S. aureus strains: USA300 NRS384, USA100, MN8, Newman, and MNHOCH were received from Network on Antimicrobial Resistance in S. aureus (NARSA) (currently BEI, beiresources.org). Previously, these strains had been well characterized for their toxin profile by western blots as well as culture supernatant toxicity in rabbit blood (33, 34). Bacterial strains were further confirmed by the hemolytic pattern on sheep blood agar. Bacterial culture supernatants were prepared in 7 ml BHI media in 50 ml Falcon tubes incubating overnight at 37°C in a shaking incubator (230 rpm). The next day, OD₆₀₀ was measured and normalized to 6.0 using BHI media. Culture supernatants were prepared by centrifuging the cultures followed by sterile filtration passing through 0.2 μm filters. Culture supernatants were aliquoted and stored at -80°C.

For challenge dose preparations, fresh BHI plates were inoculated by a glycerol stock generated from a single colony USA300 NRS384. The next day, colonies were transferred into 7 ml BHI broth and grown overnight to prepare the inoculum. On the 3^rd day, 1 liter of BHI broth was inoculated with inoculum and grown for 6 hours at 37°C shaker flask (230 rpm). Culture suspension was then centrifuged, washed in PBS, and frozen stock aliquots were prepared and stored at -80°C.

S. aureus toxins (Hla, PVL, LukAB, LukED, HlgAB, and HlgCB) and superantigens (SEA, SEB, SED, TSST-1, SEK, SEQ, and SEC1) were provided by IBT Bioservices (ibtbioservices.com) with > 98% purity. The toxins were reconstituted in PBS and stored at -80°C until use.

The HL-60 cells, from ATCC (atcc.org), Manassas, VA were cultured in RPMI media containing 82 U/ml of penicillin and streptomycin with 16% FBS. Cells were differentiated in media with 1.5% dimethyl sulfoxide (DMSO) as described previously (6, 8, 33). PBMCs from healthy volunteers were obtained from StemExpress (stemexpress.com, Cat #LE010F).

IBT-V02 vaccine consists of five rationally designed toxoid proteins (8) including Hla toxoid, PVL toxoid consisting of mutated S and F subunit proteins, and a dimeric LukAB toxoid. To generate the inactivated toxoids of SEA, SEB, and TSST-1, three-point mutations were introduced in each corresponding wild type toxin (SEA_{L48R/D70R/Y92A}, SEB_{L45R/Y89A/Y94A,} and TSST-1_{L30R/D27A/I46A}), which are the critical amino acid residues in the MHC binding sites of these toxins. These mutants had previously been tested in different animal models (35). For SEA_{L48R/D70R/Y92A}, we introduced another extra mutation H225A to generate SEA_{L48R/D70R/Y92A/H225A,} based on the published literature on the role of this amino acid to activate the T-cell response (36, 37). Individual toxoids were linked together by a flexible linker (three GGGGS repeats (3xG4S)) in order of TSST-1_{L30R/D27A/I46A}, SEB_{L45R/Y89A/Y94A,} and SEA_{L48R/D70R/Y92A/H225A}. Hence, the name TBA₂₂₅. This construct was expressed in E. coli, and the fusion protein was purified by column chromatography as described below as well as in our previously published paper (7).

The individual toxoid proteins were cloned in the vector pET24a+ and expressed in BL21 DE3 E. coli. IPTG (0.3 mM) was used to induce the expression of proteins. Cells were lysed by lysis buffer with lysozyme (1 mg/ml) followed by sonication. The supernatant containing the proteins was precipitated with ammonium sulfate. The pellet was then resuspended and desalted into capture column equilibration buffer and loaded over a Poros 50 HS cation exchange column. Fractions were collected and tested to identify the eluted fractions containing the toxoids by running and staining SDS gel against standards. Fractions were then collected and dialyzed. Each protein was thoroughly tested and stored at -80°C in storage buffer. A detailed purification method for different components of IBT-V02 has been published elsewhere (6, 7, 38). The finished products have a high level of purity, very low level of endotoxin, and expected molecular weight as characterized by SEC-HPLC, SDS-PAGE, and other QC methods described in our previous publication (8).

For rabbit immunizations, five toxoid proteins were blended, based on equal weight ratios, and used as a pentavalent immunogen. Immunogen was emulsified with Complete Freund’s Adjuvant (CFA) for prime and incomplete Freund’s Adjuvant (IFA) for boosts in a 1:1 ratio (v/v) before subcutaneous (s.c.) injection. The immunization dose was 200 µg/animal of total protein (40 µg of each toxoid) for the prime and final push, and 100 µg total protein (20 µg of each toxoid) for boosts. A total of 15 New Zealand White (NZW) specific pathogen-free (SPF) rabbits were immunized 5 times before final bleed to collect serum at Rockland under an IACUC approved protocol. Test bleeds were collected one week after each immunization. The immunization schedule is illustrated in Figure 1 .

After the terminal bleed, the serum from each rabbit was pooled and prepared for Low Endotoxin Unit (EU) purification (performed by Rockland Immunochemicals SOP 073A) using liquid chromatography (Protein A Sepharose 4 Fast Flow; performed by Rockland Immunochemicals SOP 073A). Purified rabbit IgG, meeting specification for purity and low endotoxin, was shipped to Emergent BioSolutions for final preparation of the test articles, IBT-V02-IgG and IBT-V02-F(ab’)2 (EBSI protocol PRO047450).

In general, the purified antibody was dialyzed in PBS buffers and transferred aseptically to the final container. The purified IgG was analyzed by HPLC using a Tosoh TSKgel UP‐SW3000 (4.6 mm x 30 cm) Column on an Agilent 1100, and by SDS‐PAGE under reduced and non‐reduced conditions. Part of purified IgG (9.82 mg/ml) was formulated into 10% (w/w) maltose, 0.03% (w/w) polysorbate 80 (PS80) as final formulation to provide an intact IgG hyperimmune control (IBT-V02-IgG). Another part of purified IgG was digested with pepsin to produce F(ab’)₂ plus F(ab’)₂-related immune globulin fragments. The digested material was further purified through diafiltration and anion exchange chromatography. The purified F(ab’)₂ product was diafiltrated and formulated into 10% maltose and 0.03% PS80 for the F(ab’)₂ test articles (IBT-V02-F(ab’)2). Sera collected from naïve rabbits were used as a negative control, the same purification process and formulation were used to purify control Rabbit F(ab’)₂ (RF(ab’)2). For experiments where indicated, purified intact control Rabbit IgG (R-IgG) was purchased from Rockland. Part of the IBT-V02-F(ab’)2 was further polished using EndoTrap^® HD resin and columns to meet established specifications for endotoxin levels. Two different lots of final products: lot#1 (Lot#740_POC_17_001_004) and lot #2 (Lot# 510063-20-001) were characterized using standard QC methods and the results are summarized in Table 1 .

A multiplex assay to detect serum IgG titers to cytolytic toxins and SAgs has been previously developed at IBT using the Luminex^© xMAP technology. Briefly, IBT-V02 target antigens, Hla, LukS-PV, LukF-PV, LukAB, SEA, SEB, and TSST-1, were coupled to carboxylated MagPlex microsphere beads with distinct spectral regions via a carbodiimide reaction. Individual bead stocks (bead regions) were mixed in equal ratios to generate a multiplex master mix. The master mix was co-incubated with serum samples in a 96-well plate for 2 hours at room temperature with shaking, washed three times on a magnetic plate washer and co-incubated with a secondary PE-labeled anti-rabbit IgG for detection. Plates were washed, and samples were acquired on a Luminex 200 instrument. Data were analyzed using a 4-parameter logistic (4PL) curve fit in XLFit (IDBS, Microsoft). IgG titers were expressed as the effective dilution at the point of the 4PL curve where 50% (ED₅₀) of antigen binding was detected by toxin-specific antibodies present in the serum sample.

Hla TNAs were performed as previously described (8). In brief, 8% rabbit red blood cells (RRBCs) were co-incubated with wildtype Hla ± serially diluted serum samples at 37°C. Cells were centrifuged after 30 min and absorbance determined at OD_416nm. PVL and LukAB TNAs were performed with human promyelocytic leukemia (HL-60) cells as previously described (25). In brief, differentiated HL-60 cells were incubated with either PVL or LukAB ± serially diluted serum samples at 37°C with 5% CO₂ for 3 hours, and CellTiter Glo was added to the culture to measure cell viability.

For superantigen TNAs, serum or purified antibodies were serially diluted in RPMI 1640 media (Gibco A10491), mixed with EC₉₈ concentration of superantigens, and incubated at room temperature for 1 hour in a 96-well plate (BD Falcon BD353376). PBMCs (StemExpress #LE010F) from healthy volunteers were added such that the final density was 1 × 10⁵ cells per well; plates were incubated at 37°C with 5% CO₂ for 48 hours. The supernatants were collected and IFN-γ was measured by ELISA (R&D Systems #DY285B), to calculate neutralization as described in our previous publications (7, 8). A 4-parameter logistic (4PL) curve fit in XLfit (IDBS, Microsoft) was used to analyze the data. Based on the 4PL plots, ND₅₀ (50% neutralizing dilution for serum) or NC₅₀ (50% neutralizing concentration for purified antibody) were calculated.

Female BALB/c mice were purchased from Charles River Laboratories. The age of mice for each experiment was 6-8 weeks. Mice were challenged intraperitoneally (IP) with a lethal dose of USA300 NRS384 in 200 µl volume and monitored for health, weight, and survival for 14 days post-challenge. For the full length (IBT-V02-IgG) and F(ab’)₂ (IBT-V02-F(ab’)2) comparative study, mice were administered IP (150 mg/kg) -2 hours prior to bacterial challenge once for the IBT-V02-IgG group, whereas for F(ab’)₂ control (RF(ab’)2) and IBT-V02-F(ab’)2, mice were dosed at -2, +24, 48, 72, and 96 hours post challenge in 500 µl volume. Two different challenge doses 1xLD90 and 2xLD90 were used. For IBT-V02-F(ab’)2 dose-response study, we treated mice twice a day in two different doses: (75 mg/kg per dose; 150 mg/kg/day) or (50 mg/kg per dose; 100 mg/kg/day) in two different challenge doses 1xLD90 or 1.5xLD90. Mice were maintained under pathogen-free conditions and fed laboratory chow and water ad libitum.

For the pneumonia model, mice (6-8 wks.) were anesthetized with isoflurane and were intranasally (IN) challenged with a lethal dose of S. aureus (USA300 NRS384) in 20 or 30 µl DPBS. F(ab’)₂ was administered two times per day (50 mg/kg, 75 mg/kg, or 100 mg/kg per dose) through IP route. Mice were dosed at days: D0, D1, D2, D3, and D4 post challenge in 250 µl volume per dose. Four different challenge doses were used: 5x10⁸, 4x10⁸, 3.5x10⁸, and 3x10⁸ CFU/mouse. Animals were monitored for survival, health scores, and body weight two times a day (AM and PM) within the first 4 days of challenge and then once a day until termination of the study (day 14) as described previously (39).

Data were analyzed using GraphPad PRISM v8.4.3. Statistical analysis was performed using an unpaired t-test for the comparison of area under the curve (AUC) for supernatant-based antibody neutralization assays with the controls. An overall significance level of 5% was used and the probability of a type 1 error for each test was adjusted for multiple comparisons. Kaplan-Meier curves along with log-rank (Mantel-Cox tests) were used to compare the survival rates for each treatment group to the negative control group.

IBT-V02 vaccine consists of five rationally designed proteins, seven fully attenuated toxoids, and can elicit immune responses against over 15 Staphylococcal toxins by homological cross-neutralization (8). Hla toxoid harbors two mutations (H35L and H48L), PVL toxoid consists of mutated S and F subunit proteins (LukS_mut9 and LukF_mut1), a dimeric toxoid LukAB_mut50, and a fusion protein TBA₂₂₅, which represents three superantigen toxoids fused together via a flexible linker (TSST-1, SEB and SEA). Each of these toxoids is fully attenuated, and the safety and immunogenicity had been extensively studied previously (40). Using five toxoid proteins blended at equal weight ratios as immunogen to immunize rabbits, strong antibody responses were elicited. Sera from test bleeds were collected one week after each boost ( Figure 1 ) and antibody titers to seven IBT-V02 targeted antigens (Hla, LukS-PV, LukF-PV, LukAB, SEA, SEB, and TSST-1) as determined by a multiplex Luminex assay for each rabbit are presented as the effective dilution of the sera that produced 50% maximal response (ED₅₀) ( Figure 2 ). IBT-V02 induced robust antibody responses after the first boost to all the antigens except SEA. The average ED₅₀ titer for Hla is 136,673, LukS-PV is 72,458, LukF-PV is 32,083, LukAB is 77,572, SEB is 14,369, and TSST-1 is 47,006. The magnitude of the responses to each antigen was comparable to the titers obtained from previous studies, except for SEA (8). However, the second boost did induce antibody response to SEA, and the titers steadily increased until the last terminal bleed, reaching an average ED₅₀ titer of 50,385. The antibody titers to other antigens were also gradually increased upon subsequent boosts. The terminal bleed titers against all target antigens were significantly higher than the assay internal control, which consisted of antibodies obtained by immunizing rabbits with IBT-V02 two times (IBT generated rabbit polyclonal sera). Individual rabbit responses were variable, but in close range against each antigen. There were no “non-responders” observed in the group, indicating that IBT-V02 is highly immunogenic and can successfully induce universal responsiveness with a multivalent approach.

Three types of toxin neutralization assays (TNA) were used to further examine the function of antibodies elicited by IBT-V02 immunization. The potency of antibodies against Hla was measured by the inhibition of alpha-toxin dependent rabbit red blood cell lysis. The potency against PVL and LukAB was determined by the inhibition of leukocidin-induced cytotoxicity in PMN-like HL-60 cells. Human PBMC based superantigen TNAs were carried out to study the antibody potency against SEA, SEB, and TSST-1. As observed in the Luminex titers ( Figure 2 ), the first boost elicited strong neutralizing antibodies to Hla, PVL, LukAB, SEA, SEB, and TSST-1, with few exceptions where neutralizing antibodies were not generated against SEA. The second boost increased the antibody response to all the antigens except SEA ( Figure 3 ). The response to Hla, PVL, LukAB, and TSST-1 plateaued after the second boost. The response to SEB increased by subsequent boost, and upon the last boost, the neutralization titer was significantly increased for SEA. Taken together, the binding and neutralization studies showed that highly potent antibodies against all vaccine components were generated by IBT-V02 immunization.

Next, we sought to characterize the anti-Staph hyperimmune products and compare the profiles of full-length IgG and de-speciated F(ab’)2 derived from the pooled sera of all 15 immunized rabbits. In this series of experiments, IBT-V02-IgG and IBT-V02-F(ab’)2 antibody products were tested for their binding and neutralization potencies. The Luminex EC₅₀ for antibodies against all antigens were in hundreds of ng/mL ( Supplementary Table S1 ) range, significantly lower than the control rabbit F(ab’)₂ (RF(ab’)2) generated from naïve animals. Purified antibody products also exhibited strong neutralization against toxins. The TNA NC₅₀s were in the single or double µg/mL range ( Supplementary Table S2 ). When the binding and neutralization potencies were compared among the two IBT-V02-F(ab’)2 lots and one full-length IgG (IBT-V02-IgG) lot, there was no significant difference among different antibody products, indicating that the removal of Fc region did not affect IBT-V02-F(ab’)2 in vitro potency ( Figure 4 ).

The bicomponent leukocidins produced by S. aureus include PVL, gamma hemolysins (HlgAB and HlgCB), LukED, and LukAB. The S and F subunits of bicomponent leukocidin have a high degree of homology respectively across the family except for LukAB. S. aureus also produces more than 20 types of SAgs, which have limited primary sequence identity but exhibit considerable structural homology. IBT-V02 vaccine contains toxoids for SEA, SEB, and TSST-1, which represent three divergent subgroups of SAgs. We tested the cross-neutralization potential first in terminal bleed samples from individual animals against leukotoxins ( Figure 5 ) and then in purified antibodies ( Figure 6 ). To further evaluate the toxin neutralization ability of newly purified anti-Staph hyperimmune product, we examined the cross-neutralization of IBT-V02-F(ab’)2 against leukocidins and superantigens that were not part of the vaccine components. Results showed that IBT-V02-F(ab’)2 exhibited neutralization against HlgCB and LukED when compared to control RF(ab’)₂, ( Figure 6 ) with the NC₅₀ values in the μg/ml range, so the potency was around 10 times lower than that against PVL ( Supplementary Tables S2 , S3 ). The superantigens SEC1, SED, SEK, and SEQ were used to investigate the cross-neutralization ability of IBT-V02-F(ab’)2 against superantigens, and results indicated that IBT-V02-F(ab’)2 can totally neutralize SEC1 at two concentrations tested, and neutralize SED and SEK in higher concentrations, but to a lesser extent against SEQ ( Figure 7 ). These data demonstrated that the cross-neutralizing activity of anti-Staph hyperimmune product can be achieved by a rationally designed multivalent vaccine with specific strategies targeting homology and specificity.

S. aureus strains in different clonal complexes display considerable variability in toxin profiles. The neutralizing potentials of IBT-V02-F(ab’)2 were compared in bacterial culture supernatant from different clinical strains with control RF(ab’)2 and PBS alone. Among the selected strains were the USA300 LAC (JE2), USA100, and S. aureus strains which are positive for SEB, SEA, and TSST-1 superantigens. As shown in Figure 8 , supernatant toxicity (PBS group) varies widely among the strains, and as expected, USA300 LAC exhibited the highest toxicity (rabbit blood-based hemolysis assay) followed by SEA positive Newman strain. We observed ED₅₀s of 400; 150; 64; 50; and 40 for JE2 (USA300, LAC), Newman, MNHOCH, MN8, and USA100, respectively. IBT-V02-F(ab’)2 neutralizing activities were most prominent in USA300 LAC and USA100 strains when compared with MN8, Newman, and MNHOCH in presence of 40 µg/ml when compared with Control-RF(ab’)2 ( Figure 8 ). AUC showed a significant reduction in toxicity in (IBT-V02-F(ab’)2) when compared either with PBS alone or with control RF(ab’)₂ ( Figure 8 AUC bar diagrams inset figures next to each plot). For supernatant neutralization of superantigens, PBMC based assays demonstrated that 100 µg/ml of IBT-V02-F(ab’)2 significantly decreased the amount of IFNγ production (readout of superantigenicity) when compared with either media alone or with control RF(ab’)2 ( Supplementary Figure S1 ) in all three tested strains: MN8 (TSST-1); Newman (SEA); and MNHOCH (SEB). Similarly, as proof of concept, we selected two bacterial culture supernatants, JE2 and Newman, to measure the neutralizing potency of IBT-V02-F(ab’)2 against leucotoxins. As shown in Supplementary Figure S2 , IBT-V02-F(ab’)2 significantly decreased the % lysis in HL-60 based assays. Nevertheless, cross-neutralizing activities of specific IBT-V02-F(ab’)2 were clearly demonstrated in all clinical isolates tested.

Given the equivalent in vitro potency observed for IBT-V02-F(ab’)2 and IBT-V02-IgG, we investigated whether these antibody products protected in a murine model of MRSA-induced bacteremia and could demonstrate equivalent in vivo potency. The bacteremia was established by intraperitoneal injection of USA300 strain NRS384. Upon 1 x 10⁸ CFU/mouse inoculation, high bacterial loads could be recovered from the liver, kidneys, spleen, and lungs 2 days after infection, with bacteria disseminated to blood by day 4 and could still be recovered from internal organs 6 days after infection (data not included). The LD90 was determined as 2 x 10⁸ CFU/mouse, which caused 90% mortality among the five inoculation doses tested. A single dose of 150 mg/kg of IBT-V02-IgG at 2 hours before infection and five consecutive doses of IBT-V02-F(ab’)2 at 2 hours before, and 1 day, 2 days, 3 days, and 4 days post-infection (dpi) were given to animals infected with 1xLD90 or 2xLD90 USA300 NRS384 along with the control. Upon 2xLD90 infection, all control group animals died within 24 hours, and only 10% of survival was observed for each treated group. When animals were challenged with 1xLD90 of USA300 NRS384, 90% of control group animals were dead within 48 hours; IBT-V02-IgG provided 60% survival, while IBT-V02-F(ab’)2 protected 50% of animals at the end of the 14-day study. Compared to the gradually decreased survival in the RF(ab’)2 treated group, IBT-V02-IgG treated group maintained 90% for 7 days ( Figure 9A ), which is about the half-life of the rabbit antibody in the murine model (41). When we closely examined the health scores, we found that the animals in the IBT-V02-F(ab’)2 group had lower health scores (i.e., healthier) in the afternoon after treatment compared to their conditions in the morning before treatment, when most mortality occurred. Based on previous studies, we have found that the half-life of other F(ab’)₂ hyperimmunes in mice is about 5 hours (unpublished data). In comparison, full-length or intact IgG hyperimmunes typically have a half-life of ~3-4 days in mice (unpublished data). Taken together, we hypothesized that the protection could be improved by increasing the bioavailability of IBT-V02-F(ab’)2. To test this hypothesis, animals were treated with IBT-V02-F(ab’)2 or RF(ab’)2 control twice per day, 8 hours apart for 5 consecutive days after 1xLD90 or 1.5xLD90 infection, at two treatment doses, 75 mg/kg and 50 mg/kg. Interestingly, even the lower dose at shorter treatment intervals significantly protected the animals, with 100% and 90% of survival observed against the 1xLD90 challenge ( Figure 9B ). The treatment only provided 20% survival upon 1.5xLD90 infection, but significantly increased the mean time to death (MTD) of treated groups compared to control group (Log-Rank, Mantel-Cox test). The dose and bioavailability dependent protection provided by IBT-V02-F(ab’)2 treatment strongly confirmed the in vivo potency of de-speciated anti-toxin antibodies against MRSA bacteremia. The survival data were also reflected in the health score and change in weight data ( Supplemental Figures: S3A, B ).

To further evaluate the in vivo potency of anti-toxin hyperimmune products, we investigated the protective efficacy of IBT-V02-F(ab’)2 in a mouse pneumonia model. LD90 for this model was not accurately determined due to the observation of either 100% mortality upon higher inoculation or 100% survival upon lower inoculation. Therefore, we tested the efficacy using four different interpolated doses of USA300 strain NRS384 via intranasal challenge ( Figures 10A, B ). Upon 5x10⁸ CFU/mouse challenge, only 10% of 75 mg/kg IBT-V02-F(ab’)2 treated animals survived, no significant protection was observed. Upon 4x10⁸ CFU/mouse challenge, 75 mg/kg of treatment significantly extended the MTD but didn’t significantly improve the survival rate ( Figure 10A ). When animals were infected with 3x10⁸ CFU/mouse USA300 NRS384, 100 mg/kg of IBT-V02-F(ab’)2 treatment provided an 80% survival rate, significantly higher than the control group. With the challenge dose increased to 3.5x10⁸ CFU/mouse, the 40% survival was not significant compared to control, but the treatment significantly extended the MTD ( Figure 10B ). The survival data also correlate with the health score and change in weight data ( Supplemental Figures: S4A, B ). These results further confirmed the function of IBT-V02-F(ab’)2 in stringent pneumonia models and validated the efficacy of an anti-toxin passive immunization strategy against S. aureus MRSA infection.