YjbH contributes to Staphylococcus aureus skin pathology and immune response through Agr-mediated α-toxin regulation

ABSTRACT

Staphylococcus aureus is the most common cause of skin and soft tissue infections (SSTIs) with Methicillin-Resistant S. aureus (MRSA) strains being a major contributor in both community and hospital settings. S. aureus relies on metabolic diversity and a large repertoire of virulence factors to cause disease. This includes α-hemolysin (Hla), an integral player in tissue damage found in various models, including SSTIs. Previously, we identified a role for the Spx adapter protein, YjbH, in the regulation of several virulence factors and as an inhibitor of pathogenesis in a sepsis model. In this study, we found that YjbH is critical for tissue damage during SSTI, and its absence leads to decreased proinflammatory chemokines and cytokines in the skin. We identified no contribution of YjbI, encoded on the same transcript as YjbH. Using a combination of reporters and quantitative hemolysis assays, we demonstrated that YjbH impacts Hla expression and activity both in vitro and in vivo. Additionally, expression of Hla from a non-native promoter reversed the tissue damage phenotype of the ΔyjbIH mutant. Lastly, we identified reduced Agr activity as the likely cause for reduced Hla production in the ΔyjbH mutant. This work continues to define the importance of YjbH in the pathogenesis of S. aureus infection as well as identify a new pathway important for Hla production.

Introduction

The skin is one of the first lines of defense against pathogens. It has a variety of mechanisms to combat these pathogens including toxic fatty acids, low pH, layers of dead skin cells that slough off, tissue-resident immune cells, and the ability to rapidly recruit immune cells to the site of tissue damage or infection [1,2]. Problems arise when pathogens can circumvent these defenses and cause infection. While Staphylococcus aureus is a common colonizer of the human body, it is the leading cause of skin and soft tissue infections (SSTIs) in the United States [3,4]. The emergence of lineages, such as multidrug-resistant Community-Associated Methicillin Resistant S. aureus (CA-MRSA), has made treating S. aureus infections challenging for health-care providers and MRSA alone caused over 100,000 worldwide deaths in 2019 [5]. In addition to SSTIs, S. aureus can cause life-threatening invasive infections with high morbidity and mortality. While progress has been made toward decreasing S. aureus infection, this progress has reduced in recent years [6]. While the overall incidence of SSTI is not changing drastically, infections with complications have significantly increased, with deaths post-SSTI hospitalization and healthcare-related costs also rising [7]. Considering S. aureus’ importance in human health, understanding factors contributing to disease is imperative to combating this pathogen and the hope of developing new therapeutics and treatments to fight these infections.

S. aureus is a gram-positive opportunistic pathogen that can infect almost any anatomical site on the human body. It possesses a diverse metabolic potential and a variety of virulence factors that allow this bacterium to respond rapidly to changing host niches within different tissue environments. Recently, we found that the YjbH protein is important for the production of several key S. aureus virulence factors including aureolysin and staphyloxanthin, as well as resistance to oxidative and nitrosative stress [8]. We found that a ΔyjbIH mutant has increased pathogenesis during a murine model of systemic infection. This was followed by additional reports that YjbH is important for survival in whole blood [9] and pathogenesis in a silkworm model [10,11]. YjbH mutants have altered antibiotic susceptibility profiles, including glycopeptides and β-lactams [12–14]. Recently, mutations in yjbH were identified in invasive strains of S. aureus sequenced from human infections [15]. While the role of YjbH in virulence is becoming clearer, mechanistically how this protein modulates virulence is still poorly understood. YjbH has been characterized as an adapter protein for the stress response regulator Spx in several bacteria [16–18]. Under non-stress conditions, YjbH binds Spx and targets it for degradation via the ClpXP proteolytic system [16–21]. During stress, YjbH aggregates, leading to its disengagement with Spx, which then regulates gene expression. While the full extent of the YjbH/Spx regulon is not well described in S. aureus, it interplays with several other systems including the alternate sigma factor, σ^B and the MazEF toxin-antitoxin system [8,9,18].

One central virulence factor to S. aureus pathogenesis is α-hemolysin (Hla), a pore-forming toxin that can lyse a variety of host cells [22] and activate the NLRP3 inflammasome [23]. Hla is critical for S. aureus pathogenicity in a multitude of infection models and is essential for surface tissue damage during skin infection [24–31]. While Hla expression is impacted by several regulatory networks, activation of its promoter is primarily controlled by and is a direct target of the SaeRS two-component system [32–34]. Additionally, Hla is subject to post-transcriptional control, and cannot be efficiently translated without the Agr-dependent regulatory RNA, RNAIII [35,36].

In this study, we discovered that YjbH is critical for tissue damage during a model of S. aureus skin infection. We found that in the absence of YjbH, there is a concomitant significant decrease in Hla expression and activity compared to its parent strain in vitro, and a significant attenuation in Hla production in vivo, which likely accounts for the changes in tissue damage. The impact of YjbH on Hla production likely occurs through alterations in the activity of the Agr quorum sensing system. These data further demonstrate the importance of YjbH in the ability of S. aureus to cause disease.

Materials and methods

Bacterial strains and growth conditions

Unless otherwise stated, S. aureus strains (Table 1) were grown in tryptic soy broth (TSB, BD cat. #211822) with or without agar, as needed. When necessary, media was supplemented with trimethoprim (10 µg mlL⁻¹), chloramphenicol (10 µg mL⁻¹), and/or erythromycin (5 µg mL⁻¹). Plates were grown at 37°C in a static incubator, and all cultures were grown at 37°C with a 1:10 media-to-flask volume ratio and shaken at 250 rpm. E. coli strains were grown in lysogeny broth (LB) with or without agar supplemented with ampicillin (100 µg mL⁻¹) when needed.

Table 1.

Staphylococcus aureus strains and plasmids used in this study.

Strain or Plasmid	Relevant characteristicsa	Source or reference
Bacterial Strains
RN4220	Highly transformable S. aureus	[37]
AH1263	USA300 CA-MRSA strain LAC lacking LAC-p03; wild-type strain used for these studies	[38]
JLB15	AH1263 saeR::Tn, referred to as saeR	[39]
JLB24	AH1263 hla::Tn	[40]
JLB110	AH1263 ΔyjbIH	[8]
JLB134	AH1263 ΔyjbI	[8]
JLB145	AH1263 yjbH::Tn	[8]
JLB147	AH1263 ΔyjbH	[8]
JLB323	AH1263 ΔyjbIH hla::Tn	This study
JLB316	AH1263 ΔRNAIII-agrBDCA	[41]
JLB331	AH1263 ΔsaeRS	[42]
JLB362	RN4220 Phla-lux	This study
JLB366	AH1263 Phla-lux	This study
JLB367	AH1263 ΔyjbIH Phla-lux	This study
JLB368	AH1263 ΔRNAIII-agrBDCA Phla-lux	This study
JLB371	AH1263 ΔsaeRS Phla-lux	This study
Plasmids
pCP4	yjbIH pKK22-based complementation plasmid, referred to as pyjbIH	[8]
pCP5	yjbH pKK22-based complementation plasmid, referred to as pyjbH	[8]
pCP6	yjbI pKK22-based complementation plasmid, referred to as pyjbI	[8]
pCP20/pRKC1032	Ppsmα-lacZ	[43]
pCP25	hla under control of non-native promoter	This study
pJB1017	Phla-lacZ reporter	[39]
pJB1018	Pcoa-lacZ reporter	[39]
pJB1063	lux system in pJC1112	This study
pJB1066	Hla lux reporter (Phla-lux)	This study
pJC1112	ErmR plasmid that integrates into SapI1	[44]
pKK22	TmpR complementation plasmid	[24]
pRN7023	Helper plasmid for SapI1 integration	[45]
pHC125	luxCDABEG chromosomal integration plasmid for φ11 attachment site	[46]

^aTn indicates the insertion of the Bursa aurealis transposon from the Nebraska Transposon Mutant Library.

Construction of plasmids

Primers (Table 2) were synthesized by IDT, Inc. Plasmids were first generated and confirmed in E. coli DH5α or DH5αλpir (for pKK22-based plasmids) by restriction digest using New England Biolabs enzymes. Inserts were sequenced by ACGT, Inc. to ensure no unintended mutations occurred. Next, plasmids were transferred into S. aureus RN4220 by electroporation [37], with subsequent movement between S. aureus strains performed via ϕ11-mediated transduction as previously described [47].

Table 2.

Oligonucleotides used in this study.

Name	Sequencea	Source or reference
CP14	GCTAGGGCGATGTTATCAAAGTTGG	This study
JBPRO4	CGATCCATTTAGCTCCGATTGCTTC	This study
CP114	GCGAATTCCTAATCTCTCGCATAATTGCTTATG	[43]
CP115	GAAGTCGACTAAGATTACCTCCTTTGCTTATGAGT	[43]
CP121	gttACGCGTTTAATTTGTCATTTCTTCTTTTTCCCAATCG	This study
CP126	ttaACGCGTCAAATCGTTTTAAAAAATAGAAGGATGATGAAA	This study
JBLUX1	GCACGTGAAAGATTTTGTAGTGAAAGAGATGCAAG	This study
JB22	ccgctagcCATTTTCATCATCCTTCTATTTTTTAAAACGG	[24]
JBKU145	cacgaattCTTTATGAAACAAGGAAAAGACATAGC	This study
JCO717	GTGCTTCACCAGCACCACATGCTG	[44]
JCO719	GGTATTAGTTTGAGCTGTCTTGGTTCATTGATTGC	[44]

^aPrimer sequences are provided 5’ to 3,’ lowercase indicates non-homologous bases added for cloning purposes.

To construct the hla complementation plasmid with a non-native promoter, pCP25, the hla gene and Shine Dalgarno site were amplified from AH1263 using primers CP121 and CP126. The resulting product was digested with MluI and cloned into pKK22 digested with AscI. This places hla downstream of the P_sarAP1 promoter and dfrA gene. This plasmid was then moved into JLB24 (hla::Tn), and JLB323 (ΔyjbIH hla::Tn).

The generation of the lacZ-reporter plasmid, pCP20, was partially described previously and named pRKC1032 [43]. For clarity, it is provided here with additional details. The psmα promoter was amplified from AH1263 using primers CP114 and CP115. The resulting product was digested with EcoRI and SalI and cloned into the same sites of pJB185.

Construction of mutant strains

To construct the ΔyjbIH hla::Tn mutant, ϕ11 was propagated on JLB24 (hla::Tn) and used to transduce the mutation into JLB110 (ΔyjbIH) [47]. This created the strain named JLB323 and confirmed using primers CP14 and JBPRO4.

For the lux reporter strains, a custom lux reporter construct was designed based on pHC125 and synthesized by GeneScript. The construct has luxCDEG constitutively expressed from P_cp25 and P_cap as in pHC125. They are followed by a Rho-independent transcriptional terminator and a multiple cloning site. The luxAB genes follow the multiple cloning site to allow promoter insertion. There is a second Rho-independent transcriptional terminator after luxB. This custom-designed and synthesized construct was cloned into pJC1112 to generate pJB1063, which inserts into the SapI1 site in the presence of the pRN7023 helper plasmid. Next, the hla promoter, including the RNAIII binding sequence, was synthesized and cloned into the PstI and NheI sites of pJB1063 such that the native ATG codon of Hla is the first codon of luxA, generating pJB1066. This plasmid was electroporated into S. aureus RN4220 containing pRN7023 to create JLB362 and integration was confirmed with JCO717 and JCO719. The integrated pJB1066 was transduced into AH1263. To create the reporter and control strains used for experiments, ϕ11 was used to make phage on AH1263 containing pJB1066, and transduction was performed with AH1263, JLB110 (ΔyjbIH), JLB331 (ΔsaeRS), and JLB316 (ΔRNAIII-agrBDCA) to create JLB366, JLB367, JLB371, and JLB368, respectively. Final confirmation was performed with JBLUX1, JB22, and JBKU145.

β-galactosidase assays

Assays for lacZ reporters were performed as previously described [48]. Briefly, cells from 1-mL of culture were harvested at 6 h for P_coa and P_psmα, and 8 h for P_hla, and then subsequently frozen at −80°C until the assay was performed. Cell pellets were resuspended in Z-Buffer and disrupted using 0.1 mm glass beads in 2-mL tubes using an MP Biomedicals FastPrep-24 5G homogenizer. Cell debris was pelleted, and an aliquot of cell suspension was used to determine protein content by Bradford Assay. A separate aliquot was then mixed with Z-buffer and ONPG and incubated at 37°C until slightly yellow, at which point 1 M NaCO₃ was added to stop the reaction. Absorbance at 420 nm was used to measure activity, and modified Miller units were calculated using protein content as the equilibrator.

Making of the inocula for the infection model

Inoculums for infection models were prepared as previously described with modifications [49]. Bacteria were retrieved from −80°C freezer stocks and inoculated into 3 mL of TSB supplemented with antibiotics when needed. The culture was incubated at 37°C overnight with shaking at 250 rpm. The culture was then diluted 1:100 in 10 mL TSB in a 50-mL conical tube with a loose cap and grown for 3 h at 37°C with shaking at 250 rpm. Bacteria were pelleted at 4,000 × g for 10 min, washed with Dulbecco’s phosphate-buffered saline (DPBS) without calcium and magnesium, pelleted again, and resuspended to ~4 × 10⁷ CFU per 50-μL injection. The CFU concentrations were confirmed by dilution plating. When using strains containing pCP25, inoculums were reduced by half due to high levels of Hla production.

Subcutaneous murine infection

All animal studies were conducted in strict accordance with approved protocol by the University of Kansas Medical Center Institutional Animal Care and Use Committee under ACUP number 2021–2613 and follow the ARRIVE guidelines. Nine-week-old female C57BL/6J mice (Jackson Laboratory) or SKH1 mice (Charles River Laboratories) were infected as previously described [49], with modifications. For C57BL/6J mice, the mice were depilated by mechanical shaving and Nair® and then moisturized for two subsequent days before infection. SKH1 mice were untreated. For infection, mice were anesthetized via isoflurane gas and were injected subcutaneously with ~4×10⁷ CFU of bacteria in 50 μL of DPBS (without calcium and magnesium; Corning). Mice were monitored and photographed daily for the duration of the infection. On day 4, mice were euthanized by CO₂ inhalation per the institutional protocol. Using calipers, 2.25 cm² of skin was excised around the lesion area, then bisected, and minced. The tissue was placed in two lysing matrix H tubes (MP Biomedicals) containing 1.5 mL total of Hank’s balanced salt solution (HBSS; no cations) plus 0.2% (wt/vol) human serum albumin (A5843; Sigma Aldrich) and 10 mM HEPES. The samples were homogenized using a Fastprep-24 5 G (MP Biomedicals), according to the manufacturer’s protocols for the skin. The skin samples were then serially diluted in PBS for bacterial titer enumeration on TSA plates. The remaining homogenate was clarified by centrifugation at 10,000 × g for 10 min at 4°C. cOmplete protease inhibitor (Roche) and 0.8 mM EDTA (Fisher Chemical) were added to the supernatant, and the sample was aliquoted and stored at −80°C for future cytokine and chemokine analyses. For infections using the strains containing pCP25, the lesions exceeded the given area created by the calipers, so the lesions were excised with ~3 mm of uninfected skin and weighed before homogenization. This adjustment to tissue collection was also applied to animals used in live animal imaging.

To determine surface lesion size, photos were taken with a ruler in the frame. Then, using ImageJ (NIH), the overall lesion size was measured as well as the necrosis size. Lesion size was defined as the largest observable surface lesion and included both blanching of the skin and the necrotic scab-like surface, which was used to determine necrosis alone.

For IVIS studies, infected mice were anesthetized with isoflurane and placed into an isolation imaging chamber and locked and transferred into the main imaging chamber of an IVIS Spectrum. Luminescence was monitored at 24-h intervals until euthanasia. Images were analyzed, and luminescence was quantified using Living Image 4.0 software (Perkin Elmer).

Cytokine measurement

Frozen homogenates were thawed, and cytokines and chemokines were quantified using cytometric bead array assays (BD Biosciences) following the manufacturer’s recommendations on a BD FACS Aria. A custom flex set was compiled to quantify the following cytokines: KC (CXCL1), IL-6, MIP-1α (CCL3), TNF, MIP-1β (CCL4), IL-1β, GM-CSF, RANTES, IFN-γ, G-CSF, MCP-1 (CCL2), IL-17F, and IL-1α. The BD Enhanced Flex Set was used for IL-4 and IL-17A. As previously described [49], everything was performed according to the manufacturer’s protocol, and the data analysis was performed using FCS Express Version 7.4.3.

Quantitative hemolysis assay

Cells were grown overnight at 37°C with shaking at 250 rpm. First, the OD₆₀₀ of the culture was determined using a Genesys 10S UV-Vis (Thermo Scientific) spectrophotometer for standardization. Next, the supernatant from 1 mL of culture was collected. The amount of supernatant added to the assay was corrected by OD₆₀₀ when compared to WT S. aureus and volume was balanced with supplementation of TSB. The total 200 μl solution was mixed with 200 μl 2X Hemolysis buffer (10% NaCl, 1 M CaCl₂) and 40 μl of Rabbit blood (in Citrate from Hemostat Labs). The samples were placed in a nutator and incubated at 37°C for 10 min. Unlysed cells were concentrated by centrifugation at 5,500 × g for 1 min. The supernatant was transferred to a 96-well plate, and OD₅₄₃ was measured for detection of released heme.

Sample growth, preparation, and LC-MS/MS analysis of secreted proteins

Strains were struck on TSA (with antibiotic if carrying a plasmid) from frozen stock for isolated colonies. The following day, cultures were started from isolated colonies in 3 mLs TSB (±antibiotic) and grown overnight. Cell density was determined using a spectrophotometer (Genesys 10S UV-Vis, Thermo Scientific) at OD₆₀₀. The samples were then diluted in TSB to OD₆₀₀ = 0.1 in 12.5 mL of TSB in a 125-mL flask. The flasks were then grown as previously stated above for 8 h. The samples were then pelleted by centrifugation and cell-free supernatants were collected by filtration (33 mm, 0.22 μm PES syringe filter, Fisher Scientific)

Next, the supernatants were concentrated 10× using 3k MWCO protein concentrators (0.5 mL, Pierce). The retained sample was transferred to a fresh microcentrifuge tube for protein isolation and reduced with the addition of 0.5 M TCEP to a final concentration of 5 mM followed by incubation at 37°C for 30 min. Reduced samples were alkylated with the addition of 375 mM iodoacetamide to a final concentration of 10 mM followed by incubation in the dark at room temperature for 30 min. Ice-cold acetone was added to each sample to a volume ratio of 5:1. Samples were vortexed and stored at −20°C overnight. After precipitation, samples were centrifuged at 14,000 × g at 4°C for 10 min to pellet the proteins. The supernatant was removed, and the pellet was air-dried on the benchtop for 10 min. The proteins were resuspended in 50 mM TEAB pH 8, 2 mM CaCl₂. Trypsin was added (500 ng), and the proteins were allowed to digest overnight at 37°C with shaking at 500 RPM (Thermomixer, Eppendorf). The digestion was quenched with the addition of 10% formic acid to a final concentration of 1%. Digested samples were centrifuged at 10,000 × g for 10 min to remove particulates, and the supernatant was transferred to a fresh tube and stored at −20°C until mass spectrometry analysis. Peptide concentration was measured using a Nanodrop spectrophotometer (Thermo Scientific) at 205 nm.

Samples were injected using the Vanquish Neo (ThermoFisher) nano-UPLC onto a C18 trap column (PepMap™ Neo Nano Trap, 0.3 mm × 5 mm, 5 µm particle size) using pressure loading. Peptides were eluted onto the separation column (PepMap™ Neo, 75 µm × 150 mm, 2 µm C18 particle size, ThermoFisher) before elution directly to the mass spectrometer. Briefly, peptides were loaded and washed for 5 min at a flow rate of 0.350 µL/min at 2% B (mobile phase A: 0.1% formic acid in water, mobile phase B: 80% ACN, 0.1% formic acid in water). Peptides were eluted over 100 min from 2% to 25% mobile phase B before ramping to 40% B in 20 min. The column was washed for 15 min at 100% B before re-equilibrating at 2% B for the next injection. The nano-LC was directly interfaced with the Orbitrap Ascend Tribrid mass spectrometer (ThermoFisher) using a silica emitter (20 µm i.d., 10 cm, CoAnn Technologies) equipped with a high field asymmetric ion mobility spectrometry (FAIMS) source. The data were collected by data-dependent acquisition with the intact peptide detected in the Orbitrap at 120,000 resolving power from 375 to 1500 m/z. Peptides with charge + 2-7 were selected for fragmentation by higher energy collision dissociation (HCD) at 28% NCE and were detected in the ion trap at rapid scan rate. Dynamic exclusion was set to 60 s after one instance. The mass list was shared between the FAIMS compensation voltages. FAIMS voltages were set at −45 (1.4 s), −60 (1 s), −75 (0.6 s) CV for a total duty cycle time of 3 s. Source ionization was set at +1700 V with the ion transfer tube temperature set at 305°C. Raw files were searched against the Staphylococcus aureus (strain USA300) protein database downloaded from UniProt on 09-22-2023 and a common contaminants database using Sequest in Proteome Discoverer 3.0 [50]. Abundances, abundance ratios, and p-values were exported to Microsoft Excel for further analysis.

Statistical analysis

All statistical analyses were performed using GraphPad Prism (version 10) or Proteome Discoverer 3.0 for LC-MS/MS. Individual figure legends convey the tests performed on individual experiments and data groups.

Results

YjbH contributes to tissue damage in a skin infection model

Our previous work found that a ΔyjbIH mutant had increased pathogenicity in a systemic infection model [8]. Since SSTIs are the most common type of S. aureus infection [51–55], we tested the importance of YjbIH in a mouse model of SSTI. To this end, we examined surface lesion formation during infection over the first 4 days. In contrast to the systemic infection where the absence of YjbH enhanced pathogenesis, in the skin model, the ΔyjbIH mutant had significantly smaller surface lesions (Supplementary Figure 1). This was true for the necrotic scab-like portion of the lesion (Figure 1a,c) and the total lesion (Supplementary Figure 2), which is the largest surface area that includes both blanching of the skin and the necrosis. At times, no surface lesion was apparent in ΔyjbIH mutant-infected mice, but the infection appeared as a subdermal abscess (Supplementary Figure 1). The decreased lesion size observed by the ΔyjbIH mutant could be restored with yjbIH provided on a plasmid (Figure 1c, Supplementary Figures 1, and 2a). The reduced lesion size was not the result of decreased colonization as there was no difference in bacterial titers at the site of infection (Figure 1b,d). To determine the relative contributions of YjbI and YjbH to the phenotype, we tested individual ΔyjbI and ΔyjbH mutants. YjbI had no discernable impact on lesion formation (Figure 1a, Supplementary Figures 1, 2c, and 3); however, the ΔyjbH mutant phenocopied the ΔyjbIH mutant with lower total lesion (Supplementary Figure 2b) and necrosis (Figure 1a,c). This mutant often presented as a subdermal abscess as well (Suplementary Figure 1). We observed a modest reduction in bacterial titers at the site of infection for the ΔyjbH mutant (Figure 1f). As with ΔyjbIH, the ΔyjbH mutant phenotype could be restored when yjbH was provided on a plasmid (Figure 1e,f and Supplementary 2b). To ensure the ΔyjbH mutant phenotype was not mouse-line specific, we performed the same model using SKH1 mice. Again, the ΔyjbIH mutant infection had significantly smaller lesions at 4 days post-infection (Supplementary Figure 4a,b) and no difference in bacterial titers (Supplementary Figure 4c). Together, these results demonstrate that YjbH is important for tissue damage during infection and results from factors unrelated to bacteria levels at the site of infection.

Figure 1.

YjbH contributes to tissue damage during SSTI. C57BL/6J were infected subcutaneously with wild type (WT), mutants in ΔyjbI, ΔyjbH, ΔyjbIH, and respective complement plasmids as indicated. (a,c,e) Shown as necrosis size over time. Symbols represent the mean (n = 5-8) with SEM. (b,d,f) bacterial titers enumerated on day four post-infection. Bars represent the mean with SEM. Each dot represents an individual mouse. Data is representative of at least three independent experiments. *, p < 0.05; **, p < 0.01 by two-tailed Mann-Whitney test compared to WT.

The lack of YjbH leads to a decrease in proinflammatory cytokines

Considering the reduced skin pathology observed in the ΔyjbH mutant compared to wild type, we anticipated an attenuation of the immune response. Thus, we quantified a variety of cytokines and chemokines related to inflammation, immune cell recruitment, and immune cell activation at the site of infection. From the panel measured, we found a significant decrease in proinflammatory cytokines IL-6 and TNF as well as hematopoietic growth factors GM-CSF and G-CSF that aid in the activation/potentiation of neutrophils and other leukocytes in the ΔyjbH mutant when compared to the wild-type strain (Figure 2). We also discovered that chemokines CCL2 (MCP-1), and CXCL1 (KC), which are involved in leukocyte trafficking, showed a significant attenuation in the ΔyjbH mutant. Additionally, we found a trend towards decreased IL-1β, CCL4 (MIP-1β), and CCL5 (Figure 2 & Supplementary Figure 5) though they did not reach significance. We were unable to perform statistical analyses on CCL5 (RANTES), but in contrast to wild-type infected mice, all mice infected with the ΔyjbH mutant had concentrations below the level of detection. Surprisingly, we found an increase (p = 0.07) in proinflammatory cytokines IL-17A and IL-17F (Supplementary Figure 7) which are important for the defense against extracellular pathogens and important for the host immune response during S. aureus skin infections [56–59]. Again, to show reproducibility across mouse lines, we performed the same analysis on select cytokines in our SKH1 mice. Akin to C57BL/6J mice results, SKH1 mice infected with the ΔyjbIH mutant also had significantly decreased levels of IL-6, TNF, and CXCL1. In this case, both CCL4 and IL-1β reached significance (Supplementary Figure 6). Together, these data demonstrate an altered immune environment during SSTI with the ΔyjbH mutant leading to less proinflammatory cytokines produced.

Figure 2.

Mice infected with ΔyjbH mutant have altered cytokines at the site of infection. Homogenized infected skin from mice in Figure 1e was analyzed by cytometric bead arrays. Each dot represents an individual mouse. Empty symbols are values outside the limit of quantification and were set at the max or min limit of quantification. # and ## indicate a significant difference of p < 0.05 or 0.01, respectively, compared to wild type (WT) by two-tailed Mann-Whitney test.

Activity and expression of α-hemolysin are decreased in ΔyjbIH and ΔyjbH mutants

Hla is the primary contributor to skin necrosis and hla mutants cause little to no surface lesions [26,60]. In addition, hla mutants have a modest decrease in bacterial titers at the site infection at this same time point [29], like that seen in the ΔyjbIH and ΔyjbH mutants. Therefore, we hypothesized that reduced lesion formation in the ΔyjbH mutants was due to decreased Hla production. To test this, we examined Hla activity in vitro. To this end, we performed a quantitative hemolysis assay using rabbit blood. Supernatants from the ΔyjbH and ΔyjbIH mutants had significantly reduced hemolytic activity compared to the wild type and resembled an hla mutant (Figure 3a). This phenotype could be restored by providing yjbH or yjbIH on a plasmid in the respective mutant strains. Consistent with skin necrosis resembling wild type in the ΔyjbI mutant, the hemolytic activity of this mutant was not significantly different when compared to wild type. These data demonstrate that YjbH is important for Hla activity.

Figure 3.

YjbH is important for α-hemolysin production. (a) supernatants were collected from indicated strains after overnight (15 h) growth and used in a quantitative hemolysis assay with rabbit blood. Bars represent the mean (n = 3) with SEM. (b) cells of indicated strains were collected at early stationary phase and used in a β-galactosidase assay to measure activity of a P_hla –lacZ reporter. Bars represent the mean (n = 3) with SEM. Data is representative of ≥3 independent experiments. * and ** indicate p-value of p < 0.05 and p < 0.01, respectively, by t-test.

Since Hla is subject to regulation at the transcriptional and post-transcriptional levels, reduced Hla activity could be due to either of these possibilities. Therefore, we next wanted to identify if the attenuation in activity was due to lower levels of expression or just overall activity. To test this, we used a lacZ reporter of Hla expression that accounts for both Sae- and Agr-dependent regulation of Hla. Expression was examined during the transition from exponential to stationary-phase growth when Hla is highly expressed. Like Hla activity, we found that both the ΔyjbIH and ΔyjbH mutants had significantly decreased expression from this reporter, while YjbI did not impact expression (Figure 3b). Again, expression was restored by complementation on a plasmid. Together, these data demonstrate that YjbH is important for the production of Hla, and this occurs at either the transcriptional or translational level.

YjbH promotes Agr activity

Reduced Hla expression could result from changes in SaeRS or Agr activity since the hla promoter is activated by SaeR and translation of Hla requires RNAIII, which is Agr dependent. We hypothesized that one or both systems would have reduced activity in the ΔyjbH mutant, leading to decreased Hla expression. To test SaeRS’ role in regulation, we used a lacZ-reporter plasmid for P_coa, whose expression is dependent on activated SaeR. We observed a modest but significant increase in P_coa expression in the ΔyjbH mutant, which could not account for the loss of Hla expression (Figure 4a). To measure Agr activity, we used the promoter for the α phenol-soluble modulins (psmα) operon which is a direct target of Agr activation [61]. Expression from this promoter was significantly decreased in both the ΔyjbH and ΔyjbIH mutants, indicative of reduced Agr activity (Figure 4b). The ΔyjbI mutant was similar to the wild-type strain for both reporters (Figure 4a,b). To further determine if the Agr regulon is impacted by YjbH, we performed LC-MS on the secreted proteins of wild-type and ΔyjbH strains when grown in broth culture to late exponential phase. We observed a significant change in the abundance of a number of proteins in the ΔyjbH strain compared to wild-type (Supplementary Figure 6a). This included multiple proteins that are Agr regulated [36,62,63] (Supplementary Figure 6a,b), including but not limited to Hla (2-fold), SpA (6.5-fold), and multiple leukocidins and proteases (ranging from 2–6-fold). Furthermore, ΔyjbH containing the complement plasmid more closely resembled the wild-type strain (Supplementary Figure 6b). Together, these data support a model by which Agr activity is decreased in the absence of YjbH, leading to decreased RNAIII expression and a reduction in Hla translation.

Figure 4.

The absence of YjbH leads to increased Sae activity and decreased Agr activity. Strains containing plasmids with the (a) P_coa-lacZ or (b) P_psmα-lacZ reporters for Sae and Agr activity, respectively. Samples were taken at mid-exponential phase of growth (6 h) and used in a β-galactosidase assay. Data are representative from ≥ 3 independent experiments. Bars represent the mean (n = 3 for (a) and n = 4 for (b)) with SEM. *, **, and *** indicate p < 0.05, p < 0.01, and p < 0.001 by t-test, respectively.

Controlled expression of hla alleviates YjbH effects on tissue damage

Since the ΔyjbH mutant has reduced lesion formation during skin infection and reduced Hla activity (Figure 1 and 3), we predicted that if we expressed Hla from a non-native promoter, tissue damage during infection would be unaffected by the absence of YjbH. To do this, we cloned hla under the control of the sarA P1 promoter. The plasmid was then transferred to the hla and ΔyjbH hla mutants so that the only Hla produced was from our construct. First, we confirmed that these strains had equal hemolytic activity in vitro (Figure 5a). Next, we tested whether YjbH still impacted lesion size in vivo when hla was expressed from a non-native promoter. We observed no difference in necrotic lesion formation between these strains (Figure 5b) and did not observe a difference in bacterial titers at the site of infection (Figure 5c). These data demonstrate that equalizing Hla expression between wild-type and ΔyjbH mutant strains leads to equal pathology and the restoration of bacterial titers recovered after infection. This supports the model that decreased Hla expression is the cause of reduced lesion formation in the ΔyjbH mutant.

Figure 5.

Controlled expression of α-hemolysin restores lesion formation in the ΔyjbIH mutant. (a) supernatants from overnight (15 hr) cultures of the hla mutant or ΔyjbIH hla mutant containing pCP25 (P_sarA-hla) were used in a quantitative hemolysis using rabbit blood. An hla mutant without plasmid serves as a negative control. Bars represent the mean (n = 3) with SEM and are representative of ≥3 independent experiments. “ns” and ** indicate not significant or significant (p < 0.01) by t-test. (b & c) C57BL/6J mice were infected with the hla mutant or ∆yjbIH hla mutant containing P_sarA-hla and (b) necrosis measured for 4 days and (c) bacterial titers enumerated at 4 days post-infection. Data represent the mean (n = 12-13) with SEM. For panels B & C, no comparison between strains was statistically different by two-tailed Mann-Whitney test.

Hla expression is reduced in the ΔyjbIH mutant during skin infection

Together, our data support a model whereby YjbH is important for AgrA-mediated activation of Hla expression and that this contributes to tissue damage during infection. While in vitro results are valuable, they do not always translate to in vivo responses. Therefore, we sought to monitor Hla expression during infection. To this end, we developed a bioluminescent reporter for Hla that included the native promoter, RNAIII-binding sequence, and the native Shine Dalgarno sequence. This construct (Supplementary Figure 7) was moved into wild type and ΔyjbIH as well as the negative control strains ΔsaeRS and Δagr. First, we tested these strains in vitro and showed that they behaved as expected, i.e., there was less luminescence in vitro for ΔyjbIH, ΔsaeRS, and Δagr (Figure 6a). Next, we performed our skin infection model. Bioluminescence, indicative of Hla expression, was highest at 1- and 2 days post-infection and declined afterward in the wild-type strain (Figure 6b). As expected, the bioluminescence of the Δagr and ΔsaeRS mutants was similar to background readings. The ΔyjbIH strain had significantly reduced bioluminescence, demonstrating a significant attenuation in Hla expression during skin infection. The strains behaved similarly to their parent strain (i.e. non-lux-containing strains) when it came to necrotic lesion formation (Figure 6c) and small attenuation in overall bacterial titers (Figure 6d) when compared to the WT strain. Overall, these data support the model that YjbH is important for Hla production both in vitro and in vivo and this contributes to pathology during an SSTI.

Figure 6.

YjbH is critical for α-hemolysin production during skin infection. (a) indicated strains expressing a P_hla-lux reporter were grown in TSB until late exponential phase and relative luminescence were measured using a Tecan Spark plate reader. Bars represent the mean (n = 4) with SEM. (b-d) C57BL/6J mice were infected with indicated strains expressing P_hla-lux reporter and (b) luminescence measured as photons per second using IVIS, (c) necrosis sized monitored, and (d) bacterial titers at the site of infection determined at 4 days post-infection. For b-c, data represents the mean (n = 10) with SEM. For d, each dot is an individual mouse, and the bars represent the mean (n = 10) with SEM. * p < 0.05; **, p < 0.01; ****, p < 0.0001 by t-test for panel a or two-tailed Mann-Whitney test in b-d.

Discussion

This study sheds new light on the impact of YjbH on S. aureus virulence factor regulation and disease outcomes. We observed a substantial reduction in lesion and necrosis formation when YjbH was absent compared to the wild-type strain, alongside a significant decline in specific proinflammatory cytokines. The data supports a model whereby YjbH is critical for Hla production by S. aureus, impacting disease outcomes during S. aureus skin infection. This model is supported by our findings that 1) when YjbH is absent, S. aureus produces less Hla and there is less hemolytic activity, 2) ectopic expression of Hla from a non-native promoter removes the YjbH effect on Hla activity and tissue damage, and 3) the ΔyjbH mutant has less Hla expression in vivo. Furthermore, our study revealed that this is likely mediated through the Agr quorum sensing system since Hla production is lower and P_psmα activity is reduced, which are both consistent with lower Agr activity. This is supported by proteomics demonstrating protein changes indicative of reduced Agr activity. These results underscore the pivotal role of YjbH in S. aureus’ pathogenic potential.

The impact of YjbI and YjbH on S. aureus virulence has been tested in several in vitro and in vivo models. A yjbH mutant of strain SH1000 has decreased survival in whole-human blood [9]. By contrast, a yjbI:Tn mutant showed enhanced survival when exposed to the murine macrophage-like cell line RAW 264.7 [11]. However, the transposon insertion in yjbI has a polar effect on yjbH, and this was not further explored. The finding that the yjbI::Tn would be resistant to macrophage killing is surprising since YjbH contributes to oxidative and nitrosative stress resistance [8,64]. The same yjbI::Tn mutant also had decreased pathogenicity in a silkworm model [11], which was later shown to be the result of the polar effect on yjbH [10]. Interestingly, virulence in this model was due to the role of YjbH in oxidative stress resistance. The role of YjbI and YjbH in mouse systemic models of infection remains unclear. The yjbI::Tn mutant was shown to have lower bacterial burdens in the heart and kidney, while a yjbH::Tn mutant only showed decreased titers in the heart [11]. This contrasts with our previous study, which identified no difference in the yjbI mutant and enhanced bacterial burdens in the kidneys and spleens during infection with a yjbIH mutant [8], a phenotype that was complementable. The discrepancy between these systemic infection studies could be due to either the use of different mouse strains or time points examined, as we used C57BL/6J mice and measured titers at 5–6 d.p.i. versus ICR mice at 1 d.p.i. In addition, we infected with log-phase cells, while the other study used overnight cultures. While more work is needed to fully understand how YjbIH contributes to disease outcomes, it is clear from these studies that the YjbIH system contributes to pathogenesis in a variety of infection models. Additionally, yjbH mutations have been found in clinical isolates from invasive human infections [15], lending support to increased virulence in some niches. We have found that in the absence of yjbH there are a variety of proinflammatory cytokines and chemokines that are reduced during infection. Based on our studies and others, YjbH is important to produce multiple virulence factors. Deciphering whether it is changes in secreted virulence factors or changes in oxidative and nitrosative stress resistance that are mainly responsible for changes in ΔyjbH mutant virulence awaits further investigation.

In our current study, we extended the examination of YjbI and YjbH in virulence to the skin infection model for the first time. In this context, YjbH contributes to pathogenesis. One consistent trend in different virulence studies is that YjbH is the major contributor to the observed phenotypes. Whether YjbH positively or negatively impacts infection outcomes likely results from differences in each niche. We now know that YjbH impacts the expression of a variety of virulence factors, and this occurs through multiple regulatory networks including Spx [8,9], Agr (shown here), and the alternate sigma factor B (σ^B) [8,9]. Both Agr and σ^B regulate the expression of multiple virulence factors, including toxins and proteases. This is now further supported by our proteomics data demonstrating an altered abundance of proteins known to be regulated by these systems. The PSMs are important virulence factors in S. aureus and are a direct target of Agr regulation [61]. They are of interest because they can lyse a variety of host cells and contribute to disease in several animal models [61,65–70]. Indeed, several studies have examined their contribution during skin infection and found varying levels of importance. Using our infection model, we have previously found a psmα mutant to have a modest impact on necrosis [43] which agrees with another recent study [68]. By contrast, hla mutants consistently generate little to no necrosis during S. aureus skin infection. Moreover, when ADAM10, the cellular receptor for Hla, is inhibited early during a wild-type infection, it reduces necrosis and prevents vascular damage during an SSTI model [71]. We cannot rule out a contribution of the PSMs or other virulence factors to the ΔyjbH mutant skin infection phenotype. However, our data suggests that changes in Hla expression drive altered pathology, and it is the key factor impacting tissue damage when YjbH is absent.

While not a direct component of our studies, the IVIS data from our wild-type strain provide several interesting insights into S. aureus gene expression, in vivo. First, the use of this IVIS reporter allows for the tracking of S. aureus virulence factor expression over time, which can be a powerful new tool in understanding S. aureus virulence. Second, Hla is primarily or maximally produced during the early stage of infection. This timing, perhaps not surprisingly, is when primary tissue damage is occurring. Expression of Hla decreases at 3 d.p.i. which also correlates to around peak necrotic lesion size which occurs at 3 d.p.i. (Figure 1 and 6). This implies that the inability of S. aureus to cause continued tissue damage is due to reduced Hla expression over time. Our reporter contains both the Sae-dependent hla promoter and the Agr-dependent aspects of Hla post-transcriptional regulation. Thus, it informs us that both Agr and SaeRS are important early during infection. However, these data cannot determine whether it is a decrease in one or both systems that leads to reduced Hla expression after the first 48 h of infection, i.e., which is limiting. One future goal is to develop new reporters that are singly SaeRS or Agr-dependent to test this relationship.

Using a combination of reporters and assays, our data supports a model whereby YjbH plays a critical role in Agr activation. However, the mechanism by which this occurs remains unknown. Our previous study [8] demonstrated that under the growth conditions used, the ΔyjbH mutant grows similarly to wild type. Again, we did not observe any growth concerns during this study, demonstrating that the “quorum” component of quorum sensing is not what leads to reduced Agr activity. This then questions what is short-circuiting the Agr system when YjbH is absent. One possibility is that YjbH is essential for basal Agr expression. This does not appear to be the case because the P_psmα activity in the ΔyjbH mutant is not attenuated to the point of an agr mutant (Figure 4). This would suggest that YjbH suppresses the Agr positive feedback loop. This suppression would likely be due to effects on either the AgrC kinase or AgrA response regulator. YjbH and Spx are well described to respond to redox and thiol stress. Indeed, the CxxC motif within Spx impacts its activity. Spx, in turn, induces expression of thioredoxin which reduces disulfide bonds. Interestingly, AgrA also possesses reactive cysteines that modulate its activity [72]. It is tempting to speculate whether decreased Agr activity in a ΔyjbH mutant is due to these downstream effects. Alternatively, YjbH has been found in both Bacillus and Listeria monocytogenes to bind proteins other than Spx [19,73]. Thus, ΔyjbH phenotypes could occur independent of Spx and have not been examined in S. aureus. Deciphering the mechanism by which YjbH influences Agr activity and whether this relies on Spx is the focus of our future studies.

This study adds to our understanding of YjbH in S. aureus. It demonstrates a role for YjbH in the regulation of a key virulence factor that contributes to infection in a variety of animal models. It further demonstrates that of the proteins encoded by the operon, YjbH is the major contributor to phenotypes associated with yjbIH and yjbI::Tn mutants. We demonstrate here for the first time that YjbH influences Agr activity and adds to the growing evidence that this protein has pleiotropic effects by modulating the activity of several global regulators. Given the overlap and cross-regulation of these networks in S. aureus, deciphering the mechanism by which YjbH influences them is the focus of our future studies.

Funding Statement

This work was supported by NIH NIAID award AI156251 to J.L.B. and award MF-2104-01575 from The G. Harold and Leila Y. Mathers Foundation to M.T.P.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Author contributions

AKG McReynolds was involved in study design, data generation, writing and editing manuscript. EA Pagella generated data and edited manuscript. MJ Ridder generated a key resource and performed experiments. O Rippee performed experiments. Z Clark and MJ Rekowski generated data and contributed to manuscript preparation. MT Pritchard secured funding to generate a key reagent. JL Bose was involved in study design, data generation, writing and editing manuscript, and secured funding. All authors have read and approved the final version of this manuscript.

Data availability statement

The mass spectrometry data that support the findings of this study are openly available in Mass Spectrometry Interactive Virtual Environment (MassIVE) at (https://massive.ucsd.edu/ProteoSAFe/static/massive.jsp), reference number MSV000094753 and at https://doi.org/doi:10.25345/C53X83X5J.

Supplemental data

Supplemental data for this article can be accessed online at https://doi.org/10.1080/21505594.2024.2399798