Porphyromonas gingivalis and Treponema denticola Mixed Microbial Infection in a Rat Model of Periodontal Disease

Porphyromonas gingivalis and Treponema denticola are periodontal pathogens that express virulence factors associated with the pathogenesis of periodontitis. In this paper we tested the hypothesis that P. gingivalis and T. denticola are synergistic in terms of virulence; using a model of mixed microbial infection in rats. Groups of rats were orally infected with either P. gingivalis or T. denticola or mixed microbial infections for 7 and 12 weeks. P. gingivalis genomic DNA was detected more frequently by PCR than T. denticola. Both bacteria induced significantly high IgG, IgG2b, IgG1, IgG2a antibody levels indicating a stimulation of Th1 and Th2 immune response. Radiographic and morphometric measurements demonstrated that rats infected with the mixed infection exhibited significantly more alveolar bone loss than shaminfected control rats. Histology revealed apical migration of junctional epithelium, rete ridge elongation, and crestal alveolar bone resorption; resembling periodontal disease lesion. These results showed that P. gingivalis and T. denticola exhibit no synergistic virulence in a rat model of periodontal disease.

The mechanisms of interaction between P. gingivalis and T. denticola as a consortium in the subgingival sulcus and whether they express a synergistic pathogenic potential in progressing periodontitis remain enigmatic at present [19]. Recently, we have demonstrated that P. gingivalis, T. denticola, and T. forsythia with and without F. nucleatum not only exist as a consortium that is associated with chronic periodontitis in humans but also exhibit virulence resulting in the colonization of the rat oral cavity, induction of enhanced IgG immune responses, and significant alveolar bone resorption characteristic of polymicrobial (three pathogens or more) periodontitis [25].
Monomicrobial (single pathogen) periodontal infections using P. gingivalis, T. denticola, T. forsythia, or F. nucleatum 2 Interdisciplinary Perspectives on Infectious Diseases have been studied in rats and mice [25][26][27][28][29][30]. Increasing evidence supports the concept that bacterial interactions among members of the subgingival pathogens at any time during periodontal disease progression are important. However, there are no published reports establishing a mixed microbial (two species) periodontal infection for examining the virulence between P. gingivalis and T. denticola. This study examined mixed microbial periodontal disease using P. gingivalis and T. denticola as a consortium and examined their colonization/infection characteristics, periodontal inflammation parameters, immune response patterns, induction of alveolar bone resorption, and virulence interactions.

Bacterial Strains and Inocula.
The bacteria used in this study were P. gingivalis 381 and T. denticola ATCC 35404 and these strains were grown under anaerobic conditions (85% N 2 , 10% H 2 , and 5% CO 2 ) at 37 • C in a Coy anaerobic chamber as described previously [25,31]. The P. gingivalis strain 381 was chosen due to its known role in alveolar bone resorption in adult periodontitis and its proven ability to colonize the oral cavity of rodents [25,27,28,32]. For oral monobacterial infection, P. gingivalis (2 ×10 10 cells per mL: grown for 3 days on CDC anaerobic 5% sheep blood agar plates) or T. denticola (2 ×10 10 cells per mL: grown in GM-1 broth for 48-72 hours as a log phase culture) was mixed with equal volumes of sterile 2% (w/v) low viscosity carboxymethylcellulose with PBS (CMC: Sigma), [25,26,28,33] and one mL was used for infection (10 10 cells per mL) by oral gavage [25,26,28,33]. For oral mixed microbial infection, P. gingivalis (2 ×10 10 cells per mL) was gently mixed with an equal volume of T. denticola (2 ×10 10 cells per mL), mixed gently for 1-2 min, and allowed to interact for additional 5 minutes for any interactions among these species. An equal volume of sterile 2% (w/v) CMC was added, mixed thoroughly, and one mL (5 ×10 9 cells of P. gingivalis mL, 5 ×10 9 cells of T. denticola) was administered by oral gavage and anal topical application. Rats are coprophagic in nature, and the rationale behind the anal application is that bacteria will be in the feces and will then return to the oral cavity, thereby establishing a cycle of oral reinfection [28,34].

Rat Oral Infections.
Female Sprague-Dawley rats (8 weeks old, Charles River laboratories, MA, USA) were maintained in groups and housed in microisolator cages in an AALAC facility at the University of Florida. The protocol (#E900) and all rat infection procedures used in this study were approved by the Institutional Animal Care and Use Committee of the University of Florida. Animals were fed standard powdered chow (Teklad Global 18% protein rodent diet 2918, Harlan) and water ad libitum. All rats were administered kanamycin (20 mg) and ampicillin (20 mg) daily for 4 days in the drinking water [25,35] and the oral cavity was swabbed with 0.12% (v/v) chlorhexidine gluconate (Proctor and Gamble, OH, USA) mouth rinse [25,31] to inhibit the endogenous organisms and to promote subsequent colonization of P. gingivalis and T. denticola. Rats were randomized into groups (n = 18) and monobacterial and mixed microbial inocula were administered by oral gavage and anal topical application [28,34] for 4 consecutive days per week on 4-6 alternate weeks (16-24 inoculations). Sham-infected control rats received vehicle and sterile 2% low viscosity CMC only.

Oral Microbial
Sampling. Oral microbial samples from rats were collected using sterile cotton swabs at pre-and postinfections. A total of 4-6 postinfection microbial samples were collected the following week from all the infected rats. To determine the kinetics of virulence of mono-and mixed infection-induced periodontal disease and immune responses, rats were euthanized at the beginning of the 8th (7 weeks of 16 inoculations) and 13th weeks (12 weeks of 24 inoculations). Blood was collected and sera were stored at −20 • C for IgG antibody analysis. The rats were killed, skulls were removed, autoclaved, and mechanically defleshed with a periodontal scaler. The randomly selected rat jaws (n = 3) were also suspended in 10% (v/v) buffered formalin, decalcified, tissues trimmed, and used for histomorphometry and histology.

Monitoring Bacterial
Colonization/Infection. DNA was isolated from rat oral microbial samples using a Wizard Genomic DNA purification kit (Promega, WI, USA). The standard genomic DNA for P. gingivalis, and T. denticola were also extracted following the same procedure from their respective 24-72 hours pure cultures as described previously in [25]. Subsequently, PCR was performed using 16S rRNA gene species-specific PCR oligonucleotide primers with a Bio-Rad thermal cycler as described previously in [25,31,35]. After amplification; PCR products were separated by 1.5% agarose gel electrophoresis and the bands were visualized using the UVP BioDoc-It Imaging System. The genomic DNA extracted from P. gingivalis and T. denticola served as positive control and PCR performed with no template DNA making up the negative control. Each PCR assay with the standard DNA was sensitive enough to detect 0.05 pg of DNA (P. gingivalis 30 cells; T. denticola 18 cells). Different PCR cycles from 35 to 40 were standardized to produce detectable amplicons with the least amount (0.05 pg) of template DNA.

Antibody
Analysis. Blood was collected from each rat at the time of euthanasia. Serum from monobacterial (n = 9) or mixed microbial infected rats (n = 9) at 7 weeks (4 infections) and 12 weeks (6 infections) was used to determine IgG, IgM, IgA, and IgG subclass (IgG1, IgG2a, IgG2b, IgG2c) antibody concentrations, using a standard ELISA protocol [25,[34][35][36]. Briefly, diluted infected rat serum (1 : 100 for IgG and 1 : 20 for IgM, IgA and IgG subclass) was incubated in wells of either P. gingivalis or T. denticola coated microtiter polystyrene plates (Costar, Corning, NY, USA) for 2 hours at room temperature. After washing, alkaline phosphataseconjugated goat anti-rat IgG (1 : 2000), IgM and IgA (Bethyl Laboratories, TX, USA) were added (1 : 500) and incubated for additional 2 h at room temperature on a rotator. The substrate (p-nitrophenylphosphate; Sigma, 1 mg/ml) was added to the washed plates, and the reaction was terminated by using 3 M NaOH. For IgG subclass ELISA analysis, alkaline phosphatase-conjugated goat anti-rat IgG1, IgG2a, IgG2b, and IgG2c (Bethyl Laboratories, TX, USA) were used (1 : 500). The optical density (OD) was measured at OD 405 nm using a Bio-Rad Microplate Reader. The infected rat serum antibody concentration was quantified using a gravimetric standard curve. The standard curve consisted of 8-rat IgG concentrations (Sigma, St. Louis, USA), which were coated onto wells of the microtiter polystyrene plates, detected, and developed as described above.

Morphometry Analysis of Horizontal Alveolar Bone
Resorption. The pattern of alveolar bone resorption (horizontal or vertical) induced by P. gingivalis and T. denticola were measured by both morphometric and radiograph methods, respectively. The 7 and 12 weeks mono-and mixed microbial infected jaws (n = 6) were immersed in 3% (vol/vol) hydrogen peroxide overnight, and stained with 0.1% (wt/vol) methylene blue to delineate the cementoenamel junction (CEJ) using the modified morphometric method [33,34]. The digital images of both buccal and lingual root surfaces of all molar teeth were captured under a 10 × stereo dissecting microscope (SteReo Discovery V8; Carl Zeiss), after superimposition of buccal and lingual cusps to ensure reproducibility and consistency. The line tool was used to make horizontal bone resorption measurements on all molars in each quadrant from the CEJ to the alveolar bone crest (ABC). The surface perimeters of CEJ and ABC were traced using the calibrated line tool. As the AxioVision software program was calibrated using a precise ruler, the area of the horizontal bone loss reading in mm 2 is instantly imprinted over the digital image. Two blinded investigators were used and all measurements were done two times by the same examiner at separate times and the means of the measurements were obtained for each of the four quadrants.

Radiographic Assessment of Vertical Alveolar Bone Loss.
The maxillae and mandibles were placed and stabilized with dental wax on a digital Kodak 6000 sensor (CareStream Health, USA) oriented with the axis of the teeth parallel to the sensor surface. Digital radiographs of distal and mesial surfaces of the molars were acquired with orthogonal projection geometry using an exposure time of 0.08 s at 60 kVp and 15 mA. All radiographic images were exported into the Tuned Aperture Computed Tomography workbench, calibrated for magnification using known anatomic measurements, and histograms equalized. The line tool was used to make vertical bone resorption measurements on the distal and mesial sides of each interproximal surface (2 sites per tooth) for each of the molars in each quadrant from the CEJ to the ABC (i.e., resorption) as the primary outcome parameter of the study. The summation of alveolar bone resorption in mm was tabulated and analyzed for intra and intergroup comparison [25, 31, 35].

Histomorphometric Analysis of Periodontal Tissue.
Monomicrobial and mixed microbial infected rat jaws (n = 3) were removed randomly and fixed in 10% buffered formalin. Bone was decalcified in Immunocal (Decal Chemical) for 28 days at 4 • C. The decalcified tissue was embedded in paraffin blocks, 4 μ sections prepared, stained with hematoxylin and eosin and the wholeslides were digitally scanned with a ScanScope CS system (Aperio Technologies, Vista, CA). The scanned slides were viewed with ImageScope viewing software (Aperio Technologies, Vista, CA). The inflammation at the supracrestal gingival connective tissue between the molars in each specimen at consecutive sections or levels 10 and 20 was examined based on multiple parameters including polymorphonuclear leukocytes (PMN), lymphocytes, blood vessel density, apical migration of junctional epithelium (JE), rete ridge elongation (R) and alveolar bone resorption (ABR) [37]. The number of inflammatory cells (PMNs and lymphocytes) per unit area (0.05 mm × 0.05 mm) was counted in the area of the junctional epithelium and adjacent connective tissue [37]. The migration of JE, elongation of rete ridges and resorption of alveolar bone was measured using the ImageScope software with a microgrid at a magnification of ×200. The distances from the CEJ to the coronal portion of the connective tissue attachment (apical migration), from the CEJ to the apical portion of the rete ridge; and from the CEJ to the level of the alveolar bone crest were measured [37].

Statistical Analyses. The alveolar bone resorption and
IgG antibody data were presented as means ± standard deviations (Prism 4, GraphPad software). P values were calculated using the Kruskal Walis ANOVA with Dunn's correction for multiple comparisons and Mann-Whitney Student T-test [37]. P values of .05 were considered statistically significant.

Induction of Periodontal
Disease. Prior to mono-or mixed infection; we examined all rats for P. gingivalis and T. denticola using appropriate bacterium-specific primers by PCR, and observed all rats were consistently negative for these oral pathogens. The PCR results demonstrate an appropriately sized amplicon for P. gingivalis (600 bp) present in DNA isolated from rat oral microbial samples following infection with the single-microbe inocula (data not shown). Among two monobacterial infections, P. gingivalis was found positive by PCR in all individual rats (n = 18) and was positive 1-4 times during the four to six sampling times. In addition, 50-83.3% of infected rats were positive for P. gingivalis during four (2, 3, 4, and 6) out of six sampling times. In contrast, T. denticola (860 bp) amplicons were negative for DNA isolated from rat oral microbial samples following infection with the single-microbe inocula. The rats that had been given P. gingivalis + T. denticola mixed infection showed 72% (13/18) positive amplicons for P. gingivalis and only 2 out of 18 rats were positive for amplicons of T. denticola.

Serum IgG Antibody to Oral Infections.
To provide additional documentation of oral infection and to demonstrate an immunological response to P. gingivalis and T. denticola infection, we evaluated the levels of pathogen specific IgG, IgM, IgA, and IgG subclass (IgG1, IgG2a, IgG2b, IgGc) antibodies in rat sera 7 weeks and 12 weeks post-initial infection ( Figure 1). The induction of IgG antibody response patterns was identical in both 7 and 12 weeks infected rats using either mono-or mixed infection protocols. All rats in the P. gingivalis infected group at both 7 (Figure 1(a)) and 12 weeks (Figure 1(c)) demonstrated significantly elevated IgG antibody (P < .05) compared to the levels in shaminfected control rats. Similarly, all rats infected with T. denticola produced IgG antibody that was significantly higher (P < .05) than levels of sham-infected control rats at both 7 (Figure 1(b)) and 12 weeks (Figure 1(e)) postinfection. However, P. gingivalis infection induced significantly higher IgG levels (>100-fold) than T. denticola infection in the monoinfected rats. The levels of IgG antibody in the rat serum paralleled the frequency of detection of P. gingivalis in the oral microbial samples. Interestingly, all rats in the T. denticola monobacterial infection group at 12 weeks (Figure 1(e)) induced significantly strong IgG immune response in spite of our inability to detect T. denticola DNA in the oral microbial samples.
All rats in the mixed infection groups showed elevated serum IgG antibodies to P. gingivalis and T. denticola compared to the levels in shaminfected control rats ( Figure 1). Interestingly, P. gingivalis + T. denticola infection at 7 (Figures 1(a) and 1(b)) and 12 week (Figures 1(d) and 1(f)) post-initial infection induced >100-fold stronger P. gingivalis specific IgG immune response as compared to T. denticola IgG responses. Approximately, 90% of the rats infected with mixed bacteria (16/18) presented elevated serum IgG to P. gingivalis when compared to sham-infected control rats (Figure 1(a)). Similarly, 100% of the mixed microbial infected rats (18/18) demonstrated significantly elevated serum IgG to T. denticola compared to that in sham-infected control rats (Figure 1(b)). None of the P. gingivalis and T. denticola infected rats induced IgA and IgM antibodies during 7 and 12 weeks of infection (data not shown).

IgG Subclass Responses to Infection.
Thus, we had observed that P. gingivalis and T. denticola are antigenic in the rats, which resulted in significant levels of serum IgG antibodies. In order to more fully assess the characteristics of the antibody, we determined the IgG subclass distribution of the humoral immune response following oral infection. Following 12 weeks of P. gingivalis mono-and mixed microbial infection, the IgG2b subclass levels (P < .05) were higher than the IgG1 and IgG2a antibody levels and significantly (P < .05) greater than in sham-infected control rats (Figures 1(c) and 1(d)). Similarly, in T. denticola monoand mixed microbial infection, the IgG2b subclass levels (P < .05) were higher than the IgG1 and IgG2c antibody levels and significantly greater than in sham-infected control rats (P < .05) (Figures 1(e) and 1(f)). Additionally, P. gingivalis induced higher levels of IgG subclass antibody than that induced by T. denticola.

Morphometric Evaluation of Alveolar Bone Loss.
In order to address the potential virulence between the P. gingivalis and T. denticola in periodontal disease progression in the rats, we examined the effect of infection on the maxilla and mandible alveolar bone resorption. Both mono-and mixed infection induced buccal and palatal areas of alveolar bone resorption (Table 1; Figure 2). Here, the maxillary and mandibular bone loss in the rats infected with P. gingivalis, T. denticola, and P. gingivalis + T. denticola were significantly greater (P < .05) than that of the sham-infected control group at 7 and 12 weeks ( Table 1). The maxillary and mandibular bone loss in all infected groups during 12 weeks of periodontal disease was generally greater than 7 weeks of periodontal disease. In addition, mandibular bone loss was generally higher than maxillary bone loss in both buccal and palatal surfaces ( Table 1). P. gingivalis monoinfection induced more palatal mandibular horizontal bone loss area than maxilla at 7 and 12 weeks post-initial infection. In contrast, T. denticola monoinfection generally induced greater palatal and buccal area bone loss than P. gingivalis as well as more mandibular bone loss than maxilla at 7 and 12 weeks post-initial infection. Similarly, mixed infection induced more significant bone loss in both maxilla and mandibles, palatal and buccal surfaces than monoinfection at 7 and 12 weeks post initial infection (Table 1). Furthermore, mixed infection induced more mandibular palatal surface horizontal area bone loss than maxillae.

Radiographic Evaluation of Interproximal Alveolar Bone
Loss. In order to confirm our observations of alveolar bone resorption, radiographic analysis of the maxilla and mandible was performed. Mono-and mixed infection resulted in significantly increased maxillary, mandibular, and total interproximal alveolar bone resorption at 7 weeks (data not shown) and 12 weeks of periodontal disease compared with sham-infected control rats (P < .05) (Figure 3). In addition, there was an overall loss in vertical bone height, with associated circumferential angular defects. Furthermore, mixed infection demonstrated a significant increase in maxillary, mandibular, and total vertical bone loss compared to any of the monobacterial and sham-infected control infections (P < .05) (Figure 3).

Histological and Histometrical Analysis.
In order to determine if the infection protocols induced differing levels of inflammation which could be responsible for the increased alveolar bone resorption observed, sections of the maxilla and mandible of rats infected with P. gingivalis and/or T. denticola during 7 and 12 weeks of periodontal disease were examined at consecutive levels 1, 10, and 20 for inflammation. Mixed microbial infection rats showed more significant histological changes, particularly apical migration of JE, rete ridge elongation, PMN density, lymphocytes infiltration, blood vessel density and alveolar bone loss  The error bars indicate one standard deviation from the mean. An asterisk indicates that a value is significantly different ( * P < .05) than the value for sham-infected control rats or for antibody in serum from rats infected with a different microorganism. Pg, P. gingivalis; Td, T. denticola; Cont; sham-infected control.  3 ± 0.33 b,d,e * Mean value and standard deviation from six mice per each group measured using AxioVision line tool software as described in methods section. a Significantly more than sham-infected control group (P < .05). b Significantly more than sham-infected control group (P < .01). c Significantly more than P. gingivalis infected group (P < .05). d Significantly more than P. gingivalis infected group (P < .01). e Significantly more than T. denticola infected group (P < .05). than sham-infected control animals (Table 2; Figure 4). The sham-infected control rats showed mild inflammation with fewer PMN and lymphocytes, no apical migration, and small rete ridge elongation. Similarly, mixed microbial infection induced significant apical migration of JE (P < .01) and dense inflammation in the periodontium compared to P. gingivalis monoinfected rats. In addition, mixed microbial infection rats showed no significant differences in rete ridge elongation, PMN density, lymphocytes infiltration, and blood vessel density compared to P. gingivalis and T. denticola mono-infected rats. P. gingivalis and T. denticola infected rats showed significant histological changes, specifically apical migration of JE, alveolar bone loss, PMN density, lymphocytes infiltration, and blood vessel density more than shaminfected control animals (Table 2; Figure 4). Moreover, P. gingivalis infected rats showed significant infiltration of lymphocytes (P < .05) and blood vessel density in the periodontium (P < .001) compared to mixed microbial infected rats ( Table 2).

Discussion
This paper explicitly demonstrates an experimental model for mixed microbial infections in periodontal disease, documenting colonization/infection with the P. gingivalis/T. denticola consortium of oral microorganisms, with (kinetics) induction of periodontal inflammation at 7 and 12 weeks, generation of specific systemic IgG immune responses to the infecting pathogens, and stimulation of enhanced alveolar bone resorption in rats. These results also documented, for the first time, the virulence of mixed infections with P. gingivalis + T. denticola in a periodontal disease model. Bacterial synergism in progression from periodontal health to disease has been proposed but few studies have documented bacterial synergism due to the inherent complexity of the subgingival microflora [8,38]. In the previous study [25], we had not examined the early induction of periodontal inflammation and assessment of both palatal and buccal horizontal bone loss. Here we demonstrate this early kinetics Interdisciplinary Perspectives on Infectious Diseases  Figure 3: Interproximal evaluation of alveolar bone loss. Radiographic vertical alveolar bone resorption (12 weeks of periodontal disease) in rats (n = 6) following mono infection with P. gingivalis or T. denticola and with the mixed infection. Each bar indicates the mean interproximal alveolar bone loss for distance measured between CEJ and ABC at mesial and distal sites of three molar teeth (6 sites). The error bars indicate standard deviation from the mean. An asterisk sign denotes significantly different from sham-infected control rats (P < .05). The # sign indicates significant difference comparing between mono bacterial P. gingivalis and the mixed microbial infected rats (P < .05) and & sign indicates significant difference comparing between T. denticola and the mixed microbial infected rats (P < .05). Pg, P. gingivalis; Td, T. denticola; Cont; sham-infected control. 3.7 ± 0.9 11.7 ± 3.9 c,e 5.5 ± 1.7 a 6.6 ± 1.6 b Blood vessel density (number/0.05 mm × 0.05 mm) 2.1 ± 1.0 8.4 ± 1.9 c,f 5 ± 2.5 b 4.3 ± 1.7 b * Values are presented as mean ± standard deviation (n = 5−9). † The distances from the CEJ to the coronal portion of the connective tissue attachment (apical migration of JE). ‡ The distances from the CEJ to the apical portion of the rete ridge. § The distances from the CEJ to the level of the ABC. a Significantly more than sham-infected control group (P < .05). b Significantly more than sham-infected control group (P < .01). c Significantly more than sham-infected control group (P < .0001). d Significantly more than P. gingivalis group (P < .01). e Significantly more than P. gingivalis + T. denticola group (P < .05). f Significantly more than P. gingivalis + T. denticola group (P < .0001).
in a model of periodontal disease at 7 and 12 weeks of disease.
The monobacterial infection in rats indicated that P. gingivalis exhibited the ability to colonize/infect the oral cavity with 4-6 alternate weekly infection schedules (16)(17)(18)(19)(20)(21)(22)(23)(24) inoculations) during the 7-12 weeks study establishing a chronic infection. We have shown previously that infecting rats 15-16 times with P. gingivalis, over a similar interval of the experiment, resulted in consistent detection of genomic DNA in oral microbial samples [25,31,35]. Moreover, induction of significant IgG immune responses and enhanced alveolar bone loss observed in all rats clearly documents that these rats were infected, even though the rats showed negative PCR reactions for T. denticola. We recognize the limitations in sample collection procedures, as existing with all the current techniques of microbial sampling from the oral cavity.
The serum IgG antibody levels to monoinfection during 7 and 12 weeks of periodontal disease clearly indicated that P. gingivalis is highly effective in colonization and/or is highly antigenic in the rats when compared to T. denticola. The antibody responses demonstrated substantial specificity for each of the infecting species. However, we observed an increase in serum IgG antibody to T. denticola in P. gingivalis mono-infected rats. While this "nonspecific" IgG antibody was more than sham-infected controls, it was approximately 1000-fold lower than the homologous IgG antibody response to P. gingivalis infection. This could indicate that these bacteria share some common epitopes [25]. The mixed oral infection with P. gingivalis/T. denticola elicited somewhat different profiles of serum IgG antibodies. These altered responses could be due to a lowered colonization capacity of the T. denticola within the mixed consortium challenge and/or a decreased ability to multiply in the oral cavity during the infection, thus resulting in a lower magnitude of antigenic challenge, reduced periodontal inflammation, and no robust alveolar bone loss nor virulence synergism. The predominant response following P. gingivalis infection was the IgG2b (T helper type 1) and IgG1 subclass (T helper type 2), followed by IgG2a (Th1) and undetectable level  Figure 4: Representative sections from histopathology of periodontal tissue. Comparative maxillary histology (hematoxylin and eosin) of alveolar bone sections from maxilla of rat infected with P. gingivalis and/or T. denticola at 12 weeks. (a) from sham-infected control rat displaying minimal inflammation, lack of migration of junctional epithelium (JE) and minimal inflammation in the connective tissue (CT); (b) from the P. gingivalis infected rat showing prominent epithelium hyperplasia (EH), migration of JE and increase in number of blood vessels (BV) (indicated by black arrows); (c) from the T. denticola infected rat also exhibits migration of JE and EH; and (d) from the mixed infection with P. gingivalis + T. denticola infected rat demonstrates dense inflammation (DI) (indicated by white arrows) along with migration of JE. JE indicates junctional epithelium, CT connective tissue, EH epithelial hyperplasia, BV blood vessels, DI dense inflammation. All images at 20× magnification.
of IgG2c antibody indicating a stimulation of both Th1 and Th2 activities in development of the humoral immune response to bacterial infection. Similarly, the predominant response following T. denticola monoinfection was the IgG2b subclass, followed by IgG1, IgG2c and undetectable level of IgG2a antibody in rats suggesting a mixed Th1-and Th2-responses to oral infection. In contrast, IgG1 antibody titer was much higher in mice to T. denticola infection [29]. Despite the high bacterial specific IgG antibody levels during 7 and 12 weeks of infection, there was no significant immune protection from alveolar bone loss in rats as well as in several previous studies [39,40] suggesting complex mechanisms of antibody protection. Furthermore, P. gingivalis recombinant hemagglutinin B immunization or immunized and infected rats induced IgG subclass responses (IgG1 = IgG2a > IgG2b > IgG2c) suggesting a mixed Th1 and Th2 responses and immunized rats had less alveolar bone loss indicating a protective immune response [41,42]. Similarly, immunization with P. gingivalis whole cells induced high-titer serum IgG2a (Th2), moderate-titer IgG2b (Th1) and low-titer IgG1 (Th2) responses and immunization with RgpA-Kgp cysteine proteases of P. gingivalis induced high-titer serum IgG2a (Th2) responses which restricted colonization and decreased periodontal bone loss indicating a protective immune response in the rat [34].
While differences in horizontal (palatal and buccal surface) and interproximal alveolar bone resorption levels were observed following monoinfection with P. gingivalis and T. denticola dependent upon both the differences in the sites of the samples as well as the techniques for measurements, we could not easily compare the magnitude of alveolar bone resorption between these individual bacteria. Importantly, in testing our hypothesis, oral infection with P. gingivalis/T. denticola significantly increased interproximal as well as horizontal alveolar bone loss compared to mono-infections. This increased bone loss may be related to enhancement of expression of the virulence of individual bacteria by cooperative abilities of their extracellular potent proteinases (P. gingivalis RgpA, RgpB, Kgp gingipains cysteine proteinase) (T. denticola chymotrypsin-like protease, phospholipase C, oligopeptidase, endopeptidase and cystalysin) to affect host systems through specific cleavage of cell surface receptors and the inactivation of host-defense proteins [18,20]. In addition, mixed infection with P. gingivalis + T. denticola exhibits significant virulence synergism in abscess formation and mortality in mouse pneumonia model [23] and mouse abscess model [21,22].

Conclusions
The analysis of the data have clearly shown the following: (i) mono-and mixed microbial colonization/infection of human oral pathogens in rat oral cavity during 7 weeks of periodontal disease (gained access to the oral epithelium), (ii) generation of a specific serum IgG antibody responses (as early as 7 weeks) reflecting the oral infection (engagement of systemic host response mechanisms), (iii) induction of enhanced horizontal and interproximal alveolar bone resorption in rats with mixed infection as expected (direct result of local infection), (iv) induction of inflammatory response (apical migration of JE, rete ridge elongation, crestal alveolar bone loss, PMNs) consistent with established characteristics of periodontal disease, and (v) no synergistic virulence observed with P. gingivalis/T. denticola in a rat periodontal disease model. This mixed infection model will provide an opportunity for further studies to clarify the characteristics and alterations of the host response profiles such as proinflammatory cytokines and matrix metalloproteinases in periodontal tissues that relate to osteoclastic alveolar bone loss in response to mixed infections.