Evaluation of Genetically Inactivated Alpha Toxin for Protection in Multiple Mouse Models of Staphylococcus aureus Infection

Staphylococcus aureus is a major human pathogen and a leading cause of nosocomial and community-acquired infections. Development of a vaccine against this pathogen is an important goal. While S. aureus protective antigens have been identified in the literature, the majority have only been tested in a single animal model of disease. We wished to evaluate the ability of one S. aureus vaccine antigen to protect in multiple mouse models, thus assessing whether protection in one model translates to protection in other models encompassing the full breadth of infections the pathogen can cause. We chose to focus on genetically inactivated alpha toxin mutant HlaH35L. We evaluated the protection afforded by this antigen in three models of infection using the same vaccine dose, regimen, route of immunization, adjuvant, and challenge strain. When mice were immunized with HlaH35L and challenged via a skin and soft tissue infection model, HlaH35L immunization led to a less severe infection and decreased S. aureus levels at the challenge site when compared to controls. Challenge of HlaH35L-immunized mice using a systemic infection model resulted in a limited, but statistically significant decrease in bacterial colonization as compared to that observed with control mice. In contrast, in a prosthetic implant model of chronic biofilm infection, there was no significant difference in bacterial levels when compared to controls. These results demonstrate that vaccines may confer protection against one form of S. aureus disease without conferring protection against other disease presentations and thus underscore a significant challenge in S. aureus vaccine development.


Introduction
Staphylococcus aureus is a leading cause of both nosocomial and community-acquired infections in the United States [1,2]. S. aureus infections range from superficial skin infections to life-threatening, invasive illnesses such as sepsis, osteomyelitis, and endocarditis. This bacterium colonizes the anterior nares of approximately 28% of the population [3], and infections can be caused by the colonizing strain [4][5][6]. S. aureus features an extensive and highly redundant repertoire of virulence factors that allow it to colonize and damage the host, as well as evade host immune responses and cause a variety of diseases in multiple host compartments. The profile of expressed proteins exhibited by this pathogen can vary according to growth state (planktonic vs biofilm) [7][8][9]. Therefore, the potential exists for antigens expressed in one type of infection (e.g., acute infections such as sepsis) to not be present in other types of infection, such as chronic biofilm infections.
For an S. aureus vaccine to be broadly applicable, it would need to protect against a variety of S. aureus infections. Thus far, however, most protective antigens identified in the literature were tested in one or only a few animal models of S. aureus disease that do not represent the full breadth of diseases caused by this organism. In this work, we wished to determine the protective capability of a single antigen in multiple models of S. aureus disease that encompass a broad spectrum of bacterial growth-from infection with planktonic S. aureus to biofilm infection, from acute to chronic disease, and from superficial infection limited to the skin to invasive systemic illness. For each model, we administered the antigen using the same dose, same dosing schedule, and same adjuvant in order to mimic how vaccines are actually formulated and used. We also used the same challenge strain in each model. For this study, we chose to use a genetically inactivated form of alpha hemolysin (HlaH35L) [36] as the vaccine antigen. Immunization with this antigen has previously been shown to decrease mortality and lung colonization in a mouse pneumonia model [32], decrease lesion size and lessen dermonecrosis in a murine skin infection model [30], and also lessen lethality in a peritonitis model [37]. A truncated Hla was very recently shown to elicit protection in both pneumonia and bacteremia models [38]. A heat-inactivated form of alpha-toxoid has also been shown to protect against corneal damage in a S. aureus keratitis model [39]. However, these studies used different antigen preparations, different dosing, and different adjuvants, which do not reflect the likely real world scenario in which a single vaccine formulation would be used and administered on a single dosing schedule.
In this study, we tested purified, recombinant HlaH35L in three mouse models of USA300 infection: systemic infection (IV), skin and soft tissue infection (SSTI), and chronic prosthetic implant infection (PII). Utilizing the same antigen, dosing regimen, route of immunization, adjuvant and the same bacterial challenge strain in all three models allowed for evaluation of potential variation in protective capability of the antigen against different disease forms. Results using these three divergent models indicated a differential ability of recombinant HlaH35L to protect against staphylococcal illnesses that are at different ends of the infection spectrum.

Ethics Statement
For all animal studies, protocols were reviewed and approved by the Institutional Animal Care and Use Committees (IACUC) of the Center for Biologics Evaluation and Research (Bethesda, MD; permit number 2010-04) or the University of Maryland School of Medicine (Baltimore, MD; permit number 09-07-02). All surgeries were performed under ketamine/xylazine anesthesia, and all efforts were made to minimize animal suffering. Animals were sacrificed at the time points indicated below using CO 2 inhalation.

Bacterial Strains, Plasmids, and Reagents
S. aureus USA300 strain NRS384 (NARSA) was used as a PCR template for cloning. PCR primers were generated by the CBER Core Facility (Bethesda, MD). Chemi-competent TOP10 Escherichia coli cells and the pTRChis2c expression plasmid were obtained from Invitrogen (Carlsbad, CA). Chemi-competent BL21 E. coli were generated in our laboratory as described by Mandel and Higa [40]. S. aureus strain SAP149, which is NRS384 expressing a modified lux operon from Photorhabdus luminescens on an integrated plasmid (unpublished data), was used for all challenge experiments. All reagents were obtained from Thermo-Fisher Scientific (Waltham, MA) unless noted.

Cloning, Expression, and Purification of HlaH35L
In order to generate non-toxic Hla, the gene was reconstructed by joining together PCR fragments of hla using BsaI as described by Stemmer and Morris [41]. The CAC codon at position 35 was altered to CTC to change histidine to leucine. The primers used are shown in Table 1, with the codon change indicated in the DSIE primer. A C-terminal six-histidine tag and stop codon were added in frame with the coding sequence. The PCR fragments were digested with BsaI and cloned into pTRChis2c (Invitrogen) downstream of the lac operon. Ligated vectors were then transformed into E. coli TOP10 cells. Positive transformants were selected in the presence of ampicillin, and mutants were confirmed by sequencing. Plasmids were then transformed into E. coli BL21 for gene expression.
For production and purification of recombinant HlaH35L, LB cultures were supplemented with 50 mg/ml ampicillin and inoculated 1:100 with an overnight growth of BL21/HlaH35L. Cultures were incubated at 37uC with shaking to an optical density at 600 nm (OD 600 ) of 0.5, and 1 mM IPTG was added. Incubation continued for four hours, at which point bacteria were harvested by centrifugation.
HlaH35L was purified through affinity chromatography. Briefly, bacterial pellets were resuspended in lysis buffer (50 mM NaH 2 PO 4 , 0.5 M NaCl, pH 8.0, supplemented with 1 mg/ml lysozyme) and incubated on ice for 30 minutes. Resuspended pellets were sonicated and cellular debris removed by centrifugation at 40006g for 15 minutes. Lysates were then applied to Ni-NTA columns (His-Pur, ThermoFisher) and incubated on a rotator for 60 minutes. The column was washed four times with wash buffer (50 mM NaH 2 PO 4 , 0.5 M NaCl, 20 mM imidazole, pH 8.0), and recombinant protein was eluted with 50 mM NaH 2 PO 4 , 0.5 M NaCl, 250 mM imidazole, pH 8.0. The eluted protein was concentrated and diafiltered into 20 mM Tris, 50 mM NaCl, 2 mM EDTA, pH 8.0 using AmiconH Ultra 15 ml 10,000 MWCO Centrifugal Filters (ThermoFisher). Purity was confirmed by SDS-PAGE, and protein was quantified using BCA (Pierce, Rockford, IL) following the manufacturer's instructions. Recombinant HlaH35L was stored at 270uC until use.

Immunization
For all three challenge studies, animals were immunized in the following manner: Purified, recombinant HlaH35L was thawed and subsequently diafiltered into phosphate-buffered saline (PBS, Quality Biological, Gaithersburg, MD) using AmiconH Ultra 15 ml Centrifugal Filters (10,000 MWCO). HlaH35L was then quantified by BCA and resolved via SDS-PAGE to confirm

Enzyme Linked Immunosorbant Assay (ELISA)
Ninety-six well plates (MaxiSorp, ThermoFisher) were coated with 1 mg/ml HlaH35L diluted in PBS by incubating the plates at 4uC overnight. The plates were washed three times with 0.9% NaCl, 0.05% Tween-20 and then blocked with 2% (w/v) nonfat dry milk in PBS for two hours at 37uC. The blocked plates were washed again, and sera from immunized or control mice were diluted serially in dilution buffer (PBS with 2% fetal bovine serum (Invitrogen) and 0.05% Tween- 20), and incubated at RT for two hours. After washing, anti-mouse IgG conjugated to horseradish peroxidase (HRP; KPL Inc., Gaithersburg, MD) was diluted in dilution buffer, added to the plates, and incubated at RT for two hours. Plates were washed again and reactivity was analyzed by adding ABTSH ELISA HRP Substrate (KPL, Inc.) and incubating for ten minutes at RT. The reaction was stopped through the addition of ABTSH Stop Solution (KPL), and optical densities were read at 405 nm (OD 405 ) on a VersaMax Microplate Reader (Molecular Devices, Sunnyvale, CA). Data were plotted using GraphPad Prism software (Version 5, GraphPad Software, La Jolla, CA) with a 4-parameter logistic (4PL) curve fit of OD (Y axis) versus logarithm of the reciprocal serum dilution (X axis). Titers are defined as the effective dilution that gives 50% of the maximal absorbance (ED 50 ) for each sample.

Skin and Soft Tissue Challenge
On the day of challenge, an overnight growth of S. aureus SAP149 was used to inoculate four 50 ml aliquots of Tryptic Soy Broth (TSB; MP Biomedicals, Solon, OH), supplemented with 10 mg/ml chloramphenicol (Sigma Aldrich, St. Louis, MO), at a dilution of 1:50. The cultures were grown at 37uC with shaking until an OD 600 of 0.8 was reached, and then centrifuged at 40006g for 15 minutes. All four bacterial pellets were resuspended together in 50 ml PBS, and bacterial counts were obtained using a Petroff-Hausser Counting Chamber (Hausser Scientific Partnership, Horsham, PA). The bacteria were centrifuged again at 40006g for 15 minutes and then resuspended in an appropriate volume of PBS to give a concentration of 1610 11 CFU/ml. The bacterial count was confirmed with serial dilution and plating on Tryptic Soy Agar (TSA; MP Biomedicals).
Immunized and control mice were anesthetized with 2 mg Ketamine (Ketaject, Phoenix Pharmaceutical, St. Joseph, MO) and 0.1 mg Xylazine (AnaSed, Akorn, Decatur, IL) administered IP. Images were obtained of each animal's left ear using a Nikon SMZ 745T dissecting microscope. Left ears were sterilized with 70% ethanol, and then 1610 9 CFU of S. aureus SAP149 were administered to the left ear of each mouse in a volume of 10 ml, using a Morrow Brown needle as described by Prabhakara et al. [42]. Subsets of animals were photographed on days four and seven following challenge and then sacrificed. The left ear was excised, minced, and homogenized in sterile PBS. Serial dilutions were plated on TSA. A second group of mice was imaged serially on days four, seven, and 14 in order to follow the infection progression over time in the same animals. These mice were sacrificed, and ears were harvested and plated, on day 14 postchallenge.

Intravenous (IV) Challenge
An overnight culture was started using a single hemolytic colony of S. aureus SAP149 from a one-day-old 5% blood agar plate (Becton Dickinson, Sparks, MD) containing 10 mg/ml chloramphenicol and incubated at 37uC overnight. On the day of the challenge experiment, TSB containing 10 mg/ml chloramphenicol was inoculated with a 1:100 dilution of the overnight growth. The culture was incubated at 37uC with shaking until the OD 600 reached approximately 0.63. The cell concentration calculation is based on previous experimentation in which it was determined that at an OD 600 of 0.63 the concentration of the growth is approximately 3.66610 8 CFU/ml (data not shown). The cells were then pelleted by centrifugation at 40006g for 10 minutes. The pellet was re-suspended in 1X PBS (Quality Biological) and brought to a concentration of 3610 7 CFU/ml. The challenge dose concentration was verified by serial dilution and plating on TSA. BALB/c mice were challenged via the lateral tail vein with 100 ml of the bacterial suspension two weeks following the second immunization. Two weeks following challenge, the mice were euthanized and kidneys were harvested, homogenized in PBS, and plated on TSA.

Prosthetic Implant Infection
The prosthetic implant model was conducted in a manner similar to that previously described [43]. Briefly, on the day of challenge, an overnight culture of S. aureus SAP149 was centrifuged at 40006g for 10 minutes. The pellet was washed with PBS, counted with a Petroff-Hausser Counting Chamber, and then resuspended to a concentration of 1610 6 CFU/ml. Ten C57BL/6 mice per experimental group received implants. Mice were anesthetized via IP injection of 100 mg/kg Ketamine (KetasetH; Fort Dodge Laboratories, Inc., Fort Dodge, Iowa) and 10 mg/kg Xylazine (Rugby Laboratories, Inc., Rockville Center, NY). The left leg of each mouse was cleansed with povidone iodine and rinsed with 70% ethanol before surgical implantation of an autoclaved 0.25 mm insect pin into the tibia (Fine Science Tools, Foster City, CA) as described [43,44]. The implanted pin was then inoculated with 1 ml of the S. aureus SAP149 suspension. Three weeks following implantation, mice were sacrificed and the left tibiae were harvested, cut into small pieces, and placed in 300 ml of sterile 0.85% saline per 100 mg of bone. Bones were then homogenized using a Polytron PT 1200 handheld homogenizer (Kinematica, Bohemia, NY) at 25,000 rpm, and serial 10-fold dilutions of bone homogenates were plated on TSA II, 5% defbrinated sheep's blood agar plates to enumerate viable S. aureus per gram bone, as previously described [43].

Statistical Analysis
For all challenge experiments, CFU levels were compared between HlaH35L-immunized animals and adjuvant-only control animals using GraphPad Prism. Differences in CFU levels were evaluated using the untailed Student's t test. For IV challenge experiments, the data from two identical experiments run on separate days were combined. ELISA titers were defined as the ED 50 for sera pooled within a group from each experiment. ED 50 values were compared across challenge experiments using ANOVA.

HlaH35L Immunization Decreases Disease Severity and Bacterial Colonization in a Skin and Soft Tissue Infection Model of S. aureus Infection
In order to evaluate the effect of HlaH35L immunization in an SSTI model of S. aureus infection, we challenged immunized or control mice with 1610 9 CFU S. aureus SAP149 in the left ear two weeks after the second immunization. This model was developed by Prabhakara et al. [42] and leads to a self-limited, localized infection that generates a necrotic lesion. The challenge was conducted using a Morrow Brown needle such that the bacteria do not penetrate past the upper epidermal layer of the skin. We followed several animals over time in order to monitor lesion formation and subsequent tissue damage (Figure 1). In control animals that received AlhydrogelH and CpG adjuvant without antigen, we saw that mice lost large portions of the affected ears as the infection progressed. However, in HlaH35L-immunized animals, tissue loss was minimal or did not occur (Figure 1). Animals were euthanized at days four and seven post-challenge and serial dilutions of the homogenized, affected ears were plated. Animals receiving HlaH35L with AlhydrogelH and CpG had significantly lower levels of S. aureus SAP149 present in the challenged ears than did control mice receiving adjuvant alone at both time points (Figure 2A and 2B). By day 14 post-challenge, all animals had colonization levels approaching the limit of detection ( Figure 2C); this result is not unexpected, as this SSTI model is self-limiting [42]. These results were reproducible over two independent experiments. These data indicate that HlaH35L immunization leads to a lesser infection with lower bacterial levels and nominal tissue damage when compared to controls.

HlaH35L Immunization Decreases Bacterial Colonization of Kidneys in a Systemic Model of S. aureus Infection
To test whether HlaH35L immunization lessens S. aureus systemic infection, mice were immunized with HlaH35L adjuvanted with AlhydrogelH and CpG, or adjuvant alone as a control. Mice were then challenged with 3610 6 CFU of S. aureus SAP149 via the lateral tail vein, resulting in bacteria being introduced directly into the bloodstream. When kidneys were harvested two weeks following challenge, we noted a decrease in bacterial levels in the immunized mice. The experiment was run twice; one experiment yielded statistically significant results, whereas the other did not (Figure 3). Combining data from the two experiments to increase effective sample size and therefore increase statistical power demonstrated that immunization with HlaH35L significantly reduced kidney colonization after IV challenge when compared to controls (P = 0.006).

HlaH35L Immunization does not Lessen Bacterial Colonization of Infected Tibiae in a Model of Chronic Biofilm Infection
In order to determine if HlaH35L immunization is able to decrease levels of S. aureus in a model of chronic biofilm infection, we challenged immunized and control mice by introducing a metal implant and then inoculating the implant with 1610 3 CFU S. aureus SAP149. Twenty-one days post-challenge, levels of S. aureus in the tibiae of immunized mice were determined and compared to that of mice having received AlhydrogelH and CpG alone. The experiment was performed in duplicate. There was no significant difference in bacterial colonization of affected tibiae observed between groups ( Figure 4A and B) in either experiment. These data suggest that HlaH35L immunization does not lessen severity of chronic S. aureus biofilm infection.
We should note that this model utilizes C57BL/6 mice instead of BALB/c mice, because BALB/c mice, unlike what is seen in clinical cases of S. aureus prosthetic implant infection, can spontaneously clear the infection even if a biofilm is first grown on the pin prior to implantation. In order to determine the importance of mouse strain in the other models, we performed ELISA on pools of sera from immunized mice for each of the three challenge models used in this study. We found no significant difference between the antibody titers (expressed as ED 50 s) for HlaH35L-immunized animals from the three model groups (data not shown; P = 0.874, ANOVA). Moreover, we also performed S. aureus SSTI challenge using C57BL/6 mice and found that

Discussion
While a good deal of research into potential S. aureus vaccine candidates has been conducted, the vast majority of publications detailing such investigations have described testing in a single animal model of infection. While an antigen may elicit protection against single, or even similar, disease types, broad protection against the wide variety of infections that a dynamic pathogen such as S. aureus causes may be difficult to achieve. Therefore, testing potential candidates in animal models that encompass both acute infections, e.g. skin and soft tissue infections or sepsis, and chronic infections such as osteomyelitis, would be informative.
In this study, we undertook the evaluation of a single protein antigen that has been published in several works to generate a protective response in animals against S. aureus infection and that has been included in S. aureus vaccines in clinical trials [45]. Inactivated alpha toxin has been shown to reduce bacterial populations and/or protect mice from lethality against pneumonia [32], skin infection [30], and, recently, sepsis [38] and peritoneal abscess models [37]. However, the toxin has not been tested against chronic, biofilm-associated infections. Moreover, those studies used different dosing regimens, adjuvants, and challenge strains. For this work, we eliminated potential laboratory-tolaboratory variability by using the same protein preparation for immunizations and the same challenge strain for protection studies. We also used a single dosing regimen and the same adjuvant for all studies to better mimic how a vaccine would actually be used. We found that HlaH35L combined with AlhydrogelH and CpG administered to mice that were then subjected to S. aureus SSTI or systemic challenge led to lower infection levels than did adjuvant alone. The difference between vaccine and control groups was statistically robust in the SSTI model. The difference between the two groups for the systemic infection model, while statistically significant when data from two experiments were combined, was less robust. Possibly, HlaH35L is only marginally protective in our systemic infection model, and the more moderate level of protection required a greater number of animals to reach statistical significance.
We found that the same immunization strategy did not elicit protection in a prosthetic implant infection model.   with bacterial levels constant after seven days. The infection also persists long term (.49 days), and is unable to be cleared by antibiotics or the host immune response [43]. We found that there was no significant difference in the level of S. aureus in infected tibiae from either HlaH35L-immunized or control mice. This finding may not be surprising, as microarray analyses have indicated that hla is down-regulated during biofilm growth [7,46]. It is also possible that alpha toxin may be present at some level within the mature biofilm, but that it plays no role in such an environment; alternatively, alpha toxin may not affect biofilm development or bacterial colonization of bone. In these cases, antibodies capable of toxin neutralization would not be expected to be effective in clearing the biofilm-associated bacteria. Regardless of the underlying reason why an HlaH35L vaccine was not protective in this model, this failure indicates that antigens that are chosen based on their effectiveness in other types of staphylococcal infection, such as systemic disease, may not be useful in the case of chronic, biofilm-associated infections.
Because expressed virulence factors vary among different manifestations of S. aureus disease, it is important to determine if a vaccine antigen may be broadly protective. To our knowledge, no group has described testing of individual S. aureus antigens in both acute and chronic infection models, within a laboratory using the same protein preparation, same dosing regimen, same adjuvant and the same challenge strain. Our work indicates that an antigen that is protective against one disease form may not be protective against other forms. In particular, protection against acute infection may not be predictive of protection against chronic biofilm-associated S. aureus disease. These results demonstrate that if a single antigen is to be used, care should be taken to ensure that the antigen is universally expressed, accessible to the host immune response, and necessary for bacterial persistence under all S. aureus growth conditions. An alternate strategy might be to formulate a vaccine that includes multiple protective antigens that play critical roles in both acute and chronic forms of infection. In either case, robust testing in multiple animal models may provide useful information regarding the potential of a vaccine to protect against a dynamic pathogen such as S. aureus.