Highly Conserved Surface Proteins of Oral Spirochetes as Adhesins and Potent Inducers of Proinflammatory and Osteoclastogenic Factors

Abstract

Oral spirochetes include enormously heterogeneous Treponema species, and some have been implicated in the etiology of periodontitis. In this study, we characterized highly conserved surface proteins in four representative oral spirochetes (Treponema denticola, T. lecithinolyticum, T. maltophilum, and T. socranskii subsp. socranskii) that are homologs of T. pallidum Tp92, with opsonophagocytic potential and protective capacity against syphilis. Tp92 homologs of oral spirochetes had predicted signal peptides (20 to 31 amino acids) and molecular masses of 88 to 92 kDa for mature proteins. They showed amino acid sequence identities of 37.9 to 49.3% and similarities of 54.5 to 66.9% to Tp92. The sequence identities and similarities of Tp92 homologs of oral treponemes to one another were 41.6 to 71.6% and 59.9 to 85.6%, respectively. The tp92 gene homologs were successfully expressed in Escherichia coli, and the recombinant proteins were capable of binding to KB cells, an epithelial cell line, and inhibited the binding of the whole bacteria to the cells. Antiserum (the immunoglobulin G fraction) raised against a recombinant form of the T. denticola Tp92 homolog cross-reacted with homologs from three other species of treponemes. The Tp92 homologs stimulated various factors involved in inflammation and osteoclastogenesis, like interleukin-1β (IL-1β), tumor necrosis factor alpha, IL-6, prostaglandin E₂, and matrix metalloproteinase 9, in host cells like monocytes and fibroblasts. Our results demonstrate that Tp92 homologs of oral spirochetes are highly conserved and may play an important role in cell attachment, inflammation, and tissue destruction. The coexistence of various Treponema species in a single periodontal pocket and, therefore, the accumulation of multiple Tp92 homologs may amplify the pathological effect in periodontitis.

Numerous studies examining epidemiology and virulence factors have revealed strong evidence for the implication of oral spirochetes in the etiology of periodontitis, and the presence of spirochetes in the subgingival plaque is associated with an increased severity of periodontitis (16, 38, 48). Oral spirochetes include at least 50 phylotypes (12) and account for 20 to 50% of the total microscopic count in the subgingival plaque of periodontitis patients (2, 34). Ten species of oral spirochetes have been cultivated thus far (16). Some species, like Treponema denticola, T. lecithinolyticum, and T. socranskii, are more prevalently associated with periodontitis than others and are resistant to periodontal therapy (38).

The bacterial outer membrane (OM) provides a crucial contact barrier to host cells, and outer membrane proteins (OMPs) belong to the first class of molecules involved in host-bacterium interactions. They are involved in adhesion to and in the induction of cytopathic effects on host cells and in inflammatory and immunological responses. OMPs are valuable candidates with which to study host-pathogen interactions and to identify appropriate targets for therapy and prophylaxis. It is therefore important to identify common virulence determinants that are localized at the surfaces of oral spirochetes and to characterize them in order to elucidate their roles in periodontal pathogenesis. OMPs of oral spirochetes have been reported previously to participate in the colonization of the host, cytopathogenesis, and the inflammatory responses of the host in periodontitis (8, 18, 45). Most of the OMPs identified in oral spirochetes are species specific (5, 14, 18, 19, 33, 37, 45, 53).

Tp92 (837 amino acid [aa] residues), one of the surface antigens of T. pallidum, with a molecular mass of 92 kDa, has been reported previously to have opsonic potential and to induce a protective immune response to syphilis, a T. pallidum infection (7). This protein has been suggested for use as a syphilis serodiagnostic marker and has shown 98% sensitivity and 97% specificity when tested for this capacity (50). Tp92-related proteins are also found in other genera of spirochetes, like Borrelia burgdorferi and Leptospira interrogans, and they are distributed in a wide range of gram-negative bacteria, like Chlamydia trachomatis (Omp85 analog), Neisseria spp. (Omp85), Pasteurella multocida (Oma87), Haemophilus spp. (D15), and Helicobacter pylori (D15) (43). From a functional analysis, Omp85 proteins of Neisseria spp. appear to be essential for the viability of the bacteria and to be involved in the assembly of OMPs like porins PorA and PorB, secretin PilQ, and the siderophore receptors FrpB and RmpM (51), as well as in lipid transport to the OM (21). Immunization with D15 from Haemophilus spp. was shown previously to be protective against infection with Haemophilus in animal models, and D15 has been considered as a potential vaccine candidate (49), as has Tp92 from T. pallidum (7). Antisera against Oma87 of P. multocida and D15 of H. influenzae have been found previously to be immunoprotective in animal models (1, 36). Tp92/Omp85 homologs are therefore a family of highly conserved OMPs in gram-negative bacteria with possibly conserved functions.

Periodontitis is characterized by bone resorption, which leads to tooth loss. Osteoclasts and osteoblasts play important roles in bone resorption, and osteoblastic cells regulate the differentiation, fusion, and activation of osteoclasts. Osteoclasts are activated by the receptor activator of nuclear factor-κB (NF-κB) ligand (RANKL) that is stimulated from osteoblasts/stromal cells by osteoclastogenic factors such as parathyroid hormone, 1,25(OH)₂ vitamin D₃, interleukins (interleukin-1 [IL-1], IL-6, and IL-11), tumor necrosis factor alpha (TNF-α), and prostaglandin E₂ (PGE₂) (28, 31).

This study aimed at identifying and characterizing the Tp92 homologs of four representative oral spirochetes strongly associated with periodontitis. The tp92 gene homologs were identified either by the examination of the genome sequence (in the case of T. denticola) or by PCR amplification and sequencing, and the activities of the Tp92 homologs were assessed after the cloning of the tp92 gene homologs in Escherichia coli and the purification of the recombinant proteins. The amino acid sequences of the Tp92 homologs from the oral treponemes showed high levels of homology. The immuno-cross-reactivities and the surface exposure patterns of the Tp92 homologs in oral treponemes were demonstrated with a polyclonal antibody (Ab) raised against the Tp92 homolog of T. denticola. The Tp92 homologs were then shown to bind to epithelial cells and induce the expression of proinflammatory and osteoclastogenic factors in host cells like monocytes and fibroblasts.

MATERIALS AND METHODS

Bacteria and cells.

Oral treponemes (T. denticola ATCC 33521, T. lecithinolyticum ATCC 700332, T. maltophilum ATCC 51939, and T. socranskii subsp. socranskii ATCC 35536) were cultured in an anaerobic atmosphere (10% CO₂, 5% H₂, 85% N₂) using OMIZ-Pat medium as described previously (55).

E. coli DH5α-T1 (Invitrogen, Carlsbad, CA) was used for TA cloning, and E. coli M15 (Qiagen, Valencia, CA) was used for the expression of the treponeme Tp92 homologs. E. coli bacteria containing the plasmids were cultured aerobically in Luria-Bertani broth supplemented with 50 μg of ampicillin/ml for the DH5α-T1 strain or with 100 μg of ampicillin/ml and 25 μg of kanamycin/ml for the M15 strain.

THP-1 cells (ATCC TIB-202), a human monocytic cell line, and primary cultured periodontal ligament (PDL) cells were prepared as described previously (9). THP-1 and PDL cells were cultured in RPMI 1640 medium and α-minimal essential medium supplemented with 10% fetal bovine serum (FBS), respectively, and used for stimulation with the recombinant Tp92 homologs. KB cells (ATCC CCL-17), an oral carcinoma epithelial cell line, were cultured in RPMI 1640 medium supplemented with 10% FBS and used for a bacterial binding assay. An NF-κB reporter cell line, CHO/CD14/TLR4, which expresses CD14 and Toll-like receptor 4 (TLR4) and has the CD25 reporter gene with the human E-selectin promoter, was kindly provided by S. H. Han and cultured as described previously (24). The CHO/CD14/TLR4 cells were cultured in Ham's F-12 medium (HyClone Laboratories, Logan, UT) supplemented with 10% FBS and were used to verify that prepared recombinant proteins were endotoxin free by flow cytometry analysis after treatment with the recombinant forms of the Tp92 homologs.

Sequencing of tp92 homologs.

The tp92 homolog of T. denticola (the TDE2601 open reading frame [ORF] of strain ATCC 35405) (46) in the genome sequence of the bacterium was identified using the NCBI BLAST program with the T. pallidum subsp. pallidum tp92 sequence. Nucleotide sequences of the tp92 gene and the T. denticola tp92 homolog were aligned using the program Align from EMBL-EBI, and several degenerate primers were designed for the initial amplification of the tp92 homologs of T. maltophilum and T. socranskii by PCR. In the case of T. lecithinolyticum, a portion of the tp92 homolog (400 bp) was originally identified by screening an expression library for this organism with Abs directed against the T. lecithinolyticum OM fraction. The full sequences of the tp92 homologs of T. lecithinolyticum, T. maltophilum, and T. socranskii were obtained by PCR amplification using the degenerate primers described above or by inverse PCR using specific primers designed from the partial sequences, followed by sequencing using a DNA sequencer (ABI; Applied Biosystems, Foster City, CA).

Primary structure analyses of translated proteins were performed by using ProtParam from ExPASy. Pairwise sequence comparisons were performed using Align from EMBL-EBI to analyze sequence homology. The predictions of protein subcellular localization patterns and signal peptides were analyzed using PSORTb (version 2.0) (20) and SignalP (version 3.0) (6), respectively.

Cloning and expression of tp92 homologs.

The tp92 gene homologs of T. denticola, T. lecithinolyticum, T. maltophilum, and T. socranskii were amplified from the genomic DNA by PCR and cloned without the sequences encoding the leader peptides. The sequences of the PCR primers were as follows: 5′-AAC TGA GCT CGG ATG GTA TAA TGG AAA ACC TG-3′ (SacI tagged) and 5′-AAC TCT GCA GCT ATA AAT TGG GTA TAT TGA ATG AA-3′ (PstI tagged) for T. denticola, 5′-AAC TGA GCT CGA AAG CGA TTG GTA TTA CGG A-3′ (SacI tagged) and 5′-AAC TCT GC A GTT ATT GAT TGG TAA GGT TAA AAG-3′ (PstI tagged) for T. lecithinolyticum, 5′-AAC TGG ATC CCA AAG CGA CAA TAA CTG GTA-3′ (BamHI tagged) and 5′-AAC TGA GCT CTT ATT GAT TGA CCA AAT TG-3′ (SacI tagged) for T. maltophilum, and 5′-AAC TGG ATC CGA AGA AAG CGA AGA CGA AGG-3′ (BamHI tagged) and 5′-AAC TAA GCT TTT ATC AGC GGT TCG TTA TGT TGA-3′ (HindIII tagged) for T. socranskii. The primers were designed to introduce restriction sites (underlined) and an additional four nucleotides to the 5′ ends of these restriction sites. PCR was performed in a total volume of 50 μl containing 15 pmol of each primer, 1.25 U of Ex Taq polymerase (Perkin Elmer Cetus, Foster City, CA), and 10 nmol of each deoxynucleoside triphosphate in a thermal cycler (PerkinElmer). The thermal cycle chosen included an initial denaturation step at 94°C for 4 min; 30 cycles of a denaturation step at 94°C for 1 min, an annealing step at 62°C for 1 min, and an extension step at 72°C for 1 min; and a final incubation at 72°C for 5 min. The PCR products were cloned in E. coli by using the TA cloning vector pCR2.1-TOPO, and the inserts were isolated and cloned in E. coli M15 by using the expression vector pQE-30 as described previously (33). After the induction of E. coli with 1 mM isopropyl-β-d-thiogalactopyranoside (IPTG), histidine-tagged recombinant proteins were purified by sonication, solubilization with a detergent, and subsequent renaturation, followed by affinity chromatography using nickel-nitrilotriacetic acid agarose (Qiagen, Valencia, CA) as described previously (33).

Endotoxins present in E. coli, which may potentially contaminate the recombinant proteins, were removed using polymyxin B-agarose according to the instructions of the manufacturer (Sigma Chemical Co., St. Louis, MO). The recombinant proteins were concentrated using Centricon YM-30 (Millipore, Bedford, MA), sterile filtered through 0.25-μm-pore-size membrane filters, and quantified using the bicinchoninic acid protein assay kit (Pierce, Rockford, IL).

For the mock extract used as a control for endotoxin contamination in the reagents, E. coli transformed with pQE-30 without the insert DNA was cultured and induced with IPTG. The subsequent procedures and all chemicals used were the same as those used in the preparation of the Tp92 homologs. Irrelevant N-terminally hexahistidine-tagged peptides of T. lecithinolyticum MspTL (33) and PP4 (32) were also prepared using the same procedures.

The Tp92 homologs were named Td92 for T. denticola, Tl88 for T. lecithinolyticum, Tm88 for T. maltophilum, and Tss88 for T. socrnaskii subsp. socranskii, according to their molecular weights.

Verification of endotoxin decontamination of the recombinant proteins.

TLR4-dependent NF-κB activation was analyzed to verify endotoxin decontamination of the recombinant Tp92 homologs by flow cytometry analysis. CHO/CD14/TLR4 cells were cultured to 70% confluence in Ham's F-12 medium and stimulated with the recombinant Tp92 homologs (10 μg/ml) or lipopolysaccharide (LPS; 1 μg/ml) in the presence of 2% FBS for 16 h. The cells were subjected to flow cytometry analysis using fluorescein isothiocyanate (FITC)-labeled mouse anti-human CD25 (Becton Dickinson, San Diego, CA) as described previously (24). FITC-labeled mouse immunoglobulin G1 (IgG1; Becton Dickinson) was used as an isotype-matched control.

Production of antisera against the Tp92 homolog of T. denticola.

Polyclonal Abs against the T. denticola Tp92 homolog were raised in New Zealand White rabbits by the intradermal administration of 500-μg doses of the purified proteins with complete Freund's adjuvant (500 μg [catalog no. F5881; Sigma]) followed by three subsequent booster doses of 200 μg of the proteins with incomplete Freund's adjuvant (500 μg [catalog no. F5506; Sigma]) at 2-week intervals. Antisera were collected 1 week after the final booster dose. The antisera were shown to contain high titers of specific Abs against the recombinant T. denticola Tp92 homolog, as judged by enzyme-linked immunosorbent assays (ELISAs) and immunoblotting using horseradish peroxidase (HRP)-labeled mouse anti-rabbit IgG (R&D Systems, Minneapolis, MN). Abs that cross-reacted with E. coli proteins and the histidine tag were removed by affinity chromatography using activated cyanogen bromide-activated Sepharose 4B (Amersham Pharmacia Biotech, Piscataway, NJ) according to standard protocols. For this purpose, E. coli M15 sonicates or an irrelevant histidine-tagged protein, PP4 (32), was coupled to cyanogen bromide-activated Sepharose 4B. The antisera were applied to the Sepharose, and the unbound fractions of the antisera containing the specific Abs were collected. IgG Abs (designated anti-Td92 Ab) were purified using an ImmunoPure (A Plus) IgG purification kit according to the instructions of the manufacturer (Pierce, Rockford, IL). Finally, endotoxin contamination in the IgG fraction was removed using polymyxin B-agarose (Sigma).

Preparation of the OM fraction of T. denticola.

The OM of T. denticola was isolated by a combined protocol of freeze-thaw procedures and sucrose density gradient centrifugation with a slight modification (44, 47). Briefly, T. denticola cells (80-ml culture) were harvested and washed with phosphate-buffered saline (PBS) by centrifugation at 8,000 × g for 10 min. The cell pellet was resuspended in 1 ml of 50 mM Tris-HCl buffer (pH 7.2) containing 20 mM MgCl₂, 2 mM phenylmethylsulfonyl fluoride, and DNase 1 (4 U/ml). After the cells were subjected to 40 freeze-thaw cycles, the disrupted cell suspension was centrifuged at 8,000 × g for 10 min. The supernatant containing the OM was transferred into a new tube and centrifuged at 25,000 × g for 30 min. The pellet was washed twice with 50 mM Tris-HCl (pH 7.2) and resuspended in 1 ml of 0.1 M sodium acetate buffer (pH 3.0), and the suspension was mixed gently for 2 h to remove contaminating flagella and then subjected to centrifugation at 25,000 × g for 30 min. The pellet was washed with and resuspended in 0.8 ml of 50 mM Tris-HCl buffer, and the suspension was subjected to sucrose gradient centrifugation with 25, 42, and 56% sucrose at 100,000 × g for 16 h at 4°C in an ultracentrifuge (TLA110 rotor; Beckman Instruments Inc., Fullerton, CA). The upper fraction containing the OM was removed and precipitated with methanol and chloroform. The precipitated OM fraction was resuspended in 50 μl of PBS and used to detect Td92 by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and immunoblot analysis.

Immunoblotting.

The expression of the Tp92 homologs in E. coli was verified by immunoblotting using a monoclonal mouse antihistidine Ab (Qiagen, Valencia, CA). E. coli M15 cells transformed with the recombinant plasmids were cultured in Luria-Bertani broth containing antibiotics and induced with IPTG. The E. coli lysates (20 μg of protein) were subjected to SDS-PAGE and subsequently transferred onto nitrocellulose membranes. The membranes were blocked with 2% bovine serum albumin (BSA) for 1 h and allowed to react with antihistidine Ab for 1 h. After being washed with PBS-0.2% Tween 20, the membranes were allowed to react with alkaline phosphatase-labeled anti-mouse IgG for 1 h. After being washed with PBS-0.2% Tween 20, the membranes were developed with 5-bromo-4-chloro-3-indolylphosphate (165 μg/ml) and nitroblue tetrazolium (330 μg/ml).

The reactivity of the anti-Td92 Ab was tested using the lysates of T. denticola whole cells, the OM fraction, and the recombinant Td92. Whole-cell lysates of T. denticola, the OM fraction, and Td92 were separated by SDS-PAGE and transferred onto nitrocellulose membranes. The membranes were blocked with PBS-2% BSA and allowed to react with anti-Td92 Ab for 1 h, followed by HRP-labeled goat anti-rabbit IgG for 1 h. After being washed with PBS-0.2% Tween 20, the membranes were developed with 3,3′,5,5′-tetramethylbenzidine (TMB; Sigma). To verify the purity of the OM preparation, immunoblotting using rabbit anti-FlaA (1:5,000 dilution; a kind gift from Christopher Fenno), which detects the periplasmic protein FlaA, was performed.

Immunodot blotting.

The cross-reactivity of Tp92 homologs with the anti-Td92 Ab was analyzed by immunodot blotting. The recombinant Tp92 homologs (250 ng in 1 μl) were applied onto a nitrocellulose membrane and air dried. The membrane was allowed to react with the anti-Td92 Ab, and the reaction was detected as described above. An irrelevant histidine-tagged recombinant protein, MspTL (30), and E. coli lysates were used as negative controls.

Indirect immunofluorescence microscopy.

Treponema species grown to early stationary phase were harvested, resuspended in PBS, and applied to glass slides coated with silane (Sigma). Some of the slides were used directly, and the others were used after fixation with cold acetone for 15 min and subsequent permeabilization of the OMs with 0.1% Triton X-100 in PBS for 5 min. The bacteria were incubated in PBS containing 1% BSA for 1 h and then allowed to react with the anti-Td92 Ab or anti-FlaA Ab (1:1,000 dilution) for 1 h. After being washed three times with PBS, the bacteria were allowed to react with FITC-labeled goat anti-rabbit IgG (1:50 dilution in PBS containing 1% BSA) and washed three times with PBS. To see all the bacteria applied, the slides were further stained with a fluorescent nucleic acid stain, Hoechst dye 33258 (2 μg/ml in PBS; Invitrogen), for 30 min. The cells were washed three times with PBS and observed with a fluorescence microscope (Carl Zeiss, Axioskop, Germany) at a magnification of ×1,000. Anti-FlaA Ab (1:500 dilution) was used as a control to verify that the OMs of the bacteria were intact during the immunofluorescence assay. As a control for nonspecific binding, rabbit preimmunization serum was used instead of anti-Td92 Ab.

Flow cytometry analysis.

T. denticola cells (5 × 10⁷) were fixed with acetone and permeabilized as described above. Unfixed and acetone-fixed bacteria were allowed to react with the anti-Td92 Ab or the anti-FlaA Ab (1:500 dilution) in 100 μl of PBS containing 1% BSA at room temperature for 1 h. The bacteria were washed twice with PBS containing 0.02% BSA and allowed to react with FITC-labeled goat anti-rabbit IgG (1:50 dilution in 1% BSA in PBS) at room temperature for 1 h. After being washed twice with PBS, the bacteria were analyzed using a fluorescence-activated cell sorter (FACSCalibur; Becton Dickinson, San Jose, CA). The data were obtained by using 40,000 cells. The number of bacteria was determined as described previously (30). Isotype-matched Ab (rabbit IgG) was used as a negative control.

Cell binding assay.

To determine whether the Tp92 homologs were able to bind to host cells, the binding assay was performed using KB cells, an oral carcinoma epithelial cell line, and FITC-labeled Tp92 homologs. KB cells (10⁵ cells/500 μl) were cultured in 24-well culture plates to confluence. After a change to serum-free medium, the FITC-labeled Tp92 homologs were added to the cells (at 2.5 μg per well) and the plates were incubated at 37°C under 5% CO₂ for 1 h. The cells were detached from the cell culture plates by using 200 μl of 1× trypsin-EDTA, washed with serum-containing medium, and subsequently washed three times with PBS by centrifugation at 100 × g. The fluorescence of the cells was then measured by a fluorometer (FLUOStar⁺ OPTIMA; BMG Labtechnologies Inc., Durham, NC). The level of protein binding was expressed as the percentage of bound proteins relative to the total proteins added by using the following formula: (bound fluorescence)/(total fluorescence added) × 100. One microgram of Td92 was equivalent to 10⁹ cells of T. denticola, as judged by SDS-PAGE and subsequently by the immunoblotting of aliquots containing various numbers of bacteria. The number of T. denticola cells (optical density at 600 nm [OD₆₀₀] = 0.2) corresponded to a concentration of 5 × 10⁹ cells/ml, as determined using a Petroff-Hausser counting chamber (Hausser Scientific, Hosham, PA).

To confirm that the Tp92 homologs function as adhesins, the inhibition of the binding of fluorescence-labeled T. denticola cells to epithelial cells by Td92 was tested. T. denticola cells were grown to early stationary phase, harvested, and resuspended in PBS, and the suspension was adjusted to an OD₆₀₀ of 1.0. Aliquots (1 ml) of bacteria were centrifuged and resuspended in 1 ml of PBS. The bacteria were stained with 10 μM carboxyfluorescein diacetate succinimidyl ester (CFSE; Molecular Probes, Eugene, OR) for 10 min at room temperature and washed three times with PBS, followed by storage at −20°C for 24 h to induce fluorescence stabilization. CFSE-labeled bacteria were diluted with PBS to an OD₆₀₀ of 0.2 (5 × 10⁹ bacteria/ml). KB cells were cultured in 24-well culture plates (10⁵ cells/500 μl) to confluence, washed twice with serum-free medium, and incubated with CFSE-labeled bacteria (100 μl) in the presence of Td92 (1 to 10 μg) at 37°C under 5% CO₂ for 1 h. The cells were detached and washed three times with PBS by centrifugation at 100 × g, and the fluorescence of the adherent bacteria was determined using a fluorometer as described above. An irrelevant histidine-tagged recombinant protein (PP4) was used as a negative control. The ratio of T. denticola cells to KB cells used for the binding competition assay was 5,000:1.

Cell culture and treatment with Tp92 homologs.

Human monocytic THP-1 cells (1 × 10⁶ cells/ml) and primary cultured PDL cells (5 × 10⁵ cells/ml) were cultured as described above and plated into 35-mm culture dishes in the absence of serum. The cells were then treated with 10-μg/ml concentrations of each of the recombinant Tp92 homologs for 12 to 24 h. The cells were harvested and used for RNA isolation, and the conditioned supernatants were collected and stored at −70°C for the gelatin zymography and ELISA. Cells treated with E. coli LPS (10 μg/ml) were used as a positive control, and untreated or mock extract-treated cells were used as a negative control.

Real-time RT-PCR.

RNA from the THP-1 cells and PDL cells treated with the Tp92 homologs was prepared using an easy-BLUE total extraction kit (iNtRON Biotechnology, South Korea), and cDNA was synthesized from 1 μg of RNA by using a Maxime reverse transcriptase (RT) premix kit according to the protocols of the manufacturer (iNtRON Biotechnology, South Korea). For real-time PCR, cDNA (1 μl) was mixed with 10 μl of Sybr premix Ex Taq (Takara Bio Inc., Japan) and primer pairs (4 pmol each) in a 20-μl reaction volume, and the mixture was subjected to PCR for 40 cycles of a denaturation step at 95°C for 15 s, an annealing step at 60°C for 15 s, and an extension step at 72°C for 33 s in an ABI PRISM 7500 fast real-time PCR system (Applied Biosystems, Foster City, CA). The PCR products were subjected to a melting-curve analysis to verify the presence of a single amplification product. PCR without RT was performed as a negative control. The housekeeping gene encoding glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as a reference in order to normalize expression levels and to quantify changes in gene expression levels between untreated controls and Tp92 homolog-treated cells. The change (n-fold) in the expression level of each gene was determined from the difference in cycle numbers to reach a threshold value between the values for control and Tp92 homolog-treated cells and was expressed as a log₂ ratio. The sequences of the primers for real-time RT-PCR were as follows: 5′-AGC TGT ACC CAG AGA GTC C-3′ and 5′-ACC AAA TGT GGC CGT GGT TT-3′ for the IL-1β gene, 5′-AAC CTG TCC ACT GGG CAC A-3′ and 5′-TCT GGC TCT GAA ACA AAG GAT-3′ for the IL-6 gene, 5′-GTG AAG GTG CAG TTT TGC CA-3′ and 5′-TCT CCA CAA CCC TCT GCA C-3′ for the IL-8 gene, 5′-CAG GGA CCT CTC TCT AAT CA-3′ and 5′-AGC TGG TTA TCT CTC AGC TC-3′ for the TNF-α gene, 5′-CTC TGG AGG TTC GAC GTG A-3′ and 5′-CTG CAG GAT GTC ATA GGT CA-3′ for the matrix metalloproteinase 9 (MMP-9) gene, 5′-AAG CTG GGA AGC CTT CTC TA-3′ and 5′-GTG CTG GGC AAA GAA TGC AA-3′ for the cyclooxygenase 2 (COX-2) gene, and 5′-GTG GTG GAC CTG ACC TGC-3′ and 5′-TGA GCT TGA CAA AGT GGT CG-3′ for the GAPDH gene.

ELISAs for cytokines and PGE₂.

The culture supernatants of the cells treated with Tp92 homologs were assayed to determine IL-1β, TNF-α, IL-6, IL-8, and PGE₂ levels by using the ELISA kits from R&D Systems (Minneapolis, MN).

Gelatin zymography.

To detect MMP-9 expression, a gelatin zymography analysis of the culture supernatants (25 μl) of THP-1 cells treated with the Td92 homologs was performed as described previously (9). Culture supernatants of THP-1 cells treated with 10 μg of E. coli LPS were used as positive controls. Culture supernatants of untreated cells or mock extract-treated cells were used as negative controls. To confirm that the gelatinolytic activity was based on MMP-9, but not on other proteases, the gels were incubated in reaction buffer containing 2 mM EDTA, an MMP inhibitor, after SDS-PAGE.

Statistical analyses.

Statistically significant differences between the untreated and recombinant Tp92 homolog-treated cells were evaluated using Student's t test. A P value of <0.05 was considered significant.

Nucleotide sequence accession numbers.

DNA sequences of the tp92 gene homologs have been submitted to the GenBank database under accession no. EU057146 for tl88 of T. lecithinolyticum, EU057147 for tss88 of T. socranskii subsp. socranskii, and EU057148 for tm88 of T. maltophilum.

RESULTS

Distribution of the tp92 gene homologs in oral treponemes.

From the genome sequence of T. denticola (ATCC 35405), the tp92 gene homolog was identified as ORF TDE2601, which has been reported previously to encode a putative surface antigen (46), and other tp92 gene homologs were identified by sequencing the PCR products of the genes that were amplified with degenerate and specific primers for T. lecithinolyticum, T. maltophilum, and T. socranskii subsp. socranskii. The ORFs of the tp92 homologs of T. denticola, T. lecithinolyticum, T. maltophilum, and T. socranskii subsp. socranskii were 2,457 bp (corresponding to 818 aa residues), 2,430 bp (809 aa), 2,430 bp (809 aa), and 2,442 bp (813 aa), respectively.

The translated amino acid sequences corresponding to the tp92 homologs in the oral treponemes revealed significant homology to the amino acid sequence of T. pallidum Tp92. The Tp92 homologs of four oral treponemes showed amino acid sequence identities of 37.9 to 49.3% and similarities of 54.5 to 66.9% to Tp92 (Table 1). The sequence identities and similarities of the oral treponeme homologs to one another were 41.6 to 71.6% and 59.9 to 85.6%, respectively. The Tp92 homologs showed sequence identities of 28.4 to 30.2% and 22.2 to 23.7% and similarities of 46.9 to 49.7% and 37.0 to 39.4% to homologs from two nonoral spirochetes, B. burgdorferi B31 and L. interrogans serovar Copenhageni, respectively.

TABLE 1.

Degrees of amino acid sequence homology of the Tp92 homologs

Bacterium	% Identity/% similarity of Tp92 homolog from the indicated species to Tp92 homolog from:
	T. denticola	T. lecithinolyticum	T. maltophilum	T. socranskii	T. pallidum	B. burgdorferi B31	L. interrogans serovar Copenhageni
T. denticola	100
T. lecithinolyticum	44.0/63.3	100
T. maltophilum	43.5/61.2	71.6/85.6	100
T. socranskii	41.6/59.9	47.3/66.8	46.9/67.2	100
T. pallidum	49.3/66.9	39.6/57.5	38.2/57.4	37.9/54.5	100
B. burgdorferi B31	30.2/46.9	29.2/48.6	29.0/49.7	28.4/47.2	28.6/46.8	100
L. interrogans serovar Copenhageni	22.7/37.0	23.4/38.6	23.7/39.4	22.2/38.8	22.4/37.8	21.1/38.3	100

All Tp92 homologs have N-terminal signal peptide sequences. The program SignalP identified signal sequences of 20 aa for T. denticola, 27 aa for T. lecithinolyticum, 25 aa for T. maltophilum, and 31 aa for T. socranskii subsp. socranskii. A cleavable 21-aa N-terminal signal sequence in Tp92 of T. pallidum has been identified previously (7). PSORTb predicted the localization of Tp92 homologs in four oral treponemes in the OM, with likelihoods of more than 94% (94.5% for T. denticola and 100% for T. lecithinolyticum, T. maltophilum, and T. socranskii). The predicted pIs and molecular masses of the mature Tp92 homologs were 5.79 and 91.55 kDa for T. denticola, 6.34 and 88.33 kDa for T. lecithinolyticum, 5.54 and 88.58 kDa for T. maltophilum, and 5.46 and 88.03 kDa for T. socranskii subsp. socranskii when analyzed by the ProtParam tool from ExPASy. The pI and molecular mass of Tp92 are 8.27 and 91.97 kDa. According to the molecular masses of the proteins, we designated the T. denticola, T. lecithinolyticum, T. maltophilum, and T. socranskii subsp. socranskii tp92 gene homologs td92, tl88, tm88, and tss88 and the Tp92 protein homologs Td92, Tl88, Tm88, and Tss88, respectively.

Cloning and expression of the tp92 gene homologs in E. coli.

The tp92 gene homologs were successfully expressed in E. coli by using the pQE-30 vector after PCR and TA cloning (Fig. 1). As expected from the deduced amino acid sequences, the molecular sizes of the recombinant Tp92 homologs were approximately 92 kDa for the T. denticola homolog and 88 kDa for the T. lecithinolyticum, T. maltophilum, and T. socranskii subsp. socranskii homologs. The recombinant proteins formed in the inclusion bodies were purified by solubilization, renaturation, and affinity chromatography using nickel-nitrilotriacetic acid agarose. These proteins were employed in the study for host cell stimulation after further purification through the removal of any possible E. coli LPS contamination by using polymyxin B-agarose.

FIG. 1.

Expression of the tp92 gene homologs in E. coli. The tp92 gene homologs were cloned in E. coli, and the expression of histidine-tagged recombinant proteins was analyzed by SDS-PAGE (10% polyacrylamide gel) (A) and by immunoblotting with an antihistidine Ab (B) after induction with IPTG. Lanes: Td92, Tp92 homolog of T. denticola; Tl88, Tp92 homolog of T. lecithinolyticum; Tm88, Tp92 homolog of T. maltophilum; and Tss88, Tp92 homolog of T. socranskii subsp. socranskii. −, uninduced E. coli cell lysates; +, IPTG-induced E. coli cell lysates. The positions of protein size markers (M) are indicated.

Cross-reactivity of Ab against Td92.

An IgG fraction of antisera raised against recombinant Td92 (anti-Td92 Ab) was purified and tested for reactivity with T. denticola and recombinant Tp92 homologs. An immunoblot analysis showed that the Abs reacted with a 92-kDa-protein band of T. denticola lysates and the recombinant Td92 (Fig. 2A). An immunodot blot analysis showed that the anti-Td92 Ab cross-reacted with Tl88, Tm88, and Tss88 (Fig. 2B) but that it did not react with an irrelevant histidine-tagged recombinant protein (MspTL) (33) or E. coli lysates.

FIG. 2.

Immunoreactivities of T. denticola whole-cell lysates (WC) and Tp92 homologs with anti-Td92 Ab. (A) T. denticola whole-cell lysates (10 μg) and Td92 were subjected to SDS-PAGE (8% polyacrylamide gel) followed by immunoblotting with anti-Td92 Ab. The applied amounts of Td92 were 500 ng for SDS-PAGE and 100 ng for immunoblotting. (B) The cross-reactivity of the Tp92 homologs with the anti-Td92 Ab was examined by an immunodot blot assay. The Tp92 homologs (250 ng in 1 μl) were applied to a nitrocellulose membrane and allowed to react with the anti-Td92 Ab. The detection was performed using HRP-labeled goat anti-rabbit IgG and TMB as HRP substrates. As a negative control, E. coli lysates and an irrelevant recombinant protein, MspTL, were included.

OM location of Td92 in T. denticola.

Although a computer-assisted analysis using the PSORTb program predicted that Tp92 homologs were OMPs, as described above, we empirically evaluated the surface location of T. denticola Td92 by various techniques. First, we examined the presence of Td92 in the OM fraction isolated from T. denticola by using anti-Td92 Ab. As shown in Fig. 3, the anti-Td92 Ab raised against recombinant Td92 recognized a single protein band at the position of about 92 kDa in the OM preparation. The anti-FlaA Ab detecting the periplasmic protein FlaA did not react with the OM fraction but with a 37-kDa protein in whole-cell lysates, verifying the purity of the OM fraction. Second, the abilities of intact bacteria to react with anti-Td92 Ab were assessed using an indirect immunofluorescence assay. Because of the fragile nature of the Treponema OM, anti-Fla Ab was included to control the OM integrity. As shown in Fig. 4, unfixed T. denticola cells were positive with anti-Td92 Ab but negative with anti-FlaA Ab. Acetone-fixed and permeabilized T. denticola cells were positive with both Abs, indicating that the OMs of the bacteria were intact and that Td92 was surface exposed. Furthermore, anti-Td92 Ab cross-reacted with T. lecithinolyticum, T. maltophilum, and T. socranskii cells, suggesting the surface exposure of the Tp92 homologs (data not shown). The cross-reactivity of anti-Td92 Ab as demonstrated by immunofluorescence was coincident with that observed in immunodot blot assays. Third, the flow cytometry analysis using anti-Td92 Ab showed that more than 80% of unfixed T. denticola cells were positive while a minor portion of the bacteria was detected with anti-FlaA Ab (Fig. 5A). Most of the acetone-fixed and permeabilized bacteria were positive with both Abs (Fig. 5B). Isotype-matched Ab was used as a negative control.

FIG. 3.

Identification of Td92 in the OM of T. denticola. A T. denticola OM fraction was prepared and subjected to SDS-PAGE (10% polyacrylamide gel), followed by immunoblotting with the anti-Td92 Ab and anti-FlaA Ab. The detection was performed using HRP-labeled goat anti-rabbit IgG and TMB as HRP substrates. WC, T. denticola whole-cell lysates (10 μg); OM, OM fraction (2 μg); M, protein size marker.

FIG. 4.

Immunofluorescence micrographs of T. denticola. Nonfixed (A) and acetone-fixed and permeabilized T. denticola cells (B) were allowed to react with anti-Td92 Ab or anti-FlaA Ab, followed by FITC-labeled goat anti-rabbit IgG. The same cells were further stained with Hoechst dye, a fluorescent nucleic acid stain. Images of the same fields for immunofluorescence and Hoechst dye staining were obtained. Original magnification, ×1,000.

FIG. 5.

Flow cytometry analysis of Td92 expression. Nonfixed (A) and acetone-fixed and permeabilized T. denticola cells (B) were allowed to react with anti-Td92 Ab (α-Td92 Ab) or anti-FlaA Ab, followed by FITC-labeled goat anti-rabbit IgG. The cells were then analyzed by fluorescence-activated cell sorting. Isotype-matched rabbit IgG was used as a negative control. The experiments were performed three times, and representative data are shown.

Binding of the Tp92 homologs to KB cells.

Bacterial binding to host cells is an initial step in colonization by pathogens. To determine whether Tp92 homologs function as adhesins, the capacities for binding of the Tp92 proteins to KB cells, an epithelial cell line, were examined. As shown in Fig. 6, FITC-labeled Tp92 homologs were able to bind to the cells. To test whether this binding was specific, a competitive binding inhibition assay was performed using fluorescence-labeled T. denticola cells and unlabeled Td92. As shown in Fig. 7, the binding of CFSE-labeled T. denticola cells to KB cells was inhibited by unlabeled Td92 in a dose-dependent manner, whereas an irrelevant recombinant protein (PP4) with a histidine tag did not inhibit the binding of T. denticola to KB cells.

FIG. 6.

Binding of the Tp92 homologs to KB cells. KB cells (10⁵/500 μl) were cultured in 24-well plates to confluence and incubated with FITC-labeled Tp92 homologs (2.5 μg of protein) at 37°C for 1 h. The cells were detached and washed with PBS, and the fluorescence of bound proteins was measured using a fluorometer. Protein adhesion was expressed as a percentage of bound protein relative to the total proteins added. The experiments were repeated three times, and similar results were obtained. The results are presented as means ± standard deviations for triplicate samples.

FIG. 7.

Inhibition of T. denticola binding to KB cells by Td92. KB cells (10⁵ cells/500 μl) were cultured in 24-well plates and incubated with CFSE-labeled T. denticola whole cells (100 μl of a culture with an OD₆₀₀ of 0.2, or 5 × 10⁸ bacteria) in the presence of Td92 at different concentrations for 1 h. After the cells were detached and washed with PBS, the fluorescence of bound bacteria was measured by a fluorometer. An irrelevant recombinant protein (PP4) was used as a negative control. The experiments were repeated three times, and similar results were obtained. The results are presented as means ± standard deviations for triplicate samples and are expressed as relative light units. *, statistical significance indicated by a P value of <0.05.

Host cell responses to the Tp92 homologs.

Prior to the stimulation of the host cells with the recombinant Tp92 homologs, we verified endotoxin decontamination of the recombinant proteins by using CHO/CD14/TLR4 cells that expressed CD25 by TLR4-dependent NF-κB activation. The Tp92 homologs did not increase CD25 expression in the cells, whereas LPS significantly increased the expression of CD25 (data not shown).

Using THP-1 and PDL cells, we analyzed the regulation of representative proinflammatory and osteoclastogenic factors that are known to stimulate RANKL, which is involved in osteoclastogenesis leading to bone resorption. Monocytes are among the early immune cells that react to the local presence of bacteria, and PDL is a supporting tissue linking the root surface and alveolar bone. As shown in Fig. 8, the addition of the Tp92 homologs at the concentration of 10 μg/ml significantly upregulated IL-1β, TNF-α, IL-6, and IL-8 at both the gene and protein levels in THP-1 cells as analyzed by real-time RT-PCR and ELISAs, respectively. In PDL cells, IL-6 and IL-8 were upregulated by Tp92 homologs (Fig. 8). A dose dependence analysis showed that Td92 at the concentration of 1 μg/ml significantly induced IL-8 mRNA (data not shown). The expression of COX-2 and its product, PGE₂, in both THP-1 and PDL cells was also induced by the Tp92 homologs (Fig. 9). In a gelatin zymography analysis, the Tp92 homologs induced gelatinolytic activity corresponding to a molecular mass of 92 kDa, attributed to MMP-9, in THP-1 cells (Fig. 10). This gelatinolytic activity was inhibited by 2 mM EDTA, an MMP inhibitor (data not shown). The Tp92 homologs themselves showed no gelatinolytic activity.

FIG. 8.

Upregulation of host factors by the Tp92 homologs in THP-1 and PDL cells. THP-1 cells (1 × 10⁶/ml) or PDL cells (5 × 10⁵/ml) were cultured and treated with the Tp92 homologs (10 μg/ml) or LPS (10 μg/ml) for 8 to 24 h. RNA was isolated from the cells and subjected to real-time RT-PCR (A), and the conditioned media were used for ELISAs (B). LPS-treated cells were used as a positive control. *, statistical significance indicated by a P value of <0.05 compared to untreated control cells (con).

FIG. 9.

Upregulation of COX-2 and PGE₂ by the Tp92 homologs. THP-1 cells (1 × 10⁶/ml) (A) and PDL cells (5 × 10⁵/ml) (B) were cultured and treated with the Tp92 homologs (10 μg/ml) or LPS (10 μg/ml) for 12 to 24 h. RNA was isolated from the cells, and COX-2 mRNA expression was analyzed by real-time RT-PCR. The conditioned media were used for measuring PGE₂ levels by ELISAs. LPS-treated cells were used as a positive control. *, statistical significance indicated by a P value of <0.05 compared to untreated control cells (con).

FIG. 10.

Upregulation of MMP-9 by the Tp92 homologs in THP-1 cells. THP-1 cells (1 × 10⁶/ml) were treated with the Tp92 homologs (10 μg/ml) or LPS (10 μg/ml) for 12 to 24 h. (A) RNA was isolated from the cells, and MMP-9 mRNA expression was analyzed by real time RT-PCR. (B) The conditioned media were used for analyzing MMP-9 expression by gelatin zymography. LPS-treated cells were used as a positive control. *, statistical significance indicated by a P value of <0.05 compared to untreated control cells (con).

DISCUSSION

Microbial genome projects have unraveled the genome sequences of an expanding number of pathogens, including major oral pathogens (13), and the comparison of genes among bacterial species allows for the determination of gene functions based on the similarities of corresponding amino acids and conserved functional domains. The information on bacterial genome sequences facilitates the elucidation of bacterial physiology, biosynthesis, and metabolic pathways. With comparative genomics, it is also possible to determine the virulence factors involved in pathogenesis. The enormous diversity of oral treponemes, even within a single periodontitis patient, has been a roadblock to the study of their common mechanisms of pathogenicity. Recently, the whole genome sequence of T. denticola, the most intensively studied oral spirochete, has been deciphered (46). Based on this genomic information, we have been searching for conserved surface antigens present in representative oral spirochetes associated with periodontitis to examine their functional roles and, thus, to find insights into common mechanisms of pathogenesis useful for the development of a strategy for growth inhibition or eradication. The identification and functional analysis of conserved OMPs are therefore important for understanding the common mechanisms of pathogenesis in the extremely diverse collection of oral spirochetes. In this study, we identified and characterized highly conserved surface antigens of oral spirochetes that are homologous to Tp92 in T. pallidum, which has opsonic potential (7). The Tp92 homologs of oral spirochetes showed high degrees of homology to one another, as well as to Tp92, and the Ab raised against Td92 cross-reacted with the Tp92 homologs of T. lecithinolyticum, T. maltophilum, and T. socranskii subsp. socranskii. Using anti-Td92 Ab, we could demonstrate that Td92 is an OMP with a surface-exposed epitope. The fact that anti-Td92 recognized other Treponema species included in this study suggests that Tp92 homologs of oral spirochetes have similar surface structural motifs.

In spite of high levels of homology to Tp92, the Tp92 homologs of oral spirochetes have distinct characteristics. First, the Tp92 homologs of oral spirochetes have no serine-rich region at the C terminus, which is a signature sequence in the T. pallidum Tp92. The C terminus of Tp92 encompassing the serine repeat signature sequence (245 aa) was not expressed in E. coli. However, the full-length tp92 gene homologs of oral spirochetes were expressed in E. coli, showing products with expected molecular masses. Second, the calculated pIs of the Tp92 homologs were different from that of Tp92. The pIs of the Tp92 homologs were in the acidic range between 5.46 and 6.34, while the pI of Tp92 is 8.27. Third, a cysteine residue is absent in the Tp92 homologs of oral spirochetes, and only one is present in the case of T. lecithinolyticum and T. maltophilum, while three cysteine residues are present in Tp92, implying possible intradisulfide bond formation. It will be interesting to examine whether these differences are linked to distinct functions.

T. denticola, along with Porphyromonas gingivalis and Tannerella forsythensis (“Tannerella forsythia”), belongs to the red cluster according to the classification by Socransky et al. (48), who defined bacterial groups in dental plaque biofilm with respect to their reciprocal associations and sequential formation. The red complex produces significant amounts of proteases, including trypsinlike peptidase activity, and is in contact with the gingival junctional epithelium. In this study, we observed that the Tp92 homologs adhered to epithelial cells, suggesting that the encoded proteins mediate the binding of Treponema cells to host cells and thus accelerate the pathological effects of the bacteria. Several surface proteins of T. denticola have been reported previously to attach to host cells or extracellular matrix proteins. The best studied is the 53-kDa Msp, which is able to bind epithelial cells and fibroblasts and extracellular matrix proteins like fibronectin, keratin, laminin, collagen, fibrinogen, hyaluronic acid, and heparin (14, 15). Chymotrypsinlike protease has been demonstrated previously to bind to PDL epithelial cells (18) and to P. gingivalis fimbriae, implicating it in the coaggregation of T. denticola and P. gingivalis (25). The leucine-rich-repeat protein LrrA has been demonstrated previously to attach to human epithelial cells (29). OppA (70 kDa) binds soluble plasminogen and fibronectin (19). In the present study, Td92 significantly inhibited T. denticola adhesion to epithelial cells, although not completely. This fact suggests that Td92 is one of the T. denticola adhesins.

The Tp92 homologs upregulated IL-1β, TNF-α, IL-6, IL-8, PGE₂, and MMP-9 in THP-1 cells. These factors are proinflammatory and osteoclastogenic factors and have been reported previously to be stimulated by periodontopathogens, including the members of the red complex. IL-1β and TNF-α are representative proinflammatory cytokines associated with periodontal disease characterized by a loss of connective tissue attachment and bone resorption leading to alveolar bone loss (22). IL-1β is a potent stimulator of bone resorption and an inhibitor of bone formation that enhances osteoclast differentiation and acts as a survival factor for mature osteoclasts (26, 31, 54). It has been shown previously to activate RANKL in PDL cells (40). TNF-α promotes the production of RANKL and macrophage colony-stimulating factor by stromal cells and induces osteoclast formation. In experimental periodontitis, IL-1 and TNF antagonists have been demonstrated previously to inhibit the inflammatory response and bone loss (22). IL-6, together with IL-1β and TNF-α, is a major mediator of the host response to tissue destruction and bone resorption. IL-6 binds to its receptor, existing in either a membrane-bound form (CD126) or a soluble form (soluble IL-6 receptor alpha). This complex subsequently binds to a signal-transducing protein, gp130 (42), leading to the upregulation of RANKL in osteoblasts, which is a key factor in osteoclastogenesis. IL-8 is a representative chemokine and leads inflammatory cells to infection sites, playing an important role in initiating the inflammatory reactions. PGE₂ stimulates osteoclastogenesis through the stimulation of RANKL production, the inhibition of osteoprotegerin secretion by osteoblasts, and the stimulation of IL-6 production (11, 35). PGE₂ is synergistically stimulated by IL-1 and TNF in human gingival fibroblasts (56). MMP-9 is a zinc-dependent proteinase and has been demonstrated previously to play an important role in the bone resorption process by dissolving the bone matrix, releasing the chemotactic factor on osteoclasts, and activating the cytokines involved in bone resorption (17). Although various periodontopathogens have been shown previously to induce all these host factors, only a few bacterial factors have been identified at the molecular level. LPS is a prominent molecule that induces proinflammatory cytokines and bone-resorbing factors. LPS or LPS-like molecules of periodontopathogens have been demonstrated previously to modulate biological activities, including the upregulation of IL-1, TNF-α, IL-6, IL-8, PGE₂, and MMP-9, in various host cell types (4, 10, 23, 39). These factors are found at significantly higher levels in the tissues and gingival crevicular fluids of periodontitis patients than in those of healthy subjects (52, 57). Among surface proteins, the OMP 100 of Aggregatibacter (Actinobacillus) actinomycetemcomitans and fimbrillin of P. gingivalis (41 and 67 kDa) have been demonstrated previously to induce proinflammatory cytokines in host cells, including epithelial cells and macrophages (3, 27, 41). The 67-kDa P. gingivalis fimbria protein induces osteoclast differentiation, and the blocking of IL-1, TNF, and IL-6 induced by the protein reduces osteoclast differentiation. Recently, we reported that the major surface protein MspTL of T. lecithinolyticum induces IL-1β, TNF-α, IL-6, and PGE₂ in THP-1 cells (32, 33). Although the Tp92 homologs are not abundant surface proteins like Msp of T. denticola and MspTL of T. lecithinolyticum, as judged by SDS-PAGE profiles, they are widely distributed in oral spirochetes, and the coexistence of various species of oral spirochetes in a single patient or site may result in the amplification of the cytopathological effects on host cells.

In summary, Tp92 homologs are surface-exposed OMPs that are common among the four oral Treponema species strongly associated with periodontitis. Using the recombinant forms of the Tp92 homologs that were expressed in E. coli, we have shown that these proteins have the potential to bind to host cells and stimulate host factors that contribute to inflammation and osteoclastogenesis, which are major characteristics of periodontitis. The roles of these common antigens as protective immunogens against periodontitis will be studied in the future.