Staphylococcus aureus exploits lipoic acid salvage to combat host oxidative stress

SUMMARY

Phagocytic leukocytes employ reactive oxygen species to defend against pathogenic microorganisms. The bacterial pathogen Staphylococcus aureus adapts to oxidative stress by producing antioxidant enzymes and small molecules to protect proteins, nucleic acids, and other essential cellular components. Here, we show that the lipoic acid carrier protein GcvH-L promotes S. aureus resistance to oxidative stress. The gene encoding GcvH-L lies within a conserved operon in several pathogenic microorganisms. The operon also encodes LplA2, a redox-responsive lipoyl ligase, and SirTM, an ADP-ribosyltransferase. We demonstrate that ADP-ribosylation of lipoyl-GcvH-L protects lipoic acid from oxidation and regulates its transfer from GcvH-L to enzyme complexes needed for central metabolism. A ΔgcvH-L mutant is attenuated during infection and is more sensitive to phagocyte respiratory burst, phenotypes that are abrogated in NADPH oxidase-deficient mice. Thus, ADP-ribosylation and lipoylation converge on GcvH-L to promote S. aureus resistance to oxidative stress.

In brief

Acosta et al. find that ADP-ribosylation of the lipoic acid carrier protein GcvH-L promotes Staphylococcus aureus resistance to oxidative burst and facilitates infection. ADP-ribosylation leads to a structural shift within GcvH-L that safeguards lipoic acid sulfhydryls from detrimental oxidation and regulates its transfer to enzyme complexes required for growth.

Graphical Abstract

INTRODUCTION

The innate immune system controls bacterial infection through the coordinated actions of phagocytic leukocytes.^1,2 These cells generate large amounts of antimicrobial free radicals, including reactive oxygen species (ROS), via the activity of nicotinamide adenine dinucleotide phosphate (NADPH) oxidase in a process known as respiratory burst.^3,4 NADPH oxidase dysfunction renders individuals highly susceptible to infection with bacteria and fungi, underscoring the relevance of respiratory burst to infection and its role as a potent antimicrobial.^3,5–7 Bacterial pathogens have adapted in several ways to evade the detrimental effects of respiratory burst. For example, Staphylococcus aureus activates cellular repair pathways and synthesizes enzymes and small molecules that neutralize ROS.^8–10 In addition to neutralizing ROS, other small molecules play crucial roles in regulating intracellular redox homeostasis.^8,11 Low-molecular-weight thiols like coenzyme A and bacillithiol protect sulfur-containing proteins by interacting with their thiol groups, mitigating damage and facilitating the recovery of enzymatic function after stress.^12–15 Thus, the bacterial response to oxidative stress entails both neutralizing ROS and preventing oxidation of preexisting cellular components.¹⁶

Lipoic acid (LA) is an essential metabolic cofactor with antioxidant properties in vitro due to redox-sensitive sulfhydryls inserted at carbons 6 and 8 of its precursor, octanoic acid, a medium-chain fatty acid.¹⁷ In S. aureus, LA is covalently bound to a conserved lysine of the E2 subunits of pyruvate dehydrogenase (PDH), 2-oxoglutarate dehydrogenase (OGDH), branched-chain 2-oxoacid dehydrogenase (BCODH), as well as the H protein of the glycine cleavage system (GcvH) and a GcvH homolog of unknown function (GcvH-L).^18–20 S. aureus ensures LA bioavailability through de novo biosynthesis and salvage. Impairment in these pathways results in pronounced defects in glycolysis, tricarboxylic acid cycle activity, and branched-chain fatty acid metabolism, ultimately reducing bacterial fitness during infection.²⁰ Furthermore, the release of cytoplasmic lipoyl-E2-PDH to the extracellular space and de novo LA synthesis by S. aureus directly hampers macrophage activation and subsequent ROS-mediated killing.^10,21 In Mycobacterium tuberculosis, LA on E2 subunits of metabolic complexes accepts electrons from the oxidation of the thioredoxin-like protein AhpD and a peroxiredoxin, AhpC, to promote redox balance.^22–24 These findings argue for a direct link between LA and redox homeostasis. Nevertheless, the benefits of LA for oxidative stress resistance during infection or sterile inflammation are not clear, especially considering that the bioavailability of free LA in the host is low. This limited LA bioavailability during infection and its role as a critical cofactor in central carbon metabolism also raises the question of whether bacteria have developed mechanisms to prevent detrimental oxidation of LA sulfhydryls.^25–29

We previously determined that S. aureus requires LA salvage to establish infection.²⁰ LA salvage in S. aureus is mediated by two ligases, LplA1 and LplA2 (Figure 1A).^20,30 At the genetic level, the lplA1 gene exists as a single open reading frame, and its activity predominates under standard growth conditions, while lplA2 is part of a five-gene operon (Figure 1B).²⁰ The lplA2 operon contains open reading frames encoding a sirtuin, SirTM, which has ADP-ribosylating activity; an ADP-ribosyl hydrolase, Macro; GcvH-L; and a luciferase-like monooxygenase (LLM).^19,20,31 GcvH-L can participate in LA salvage, where it is used as a substrate for LA transfer to E2 subunits, but its position within a multi-gene operon argues for potential unknown activities (Figure 1A).^30,32 In a prior biochemical study, Rack et al. found that SirTM ADP-ribosylates GcvH-L of S. aureus and Streptococcus pyogenes only after prior lipoylation by LplA2, implying a crosstalk between the two posttranslational modifications (PTMs); however, no physiological effect was uncovered in either bacterium.¹⁹ ADP-ribosylation is predominantly studied in eukaryotes, where it modulates many cellular processes, including immune signaling and oxidative stress responses.^33,34 It is also exploited by several bacterial pathogens, which secrete toxins and effectors that ADP-ribosylate host proteins to promote pathogen survival.^35–38 In contrast, the roles of ADP-ribosylation within bacteria are comparatively unexplored,^39,40 and the biological significance of the convergence of lipoylation and ADP-ribosylation on GcvH-L is not known; however, the literature on ADP-ribosylation in eukaryotes implies a potential role in stress responses.

Figure 1. The lplA2 operon mediates resistance to ROS.

(A) Model of the LA salvage pathway in S. aureus.

(B) Representation of the genetic organization of the lipoyl-protein ligase genes.

(C) Growth curves of the indicated strains in RPMI medium for 12 h.

(D) α-LA immunoblots of whole-cell lysates of the indicated strains after 9 h growth in RPMI medium (top) and Coomassie-stained gel (bottom). An arrow denotes the position of E2-PDH for comparison.

(E) Susceptibility of the indicated strains to 30 μM NaOCl (n = 9 biological replicates).

(F) Susceptibility of the indicated strains to 30 μM NaOCl (n = 5 biological replicates).

(G) Susceptibility of the indicated strains to 15 mM H₂O₂ (n = 4 biological replicates).

(H and I) The minimum inhibitory concentration of the indicated strains in cefazolin and trimethoprim-sulfamethoxazole (n = 4 biological replicates). y axis (I), trimethoprim and sulfamethoxazole concentrations.

All biological replicates were conducted in technical triplicate. Boxplots indicate the median and quartiles, and whiskers indicate the range. Statistical significance was determined by one-way ANOVA with Dunnett’s multiple comparisons post hoc test. **p < 0.01, ***p < 0.001, ****p < 0.0001.

We discovered that ADP-ribosylation of GcvH-L is critical for S. aureus resistance to oxidative stress in vitro and in vivo. ADP-ribosylation protects LA from oxidation and regulates its transfer to E2 subunits of major metabolic enzymes, allowing rapid redistribution of the functional cofactor. GcvH-L is required for bacterial survival in the presence of phagocytic leukocytes as well as for infection in an NADPH oxidase-dependent manner. Furthermore, biochemical studies indicate that operon-encoded LplA2 is a redox-sensitive lipoyl ligase with improved activity in oxidizing conditions. Altogether, we propose that the components of the lplA2 operon constitute a redox-sensitive molecular switch that responds to the oxidative state of the cell to protect LA from oxidative damage and promote rapid cellular recovery by regulating the delivery of reduced LA to metabolic enzymes.

RESULTS

The lplA2 operon mediates resistance to oxidative stress

To investigate the role of the lplA2 operon in the oxidative stress response of S. aureus, we generated in-frame deletions of each gene as well as the monocistronic lplA1 gene in the methicillin-resistant S. aureus strain LAC (Figure 1B).⁴¹ All strains grew in LA-deficient medium and had similar lipoylation on metabolic E2 subunits and GcvH (Figures 1C and 1D). After a challenge with 30 μM sodium hypochlorite (NaOCl), the ΔgcvH-L, ΔsirTM, and ΔlplA2 mutants each had a 95- to 1000-fold reduction in survival relative to the wild-type (WT) strain. The Δllm mutant had a marginal but statistically significant reduction in colony-forming units (CFUs) with significant variability between experiments. No significant differences were observed with the Δmacro and ΔlplA1 strains relative to the WT (Figure 1E). We generated single-copy chromosomal complementation strains from the two mutants that exhibited consistent susceptibility, ΔgcvH-L and ΔsirTM, and tested resistance to 30 μM NaOCl. The CFUs recovered from both complement strains resembled those of the WT strain (Figure 1F). The ΔgcvH-L and ΔsirTM mutants were also more sensitive (10-fold reduction in CFUs) to exposure to 15 mM H₂O₂ (Figure 1G) but not to unrelated antimicrobial agents, including the antibiotics cefazolin and trimethoprim-sulfamethoxazole (Figures 1H and 1I). These data imply a specific role of the lplA2 operon in promoting oxidative stress resistance.

Resistance to oxidative stress requires PTMs on GcvH-L

GcvH-L is lipoylated by LplA2 and ADP-ribosylated by SirTM, underscoring a potential central role of these PTMs in the function of GcvH-L and oxidative stress resistance.^19,30 We complemented the ΔgcvH-L mutant with a constitutively expressed gcvH-L allele that lacks the LA modification (gcvH-L(K56A)), lacks the mono-ADP-ribose (MAR) modification (gcvH-L (D27A)), or contains the wt allele (gcvH-L) (Figure 2A).^19,30 Immunoblot analysis confirmed that all strains produced equal amounts of GcvH-L with no growth defect (Figures 2A, 2B, and S1A). The two GcvH-L variants migrated differently compared to the WT protein during SDS-PAGE. We attribute this to the loss of charged amino acids and resistance of GcvH-L to denaturation in SDS (see below).^42–44 We assessed the ability of these strains to resist 30 μM NaOCl and 15 mM H₂O₂ and found that the ΔgcvH-L + gcvH-L(K56A) and ΔgcvH-L + gcvH-L(D27A) strains remained sensitive to ROS, as evidenced by 100- to 1,000-fold fewer CFUs compared to the ΔgcvH-L + gcvH-L strain (Figures 2C and 2D). These results indicate that the PTMs on GcvH-L are required to promote oxidative stress resistance in S. aureus.

Figure 2. PTMs on GcvH-L promote resistance to oxidative stress.

(A) Representation of the amino acids involved in the attachment of LA (K56) and mono-ADP-ribose (MAR) (D27) to GcvH-L. Shown are immunoblots against GcvH-L and LA from whole-cell lysates of the Δsbi Δspa ΔgcvH-L mutant containing pJC1111-6xhis-gcvH-L, pJC1111-6xhis-gcvH-L (K56A), and pJC1111-6xhis-gcvH-L(D27A).

(B) Growth of the indicated strains (described in A) in RPMI medium for 12 h.

(C and D) Susceptibility of the same strains to 30 μM NaOCl (n = 5 biological replicates) and 15 mM H₂O₂ (n = 4 biological replicates). All biological replicates were conducted in technical triplicates. Boxplots indicate the median and quartiles, and whiskers indicate the range. ****p < 0.0001 by one-way ANOVA with Dunnett’s post hoc test.

(E) Immunoblot for endogenous GcvH-L from whole-cell lysates of Δsbi Δspa and Δsbi Δspa ΔgcvH-L strains, untreated or treated with 2 mM NaOCl (top) and Coomassie-stained gel of the same samples (bottom).

(F) Immunoblots of GcvH-L isolated from a Δsbi Δspa ΔgcvH-L Δmacro ΔsirTM mutant containing pJC1111-6xhis-gcvH-L, pJC1111-6xhis-gcvH-L (K56A), or pJC1111-6xhis-gcvH-L (D27A). Bacterial cultures were treated with 2 mM NaOCl, and purified proteins were resolved under non-reducing and reducing conditions. A pink arrow signifies oxidation of the lipoyl moiety on GcvH-L. All immunoblots are representative of at least three independent experiments.

The LA on GcvH-L is oxidized during oxidative stress

The lplA2 operon is poorly expressed under standard culture conditions.^20,30 Prior transcriptome studies in S. aureus noted modest (2- and 4-fold) upregulation of the lplA2 operon under oxidative and nitrosamine stress, respectively.^45–47 We performed protein enrichment using trichloroacetic acid and immunoprecipitation studies from whole-cell lysates of S. aureus cultures, untreated and treated with 2 mM NaOCl, and assessed natively expressed GcvH-L abundance by immunoblot. We observed more GcvH-L in NaOCl-treated cultures compared to untreated cultures (Figures 2E and S1B); however, we could not detect lipoylation or ADP-ribosylation (Figure S1B), suggesting that protein levels were below the limit of detection of the α-ADPr and α-LA antibodies. Indeed, immunoblot analysis of recombinant ADP-ribosylated lipoyl-GcvH-L (MAR-GcvH-L) and lipoyl-GcvH-L revealed that the α-GcvH-L antibody detects GcvH-L at concentrations that are well below the limit of the α-ADPr and α-LA antibodies (Figures S1C and S1D). To improve the detection of lipoylation on GcvH-L and evaluate the potential oxidation of its lipoyl moiety in vivo, we constitutively expressed His₆-GcvH-L, His₆-GcvH-L(K56A), and His₆-GcvH-L(D27A) in a ΔgcvH-L Δmacro ΔsirTM mutant strain. We used nickel-nitrilotriacetic acid resin to enrich proteins from cultures treated with 2 mM NaOCl or mock treated, and isolated proteins were resolved under reducing and non-reducing conditions. As expected, GcvH-L(K56A) was not lipoylated, whereas GcvH-L was lipoylated (Figures 2A and S1E). Immunoblot analysis using an α-GcvH-L antibody showed that NaOCl treatment resulted in the appearance of a higher-molecular-weight GcvH-L species for both GcvH-L and GcvH-L(D27A) under non-reducing conditions. These higher-molecular-weight species were not present under reducing conditions (Figure 2F). In contrast, GcvH-L (K56A) migration patterns did not change under either condition, indicating that this effect depended on the presence of the lipoyl moiety (Figure 2F). Immunoblots against LA confirmed the presence of two species for GcvH-L under non-reducing conditions, whereas the higher-molecular-weight species were not present under reducing conditions (Figure S1E). Together, these results indicate that GcvH-L protein levels are increased during oxidative stress and that the lipoyl moiety can be oxidized in the presence of NaOCl in live cells.

SirTM ADP-ribosylates and interacts with GcvH-L

To determine whether SirTM ADP-ribosylates GcvH-L in live S. aureus, we generated a ΔgcvH-L Δmacro ΔsirTM mutant strain harboring the integrative plasmid pJC1111-P_Help-6xhis-gcvH-L or pJC1111-P_Help-6xhis-gcvH-L and pOS1-P_hrtAB-sirTM, with expression of sirTM driven by the hrtAB hemin-inducible promoter.⁴⁸ All strains had equivalent growth characteristics in 2 μM hemin (Figure S2A). GcvH-L isolated from the SirTM-expressing strain migrated as two species on an SDS-PAGE gel, one of which was positive for ADP-ribosylation by immunoblot (Figure 3A). To verify that ADP-ribosylation of GcvH-L is a direct result of SirTM activity, we purified recombinant His₆-SirTM and His₆-GcvH-L from E. coli and conducted in vitro ADP-ribosylation assays. The reaction containing SirTM and GcvH-L resulted in the ADP-ribosylation of approximately 50% of GcvH-L, as evidenced by a reduction in electrophoretic mobility and positive ADPr signal by immunoblot (Figures 3B and S2B). These results demonstrate that SirTM can ADP-ribosylate GcvH-L in S. aureus and reaffirm ADP-ribosyl transferase activity in vitro.¹⁹ We next pursued evidence for a SirTM GcvH-L interaction. Size exclusion chromatography revealed that SirTM and lipoyl-GcvH-L co-eluted in a single peak at ~9 mL (Figure 3C). Comparison to protein standards suggested an ~300 kDa complex. In contrast, SirTM alone eluted in three distinct oligomeric peaks (~278, ~175, and ~72 kDa), potentially reflecting octameric, pentameric, and dimeric states, respectively, while lipoyl-GcvH-L alone eluted as a single monomeric peak at ~22 kDa (Figures 3C and S2C). Immunoblot analysis of the elution fractions confirmed co-elution of SirTM and GcvH-L at 9 mL (Figure 3D). Together, these results suggest a potentially stable protein-protein interaction between GcvH-L and SirTM that leads to ADP-ribosylation.

Figure 3. SirTM, Macro, LplA2, and LplA1 are sensitive to redox state.

(A) ADP-ribosylation of GcvH-L in live bacteria. The Δsbi Δspa ΔgcvH-L Δmacro ΔsirTM strain (ΔΔΔ), bearing pJC1111-6xhis-gcvH-L (+gcvH-L), empty pOS1, or pJC1111-6xhis-gcvH-L and pOS1-P_hrtAB-sirTM (+gcvH-L + sirTM), was grown in RPMI medium supplemented with 2 μM hemin for 16 h before isolating His₆-GcvH-L.

(B) In vitro ADP-ribosylation assays.

(C) Size exclusion analyses of SirTM (blue line), GcvH-L (cyan line), or a 1:1 mixture of GcvH-L and SirTM (gray line). Numbers on the graph indicate approximate molecular weight (kDa) compared to a standard curve.

(D) Immunoblot analysis of the elution fractions from the size exclusion analyses.

(E and F) Effect of H₂O₂ titration on SirTM activity (ADP-ribosylation) (E) and Macro (ADP-ribosyl hydrolase) activity (F) on GcvH-L. Reactions containing 40 mM DTT were supplemented with increasing H₂O₂, followed by Coomassie staining and α-ADPr immunoblot.

(G and H) LplA2 (G) and LplA1 (H) activities were evaluated in the presence of 10 mM H₂O₂ or 20 mM DTT. Lipoylation of GcvH-L was determined by a downward gel shift after resolving proteins on a 15% SDS-PAGE gel and by α-LA immunoblots.

All immunoblot and SDS-PAGE gel images are representative of at least three independent experiments.

lplA2-operon encoded enzymes are redox sensitive

SirTM and GcvH-L promote S. aureus resistance to ROS (Figures 1 and 2). In addition, SirTM has four conserved cysteines with thiols that presumably coordinate zinc in the active site.^19,49 Macro, the lplA2-locus encoded ADP-ribosylhydrolase, contains 7 cysteines, including a di-cysteine motif at its C terminus.³¹ Thus, SirTM and Macro have high potential to modulate enzymatic activity in response to redox state, as enzymes that contain cysteines have reactive thiol groups that are susceptible to oxidation and reduction, acting as redox switches^50–54 (Figure S2D). We tested whether redox state might regulate SirTM activity using in vitro ADP-ribosylation assays in the presence of several oxidizing and reducing agents. GcvH-L was most extensively ADP-ribosylated after incubation with the reducing agent DTT, with optimal activity at concentrations above 10 mM (Figures S2E and S2F). In contrast, adding excess H₂O₂ blocked ADP-ribosylation (Figure 3E). This inhibitory effect of H₂O₂ was also seen for Macro hydrolase activity on MAR-GcvH-L (Figure 3F), indicating that SirTM and Macro activities are potentially responsive to the local redox environment.

The lipoyl ligases of the parasite Plasmodium falciparum rely on a reducing environment to lipoylate E2 subunits in the mitochondria.^29,55 Since S. aureus produces two LA ligases, we speculated that redox state might differentiate the activity of LplA2, given its direct involvement in oxidative stress resistance and the presence of four potential reactive cysteine residues (Figures 1D and S2D). We conducted lipoylation assays with 10 mM H₂O₂ or 20 mM DTT using recombinant apo His₆-GcvH-L, His₆-LplA1, and His₆-LplA2 from E. coli. Lipoylation was evaluated by immunoblot using an α-LA antibody and by assessing downward shifts in migration of lipoyl-GcvH-L compared to apo GcvH-L.³⁰ LplA2-mediated lipoylation was more efficient in the presence of H₂O₂ relative to DTT, with complete lipoylation occurring within 10 min, whereas the reaction with DTT required more than 40 min to completely lipoylate apo GcvH-L (Figures 3G and S2G). In contrast, LplA1-mediated lipoylation was reduced in the presence of H₂O₂ relative to DTT. The reaction with H₂O₂ required more than 40 min to fully lipoylate apo GcvH-L, while the reaction with DTT achieved full lipoylation within 20 min (Figures 3H and S2G). These findings suggest that LplA1 and LplA2 are redox-sensitive enzymes with optimal activities under reducing and oxidizing conditions, respectively.

ADP-ribosylation on GcvH-L protects LA from oxidation

Structural modeling of GcvH-L shows close spatial proximity between the residues required for ADP-ribosylation (D27) and lipoylation (K56) on GcvH-L,⁵⁶ where the lipoyl moiety is located near a predicted hydrophobic cavity that is potentially suitable to accommodate the cofactor (Figures 4A and 4B). We reasoned that ADP-ribosylation could provide structural cues that might protect the lipoyl moiety on GcvH-L from oxidation. We purified recombinant apo His₆-GcvH-L, generated lipoyl-GcvH-L, and assessed the oxidation state of LA in vitro. Lipoyl-GcvH-L treated with 10 mM H₂O₂ had reduced electrophoretic mobility under non-reducing conditions relative to untreated lipoyl-GcvH-L. This reduced mobility depended on the presence of the lipoyl moiety, as apo GcvH-L did not change migration pattern (Figure S3A). To determine whether ADP-ribosylation facilitates protection of the lipoyl moiety of GcvH-L from oxidation, we assessed the oxidation state of LA on lipoyl-GcvH-L, MAR-GcvH-L, and lipoyl-GcvH-L(D27A) using the thiol-reactive compound 4-acetamido-4′-maleimidylstilbene-2,2′-disulfonic acid (AMS), which reduces electrophoretic migration of proteins in SDS-PAGE gels upon reacting with free thiols.⁵⁷ Untreated lipoyl-GcvH-L and lipoyl-GcvH-L(D27A) displayed a shift in molecular weight after treatment with AMS, indicative of the interaction of AMS with free thiols (Figures 4C and S3B). In contrast, most lipoyl-GcvH-L and lipoyl-GcvH-L(D27A) treated with 10 mM H₂O₂ did not shift on account of thiol oxidation. H₂O₂-treated MAR-GcvH-L migrated near the untreated control (Figures 4C and S3B), indicating that ADP-ribosylation on GcvH-L facilitates protection of the lipoyl moiety from oxidation. In an aqueous 8 M urea solution, proteins unfold with the subsequent exposure of the hydrophobic residues toward the solvent.⁵⁸ We reasoned that the protective effect of ADP-ribosylation on LA might coincide with structural changes in GcvH-L that preclude access of oxidizing agents to bound LA in the hydrophobic cavity. To test this possibility, we first fully denatured GcvH-L in 8 M urea at 90°C, as evidenced by complete exposure of its His₆ epitope tag relative to non-treated GcvH-L (Figure S3C). Denaturing MAR-GcvH-L prior to H₂O₂ treatment did not result in an AMS-dependent shift in molecular weight compared to the untreated control, indicating that ADP-ribosylation no longer protected the sulfhydryls of the lipoyl moiety from oxidation (Figures 4D and S3D). These data indicate that ADP-ribosylation promotes protection of LA from oxidation in a manner that depends on the structural properties of GcvH-L.

Figure 4. ADP-ribosylation on GcvH-L protects its lipoyl moiety from oxidation.

(A) AlphaFold-generated model of S. aureus GcvH-L threaded on GcvH-L from S. pyogenes (PDB: 5a35). Lysine 56 and aspartate 27 are highlighted in yellow and magenta, respectively.

(B) GcvH-L contains a hydrophobic pocket near lysine 56. The model illustrates the relative hydrophobicity of the pocket, where the most hydrophobic residues are shown in red, and less hydrophobic amino acids are displayed in gray.

(C) Determination of the redox state of LA. MAR-GcvH-L, lipoyl-GcvH-L, and lipoyl-GcvH-L (D27A) were treated with 10 mM H₂O₂ and labeled with AMS.

(D) Determination of the redox state of LA by AMS labeling non-denatured and denatured lipoyl-GcvH-L compared to MAR-GcvH-L. Purified proteins were treated with 8 M urea, heated at 90°C for 30 min prior to treatment with 10 mM H₂O₂, and labeled with AMS. Blots were probed with α-ADPr and α-GcvH-L antibodies. A green arrow signifies the shift of MAR-GcvH-L upon AMS labeling.

(E and F) NaOCl (30 μM, n = 6 biological replicates) and H₂O₂ (15 mM H₂O₂, n = 7 biological replicates) susceptibility of the Δsbi Δspa ΔgcvH-L Δmacro ΔsirTM strain transformed with (1) pJC1111-6xhis-gcvH-L (+gcvH-L), (2) pOS1-P_hrtAB-sirTM (+sirTM), or (3) pJC1111-6xhis-gcvH-L+ pOS1-P_hrtAB-sirTM (+gcvH-L + sirTM). All biological replicates were conducted in technical triplicates. Boxplots indicate the median and quartiles, and whiskers indicate the range. Statistical significance was determined by Kruskal-Wallis test (NaOCl) or one-way ANOVA (H₂O₂) with Dunn’s or Dunnett’s multiple comparisons post hoc test. **p < 0.01, ***p < 0.001, ****p < 0.0001.

(G) Immunoblot for endogenous GcvH-L from whole-cell lysates of Δsbi Δspa, Δsbi Δspa ΔsirTM, and Δsbi Δspa ΔgcvH-L (negative control) lysates, untreated or treated with 2 mM NaOCl (top) and a Coomassie-stained gel of the same samples (bottom).

All immunoblot and SDS-PAGE gel images are representative of at least three independent experiments.

To determine whether the protective effect of SirTM-dependent ADP-ribosylation of GcvH-L is relevant in vivo, we evaluated the susceptibility of the ΔgcvH-L Δmacro ΔsirTM mutant strain expressing (1) sirTM, (2) 6xhis-gcvH-L, or (3) both 6xhis-gcvH-L and sirTM to 30 μM NaOCl and 15 mM H₂O₂. S. aureus expressing sirTM and gcvH-L exhibited significantly greater resistance to NaOCl and H₂O₂ compared to the strains expressing only sirTM (Figures 4E and 4F). Further evidence of a protective effect of SirTM on GcvH-L in vivo included its stabilization in the presence of 2 mM NaOCl, where the ΔsirTM mutant had significantly reduced cytosolic GcvH-L compared with the WT strain following NaOCl treatment (Figure 4G). Despite multiple attempts, we could not detect endogenous MAR-GcvH-L or lipoyl-GcvH-L, likely due to the low concentrations of GcvH-L in S. aureus during growth in culture. These data further support the relevance of SirTM and GcvH-L to oxidative stress resistance.

ADP-ribosylation on GcvH-L controls the transfer of LA to E2 subunits of metabolic enzyme complexes

GcvH-L serves as a lipoyl donor to the E2 subunits of metabolic enzymes, with the amidotransferase LipL facilitating the transfer of LA from the H-protein to E2 subunits.³² When WT S. aureus was treated with 2 mM NaOCl, we noted reduced lipoylation of the E2 subunits (Figure S4A), indicating that oxidants reduce lipoylation of E2 subunits in vivo. We suspected that ADP-ribosylation of lipoyl-GcvH-L might regulate LA transfer to E2 subunits. We purified recombinant His₆-GcvH-L, His₆-GcvH-L(D27A), His₆-SirTM, His₆-Macro, His₆-LipL, and all apo E2 subunits (His₆-E2-OGDH, His₆-E2-PDH, and His₆-E2-BCODH) from E. coli (Figure S4B) and conducted LipL-mediated LA transfer assays (lipoyl-GcvH-L to E2 subunits) using MAR-GcvH-L or lipoyl-GcvH-L as substrates. We saw significantly reduced LA transfer to E2-OGDH, E2-PDH, and E2-BCODH when lipoyl-GcvH-L was ADP-ribosylated (Figures 5A–5C) but not when lipoyl-GcvH-L or lipoyl-GcvH-L(D27A) were used as substrates for LipL (Figures 5A–5C and S4C). Transfer efficiency was fully restored after adding the ADP-ribosyl hydrolase Macro, which removed the ADP-ribose (Figure 5D). These results indicate that ADP-ribosylation regulates the transfer of LA to E2 subunits in vitro.

Figure 5. ADP-ribosylation on GcvH-L prevents LipL-dependent lipoyl transfer to E2 subunits.

(A–C) LipL-dependent lipoylation of E2-OGDH, E2-PDH, and E2-BCODH using MAR-GcvH-L or lipoyl-GcvH-L (Lipoyl-GcvH-L) as a lipoyl donor.

(D) LipL-dependent lipoylation of E2-OGDH subunits after hydrolysis of ADP-ribose by Macro.

(E) Growth curves of the indicated strains in RPMI medium or RPMI medium supplemented with branched-chain carboxylic acids (BCCAs) to bypass LA auxotrophy.

(F) Immunoblot with α-LA and α-GcvH-L antibodies of whole-cell lysates derived from the indicated strains grown in RPMI medium (WT and ΔgcvH ΔsirTM) and RPMI medium + BCCA (ΔgcvH) (top) and Coomassie-stained gel of the same samples (bottom).

All immunoblot and SDS-PAGE gel images are representative of at least three independent experiments.

Low levels of LA are needed to support the growth of an S. aureus LA auxotroph (ΔlipA).^20,21 Indeed, we found that as little as 0.5 nM LA was sufficient for ΔlipA mutant growth (Figure S4D); thus, even a small reservoir of LA on GcvH-L could likely sustain LA salvage and bacterial growth if transfer is enabled. S. aureus encodes two H proteins: GcvH and GcvH-L.²⁰ A ΔgcvH mutant is unable to grow in LA-deficient medium, suggesting that GcvH-L is not used as a source of LA under these conditions.³⁰ To explore whether SirTM regulates transfer of LA from GcvH-L to E2 subunits in vivo, we assessed growth of WT, ΔgcvH, and ΔgcvH ΔsirTM strains in RPMI medium. The ΔgcvH ΔsirTM mutant grew after 16 h of incubation, whereas the ΔgcvH mutant did not (Figure 5E). The same strain had improved growth over the ΔgcvH mutant in a medium that bypasses the strict requirement of LA for growth (supplementation with branched-chain carboxylic acids [BCCAs]). Immunoblot analyses from whole-cell lysates showed more GcvH-L and clear evidence of lipoylation of the E2 subunits in the ΔgcvH ΔsirTM mutant (Figure 5F). These results indicate that SirTM can regulate LA transfer to E2 subunits in vivo.

GcvH-L is required for infection and resistance to ROS in vivo

We next determined whether GcvH-L is required for oxidative stress resistance and survival during infection. Mice were infected in the peritoneal cavity with WT, ΔgcvH-L, ΔgcvH-L+ gcvH-L (K56A), and ΔgcvH-L + gcvH-L strains, and bacteria were enumerated 24 h post infection. The ΔgcvH-L mutant was attenuated with 14 times fewer bacteria in the kidneys and 100 times fewer in the lavage fluid relative to the WT and the ΔgcvH-L + gcvH-L strain (Figure 6A). Consistent with the role of lipoylation on GcvH-L for oxidative stress resistance, a ΔgcvH-L+ gcvH-L(K56A) strain was attenuated to the same degree as the ΔgcvH-L mutant (Figure 6A). The ΔgcvH-L mutant was also attenuated in a bloodstream infection model, with 10 times fewer bacteria in the kidneys than the WT or the ΔgcvH-L + gcvH-L strain (Figure 6B). To determine whether the attenuation of the ΔgcvH-L strain was linked to increased sensitivity to oxidative stress, we employed Cybb⁻ mice, which lack the gp91^phox subunit of the NADPH oxidase and are used in models of X-linked chronic granulomatous disease.⁷ We first infected WT C57BL/6 mice and recapitulated the results with Swiss Webster mice, wherein the ΔgcvH-L strain had a 100-fold reduction in CFUs compared to mice infected with the WT (Figure 6C). At this infectious dose (1 × 10⁸ CFUs), Cybb⁻ mice rapidly succumbed to infection prior to 24 h, precluding assessment of CFUs. At a reduced inoculum of 10⁷ CFUs, the ΔgcvH-L mutant remained attenuated relative to the WT (Figure 6C). In contrast, the virulence defect of a ΔgcvH-L mutant was completely abrogated in Cybb⁻ mice (Figure 6C), supporting the notion that GcvH-L promotes resistance to respiratory burst in vivo.

Figure 6. GcvH-L mediates the resistance to respiratory burst in vivo.

(A) Bacterial recovery of the indicated strains from the kidneys and peritoneal cavity of Swiss Webster mice 24 h after intraperitoneal injection (n = 24). *p < 0.05 and ***, p < 0.001, calculated by the Kruskal-Wallis test with Dunn’s multiple comparisons test.

(B) Bacterial recovery of the indicated strains from the kidneys of C57BL/6 mice 24 h after bloodstream infection (n = 16). *p < 0.05 and **p < 0.01 by Kruskal-Wallis test with Dunn’s multiple comparisons test.

(C) Bacterial recovery of the indicated strains from the peritoneal cavity of C57BL/6 and Cybb⁻ mice 24 h after intraperitoneal injection with 1 × 10⁸ CFUs (n = 20) or 1 × 10⁷ CFUs (n = 10). **p < 0.01, ***p < 0.001 by unpaired t test (1 × 10⁸ CFUs) or Mann-Whitney test (1 × 10⁷ CFUs).

(D) Outgrowth of the indicated strains after infection with isolated macrophages from Swiss Webster mice.

(E) Outgrowth of the indicated strains after infection with isolated macrophages from Swiss Webster mice that were treated with vehicle control (water) or the NADPH oxidase inhibitor gp91ds-tat (50 μM).

(F) Outgrowth of the indicated strains after infection with isolated macrophages from C57BL/6 and Cybb⁻ mice. **p < 0.01, ***p < 0.001, ****p < 0.0001 by two-way ANOVA with Tukey’s post hoc test (D–F). Data are represented as individual values with the median indicated as a solid line.

GcvH-L mediates resistance to respiratory burst of macrophages

To test whether GcvH-L promotes resistance to ROS produced by phagocytic leukocytes, we conducted ex vivo infection experiments with macrophages (F4/80⁺) isolated from the peritoneal cavity of mice that were primed with the WT S. aureus 72 h before cell collection. ROS produced by F4/80⁺ cells restrict S. aureus outgrowth in ex vivo infection assays.^10,21 The isolated macrophages inhibited the outgrowth of the ΔgcvH-L, ΔgcvH-L + gcvH-L (K56A), and ΔgcvH-L + gcvH-L (D27A) strains more efficiently than those infected with the WT and ΔgcvH-L + gcvH-L strains (Figure 6D). The addition of the NADPH oxidase inhibitor, gp91ds-tat, to macrophages restored outgrowth rates of the ΔgcvH-L mutant to levels that were comparable to the WT and ΔgcvH-L + gcvH-L strains (Figure 6E). Consistent with in vivo infection and NADPH oxidase inhibitor studies, macrophages isolated from Cybb⁻ mice were unable to restrict the growth of the ΔgcvH-L strain relative to macrophages obtained from C57BL/6 mice, and all three strains had similar rates of outgrowth (Figure 6F). Collectively, these results indicate that GcvH-L and its modifications with LA and ADP-ribose are required for S. aureus to resist the respiratory burst of macrophages.

DISCUSSION

Here, we provide evidence arguing that the convergence of ADP-ribosylation and lipoylation on GcvH-L constitutes a molecular switch that confers resistance to ROS in S. aureus and promotes infection by evading the deleterious consequences of the respiratory burst of phagocytic leukocytes. Our results support a model wherein S. aureus exploits redox-responsive accessory LA salvage machinery to establish a reservoir of reduced LA that is stalled at GcvH-L during periods of stress. Mechanistically, we provide supportive evidence that argues that SirTM ADP-ribosylation or interaction with lipoyl-GcvH-L leads to shielding of LA from detrimental oxidation, stabilization of lipoyl-GcvH-L protein, and control of LA redistribution to E2 subunits of critical metabolic enzymes (Figure 7). This discovery underscores the relevance of the lplA2 operon during S. aureus infection and expands upon the more simplistic model that LA salvage pathways in bacteria are functionally redundant.^19,20,30,59 Because one single mechanism of ROS resistance is often not sufficient to withstand prolonged periods of ROS exposure, S. aureus uses an integrated cellular response that neutralizes free radicals, repairs cellular damage, and protects preexisting essential cellular components such as LA from detrimental oxidation.^2,16,60,61 There are currently few mechanistic studies of the physiological effects of endogenous ADP-ribosylation in bacteria.^62–64 Our findings highlight a compelling example of bacterial regulation of oxidative stress responses and LA salvage via ADP-ribosylation.

Figure 7. Proposed model of GcvH-L-mediated resistance to the respiratory burst of phagocytic leukocytes.

(A) Upon infection, S. aureus is phagocytosed by macrophages and other phagocytic leukocytes; it resists respiratory burst in the phagolysosome by (1) inducing LplA2-dependent lipoylation of GcvH-L. (2) LA readily becomes oxidized unless (3) ADP-ribosylation coupled with interactions with SirTM safeguards lipoyl-GcvH-L from detrimental oxidation and prevents LA transfer to E2 subunits.

(B) Upon overcoming respiratory burst, S. aureus escapes from the phagosome, and replication is initiated by (4) Macro-dependent ADP-ribosylhydrolase activity on GcvH-L, allowing (5) LipL-mediated transfer of LA from GcvH-L to E2 subunits and reinitiation of lipoylation-dependent metabolic pathways.

The AlphaFold model of S. aureus GcvH-L closely aligns with the crystal structure of S. pyogenes GcvH-L and emphasizes the spatial proximity between the residues required for ADP-ribosylation (D27) and lipoylation (K56) (Figure 4A).⁵⁶ Furthermore, we noted the presence of a hydrophobic pocket near the lipoylation site (Figure 4B). We suspect that ADP-ribosylation coupled with interactions with SirTM may position the lipoyl moiety within the hydrophobic pocket. Consequently, LA would be protected from detrimental oxidation by limiting the access of oxidants (Figures 4 and 3C). Evidence to support this molecular mechanism comes from size exclusion analyses with SirTM and GcvH-L (Figures 3C and 3D) and biochemical data, wherein denaturation of GcvH-L before H₂O₂ treatment completely eliminates the ability of ADP-ribosylation to protect the lipoyl moiety from oxidation (Figure 4D). The predicted localization of ADP-ribose within ~8 Å of the lipoyl moiety⁵⁶ likely also blocks substrate recognition by LipL, thereby preventing LA transfer to the E2 subunits during periods of redox stress (Figure 5). Notably, we provide several lines of evidence to support the proposed role of SirTM-dependent ADP-ribosylation of GcvH-L in oxidative stress resistance in S. aureus: (1) ΔsirTM and ΔgcvH-L mutants were sensitive to ROS (Figures 1E–1G), (2) ΔgcvH-L + gcvH-L (D27A) did not restore ROS resistance (Figures 2C and 2D), (3) co-expression of sirTM and gcvH-L increased ROS resistance (Figures 4E and 4F), and (4) deleting the sirTM gene reduced GcvH-L abundance during oxidative stress (Figure 4G). Additionally, deletion of sirTM in an LA auxotroph strain, ΔgcvH, enabled lipoyl transfer to the E2 subunits, likely via GcvH-L (Figures 5E and 5F).

LA is well known for its ability to scavenge ROS.⁶⁵ LA supplements have been linked to the relief of inflammatory symptoms in patients with oxidative stress-related diseases, such as diabetic neuropathy, osteoarthritis, and Alzheimer’s disease.^66–68 In microbial pathogens, the LA attached to E2 subunits and GcvH is a natural reducing agent for antioxidant proteins such as thioredoxins and peroxidases, where it supports enzymatic activity.^23,69–71 However, the mechanism by which GcvH-L protects against ROS in S. aureus differs from this conceptually simpler model of ROS scavenging. For ROS scavenging to be effective, the scavenging molecules must be in high cellular concentrations to neutralize ROS,⁷² whereas our data indicate that GcvH-L concentrations in the bacterial cell are low (Figures 2 and S1). In addition, the expression of gcvH-L(D27A), which is lipoylated but not ADP-ribosylated, does not restore the resistance to ROS, arguing that it is unlikely that lipoyl-GcvH-L acts as a ROS scavenger (Figure 2).

Impairment in 2-oxoacid dehydrogenase activity reduces bacterial fitness due to redox imbalance, compromised energy conservation, and decreased synthesis of metabolic intermediates required for growth.^59,73 Our data suggest that S. aureus protects LA from oxidative damage and regulates its transfer because its attachment to E2 subunits is crucial for the activity of the 2-oxoacid dehydrogenases required for central carbon metabolism and cell viability.⁷⁴ In support of this point, LA attached to E2 subunits is also protected from detrimental oxidation in eukaryotes.^75,76 The LA on GcvH is targeted by the toxic subproduct of lipid peroxidation, 4-hydroxy-2-nonenal, which forms a covalent linkage with the thiol groups of LA, compromising glycine-dependent respiration.^77,78 Eukaryotic cells counteract the deleterious oxidation of LA via the small antioxidant molecule glutathione, which interacts with the lipoyl moiety of the E2-OGDH, preventing its oxidation and facilitating recovery of enzyme activity once the stress is relieved.^14,79 In S. aureus, LA is safeguarded from detrimental oxidation by harnessing the genes of the lplA2 operon and allowing the cell to preserve a pool of reduced LA for rapid redistribution post stress, especially in states where de novo synthesis is compromised (de novo LA synthesis requires iron-sulfur cluster-containing enzymes that are also sensitive to ROS).^80,81

SirTM is a member of the class M ADP-ribosylating sirtuins known for their association with Macro domain proteins.¹⁹ M sirtuins are found in pathogenic microorganisms, suggesting roles in adapting to stresses imposed by the host.^19,62,82 While research on M sirtuins is limited in bacteria, the homologs in eukaryotes serve as cellular regulators that control several processes, such as cell cycle, DNA repair, metabolism, antiviral immunity, and tolerance to oxidative stress.^19,83,84 ADP-ribosylation significantly increases during oxidative stress in eukaryotic organisms, where several sirtuins with ADP-ribosylation activity are necessary to resist oxidative stress.^85–89 We suspect that, while ADP-ribosylating sirtuins in eukaryotes and bacteria both affect responses to oxidative stress, they likely do so in different ways. For instance, deletion of the sirTM-macro homolog, macrodomain-fused SirTM 1 (Mfs1), in Candida albicans enhances resistance to H₂O₂,¹⁹ which contrasts with our findings. In S. aureus, SirTM-mediated ADP-ribosylation on GcvH-L is required to resist oxidative stress, yet there is no evidence that Mfs1 ADP-ribosylates a GcvH or GcvH-like protein in C. albicans, nor is there a link to lipoylation. Further, the literature suggests that Mfs1 homologs may also be involved in other forms of stress, including metabolic stress,^90,91 suggesting that macrodomain-associated sirtuins in fungi could modify entirely different targets in the cell. It remains to be determined in what capacity ADP-ribosylation regulates other physiological processes in S. aureus or other pathogens that harbor M-type sirtuins. Thus far, our studies have centered on direct ADP-ribosylation of GcvH-L by SirTM, though it is conceivable that other proteins are ADP-ribosylated with unknown regulatory consequences.³¹ Future work will test this possibility.

Our biochemical assays demonstrated that SirTM, Macro, and LplA2 activities are regulated by the redox state of the environment (Figures 3 and S2). Support for the idea that redox chemistry can dictate lipoyl ligase activity comes from prior studies in P. falciparum, where a reducing environment was required for lipoylation of mitochondrial E2 subunits.^29,55 Thus, like in Plasmodium, the redox state is likely to result in dynamic changes to lipoylation and ADP-ribosylation frequency on GcvH-L. However, further studies are necessary to thoroughly assess how the activities of SirTM, Macro, and LplA2 in S. aureus change before, during, and after stress. Indeed, operon-encoded LLM contains two cysteines, potentially expanding the relevance of the redox state to proteins encoded within the lplA2 operon even further. A previous study found that GcvH-L directly interacts with Macro and the divergently encoded old yellow enzyme (OYE), a predicted oxidoreductase. These interactions required lipoylation of GcvH-L.¹⁹ The redox dependency of SirTM, Macro, and LplA2, along with this link between lipoyl-GcvH-L and OYE, implies the potential for OYE to regulate the activity of proteins encoded in the lplA2 operon. We are investigating this possibility.

LA biosynthesis and salvage pathways in S. aureus promote pathogenesis in a tissue-dependent manner.^10,20,21 Additional studies show that several microbial pathogens exclusively use the LA salvage pathway to acquire this cofactor during infection.^{25,27–29,55,92} LA salvage in S. aureus relies on the activity of two semi-redundant ligases (LplA1 and LplA2), facilitating kidney colonization during systemic infection.²⁰ LplA1 lipoylates GcvH-L and GcvH but not E2 subunits, whereas LplA2 lipoylates GcvH-L and E2 subunits in vitro, implying some divergence in function, although it is not clear whether this distinction in lipoylation targets occurs in vivo.^19,30 Nevertheless, the current study further expands on the idea that LplA1 and LplA2 are not functionally redundant enzymes. The opposing redox dependency of the two enzymes for optimal activity suggests a division of labor where LplA2 is functional under conditions of heightened ROS exposure, potentially to ensure that an adequate amount of LA is stored on GcvH-L (Figures 3G and 3H).

The attenuation of the ΔgcvH-L strain in WT mice and the abrogation of this defect in Cybb⁻ mice provide compelling evidence for a direct link between GcvH-L and resistance to respiratory burst in vivo (Figure 6). Our ex vivo studies with ROS-deficient macrophages corroborate these points (Figures 6D–6F). Moreover, the lack of complementation in ΔgcvH-L+ gcvH-L (K56A) during infection establishes the essential role of the lipoyl moiety on GcvH-L in promoting pathogenesis (Figure 6A). A notable feature of the bacterial response to oxidative stress is the precise control of central carbon metabolism to prevent the detrimental accumulation of endogenous ROS.² ROS targets enzymes in central carbon metabolism, including lipoyl-E2-PDH and lipoyl-E2-OGDH, resulting in energy depletion and growth arrest.^75,76,93,94 This is consistent with our results, which show that treating S. aureus with NaOCl reduces E2 subunit lipoylation levels (Figure S4A). Under standard growth conditions, GcvH-L is expressed at low levels; thus, it is conceivable that GcvH-L alone might not be sufficient to support restoration of LA-dependent growth after stress exposure (Figures 5E and 5F). However, the amount of LA required to restore an LA auxotroph, ΔlipA, is minimal (Figure S4D), suggesting that even low levels of GcvH-L could suffice for S. aureus to redistribute LA in restricted environments. We postulate that GcvH-L is poised as a minimal source of LA, allowing dynamic redistribution of LA on E2-PDH, E2-OGDH, and E2-BCODH, thereby preserving metabolic homeostasis during infection (Figure 7).^2,10,75,93

Collectively, this work supports a mechanism where SirTM-dependent ADP-ribosylation of GcvH-L shields LA and regulates its redistribution to improve fitness in response to respiratory burst. This discovery provides insight into the interface between microbial pathogens and this crucial component of innate immunity and highlights how bacteria modulate vital cellular processes through converging PTMs.

Limitations of this study

The technical limitations of this study include the low endogenous expression of the lplA2 operon in live S. aureus under standard laboratory culture conditions and the comparatively weak affinity of the ADP-ribose and LA antibodies. These limitations preclude us from conducting a detailed assessment of the proposed dynamic balance between ADP-ribosylation and macro-dependent ADP-ribosylhydrolase activity during oxidative stress. Furthermore, S. aureus ensures the bioavailability of LA through both de novo biosynthesis and two salvage ligases, complicating efforts to determine the precise division of labor between de novo synthesis and LplA1 or LplA2 ligases during infection. Despite these limitations, we remain confident that our model approximates what occurs in vivo, given (1) the sensitivity of the ΔsirTM and ΔgcvH-L mutants to ROS, (2) the requirement of SirTM for ADP-ribosylation in vitro and in vivo, (3) the attenuation of the ΔgcvH-L mutant during infection that is abrogated in Cybb⁻ mice, (4) the protective effect of inducing SirTM expression during oxidative stress, and (5) the SirTM-dependent restoration of growth and lipoylation to an LA auxotroph.

RESOURCE AVAILABILITY

Lead contact

Requests for further information, resources, and reagents should be directed to and will be fulfilled by the lead contact, Francis Alonzo (falonzo@uic.edu).

Materials availability

The bacterial strains or plasmids generated in this study will be provided upon request.

Data and code availability

• This paper does not report original code.

• Data reported in this paper will be shared by the lead contact upon request.

STAR★METHODS

EXPERIMENTAL MODEL AND SUBJECT DETAILS

Ethics statement

All animal experiments were performed in ABSL2 facilities with protocols approved by the University of Illinois Chicago, Animal Care Committee (Protocol #22–078), following guidelines set forth by the USDA and PHS Policy on Humane Care and Use of Laboratory Animals under the guidance of the Office of Laboratory Animal Welfare (OLAW). The University of Illinois Chicago has an Animal Assurance on file with the Public Health Service (#A3460–01) approved through (04/30/2027), is a fully AAALAC International accredited institution (#000186, certification dated (03/09/2021), and is a USDA registered/licensed institution (#33-R-0018).

Mouse experiments

Cohorts of seven-week-old female Hsd: ND4 Swiss Webster outbred mice were purchased from Envigo. Four-to-seven-week-old C57BL/6 and Cybb⁻ male mice were purchased from Jackson Laboratory. Mice were housed in groups of 4 in ventilated cages and provided with food and water ad libitum. Paper bedding was used in cages housing Cybb⁻ mice. Mice were anesthetized with a ketamine xylazine-cocktail (100 mg/kg ketamine and 10mg/kg xylazine) administered IP before infections.

Bacterial strains and growth conditions

All bacterial strains used in this study are listed in the key resources table. Escherichia coli strains were grown in Lysogeny Broth (LB) (BD), and S. aureus strains were grown in Trypticase Soy Broth (TSB) (BD) or Roswell Park Memorial Institute medium (RPMI) (Corning) supplemented with 1% casamino acids (VWR). Unless stipulated, strains were grown at 37°C, shaking at 200 rpm with tubes held at a 45° angle and maintaining a tube-to-medium ratio of 1:5. When needed, cultures were supplemented with the following agents for selection: ampicillin (100 μg/mL), chloramphenicol (10 μg/mL), cadmium chloride (0.1 mM), anhydrous tetracycline (1 μg/mL) (AnTet) (Acros Organics), and sodium citrate (10 mM) (Sigma). For experiments requiring the bypass of LA requirements for growth, branched-chain carboxylic acids (BCCA) were added to RPMI media at the following concentrations: 10.8 mM isobutyric acid, 9.2 mM 2-methylbutyric acid, and 9 mM isovaleric acid, as well as 10 mM sodium acetate (Sigma).

KEY RESOURCES TABLE.

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
Anti-Lipoic Acid Rabbit pAb antibody	Sigma-Aldrich	Cat # 437695; RRID: AB_212120
Anti-GcvH-L antibody	Pacific Immunology	N/A
Mono-ADP-Ribose antibody | AbD33204	Bio-Rad	Cat # HCA354
Anti-6X His tag® antibody [HIS.H8]	Abcam	Cat # ab18184; RRID: AB_444306
Anti-PBP2a of MRSA antibody, Mouse monoclonal	Sigma-Aldrich	Cat # SAB4200853
Goat anti-Rabbit IgG (H + L) Secondary antibody, AP	Invitrogen	Cat # 65-612-2
Goat anti-Rabbit IgG (H + L) Secondary antibody, HRP	Invitrogen	Cat # 65-6120; RRID: AB_2533967
Goat anti-Mouse IgG (H + L) Secondary antibody, AP	Invitrogen	Cat # PI31320
IgG from human serum	Sigma-Aldrich	Cat #I2511; RRID: AB_1163604
Mouse Swiss Webster serum	Innovative Research Inc	Cat # 50-203-5738
TruStain fcX(TM) (anti-mouse CD16/32) antibody (clone 93)	BioLegend	Cat # 101320; RRID: AB_1574975
Biotin anti-mouse F4/80 antibody (clone BM8)	BioLegend	Cat # 123106; RRID: AB_893501
Bacterial and virus strains
DH5α E. coli	New England Biolabs	Cat #C2989K
IM08B E. coli	Monk et al.95	N/A
lysY/Iq E. coli	New England Biolabs	Cat #C3013I
lysY/Iq ΔlipA::kan	Grayczyk et al.21	FA-E1344
E. coli lysY/Iq ΔlipA::kan + pET15b-6xhis-gcvH-L	Laczkovich et al.30	FA-E1383
E. coli lysY/Iq + pET15b-6xhis-gcvH-L(D27A)	This study	FA-E1810
E. coli lysY/Iq + pQ30-6xhis-sirTM	This study	FA-E3181
E. coli lysY/Iq + pQ30-6xhis-macro	This study	FA-E3336
E. coli lysY/Iq ΔlipA::kan + pET15b-6xhis-lplA1	Laczkovich et al.30	FA-E1284
E. coli lysY/Iq ΔlipA::kan + pET15b-6xhis-lplA2	Laczkovich et al.30	FA-E1278
E. coli lysY/Iq ΔlipA::kan + pET15b-6xhis-e2pdh	Laczkovich et al.30	FA-E1359
E. coli lysY/Iq ΔlipA::kan + pET15b-6xhis-e2ogdh	Laczkovich et al.30	FA-E1363
E. coli lysY/Iq ΔlipA::kan + pET15b-6xhis-e2bcodh	Laczkovich et al.30	FA-E1367
S. aureus USA300 strain LAC (WT)	Boles et al.41	N/A
NCTC8325 derivative RN4220	Fairweather et al.96	N/A
S. aureus RN9011	Chen et al.97	N/A
LAC Δllm	This study	FA-S1441
LAC ΔgcvH-L	Laczkovich et al.30	FA-S1435
LAC Δmacro	This study	FA-S2637
LAC ΔsirTM	This study	FA-S1592
LAC ΔlplA2	Zorzoli et al.20	FA-S1597
LAC ΔlplA1	Zorzoli et al.20	FA-S841
LAC ΔgcvH-L+gcvH-L	Laczkovich et al.30	FA-S1498
LAC ΔsirTM+sirTM	This study	FA-S1654
LAC ΔgcvH-L + pJC1111-6xhis-gcvH-L	This study	FA-S3375
LAC ΔgcvH-L + pJC1111-6xhis-gcvH-L(K56A)	This study	FA-S3377
LAC ΔgcvH::kan	Zorzoli et al.20	FA-S1038
LAC Δe2pdh::kan	Zorzoli et al.20	FA-S1041
LAC ΔlipA	Zorzoli et al.20	FA-S831
LAC ΔgcvH:: kan ΔsirTM	This study	FA-S2141
LAC Δsbi Δspa	This study	FA-S2365
LAC Δsbi Δspa ΔgcvH-L	This study	FA-S2450
LAC Δsbi Δspa ΔsirTM	This study	FA-S3547
LAC Δsbi Δspa ΔgcvH-L Δmacro ΔsirTM	This study	FA-S3106
LAC Δsbi Δspa ΔgcvH-L Δmacro ΔsirTM + pJC1111- 6xhis-gcvH-L	This study	FA-S3110
LAC Δsbi Δspa ΔgcvH-L Δmacro ΔsirTM + pJC1111-6xhis-gcvH-L(K56A)	This study	FA-S3112
LAC Δsbi Δspa ΔgcvH-L Δmacro ΔsirTM + pJC1111-6xhis-gcvH-L(D27A)	This study	FA-S3114
LAC Δsbi Δspa ΔgcvH-L Δmacro ΔsirTM + pOS1	This study	FA-S3186
LAC Δsbi Δspa ΔgcvH-L Δmacro ΔsirTM + pOS1-sirTM	This study	FA-S3130
LAC Δsbi Δspa ΔgcvH-L Δmacro ΔsirTM + pJC1111-6xhis-gcvH-L	This study	FA-S3110
LAC Δsbi Δspa ΔgcvH-L Δmacro ΔsirTM + 6xhis-gcvH-L + pOS1-sirTM	This study	FA-S3119
Chemicals, peptides, and recombinant proteins
RPMI 1640	Corning	Cat # 50-020-PC
RPMI 1640 1X	Corning	Cat # 10-040-CV
LB Broth, Miller (Luria-Bertani)	BD	Cat # 244610
Tryptic Soy Broth (Soybean-Casein Digest Medium)	BD	Cat # 211822
Casamino acids	VWR	Cat # 97063-378
Ampicillin sodium	Gold Biotechnology Inc	Cat # A-301-100
Chloramphenicol	VWR	Cat # 97061-244
Cadmium chloride hydrate	VWR	Cat # AA20129-14
Anhydrotetracycline hydrochloride	Thermo Scientific	Cat # AC233131000
Sodium citrate tribasic dihydrate	Sigma-Aldrich	Cat #S4641
Isobutyric acid	Sigma-Aldrich	Cat #I1754
Isovaleric acid	Sigma-Aldrich	Cat # 129542
(±)-2-Methylbutyric acid	VWR	Cat # AAA11546-AP
Sodium acetate, anhydrous ACS	VWR	Cat # 97061-994
Phusion High-Fidelity DNA Polymerase	Fisher Scientific	Cat #F530L
Tris(hydroxymethyl)aminomethane (TRIS, Trometamol), Ultrapure	VWR	Cat # 0497-5KG
Sodium chloride, high purity 5 KG	VWR	Cat # 97063-368
Gelatin, food grade, type A	MP Biomedicals™	Cat # 02901771-CF
Calcium chloride	VWR	Cat # 97062-590
Sodium hypochlorite solution, reagent grade, available chlorine 10-15%	Sigma-Aldrich	Cat # 425044-250ML
Hydrogen Peroxide	Sigma-Aldrich	Cat # 386790-100ML-M
Catalase	MP Biomedicals™	Cat # 02190311-CF
Trimethoprim,≥98% (HPLC)	Sigma-Aldrich	Cat #T7883-5G
Sulfamethoxazole	Sigma-Aldrich	Cat #S7507-10G
Cefazolin Sodium Salt	Sigma-Aldrich	Cat #C5020-100MG
Immobilon-PSQ PVDF Membrane	Sigma-Aldrich	Cat # ISEQ00010
Bovine Serum Albumin	Gold Biotechnology Inc	Cat # A-420-500
Acrylamide: Bis-Acrylamide 29:1 (40% Solution/Electrophoresis)	Fisher Scientific	Cat # BP1408-1
Sodium dodecyl sulfate (SDS)	VWR	Cat # 97064-474
PageRuler™ Prestained Protein Ladder, 10 to 180 kDa	Thermo Scientific	Cat # 26616
Invitrogen™ Novex™ Tris-Glycine Mini Protein Gels, 4-20%, 1.0 mm, WedgeWell™ format	Fisher Scientific	Cat # XP04205BOX
Electron Microscopy Sciences Genie™ Media Disruptor Beads 0.1 mm/Bacteria 8 OZ	Electron Microscopy Sciences	Cat # 50-341-02
Lysostaphin	AMBI PRODUCTS LLC	Cat # LSPN-50
Phosphate-Buffered Saline (PBS) 1X	Corning	Cat # 21-040-CM
Ni-NTA Agarose	Qiagen	Cat # 30230
Pierce™ Protein A Agarose	Thermo Scientific	Cat # 20333
DTT (Dithiothreitol)	Gold Biotechnology Inc	Cat # DTT10
IPTG	Gold Biotechnology Inc	Cat #I2481C
PMSF Protease Inhibitor	Thermo Scientific	Cat # 36978
Imidazole	Fisher Scientific	Cat #O3196-500
Gelcode™ Blue Safe Protein Stain	Thermo Scientific	Cat # 24596
Adenosine-5′ -triphosphate disodium salt (ATP)	Fisher Scientific	Cat # AAJ6112506
Zinc Chloride, Anhydrous	VWR	Cat #N882-100G
Zinc Acetate	Sigma-Aldrich	Cat # 383317-25G
Lipoic Acid	Sigma-Aldrich	Cat #T5625-5G
NAD Cofactor	Sigma-Aldrich	Cat # 20-221
Trichloroacetic Acid (TCA)	Fisher Scientific	Cat # A322-500
ENrich™ SEC 650 10 × 300 Column	Bio-Rad	Cat # 7801650
Superdex™ 75 Increase 10/300 GL	Cytiva	Cat # 29148721
Amicon® Ultra Centrifugal Filter, 50 kDa MWCO	Sigma-Aldrich	Cat # UFC8050
Gel Filtration Standard	Bio-Rad	Cat # 1511901
4-acetamido-4′ -maleimidylstilbene- 2,2′ -disulfonic acid, disodium salt	Thermo Scientific	Cat # A485
Urea	Sigma-Aldrich	Cat #U5128-5KG
Fetal Bovine Serum	VWR	Cat # 97068-085
Gp91 ds-tat	Anaspec	Cat # AS-63821
Critical commercial assays
QIAprep Spin Miniprep Kit	Qiagen	Cat # 27106
QIAGEN Plasmid Midi Kit	Qiagen	Cat # 12143
DNeasy Blood & Tissue Kit	Qiagen	Cat # 69506
QIAquick Gel Extraction Kit	Qiagen	Cat # 28706
SuperSignal™ West Femto Maximum Sensitivity Substrate	Thermo Scientific	Cat # PI34095
Pierce™ BCA Protein Assay Kit	Thermo Scientific	Cat # PI23250
Clean-Blot™ IP Detection Reagent (HRP)	Thermo Scientific	Cat # PI21230
QIAprep Spin Miniprep Kit	Qiagen	Cat # 27106
QIAGEN Plasmid Midi Kit	Qiagen	Cat # 12143
DNeasy Blood & Tissue Kit	Qiagen	Cat # 69506
QIAquick Gel Extraction Kit	Qiagen	Cat # 28706
SuperSignal™ West Femto Maximum Sensitivity Substrate	Thermo Scientific	Cat # PI34095
Pierce™ BCA Protein Assay Kit	Thermo Scientific	Cat # PI23250
Clean-Blot™ IP Detection Reagent (HRP)	Thermo Scientific	Cat # PI21230
Experimental models: Organisms/strains
C57BL/6	Jackson Laboratory	Cat # 000664
C57BL/6 Cybb−	Jackson Laboratory	Cat # 002365
Hsd:ND4 Swiss Webster mice	Envigo	Cat # 032
Oligonucleotides
See Table S1	-	-
Recombinant DNA
Plasmid: pET15b	Laboratory of Victor Torre (Novagen)	N/A
Plasmid: pQE30	QIAGEN N.V.	Cat # 32915
Plasmid: pIMAY	Monk IR et al.98	N/A
Plasmid: pJC1111	Chen et al.97	N/A
Plasmid: pOS1	Laboratory of Victor Torres	N/A
Software and algorithms
GraphPad Prism, Version 10	GraphPad Software Inc.	https://www.graphpad.com/
Excel	Microsoft Corporation	N/A
Word	Microsoft Corporation	N/A
Adobe Illustrator, Version 28.7.1	Adobe Inc	https://www.adobe.com/
The PyMOL Molecular Graphics System, Version 3.0	Schrödinger, LLC	https://www.pymol.org/
Snap Gene Viewer, Version 7.2.1	Dotmatics	https://www.snapgene.com/
ImageJ	NIH	https://imagej.net/ij/
ChromLab Software 6.1.29	Bio-Rad Laboratories Inc	https://www.bio-rad.com/
Image Lab™, Version 6.0.1	Bio-Rad Laboratories Inc	https://www.bio-rad.com/

METHOD DETAILS

Generation of in-frame deletion mutants

S. aureus isogenic mutants were generated according to the methodology described by Zorzoli et al. 2016.²⁰ Briefly, 500 bp upstream and downstream flanking regions adjacent to the start and the stop codon of each gene were amplified using genomic DNA as a template and the oligonucleotides listed in Table S1. Upstream and downstream amplicons were purified and used as templates for splicing by overlap extension (SOE) PCR to generate a final amplicon of approximately 1 Kb. This amplicon was ligated into the temperature-sensitive plasmid, pIMAY, and was subsequently transformed into E. coli IM08B.⁹⁵ Recombinant plasmids were verified by Sanger sequencing before transformation into the final recipient strains of S. aureus. Recombinant plasmids were transformed into S. aureus LAC via electroporation, and transformants were selected by growing on TSA + chloramphenicol at 30°C, which is a permissive temperature for plasmid replication.^98,99 For allelic exchange, the recombinant plasmid was first integrated into the bacterial chromosome by streaking a single colony on TSA + chloramphenicol and incubating at 37°C, a temperature that restricts pIMAY replication.^98,99 To stimulate plasmid excision, a single colony from a plasmid integrant was inoculated into 3 mL of TSB without chloramphenicol in a 15 mL conical tube and incubated at 30°C overnight. The bacterial culture was serially diluted in fresh TSB, and the resulting dilutions were plated on TSA supplemented with AnTet, which induces the expression of the secY antisense RNA for counter-selection. AnTet^R chloramphenicol-sensitive colonies were selected and the gene deletion was confirmed by PCR of genomic DNA and Sanger sequencing.

Generation of complement strains and phage transduction

Complement strains were generated using the plasmid pJC1111, which integrates at the attachment site of the S. aureus pathogenicity island 1 (SaPI1).⁹⁷ Open reading frames were amplified from S. aureus LAC genomic DNA, while the constitutive P_HELP promoter was amplified from the pIMAY plasmid using the primers listed in Table S1. The resulting amplicons were fused by SOE PCR, digested with PstI and SacI endonucleases, and ligated into pJC1111. The recombinant plasmid was subsequently transformed via electroporation into S. aureus RN9011, which harbors the pRN7203 plasmid that expresses the SaPI-1 integrase.^96,97 The integrated plasmids were then moved from the S. aureus RN9011 to the final recipient strain via phage transduction.

To collect phage, a single colony of the donor strain was inoculated in 3 mL of TSB in a 15 mL conical tube and grown overnight at 37°C with shaking. The bacterial culture was diluted in fresh TSB to a final OD₆₀₀ of 0.2. In a 15 mL conical tube, 5 mL of the standardized bacterial suspension was mixed with 5 mL of TMG buffer (10 mM Tris pH 7.5, 5 mM MgCl₂, 0.01% gelatin (v/v)) supplemented with 5 mM CaCl₂, and 100 μL of Φ₁₁ phage was added to the mixture. The tube was incubated at room temperature overnight or until the phage completely lysed the bacteria. Following incubation, the phage lysate was filtered through 0.45 μM and 0.22 μM syringe filters. For transduction, 1 mL overnight culture of the recipient strain grown in TSB was supplemented with 10 mM CaCl₂ and mixed with 100 μL phage lysate in a 15 mL conical tube. The tube was subsequently incubated at 37°C with shaking for 20 min. After incubation, the bacterial suspension was centrifuged at 7,000 g for 10 min, the supernatant was removed, and the pellet was washed thrice by centrifugation and resuspension with 1 mL of sterile 40 mM sodium citrate solution in ddH₂O. The bacterial pellet was resuspended in 40 mM sodium citrate and 100 μL of the bacterial suspension was plated on TSA supplemented with 0.1 mM CdCl₂ and 10 mM sodium citrate. Plates were incubated at 37°C for 24–48 h, and positive colonies were patched on a new plate to confirm the resistance to CdCl₂, followed by confirmation of pJC1111 integration by isolating genomic DNA, performing a PCR, and Sanger sequencing using pJC1111 primers (Table S1).

Generation of 6x-Histidine tagged SirTM, macro, and GcvH-L (D27A) expression plasmids for purification from Escherichia coli

The sirTM and macro genes were amplified from genomic DNA of S. aureus strain LAC using the primers listed in Table S1. The resulting amplicons were digested with KpnI and PstI endonucleases and ligated into the pQE30 (Qiagen) expression vector. The resulting plasmids were transformed into E. coli lysY/I^q, and transformants were selected on LB agar with 100 μg/mL ampicillin. The pET15b-gcvH-L(D27A) plasmid was generated by site-directed mutagenesis. A two-nucleotide change was introduced in the 5′-phosphorylated forward primer gcvH-L-D27A-F (key resources table). The 5′-phosphorylated reverse primer (gcvH-L-D27A-R) was designed to align the 5′ ends back-to-back with the forward primer. The entire plasmid was amplified using Phusion High-Fidelity DNA Polymerase (Thermo Scientific), followed by treatment with the methylation-dependent restriction enzyme DpnI (NEB) for 1 h at 37°C to eliminate the original plasmid template. The blunt-end PCR products were then ligated with T4 DNA ligase (NEB) overnight at 16°C. The ligated plasmid was transformed into lysY/I^q E. coli, and transformants were selected on LB agar with 100 μg/mL ampicillin. All plasmids were purified and sequenced to confirm the presence of the desired mutation.

Generation of 6xHis-GcvH-L, 6xHis-GcvH-L (K56A), and 6xHis-GcvH-L (D27A) strains

The 6xhis-gcvH-L, 6xhis-gcvH-L (K56A), and 6xhis-gcvH-L (D27A) genes were cloned into the pJC1111 plasmid under the control of the P_HELP promoter. A 6xHis tag was added to the N terminus of GcvH-L by including the sequence of the tag in the forward primer (Table S1). All three gcvH-L variants were amplified using his-gcvH-L-pJC1111-P3 and his-gcvH-L-pJC1111-P4-KpnI primers (Table S1) and the following DNA templates: (i) S. aureus LAC genomic DNA for 6xhis-gcvH-L, (ii) pJC1111-gcvH-L (K56A) for 6xhis-gcvH-L (K56A),³⁰ and (iii) pET15b-gcvH-L (D27A) for 6xhis-gcvH-L (D27A). Primers his-sirTM-pJC1111-P1-PstI and his-gcvH-L-pJC1111-P2 were used to amplify the P_HELP promoter from the pIMAY. The amplicons were fused by SOE PCR, digested with PstI and KpnI endonucleases, and ligated into pJC1111. Plasmids were then transformed into E. coli and subsequently verified by Sanger sequencing. The confirmed plasmids were transformed into the RN9011 strain to allow integration of the pJC1111 plasmid into the chromosome, followed by transduction into the following S. aureus recipient strains: Δsbi Δspa ΔgcvH-L and Δsbi Δspa ΔgcvH-L ΔsirTM Δmacro.

S. aureus growth curves

A single colony was inoculated into 3 mL of RPMI medium in a 15 mL conical tube and incubated overnight at 37°C with shaking. The next day, the strains were diluted 1:1,000 in fresh media, and 150 μL of the resulting bacterial suspension was added into a flat-bottom 96-well plate (Costar). The Δsbi Δspa ΔgcvH-L Δmacro ΔsirTM strains transformed with (i) pJC1111-6xhis-gcvH-L (+gcvH-L), (ii) pOS1-P_hrtAB-sirTM (+sirTM), or (iii) pJC1111-6xhis-gcvH-L+ pOS1-P_hrtAB-sirTM (+gcvH-L + sirTM) were cultured in RPMI supplemented with 2 μM hemin overnight. The following day, the strains were diluted 1:100 in fresh RPMI supplemented with 2 μM hemin. For experiments involving LA auxotrophs, ΔgcvH and ΔlipA, strains were initially grown overnight in RPMI supplemented with BCCA. The next morning, prior to dilution into fresh media, cells were harvested by centrifugation at 7,000 g for 10 min and washed four times with 3 mL of RPMI to remove any residual BCCA. After washing, cells were resuspended in 3 mL RPMI, and the bacterial strains were diluted 1:1,000 in fresh RPMI or RPMI supplemented with BCCA. Plates were incubated in a Tecan Spark plate reader at 37°C with orbital shaking at 3 mm with a frequency of 180 rpm. Bacterial growth was monitored by reading OD_600nm every hour.

Evaluation of bacterial susceptibility to sodium hypochlorite (NaOCl)

S. aureus strains were struck from −80°C stocks onto TSA and incubated overnight at 37°C. The next day, a single colony from the plate was inoculated into 10 mL of RPMI medium in a 50 mL conical tube and incubated for 16 h at 37°C with shaking. After incubation, cells were diluted to a final concentration of ~3 × 10⁶ CFU/ml in 10 mL of PBS, and the standardized bacterial suspension was then divided into two 15 mL conical tubes, each containing 5 mL. One tube was treated with NaOCl (Sigma) at a final concentration of 30 μM, while the other tube remained untreated. Both tubes were incubated for 20 min at 37°C with shaking. Non-treated and NaOCl-treated bacterial suspensions were serially diluted in 1X PBS and plated onto LB agar. Plates were incubated at 37°C for 20 h for CFU enumeration. Results are presented as percentage survival, which was determined by dividing the CFU recovered from the treated sample by the CFU in the untreated sample and multiplying by 100.

Evaluation of bacterial susceptibility to hydrogen peroxide (H₂O₂)

The growth conditions and inoculum standardization used for testing the susceptibility to H₂O₂ were the same as those described for NaOCl. For the experiment testing the Δsbi Δspa ΔgcvH-L Δmacro ΔsirTM strains transformed with (i) pJC1111-6xhis-gcvH-L (+gcvH-L), (ii) pOS1-P_hrtAB-sirTM (+sirTM), or (iii) pJC1111-6xhis-gcvH-L+ pOS1-P_hrtAB-sirTM (+gcvH-L + sirTM), the three strains were grown in RPMI supplemented with 2 μM hemin and the caps of the 50 mL conical tubes were securely sealed to minimize gas exchange and mitigate hemin toxicity during H₂O₂ treatment. Standardized bacterial suspensions were treated with H₂O₂ (Sigma) at a final concentration of 15 mM, and the tubes were incubated at 37°C for 1 h with shaking. After 1 h, the H₂O₂ was neutralized by adding 210 U of catalase (MP Biomedicals) to each tube. The tubes were incubated for 5 min at room temperature, then vortexed for 1 min. The neutralized samples were serially diluted in 1 X PBS and plated on LB agar. Plates were incubated at 37°C for 20 h for CFU enumeration. Results are presented as percentage survival, which was determined by dividing the CFU recovered from the treated sample by the CFU in the untreated sample and multiplying by 100.

Determination of minimum inhibitory concentrations (MIC)

The MIC was determined following the microdilution technique described by Andrews et al. with minor modifications.¹⁰⁰ Bacterial cultures grown in RPMI medium overnight were adjusted to an OD₆₀₀ of 0.3 and subsequently diluted 1:100 in fresh medium to obtain a final bacterial concentration of approximately 1 × 10⁶ CFU/mL. 75 μL of triplicate standardized bacterial suspensions were mixed with 75 μL of fresh RPMI, which contained antibiotic concentrations in a 2-fold dilution series in a 96-well plate. The 96-well plate was incubated for 16 h at 37°C without shaking under humid conditions to avoid media evaporation. The MIC was defined as the lowest concentration of antibiotic at which bacterial growth was inhibited.

Immunoblot and antibodies

All protein samples were resolved in 12 or 15% SDS-PAGE, or Novex Tris-Glycine Mini gradient (4–20%) gels. After SDS-PAGE, the separated proteins were transferred onto 0.45 μM PVDF membranes (Sigma) using a wet transfer system (Biorad) and Towbin buffer (25 mM Tris, 200 mM glycine, 20% methanol (vol/vol) pH 8.3), applying 100 V for 1 h. The tank was kept in an ice bath during the transfer process to prevent overheating. After transfer, the membrane was blocked for 1 h with 5% bovine serum albumin (BSA) (Goldbio) dissolved in tris-buffered saline containing 0.2% Tween (TBST). The blocked membrane was incubated with the primary antibody for 1 h with agitation. The following antibodies and dilutions were used in this study: (i) rabbit polyclonal α-GcvH-L (Pacific Immunology) (1:40,000), (ii) rabbit polyclonal α-LA (Sigma) (1:5,000), (iii) mouse monoclonal α-PBP2a antibody (1:10,000), (iv) human/rabbit chimera α-mono-ADP-ribose monoclonal antibody (Biorad) (1:7500), and (v) mouse monoclonal α-6X-His (Abcam) (1:10,000). After incubation, the membrane was washed three times in TBST for 10 min each, followed by addition of secondary antibody for 1 h with agitation. The following antibodies and dilutions were used: (i) goat α-rabbit IgG alkaline phosphatase (AP) or Horseradish peroxidase (HRP) (Thermo Scientific) (1:5,000), and (ii) goat α-mouse (AP) (Thermo Scientific) (1:10,000). The membrane was subsequently washed three times in TBST for 10 min. Immunoblots were developed using a SuperSignal West Femto chemiluminescence kit (Thermo Scientific) or using colorimetric detection by adding 66 μL of nitro-blue tetrazolium (NBT) (50 mg NBT in 1 mL 70% dimethylformamide (DMF)/30% H₂O) and 35 μL of 5-bromo-4-chloro-3′-indolyphosphate (BCIP) (50 mg BCIP in 1 mL DMF) to 10 mL AP Buffer (100 mM Tris, 100 mM NaCl, 5 mM MgCl_2, pH 9.5).

Determination of protein lipoylation in whole cell lysates

S. aureus WT and mutant strains in the LA salvage pathway were cultivated overnight in 15 mL conical tubes containing 3 mL of RPMI medium at 37°C with shaking. Overnight cultures were diluted 1:100 in 6 mL fresh RPMI in 15 mL conical tubes and incubated for 9 h with shaking. Bacteria were pelleted at 7,000 g for 10 min, the supernatants were removed, and cell pellets were resuspended in 0.4 mL 1X PBS. Bacterial suspensions were transferred to screw cap microcentrifuge tubes (Fisher Scientific), which were preloaded with 250 μL of 0.1 mm glass beads (Electron Microscopy Sciences). Tubes were then placed in a Fast Prep-24 5G (MP Biomedicals) bead disruption system. Two sequential cell lysis steps were performed, one at 5.0 speed for 20 s and the second at 4.5 speed for 20 s. Samples were kept on ice for 5 min between the two lysis steps. After lysis, samples were centrifuged at 20,000 g for 20 min, and 150 μL of the supernatant was collected in 1.5 mL centrifuge tubes containing 50 μL of 4X SDS sample buffer (0.2 M Tri-HCl [pH 6,8]; 8% SDS; 5.5 M glycerol; 0.02 M EDTA; 0.6 M β-mercaptoethanol; and 6 mM Bromophenol blue). Samples were boiled for 10 min and stored at −20°C before use. The protein extracts were resolved in a Novex Tris-Glycine Mini gradient gel, 4–20% (Invitrogen), followed by transfer and detection of lipoylation via immunoblot, as described above.

Pulldown of 6x-Histidine tagged GcvH-L from S. aureus strains under oxidative stress

S. aureus Δsbi Δspa ΔgcvH-L Δmacro ΔsirTM strain bearing pJC1111-6xhis-gcvH-L, pJC1111-6xhis-gcvH-L (K56A), or pJC1111-6xhis-gcvH-L (D27A) was grown overnight in 3 mL of RPMI in 15 mL conical tubes. The following day, 0.6 mL of bacterial culture was subcultured into 500 mL flasks containing 60 mL of fresh media and incubated for 4 h at 37°C with shaking. Before adding NaOCl (Sigma), 30 mL of bacterial culture was withdrawn, pelleted at 7,000 g for 10 min, and stored at −80°C (untreated). The remaining 30 mL of culture was treated with 2 mM NaOCl for 2 h at 37°C with shaking. The bacterial culture was then pelleted and washed thrice with 30 mL of fresh media by centrifuging tubes at 7,000 g for 10 min. After the final wash, cells were pelleted and stored at −80°C.

Each cell pellet from the two conditions was resuspended in 1 mL of 1X PBS containing 10 μg mL⁻¹ lysostaphin, and the bacterial suspension was transferred to a screw cap microcentrifuge tube (Fisher Scientific), which was preloaded with 250 μL of 0.1 mm glass beads (Electron Microscopy Sciences). Tubes were incubated at 37°C for 30 min, followed by mechanical disruption using a Fast Prep-24 5G (MP Biomedicals) bead disruption system. The mechanical disruption was done in two sequential steps: (i) at 5.0 speed for 20 s and (ii) at 4.5 speed for 20 s, with a 5 min incubation on ice between each step. Lysates were centrifuged at 20,000 g for 30 min, and the supernatant containing soluble material was collected. A 75 μL aliquot of the lysate was taken, mixed with 25 μL of 4X SDS sample buffer, and boiled for 10 min at 95°C as the input control. The remaining whole cell extract was added to a 1.5 mL microcentrifuge tube containing 50 μL nickel-NTA resin (Qiagen) that was equilibrated with 1 X PBS and incubated at 4°C for 16 h on a rotisserie. After incubation, tubes were centrifuged at 500 g for 5 min, the supernatant was discarded, and the nickel-NTA resin was subsequently washed three times by centrifugation at 500g for 5 min and resuspended in 1 mL 1X PBS. After the final wash, the resin was resuspended in 100 μL 4X sample buffer, followed by boiling for 10 min at 95°C. Samples were resolved in 15% SDS-PAGE gels.

Evaluation of endogenous GcvH-L levels from S. aureus cultures

Δsbi Δspa, Δsbi Δspa ΔsirTM, and Δsbi Δspa ΔgcvH-L strains were cultured in 10 mL of LB broth overnight. The following day, 2.5 mL of each bacterial culture was inoculated into 250 mL of fresh LB broth supplemented 500 nM LA, and cultures were incubated for 4 h prior to treatment with NaOCl. A volume of 125 mL from each culture was withdrawn and pelleted by centrifugation at 7,000 g for 10 min (untreated samples), while the remaining 125 mL was treated with 2 mM NaOCl for 2 h. After this treatment, bacterial cultures were washed three times with 1X PBS, pelleted by centrifugation at 7,000 g for 10 min (treated samples), and bacterial pellets were stored at −80°C overnight. The next day, bacterial pellets were resuspended in 30 mL of 1X PBS, and 10 μg/mL lysostaphin was added to each tube. The suspensions were incubated for 30 min at 37°C, followed by sonication on an ice bath for 3 min, with intervals of 20 s ON and 30 s OFF at an amplitude of 340 W. Whole cell lysates were centrifuged at 20,000 g for 30 min, and the supernatant was collected and passed through a 0.22 μm syringe filter. The clarified lysates were then loaded into an Amicon ultrafiltration device (Sigma) with a 50 kDa cutoff, and the flow-through was collected for protein precipitation using trichloroacetic acid (TCA) at a final concentration of 10%. The samples were incubated overnight at 4°C and then centrifuged at 20,000 g for 40 min at 4°C. Following centrifugation, the supernatants were discarded, and the pellets were washed twice with cold acetone. After the final wash, the acetone was carefully removed, and the pellets were air-dried by leaving the tubes uncapped and incubating them at 37°C for 15 min. The TCA-precipitated pellets were resuspended in 300 μL of an aqueous solution of 8M urea, mixed with 4X sample buffer, and separated using 15% SDS-PAGE gels. GcvH-L levels were assessed via immunoblot analysis, and a Coomassie-stained gel of the samples was used to ensure equal protein loading.

Immunoprecipitation using anti-GcvH-L antibody

A Δsbi Δspa mutant of S. aureus was used in these experiments to prevent any non-specific binding of antibodies.¹⁰¹ The growth conditions and sample collection of Δsbi Δspa and Δsbi Δspa ΔgcvH-L strains were the same as those outlined in the previous section. Endogenous GcvH-L was immunoprecipitated using Protein A agarose beads (Thermo Scientific). Briefly, 50 μL of Protein A agarose beads were added into a 1.5 mL microcentrifuge tube and washed three times by centrifuging at 500 g for 10 min and resuspending the resin in 1 mL of 1X PBS. After the last wash, the resin was resuspended in 500 μL of 1X PBS, and 10 μL of α-GcvH-L antibody was added into the tube and incubated on a rotisserie at room temperature for 2 h. The resin was washed three times with 1 mL 1X PBS, as previously described, and the supernatant was removed after the final wash. 900 μL of the whole cell lysate from S. aureus was added, the mixture was incubated for 3 h at room temperature, and the resin was washed thrice with 1 mL 1X PBS. After the final wash, the resin was resuspended in 100 μL of 4X sample buffer without β-mercaptoethanol and boiled for 10 min. Samples were resolved on a 15% SDS-PAGE gel, followed by immunoblot with α-GcvH-L antibody. The membrane was then incubated for 1 h with a clean-Blot IP detection reagent (Thermo Scientific) at a 1:300 dilution in TBST, and immunoblots were developed using a SuperSignal West Femto chemiluminescence kit (Invitrogen ).

Purification of 6x-Histidine-tagged proteins

All recombinant proteins were expressed in E. coli and purified using nickel affinity chromatography as described by Laczkovich et al. with some modifications.³⁰ Plasmids encoding recombinant 6xHis-SirTM, 6xHis-LplA1, 6xHis-LplA2, 6xHis-LipL, and 6xHis-GcvH-L (D27A) were transformed into E. coli lysY/I^q, while 6xHis-GcvH-L expressing plasmid was transformed into ΔlipA:kan E. coli lysY/I^q to ensure the protein would be purified in its apo form.³⁰ Freshly transformed colonies were inoculated into 10 mL of LB broth supplemented with 100 μg/mL ampicillin and incubated overnight at 37°C with shaking. Bacterial cultures were diluted 1:100 into fresh LB broth +100 μg/mL ampicillin and incubated with shaking until they reached an OD₆₀₀ of 0.3, which took approximately 3 h. Once bacterial cultures reached the desired OD₆₀₀, protein expression was induced with 0.1 mM IPTG at 16°C overnight with shaking. Following induction, bacteria were collected by centrifugation at 7,000 g for 20 min, and pellets were stored at −80°C before lysis.

For cell lysis, bacterial pellets were thawed at 37°C and resuspended in lysis buffer (25 mM imidazole, 50 mM Tris-HCl, 300 mM NaCl, 1 mM phenylmethylsulfonyl fluoride (PMSF), pH 8). The bacterial suspension was sonicated on an ice bath for 10 min with intervals of 10 s ON and 50 s OFF at an amplitude of 340 W. Cell debris was removed by centrifugation at 20,000 g for 30 min, and the supernatant was collected and passed through a 0.22 μm syringe filter. The clarified lysates were then mixed with 1 mL of nickel-NTA resin (Qiagen), preequilibrated with lysis buffer, and incubated on a rotisserie at 4°C for 1 h. After incubation, the mixture was poured into a glass chromatography column, and the resin was washed with 50 mL of lysis buffer, followed by elution of the bound protein with 3 mL elution buffer (500 mM imidazole, 50 mM Tris-HCl, 300 mM NaCl, pH 8).

All purified protein was dialyzed using dialysis cassettes (Thermo Scientific) with a molecular weight cutoff ranging from 10 kDa to 30 kDa. Dialysis was carried out at 4°C in buffer 1 (100 mM imidazole, 50 mM Tris-HCl, 300 mM NaCl, pH 8) for 3 h, followed by incubation in buffer 2 (25 mM imidazole, 50 mM Tris-HCl, 300 mM NaCl, pH 8) for 3 h, and buffer 3 (50 mM Tris-HCl, 300 mM NaCl, pH 8) overnight. After dialysis, protein concentration was measured using a bicinchoninic acid (BCA) kit (Thermo Scientific), and protein purity was evaluated by running the proteins on SDS-PAGE gels followed by Coomassie staining. If necessary, proteins were further purified using fast protein liquid chromatography (FPLC) on an ENrich SEC 650 10 × 300 mm column (Bio-Rad) size exclusion column. Purified proteins were stored at −80°C before use.

Purification of 6x-Histidine-apo-E2 subunits from S. aureus

Apo-6xHis-E2PDH, apo-6xHis-E2OGDH, and apo-6xHis-E2BCODH were expressed in a ΔlipA:kan E. coli lysY/I^q strain and purified as previously described.³²

LplA1 and LplA2 lipoylation assays

The in vitro lipoylation assays with LplA1 and LplA2 were performed as described by Laczkovich et al., with some modifications.³⁰ The reaction was carried out in a 1.5 mL microcentrifuge tube with a final volume of 50 μL. The lipoylation reaction contained the following components at the specified concentrations: (i) 6 mM ATP, (ii) 1 mM DTT, (iii) 1 mM ZnCl₂, (iv) 2 mM LA, (v) 1 μM purified LplA1 or LplA2, and (vi) 10 to 20 μM of apo-GcvH-L or apo-GcvH-L (D27A). All components were diluted in Tris-buffered saline (50 mM Tris-HCl, 300 mM NaCl, pH 8.0). The reaction mixture was subsequently incubated in a thermomixer for 2 h at 37°C with shaking at 300 rpm. Lipoylation was evaluated via immunoblot, and lipoyl GcvH-L was further purified on an ENrich SEC 650 10 × 300 mm column (Biorad) for downstream applications. For experiments that analyzed the effect of redox conditions on LplA1 and LplA2 activity, 10 mM H₂O₂ or 20 mM DTT (final concentration) was added to the reaction mixtures, and 5 μL of the sample was taken from each tube at 10, 20, 30, and 40 min. Reactions were stopped by adding 25 μL of 4X sample buffer supplemented with 5 mM DTT and boiling for 10 min. Protein samples were resolved in 15% SDS-PAGE gels, where one gel was stained with Coomassie and the other used for immunoblot analysis. Lipoylation of GcvH-L was determined by comparing the migration patterns in SDS-PAGE gels between the lipoylated and the apo proteins and the detection of LA via immunoblot. Lipoyl-GcvH-L was quantified by measuring the intensity and area of protein bands from Coomassie-stained gels in three independent experiments using ImageJ.

In vitro ADP-ribosylation of GcvH-L

ADP-ribosylation reactions were carried out in a 1.5 mL microcentrifuge tube with a final volume of 50 μL in Tris-buffered saline (50 mM Tris-HCl, 300 mM NaCl, pH 8.0). Each reaction contained the following components at the specified concentrations: (i) 1 mM zinc acetate, (ii) 0.4 mM NAD+, (iii) 1 mM DTT, (iv) 1 μM lipoyl GcvH-L, and (v) 1 μM SirTM. Reactions were incubated in a thermomixer at 30°C for 2 h with shaking at 300 rpm. For studies that assessed the role of redox conditions in the activity of SirTM, 1 mM of DTT was used as a baseline, and a single reaction was supplemented with one of the following oxidizing or reducing reagents: (i) 10 mM H₂O₂, (ii) 0.25 mM NaOCl, or (iii) 10 mM DTT. Reactions were stopped by adding 4X sample buffer and boiling at 95°C for 10 min. The samples were resolved in a 15% SDS-PAGE gel, and the ADP-ribosylation of GcvH-L was evaluated by comparing the migration patterns between the MAR-GcvH-L and lipoyl-GcvH-L and by immunoblot against mono-ADP-ribose.

Size exclusion analysis of SirTM, GcvH-L, and SirTM-GcvH-L

Size exclusion analysis of purified SirTM, GcvH-L, and the SirTM-GcvH-L complex was performed using a Superdex 75 Increase 10/300 GL column from Cytiva. Tris-buffered saline (50 mM Tris-HCl, 300 mM NaCl, pH 8.0) was used as the buffer for all runs. For the analysis of the SirTM-GcvH-L complex, the buffer was supplemented with NAD+ (0.4mM) and Zinc Acetate (1 mM) before proceeding with size exclusion separation. The column was equilibrated with five column volumes of buffer, followed by sequential addition of 500 μL of 1 μM GcvH-L, SirTM, or SirTM + GcvH-L to the column. Protein elution was monitored by measuring the absorbance at 280 nm, and 18 fractions of 1 mL each were collected. The presence of protein(s) in the detected peaks was validated by analyzing all elution fractions by immunoblot against α-GcvH-L and α-6xHis for SirTM. Additionally, the elution fractions of the individual proteins were assessed by SDS-PAGE to confirm protein purity. Gel filtration standards were loaded on the column to generate a standard curve of molecular weights using ChromLab software to calculate.

Determination of the redox state of the LA on GcvH-L

The redox state of the lipoyl moiety on MAR-GcvH-L and GcvH-L D27A was determined according to the protocol established by Denoncin et al..⁵⁷ Purified GcvH-L and GcvH-L D27A were used in ADP-ribosylation reactions in a 1.5 mL microcentrifuge tube with a final reaction volume of 100 μL. The reaction mixture contained 1 mM zinc acetate, 0.4 mM NAD+, 20 mM DTT, 1 μM lipoyl GcvH-L or GcvH-L-(D27A), and 5 μM SirTM, in Tris-buffered saline (50 mM Tris-HCl, 300 mM NaCl, pH 8.0). Reactions were incubated in a thermomixer at 30°C for 2 h with shaking at 300 rpm. The samples were then dialyzed in a MINI unit with a molecular weight cutoff of 2 kDa (Thermo Scientific). The dialysis unit was loaded with 100 μL of the protein mixture, placed in a 15 mL conical tube containing 15 mL of Tris-buffered saline, and incubated for 2 h at 4°C followed by transfer into fresh buffer three times to ensure complete removal of the DTT. After dialysis, samples were divided into two groups; one group was treated with 10 mM H₂O₂ final concentration, while the other group remained untreated, followed by incubation for 2 h at 30°C with shaking at 300 rpm. After the treatment, the proteins were precipitated by adding 5.5 μL of 100% stock solution of TCA to achieve a final concentration of 10%, followed by incubation on ice for 30 min, and centrifugation at 20,000 g for 30 min at 4°C. Supernatant was removed, and the pellet was washed once with 200 μL of ice-cold acetone. Samples were centrifuged at 20,000 g for 10 min at 4°C, and the supernatant was removed carefully without disrupting the pellet. Protein pellets were air-dried by incubating the tubes with an open cap at 37°C for 15 min. Free thiol labeling was carried out by resuspending the protein pellets in 30 μL of 20 mM 4-acetamido-4′-maleimidylstilbene-2,2′-disulfonic acid, disodium salt (AMS) (Thermo Scientific) solution, which was previously dissolved in a buffer containing 50 mM Tris–HCl, pH 7.5, 0.1% SDS. Protein pellets for unlabelled controls were diluted in buffer without AMS (50 mM Tris–HCl, pH 7.5, 0.1% SDS). After resuspension, microcentrifuge tubes were vortexed vigorously for 1 min, followed by incubation in a thermomixer at 37°C for 1 h with shaking at 1400 rpm. Labeled proteins were mixed with 270 μL of 4X sample buffer without β-mercaptoethanol, and the tubes were boiled for 15 min. Protein samples were resolved in a 15% SDS-PAGE gel, followed by immunoblot analysis.

Assessing the redox state of LA on GcvH-L using denatured protein

To investigate the role of the GcvH-L structure in the protective effects of ADP-ribosylation on lipoic acid oxidation, we fully linearized GcvH-L using urea and heat before or after treating it with H₂O₂. ADP-ribosylation of GcvH-L and dialysis were performed as described in the previous section. For the samples denatured before H₂O₂ treatment, the MINI unit was transferred to a Tris-buffered saline solution containing 8M urea (Sigma) after the last step of dialysis. Samples were incubated for 2 h at 4°C, followed by heating at 90°C for 30 min. For samples denatured after H₂O₂ treatment, H₂O₂-treated protein was loaded into a new MINI unit and placed in a 15 mL conical tube containing 15 mL of Tris-buffered saline supplemented with 8M urea. Samples were incubated for 2 h at 4°C, followed by transferring into fresh buffer containing 8M urea twice to ensure the complete removal of the H₂O₂. After dialysis, the protein was heated at 90°C for 30 min. Denatured GcvH-L was precipitated with 10% TCA and labeled with AMS, as described in the previous section.

Lipoyl transfer assays

The transfer of the lipoyl group from GcvH-L/ADPr-GcvH-L to the E2 subunits via LipL was carried out in a two-step reaction. The first reaction involved the ADP-ribosylation of lipoyl-GcvH-L, followed by the assessment of lipoyl transfer via LipL to E2 subunit acceptors. ADP-ribosylation was performed in a 1.5 mL microcentrifuge tube with a 50 μL reaction volume. The following components were added to the reaction mixture to ensure 100% ADP-ribosylation of GcvH-L: (i) 1 mM zinc acetate, (ii) 0.4 mM NAD+, (iii) 20 mM DTT, (iv) 1 μM lipoyl GcvH-L, and (v) 5 μM SirTM in Tris-buffered saline (50 mM Tris-HCl, 300 mM NaCl, pH 8.0). Non-MAR GcvH-L controls contained all components except for NAD+ and zinc acetate. Reactions were then incubated in a thermomixer at 30°C for 2 h with shaking at 300 rpm, followed by dialysis in a MINI unit with a molecular cutoff of 2 kDa (Thermo Scientific) to ensure the complete removal of all cofactors. For the second reaction, 30 μL of the dialyzed mixture was mixed with 1 μM apo-E2 subunits, 0.1 μM LipL, 1 mM MgCl₂, and Tris-buffered saline in a final reaction volume of 50 μL in a 1.5 mL microcentrifuge tube. The reaction mixtures were incubated in a thermomixer for 30 min at 37°C with shaking at 300 rpm. Samples were withdrawn every 10 min, and the reaction was stopped by adding 4X sample buffer and boiling for 10 min. The samples were resolved on 12% and 15% SDS-PAGE gels, followed by Coomassie staining and immunoblotting to determine protein lipoylation and ADP-ribosylation.

Murine infection models

S. aureus strains were struck from frozen stocks on TSA and grown overnight at 37°C. The next day, a single colony was inoculated into 10 mL of RPMI in a 50 mL conical tube and grown for 16 h at 37°C with shaking. The following morning, bacteria were harvested by centrifugation at 7,000 g for 10 min and then washed thrice with 1X PBS. Bacterial suspensions were subsequently normalized in 1X PBS to a final OD₆₀₀ of 0.3 (~1 × 10⁸ CFU/mL) for retroorbital inoculation and OD₆₀₀ of 1.0 (~1 × 10⁹ CFU/mL) for intraperitoneal inoculation. Before infection, OD-normalized bacterial suspensions were subjected to serial dilution in 1X PBS and plated on LB agar to evaluate the accuracy of the inoculum. For retroorbital inoculations, five-to-six-week-old female C57BL/6 mice (JAX) were anesthetized intraperitoneally with ketamine/xylazine (100/10 mg/kg) followed by inoculation with 100 μL of bacterial suspension (~1.0 × 10⁷ CFU) directly into the bloodstream via retro-orbital sinus injection. For intraperitoneal inoculations, seven-week-old female Swiss Webster mice (Envigo) were infected with 100 μL normalized bacterial suspension (~1.0 × 10⁸ CFU) directly into the peritoneal cavity. C57BL/6 male mice (JAX) were injected with ~1.0 × 10⁸ CFU or ~1.0 × 10⁷ CFU, while B6.129S-Cybb^™1Din/J (Cybb⁻) (JAX) mice were injected with ~1.0 × 10⁷ CFU. All mice were euthanized by CO₂ narcosis at 24 h postinfection. In the case of retroorbital infection, kidneys were collected and homogenized in 5 mL of 1X PBS. Organ homogenates were serially diluted in 1 X PBS, and dilutions were plated onto LB agar for CFU enumeration. For the peritoneal infection, a lavage of the peritoneal cavity was performed with 7 mL of 1X PBS using a sterile syringe and an 18-gauge needle. The kidneys were also isolated and homogenized in 5 mL 1X PBS. Lavage fluid and kidney homogenates were serially diluted in 1X PBS and plated on LB agar, followed by incubation at 37°C overnight to enumerate CFU.

ex vivo macrophage survival assays

The outgrowth of S. aureus after phagocytosis by F4/80+ cells was performed as described by Grayczyk et al. with minor modifications.¹⁰ Briefly, seven-week-old female Swiss Webster (Envigo) or seven-week-old male C57BL/6 mice (JAX) were intraperitoneally infected with approximately 1.0 × 10⁸ CFU of the S. aureus WT strain. Four-to-seven-week-old male B6.129S-Cybb^™1Din/J (Cybb⁻) (JAX) mice were infected with 1.0 × 10⁷ CFU. Animals were euthanized after 72 h for WT mice and 24 h for knockout mice (Cybb⁻). Peritoneal cells were isolated by lavage of the peritoneal cavity with 7 mL of 1X PBS containing 100 μg/mL Pen/Strep and 50 μg/mL gentamicin. Cell suspensions were incubated on ice for 30 min, followed by washing three times with antibiotic-free RPMI medium (Corning) containing 10% heat-inactivated Fetal Bovine Serum (FBS). The washed cell pellet was resuspended in fluorescence-activated cell sorting (FACS) buffer (1X PBS containing 2% heat-inactivated FBS) for magnetic bead sorting. F4/80+ cells were isolated from total peritoneal cells using a BD IMag cell separation system (BD Biosciences).^10,21 The total peritoneal cells were first labeled with anti-CD16/CD32 (1 μg of antibody per million cells) (BioLegend) for 30 min on ice, followed by the addition of biotinylated anti-F4/80 antibodies (0.25 μg per million cells) (BM8) (BioLegend) and incubation for 30 min on ice. After labeling, 50 μL per ten million cells of streptavidin-conjugated magnetic beads (BD Biosciences) were added, and the samples were incubated for 30 min at 4°C. The unbound fraction was carefully removed, and the magnetic beads containing cells were washed three times with FACS buffer. Sorted cells were resuspended in RPMI medium and stored overnight at 4°C before use.

The next day, overnight cultures of the bacterial strains grown in RPMI were washed three times with 1X PBS, normalized to 1.0 × 10⁸ CFU/mL, and opsonized with 10% whole mouse serum for 30 min at 37°C. The opsonized bacteria were washed thrice with PBS and resuspended in RPMI medium. 1.0 × 10⁶ F4/80+ cells were mixed with opsonized bacteria in a 1.5 mL microcentrifuge tube at a multiplicity of infection (MOI) of 0.1, and the cell mixture was incubated for 30 min at 37°C on a rotisserie. After incubation, F4/80+ cells were washed 3 times with RPMI supplemented with 10% heat-inactivated FBS, followed by resuspension in 1 mL of RPMI and incubation at 37°C on a rotisserie. For the experiments in which the NADPH oxidase inhibitor, gp91ds-tat (Anaspec), was used, F4/80+ cells were treated with 50 μM inhibitor in serum-free RPMI for 1 h before adding the bacteria. CFUs were enumerated at the indicated time points by removing a 100 μL aliquot, lysing F4/80+ cells with 1% saponin (Sigma) for 15 min on ice, serially diluting, and plating dilutions on LB agar.

QUANTIFICATION AND STATISTICAL ANALYSIS

All statistical details are provided in the figure legends, including the statistical tests and the exact sample size (n) used in each experiment. Data are presented in three different formats: (i) box-and-whisker plots with individual values, median, quartiles, and range; (ii) individual values and median indicated; and (iii) scatterplots with bars indicating the mean ± SEM. All statistical analyses were carried out using GraphPad Prism V10 software. Before performing statistical analyses, a D’Agostino and Pearson test or a Shapiro-Wilk test was used to determine whether the dataset followed a normal (parametric) or skewed (nonparametric distribution). For parametric tests, either a one-way or two-way ANOVA was conducted, followed by multiple comparisons using Dunnett’s or Tukey’s test. The statistical significance between two groups was assessed using an unpaired t test. For nonparametric tests, the Kruskal-Wallis test with Dunn’s post hoc analysis was used for three or more groups, while the Mann-Whitney test was employed to assess significance between two groups. Significant differences were considered when the p-value was smaller than 0.05. Statistical significance: *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001.

Supplementary Material

SUPPLEMENTAL INFORMATION

Supplemental information can be found online at https://doi.org/10.1016/j.celrep.2025.116095.

Highlights.

• Staphylococcus aureus harnesses lipoic acid salvage to resist oxidative stress in vivo

• Resistance to oxidative stress requires lipoylation and ADP-ribosylation on GcvH-L

• ADP-ribosylation protects lipoic acid from detrimental oxidation and regulates its transfer