Bacterial strains used in this study.

The WT S. aureus strain used in these studies was JE2, a laboratory derivative from the community-acquired, methicillin-resistant USA300 LAC [28]. SAUSA300_0200-ggt mutant strains were generated via transduction of the transposon-inactivated gene from the Nebraska Transposon Mutant Library strain into JE2 using previously described techniques [28,70]. Bacterial strains used in this study are presented in S1 Table.

A strain harboring an in-frame deletion of gisABCD-ggt (Δgis) was constructed using a previously described allelic exchange methodology for S. aureus [71]. One kb upstream of SAUSA300_0200 and one kb downstream of ggt were amplified using primers listed in S2 Table and cloned into pKOR1-mcs. pKOR1-mcs was confirmed to have correct 1kb homology sequences by Sanger sequencing. The deletion strain was screened for hemolysis on blood agar plates and displayed WT-like hemolysis.

Isolation of clinical isolates.

Four clinical isolates of S. aureus were obtained from de-identified specimens at a regional hospital clinical laboratory. Three abscess isolates were confirmed to be methicillin-resistant (strains 1055–1057) and the other was a methicillin-susceptible bone isolate (strain 1059). Identification and minimum inhibitory concentration assays were performed following Clinical and Laboratory Standards Institute approved methods. After initial isolation, subcultures were grown on tryptic soy agar (TSA, Remel) overnight.

Sulfur source-dependent proliferation analysis.

Chemically defined (PN) medium was prepared as previously described [4,27]. PN medium was supplemented with 5 mg mL^-1 glucose for this work. Prior to inoculation in PN S. aureus was cultured in tryptic soy broth (TSB, BD Bacto) overnight, washed with phosphate-buffered saline (PBS), and resuspended in PN medium to an OD₆₀₀ equal to 1. Round bottom 96-well plates containing PN supplemented with 5 mg mL^-1 glucose and the indicated sulfur sources were inoculated with S. aureus strains at an initial inoculum of OD₆₀₀ 0.01. Growth analysis was carried out in a Biotek Epoch 2 plate reader set to 37° C with continuous shaking for the indicated time. PN was modified to test sulfur source-dependent proliferation of S. epidermidis and S. aureus by replacing MgSO₄ with MgCl₂ and omitting Met, resulting in PN_mod. S. aureus and S. epidermidis growth curves were performed as described above in PN_mod supplemented with the indicated sulfur sources for 25 h. Sulfur sources were purchased from Millipore Sigma and GSH solutions were freshly prepared prior to each trial to limit oxidation. Alternatively, to ensure CSSC, GSH, and GSSG were maintained in their respective reduced or oxidized forms, stock solutions were prepared by weighing the appropriate amount of the chemical aerobically and immediately transferring it to an anaerobic chamber (Coy) with a 95%:5% nitrogen:hydrogen atmosphere. Sulfur sources were then resuspended in either anaerobically acclimated water (GSH and GSSG) or anaerobically acclimated 1 N HCl (CSSC). Anaerobic proliferation was monitored within the Coy chamber using a Biotek Epoch 2 plate reader.

Isolation of GSSG-proliferation impaired transposon mutants.

Mutant strains that comprise the Nebraska transposon mutant library (NTML) were transferred from flat bottom 96-well plates archived at -80° C into flat bottom 96-well plates containing TSB supplemented with 10 μg mL^-1 erythromycin using a 96-pronged replica plater (Millipore Sigma). Mutant cells were cultured overnight at 37° C and subcultured using a 1:150 dilution into a fresh round bottom 96-well plate containing 150 μL PN media supplemented with 25 μM GSSG instead of 25 μM CSSC as the primary sulfur source. Proliferation of each mutant within the 96-well plate was assessed by monitoring OD₆₀₀ over 22 h using a H1 Biotek plate reader set to 37° C with continuous shaking. This process was repeated for all twenty 96-well plates that comprise the NTML. To account for plate-to-plate variation, proliferation was averaged across all 96 transposon (Tn) mutant strains within a given 96-well plate. Tn mutants that exhibited decreased proliferation equal to at least two standard deviations below the 96-well plate average across mid-exponential and stationary phase were isolated by streak plating onto TSA supplemented with 10 μg mL^-1 erythromycin. Isolated Tn mutants were validated in a secondary screen that compared proliferation in PN media supplemented with 25 μM GSSG to WT JE2 in technical and biological triplicate using round bottom 96-well plates and monitoring OD₆₀₀ over 22 h using a H1 Biotek plate reader set to 37° C with continuous shaking. Growth in TSB was used as a positive proliferation control. The following Tn mutants displayed significantly reduced growth compared to WT JE2 in 25 μM GSSG supplemented PN but not TSB: NE392 (SAUSA300_0200::Tn), NE541 (SAUSA300_0201::Tn), NE457 (SAUSA300_0202::Tn), NE215 (SAUSA300_0203::Tn), NE254 (ggt::Tn) (S1 Table). These mutants were selected for two additional steps of identification and validation. First, the location of the bursa aurealis Tn insertion for each mutant strain was verified using a previously described inverse PCR method followed by Sanger sequencing [28]. Second, the Tn mutations were backcrossed into the WT JE2 parental strain using a previously described phage transduction method [28,70] and the resulting backcrossed mutants were analyzed for proliferation as mentioned above in PN supplemented with 25 μM GSSG or the indicated sulfur source (S1 Table).

Liquid Chromatography-Tandem Mass spectrometry quantification of S. aureus-associated GSSG.

WT JE2, Δgis and ggt::Tn strains were cultured in a 1 L Erlenmeyer flask containing 250 mL PN supplemented with 25 μM CSSC (4:1 flask-to-volume ratio). Cultures were incubated at 37° C with 225 rpm to an OD₆₀₀ of 0.4–0.5. At this time the three cultures (WT JE2, Δgis and ggt::Tn) were split into two 50 mL conical tubes each containing 50 mL aliquots of culture, the cells washed once in PBS, and resuspended in 50 mL of prewarmed PN media supplemented with either 0 μM or 25 μM GSSG. The six 50 mL conical tubes were incubated at 37° C without shaking for 5 min. Cells were centrifuged, washed twice with 10 mL PBS, and resuspended in 500 μL HPLC grade methanol containing [²H₁₀] GSSG (Toronto Research Chemicals, Canada) as the internal standard. Cells were lysed via bead beating using 0.1 μm zirconia beads (BioSpec, Bartlesville, OK) in a Bullet Blender Storm 24 (Next Advanced, Troy, NY) at maximum speed for 2 min. The lysate was collected and centrifuged at maximum speed for 10 min to remove insoluble debris. Metabolite-containing supernatants for each of the six samples were dried using a Savant DNA 120 SpeedVac concentrator (Thermo Scientific) and were stored overnight at -20° C. Samples were reconstituted 500 μL of 0.1% formic acid in preparation for mass spectrometry.

Metabolites from a 10 μL aliquot from each sample were separated in reversed-phase mode using an Acquity HSS T3 column (1.8 μm 100 X 2.1 mm²; Waters, Milfor, MA). Solvent A was 0.1% formic acid in water and solvent B was methanol. The mobile phase gradient was as follows: 0 min−100% A, 1 min − 100% A, 5 min − 60% A, 7 min − 1% A, 8 min − 1% A; 8.01 min − 100% A, and 10 min − 100% A with a flow rate of 0.3 mL/min. Mass spectrometry detection was performed using a Xevo G2-XS quadrupole time-of-flight (QTof) by positive ion electrospray ionization. GSSG detection was based on the retention time of 2.96 min to 3.00 min and mass accuracy using MassLynx Version 4.2 (Waters). GSSG calibration curves were generated with a six-point curve of serially diluted unlabeled GSSG standards with the corresponding concentration of [²H₁₀] GSSG.

Expression and purification of Ggt and GisA.

Open reading frames corresponding to the ggt or gisA genes were PCR amplified from the S. aureus JE2 genome with primers listed in S2 Table. Subsequently, the genes were cloned into the pET28b expression vector using NEB HiFi Gibson assembly kit (NEB, New England, MA) after the plasmid has been linearized with NcoI-HF and XhoI-HF. The assembly mixture was transformed into E. coli, cells were recovered in lysogeny broth (LB), and plated onto LB agar (Fisher) containing 50 μg mL^-1 kanamycin and 5 mg mL^-1 glucose. Plasmids were confirmed using Sanger sequencing and transformed into an E. coli NEB strain 3016 slyD mutant [72]. Transformed E. coli were cultured in LB with 50 μg mL^-1 kanamycin overnight at 37° C with shaking at 225 rpm, sub-cultured 1:50 into 500 mL LB with 50 μg mL^-1 kanamycin in a 2 L flask and grown to an OD₆₀₀ of 0.4–0.7. Ggt or GisA protein expression was induced by addition of 200 μM isopropyl-1-thio-β-D-galactopyranoside (IPTG) and the culture was separated into five 500 mL flasks containing 100 mL of culture and incubated for 4 h at 27° C and 225 rpm shaking. After induction, cells were centrifuged at 10,000 x g for 10 min at 4° C and washed with PBS. Resulting GisA and Ggt induction pellets were resuspended in 40 mL of buffer containing 50 mM tris, 200 mM KCl, 20 mM imidazole at pH 8, or 40-mL buffer containing 50 mM tris, 500 mM NaCl, 20 mM imidazole at pH 8, respectively. Cells were lysed via five consecutive cycles through a fluidizer set to 20,000 psi. Lysates were then centrifuged at 15,000 x g for 15 min to remove intact cells and the resulting supernatant was retained. To purify the target proteins, Ni-NTA chromatography was used. Purification was performed by incubating the cleared lysate with 1 mL Ni-NTA resin (Qiagen, Hilden, Germany) on a rotating platform at 4° C for 2 h. Protein was eluted with 50 mM tris 400 mM imidazole. Buffers used to purify GisA contained 200 mM KCl while buffers used to purify Ggt contained 500 mM NaCl. Each buffer contained 1x protease inhibitor cocktail (Millipore-Sigma). The GisA elutant was dialyzed using 10 mM tris, 200 mM KCl at pH 7.5 as the dialysis buffer for 18 h. The Ggt elutant was dialyzed using 10 mM tris, 150 mM at pH 7.0 as the dialysis buffer for 18 h. Both elutions were concentrated using a 10 kDa molecular weight cutoff protein concentrators. Purification was confirmed via electrophoresis using 12% SDS-PAGE gels. Protein concentrations were determined with the bicinchoninic acid (BCA) protein kit (Pierce ThermoFisher).

Quantitation of Ggt enzyme kinetics.

γ-glutamyl transpeptidase reactions contained 5 μg recombinant Ggt, reaction buffer (10 mM tris with 150 mM NaCl), and the indicated concentrations of GSH and GSSG dissolved in reaction buffer. Reactions proceeded for 30 min at 37° C after which samples were incubated at 80° C for 5 min to stop the reaction. Samples were dried using a roto-vac speed vacuum and stored at -80° C until they were hydrated via resuspension in water, derivatized with carboxybenzyl (CBZ), and applied to a Waters Xevo TQ-S triple quadrupole mass spectrometer as previously described [73]. Peak processing was performed by MAVEN, and the signal was normalized to a ¹³C-glutamine internal standard [74]. An external glutamate standard curve was generated using the same chromatographic conditions, and the signal was normalized to a ¹³C-glutamine internal standard. A fit equation to the standard curve was employed to quantify glutamate within the samples. Glutamate released per min was calculated and data were fit to the Michaelis-Menten equation using GraphPad Prism. Data represent the average of glutamate quantified from four independent protein purifications.

Fractionation and western blot analysis of His-tagged S. aureus Ggt.

The His-tagged ggt ORF was amplified from pET28b::ggt and cloned into pOS1 P_lgt digested with NdeI and HindIII using Gibson assembly to generate Ggt-His (His +). The ggt ORF was amplified from JE2 genomic DNA and pOS1 P_lgt digested with NdeI and HindIII using Gibson assembly to generate the untagged Ggt (His -). Plasmids were confirmed by Sanger sequencing and transformed from E. coli DH5α into S. aureus RN4220 via electroporation. Plasmids were purified from RN4220 and transformed into JE2 ggt::Tn. An empty vector control strain was generated by transforming JE2 and ggt::Tn with pOS1 P_lgt. To assess Ggt localization, S. aureus ggt::Tn pOS1 P_lgt::ggt (His -) and ggt::Tn pOS1 P_lgt::ggt-His (His +) cultures were prepared as previously described in the proliferation analysis section by sub-culturing into three, 250 mL flasks each containing 100 mL PN supplemented 25 μM GSSG and 10 μg mL^-1 chloramphenicol at a starting OD₆₀₀ equal to 0.1. Cells were cultured for 4 h at 37° C and 225 rpm shaking to mid-exponential phase. At this time cells were collected via centrifugation, the supernatant recovered, and the cell pellet washed with PBS. A total of 50 mL of supernatant was precipitated with trichloracetic acid (TCA, final percent of 10% v/v), incubated overnight at 4° C, pelleted, and the resulting pellet washed twice with 95% ethanol. Whole cell lysates (WCL) were prepared by resuspending the washed cell pellet in membrane buffer (50 mM Tris-HCl pH 7.0, 10 mM MgCl₂, 60 mM KCl) and transferring the suspension to a 2 mL bead beating tube containing 500 μL volume of 0.1 mm zirconia/silica beads (BioSpec Bartlesville, OK). Cells were bead beaten using a Mini-beadbeater 16 (BioSpec) thrice for 1 min and centrifuged 4,000 x g for 10 min to remove cells that were not lysed. Cell wall and lysed protoplast fractions were generated from whole cells cultured as described above. Staphylococci were pelleted via centrifugation, resuspended in TSM (100 mM Tris-HCl pH 7.0, 500 mM sucrose, 10 mM MgCl₂), and incubated with 100 μg lysostaphin for 1 h at 37° C. Protoplasts were recovered by centrifugation at 13,000 x g for 15 min. The supernatant containing the cell wall fraction was collected and the protoplasts were washed twice with TSM prior to resuspension in membrane buffer and lysed via bead beating as previously described. Intact protoplasts were removed via centrifugation at 4,000 x g for 10 min, generating the protoplast lysate. Protein concentrations of the whole cell lysate (WCL), cell wall, and protoplast lysate were determined using the BCA method (Pierce ThermoFisher). The cell wall fraction was further concentrated using the TCA precipitation method previously described for the supernatant. WCL and lysed protoplast fractions were mixed 1:1 by volume with 2x Laemmli buffer, boiled for 10 min, and an equivalent of 40 μg was loaded onto a 12% SDS-PAGE gel. Concentrated cell wall pellet was resuspended in 1x Laemmli buffer, boiled for 10 min, and 40 μg was loaded onto a 12% SDS-PAGE gel. PAGE was performed using Tris-glycine running buffer and samples were transferred at 65 volts for 1 h to a PVDF membrane (GVS North America) at 4° C. Membranes were incubated overnight in phosphate buffered saline tween-20 (PBST) with 3% bovine serum albumin (BSA) at 4° C with agitation. An αHis mouse antibody (Millipore) was used as the primary antibody at a 1:4,000 dilution in PBST supplemented with 5% BSA and incubated for 1 h with shaking. The membrane was washed thrice with PBST. An α-mouse IgG conjugated to horseradish peroxidase (HRP) was used as the secondary antibody at a dilution of 1:4,000 (Sigma-Aldrich). To assess supernatant and WCL fraction a rabbit anti-α-hemolysin (αHla) primary antibody (Sigma-Aldrich) was used at a 1:8,000 dilution. To control for proper cell wall fractionation a mouse anti-protein A (αSpa) primary antibody (Sigma-Aldrich) was used at 1:6,000 dilution. Membranes were washed thrice in PBST after incubation with either a 1:10,000 diluted horseradish peroxidase (HRP)- linked goat anti-rabbit IgG (Sigma) or 1:5,000 diluted HRP-linked goat anti-mouse IgG (Millipore) secondary antibodies. Membranes were developed using the ECL Prime kit (Cytiva, Marlborough, MA) and imaged using Amersham Imager 600 (GE Healthcare, Amersham, Buckinghamshire, UK). Fractionation was repeated with three independent sets of cultures and one set of immunoblots is presented.

Proteinase K protection of S. aureus Ggt.

S. aureus ggt::Tn pOS1 P_lgt::ggt and ggt::Tn pOS1 P_lgt::ggt-His were cultured as described in the western blot analysis section. Cells were collected via centrifugation after normalizing OD₆₀₀. The resulting pellet was washed with PBS, resuspended in TSM, and incubated with 100 μg lysostaphin for 1 h at 37° C prior to treatment with 10 μg mL^-1 proteinase K (Thermo-Scientific) for 30 min at 37° C. At this time 5 mM phenylmethylsulfonyl fluoride was added to halt Proteinase K activity. Protoplasts were isolated from the cell wall fraction via centrifugation at 13,000 x g at 4° C for 15 min and washed twice with TSM. The cell wall fraction protein concentration was determined using BCA and was TCA precipitated as previously described. Protoplasts were lysed, and protein concentrated using the Wizard Genomic DNA Purification Kit (Promega) nuclei lysis buffer and protein precipitation solution, respectively. Cell wall and protoplast protein pellets were reconstituted in Laemmli buffer and immunoblotted as described above.

Identification of GisABCD-Ggt homologues across bacteria.

The USA300_FPR3757 (assembly GCF_000013465.1) Ggt protein sequence (ABD22038.1) was used as the query protein for homology searches with MolEvolvR using DELTA-BLAST and the NCBI RefSeq database [36,75–77]. Data were filtered to include only Firmicutes that encoded Ggt homologues containing a glutamyl transpeptidase domain. Only genomes harboring Ggt homologues were used to query USA300_FPR3757 GisABCD. Percent similarities to the S. aureus GisABCD-Ggt protein amino acid sequences were used to generate a heatmap. The heatmap and hierarchical clustering of similar protein profiles were generated using the R package, pheatmap.

16S rRNA phylogenic analysis of Staphylococcal species.

16S ribosomal RNA DNA sequences were retrieved from NCBI for the following strains: S. aureus USA300 FPR3757, S. epidermidis RP62s, S. simiae NCTC 13838, S. schweitzeri NCTC13712, and S. argenteus 58113. Parsimonious reconstruction was conducted using kSNP4 using default parameters and selecting Macrococcus caseolyticus FDAARGOS_868 as the outgroup [78,79].

Quantitation of S. aureus and S. epidermidis competition.

S. aureus, S. aureus Δgis, and S. epidermidis were cultured overnight in TSB, pelleted, washed in PBS, and normalized to the same OD₆₀₀ in PN_mod. Strains were mixed in a 1:1 ratio (v/v) and inoculated into 5 mL PN_mod. PN_mod was supplemented with 25 μM GSSG, 50 μM GSH, 750 μM GSH, or 50 μM Met. Dilution plating of the freshly mixed co-culture was plated onto mannitol salt agar (MSA) to quantify initial counts of each organism. Cultures were incubated for 24 h at 37° C with 225 rpm shaking after which the cultures were dilution plated onto MSA and allowed to grow for 48 h at 35° C. S. aureus ferments mannitol and appears yellow on MSA, while S. epidermidis does not and maintains a pink color; consequently, yellow and pink colored colonies were enumerated to assess quantities of each organism. Competitive indices (CI) were calculated by dividing the S. aureus to S. epidermidis output ratio by the S. aureus to S. epidermidis input ratio. A CI greater than one indicates more S. aureus than S. epidermidis while a CI less than one signifies greater quantities of S. epidermidis compared to S. aureus.