|Home | About | Journals | Submit | Contact Us | Français|
Pneumococcal surface protein A (PspA) is highly immunogenic and can induce a protective immune response against pneumococcal infection. PspA is divided into two major families based on serological variability: family 1 and family 2. To provide broad protection, PspA proteins from pneumococcal strains Rx1 (family 1) and EF5668 (family 2) were combined to form two PspA fusion proteins, PspA/Rx1-EF5668 and PspA/EF5668-Rx1. Each protein was fused to a type II secretion signal and delivered by a recombinant attenuated Salmonella vaccine (RASV). Both PspA/Rx1-EF5668 and PspA/EF5668-Rx1 were synthesized in the RASV and secreted into the periplasm and supernatant. The fusion proteins reacted strongly with both anti-PspA/Rx1 and anti-PspA/EF5668 antisera. Oral immunization of BALB/c mice with RASV synthesizing either PspA fusion protein elicited serum immunoglobulin G (IgG) and mucosal IgA responses against both families of PspA. Analysis of IgG isotypes (IgG2a and IgG1) indicated a strong Th1 bias to the immune responses to both proteins. Sera from mice immunized with RASV synthesizing PspA/Rx1-EF5668 bound to the surface and directed C3 complement deposition on representative strains from all five PspA clades. Immunization with RASV synthesizing either protein protected mice against intraperitoneal challenge with Streptococcus pneumoniae WU2 strain (family 1), intravenous challenge with S. pneumoniae 3JYP2670 strain (family 2), and intranasal challenge with S. pneumoniae A66.1 (family 1). The PspA/Rx1-EF5668 protein elicited significantly greater protection than PspA/EF5668-Rx1, PspA/Rx1, or PspA/EF5668. These results indicate an RASV synthesizing a PspA fusion protein representing both PspA families constitutes an effective antipneumococcal vaccine, extending and enhancing protection against multiple strains of S. pneumoniae.
Streptococcus pneumoniae is a human pathogen causing significant morbidity and mortality worldwide, especially in developing countries. It causes respiratory infections, otitis media, sinusitis, and invasive diseases such as pneumonia, meningitis, and bacteremia. S. pneumoniae causes more than 1 million deaths worldwide every year among children under 5 years of age (8, 11, 20). The current 23-valent capsular polysaccharide vaccine elicits immunity in individuals greater than 2 years of age, and the current conjugate polysaccharide-protein pneumococcal vaccine provides protection for those under the age of 2 years (23, 26, 33). However, protection is restricted to only the limited number of serotypes included in the vaccine formulation (26), and the expensive production costs limit its use in developing countries. Moreover, serotype replacement has been observed in vaccinated populations and an increase in infections by pneumococcal serotypes not included in the 7-valent conjugated polysaccharide vaccine has been described recently (29, 56). In some countries, as many as 66% of childhood strains would not be covered (26, 45). Treatment of pneumococcal diseases has become more challenging due to the increase in multiple-drug-resistant pneumococcal strains (58). These issues reinforce the need for more affordable, broadly protective strategies for immunization against pneumococcal infection.
Several pneumococcal proteins have been under investigation as potential vaccine candidates, including pneumococcal surface protein A (PspA) (7, 10, 14), pneumococcal surface protein C (PspC) (12), and pneumolysin (1, 50). PspA is a virulence factor expressed by all clinical S. pneumoniae isolates. It consists of five domains: (i) a signal peptide, (ii) an α-helical and charged domain that bears a strong 7-residue periodicity typical of coiled-coil proteins (amino acids [aa] 1 to 288), (iii) a proline-rich region (aa 289 to 370) which spans the cell wall and is highly conserved in all S. pneumoniae strains, (iv) a choline-binding domain consisting of 10 20-aa repeats (aa 371 to 571) that anchors the protein to the cell surface, and (v) a C-terminal 17-aa tail (aa 572 to 589) (Fig. (Fig.1).1). The α-helical region is variable in length and amino acid sequence, but the antibodies against this region are protective and cross-reactive. PspA proteins have been grouped into three families encompassing six different clades based on the C-terminal 100 aa of the α-helical region (28). Family 1 is comprised of clades 1 and 2, family 2 is comprised of clades 3, 4, and 5, and family 3 consists of clade 6. S. pneumoniae strains expressing family 1 or 2 PspA proteins constitute 98% of clinical isolates (27, 28, 53). To accommodate this variability, it was proposed that a combination of two PspA antigens, one from PspA family 1 and one from PspA family 2, would elicit protection against the vast majority of S. pneumoniae strains (27, 28, 47). In addition to the α-helical region, the proline-rich domain has been shown to encode protective epitopes (S. Hollingshead, unpublished observation). This region of the protein is highly conserved compared to the α-helical region, making inclusion of the proline-rich domain important to achieve broad protection (4, 9, 28).
Complement-mediated opsonin-dependent phagocytosis is an important defense mechanism against pneumococcal infections. It activates both the classical and alternative complement pathways, depositing C3b on the pneumococcal surface (13, 34, 35). PspA inhibits complement activation (60), and anti-PspA antibodies can overcome this effect (53), leading to increased C3 deposition on the bacterial surface and enhanced clearance. Anti-PspA-directed C3 complement deposition has been correlated with protection against S. pneumoniae challenge in mice (19). Therefore, measurement of C3 complement deposition on the pneumococcal surface directed by sera from vaccinated individuals could be an important correlate of protection.
Previous work in our laboratory demonstrated that recombinant avirulent Salmonella enterica serovar Typhimurium vaccines (RASVs) can be used to deliver PspA cloned from S. pneumoniae strain Rx1 (family 1) and induce protection in mice against challenge with homologous family 1 S. pneumoniae strain WU2 (38-40, 48, 64). Using RASV to deliver antigens has many advantages, including low-cost vaccine production, needle-free delivery, and induction of strong mucosal immunity (16, 18). In this article, gene fragments encoding the α-helix domain of PspA from family 1 strain Rx1 and the α-helix domain and proline-rich region of family 2 strain EF5668 were used to construct gene fusions encoding two PspA fusion proteins, PspA/Rx1-EF5668 and PspA/EF5668-Rx1. These gene fusions were expressed and delivered by an RASV strain designed to regulate antigen expression by the availability of arabinose, resulting in regulated delayed antigen synthesis, to enhance and extend protection against S. pneumoniae clinical strains.
The bacterial strains and plasmids used in this study are listed in Table Table1.1. Escherichia coli and S. Typhimurium cultures were grown at 37°C in LB broth or on LB agar plates (5). When required, antibiotics were added to culture media at the following concentrations: ampicillin, 100 μg/ml; kanamycin, 50 μg/ml; and tetracycline, 12.5 μg/ml. Diaminopimelic acid was added (50 μg/ml) for the growth of Asd− strains (38). S. pneumoniae cells were maintained as frozen stocks (−80°C) in Todd-Hewitt broth supplemented with 0.5% yeast extract (THY) and 10% glycerol. S. pneumoniae was cultured on brain heart infusion agar containing 5% sheep blood or in THY in an anaerobic container (10).
All cloning procedures were performed with E. coli strain χ6212 grown in LB medium. DNA fragments encoding portions of the N-terminal regions of EF5668 pspA were amplified by PCR using primers 1 (5′-CAGGAATTCAACCAGTCTAAAGCTGAGAAAGAC-3′) and 2 (5′-GATAAGCTTATTATTGGTGCAGGAGCTGGTTGC-3′) from S. pneumoniae EF5668 to form pYA4325. The pspA gene (aa 4 to 417) of S. pneumoniae EF5668 was codon optimized for better expression in Salmonella, specifically codons 51 (AGA to CGT), 57 (ATA to ATC), 80 (AGA to CGT), 87 (ATA to ATC), 105 (CGA to CGT), 151 (ATA to ATC), 192 (AGA to CGT), and 231 (ATA to ATC), and cloned into plasmid pYA3493 to form pYA4326 (Table (Table1).1). Codon-optimized EF5668 pspA was PCR amplified by primers 2 (described above) and 3 (5′-TGACTGCAGAGTCTCTTCTTCATCTCCATCAGG-3′) using pYA4326 as the template. The resulting PCR product, encoding aa 4 to 417 of EF5668 PspA, and plasmid pYA3802, which encodes aa 3 to 285 of Rx1 PspA, were digested with PstI and HindIII and ligated to form pYA4432. EF5668 pspA was PCR amplified by primers 1 (described above) and 4 (5′-GATGAATTCTGGTGCAGGAGCTGGTTGCTC-3′). The resulting PCR product and plasmid pYA4088 were digested with EcoRI and ligated to form pYA4550 (Fig. (Fig.2).2). Transformations of E. coli and Salmonella were done by electroporation (Bio-Rad, Hercules, CA). Transformants containing Asd+ plasmids were selected on LB agar plates without diaminopimelic acid. Only clones containing the recombinant plasmids were able to grow under these conditions (17, 22). All constructs were confirmed by DNA sequencing. Nucleotide sequencing reactions were performed by the sequencing lab at Arizona State University using ABI Prism fluorescent Big Dye terminators according to the instructions of the manufacturer (PE Biosystems, Norwalk, CT).
Synthesis of PspA in Salmonella vaccine strains was evaluated by Western blotting essentially as described previously (64), except that PspA/EF5668-specific antibody raised in rabbits injected with a purified His-tagged PspA/EF5668 (aa 79 to 353) was used for some assays. Protein stability of PspA fusions was evaluated as follows. χ9241(pYA4432) and χ9241(pYA4550) were grown overnight in LB broth at 37°C. The overnight cultures were diluted 1:20 into fresh medium the next day and grown at 37°C to an optical density at 600 nm (OD600) of 1.0. The culture was split into two tubes. Chloramphenicol was added to one tube to a final concentration of 100 μg/ml, and incubation of both tubes was continued. One-milliliter samples were taken at 1, 2, 3, 4, 6, and 18 h, and PspA levels were evaluated by Western blot analysis.
Periplasmic proteins were isolated by a lysozyme-osmotic shock method (61), and cell fractions were prepared and analyzed as previously described (64). To evaluate protein secretion, supernatant samples were taken 3 and 6 h after dilution of the overnight culture and evaluated by Western blotting. Purification of recombinant His-tagged PspA/Rx1 (aa 1 to 302) and His-tagged PspA/EF5668 (aa 4 to 417) for analysis by enzyme-linked immunosorbent assay (ELISA) was performed as previously described (64).
Inbred 7-week-old female BALB/c mice were deprived of food and water for 6 h before oral immunization. The recombinant Salmonella strains χ9241(pYA4088) (64), χ9241(pYA4432), χ9241(pYA4326), and χ9241(pYA4550) were grown in LB with 0.05% arabinose to an OD600 of 0.8. Cultures were centrifuged at 4,000 × g at room temperature and suspended in buffered saline containing 0.01% gelatin (BSG) (15) to a final concentration of 5 × 1010 CFU/ml. Twenty microliters (1 × 109 CFU) was orally administered to BALB/c mice on days 1, 7, and 42. RASV strain χ9241(pYA3493) was used as the vector control. Food and water were returned to the mice after 30 min. Blood samples were taken by submandibular bleeding at 2, 4, 6, 7, and 8 weeks after primary immunization. After incubation at 37°C for 60 min, blood was centrifuged at 4,000 × g for 5 min. The serum was removed and stored at −70°C. Vaginal secretion specimens were collected in a 50-μl BSG wash and stored at −20°C (38, 66).
The ELISA used to assay antibodies against PspA in vaginal secretions and serum was essentially performed as described previously (38).
To assess the ability of the RASVs to cross-protect the immunized mice against different families of S. pneumoniae, immunized and control mice were challenged intraperitoneally (i.p.) with 2 × 104 CFU of family 1 strain WU2 (200 50% lethal doses [LD50s]) (38, 39, 64) or intravenously (i.v.) with 1 ×106 CFU of family 2 strain 3JYP2670 (100 LD50s) (54) in 200 μl of BSG. To evaluate protection against intranasal (i.n.) challenge, 1 ×108 CFU of S. pneumoniae family 1 strain A66.1 (20 LD50s) (2, 62) in 20 μl of BSG was administered. All challenges were done 2 weeks after the final boost. Mortality was monitored for 3 weeks following pneumococcal challenge.
Sera used for these assays were taken from mice 7 weeks after the primary immunization. To assess antibody binding, S. pneumoniae strains were grown in THY media to a concentration of 1 × 108 CFU/ml and harvested by centrifugation at 2,000 × g for 2 min. The pellets were washed once with phosphate-buffered saline (PBS), resuspended in the same buffer, and incubated in the presence of 20% pooled sera from immunized mice for 30 min at 37°C. After another wash with PBS, the samples were incubated with 100 μl of fluorescein isothiocyanate (FITC)-conjugated goat anti-mouse immunoglobulin G (IgG) Fc (MP Biomedicals) diluted 1:1,000 on ice for 30 min in the dark. Samples were analyzed with a Cytomics FC 500 (Beckman Coulter, Inc., Fullerton, CA).
For the complement deposition assay, we used a modified version of the method described by Ren et al. (53). Complement in sera from immunized mice was inactivated by incubation of sera at 56°C for 30 min. Bacterial pellets were washed once, centrifuged, and resuspended in PBS. Samples (80 μl) were incubated in the presence of complement-depleted anti-PspA sera at a final concentration of 10% for 30 min at 37°C. Bacteria were then washed once with PBS, resuspended in 90 μl of PBS-bovine serum albumin buffer, and incubated in the presence of fresh-frozen naïve BALB/c mouse serum at 37°C for 30 min. After another wash with PBS, the samples were incubated with 100 μl of FITC-conjugated goat antiserum to mouse complement C3 (MP Biomedicals) at a dilution of 1:1,000 on ice for 30 min in the dark. Finally, the bacteria were washed two more times with PBS, resuspended in 1% formaldehyde, and stored at 4°C in the dark until analysis with a Cytomics FC 500 (BD Biosciences).
An analysis of variance (SPSS Software), followed by Tukey's method, was used to evaluate differences in antibody titer, discerned to 95% confidence intervals. The Kaplan-Meier method (GraphPad Prism; GraphPad Software) was applied to obtain the survival fractions following i.p., i.v., or i.n. challenge of orally immunized mice.
We constructed two protein fusions combining the α-helical domain of PspA from Rx1 with the proline-rich and α-helical domains of PspA from EF5668 (Fig. (Fig.1).1). In one protein, PspA/Rx1-EF5668, the PspA/Rx1 α-helical domain lies at the amino-terminal end, and in the other protein, PspA/EF5668-Rx1, the EF5668 α-helical domain lies at the amino-terminal end. The genes encoding these two fusion proteins were cloned into Asd+ expression plasmid pYA3493 to yield plasmids pYA4432 and pYA4550 (Fig. (Fig.2).2). Each PspA fusion protein was fused to the type 2 secretion signal from β-lactamase (bla SS) to direct protein secretion to the periplasm and outside the cell. Fusions of this type have been shown to elicit higher antibody titers against the fused antigen and elicit greater protection in mice than when the antigen is expressed only in the cytoplasm (37). These two plasmids and plasmids pYA4326, carrying bla SS pspA/EF5668, and pYA4088, carrying bla SS pspA/Rx1, were moved into attenuated S. Typhimurium strain χ9241 by electroporation.
The amounts of PspA antigens produced in S. Typhimurium χ9241 harboring pYA4088, pYA4432, pYA4550, or pYA4326 were evaluated by Western blotting (Fig. (Fig.3).3). Cell lysates from the strains carrying pYA4432 (PspA/Rx1-EF5668) and pYA4550 (PspA/EF5668-Rx1) reacted strongly with both anti-PspA/Rx1 and anti-PspA/EF5668 antisera, whereas PspA/EF5668 reacted only with anti-PspA/EF5668 and PspA/Rx1 reacted only with anti-PspA/Rx1 (Fig. (Fig.3).3). However, some PspA-positive bands were smaller than the expected size of the full-length fusion protein, indicating that proteolytic degradation had occurred. While the presence of these extra bands is not ideal, we have seen similar results in extracts from E. coli synthesizing these same proteins, but not in full-length PspA from S. pneumoniae extracts (unpublished results). To examine this further, we performed a protein stability test. Our results showed that the full-length PspA fusion proteins were stable over the 16-h course of the experiment (data not shown). Some of the shorter cross-reactive peptides disappeared after 16 h, but most of the short fragments were stable. The presence of so many stable bands makes it unlikely that they are all the result of proteolysis. It is more likely that the observed fragments are due to premature transcriptional or translational stops.
Secretion of antigens synthesized in RASV enhances immune responses and protection against challenge (38, 64). We have previously shown that the PspA/Rx1 synthesized in χ9241(pYA4088) is secreted into the periplasm and supernatant (64). To examine secretion of the PspA fusion proteins, subcellular fractions, including cytoplasm, periplasm, and culture supernatants from RASV strain χ9241 harboring pYA4432 (PspA/Rx1-EF5668) or pYA4550 (PspA/EF5668-Rx1) were prepared by a modification of the lysozyme-osmotic shock method (38). PspA/Rx1-EF5668 and PspA/EF5668-Rx1 were detected in the periplasmic fraction and culture supernatant, indicating that the bla SS signal sequences can facilitate the secretion of both PspA fusion proteins (data not shown). Approximately 50% of the fusion proteins PspA/Rx1-EF5668 and PspA/EF5668-Rx1 was found in the periplasmic fraction and supernatant, consistent with previous results with PspA/Rx1 (64).
Strain χ9241 carries the ΔrelA198::araC PBAD lacI TT mutation in which expression of lacI is under the control of the arabinose-regulated araC PBAD promoter. The LacI-repressible Ptrc promoter drives pspA expression in all of the pspA-expressing plasmids used in this study (Table (Table11 and Fig. Fig.2).2). This combination of chromosomal lacI expression and Ptrc transcribed antigen genes has been termed “delayed antigen synthesis” (64). We confirmed that synthesis of all of the PspA constructs is regulated by arabinose availability by Western blot analysis, as shown previously for χ9241(pYA4088) (64) (data not shown).
To investigate the immunogenicity of each of the PspA proteins delivered by RASV, we orally inoculated groups of BALB/c mice with three doses of S. Typhimurium χ9241(pYA4088), χ9241(pYA4432), χ9241(pYA4326), or χ9241(pYA4550) on days 1, 7, and 42. Serum immunoglobulin G (IgG) responses to PspA/Rx1 and PspA-EF5668 from immunized mice were measured by ELISA (Fig. (Fig.4).4). IgG responses to PspA were observed after 2 weeks postimmunization and increased over time. Maximal anti-PspA IgG levels were detected at 6 to 8 weeks post-primary immunization, similar to previous results (38). All of the vaccine groups had significantly higher anti-PspA/Rx1 antibody titers than mice immunized with the vector control strain χ9241(pYA3493) and PBS control mice. Mice immunized with strain χ9241(pYA4088) (PspA/Rx1) or χ9241(pYA4432) (PspA/Rx1-EF5668) achieved higher anti-PspA/Rx1 IgG titers than mice immunized with χ9241(pYA4550) (PspA/EF5668-Rx1) or χ9241(pYA4326) (PspA/EF5668) (P < 0.05). The endpoint titers of mice immunized with χ9241(pYA4432) (PspA/Rx1-EF5668) at 8 weeks were not significantly different from those for mice immunized with χ9241(pYA4088) (PspA/Rx1) (P > 0.05).
All PspA-vaccinated mice produced antibody that reacted with PspA/EF5668 (Fig. (Fig.4).4). The anti-PspA/EF5668 titers in mice immunized with χ9241(pYA4326) (PspA/EF5668), χ9241(pYA4432) (PspA/Rx1-EF5668), or χ9241(pYA4550) (PspA/EF5668-Rx1) were not significantly different from each other but were significantly higher than those in mice immunized with χ9241(pYA4088) (PspA/Rx1) (P < 0.05).
No anti-PspA IgG was detected in sera obtained from mice immunized with the vector control or PBS. The anti-Salmonella outer membrane protein IgG responses in all groups including the vector control were similar both in kinetics and titer at 8 weeks and were not significantly different (P > 0.05) (data not shown). These results indicate that PspA fusion protein Rx1-EF5668(pYA4432) delivered by strain χ9241 induced high antibody titers against both PspA/Rx1 and PspA/EF5668.
The immune responses to Salmonella fusion PspA were further examined by measuring the levels of IgG isotype subclasses IgG2a and IgG1 in serum 7 weeks after primary immunization and 1 week after the final boost. Th1 helper cells direct cell-mediated immunity and promote IgG class switching to IgG2a, and Th2 cells provide potent help for B-cell antibody production and promote IgG class switching to IgG1 (49, 57). The IgG2a titers to PspA in all groups were higher than IgG1 titers, indicating that all of the Salmonella vaccines induced a strong Th1 response against PspA/Rx1 or PspA/EF5668 (Fig. (Fig.5).5). Th1-type dominant immune responses are frequently observed after immunization with attenuated Salmonella (46), but inclusion of a sopB mutation shifts it to a mixed Th1-Th2 response (40).
IgA against PspA/Rx1 and PspA/EF5668 was detected in vaginal fluids from all mice immunized with Salmonella expressing pspA fusions (Fig. (Fig.5C).5C). Mice have a common mucosal system (44) that facilitates the production of antigen-specific antibody responses at mucosal sites distant from the site of mucosal immunization, including both the upper respiratory and genital tract (30). For example, RASV expressing pspA can be administered by oral, i.n., intravaginal, or i.p. routes and elicit strong mucosal responses against PspA in vaginal secretions (59). Conversely, immunization of mice either orally or i.n. with attenuated Salmonella expressing heterologous antigens results in the production of antigen-specific antibodies at distant mucosal sites (31, 36, 63), although there can be antigen-dependent differences in the magnitude of the responses between sites—typically <10-fold (31, 36, 59, 63). Therefore, we used vaginal washes as a surrogate for nasal secretions, as it is a convenient way to obtain multiple samples from the same animal and it allowed us to keep the animal alive for challenge studies. The IgA responses showed strong PspA family dependence. The RASV synthesizing PspA/Rx1 induced a strong IgA response against PspA/Rx1 and a weak response against PspA/EF5668. Similarly, the RASV synthesizing PspA/EF5668 induced a stronger IgA response against PspA/EF5668 than it did against PspA/Rx1. Both fusion proteins elicited similar IgA responses: a strong response against PspA/Rx1, similar to that in mice immunized with PspA/Rx1 only, and a weaker response against PspA/EF5668.
To investigate the ability of anti-PspA antibody to bind intact pneumococci, week 8 sera from PspA-immunized mice were incubated with S. pneumoniae expressing pspA from clades 1 to 5, which includes families 1 and 2 (Table (Table1)1) (4, 28). After incubation with an FITC-labeled secondary anti-mouse IgG antibody, the percentage of fluorescent bacteria in each group was measured by flow cytometry (Fig. (Fig.6A6A).
The anti-PspA antibodies showed family dependence on binding to the surface of S. pneumoniae. Sera from mice immunized with χ9241(pYA4088) (PspA/Rx1) bound to family 1 strains L81905 (clade 1; 53.9%) and D39 (clade 2; 37.4%) more strongly than sera from mice immunized with χ9241(pYA4326) (PspA/EF5668), expressing family 2 pspA, but also bound reasonably well to family 2 strains EF3269 (clade 3; 35.8%), and ATCC 6303 (clade 5; 30.6%). Only weak surface binding was observed to strain 3JYP2670 (clade 4; 15.0%).
Sera from mice immunized with χ9241(pYA4326) (PspA/EF5668) bound family 2 strains more strongly than sera from mice immunized with χ9241(pYA4088) (PspA/Rx1). Conversely sera from mice immunized with RASV synthesizing PspA/EF5668 did not bind PspA family 1 strains L81905 and D39 as well as sera from mice immunized with χ9241(pYA4088) (PspA/Rx1). Interestingly, while binding to L81905 (clade 1; 13.6%) was weak, we observed some binding to D39 (clade 2; 24.0%).
Sera from mice immunized with χ9241(pYA4432) (PspA/Rx1-EF5668) showed strong surface binding to both family 1 strains L81905 (76.9%) and D39 (80.7%) and family 2 strains EF3269 (96.8%), 3JYP2670 (79.4%), and ATCC 6030 (43.4%). Sera from mice immunized with χ9241(pYA4550) (PspA/EF5668-Rx1) also showed strong surface binding to the family 1 strains L81905 (47.8%) and D39 (62.1%) and to the family 2-bearing strains EF3269 (62.1%) and 3JYP2670 (67.8%), although binding to strain ATCC 6030 (22.4%) was only about half the level observed for χ9241(pYA4432) immune sera. Surface binding by anti-PspA/Rx1-EF5668 antibody was always greater than binding by anti-PspA/EF5668-Rx1 sera.
Complement-mediated opsonin-dependent phagocytosis is an important defense mechanism against pneumococcal infections. C3 complement deposition is the key process leading to complement activation (34, 35), so we determined the ability of sera from immunized and control mice to direct complement deposition on the surface of S. pneumoniae strains from each clade. Pneumococci were incubated with decomplemented immune mouse sera, washed, incubated with 10% fresh-frozen control mouse serum, washed, and labeled with FITC-conjugated goat anti-mouse C3. The percentage of bacteria coated with C3 was determined by flow cytometry (Fig. (Fig.6B6B).
Antibodies induced against PspA/Rx1 increased by approximately twofold or greater the percentage of C3-positive cells for pneumococcal strains L81905 (clade 1; 98.7%), D39 (clade 2; 69.3%), EF3269 (clade 3; 40.2%), and ATCC 6303 (clade 5; 28.0%) compared to control sera. No increase was observed for strain 3JYP2670 (clade 4) compared to the control.
Anti-PspA/EF5668 serum did not enhance C3 deposition on the clade 1 (L81905; 24.9%) strain compared to the control. This serum increased the percentage of C3-positive cells by two- to fivefold for clade 2 (D39; 54.4%), clade 3 (EF3269; 59.5%), clade 4 (3JYP2670; 66.9%), and clade 5 (ATCC 6303; 31.9%) strains.
Antibodies raised against both fusion PspA/Rx1-EF5668 and fusion PspA/EF5668-Rx1 (Fig. (Fig.6)6) strongly augmented the percentage of cells with surface-bound C3 on strains expressing family 1 and 2 PspAs. Anti-PspA/Rx1-EF5668 serum and anti-PspA/EF5668-Rx1 serum behaved similarly in this assay, inducing a three- to fivefold enhancement of C3 deposition on all five test strains, except for the case of anti-EF5668-Rx1, in which the enhancement on clade 2 strain D39 was less than twofold. This result was surprising, since this serum bound avidly to the surface of strain D39 (Fig. (Fig.6A).6A). In each case, C3 deposition directed by anti-PspA/Rx1-EF5668 serum was slightly greater than that by anti-PspA/EF5668-Rx1 serum in all PspA clades except clade 3.
To determine whether the PspA fusions delivered by RASV provided protection across S. pneumoniae families, we challenged immunized mice with strains from each family. One group of orally immunized BALB/c mice was challenged i.p. with 200 LD50s of S. pneumoniae WU2 (family 1). All RASVs synthesizing PspA provided significant protection against family 1 pneumococcal challenge compared with vector and PBS controls (P < 0.01) (Fig. (Fig.7A).7A). While the PspA/EF5668 vaccine, χ9241(pYA4326), was cross-protective, it was the least efficacious of the vaccine strains tested and showed significantly lower protection than PspA/Rx1 and two fusion PspAs (P < 0.05). Notably, the RASV synthesizing PspA/Rx1-EF5668, χ9241(pYA4432), had the greatest efficacy, providing significantly greater protection than any of the other RASVs (P < 0.01).
Another group of immunized mice were challenged i.v. with 100 LD50s of S. pneumoniae 3JYP2670 (family 2) (Fig. (Fig.7B).7B). All vaccine strains expressing pspA provided significant protection against family 2 pneumococcal challenge compared with the controls (P < 0.01). As seen in the previous challenge experiment, the level of protection was influenced by PspA family. Strain χ9241(pYA4088), which synthesizes family 1 PspA/Rx1, provided the weakest level of protection, and protection was significantly lower than that provide by PspA/EF5668 and two fusion PspAs (P < 0.05). In addition, as in the previous experiment, strain χ9241(pYA4432) (PspA/Rx1-EF5668) provided significantly greater protection than the other vaccine strains (P < 0.01).
To evaluate protection in a pneumonia model, mice orally immunized with χ9241(pYA4432) (PspA/Rx1-EF5668), χ9241(pYA4550) (PspA/EF5668-Rx1), or χ9241(pYA3493) (control) were challenged i.n. with 20 LD50s of S. pneumoniae strain A66.1 (family 1). RASV synthesizing either PspA fusion protein provided complete protection that was significantly greater than those of the vector-only and PBS controls (P < 0.01) (Fig. (Fig.7C7C).
Taken together, these results show that RASV strain χ9241 expressing PspA fusions combining family 1 and family 2 proteins provided protection against family 1 and family 2 pneumococcal challenges. The PspA fusion (Rx1-EF5668) provided significantly greater protection against challenge with both family 1 and family 2 strains by two of the three challenge routes.
The pspA gene has a mosaic structure and shows some antigenic diversity among strains, leading to a grouping of most S. pneumoniae strains, based on variations in the α-helical region of the protein, into two families consisting of five clades (28). It has been proposed that the inclusion of the α-helical regions from both families would provide protection against nearly all S. pneumoniae strains (19, 27, 32, 47, 55). In this article, we investigated the potential of two PspA fusion proteins comprised of PspA fragments from families 1 and 2 delivered by an RASV to elicit serum antibodies able to bind to and provide protection against challenge by both family 1 and family 2 strains. Sera against the single PspA fragments—PspA/Rx1 and PspA/EF5668—reacted more strongly with strains within the same family than with strains from the other family (Fig. (Fig.6A).6A). However, some cross-reactivity between families was observed. Antibodies induced against the fusions PspA/Rx1-EF5668 and PspA/EF5668-Rx1 were strongly reactive with strains from both PspA families.
The proline-rich domain is highly conserved in all of the pneumococci. This domain was originally believed to be localized in the cell wall due to its similarity to other proteins from gram-positive bacteria (65). It was subsequently suggested that immunization with the proline-rich domain may protect mice against pneumococcal challenge (4, 9, 28). The proline-rich domain may also carry protective epitopes that may cross-protect against a variety of S. pneumoniae strains (S. Hollingshead, personal communication). In this study, the proline-rich domain from the S. pneumoniae EF5668 PspA protein was included in fusion construction to extend protective coverage. In preliminary experiments, a fusion PspA protein without the proline-rich domain was poorly expressed in Salmonella and immunization with an RASV expressing this fusion gene provided only poor protection in mice (12). This suggests that the inclusion of longer PspA fragments (α-helix and proline-rich region), containing more conserved regions, not only is important for broadening cross-protection against strains of different PspA clades but also enhances fusion gene expression in Salmonella.
In a recent study, Darrieux et al. constructed two family 1-family 2 fusion proteins and PspA subclones including the α-helical regions and proline-rich regions from family 1 and family 2 strains (19). The proteins were purified from E. coli. Groups of mice were injected with three doses of each protein. Serum was evaluated for surface binding and complement deposition. The results in that study were similar to our results with regard to family-specific responses against clade 1 and clade 4 strains. However, in our study, we observed cross-clade binding and complement deposition against clade 2 and 3 strains that were not observed in the previous study. (Clade 5 strains were not examined.) This difference in results could be due to differences in fusion protein structure and amino acid sequence or due to the method of delivery: mucosal immunization with an RASV versus parenteral injection. We note that their fusion proteins were derived from different S. pneumoniae strains from those used here. Mice immunized with three doses of the purified proteins were partially protected against challenge with S. pneumoniae strains A66.1 (family 1, clades 1 and 2), 3JYP2670 (family 2, clade 4), and 679/99 (family 2, clade 3). However, neither protein protected well against all three strains. Our results showed that both fusion proteins provided significant protection against challenge, regardless of challenge strain or challenge route (Fig. (Fig.7).7). In addition, one of the fusion proteins we used, PspA/Rx1-EF5668, elicited a robust immune response, eliciting serum antibodies that bound avidly to strains from all five clades tested (Fig. (Fig.6A),6A), efficiently directed complement deposition on these strains (Fig. (Fig.6B),6B), and provided significantly greater protection to challenge than the other PspA proteins tested (Fig. (Fig.7).7). These results indicate that the PspA fusion protein maintained enough structural epitopes so that antibodies against them could bind to native PspA on intact bacteria. The binding of antibody to the bacteria was consistent with the ELISA results (Fig. (Fig.4).4). Taken together, these results indicate that PspA fusion delivery by RASV may lead to a more broadly protective immune response than parenteral injection. It will be interesting to determine whether this can be confirmed in future studies using identical protein fusions.
In Western blots of the RASV strains synthesizing fusion proteins PspA/Rx1-EF5668 and PspA/EF5668-Rx1, we observed a number of bands smaller than the expected 107-kDa size of the intact protein (Fig. (Fig.3).3). There were small bands reacting with anti-PspA/Rx1 and anti-PspA/EF5668 antibodies, although we could not determine whether any bands other than the size of the full-length protein reacted with both antibodies. It is not clear whether the small bands were due to limited proteolysis or to peptides truncated by early transcriptional or translational termination. Our results are consistent with the possibility that some of the extra bands are due to truncated protein synthesis, although it is likely that some bands are also due to proteolysis. Surprisingly, most of these short proteins were stable and secreted, indicating that they may have contributed to immunogenicity, since these vaccine strains were able to induce a strong, protective immune response in immunized mice.
C3 complement deposition on the bacterial surface is important for complement-mediated opsonin-dependent phagocytosis (53). Therefore, we investigated whether antibodies against fusion PspA could enhance C3 complement deposition on the pneumococcal cell surface. The ability of anti-PspA antibodies to increase complement deposition was dependent on the PspA family in the bacterium, although cross-reaction was observed for some strains. Antibody against fusion PspA/Rx1-EF5668 and PspA/EF5668-Rx1 led to efficient C3 complement deposition on the surface of all strains tested, regardless of family or clade (Fig. (Fig.6B).6B). All of the Salmonella vaccine groups induced a strong Th1 response in which the anti-PspA IgG2a/IgG1 ratio was fourfold or greater (Fig. (Fig.5).5). IgG2a is the isotype with the greatest capacity to mediate complement deposition onto the surface of bacteria (21, 24), and an increase in anti-PspA IgG2a has been correlated with increased C3 deposition on the S. pneumoniae cell surface (21). Therefore, our data indicate that the RASVs synthesizing PspA elicit a strong anti-PspA IgG2a response, precisely what is needed to direct complement deposition in the pneumococcal surface.
Immunization with RASV synthesizing single PspAs worked best against challenge with strains expressing pspA of the same family. PspA/Rx1 (family 1) and PspA/EF5668 (family 2) provided the best protection against pneumococcal strains WU2 (family 1) and 3JYP2670 (family 2), respectively. However, immunization with fusion PspA/Rx1-EF5668 and PspA/EF5668-Rx1 led to greater protection against challenge with both pneumococcal strains WU2 (family 1) and 3JYP2670 (family 2). Fusion PspA/Rx1-EF5668 provided significantly greater protection against two pneumococcal family strains than the other vaccines in both the i.p. (family 1) and i.v. (family 2) challenges (Fig. 7A and B). Both fusion proteins delivered by RASV, PspA/Rx1-EF5668 and PspA/EF5668-Rx1, induced complete protection against i.n. challenge with family 1 pneumococcal strain A66.1.
We observed a strong correlation between the anti-PspA serum titers, pneumococcal surface binding, and C3 complement deposition and survival against a challenge with different pneumococcal strains, suggesting that it is the capacity for these antibodies to recognize PspA and direct complement deposition that is the mechanism responsible for protection against a pneumococcal challenge. We conclude that delivering fusion PspA/Rx1-EF5668 by RASV provides a significant step toward extending and enhancing protection against all S. pneumoniae strains. However, despite the correlation between antibody responses and protection we observed, there is strong evidence that there is an antibody-independent component to protection mediated by CD4+ cells (43). Consistent with this notion, previous work in our laboratory showed that introduction of a ΔsopB mutation into strain χ9241 expressing pspA enhanced the population of PspA-responsive CD4+ cells with a concomitant increase in protective efficacy against S. pneumoniae challenge (40). Therefore, it is possible that introduction of a ΔsopB mutation into the RASV synthesizing a PspA fusion protein would serve to further enhance protective efficacy. Notably, the ΔsopB RASV expressing pspA also elicited higher anti-PspA IgG2a titers than an isogenic SopB+ RASV in immunized mice. We plan to fully examine the induction and importance of cell-mediated immunity in future studies, including the use of mice lacking CD4+ cells.
We evaluated the efficacy of RASV synthesizing single or fusion PspAs using three routes of infection. While the i.n. route better mimics the natural route of infection, i.p. and i.v. challenge routes model invasive disease that, while less frequent overall than pneumonia, is prevalent among infants and young children (6, 25). We chose to use both i.p. and i.v. routes of infection because there is evidence that some pneumococcal virulence genes, such as pspC, are differentially regulated depending on whether challenge is administered i.p. or i.v. (51). We also evaluated our vaccine by the i.n. route to model the natural route of infection. Immunization with strain χ9241(pYA4088) elicited protective immune responses against all three challenge routes (Fig. (Fig.7).7). Another interesting and important question is whether this vaccine can protect against colonization of the nasopharyngeal tract by S. pneumoniae. The effect of vaccination on colonization has been evaluated for several vaccine formulations (3, 41, 42). However, the desirability of this endpoint is controversial, as eliminating S. pneumoniae may provide a niche for other pathogenic organisms such as Staphylococccus aureus (52). Therefore, the goal for this study was to evaluate the capacity of our vaccine to prevent frank disease. Eventually, we plan to develop a single RASV strain for humans that coexpresses genes encoding a number of protective antigens, including pspA. The final vaccine formulation will be thoroughly characterized for its efficacy in the prevention of colonization, pneumonia, and sepsis by S. pneumoniae.
We thank Susan K. Hollingshead and David Briles from the University of Alabama at Birmingham for providing all of the pneumococcal strains and suggestions on challenge methods, Soo-Young Wanda and Karen Brenneman for suggestions on DNA construction and the animal experiments, Vidya Ananthnarayan and Qing Liu for protein purifications, Jacquelyn Kilbourne for expert assistance for animal testing, Tae-Ho Lee and Xiangmin Zhang for performing some gene codon optimization, and Karen Ellis for expert technical assistance.
This research project was supported by grants from the National Institutes of Health (R01 AI056289) and the Bill and Melinda Gates Foundation (#37863).
Editor: J. N. Weiser
Published ahead of print on 17 August 2009.