A redundant isoprenoid biosynthetic pathway supports Staphylococcus aureus metabolic versatility

ABSTRACT

Isoprenoids are ubiquitous molecules that serve as fundamental building blocks for life. In bacteria, isoprenoids are precursors for carotenoid pigments, respiratory cofactors, and essential sugar carrier lipids, such as lipid II. Isoprenoid synthesis initiates via condensation of the five-carbon (C₅) precursors, isopentenyl diphosphate (IPP) and dimethylallyl diphosphate (DMAPP). This initial reaction condenses one DMAPP and two IPPs, resulting in C₁₅ farnesyl diphosphate (FPP), an intermediate that is sequentially elongated with IPP. FPP is thought to be synthesized exclusively by the prenyl diphosphate synthase (PDS), IspA. In Staphylococcus aureus, ispA mutants lack the golden carotenoid pigment, staphyloxanthin. The fact that ispA can be inactivated in S. aureus and other bacteria is surprising given the reliance of lipid II on FPP and supports the hypothesis that an additional enzyme produces the critical isoprenoid precursor. We isolated pigmented ispA suppressor mutants harboring single-nucleotide polymorphisms within a second PDS-encoding gene, hepT, suggesting that HepT and IspA have overlapping roles in S. aureus isoprenoid synthesis. Subsequent work determined that IspA and HepT support metabolic versatility, as a hepT ispA double mutant fails to aerobically respire partially due to a lack of prenylated heme cofactors. The finding that a hepT ispA double mutant is viable supports a model whereby a third PDS compensates in the absence of ispA and hepT to produce lipid II precursors. Lastly, we show that ispA and hepT mutants exhibit colonization defects in a murine model of systemic infection, demonstrating that isoprenoid biosynthesis is a potential drug target for combating S. aureus.

IMPORTANCE

Isoprenoid synthesis is an essential process that is presumed to be initiated by the prenyl diphosphate synthase (PDS), IspA. However, our understanding of this pathway is incomplete considering that ispA mutants have been described in several bacterial species, leaving the mechanism for isoprenoid synthesis initiation uncertain in these genetic backgrounds. Using the opportunistic pathogen Staphylococcus aureus, we demonstrate that a second PDS, HepT, supports the production of isoprenoid-dependent molecules in the absence of IspA. Importantly, we show that mutants deficient for either IspA or HepT display colonization defects in a murine model of systemic infection. Furthermore, the simultaneous mutation of hepT and ispA is tolerated in S. aureus and suggests the presence of a third PDS capable of initiating isoprenoid synthesis. This study establishes PDSs as viable targets for the treatment of S. aureus infections and provides novel insights into the redundant nature of isoprenoid synthesis in this pathogen.

INTRODUCTION

Isoprenoid synthesis is a highly conserved process present in virtually all living organisms and begins with the production of the universal five-carbon (C₅) precursors, isopentenyl diphosphate (IPP) and dimethylallyl diphosphate (DMAPP) (1, 2). Depending on the organism, IPP and DMAPP can be generated by either the methyl erythritol phosphate pathway, which is active in most bacteria, or the mevalonate pathway utilized by animals, archaea, and some gram-positive bacteria (3, 4). An initial condensation reaction of one DMAPP and one IPP produces geranyl diphosphate (GPP), to which a second IPP is immediately added to generate farnesyl diphosphate (FPP), a C₁₅ isoprenoid that serves as a substrate for downstream cellular pathways (2). This initial condensation reaction and subsequent elongation reactions that add IPP units in a stepwise fashion are carried out by enzymes, called prenyl diphosphate synthases (PDSs) (2). Generally, bacteria harbor three PDSs: a short-chain PDS that catalyzes the initial condensation reaction between IPP and DMAPP to generate C₁₀ GPP and then FPP; a medium-chain PDS that uses FPP and several additional IPP to produce medium-chain length isoprenoids; and a long-chain PDS that also uses FPP and IPP to generate long-chain isoprenoids (5–11). Medium-chain isoprenoids range in length from C₃₅ to C₅₀ and support the production of quinone respiratory cofactors (12, 13). Long-chain isoprenoids are C₅₅ or greater and in bacteria include undecaprenyl phosphate (Und-P), which serves as a scaffold onto which glycan units are added to generate lipid II, an indispensable metabolite required for generating peptidoglycan (13, 14). Lipid II is essential for cell viability and the target of several classes of antibiotics (14–17). Therefore, a greater understanding of the isoprenoid synthesis in bacteria will reveal fundamental mechanisms by which critical isoprenoid precursors are allocated to support cell envelope maintenance and metabolism.

In the gram-positive pathogen Staphylococcus aureus, FPP is a substrate for three reactions: condensation, elongation, or prenylation, the covalent addition of an isoprenoid directly to a substrate (Fig. S1). The condensation of two FPPs generated by the short-chain PDS IspA is catalyzed by CrtM to generate dehydrosqualene, a C₃₀ precursor for staphyloxanthin, the membrane-localized carotenoid antioxidant that gives the pathogen its distinctive golden color (18). As such, ispA and crtM mutants are not pigmented (19, 20). Elongation reactions add IPP to FPP in a stepwise fashion, and at least two elongation reactions are active in S. aureus. The first is catalyzed by UppS, the enzyme that produces long-chain undecaprenyl diphosphate (Und-PP). Und-PP is the precursor of the sugar carrier lipid, Und-P, that transports polysaccharides across the cytoplasmic membrane. Extracytoplasmic saccharide-based polymers generated by S. aureus include peptidoglycan, wall teichoic acid (WTA), and in some strains, capsular polysaccharide (CP) (14, 21, 22). Importantly, each of these polysaccharide cell surface structures is either essential (peptidoglycan) or contributes to virulence (WTA and CP), making UppS an essential enzyme and a high-priority target for therapeutic intervention (14, 23–26). Given the essentiality of UppS, it is surprising that the enzyme which produces the FPP substrate for its activity, IspA, is dispensable. Mutants lacking functional ispA have been generated in several different bacterial species, including S. aureus, Escherichia coli, Pseudomonas aeruginosa, and Bacillus subtilis (27–31). In S. aureus, ispA mutants produce reduced quantities of FPP compared to wild type (WT), indicating that at least one other PDS can produce FPP (27). Clarifying the PDS enzymes that generate UppS substrates will provide new knowledge for how the critical peptidoglycan lipid carrier is synthesized.

The second elongation is predicted to be carried out by HepT. B. subtilis HepT and the E. coli HepT homolog, IspB, produce the medium-chain isoprenoids used in menaquinone (MK) and ubiquinone (UQ) syntheses (8, 29, 32–35). MK and UQ are essential electron carriers in the electron transport chain. S. aureus produces MK, of which three species are synthesized that differ in the length of the isoprenoid group: MK-7, MK-8, and MK-9 (36, 37). The number refers to the C₅ isoprenoid moieties; thus, MK-7, MK-8, and MK-9 have chain lengths of C₃₅, C₄₀, and C₄₅, respectively. Of note, some strains of S. aureus harbor HepT alleles that result in the production of a C₅₀ MK-10 (36), indicating that different hepT alleles influence the length of its isoprenoid product. These findings support a model whereby S. aureus HepT is required to generate MK; however, direct experimental evidence for this function is lacking.

MK is a required cofactor for aerobic respiration, and S. aureus mutant strains impaired for MK synthesis are restricted to fermentative metabolism. This leads to a distinctive phenotype, called the small colony variant (SCV), which coincides with increased antibiotic resistance (37–40). In fact, MK auxotrophs are among the most common types of SCVs isolated from cystic fibrosis patients (41). Aminoglycoside treatment selects for SCVs because the antibiotic requires the proton motive force (PMF) to enter the cell (42–44). In response, S. aureus develops inactivating mutations in the MK synthesis pathway, switching to fermentation and reducing the PMF (45). Notably, an S. aureus hepT mutant exhibited increased pigmentation, but growth phenotypes were not reported (20); therefore, it is unclear whether the mutation of hepT induces the SCV phenotype. Additionally, whether HepT participates in other isoprenoid-dependent pathways is uncertain, as HepT in B. subtilis and IspB in E. coli, Acinetobacter baumannii, and Corynebacterium glutamicum are essential, implying medium-chain PDSs may function beyond MK and UQ syntheses (35, 46–48). In fact, loss of heptaprenyl diphosphate synthase activity in B. subtilis was only tolerated when the mutant was cultured in a medium that supports cell wall-free L-form growth, revealing a potential role for HepT in cell wall synthesis (49). The fact that HepT mutants are viable in S. aureus suggests that unique or redundant isoprenoid biosynthetic pathways exist in this species and presents an opportunity to study the physiological consequences of HepT inactivation, which has not been possible in other bacterial species.

Prenylation, the third FPP-dependent reaction, also supports aerobic respiration through the covalent attachment of FPP to heme b to produce the modified heme cofactors, hemes o and a. Both prenylated heme cofactors are capable of supporting the activity of the major terminal oxidase, QoxABCD (50–52). Therefore, prenylation is essential for one of the two enzymes that perform the last step of aerobic respiration, the reduction of O₂ to H₂O (51). A second terminal oxidase, CydAB, functions in addition to QoxABCD, leading to a branched respiratory chain. The CydAB activity is not dependent on heme o or a (51, 53). Whether IspA produces FPP needed to generate heme o or a is unclear. The ispA mutant exhibits slightly increased resistance to aminoglycosides gentamicin and kanamycin, as well as a reduced ATP production, suggesting that the electron transport chain activity is attenuated (27). However, whether IspA provides the isoprenoid precursors required to generate the modified heme cofactors needed for the QoxABCD activity has not been directly demonstrated. Overall, resolving the contributions of IspA and HepT toward generating isoprenoid metabolites in S. aureus will provide novel insight into three major pathways: carotenoid pigment production, cell envelope maintenance, and respiration.

In this study, we show that isoprenoid synthesis supports metabolic versatility in S. aureus by demonstrating that IspA and HepT preferentially contribute to carotenoid pigment and MK synthesis, respectively. Either enzyme is sufficient for the generation of hemes o and a, indicating that both can provide the FPP precursor needed for heme prenylation. The basis for this conclusion is the finding that a hepT ispA mutant is viable but restricted to fermentative metabolism despite supplementation with the C₂₀ MK analog, MK-4. These results support a model whereby a third enzyme generates FPP that is specifically used to produce Und-PP for peptidoglycan synthesis. Additional experimentation focused on chemically complementing MK deficiency using MK-4 and demonstrated that the shorter MK analog promotes QoxABCD but not CydAB activity. This result suggests that the MK isoprenoid chain length plays an important role in the CydAB function, indicating that different MK species modulate the activity of the branched respiratory chain. Lastly, we show that ispA or hepT mutants exhibit colonization defects in a murine model of systemic infection, providing evidence that the isoprenoid biosynthetic pathway is a potential drug target. Overall, these results support a revised model of S. aureus isoprenoid synthesis whereby precursors generated by specific enzymes are used to produce staphyloxanthin, MK, or Und-P, but prenylation of heme can be achieved using precursor from either IspA or HepT. These data show that redundancy within the isoprenoid synthesis pathway of S. aureus supports the metabolic versatility and virulence of this important human pathogen.

RESULTS

Pigmented ispA suppressor mutants contain nonsynonymous gain-of-function mutations in hepT

A previous study reported that ispA mutants lack pigmentation and exhibit increased resistance to gentamicin (27). We sought to recapitulate these results using an agar-based assay whereby 10⁸ CFU/mL ispA::Tn mutant cells were spread onto a tryptic soy agar (TSA) plate supplemented with 3 µg/mL gentamicin, yielding at least 100 isolated resistant colonies per plate. Across three independent trials, the total quantity of gentamicin-resistant colonies ranged between 10³ and 10⁴ CFU/mL. Unexpectedly, pigmented resistant colonies could be observed at a rate of 0.04% among the nonpigmented resistant colonies across three independent biological replicates. Pigmented colonies demonstrated WT-like levels of staphyloxanthin production, indicating the presence of suppressor mutations (Fig. 1A). Whole-genome sequencing analysis revealed nonsynonymous mutations in the gene encoding heptaprenyl diphosphate synthase, hepT (Table S1). Notably, three resistant pigmentation suppressor mutants (ispA::Tn¹¹⁵, ispA::Tn¹⁴⁴, and ispA::Tn¹⁶⁴) isolated from three different trials shared the same A72E amino acid change five residues preceding the first aspartate-rich motif (FARM) (Fig. 1B). This amino acid position is highly conserved in IspA and HepT homologs (54) and has been implicated in determining the chain length of the isoprenoid product (55). The nonpigmented gentamicin-resistant control isolate harbored an amino acid change 93 amino acids downstream of the HepT FARM domain (A165T). Given the putative activity of S. aureus HepT, it is likely that a hepT mutant does not produce MK and therefore cannot perform respiration, resulting in the observed gentamicin resistance. Consistent with this, hepT::Tn has a slower growth rate compared to WT (Fig. 1C), as it likely relies on fermentation to generate energy. Growth rate analysis of ispA::Tn gentamicin-resistant colonies showed similar doubling times to hepT::Tn (Fig. 1C), suggesting that the hepT mutations likely lead to reduced production of the medium-chain isoprenoids necessary for MK synthesis. In fact, the altered FARM domain of HepT^A72E increases in similarity to the FARM domain of other PDSs that are known to synthesize short-chain isoprenoids (55) and might be performing an IspA-like role in producing FPP needed for pigmentation.

Fig 1.

Pigmented ispA::Tn suppressors harbor SNPs within hepT. (A) Indicated strains were streaked on TSA and incubated overnight. (B) hepT is predicted to be operonic with SAUSA300_1361 and ubiE but not ndk. Multiple alignment of the HepT amino acid sequence encoded by the indicated strains. Conserved aspartic acid residues that comprise the first aspartic acid-rich motif (FARM) domain are highlighted in yellow. Black-highlighted residues represent the amino acid position that was altered due to acquired mutations in the ispA::Tn gentamicin-resistant mutants. (C) Doubling times of the indicated strains determined after incubation in TSB. Error bars represent one standard deviation from the mean. Statistical significance was determined via one-way ANOVA. * and ** represent P-values of <0.05 and <0.01, respectively.

Disruption of isoprenoid synthesis leads to downstream perturbations in the abundance of isoprenoid-derived metabolites

To determine the roles of IspA and HepT in the generation of downstream isoprenoid-containing metabolites, MK species undecaprenyl phosphate (Und-P) and undecaprenol (Und-OH) were quantified via high-performance liquid chromatography (HPLC). The primary MK species produced by WT was MK-8, followed by MK-7 and MK-9 (Fig. 2A), which is consistent with previous studies (36, 37). However, low levels of MK-5 and MK-6 were also detected in WT, which, to the best of our knowledge, is the first report of S. aureus producing these MK species. The mutation of ispA resulted in increased production of longer-chain MKs with a significantly higher production of MK-8 and MK-9 but a decreased production of MK-7, while MK-5 and MK-6 were not altered compared to WT. Notably, hepT::Tn did not produce detectable levels of any MK species. Consistent with the hypothesis that the HepT^A72E variant generates shorter-chain isoprenoids, ispA::Tn¹¹⁵ produced significantly higher levels of MK-5 compared to WT, whereas MK-8 and MK-9 were not detected, and MK-7 was also significantly reduced (Fig. 2A). Additionally, hepT::Tn contained significantly lower amounts of the lipid II precursor Und-P compared to WT (Fig. 2B). All of the mutants also demonstrated significantly lower quantities of Und-OH compared to WT, indicating both IspA and HepT may contribute to the production of precursors necessary for long-chain isoprenoid biosynthesis.

Fig 2.

Mutating the PDS enzymes IspA and HepT alters the abundance of medium- and long-chain isoprenoid-containing metabolites. (A) High-performance liquid chromatography (HPLC) analysis of MKs extracted from whole cells. Error bars represent one standard deviation from the mean. Statistical significance was determined by two-way ANOVA with Tukey correction. * and **** represent P-values of <0.05 and <0.0001, respectively. n.d. (not detected) represents samples for which no MK was detected. (B) HPLC analysis of undecaprenyl phosphate (Und-P) and undecaprenol (Und-OH) extracted from whole cells. Statistical analysis was determined by two-way ANOVA with Tukey correction. * and **** represent P-values of <0.05 and <0.0001 respectively. Data presented in panels A and B are the average of three independent biological replicates performed in triplicate.

Simultaneous inactivation of hepT and ispA induces a small colony variant phenotype that is unresponsive to MK-4 supplementation

A consequence of inactivating ispA in S. aureus is reduced quantities of FPP, supporting a model whereby a second enzyme generates this indispensable precursor (27). Given that IspA and HepT are the only two enzymes known to contain polyprenylsynthetase FARM domains in S. aureus, we hypothesized that the simultaneous inactivation of both enzymes would result in synthetic lethality. Given that hepT is predicted to be the terminal gene in a three-gene operon and is in close proximity to ndk (Fig. 1B), an in-frame deletion was generated to reduce the possibility of polar effects. The ΔhepT mutant was subsequently transduced with φ85 propagated on the ispA::Tn mutant to yield the ΔhepT ispA::Tn double mutant. Surprisingly, ΔhepT ispA::Tn was viable but displayed a significantly increased doubling time compared to the WT and ΔhepT (Fig. 3A). The supplementation of the growth medium with MK-4 decreased the doubling time of ΔhepT and a menE::Tn control SCV to WT-like levels, indicating that respiration was restored in these mutants. This result shows that the slower growth rate observed in these mutants is due to a lack of MK that restricts metabolism to fermentation. However, the MK-4 supplementation failed to reduce the doubling time of ΔhepT ispA::Tn, indicating that this strain is unable to aerobically respire. Genetic complementation of ΔhepT ispA::Tn with plasmid-encoded hepT or ispA restored pigmentation, doubling time, and stationary-phase optical density at 600 nm (O.D.₆₀₀) to that of the respective single mutant (Fig. S2). These results confirm that the SCV phenotype observed in ΔhepT ispA::Tn is not due to second site mutations. Having the ΔhepT ispA::Tn mutant in hand allowed us to investigate the effects of the hepT^A72E allele independent of either ispA or hepT. Consistent with the isolation of the hepT^A72E allele in the pigmentation suppressor mutants, the allele partially restored pigmentation of ΔhepT ispA::Tn (Fig. S3A). However, the doubling time and the stationary-phase O.D.₆₀₀ remained similar to the empty vector control (Fig. S3B and C). These results demonstrate that expressing HepT^A72E does not complement the growth of the double mutant and provide an explanation for why this allele increases gentamicin resistance.

Fig 3.

Simultaneous inactivation of hepT and ispA impairs aerobic respiration. (A) Doubling times of the indicated strains in TSB supplemented with or without 12.5 µM MK-4. Error bars represent one standard deviation from the mean. Statistical significance was determined via one-way ANOVA. **, ***, and **** represent P-values of <0.01, <0.001, and <0.0001, respectively. (B) The concentration of excreted l-lactate was determined from cell-free supernatants of the indicated strain. Statistical significance was determined by two-way ANOVA with Bonferroni correction. Error bars represent one standard deviation from the mean. *** and **** represent P-values of <0.001 and <0.0001, respectively. (C) Anaerobic stationary-phase optical density at 600 nm after 24 h incubation of the indicated strains cultured in TSB, TSB supplemented with MK-4, TSB supplemented with KNO₃, or TSB supplemented with both KNO₃, and MK-4. Error bars represent one standard deviation from the mean. Statistical significance was determined via two-way ANOVA comparing strains supplemented with MK-4 and KNO₃ to strains lacking supplementation (TSB) or supplemented with MK-4. **** represents a P-value of <0.0001.

Given the hyperpigmentation phenotype of ΔhepT and the importance of ispA in staphyloxanthin production, we tested whether loss of the antioxidant carotenoid pigment was responsible for the respiration defect. A ΔhepT crtM::Tn double mutant was constructed and exhibited a WT-like colony size compared to the SCV variant phenotype of ΔhepT ispA::Tn (Fig. S4A). Additionally, MK-4 supplementation restored the doubling time and the stationary-phase O.D.₆₀₀ of ΔhepT crtM::Tn (Fig. S4B and C), demonstrating that loss of pigmentation is not responsible for the ΔhepT ispA::Tn SCV phenotype.

S. aureus predominantly produces lactate as a major fermentation product (53, 56). To confirm that ΔhepT ispA::Tn is restricted to fermentation, we quantified the lactate production of strains grown in tryptic soy broth (TSB) or TSB supplemented with MK-4. WT and ispA::Tn produced low amounts of l-lactate, regardless of MK-4 supplementation (Fig. 3B), indicating that respiration is the primary energy-generating pathway in these strains. The menE::Tn control SCV produced high levels of l-lactate, which were significantly reduced upon MK-4 supplementation. This is similar to ΔhepT and demonstrates that MK-4 induces respiration in these mutants. However, ΔhepT ispA::Tn produced high levels of l-lactate with or without MK-4 supplementation (Fig. 3B). Together, these results demonstrate that ΔhepT ispA::Tn is not capable of aerobic respiration.

Next, we sought to determine whether the failure of ΔhepT ispA::Tn to respond to MK supplementation is due to a general respiration deficiency or if it is specific to aerobic respiration. To test this, ΔhepT ispA::Tn was cultured anaerobically in the presence or absence of MK-4 and the alternative terminal electron acceptor, nitrate (KNO₃), as S. aureus is capable of utilizing nitrate for anaerobic respiration (57, 58). WT, ispA::Tn, ΔhepT, ΔhepT ispA::Tn, and menE::Tn proliferated to similar end-point optical densities (Fig. 3C). The addition of the terminal electron acceptor, KNO₃, increased the stationary-phase, end-point optical density of WT and the ispA:Tn mutant, which produce MK and are, therefore, capable of anaerobically respiring. Interestingly, the addition of KNO₃ had an inhibitory effect on ΔhepT, ΔhepT ispA::Tn, and menE::Tn. Nitrate is often used as a preservative in food to prevent microbial growth (59), and we suspect nitrate may be inhibiting the proliferation of strains that cannot utilize this compound to anaerobically respire. Supplementation with both MK-4 and KNO₃ increased the stationary-phase, end-point optical density of all strains, including ΔhepT ispA::Tn (Fig. 3C). Together, these data show that ΔhepT ispA::Tn is capable of anaerobic respiration via supplementation with MK-4, revealing that the SCV phenotype is specific to aerobic respiration.

QoxABCD activity is impeded in the ΔhepT ispA::Tn double mutant due to loss of prenylated heme cofactors

S. aureus utilizes a branched aerobic respiratory chain consisting of the QoxABCD and CydAB terminal oxidases, both of which require heme for their activity (51, 53). The finding that ΔhepT ispA::Tn respires anaerobically but not aerobically supports the hypothesis that both QoxABCD and CydAB are impeded in this strain. The QoxABCD activity is dependent on prenylated heme cofactors, which are generated via the addition of FPP onto heme b by CtaB to produce heme o. Heme o is further modified by the addition of a carbonyl group by CtaA to produce heme a (Fig. 4A). While QoxABCD requires prenylated hemes (heme o or a) for its activity, CydAB functions with heme b as the sole heme cofactor (51). To determine whether prenylated heme cofactors are produced in ΔhepT ispA::Tn, hemes were extracted from overnight cultures and quantified via HPLC. In each strain, heme b was detected and used as an internal standard to normalize heme o and a abundance across samples. The mutation of ispA leads to a significant decrease in heme o and a abundance compared to WT (Fig. 4B and C). Interestingly, the disruption of MK synthesis via the mutation of menE also leads to a significant decrease in hemes o and a compared to WT and suggests that modified heme cofactor production is reduced during fermentative growth. Supplementation of the menE::Tn mutant with MK-4 partially restores heme a abundance, but not that of heme o. Similarly, ΔhepT also exhibits decreased amounts of prenylated hemes compared to WT, of which only heme a was restored to WT-like levels upon MK-4 supplementation. Given that heme a synthesis is dependent on heme o, it is surprising that MK-4 supplementation disproportionately increases heme a abundance in menE::Tn and ΔhepT. The regulation of prenylated heme synthesis is not known, but these results imply that MK-4 influences the activities of CtaB and CtaA in MK-deficient strains. Notably, ΔhepT ispA::Tn does not produce detectable quantities of heme o or a (Fig. 4B and C), indicating that both HepT and IspA are capable of contributing to prenylated heme production. These findings demonstrate that the QoxABCD activity in the ΔhepT ispA::Tn double mutant is impaired due to a lack of prenylated heme cofactors.

Fig 4.

HepT and IspA contribute to the production of prenylated heme cofactors. (A) An illustration of the prenylated heme synthesis pathway beginning with heme b. Heme b farnesylated by CtaB generating heme o. Heme o is modified by the addition of a carbonyl group to produce heme a. (B and C) High-performance liquid chromatography analysis of hemes extracted from whole cells. Heme b was used as an internal standard to which hemes o (B) and a (C) were normalized for each sample. Data are the average of three independent biological replicates performed in triplicate. Statistical significance was determined via two-way ANOVA. *** and **** represent P-values of 0.001 and 0.0001, respectively.

CydAB function is not stimulated by MK-4

The inability to produce prenylated heme cofactors accounts for impairment of QoxABCD in ΔhepT ispA::Tn, but both CydAB and QoxABCD must be inactive to induce the SCV phenotype. Therefore, we reasoned that the CydAB activity is not stimulated by MK-4. We showed that S. aureus produces five MK species: MK-5, MK-6, MK-7, MK-8, and MK-9. Previous experiments used MK-4, an MK analog that is not natively produced by S. aureus, to chemically complement MK deficiency (Fig. 3A and B). To determine whether MK-4 serves as an electron carrier substrate for CydAB, a ΔmenB qoxA::Tn double mutant was generated. In this strain, respiration is dependent on CydAB and exogenous MKs. The MK-4 supplementation of ΔmenB qoxA::Tn did not promote significant proliferation after 24 h of incubation compared to the un-supplemented condition, which is similar to the phenotype observed for ΔhepT ispA::Tn (Fig. 5A). Importantly, both ΔmenB and ΔmenB cydA::Tn, a strain restricted to QoxABCD for aerobic respiration, exhibited increased proliferation when supplemented with MK-4 (Fig. 5A). To determine if the shortened isoprenoid affects whether MK-4 can promote the CydAB activity, the experiment was repeated using the MK analog menadione (MD), which lacks an isoprenoid moiety. While MD also increased the proliferation of ΔmenB and ΔmenB cydA::Tn, it failed to enhance the proliferation of ΔmenB qoxA::Tn, mimicking the ΔhepT ispA::Tn phenotype (Fig. 5B). Importantly, supplementation with either MK-4 or MD increased the ΔhepT proliferation, indicating that the mutation of hepT alone is not responsible for the respiration phenotype observed in ΔhepT ispA::Tn. These data suggest that while ΔhepT ispA::Tn exhibits a defect in isoprenoid biosynthesis, its inability to perform aerobic respiration stems from the electron carrier preference of CydAB, which cannot effectively utilize MK-4. To provide further evidence of this, ΔispA qoxA::Tn and ΔhepT qoxA::Tn were constructed. The ΔispA qoxA::Tn mutant exhibited similar colony size and doubling time to the WT, indicating that respiration is intact in this strain (Fig. S5A and B). However, ΔhepT qoxA::Tn displayed the SCV phenotype that also fails to respond to the MK-4 supplementation, the same phenotype exhibited by ΔhepT ispA::Tn (Fig. S5C and D). Together, these data show that the CydAB terminal oxidase cannot utilize MK-4 or MD as electron carriers.

Fig 5.

MK-4 and menadione fail to stimulate aerobic respiration in cells restricted to CydAB-dependent respiration (A and B). O.D.₆₀₀ values were collected after 24 h of incubation in TSB supplemented with either 12.5 µM MK-4 (A) or 2.5 µM menadione (MD) (B). Error bars represent one standard deviation from the mean. Statistical significance was determined via two-way ANOVA with Bonferroni correction. **** represents a P-value of <0.0001. Data are the average of three independent biological replicates performed in triplicate.

PDS activity supports host colonization

Isoprenoid biosynthesis contributes to three cellular pathways: carotenoid pigment production, cell envelope synthesis, and respiration. Importantly, each of these pathways contributes to virulence or host colonization (19, 38, 53, 60, 61). Therefore, we sought to determine whether the disruption of PDS function impacts fitness during infection. The heart, liver, and kidneys of systemically infected mice were harvested 96 h post-infection, and bacterial burdens were quantified. The ispA::Tn mutant exhibited significantly reduced abundance in the heart, liver, and kidneys, indicating that ispA plays a role in host colonization (Fig. 6A). Additionally, hepT::Tn displayed significantly reduced burdens in the heart and liver of systemically infected mice (Fig. 6B), a phenotype that is similar to an MK-deficient menC::Tn mutant. The similarity between these results suggests the colonization defect observed for hepT::Tn could be due to a loss of MK production, but further experimentation is required to rule out the contribution of HepT to the other isoprenoid pathways.

Fig 6.

Disruption of isoprenoid synthesis impairs S. aureus colonization across multiple organs during systemic infection. (A) Bacterial burdens of WT or ispA::Tn in the heart, liver, and kidneys quantified after 96 h of infection and represented as colony forming units (CFU) per milliliter (CFU/mL). Statistical significance was determined via unpaired Mann-Whitney test. (B) Bacterial burdens of WT, hepT::Tn, and menC::Tn after 96 h of infection represented as CFU/mL. Prior to infection, hepT::Tn and menC::Tn were supplemented with 12.5 µM MK-4 to achieve a WT-like inoculum. Statistical significance was determined via one-way ANOVA with Tukey correction. (A and B) Error bars represent one standard deviation from the mean. *, **, and *** represent P-values of <0.05, <0.01, and <0.001, respectively.

DISCUSSION

Addressing how bacterial isoprenoid synthesis is initiated in the absence of the short-chain PDS, IspA, is essential to understanding this pathway and its products. Our data shed light on this inquiry by demonstrating that functional redundancy sustains isoprenoid synthesis in S. aureus. We found that the medium-chain PDS HepT functions in three pathways: it is essential for MK synthesis but also plays a role in the lipid II cycle and supports production of prenylated hemes. While it has been hypothesized that HepT synthesizes the isoprenoid moieties used for MK synthesis (32, 37, 62), our results experimentally verify this function. We demonstrate that HepT inactivation results in the loss of MK-5, MK-6, MK-7, MK-8, and MK-9, providing direct evidence that HepT is needed for the synthesis of these MK species. Cells lacking MK exhibit a distinct colony phenotype, called the small colony variant (37, 41, 63, 64). SCV proliferation is limited due to a restricted metabolism that relies on lactic acid fermentation. Consistent with the lack of MK, the hepT mutant growth is impaired compared to WT, and the cells produce lactate, indicating that fermentation is the primary means by which the mutant cells generate energy. We also demonstrate the MK analog, MK-4, chemically complements proliferation and reduces lactate production. Together, these results support the conclusion that HepT is necessary for MK synthesis and respiration.

The isolation of pigmented gentamicin-resistant ispA mutants harboring nonsynonymous point mutations in hepT further supports the importance of the medium-chain PDS in aerobic respiration. Previous results demonstrated that an ispA mutant displays increased gentamicin resistance (27). Gentamicin is an aminoglycoside antibiotic, which relies on the PMF to enter a bacterial cell. Consequently, gentamicin resistance has a known correlation with the respiratory status of the cell (63, 65, 66). This led Krute et al. to conclude that aerobic respiration is impaired in ispA mutants likely due to decreased synthesis of MK and prenylated heme cofactors (27). However, the parental ispA mutant produces MK-8 and MK-9 and generates WT levels of lactate, demonstrating that respiration is intact in this mutant. Conversely, ispA mutants harboring the hepT^A72E allele exhibit significantly reduced MK pools and decreased growth kinetics. Previous work established that MK mutants are gentamicin-resistant (67). Together, these facts indicate that aerobic respiration is active in the ispA mutant but reduced in cells encoding the hepT^A72E allele, providing a mechanism for gentamicin resistance. Why an ispA mutant is more resistant to gentamicin than the WT remains a mystery. Also, it is unclear whether pigmentation is directly selected for or is an artificial consequence of the A72E substitution. Our results are consistent with previous reports that demonstrate this mutation alters the first aspartate-rich motif, increasing the similarity between HepT and short-chain PDSs like IspA (55). This change allows the cells to produce staphyloxanthin and short-chain MK-5, MK-6, and MK-7 but not the longer-chain isoprenoids needed for MK-8 or MK-9. Notably, the complete inactivation of hepT via transposon insertion (hepT::Tn) or in-frame deletion (ΔhepT) also leads to hyperpigmentation. SCVs display a nonpigmented phenotype when grown on solid media (64, 68), and our data demonstrate that mutating hepT leads to the SCV phenotype. It is interesting, then, that hepT mutants are hyperpigmented. This may be explained by an altered metabolite flux toward short-chain staphyloxanthin and away from medium-chain isoprenoids when hepT is mutated. The altered expression of genes in the staphyloxanthin and MK biosynthetic pathways could also explain the hyperpigmentation phenotype. Future investigations quantifying staphyloxanthin precursors with concomitant transcriptional quantification of genes in this pathway and MK synthesis will determine whether metabolite reallocation and altered gene expression promote the hyperpigmentation phenotype of the hepT mutant.

The isolation of pigmented ispA mutants encoding missense hepT mutations led us to hypothesize that HepT is also capable of condensing IPP and DMAPP to initiate isoprenoid synthesis. In keeping with this, we predicted that the simultaneous inactivation of ispA and hepT would be synthetically lethal, as isoprenoid synthesis would not be initiated in this genetic background. Additional rationale for this prediction is provided by a biochemical investigation of HepT and HepT homologs. B. subtilis HepT is capable of condensing IPP and DMAPP, although the reaction is inefficient (33). A similar result was described for the E. coli HepT homolog IspB, which is also capable of using IPP and DMAPP as substrates (29). The fact that IspB is essential in E. coli, A. baumannii, and C. glutamicum under standard growth conditions supports a role for IspB beyond quinone synthesis (35, 46–48). HepT is also a requirement for B. subtilis viability under standard laboratory conditions, but this species can tolerate loss of HepT activity when cultured in medium that supports proliferation of L-form cells that lack peptidoglycan (49). This finding implies that HepT also supports peptidoglycan synthesis. Our results demonstrate that S. aureus is amenable to genetic inactivation of hepT in the presence or absence of ispA. However, ΔhepT ispA::Tn lacked pigmentation and did not respond to chemical complementation with the MK analog, MK-4. Subsequent experimentation demonstrated that ΔhepT ispA::Tn is impaired for aerobic respiration due to inactivity of both terminal oxidases, CydAB and QoxABCD. Terminal oxidases accept electrons from the quinone pool and reduce oxygen, thereby performing the last step of aerobic respiration. Our data show that QoxACBD impairment can be attributed to loss of prenylated heme cofactors, while the inability of MK-4 to stimulate respiration in ΔhepT ispA::Tn is dependent on CydAB. The finding that MK-4 fails to chemically complement ΔmenB qoxA::Tn conclusively shows that CydAB is restricted to the use of long-chain MK, presumably MK-7 or greater. The functional role of short-chain MKs (MK-5 and MK-6) in supporting respiration is unclear given these MK species are produced at significantly reduced quantities compared to the longer-chain MKs (MK-7, MK-8, and MK-9). However, given the apparent preference of terminal oxidases for the isoprenoid chain length of MKs, further investigation should be carried out to precisely define correlations between MK tail length, CydAB terminal oxidase activity, and the composition of the MK pool on cellular respiration.

Investigating ΔhepT ispA:Tn in the context of the single mutants demonstrated that HepT uniquely contributes to MK production, while the precursor for staphyloxanthin is provided exclusively by IspA. Interestingly, either enzyme can contribute to prenylated heme production. Both single mutants exhibit decreased prenylated heme cofactors compared to WT, but levels of heme o in the hepT mutant suggest a preference for HepT. The finding that MK-4 supplementation increases heme a levels but not those of heme o in the ΔhepT and menE::Tn mutants suggests that MK influences the final product of heme prenylation. The synthesis of heme a is dependent on heme o; therefore, an abundance of heme a indicates that flux to heme a is increased in MK-4-supplemented ΔhepT and menE::Tn mutant cultures. Whether the activities of CtaA, CtaB, and CtaM are influenced by MK is currently unknown. Also, QoxABCD can be populated with either heme o or a, and how each heme cofactor might affect interactions with MK species has not been elucidated (51). Addressing these questions will define mechanisms that control heme prenylation and provide additional insights into how respiration is modulated in S. aureus. These results reveal that HepT can function beyond MK synthesis. Consistent with this, we monitored levels of the lipid II precursors Und-P and Und-OH, which revealed decreased abundance of both metabolites in the hepT mutant. Though Und-OH does not participate directly in the lipid II cycle (14), previous reports support a model whereby it serves as a reservoir for rapid conversion to Und-P via undecaprenol kinase when higher levels of Und-P are needed (69). In the hepT mutant, a depleted Und-OH reservoir implies the cell is attempting to maintain Und-P levels and demonstrates that HepT plays a role in maintaining Und-P. Overall, the finding that ΔhepT ispA::Tn is viable provides strong evidence that another enzyme is capable of condensing IPP and DMAPP to produce the FPP needed for essential lipid II. Based on this supposition and our results, a model can be considered whereby IspA, HepT, and the unknown PDS each produce FPP that is restricted to generate staphyloxanthin, MK, and lipid II, respectively (Fig. S6). However, our findings cannot rule out the possibility that FPP generated by the uncharacterized PDS can also be used as a HepT substrate. The fact that HepT homologs are capable of condensing IPP and DMAPP in vitro, albeit at a limited rate, implies that S. aureus HepT is a viable source of FPP (29, 33). In fact, Krute et al. also hypothesized that HepT, UppS, or both can produce FPP to support the growth of ispA mutant cells (27). However, studies of UppS substrate specificity in other bacterial species show that DMAPP is not used as a substrate by this enzyme (9, 70). Discerning between these possibilities requires further biochemical characterization of purified recombinant S. aureus HepT and UppS to establish substrate specificities and reaction kinetics. Nonetheless, our work demonstrates that in addition to its importance in producing MK, HepT plays a multifunctional role by contributing to the isoprenoid-dependent heme cofactor and lipid II precursor pathways.

Staphyloxanthin, MK, terminal oxidase activity, and aerobic respiration have been shown to be critical factors for S. aureus pathogenesis (19, 38, 53, 71). To determine the contributions of IspA and HepT to S. aureus host colonization, we used the systemic mouse model of infection. This analysis showed that the hepT mutant exhibited colonization defects in the heart and liver, which was similar to the MK-deficient menC mutant. This finding suggests that the loss of MK is a factor in the colonization defects displayed by the hepT mutant. However, further investigation is required to determine whether perturbations of the other HepT-associated isoprenoids also contribute to the decreased virulence of the mutant. The ispA mutant showed broader colonization defects, with reduced bacterial burden not only in the heart and liver but also in the kidneys. The disruption of isoprenoid biosynthesis has been previously established to reduce virulence, as mice treated with a chemical inhibitor of UppS exhibit increased survival compared to untreated mice in a model of systemic infection (72). Our data demonstrate that inhibiting isoprenoid synthesis at an earlier step in this pathway is an alternative strategy for reducing S. aureus infection.

S. aureus is a common nosocomial pathogen that is frequently resistant to antibiotics, presenting a clinical challenge (73). Furthermore, the presence of antibiotic-resistant S. aureus in hospital settings can complicate surgical recovery and lead to increased mortality (74). By identifying biological pathways that support survival in the host, we reveal potential targets that could be used for drug development and therapeutic intervention. In this study, we showed that the genetic disruption of isoprenoid biosynthesis impacted host colonization and impeded several downstream pathways that are known to support pathogenesis, revealing the potential of this pathway as a drug target. Additionally, our data support a redundant model of isoprenoid biosynthesis and showed that a high level of disruption in this pathway is tolerated by S. aureus but limits the pathogen’s metabolic versatility. As isoprenoid biosynthesis is a highly conserved biological process that is active in other bacterial pathogens, developing strategies for targeting bacterial isoprenoid biosynthesis may impact treatment of a wide range of microbial pathogens.

MATERIALS AND METHODS

Bacterial strains, plasmids, and culture conditions

S. aureus strain JE2 was used as the WT strain in this study. All S. aureus mutants were generated in the JE2 WT background as listed in Table 1. Transposon mutagenesis was carried out by propagating φ85 on the appropriate strain from the Nebraska Transposon Mutant Library and transduced into a recipient strain as described previously (20, 75). However, for generating the ΔhepT ispA::Tn double mutant, we found no colonies were obtained when plating on TSA (Remel) supplemented with 10 µg/mL erythromycin and 40 mM sodium citrate. Instead, plating on TSA supplemented with 10 µg/mL erythromycin without sodium citrate yielded colonies. In-frame deletions were generated via an allelic exchange protocol with the pKOR1mcs-ΔhepT plasmid as described previously and confirmed via PCR (76, 77). The previously generated pKOR1-ΔmenB plasmid was used to generate the ΔmenB deletion mutant in the JE2 background (37). Cloning and PCR confirmation of mutants were performed using the primers listed in Table S2. All overnight cultures were started from single colonies in 5 mL TSB (Fisher Scientific) and incubated overnight at 37°C at 225 rpm unless stated otherwise. Mutants deficient for MK synthesis were supplemented with 12.5 µM MK-4 in overnight cultures. All strains harboring the pOS1 plasmid were grown in the presence of 10 µg/mL chloramphenicol. Genes were cloned into pOS1 via restriction ligation or Gibson assembly, as denoted in Table S2, transformed into E. coli DH5α, and passaged through the restriction minus modification plus S. aureus strain RN4220 prior to being transformed into the respective JE2 strain.

TABLE 1.

Strains used in this study

Strain	Description	Reference
RN4220	Restriction-deficient, methylation-proficient cloning intermediate	(78)
JE2	USA300, wild type (WT)	(20)
ispA::Tn	ispA transposon from NE1447 backcrossed into WT JE2	This study
hepT::Tn	hepT transposon from NE1920 backcrossed into WT JE2	This study
ΔhepT	In-frame deletion of hepT constructed using pKOR1-ΔhepT	This study
ΔhepT ispA::Tn	ispA transposon from NE1447 transduced into ΔhepT	This study
menE::Tn	Tn917 inserted into menE backcrossed into WT JE2	(71)
ΔmenB	In-frame deletion of menB constructed using pKOR1-ΔmenB	This study
menC::Tn	Tn917 inserted into menC backcrossed into WT JE2	(71)
qoxA::Tn	qoxA transposon from NE92 backcrossed into WT JE2	This study
cydA::Tn	cydA transposon from NE117 backcrossed into WT JE2	This study
ΔmenB qoxA::Tn	qoxA transposon from NE92 transduced into ΔmenB	This study
ΔmenB cydA::Tn	cydA transposon from NE117 transduced into ΔmenB	This study
crtM::Tn	crtM transposon from NE2499 backcrossed into WT JE2	This study
ΔhepT crtM::Tn	crtM transposon from NE2499 transduced into ΔhepT	This study
ΔispA	In-frame deletion of ispA constructed using pKOR1-ΔispA	This study
ΔispA qoxA::Tn	qoxA transposon from NE92 transduced into ΔispA	This study
ΔispA cydA::Tn	cydA transposon from NE117 transduced into ΔispA	This study
ΔhepT qoxA::Tn	qoxA transposon from NE92 transduced into ΔhepT	This study
ΔhepT cydA::Tn	cydA transposon from NE117 transduced into JE2 ΔhepT	This study
ispA::Tn suppressor 115	Gentamicin-resistant pigmented ispA::Tn isolate that harbors the HepTA72E allele	This study
ispA::Tn suppressor 144	Gentamicin-resistant pigmented ispA::Tn isolate that harbors the HepTA72E allele	This study
ispA::Tn suppressor 164	Gentamicin-resistant pigmented ispA::Tn isolate that harbors the HepTA72E allele	This study
ispA::Tn suppressor 165	Gentamicin-resistant nonpigmented ispA::Tn isolate that harbors the HepTA165T allele	This study
JE2 pOS1	Empty pOS1 plasmid vector	This study
ispA::Tn pOS1	Empty pOS1 plasmid vector	This study
ispA::Tn pOS1-ispA	ispA::Tn harboring a pOS1 plasmid encoding the ispA gene under the control of the Plgt promoter	This study
ΔhepT pOS1	Empty pOS1 plasmid vector control	This study
ΔhepT pOS1-hepT	ΔhepT harboring a pOS1 plasmid encoding the hepT gene under the control of the Plgt promoter	This study
ΔhepT ispA::Tn pOS1	Empty pOS1 plasmid vector control	This study
ΔhepT ispA::Tn pOS1-hepT	ΔhepT ispA::Tn harboring a pOS1 plasmid encoding the hepT gene under the control of the Plgt promoter	This study
ΔhepT ispA::Tn pOS1-ispA	ΔhepT ispA::Tn harboring a pOS1 plasmid encoding ispA under the control of the Plgt promoter	This study
ΔhepT ispA::Tn pOS1-hepTA72E	ΔhepT ispA::Tn harboring a pOS1 plasmid encoding the hepTA72E allele from ispA::Tn suppressor 115 under the control of the Plgt promoter	This study

Isolation of gentamicin-resistant pigmented ispA::Tn suppressor mutants

An overnight culture of ispA::Tn was prepared from a single colony in TSB. A 1:10 dilution of the 10⁹ CFU/mL overnight culture was made and spread onto a TSA plate supplemented with 3 µg/mL gentamicin. The plate was incubated at 37°C, and pigmented colonies were observed after ~72 h of incubation. Pigmented colonies were isolated on TSA, and a single colony was used to make an overnight culture to ensure a clonal population was obtained. The overnight cultures were used to make frozen archival stocks, and cells used for subsequent experimentation involving the gentamicin-resistant pigmented ispA::Tn suppressors were recovered from these stocks. Table S1 lists single nucleotide polymorphisms (SNPs) present in each suppressor isolate.

Whole-genome sequencing and analysis

Genomic DNA was extracted from 1 mL of the overnight cultures using a Wizard Genomic DNA Purification Kit (Promega). Genomic DNA was sequenced via Illumina sequencing at the Duke University Sequencing and Genomic Technologies core facility. Genomic analyses, including paired-end read trimming, mapping, and SNP calling, were performed using Geneious Prime version 2024.0.5. The genome of S. aureus strain USA300_FPR3757 (GenBank accession number: CP000255.1) was used as a reference genome to which sequencing data were mapped.

Quantification of isoprenoid-derived metabolites via high-performance liquid chromatography mass spectrometry

MKs and C₅₅ isoprenoids were extracted and analyzed as described previously with slight modifications (29). S. aureus cultures were grown in 50 mL of TSB at 37°C with shaking at 225 rpm until stationary phase. Cultures were pelleted and resuspended in 2 mL of methanol:0.3% NaCl (10:1, v/v). Vitamin K₁, solanesol, and solanesyl phosphate were added as internal standards. To extract MKs and Und-OH, hexane was added to the cell suspension and vortexed, and the upper phase was collected. The remaining Und-OH and Und-P in the aqueous layer were treated with alkali by adding 1 mL 60% KOH and boiled for 60 min. Diethyl ether was added and vortexed, then collected after phase separation. The collected diethyl ether was washed with 5% acetic acid.

The hexane solution containing MKs and polyprenols was loaded onto a 0.4 g column of neutral alumina (grade III). MKs were eluted with 2.4% diethyl ether in hexane, and polyprenols were eluted with 10% diethyl ether in hexane. MKs and polyprenols were run on an LCMS-2010 (Shimadzu Co.) using an STR ODS_II column (Shinwa Chemical Ind. Ltd.). A 2-propanol:methanol (1:1, v/v) mixture was used as the mobile phase at a flow rate of 0.1 mL/min. The diethyl ether extract was divided in half. One half was dried under nitrogen gas, and the remaining residue was resuspended in hexane. The hexane solution was loaded onto a neutral alumina column, and polyprenols were eluted with 10% diethyl ether in hexane as described above. The other half of the diethyl ether extract was dried under nitrogen gas, and the remaining residue was resuspended in chloroform:methanol (2:1, v/v). The chloroform:methanol (2:1, v/v) solution containing polyprenyl phosphates derived by alkaline hydrolysis was loaded onto an ion exchange cartridge (Supelclean LC-NH₂). Polyprenyl phosphates were eluted with a chloroform:methanol:water (2:0.9:0.1, v/v/v) solution containing 0.1 M ammonium acetate and analyzed via HPLC using a STR ODS_II column with 2-propanol:methanol (1:1, v/v) containing 5 mM phosphoric acid.

Growth curve analysis and end-point optical density reading

Overnight cultures were pelleted and resuspended in 137 mM NaCl, 2.7 mM KCl, 10.1 mM Na₂HO₄P, and 1.8 mM KH₂O₄P phosphate-buffered saline pH 7.4 (PBS) and placed on ice. The optical density at 600 nm (O.D.₆₀₀) was measured for resuspended cultures and normalized to an O.D.₆₀₀ equal to 1.0 (or ~3.5 × 10⁹ CFU/mL). A volume of 150 µL growth medium was dispensed into each well of a 96-well plate, and 1.5 µL of the normalized culture (1:100 dilution) was used to inoculate each well. The plate was incubated in a Stratus (Cerillo) plate reader at 37°C with shaking at 300 rpm. To calculate the doubling time, the following equation was used: doubling time = (Ln(O.D.₂) – Ln(O.D.₁)) / (T₂ – T₁), where O.D.₁ represents the first O.D.₆₀₀ reading at the start of the mid-log phase, and O.D.₂ represents the last O.D.₆₀₀ reading of the mid-log phase. T₁ and T₂ represent the time points at which O.D.₁ and O.D.₂ were collected, respectively. For growth analysis where the end-point, stationary-phase O.D.₆₀₀ was collected, a 96-well plate was prepared and incubated as described above. After the designated incubation time, the plate was removed from the Stratus plate reader, and each well was carefully pipetted up and down to ensure all cells were resuspended. A single time point was then collected by measuring the O.D.₆₀₀ using an H1 Synergy (Biotek) plate reader.

Measuring l-lactate production

The EnzyChrom l-Lactate Assay Kit (ECLC-100, BioAssay Systems) was used to assess the l-lactate production in S. aureus cultures. Overnight cultures were pelleted and normalized to an O.D.₆₀₀ of 1.0 in PBS and placed on ice. Normalized cultures were subcultured 1:100 in 5 mL fresh TSB and incubated at 37°C with shaking at 225 rpm for 15 h. Cultures were then pelleted, and the supernatant was collected, sterile-filtered, and stored at −20°C. Sterile supernatants were diluted 1:10 in sterile Milli-Q water, and 20 µL was dispensed into a well of a 96-well plate. The reaction buffer was prepared by mixing the following reagents from the l-lactate assay kit: 60 µL assay buffer, 1 µL enzyme A, 1 µL enzyme B, 10 µL NAD, and 14 µL MTT. In a 96-well plate, 20 µL of the diluted supernatant was added to a well for each sample. An 80 µL volume of reaction buffer was added to each well and mixed briefly by pipetting up and down. The initial O.D.₅₆₅ was measured immediately after mixing using an H1 Synergy plate reader, and then the plate was incubated at room temperature for 20 min. The O.D.₅₆₅ was measured again, and the initial O.D.₅₆₅ measurement was subtracted from the second measurement to yield the sample values. The sample values were compared to a standard curve (prepared as described in the l-lactate assay kit) to determine l-lactate concentrations.

Heme quantification via HPLC

In a 125 mL Erlenmeyer flask, 20 mL of TSB was inoculated with a single colony and grown overnight at 37°C with shaking at 225 rpm. Cultures were pelleted and resuspended in 200 µL molecular biology grade water (Cytiva: SH30538.03). A volume of 100 µL of the cell suspension was transferred to a 1.5 mL Eppendorf tube. An acid:acetone solution was prepared, consisting of five parts 12 M HCl to 95 parts acetone. A 150 µL volume of acid acetone was added to the cell suspension and mixed briefly by vortexing. The mixture was incubated on ice for 10 min and vortexed for 20 s. The mixture was further incubated for 10 min and then vortexed for 20 s. The cells were pelleted, and the supernatant was collected. The supernatant was centrifuged again, and the resulting supernatant was transferred to an HPLC injection vial.

Hemes were analyzed with the Agilent 1260 Infinity HPLC System using the InfinityLab Poroshell EC-C18 reverse-phase column (Agilent: 699975-902) coupled with a diode array detector (Agilent: G1315D). A 10 µL volume of sample was injected, and hemes were separated using a gradient of solvents A (0.1% v/v trifluoroacetic acid in Milli-Q water) and B (0.1% v/v trifluoroacetic acid in acetonitrile). The gradient progressed as follows: 25% solvent B (0.00–2.67 min), 25–55% solvent B (2.67–4.33 min), 55–75% solvent B (4.33–11.00 min), 75–100% solvent B (11.00–12.67 min), 100% solvent B (12.67–20.00 min), and a gradient returning to 25% solvent B (20.00–23.00 min). Using this method, heme b elutes at 6.8 min, heme a at 10.8 min, and heme o at 12 min. Hemes were detected by measuring absorbance at 400 nm and identified by comparing retention time and the absorbance maximum (79). HPLC chromatograms were used to determine the area under the peak for each of the heme species. Normalized heme abundance was determined by dividing the peak area of hemes o and a by the peak area of heme b.

Systemic mouse infections

Overnight cultures were subcultured 1:100 in 5 mL fresh TSB and incubated at 37°C with shaking at 225 rpm for 3 h. MK-4 was used to supplement menC::Tn and hepT::Tn cultures to induce respiration and promote growth to WT and ispA::Tn mutant levels. Cultures were pelleted at 4°C and washed once in 12 mL of Dulbecco’s PBS (DPBS, Sigma-Aldrich: D8537). Resuspended cultures were pelleted again and normalized to an O.D.₆₀₀ of 0.4 in DPBS and kept on ice. Eight-week-old female BALB/c mice (Jackson Laboratories) were retro-orbitally infected with 10⁷ CFU of the indicated S. aureus strain. Mice were euthanized via CO₂ inhalation 96 h post-infection, and the heart, liver, and kidneys were harvested. The heart and kidneys were homogenized via bead beating (1.5 mL RINO lysis beads, NextAdvance, Inc.) in 500 μL DPBS using Bullet Blender Storm24 (NextAdvance, Inc.). Once homogenized, 500 μL DPBS was added to the lysate. The livers were manually homogenized in a Whirl-pak (Nasco) containing 1 mL DPBS. Organ homogenates were serially diluted in DPBS and plated on TSA for enumeration. All infections were performed at Michigan State University under the principles and guidelines described in the Guide for the Care and Use of Laboratory Animals described in protocol PROTO202200474, which received approval from the Michigan State University Institutional Animal Care and Use Committee.

SUPPLEMENTAL MATERIAL

The following material is available online at https://doi.org/10.1128/mbio.00353-25.

Supplemental figures and tables. mbio.00353-25-s0001.docx.

Figure S1 to S6 and Table S1 and S2.

ASM does not own the copyrights to Supplemental Material that may be linked to, or accessed through, an article. The authors have granted ASM a non-exclusive, world-wide license to publish the Supplemental Material files. Please contact the corresponding author directly for reuse.