A Staphylococcal Glucosaminidase Drives Inflammatory Responses by Processing Peptidoglycan Chains to Physiological Lengths

ABSTRACT

The peptidoglycan of Staphylococcus aureus is a critical cell envelope constituent and virulence factor that subverts host immune defenses and provides protection against environmental stressors. Peptidoglycan chains of the S. aureus cell wall are processed to characteristically short lengths by the glucosaminidase SagB. It is well established that peptidoglycan is an important pathogen-associated molecular pattern (PAMP) that is recognized by the host innate immune system and promotes production of proinflammatory cytokines, including interleukin-1β (IL-1β). However, how bacterial processing of peptidoglycan drives IL-1β production is comparatively unexplored. Here, we tested the involvement of staphylococcal glucosaminidases in shaping innate immune responses and identified SagB as a mediator of IL-1β production. A ΔsagB mutant fails to promote IL-1β production by macrophages and dendritic cells, and processing of peptidoglycan by SagB is essential for this response. SagB-dependent IL-1β production by macrophages is independent of canonical pattern recognition receptor engagement and NLRP3 inflammasome-mediated caspase activity. Instead, treatment of macrophages with heat-killed cells from a ΔsagB mutant leads to reduced caspase-independent cleavage of pro-IL-1β, resulting in accumulation of the pro form in the macrophage cytosol. Furthermore, SagB is required for virulence in systemic infection and promotes IL-1β production in a skin and soft tissue infection model. Taken together, our results suggest that the length of S. aureus cell wall glycan chains can drive IL-1β production by innate immune cells through a previously undescribed mechanism related to IL-1β maturation.

INTRODUCTION

Methicillin-resistant Staphylococcus aureus remains one of the most significant burdens to human health, with ~750,000 associated deaths worldwide reported in 2019; a third of which were directly attributed to antimicrobial resistance (AMR) (1). Many therapeutics that target cell wall synthesis are ineffective on account of AMR, underscoring the importance of peptidoglycan (PGN) synthesis for microbial survival. The mesh-like layer of PGN not only provides environmental protection and structural integrity for S. aureus, but it also serves as an important pathogen-associated molecular pattern (PAMP) for initiation of inflammatory responses by host innate immune cells, like macrophages and dendritic cells (2).

S. aureus peptidoglycan is composed of glycan strands of repeating disaccharide units of N-acetylmuramic acid (NAM) and N-acetylglucosamine (NAG). Individual glycan strands are linked by pentaglycine cross bridges formed between stem peptides attached to NAM residues. The length of glycan strands varies greatly between bacterial species, but S. aureus produces characteristically short glycan chains with average lengths of 3 to 10 disaccharide units (3, 4). Glycan chain length is primarily dictated by the glucosaminidase SagB and is facilitated by surface protein determinant C (SpdC), an abortive infectivity (Abi) domain-containing membrane protein that aids in the positioning of SagB for glycan chain processing (5–7). Mutation of sagB leads to a cell wall with elongated glycan chains that often exceed 50 disaccharides in length (4). Consequently, a ΔsagB mutant has a partially disrupted protein secretion profile, increased susceptibility to the β-lactam antibiotic oxacillin, and increased resistance to digestion with the endopeptidase lysostaphin (8). Additionally, a ΔsagB mutant is attenuated in a murine systemic infection model and has reduced survival in human macrophages relative to the wild type (WT) (9). Thus, processing of glycan chains by SagB is critical for S. aureus physiology and virulence.

Interleukin-1β (IL-1β) is a critical cytokine for the host inflammatory response to S. aureus (10–12). Production and release of IL-1β by innate immune cells occur in response to peptidoglycan recognition (13–16). Initially produced as a 31-kDa inactive precursor, pro-IL-1β is processed into mature IL-1β and released to induce inflammation (17). Maturation of IL-1β is best described in the context of the NLRP3 inflammasome and classically occurs in two stages. The first stage, termed “priming,” involves transcriptional upregulation of the inflammasome components and pro-IL-1β in response to PAMP engagement of pattern recognition receptors (PRRs), such as Toll-like receptors (TLRs). Most evidence supports the idea that S. aureus peptidoglycan is recognized via surface receptor TLR2 and cytosolic receptor nucleotide-binding oligomerization domain 2 (NOD2), although several studies argue against the importance of TLR2 on account of contaminating lipoproteins and structural differences between purified peptidoglycan preparations (13, 14, 18–31). A second “activation” signal is induced by diverse stimuli and causes inflammasome assembly and subsequent caspase-1 activation, which facilitates cleavage of pro-IL-1β into mature IL-1β. As IL-1β lacks a secretion signal, release from the cell can occur via multiple unconventional secretion mechanisms but largely relies on caspase-1-mediated activation and oligomerization of gasdermin D (GSDMD), which forms pores that serve as conduits for IL-1β release (32). However, some reports have identified alternative mechanisms of IL-1β release in response to bacterial pathogens that stray from this classical model (14, 33–35).

Prior studies suggest that phagocytosis and phagolysosomal degradation of S. aureus facilitate the release of cell wall PAMPs and precede IL-1β production by macrophages (13, 36, 37). However, S. aureus resists degradation by host lysozyme via O-acetylation of NAM residues in glycan chains by the O-acetyltransferase OatA. Mutation of oatA renders S. aureus susceptible to lysozyme digestion and subsequently bolsters IL-1β production (13). Reducing the degree of cross-linking between glycan chains by culturing S. aureus in the presence of β-lactam antibiotics enhances lysosomal degradation and IL-1β production by macrophages (16). In addition, peptidoglycan fragments generated by lysosomal degradation are released into the cytosol, where they are sensed by intracellular receptors, such as NOD2. Minimally, the NAG monosaccharide of peptidoglycan can activate the NLRP3 inflammasome by triggering the release of the glycolytic enzyme hexokinase from the mitochondrial membrane (14). Furthermore, while it is known that disaccharide peptidoglycan species induce macrophage il-1b transcription more robustly than monosaccharide species, comparatively less is known about peptidoglycan composition and its impact on immunological responses to S. aureus (38).

In this study, we sought to investigate the role of staphylococcal glucosaminidases in shaping the host immune response. We found that one of these enzymes, SagB, is critical for production of IL-1β by innate immune cells in response to heat-killed S. aureus and purified peptidoglycan. Furthermore, reestablishing glycan chain length with recombinant SagB restored IL-1β production by macrophages. SagB-dependent IL-1β production appears to be due to caspase-independent cleavage of pro-IL-1β. Altogether, these results provide foundational evidence to suggest that the short glycan chains of S. aureus peptidoglycan are sufficient to drive IL-1β production by innate immune cells.

RESULTS

SagB is a mediator of IL-1β production by innate immune cells.

To test the role of staphylococcal glucosaminidases in the activation of the innate immune response, we applied heat-killed preparations (HKSA) of S. aureus transposon insertion mutants of each of the four known glucosaminidase genes (atl::Tn, scaH::Tn, sagA::Tn, and sagB::Tn) to bone marrow-derived macrophages (BMMs) and quantified the cytokines and chemokines produced. We found that HKSA from a sagB::Tn mutant led to significantly less IL-1β production than HKSA derived from the WT strain (JE2) (Fig. 1A). In contrast, production of other proinflammatory cytokines, such as IL-6, KC, and tumor necrosis factor (TNF), was unperturbed (Fig. 1B). The atl::Tn, scaH::Tn, and sagA::Tn mutants had no impact on IL-1β, IL-6, KC, or TNF production (Fig. 1A and B). We subsequently generated a ΔsagB in-frame deletion mutant and a ΔsagB + sagB complementation strain in the USA300 LAC background and found that addition of HKSA from a ΔsagB mutant to BMMs did not induce production of IL-1β, whereas HKSA from the ΔsagB + sagB complementation strain restored IL-1β levels to WT levels (Fig. 1C). Levels of BMM production of IL-6, KC, and TNF were identical upon addition of HKSA from all strains tested (Fig. 1D). Viability of macrophages was similar between all HKSA treatments, indicating that the change in IL-1β production is not due to a difference in inflammatory cell death (Fig. 1E). These data indicate SagB is the main glucosaminidase that drives IL-1β production by primary macrophages treated with HKSA.

FIG 1.

SagB is the primary glucosaminidase that promotes IL-1β production by macrophages. BMMs were incubated with HKSA of the indicated strains (MOI of 100) overnight, and supernatants were collected to quantify IL-1β (A and C), IL-6, KC, and TNF (B and D) by CBA. (E) BMM viability was determined using an EtBr staining assay. The data shown are from one of at least three experiments done in triplicate. Statistical significance was determined by one-way ANOVA with Tukey’s post hoc test. **, P < 0.01; ****, P < 0.0001.

Characterization of SagB-mediated production of IL-1β by innate immune cells.

To test if SagB-mediated production of IL-1β was restricted to BMMs, we generated bone marrow dendritic cells (BMDCs) and applied HKSA from the WT, ΔsagB, and ΔsagB + sagB strains. BMDCs treated with HKSA from a ΔsagB mutant produced significantly lower levels of IL-1β, while IL-6 production remained high (Fig. 2A). Additionally, we assessed if the cellular material from a ΔsagB mutant was blocking or failing to activate innate cell production of IL-1β. Macrophages were cotreated with WT (multiplicity of infection [MOI] of 100) and ΔsagB HKSA in 10-fold serial dilutions (MOI of 100, 10, and 1) to evaluate if the presence of cellular material from a ΔsagB mutant blocks IL-1β production in response to WT HKSA. When WT and ΔsagB HKSA were mixed 1:1, 10:1, or 100:1, the levels of IL-1β produced were equivalent to that of macrophages stimulated with WT HKSA alone (Fig. 2B). In keeping with what was observed for HKSA, we found that infection of BMMs with a live ΔsagB mutant strain led to reduced IL-1β production early after infection (Fig. 2C). Thus, the presence of SagB is sufficient to promote IL-1β production by BMMs and BMDCs after treatment with live or heat-killed S. aureus. Furthermore, these data suggest that SagB activity facilitates some aspect of IL-1β production or maturation that is impeded in a ΔsagB mutant.

FIG 2.

SagB is critical for IL-1β production by macrophages and dendritic cells. (A) BMDCs were incubated with HKSA of the indicated strains at an MOI of 100, and IL-1β and IL-6 were measured via CBA. (B) HKSA from indicated strains was added to BMMs in mixed ratios overnight, and supernatants were collected to quantify IL-1β. (C) BMMs were infected with the indicated strains of live S. aureus (MOI of 5) for 4 h, at which point extracellular bacteria were removed by washing and gentamicin treatment. One hour later, supernatants were collected for CBA analysis. (D) BMMs were incubated with HKSA of the indicated strains at an MOI of 100, and IL-1β and IL-6 were measured. The data shown are from one of at least three experiments performed in triplicate. Statistical significance was determined by one-way ANOVA with Tukey’s post hoc test. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

Recent studies suggest that the Abi domain membrane protein SpdC interacts with and promotes the function of SagB (5–7). To investigate if SpdC promotes SagB-dependent production of IL-1β by BMM, we generated ΔspdC and ΔspdC ΔsagB mutants, applied HKSA from these mutants to BMMs, and measured IL-1β production. Treatment with HKSA from a ΔspdC mutant led to a modest but statistically significant decrease in IL-1β production by BMMs, whereas treatment with HKSA from ΔsagB and ΔspdC ΔsagB mutants led to complete loss of IL-1β production (Fig. 2D). In contrast, IL-6 levels remained unchanged. These data suggest that IL-1β production by macrophages is primarily dependent on the activity of SagB, whereas SpdC contributes to a moderate degree, in agreement with its proposed role in promoting SagB efficiency (5–7).

SagB is required for virulence during systemic infection and IL-1β production in skin and soft tissue infection.

Since treatment with a ΔsagB mutant is unable to promote IL-1β production from BMMs and BMDCs relative to WT S. aureus in vitro, we performed murine infection experiments to determine if this phenotype impacts infection or IL-1β production in vivo. Using a skin and soft tissue infection (SSTI) model, we found that infection with the ΔsagB mutant led to similar CFU recovery to infection with WT S. aureus (Fig. 3A). However, homogenized abscesses from mice infected with a ΔsagB mutant had significantly reduced tissue IL-1β and overall abscess area compared to mice infected with WT or the ΔsagB + sagB strain (Fig. 3B and C). Consistent with prior studies, mice infected in the bloodstream with a ΔsagB mutant were attenuated for infection in the kidneys (9) (Fig. 3D) as well as the heart (Fig. 3E). Mice infected with a ΔsagB mutant exhibited a trend toward lower levels of IL-1β in the kidneys than the WT and ΔsagB + sagB strains at 96 h, but differences were not statistically significant (Fig. 3F). We detected no IL-1β in the hearts regardless of the infecting strain (Fig. 3G). Taken together, these data indicate that a ΔsagB mutant is attenuated for systemic infection and has diminished IL-1β production and abscess lesion size in an SSTI infection model, consistent with in vitro observations.

FIG 3.

SagB is required for systemic infection and promotes IL-1β production during skin and soft tissue infection. (A) Four-week-old Swiss Webster mice were injected intradermally with 1 × 10⁷ CFU/mL of the indicated strains, and the number of CFU per abscess was enumerated after 120 h postinfection (n = 24). (B) IL-1β from abscess homogenates was quantified via CBA. (C) Abscess area (square millimeters) for mice intradermally infected with the indicated strains. (D and E) Mice were also systemically infected with 1 × 10⁷ CFU/mL of the indicated strains, and at 96 h postinfection, kidneys (D) and hearts (E) were harvested for enumeration of CFU per organ (n = 32). (F and G) IL-1β was quantified in kidney (F) and heart (G) homogenates via CBA. For CFU data, statistics were calculated using a nonparametric one-way ANOVA with Kruskal-Wallis multiple-comparison posttest. For CBA data, statistical significance was determined by one-way ANOVA with Tukey’s post hoc test. *, P < 0.05; ***, P < 0.005.

SagB-processed peptidoglycan is sufficient for IL-1β maturation.

Thus far, our data implicate SagB in the ability of S. aureus to induce IL-1β production by innate immune cells. Given that SagB is the primary glucosaminidase required for processing glycan chains of the cell wall to shortened lengths, we posited that the enrichment of elongated glycan chains within the cell wall of a ΔsagB mutant might hinder the ability to induce IL-1β production by macrophages. To test this possibility, we first purified peptidoglycan from WT and ΔsagB mutant strains and subsequently digested the peptidoglycan with purified recombinant SagB and/or lysostaphin to hydrolyze peptide cross bridges and solubilize the insoluble material. In agreement with previous work, analysis of peptidoglycan by reverse-phase high-performance liquid chromatography (RP-HPLC) revealed an enrichment of long glycan chains (broad peak of >100 min) and significant reductions in short glycan chains (peaks at <100 min) for a ΔsagB mutant (8) (Fig. 4A). As a control, we also performed digestions with the purified glucosaminidase domain of Atl, which is known to cleave glycan strands to their smallest subunits (8) (Fig. 4B). Treatments of WT and ΔsagB peptidoglycan with both Atl and lysostaphin yielded small glycan fragments (peaks at 25 min) (Fig. 4B). Upon treatment with SagB, the glycan elution profiles of WT and ΔsagB mutant peptidoglycan were enriched for glycans that eluted between 50 and 100 min (Fig. 4C). Thus, treatment of peptidoglycan from a ΔsagB mutant with recombinant SagB yields short glycan chains of similar length to those typically found in the S. aureus cell wall and that are distinct in length from the smallest peptidoglycan subunits produced by Atl (8) (Fig. 4B and C).

FIG 4.

SagB processing of peptidoglycan is sufficient to promote IL-1β production by macrophages. (A to C) Purified peptidoglycan from the WT (black) or ΔsagB mutant (green) was treated with lysostaphin (0.1 mg/mL) (A), recombinant Atl (0.5 mg/mL) and then lysostaphin (B), or recombinant SagB (0.5 mg/mL) and then lysostaphin (C). Solubilized glycan chains were reduced with sodium borohydride, separated by RP-HPLC, and detected at 206 nm. (D) Purified peptidoglycan from WT or the ΔsagB mutant was subjected to the indicated treatments, and soluble peptidoglycan (~6 to 8.5 μg/mL) was added to BMMs overnight. IL-1β was measured via CBA. (E) Undigested particulate peptidoglycan (~250 μg) was added to macrophages overnight, and IL-1β was measured. Statistical significance was determined by one-way ANOVA with Tukey’s post hoc test. **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

To test if SagB-processed peptidoglycan was sufficient for IL-1β production by BMMs, we digested WT or ΔsagB mutant peptidoglycan with recombinant lysostaphin or SagB plus lysostaphin and subsequently added the peptidoglycan to BMMs (Fig. 4D). Addition of soluble (lysostaphin-treated) WT peptidoglycan to BMMs led to production of IL-1β, while ΔsagB mutant peptidoglycan did not, similar to what was observed for live and HKSA (Fig. 1 and 2). However, upon incubation with recombinant SagB, the ΔsagB mutant peptidoglycan induced IL-1β production from BMMs at levels that were comparable to that of WT peptidoglycan (Fig. 4D). Additionally, we added insoluble peptidoglycan from the WT and the ΔsagB mutant with no prior enzymatic treatments to BMMs and found a similar dependency on SagB for IL-1β production (Fig. 4E). These data argue that processing of S. aureus peptidoglycan by SagB is critical for the induction of IL-1β by macrophages and that peptidoglycan alone is sufficient to induce the response.

Macrophage phagocytic capacity and lysozyme resistance do not drive SagB-dependent IL-1β production by BMM.

Thus far, we have established that SagB-dependent processing of glycan chains is sufficient for BMM production of IL-1β. Prior studies demonstrated that mutation of sagB also conferred increased resistance to the endopeptidase lysostaphin as well as increased rigidity of the cell wall (4, 8). However, others reported that a sagB::kan mutant exhibits increased sensitivity to macrophage killing in vitro (9). Thus, it is plausible that mutation of sagB may affect the ability of macrophages to phagocytose and degrade S. aureus in the lysosome, thereby preventing the release of peptidoglycan fragments important for activation of IL-1β production. Deletion of the gene encoding the O-acetyltransferase (OatA) renders S. aureus highly susceptible to lysozyme, a major degradative enzyme in the phagolysosome of phagocytic leukocytes (13, 36, 37). To test if susceptibility to enzymatic degradation in the phagolysosome drives SagB-dependent IL-1β production, we generated ΔoatA and ΔsagB ΔoatA mutants and evaluated IL-1β production by macrophages after incubation with HKSA of each strain. While both strains were rendered equally susceptible to lysozyme (Fig. 5A), the ΔsagB mutant and the ΔsagB ΔoatA mutant similarly failed to induce IL-1β production by macrophages (Fig. 5B). In line with previous reports, the ΔoatA mutant stimulated slightly higher IL-1β levels than the WT strain (13). Together, these data suggest SagB-dependent IL-1β production by HKSA-treated BMMs is not solely attributed to disrupted degradative capacity, although it remains possible that enhanced degradative susceptibility unrelated to lysozyme resistance could contribute to IL-1β production.

FIG 5.

SagB-dependent production of IL-1β by macrophages is independent of lysozyme sensitivity and phagocytosis. (A) The indicated strains of S. aureus were grown in TSB with or without the addition of lysozyme (1.5 mg/mL) at mid-exponential phase, and OD₆₀₀ was measured. (B) BMMs were stimulated with HKSA (MOI of 100) from the indicated strains, and supernatants were collected for IL-1β quantification by CBA. (C and D) BMMs were pretreated with the phagocytosis inhibitor cytochalasin D (6 μM) or dynasore (40 μM) 1 h prior to incubation with HKSA (MOI of 100). IL-1β (C) and IL-6 (D) were measured by CBA. The data shown are from one of at least three experiments performed in triplicate. Statistical significance was determined by one-way ANOVA with Tukey’s post hoc test. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

To test if phagocytosis is a prerequisite for SagB-dependent IL-1β production by BMMs, we pretreated macrophages with inhibitors of actin and dynamin polymerization, cytochalasin D and dynasore, respectively, prior to incubation with HKSA. Compared to untreated BMM, those pretreated with inhibitors produced similar amounts of IL-1β when incubated with HKSA from the WT or the ΔsagB + sagB complementation strain (Fig. 5C). BMMs incubated with ΔsagB mutant HKSA produced no IL-1β regardless of inhibitor pretreatment. In contrast, production of IL-6 was fully inhibited by pretreatment with the phagocytosis inhibitors, reinforcing the importance of these processes to the general inflammatory response and controlling for the activity of the inhibitors (Fig. 5D). Thus, SagB-dependent production of IL-1β by BMMs is likely not driven by differences in phagocytosis.

The priming stages of inflammasome activation are not SagB dependent.

To probe the underlying mechanism behind the reduced IL-1β production upon treatment of BMMs with a ΔsagB mutant, we interrogated the involvement of PRR signaling pathways previously implicated in peptidoglycan recognition. BMMs were differentiated from TLR2^−/−, NOD2^−/−, and MyD88^−/− knockout (KO) mice, and inflammasome activation was assessed in response to HKSA. We found no defect in IL-1β production by either TLR2^−/− or NOD2^−/− BMMs in response to WT S. aureus or the ΔsagB + sagB complementation strain (Fig. 6A to C). Consistent with its role as a major adapter protein required for TLR signaling, MyD88^−/− BMMs produced reduced amounts of IL-1β relative to the WT C57BL/6J BMMs, yet SagB dependence for IL-1β production remained (Fig. 6D). The integrity of TLR2^−/− and MyD88^−/− BMMs was validated using a known ligand of TLR2 (Pam3CSK4) (Fig. 6E). The synthetic ligand control for NOD2, MDP, was not included as it was previously demonstrated to inefficiently induce BMM proinflammatory cytokine production, including IL-1β (13, 14). Since the SagB-dependent production of IL-1β by BMMs does not appear to be driven directly by PRR engagement, we validated this point by measuring il-1b transcription, which should be increased upon engagement of PRRs. Indeed, treatment with WT and ΔsagB HKSA lead to identical increases in transcription of il-1b (Fig. 6F). Taken together, these data suggest that priming stages of inflammasome activation do not depend on SagB.

FIG 6.

SagB-dependent IL-1β production is independent of canonical peptidoglycan PRR recognition. BMMs from WT C57BL6/J (A), TLR2^−/− (B), NOD2^−/− (C), or MyD88^−/− (D) mice were incubated with HKSA (MOI of 100) of the indicated strains overnight, and IL-1β in supernatants was quantified by CBA. (E) TLR2 agonist Pam3CSK4 (100 ng/mL) was added to WT and KO BMMs, and IL-6 in supernatants was quantified by CBA. (F) BMMs were treated with HKSA (MOI of 100) of the indicated strains for 3 h in triplicate, and RNA was isolated for cDNA synthesis and qPCR. Fold change expression of il-1b was calculated relative to unstimulated macrophages. The data shown are from one of at least three experiments performed in triplicate. Statistical significance was determined by one-way ANOVA with Tukey’s post hoc test. ***, P < 0.001; ****, P < 0.0001.

SagB-dependent IL-1β production is independent of NLRP3 and caspase-1.

The NLRP3 inflammasome is believed to drive IL-1β production in response to S. aureus peptidoglycan degradation products, yet the mechanisms underlying this activation remain to be fully characterized (13, 14, 29). To determine if the NLRP3 inflammasome is responsible for SagB-dependent IL-1β production, we collected BMMs from NLRP3^−/− and CASP1^−/− KO mice, as well as ASC^−/− KO mice, which are deficient in an adapter molecule for multiple inflammasomes. Upon treatment with HKSA, each of these KO cells produced equivalent amounts of IL-1β (Fig. 7A to C). However, sequential treatment of BMMs with known inducers of the NLRP3 inflammasome, lipopolysaccharide (LPS) and nigericin, failed to elicit IL-1β production (Fig. 7D). Remarkably, these data suggest that SagB-dependent IL-1β production is independent of the NLRP3 inflammasome and caspase-1.

FIG 7.

IL-1β production is independent of NLRP3 and caspase-1. BMMs from NLRP3^−/− (A), ASC^−/− (B), or CASP1^−/− (C) mice were incubated with HKSA (MOI of 100) of the indicated strains overnight, and IL-1β in supernatants was quantified by CBA. (D) BMMs were sequentially incubated with LPS (100 ng/mL) and nigericin (10 μg/mL), and IL-1β in supernatants was quantified by CBA. (E and F) Pan-caspase inhibitors Q-VD-OPh (10 μM) and Z-VAD-FMK (100 μM) were added to BMMs 1 h prior to addition of HKSA (MOI of 100) (E) or control LPS-plus-nigericin treatments (F). Supernatants from treated macrophages were collected, and IL-1β was measured via CBA. The data shown are from one of at least three experiments performed in triplicate. Statistical significance was determined by one-way ANOVA with Tukey’s post hoc test. **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

While caspase-1 is the most common protease involved in maturation of IL-1β, other caspases such as caspase-8 have been described to have processing capability (39). To determine if SagB-dependent IL-1β production requires the activity of an alternative caspase, we pretreated macrophages with two different pan-caspase inhibitors, Q-VD-OPh and Z-VAD-FMK, followed by incubation with HKSA. Compared to macrophages without inhibitor treatment, those treated with either inhibitor had only slightly reduced IL-1β levels after incubation with WT HKSA (Fig. 7E), while responses to LPS plus nigericin were completely blocked (Fig. 7F). In all instances, the ΔsagB mutant HKSA stimulated no IL-1β production. Taken together, these data suggest that the IL-1β produced in response to HKSA does not require NLRP3 and caspase-1 and may not require other caspases.

SagB activity drives IL-1β maturation and release.

While our data argue against a role for NLRP3 and caspase in SagB-dependent IL-1β production, they do not rule out the possibility that SagB might drive some previously undefined aspect of IL-1β maturation. Thus, we considered the possibility that there may be a deficiency in processing of pro-IL-1β to mature IL-1β that occurs via unconventional means. To test this possibility, we collected macrophage lysates and supernatants following stimulation with HKSA from the WT or a ΔsagB mutant and performed Western blots for IL-1β to compare the amounts of pro-IL-1β in the cell lysate to mature IL-1β in the supernatant. We observed an accumulation of pro-IL-1β in the cell lysates of macrophages stimulated with HKSA from the ΔsagB mutant (Fig. 8A) and a concomitant decrease in mature IL-1β in the supernatant (Fig. 8B and C) compared to those stimulated with WT S. aureus. Consistent with prior evidence suggesting IL-1β production is independent of NLRP3 and caspase-1 (Fig. 7), we found significant decreases in mature IL-1β in the supernatant of NLRP3^−/− and CASP1^−/− macrophages stimulated with HKSA from a ΔsagB mutant but not WT S. aureus (Fig. 8D). Thus, macrophages treated with the ΔsagB mutant are defective for maturation and release of mature IL-1β.

FIG 8.

SagB promotes noncanonical cleavage of pro-IL-1β to mature IL-1β. (A and B) BMMs were stimulated with HKSA (MOI of 100), and lysates and supernatants were collected. (A) Lysate preparations were immunoblotted for pro-IL-1β and β-actin. (B) IL-1β was precipitated from supernatants, and mature IL-1β was detected by immunoblotting. (C) IL-1β was quantified via CBA from supernatant prior to immunoprecipitation. (D) IL-1β was precipitated from supernatants of the indicated macrophages, and mature IL-1β was detected by immunoblotting. The data shown are from one of at least three experiments performed in triplicate. Statistical significance was determined by one-way ANOVA with Tukey’s post hoc test. ****, P < 0.0001.

DISCUSSION

In this study, we explored how processing of peptidoglycan by S. aureus impacts host innate immune responses. We found that the staphylococcal glucosaminidase SagB is critical for production of IL-1β by macrophages and dendritic cells in response to heat-killed and live S. aureus. Furthermore, we demonstrated that purified SagB-processed peptidoglycan is sufficient for stimulation of IL-1β production by macrophages, whereas unprocessed elongated glycan chains from a ΔsagB mutant are not. To our surprise, this SagB-dependent IL-1β response is not driven by PRR signaling, activation of the NLRP3 inflammasome, or caspase-1. Instead, SagB-dependent processing of peptidoglycan triggers an unconventional pro-IL-1β cleavage event that leads to release of mature IL-1β.

Peptidoglycan recognition occurs through myriad mechanisms, many of which are still being identified and have yet to be fully characterized. During infection, the innate immune system encounters peptidoglycan in a variety of different forms. These include soluble or insoluble peptidoglycan fragments released during cell division or phagosomal degradation as well as intact sacculi from live or dead bacterial cells. Here, we found that SagB processing of staphylococcal peptidoglycan is important for immune recognition in multiple scenarios. IL-1β production by macrophages was compromised in the absence of SagB for both live and heat-killed cells as well as purified soluble and insoluble peptidoglycan (Fig. 1 and 4). However, in contrast with prior reports, the mechanism by which SagB affects peptidoglycan recognition does not seem to involve canonical peptidoglycan PRRs. Previous studies characterizing how peptidoglycan is recognized have largely focused on TLR2 signaling, but there is still uncertainty as to how TLR2 recognizes peptidoglycan. In addition, debate over the source and purity of peptidoglycan used in prior studies has complicated interpretations (19, 40). Despite these discrepancies, we found that SagB-dependent IL-1β production by macrophages in response to HKSA does not require TLR2 (Fig. 6B) (19, 21, 31, 40). However, MyD88^−/− macrophages produced less IL-1β than WT macrophages when stimulated with HKSA (Fig. 6D), suggesting that there may be some input from an alternative PRR upstream of MyD88 signaling. Others have observed a similar trend in an S. aureus subcutaneous infection model where TLR2-deficient mice exhibited no defect in IL-1β production compared to WT mice, whereas MyD88-deficient mice produced less IL-1β (12). MyD88- and IL-1R-deficient mice also had significantly larger lesions than TLR2-deficient mice, reinforcing our data and suggesting that TLR2 signaling is not necessarily required for IL-1β production in response to S. aureus (12). Absence of NOD2 also had no effect on SagB-dependent IL-1β production (Fig. 6C), which is corroborated by other reports suggesting that this receptor is not required (13). Although NOD2 and TLR2 are the most well-characterized receptors for peptidoglycan, recent studies have identified unconventional modes of recognition. For example, the glycolytic enzyme hexokinase is an unconventional PRR that binds to NAG, leading to activation of the NLRP3 inflammasome (14). These results suggest there are likely other noncanonical peptidoglycan recognition pathways which may also be at play in this study.

Previous studies indicate that phagocytosis and lysozyme-based digestion of peptidoglycan are coupled to NLRP3 inflammasome activation and IL-1β production (13, 16, 36, 37). However, inhibiting phagocytosis and rendering S. aureus more susceptible to lysozyme degradation by ablating O-acetylation did not dramatically affect IL-1β production (Fig. 5). Additionally, SagB-processed peptidoglycan was found to induce IL-1β maturation in an inflammasome-independent manner (Fig. 7). Indeed, treatment of NLRP3^−/−, CASP1^−/−, and ASC^−/− BMMs with HKSA (Fig. 7A to C) as well as treatment with pan-caspase inhibitors (Fig. 7E) did not impact IL-1β production. Macrophages derived from CASP1^−/− mice are also deficient in caspase-11, suggesting caspase-11 is also not involved in SagB-dependent IL-1β responses. In contrast, NLRP3 inflammasome-deficient macrophages (Fig. 7D) and macrophages treated with pan-caspase inhibitors (Fig. 7F) were unable to produce IL-1β when stimulated with the classical inflammasome activators LPS and nigericin, highlighting the divergence in recognition mechanisms involved. Inflammasome-independent IL-1β production has been observed in mice during Mycobacterium tuberculosis and Bordetella pertussis infection (33, 34). However, the protease or proteases involved in caspase-independent IL-1β cleavage are not known. There is evidence to suggest that macrophage or neutrophil-derived serine proteases, such as cathepsins (41, 42), neutrophil elastase (43), granzyme A (44, 45), and proteinase-3 (46), have IL-1β processing capability, but the circumstances where each of these enzymes are used is unclear. Other reports have also shown that bacterium-derived proteases from Staphylococcus epidermidis and other pathogens have IL-1β processing capabilities (35, 47, 48). It is possible that a staphylococcal protease may contribute to alternative IL-1β cleavage in a SagB-dependent manner during live infection in vitro (Fig. 2C) or in vivo (Fig. 3); however, our data with HKSA (Fig. 1 and 2) and purified peptidoglycan (Fig. 4) suggest that any alternative protease is likely macrophage derived. Additionally, it is possible that performing additional infection studies with KO BMMs using live, pore-forming toxin-producing S. aureus may lead to different outcomes than what we observed with HKSA. However, the fact that we saw SagB-dependent IL-1β production during live infection of WT BMMs at early time points (Fig. 2C) supports the hypothesis that SagB-processed peptidoglycan is, at minimum, relevant during early recognition of S. aureus.

In agreement with previous work, we found that a ΔsagB mutant is attenuated for colonization of the kidneys (9) (Fig. 3D) as well as the heart (Fig. 3E) during murine systemic infection. There were significant differences in CFU at 96 h postinfection and modest reductions in the amount of IL-1β found in kidney homogenates at this time point (Fig. 3F). Thus, it remains to be determined if the decreased CFU found in these distal sites is due to an increased susceptibility of the ΔsagB mutant to immune-mediated clearance, decreased fitness in vivo, or both. Surveying earlier time points in infection will address potential differences in infection kinetics for the ΔsagB mutant and possibly reveal additional changes in IL-1β as it is known that this cytokine plays an important protective role early in S. aureus infections (43, 49, 50). Nevertheless, in an SSTI model we found that a ΔsagB mutant colonizes the skin similarly to the WT (Fig. 3A) but elicits less IL-1β after 120 h (Fig. 3B), which coincides with reduced abscess size (Fig. 3C). Thus, although SagB may not be required for infection of the skin, it is conceivable that the difference in elicited IL-1β reduces pathogenic inflammation in the skin (11, 50, 51). We will explore these possibilities in future experiments to better understand the role that SagB plays in the innate immune response in vivo.

While our results indicate that processing of S. aureus peptidoglycan by SagB is a prerequisite for IL-1β production by macrophages, the optimal and minimum glycan chain lengths required for this response are currently unknown. Minimally, NAG produced by phagosomal degradation can trigger IL-1β production through hexokinase, which is a PRR-independent but NLRP3-dependent pathway (14). In contrast, our data suggest that shorter glycan chains produced by SagB activate an alternative NLRP3-independent pathway of IL-1β production. We and others have noted that extended incubation of purified peptidoglycan with SagB does not affect the resulting muropeptide profile suggesting that SagB generates glycan chains within a range of specific lengths as opposed to the smallest possible disaccharide units (8). Prior studies showed that interaction with SpdC helps facilitate SagB cleavage of glycan chains to physiological lengths, although SagB can function without SpdC (5, 7). On average, the glycan chain length dictated by SagB has been shown to be between 3 and 10 disaccharides (3, 4). Thus, we reason that the optimal glycan chain length that stimulates IL-1β is within this range, although it is possible that slight variations in length or stem peptide composition could affect recognition. Macrophages have been shown to respond differently to synthetic mono- or disaccharide peptidoglycan fragments with various stem peptide lengths, where more robust activation was observed in response to fragments with longer peptides (38). Little is known about how these peptidoglycan characteristics modulate innate immune recognition. Thus, interrogating the immunostimulatory capacity of the different SagB-generated peptidoglycan fragments will be critical for a more complete understanding of the IL-1β response to S. aureus.

Taken together, this work substantiates peptidoglycan chain length as an important variable that contributes to innate immune responses to S. aureus. The findings have broad implications for our understanding of innate immune recognition of staphylococci—and potentially other pathogens. There is a high degree of variability in structure, average glycan chain length, and stem peptide composition between Gram-positive pathogens, representing a valuable topic of future study on the nuances in innate immune responses to high-priority pathogens.

MATERIALS AND METHODS

Bacterial strains and culture conditions.

All bacterial strains used in this study are listed in Table 1. E. coli strains were cultured in Miller’s lysogeny broth (LB) or on agar plates (BD Biosciences) with antibiotics added as needed at the following concentrations: 100 μg/mL ampicillin and 10 μg/mL chloramphenicol. S. aureus strains were grown in tryptic soy broth (TSB) or on agar plates (BD Biosciences) supplemented with antibiotics as necessary at the following concentrations: 10 μg/mL chloramphenicol, 0.1 mM cadmium chloride, 3.5 μg/mL erythromycin, and 1 μg/mL anhydrous tetracycline. All bacterial strains were grown at 37°C shaking at 200 rpm unless noted otherwise.

TABLE 1.

Bacterial strains used in this study

Strain	Description	Annotation	Source
E. coli
DH5α	DH5α	DH5α	New England Biolabs
lysY/Iq strain	T7 expressing lysY/Iq	lysY/Iq	New England Biolabs
IM08B	SA08BΩPN25-hsdS (CC8-1) (SAUSA300_0406) of NRS384 integrated between essQ and cspB genes	E. coli IM08B	52
FA-E2375	E. coli strain constitutively expressing glucosaminidase domain of Atl	lysY/Iq pET15b-AtlGL	This study
FA-E2373	E. coli strain constitutively expressing SagB lacking single transmembrane domain	lysY/Iq pET15b-SagB	This study
S. aureus
USA300 LAC	S. aureus USA300 strain LAC (AH-1264); plasmid cured	LAC (WT)	59
RN4220	Restriction-deficient S. aureus strain for plasmid passage	RN4220	60
JE2	Parental strain of Nebraska Transposon Mutant Library	JE2	61
FA-S2297	LAC with in-frame deletion of sagB	ΔsagB	This study
FA-S2314	FA-S2297 complemented with pJC1111-sagB	ΔsagB + sagB	This study
FA-S2398	LAC with in-frame deletion of spdC	ΔspdC	This study
FA-S2408	LAC with in-frame deletion of sagB and spdC	ΔsagB ΔspdC	This study
FA-S2348	LAC with in-frame deletion of oatA	ΔoatA	This study
FA-S2375	LAC with in-frame deletions of sagB and oatA	ΔsagB ΔoatA	This study
NE460	JE2 atl::erm transposon insertion mutant	atl::erm	61
NE1190	JE2 sagA::erm transposon insertion mutant	sagA::erm	61
NE1909	JE2 sagB::erm transposon insertion mutant	sagB::erm	61
NE1353	JE2 scaH::erm transposon insertion mutant	scaH::erm	61

Generation of S. aureus mutant strains.

Oligonucleotide pairs were designed to amplify ~500 nucleotides upstream and downstream of the start and stop codons of each target gene for deletion using S. aureus genomic DNA as the template (Table 2). Upstream and downstream amplicons were purified and used as the templates for splicing by overlap extension (SOE) PCR to yield the final amplicon of ~1,000 nucleotides. Each amplicon was subcloned into pIMAY using restriction endonucleases KpnI and SacI. Ligated plasmids were transformed into E. coli IM08B, and transformants were verified via PCR (52). Plasmids were purified from IM08B strains and were subsequently transformed into S. aureus electrocompetent cells. Mutagenesis was performed as previously described (53, 54).

TABLE 2.

List of oligonucleotides used in this study

Name	Sequence (5′→3′)
pET15b-SagB_FWD	AAAACATATGTCCGATCAGATATTTTTCAAACA
pET15b-SagB_REV	AAAAGGATCCTTACTTATTCAAATGTTTACTGTCAT
pET15b-AtlGL_FWD	AAAGGATCCGGGTTTACAATATAAACCACAAGTACAACGTG
pET15b-AtlGL_REV	AAAAGGATCCTTATTTATATTGTGGGATGT
d1720SOE1	ATATGGTACCCAATGGATTACGCACATTTA
d1720SOE2	AACAACTCGTAGCTTATCAAAATCCACACCTCTTAGGTCATT
d1720SOE3	AATGACCTAAGAGGTGTGGATTTTGATAAGCTACGAGTTGTT
d1720SOE4	ATATGAGCTCATCAGATTTAATCGATAATAA
UniCompSOE1	ATATTCTAGAATCCCATTATGCTTTGGCA
1720compSOE2	TTTCTTGTGTTTATTCATGGGTTTCACTCTCCTTCT
1720compSOE3	AGAAGGAGAGTGAAACCCATGAATAAACACAAGAAA
1720compSOE4	TATAGAGCTCAGTAATTCTTTATCGATGTCCTGC
Del_oatA_F1	ATATGGTACCTCGATGCAATAATAATTGATATGACAATCT
Del_oatA_R1	GTAATATTTCAAAAGTTTAGTGCATCAAAGTTAATAAACGCCCCATTTATTTTTCTCTA
Del_oatA_F2	TAGAGAAAAATAAATGGGGCGTTTATTAACTTTGATGCACTAAACTTTTGAAATATTAC
Del_oatA_R2	ATATGAGCTCAAGTGTAATGATAATTACCTTGTGCATC
Del_spdC_SOE1	ATATGGTACCATCGCATCTGGTCCTTCTTTT
Del_spdC_SOE2	GTACTAGCAAGCGCTTTGTTATATATGTAACCTCCATTAGGT
Del_spdC_SOE3	ACCTAATGGAGGTTACATATATAACAAAGCGCTTGCTAGTAC
Del_spdC_SOE4	ATATGAGCTCATGGTGCTGATGAAGTAGATC
Ms-IL-1b-F	ATCAACCAACAAGTGATATTCTCCAT
Ms-IL-1b-R	GGGTGTGCCGTCTTTCATTAC
Ms-b-actin-F	GTGACGTTGACATCCGTAAAGA
Ms-b-actin-R	GCCGGACTCATCGTACTCC

Bacterial growth curves.

Cells of the WT, ΔsagB, ΔoatA, and ΔsagB ΔoatA strains were grown overnight at 37°C with shaking in 200 μL TSB in triplicate wells of a flat-bottom 96-well plate (Fisher). The next day, strains were subcultured 1:100 into 198 μL TSB in a new flat-bottom 96-well plate and were grown at 37°C with shaking in a SpectraMax iD3 microplate reader (Molecular Devices) with the optical density at 600 nm (OD₆₀₀) measured every 15 min for 24 h. For lysozyme sensitivity assays, 1.5 mg/mL lysozyme (Sigma) was added at mid-exponential phase, followed by measurement of the OD₆₀₀ every 15 min for 24 h.

Cell culture conditions.

L929 mouse fibroblasts were cultured at 37°C in 5% CO₂ in minimal essential medium (MEM) (Corning) plus 10% heat-inactivated fetal bovine serum (FBS) (VWR) plus 1 mM sodium pyruvate (Corning) plus 1 mM HEPES buffer (Corning) plus 2 mM l-glutamine (Corning) plus 100 μg/mL penicillin/streptomycin (Pen/Strep) (Corning). Murine bone marrow-derived macrophages (BMMs) were differentiated from bone marrow progenitor cells of four- to 6-week old female or male C57BL/6J (WT, TLR2^−/−, NOD2^−/−, MyD88^−/−, NLRP3^−/−, ASC^−/−) and C57BL/6NJ (CASP1^−/−) mice and cultured at 37°C in 5% CO₂ in bone marrow macrophage (BMM) medium (Dulbecco’s modified Eagle’s medium [DMEM]) (Corning) plus 1 mM sodium pyruvate plus 1 mM HEPES buffer plus 2 mM l-glutamine plus 20% heat-inactivated FBS plus 30% L929 fibroblast supernatant plus 100 μg/mL Pen/Strep plus 50 μM β-mercaptoethanol (Amresco). Murine bone marrow-derived dendritic cells were differentiated from bone marrow progenitor cells of 4- to 6-week-old female C57BL/6J mice and cultured at 37°C in 5% CO₂ in complete RPMI (RPMI 1640) (Corning) plus 25 mM HEPES buffer plus 10% heat-inactivated FBS plus 50 μM β-mercaptoethanol plus 100 μg/mL Pen/Strep.

Isolation of bone marrow-derived macrophages.

Primary bone-marrow macrophages (BMMs) were derived from bone marrow progenitor cells isolated from femurs and tibias of 4- to 6-week-old C57BL/6J WT (Jax no. 000664), TLR2^−/− (Jax no. 004650), NOD2^−/− (Jax no. 005763), MyD88^−/− (Jax no. 009088), NLRP3^−/− (Jax no. 021302), ASC^−/−, and CASP1^−/− (also deficient in caspase-11) (Jax no. 016621) mice. Progenitor cells were seeded at 5 × 10⁶ cells/plate into 100- by 26-mm petri dishes containing 15 mL BMM medium with 100 μg/mL Pen/Strep. Following 3 days of incubation at 37°C, 5% CO₂, 10 mL of fresh BMM medium was added and the macrophages were allowed to differentiate for another 6 days. Differentiated macrophages were removed from plates by being washed with 1× PBS and incubated with 10 mL 1× PBS at 4°C for 30 min. Cells were removed from the bottom of the plate by manual pipetting and centrifuged at 1,500 rpm. BMMs were resuspended in BMM medium containing 10% dimethyl sulfoxide (DMSO) at 1 × 10⁷ cells/mL and stored in liquid nitrogen until use. To thaw macrophages for use, the frozen vial was incubated at 37°C until thawed and added dropwise to 9 mL fresh BMM medium, pelleted by centrifugation at 1,500 rpm at 4°C for 5 min, and resuspended in 10 mL BMM medium. The cell suspension was split into two 100- by 26-mm petri dishes with 10 mL BMM medium containing 100 μg/mL Pen/Strep and cultured for 2 to 3 days at 37°C in 5% CO₂ before use.

Isolation of bone marrow-derived dendritic cells.

Primary bone marrow-derived dendritic cells (BMDCs) were isolated following a previously published protocol (55). BMDCs were derived from bone marrow progenitor cells isolated from femurs and tibias of four- to 6-week-old C57BL/6J mice. Progenitor cells were collected by flushing the bones with complete RPMI with a 27G needle followed by centrifugation at 1,500 rpm at 4°C for 5 min. After discarding the supernatant, red blood cells were lysed with 1 mL ACK lysing buffer (Lonza) for 30 s and immediately diluted in 50 mL of ice-cold 1× PBS, followed by centrifugation at 1,500 rpm at 4°C. The cells were washed once more with 50 mL 1× PBS and resuspended in 10 mL of complete RPMI for cell counting. A total of 2 × 10⁶ progenitor cells were plated in 100- by 26-mm petri dishes with 20 mL complete RPMI supplemented with 20 ng/mL granulocyte-macrophage colony-stimulating factor (GM-CSF) (BioLegend) and incubated at 37°C in 5% CO₂. After 3 days, an additional 20 mL of complete RPMI with 20 ng/mL GM-CSF was added. On day 6, 20 mL was removed from each plate and centrifuged at 1,500 rpm, and pelleted cells were resuspended in 20 mL complete RPMI with GM-CSF and IL-4 (BioLegend) added at a final concentration of 10 ng/mL for each cytokine. At day 8, this step was repeated, but with 5 ng/mL GM-CSF and 10 ng/mL IL-4. Differentiated dendritic cells were isolated after 10 days by centrifuging the medium at 1,500 rpm to collect the floating cells, followed by gently rinsing each plate with warmed 1× PBS to obtain the lightly adherent cells. BMDCs were counted and resuspended at 5 × 10⁶ cells/mL in complete RPMI containing 10% DMSO for storage in liquid nitrogen until use. To thaw BMDCs for use, the frozen vial was incubated at 37°C until thawed and added dropwise to 9 mL fresh complete RPMI, pelleted by centrifugation at 1500 rpm at 4°C for 5 min, and resuspended in 10 mL complete RPMI. The cell suspension was then split into two 100- by 26-mm petri dishes with 10 mL complete RPMI containing 100 μg/mL Pen/Strep and cultured for 2 to 3 days at 37°C in 5% CO₂ before use.

BMM and BMDC activation assay.

BMMs and BMDCs were seeded onto treated polystyrene flat bottom 96-well plates (Corning) at 100,000 cells/well in 90 μL of BMM medium or complete RPMI, respectively, and were allowed to rest for at least 4 h before treatments. For the heat-killed S. aureus (HKSA) preparation, 5-mL overnight cultures of each strain were grown in TSB at 37°C shaking at 200 rpm. The next day, 1:100 subcultures were inoculated in 5 mL TSB and grown for 9 h at 37°C with shaking at 200 rpm. After growth, strains were pelleted by centrifugation at 3,900 rpm for 10 min. Bacterial pellets were washed three times with 10 mL 1× PBS and normalized to an OD₆₀₀ of 1.0 (1 × 10⁹ cells/mL). OD-normalized strains were boiled on a heat block for 30 min at 100°C and allowed to cool on ice. For treatments of BMMs and BMDCs, 10 μL of HKSA (MOI of 100) of each strain was added to triplicate wells and incubated at 37°C in 5% CO₂ overnight unless noted otherwise. In phagocytosis inhibition experiments, 6 μM cytochalasin D (Sigma) or 40 μM dynasore (Sigma) was added to wells 1 h prior to treatment with HKSA. In experiments interrogating the involvement of alternative caspases, BMMs were pretreated for 30 min with pan-caspase inhibitor Q-VD-OPh (10 μM) (Cayman Chemical) or Z-VAD-FMK (100 μM) (Invivogen) before treatment with HKSA. As a control for inflammasome activation, WT, NLRP3^−/−, and Casp1^−/− BMMs were treated with 100 ng/mL LPS (Sigma) for 3 h, followed by 10 μg/mL nigericin (Sigma) for 30 min. Pattern recognition receptor knockout macrophages were verified by performing control treatments of TLR2^−/− and MyD88^−/− mice with Pam3CSK4 (100 ng/mL) (Invivogen). Following BMM or BMDC activation, supernatants were collected and used immediately in the cytometric bead array (CBA) or stored at −80°C until use.

To quantify the cytokines and chemokines produced from activated BMMs or BMDCs, a custom CBA flex set (BD Biosciences) was used following the manufacturers specifications. Supernatants were incubated with beads in a V-bottom 96-well plate (Corning) for 1.5 h at 25°C at 600 rpm on a Thermo Mixer C (Eppendorf). Samples were then washed with CBA wash buffer, and data were collected on an LSR Fortessa (BD Biosciences) and analyzed using FCAP Array software v.3.0 (BD Biosciences). Data were graphed using GraphPad Prism.

Viability assay.

BMM viability was determined using an ethidium bromide (EtBr) staining assay. BMMs were washed with 1× PBS, and EtBr (50 μg/mL) was added to each well. BMMs were incubated at 37°C in 5% CO₂ for 30 min, and fluorescence was measured using a Molecular Devices SpectraMax iD3 plate reader (excitation, 530 nm; emission, 590 nm). Percentage of survival was calculated by normalizing fluorescence relative to unstimulated BMMs.

BMM infection.

BMMs were seeded in 96-well plates in BMM medium as described above. S. aureus strains were prepared and OD₆₀₀ normalized similarly to HKSA as described above, without subsequent boiling. Live S. aureus cells from WT, ΔsagB mutant, or ΔsagB + sagB complementation strains were added to BMMs (MOI of 5) followed by centrifugation at 1,500 RPM for 5 min to synchronize the infection and subsequently incubated for 4 h at 37°C in 5% CO₂. After incubation, the medium was removed, and BMMs were washed three times with 1× PBS, followed by the addition of fresh medium supplemented with gentamicin (100 μg/mL) and incubation for 30 min at 37°C in 5% CO₂. BMMs were washed twice with 1× PBS, and the medium was replaced with fresh BMM medium supplemented with gentamicin (4 μg/mL) and lysostaphin (0.8 μg/mL), followed by incubation at 37°C in 5% CO₂ for an additional 1 h, at which point the supernatant was collected for quantification of IL-1β as described above.

Purification of recombinant SagB and Atl.

The expression vectors used to isolate N-terminally His₆-tagged SagB and the glucosaminidase domain of Atl were constructed in pET15b as previously described (8). Residues 91 to 855 of SagB were amplified via PCR to exclude the single transmembrane domain. Constructs were verified by DNA sequencing.

The E. coli lysY/I^q strains harboring pET15b-SagB_91-855 or pET15b-Atl were cultured in 1 L LB to an OD₆₀₀ of 0.2 to 0.3 at 37°C with shaking at 200 rpm, and protein expression was induced by adding 0.5 mM IPTG (isopropyl-β-d-thiogalactopyranoside) (GoldBio). Cultures were grown for 3 h after addition of IPTG and then pelleted by centrifugation at 8,000 rpm for 10 min at 4°C. Pellets were resuspended in 1× PBS plus 20 mM imidazole (Fisher), and cells were lysed at a constant rate of 10 s per pulse with an output of 340 W in 1-min intervals for 10 min total on ice using a Branson SFX500 digital sonifier. Cellular debris was removed from lysates by centrifugation at 12,000 rpm at 4°C for 30 min, followed by filtration with a 0.45-μm-pore syringe filter (Celltreat). Clarified lysates were incubated with 1 mL PBS-equilibrated nickel-nitrilotriacetic acid (NTA) resin slurry (Qiagen) on a rotisserie for 1 h at 4°C. Lysates with resin were poured into a glass chromatography column, and the resin was washed with 20 column volumes of 1× PBS plus 20 mM imidazole, followed by elution with 10 mL 1× PBS plus 500 mM imidazole. Eluates were subsequently incubated with Pierce high-capacity endotoxin removal resin overnight at 4°C to remove any residual LPS (Thermo Scientific). Samples were dialyzed using 10-kDa molecular-weight-cutoff regenerated cellulose dialysis tubing (Spectrum). Dialysis was performed stepwise by incubating protein at 4°C in 1× PBS containing 250 mM imidazole for 4 h, 100 mM imidazole for 4 h, and then 1× PBS with no imidazole overnight. Protein was stored at −80°C in 1× PBS plus 10% glycerol.

Peptidoglycan purification from S. aureus.

WT and ΔsagB strains were subcultured 1:100 in 1 L TSB and grown at 37°C shaking at 200 rpm until cultures reached an OD₆₀₀ of 0.6. Cells were pelleted by centrifugation at 8,000 rpm, and supernatant was discarded. Pellets were resuspended in 280 mL sterile water, followed by the addition of 70 mL 20% SDS. Cell suspensions were boiled for 30 min and allowed to cool to room temperature before being pelleted by centrifugation at 8,000 rpm for 10 min. Cells were washed at least five times with 350 mL water to remove SDS, followed by being resuspended in 40 mL of water and distributed into 15-mL conical tubes containing 3 mL of 0.1-mm glass beads (Electron Microscopy Sciences). Cells were lysed using an MP FastPrep-24 homogenizer in 10 pulses of 1 min at a speed of 6 m/s with a 5-min incubation on ice between pulses. After the final pulse, beads were allowed to sediment by gravity and lysates were transferred to a clean 50-mL conical tube. Beads were pulsed once more with 1 mL water to collect the remaining cellular debris, which was subsequently added to the collected lysates. Lysates were centrifuged at 10,000 rpm for 10 min and washed once with 20 mL water. Pellets were then resuspended in 25 mL 50 mM Tris-HCl plus 10 mM CaCl₂ plus 20 mM MgCl₂ (pH 7.5), 100 μg/mL amylase (Sigma), 10 μg/mL DNase I (Sigma), and 50 μg/mL RNase A (VWR) were added, and the mixture was incubated on a rotisserie at 37°C for 4 h. After treatment, 1 mg/mL trypsin was added, followed by incubation at 37°C for an additional 16 h. Insoluble cellular material was collected by centrifugation at 10,000 rpm for 10 min and washed once with water. Pellets were resuspended in 28.5 mL water and 1.5 mL 20% SDS and boiled for 15 min to inactivate enzymes, followed by being washed at least five times with 25 mL water to remove SDS. Washes were then performed in the following order: one wash with 10 mL 8 M LiCl, one with 25 mL 100 mM EDTA, two washes with 25 mL water, one wash with 25 mL acetone, and two final washes with 25 mL water. Cellular material was resuspended in 10 mL 49% hydrofluoric acid (Sigma) and incubated at 4°C for 48 h. Peptidoglycan (PGN) was recovered by centrifugation at 12,000 rpm for 45 min at 4°C and washed twice with 20 mL water, once with 20 mL 100 mM Tris-HCl (pH 7.5), and then twice more with 20 mL water. The PGN pellet was then resuspended in 20 mL 100 mM (NH₄)₂CO₃, 250 μg/mL alkaline phosphatase was added, and the mixture was incubated at 37°C for 16 h on a rotisserie. The next day, the PGN was boiled for 5 min to inactive the enzyme and then washed three times with water. Purified PGN was stored in sterile water at 4°C until use.

Digestion of purified peptidoglycan.

For HPLC experiments, purified PGN from the WT or ΔsagB mutant (50 mg) was resuspended in 500 μL 100 mM sodium phosphate buffer (pH 5.0) and sequentially incubated with 0.5 mg/mL SagB or Atl and/or 0.1 mg/mL lysostaphin at 37°C for 24 h per enzyme with shaking at 650 rpm on an Eppendorf Thermomixer C. For macrophage stimulation experiments, 10 mg of peptidoglycan was resuspended in 100 μL buffer and digested as described above. Digests were boiled for 10 min to inactivate enzymes and then centrifuged at 13,000 rpm to collect solubilized PGN. Pellets were weighed to calculate the amount solubilized. PGN was also quantified using a modified phenol sulfuric acid assay (56–58). Briefly, 50 μL of solubilized PGN was mixed with 150 μL sulfuric acid in a flat-bottom polystyrene 96-well plate (Fisher) at 68°C for 1.5 h. After incubation, the plate was centrifuged briefly at 1,500 rpm, and absorbance was read at 315 nm on a Molecular Devices SpectraMax iD3 plate reader. Concentrations of soluble PGN (sPGN) samples were calculated by comparing the absorbances to those of xanthan standards. Prior to addition of sPGN (~6 to 8.5 μg/mL) to macrophages to assess activation of cytokine production, the pH was adjusted from 5 to 7 using 1 N NaOH, and concentrations were normalized based on the quantification described above.

High-performance liquid chromatography of digested peptidoglycan.

Glycan chains were reduced by adding 200 μL of 10 mg/mL sodium borohydride in borate buffer to 200 μL sPGN in a 15-mL conical tube. Tubes were incubated with the caps off at room temperature in a fume hood for 15 min, at which point the reaction was quenched by the addition of 10 μL phosphoric acid, bringing the sample to a pH of 2. Reduced glycan chains were then separated by reverse-phase high-performance liquid chromatography (RP-HPLC) using an Agilent Technologies 1260 Infinity system. Fifty microliters of samples was injected onto a 150- by 4.60-mm C₁₈ column (Kinetex 5 μm EVO C₁₈; Phenomenex) and eluted at a flow rate of 0.5 mL/min with an isocratic hold for 5 min of 5% (vol/vol) methanol, followed by a linear gradient to 30% in 100 mM sodium phosphate buffer (adjusted to pH 2.5 with phosphoric acid) over 150 min. Eluted glycan chains were detected by absorption at 206 nm.

Western blots.

BMMs were seeded at 100,000 cells/well in Opti-MEM (Gibco) and stimulated with HKSA (MOI of 100) as previously described. Cell lysates and supernatants were collected from BMMs and pooled after overnight stimulation. For lysates, cells were washed twice with 1× PBS and then incubated with 100 μL lysis buffer (20 mM HEPES [pH 7.5], 150 mM NaCl, 1 mM EDTA, 10% glycerol, 1% Triton X-100) plus 1× Complete mini-protease inhibitors (Roche) on ice for 10 min. Cell debris was pelleted by centrifugation, and lysate was mixed with 6× loading dye and boiled for 10 min. For supernatants, IL-1β was immunoprecipitated using protein A agarose beads (Thermo Scientific) loaded with 5 μg anti-IL-1β (Abcam ab97220). BMM supernatants were incubated with 50 μL loaded beads at 4°C for 6 h on a rotisserie. Beads were then collected and washed three times with 1× PBS. After the final wash, beads were resuspended in 100 μL 1× PBS, mixed with 6× loading dye, and boiled for 10 min. Lysates and supernatant samples were resolved in a NuPAGE 4 to 12% Bis-Tris gel at 150 V and transferred to 0.2-μm-pore polyvinylidene difluoride (PVDF) membrane in CAPS (N-cyclohexyl-3-aminopropanesulfonic acid) buffer (pH 11). For lysate samples, Western blots were probed with 1:1,000 anti-IL-1β (R&D; AF-401-NA) in StartingBlock Tris-buffered saline (TBS) buffer (Thermo Scientific) overnight at 4°C, washed 3× with TBST (0.1% Tween 20 [Amresco] in TBS), and then incubated with 1:50,000 horseradish peroxidase (HRP)-conjugated rabbit anti-goat IgG (H+L) (Invitrogen) for 1 h, followed by 3× TBST washes. As a loading control, blots were probed for β-actin using 1:50,000 anti-β-actin (R&D; MAB8929) and 1:50,000 HRP-conjugated goat-anti-mouse IgG (H+L) (Invitrogen) for 1 h in TBST. For supernatant samples, blots were probed similarly as described above but with 1:1000 anti-IL-1β (Abcam) and 1:5,000 Clean-Blot IP detection reagent (Thermo Scientific) in TBST, respectively. Blots were developed using a SuperSignal West Femto chemiluminescence kit (Thermo Scientific) and visualized using a Bio-Rad Gel Doc XR imaging system.

qPCR analysis.

Triplicate wells of 100,000 BMMs/well in a 96-well plate were stimulated with WT, ΔsagB mutant, or ΔsagB + sagB complementation strain HKSA (MOI of 100) for 3 h at 37°C. Following activation, lysates from triplicate wells were combined, RNA was purified using an RNeasy Micro kit (Qiagen), and cDNA was synthesized with the GoScript reverse transcription System (Promega). Expression of IL-1β and mouse β-actin was quantified by quantitative PCR (qPCR) using SYBR green (Bio-Rad) and the primer pairs in Table 2. Fold change of induction relative to unstimulated macrophages was calculated using the threshold cycle (2^−ΔΔCT) method.

Murine infection models.

Overnight cultures of S. aureus strains were inoculated in TSB and grown at 37°C. The overnight cultures were then subcultured 1:100 in 15 mL TSB and grown for 3 h at 37°C. Cultures were pelleted by centrifugation at 3,900 rpm at 4°C, and cells were washed three times with 1× PBS. Bacterial cell suspensions were normalized to an OD₆₀₀ of 0.32 to 0.33 (~1 × 10⁸ CFU/mL), and dilutions were plated onto tryptic soy agar (TSA) plates to ensure accuracy of inocula. Six- to 8-week-old female Swiss Webster mice (Envigo) were anesthetized with 2,2,2-tribromoethanol (Avertin) (250 mg/kg) (Sigma) via intraperitoneal injection, followed by inoculation with 100 μL OD-normalized S. aureus in PBS (~1.0 × 10⁷ to 2.0 × 10⁷ CFU) directly into the bloodstream via retro-orbital sinus injection. Mice were monitored daily, and after 96 h postinfection, mice were immediately euthanized by CO₂ narcosis and heart and kidneys were isolated. Organs were homogenized and serially diluted onto TSA plates and incubated at 37°C overnight to enumerate CFU. For the skin infection model, bacterial suspensions were mixed 1:1 with sterile Cytodex 3 microcarrier beads (Sigma), and 200 μL of this mixture (1 × 10⁷ CFU/mL) was injected intradermally into anesthetized mice on each side of a shaved flank region using a 25G needle. After 120 h postinfection, mice were euthanized and abscesses were excised, homogenized, and plated on TSA to enumerate CFU. Homogenates were also clarified via centrifugation and stored at −80°C until used for CBA quantification of cytokines.

Ethics statement.

All experiments were performed according to an IACUC-approved protocol (IACUC no. 20-025) following the guidelines of the Office of Laboratory Animal Welfare, USDA and PHS policy, and the ethical standards of the Institutional Biosafety Committee and the Institutional Animal Care and Use Committee (IACUC) at the Loyola University Chicago Health Sciences Division. The institution is approved by Public Health Service (D16-0074), is fully accredited by the AAALAC International (000180, certification dated November 2016), and is registered/licensed by the USDA (33-R-0024 through 24 August 2023).

Statistical analyses.

Statistical analyses were performed using Prism 9 GraphPad Software. All experiments were performed at least three independent times. For all in vitro macrophage and dendritic cell cytokine data, statistical significance (P < 0.5) was determined from representative experiments performed in triplicate by one-way analysis of variance (ANOVA) with Tukey’s post hoc test. For in vivo infection data, a nonparametric Kruskal-Wallis test was used. The number of animals (n) used in each experiment is indicated in the figure legends.