rPNAG Is Physically Associated with OMVs.

Previous studies have established that the majority of PNAG produced by cells is tightly bound to the outer surface (7, 23). To determine whether rPNAG produced by our engineered E. coli was similarly associated with OMVs, we first performed density gradient ultracentrifugation, which leverages the fact that liposomal nanoparticles, like OMVs, are less dense than soluble macromolecules, like proteins and polysaccharides, due to their lipid content. Western blotting and Coomassie staining of the fractions obtained from density-based separation revealed that rPNAG, total proteins, and an OMV-specific protein marker, namely OmpA, all comigrated to similar fractions (SI Appendix, Fig. S1), reminiscent of the gradient profiles seen previously for intact OMVs and OMV-associated cargo, including recombinantly expressed proteins and oligosaccharides (18, 20, 21, 28). To more precisely examine rPNAG localization, we performed immune-EM (IEM) using mAb F598 and a colloidal gold-conjugated anti-human secondary antibody. Whereas OMVs derived from JC8031 and JC8031 ΔpgaC cells failed to react with F598, JC8031 cells overexpressing plasmid-encoded pgaABCD exhibited strong F598 antibody labeling in circular patterns consistent with the observed diameters of OMVs (Fig. 2D). It should also be pointed out that there was no detectable immunostaining of these rPNAG-containing OMVs with the human IgG1 mAb F429, which has identical antibody constant regions as F598 but with variable regions that provide specific binding to the alginate antigen from Pseudomonas aeruginosa (29). Taken together, these results provide strong support for our hypothesis that cell surface-bound rPNAG becomes a constituent of released OMVs, giving rise to glycOMVs displaying rPNAG polysaccharides (rPNAG-glycOMVs).

Mice Vaccinated with glycOMVs Produce rPNAG-Specific Antibodies.

Next, we investigated the immunogenicity and IgG isotype distribution in sera of mice receiving engineered rPNAG-glycOMVs. Female BALB/c mice immunized s.c. with rPNAG-glycOMVs developed robust IgG responses to rPNAG that increased during the weeks postimmunization (Fig. 3A). Maximum IgG titers were achieved 8 wk after the first immunization and were much higher than the titers measured for control mice immunized with empty OMVs or PBS (Fig. 3A). A similar response was observed in mice immunized with a conjugate vaccine composed of a synthetic 5-mer of fully deacetylated GlcNAc monosaccharides linked to the carrier protein tetanus toxoid (5GlcNH₂-TT) (8) (Fig. 3B). It is noteworthy that the IgG titers developed in mice immunized with an identical dose of unconjugated rPNAG were on par with the PBS control (Fig. 3A), consistent with the weak immunogenicity of glycans alone (9). In addition to IgGs, the rPNAG-glycOMVs also stimulated PNAG-specific IgM but not IgA antibodies (SI Appendix, Fig. S2A). It is also worth mentioning that the titers of serum antibodies against OmpA, an outer membrane protein antigen present in both empty and PNAG-containing OMV preparations, were indistinguishable (SI Appendix, Fig. S2B), confirming that an equivalent amount of OMVs was used in the immunizations.

Open in a separate window

Fig. 3.

Immunization with glycOMVs yields rPNAG-specific IgGs. ELISA-determined IgG antibody titers to rPNAG in sera of immunized BALB/c mice at 6 wk (gray bars) and 8 wk (white bars) after the first immunization (t = 0; black bars). Groups of mice (n = 8) were immunized as follows: (A) PBS, unconjugated PNAG derived from E. coli Top10 cells carrying pUCP18Tc-pga (rPNAG), OMVs from plasmid-free JC8031 ΔpgaC cells (empty OMVs), OMVs from JC8031 cells carrying pUCP18Tc-pga (rPNAG-glycOMVs), or OMVs from JC8031 carrying pUCP18Tc-pga and pTrc-IcaB (rdPNAG-glycOMVs); and (B) PBS and 5GlcNH₂-TT conjugate. Mice immunized with OMVs were injected s.c. at t = 0 and boosted at 3 and 6 wk with 10 µg total protein as measured by total protein content in OMV fraction. Mice immunized with 5GlcNH₂-TT were injected s.c. at t = 0 and boosted at 3 and 6 wk with 10 µg purified conjugate, which was administered with incomplete Freund’s adjuvant. Purified rPNAG isolated from E. coli Top10 cells carrying pUCP18Tc-pga was used as immobilized antigen. Bars represent means, and error bars indicate SD of mean IgG titers. One-way ANOVA allowed for rejection of the null hypothesis that all immunization groups have the same mean IgG titer (P < 0.001). *Statistical significance (P < 0.05) by Tukey–Kramer honest significance difference (HSD); **Statistical significance (P < 0.01) by Tukey–Kramer HSD.

IgG titers were further characterized by analysis of IgG1 and IgG2a titers, wherein mean IgG1 to IgG2a antibody ratios serve as an indicator of a Th1- or Th2-biased immune response. Mice immunized with rPNAG-glycOMVs or 5GlcNH₂-TT developed titers of both rPNAG-specific IgG1 and IgG2a antibodies that were much higher than the empty OMV and PBS control groups (SI Appendix, Fig. S3). Interestingly, whereas 5GlcNH₂-TT elicited a response that was strongly Th2 biased (IgG1 >> IgG2a), which is typical of most conjugate vaccines (30), including dPNAG-containing conjugates (6), the rPNAG-glycOMVs triggered a much more balanced Th1/Th2 response, which is consistent with previously reported responses to glycOMVs displaying different oligosaccharides (20, 21). Taken together, these results indicate that the display of rPNAG on the exterior of OMVs significantly increased the immunogenicity of this polysaccharide antigen.

IcaB Overexpression Augments the Immunogenicity of glycOMVs.

Previous studies showed that a conjugate vaccine composed of native PNAG or dPNAG was able to stimulate production of IgG antibodies in immunized animals (6). However, immunization with the highly acetylated PNAG conjugate resulted in significantly higher antibody titers to PNAG than to dPNAG, whereas the poorly acetylated dPNAG conjugate triggered strong antibody titers to both antigens (6). This specificity is important, as the antibodies elicited by the dPNAG conjugate were superior in terms of opsonic killing activity and protective efficacy against S. aureus and E. coli infection in animals (6, 31). Here, we observed that antibodies raised to rPNAG-glycOMVs were specific for the highly acetylated PNAG as determined by measuring ELISA titers in the different sera to either native PNAG purified from S. aureus, having high acetylation levels, or a chemically prepared dPNAG antigen, having only a small percentage of the amino groups substituted with acetate (SI Appendix, Fig. S4A). This result was not surprising given that overexpression of pgaABCD in E. coli gives rise to a PNAG glycopolymer that contains only about 3% dPNAG (4).

In an effort to increase PNAG deacetylation in glycOMVs, E. coli cells carrying the rPNAG-producing plasmid pUCP18Tc-pga were cotransformed with a second plasmid for heterologously expressing the PNAG deacetylase, IcaB, from S. aureus. Whereas the E. coli deacetylase, PgaB, is a lipoprotein, the IcaB enzyme is membrane associated through binding to peptidoglycan (7). To facilitate IcaB expression in E. coli, we removed its native secretion signal and replaced it with an E. coli Sec-dependent signal peptide for high-level periplasmic expression. Western blot analysis of both periplasmic and whole-cell fractions derived from this strain showed that IcaB was present as a soluble protein in the periplasm and that the majority of it was associated with the membrane containing whole-cell fraction (SI Appendix, Fig. S5A). Importantly, OMVs isolated from cells overexpressing pgaABCD and icaB [recombinant deacetylated poly-N-acetyl-d-glucosamine (rdPNAG)-glycOMVs] were recognized by mAb F598 to an extent that was indistinguishable from that observed for rPNAG-glycOMVs (SI Appendix, Fig. S5B), indicating that these OMVs were similarly decorated with PNAG. Unfortunately, chemical characterization using GC/MS or NMR on preparations obtained from these bacteria was unsuccessful, a well-documented problem (5) that prevented us from definitively confirming the degree of deacetylation for PNAG produced in IcaB-expressing cells. Underlying the barrier to chemical characterization of PNAG is the finding that, unlike bacterial CPSs that can be readily isolated from culture supernatants and microbial cell surfaces, almost all of the PNAG is tightly bound to the cell surface and thus, must be extracted for purification (7, 23). As a result, the actual yields are markedly reduced by the difficulty in extracting PNAG from the cell surface as well as additional significant losses caused by the insolubility of PNAG at neutral pH and the complex purification procedures needed to eliminate tightly associated contaminating bacterial antigens (6, 23). Nonetheless, indirect evidence for increased deacetylation was obtained from immunization experiments. First, the rPNAG antigen-specific antibody titers developed in mice immunized with rdPNAG-glycOMVs were significantly higher than those obtained with rPNAG-glycOMVs, rivaling the titers observed for immunization with the 5GlcNH₂-TT conjugate (Fig. 3). Second, antibodies in mice immunized with rdPNAG-glycOMVs were directed to epitopes on both highly and poorly acetylated PNAG antigens, which was in stark contrast to the rPNAG-glycOMV–induced antibodies that mainly reacted with acetyl-containing epitopes (SI Appendix, Fig. S4A). It is also noteworthy that the rdPNAG-glycOMVs stimulated PNAG-specific IgM antibodies and unlike the more acetylated rPNAG-glycOMVs, IgA antibodies as well (SI Appendix, Fig. S2A).

Detoxified glycOMVs Retain Immunogenicity.

One concern with the use of bacterial OMVs as a vaccination platform is the reactogenicity caused by the presence of LPS on the membrane surface. To circumvent this issue, we previously remodeled the lipid A structure of E. coli strain JC8031 at the genetic level by knocking out the acyl transferase, LpxM, which causes the strain to produce a pentaacylated form of lipid A that has significantly reduced toxicity (32). The resulting strain, JH8033, yielded OMVs that were significantly less toxic as measured by hTLR4 activation, while retaining full immunostimulatory qualities (21). OMVs harvested from JH8033 cells overexpressing pgaABCD and icaB reacted strongly with mAb F598 (SI Appendix, Fig. S5B), indicating that the abundance of rPNAG on the exterior of OMVs was not diminished by remodeling of lipid A. Importantly, the IgG antibody titers developed in mice immunized with rdPNAG-glycOMVs harboring wild-type lipid A vs. remodeled lipid A were nearly identical (SI Appendix, Fig. S4B), indicating that reducing the reactogenicity by lipid A modification does not adversely affect the immunogenicity of rPNAG-containing glycOMVs.

Opsonic Killing Activity of glycOMV-Induced Antibodies Against S. aureus.

Encouraged by the robust polysaccharide-specific IgG responses induced by glycOMVs, we next screened mouse sera for opsonic killing activity in the presence of human polymorphonuclear leukocytes (PMNs) and autologous human complement. While PNAG is known to be present on the surfaces of many important Gram-negative and Gram-positive bacterial pathogens (5), we focused initially on S. aureus, because PNAG production is most extensively characterized in this bacterium. Mice immunized with rdPNAG-glycOMVs clearly had opsonically active antibodies that mediated killing of S. aureus CP8 strain MN8, whereas mice receiving empty OMVs or PBS had significantly less killing activity (Fig. 4A). It is noteworthy that the opsonic killing activity measured for these antibodies exceeded that of the serum antibodies developed in mice immunized with the 5GlcNH₂-TT conjugate and was on par with that measured for the positive control mAb F598 (Fig. 4A). Importantly, there was no significant bactericidal activity observed when key components of the assay were missing or inactive (Fig. 4A).

Open in a separate window

Fig. 4.

glycOMVs elicit opsonically active antibodies and protective immune responses against S. aureus. (A) Opsonic killing activity against S. aureus CP8 strain MN8 by different dilutions of serum from BALB/c mice (n = 8 except for 5GlcNH₂-TT, where n = 4) immunized with PBS, empty OMVs, rdPNAG-glycOMVs, or 5GlcNH₂-TT. Bars represent means, and error bars indicate SD of the mean. The mAb F598 served as positive control and was tested at 10 μg/mL (black bars) and 5 μg/mL (gray bars). The mAb F429, which is specific for the P. aeruginosa alginate antigen, was tested at 10 μg/mL and served as a negative control to which all data were normalized. Negative control serum from mice receiving PBS was only tested at a dilution of 1:30. All of the sera were tested in the absence of either human PMNs (no PMNs) or human complement (no C′) or in the presence of heat-inactivated complement (HIC′), and the average of this entire dataset is shown. (B) Kaplan–Meier survival analysis of groups of BALB/c mice (n = 16 except for 5GlcNH₂-TT, where n = 12, and both detoxified OMV groups, where n = 8) that were immunized with PBS, empty OMVs, rPNAG-glycOMVs, rdPNAG-glycOMVs, detoxified rdPNAG-glycOMVs, or 5GlcNH₂-TT and subsequently challenged at 56 d after the first immunization with 1 × 10⁹ cfu/mL (10 × mLD₅₀) of S. aureus strain Rosenbach in 200 μL sterile PBS via tail vein injection. Mice immunized with OMVs were injected s.c. at t = 0 and boosted at 3 and 6 wk with 10 µg total protein as measured by total protein content in the OMV fraction. Mice immunized with 5GlcNH₂-TT were injected s.c. at t = 0 and boosted at 3 and 6 wk with 10 µg purified conjugate, which was administered with incomplete Freund’s adjuvant. The groups receiving rdPNAG-glycOMVs and detoxified rdPNAG-glycOMVs showed statistically significant protection over the PBS group (P < 0.05) as determined by a log rank test.

Protective Efficacy of glycOMVs Against S. aureus Challenge in Mice.

To determine whether the opsonic killing activity measured above translated to a protective immune response in animals, immunized BALB/c mice were challenged at 56 d after the initial dose with 10 times the modified LD₅₀ (mLD₅₀) of S. aureus strain Rosenbach via tail vein injection. Mice immunized with rdPNAG-glycOMVs or their detoxified counterpart had the highest survival rates (~60%), while the survival rates of those receiving the 5GlcNH₂-TT conjugate were slightly lower (42%). That there was no significant difference in survival between the two rdPNAG-containing OMV-treated groups indicates that lipid A remodeling can effectively reduce the reactogenicity of PNAG-containing OMV candidates without compromising their protective activity. As expected, mice immunized with PBS or empty OMVs had the lowest survival rates (<20%) (Fig. 4B). Importantly, these survival rates correlated strongly with the measured PNAG-specific IgG titers and opsonic killing activity developed in these mice. It is also noteworthy that rPNAG-glycOMVs, which putatively contained mostly acetylated PNAG as discussed above, afforded a level of protection that was on par with the 5GlcNH₂-TT conjugate, which was completely deacetylated. This result was particularly interesting, because previous studies have shown that passive administration of antibodies raised against the highly acetylated, native PNAG antigen is unable to mediate clearance of S. aureus from the blood or protect against a high-dose lethal infection (6, 8).

Serum Bactericidal Activity of glycOMV-Induced Antibodies Against F. tularensis.

In light of the fact that diverse Gram-negative and Gram-positive bacterial pathogens produce surface-bound PNAG (5), we sought to determine whether PNAG-containing OMVs could provide cross-species protection. We focused our attention on F. tularensis, a Gram-negative, nonspore-forming bacterium and the causative agent of tularemia, which had not previously been confirmed for surface-associated PNAG production. Our own genetic analysis of published genomes failed to identify genetic loci in F. tularensis similar to the known four-gene ica or pga loci in S. aureus and E. coli, respectively; however, a number of ica/pga-negative organisms have been confirmed to produce PNAG by biosynthetic enzymes encoded in related but currently unidentifiable genetic loci (5). Therefore, we proceeded to experimentally interrogate the surface of the F. tularensis subsp. holarctica live vaccine strain (LVS) for PNAG expression using a sensitive and specific immunochemical fluorescence analysis developed previously (5). Indeed, PNAG was detected by confocal microscopy on the surface of F. tularensis LVS cells using mAb F598 directly conjugated to AlexaFluor 488 (green fluorescence) (SI Appendix, Fig. S6). To confirm that mAb F598 was binding to PNAG, cells were subjected to digestion with the PNAG-degrading enzyme dispersin B or a related degradative control enzyme, chitinase, that specifically hydrolyzes the β-(1→4)–linked GlcNAc chitin molecule or were instead treated with sodium metaperiodate. Chitinase-resistant but dispersin B- and periodate-sensitive binding of mAb F598 was observed, thereby showing PNAG-specific labeling of cells (SI Appendix, Fig. S6). Nearly identical immunochemical labeling results were observed for two additional subspecies of F. tularensis, namely F. tularensis subsp. novicida U112 and F. tularensis subsp. tularensis strain Schu S4 (SI Appendix, Fig. S7).

Next, we investigated F. tularensis killing activity using a serum bactericidal assay (SBA). The serum antibodies developed in mice immunized with rdPNAG-glycOMVs exhibited potent concentration-dependent bactericidal activity against F. tularensis LVS that rivaled the activity measured for serum antibodies induced by the 5GlcNH₂-TT conjugate or the positive control mAb F598 (Fig. 5A). It should be noted, however, that mice immunized with 5GlcNH₂-TT had highly variable bactericidal activity, with antibodies from sera from two mice showing virtually no bactericidal activity and antibodies from sera from two others showing complete killing at all concentrations tested. Importantly, mice receiving PBS generated little to no bactericidal activity above the background measured in control tubes that contained bacteria but lacked complement or included heat-inactivated complement, whereas mice receiving empty OMVs showed a moderate level of bactericidal activity under the conditions tested (Fig. 5A). This latter activity was unexpected and indicated that some of the antibodies raised to features present in the empty OMVs were cross-reactive with F. tularensis LVS.

Open in a separate window

Fig. 5.

GlycOMVs elicit bactericidal antibodies and protective immunity against F. tularensis LVS. (A) Serum bactericidal activity against F. tularensis LVS by antibodies in the serum of BALB/c mice (n = 8 except for 5GlcNH₂-TT, where n = 4) immunized with PBS, empty OMVs, rdPNAG-glycOMVs, or 5GlcNH₂-TT. Survival data are derived from standard SBA, where dilutions of serum from immunized mice were tested against F. tularensis LVS Iowa in the presence of baby rabbit complement. All data were normalized to the killing measured for baby rabbit complement alone. Bars represent means, and error bars indicate SD of the mean. The mAb F598 served as positive control and was assayed according to the same dilution scheme, with 1:1 being equivalent to 1 mg/mL. Negative controls included mAb F598 in the absence of human complement (no C′) or in the presence of heat-inactivated complement (HIC′). (B) Kaplan–Meier survival analysis of groups of BALB/c mice (n = 5) that were immunized with PBS, detoxified empty OMVs, detoxified rdPNAG-glycOMVs, or 5GlcNH₂-TT and subsequently challenged at 70 d after the first immunization with F. tularensis LVS Iowa via i.p. injection of 500 cfu (~500 × LD₅₀). Mice immunized with OMVs were injected s.c. at t = 0 and boosted at 2 and 4 wk with 10 µg total protein as measured by total protein content in OMV fraction. Mice immunized with 5GlcNH₂-TT were injected s.c. at t = 0 and boosted at 2 and 4 wk with 10 µg purified conjugate, which was administered with incomplete Freund’s adjuvant. The groups receiving detoxified rdPNAG-glycOMVs showed statistically significant protection over the PBS and detoxified empty OMV groups (P < 0.05) as determined by a log rank test. The 5GlcNH₂-TT group was not significantly higher than the PBS group or significantly lower than the detoxified rdPNAG-glycOMVs group.

OMV Preparation.

OMVs were prepared as described previously (18) using JC8031 or its isogenic derivatives with or without plasmids pUCP18Tc-pga and/or pTrc-IcaB. Cells were selected on LBA supplemented with the appropriate antibiotic as needed, from which single colonies were selected and grown overnight in liquid LB media. Saturated cultures were subcultured 1:100 in LB supplemented with appropriate antibiotics and grown to midlog phase (OD₆₀₀ ~ 0.6), at which point pgaABCD and/or icaB expression was induced with 0.1 mM IPTG for 16–20 h at 37 °C. Cells were pelleted via centrifugation at 10,000 × g for 10 min. Cell-free culture supernatants were collected postinduction and filtered through a 0.2-µm filter. OMVs were isolated by ultracentrifugation (TiSW28 rotor; Beckman-Coulter) at 141,000 × g for 3 h at 4 °C and resuspended in sterile PBS. OMVs were quantified by the bicinchoninic acid (BCA) total protein assay (Pierce) using BSA as the protein standard. OMVs were subjected to density gradient ultracentrifugation as described previously (18, 20, 21) and in SI Appendix, SI Materials and Methods.

Dot Blot, SDS/PAGE, and Western Blot Analysis.

For dot blot analysis, OMV-containing fractions were diluted to 200 μg/mL total protein in PBS and incubated with or without 50 μg/mL dispersin B (Kane Biotech, Inc.) at 37 °C for 24 h. Samples were spotted directly onto nitrocellulose membranes followed by blocking with 5% milk in Tris-buffered saline (TBS) and probing with antibodies as described below for Western blot analysis. For SDS/PAGE, OMV-containing fractions were boiled for 15 min in loading buffer containing β-mercaptoethanol. After cooling to room temperature, samples were loaded into 12% polyacrylamide gels (BioRad) and separated electrophoretically. To visualize proteins, SDS/PAGE gels were stained with Coomassie Brilliant Blue R-250 (BioRad). For Western blot analysis, subcellular fractions, including soluble periplasm, whole-cell lysate, and OMV fractions, were separated by SDS/PAGE as described above and subsequently transferred to a PVDF membrane via semidry transfer. Membranes were blocked with 5% milk in TBS and then probed with the following primary antibodies: HRP-conjugated mouse anti-His antibody (Abcam) for the detection of IcaB, human mAb F598 for the detection of PNAG and dPNAG, and mouse anti-OmpA. Anti-human and anti-mouse HRP-conjugated secondary antibodies (Promega) were used in combination with mAb F598 and anti-OmpA, respectively. Signals were visualized using HRP substrate Clarity ECL (BioRad) and were imaged with a ChemiDoc XRS+ Imaging System (BioRad). Structural analysis of vesicles was performed via TEM as described previously (17) and in SI Appendix, SI Materials and Methods. Confocal microscopy and immunochemical detection of PNAG are also described in SI Appendix, SI Materials and Methods.

Purification of rPNAG from E. coli.

Top10 E. coli cells, which do not natively produce PNAG, were transformed with pUCP18Tc-pga and cultured to midlog phase. Production of rPNAG was stimulated with IPTG, and the cells were cultured for an additional 16 h, after which they were pelleted via centrifugation at 10,000 × g for 10 min. The supernatant was recovered and filtered through a 0.2-µm filter before being concentrated 25× in a 3-kDa molecular weight cutoff concentration column (Millipore). rPNAG was then precipitated from solution via the addition of 2 vol of ethanol. Insoluble rPNAG was recovered by gentle centrifugation (<1,000 × g for <5 min), and the supernatant was discarded. The pelleted rPNAG was resuspended in PBS (1:50 original supernatant volume) and incubated overnight in 10 µg/mL lysozyme and 10 µg/mL DNaseI at 37 °C. The peptidoglycan- and DNA-digested rPNAG was then subjected to 100 µg/mL proteinase K (Promega) treatment overnight at 37 °C. The isolated rPNAG was then dialyzed into water to remove digested peptides and nucleotides using a 5-kDa Slide-A-Lyzer cassette (Thermo-Fisher) before being lyophilized. The purity of this sample was evaluated using a Bradford total protein assay and was determined to have <1% contaminating protein by mass. The method of purification of PNAG from the S. aureus MN8m strain and the chemical characterization of the resulting PNAG antigen were performed as described previously (34) and in SI Appendix, SI Materials and Methods.

ELISA.

PNAG-specific antibodies produced in immunized mice were measured via indirect ELISA using a modification of a previously described protocol (17). Briefly, sera were isolated from the collected blood draws after centrifugation at 2,200 × g for 10 min and stored at −20 °C; 96-well plates (Maxisorp; Nunc Nalgene) were coated with purified rPNAG (25 μg/mL in PBS, pH 7.4) and incubated overnight at 4 °C. The next day, plates were washed three times with PBST (PBS, 0.05% Tween-20, 0.3% BSA) and blocked overnight at 4 °C with 5% nonfat dry milk (Carnation) in PBS. Samples were serially diluted by a factor of five in triplicate between 1:100 and 1:7,812,500 in blocking buffer and added to the plate for 2 h at 37 °C. Plates were washed three times with PBST and incubated for 1 h at 37 °C in the presence of one of the following HRP-conjugated antibodies (all from Abcam and used at 1:10,000 dilution): goat anti-mouse IgG, anti-mouse IgG1, anti-mouse IgG2a, anti-mouse IgA, and anti-mouse IgM. After three additional washes with PBST, 3,3′-5,5′-tetramethylbenzidine substrate (1-Step Ultra TMB-ELISA; Thermo-Fisher) was added, and the plate was incubated at room temperature for 30 min. The reaction was halted with 2 M H₂SO₄, and absorbance was quantified via microplate spectrophotometer (Molecular Devices) at a wavelength of 450 nm. Serum antibody titers were determined by measuring the lowest dilution that resulted in signal 3 SDs above no serum background controls. Statistical significance was determined using ANOVA and the Tukey–Kramer post hoc honest significant difference test and compared against the PBS and empty OMV control cases. OmpA-specific serum IgG antibodies were measured in a similar way with the following modifications: purified OmpA (ProSpec) was coated on high-binding 96-well plates (Corning), and sera from individual mice in each group were pooled and serially diluted by a factor of 10 in triplicate from 1:100.

Opsonization Phagocytosis and Killing Assay.

Opsonization phagocytosis and killing assay was performed as described previously using S. aureus CP8 strain MN8 (6). Bacteria were grown overnight on tryptic soy blood agar and suspended to a final concentration of 2 × 10⁷ cfu/mL in sterile gelatin veronal buffer (GVB; Boston BioProducts). Serum samples taken 10 wk after immunization from unchallenged mice were heat inactivated via incubation at 56 °C for 30 min before use. Human mAb F598 to PNAG was used as a positive control for bactericidal activity, while mAb F429 to P. aeruginosa alginate was used as a negative control. Serum samples and mAbs were diluted 1:30 in sterile GVB and then serially diluted twofold to a maximum dilution factor of 1:120 (10 µg/mL = 1:30 dilution of mAbs). Neutrophils were isolated from donor blood samples using the EasySep magnetic isolation kit (StemCell) and suspended to a final concentration of 10⁷ cells per 1 mL in GVB. Serum was also isolated from donor blood via centrifugation at 2,200 × g for 10 min to obtain autologous complement for use in the assay. Complement was diluted to 40% (vol/vol) in GVB and adsorbed to S. aureus MN8 at a final concentration of ~10⁹ cfu/mL for 30 min at 4 °C to remove any PNAG binding antibodies from the complement. Adsorbed complement was sterilized through a 0.5-µm filter and stored on ice before use. A fraction of adsorbed complement was heat inactivated at 56 °C for 30 min to serve as a control. Equal volumes (100 µL) of serum, complement, neutrophils, and bacterial dilution were combined and mixed end-over-end in 2.0-mL round-bottom microcentrifuge tubes for 3 h at 37 °C. Samples were diluted 1:10 and 1:100 in TBS with 0.05% Tween, and 10 µL of each dilution was spotted in duplicate on plates and allowed to trail down one-half the length of the plate to facilitate colony counting. Plates were then incubated overnight at 37 °C, and colonies were counted the next day. Percentage of bactericidal activity was calculated by comparing the mean number of cfu in an experimental group with the number of cfu determined for the mAb F429 control. Values were reported as averages of bactericidal activity from individual mice in each immunization group, and the error bars are the SD.

SBA.

SBA was performed using a modified version of previously reported protocols (5, 50). An overnight culture of F. tularensis LVS Iowa was used to make a suspension of 10⁵ cfu/mL in sterile PBS. Serum samples taken 10 wk after immunization from unchallenged mice were heat inactivated via incubation at 56 °C for 30 min before use. Human mAb F598 (1 mg/mL = 1:1 dilution of mAb) was used as a positive control for bactericidal activity. Serum samples and mAb F598 were diluted 1:2 in sterile PBS and then serially diluted threefold to a maximum dilution factor of 1:54 before being diluted 1:5 in the assay (final concentration range 1:10–1:270). Naïve baby rabbit complement (Cedarlane) or heat-inactivated complement (56 °C for 30 min) was used at a final concentration of 6% (vol/vol). Reactions were incubated at 37 °C for 30 min before 10 µL of each reaction was spotted on cysteine heart agar + 9% sheep’s blood and allowed to trail down one-half the length of the plate to facilitate colony counting. Plates were then incubated overnight at 37 °C, and colonies were counted the next day. Percentage of bactericidal activity was calculated by comparing the number of cfu in a spot with the number of cfu in the spot of the “no complement” control. Values shown are averages of bactericidal activity from individual mice in each immunization group, and the error bars are the SD.

We thank Roland Lloubes, Karen Elkins, and Wilfred Chen for strains and antibodies used in this work. We acknowledge the use of the Cornell Center for Materials Research Shared Facilities, which are supported through National Science Foundation (NSF) Materials Research Science and Engineering Center Program DMR-1120296. This material was based on work supported by NSF Graduate Research Fellowships (to T.C.S. and K.B.W.); Samuel C. Fleming Family Graduate Research Fellowships (to T.D.M. and K.B.W.); NIH Grants EB005669-01 (to D.P. and M.P.D.), AI044642 (to B.D.J.), AI057160 (to B.D.J.), EY016144 (to G.B.P.), and GM088905-01 (to M.P.D.); Project 14 of the Midwest Regional Center of Excellence for Biodefense and Emerging Infectious Disease Research (B.D.J.); and NSF Grants CBET-1159581 (to M.P.D.) and CBET-1264701 (to M.P.D.). This work was also supported by unrestricted support from Alopexx Vaccine, LLC (C.C.-B.).