Antibiotic Treatment of Staphylococcus aureus Infection Inhibits the Development of Protective Immunity

ABSTRACT

Recurrent Staphylococcus aureus infections are common, suggesting a failure to elicit protective immunity. Given the emergence of antibiotic resistance, a vaccine is urgently needed, but there is no approved vaccine for S. aureus. While antibiotics are routinely used to treat S. aureus infections, their impact on the development of protective immunity is not understood. Using an established mouse model of S. aureus skin and soft tissue infection (SSTI), we observed that antibiotic therapy effectively resolved infection but failed to elicit protection against secondary (2°) SSTI. Key contributors to protective immunity, toxin-specific antibodies and interleukin-17A (IL-17A)-producing T cells, were not strongly elicited in antibiotic-treated mice. Delaying antibiotic treatment failed to resolve skin lesions but resulted in higher antibody levels after infection and strong protection against 2° SSTI, suggesting that the development of protective immunity requires a longer period of antigen exposure. We next investigated if combining α-hemolysin (Hla) vaccination with antibiotics during primary infection would both treat infection and generate durable protective immunity. This “therapeutic vaccination” approach resulted in rapid resolution of primary infection and protection against recurrent infection, demonstrating that concurrent vaccination could circumvent the deleterious effects of antibiotic therapy on elicited immune responses. Collectively, these findings suggest that protective immunity is thwarted by the rapid elimination of antigen during antibiotic treatment. However, vaccination in conjunction with antibiotic treatment can retain the benefits of antibiotic treatment while also establishing protective immunity.

INTRODUCTION

Staphylococcus aureus is the most common cause of skin and soft tissue infections (SSTIs) globally (1). S. aureus also frequently causes severe invasive infections, such as bacteremia, osteoarticular infections, pneumonia, and septic shock (2). The emergence of antibiotic resistance among infecting S. aureus isolates, most notably methicillin-resistant S. aureus (MRSA), has complicated treatment of these infections (3). Historically a leading cause of infections in hospitals, MRSA increasingly causes community-associated infections in otherwise healthy individuals (2–4). Recurrent S. aureus infections occur in up to 50% of children and adults within a year after SSTI (5, 6), suggesting that infection is not sufficient to elicit protective immunity against subsequent infections.

There is no licensed vaccine to prevent S. aureus infections, and several candidate vaccines have failed in clinical trials (7, 8). These failures highlight uncertainties in developing a successful vaccine, including a lack of understanding of the mechanisms of protective immunity, difficulty in identifying ideal vaccine antigens, and debate regarding the appropriate population to target for vaccination (9–12). The pore-forming toxin α-hemolysin (Hla) is a promising target because it is critical for pathogenesis and is highly protective in preclinical models, and Hla-specific antibody levels are correlated with protection against recurrent infection in children (6, 13, 14). In the absence of preventative strategies, antibiotics remain the mainstay of treatment of S. aureus SSTI. However, the impact of antibiotic treatment of S. aureus SSTI on the development of protective immunity is not clear.

We reported a mouse model of recurrent SSTI in which primary infection of BALB/c mice elicits protection against secondary infection (13). This protection was characterized by toxin-specific antibody and Th17 responses. In the current study, we investigated the effects of antibiotic treatment of SSTI on the development of protective immunity. We found that although early antibiotic treatment effectively resolved SSTI in BALB/c mice, treatment inhibited the development of protective antibody and T-cell responses following primary infection and subsequent protection against secondary infection. Importantly, early vaccination with an Hla-based vaccine during treatment restored protection against recurrent infection.

RESULTS

Antibiotic treatment improved resolution of S. aureus SSTI.

We modified our established mouse model of primary S. aureus SSTI (15) to determine whether treatment with the glycopeptide antibiotic vancomycin would resolve infection in BALB/c mice. We tested two doses—15 mg/kg of body weight (low dose) and 150 mg/kg (high dose)—each administered 30 min after inoculation with a bioluminescent S. aureus isolate. Mice treated with high-dose vancomycin did not develop dermonecrotic skin lesions (Fig. 1A and C). In contrast, mice treated with low-dose vancomycin developed dermonecrotic skin lesions, but the lesions were smaller than those of untreated mice. We observed rapid clearance of bacteria from the skin lesions by bioluminescence in both antibiotic groups, compared with untreated mice, but clearance was superior in mice treated with the high dose (Fig. 1B and D). These findings established that treatment of S. aureus SSTI shortly after inoculation with a single high dose of vancomycin effectively treated infection by enhancing bacterial clearance and preventing the development of dermonecrosis, so for the following experiments, we used the high-dose treatment.

FIG 1.

Antibiotic treatment decreased lesion size and bacterial burden following 1° SSTI. Mice treated with either high- or low-dose vancomycin had significantly smaller dermonecrotic skin lesions (A) and bacterial burdens (bioluminescence) (B) at the site of infection. Shown are pictures of representative (C) skin lesions and (D) bioluminescence from each group on different days post-primary SSTI. The control group was injected with PBS for the primary infection (consistent for all figures). n = 10/group. Data are presented as the mean ± standard error of the mean (SEM). Data were pooled from two independent experiments. Data were compared using two-way ANOVA with Tukey’s multiple-comparison test. **, P < 0.01; ****, P < 0.001; ns, not significant.

Antibiotic treatment of 1° SSTI inhibited development of protection against 2° SSTI.

In BALB/c mice, primary (1°) SSTI strongly protects against secondary (2°) SSTI (13). We hypothesized that antibiotic treatment after 1° SSTI, by virtue of rapid bacterial clearance, would prevent the development of protective immunity. To test this, we reinfected BALB/c mice 6 weeks following primary SSTI inoculation (±high-dose vancomycin treatment). Following reinfection, mice previously treated with high-dose vancomycin developed large dermonecrotic lesions, whereas untreated mice did not develop dermonecrosis, although subcutaneous abscesses were observed (Fig. 2A). In fact, the lesions of mice previously treated with high-dose vancomycin were similar in size to those in mice with primary SSTI (control group), indicating that high-dose vancomycin completely inhibits protection against secondary SSTI.

FIG 2.

Antibiotic treatment of 1° SSTI inhibited protective immunity against 2° SSTI. (A) Mice treated with antibiotics after 1° SSTI had larger dermonecrotic lesions following 2° SSTI than mice without antibiotic treatment. Vancomycin-treated mice produced significantly less antibodies against LukE (B), LukS-PV (C), and Hla (D) after 1° SSTI compared to the untreated group. Following 1° SSTI, untreated mice elicited strong IL-17A (E)- and IFN-γ (F)-secreting T-cell responses against LukE, LukS-PV, HLA, and HKSA. However, LukE-, LukS-PV-, HLA-, HKSA-specific IL-17A-secreting T-cell responses were significantly reduced in the vancomycin-treated group. Hla-, LukE-, and LukS-PV-specific IFN-γ-secreting T-cell responses were also reduced in the vancomycin-treated group, though not significantly. In panel A, n = 9/group except for the control (PBS→SSTI) (n = 5), in panels B to D, n = 10/group, and in panels E and F, n = 4/group except for the control (PBS) (n = 2). Data were pooled from two independent experiments (A to D) or one representative experiment of two repeats (E and F). Data were compared using two-way ANOVA with Tukey’s multiple-comparison test (A), one-way ANOVA with Tukey’s multiple-comparison test (B to D), and unpaired t test (E and F). Data are presented as mean ± SEM. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001; ns, not significant.

Previous studies have shown that protective immunity elicited by S. aureus is mediated by antibody and interleukin-17A (IL-17A) T-cell responses (3, 13, 14, 16–18). Therefore, we examined the antibody and T-cell responses in vancomycin-treated mice after primary infection. We hypothesized that the lack of protection in vancomycin-treated mice would be accompanied by decreased antibody and T-cell responses. Treatment of primary SSTI with vancomycin resulted in significantly lower antibody levels against leukocidin E (LukE), Panton-Valentine leukocidin S (LukS-PV), and Hla, compared with no treatment (Fig. 2B to D). In fact, vancomycin-treated mice had essentially undetectable antibody levels against LukE, LukS-PV, and Hla, similar to naive mice, indicating that antibiotic-mediated clearance of bacteria prevented the development of protective antibody responses. Primary SSTI in untreated mice elicited strong IL-17A- and interferon gamma (IFN-γ)-secreting T-cell responses against LukE, LukS-PV, Hla, and heat-killed S. aureus (HKSA) (Fig. 2E and F). However, LukE-, LukS-PV-, Hla-, and HKSA-specific IL-17A-secreting T-cell responses were significantly reduced in the vancomycin-treated mice. There was also a trend toward weaker Hla-, LukE-, and LukS-PV-specific IFN-γ-secreting T-cell responses in the antibiotic-treated mice, but the differences were not significant. Taken together, these results demonstrate that antibiotic treatment of S. aureus SSTI interferes with the development of protective immunity by inhibiting toxin-specific antibody and T-cell responses.

Delaying antibiotic treatment of S. aureus SSTI reduced efficacy.

We next investigated the impact of delayed antibiotic treatment on the resolution of primary SSTI in BALB/c mice. Delaying treatment to 3 h postinfection (hpi) resulted in small dermonecrotic lesions that were significantly smaller and resolved faster than those in untreated mice (Fig. 3A and C). In addition, lower bioluminescence was observed in the 3-hpi group than in untreated mice, indicating a smaller bacterial burden (Fig. 3B and D). Interestingly, mice treated 24 hpi developed large dermonecrotic lesions, similar in size to those in untreated mice. Bioluminescence in the 24-hpi group was similar to that in untreated mice, demonstrating that treatment 24 hpi did not lower the bacterial burden compared to that in untreated mice. Taken together, these findings demonstrate that delaying antibiotic treatment fails to resolve 1° SSTI. However, delaying treatment to 3 h postinfection, although not as effective as treatment immediately (30 min) after infection, decreased lesion size and enhanced bacterial clearance.

FIG 3.

Delayed antibiotic treatment resulted in larger lesion size and less effective bacterial clearance. (A and B) Mice treated with antibiotics 30 min post-1° SSTI had no dermonecrosis, but did have subcutaneous abscesses. Mice treated with antibiotics 3 h after infection had small dermonecrotic skin lesions and increased bioluminescence at the site of infection compared with untreated mice or mice treated 30 min postinfection. Antibiotic treatment 24 h following 1° SSTI resulted in dermonecrotic skin lesions (A) and bacterial burdens (B) similar in size to those of untreated mice. Shown are pictures of representative (C) skin lesions and (D) bioluminescence from each group on different days post-primary SSTI. n = 8/group. Data are presented as mean ± SEM. Data were pooled from two independent experiments. Data were compared using two-way ANOVA with Tukey’s multiple-comparison test. *, P < 0.05; ****, P < 0.0001; ns, not significant.

Delayed antibiotic treatment after 1° SSTI resulted in protection against 2° SSTI.

We next sought to understand if delaying treatment would enable the development of protective immunity against 2° SSTI. Mice that received vancomycin 3 hpi had smaller dermonecrotic lesions following 2° SSTI than those that were treated 30 min after infection (Fig. 4A). In contrast, mice that were treated 24 hpi were more strongly protected against 2° SSTI and developed lesions similar in size to those of untreated mice (Fig. 4A). Although not completely protected against 2° SSTI, the 3-hpi group still developed protective immunity.

FIG 4.

Delaying antibiotic treatment led to protection against 2° SSTI. (A) Mice treated with antibiotics 24 h after 1° SSTI developed small lesions following 2° SSTI that were similar in size to the lesions of untreated mice. Mice previously treated 3 h postinfection (hpi) developed larger lesions than were seen in untreated mice after 2° SSTI, but which were smaller than the lesions of mice treated 30 min postinfection (mpi) and control mice. Mice treated 24 hpi produced significantly more antibodies against LukE (B), LukS-PV (C), and Hla (D) after 1° SSTI than the mice treated 30 mpi. Treatment 3 hpi resulted in increased antibody levels against Hla (D) compared to 30 mpi, but not against LukE and LukS-PV. n = 8/group. Data are presented as mean ± SEM. Data were pooled from two independent experiments. Data were compared using one-way (B to D) and two-way (A) ANOVA with Tukey’s multiple-comparison test. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001; ns, not significant.

The protection observed against 2° SSTI in mice treated with delayed antibiotics during 1° SSTI suggested that protective immune responses were generated after 1° SSTI. To test this hypothesis, we measured antibody levels against LukE, LukS-PV, and Hla following the resolution of 1° SSTI. Delaying antibiotic treatment to 24 h resulted in high antibody levels against all three antigens, similar to untreated mice (Fig. 4B to D). The mice treated 3 hpi had higher antibody levels against Hla, compared with mice that received treatment in 30 min, but anti-Hla levels were lower than in untreated mice. There were no significant increases in antibody levels against LukE and LukS-PV in the 3-hpi group. Delaying treatment of antibiotics to 3 hpi resulted in an intermediate level of protection against 2° SSTI between mice that were rapidly treated and those in which treatment was delayed 24 hpi. The incomplete development of protective immunity following antibiotic treatment 3 hpi suggests that a longer duration of antigen exposure is necessary to develop protection.

“Therapeutic vaccination” effectively treated infection and established protective immunity.

To reach optimal protection against both primary and recurrent infections, we investigated if combining α-hemolysin (Hla) vaccination with antibiotic treatment during primary infection would both treat infection and generate durable protective immunity. Three hours post-1° SSTI, we treated mice with vancomycin in addition to vaccinating mice with a dose of Hla_H35L (19) and boosting with Hla_H35L 3 weeks following 1° SSTI. This “therapeutic vaccination” approach resulted in decreased lesion size and decreased bacterial burden compared to untreated mice following 1° SSTI (Fig. 5A and B). Mice that only received the vaccination developed slightly larger lesions after 1° SSTI, compared with untreated and antibiotic-treated mice (Fig. 5A), but there were no significant differences in the bacterial CFU between the vaccinated and untreated groups (Fig. 5B). Following the resolution of 1° SSTI, we measured antibody levels against LukE, LukS-PV, and Hla. Both groups that received antibiotic treatment had decreased antibody levels against LukE and LukS-PV compared to untreated groups. The vaccinated groups had a 2-fold increase in antibody levels against Hla, compared with unvaccinated mice (Fig. 5E), suggesting that robust protection against 2° SSTI would be observed in vaccinated mice. As expected, mice receiving vaccination did not develop lesions following 2° SSTI (Fig. 5F). In addition, significantly fewer bacteria were recovered from the vaccinated groups compared with the control group as well as the treated group that did not receive the vaccine (Fig. 5G). There was no significant difference in bacteria recovered between the untreated group and the vaccinated groups. Vaccination alone generated protective antibodies following the primary infection and effectively cleared the bacteria during the secondary infection. However, antibiotic treatment was necessary to reduce infection severity after 1° SSTI. Thus, “therapeutic vaccination” is optimal as it generated a robust protective immunity against further S. aureus infection in addition to effectively treating primary infection.

FIG 5.

Combining vaccination (Vax) with antibiotics (vancomycin [Vanco]) treated 1° SSTI and generated protection against 2° SSTI. Mice treated with antibiotics after 1° SSTI (+Vanco/−Vax and +Vanco/+Vax) have significantly decreased dermonecrotic lesions (A) and bacterial luminescence (B) compared with untreated mice (−Vanco/−Vax). Untreated mice that received the vaccination (−Vanco/+Vax) developed similar-sized lesions to untreated mice and had similar levels of bacterial luminescence. Six weeks post-1° SSTI, vaccinated-only mice (−Vanco/+Vax) produced significantly higher antibody levels against LukE (C), LukS-PV (D), and Hla (E) than antibiotic-only-treated mice (+Vanco/−Vax). Mice receiving concurrent vaccination and treatment (+Vanco/+Vax) produced significantly higher antibody responses against Hla compared to both untreated and antibiotic-treated mice (+Vanco/−Vax). After 2° SSTI, both groups of vaccinated mice did not develop dermonecrotic lesions (F) and had significantly less bacteria recovered than control and antibiotic-only-treated mice (G). n = 10/group. Data are presented as mean ± SEM. Data were pooled from two independent experiments. Data were compared using one-way (C to E and G) and two-way (A, B, and F) ANOVA with Tukey’s multiple-comparison test. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001; ns, not significant.

DISCUSSION

Infections caused by S. aureus remain an important cause of hospitalization and morbidity in otherwise healthy children (2, 4). We adapted our established mouse model of S. aureus SSTI by adding treatment with the antibiotic vancomycin to understand the impact of treatment on the development of protective immunity. We found that treatment with high-dose vancomycin shortly after infection effectively resolved 1° SSTI, but treated mice did not develop protection against 2° SSTI normally observed in this model. Delaying antibiotic treatment resulted in less effective clearance of primary infection but allowed for the development of protective immunity against subsequent infections. We developed a strategy to resolve primary infection while also generating robust protection by “therapeutic vaccination,” in which infected mice were concurrently treated with vancomycin and vaccinated with an inactivated α-hemolysin (Hla_H35L). This approach resulted in rapid resolution of 1° SSTI and protection against 2° SSTI, suggesting that this may be an effective strategy during acute infection to both treat infection and prevent recurrence.

Primary infection in mice treated with vancomycin 30 min postinfection (mpi) did not elicit protection against secondary infection. We and others have reported that protective immunity against S. aureus is dependent on antibodies and IL-17A-producing T cells directed against key S. aureus toxins, including Hla, LukE, and LukS-PV (3, 13, 14, 16, 17). Primary infection in vancomycin-treated mice failed to elicit strong antibody responses against Hla, LukE, and LukS-PV and Hla-, LukE- and LukS-PV-specific Th17 responses, suggesting antibiotic treatment prevents protective immunity by impairing both antibody and T-cell-mediated immunity. Antibiotic treatment of other bacterial infections, including Salmonella and Chlamydia, results in a similar inhibition of CD4⁺ T-cell responses (20–22). However, Th1 CD4 T cells mediate protection in Salmonella and Chlamydia infections, whereas Th17 CD4 T cells mediate protection in S. aureus infections, indicating CD4⁺ T-cell-mediated protective immunity are broadly impacted by antibiotic treatment. Benoun et al. (22) suggested that antibiotic treatment may free antigens from dead bacteria but not create a sustained period of antigen presentation, negatively impacting development of CD4 memory responses. Reinforcing this conclusion, previous research has shown that antigen persistence is required for differentiation and proliferation of CD4⁺ T-cells (23, 24). Delaying vancomycin treatment to 3 h post-S. aureus 1° SSTI partly recovered antibody levels, and delaying treatment to 24 h completely recovered antibody levels, resulting in complete protection against 2° SSTI. This suggests that early treatment effectively cleared bacteria and eliminated antigen persistence, but delayed treatment resulted in longer antigen persistence leading to the subsequent generation of robust protective immunity. Similarly, delaying antibiotic intervention in Salmonella infection resulted in development of protection (21). However, this protection required 1 to 2 weeks of antigen exposure, suggesting the duration of antigen persistence needed to develop protection differs depending on what antigen is involved. Together, our results build upon current research proposing that antibiotic treatment, by rapidly eliminating bacterial antigen, interrupts the development of CD4⁺ memory T-cells.

Although delaying antibiotic treatment resulted in development of protection against 2° SSTI, delayed treatment was less effective in treating 1° SSTI. In contrast to treatment directly after infection (30 mpi), delaying treatment to 3 or 24 hpi resulted in dermonecrotic lesions. Delaying treatment to 24 hpi did not impact lesion size compared to that in untreated mice, demonstrating that, in this model, delayed therapy ineffectively treated dermonecrosis. This is consistent with strong evidence that early antibiotic therapy is critical for resolving systemic bacterial infections in humans (25, 26), although it is unclear whether rapid treatment is equally important in SSTI. The quick window for antibiotic treatment effectiveness in this model suggests that this mechanism involves the innate immune system, which is the first line of defense against S. aureus SSTIs and involves neutrophils, interleukin-1β (IL-1β), and pattern recognition receptors (3). We reported that early neutrophil recruitment and toxin neutralization are critical to determining the fate of infection within the first 24 h of infection (27). Our current findings further support the idea that, in this model, the fate of skin and soft tissue infections is determined within the first 24 h after infection. Future work will examine the impact of antibiotic treatment on local innate immune responses.

The pore-forming toxin Hla is a promising vaccine candidate against S. aureus infection. Active or passive immunization with an Hla mutant (Hla_H35L) strongly protected mice against S. aureus SSTI (14). Hla-specific antibody responses also partially mediate protection against recurrent S. aureus SSTI in mice (13). In addition, anti-Hla IgG levels in children correlated with a lower risk of recurrent S. aureus infections, suggesting that the importance of Hla-specific immunity is not limited to mouse models (6). Therefore, to both treat infection and generate durable protective immunity, we combined Hla_H35L vaccination with vancomycin treatment during primary infection. This “therapeutic vaccination” approach resulted in rapid resolution of primary dermonecrosis compared with untreated mice. Importantly, therapeutic vaccination, even in vancomycin-treated mice, elicited anti-Hla antibody and strongly protected against recurrent infection. Thus, our findings demonstrate that concurrent vaccination, by providing a prolonged antigen stimulus, could circumvent the deleterious effects of antibiotic therapy on elicited immune responses in S. aureus SSTI. A recent report demonstrated increased survival of mice in invasive S. aureus infections (sepsis and pneumonia) with combined treatment of antibiotics and a monoclonal antibody cocktail, including Hla (28). Therefore, this approach may be effective against multiple S. aureus infectious syndromes. Because rapid administration of antibiotics is a cornerstone of effective treatment of these potentially devastating infections, we propose that therapeutic vaccination is a promising strategy that will enable both effective treatment of ongoing infection and generation of protective immunity against recurrent infection. This is particularly important because recurrent infections occur in at least 50% of children within a year after SSTI (5, 6). Moreover, because there has been much debate about a suitable target population for any candidate vaccine, these findings suggest that children with SSTI, who by definition are a high-risk population, may represent an ideal target population that is both easily identified and readily amenable to vaccination.

The primary limitation to this study is that the relevance of the mouse model to human infection is uncertain. Whereas the mice used in these studies had no known previous exposure to S. aureus, nearly all humans have evidence of some exposure to the bacterium; therefore, whether preexisting immunity would impact these responses is not clear. In addition, we elected to treat mice shortly after infection (3 to 24 h), but most individuals present for medical care days after the onset of symptoms. This raises the question of whether more delayed treatment in clinical practice would also inhibit protective immune responses. Moreover, we tested only vancomycin in our mouse models. Whether these findings would extend to other antibiotics commonly used to treat S. aureus infections is not known. Finally, whereas S. aureus SSTI elicits protective immunity in mouse models, the frequency of recurrent infection in humans suggests that infection does not reliably elicit protection in humans. Clearly, future studies in human populations are necessary in order to enable translation of our findings into clinical practice.

In conclusion, our findings advance the understanding of the impact of antibiotic treatment on protective immunity. We demonstrated that by minimizing the period of antigen exposure, antibiotics inhibit the protective IL-17A T-cell and antibody responses. Vaccination with Hla in conjunction with antibiotic treatment extends the period of antigen exposure, thus overcoming the inhibitory effects on the development of protective immunity, providing evidence for the efficacy of vaccination at the time of antibiotic treatment.

MATERIALS AND METHODS

Bacterial preparation.

The virulence of the S. aureus USA300 isolate 923 in mouse models has been reported (29). A bioluminescent 923 isolate was a gift from Susan Daum (University of Chicago [currently at NIH Center for Scientific Review]); this was constructed by cloning the luxABCDE operon from pXen1 into 923 under the control of the pdh (pyruvate dehydrogenase) promoter for constitutive expression. Bacterial isolates were recovered from frozen stock cultures by plating on tryptic soy agar (TSA) overnight at 37°C. A bioluminescent colony was selected, inoculated in tryptic soy broth (TSB), and grown overnight in a shaking incubator at 250 rpm at 37°C. On the day of infection, the overnight culture was diluted 1:100 in tryptic soy broth. The culture was harvested 3 h after incubation in 37°C, and the bacteria were pelleted by centrifugation, washed, and resuspended in sterile phosphate-buffered saline (PBS) to an adjusted concentration of 1.5 × 10⁷ CFU/50 μL.

Mouse model of S. aureus SSTI.

Animal experiments were approved by the Animal Care and Use Committee at the Abigail Wexner Research Institute at Nationwide Children’s Hospital. As previously reported (13), BALB/c mice were sedated, shaved, followed by application of Nair (Carter-Wallace, Inc., New York, NY) to remove hair, and inoculated with 50 μL (∼1.5 × 10⁷ CFU) of S. aureus or PBS for a negative control by subcutaneous injection. Following infection, skin lesions were photographed and measured daily for 2 weeks, and the bacterial burden at the site of infection was quantified by measuring bioluminescence using Xenogen IVIS Spectrum (PerkinElmer). Secondary infection was performed 6 weeks following day 0 of primary infection. Groups of mice were treated with 15 mg/kg vancomycin (low dose), 150 mg/kg vancomycin (high dose), or PBS for 30 min, 3 h, or 24 h post-1° infection via intraperitoneal injection. For the 3-h-postinfection (hpi) group and 24-hpi group, antibiotic treatment was continued once a day for 7 days.

Vaccination.

A plasmid for expression and purification of Hla_H35L was a gift from Juliane Bubeck Wardenburg (Washington University—St. Louis) (30). The vaccine was prepared by adding 10 μg of Hla_H35L protein adjuvanted with Al(OH)₃ at a final concentration of 0.1% in a total volume of 200 μL. Mice were vaccinated subcutaneously on the back 3 h after infection, and a booster dose was administered 3 weeks following infection.

Antibody quantification.

Serum was collected from BALB/c mice 4 to 6 weeks after primary infection. Indirect enzyme-linked immunosorbent assay (ELISA) was performed to quantify total IgG antibodies against alpha-hemolysin (HLA), leukocidin E (LukE), or Panton-Valentine leukocidin S (LukS-PV) (19). Ninety-six-well plates (Costar, Corning, Inc.) were coated with purified antigen (5 μg/mL), followed by incubation with mouse serum, diluted 1:50 with PBS, and subsequently with alkaline phosphatase-conjugated anti-mouse IgG (1:5,000, AffiniPure; Jackson ImmunoResearch). The plates were washed and incubated with p-nitrophenyl phosphate (Sigma), and the absorbance was measured at 405 nm using BioTek Synergy HTX plate reader.

Quantification of T-cell responses.

Splenocytes from BALB/c mice were excised 4 weeks after primary infection and were plated in a 96-well enzyme-linked immunosorbent spot (ELISpot) plate, previously coated with anti-IL-17A or IFN-γ (19). Cells were incubated with LukE, LukS-PV, HLA (1 μg/mL), or heat-killed S. aureus (HKSA; 5 × 10⁵ CFU/well) for 24 h, followed by washing and incubation with biotin-conjugated anti-IL-17A or IFN-γ antibodies. The plates were washed and incubated in avidin-horseradish peroxidase (HRP) antibody. Subsequently, spots were developed using the 3-amino-9-ethylcarbazole (AEC) substrate kit (BD Biosciences) and counted using an Immunospot series 1 analyzer (Cellular Technology).

Lesion CFU.

The skin lesions (or subcutaneous abscesses) were excised 7 days following secondary SSTI and homogenized in order to quantify the entire bacterial burden in the skin and subcutaneous tissue. The homogenates were serially diluted, and four dilutions (10⁻², 10⁻³, 10⁻⁴, and 10⁻⁵) were plated on TSA plates. The plates were incubated overnight at 37°C. The colonies were counted the following day, and the numbers of CFU/lesion were quantified.

Data analysis.

Data were compared using one-way analysis of variance (ANOVA) with Tukey’s multiple-comparison test, unpaired t test, and two-way ANOVA with Tukey’s multiple-comparison test where appropriate. Differences were considered significant when P was <0.05. All data were analyzed using GraphPad Prism.