SarS Is a Repressor of Staphylococcus aureus Bicomponent Pore-Forming Leukocidins

ABSTRACT Staphylococcus aureus is a successful pathogen that produces a wide range of virulence factors that it uses to subvert and suppress the immune system. These include the bicomponent pore-forming leukocidins. How the expression of these toxins is regulated is not completely understood. Here, we describe a screen to identify transcription factors involved in the regulation of leukocidins. The most prominent discovery from this screen is that SarS, a known transcription factor which had previously been described as a repressor of alpha-toxin expression, was found to be a potent repressor of leukocidins LukED and LukSF-PV. We found that inactivating sarS resulted in increased virulence both in an ex vivo model using primary human neutrophils and in an in vivo infection model in mice. Further experimentation revealed that SarS represses leukocidins by serving as an activator of Rot, a critical repressor of toxins, as well as by directly binding and repressing the leukocidin promoters. By studying contemporary clinical isolates, we identified naturally occurring mutations in the sarS promoter that resulted in overexpression of sarS and increased repression of leukocidins in USA300 bloodstream clinical isolates. Overall, these data establish SarS as an important repressor of leukocidins and expand our understanding of how these virulence factors are being regulated in vitro and in vivo by S. aureus.


RESULTS
Screen to identify transcriptional regulators that alter S. aureus leukocidin promoter activity. To identify transcriptional factors that regulate S. aureus leukocidins, we conducted a screen to measure lukSF-PV promoter activity in transposon mutants from the Nebraska transposon library (30), which was constructed in S. aureus strain JE2. JE2 is in the USA300 background, which is the most prevalent lineage in the current epidemic of community-associated methicillin-resistant S. aureus (MRSA) (31). To focus our search on genes that could be directly involved in virulence regulation, we utilized a previously described regulator sublibrary (32), which includes 250 mutants in genes likely to be involved in transcription and translation. We transduced these mutants with a plasmid containing a lukSF-PV promoter fusion driving the expression of the click beetle red luciferase (CBR-luc) (33). lukSF-PV was selected for this screen due to its high levels of expression in vitro (32), which would allow us to more easily observe change in expression between the transposon mutants and the wild-type (WT) strain. Following overnight growth, the reporter library was subcultured for 5 h in deep-well 96-well plates to stationary phase. We then added D-luciferin and measured luminescence as a readout for promoter activity (Fig. 1A).
Compared to the wild-type JE2 strain, we observed increased expression of the lukSF-PV reporter in the rot mutant (rot::bursa) and decreased expression in the saeS mutant (saeS::bursa), validating the screen. Altogether, the screen identified 27 mutants that had at least a 2-fold increase in promoter activity compared to the wild-type strain and 10 mutants that had at least 2-fold less promoter activity than wild-type JE2 ( Fig. 1B; see also Table S1A in the supplemental material). We next determined which mutants had the largest differences in promoter activity and focused on genes that had features that would suggest that they were involved in the direct repression or activation of the toxin, such as DNA-binding motifs. This analysis resulted in 8 putative repressors and 10 putative activators of lukSF-PV promoter activity (Fig. 1C). Of note, all these mutants had upregulation or downregulation of promoter activity similar to that of either the rot::bursa strain or the saeS::bursa strain, respectively.
The mutant strain that had the highest level of lukSF-PV promoter activity was the sarS::bursa strain. SarS (also known as SarH1) is a member of the SarA family of S. aureus regulators (34,35). The regulation of sarS is highly complex: sarS is repressed by SarA, MgrA, and Agr (34)(35)(36), while it is activated by SarT (37), Rot (38), and TcaR (39) and posttranscriptionally activated by gdpS (40). SarS is an activator of spa (encodes protein A), which is located directly downstream of the sarS locus (34). This activation is part of a regulatory cascade involving Agr-mediated regulation of SarT, which, in turn, activates SarS to upregulate spa expression (25). SarS is also an activator of the protease ScpA (41). Additionally, SarS has been shown to be a repressor of hla, the gene encoding alpha-toxin (34), as well as other virulence factors, including the protease-encoding genes aur, ssp, and spl (34,41,42), and the exfoliative toxin-encoding gene eta (43). Lastly, SarS has been shown to bind directly to the promoter regions of hla, spa, and ssp (34).
To further validate the high levels of lukSF-PV promoter activity that we observed in the sarS::bursa strain, we transduced the transposon into the USA300 AH-LAC background (44) and measured toxin transcript levels of all five leukocidins and alpha-toxin via reverse transcription-quantitative PCR (qRT-PCR), at both the exponential and stationary phases (Fig. 1D). We observed a significant increase in transcript levels of lukED and lukSF-PV in the sarS::bursa strain compared to the wild type at both time points, further establishing SarS as a repressor of leukocidins. Of note, we observed only a modest increase in hla transcript in sarS::bursa at both time points.
SarS represses lukED and lukSF-PV expression and production. To further investigate the role of SarS in toxin regulation, we generated a pair of isogenic sarS deletion promoter activity compared to the wild type, measured by luminescence. Statistical analysis was performed using 2-way ANOVA with multiple comparisons. Error bars show standard errors of the mean. (D) Log 2 fold change of pore-forming toxin transcript levels measured by qRT-PCR in the sarS::bursa strain compared to wild-type AH-LAC in exponential and stationary phases of growth (n = 3). Statistical analysis was performed using 2-way ANOVA with multiple comparisons. Error bars show standard errors of the mean. *, P , 0.05; **, P , 0.01; ***, P , 0.001; ****, P , 0.0001. ns, not significant. and complementation strains. The DsarS strain was made by phage transducing the sarS::erm mutation, described by Cheung et al. (35), into AH-LAC, while the DsarS::sarS complementation strain was created using pIMAY* (45), in which sarS::erm was replaced with wild-type sarS. As with the transposon strain, we observed a significant increase in lukED and lukSF-PV transcripts in the sarS deletion strain compared to the wild-type strain ( Fig. 2A). Again, we detected a modest increase in hla transcripts as with the transposon strain. Importantly, the increased toxin levels were repressed to wild-type levels in the complementation strain. We observed that lukED and lukSF-PV transcripts were more highly upregulated in the DsarS strain than the hla transcript, suggesting that SarS represses these toxins to a greater extent than it represses hla.
We next investigated the impact of SarS on toxin production. We observed altered exoprotein profiles when sarS was deleted and noticed increased abundances of proteins that corresponded to the sizes of the two components of the leukocidins (;35 kDa and ;43 kDa) (Fig. 2B). Of note, these changes were restored to wild-type levels in the complementation strain. To directly evaluate the impact of deleting sarS on the production of leukocidins, we performed Western blotting using toxin-specific antibodies. We observed higher levels of LukE, LukF-PV, and alpha-toxin at both 3 and 5 h in the DsarS strain, a phenotype that was complemented by restoring sarS (Fig. 2C). Altogether, these data demonstrate that SarS represses lukED and lukSF-PV expression, leading to decreased toxin production. SarS represses toxins dually through activation of Rot and direct binding of toxin promoters. To determine how SarS represses the leukocidins, we first investigated if SarS was regulating any of the known regulators of toxin expression in S. aureus. This includes RNAIII, the effector molecule of the Agr system, the repressor of toxins Rot, and the activator of leukocidins, the SaeRS system (Fig. 3A). We measured the promoter activities using green fluorescent protein (GFP) fusions of these regulators in the wild type and sarS deletion strains. Additionally, a rot translational fusion was tested in which the promoter and the first 35 amino acids of the coding sequence were fused to GFP. We observed a repression of both the rot transcription and rot translation fusions in the DsarS strain, demonstrating that SarS is an activator of rot expression (Fig. 3B). In contrast, sae promoter activity was only minimally increased when sarS was deleted in late stationary phase. We observed no detectable change in the activation of the rnaIII promoter in the sarS mutant background. To further investigate the impact of SarS on Rot and SaeR production, we performed Western blotting. We observed that in the absence of sarS, there was a notable decrease in the amount of Rot that S. aureus was producing at both the exponential and stationary phases (Fig. 3C). Complementation studies rescued Rot production, albeit to a slightly lower level than with the wild-type strain. These findings agree with the observed repression of the rot transcription and translation fusions in the DsarS strain. We saw no notable difference in the amount of SaeR in the wild-type, DsarS, or DsarS::sarS strain, which corroborated the GFP reporter data.
Having found that SarS can indirectly repress leukocidins through the activation of Rot, we wanted to determine if SarS could also directly repress toxins through the binding of their promoter regions. We performed DNA-binding experiments between the promoters of lukED, lukSF-PV, and hla and purified SarS. Biotinylated DNA probes were conjugated to streptavidin magnetic beads, and following incubation with Histagged SarS, binding was detected via immunoblotting. We included the promoter for purA as a control to which SarS was not expected to bind based on the literature (34,35). We found that SarS bound more strongly to the lukED, lukSF-PV, and hla promoters than to the negative-control purA promoter (Fig. 3D). Moreover, while nonbiotinylated purA DNA was able to only modestly outcompete binding of SarS to the lukSF-PV promoter, the SarS binding was competed off with nonbiotinylated lukSF-PV promoter DNA (Fig. 3E). Notably, SarS was able to bind to the lukSF-PV promoter significantly better than the lukAB or hlgA promoters (Fig. 3F), supporting the idea that SarS differentially regulates the leukocidins. Altogether, these results suggest that SarS can directly interact with the promoters of leukocidins, which may contribute to their repression.
Deletion of sarS enhances S. aureus virulence. Next we wanted to determine the role of SarS in the virulence potential of USA300. We measured the cytotoxicity of supernatants from the wild-type, sarS deletion, and complementation strains on primary human neutrophils. When sarS was deleted, there was significantly more killing of the neutrophils than with the wild-type strain, which we attributed to there being more toxins present in the supernatant (Fig. 4A). This phenotype was fully complemented in the DsarS::sarS strain. We also infected neutrophils with live wild-type, DsarS, and DsarS::sarS strains and measured neutrophil killing by the bacteria. Similar to the intoxication data from supernatants, we observed a significant increase in neutrophil death when the cells were infected with the DsarS strain compared to the wildtype or complementation strain (Fig. 4B).
To further understand the role of SarS in pathogenesis, an in vivo model of murine intraperitoneal infection was utilized. Mice were infected with the wild-type, DsarS, and DsarS::sarS strains and monitored for survival. Significantly more mice infected with the DsarS strain succumbed to the infection than the mice infected with the wildtype strain, a phenotype that was complemented by introducing a wild-type copy of sarS (DsarS::sarS) (Fig. 4C). Altogether, these data demonstrate that sarS is critical for USA300 pathogenesis.
Natural mutations in the sarS promoter result in altered sarS expression and cytotoxicity. The identification of SarS as a potent repressor of leukocidins prompted us to examine if clinical isolates differ in their sarS sequences, which could impact S. aureus pathogenesis. We analyzed the genomes of a series of contemporary USA300 bloodstream isolates whose cytotoxicity against human neutrophils had previously been characterized (46). In doing this, we discovered several low-cytotoxicity isolates that harbored mutations in a 16-bp region within the sarS promoter (Fig. 5A). The mutations localized downstream of the putative translational start site and the SarT binding sites and upstream of the gdpS mRNA binding site. This encompasses a stretch of DNA that contains both the promoter and the 59 untranslated region (UTR) of sarS. We hypothesized that these mutations resulted in an overexpression of sarS, which contributed to the observed decrease in cytotoxicity. To determine if this was the case, GFP reporter strains were constructed containing either these mutant promoters/59 UTRs (P1 to P4), the USA300 reference strain promoter/59 UTR (WT), or just the native sarS ribosomal binding sequence (RBS). We observed that several of the promoter/59 UTR mutations led to a greater GFP signal, suggesting that indeed these mutations resulted in higher expression of sarS than the wild-type nucleotide sequence (Fig. 5B).
Next, we constructed strains in which DsarS was complemented with sarS driven by either the wild-type or the P1 promoter, both located at the pathogenicity island SaPI1 (47). The P1 promoter was chosen because it showed the largest increase in GFP signal compared to the wild-type promoter in the reporter assay. To test that the P1 promoter resulted in increased sarS expression, qRT-PCR was performed. We observed that the P1 mutation resulted in a significant increase in sarS mRNA compared to the wild-type and wild-type complemented strains (Fig. 5C).
We also examined toxin expression by qRT-PCR and observed a decrease in lukED, lukSF-PV, and hla transcript levels in the DsarS SaPI1::sarS P1 strain compared to the DsarS SaPI1::sarS WT strain (Fig. 5D). Consistent with the expression levels, Western blots for LukE, LukF-PV, and alpha-toxin revealed less production of all three toxins in the DsarS SaPI1::sarS P1 compared to the DsarS SaPI1::sarS WT strain (Fig. 5E).
We next examined how the sarS P1 promoter mutation impacted S. aureus cytotoxicity toward immune cells. We collected supernatants from the complemented strains and intoxicated primary human neutrophils. Supernatants from the DsarS SaPI1::sarS Eight-week-old female B6 mice were infected intraperitoneally with wild-type AH-LAC or DsarS or DsarS::sarS S. aureus at a concentration of 1 Â 10 8 CFU/mouse and monitored for survival (n = 10 mice per group). Statistical analysis was performed using log rank Mantel-Cox test. *, P , 0.05; **, P , 0.01; ***, P , 0.001. P1 strain were significantly less cytotoxic than supernatants from the DsarS SaPI1::sarS WT strain (Fig. 5F), demonstrating that the decrease in leukocidin production due to the sarS promoter mutations that occur in the bloodstream resulted in altered virulence patterns.
USA300 isolates harbor mutations in the sarS promoter. Having established that natural mutations in the sarS promoter can impact USA300 virulence, we next investigated if mutations in the sarS promoter/59 UTR were widespread. We examined the sarS intergenic region and coding sequence across the USA300 genomes in GenBank for the presence of mutations (Fig. 6A). Mutations were broadly observed in the intergenic region (;35% of isolates), and mutations downstream of the 235 RNA polymerase binding site were also identified (;4% of isolates). These analyses also revealed that mutations in the sarS intergenic region are more pervasive than nonsynonymous mutations in the sarS gene (;35% versus ;2%) (Fig. 6B). Interestingly, we identified mutations in the 16-bp region of sarS DNA that harbors the P1, P2, P3, and P4 mutations (Fig. 6C). Additionally, there are mutations at the exact site of the P1, P3, and P4 mutations, further supporting the idea that mutations that alter sarS expression are being selected for in broader USA300 populations. These data indicate that mutations in the sarS promoter are not unique to our data set but widespread. The recurrence of Reporter strains in which GFP production was driven by the wild-type AH-LAC promoter or the mutant promoters (n = 4). (C) The wild-type and P1 promoters were cloned into the SaPI1 site in the DsarS::sarS strain and sarS transcript level was measured by qRT-PCR at exponential phase (n = 4). (D) Transcript levels of lukED, lukSF-PV, and hla were measured in the strains by qRT-PCR at exponential phase (n = 4). (E) Immunoblotting was performed for LukE, LukF-PV, and alpha-toxin. (F) Supernatants (2.5%) from the DsarS SapI1::sarS P1 strains were more cytotoxic than supernatants from the DsarS SapI1::sarS wild-type strain, as measured by CellTiter. Each data point represents an individual donor (n = 6). For all bar graphs, statistical analysis was performed using 2-way ANOVA with multiple comparisons. Error bars show standard errors of the mean. *, P , 0.05; **, P , 0.01; ***, P , 0.001; ****, P , 0.0001. Infection and Immunity these mutations across an extensive population suggests that expression of sarS could be a target of evolution in USA300.

DISCUSSION
The bicomponent pore-forming leukocidins are a critical component of the S. aureus virulon (6,48,49), but a complete understanding of the networks that regulate their expression is lacking. In this work, we describe a screen that combined transcriptional reporters with transposon mutants to identify novel regulators of leukocidins. Among the mutants identified, we characterized and established SarS as a key repressor of lukED and lukSF-PV expression ( Fig. 1 and 2). We demonstrated the impact of SarS in ex vivo and in vivo infection models and identified activation of Rot and direct binding to toxin promoters as two mechanisms by which SarS regulates toxins ( Fig. 3 and 4). Additionally, we found that mutations in the sarS promoter occur naturally in clinical USA300 isolates (Fig. 5), suggesting selective pressure to alter sarS expression during infection.
In terms of toxin regulation, SarS was initially identified as a repressor of hla, an effect that was dependent on the absence of sarA (34). Throughout this study we did not see a profound upregulation of hla in the absence of sarS, which could be because sarA was present in all strains used. While this initial study helped us understand the role of SarS in virulence regulation and shed light on how the Agr system and SarA mediate regulation of virulence factors (34), we now show that SarS is also a regulator of lukED and lukSF-PV.
It has been shown that the leukocidin promoters are differentially regulated in vitro (32), with the lukSF-PV and lukAB promoters being the most active during postexponential phase. This differential promoter activation could be in part due to differential regulation by repressors and activators such as SarS. Interestingly, we observed differential regulation of the leukocidins by SarS, in which lukED and lukSF-PV are highly repressed while the other bicomponent pore-forming leukocidins are not significantly regulated by SarS. It stands to reason that a pathogen that is able to successfully infect numerous environmental niches within the human body must be able to fine-tune its virulence factors to adapt to whatever tissue it is inhabiting (28). The ability of S. aureus to differentially regulate the expression of the leukocidins would therefore be beneficial to the adaptability of the pathogen.
We found that SarS can repress the leukocidins both indirectly through the activation of Rot and directly through the binding of the toxin promoters. It holds true to a trend seen in S. aureus virulence regulation, where both direct and indirect regulation of target genes is often observed. Such is the case with Rot, which can both directly and indirectly repress leukocidins through the binding of promoters and the repression of the SaeRS system (15,16,18). The observation that Rot and SarS are closely intertwined in their regulation is interesting, as Rot has been shown to be an activator of SarS (38). However, Hsieh et al. found rot to be upregulated in the absence of sarS, suggesting that SarS is a repressor of Rot (50). The experiments conducted by Hsieh et al. were performed in a different background, a derivative of S. aureus strain NCTC 8325 that was cured of all prophages (50,51), whereas the experiments conducted in our study were performed in a derivative of strain LAC, a representative of the CA-MRSA USA300 lineage. The many differences between these strains could account for the contrasting phenotypes observed with regard to rot regulation by SarS.
An interesting finding highlighted in this work is that USA300 clinical isolates harbor mutations in the sarS promoter that may lead to altered expression of sarS and SarS Regulates MRSA Pore-Forming Leukocidins Infection and Immunity decreased toxin expression and cytotoxicity. This finding sheds light on the importance of studying noncoding mutations to further understand S. aureus pathogenesis. These strains have likely evolved to thrive in patients, following an interesting trend seen in hospital-associated bloodstream infection: regulators of virulence factors are mutated to result in what may be less virulent strains in vitro. For example, the agr locus in hospital-associated MRSA isolates often have mutations that lead to attenuated agr activity (52)(53)(54). These observations highlight the need to understand how bacteria adapt under selective pressure in humans. The work presented herein demonstrates that one way S. aureus may adapt to infection is by altering cytotoxicity through the acquisition of mutations that impact sarS expression. There may be a give and take, a sacrifice of one type of virulence for the prioritization of other, that S. aureus must balance in order to become a successful pathogen. Alternatively, the bacterium may be adapting to better infect a certain type of host or to become more transmissible or have enhanced colonization. Hospitalized individuals often have impaired immune systems, and therefore, it may be advantageous for a nosocomial pathogen to downregulate certain virulence factors as to not overwhelm its host. Additionally, upregulation of virulence factors that aid in the adherence to medical devices can be beneficial in the hospital setting as well. While one role of SarS is to repress toxins, another key role is to activate the expression of spa. Protein A is a key virulence factor that promotes immune suppression and protects S. aureus from opsonophagocytic killing (55,56). Therefore, S. aureus may upregulate spa expression in vivo through sarS promoter mutations at the expense of toxin production. Altogether, the findings presented herein highlight the complex nature of toxin regulation in S. aureus. Additional studies are necessary to better understand the regulators, location, and host-derived signals at play during infection. We posit that such information, together with studies that incorporate contemporary clinical isolates that naturally differ in toxin regulation, are key to enable us to assemble a more complete picture of the regulation of toxins during infection and perhaps highlight potential opportunities for intervention.

MATERIALS AND METHODS
Ethics statement. Buffy coats were obtained from anonymous blood donors with informed consent from the New York Blood Center. All animal experiments were reviewed and approved by the Institutional Animal Care and Use Committee of New York University Langone Medical Center (NYULMC). All experiments were performed according to NIH guidelines, the Animal Welfare Act, and U.S. federal law.
Bacterial growth conditions. S. aureus strains (Table 1) were routinely grown at 37°C on tryptic soy agar (TSA) or in tryptic soy broth (TSB). Escherichia coli bacteria were grown in Luria-Bertani broth. Agar and broth were supplemented with antibiotics as needed to the following concentrations: erythromycin to 5 mg/ mL, chloramphenicol to 10 mg/mL (Cm10), and tetracycline to 4 mg/mL. Liquid cultures were grown in 5 mL of growth medium in 15-mL conical tubes and incubated at a 45°angle with shaking at 180 rpm, unless otherwise specified. For all experiments involving the growth of S. aureus, a 1:100 dilution of overnight cultures was subcultured into fresh medium.
Construction of bacterial strains and plasmids. For all the strains and oligonucleotides used in this study, see Tables 1 and 2.
(i) Reporter strains. (a) AH-LAC PlukSF_luc and regulatory library plus PlukSF_luc. The backbone of the PlukSF_luc plasmid originated from the plasmid pHC123 (33) and was cut at the SalI and KpnI cut sites before being ligated with the lukSF intergenic region and being transformed into DH5a and electroporated into AH-LAC. The primers pHC123_lukSF_F and pHC123_lukSF_R were used. The promoter reporter library was generated by phage transduction using phage 80a lysate from the AH-LAC PlukSF_luc strain.
The regulatory library was grown overnight in 400 mL of TSB in a round-bottom deep-well plate. In the morning, 390 mL of fresh TSB was inoculated with 10 mL of the overnight culture and grown at 120 rpm until an optical density at 600 nm (OD 600 ) of 1 was reached. Next, 5 mL of 1 M CaCl 2 and 100 mL of phage lysate were added to each well, and the plate was left at room temperature for 20 min. We added 40 mL of 1 M sodium citrate, and 10 mL was spot platted onto TSA plus Cm10 and grown overnight. Colonies were picked from this plate and grown overnight, and then 50 mL of the overnight culture was added to 50 mL of 20% glycerol and frozen down for further use.
(b) pSarS-WT-sGFP, pSarS-P1-sGFP, pSarS-P2-sGFP, pSarS-P3sGFP, pSarS-P4-sGFP, and pSarS-RBS-sGFP. To make sarS promoter reporter strains, the vector containing sgfp (superfolding GFP) was amplified from pOS1sGFP-PsarA-sodRBS (19). The vector was digested with SmaI (inside the SarS promoter sequence) and amplified in two pieces for Gibson assembly so that just the backbone and sgfp were present (no promoter or RBS). This was done using primers pOS1_1_F and pOS1_1_R for one piece and pOS1_2_F and pOS1_2_R for the other. For inserts PWT and P1 to P4, the entire intergenic region between sirC and sarS, with homology to pOS1, was amplified using primers pSarS_pOS1_F and pSarS_pOS1_R. Genomic DNA for amplification was used as follows: PWT, AH-LAC; P1, ER00594; P2, ER02658; P3, ER04127; and P4, ER05167. For RBS control, the oligonucleotides sarS_RBS_pOS1_F and sarS_RBS_pOS1_R were resuspended at 100 mM in annealing buffer (10 mM Tris [pH 7.5 to 8.0], 50 mM NaCl, 1 mM EDTA) and placed in a 94°C heat block for 2 min. The heat block was then turned off and cooled gradually (45 to 60 min), allowing for the oligonucleotides to anneal. For Gibson assembly, 2Â Gibson master mix (Fisher Scientific), the genomic insert, and the vector backbone were incubated at 50°C for 1 h. Following Gibson assembly, plasmids were heat transformed into E. coli IM08B and electroporated into AH-LAC.
(c) Psae-GFP, Prot-GFP, Prot35-GFP, and PrnaIII-GFP. Construction of the regulator promoter reporter strains was as previously described (32). The plasmids containing the GFP reporter fusions were electroporated into AH-LAC DsarS and selected for with 10 mg/mL of chloramphenicol.
(iii) Complementation strains. AH-LAC DsarS::sarS was made by cloning the sarS allele into the pIMAY* plasmid (45) and integrating it into the original site in the chromosome. The following primers were used to amplify upstream and downstream of sarS with 25 bp of homology to pIMAY* for Gibson cloning: Up_sarS_pIMAY_F and Down_sarS_pIMAY_R. For Gibson assembly, 2Â Gibson master mix (Fisher Scientific), the genomic insert, and the vector backbone were incubated at 50°C for 1 h. The plasmid was transformed into IMO8B and then electroporated into AH-LAC.
DsarS SaPI1::sarS wild type and DsarS SaPI1::sarS P1 were created by complementing the AH-LAC DsarS:: sarS strain using the pJC1111 plasmid (47), which integrates a single copy of the complementing gene into the SaPI1 site. sarS, along with the upstream and downstream intergenic regions, was amplified from AH-LAC (wild-type allele) or the clinical isolate containing the P1 mutation (ER00594). The following primers were used to amplify upstream and downstream of sarS, with cut sites at KpnI and XbaI, and 25 bp homology to pJC1111 for cloning with Gibson: Up_sarS_pJC1111_F and Down_sarS_pJC1111_R. For Gibson assembly, 2Â a VJT # is the number given to the strain in our lab database. Inclusion of this number makes it easier to find strains should other lab reach out and request them.

SarS Regulates MRSA Pore-Forming Leukocidins
Gibson master mix (Fisher Scientific), the genomic insert, and the vector backbone were incubated at 50°C for 1 h. Following Gibson assembly, the plasmids were transformed into DH5a and electroporated into AH-LAC DsarS.
Transcriptional regulator screen. The library was grown in a 1-mL volume in round-bottom deepwell block plates (Axygen; P-DW-20-C) in TSB plus 10 mg/mL of chloramphenicol. The following morning, the bacteria were subcultured 1:100 in a 1-mL volume in round-bottom deep-well block plates in TSB plus 5 mg/mL of chloramphenicol. The plates were incubated at 37°C with shaking at 180 rpm for 5 h. Next, 100 mL was removed and added to a clear-bottom round plate to measure OD and a whitebottom flat plate (Corning) to measure luminescence, which was assessed with a PerkinElmer EnVision A total of 0.15 mg of luciferin was added to each well of the plate in a 10-mL volume, the plate was incubated in the dark for 30 min, and then luminescence was measured. All luminescence values were OD normalized. RNA isolation. Bacteria were subcultured for 3 and 5 h in 5 mL of TSB before being spun down and resuspended in 1 mL of RNA Stat-60 (Amsbio). Samples were bead beated in lysing matrix B tubes (MP Bio) using the FastPrep and spun down for 10 min at 12,000 Â g. The upper layer was collected and 200 mL of chloroform was added. The samples were incubated at room temperature for 3 min before being spun down for 15 min at 12,000 Â g. The aqueous phase was removed, and 0.5 mL of isopropanol was added to precipitate RNA. RNA was washed twice with 70% ethanol, air dried, and resuspended in RNase-free water. RNA (10,000 ng) was DNase treated (TURBO DNA-fee kit; Invitrogen Ambion).
qRT-PCR. DNase-treated samples were converted into cDNA with the SuperScript III first-strand synthesis kit (Thermo Scientific). Next, 1 mL of cDNA was added to TaqMan probes and universal PCR master mix. Transcript levels were measured using the QuantStudio 3 system. Genes were normalized to the rpoB housekeeping gene and reported as threshold cycle (2 2DCT ). Probes used were PrimerTime 59 Hex and 39 BHQ-1.
Exoprotein isolation, Coomassie staining, and immunoblotting. The proteins in the culture supernatants of bacteria grown for either 3 or 5 h were precipitated and analyzed as previously described (16). Briefly, following OD normalization, the cultures were spun down and 1.4 mL of supernatant was added to 140 mL of trichloroacetic acid (TCA) and left overnight at 4°C. The precipitated proteins were sedimented, washed, dried, resuspended in 8 M urea, and left at room temperature for 30 min. Next, 2Â SDS loading buffer was added and the mixture was boiled for 10 min. Proteins were separated on a 12% SDS-PAGE gel, transferred to nitrocellulose membranes, and probed with indicated primary antibodies.
For the Rot and SaeR Western blots, 5-mL cultures were grown and spun down and the pellet was resuspended in lysis buffer (50 mM Tris, 10 mM MgCl 2 , 1 mM CaCl 2 ) with 5% lysostaphin, 1% HALT protease inhibitor, 1% RNase A, and 4% DNase and incubated at 37°C for 30 min. The samples were boiled for 10 min and then spun down at 15,000 Â g for 20 min. The top layer was filtered and 4Â SDS sample buffer was added before boiling the samples again.
GFP regulator reporter assay. GFP assay with the regulator promoters was performed as previously outlined by Balasubramanian et al. (32). Briefly, strains containing reporter plasmids were grown overnight in TSB with 10 mg/mL of chloramphenicol in round-bottom 96-well plates. The following morning, cultures were subcultured in 96-well black flat-bottom plates (Corning) and grown for the indicated time. Fluorescence was measured with a PerkinElmer EnVision 2130 multilabel reader.
SarS purification. Full-length sarS was amplified from AH-LAC, adding XhoI and BamHI cut sites, and cloned into pET15b. The plasmid was transformed into BL21-DE3g. The expression strain was grown in 400 mL of LB broth with 100 mg/mL of ampicillin at 37°C and 250 rpm until the cultures reached an OD 600 of 0.6. The culture was induced with 1 mM isopropyl-b-D-thiogalactopyranoside (IPTG) and grown for 4 h at 37°C and 250 rpm. The culture was spun down and resuspended in 20 mM Tris (pH 7.5), 300 mM NaCl, and 10% glycerol. Cells were lysed in 1Â protease inhibitor (Pierce protease inhibitor cocktail) and sonicated on ice. Bugbuster (Millipore) at 1Â was added, and the cultures were incubated at room temperature for 35 min. Following incubation on ice for 15 min and centrifugation, the supernatant was filtered through a 0.2-mm filter. The protein was purified using a HisTrap HP column on an AKTA system, eluting with a linear gradient in elution buffer (20 mM Na 2 HPO 4 , 500 mM NaCl, 400 mM imidazole [pH 7.4]). Purified protein was dialyzed in 10% glycerol in TSB.
Promoter pulldown assays. The promoter pulldown assay protocol was adapted from that of Sause et al. (57). We generated PCR products for each promoter using oligonucleotides containing biotinylated labels and purified the PCR products (Qiagen). M-280 streptavidin Dynabeads (Invitrogen; 11205D) were washed and resuspended in wash buffer (2 M NaCl, 1 mM EDTA, 10 mM Tris [pH 7.5]). The beads were incubated with 800 fmol of each DNA fragment for 30 min at room temperature on a rotisserie. The samples were washed three times with wash buffer before being resuspended in binding buffer (25 mM Tris-HCl, 0.1 mM EDTA, 75 mM NaCl, 10% glycerol, 1 mM dithiothreitol [DTT; pH 7.5]). SarS at a concentration of 100 nM and 5 mg of poly(dGÁdC) were added to the beads and mixed on a shaking platform for 15 min at 30°C and 550 rpm. For competition experiments, 800 fmol of nonbiotinylated DNA was mixed with the DNA-conjugated beads before the addition of SarS. Following incubation, beads were washed twice with binding buffer, resuspended in 1Â SDS sample buffer, and boiled for 10 min. Following electrophoresis, gels were transferred to a nitrocellulose membrane and probed with an anti-His antibody (1:2,000; Cell Sciences; CSI20563B) and detected with a fluorescent Alexa Fluor 680-conjugated anti-mouse antibody (1:25,000; Invitrogen).
Cytotoxicity assay and extracellular hPMN infections. Human polymorphonuclear leukocytes (hPMNs) were isolated using a Ficoll-Paque method and cytotoxicity assays were performed as described by Reyes-Robles et al. (58). Briefly, PMNs were seeded at 2 Â 10 5 cells per well in RPMI medium without phenol red (Fisher Scientific) supplemented with 10% heat-inactivated fetal bovine serum (FBS; Gemini Bio-Products). Either 1.25% or 2.5% supernatant was added (percentage is the percentage of total reaction volume that was culture supernatant). To measure cell viability, the metabolic dye CellTiter (Promega) was added at a final concentration of 10% per well and incubated for 2 h at 37°C and 5% CO 2 . Absorbance at 492 nm was measured using the PerkinElmer EnVision plate reader. For extracellular infections, bacteria were subcultured for 3 h in 5 mL of TSB, washed twice with PBS, and normalized to an OD 600 of 1. hPMNs were added to a flat-bottom tissue culture treated plate at 2 Â 10 5 and incubated at room temperature for 30 min. Bacteria were added at a multiplicity of infection (MOI) of 100. After a 2-h infection at 37°C and 5% CO 2 , hPMN viability was determined by lactate dehydrogenase (LDH) release (CytoTox-ONE homogenous membrane integrity assay; Promega), measured using the PerkinElmer EnVision plate reader.
Animal housing conditions. Animals received PicoLab rodent diet 20 (LabDiet) and acidified water and were housed under normal lighting cycle conditions (12 h on/12 h off) and a temperature of 70°F.
Murine intraperitoneal infection. Bacteria were subcultured for 3 h in 5 mL of TSB, washed twice in PBS, and OD normalized. Eight-week-old C57BL/6J female mice (Jackson Laboratory) were injected with 200 mL of bacteria intraperitoneally. Mice were monitored multiple times a day for 48 h and euthanized upon severe signs of mortality or excessive weight loss.
Analysis of USA300 genomes in GenBank. A total of 45,270 S. aureus genomes from human biosamples were downloaded in April 2022 using NCBI data sets version 13.7.0. Genomes were deduplicated such that only the first genome per biosample was retained for further analysis. In addition, biosamples and genomes originating from the Mount Sinai Health System were removed, leaving 26,792 genomes after filtering. The multilocus sequence type (MLST) of each genome was determined using mlst (https://github.com/ tseemann/mlst) and PubMLST (59) and used to assign the USA group of each isolate as defined by David et al. (60). To obtain a comprehensive list of mutations seen in sarS and the sarS promoter, the sequence of USA300 FPR3757 (GenBank accession number NC_007793) was used as the wild type and aligned using BLASTn. Subsequently, each of the 26,792 GenBank genomes was compared (61). For all the USA300 GenBank isolates, core gene MLST (cgMLST) schemes for each were determined using chewBBACA (version 2.8.5) (62) and the cgMLST scheme available at https://cgmlst.org/ncs (63). cgMLST trees were visualized using GrapeTree (64).
sGFP reporter assay. Bacteria were subcultured from overnight cultures 1:100 in 5 mL of TSB plus 5 mg/ mL of chloramphenicol. After 5 h of growth, cultures were spun down and resuspended in PBS. Samples were serially diluted 1:2 Â 6 (samples were serially diluted 1:2 six times) in 100 mL in a 96-well flat-bottom plate. GFP signal was determined using a PerkinElmer plate reader. Background signal was subtracted from PBS-only controls.
Statistical methods. Prism software (GraphPad, Inc.) was used to perform statistical analysis. Twoway analysis of variance (ANOVA) was utilized for qRT-PCR, promoter pulldown assays, GFP reporter assay, cytotoxicity, hPMN infection, luminescence screen hits, and sGFP assays. The statistical significance of difference between survival curves was determined by the log rank Mantel-Cox test.

SUPPLEMENTAL MATERIAL
Supplemental material is available online only. SUPPLEMENTAL FILE 1, XLSX file, 0.03 MB.