The novel protein ScrA acts through the SaeRS two-component system to regulate virulence gene expression in Staphylococcus aureus.

Staphylococcus aureus is a Gram-positive commensal that can also cause a variety of infections in humans. S. aureus virulence factor gene expression is under tight control by a complex regulatory network, which includes, sigma factors, sRNAs, and two-component systems (TCS). Previous work in our laboratory demonstrated that overexpression of the sRNA tsr37 leads to an increase in bacterial aggregation. Here, we demonstrate that the clumping phenotype is dependent on a previously unannotated 88 amino acid protein encoded within the tsr37 sRNA transcript (which we named ScrA for S. aureus clumping regulator A). To investigate the mechanism of action of ScrA we performed proteomics and transcriptomics in a ScrA overexpressing strain and show that a number of surface adhesins are upregulated, while secreted proteases are downregulated. Results also showed upregulation of the SaeRS TCS, suggesting that ScrA is influencing SaeRS activity. Overexpression of ScrA in a saeR mutant abrogates the clumping phenotype confirming that ScrA functions via the Sae system. Finally, we identified the ArlRS TCS as a positive regulator of scrA expression. Collectively, our results show that ScrA is an activator of the SaeRS system and suggests that ScrA may act as an intermediary between the ArlRS and SaeRS systems.

INTRODUCTION

Staphylococcus aureus is a gram‐positive opportunistic pathogen capable of causing a wide variety of diseases ranging from minor skin and soft tissue infections to life‐threatening endocarditis and bacterial septicemia (Tong et al., 2015). S. aureus is regarded as a human commensal with approximately 27% of the population being colonized in the nares (Wertheim et al., 2005). Colonization with S. aureus is usually asymptomatic, however, colonized individuals are at greater risk for S. aureus invasive infections (Kluytmans et al., 1997; Wertheim et al., 2005). The versatility of S. aureus, in terms of lifestyle and severity of infection, is due in part to the arsenal of virulence factors encoded by the bacterium. Amongst these virulence factors are adhesins, toxins, exoenzymes, and immune evasion proteins. Precise, coordinated expression of the genes encoding these virulence factors is critical for S. aureus to cause disease, and a complex network of regulatory circuits has been implicated in the regulation of virulence gene expression.

Two‐component signal transduction systems (TCS) are well‐characterized virulence regulators in S. aureus. There are 16 two‐component systems encoded on the S. aureus genome, many of which are known to directly regulate virulence determinants (e.g. the Agr, Sae, and Arl systems) (Haag & Bagnoli, 2015; Jenul & Horswill, 2019). The exoprotein expression (Sae) TCS is known to regulate virulence factors with functions such as adhesion, hemolysis, and proteolysis, (including hla, hlgA/B/C, sspA, aur, and fnbA) (Jenul & Horswill, 2019; Liu et al., 2016; Rogasch et al., 2006), through the use of two classes of target binding sites. Class I targets, also known as low‐affinity targets, are known to be upregulated in strain Newman, which has increased basal levels of SaeS kinase activity, and includes genes such as coa, eap, and sbi (Liu et al., 2016; Rogasch et al., 2006). Class II targets, also known as high‐affinity targets, are insensitive to changes in basal kinase activity and include genes such as hla (Jeong et al., 2012; Liu et al., 2016; Mainiero et al., 2010). The SaeRS system is activated by a range of stimuli, including calprotectin, hydrogen peroxide, and human neutrophil proteins 1–3, all of which play a role in the human immune response (Cho et al., 2015; Geiger et al., 2008). The ArlRS system has been established to regulate genes involved in a variety of cellular functions including ebh and sdrD (involved in adhesion), virulence factors such as nuc, lukA, esxA, and transcriptional regulators such as sarV, and mgrA (Crosby et al., 2020). The regulation of adhesins by ArlRS is critical to endovascular infection (Kwiecinski et al., 2019). One known signal for ArlRS activation is disruption of glycolysis, specifically through a reduction in manganese (Párraga Solórzano et al., 2019), suggesting that ArlRS may be activated in nutrient‐poor environments.

Although two‐component systems, and stand‐alone regulators (such as the Sar family), have been the primary focus of gene regulation studies in S. aureus, recently, sRNAs have emerged as potent regulators of processes such as virulence, biofilm formation, and stress response (Bronesky et al., 2016; Lalaouna et al., 2019; Romilly et al., 2012). In addition to carrying out regulatory roles on their own, several sRNAs have also been shown to carry open reading frames encoding small peptides (Janzon et al., 1989; Nielsen et al., 2011), which can carry out functions independent of the sRNA. The best‐studied example of this type of molecule in S. aureus is the delta toxin Hld, an α‐type phenol soluble modulin (PSM), which is encoded on the sRNA RNAIII. While RNAIII coordinates the regulation of a variety of S. aureus virulence genes (as an RNA molecules), Hld also functions as a hemolysin. Small proteins/peptides are likely widespread in bacterial genomes but are not annotated due to their small size (Garai & Blanc‐Potard, 2020; Miravet‐Verde et al., 2019; Schilcher et al., 2020).

Previous work by our group generated updated S. aureus annotation files to include previously identified sRNA molecules (Carroll et al., 2016). In the same study, we identified 39 novel putative sRNAs, named tsr1‐39, several of which demonstrated altered expression in human serum compared to TSB (Carroll et al., 2016). A number of the newly identified tsr transcripts had the potential to encode small proteins or peptides. Follow‐up work by our group revealed that three of the tsr transcripts (tsr21, tsr22, and tsr37) encode small proteins (Sorensen et al., 2020). During this analysis, we observed that overexpression of a histidine tagged protein, encoded on the tsr37 transcript, led to a dramatic increase in cellular aggregation in the absence of human serum.

In this study, we investigate the biological role and function of the tsr37 encoded protein (herein renamed ScrA for Staphylococcal clumping regulator A). We show that ScrA contributes to S. aureus auto‐aggregation, leading to increased clumping in planktonic cultures, and to increased biofilm formation. Whole‐cell transcriptomics and mass spectrometry of secreted proteins revealed the expression of several virulence factors is altered upon ScrA overexpression, many of which are part of the SaeRS regulon. We go on to demonstrate that ScrA‐mediated phenotypes require a functional SaeRS system, strongly suggesting that ScrA activates the Sae system. Finally, we demonstrate a role for the ArlRS TCS in activating expression of scrA and show that the SaeRS system has a negative influence on scrA expression. Based on these results we hypothesize that ScrA acts as an intermediary linking the ArlRS and SaeRS systems.

RESULTS

Investigation of the locus

Previously, we demonstrated that overexpressing a 6x‐his tagged version of the ScrA protein resulted in increased autoaggregation of S. aureus cells (Sorensen et al., 2020). The construct used in the previous study (pScrA‐His) consisted of the ScrA open reading frame fused to six histidine residues, and an ~300 bp upstream region, leaving the protein under the control of its native promoter (Figure 1a). Since the scrA transcript was truncated in this construct, we constructed a second scrA overexpression plasmid containing the entire scrA gene, (pScrA), which also expressed scrA under control of its native promoter (but did not contain a his tag on the ScrA open reading frame) (Figure 1a). The MRSA252 and NCTC8325 genomes contain annotations for a gene immediately downstream of scrA, which is not annotated in USA300. Previously published data by Mäder et al. (2016) suggested that ScrA and the downstream gene (which we have designated scrB) are encoded on a polycistronic transcript (Figure 1b). To investigate if scrB contributes to the clumping phenotype observed, we constructed a third overexpression plasmid (pScrAB) containing both the scrA and scrB genes under control of their native promoter (s) (Figure 1a). To investigate the operon structure of scrA and scrB, a Northern blot was performed using RNA samples from wild‐type S. aureus containing either pMK4 (empty vector control) or pScrAB. An scrA mutant (containing the empty vector pMK4) was included as a negative control. Using a probe that encompasses the 267 nt scrA open reading frame (Figure 1a, red bar) we detected an ~480 nt band in the wild type and ScrAB overexpressing strains (Figure 1c), while no bands were detected in the scrA mutant. We were unable to detect any scrB transcript when probing 20 μg of total RNA with a 309 nt scrB‐specific probe (data not shown). No band corresponding in size to polycistronic scrAB transcript was detected. The size of the scrA transcript detected (~480 nt) is consistent with a transcript encompassing the scrA coding sequence (267 nt) plus approx. 200 nt of untranslated sequence, strongly suggesting that scrA is monocistronic. Nonetheless, we cannot rule out the possibility that the scrA transcript detected arose as a result of processing a longer scrAB transcript.

FIGURE 1

Predicted transcript architecture of the scrAB locus. (a) Three overexpression plasmids were constructed to express either a his tagged ScrA (pScrA‐his), scrA (pScrA), or scrA and scrB (pScrAB). (b) Data previously published by Mäder et al. (2016) suggest that scrAB is encoded on a polycistronic transcript. (c) Northern blot of wild‐type S. aureus containing the pMK4 empty vector (WT pMK4), wild‐type S. aureus containing pscrAB (WT pScrAB), and the scrA mutant (scrA). Blots were probed with a riboprobe antisense to the scrA open reading frame. Blots were loaded with either 10 or 20 μg of total RNA as indicated. Red bar in panel a indicates the sequence used to generate the scrA northern probe, while the scrB probe sequence in indicated by a green bar

ScrA‐induced clumping is not strain specific and leads to aggregation in both planktonic and static cultures

During our initial investigation into ScrA function, bacterial cells were observed to spontaneously clump in the absence of human serum, a phenotype that can be quantified by measuring OD600 before and after static incubation (Sorensen et al., 2020). To investigate if this clumping was specifically due to scrA overexpression, or an artifact resulting from overexpressing a truncated form of scrA, we subjected strains containing the three overexpression constructs (pScrA‐his, pScrA, and pScrAB), to a clumping assay in the absence of human serum. OD600 was determined before and after incubation, and all three scrA overexpression strains showed an increase in clumping over the pMK4 empty vector control. The pScrA‐his containing strain demonstrated ~40% clumping after a 2 h incubation, while pScrA and pScrAB containing strains showed ~80% clumping (Figure 2a). The similarity in clumping between pScrA and pScrAB strains, and the absence of scrB in the pScrA plasmid, indicates that ScrA is primarily responsible for the observed clumping phenotype.

FIGURE 2

Overexpression of scrA induces clumping. (a) S. aureus containing the pMK4 empty vector as well as the three overexpression constructs were grown overnight and were left static for 2 h. The initial and final OD600 of the top 100 μl were used to calculate clumping. (b) The scrAB overexpression plasmid was transduced into S. aureus strains SH1000, Newman, and UAMS‐1, and clumping was assessed. An increase in clumping was observed in each background. (c) Individual overexpression of either scrA, scrB, or scrAB was performed in plasmid pCN51 to determine the contribution of each individual transcript to clumping. Overexpression was driven by a cadmium‐inducible promoter. Increased clumping was only observable when overexpressing scrA. (d) A biofilm formation assay was performed over time with wild‐type S. aureus containing the pMK4 empty vector (pMK4) and the scrAB overexpressing strain (pScrAB). A statistically significant increase in biofilm formation was observed in the scrAB overexpressing strain. Experiments were performed for a minimum of three times for panels a, b, c, and d, respectively. Error bars represent standard deviation. Statistical significance was determined using an ordinary one‐way ANOVA and Tukey's multiple comparison for panels a–c. Student's t‐test was used at each time point for panel d; *p < 0.05, **p < 0.01, ***p < 0.005, ****p < 0.001

Due to the presence of ScrA and ScrB homologs in MRSA252 and NCTC8325, we next sought to determine if ScrA function was restricted to AH1263 or extended to additional S. aureus backgrounds. To investigate this, we introduced the pScrAB overexpression plasmid into the S. aureus backgrounds SH1000, Newman, and UAMS‐1, along with the empty vector control pMK4. While the background rate of clumping varied by strain, an increase in clumping was observed when overexpressing pScrAB, in all S. aureus backgrounds tested (Figure 2b), demonstrating that the ScrA clumping phenotype is not limited to the S. aureus USA300 lineage (AH1263).

Our initial assay using pScrA and pScrAB suggests that ScrA is responsible for clumping. To confirm this hypothesis, and determine the role of each gene (scrA and scrB) in clumping, we utilized the cadmium‐inducible promoter in plasmid pCN51, to overexpress either scrA alone, scrB alone, or both scrA and scrB. Wild‐type S. aureus containing the pCN51 plasmids were subjected to a clumping assay after overnight growth with 10 μM cadmium chloride to induce expression. When compared with the pCN51 empty vector control, pCN51_ScrAB demonstrated ~80% clumping, similar to levels observed when under control of the native scrA promoter on pMK4 (Figure 2c). Similar levels of clumping were observed using the pScrA construct (~80% clumping). In contrast, strains containing the pScrB construct demonstrated no significant difference compared to the empty vector control (Figure 2c). These results clearly demonstrate that ScrA is solely responsible for the observed clumping phenotype. However, since the operon structure of scrAB is unclear, to ensure optimal production of ScrA all further experiments were performed with the pScrAB construct under control of its native promoter as our primary overexpressor plasmid.

It is well established that interactions between bacterial surface proteins can mediate the initial phases of biofilm formation (Jin et al., 2019). To investigate if the clumping phenotype observed in planktonic cultures overexpressing ScrA influences biofilm formation, we used our ScrA overexpressing strain in a biofilm assay. Experiments were performed in 24‐well plates coated with human serum and inoculated with either the empty vector control or ScrA overexpressing strain. The wells were washed at 1 h intervals (from 1 to 5 h), and biofilm quantity was determined by crystal violet retention. Results show that the ScrA overexpressing strain developed more robust biofilms when compared with the empty vector control (Figure 2d). This suggests that ScrA overexpression induces clumping not only to other bacterial cells (as is likely in planktonic cultures), but also induces adhesion to host serum proteins.

ScrA‐mediated clumping is due to an encoded small protein

While the scrA gene was observed to influence clumping, it was unclear if the effector was the scrA transcript itself acting as a small RNA, or the small protein encoded within. To determine which molecule is the effector, we constructed 2 scrA expression plasmids containing nonsense mutants. A TAA stop codon was introduced into the pCN51_ScrA overexpression plasmid at either the 3rd (pCN51_ScrA_NSAA3) or 8th (pCN51_ScrA_NSAA8) codon, and expression induced by the addition of cadmium chloride. In a clumping assay, neither nonsense mutant was observed to have increased clumping (Figure 3a), suggesting that the observed clumping phenotype is due to the encoded ScrA protein. Expression of scrA was confirmed to be similar between the nonsense mutants and native sequence by RT‐qPCR (data not shown).

FIGURE 3

Role of the ScrA protein in clumping. (a) Serum‐free clumping assay performed using S. aureus containing the empty vector (pCN51), overexpressing native ScrA (pCN51_ScrA), overexpressing ScrA with a nonsense mutation at amino acid 3 (pCN51_ScrA_NSAA3), or overexpressing ScrA with a nonsense mutation at amino acid 8 (pCN51_ScrA_NSAA8). (b) Serum‐free clumping assay performed on S. aureus containing either the empty vector (pCN51, pMK4), overexpressing full‐length scrA (pCN51_ScrAB, pMK4_ScrAB), overexpressing the ScrA C‐terminal tail (pCN51_ScrA‐CTD), or overexpressing the ScrA transmembrane domain (pMK4_ScrA‐TM). Experiments were performed for a minimum of three times. Error bars represent standard deviation. Statistical significance was determined using an ordinary one‐way ANOVA and Tukey's multiple comparison. *p < 0.05, **p < 0.01, ***p < 0.005, ****p < 0.001

Computer modeling of ScrA predicts 2 distinct domains, a transmembrane helix, and a C‐terminal domain. To determine which domains are responsible for clumping, we overexpressed a truncated ScrA consisting of only the transmembrane domain driven by its native promoter in pMK4. Additionally, we constructed a plasmid overexpressing the ScrA C‐terminal domain driven by a cadmium‐inducible promoter in pCN51. Clumping assays showed a modest increase in clumping when the transmembrane domain alone was expressed, although the level of clumping observed was significantly lower than that observed when overexpressing the full‐length protein (Figure 3b). Overexpression of the C‐terminal domain alone did not result in increased clumping. Together, these data suggest that the clumping phenotype observed is mediated by a protein encoded by the scrA gene in which the transmembrane domain is essential to its function.

ScrA influences abundance and secreted virulence factors

Overexpression of ScrA leads to increases in clumping and biofilm formation, however, the exact mechanism underlying these processes remains unclear. The clumping phenotype could be mediated directly by ScrA, however, we deemed this unlikely due to the small size of the ScrA protein and it's predicted localization in the cell membrane. Alternatively, ScrA could mediate clumping indirectly via some additional S. aureus factor (s). To investigate how ScrA influences clumping we first performed whole transcriptomics by RNA sequencing (RNA‐seq) to determine how ScrA overexpression alters the bacterial transcriptome (compared to empty vector control). Strains were grown to mid‐exponential phase, RNA extracted, and RNA‐seq performed. Raw data were analyzed and visualized by volcano plot (Figure 4a, Table S1). To identify genes significantly altered upon ScrA expression, a differential expression analysis was performed using the following parameters; 2‐fold change in expression, a p < 0.05, and a mean reads per kilobase of transcript per million mapped reads (RPKM) >10 in either strain (Figure 4a, red dots). In total, 353 genes were altered with 151 upregulated upon ScrA overexpression and 201 downregulated. The 10 most upregulated and 10 most downregulated genes (by abundance) are listed in (Figure 4b). The two most highly upregulated protein coding genes were the surface adhesins coa and empbp at 25.5‐ and 25.4‐fold upregulated, respectively. The map gene, encoding the secreted adhesin Map was also highly upregulated (20.5‐fold). Upregulation of these genes could explain the increase in biofilm formation observed (Figure 2d), as their gene products encode proteins capable of binding host factors. Interestingly saeS which encodes the sensor kinase of the SaeRS two‐component system, was upregulated 7.76‐fold. The response regulator saeR was also found to be upregulated (7.18‐fold), while the saeP (SAUSA300_0692) and saeQ (SAUSA300_0693) genes were upregulated 3.32‐ and 5.77‐fold, respectively (Table S1). SaeRS is known to positively autoregulate its own expression (Liu et al., 2016), suggesting that overexpression of ScrA may lead to SaeRS activation. To confirm and validate the RNA‐seq data, we performed RT‐qPCR on two genes shown to be upregulated upon ScrA overexpression (i.e., coa and saeS). Results confirmed the RNA‐seq data showing that both genes were upregulated in the ScrAB overexpression strain, (Figure S1).

FIGURE 4

Transcriptomic and proteomic analysis of ScrA overexpressing strain. (a) RNA sequencing was performed on 3 h cultures of S. aureus containing the pMK4 empty vector and the scrAB overexpressing strain. Differential expression analysis was performed, and the results were visualized on a volcano plot. Significance was determined using Student's t‐test. Log2 fold change is shown on the x axis, while −log10 p is shown on the y axis. Genes indicted by red circles displayed a fold change >2 and p value was <0.05. (b) The 10 genes demonstrating the highest fold increase and decrease in expression in the scrAB overexpressing strain. Several adhesions, including coa, empbp, and sbi, were identified as being upregulated in a scrAB overexpresser. (c) The secreted protein profiles of S. aureus containing the pMK4 empty vector and the scrAB overexpressing strain were analyzed by mass spectrometry proteomics. Differential expression analysis was performed, and the results were visualized on a volcano plot. Significance was determined using Student's t‐test. Log2 fold change is shown on the x axis, while −log10 p is shown on the y axis. Genes indicated by red circles displayed a fold change >2 and p value was <0.05. (d) the 10 proteins demonstrating the highest fold increase and decrease in abundance in the scrAB overexpressing strain

The SaeRS system is a global regulator of secreted virulence factors in S. aureus (Giraudo et al., 1997; Liu et al., 2016; Mainiero et al., 2010), therefore we next investigated changes to the S. aureus secreted protein profile (secretome) upon ScrA overexpression. Overnight cultures (scrA overexpressor and empty vector control) were pelleted via centrifugation and the supernatants were TCA precipitated to concentrate proteins. Liquid chromatography coupled to mass spectrometry was utilized to identify proteins and determine their abundance. Data were visualized by volcano plot (Figure 4c) and differential expression analysis was performed using similar criteria as those employed for RNAseq (2‐fold change in expression, a p < 0.05) (Table S2). 135 proteins were altered with 50 upregulated upon ScrA overexpression and 85 downregulated. The 10 most upregulated and 10 most downregulated proteins by abundance are listed in Figure 4d. Notably, the Von Willebrand binding protein (vWbp) and extracellular matrix binding protein (Empbp) were 112‐fold and 20‐fold upregulated, respectively, while numerous superantigen‐like proteins were also increased upon ScrA overexpression (Figure 4c,d). Proteases SplF, Aur, SplE, and SplC were downregulated. Another target of SaeRS, gamma hemolysin component A (HlgA) was found to be 24‐fold upregulated, while gamma hemolysin component B (HlgB) was 7.5‐fold upregulated. Collectively, these data are broadly consistent with both SaeRS activation, and the observed clumping and biofilm phenotypes observed in an ScrA overexpresser.

ScrA overexpression leads to an increase in HlgA‐mediated hemolysis

As noted above, increases in HlgA and HlgB proteins were observed upon scrA overexpression. HlgA and HlgB form a heterodimer, which is capable of lysing human erythrocytes. Therefore, to determine if changes in HlgAB levels were biologically significant we examined the consequence of scrA overexpression on the hemolytic activity of S. aureus. Human erythrocyte lysis assays were performed using the ScrA overexpressing strain and empty vector control. Results show an approx. 7‐fold increase in hemolytic activity in the ScrA overexpressing strain compared to the empty vector control (Figure 5). The αPSMs are toxic peptides produced by S. aureus and are potent cytotoxins, particularly against human erythrocytes (Giraudo et al., 1997; Wang et al., 2007; Zapf et al., 2019). No difference in αPSM levels was observed in the secretomic analysis, suggesting that the differences observed in hemolysis are HlgAB mediated. This result supports the proteomic data and suggests that the increase in gamma hemolysin production observed in the ScrA overexpressing strain manifests as a biologically meaningful increase in activity.

FIGURE 5

ScrAB overexpressing leads to increased hemolytic activity against human erythrocytes. Cultures of S. aureus containing the pMK4 empty vector (WT pMK4_EV) and the scrAB overexpressing strain (WT pScrAB) were grown for 15 h and hemolysis assays were performed with cell‐free culture supernatants. An ~7 fold increase in hemolytic activity was observed in the scrAB overexpressing strain compared to the empty vector control. Experiments were performed four times. Error bars represent standard deviation. Significance was determined using Student's t‐test *p < 0.05, **p < 0.01, ***p < 0.005, ****p < 0.001

SaeRS is essential for ScrA function

The phenotypes observed above (Figure 2) and transcriptomic/proteomic data (Figure 4) strongly suggest the SaeRS system is activated in response to ScrA overexpression. To determine if SaeRS is essential for the observed ScrA‐mediated phenotypes, we overexpressed ScrA in saeR and saeS mutant backgrounds and examined clumping of each strain compared to empty vector controls. As previously observed, overexpression of ScrA led to increased clumping in the wild‐type background, however, overexpression of ScrA did not lead to any measurable increase in clumping in either the saeR or saeS mutant strains (Figure 6a), indicating that ScrA‐mediated clumping requires both SaeR and SaeS. We next investigated the requirement for SaeRS in ScrA‐mediated biofilm formation. A biofilm assay was performed using wild‐type S. aureus (AH1263) and the saeR mutant, containing either the empty vector or ScrA overexpression plasmid (Figure 6b). As previously observed, ScrA overexpression in the wild‐type background led to a more robust biofilm, while ScrA overexpression in an saeR mutant showed no increase over the empty vector control. These results strongly suggest that ScrA is functioning via the SaeRS system, either directly or indirectly. To confirm that the abrogation of ScrA‐mediated phenotypes in sae system mutants is specifically due to the inactivation of the SaeRS system, we overexpressed ScrA in an agrA mutant (AgrA is the response regulator of the Agr two‐component system). ScrA overexpression in the agrA background led to an increase in clumping similar to that observed in the wild‐type background (Figure S2), confirming that ScrA‐mediated phenotypes specifically require an intact SaeRS system.

FIGURE 6

The SaeRS system is required for ScrA‐mediated phenotypes. (a) Clumping assays were performed in wild‐type S. aureus, saeR, and saeS mutants overexpressing scrAB. No increase in clumping was observed when either saeR or saeS was disrupted. (b) Biofilm assays were performed in wild‐type S. aureus and an saeR mutant overexpressing scrAB. No increase in biofilm formation was observed when ScrA was overexpressed in the saeR mutant. Experiments were performed four times. Error bars represent standard deviation. Significance was determined using a standard one‐way ANOVA and Tukey's multiple comparison. *p < 0.05, **p < 0.01, ***p < 0.005, ****p < 0.001

ScrA expression is positively regulated by ArlR and negatively regulated by SaeR

Previous work by Rapun‐Araiz et al. (2020), overexpressed individual two‐component system (TCS) response regulators and used RNA‐seq to investigate the specific regulons and direct targets of each TCS in S. aureus. We re‐examined the data generated in this study to determine expression values (in each TCS overexpressing strain) for scrA, which was absent from the USA300 genome file used as a reference in the study. To generate values for scrA we utilized an updated USA300 reference genome, previously generated by our group, that contains annotations for sRNA genes (scrA was originally annotated as tsr37/SAUSA300s301 in this study; Carroll et al., 2016; Sorensen et al., 2020. The goal of this analysis was to determine if scrA expression was under the control of any TCS in S. aureus. Results show that for 15 of the 16 TCS in S. aureus there was no significant variation in scrA expression when a constitutively active form of the response regulator was expressed (Figure 7a). However, a 25‐fold increase in scrA expression was observed when the ArlR response regulator was overexpressed (Figure 7a). These data strongly suggest that scrA is positively regulated by the ArlRS TCS.

FIGURE 7

ArlR positively regulates scrA expression. (a) Reanalysis of data previously published by Rapun‐Araiz et al. (2020) demonstrated that constitutive expression of ArlR resulted in an approx. 25‐fold increase in scrA expression. No other TCS response regulator increased scrA expression when constitutively activated. (b) Quantification of scrA transcript abundance in S. aureus WT, arlR, and saeR mutant strains containing the scrAB overexpression plasmid following 3, 6, and 9 h growth. Abundance of scrA transcript in each strain was determined by RT‐qPCR and normalized against the value in the WT strain at each timepoint. A significant decrease in scrA expression was observed in the arlR mutant at both 3 and 6 h, but there was no significant difference by 9 h. Interestingly disruption of saeR resulted in increased scrA expression at 3, 6, and 9 h of growth. (c) Clumping assays were performed using WT S. aureus and the arlR mutant containing the scrAB overexpression plasmid. Overexpressing scrAB in the arlR mutant led to a significant increase in clumping, but not to the same extent observed in the WT strain. The increase in clumping in the arlR mutant was significantly lower than that in the WT background. Experiments were performed four times. Error bars represent standard deviation. Significance was determined using a Student's t‐test for panel B and an ordinary one‐way ANOVA and Tukey's multiple comparison for panel C. *p < 0.05, **p < 0.01, ***p < 0.005, ****p < 0.001

To further explore the Arl‐ScrA‐Sae regulatory pathway, and confirm that ArlR positively regulates scrA, we performed RT‐qPCR to examine scrA transcript levels following 3‐, 6‐, and 9‐ h of growth in TSB. RNA for RT‐qPCR was isolated from WT S. aureus and an arlR mutant containing the ScrA overexpression plasmid (in which scrA is under the control of its native promoter). We also included the saeR mutant (containing the ScrA overexpression plasmid) in the analysis to investigate if SaeR was downstream of ScrA in the regulatory pathway. At 3 h scrA expression was reduced in the arlR mutant ~3‐fold relative to the wild type (Figure 7b), which is consistent with ArlR positively regulating scrA. This reduction was also observed at 6 h, although the effect was slightly less that 2‐fold, and by 9 h no significant difference was observed (Figure 7b). Surprisingly, scrA expression in the saeR mutant, was increased ~3.5‐fold at 6 and 9 h. This increase is suggestive of a negative feedback loop by SaeR on scrA expression and the absence of SaeR leads to derepression and increased transcription of scrA. Analysis of the promoter region of scrA showed no canonical SaeR‐binding site, suggesting that either SaeR is binding to an alternative sequence or that Sae‐mediated repression of scrA is indirect and facilitated by a downstream target of SaeR.

Finally, since ArlR positively influences ScrA expression, we sought to determine if the absence of arlR leads to decreased ScrA activity. To investigate this, we overexpressed ScrA (or the empty vector) in wild‐type AH1263, and the arlR mutant and subjected the strains to a clumping assay (Figure 7c). We observed an ~80% increase in clumping when ScrA was overexpressed in the wild‐type strain, consistent with previous assays. In the arlR mutant background, an increase in clumping was observed, however, the increase was less than that observed in the WT (~50% increase in clumping). This further suggests a positive regulatory role of ArlR on scrA. However, the increase in clumping also suggests that scrA expression is not entirely dependent on arlR and additional unidentified regulators of scrA expression may exist in the cell. This is consistent with the RT‐qPCR data taken at 9 h (Figure 7b).

ScrA overexpression leads to increased Sae‐dependent membrane stability

The ScrA protein is 88 amino acids in length and contains one predicted transmembrane helix. Overexpression of a protein containing a transmembrane helix, such as ScrA, and subsequent insertion of the protein into the cell membrane, could potentially destabilize the membrane, and therefore, while unlikely, it is possible that this nonspecific membrane destabilization could cause all of the phenotypes outlined in this study. To investigate if the ScrA‐mediated phenotypes observed in this study are due to alterations in membrane stability (as a result of increased accumulation of ScrA in the membrane), we examined membrane integrity using the nonmembrane permeable dye propidium iodide, which is fluorescent when it enters through damaged membranes and intercalates into DNA. Cells were harvested by centrifugation, washed, and incubated with propidium iodide. Overexpression of ScrA in WT S. aureus led to increased permeability of the membrane suggesting some alterations in membrane stability arise from ScrA overexpression (Figure 8). However, when ScrA was overexpressed in the saeR mutant, no increase in permeability was observed. This result indicates that the increase in membrane instability is not directly attributable to the production of ScrA (which is high in the saeR mutant, Figure 7b), but rather it is a result of the activation of the SaeRS system that accompanies ScrA overexpression. Consistent with this, an increase in membrane instability was also observed, albeit to a lesser degree, in the arlR mutant. This result mirrors the results from the clumping assay (Figure 7c) again suggesting that membrane instability is not due to ScrA per se, but is instead due to the function of downstream targets, affected by ScrA overexpression.

FIGURE 8

Overexpression of scrAB leads to membrane instability in an SaeRS‐dependent manner. Propidium iodide staining was used to measure membrane stability. Increased fluorescence is indicative of greater instability. Overexpression of scrAB in the wild‐type strain resulted in a significant decrease in membrane stability, while no difference in membrane stability was observed following scrAB overexpression in the saeR mutant strain. Membrane instability was also increased following scrAB overexpression in the arlR mutant although not to the same degree as in WT S. aureus. Error bars represent standard deviation. Significance was determined using an ordinary one‐way ANOVA and Tukey's multiple comparison. *p < 0.05, **p < 0.01, ***p < 0.005, ****p < 0.001

DISCUSSION

In this study, we demonstrate that overexpression of the ScrAB locus results in changes to the global expression of adhesins, proteases, hemolysins, and the two‐component system SaeRS. Observed changes in the transcriptome and secretome are consistent with, and suggestive of, overstimulation of the SaeRS system. ScrAB‐mediated phenotypes are dependent on the presence of an intact SaeRS system, with observed clumping, biofilm, and hemolysis phenotypes abrogated when saeR or saeS was disrupted by a transposon insertion (Figure 6). Much of our understanding of the SaeRS regulon comes from S. aureus strain Newman, which shows increased basal kinase activity due to an L18P mutation (Rogasch et al., 2006). Several genes known to be altered in strain Newman, [including saeS, saeR, efb, embpb, splC, sbi, aur, sspA, hlgA, and hlgB (Rogasch et al., 2006)), were also altered in our transcriptomic and secretomic data (Figure 4) strongly suggesting that the SaeRS system is being stimulated. Changes in HlgA were demonstrated to be biologically significant, as overexpression of pScrAB led to an ~7‐fold increase in hemolytic activity (Figure 5). In addition to showing that ScrA promotes the SaeRS system, we have demonstrated that ArlR positively regulates scrA expression (Figure 7). Disruption of the ArlRS system resulted in a significant decrease in scrA expression. While it is possible that ArlR directly activates scrA expression, ArlR is also known to function indirectly through the global regulator MgrA. Activity of the scrA promoter may be mediated by direct binding of ArlR, or through binding by MgrA. We are currently investigating the molecular mechanism through which the ArlRS system influences scrA expression and in turn how this impacts the SaeRS system.

While it appears that ScrA acts through SaeRS, the exact mechanism of action of ScrA activity remains unclear. The SaeRS system is complex, being comprised of four proteins, SaePQRS. SaeR and SaeS are the response regulator and histidine kinase, respectively, while SaeP and SaeQ are accessory factors that modulate the phosphatase activity of SaeS (Jeong et al., 2012). ScrA may act by stimulating the kinase activity of SaeS, or alternatively, it could repress phosphatase activity, by inhibiting SaePQ function on SaeS. If ScrA acts directly to stimulate SaeS kinase activity then it may function in a manner similar to Human Neutrophil peptide 1 (HNP‐1), which is known to activate SaeS kinase activity (Geiger et al., 2008). Furthermore, as previously mentioned, a point mutation (L18P) found in a transmembrane helix of SaeS in strain Newman has been shown to increase basal kinase activity of SaeS (Adhikari & Novick, 2008; Liu et al., 2016). It is tempting to speculate that interaction with this transmembrane helix by the intramembrane portion of ScrA, may alter the conformation of the helix in a manner similar to the L18P mutation, thus increasing kinase activity. RT‐qPCR suggests that the SaeRS system has a negative feedback loop on scrA expression. This could serve as a circuit breaker and prevent over‐activation of the Sae system by ScrA and/or the ArlRS system. While it is possible that SaeR is directly mediating this repression, the lack of a canonical SaeR binding site suggests that negative feedback is indirect, mediated by a downstream target of SaeR. Interestingly, SaeR was not indicated as a repressor of scrA in our analysis of the reconstructed TCS data by Rapun‐Araiz et al. (2020). This may be due to the low basal level of scrA expression and the absence of the activator ArlR in the SaeR overexpressing strain (the background strain for the experiments lacked all TCS response regulators other than the specific one being overexpressed). Thus, it is likely that the basal level of scrA transcription was low enough that additional repression would not lead to detectable changes in expression.

Throughout this study, we utilized an scrAB overexpressing plasmid, and while we have demonstrated that the phenotypes observed appear to be ScrA‐dependent, we acknowledge the exact contribution of ScrA and ScrB to each phenotype is unclear. We have demonstrated that clumping is influenced by ScrA exclusively. However, the role of each protein in additional ScrA‐mediated phenotypes (e.g. hemolysis, biofilm formation, Sae activation) was not investigated. While we consider it likely that ScrA is the primary factor responsible for these phenotypes further work is necessary to conclusively eliminate any contribution from ScrB in regulation of S. aureus virulence factor expression. Furthermore, the biological significance of these ScrAB‐mediated phenotypic changes to the S. aureus cell and S. aureus pathogenesis requires further study. It has been established that S. aureus binding to Von Willebrand Factor (VWF), (mediated by vWbp), facilitates adherence to blood vessel walls and influences the establishment of endocarditis (Claes et al., 2017; Liesenborghs et al., 2019). Identification of vWbp being 112‐fold upregulated in the ScrA overexpression strain (Figure 4d), suggests that ScrA may play a role in endocarditis. Experiments are ongoing in our laboratory to determine the contribution of ScrA to S. aureus infection.

This study has demonstrated that overexpression of the scrAB locus leads to SaeRS activation and substantial changes in both the transcriptome and secretome. These changes lead to alterations in cellular clumping, biofilm formation, and hemolytic activity against human erythrocytes. RT‐qPCR revealed a positive regulatory role of ArlR on scrA expression, while SaeR appears to repress scrA expression. Collectively, these observations led us to hypothesize a working model whereby ScrA is acting as an intermediary between the ArlR and SaeRS systems (Figure 9). We hypothesize that activation of the ArlRS system, leads to increased expression of scrA by the ArlR response regulator (or indirectly through another regulator such as MgrA). ScrA then acts, either directly or through an unknown intermediary, on the Sae system, stimulating SaeS kinase activity. This activation leads to changes in the SaeRS regulon, including activating adhesins and hemolysins while repressing proteases, resulting in increased cellular aggregation, biofilm formation, and hemolysis. Following activation, SaeRS acts to repress scrA either directly or indirectly. This function of ScrA may represent a previously unidentified functional link between the ArlRS and SaeRS two‐component systems in S. aureus.

FIGURE 9

Hypothetical working model of ArlRS‐ScrA‐SaeRS action. ArlR positively regulates scrA expression, either directly or indirectly. The ScrA protein inserts into the cell membrane which in turn activates the SaeRS system, causing increased expression of Sae regulon genes. The increase in Sae system activity has a negative feedback on scrA expression either directly through SaeR or indirectly via a downstream Sae regulon gene

MATERIALS AND METHODS

Strains and strain construction

All bacterial strains and plasmids used in this study are listed in Table 1. All oligonucleotides used are listed in Table 2. Transposon mutants were acquired from the Network on Antimicrobial Resistance in Staphylococcus aureus (NARSA) (Fey et al., 2013) and transduced into USA300 AH1263. Phage transduction of both transposon mutations as well as plasmids utilized bacteriophage Φ11. Transposon presence was confirmed for agrB, arlR, and saeR, via PCR utilizing primer pairs #0049/#0052, #0055/#0056, and #0057/#0058, respectively. Construction of overexpression plasmids utilized USA300 AH1263 genomic DNA as a template to amplify fragments for insertion as follows. Primer pair #0831/#0902 was used to generate a fragment used in the construction of pRKC0751, #0832/#0902 generated a fragment for the construction of the pRKC0752 plasmid. Primers #1045/#1046 generated a fragment for pRKC993. Primers #1045/#1069 generated a fragment for pRKC1033. Primers #1070/#1071 generated a fragment for pRKC1034. The insert and backbone were restriction enzyme digested, the backbone was treated with alkaline phosphatase and the insert was ligated into the plasmid with T4 DNA ligase. Constructed plasmids were transformed into Escherichia coli strain DH5α. The creation of the scrA nonsense mutants was performed using an in vivo assembly (IVA) method as previously described (García‐Nafría et al., 2016) using primer #1451/#1452 for pCN51_AA3 and #1453/#1454 for pCN51_AA8. Screening and sequencing to confirm insertion of the fragments into the vector utilized #0045/#0046 for pMK4 and #0230/#0231 for pCN51. Plasmids were introduced into S. aureus strain RN4220 by electroporation and phage transduced to strain AH1263.

TABLE 1

Strains and plasmids used in this study

Strain or plasmid	Characteristics	Reference or source
S. aureus
AH1263	USA 300 LAC isolate cured of plasmid LAC‐p03	Boles et al. (2010)
RN4220	Restriction‐deficient transformation recipient	Kreiswirth et al. (1983)
SH1000	rbsU repaired laboratory strain	Horsburgh et al. (2002)
UAMS‐1	Osteomyelitis clinical isolate	Gillaspy et al. (1995)
Newman	Laboratory strain	Carroll et al. (2012)
RKC600	SH1000 pMK4	Zapf et al. (2019)
RKC602	UAMS‐1 pMK4	Zapf et al. (2019)
RKC604	Newman pMK4	Zapf et al. (2019)
NE1622	JE2 saeR::Bursa, NARSA transposon mutant	Fey et al. (2013)
NE1296	JE2 saeS::Bursa, NARSA transposon mutant	Fey et al. (2013)
NE1684	JE2 arlR::Bursa, NARSA transposon mutant	Fey et al. (2013)
NE1532	JE2 agrA::Bursa, NARSA transposon mutant	Fey et al. (2013)
RKC681	AH1263 pRKC679	This study
RKC684	AH1263 saeR::Bursa	This study
RKC759	AH1263 pRKC751	This study
RKC760	AH1263 pRKC752	This study
RKC854	SH1000 pRKC752	This study
RKC856	Newman pRKC752	This study
RKC857	UAMS‐1 pRKC752	This study
RKC878	AH1263 saeR::Bursa pscrAB	This study
RKC908	AH1263 saeR::Bursa pMK4_EV	This study
RKC1009	AH1263 pRKC1009	This Study
RKC1058	AH1263 pRKC1058	This Study
RKC1066	AH1263 saeS::Bursa pscrAB	This study
RKC1067	AH1263 saeS::Bursa pMK4_EV	This study
RKC1005	AH1263 pRKC993	This study
RKC1039	AH1263 pRKC1033	This study
RKC1040	AH1263 pRKC1034	This study
RKC809	AH1263 arlR::Bursa pMK4	This study
RKC0811	AH1263 arlR::Bursa pRKC752	This study
RKC0772	AH1263 agrA::Ery pRKC752	This study
RKC0694	AH1263 agrA::Ery	This study
RKC1229	AH1263 pRKC1226	This study
RKC1230	AH1263 pRKC1127	This study
E. coli
DH5α	Cloning strain	Invitrogen
DH10β	Cloning strain	Invitrogen
Plasmids
pMK4	Gram‐positive shuttle vector (CMR)	Sullivan et al. (1984)
pCN51	Cadmium inducible promoter (EryR)	Charpentier et al. (2004)
pRKC679	pMK4_pscrA_6x‐his (vector overexpressing truncated his tagged scrA from its native promoter)	Sorensen et al. (2020)
pRKC751	pMK4_scrA (vector overexpressing full length scrA from its native promoter)	This Study
pRKC752	pMK4_scrAB (vector overexpressing scrAB from its native promoter)	This Study
pRKC993	pCN51_scrAB (vector overexpression scrAB from a cadmium inducible promoter)	This Study
pRKC1009	pMK4_scrA_TM (vector overexpressing the transmembrane domain of ScrA)	This Study
pRKC1033	pCN51_scrA (vector overexpressing scrA from a cadmium inducible promoter)	This Study
pRKC1034	pCN51_scrB (vector overexpressing scrB from a cadmium inducible promoter)	This Study
pRKC1058	pCN51_scrA_C‐tail (vector overexpressing the c‐terminal tail of ScrA from a cadmium inducible promoter)	This Study
pRKC1226	pCN51_scrA_NSAA3 (vector overexpressing scrA with a TAA nonsense mutation at amino acid 3 from a cadmium inducible promoter)	This Study
pRKC1227	pCN51_scrA_NSAA8 (vector overexpressing scrA with a TAA nonsense mutation at amino acid 8 from a cadmium inducible promoter)	This Study

TABLE 2

Oligonucleotides used in this study

Name	Sequence	Details
#0045	GTAAAACGACGGCCAGTG	M13 forward primer
#0046	GGAAACAGCTATGACCATG	M13 reverse primer
#0049	CCAACATTACAAGAGGTTGAACAAGC	agr R oligo
#0052	CGTATAATGACAGTGAGGAGAGTGG	agrB F
#0055	GGTAACAATAATCCAGTTATTGC	arlR F
#0056	CGTAATATGAGGTGTACAAATGACG	arlR R
#0057	CTAATTGATAACACCATTATCGG	saeR F
#0058	GAGGTCGTAAGAACAGAGG	saeR R
#0230	GGTGATGAACATATCAGGCAGA	pCN51 F
#0231	TGATATCAAAATTATACATGTCAACGA	pCN51 R
#0669	AAAACTGCAGAAAATTAATGCGATGATTTTTAGC	scrA His F
#0670	CGGATCCTTAATGATGATGATGATGATGATCTTTTGTCATGAAATAAATGGG	scrA His R
#0831	CGGATCCCCTGATAGAATATAATGTACTGTC	scrA rev
#0832	CGATCCCGCATAAATGATTCTATATTAATGC	scrAB rev
#0902	CAAGAGCTCcAAAATTAATGCGATGATTTTTAGC	scrAB F
#0972	AAAATCCGACAGTTCCAACG	scrA qPCR F
#0973	TGGGATGAATATCACGACTAGAAG	scrA qPCR R
#1045	ACGCgtcgacTTTTAGAAAGGATGTGAAA	scrAB into pCN51 F
#1046	GgaattcGCATAAATGATTCTATATTAATG	scrAB into pCN51 R
#1052	CGggatccTTAATGATGATGATGATGATGTTCATAATGTTTTGCATATTG	Truncated scrA into pMK4
#1069	GgaattcTTAATCTTTTGTCATGAAATAA	scrA into pCN51
#1070	ACGCgtcgacTATATTCTATCAGGAAGGTG	scrB into pCN51 F
#1071	GgaattcTTATGGGTATTTTGTAATTTTATAA	scrB into pCN51 R
#1129	ACGCgtcgacAGAAAGGATGTGAAATAATGCAATATGCAAAACATTATGA	scrA c‐tail into pCN51 F
#1130	GgaattcTTAATCTTTTGTCATGAAATAAATG	scrA c‐tail into pCN51 R
#1451	ATATTTTAGAAAGGATGTGAAATAATGAAATAATCTAAACAAATACTTTT GATTATGGGC	scrA TAA nonsense mutation at codon 3 in pCN51 F
#1452	TTTCATTATTTCACATCCTTTCTAAAATAT	scrA TAA nonsense mutation at codon 3 in pCN51 R
#1453	TGTGAAATAATGAAAGGCTCTAAACAAATATAATTGATTATGGGCATTATATC TCTTATTGT	scrA TAA nonsense mutation at codon 8 in pCN51 F
#1454	TATTTGTTTAGAGCCTTTCATTATTTCAC	scrA TAA nonsense mutation at codon 8 in pCN51 R
#1505	ATGAAAGGCTCTAAACAAATACTTTTGATTATGGGCATTA	scrA riboprobe F
#1506	TAATACGACTCACTATAGGGTTAATCTTTTGTCATGAAATAAATGGGATGAATATCAC GACTAGAAGTAATGTTA	scrA riboprobe R
#1507	TATATTCTATCAGGAAGGTGCAACAATGACC	scrB riboprobe F
#1508	TAATACGACTCACTATAGGGTTATGGGTATTTTGTAATTTTATAAAAG CAAACGTAGAATAATGCGATAAGTAATAATGC	scrB riboprobe R

Bacterial growth conditions

S. aureus cultures were routinely grown at 37°C with shaking in tryptic soy broth (TSB). E. coli cultures were grown at 37°C with shaking in lysogeny broth (LB). Where indicated, antibiotics were used at the following concentrations: Chloramphenicol (10 μg/ml), erythromycin (5 μg/ml), lincomycin (25 μg/ml), ampicillin (100 μg/ml). Where indicated cultures were synchronized as follows. Five milliliters of overnight starter culture was diluted 1:100 into 10 ml of fresh, prewarmed TSB and grown for 3 h to mid‐exponential phase. Cultures were then diluted into 25 ml of fresh TSB in a 250 ml flask to a starting OD600 of 0.05. Cultures were then grown for the indicated time.

Bioinformatics analysis

CLC genomic workbench (Qiagen) was used for the analysis of RNA‐seq data as described previously (Carroll et al., 2014). The previously published updated USA300 genome file was used as a reference for all RNA‐seq analysis (Carroll et al., 2016). Volcano plots were generated as previously described (Briaud et al., 2021) using Rstudio V1.4.1106 (Team R Development Core, 2018) and the EnhancedVolcano package V1.8.0 with cutoffs set at log2 fold change >1 and –log10 p‐value >1.3 (p‐value <0.05).

TCA precipitation

Nine milliliters of supernatant was combined with 1 ml of trichloroacetic acid and incubated at 4°C for 24 h. Precipitated proteins were pelleted at 4°C and washed three times with ice‐cold acetone. The protein pellet was resuspended in 500 μl of 8 M urea unless otherwise noted.

Cell clumping assay

Five milliliters of overnight cultures was grown at 37°C in a 15 ml conical tube. One milliliter of culture was transferred to a 1.7 ml microcentrifuge tube and pelleted. The cells were resuspended in phosphate‐buffered saline (PBS) and thoroughly dispersed via vortex mixing and pipetting. One hundred microloters was removed from each tube and used to determine the initial OD600. Suspensions were incubated statically at room temperature for 2 h. One hundred microliters of solution was removed from the top of the suspension and the OD600 was measured. The percent clumping was calculated as the percent reduction in OD600 after 2 h of incubation.

Biofilm assay

Biofilm assays were modified from a previously described protocol by Cue et al. (2015). In short, each well in a 24‐well cell culture‐treated plate was coated with human serum as follows: 100 μl of 2% human serum was pipetted into each well and the plate was incubated overnight at 4°C. Human serum was aspirated from each well prior to inoculation with bacterial strains. Quadruplicate overnight bacterial cultures were diluted 1:100 into TSB supplemented with 0.5% dextrose and 3% NaCl. One milliliter of diluted culture was added to each well. Inoculated plates were incubated statically at 37°C for up to 6 h. At each timepoint, plates were removed and washed with 2x with PBS. Plates were baked at 60°C overnight to adhere biofilm to the wells. Biofilm was stained with 0.05% crystal violet and washed 2x with PBS. Biofilms were destained with acetic acid and the quantity of biofilm present in the wells was determined by measurement of OD595.

Sample preparation for proteomics

Synchronized 16 h cultures of S. aureus were centrifuged and the supernatant was filter sterilized. Secreted proteins were TCA precipitated as outlined above. Samples were prepared by filter‐assisted sample preparation (FASP). Samples were resuspended in 4% (w/v) SDS, 100 mM Tris pH 7.4, 100 mM DTT, with protease inhibitor cocktail (ThermoFisher Scientific), clarified by centrifugation at 17,000 × g for 10 min, and protein concentration was determined by Pierce 600 nm protein assay (ThermoFisher Scientific). Samples were then standardized to 100 μg and reduced at 37°C for 1 h. Urea was added to a final concentration of 6 M with 20 mM Tris pH 8.5, and samples were placed in a 30 kDa Mw protein concentrator column (Millipore Sigma). All centrifugation steps performed from this point were performed at 12,000 × g for 3–5 min until column was almost empty. Three washes were performed with 8 M urea, 20 mM Tris pH 8.5 (urea buffer), prior to alkylation with 10 mM iodoacetamide in urea buffer, and incubation in the dark at room temperature for 30 min. Washes were performed as above, followed by three more washes with 100 mM triethylammonium bicarbonate pH 8 (TEAB). Trypsin was added in TEAB at 1:100 trypsin: protein (1 μg), and incubated at 37°C for 18 h. Digested samples were eluted by centrifugation, desalted using C18 columns (Waters), and resuspended in 2% ACN 0.1% formic acid.

Mass spectrometry and data analysis

Digested peptides (5 μl) were separated on a 50 cm Acclaim™ PepMap™ 100 C18 reversed‐phase high‐pressure liquid chromatography (HPLC) column (Thermo Fisher Scientific) using an Ultimate3000 UHPLC (Thermo Fisher Scientific) with a 60 (in‐gel digest) or 180 (whole proteome) min gradient (2% to 32% acetonitrile with 0.1% formic acid). Peptides were analyzed on a hybrid Quadrupole‐Orbitrap instrument (Q Exactive Plus; Thermo Fisher Scientific) using data‐dependent acquisition in which the top 10 most abundant ions were selected for MS/MS analysis. Raw files were searched against the S. aureus USA300 proteome (UniProt ID: UP000001939) using MaxQuant (Cox & Mann, 2008) (www.maxquant.org) and the integrated Andromeda search engine. Digestion was set as trypsin/P, variable modifications included oxidation (M) and acetylation (protein N‐term), and carbamidomethylation (C) was fixed. Label‐free quantification was used, with peptides matched between runs. Other settings were left as defaults. The resulting protein groups files were further processed using Perseus (Tyanova et al., 2016), and for whole proteome experiments, this included an imputation step with default settings. Unpaired t‐test with Welch's correction was used to establish significant changes in protein abundance (LFQ intensity) between strains. Proteins with a p‐value less than 0.05 and a fold change greater than 2 up or down were considered significant. The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE (REF ‐ PubMed ID: 30395289) partner repository with the dataset identifier PXD028409.

Human erythrocyte lysis assay

Measurement of human erythrocyte lysis was performed as previously described (Keogh et al., 2019). In short, 16 h cell‐free supernatants were diluted 1:2 in a reaction buffer consisting of 40 mM CaCl2 and 1.7% NaCl. Twenty‐five microliters of whole human blood was added to the solution. Samples were incubated at 37°C while rotating. Intact erythrocytes were pelleted at 5500 × g and the OD543 of the supernatant was recorded. Samples were averaged to either the WT strain or empty vector as indicated.

RNA isolation

Samples for RNA‐sequencing were prepared as follows. Triplicate cultures were synchronized and grown for 3 h. Five milliliters of bacterial culture was pelleted and washed with ice‐cold PBS. Pellets were stored at −80°C until use. Isolation of RNA was performed as previously described (Hussein et al., 2019) using a slightly modified protocol for the RNeasy mini prep kit (Qiagen). RNA samples were treated with a Turbo DNA Free Kit (Ambion). RNA integrity was confirmed via Bioanalyzer (Agilent 2100 Bioanalyzer) and all samples had RIN values >9. RNA was stored at −80°C until use.

RNA sequencing

Ribosomal RNA depletion was performed on each sample using a Staphylococcus aureus‐specific riboPOOL rRNA removal kit (siTOOLs Biotech). RNAseq libraries were generated from the rRNA‐depleted samples by the Ohio University Genomics Facility. Libraries were created using the Illumina TruSeq Stranded mRNA kit (Illumina Cat. # 20020594) per the manufacturer's instruction starting with the Fragment, Prime, and Finish step to accommodate non‐polyadenylated, rRNA‐depleted bacterial RNA. Resultant libraries were sequenced on the Illumina MiSeq using Illumina MiSeq Reagent Kit v3 (150 cycle) (Illumina Cat# MS‐102‐3001) to generate 75 bp paired‐end reads. RNA‐seq data are deposited in the Gene Expression Omnibus (GEO) database under the accession number GSE184753.

Propidium iodide assay

Bacterial cultures were pelleted and resuspended to an OD600 of 1.5. One milliliter of culture was washed with PBS and 100 μl was transferred to a 96 well plate. One microliter of propidium iodide was added to the remaining 900 μl of culture and allowed to incubate statically at room temperature for 5 min. One hundred microliters of culture was then moved to a 96 well plate. Fluorescence was measured with an excitation wavelength of 535 nm and an emission wavelength of 617 nm.

Reverse transcriptase‐quantitative PCR (RT‐qPCR)

Biological quadruplicates were grown and 1 μg total RNA was reverse transcribed using the iScript cDNA synthesis kit (BioRad) per manufacturer instruction. cDNA was diluted 10 times and qPCR was performed using iTaq Universal SYBR Green Supermix (BioRad) in technical duplicates. The housekeeping gene gyrB was used as an endogenous control in all reactions. Amplification and analysis were performed as previously described (Fris et al., 2017).

Northern blot

RNA was isolated from cultures grown for 3 h as described above. The quantity and purity of the RNA were determined by a bioanalyzer nanochip. Either 10 or 20 μg of RNA were loaded onto a formaldehyde agarose gel and electrophoresed for 1 h 15 m at 120 V. The gel was transferred to a nylon membrane by capillary transfer and RNA was UV cross‐linked to the membrane. The ladder and rRNA bands were visualized by staining with a 0.04% methylene blue and 0.5 M sodium acetate solution. To detect the scrAB transcript(s) a riboprobe was synthesized as follows: a PCR fragment encompassing either the scrA or scrB open reading frame was synthesized containing a T7 promoter driving antisense expression of scrA or scrB. This fragment was used as template in an in vitro transcription reaction to generate an antisense riboprobe. The probe was synthesized using α‐P32‐labeled cytosine triphosphate. The membrane was prehybridized for 2 h at 68°C prior to the addition of the probe. The probe was allowed to hybridize overnight at 68°C. The membrane was washed with 2X SSC, 1X SSC, and 0.5X SSC, for 15 min each at 68°C. The membrane was exposed to a phosphor imaging screen overnight and visualized using a phosphor imager.

ETHICS APPROVAL

Human blood was obtained in accordance with procedures approved by the Ohio University Institutional Review Board. Blood was obtained from anonymous donors at Ohio University.

Figure S1

Click here for additional data file.

Figure S2

Click here for additional data file.

Table S1

Click here for additional data file.

Table S2

Click here for additional data file.

Supinfo

Click here for additional data file.