Evaluation of a monovalent companion animal periodontal disease vaccine in an experimental mouse periodontitis model

Vaccine 23 (2005) 3148–3156

Evaluation of a monovalent companion animal periodontal disease vaccine in an experimental mouse periodontitis model John Hardham a, ∗ , Maryke Reed a , Jason Wong a , Kendall King a , Bob Laurinat a , Cornelia Sfintescu b , Richard T. Evans b b

a Pfizer Inc., Veterinary Medicine Research and Development, 301 Henrietta Street, Kalamazoo, MI 49001, USA University at Buffalo, School of Dental Medicine, Department of Oral Biology, 219 Foster Hall, Buffalo, NY 14214, USA

Received 2 September 2004; received in revised form 17 December 2004; accepted 21 December 2004 Available online 19 February 2005

Abstract Periodontal disease in companion animals is clinically similar to that of human periodontal disease. Despite the usage of veterinary procedures and antibiotic therapy, the disease still remains as one of the most highly prevalent disorders seen by veterinarians. The goal of this study was to evaluate the immunogenic properties and vaccine performance of a monovalent canine periodontal disease vaccine in the mouse oral challenge model of periodontitis. Mice vaccinated subcutaneously with inactivated, whole-cell bacterin preparations of Porphyromonas gulae displayed both high titers of anti-P. gulae specific antibodies and significantly reduced alveolar bone loss in response to homologous, heterologous, and cross-species challenge. Based on the results of these studies, a periodontal disease vaccine may be a useful tool in preventing the progression of periodontitis in animals. © 2005 Elsevier Ltd. All rights reserved. Keywords: Periodontitis; Porphyromonas; Veterinary

1. Introduction The study of periodontal disease in humans has been of considerable interest throughout the past decade. Much progress has been made in the understanding of the disease etiology and the intricate interrelationships between the host and periodontal pathogens [1,2]. In addition, there has been steady progress in elucidating the disease mechanisms and virulence factors of the prime periodontal pathogen, Porphyromonas gingivalis [3–5]. Several laboratories have utilized a variety of small animal and non-human primate periodontal disease models to demonstrate efficacy of subunitor bacterin-based vaccines [6,7]. In contrast, there is comparatively little information available regarding periodontal disease in companion animals. It has been estimated that approximately 80% of dogs and cats demonstrate some degree of periodontal disease by four years of age [8]. One con∗

Corresponding author. Tel.: +1 269 833 0738; fax: +1 269 833 7636. E-mail address: [email protected] (J. Hardham).

0264-410X/$ – see front matter © 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.vaccine.2004.12.026

tributing factor to this high percentage is the lack of routine oral care. Both the progression and etiology of periodontal disease in companion animals appears to roughly parallel the disease path in humans [9]. Black pigmenting anaerobic bacteria (BPAB) have been isolated from the periodontal pockets of dogs and cats [10–15], sheep [16,17], and several wild animals [18]. While little has been done to definitively speciate many of these isolates, they appear to be predominantly Porphyromonas spp. [10,15]. However, distinct differences have been noted between human and canine Porphyromonas spp. Isogai et al. [15] and Harvey et al. [12] utilized biochemical testing to identify P. gingivalis-like organisms in the gingival crevicular spaces of dogs with periodontal disease. Harvey et al. [12] noted that “human” P. gingivalis isolates were catalase-negative whereas “veterinary” P. gingivalis isolates were catalase-positive. Other Porphyromonas spp. of veterinary origin have been identified, including Porphyromonas gulae, Porphyromonas canoris, Porphyromonas gingivicanis, and Porphyromonas crevioricanis [18–20]. Fournier et al. [18] recently suggested that many of the previously iden-

J. Hardham et al. / Vaccine 23 (2005) 3148–3156

tified veterinary P. gingivalis-like bacteria should be reclassified as P. gulae. Our laboratory has recently isolated bacteria from other genera from dog periodontal pockets. Bacteroides denticanium was recently identified and was found to be capable of inducing periodontal disease in mice [21]. In addition, we have identified a novel genospecies in the Porphyromonadaceae family, tentatively classified as Porphyromonas denticanis [21]. Data generated in our laboratory has indicated that the most frequently isolated BPAB in dog and cat periodontal pockets are P. gulae, Porphyromonas salivosa, and P. denticanis [21]. Each of these isolates was demonstrated to be pathogenic in the mouse model of periodontal disease [21]. As also noted by Fournier et al. [18], P. gingivalis was not frequently isolated from dog and cat periodontal pockets. The high incidence of periodontal disease in companion animals, the low compliance rate of oral care, the clinical difficulties associated with periodontal disease treatment, and potential systemic complications warrant the development of an efficacious vaccine. In this report, we assessed the performance of a P. gulae bacterin in the mouse oral challenge model of periodontal disease.

2. Materials and methods

3149

suspended using a disposable spreader. The cell concentration was then adjusted to approximately 2 × 1010 cells/ml with supplemented (1/2) BHI medium. An equal volume of 2% (w/v) carboxymethyl cellulose (CMC; Sigma) was added to yield a final cell concentration of approximately 1 × 1010 cells/ml in 1% CMC. 2.2. Bacterial fermentation Bacteria used to formulate the vaccines were propagated using New Brunswick BioFlo 3000 fermentors with vessels having a 5-l working volume. Batch or partial fed-batch fermentations utilized modified PYG medium. The fermentation vessels were poised pre-inoculation by sparging sterile medium with anaerobic gases to ensure a reduced environment. Fermentors were seeded either directly with thawed vials of seed stock or with liquid cultures grown in a Bactron IV anaerobe chamber. Volumetric rates of inoculum varied from 0.02 to 8% (v/v). Sparging was continued during fermentation to both ensure reduced medium as well as a means of pH control. Dilute ammonium hydroxide was added via a pump-assisted delivery mechanism to further control the pH. The fermentation vessels were temperature controlled at 37 ◦ C under moderate agitation. Fermentations were terminated between 42.75 and 87 h post-inoculation.

2.1. Bacteria 2.3. Bacterial inactivation P. gulae strain B43, P. salivosa strain B104, P. denticanis strain B106, and B. denticanium strain B78 were isolated from the gingival crevicular fluid of dogs with periodontal disease (Table 1) [21]. P. gulae strain B69 was isolated from the gingival crevicular fluid of a cat with periodontal disease [21]. Broth cultures of these bacteria were prepared by inoculating a 1 ml seed stock into 200 ml of modified Phytone Yeast Glucose (PYG) medium [21] followed by 48 h of incubation at 37 ◦ C in a Bactron IV anaerobic chamber maintained in an anaerobic state (5% H2 , 5% CO2 , 90% N2 ). Challenge material was prepared by inoculating halfstrength brain heart infusion (1/2BHI) agar plates (containing 10 mg/ml (w/v) yeast extract (Becton Dickinson), 0.15 mg/ml (w/v) hemin (Sigma)), and 0.075 ml/ml (v/v) menadione (Sigma) with 100 ␮l of the appropriate frozen stock and incubating for 48 h at 37 ◦ C in an anaerobic environment. Following incubation, 1 ml of supplemented (1/2) BHI medium was applied to the surface of the plates and the cells were

Fermentor grown P. gulae B43 broth cultures were inactivated using formalin-formaldehyde, heat, or aeration. Inactivation with formalin-formaldehyde was accomplished by the addition of 0.4% (v/v) formalin-formaldehyde (J.T. Baker; Phillipsburg, NJ) to a freshly harvested bacterial culture followed by incubation at room temperature for 24 h. Heat inactivation was carried out by incubating the freshly harvested culture at 60 ◦ C for 30 min with gentle agitation. Since P. gulae is an obligate anaerobe, exposure to oxygen for extended periods can kill the bacterial cells. A freshly harvested bacterial culture was sparged with air at 0.1 standard liters per minute at room temperature for 48 h with gentle agitation. Fermentor cultures at time of inactivation were determined to have optical density readings, taken at 600 nm, between 2.02 and 3.20. Total cell counts per milliliter of inactivated fermentor cultures were between 3.6 × 109 and 4.4 × 109 cells/ml.

Table 1 Bacterial strains utilized in this study Bacterial strain

Host

Tooth

Pocket depth

P. gingivalis ATCC 53977 P. gulae B43 P. gulae B69 B. denticanium B78 P. salivosa B104 P. denticanis B106

Human Dog Cat Dog Dog Dog

NAa NA Upper left canine Upper left pre-molar Lower left first molar Lower left first molar

NA NA 2 5 4 4

a

NA, not available.

3150

J. Hardham et al. / Vaccine 23 (2005) 3148–3156

2.4. Vaccine formulation

2.7. Detection of serum antibody responses

Formalin-, heat-, or aeration-inactivated cultures of P. gulae B43 were formulated into experimental vaccines based on the cell counts. Briefly, bacterial cells were harvested from the inactivated cultures by centrifugation at 4 ◦ C. The cell pellets were resuspended in Delbecco’s Phosphate Buffered Saline Solution (Invitrogen Corp.; Carlsbad, CA) to a final concentration of 2 × 1010 cells/ml and stored at 4 ◦ C until formulated. Vaccines were prepared by mixing the inactivated bacterial cells with either monophosphoryl lipid A and trehalose dicorynomycolate (MPL + TDM) adjuvant (Corixa; Seattle, WA) or 50 ␮g/ml (w/v) of Quil A (E.M. Sergeant; Clifton, NJ) and 50 ␮g/ml (w/v) cholesterol (Fabrichem; Fairfield, CT). All vaccines were formulated to contain a final dose of 1 × 1010 cells/ml. All vaccines were stored at 4 ◦ C until needed.

Serum antibody responses were determined by enzymelinked immunosorbant assay (ELISA). Various Porphyromonas antigen preparations including whole cell, formalininactivated P. gulae B43, recombinant P. gulae B43 FimA, or recombinant P. gulae B43 OprF were coated on Nunc MaxiSorp plates at 250 ng/well in 100 ␮l of Borate buffer, covered and incubated overnight at 4 ◦ C. Plates were washed three times with PBS-T (0.05%, v/v, Tween 20 in phosphate buffered saline), blocked with 200 ␮l of 5% (v/v) skim milk in PBS-T, covered and incubated for 1 h at 37 ◦ C in 5% CO2 . Serial dilutions of the serum samples were prepared in PBST starting at 1:100. After washing plates three times with PBS-T, the serially diluted serum samples were added to the plate in duplicate. Plates were covered and incubated for 1 h at 37 ◦ C in 5% CO2 . After three washes with PBS-T, 100 ␮l of Horse Radish Peroxidase-conjugated cocktail (goat antimouse IgG (H&L) and goat anti-rabbit IgG (H&L) diluted 1:1000 in PBS-T) was added. The plates were covered and incubated for 1 h at 37 ◦ C in 5% CO2 . After washing with PBS, the plates were developed using 100 ␮l/well of 2,2 -azinodi-3-ethyl-benzthiazoline-6-sulfonate (ABTS), and the absorbance read at 405/490 nm in an ELISA plate reader. Titers were determined to be the reciprocal of the dilution of the sample whose optical density was equal to 30% of the positive control. The positive control serum was a mouse serum pool previously determined to be reactive with P. gulae B43, rFimA, and rOprF.

2.5. Mice Three-week-old, age-matched male Balb/cCyJ mice (Jackson Laboratories; Bar Harbor, ME) were utilized for all vaccine studies. For uniformity in alveolar bone loss measurements, all mice were from litters born within 3 days of each other. Animals were housed in positive pressure barrier cage units. Food pellets, standard for the species, and water were provided ad libitum throughout the experiment. The bedding material utilized was granular Bed O’Cobs to minimize impaction in the gingival tissues. After receipt, all animals were acclimatized for 5–7 days, during which time the third molar eruption occurred. Thus, experimental variability due to alveolar bone remodeling was reduced. To lower the number of resident competing flora prior to infection, all animals were placed on a mixture of sulfamethoxazole and trimethoprim (10 ml/l drinking water; approximately 2 mg/ml, w/v, and 0.4 mg/ml, w/v, respectively) for 10 days. All animals were returned to normal drinking water for a 5-day washout period prior to infection.

2.8. Mouse oral challenge On day 28 of each vaccination/challenge study, groups of 16 mice were individually challenged by gavage either with a 0.5 ml suspension containing approximately 5.0 × 109 bacteria in 1% CMC or sterile saline. All challenges were repeated on day 30 and day 32. The study concluded on day 70, at which time all animals were humanely euthanized.

2.6. Immunization and sample collection

2.9. Tissue samples

Mice were immunized subcutaneously with 0.2 ml of sterile saline, saline plus adjuvant, or experimental vaccine on days 0 and 14 (or day 21 for antibody titer experiments). Each dose was divided into two 0.1 ml aliquots, which were delivered to the right and left scapular regions. Serum was collected prior to challenge and at the conclusion of the study. Oral swabs were taken at these same time points to determine the presence of the infecting organism. Oral swabs were suspended in 1 ml of supplemented (1/2) BHI medium, vortexed for 30 s, and 0.1 ml was plated on supplemented (1/2) BHI agar. The plates were incubated at 37 ◦ C in an anaerobic environment for 7 days. The BPAB were then enumerated. Since mice do not normally harbor BPAB, the total number of BPAB would be indicative of the level of challenge material remaining.

Following euthanasia on day 70, the mouse jaws were isolated and defleshed as previously described [22]. The degree of horizontal alveolar bone loss (cementoenamel junction to alveolar bone crest (CEJ–ABC) distance) was assessed microscopically at 14 separate sites along the maxillary buccal surface of each jaw as previously described [22]. The assessments were performed three times to minimize operator error. The alveolar bone loss was expressed as average bone loss/site/jaw in mm. 2.10. Data and statistics Personnel were blinded with respect to all study groups and samples. The study group sizes were determined using SamplePower version 2.0 (SPSS Inc.; Chicago, IL). Statisti-

J. Hardham et al. / Vaccine 23 (2005) 3148–3156

3151

Table 2 Serological response of mice vaccinated with the P. gulae B43 vaccinea Vaccine

Pre-vaccination titerb

Post-vaccination titer

Saline P. gulae B43 bacterin

50 50

50 38,945

a The vaccine was composed of formalin-inactivated P. gulae B43 at 1 × 1010 cells/ml, 50 ␮g/ml of QuilA, and 50 ␮g/ml cholesterol as the adjuvant. Mice were immunized as described. b A whole-cell P. gulae B43 ELISA was used to assess the serological response. The lowest dilution measured was 1:100. If a sample was negative at this dilution, then it was scored as a 50.

cal analysis of all data was performed using SigmaStat version 3.0 and SigmaPlot version 8.02 packages (Systat Software, Inc.; Point Richmond, CA).

3. Results 3.1. Induction of a P. gulae specific antibody response following immunization In order to evaluate the ability of a P. gulae B43 bacterin to induce a serum antibody response, two groups of mice were subcutaneously immunized with either saline or 2 × 109 formalin-inactivated P. gulae cells with 50 ␮g/ml QuilA and 50 ␮g/ml cholesterol. The mice received their primary immunization on day 0 followed by a boost on day 21. Serum was drawn on day 28. The immune responses were measured using an ELISA measuring antigen specific total serum immunoglobulin, IgG, IgM, and IgA. Table 2 shows the average pre- and post-vaccination titers for each group. All mice mounted a robust immune response against the homologous P. gulae B43 and demonstrated an approximate 3-log increase in titer. 3.2. Assessment of various inactivating agents Since the method of inactivating a bacterial culture can potentially have a significant impact on the performance of a vaccine, we sought to determine which of three inactivating agents would be best suited for a companion animal periodontal disease vaccine. The performance of three different vaccines prepared with P. gulae B43 cells inactivated with formalin, heat, or aeration was assessed using the mouse oral challenge model [22]. All vaccines were prepared to the same final cell concentration (1 × 1010 cells/ml) to minimize variability due to antigen load. The vaccines were adjuvanted with the RIBI MPL + TDM adjuvant (Corixa Corp.; Seattle, WA). Fig. 1 shows the average net bone loss of each of the five study groups. Mice vaccinated with the formalin-, heat-, and aeration-inactivated vaccines demonstrated a significant reduction (P = 0.050, 0.057, and 0.002, respectively) in bone loss resulting from oral challenge with virulent P. gulae B43. Table 3 shows the post-challenge recovery of the challenge organism. Relative to the non-vaccinated/challenged group, each of the vaccinated groups showed a reduction in the num-

Fig. 1. Comparison of various inactivating agents in the mouse oral challenge model of periodontitis. Groups of mice were immunized with either saline or various P. gulae B43 bacterins and challenged with virulent P. gulae B43. Forty-two days following infection, the mice were euthanized, the jaws defleshed, and the CEJ–ABC distance was determined. The net bone loss is shown (in mm). The net bone loss is the alveolar bone loss above and beyond that of the negative control group. The net bone loss for the positive (+) and negative (−) control groups are indicated. Standard error bars are shown.

ber of BPAB that were recovered 42-days post-challenge. The formalin inactivation process was chosen for further studies. 3.3. Protection against challenge with homologous and heterologous Porphyromonas spp. In both humans and companion animals, there are several different species of Porphyromonas that have been isolated from periodontal pockets. Several of these species have also been demonstrated to be capable of inducing bone loss in experimental models [21]. Therefore, we assessed the ability of the formalin-inactivated P. gulae B43 bacterin to protect mice from challenge with either homologous or heterologous Porphyromonas spp. Four groups of 16 mice were vaccinated with saline and three groups of 16 mice were vaccinated with the formalin-inactivated P. gulae B43 bacterin. The groups of mice were subsequently challenged with saline, P. gulae B43, P. gulae B69, or P. salivosa B104. Fig. 2 shows representative photos of the post-infection jaws. Sham vaccinated mice that are challenged with P. gulae B69 have an extended CEJ–ABC distance (Fig. 2b) while the P. gulae B43 bacterin vaccinated/P. gulae B69 challenged mice (Fig. 2c) display a CEJ–ABC distance similar to that of sham challenged mice Table 3 BPAB colony forming units recovered following vaccination and homologous challenge with virulent P. gulae B43 Vaccine

Challenge material

BPAB CFUs recovereda

Saline Saline P. gulae B43 bacterin (formalin) P. gulae B43 bacterin (heat) P. gulae B43 bacterin (aeration)

Saline P. gulae B43 P. gulae B43

Negative 4+ 1+

P. gulae B43

1+

P. gulae B43

1+

a Negative, no BPAB recovered; 4+, >100 BPAB per plate; 3+, 20–100 BPAB per plate; 2+, 10–20 BPAB per plate; 1+, <10 BPAB per plate.

3152

J. Hardham et al. / Vaccine 23 (2005) 3148–3156

Fig. 2. Representative photographs of mouse jaws from the heterologous vaccine challenge study. Groups of mice were immunized with saline (A and B) or the formalin-inactivated P. gulae B43 bacterin (C) and challenged with either saline (A) or P. gulae B69 (B and C). The cementoenamel junction (CEJ) and the alveolar bone crest (ABC) are indicated.

(Fig. 2a). Mice vaccinated with the formalin-inactivated P. gulae B43 bacterin demonstrated an 83.9% reduction in alveolar bone loss (P = 0.041) when challenged with homologous P. gulae B43 compared to saline-vaccinated and challenged mice (Fig. 3a). P. gulae B43-vaccinated mice showed 40.7% (P = 0.043, Fig. 3b) and 64.6% (P = 0.061, Fig. 3c) reductions in alveolar bone loss when challenged with heterologous P. gulae B69 or P. salivosa B104, respectively, compared to saline-vaccinated and challenged mice.

mice demonstrated a significant reduction (P = 0.004) in bone loss when challenged with B. denticanium B78, indicating that a degree of cross-protection was afforded (Fig. 4a). However, there was no significant reduction in bone loss of P. gulae B43-vaccinated mice that were challenged with P. denticanis B106 (Fig. 4b). Hence, the immune response generated in response to vaccination with P. gulae B43 was not protective against challenge with P. denticanis B106.

3.4. Protection against challenge with heterologous Bacteroides spp. and novel Porphyromonas spp.

4. Discussion

The ability of the formalin-inactivated P. gulae B43 bacterin to cross-protect against other genera important in companion animal periodontal disease was assessed using the mouse oral challenge model of periodontal disease. Three groups of 16 mice each were vaccinated with saline and two groups of 16 mice were vaccinated with the formalininactivated P. gulae B43 bacterin. Following the vaccination period, mice were challenged with either saline, B. denticanium B78, or P. denticanis B106. Fig. 4 shows the bone loss observed 42-days after challenge. P. gulae B43-vaccinated

The performance of monovalent P. gulae B43 whole-cell bacterin vaccines in the mouse oral challenge periodontal disease model was investigated. While there have been several publications regarding the efficacy of human P. gingivalis vaccines in various small animal and non-human primate models, this is the first example of an animal-origin clinical isolate being evaluated as a vaccine for veterinary periodontal disease. Several non-human model systems have been developed for assessing vaccine performance against periodontal disease pathogens. The two most widely used rodent models

J. Hardham et al. / Vaccine 23 (2005) 3148–3156

3153

Fig. 4. Impact of vaccination on alveolar bone loss due to challenge with B. denticanium and P. denticanis. Groups of mice were immunized with the P. gulae B43 bacterin and challenged with (A) B. denticanium B78 or (B) P. denticanis B106. The CEJ–ABC distance (in mm) (A) and the net bone loss (in mm) (B) are shown. The values for the positive (+) and negative (−) control groups are indicated. Standard error bars are shown.

Fig. 3. Impact of vaccination on alveolar bone loss due to challenge with P. gulae and P. salivosa. Groups of mice were immunized with the P. gulae B43 bacterin and challenged with (A) P. gulae B43, (B) P. gulae B69, or (C) P. salivosa B104. The net bone loss (in mm) for the positive (+) and negative (−) control groups are indicated. Standard error bars are shown.

are the abscess (or chamber) model [23] and the oral challenge model [22]. The abscess model is short in duration (several days to a few weeks) and hence easily performed. The main drawback of this model is that it does not assess the disease process in the oral cavity where the influences of the oral biota are considerable. The mouse oral challenge model is perhaps the most widely utilized small animal model for assessment of periodontal disease vaccines. While this model does not allow for clinical parameters to be measured (bleeding on probing, probing depth, attachment loss, etc.), it does directly measure the impact of a periodontal pathogen on the alveolar bone height. For this reason, it is deemed a more appropriate model for initial assessment of periodontal disease vaccines. The Macaca fascicularis ligature model [24–26] has been utilized to study periodontal disease in the non-human primate oral cavity. In this model, the alveolar bone height is also measured. One significant drawback of the M. fascicularis model is that it incorporates a silk ligature, which is capable of inducing bone loss in the absence of a supra-bacterial challenge material. For our

initial studies, we chose to utilize the mouse oral challenge model. There have been numerous studies demonstrating various degrees of efficacy of both whole-cell and subunit P. gingivalis vaccines [27,7]. Ebersole et al. [24] and Persson et al. [25] demonstrated that immunization of monkeys with formalin-inactivated P. gingivalis markedly reduced the progression of ligature-induced alveolar bone loss. In addition, Roberts et al. [28] have shown that monkeys immunized with formalin-inactivated P. gingivalis have reduced PGE-2 levels compared to sham-vaccinated animals during ligatureinduced periodontitis. Several P. gingivalis subunit antigens have been tested in various animal models and demonstrated varying degrees of efficacy. Some of the antigens tested include FimA [29–31], the cysteine and/or arginine protease complexes/HA2 domain [7,32–37], crude cell envelope and cell wall preparations, OprF [38,39], a 40 kDa outer membrane protein [40], capsular polysaccharide [41], and HagB [42,43]. Based on this information, a whole-cell bacterin approach appears to offer the best opportunities for both homologous and heterologous vaccine efficacy. There are differing opinions on whether mucosal or systemic antibody responses are important for protection against periodontal disease. For the most part, immunization with appropriate antigens leading predominantly toward either type of humoral immune response has resulted in significant protection [30,25,31]. The oral location of the infectious process would lend itself toward a mucosal route of protection. Secretory IgA antibodies present in the saliva have been hypothesized to prevent initial colonization of P. gingivalis [44,31].

3154

J. Hardham et al. / Vaccine 23 (2005) 3148–3156

However, access of saliva to the periodontal pocket is limited. Gingival crevicular fluid, a serum exudate that contains serum IgG and IgM antibodies, composes the majority of the liquid in the periodontal pocket [45]. Therefore, subcutaneous immunization may lead to the production of specific antibodies that are capable of entering the gingival crevicular space and preventing initial penetration of periodontal pathogens and subsequent tissue destruction. We chose to utilize a subcutaneous immunization route for our studies since the systemic approach offers the potential to prevent disease, treat existing disease, and potentially inhibit systemic spread of these periodontal pathogens. The vast majority of experimental vaccines developed for periodontal disease have utilized human P. gingivalis isolates. While the similarity to veterinary isolates is arguable, the specific species associated with veterinary disease are somewhat different. Previous studies in our laboratory identified P. gulae, P. salivosa, and P. denticanis as the three most frequently isolated periodontal pathogens in companion animals [21]. Herein, we describe the first example of a veterinary vaccine for periodontal disease using P. gulae. Immunization of mice with the P. gulae B43 bacterin was shown to inhibit the progression of alveolar bone loss due to infection with homologous P. gulae B43 and heterologous P. gulae B69 (Figs. 1–3). An associated reduction in the relative numbers of P. gulae recovered post-challenge was also observed (Table 3). In addition, the bacterin was able to crossprotect against challenge with P. salivosa, the second most frequently isolated veterinary periodontal pathogen (Fig. 3c). Since other genera of bacteria have also been shown to be important in periodontal disease, we sought to determine if the P. gulae B43 bacterin could cross-protect against other related genera. Immunization with the P. gulae B43 bacterin afforded mice protection against challenge with B. denticanium B78, yet failed to inhibit bone loss in response to challenge with P. denticanis B106 (Fig. 4). It should be noted that P. denticanis is relatively divergent from other Porphyromonas spp. and may well be split into a separate genus. The cross-genera reactivity observed with Bacteroides spp. is not without merit. Vasel et al. [46] demonstrated that there is considerable antigenic cross-reactivity between P. gingivalis and B. forsythus (now Tannerella forsythensis). Our data support cross-reactivity between P. gulae and B. denticanium. The monovalent P. gulae B43 bacterin was capable of protecting mice against challenge with two of the top three most frequently isolated periodontal pathogens from companion animals. Since there are a number of Porphyromonas spp. that have been isolated from the oral cavity of companion animals, it would be advantageous for a companion animal periodontal disease vaccine be efficacious against as many of these periodontal pathogens as possible. Hence, future studies may, by necessity, focus on development of a multivalent bacterin composed of the three most frequently isolated companion animal periodontal pathogens. In humans, oral malodor (halitosis) is a well-recognized symptom of periodontal disease [47–56]. P. gingivalis is ca-

pable of producing volatile sulfur compounds, such as hydrogen sulfide and methyl mercaptan, that are associated with halitosis [53,57–59]. Companion animal halitosis has also been demonstrated [60,61] and has been recognized as one of the most frequent complaints of owners to their veterinarians [62]. Rawlings and Culham [63] have shown that a reduction in periodontal disease status correlated with a reduction in halitosis in dogs. Therefore, a potential benefit that may be realized by vaccinating companion animals with a periodontal disease vaccine is the reduction of halitosis. These results indicate that a vaccine for combating companion animal periodontitis is feasible. Future efforts will be focused on developing canine and feline periodontal disease models that are non-ligature induced and capable of inducing alveolar bone loss in response to challenge with a single bacterial species. Once such host animal models have been developed, the performance of the vaccines discussed herein will be evaluated. All procedures in this study were approved by the Institutional Animal Care and Use Committee and conducted in compliance with the Guide for Care and Use of Laboratory Animals, as well as with all internal company policies and guidelines.

Acknowledgements We would like to thank Raja Krishnan and David Lowery for critical review of this manuscript.

References [1] Amano A. Molecular interaction of Porphyromonas gingivalis with host cells: implication for the microbial pathogenesis of periodontal disease. J Periodontol 2003;74:90–6. [2] Baker PJ, Dixon M, Evans RT, Dufour L, Johnson E, Roopenian DC. CD4+ T cells and the proinflammatory cytokines gamma interferon and interleukin-6 contribute to alveolar bone loss in mice. Infect Immun 1999;67:2804–9. [3] Cutler CW, Kalmar JR, Genco CA. Pathogenic strategies of the oral anaerobe, Porphyromonas gingivalis. Trends Microbiol 1995;3:45–51. [4] Imamura T. The role of gingipains in the pathogenesis of periodontal disease. J Periodontol 2003;74:111–8. [5] Lamont RJ, Jenkinson HF. Life below the gum line: pathogenic mechanisms of Porphyromonas gingivalis. Microbiol Mol Biol Rev 1998;62:1244–63. [6] Page RC, Houston LS. Prospects for vaccination against plaquerelated oral diseases. In: Dental plaque revisited. London: BioLine; 1999. p. 563–85. [7] Page RC. Vaccination and periodontitis: myth or reality. J Int Acad Periodontol 2000;2:31–43. [8] Harvey CE, Emily PP. Hyperplastic gingivitis. In: Harvey CE, Emily PP, editors. Small animal dentistry. St. Louis, MO: Mosby-Year Books; 1993. p. 104. [9] Hennet PR, Harvey CE. Natural development of periodontal disease in the dog: a review of clinical, anatomical and histological features. J Vet Dent 1992;9:13–9.

J. Hardham et al. / Vaccine 23 (2005) 3148–3156 [10] Allaker RP, de Rosayro R, Young KA, Hardie JM. Prevalence of Porphyromonas and Prevotella species in the dental plaque of dogs. Vet Rec 1997;140:147–8. [11] Boyce EN, Ching RJ, Logan EI, Hunt JH, Maseman DC, Gaeddert KL, et al. Occurrence of gram-negative black-pigmented anaerobes in subgingival plaque during the development of canine periodontal disease. Clin Infect Dis 1995;20(Suppl. 2):S317–9. [12] Harvey CE, Thornsberry C, Miller BR. Subgingival bacteria—comparison of culture results in dogs and cats with gingivitis. J Vet Dent 1995;12:147–50. [13] Harvey CE. Periodontal disease in dogs. Etiopathogenesis, prevalence, and significance. Vet Clin North Am Small Anim Pract 1998;28:1111–28, vi. [14] Hennet PR, Harvey CE. Anaerobes in periodontal disease in the dog: a review. J Vet Dent 1991;8:18–21. [15] Isogai H, Kosako Y, Benno Y, Isogai E. Ecology of genus Porphyromonas in canine periodontal disease. J Vet Med 2000;46:467–73. [16] Duncan WJ, Persson GR, Sims TJ, Braham P, Pack AR, Page RC. Ovine periodontitis as a potential model for periodontal studies. Cross-sectional analysis of clinical, microbiological, and serum immunological parameters. J Clin Periodontol 2003;30:63–72. [17] Dreyer WP, Basson NJ. Porphyromonas gingivalis as putative pathogen in ovine periodontitis. J Dent Assoc S Afr 1992;47:513–6. [18] Fournier D, Mouton C, Lapierre P, Kato T, Okuda K, Menard C. Porphyromonas gulae sp. nov., an anaerobic, Gram-negative coccobacillus from the gingival sulcus of various animal hosts. Int J Syst Evol Microbiol 2001;51:1179–89. [19] Hirasawa M, Takada K. Porphyromonas gingivicanis sp. nov. and Porphyromonas crevioricanis sp. nov., isolated from beagles. Int J Syst Bacteriol 1994;44:637–40. [20] Love DN, Karjalainen J, Kanervo A, Forsblom B, Sarkiala E, Bailey GD, et al. Porphyromonas canoris sp. nov., an asaccharolytic, black-pigmented species from the gingival sulcus of dogs. Int J Syst Bacteriol 1994;44:204–8. [21] Hardham JM, Dreier K, Wong J, Sfintescu C, Evans RT. Pigmentedanaerobic bacteria associated with canine periodontitis. Vet. Microbiol., in press. [22] Baker PJ, Evans RT, Roopenian DC. Oral infection with Porphyromonas gingivalis and induced alveolar bone loss in immunocompetent and severe combined immunodeficient mice. Arch Oral Biol 1994;39:1035–40. [23] Genco CA, Kapczynski DR, Cutler CW, Arko RJ, Arnold RR. Influence of immunization on Porphyromonas gingivalis colonization and invasion in the mouse chamber model. Infect Immun 1992;60:1447–54. [24] Ebersole JL, Brunsvold M, Steffensen B, Wood R, Holt SC. Effects of immunization with Porphyromonas gingivalis and Prevotella intermedia on progression of ligature-induced periodontitis in the nonhuman primate Macaca fascicularis. Infect Immun 1991;59:3351–9. [25] Persson GR, Engel D, Whitney C, Darveau R, Weinberg A, Brunsvold M, et al. Immunization against Porphyromonas gingivalis inhibits progression of experimental periodontitis in nonhuman primates. Infect Immun 1994;62:1026–31. [26] Persson GR, Engel LD, Whitney C, Weinberg A, Moncla BJ, Darveau RP, et al. Macaca fascicularis as a model in which to assess the safety and efficacy of a vaccine for periodontitis. Oral Microbiol Immunol 1994;9:104–11. [27] Nakagawa T, Saito A, Hosaka Y, Ishihara K. Gingipains as candidate antigens for Porphyromonas gingivalis vaccine. Keio J Med 2003;52:158–62. [28] Roberts FA, Houston LS, Lukehart SA, Mancl LA, Persson GR, Page RC. Periodontitis vaccine decreases local prostaglandin E2 levels in a primate model. Infect Immun 2004;72:1166–8. [29] Evans RT, Klausen B, Genco RJ. Immunization with fimbrial protein and peptide protects against Porphyromonas gingivalis-induced periodontal tissue destruction. Adv Exp Med Biol 1992;327:255– 62.

3155

[30] Evans RT, Klausen B, Sojar HT, Bedi GS, Sfintescu C, Ramamurthy NS, et al. Immunization with Porphyromonas (Bacteroides) gingivalis fimbriae protects against periodontal destruction. Infect Immun 1999;60:2926–35. [31] Sharma A, Honma K, Evans RT, Hruby DE, Genco RJ. Oral immunization with recombinant Streptococcus gordonii expressing Porphyromonas gingivalis FimA domains. Infect Immun 2001;69:2928–34. [32] DeCarlo AA, Huang Y, Collyer CA, Langley DB, Katz J. Feasibility of an HA2 domain-based periodontitis vaccine. Infect Immun 2003;71:562–6. [33] Gibson III FC, Genco CA. Prevention of Porphyromonas gingivalisinduced oral bone loss following immunization with Gingipain R1. Infect Immun 2001;69:7959–63. [34] Kuboniwa M, Amano A, Shizukuishi S, Nakagawa I, Hamada S. Specific antibodies to Porphyromonas gingivalis Lys-gingipain by DNA vaccination inhibit bacterial binding to hemoglobin and protect mice from infection. Infect Immun 2001;69:2972–9. [35] Moritz AJ, Cappelli D, Lantz MS, Holt SC, Ebersole JL. Immunization with Porphyromonas gingivalis cysteine protease: effects on experimental gingivitis and ligature-induced periodontitis in Macaca fascicularis. J Periodontol 1998;69:686–97. [36] Rajapakse PS, O’Brien-Simpson NM, Slakeski N, Hoffmann B, Reynolds EC. Immunization with the RgpA-Kgp proteinase–adhesin complexes of Porphyromonas gingivalis protects against periodontal bone loss in the rat periodontitis model. Infect Immun 2002;70:2480–6. [37] Yonezawa H, Ishihara K, Okuda K. Arg-gingipain A DNA vaccination induces protective immunity against infection by Porphyromonas gingivalis in a murine model. Infect Immun 2001;69:2858–64. [38] Ross BC, Czajkowski L, Hocking D, Margetts M, Webb E, Rothel L, et al. Identification of vaccine candidate antigens from a genomic analysis of Porphyromonas gingivalis. Vaccine 2001;19:4135–42. [39] Ross BC, Czajkowski L, Vandenberg KL, Camuglia S, Woods J, Agius C, et al. Characterization of two outer membrane protein antigens of Porphyromonas gingivalis that are protective in a murine lesion model. Oral Microbiol Immunol 2004;19:6–15. [40] Namikoshi J, Otake S, Maeba S, Hayakawa M, Abiko Y, Yamamoto M. Specific antibodies induced by nasally administered 40-kDa outer membrane protein of Porphyromonas gingivalis inhibits coaggregation activity of P. gingivalis. Vaccine 2003;22:250–6. [41] Gonzalez D, Tzianabos AO, Genco CA, Gibson III FC. Immunization with Porphyromonas gingivalis capsular polysaccharide prevents P. gingivalis-elicited oral bone loss in a murine model. Infect Immun 2003;71:2283–7. [42] Katz J, Black KP, Michalek SM. Host responses to recombinant hemagglutinin B of Porphyromonas gingivalis in an experimental rat model. Infect Immun 1999;67:4352–9. [43] Zhang P, Yang Q, Marciani DJ, Martin M, Clements JD, Michalek SM, et al. Effectiveness of the quillaja saponin semi-synthetic analog GPI-0100 in potentiating mucosal and systematic responses to recombinant HagB from Porphyromonas gingivalis. Vaccine 2003;21:4459–71. [44] Sharma A, Honma K, Sojar HT, Hruby DE, Kuramitsu H, Genco RJ. Expression of saliva-binding epitopes of the Porphyromonas gingivalis FimA protein on the surface of Streptococcus gordonii. Biochem Biophys Res Commun 1999;258:222–6. [45] Ebersole JL. Humoral immune responses in gingival crevice fluid: local and systemic implications. Periodontol 2000 2003;31:135–66. [46] Vasel D, Sims TJ, Bainbridge B, Houston L, Darveau R, Page RC. Shared antigens of Porphyromonas gingivalis and Bacteroides forsythus. Oral Microbiol Immunol 1996;11:226–35. [47] Awano S, Gohara K, Kurihara E, Ansai T, Takehara T. The relationship between the presence of periodontopathogenic bacteria in saliva and halitosis. Int Dent J 2002;52(Suppl. 3):212–6. [48] Coil JM, Tonzetich J. Characterization of volatile sulphur compounds production at individual gingival crevicular sites in humans. J Clin Dent 1992;3:97–103.

3156

J. Hardham et al. / Vaccine 23 (2005) 3148–3156

[49] De Boever EH, De Uzeda M, Loesche WJ. Relationship between volatile sulfur compounds. BANA-hydrolyzing bacteria and gingival health in patients with and without complaints of oral malodor. J Clin Dent 1994;4:114–9. [50] Figueiredo LC, Rosetti EP, Marcantonio Jr E, Marcantonio RA, Salvador SL. The relationship of oral malodor in patients with or without periodontal disease. J Periodontol 2002;73:1338–42. [51] Kostelc JG, Zelson PR, Preti G, Tonzetich J. Quantitative differences in volatiles from healthy mouths and mouths with periodontitis. Clin Chem 1981;27:842–5. [52] Miyazaki H, Sakao S, Katoh Y, Takehara T. Correlation between volatile sulphur compounds and certain oral health measurements in the general population. J Periodontol 1995;66:679–84. [53] Nakano Y, Yoshimura M, Koga T. Correlation between oral malodor and periodontal bacteria. Microbes Infect 2002;4:679–83. [54] Persson S. Hydrogen sulfide and methyl mercaptan in periodontal pockets. Oral Microbiol Immunol 1992;7:378–9. [55] Ratcliff PA, Johnson PW. The relationship between oral malodor, gingivitis, and periodontitis. A review. J Periodontol 1999;70:485–9.

[56] Yaegaki K, Sanada K. Biochemical and clinical factors influencing oral malodor in periodontal patients. J Periodontol 1992;63: 783–9. [57] Nakano Y, Yoshimura M, Koga T. Methyl mercaptan production by periodontal bacteria. Int Dent J 2002;52:217–20. [58] Tonzetich J, McBride BC. Characterization of volatile sulphur production by pathogenic and non-pathogenic strains of oral Bacteroides. Arch Oral Biol 1981;26:963–9. [59] Yoshimura M, Nakano Y, Yamashita Y, Oho T, Saito T, Koga T. Formation of methyl mercaptan from l-methionine by Porphyromonas gingivalis. Infect Immun 2000;68:6912–6. [60] Hennet P, Delille B, Davot JL. Oral malodor in dogs: measurement using a sulfide monitor. J Vet Dent 1995;12:101–3. [61] Rawlings JM, Culham N. Studies of oral malodor in the dog. J Vet Dent 1998;15:169–73. [62] Culham N, Rawlings JM. Oral malodor and its relevance to periodontal disease in the dog. J Vet Dent 1998;15:165–8. [63] Rawlings JM, Culham N. Halitosis in dogs and the effect of periodontal therapy. J Nutr 1998;128:2715S–6S.