The Impacts of msaABCR on sarA-Associated Phenotypes Are Different in Divergent Clinical Isolates of Staphylococcus aureus

The staphylococcal accessory regulator (sarA) plays an important role in Staphylococcus aureus infections, including osteomyelitis, and the msaABCR operon has been implicated as an important factor in modulating expression of sarA. Thus, we investigated the contribution of msaABCR to sarA-associated phenotypes in the S. aureus clinical isolates LAC and UAMS-1. Mutation of msaABCR resulted in reduced production of SarA and a reduced capacity to form a biofilm in both strains.

M utation of the staphylococcal accessory regulator (sarA) attenuates the virulence of divergent clinical isolates of Staphylococcus aureus in animal models of bacteremia, postsurgical osteomyelitis, and infective endocarditis (1)(2)(3). It also limits biofilm formation in vitro and in vivo to a degree that can be correlated with increased antibiotic susceptibility (2,(4)(5)(6). The effector molecule of the sarA regulatory system is a 15-kDa protein that has been shown to impact the production of multiple S. aureus virulence factors at a transcriptional level and by modulating the stability of mRNA (7)(8)(9)(10)(11)(12). We have also demonstrated that an important factor contributing to the reduced virulence of sarA mutants, and their reduced capacity to form a biofilm, is the increased production of extracellular proteases and resulting decrease in the accumulation of multiple S. aureus proteins, including both surface-associated and extracellular virulence factors (1,(13)(14)(15)(16)(17).
Thus, the sarA regulatory locus impacts both the production and the accumulation of S. aureus virulence factors, and this collectively makes an important contribution to diverse phenotypes that contribute to pathogenesis. This makes sarA a potential therapeutic target, and efforts have been made to exploit sarA in this regard (17)(18)(19).
However, S. aureus regulatory circuits are complex and highly interactive (20), and mutation of other S. aureus regulatory loci within this circuit has also been shown to increase protease production to a degree that limits biofilm formation (21)(22)(23)(24)(25).
Among these other loci is msa (modulator of sarA), mutation of which was originally reported to limit the expression of sarA and the production of SarA itself (26). The msa gene was identified in the 8325-4 strain RN6390 by a transposon insertion in the open-reading frame SA1233 as designated in the N315 genome, but it was subsequently shown to be part of a four-gene operon now designated msaABCR (27). Genes within the msa operon encode a putative protein (MsaA) with no known function, a DNA binding protein (MsaB) shown to act as a transcription factor that regulates expression of numerous genes, and a regulatory RNA (msaC) and an antisense RNA (msaR) complementary to msaB (27). As would be expected based on the phenotypes of sarA mutants (3,4,13,15,16,28) and the role of msaABCR in enhancing expression of sarA, mutation of msaABCR (hereinafter referred to as msa) has been correlated with increased protease production and a decreased capacity to form a biofilm (25,27,29).
Mutation of msa was also reported to result in decreased expression of the accessory gene regulator (agr) in the 8325-4 strain RN6390 but to have the opposite effect in the clinical isolate UAMS-1 (26). Expression levels of the well-characterized agr-regulated genes encoding alpha toxin (hla) and protein A (spa) also differed between these two strains, while expression of the genes encoding aureolysin (aur) and SspA (sspA) was increased in both strains. Differences between these two strains have also been observed in the phenotype of their isogenic sarA mutants (30,31). Such reports are not surprising given that RN6390 has a mutation in rsbU that impacts the sigB regulatory pathway (32), which has also been shown to impact expression of both agr and sarA as well as protease production (33,34). However, significant differences also exist among clinical isolates, and to date, such strain-dependent differences have not been adequately investigated. Thus, the overall impact of msa in divergent clinical isolates, and the extent to which it is dependent on its interaction with sarA, remains unclear. In this report, we addressed these issues by generating msa, sarA, and msa sarA mutants in the methicillin-resistant USA300 strain LAC and the methicillin-sensitive USA200 strain UAMS-1 and assessed the impact these mutations had on well-defined phenotypes associated with their isogenic sarA mutants.

RESULTS AND DISCUSSION
Impacts of msa on sarA expression. Using an anti-SarA antibody (35), we first assessed the production of SarA in msa mutants generated in LAC and UAMS-1 by Western blotting. Experiments were performed using whole-cell lysates prepared from equal numbers of CFU harvested from cultures in the mid-, late-, and post-exponential growth phases. The results were comparable in both strains (Fig. 1) and confirmed that mutation of msa results in reduced production of SarA, particularly during the mid-and late-exponential growth phases. However, while the differences in the abundance of SarA were in most cases statistically significant, they were also modest in that the amount of SarA present in lysates prepared from LAC and UAMS-1 msa mutants was consistently Ͼ50% of that observed in the isogenic parent strain irrespective of growth stage. This is consistent with transcriptional analysis, which demonstrated that mutation of msa results in a modest but statistically significant decrease in the levels of sarA transcripts in both LAC and UAMS-1 compared to that in the isogenic parent strain (Table 1). These studies also confirmed that this transcriptional phenotype could be genetically complemented. These results are consistent with the hypothesis that msa functions upstream to modulate the expression of SarA.
Impact of msa on biofilm formation. Thus, the important question becomes whether the reduction in the amount of SarA observed in msa mutants is phenotypically relevant. One of the primary phenotypes that define sarA mutants in divergent clinical isolates, including LAC and UAMS-1, is the reduced capacity to form a biofilm (36). Using a well-established microtiter plate assay (28), we confirmed that mutation of msa limits biofilm formation in both LAC and UAMS-1 but to a limited extent compared to that of the isogenic sarA mutants (Fig. 2). The relative impact of mutating msa versus sarA was confirmed by demonstrating that concomitant mutation of both msa and sarA limited biofilm formation to a level comparable to that observed in the isogenic sarA mutant and well below that observed in the corresponding msa mutant (see Fig. S1 in the supplemental material). These results are also consistent with the hypothesis that msa is upstream of SarA and the observation that mutation of msa had only a modest impact on the accumulation of SarA, but they also suggest that the reduced amount of SarA observed in msa mutants is phenotypically relevant in the context of biofilm formation.
If this is true, then restoring the production of SarA in an msa mutant should restore biofilm formation. To investigate this, we introduced the same plasmid (pSARA) used to genetically complement the sarA mutation into an msa mutant. Western blot analysis confirmed that the accumulation of SarA was restored in both LAC and UAMS-1 msa mutants (Fig. 3). Introducing pSARA also restored biofilm formation in a LAC msa mutant but not in a UAMS-1 msa mutant (Fig. 2). The reasons for this strain-dependent difference are unclear, but these results suggest that msa limits biofilm formation in UAMS-1 owing to a sarA-independent regulatory effect.  Impact of msa on protease production. To investigate the mechanistic basis for these biofilm phenotypes, we examined the relative impact of mutating sarA and msa on the production of extracellular proteases. This was based on our previous demonstration that the increased production of extracellular proteases plays a key role in defining the biofilm-deficient phenotype of S. aureus sarA mutants (1). In LAC, mutation of msa resulted in a statistically significant increase in overall protease activity as FIG 2 Impacts of msa and sarA on biofilm formation. Biofilm formation was assessed with LAC, UAMS-1, their sarA and msa mutants, and mutants complemented with sarA ( S ) or msa ( M ). Bar charts represent cumulative results from at least two biological replicates, each of which included five experimental replicates. Error bars indicate standard errors of the means. *, statistical significance relative to the isogenic parent strain; **, statistical significance relative to the isogenic sarA mutant.

FIG 3
SarA accumulation in sarA-and msa-complemented mutants. SarA accumulation was assessed by Western blotting of whole-cell lysates prepared from mid-exponential-phase cultures of LAC, UAMS-1, their sarA and msa mutants, and mutants complemented with sarA ( S ) or msa ( M ). Bar charts illustrate densitometry based on at least two experimental replicates. Densitometry was performed using samples prepared from cells obtained at mid-exponential growth phase (standardized to OD 560 of 1.5). Error bars indicate standard errors of the means. *, statistical significance relative to the isogenic parent strain; **, statistical significance relative to the isogenic sarA mutant. assessed using both casein-and gelatin-based fluorescence resonance energy transfer (FRET) assays, although the impact was more evident in the casein-based assay than in the gelatin-based assay (Fig. 4). This was not true in a LAC sarA mutant, where the impact of mutating sarA on protease production was readily evident in both assays (Fig.  4). Additionally, restoring SarA production in a LAC msa mutant decreased protease production, in the case of the casein-based assay, to wild-type levels. As might be expected based on the relative sensitivity of the two assays, this was most evident when assessed using the casein-based assay. However, mutation of msa in UAMS-1 did not have a significant impact on overall protease activity as assessed using either casein-or gelatin-based FRET assays (Fig. 4). As in LAC, mutation of sarA in UAMS-1 resulted in a statistically significant increase in protease production in both protease assays. These results are also consistent with the hypothesis that the impact of mutating msa on biofilm formation in UAMS-1 occurs via a sarA-independent regulatory effect.
This strain-dependent difference was also apparent in assays employing gfp transcriptional reporter constructs generated with the promoters from each of the genes and/or operons encoding S. aureus extracellular proteases (aur, splA-F, sspABC, and scpAB). Specifically, expression levels from all four reporters were significantly increased in a LAC msa mutant but not to the level observed in the isogenic sarA mutant (Fig. 5). In contrast, fluorescence was not increased to a significant extent in a UAMS-1 msa mutant with any reporter other than the scp::gfp, and even then, the increase was modest by comparison to fluorescence levels observed with the same reporter in the LAC msa mutant and with all four reporters in the UAMS-1 sarA mutant (Fig. 5). These results suggest that the strain-dependent impact of msa on protease production is mediated at a transcriptional level.
These results also suggest the possibility of a cause-and-effect relationship between increased protease production and decreased biofilm formation in a LAC msa mutant. Indeed, there was an inverse and proportional relationship between protease production and biofilm formation in LAC and its isogenic sarA, msa, and sarA msa mutants (see Fig. S2). However, this inverse relationship was not apparent in a UAMS-1 msa mutant. Mutation of msa in LAC also resulted in the decreased accumulation of both Hla and extracellular protein A (eSpa) (Fig. 6). In contrast, in UAMS-1, which does not produce Hla, the accumulation of eSpa was greatly reduced in a sarA mutant but not in the isogenic msa mutant. The reduced accumulation of eSpa observed in a LAC msa mutant was reversed by eliminating the production of extracellular proteases, while in a UAMS-1 msa mutant, the abundance of eSpa was not affected by the inability to produce these proteases (Fig. 6).
These results demonstrate that mutating msa results in a significant increase in protease production in LAC but not in UAMS-1. SDS-PAGE analysis of conditioned medium (CM) from overnight cultures confirmed the decreased accumulation of highmolecular-weight (HMW) proteins in a LAC msa mutant and that this was reversed by eliminating the production of extracellular proteases (Fig. 7). As would be expected based on the results discussed above, this effect was not apparent in a UAMS-1 msa mutant. In contrast, mutation of sarA limited the accumulation of HMW proteins in CM in both LAC and UAMS-1, and in both cases, this was reversed by eliminating the ability of these mutants to produce extracellular proteases (Fig. 7).  Impact of msa on PIA production. To examine other possibilities, we assessed the production of the polysaccharide intracellular adhesin (PIA) in msa and sarA mutants. PIA is known to contribute to biofilm formation, and it has been suggested that it plays a particularly important role in methicillin-sensitive strains such as UAMS-1 (37). However, we were unable to detect PIA above background levels in LAC, UAMS-1, or their isogenic sarA and msa mutants (see Fig. S3).
Impact of msa on extracellular nuclease. Extracellular DNA and the production of extracellular nucleases have also been implicated in biofilm formation in both methicillin-resistant and methicillin-sensitive strains (38). S. aureus produces at least two nucleases, one of which (Nuc1) is a secreted extracellular protein, while the other (Nuc2) remains bound to the cell surface (39). Mutation of sarA in UAMS-1 has been shown to result in the increased production of these nucleases, and at least under in vitro conditions, this has been shown to limit biofilm formation (40). Based on this, we examined the impact of mutating msa on nuclease production with a specific focus on the Nuc1 extracellular nuclease. This was facilitated by the availability of an anti-Nuc1 antibody (16), which allowed us to investigate this issue using Western blots of CM harvested from overnight cultures of each strain. It is important to recognize that Nuc1 is produced in two forms, the smaller of which (NucA) is proteolytically derived from the larger (NucB), and both of which are enzymatically active (41).
Relative to the parent strain, Nuc1 was present in increased amounts in a UAMS-1 sarA mutant, and all of the Nuc1 present that could be detected by Western blotting was present in the smaller NucA form (Fig. 8). This suggests that the increased  production of extracellular proteases in a UAMS-1 sarA mutant can be correlated with the absence of NucB. This was confirmed in Western blots with CM from a sarA mutant unable to produce these proteases, in which case, all of the Nuc1 detected was in the NucB form. Moreover, the overall abundance of Nuc1 was increased in the proteasedeficient UAMS-1 sarA mutant compared to that in the sarA mutant (Fig. 8). The abundance of Nuc1 was also increased in a UAMS-1 msa mutant, and in this case, both NucA and NucB were detectable by Western blotting. While the overall amount of Nuc1 was not increased in a protease-deficient UAMS-1 msa mutant, all of the Nuc1 present was in the larger NucB form. This could be interpreted to suggest that mutation of msa does result in an increase in protease production in UAMS-1 that is phenotypically apparent, but we believe this would be an overinterpretation in that, unlike in the isogenic protease-deficient sarA mutant, the amount of Nuc1 did not increase appreciably in the UAMS-1 protease-deficient msa mutant (Fig. 8).
The increased abundance of Nuc1 observed in a UAMS-1 sarA mutant was not apparent in a LAC sarA mutant, but it was apparent in the isogenic msa mutant (Fig. 8). Unlike in the UAMS-1 msa mutant, all of the Nuc1 detectable by Western blotting in the LAC msa mutant was present in the smaller NucA form. This is consistent with the observation that mutating msa had a significant impact on protease production in LAC but not in UAMS-1. As with the UAMS-1 protease-deficient sarA and msa mutants, only NucB was detected in CM from the protease-deficient LAC sarA and msa mutants (Fig.  8). As with a UAMS-1 msa mutant, eliminating protease production in a LAC msa mutant limited proteolytic processing of Nuc1 but did not appreciably alter the overall amount. In contrast, the abundance of NucB was also enhanced in a protease-deficient LAC sarA mutant compared to that in the isogenic sarA mutant itself. These results demonstrate that the production of Nuc1 is increased in LAC and UAMS-1 sarA and msa mutants. They also indicate that the abundance of Nuc1 is limited by increased protease production in sarA mutants generated in both strains but that this is not the case in LAC msa mutants. However, the impact of msa on protease production was still evident in a LAC msa mutant, in that all of the Nuc1 present was present in the smaller NucA form (Fig. 8).
Impacts of protease and nuclease production on biofilm formation. Given these overlapping protease and nuclease phenotypes, we directly examined the impacts of eliminating the production of extracellular proteases or Nuc1 on the biofilm-deficient phenotype of LAC and UAMS-1 sarA and msa mutants. In both strains, eliminating the ability to produce extracellular proteases enhanced biofilm formation in both sarA and msa mutants to levels comparable to those observed in the isogenic parent strain (Fig.  9). This could be interpreted to suggest that the increased production of extracellular proteases limits biofilm formation in msa mutants, even in UAMS-1. However, it is important to note that eliminating protease production also enhanced biofilm formation in UAMS-1 itself to a greater extent than in LAC (Fig. 9). In fact, the increase in biofilm formation observed in a protease-deficient derivative of UAMS-1 was comparable to that observed in the UAMS-1 msa mutant, and this was not the case in the same derivatives of LAC. Thus, we believe these results are also consistent with the conclusion that the increased production of extracellular proteases limits biofilm formation in a LAC msa mutant but not in a UAMS-1 msa mutant.
Biofilm formation was also enhanced in LAC and UAMS-1 msa mutants unable to produce Nuc1, but once again, these results must be interpreted with caution, because eliminating the production of Nuc1 also enhanced biofilm formation in the LAC and UAMS-1 parent strains (Fig. 9). As with protease production, the increase in biofilm formation observed in the nuclease-deficient UAMS-1 msa mutant was less than that observed in the nuclease-deficient LAC msa mutant, and this was reflected in the relative impact of eliminating Nuc1 production on biofilm formation (Fig. 9). In contrast, eliminating the production of Nuc1 did have a significant impact on biofilm formation in a UAMS-1 sarA mutant but not in a LAC sarA mutant (Fig. 9). This is consistent with the observation that mutation of msa resulted in an increase in the abundance of Nuc1 in a UAMS-1 sarA mutant but not in a LAC sarA mutant, although as previously discussed, protease production was shown to limit the abundance and processing of Nuc1 in sarA mutants generated in both strains.
Impact of msa on staphyloxanthin production. All of the results discussed above are consistent with a model in which msa functions upstream to enhance the production of SarA but also demonstrate that the impact of mutating msa on sarA-associated phenotypes is strain dependent. There are also reports implicating mutation of msa in LAC in phenotypes that have not been previously associated with sarA. One of these is that mutation of msa in LAC results in the reduced production of staphyloxanthin (27), which has been implicated as an important virulence factor in S. aureus (42). We examined this in LAC and UAMS-1 sarA and msa mutants, and the results confirmed that mutation of msa in LAC results in a statistically significant reduction in the production of staphyloxanthin (Fig. 10) and consequently reduced pigmentation of colonies on agar plates (data not shown). Importantly, unlike the relative impacts of mutating sarA and msa on biofilm formation and protease production, the impact of mutating msa exceeded that of mutating sarA in this regard, thus suggesting that the impact of mutating msa on staphyloxanthin production is primarily independent of its impact on sarA. In UAMS-1, the results of these assays provided an even more striking contrast. Specifically, staphyloxanthin production was increased in a UAMS-1 sarA mutant but decreased in the isogenic msa mutant (Fig. 10). Although the decrease observed in a UAMS-1 msa mutant was not statistically significant, this contrast nevertheless makes it evident that the impact of mutating msa on staphyloxanthin production in UAMS-1 is independent of its impact on sarA.
Impact of msa in osteomyelitis. The results discussed above provide insight into the impact of msa on sarA-associated phenotypes in divergent clinical isolates of S. aureus. However, they also suggest, specifically, with respect to our staphyloxanthin assays, that msa serves regulatory functions that are independent of its impact on sarA. Moreover, all of these results are based on in vitro assays that do not necessarily reflect the unique microenvironment of the bone. Thus, we wanted to directly assess the relative contribution of msa and sarA to virulence in our murine osteomyelitis model (3,43). As previously reported (3), mutation of sarA limited virulence in both strains as assessed by reactive bone formation and cortical bone destruction, although in this experiment, the reduction in cortical bone destruction observed with the UAMS-1 sarA mutant did not reach statistical significance (Fig. 11). By comparison, mutation of msa had only a modest impact on virulence in LAC, particularly in the context of cortical bone destruction, and it had no significant impact in UAMS-1 in either reactive bone formation or cortical bone destruction.
Conclusions. Most reports describing the impact of S. aureus regulatory loci on clinically relevant phenotypes, including virulence, are based on examination of single loci in a single strain, and this makes it difficult to reach conclusions regarding the relative potential of different regulatory loci as therapeutic targets. We have attempted to address this by directly comparing different regulatory mutants generated in divergent clinical isolates of S. aureus using both in vitro and in vivo assays (3,4,44). The results of these studies have led us to focus on sarA and to hypothesize that a primary factor contributing to the impact of mutating sarA on virulence and virulenceassociated phenotypes is the increased production of extracellular proteases and the limitation this imposes on the accumulation of both surface-associated and extracellular virulence factors (1,16). To date, we have not included the msaABCR operon in these studies, and it is important to do so given that msa has been shown to function upstream of sarA and to impact sarA-associated phenotypes, including biofilm formation and protease production (25)(26)(27)29). This raises the possibility that msa could also be a viable therapeutic target. Experimentally addressing this possibility was the focus of the experiments we report. However, the results we report lead us to conclude that this is not the case for two reasons. First, even in the genetically and phenotypically divergent clinical isolates LAC and UAMS-1, the impacts of mutating msa on biofilm formation and virulence in our osteomyelitis model were limited in comparison to FIG 10 Staphyloxanthin production in sarA and msa mutants. Pigment was extracted from standardized samples of bacteria grown to stationary phase and measured at an absorbance of 465 nm. Bar charts represent cumulative results from at least four biological replicates, each of which included three experimental replicates. Error bars indicate standard errors of the means. *, statistical significance relative to the isogenic parent strain; **, statistical significance relative to the isogenic sarA mutants. those of mutating sarA. Second, the relative impacts of mutating msa differed between these two strains with respect to both of these phenotypes. This emphasizes the need for direct comparative studies like those we report, particularly given the complexity of S. aureus regulatory circuits and the diversity among S. aureus strains as represented by the USA300 isolate LAC and the USA200 strain UAMS-1.
Preparation of S. aureus conditioned medium. To prepare conditioned medium (CM), cultures of each strain were grown overnight (16 h) in TSB at 37°C with constant shaking. The optical density at 560 nm (OD 560 ) of each culture was determined, and fresh TSB was added to standardize each culture to an equivalent optical density. Cells were then removed by centrifugation and CM prepared by filter sterilization. Samples were stored at Ϫ80°C until used.
Preparation of whole-cell lysates. Whole-cell lysates were prepared as previously described with minor modification (45). Briefly, strains were cultured at 37°C in TSB with constant shaking and a medium-to-flask ratio of 0.5. Bacterial cells from a volume of each culture calculated to obtain an equivalent number of cells were harvested by centrifugation at an OD 560 of approximately 1.5, 4.0, and 10.0, which correspond to the mid᎑exponential, late᎑exponential, and post᎑exponential growth phases, respectively. Cells were washed with sterile phosphate-buffered saline (PBS) and resuspended in 750 l of TEG buffer (25 mM Tris᎑HCl [pH 8.0], 25 mM EGTA). Cell suspensions were stored at Ϫ20°C until all samples had been collected, at which point samples were thawed on ice, transferred to Fastprep Lysing Matrix B tubes, and lysed in a FastPrep-24 benchtop homogenizer (MP Biomedicals) using two 40-s intervals at a rate of 6.0 m/s interrupted by a 5-min interval during which the homogenates were chilled on ice. After centrifugation at 15,000 ϫ g for 10 min at 4°C, supernatants were harvested and stored at Ϫ80°C.

FIG 11
Impacts of sarA and msa on the virulence of LAC and UAMS-1 in an osteomyelitis model. Images were analyzed for cortical bone destruction and reactive (new) bone formation in C57BL/6 mice infected with LAC, UAMS-1, or their isogenic sarA and msa mutants. Values are presented as volumes relative to those in mock-infected mice which underwent the surgical procedure but were injected only with sterile PBS. At least ten mice were analyzed for each mutant or respective parent strain. *, statistical significance relative to the isogenic parent strain; **, statistical significance relative to the isogenic sarA mutant.
Western blotting. SarA Western blotting was performed with an anti-SarA antibody and appropriate secondary antibodies, as previously described (1,15,16). Western blots included at least two biological replicates. Densitometric values were obtained with a Bio-Rad ChemiDoc MP imaging system and Image Lab software (Bio-Rad Laboratories).
RNA isolation and real-time qPCR. Overnight cultures of S. aureus were diluted 1:10 in fresh TSB and incubated at 37°C with shaking (200 rpm) for 2 h. The cells were then normalized to an OD 600 of 0.05 in 25 ml TSB in a 125-ml conical flask and incubated at 37°C with shaking (200 rpm). The cells were collected at mid-exponential growth phase. Total RNA was isolated from cells using a Qiagen RNeasy Maxi column (Qiagen), as previously described (27). The quality of total RNA was determined by NanoDrop spectrometer readings, and 1 g RNA was used to synthesize cDNA using iScript Reverse Transcription Supermix for reverse transcription-quantitative PCR (RT-qPCR) (Bio-Rad). RT-qPCR was performed using iTaq Universal SYBR Green Supermix (Bio-Rad) as described previously (27). The constitutively expressed gyrase A (gyrA) gene was used as an endogenous control gene and was included in all experiments. The following primer sequences were used to measure sarA expression: RT-sarA-F, TTTGCTTCAGTGATTCGTTTATTTACTC, and RT-sarA-R, GTAATGAGCATGATGAAAGAACTGTATT. Analysis of expression of each gene was conducted based on at least three biological replicates.
Static in vitro biofilm assay. Biofilm formation was assessed in vitro using a microtiter plate assay as previously described (28). Briefly, sterile 96-well microtiter plates were coated with 100 l of 20% carbonate-bicarbonate-reconstituted human plasma (Sigma) and incubated overnight at 4°C. Bacterial cultures were grown overnight in TSB supplemented with 3% sodium chloride and 0.5% glucose (biofilm medium [BFM]) at 37°C. Cultures were standardized to an OD 560 of 0.05 in fresh BFM. Plasma was gently aspirated, and the microtiter plate was inoculated with 200 l of standardized culture per well. The plate was incubated statically overnight at 37°C. Wells were gently washed three times with 200 l PBS, fixed with 200 l 100% ethanol (EtOH), stained with 200 l Gram's crystal violet, and finally washed three times with 250 l PBS. The stain was eluted with 100 l 100% EtOH for 10 min, the eluent was diluted into a new 96-well plate, and the absorbance was measured at 595 nm with a FLUOstar Omega microplate reader (BMG Labtech).
Total protease activity. Total protease activity of CM was assessed using the FRET-based protease fluorescent detection kit (Sigma) and the EnzChek gelatinase/collagenase assay kit (Thermo Fisher Scientific), both according to the manufacturer's instructions.
Protease reporter assay. Strains carrying each protease reporter (pCM13, pCM15, pCM16, or pCM35) were cultured in TSB overnight as detailed above. Cultures were then standardized to an OD 560 of 10.0. Two hundred microliters of each standardized culture was then aliquoted in triplicates into a black clear-bottomed 96-well plate, and the mean fluorescence intensity (MFI) was measured with a FLUOstar Omega microplate reader (excitation, 485 nm; emission, 520 nm) (BMG Labtech).
PIA immunoblot. Production of the polysaccharide intercellular adhesin (PIA) was assessed as previously described with minor modifications (44). Specifically, cultures were grown overnight in BFM. After standardization to OD 560 of 5.0, cells were harvested by centrifugation and resuspended in 60 l 0.5 M EDTA. Cell suspensions were boiled for 5 min followed by centrifugation (14,000 ϫ g for 2 min). Forty microliters of the supernatant was then incubated for 30 min at 48°C with 1 l proteinase K (10 mg/ml). Twenty microliters of Tris-buffered saline (20 mM Tris-HCl, 150 mM NaCl [pH 7.4]) was added to each sample, which was then stored at Ϫ20°C. For analysis, 2 l of each sample was spotted directly to a dry nitrocellulose membrane, and PIA was detected by using an anti-PIA antibody as previously described (44). Characterization of exoprotein profiles. Exoprotein profiles were examined as previously described (1). CM harvested as described above was resolved by SDS-PAGE using 4% to 12% gradient Novex Bis-Tris Plus gels (Life Technologies). Proteins were visualized by staining with SimplyBlue SafeStain (Life Technologies). Images were obtained using a Bio-Rad ChemiDoc MP imaging system (Bio-Rad Laboratories).
Staphyloxanthin production. The relative production of staphyloxanthin was assessed using bacterial cells harvested from overnight cultures as previously described (27). Briefly, cells were harvested and standardized to an OD 560 of 10.0 and washed twice with sterile water. Cells were then resuspended in 1.0 ml of 100% methanol and heated at 55°C for 5 min with occasional vortexing. The cells were removed by centrifugation at 15,000 ϫ g for 1 min, and 100 l of supernatant was placed into a 96-well microtiter plate in triplicates. Absorbance values were read on a FLUOstar Omega microplate reader (BMG Labtech) at 465 nm and background corrected with a methanol blank.
Murine model of posttraumatic osteomyelitis. The murine model of acute posttraumatic osteomyelitis was performed as previously described (43). Prior to surgery, 8-to 10-week-old C57BL/6 mice received 2.0 mg/kg of body weight meloxicam via subcutaneous injection and were then anesthetized with isoflurane for the duration of the surgery. For each mouse, an incision was made above the right hind limb. The periosteum was pulled apart with forceps, and using a 21-gauge Precision Glide needle (Becton, Dickinson), a 1-mm uni-cortical bone defect was made at the lateral midshaft of the femur. A bacterial inoculum of 1 ϫ 10 6 CFU in 2 l of PBS was delivered into the intramedullary canal. The periosteum and skin were then closed with sutures, and the mice were allowed to recover from anesthesia. Infection was allowed to proceed for 14 days thereafter, at which time the mice were euthanized and the right femur was removed and subjected to micro-computed tomography (micro-CT) analysis. All experiments involving animals were reviewed and approved by the Institutional Animal Care and Use Committee of the University of Arkansas for Medical Sciences and were performed according to NIH guidelines, the Animal Welfare Act, and U.S. federal law.
Micro-computed tomography. The analysis of cortical bone destruction and new bone formation was performed using micro-CT imaging with a Skyscan 1174 micro-CT (Bruker), and scans were analyzed using the manufacturer's analytical software. Briefly, axial images of each femur were acquired at a resolution of 6.7 m at 50 kV and 800 A through a 0.25-mm aluminum filter. Bones were visualized using a scout scan and then scanned in three sections as an oversize scan to image the entire femoral ΔmsaABCR, pSARA length. The volume of cortical bone was isolated in a semiautomated process per the manufacturer's instructions. Briefly, cortical bone was isolated from soft tissue and the background by global thresholding (low threshold, 89; high threshold, 255). The processes of opening, closing, dilation, erosion, and despeckling were configured using the bones from sham-treated controls to separate the new bone from the existing cortical bone, and a task list was created to apply the same process and values to all bones in the data set. After processing of the bones using the task list, the volume of interest (VOI) was corrected by drawing inclusive or exclusive contours on the periosteal surface. Cortical bone destruction analysis consisted of approximately 1,800 slices between anatomical landmarks at opposing ends of the femur. Destruction was determined by subtraction of the volume of infected bones from the average bone volume from sham-treated controls. Reactive new bone formation was assessed by first isolating the region of interest (ROI) that contained only the original cortical bone (as described above). After cortical bone isolation, the new bone volume was determined by subtracting the cortical bone volume from the total bone volume. All calculations were performed on the basis of direct voxel counts. Statistical analysis. To allow for statistical comparison across biological and experimental replicates, the results obtained for each experimental replicate with each strain were averaged across all biological replicates. For densitometric analyses of Western blots, protease assays, biofilm assays, and pigmentation assays, results observed with the isogenic wild-type strain were set to 1.0, and these averages were then plotted relative to the results observed with this strain. For protease reporter assays and micro-CT analysis, absolute values were plotted for all replicates obtained with each strain. Analysis of variance (ANOVA) models with Dunnett's posttest adjustment were used to assess statistical significance. P values of Յ0.05 were considered to be statistically significant. Statistical analyses were performed using the statistical programming language R version 3.3.3 (R Foundation for Statistical Computing, Vienna, Austria), SAS 9.4 (SAS Institute Inc., Cary, NC), and GraphPad Prism 5.0 (GraphPad Software, La Jolla, CA).

SUPPLEMENTAL MATERIAL
Supplemental material is available online only. SUPPLEMENTAL FILE 1, PDF file, 0.1 MB. SUPPLEMENTAL FILE 2, PDF file, 0.1 MB. SUPPLEMENTAL FILE 3, PDF file, 0.1 MB.