Capsular Polysaccharide-Fimbrial Protein Conjugate Vaccine Protects against Porphyromonas gingivalis Infection in SCID Mice Reconstituted with Human Peripheral Blood Lymphocytes

The effect of immunization with either a Porphyromonas gingivalis fimbrial protein, a capsular polysaccharide, or a capsular polysaccharide-fimbrial protein conjugate vaccine were compared in hu-PBL-SCID mice. A significantly higher human immunoglobulin G antibody response and the highest degree of in vivo protection against bacterial challenge was observed in the group immunized with the conjugate vaccine. It was concluded that capsular polysaccharide-fimbrial protein conjugate from P. gingivalis could potentially be developed as a vaccine against periodontal infection by P. gingivalis.Porphyromonas gingivalis has been implicated as one of the major periodontal pathogens, and specific humoral and cell-mediated immune reactions to this organism have been demonstrated in periodontal diseases (13, 16). Attempts to induce protection against experimental infection with P. gingivalis by active immunization procedures have been studied by immunization with selected cell wall fractions, outer membrane proteins, and capsular polysaccharides (CPS) of P. gingivalis (8, 14). While most of these approaches afforded significant levels of protection (8, 14), problems such as maintaining functional levels of specific antibodies for extended periods of time (immune memory) (9), the multiple antigenicity of various pathogenic organisms, and the inability to activate T-cell-dependent immune responses (12) remain to be overcome. One strategy may be to develop a conjugation vaccine composed of CPS coupled with an outer membrane protein of P. gingivalis which can function as an immunodominant antigen as well as a carrier protein to activate T-cell-dependent immune responses.An additional area of improvement in vaccine strategies is the development of an adequate animal model system for simulating humanized antibody responses. Conventional animal models have disadvantages, since the animals may be qualitatively different from humans with respect to oral microbial environments and histological components in the development of periodontal lesions, and the nature of animal immune functions differs from that of human immune responses. Also, immunogenetic makeup (i.e., immunoglobulin [Ig] allotypes) and control over Ig class and subclass responses differ in animals and humans. Recently, mice with severe combined immunodeficiency (SCID) were identified (3, 15). The SCID mice lack functional T and B cells due to a mutation affecting the recombinase system that impairs the rearrangement of antigen receptor genes in both T and B cells. As postimmunization levels of IgG subclasses in vaccinees or in hu-PBL-SCID mice were closely associated with human Ig allotypes (5, 10, 11), we reconstituted the SCID mouse phenotype with human peripheral blood lymphocytes (PBL) whose Ig allotypes were positive for the phenotype fnb. As an extension of our previous experiments (5), we evaluated the protective effect of a newly developed polysaccharide-fimbrillin (FIM) protein conjugate vaccine with hu-PBL-SCID mice.Twenty-six SCID mice (C.B.-17-scid; Charles River Japan, Inc., Kanagawa, Japan) initially examined for IgG against P. gingivalis whole cells were reconstituted with 0.5 ml of human PBL (8 × 10⁷/ml) from periodontally healthy donors who were positive for the Ig phenotype fnb (either agfnb or axgfnb). Two weeks after reconstitution with human PBL, the expression of human Ig allotype markers was identified by the hemagglutination inhibition assay described previously (4).CPS of P. gingivalis 53977 was prepared by a modification of the method previously described (14). Briefly, bacterial cells were suspended in water (0.2 to 0.4 g [wet weight]/ml) and extracted with an equal volume of 90% phenol for 20 min at 65 to 68°C. The aqueous phase was obtained by centrifugation at 4,000 × g and dialyzed against distilled water with Spectrapor 1 tubing. The dialyzed solution was brought to 0.15 M sodium chloride, 4 mM MgCl₂, 1 mM CaCl₂, and pH 7.5 with Tris-HCl, treated with RNase A (0.04 mg/ml) and DNase I (0.01 mg/ml) (Sigma, St. Louis, Mo.) for 2 h at 37°C, treated with proteinase K (0.04 mg/ml) for 1 h at 60°C, dialyzed against dH₂O, and lyophilized. The lyophilized extract was dissolved in 0.05 M Tris-HCl buffer (pH 9.5) containing 0.3% deoxycholate and 0.001 M trisodium EDTA, applied to a column of Sephacryl S-400 HR (1.0 by 47 cm) (Pharmacia, Piscataway, N.J.), and eluted with the deoxycholate-containing buffer. Fractions were assessed for lipopolysaccharide and CPS by double immunodiffusion in agarose, for LPS contamination by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, and for protein contamination. Fractions containing only CPS were pooled, sodium chloride was added to 0.15 M, and CPS was precipitated with 4 volumes of 95% ethanol. The precipitates were isolated by centrifugation, dissolved, dialyzed, and lyophilized.Fimbriae of P. gingivalis 381 were purified as follows (17). Briefly, cells were harvested by centrifugation and suspended in 20 mM Tris-HCl (pH 7.4)–0.15 M NaCl–10 mM MgCl₂ by repeated pipetting. The suspension was agitated by magnetic stirrer for 30 min, and the supernatant was obtained after centrifugation at 8,000 × g for 20 min. Ammonium sulfate was added to 40% saturation, precipitated proteins were collected by centrifugation, and the precipitate was dissolved in 20 mM Tris-HCl (pH 8.0) and dialyzed against 20 mM Tris-HCl (pH 8.0). The dialysate was clarified by centrifugation at 8,000 × g for 20 min and applied to a column of DEAE-Sepharose CL-6B (1.5 by 16 cm) (Pharmacia) equilibrated with the above-described buffer. The column was washed with 20 mM Tris-HCl, pH 8.0, and eluted with a linear gradient of 0 to 0.3 M NaCl. The 43,000-molecular-weight protein (43K protein) band was not detected in fractions eluted after 0.17 M NaCl. Fractions containing the 43K protein were concentrated by ammonium sulfate precipitation and dialyzed against 3 mM Tris-HCl (pH 8.0) or 3 mM sodium bicarbonate (pH 8.0).The carboxylate group of the P. gingivalis CPS was conjugated to free amino residues of either cationic bovine serum albumin (CPS-BSA) or 43-kDa P. gingivalis fimbrillin (CPS-FIM) via a 1-ethyl-3-(dimethylaminopropyl)-carbodiimidide (EDC) intermediate. A total of 1.5 mg of CPS was dissolved in 0.5 ml of EDC conjugation buffer (Pierce, Rockford, Ill.), and 2 mg of BSA (SuperCarrier; Pierce) or 2 mg of fimbrial protein dissolved in 0.2 ml of water was mixed and added to the EDC (10 mg in 1 ml of water) and reacted for 2 h at room temperature. The mixture was separated from carbodiimidide by gel filtration on a Sephacryl S-300 column (1.0 by 10 cm) with 0.9 M sodium chloride and 0.083 M sodium phosphate, pH 7.2. Fractions were screened by immunodiffusion in agarose gel with rabbit antisera to CPS, fimbrial protein, or BSA.Only mice with no detectable amount of murine IgG against P. gingivalis were included in the study. Two weeks following PBL reconstitution, mice were examined for the expression of human Ig allotypes. Three of 26 mice were found to be leaky; a total of 23 mice were used in the experiment. Group I (n = 6) was immunized with FIM, group II (n = 6) was immunized with the CPS-BSA vaccine, group III (n = 6) was immunized with the CPS-FIM conjugate vaccine, and group IV (control, n = 5) was immunized with BSA. Each immunization procedure consisted of two intraperitoneal injections (at 2-week intervals) with 0.2 ml of immunogen adjuvant (Imject Alum; Pierce) mixture. The final amount of immunogen was 10 μg. Two weeks after the final immunization, mice were challenged by two dorsal subcutaneous injections (0.1 ml each) of whole P. gingivalis 53977 cells (1 × 10¹¹/ml) and evaluated for protective effects for 3 weeks based on the following criteria: general appearance, cachexia, weight loss, size and nature of localized abscess formation, development and size of secondary lesion, and death. Preimmune, postimmune (2 weeks following final immunization), and postinfection (3 weeks following infection) total IgG and IgG subclass antibody titers were determined by enzyme-linked immunosorbent assay (ELISA) with an alkaline phosphatase assay system. Microtiter plates (96 well) were coated with 0.1 ml of antigen (10 μg/ml) diluted in 0.01 M phosphate buffer (pH 7.2). After overnight incubation at 4°C, the plates were washed three times with phosphate-buffered saline (PBS) containing 0.1% Tween 20. A total of 0.05 ml of mouse serum samples diluted in PBS containing 0.1% Tween 20 was added to each well and incubated for 2 h at room temperature. The plate was washed three times with PBS containing 0.1% Tween 20, and then 0.1 ml of four mouse anti-human IgG subclasses (affinity-purified monoclonal antibody, γ-chain-specific, IgG1; 8c/6-39, IgG2; HP-6014, IgG3; HP-6050, IgG4; HP-6025; Sigma) diluted in PBS containing 0.1% Tween 20 were added to each well and incubated for 2 h at room temperature. After being washed three times with PBS containing 0.1% Tween 20, 0.05 ml of goat anti-mouse IgG (heavy- and light-chain specific, affinity purified, alkaline phosphatase conjugated; Calbiochem, Basel, Switzerland) diluted in PBS containing 0.1% Tween 20 were added to each well and incubated overnight at room temperature. After the plates were washed, 0.1 ml of nitrophenyl phosphate (1 mg/ml in diethanolamine buffer, pH 9.8) was added to each well and incubated for 60 min, and 0.1 ml of 3 N NaOH was added to stop the color reaction. For total IgG antibody measurements, goat anti-human IgG (affinity purified, γ-chain specific, alkaline phosphatase conjugated; Calbiochem) was used. Optical densities were plotted as a function of serum dilution factor, regression analysis was performed, and reciprocals of the serum dilution factors at the x axis intersection of an optical density of 0.2 were expressed in ELISA units for each sample. For the comparison of antibody levels between groups or intervals, Student’s t test was done.Except for those mice which were found to be leaky (n = 3), all mice (n = 23) expressed human Ig allotypes, either axgfnbt or agfnbt, according to the donors’ allotypes, confirming our previous observation (5). IgG and IgG subclass titers are summarized in Table ​Table1.1. Both postimmune and postinfection IgG levels against whole cells increased significantly compared to baseline values in all groups. IgG levels to whole cells in group III were significantly higher than those of groups I, II, or IV throughout the experimental period. In groups I and III, IgG4 subclass antibody titers were higher at the postinfection phase than at the postimmunization phase. Our previous studies have shown that early-onset periodontitis patients, whose haplotype fnb frequency was significantly higher than that of the race- and age-matched control group, had significantly higher IgG2 and IgG4 levels to P. gingivalis (4, 6). It was reasoned that the conversion of IgG1-restricted responses to IgG4-restricted responses with prolonged proteineous antigenic stimulation might be also responsible for the higher IgG4 subclass levels (1). IgG2 subclass responses were elevated to the polysaccharide antigen, while IgG1 responses were elevated to the fimbrial antigen, similar to previous results following bacterial infection or vaccination (2, 7, 10, 11). While all mice immunized with the conjugate vaccine survived the high-dose bacterial challenge (2 × 10¹⁰ cells of P. gingivalis 53977), one-third of the mice in the other two experimental groups and all of the mice in the control group died. The magnitude of the humoral antibody response was highest and the in vivo protective effect was greatest with minimal weight loss in group III (Tables ​(Tables11 and ​and2).2). This observation implies that through the use of a sophisticated conjugational vaccine incorporating two kinds of immunodominant antigens (i.e., outer membrane fimbrial protein and CPS), protection against P. gingivalis infection can be enhanced. Moreover, the SCID mice reconstituted with PBL from donors of the same IG allotypes provided a genetics-based simulation model for investigating humanized antibody responses to various vaccine formulas. By establishing T-cell hybridomas from hu-PBL-SCID mice, we are attempting to identify antigenic epitopes for T-cell clonal activation and to identify heavy- and light-chain variable gene usage of the antibodies from the conjugate vaccine.

TABLE 1

Baseline, postimmunization, and postinfection IgG and IgG subclass levels for each mouse group (ELISA unit ± standard deviation)

Treatment and group	Mean Ig level ± SD
	IgGd	1	2	3	4	FIM	CPS
Baseline
I	27 ± 5
II	29 ± 6
III	29 ± 4
IV (control)	30 ± 5
Postimmune
I	232 ± 26	158 ± 17	21 ± 11	13 ± 11	11 ± 10	209 ± 20	25 ± 10
II	216 ± 22	41 ± 10	102 ± 22	20 ± 12	4 ± 5	77 ± 21	113 ± 26
III	298 ± 23c	131 ± 19	140 ± 31	31 ± 16	41 ± 17	123 ± 29	139 ± 13
IV (control)	45 ± 14	NDa	ND	ND	ND	21 ± 11	15 ± 9
Postinfection
I	286 ± 34	98 ± 13	28 ± 12	16 ± 9	124 ± 36	219 ± 30	36 ± 9
II	256 ± 27	65 ± 11	154 ± 23	31 ± 10	18 ± 12	81 ± 12	166 ± 24
III	451 ± 45c	108 ± 17	166 ± 21	34 ± 13	154 ± 26	251 ± 22	135 ± 19
IV (control)	NAb	NA	NA	NA	NA	NA	NA

Open in a separate window^aND, not determined due to undetectability. ^bNA, not available due to death. ^cIgG levels to whole cells were significantly higher than groups I, II, or IV (P < 0.05). ^dIgG levels to whole cells in postimmune and postinfection phases were significantly higher than baseline levels (P < 0.05). 

TABLE 2

Clinical course of postinfection mice from different immunization groups

Group no.	Primary lesion		Secondary lesion		Cachexia	Death (no. of mice/total)
	Onset (day)	Size (mm)	Onset (day)	Size (mm)
I	2	4 × 5	5	7 × 11	Moderate	2/6
II	2	4 × 6	5	6 × 13	Moderate	2/6
III	2	2 × 4	7	3 × 5	Minimal	0/6
IV	2	5 × 6	3	12 × 16	Severe	5/5

Open in a separate window