Staphylococcus aureus susceptibility to complestatin and corbomycin depends on the VraSR two-component system

ABSTRACT The overuse of antibiotics in humans and livestock has driven the emergence and spread of antimicrobial resistance and has therefore prompted research on the discovery of novel antibiotics. Complestatin (Cm) and corbomycin (Cb) are glycopeptide antibiotics with an unprecedented mechanism of action that is active even against methicillin-resistant and daptomycin-resistant Staphylococcus aureus. They bind to peptidoglycan and block the activity of peptidoglycan hydrolases required for remodeling the cell wall during growth. Bacterial signaling through two-component transduction systems (TCSs) has been associated with the development of S. aureus antimicrobial resistance. However, the role of TCSs in S. aureus susceptibility to Cm and Cb has not been previously addressed. In this study, we determined that, among all 16 S. aureus TCSs, VraSR is the only one controlling the susceptibility to Cm and Cb. Deletion of vraSR increased bacterial susceptibility to both antibiotics. Epistasis analysis with members of the vraSR regulon revealed that deletion of spdC, which encodes a membrane protein that scaffolds SagB for cleavage of peptidoglycan strands to achieve physiological length, in the vraSR mutant restored Cm and Cb susceptibility to wild-type levels. Moreover, deletion of either spdC or sagB in the wild-type strain increased resistance to both antibiotics. Further analyses revealed a significant rise in the relative amount of peptidoglycan and its total degree of cross-linkage in ΔspdC and ΔsagB mutants compared to the wild-type strain, suggesting that these changes in the cell wall provide resistance to the damaging effect of Cm and Cb. IMPORTANCE Although Staphylococcus aureus is a common colonizer of the skin and digestive tract of humans and many animals, it is also a versatile pathogen responsible for causing a wide variety and number of infections. Treatment of these infections requires the bacteria to be constantly exposed to antibiotic treatment, which facilitates the selection of antibiotic-resistant strains. The development of new antibiotics is, therefore, urgently needed. In this paper, we investigated the role of the sensory system of S. aureus in susceptibility to two new antibiotics: corbomycin and complestatin. The results shed light on the cell-wall synthesis processes that are affected by the presence of the antibiotic and the sensory system responsible for coordinating their activity.

T he indiscriminate use of antibiotics has led to the emergence of numerous multi drug-resistant (MDR) bacteria (1)(2)(3)(4)(5).Two main actions to combat MDR are rigorous policy interventions ensuring the judicious use of antibiotics and a renewal effort for discovering new antimicrobial substances.Recently, two members of a new functional class of natural glycopeptide antibiotics produced by Streptomyces have been described (6).Both antibiotics, named complestatin (Cm) and corbomycin (Cb), are low molecu lar weight linear peptides biosynthesized by nonribosomal peptide synthetases.The mechanism of action of both Cm and Cb is based on their ability to bind the peptidogly can (PG), the structural unit of the bacterial cell wall.Such binding blocks the access of autolysins, enzymes required to break down the peptidoglycan for the insertion of new precursors during peptidoglycan growth and for cell separation after division (7).Autolysins are a highly redundant family of enzymes with hydrolytic functions that include: (i) amidases that cleave the amide bond between N-acetylmuramic (NAM) acid and the L-alanine residue at the N-terminal of the stem peptide, (ii) glycosidases that cleave the glycosidic linkages between N-acetylmuramic acid and N-acetylglucosamine (NAG), and (iii) peptidases that cleave amide bonds between amino acids within the peptidoglycan chain (8,9).The finding that Cm and Cb impact the activity of a broad range of autolysins instead of selectively targeting a specific type is relevant because inhibition of a single enzyme is usually compensated by other members of the family (10,11).Peptidoglycan degradation must be tightly regulated and constantly balanced with peptidoglycan synthesis to maintain cell-wall homeostasis (12,13).Many bacteria use two-component systems (TCSs) that allow bacteria to sense and respond to changes in the environment (14,15) to modulate peptidoglycan biosynthesis and remodeling in response to a variety of signals, including the presence of cell-wall-acting antibiotics (16)(17)(18).Presently, there is no comprehensive knowledge of whether and how signal transduction pathways are involved in Cm and Cb sensitivity.However, such knowledge will be valuable to uncover differences in peptidoglycan remodeling under different environmental conditions and anticipate how antibiotic resistance might emerge (6).
One pathogen for which it is necessary to develop new antibiotics because of its alarmingly high resistance levels to current antimicrobials is Staphylococcus aureus (19,20), a Gram-positive, coagulase-positive, and non-motile bacterium responsible for causing life-threatening infections including endocarditis, sepsis, osteomyelitis, and pneumonia (21).One-third of the general population is colonized by S. aureus, consti tuting a risk factor for infection by this pathogen (22).Importantly, both Cm and Cb are active against methicillin-resistant (MRSA), vancomycin-intermediate, and daptomy cin-resistant S. aureus strains (6).In S. aureus, the activity and/or expression of several autolysins is controlled by various TCSs (23)(24)(25).Furthermore, different TCSs of S. aureus have already been implicated in MDR through modification of the cell wall, efflux mechanisms, and inhibition of drug uptake (26,27): WalKR and VraSR play a role in the resistance to vancomycin (28)(29)(30)(31), BraRS affects the susceptibility to bacitracin and nisin (32), GraRS affects the susceptibility to cationic antimicrobial peptides (33), ArlRS regulates resistance to ceftaroline (34), and HptSR is related with resistance to fosfomy cin (35).The absence of any of these TCSs increases susceptibility to the correspond ing antibiotic.Thus, they represent targets whose inhibition might have antimicrobial activity or may resensitize S. aureus to currently ineffective antibiotics.
The present work addresses whether any TCS network of S. aureus is associated with the susceptibility to Cm and Cb.To this end, we have interrogated a collection of 15 S. aureus mutants in each TCS as well as a mutant lacking all 15 non-essential TCSs and its derivatives containing a single TCS for their susceptibility to Cm and Cb (24,36).Our analyses identified VraSR as the only TCS involved in S. aureus susceptibility to both antibiotics.Exploration of specific VraSR-regulated proteins involved in cell-wall remodeling determined that the absence of the autolysin SagB or its interacting partner SpdC increases resistance to Cm and Cb.Subsequent characterization of the peptidogly can composition suggested that such a rise in resistance might be due to a significant increase in the relative amount of peptidoglycan per optical density (OD) of culture and its degree of cross-linkage identified in ΔsagB and ΔspdC mutant cells.

Gene deletion and mgt overexpression
Generation of deletion mutants was performed as described in reference (38) with some modifications.Briefly, two fragments of at least 500 bp that flanked the left and right of the gene targeted for deletion (ssaA, isaA, spdC, and sagB) were amplified by PCR using primers AB and CD (Table S3) and chromosomal DNA from S. aureus MW2 as a template.PCR products (AB and CD fragments) were purified and used as templates in an overlapping PCR carried out with primers AD (Table S3).The corresponding PCR products were cloned into pJET 1.2 Blunt vector and then subcloned into the shuttle vector pMAD (38), generating plasmids pMAD::ssaA, pMAD::isaA, pMAD::spdC, and pMAD::sagB.pMAD plasmids were purified from E. coli IM01B and transformed into the MW2 wild type or ΔvraSR mutant strain by electroporation.Homologous recombination experiments were performed as described (38).Erythromycin-sensitive white colonies, which did not further contain the pMAD plasmid, were tested by PCR using primers D and E (Table S3).
To construct the pCN51::mgt plasmid, the mgt gene was amplified by PCR with primers mgt_Fw and mgt_Rv (Table S3) and chromosomal DNA from S. aureus MW2 as a template.The PCR product was purified and cloned into pJET 1.2 Blunt vector and then subcloned into plasmid pCN51 (39) digested with BamHI and EcoRI.The resulting plasmid, pCN51::mgt, was purified from E. coli IM01B and transformed into the ΔvraSR mutant strain by electroporation.

Antibiotic susceptibility testing
The minimum inhibitory concentration (MIC) was determined following the EUCAST (European Commitee on Antimicrobial Susceptibility Testing) reading guide for broth microdilution.Briefly, twofold dilutions of antibiotics in U-shaped bottom 96-well microplates (TC Microwell 96U Nunclon, Thermo Fisher) containing MHB were prepared.Overnight cultures of the bacteria were adjusted to 5 × 10 5 CFU mL −1 , and 100 µL was used to inoculate the same volume of MHB containing the antibiotics.Plates were incubated at 37°C for 24 h, and the lowest concentration inhibiting visible bacterial growth was recorded as MIC.To assess antibiotic susceptibility over time at a specific antibiotic concentration, overnight cultures of bacterial strains were diluted to an optical density at 595 nm (OD 595 ) of 0.1 in fresh MHB medium.A volume of 5 µL of the adjusted cultures was used to inoculate 195 µL of MHB (three replicates of each) containing complestatin, corbomycin, or bacitracin at the desired concentration, using 96-well plates (Thermo Scientific).Growth kinetics were assayed using a Synergy H1 hybrid multimode microplate reader (Biotek).Growth data (OD 595 ) were collected every 30 min for 20 h at 37°C with shaking (fast orbital shaking; 425 rpm).MHB was supplemented with 0.1 µg mL −1 of anhydrotetracycline to induce gene expression from the pRMC2 plasmid.In the case of strains carrying the pCN51 inducible plasmid, experiments were carried out without cadmium supplementation since pCN51 shows a basal expression in the absence of cadmium.In vitro evolutionary studies to select mutants resistant to the antibiotics were performed as previously described (6).

Peptidoglycan isolation and analysis
To purify the PG of the strains under study, an established protocol was followed with some modifications (40).Briefly, 200 mL cultures were grown to an OD 595 of 0.5, and cell pellets were mechanically lysed using glass beads (0.1 mm diameter) and a Mini-Bead beater.The supernatants were ultracentrifuged, and resulting pellets were resuspended in 1 mL Tris-HCl 100 mM pH 7.5 and treated with 40 µL MgSO 4 1 M, 2 µL RNase A (500 µg mL −1 ), and 1 µL DNase I (100 µg mL −1 ) for 2 h and then with 50 µL CaCl 2 and 100 µL trypsin (2 mg mL −1 ) overnight.Samples were centrifuged at 14,000 rpm, and the pellet was treated with 1 mL LiCl 8 M, 1 mL EDTA (ethylene diamine tetra-acetic acid) 100 mM, and 1 mL of acetone.After three washes with cold MilliQ water, pellets were resuspended in 1 mL of hydrofluoric acid 48% and incubated with shaking during 48 h at 4°C to remove teichoic acids.Finally, pellets were washed four times with cold MilliQ water and resuspended in phosphate buffer 50 mM pH 4.9.PGs were digested with 4 µL muramidase (mutanolysin from Streptomyces globisporus ATCC 21553 at 1 mg mL −1 ) for 24 h at 37°C.Before the analysis by liquid chromatography, digested PGs were reduced as previously described using borate buffer 0.5 M pH 9 and NaBH 4 during 30 min.Finally, the pH of the samples was adjusted to 3 by addition of 25% orthophosphoric acid.Peptidoglycan profiles were obtained by separating the different muropeptides using a UPLC (Ultra Performance Liquid Chromatography) system (Waters Corp.).The previously described organic method (41) was modified to achieve a better separation of the peaks changing the percentage of buffer A for the gradient as follows: 98% at minutes 0 to 1; 90% at minutes 15 to 15.10; 85% at minutes 18 to 18.10; 0% at 37 min; and 98% at min 20.10 to 25. Muropeptides were detected by measuring the absorbance at 204 nm.
Peak identities were assigned by mass spectrometry using an UPLC system coupled to a Xevo G2/XS Q-TOF mass spectrometer (Waters Corp.).Chromatographic separation and mass analysis were performed as previously described (42) using an ACQUITY UPLC BEH C18 Column (Waters Corp.) heated at 45°C.An amount of 0.1% formic acid in Milli-Q water and 0.1% formic acid in acetonitrile were used as eluents.The QTOF-MS (quadru pole time-of-flight mass spectrometry) instrument was operated in positive ionization mode, and MS e was performed for acquisition of the data using the following parame ters: capillary voltage at 3.0 kV, source temperature to 120°C, desolvation temperature to 350°C, sample cone voltage at 40 V, cone gas flow 100 L/h, desolvation gas flow 500 L/h, and collision energy (CE); low CE: 6 eV and high CE ramp: 15-40 eV.Mass spectra were acquired at a speed of 0.25 s/scan.The scan was in a range of 100-2,000 m/z.Data acquisition and processing were performed using UNIFI software package (Waters Corp.).
Structural characterization of muropeptides was determined based on their mass spectrometric data and tandem mass spectrometric fragmentation pattern, matched with peptidoglycan composition and structure reported previously (43,44).Relative amount of each muropeptide was calculated by dividing the peak area of a muropeptide by the total area of the chromatogram.The relative amount of peptidoglycan per OD of culture was calculated by integrating the total area of the chromatogram obtained from cell cultures normalized to the OD 595 , and the degree of cross-linkage comparing the area of all cross-linked muropeptides (dimers, trimers, tetramers, pentamers, and hexamers) with the total area of the chromatogram.All peptidoglycan analyses were performed using biological triplicates, and a representative chromatogram is shown for each strain.

The VraSR TCS controls the susceptibility of S. aureus to the antibiotics Cm and Cb
To investigate if the TCS network is involved in the susceptibility of S. aureus to Cm and Cb, we first determined the MIC of both antibiotics for the wild-type strain MW2 and a multiple TCS mutant, MW2 ΔXV strain, lacking all 15 non-essential TCSs (36).The MIC of Cm was 2 µg mL −1 and 1.25 µg mL −1 for the wild-type and ΔXV strains, respectively, whereas the MIC of Cb was 3 µg mL −1 for the wild-type and 2 µg mL −1 for the ΔXV strain.To confirm differences in susceptibility between the wild-type and ΔXV mutant, strains were incubated in MHB supplemented with Cm or Cb at the ΔXV MIC for 20 h at 37°C, and growth was recorded over time.Results showed that, in contrast to the wild-type strain, ΔXV growth was completely halted by both antibiotics (Fig. 1A), indicating that one or various TCSs control the susceptibility to Cm and Cb in S. aureus.
We then used a collection of 15 single mutants in each non-essential TCS and a collection of strains, derivative of S. aureus ΔXV, each containing a single TCS (complete histidine kinase and response regulator pair) (Table S1).These collections are valuable tools to analyze the role of TCSs in a specific phenotype and to investigate the selfsufficiency of each TCS signaling pathway without the interference caused by other members of the network.To first validate if the use of these collections allows the identification of unique TCSs involved in S. aureus antibiotic resistance, we examined the susceptibility of the S. aureus wild type, ΔXV, and both collections of strains to the antibiotic bacitracin, which is known to be controlled mainly by the BraRS (bacitracin resistance associated) TCS.Incubation in MHB supplemented with 1 µg mL −1 of bacitracin for 20 h at 37°C showed that the mutant in braRS was the only one exhibiting sensitivity to bacitracin equal to that of ΔXV strain (Fig. 1B).Moreover, ΔXV complemented with the BraRS system was the only ΔXV derivative expressing individual TCSs showing increased resistance to bacitracin (Fig. 1B).
Once we validated our experimental approach and to determine which TCS or TCSs is/are implicated in Cm and Cb susceptibility, we incubated the wild-type, ΔXV, and single TCS mutants in MHB under the presence of either Cm or Cb at the ΔXV MIC.The results showed that among all single mutants, ΔvraSR was the only one presenting a growth arrest similar to the ΔXV strain, suggesting that the VraSR TCS is the main system controlling susceptibility to both Cm and Cb (Fig. 2A).Furthermore, when the growth of the 15 ΔXV complemented strains was evaluated, increased resistance to Cm and Cb was only observed in the case of the ΔXV derivative expressing vraSR (Fig. 2B).Altogether, these results indicated that among all non-essential TCSs of S. aureus, the VraSR signal transduction pathway is the main one responsible for regulating the susceptibility to Cm and Cb, and also suggested the involvement of similar molecular mechanisms in the control of resistance to both antibiotics.
Since the ΔXV strain lacks all non-essential TCSs but maintains the walRK essential TCS in its genome, we analyzed a plausible role of WalRK in Cm and Cb susceptibility.To do so, we used a ΔXV strain complemented with a constitutively active form of the WalR response regulator (WalR D52E, harboring a mutation of the phosphorylation reception residue aspartic acid to a phosphomimetic residue glutamate) (Fig. 3A).Also, as a control, we analyzed a ΔXV strain complemented with a constitutively active form of VraR (VraR D55E).Overproduction of WalR D52E and VraR D55E has previously been shown to activate the WalRK and VraSR signal transduction pathways, respectively (24).As expected, the ΔXV strain expressing the constitutively active form of VraR showed enhanced resistance to Cm and Cb compared to ΔXV.On the contrary, complementation with WalR D52E did not affect Cm and Cb susceptibility.
To confirm that the role of the VraSR signaling pathway in resistance to Cm and Cb is not strain dependent and specific to the MRSA background, we tested the susceptibility of a wild-type methicillin-sensitive S. aureus strain (S. aureus 15981) (45) and its ΔvraSR mutant (46) to these two antibiotics (Fig. 3B).The 15981 ΔvraSR mutant could not grow in MHB supplemented with 1.25 µg mL −1 of Cm or 2 µg mL −1 of Cb.In contrast, comple mentation of the ΔvraSR mutant with a plasmid carrying the vraSR TCS restored Cm and Cb sensitivity to wild-type levels.
Overall, our results point toward VraSR as the only TCS controlling the susceptibility of S. aureus to the antibiotics Cm and Cb, and suggest that specific components of the VraSR regulon may be responsible for this control.

Hypersusceptibility of the vraSR mutant to corbomycin and complestatin is reversed in a double vraSR spdC mutant
By binding to peptidoglycan, Cm and Cb block the action of autolysins, which are required to remodel the cell wall during growth (6).With this in mind and to understand how the VraSR TCS regulates the susceptibility to Cm and Cb, we analyzed the already published complete VraSR regulon looking for genes encoding proteins involved in cellwall synthesis or remodeling and linked to autolysis (24).Overexpression of VraR D55E results in the downregulation of ssaA, isaA, and spdC and the upregulation of mgt (Table S4).Thus, higher levels of SsaA, IsaA, and SpdC and lower levels of MGT (monofunctional glycosyltransferase) are expected in a vraSR mutant compared to the wild-type strain.SsaA (staphylococcal secretory antigen A protein) is involved in the susceptibility to macrolide-lincosamide-streptogramin B antibiotics (47) and shows similarity to the SsaALP autolysin (48).IsaA is a putative lytic transglycosylase able to cleave peptidogly can (49).SpdC (formerly LyrA) (50) forms a complex with the cell-wall hydrolase SagB, and the complex has been proposed to release nascent peptidoglycan strands from the cell membrane to complete integration into the cell-wall matrix (51,52).Last, MGT is a monofunctional transglycosylase that catalyzes the elongation of peptidoglycan chains in a metal ion-dependent manner (53).
To determine if any of these proteins is involved in the hypersusceptible phenotype to Cm and Cb observed for the vraSR mutant, we constructed double mutants in vraSR and ssaA, isaA, or spdC genes, and also overexpressed mgt in the ΔvraSR background (Fig. 4A).The growth of these strains in MHB supplemented with either Cm or Cb at the ΔXV MIC showed a significant resistance recovery to both antibiotics in the case of the ΔvraSR ΔspdC mutant, while the rest of double mutants and the ΔvraSR overexpressing MGT were as susceptible to the antibiotics as the ΔvraSR mutant strain.These results indicated that abolishing SpdC production in the ΔvraSR mutant leads to an increase in resistance to Cm and Cb, and suggested that the hypersusceptible phenotype of ΔvraSR might be due to its increased SpdC levels.

Both SpdC and SagB play a role in S. aureus susceptibility to Cm and Cb
SpdC is a membrane protein homolog of eukaryotic CAAX proteases that interacts with SagB, a membrane-associated N-acetylglucosaminidase that cleaves glycan strands of peptidoglycan to achieve the physiological length (51)(52)(53).By forming a complex, SpdC scaffolds SagB to orient its active site for cleaving glycan strands; thus, spdC and sagB mutants showed similar phenotypes to cell-wall-disrupting agents (resistance to lysostaphin and inhibition of teichoic acids by tunicamycin) (52).Taking this into account, we next sought to investigate the role of the SpdC/SagB complex in Cm and Cb suscepti bility (Fig. 4B).To do so, we first aimed at deleting sagB in the ΔvraSR mutant to test the susceptibility to Cm and Cb of the double mutant ΔvraSR ΔsagB.However, we were unable to construct this double mutant, suggesting that VraSR and SagB are a synthetic lethal pair in S. aureus.Next, we constructed single mutants in spdC and sagB and evaluated their susceptibility to Cm and Cb.Notably, the MIC of Cm and Cb for both mutants raised from 2 μg mL −1 to 3 μg mL −1 and from 3 μg mL −1 to 4 μg mL −1 , respec tively, when compared to the wild-type strain.Accordingly, the ΔspdC and ΔsagB single mutants were able to grow in MHB supplemented with 2.5 μg mL −1 of Cm or 3.5 μg mL −1 of Cb, while the growth of the wild-type strain was inhibited under these conditions (Fig. 4B).
These results indicated that the SpdC/SagB complex is a primary determinant of Cm and Cb susceptibility in S. aureus.
To identify other physiological pathways that are affected by Cm and Cb, we performed a molecular evolution experiment, in which mutants with reduced suscepti bility to antibiotics were selected.Among the selected mutants with moderate resistance to Cm and Cb, there were two independent mutants in the spdC gene and one mutant in the vraG gene (Table 1; Table S5).We also identified mutations in the two-component graSR system, three independent mutants in a protein responsible for the glycosylation of wall teichoic acid polymers (tarS), mutants in purine metabolism (purR and purL), and an autolysin.Importantly, many of these genes have already been linked to resistance to cationic peptides (vraG and graS) (54) and methicillin (tarS) (55).These results confirm the role of spdC and VraSR in susceptibility to Cm and Cb, and identify other molecu lar pathways related to cell-wall metabolism that also influence susceptibility to these antibiotics.

Changes in the peptidoglycan structure provide resistance to Cm and Cb
The bacteriostatic activity of Cm and Cb has been described to be the result of the binding of these molecules to the peptidoglycan and the subsequent blocking of the activity of autolysins.In this context, it is difficult to associate the lack of the hydrolase activity of the SagB autolysin with the increase of Cm and Cb resistance observed for ΔsagB and ΔspdC mutants, as a decrease in peptidoglycan hydrolase activity would be expected to exacerbate and not decrease the detrimental effect of these antibiotics.Since Cb and Cm are cell-wall-acting antibiotics, we wondered if the lack of SagB activity causes changes in the peptidoglycan structure that might confer resistance to these substances.Previous work described changes in the peptidoglycan of sagB mutants, but the analysis was limited to the description of the length of the glycan chains (56).To determine if there are changes in the structure of the peptidoglycan of the sagB and spdC mutants that could be responsible for the increase of resistance observed for the glycopeptide antibiotics under study, we analyzed the muropeptide composition of these mutant strains (Fig. S1).Murein sacculi were digested with muramidase, an enzyme that splits the β1-4 bonds between NAM acid and NAG producing muropeptide subunits and leaving the bridges between the peptide stems intact (57).The comparative analysis of the muropeptide profiles of the wild-type and mutant strains (Fig. 5A and B) revealed a significant increase in the relative amount of peptidoglycan per OD of culture and an increase of cross-linked muropeptides in both ΔsagB and ΔspdC strains (Fig. 5C and  D).The antimicrobial activity of Cm and Cb is related to the impairment of cell-wall homeostasis.Thus, the structural changes in the peptidoglycan of S. aureus taking place when SagB is not active (by deleting sagB or its activator spdC) might be responsible for the increase of resistance to these glycopeptides observed in ΔspdC and ΔsagB mutants.

DISCUSSION
In this report, we showed that VraSR is the only TCS whose absence leads to a significantly enhanced susceptibility to Cm and Cb in S. aureus.Two findings support this conclusion: on the one hand, when the 15 non-essential TCSs were systematically disrupted in an S. aureus wild-type strain, only the absence of VraSR caused a significant increase in the susceptibility to both antibiotics.On the other hand, when an S. aureus strain deficient in the 15 non-essential TCSs was complemented with each TCS, an almost complete restoration of the wild-type susceptibility to Cm and Cb was only achieved when vraSR was overexpressed.This strategy was validated using bacitracin as a query against the same two collections of strains.In agreement with previous studies, BraRS was identified as the unique TCS responsible for bacitracin susceptibility in S. aureus (32).The initial report describing Cm and Cb concluded that these antibiotics block the activity of autolysins, irrespective of the enzyme family (6).S. aureus encodes at least 16 peptidoglycan hydrolases, including five glucosaminidases (Atl, SagA, SagB, ScaH, SsaA), two lytic transglycosylases (IsaA, SceD), four putative amidases (Atl, Sle1, LytN, EssH), and five putative endopeptidases (LytN, LytH, LytM, LytU, EssH).How these enzymes interact with each other and modulate their activity is not well understood, but it is well estab lished that their expression needs to be tightly regulated to coordinate peptidoglycan biosynthesis and degradation to prevent cell lysis (12,58).This regulation can occur at transcriptional and post-translational levels through protein-protein interactions, conformational changes, and structural dynamics that affect the activity of PG synthesis enzymes (59).At the transcriptional level, several S. aureus TCSs strongly contribute to regulating autolysins expression.For instance, WalRK positively controls the expression of Atl, LytM, SsaA, IsaA, SceD, EssH, and various CHAP (Cysteine, Histidine-dependent Amidohydrolases/Peptidases)-containing putative peptidoglycan hydrolases (SA0620, SA2097, and SA2353) (23,60).Consequently, cells without WalRK have reduced peptido glycan hydrolytic activity but continued peptidoglycan synthesis (60).GraSR also positively regulates the expression of Atl, AsaA, IsaA, ScdE, and various CHAP-containing putative peptidoglycan hydrolases (SA2353, SA0620, SA2097, and SA2332) (61).ArlRS represses the expression of LytN and EssH (62,63).Other TCSs, such as AgrCA and SrrAB, also control the autolysis rate by regulating the expression of LytM and IsaA, respectively (49).A special case is LytRS, which negatively regulates peptidoglycan hydrolytic activity by regulating the expression of the lrgAB operon, which apparently controls the trans port of autolysins across the membrane (64).Despite all these TCSs having a profound effect on the regulation of the expression and/or activity of several autolysins, in the present study, we have demonstrated that their deletion or constitutive activation, in the case of WalR, does not significantly affect the susceptibility of S. aureus to Cm and Cb and that VraSR is the only TCS responsible for tolerance to both antibiotics.The VraSR regulon was originally identified by comparing the transcriptional response of a wild-type and its isogenic vraSR mutant strain after vancomycin treatment (65).The results revealed that VraSR regulates several genes associated with cell-wall peptidoglycan synthesis, such as pbp2, sgtB, murZ, fmtA, and teicoplanin-resistance-related proteins (tcaA/tcaB).More recently, we elucidated all TCSs' regulons by complementing a strain devoid of the complete non-essential TCS network (ΔXV mutant) with the constitutively active form of each response regulator.Results established that three autolysins were downregulated (ssaA, isaA, and spdC/sagB), whereas one of them (mgt) was upregulated in the ΔXV strain overexpressing VraR D55E (24).Interestingly, all these autolysins were also shown to be regulated by at least another TCS, while deletion of spdC was enough to restore the susceptibility of the vraSR mutant to Cm and Cb to almost wild-type levels.SpdC is an integral membrane protein that forms a complex with SagB and orients the active site of SagB for efficient cleavage of the nascent peptidoglycan to produce free oligomers that can undergo further peptidoglycan elongation.As spdC and sagB mutants are function ally related, we made several attempts to mutate sagB in the vraSR mutant.However, we were unable to generate the double vraSR/sagB mutant.Thus, we generated single mutants in spdC and sagB.Both single mutants showed decreased susceptibility to Cm and Cb compared to the wild-type strain.All in all, these findings do not allow us to firmly conclude that the greater susceptibility to Cm and Cb shown by the vraSR mutant is due to the overproduction of SpdC, but they reveal that spdC and/or sagB deficiency decreases S. aureus susceptibility to both antibiotics.Further evidence for the contribu tion of SpdC/SagB to Cd and Cm susceptibility comes from in vitro evolutionary experi ments in which independent mutants in SpdC were selected.The same experiment identifies GraSR and VraG mutants with increased antibiotic resistance.The selection of these mutants is interesting because it has been described that the interaction of VraG with GraS is necessary for GraS to recognize cationic peptides (54).
While it might seem contradictory that the lack of activity of SagB autolysin recovers the resistance to the autolysin-inhibiting antibiotics Cm and Cb, certain autolysins have been previously reported to strengthen the bactericidal or bacteriostatic activity of other antibiotics (8).The structural changes observed in the cell wall when SagB is inactive (i.e., increase in the relative amount of peptidoglycan per OD of culture and cross-linkage) could help to maintain cell-wall homeostasis in the presence of Cm and Cb, providing resistance to these antibiotics.For the specific case of Cm and Cb, remodeling the peptidoglycan structure might directly interfere with the activity of other autolysins: S. aureus hypocross-linked peptidoglycans lead to an increased in vitro hydrolysis of cell walls by autolytic enzymes (66), and some peptidoglycan hydrolases have been described to present substrate specificity (39).Additionally, although in our experiments we did not see a significant decrease in the relative amount of peptidogly can in the ΔvraSR mutant in comparison with the wild-type strain (Fig. 5C), VraSR has been shown to positively regulate cell-wall peptidoglycan synthesis (65).In this scenario, the hypersusceptibility to Cm and Cb of the ΔvraRS mutant could also be alleviated by the increase in peptidoglycan resulting from the deletion of SpdC or SagB.It remains uncertain whether these effects are direct or instead affect specific downstream cell-wall components.On the other hand, decreased transcription and enzymatic activities of several key autolytic enzymes of S. aureus lead to a reduction of autolysis induced by some cell-wall-acting antibiotics, thus the single loss of function of the SagB autolysin may also be behind the Cm and Cb resistance phenotype observed for spdC and sagB mutants.Future investigations should aim to elucidate the mechanistic role of SagB and SpdC in antibiotic resistance.

Conclusions
These data contribute to our understanding of which TCS is responsible for controlling homeostatic cell-wall remodeling that is perturbed by the presence of Cm and Cb.Several TCSs monitor the correct expression of autolysins to avoid either insufficient activity that would prevent cell-wall expansion or excessive activity that would cause lysis of the bacteria.Our results indicate that the regulatory activity of TCSs does not overlap.The VraSR system is the only TCS whose absence affects the susceptibility of S. aureus to these antibiotics.This phenotype is related to the overproduction of SpdC and its regulatory activity on SagB.However, other VraRS regulon genes must be involved in this phenotype because SpdC expression is controlled by other TCSs, and still only the absence of VraRS affects Cm and Cb susceptibility.
We can envision at least two limitations in this study.Some TCSs may not be in the phosphorylated state under the tested conditions.It is important to note that the cognate signals for most S. aureus TCSs are still unknown.This opens the possibility that other TCSs might also affect the susceptibility to Cm and Cb.We did not consider phosphorylation through serine/threonine kinases and phosphatases, which play major roles in regulating cell-wall synthesis and antibiotic susceptibility.S. aureus has one serine/threonine kinase (Stk1 or PknB) and a cognate phosphatase Stp1.These questions should be addressed in future studies.

FIG 1
FIG 1 Absence of all non-essential TCSs increases S. aureus susceptibility to complestatin and corbomycin.(A) The optical density at 595 nm was recorded during growth of the wild-type MW2 strain and the ΔXV mutant, lacking all 15 non-essential TCSs, in MHB supplemented with Cm (left panel) or Cb (right panel) at the ΔXV MIC for 20 h at 37°C with shaking.(B) Comparison of the susceptibility to bacitracin (1 µg mL −1 ) of the wild-type MW2 strain and the ΔXV mutant with that of a collection of 15 MW2 single mutants in each non-essential TCS (left panel) and with that of a second collection of 15 strains, derivative of ΔXV, each complemented with a plasmid expressing a single TCS (right panel).In the latter case, the wild-type and ΔXV strains carried an empty pCN51 plasmid.Use of both collections allows the identification of BraRS as the specific TCS that controls S. aureus susceptibility to bacitracin.The average and SD of three technical replicates from one representative experiment of at least three independent experiments are shown.

FIG 2
FIG 2 The VraSR TCS is the only non-essential TCS involved in S. aureus susceptibility to Cm and Cb.(A) Comparison of the susceptibility to Cm (left panel) or Cb (right panel) of the wild-type MW2 strain and the ΔXV mutant with that of a collection of 15 MW2 single mutants in each non-essential TCS.The ΔvraSR mutant strain is the only mutant unable to grow in the presence of the antibiotics.(B) Results obtained with a second collection of 15 strains, derivative of ΔXV, each complemented with a plasmid expressing a single TCS.In this case, the wild-type and ΔXV strains carried an empty pCN51 plasmid.The ΔXV derivative expressing vraSR (ΔXV + vra) is the only one showing a significantly increased resistance to Cm and Cb.The optical density at 595 nm was recorded during growth in MHB supplemented with the antibiotics at the ΔXV MIC for 20 h at 37°C with shaking.Average and SD of three technical replicates from one representative experiment of at least three independent experiments are shown.

FIG 3
FIG 3 Complementation of the ΔXV mutant with a constitutively active form of WalR does not increase resistance levels of ΔXV to complestatin and corbomycin.(A) The optical density at 595 nm was recorded during growth of the wild-type MW2 strain, the ΔXV mutant, and the ΔXV strain complemented with a constitutively active form of the response regulator WalR* (WalR D52E) or the response regulator VraR* (VraR D55E), in MHB supplemented with Cm (left panel) or Cb (right panel) at the ΔXV MIC for 20 h at 37°C with shaking.The wild-type and ΔXV strains carried an empty pRMC2 plasmid.The experiments were performed in the presence of anhydrotetracycline at a concentration of 0.1 µg mL −1 to induce expression from the pRMC2 plasmid.The wild-type 15981 strain and the 15981 ΔvraSR mutant carried an empty pCN51 plasmid.The ΔvraSR mutant is unable to grow in the presence of both antibiotics.(B) The VraSR TCS determines Cm and Cb susceptibility in the methicillin-sensitive S. aureus 15981 strain.The optical density at 595 nm was recorded during growth of the wild-type 15981 strain, the 15981 ΔvraSR mutant, and the ΔvraSR mutant complemented with a pCN51 plasmid carrying the complete vraSR TCS (15981 Δvra + vra), in MHB supplemented with Cm (left panel) or Cb (right panel) at the MW2 ΔXV MIC for 20 h at 37°C with shaking.Average and SD of three technical replicates from one representative experiment of at least three independent experiments are shown.

FIG 4
FIG 4 Role of SpdC and SagB in S. aureus susceptibility to complestatin and corbomycin.(A) Comparison of the susceptibility to Cm (left panel) or Cb (right panel) of the wild-type MW2, ΔvraSR, ΔvraSR complemented with a plasmid expressing the mgt gene (Δvra + mgt) and the double mutants ΔvraSR ΔspdC,ΔvraSR ΔssaA, and ΔvraSR ΔisaA.Deletion of spdC in the ΔvraSR background increases resistance to both Cm and Cb.The optical density at 595 nm was recorded during growth in MHB supplemented with Cm or Cb at the ΔXV MIC for 20 h at 37°C with shaking.(B) Comparison of the susceptibility to Cm (left panel) or Cb (right panel) of the wild-type MW2, and the single mutants ΔspdC and ΔsagB.The optical density at 595 nm was recorded during growth for 20 h, at 37°C, in MHB supplemented with Cm or Cb at a concentration above the MIC of the wild-type strain.Average and SD of three technical replicates from one representative experiment of at least three independent experiments are shown.

FIG 5
FIG 5 Changes in the peptidoglycan structure of spdC and sagB mutants provide resistance to Cm and Cb.(A) Peptidoglycan profile obtained for S. aureus MW2 ΔspdC mutant strain grown in TSB medium and identity of muropeptides (see Fig. S1).M, disaccharide NAG-NAM; numbers, length of stem peptides or glycine bridges.(B) Representative peptidoglycan chromatograms obtained for the wild-type, ΔvraSR, ΔvraSR ΔspdC, ΔspdC, and ΔsagB mutant strains grown in TSB medium.Obtained chromatograms (from three biological replicates) were used to calculate the relative amount of peptidoglycan per OD (C) and the degree of cross-linkage (D).Error bars in (C) and (D) correspond to the standard deviation of three biological replicates.ns, not significant; *P < 0.05, unpaired t test.

TABLE 1
Mutants of S. aureus with moderate resistance to complestatin and corbomycin a