Drug-like Fragments Inhibit agr-Mediated Virulence Expression in Staphylococcus aureus

In response to the increasingly problematic emergence of antibiotic resistance, novel strategies for combating pathogenic bacteria are being investigated. Targeting the agr quorum sensing system, which regulates expression of virulence in Staphylococcus aureus, is one potentially useful approach for combating drug-resistant pathogens that has not yet been fully explored. A previously published study of a fragment screen resulted in the identification of five compound fragments that interact with the DNA-binding domain of the response regulator AgrA from S. aureus. We have analyzed the ability of these compounds to affect agr-mediated virulence gene expression in cultured S. aureus cells. Three of the compounds demonstrated the ability to reduce agr-driven transcription at the P2 and P3 promoters of the agr operon and increase biofilm formation, and two of these compounds also showed the ability to reduce levels of secreted toxins. The finding that the compounds tested were able to reduce agr activity suggests that they could be useful tools for probing the effects of agr inhibition. Furthermore, the characteristics of compound fragments make them good starting materials for the development of compound libraries to iteratively improve the inhibitors.

Staphylococcus aureus is a dangerous human pathogen and a leading cause of endocarditis, bone and joint infections, pulmonary infections, and bacteremia 1 . S. aureus infections have become increasingly difficult to treat due to the growing prevalence of antibiotic-resistant strains. Methicillin-resistant S. aureus (MRSA) strains such as USA300 have become the predominant source of soft-tissue infections in the USA 2,3 . MRSA infections are often treated with vancomycin as a last-resort antibiotic; however, strains resistant to vancomycin have been reported 4,5 .
Although clinical observation of vancomycin resistance in infections has been relatively limited, the threat highlights the urgent need for novel antibiotic therapies 6 .
In response to the problem of increasing antibiotic resistance, targeting bacterial virulence rather than viability has been proposed. Because virulence expression and regulation are important for the establishment and maintenance of an infection but are otherwise non-essential, it has been argued that targeting virulence might be less likely to lead to the development of resistance 7,8 . While these potential advantages make the idea of targeting virulence extremely appealing, this strategy remains largely untested.
In S. aureus, the agr quorum sensing system plays a major role in the regulation of virulence 9 . The agr system coordinates the timing of the transition to an invasive mode that entails increased production of virulence factors and a reduction in surface proteins 10 . Infection models have shown that disruption of the timing of agr activation leads to the attenuation of an infection 11 . The importance of agr-mediated expression of virulence genes has also been demonstrated in several infection models where agr-deficient strains generate significantly milder infections than their wild-type (WT) counterparts [12][13][14][15] .
The agr operon consists of four genes: agrB, agrD, agrC, and agrA that encode the components of the quorum sensing system 16 . Transcription of the operon is driven by the P2 promoter, which is activated by the response regulator AgrA in an autoregulated fashion. agrD encodes a 46-amino acid pro-peptide that is processed and secreted from the cell by the transmembrane endopeptidase AgrB 17,18 . The mature secreted AgrD is the auto-inducing peptide (AIP) of the quorum sensing system, which, after building up to sufficient extracellular concentrations, is capable of activating the receptor histidine kinase AgrC 19 . Activated AgrC promotes the transfer of a phosphoryl group to the response regulator AgrA, which in turn activates transcription at the P2 promoter, completing the auto-regulatory loop 20 . Phosphorylated AgrA also promotes transcription at the P3 promoter, leading to expression of RNAIII, a 514-nucleotide RNA molecule that both serves as the transcript for the hld gene encoding δ-hemolysin and functions as a regulatory RNA 21,22 . RNAIII plays a central role in effecting the transition to a virulent mode as it serves to enhance the expression of genes encoding toxins such as hla (α-hemolysin) while reducing the expression of genes encoding surface proteins, such as spa (protein A). The down-regulation of adhesion molecules upon the activation of the agr system is accompanied by the increased expression of enzymes capable of dissolving biofilm matrices, such as nucleases and proteases. Thus, increased agr activity results in the inhibition of biofilm formation as well as facilitating the dispersal of bacteria from pre-formed biofilms 23,24 .
AgrA is a response regulator of the LytTR family, characterized by a DNA-binding domain that is relatively uncommon among bacteria and absent from higher organisms 25 . LytTR domains are typically found in transcription factors that regulate virulence gene expression 26 . A previously conducted fragment screen against the AgrA LytTR domain identified five compounds that interacted with the DNA-binding domain at a common site that overlapped the DNA-binding surface. Three compounds were shown to inhibit interactions of the AgrA DNA-binding domain with its target DNA in an in vitro assay 27 . Drug-like fragments, which are smaller than typical small-molecule drugs and thus bind with relatively low affinity, are considered to be good starting points in drug-development pipelines 28 . We aimed to test the hypothesis that the previously identified fragments, which target a DNA-interaction surface of AgrA, would inhibit AgrA activity in S. aureus cells. Here we present data demonstrating that several of the compounds identified in the original screen substantially reduce virulence gene activation in S. aureus in ways that are consistent with inhibition of the agr system. These data suggest that these molecules are not only useful for the study of effects of agr inhibition but also have potential as starting molecules for the design of improved inhibitors.

Treatment with inhibitors results in decreased activation of the P3 promoter. Phosphorylated
AgrA drives transcription at the P3 promoter, leading to expression of RNAIII 22 . To test the ability of the inhibitory compounds to disrupt the activation of transcription by AgrA, we employed a cell-based reporter assay using a plasmid containing the gfp gene under the transcriptional control of the P3 promoter 29 . Untreated WT cultures grown for 8 h demonstrated robust expression of GFP, indicating strong transcription of RNAIII (Fig. 1a). In contrast, cultures grown in the presence of compounds 1-4 at a concentration of 120 μM resulted in reduced expression of GFP (Fig. 1a). This concentration of the compounds did not inhibit growth ( Supplementary Fig. S1). Treatment with compound 5 did not significantly alter reporter activity. The largest reduction in GFP expression was achieved with compound 1, although levels were not reduced to those of the Δagr::tet strain, henceforth referred to as the Δagr strain. None of the compounds had significant effects on the Δagr strain (p > 0.05), suggesting that the compounds exert their inhibitory effects via the agr system. Compounds 1-3, which showed the greatest inhibitory effects, were chosen for further analyses. Treatment with compounds 1-3 had no effects on GFP expression in strains containing transcriptional reporter plasmids for dps and recA, two genes that are not known to be under agr regulation 30 (Supplementary Fig. S2). These data suggest that the compounds do not act generally to reduce GFP expression and are consistent with the interpretation that the compounds reduce transcription from the P3 promoter through the agr system. Treatment with inhibitors reduces levels of transcripts directly regulated by AgrA. AgrA directly regulates the transcription of several genes involved in virulence regulation. It promotes the transcription of genes encoding the phenol-soluble modulin (PSM) α and β proteins 30 , the agrBDCA operon at the P2 promoter, and the regulatory RNA, RNAIII, at the P3 promoter. Inhibition of AgrA is therefore expected to result in reduced cellular levels of these AgrA-regulated transcripts. Quantitative real-time PCR was performed to determine the transcript levels of psmα1, psmβ1, agrA, and RNAIII in cultures grown with each compound at a concentration of 120 µM for 8 h. Levels of each of the tested transcripts were significantly lower in cultures treated with compounds 1-3 compared to those in untreated cultures (between 1.6-22 fold decrease in expression) (Fig. 1b). The reduction of transcript levels upon treatment with the compounds is consistent with the interpretation that the compounds interfere with regulation of specific transcripts by AgrA.
Treatment with inhibitors results in reduced cellular levels of AgrA. AgrA drives transcription of the agrBDCA operon at the P2 promoter. Inhibition of AgrA would therefore be expected to result in reduced cellular levels of AgrA. In order to test whether the compounds reduced AgrA levels, western blots were employed to measure levels of AgrA in a Δspa strain. The Δspa strain was used to avoid interference from the immunoglobulin-binding protein A with western blotting. Compounds were added to cultures upon reaching an OD of ~2.5, designated as time zero, at which AgrA levels were moderate but detectable. Samples were collected after an additional 8 h of growth, a point expected to exhibit high levels of AgrA, induced by high cell density ( Supplementary Fig. S3). In untreated samples, levels of AgrA approximately doubled in the 8-h growth window (Fig. 2a). However, treatment with compounds 1-3 resulted in a concentration-dependent reduction in levels of AgrA. Importantly, levels of AgrA from cultures treated with each of the three compounds at concentrations of 250 μM did not increase significantly during the 8-h growth period, indicating nearly complete inhibition of new AgrA synthesis.
An alternate explanation for the reduction in AgrA levels is that treatment with the compounds led to an increase in the turnover of AgrA. To examine the effects of the compounds on protein stability, cultures were pre-treated with the translational inhibitor erythromycin prior to treatment with compounds 1-3. Cultures were www.nature.com/scientificreports www.nature.com/scientificreports/ sampled immediately prior to the addition of erythromycin, 20 min post addition of the compounds, and finally after 8 h of incubation. In the absence of erythromycin, levels of AgrA increased over time, while cultures grown with erythromycin maintained constant levels of AgrA (Fig. 2b). The addition of compounds 1-3 to samples pre-treated with erythromycin resulted in no significant difference in the levels of AgrA. These results are consistent with the interpretation that the compounds disrupt agr activity by interfering with the ability of AgrA to activate agrBDCA transcription.
Treatment with inhibitors results in decreased production of exoproteins, reduced hemolytic activity, and altered levels of the spa and hla transcripts. The agr system regulates virulence in part by increasing the expression of exoproteins such as hemolytic toxins 31 . Inhibition of AgrA activity would therefore be expected to decrease exoprotein production and lead to decreased hemolytic activity. The presence of exoproteins was analyzed in the spent media supernatant of cultures grown over a period of 8 h. Exoprotein levels of the culture supernatant obtained from WT cultures treated with compounds 1 and 2 were visibly decreased relative to those from untreated cultures when analyzed using SDS-PAGE (Fig. 3a). In addition, differences in the relative intensities of some bands were observed. No appreciable change in exoprotein profiles was noted in the Δagr strain upon treatment with the compounds.
To quantify the differences in levels of secreted proteins upon treatment with compounds, total protein levels in the culture media were measured using a Bradford protein assay. Secreted proteins were significantly reduced (p < 0.05) in WT cultures treated with compound 1 and compound 2, but not compound 3 (Fig. 3b). No appreciable change in protein levels was observed when Δagr cultures were treated with compounds ( Supplementary  Fig. S5). Similarly, treatment of the cultures with compounds 1 and 2, but not 3, resulted in a significant decrease in the hemolytic activity of culture supernatants (p > 0.05) (Fig. 3c).
While expression of RNAIII alters expression of protein A and α-hemolysin via modulation of translation 22 , down-stream effects of activation of agr system also lead to altered transcription of the spa and hla genes 30,32 . The effects of the compounds on the transcription of the spa and hla genes were assessed using quantitative real-time Bars represent OD-normalized fluorescence averaged from three separate experiments relative to the untreated sample (normalized to 100) with error bars indicating standard deviations. Statistical significance relative to the sample without compound was determined using a Student's t-test (*p < 0.05, **p < 0.01). (b) Compounds reduce transcript levels of genes directly regulated by AgrA. mRNA was isolated for qPCR analysis after cultures were grown for 8 h in the presence of the indicated compound (Cpd.) at a concentration of 120 μM or DMSO alone. Bars represent the fold change in mRNA level from treated cultures relative to untreated cultures averaged from three separate experiments. Statistical significance and 95% confidence intervals (displayed as error bars) were determined using REST2009 software 59 (*p < 0.05, **p < 0.01).

Treatment with inhibitors promotes biofilm formation.
Activation of the agr system inhibits the formation of biofilms and promotes the dispersal of biofilm matrices. Consequently, strains with inactive agr systems demonstrate increased formation of biofilms 23 . Therefore, inhibition of agr activity was expected to promote biofilm formation. To assess levels of biofilm formation, cultures were incubated in 96-well plates without shaking. While untreated cultures developed moderate biofilms, treatment with compounds 1-3 resulted in increased biofilm formation in a concentration-dependent manner (Fig. 4a-c). Interestingly, treatment with compound 1 at a concentration of 125 µM resulted in biofilm formation at a level comparable to that of the Δagr strain, suggesting that this compound is capable of completely reversing AgrA-mediated repression of biofilm formation (Fig. 4d).
Treatment with inhibitors results in decreased production of exoproteins in strains of different agr types. Across different strains of S. aureus, a region of hyper-variability exists within the agr operon, encompassing the latter half of agrB, agrD and the first half of agrC 33 . The resultant differences in AgrB, AgrD, and the N-terminal domain of AgrC allow for the production (in the case of AgrB and AgrD) and recognition (in the case of AgrC) of different autoinducing peptides. Based on these differences, S. aureus strains are classified as  The concentration of total secreted protein was determined using a Bradford protein assay. Levels of agrdependent protein secreted by WT in the absence and presence of compounds were estimated by subtracting the level of secreted protein in Δagr from the average levels in WT determined from three replicates. Bars represent values relative to the untreated sample (normalized to 1) with errors bars representing standard deviations. (c) Hemolytic activity was assessed after 8 h of growth by adding dilutions of filtered culture media to defibrinated rabbit blood and measuring liberated hemoglobin. HA 50 values were calculated and are shown with error bars representing the standard error of the mean. (d,e) Compounds alter transcript levels for spa (e) and hla (f). mRNA was isolated for qPCR analysis after cultures were grown for 8 h in the presence of the indicated compound at a concentration of 120 μM or DMSO alone. Bars represent the fold change in mRNA level from treated cultures relative to untreated cultures averaged from three separate experiments. Statistical significance and 95% confidence intervals (displayed as error bars) were determined using REST2009 software (*p < 0.05, **p < 0.01).
www.nature.com/scientificreports www.nature.com/scientificreports/ one of four agr types, numbered I-IV 34,35 . In contrast to agrB-C, the sequence of agrA is highly conserved within S. aureus, and therefore inhibition of AgrA would be expected to affect strains of all agr types. The USA300 LAC strain belongs to group I. We examined exoprotein levels in type II, III, and IV strains. Treatment with compounds 1 and 2 resulted in decreased production of exoproteins in N315 (agr type II), MW2 (agr type III), and MN EV (agr type IV) strains (Fig. 5). These data are consistent with the proposed mechanism whereby the inhibitors are functioning primarily through inhibition of AgrA.
Compounds interact with AgrA with sub-millimolar affinity. Binding of the compounds to AgrA C was previously demonstrated by NMR WATERGATE W5 LOGSY and chemical shift perturbation analyses 27 , however, binding affinities were not directly determined. Isothermal titration calorimetry (ITC) was used to measure the affinity of compounds 1-3 for AgrA C . Isotherms were easily fitted to one-binding-site models. Fitting of the isotherms was improved by fixing the stoichiometry to N = 1. Fitting of the isotherm generated by compound 1 resulted in a measured ΔH of −1154 ± 487.0 cal mol −1 , ΔS of 23.4 cal mol −1 deg −1 , and a K d of 485 ± 39.3 μM (Fig. 6a). Fitting of the isotherm generated by compound 2 yielded a measured ΔH of −423.4 ± 25.19 cal mol −1 , ΔS of 14.0 cal mol −1 deg −1 , and a K d of 417 ± 62.9 μM (Fig. 6b). Fitting of the isotherm generated by compound 3 yielded a ΔH of −340.0 ± 35.53 cal mol −1 , ΔS of 17.0 cal mol −1 deg −1 , and a K d of 110 ± 45.6 μM (Fig. 6c). The relatively large errors associated with the fitting of these parameters can be attributed to difficulties in analyzing low-affinity interactions. www.nature.com/scientificreports www.nature.com/scientificreports/

Discussion
Inhibition of virulence has been proposed as a strategy for combating bacterial infections because it exploits previously unexplored targets for inhibition and also has the potential to limit selection for resistance. S. aureus is one pathogen for which virulence inhibition is especially appealing as it produces an extensive array of harmful virulence factors 36 and has a history of antibiotic resistance that dates back to the initial introduction of penicillin 37 . Within S. aureus, the agr quorum sensing system has been specifically identified as a potential target for therapeutic development because of the central role the system plays in regulating virulence gene expression. The potential of this strategy is supported by studies using certain infection models that demonstrate attenuated infections with agr-deficient strains 11,13,14 . Despite the recent interest in the strategy of targeting virulence, both the effectiveness of using a therapeutic agent to target virulence and the reduction in selective pressure caused by such agents remain unproven.
Because of the unusual three-dimensional fold of the AgrA DNA-binding domain and the key role that AgrA plays in both virulence activation (by promoting expression of toxins) and regulation (by activating the quorum sensing mechanism), AgrA has become a target of interest for inhibiting virulence 38,39 . Our previous fragment screen against the AgrA DNA-binding domain identified five compounds with the potential to act as inhibitors of Figure 5. Treatment with compounds reduces exoprotein expression among S. aureus strains of agr-types II-IV. The concentration of total protein secreted into culture media was determined using a Bradford assay, and levels were normalized to those of untreated samples. Statistical significance relative to the untreated samples was determined using a Student's t-test (*p < 0.05). www.nature.com/scientificreports www.nature.com/scientificreports/ AgrA 27 . While these compounds were conclusively shown to interact with the LytTR domain of AgrA, they were unproven in their abilities to reduce agr activity in S. aureus cells.
AgrA functions as a transcription factor, and thus the immediate effect of inhibiting AgrA would be to alter the levels of AgrA-regulated transcripts. The presented studies demonstrate that compounds 1-4 significantly reduced the expression of GFP driven by the P3 promoter, and compounds 1-3 reduced levels of transcripts of agrA, RNAIII, psmα, and psmβ, strongly suggesting that the compounds are acting to interfere with AgrA-regulated transcription. However, while the reduction in transcript levels were significant, they were not comparable to levels seen in the Δagr strain, which were below the threshold of detection under the tested conditions. Optimization of the compounds will therefore be required to achieve complete inhibition of AgrA.
The expression of RNAIII in particular is of central importance for the transition to virulence activation 40,41 . RNAIII coordinates virulence activation via interaction with multiple mRNA targets, resulting in both the direct regulation of some mRNAs and indirect regulation via modulation of expression of regulators such as Rot and MgrA 32,42-44 . Because RNAIII expression impacts several layers of virulence regulation, it substantially alters the expression profile of S. aureus, with studies suggesting that approximately 70 extracellular proteins are regulated by RNAIII 45 . Therefore, for an AgrA inhibitor to function as an inhibitor of virulence activation, expression of RNAIII-regulated virulence factors must be reduced. Consistent with decreased production of RNAIII, treatment with compounds 1-3 reduced levels of secreted proteins and hla transcript levels, and treatment with compounds 1 and 2 reduced hemolytic activity. Furthermore, treatment with compounds 1-3 promoted biofilm formation and increased levels of spa transcript. These effects are all consistent with a reduction in RNAIII production. However, the reductions in these downstream effects were not directly correlated with the reduction in RNAIII expression, likely reflecting the responsiveness of individual RNAIII-regulated systems to the level of RNAIII. In addition, compounds 1-3 were capable of significantly reducing the expression of AgrA itself, supporting the interpretation that the compounds were acting by disrupting transcription at both the P2 and P3 promoters. By reducing both the autoregulatory and virulence factor production functions of AgrA, the potency of virulence inhibition could potentially be enhanced. Together, these results suggest that the compounds interfere with virulence gene activation in cultured S. aureus cells in a manner that is consistent with inhibition of the agr system.
The compounds examined in this study reduced several different aspects of agr-mediated virulence factor activation, suggesting that they could serve as useful tools for experimental validation of the strategy of targeting the agr system to combat S. aureus infections. Pressing questions regarding the strategy of targeting the agr system to combat S. aureus infections remain unanswered. One of the most problematic issues is the promotion of biofilm formation that is associated with a reduction in agr activity. Formation of biofilms is associated with an increase in antibiotic resistance 46 , and agr deficiency has been associated with types of infections for which biofilm formation is especially prevalent 47 . It is possible that the therapeutic efficacy of agr inhibition would depend on the type of infection being treated. Furthermore, the nature of the agr quorum sensing system suggests that the activation of virulence gene expression must be precisely timed in order to be effective. In fact, specifically timed transient inhibition has been shown to attenuate model infections 11 . In light of these temporal requirements, it is likely that inhibition at a point either too late or too early during an infection might reduce or eliminate the efficacy of agr inhibition. Discovering the ideal timing of virulence factor inhibition is important for analyzing the effectiveness of the strategy. Finally, the question of whether or not inhibiting virulence results in selective pressure that leads to the development of resistance as seen with traditional antibiotics is still unknown. These questions may be more easily answered using chemical inhibitors that allow variation in both timing and dose, rather than by use of genetic mutations. Compound 1, which demonstrated the strongest inhibitory activity among the tested compounds, reduced agr activity in many assays to levels close to those of an agr mutant, indicating that compound 1 may be almost as effective as using mutant strains while retaining the flexibility of using chemical inhibition.
Across all of the assays that were employed, a consistent pattern emerged where compound 1 diplayed the most potent inhibition followed by compounds 2 and 3. It is interesting that that the affinity of compound 1 determined by ITC analyses was significantly lower than that of compounds 2 or 3. The discrepancy between the strength of binding (where compounds 2 and 3 bound tighter than compound 1) and the results from cellular assays (where compound 1 consistently demonstrated stronger inhibition than both compounds 2 and 3) is likely explained by other characteristics of compound 1 that allow it to function better as an inhibitor either during uptake or within the cellular environment. Compounds 1-3 all bind with K d 's in the range of 10 −4 M. These modest binding affinities are typical of small compounds identified in fragment screens, with the expectation that affinities can be greatly increased as the compounds are built out 28 .
Recently, several attempts to develop inhibitors of the agr system have been pursued using different strategies. Approaches using AIP analogs 48 , identifying natural product inhibitors 49 , and using traditional chemical inhibition [50][51][52][53][54][55] have yielded promising results. In particular, two compounds, Savirin 54 and ω-Hydroxyemodin (OHM) 55 , are suspected of inhibiting the LytTR domain of AgrA and are likely to behave similarly to the compounds reported in the present study. Both Savirin and OHM were shown to reduce agr activity within S. aureus and were also effective at reducing the severity of infections in mouse models 54,55 . Interestingly, OHM shares structural features with compound 1: both the xanthene base of compound 1 and the anthraquinone OHM feature a similar three-ringed foundation. However, without experimental data to determine how each compound interacts with the AgrA LytTR domain, the importance of the structural similarities of the compounds to their inhibitory functions cannot be assessed.
All five of the compounds tested in these studies originated from a fragment screen library. Fragments are designed to explore large areas of chemical space and to serve as good starting points for the development of therapeutics as they are small in size and amenable to further chemical modifications. The three compounds that we have shown to substantially, albeit incompletely, inhibit agr activity in cultured S. aureus cells are logical starting points for further development. However, it should be noted that our previous studies indicate that all five of the compounds bind to AgrA at a similar site, targeting a region of the LytTR domain that overlaps with www.nature.com/scientificreports www.nature.com/scientificreports/ a surface involved in DNA binding. Thus all compounds might be considered as potential candidates for future pursuit. Just as the affinities of compound 1 and compound 3 are not correlated with the strength of inhibition in cells, the differences observed in the potency of inhibition across all compounds could be due to many different characteristics that are likely to be altered as the size and the complexity of the fragments are increased to enhance affinity, specificity, and intracellular inhibitory activity. Iterations of these compounds with improved inhibitory characteristics would be appropriate for testing in an infection model.
Bacterial growth conditions. The S. aureus strains used in this study are listed in Table 2. The wild-type strain, S. aureus USA300 strain LAC, has been previously described 56 . S. aureus cells were cultured at 37 °C in flasks or tubes such that the ratio of the culture vessel headspace to media volume was 10:1 with shaking at 200 rpm. Alternatively, cells were cultured in 200-μL volumes in 96-well plates. Cells were cultured in Trypticase Soy Broth (TSB) or TSB supplemented with agar (15 g L −1 ; Difco Bacto agar). Where required, the media were supplemented with chloramphenicol (10 μg mL −1 ) or erythromycin (3.3 μg mL −1 ).

Growth inhibition assays.
Overnight cultures were inoculated into 20-mL tubes containing 5 mL of either TSB, TSB supplemented with 5% v/v DMSO, or 5% v/v DMSO with the indicated concentration of compound to an OD (600 nm) of 0.05. The OD was measured at regular intervals to generate growth curves. GFP reporter assays. Overnight cultures were inoculated into 2 mL of TSB supplemented with 0.5% w/v glucose, 5% v/v DMSO, 120 μM compound, and antibiotics appropriate for plasmid retention to an OD (600 nm) of 0.1. Following 8 h of growth, fluorescence and OD were measured using a Thermo VarioSkan plate reader. Green Fluorescent Protein (GFP) fluorescence was measured by excitation at 485 nm and emission at 535 nm.
Data were analyzed by normalizing GFP fluorescence to OD (600 nm) and then subtracting the normalized fluorescence of a control strain carrying a non-fluorescent vector (JMB1242, a WT background strain containing the pCM28 vector 57 ) from values for experimental cultures to account for background signal. Normalized fluorescence values from three triplicates were averaged and compound-treated samples were compared to those of untreated (DMSO only) samples using a Student's t-test to determine statistical significance. www.nature.com/scientificreports www.nature.com/scientificreports/ Real-Time Quanitative PCR. Overnight cultures were inoculated into 2 mL of TSB supplemented with 0.5% w/v glucose, 5% v/v DMSO, and 120 μM compound. Following 8 h of growth, aliquots corresponding to 1.0 OD·mL were collected. The samples were centrifuged at 16,000 × g for 1 min, pellets were resuspended in RNAprotect Bacteria Reagent (Qiagen, Germany) and incubated at room temperature for 5 min. Cells were pelleted by centrifugation at 16,000 × g for 1 min, RNAprotect Bacteria Reagent was discarded, and samples were stored at −80 °C. Cells were resuspended and washed twice with a lysis buffer consisting of 50 mM Tris, 150 mM NaCl at pH 8.0. To lyse the resupsended cells, lysostaphin was added to 15 µg mL −1 and cells were incubated at 37 °C for 30 min. RNA was extracted using TRIzol (Invitrogen, CA); contaminating DNA was degraded using a Turbo DNA-free kit (Invitrogen, CA); and cDNA was generated using a High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems, CA). Primers for qPCR were designed using Primer3Plus 58 and are listed in Table 3. qPCR was performed using a GoTaq qPCR Master Mix kit (Promega, WI) and a QuantStudio 3 Real-Time PCR System (Applied Biosystems, CA). Data were processed and analyzed with REST2009 software 59 .

Western blotting.
To determine AgrA levels, overnight cultures of USA300 LAC spa::kan were inoculated into 25 mL of growth media to an OD of 0.05. Cultures were grown to an OD of ~2.5 at which time aliquots corresponding to 1.3 OD·mL were collected. The samples were centrifuged at 16,000 × g for 1 min and pellets were washed with 1x PBS before freezing and storage at −20 °C. The cultures were then split and compounds were added to a final concentration of 120 μM and 5% v/v DMSO. Samples were again collected after 8 h of growth and processed as described above.
Cell-free extracts were prepared by resuspending cell pellets in 100 μL of a pH 7.6 lysis buffer consisting of 20 mM Tris-HCl, 0.5 mM CaCl 2 , 50 mM NaCl, 40 μg mL −1 DNase I, and 20 μg mL −1 lysostaphin. The lysis mixture was incubated at 37 °C for 1 h, after which 4x SDS-PAGE sample loading buffer spiked with purified AgrA C (for use as a loading control; purification previously described 38 ) was added to a final concentration of 1x. Samples were analyzed by electrophoresis using a 15% polyacrylamide SDS-PAGE gel. Protein was transferred to a nitrocellulose membrane, labeled with anti-AgrA C rabbit antisera and Cy5 goat anti-rabbit IgG secondary antibody (Life Technologies, CA), and visualized using a FluorChem Q Imager (Supplementary Fig. S4). Quantitation of band densities was performed using ImageJ software 60 .  www.nature.com/scientificreports www.nature.com/scientificreports/ Isolation of culture supernatants for exoprotein analysis. Culture supernatants for exoprotein analyses (exoprotein profiles, protein quanitation, and hemolytic activity) were obtained simultaneously as described previously 61,62 . Overnight cultures of USA300 LAC and USA300 agr::tet were inoculated into tubes containing 5 mL of fresh TSB supplemented with 0.5% w/v glucose, 120 μM compound and 5% v/v DMSO to an OD of 0.1. After 8 h of growth, cell densities were normalized by diluting with fresh media. Cultures were centrifuged at 4,000 × g for 20 min to pellet cells and the supernatants were passed through 0.2-μm filters and held at 4 °C or frozen at −80 °C for long-term storage.
Exoprotein profiling assay. For SDS-PAGE analysis, supernatants were concentrated by precipitation with trichloroacetic acid by addition to a final concentration of 10% w/v as previously described 61 . Samples were incubated on ice for 30 min before centrifugation at 4,000 × g for 20 min. Pellets were washed twice with cold acetone, resuspended in 0.025 the original sample volume of 1x SDS-PAGE loading buffer, and stored at −20 °C. Aliquots of triplicates were pooled, analyzed by 15% polyacrylamide SDS-PAGE, and visualized with Coomassie Brilliant Blue. Images were captured using a FluorChem Q Imager.
Quantitative analysis of secreted protein. Total protein secreted was determined using a Coomassie reagent-staining assay. 5 μL of the filtered supernatant was pipetted into a clear flat-bottom UV plate and 250 μL of 1x Advanced Protein Assay Reagent (Cytoskeleton Inc., CO) was added and mixed. The plate was equilibrated at room temperature for 10 min and absorbances at 595 and 450 nm were measured. The data were processed using the ratio of the absorbance at 595 nm to that at 450 nm for all samples 63 . Ratios for the USA300 LAC agr::tet samples were subtracted from the other samples to correct for non-agr related exoproteins. The averages of triplicates for samples from cultures treated with compounds were compared to untreated compounds with significance determined by a Student's t-test.
Hemolytic activity assay. Hemolytic activity was assessed using a modification of the protocol described by Sully et al. 54 . Defibrinated rabbit blood (Hemostat Laboratories) was washed by centrifuging cells at 1000 × g for 5 min followed by gently resuspending in ice-cold 1x PBS, and repeating until the supernatant was clear. Blood cells suspended in 1x PBS or dH 2 O were used to determine baselines for no lysis or complete lysis, respectively. The washed red blood cells were added to the plate containing extracts diluted in 1x PBS and controls to achieve final concentrations of 1% rabbit blood and dilutions of culture extract from 1:4 to 1:256. Reactions were mixed by gentle pipetting and incubated at 37 °C for 1 h. After incubation, the plates were centrifuged at 1000 × g for 5 min at 4 °C. 100-µL aliquots of the supernatants were transferred to a 96-well UV-transparent plate and the absorbance at 415 nm was measured. Values of the PBS control were subtracted from the experimental data, and the difference was normalized to the absorbance values from the dH 2 O wells. These values were plotted versus the concentration of extract to generate an activity curve, which was fitted to a four-parameter logistic model using SigmaPlot 10 (Systat Software, Inc., CA). The calculated EC 50 from the fitting was then inverted to generate the HA 50 . HA 50 values from cells treated with compounds were compared to those from untreated samples (DMSO only) with statistical significance determined by Student's t-tests.
Biofilm formation assay. Biofilm formation was assessed upon abiotic surfaces as described previously 64,65 .
Strains were cultured for ~18 h and diluted into tryptic soy broth (TSB) supplemented with glucose and NaCl (biofilm inducing media) 66 . Where mentioned, compounds were added at the time of inoculation. The cultures were transferred to a treated 96-well tissue culture plate and incubated without shaking at 37°C for 22 h. Subsequently, the culture supernatants were discarded and the plate was washed with water, the biofilms were heat-fixed, stained with crystal violet (0.1% w/v), and destained with acetic acid (33% v/v). The absorbance (A 570 ) of the destained eluate was measured relative to eluate from a cell-free sample and normalized to the OD (A 600 ) of the cells at the time of harvest.
AgrA in vivo stability assay. Overnight cultures of USA300 LAC spa::kan were inoculated into 25-mL cultures to an OD of 0.05. The cultures were grown to an OD of 2.0 and samples were collected in volumes equivalent to 1.3 OD·mL. Cells from these aliquots were harvested by centrifugation at 18,000 × g for 1 min, washed with 1x PBS, and pellets were frozen at −20 °C for storage. The remaining cultures were split and erythromycin was added to a final concentration of 10 µM to all cultures except a control. After an incubation of 20 min, samples were again removed and processed. Compounds were added to the remaining cultures to a final concentration of 120 μM and 5% v/v DMSO. Samples were collected after 8 h of growth. All samples were processed for western blotting and analyzed as described above.
Isothermal titration calorimetry. Purified AgrA C was dialyzed into 20 mM sodium citrate buffer, 250 mM NaCl, and 1 mM tris(2-carboxyethyl)phosphine at pH 6.0. The dialysate was retained and used to prepare the compounds. DMSO was added to 5% v/v with an AgrA C concentration of ~375 μM for use with compounds 1 and 2 and 250 μM for use with compound 3. Compounds 1 and 2 were prepared to concentrations of 10 mM and compound 3 was prepared to 5 mM in the above citrate buffer supplemented with DMSO. Each compound was titrated into the protein solution using a MicroCal iTC 200 Microcalorimeter (Malvern Instruments, UK) at 25 °C. The data were processed, analyzed, and fitted using the Origin 7.0 MicroCal module (OriginLab, MA).

Data Availability
All data generated or analysed during this study are included in this published article.