaging System and living image software (Xenogen, Alameda, CA). Bioluminescent image data are presented on a pseudocolor scale (blue representing least intense and red representing the most intense signal) overlaid onto a gray-scale photographic image. Using the image analysis tools in living image software, circular analysis windows (of uniform area) were overlaid onto abdominal regions of interest (as depicted in Fig. 5a) and the corresponding bioluminescence values (total flux) were measured and plotted vs. time after infection. For all infections, challenge dose was confirmed by plating serial dilutions of inoculum on TSA and counting ensuing colonies after overnight culture.

QUANTIFICATION AND STATISTICAL ANALYSIS

For each experiment, the total number of replicates (N) and the statistical tests performed can be found in the figure legends. With the exception of flow cytometry data, all analyses were performed using GraphPad PRISM software version 7 (GraphPad Software, Inc. La Jolla, CA). Where multiple comparisons were used, a correction was performed to generate the highest power statistical test, as recommended by the GraphPad PRISM software. Throughout this work, significance was defined as p < 0.05. To compare the quorum quenching abilities of coagulase-negative strains against the four S. aureus agr types (Fig. 1B), a two-way ANOVA with multiple comparisons (Dunnett’s correction) was performed. For multiple comparisons, within every agr type, each treated condition was compared to the untreated control of that agr type. To compare the quorum inhibitory effects of S. caprae spent media treated with various disruptions (Fig. 2A), data from each experiment were independently normalized to the untreated S. caprae supernatant control, then pooled and analyzed by a one-way ANOVA with multiple comparisons (Dunnett’s correction). For multiple comparisons, each treatment was compared to the untreated S. caprae supernatant control. To compare each S. caprae spent media-treated strain (wild type or constitutive agrC) with its respective untreated control (Fig. 2B), EC50’s of the four strain conditions were independently calculated for each experiment using a four-parameter logistic curve. The pooled EC₅₀’s were then analyzed by a two-way ANOVA with each untreated vs. treated pair compared (Sidak multiple comparison test). To quantify the quorum quenching effects of engineered spent media (Fig. 2C), a one-way ANOVA with multiple comparisons (Dunnett’s correction) was performed on the dataset as a whole, with each value compared to the untreated control. To quantify activity of the synthetic peptides (Fig. 3A–D), a four parameter logistic curve was used in PRISM to independently calculate IC₅₀ for each experiment. In figures four through seven, unless otherwise noted, statistics are reported as two-tailed, unpaired Student’s T tests with each treatment compared to the vehicle control.

DATA AND SOFTWARE AVAILABILITY

S. caprae sequencing data is deposited in NCBI GenBank as MG159799.

Highlights.

• Staphylococcus caprae autoinducer (AIP) inhibits agr quorum sensing in S. aureus

• Mass spectrometry identified the S. caprae AIP as an 8-residue thiolactone peptide

• S. caprae AIP attenuates MRSA-induced necrosis and burden in a skin infection model

• S. caprae directly competes with MRSA during skin colonization and infection