INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Adult periodontitis is a chronic infection of the periodontium that results in periodontal destruction and alveolar bone loss (11, 32). Porphyromonas gingivalis, a gram-negative and black-pigmented anaerobe, has been etiologically associated with various types of periodontal diseases including adult periodontitis (1, 2). The organism expresses a number of potential virulence factors, which have been implicated in the pathogenesis of adult-onset periodontitis (18). P. gingivalis fimbriae have been reported to exhibit a wide variety of biological and immunological activities and are recognized as a major virulence factor in the infection and pathogenesis of this organism (5, 17, 24).

Several studies have demonstrated a divergence in in vitro pathogenicities among strains of P. gingivalis when evaluated by subcutaneous injection of the organism into rodents (15, 16, 25, 33). Other approaches have included serological and genetic typing to examine the relationship of P. gingivalis and its periodontal pathogenicity (10, 16, 17, 19, 26, 29, 30). However, no clear relationship between experimental dermal infections and oral infections was demonstrated regarding the virulence capability of P. gingivalis.

Recently, we examined the distribution of P. gingivalis in periodontitis patients using genotyping of the fimA gene, which encodes fimbrillin (FimA), a subunit protein of fimbriae of this organism (6). These results indicated that the occurrence of the organisms with different fimA genotype distributions showed a clear relationship with periodontal destruction, and P. gingivalis with type II fimA was found to be significantly predominant in severe periodontitis patients. The investigation also revealed the existence of a P. gingivalis strain(s) untypeable by our PCR assay, which enables us to divide fimA genes into four types, and the prevalence of the untypeable strain(s) was 6.3% of the dental plaque samples from periodontitis patients. These data indicate that unknown fimA genes could exist within the P. gingivalis strains, and those untypeable organisms may affect the development and progression of periodontitis. In this study, we cloned a new type of the fimA gene from the untypeable fimA specimens, and we isolated a P. gingivalis strain carrying a new fimA gene.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Bacterial strains. P. gingivalis strains ATCC 33277 (fimA type I), HW24D1 (fimA type II), 6/26 (fimA type III), and HG564 (fimA type IV) were selected from our culture collections. These organisms were grown in GAM broth (Nissui, Tokyo, Japan) supplemented with 5 µg of hemin per ml and 1 µg of menadione per ml anaerobically (80% N₂, 10% H₂, and 10% CO₂) at 37°C. For the binding assay, these organisms were grown anaerobically in tryptic soy (TS) broth (Difco, Detroit, Mich.) supplemented with hemin and menadione. Escherichia coli XL10-Gold (Stratagene, La Jolla, Calif.) was cultured in Luria-Bertani medium or on a Luria-Bertani agar plate supplemented with 100 µg of ampicillin per ml for cloning and sequencing of the cloned gene. For isolation of P. gingivalis from clinical specimens, TS agar (Difco) supplemented with 5% rabbit blood, hemin, and menadione (TS blood agar) was used.

Clinical specimens. Subgingival plaque samples were collected from patients with periodontitis. These subjects were enrolled with informed consent (6). The bacterial genomic DNA from plaque samples was isolated with a DNA isolation kit according to the manufacturer's instructions (Puregene; Gentra Systems, Minneapolis, Minn.), and the isolated DNA was dissolved in 100 µl of TE (10 mM Tris HCl [pH 8.0] and 1 mM EDTA) buffer. The clinical parameters of the patients were described in our previous study (6).

Cloning of a new fimA gene by PCR. Among 73 samples which were positive for the P. gingivalis 16S rRNA gene in our previous study (6), five clinical specimens were found to contain an untypeable fimA gene(s). Thus, these five samples were used as templates for a new fimA gene. Oligonucleotides M11 (AATCTGAACGAACTGCGACGCTAT) and M12 (CTCCCTGTATTCCGAATATAGAC) were designed for amplification of the fimA gene with the open reading frame (ORF) and promoter region according to the sequences of the fimA genes reported previously (14). The PCR amplification was performed in a total volume of 50 µl consisting of 0.2 µM (each) primer, 5 µl of template DNA, and 2.5 U of ExTaq (Takara Shuzo, Otsu, Japan) according to the manufacturer's instructions. The amplification reaction was performed in a model 9700 thermal cycler (PE Applied Biosystems, Branchburg, N.J.) with the following cycling parameters: an initial denaturation at 95°C for 5 min; 30 cycles consisting of 94°C for 30 s, 55°C for 30 s, and 72°C for 1 min; and a final extension at 72°C for 7 min. The PCR products were separated by electrophoresis using a 1% agarose gel, and the amplified DNA (about 1.3 kb) was extracted using QIAEX (Qiagen, Düsseldorf, Germany). The DNA was directly cloned into pGEM-T vector (Promega, Madison, Wis.). The nucleotide sequence of the cloned gene was determined by using a dye-terminator reaction with a model 310 Genetic Analyzer (PE Applied Biosystems).

Data analysis of nucleotide sequence and amino acid sequence. Data analyses of nucleotide sequences and deduced amino acid sequences were performed with GeneWorks software (IntelliGenetics, Mountain View, Calif.). Multiple alignment analysis and construction of a phylogenetic tree were performed with CLUSTAL W in the DNA Data Bank of Japan (DDBJ; Mishima, Japan) (28). The sequence data for the fimA genes of P. gingivalis 381 (type I fimA), ATCC 33277 (type I), BH18/10 (type I), HW24D1 (type II), OMZ314 (type II), OMZ409 (type II), ATCC 49417 (type II), 6/26 (type III), and HG564 (type IV) were obtained from DDBJ under accession no. D17794, D17795, D17796, D17797, D17798, D17799, D17800, D17801, and D17802, respectively.

Prevalence of fimA type-specific P. gingivalis in periodontitis patients. Supra- and subgingival plaque samples were taken from a total of 93 periodontitis patients in our previous study (6). The plaque samples were then processed to isolate bacterial DNA as described previously (6). The isolated DNA was dissolved in 100 µl of TE buffer at 65°C for 10 min and then stored at 20°C until use. The detection of P. gingivalis and fimA typing were performed as described previously (6). The specificities and sensitivities of the fimA genotype-specific sets (types I to IV) for the primers have been previously demonstrated (6). A ubiquitous primer set that matches almost all bacterial 16S rRNA genes was used as a positive control, and the P. gingivalis species-specific primers (16S rRNA) were used for fimA typing. All primers were purchased from Amersham Pharmacia Biotech (Tokyo, Japan). PCR was performed as described previously (6). To demonstrate the presence of P. gingivalis with a new, genotype V fimA gene, type-specific primers were constructed as follows; fimV-f, AACAACAGTCTCCTTGACAGTG; fimV-r, TATTGGGGGTCGAACGTTACTGTC. The specificities of the prospective primers were tested by the program Amplify (21), and no amplification was detected for any of the strains listed in our previous report (6) other than the prospective positive samples (data not shown).

Isolation of P. gingivalis with a new type of fimA gene. Subgingival plaque samples containing P. gingivalis with the type V fimA gene were spread onto TS blood agar following dilution with phosphate-buffered saline and cultured anaerobically for 5 days at 37°C. Black-pigmented colonies were directly screened by using PCR with the specific primers (6). The clinical isolates were further examined by Gram staining and for anaerobic growth, the inability to ferment glucose, and the production of indole. The obtained isolate was maintained in GDO medium (group of difficult organisms medium; Nissui).

Analysis of genes of P. gingivalis strains. Southern blot analysis was performed as described by Ausubel et al. (7). Total genomic DNAs from lysozyme lysates of the organisms were digested with EcoRI, BamHI, or HindIII (New England Biolabs, Beverly, Mass.), and the DNA fragments separated by 1% agarose gel electrophoresis were transferred to a nylon membrane (GeneScreen; NEN, Boston, Mass.). The fimA gene (1.3 kb) was amplified from genomic DNA of P. gingivalis ATCC 33277 with the M11 and M12 primers. The sod (superoxide dismutase; 575-bp) and kgp (Lys-gingipain; 560-bp) genes were amplified from the genomic DNA of the organism. These genes were then radiolabeled with [-³²P]dCTP (Amersham Pharmacia Biotech) by using the BcaBest DNA labeling kit (Takara Shuzo). The blotted membranes were prehybridized and hybridized according to the protocol described by the manufacturer (NEN). After hybridization, the membranes were washed in 1× SSPE (180 mM NaCl, 10 mM NaPO₄, 1 mM EDTA, pH 7.4) at room temperature twice and washed in 0.1× SSPE for 20 min at 65°C. The washed membranes were exposed to Kodak BioMax MR films for 2 h at 70°C with an intensifying screen.

Immunoreactivity of FimA. Bacterial cells were harvested by centrifugation at 5,000 × g for 10 min and washed twice with ice-cold phosphate-buffered saline. The cells were resuspended in sodium dodecyl sulfate (SDS) gel loading buffer (7) and boiled for 5 min. The cellular proteins were separated by SDS-12% polyacrylamide gel electrophoresis and were transferred onto a polyvinylidene difluoride membrane (Immobilon; Millipore, Bedford, Mass.) at 24 mA for 1.5 h. The transferred protein bands were reacted with rabbit antibodies to recombinant FimA (rFimA) derived from strain 381 (type I fimA) and strain HG564 (type IV fimA) antibodies, respectively. The reactions were visualized using an alkaline phosphatase-conjugated anti-rabbit immunoglobulin G antibody (New England Biolabs) and 5-bromo-4-chloro-3-indolylphosphate-nitroblue tetrazolium substrate (Moss Inc., Pasadena, Md.).

Assay for binding of P. gingivalis cells to saliva-coated hydroxyapatite (sHA) beads. Abilities of P. gingivalis cells to bind sHA beads were examined as described previously (3). Briefly, HA beads (3 mg) were incubated with 300 µl of clarified human whole saliva in silicon-coated borosilicate tubes at 25°C for 18 h. For determination of specific binding of salivary proteins, HA beads (3 mg) were incubated with buffered KCl (pH 6.8) for the same periods. The sHA and uncoated HA beads were washed twice with buffered KCl and incubated with 10⁶ to 10^{8 3}H-labeled cells at room temperature for 1 h. After incubation, the reaction mixture was layered on 1.5 ml of 100% Percoll (Amersham Pharmacia Biotech) in a new siliconized borosilicate tube to separate unbound cells. The radioactivities of the cell-bound beads were determined in a liquid scintillation counter (model 1401; LKB, Uppsala, Sweden). The assay was performed in triplicate and repeated three times. One-way analysis of variance and the Tukey-Kramer test were used for statistical analysis of comparison of binding abilities (P < 0.01).

Nucleotide sequence accession number. The type V fimA gene sequence is available from GenBank (accession no. AB027294).

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Cloning of a new type of fimA. Cloning of a new fimA gene was carried out using a pair of PCR primers, M11 and M12. Five out of 73 plaque samples containing P. gingivalis that possessed untypeable fimA genes were used as templates. An expected-size DNA fragment was amplified by PCR from three of these samples. After cloning of PCR products into pGEM-T vector, the nucleotide sequences were determined for at least five individual clones from each sample. The sequences of all the clones were found to be identical (data not shown). The multiple alignment analysis showed that this gene fragment has sequence homology with type I fimA (strain ATCC 33277) (49%), type II fimA (strain HW24D1) (49%), type III fimA (strain 6/26) (52%), and type IV fimA (strain HG564) (59%). This gene fragment contained a 10 region (35 to 40) and a 35 region (58 to 63) upstream of the putative initiation codon, which are essential for expression of the fimA gene of P. gingivalis (34). This promoter sequence was found to be conserved among all fimA types. The comparison of the deduced amino acid sequences of type V fimA and other types of fimA is shown in Fig. 1. Type V FimA has an 18-amino-acid putative signal sequence, and this region was very similar to that of type IV FimA (strain HG564). However, the sequence of type V FimA was considerably different from those of other strains. The percent homologies of type V FimA with type I FimA (strain ATCC 33277), type II (HW24D1), type III (6/26), and type IV (strain HG564) were 42, 41, 40, and 52%, respectively. The phylogenetic tree showed that type IV fimA (strain HG564) and type V fimA made a cluster, but the branch length between type IV and type V (0.466) was longer than those between type I and type II (0.231) and type I and type III (0.270) (Fig. 2).

View larger version (99K): [in this window] [in a new window]   FIG. 1.   Comparison of predicted amino acid sequences for FimAs encoded by the fimA genes of various P. gingivalis strains and type V fimA gene. Amino acid identities are shown by asterisks. Hyphens are used to indicate the positions of gaps in the multiple alignment. The putative signal peptides are underlined. The number of amino acids and the molecular weight of the FimA of each strain are given. The alignment of the deduced amino acid sequences was performed with the CLUSTAL W program of the DNA Data Bank of Japan.

View larger version (15K): [in this window] [in a new window]   FIG. 2.   Evolutionary relationships based on synonymous site variation in the fimA gene of P. gingivalis. The neighbor-joining method was used to construct the phylogenetic tree, using CLUSTAL W (DNA Data Bank of Japan) and TreeView software (http://taxonomy.zoology.gla.ac.uk/rod/treeview.html).

Isolation and characterization of P. gingivalis carrying the type V fimA gene. The distribution of P. gingivalis with type V fimA was examined in plaque samples from 73 periodontitis patients carrying P. gingivalis by PCR using the type V fimA-specific primers and the four previous fimA type-specific primers (6). Type V fimA P. gingivalis was detected, solely or in combination with other types, in 16.4% of the samples. The total incidence of type V fimA was higher than that of type I but was almost the same as that of type IV (Table 1).

View larger version (56K): [in this window] [in a new window]   FIG. 3.   Southern blot analyses of P. gingivalis specific genes. Genomic DNA was isolated from each strain; digested with EcoRI, BamHI, and HindIII; and then separated on a 1% agarose gel. After blotting to a nylon membrane, P. gingivalis specific genes were probed with 32P-labeled fimA, sod, and kgp gene fragments from strain ATCC 33277. Lanes: 1, ATCC 33277; 2, HW24D1; 3, 6/26; 4, HG564; 5, HNA-99.

View larger version (43K): [in this window] [in a new window]   FIG. 4.   Western blot analyses of the FimAs in five type-representative strains of P. gingivalis. Whole-cell lysates of P. gingivalis (2 × 107 cells) were separated by SDS-12% polyacrylamide gel electrophoresis. After electrophoresis, gels were transferred to polyvinylidene difluoride membranes and FimA was detected with antibodies to rFimA (381, type I fimA [A], and HG564, type IV fimA [B]). Lanes: M, prestained protein marker; 1, P. gingivalis ATCC 33277; 2, P. gingivalis HW24D1; 3, P. gingivalis 6/26; 4, P. gingivalis HG564; 5, P. gingivalis HNA-99. Arrows indicate FimA.

Binding of P. gingivalis cells to salivary components. To determine whether the variation of FimA types could be related to functional ability, levels of binding of P. gingivalis organisms to HA beads coated with whole saliva were compared. As shown in Fig. 5, P. gingivalis strains ATCC 33277 (type I) and HW24D1 (type II) strongly bound to sHA beads, while strains 6/26 (type III), HG564 (type IV), and HNA-99 (type V) showed weak binding capabilities.

View larger version (27K): [in this window] [in a new window]   FIG. 5.   Binding of P. gingivalis representing the five different fimA types to HA and sHA beads. Three milligrams of HA beads equilibrated with buffered KCl or clarified whole human saliva was added to a siliconized borosilicate tube and incubated with different numbers of the 3H-labeled P. gingivalis (106 to 108) cells in a total volume of 300 µl with a gentle, oscillating motion for 1 h at room temperature. The mixture was layered on 100% Percoll to separate unbound cells from the bead-bound cells. After washing, radioactivity of the bead-bound cells was quantitated with a liquid scintillation counter. The results are shown as the mean values of triplicate samples from three individual experiments. One-way analysis of variance and the Tukey-Kramer test were used for the comparison of the binding abilities of P. gingivalis cells (*, P < 0.0001).

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Since fimbriae appear to be important for attachment and invasion by P. gingivalis, characterization of fimbriae and elucidation of the attachment mechanisms have been extensively attempted (17). Here, we have successfully cloned a type V fimA gene, a new genotype of the fimbrillin gene of P. gingivalis, and have isolated the organism carrying this gene. P. gingivalis FimAs were classified into four types on the basis of the first 20 N-terminal amino acids (20) and variations in the nucleotide sequences of the fimA gene (14). As shown in Results, the multiple alignment analysis showed that our cloned gene fragment has sequence homology of 49 to 52% with other types of fimA genes. The highly homologous sequence strongly suggests that it is a new type of P. gingivalis fimA gene, and we designated it type V fimA.

P. gingivalis possessing the type V fimA gene was detected with a higher frequency (total, 16.4%) than were type I fimA organisms (13.7%) in the present cohort. This finding suggest that the present type of P. gingivalis is involved in the etiology of periodontitis. A wide variety of studies suggested that P. gingivalis fimbriae are major virulence factors and are possible candidates for use in a vaccine (17). Thus, we attempted further characterization of the isolate in this study.

Interestingly, analyzing the promoter region of fimA shows that the type V fimA gene shares the identical 5' upstream (promoter region) and 3' downstream regions of the ORF with the other types of fimA genes (data not shown). This promoter region is essential for the expression of FimA (34), and the gene expression is tightly regulated by environmental conditions (4, 33); these observations also support the view that it is a new type of fimA gene in P. gingivalis. In addition, the phylogenetic tree revealed that the genetic distances of the fimA gene among P. gingivalis strains were less than 0.5. Boyd et al. (9) reported that the fimA gene of Salmonella enterica was hypervariable among natural isolates and that the genetic distance between subspecies IV and VII was 0.87, still suggesting a common ancestor. In this context, the genetic distance of 0.5 strongly suggests that the fimA gene clusters of P. gingivalis strains were derived from a common ancestor, even though the branch lengths between the tested strains are different (Fig. 2).

Southern blot analysis revealed that the hybridization patterns resembled each other for all strains when probed with the type I fimA gene, while the hybridization intensity was weak with strain HG564 (type IV) and only faint bands were observed for strain HNA-99 (type V) (Fig. 3). Loos and Dyer (21) used probe 1 (855-bp fragment of the internal fimA₃₈₁ coding sequence) and probe 2 (2.5-kb fragment including the ORF as well as about 800 bp flanking each side of the ORF) for restriction fragment length polymorphism analyses. The results showed that probe 1 hybridized only rarely with the fimA gene of strains W50, W12, 9-14K-1, AJW-1, and HG564, while probe 2 hybridized with all samples. Our fimA probe included, in addition to the fimA sequence, a short-ended 200-bp flanking sequence, which could have been the cause for the weak hybridization with the samples of strains HG564 and HNA-99. On the other hand, other P. gingivalis-specific genes, sod and kgp, could hybridize with all the test samples with similar signal intensities (Fig. 3). These results suggest that the fimA diversity is most likely generated through mutation and genetic exchange within the ORF but not in the promoter region of the fimA gene. The process of antigenic variation through gene cassette recombination has been shown for the pathogenic Neisseria species (20); however, multiple fimA alleles within a single strain were not observed for P. gingivalis strains (Fig. 3). Furthermore, the phase variation of type 1 fimbriation in E. coli is controlled by the site-specific DNA inversion of recombinases (8), but the site-specific inversion was not found in phase variation of fimA gene expression (34). These observations also supported the concept that the genetic variation of fimA has occurred within the ORF in P. gingivalis; however, the mechanism of genetic exchange of the fimA gene is still unknown. Further studies are needed for analysis of the genetic diversity of the fimA gene of P. gingivalis.

Western blot analysis using antisera against rFimA of strains 381 and HG564 indicated that the immunoreactivity of type V FimA was significantly different from those of other strains. It should be noted that a 41-kDa protein from strain HNA-99 was very faintly reactive with the anti-HG564 rFimA antibody. Our previous study revealed that the fimA gene of HG564 was considerably different from that of 381 and that the type IV rFimA did not clearly react with anti-rFimA (type I) antibody (14). These results indicate that the antigenic variation of fimbriae of P. gingivalis may depend on some specific epitope of FimA. It was reported that the immune response against FimA was enhanced by immunization with the synthetic peptide FP381(201-221) in guinea pigs (26) and that the synthetic peptide PgF-P8 could induce a protective immune response in a chamber infection model using mice (12). These antigenic epitopes are conserved among type I, II, and III fimA strains but not in the type IV and V organisms. The differences in immunoreactivity may influence the pathogenicity of P. gingivalis strains. In addition, serum antibody responses against type V FimA in periodontitis patients are not yet known. The purification of type V FimA and the antibody responses to type V FimA are now under investigation in our laboratory.

The activity of P. gingivalis HNA-99 binding to salivary proteins was almost 50% of that of P. gingivalis ATCC 33277 (Fig. 5). The site of active binding of P. gingivalis fimbriae to salivary proteins has been examined by using whole saliva-coated HA beads (3). The C-terminal region between amino acids 226 and 337 of FimA of strain 381 was essential for the binding of the organism to salivary components (20), and synthetic peptide FimA (amino acids 266 to 286) could inhibit the P. gingivalis whole-cell binding to whole saliva (23). These regions are conserved between type I and type II FimA, but only 10 amino acids are conserved in type IV and type V FimAs. The differences in amino acid compositions of the C-terminal region may reflect the ability of FimA to bind salivary components.

Taking all data into consideration, we conclude that the type V fimA is a new type of P. gingivalis fimA gene. However, the immunoreactivity of type V FimA is significantly different from those of other known types of FimA. Since the fimA genotyping assay enables us to differentiate disease-associated and nonassociated clones of P. gingivalis, further studies are needed to establish the association of type V P. gingivalis with periodontal diseases.