Treponemas. The following treponemas were obtained from the American Type Culture Collection (ATCC) unless otherwise noted: T. denticola ATCC 35405, T. denticola ATCC 33521, T. denticola ATCC 35404, T. denticola GM-1 (gift from Denée Thomas, University of Texas, San Antonio), T. phagedenis ATCC 51274, T. vincentii ATCC 35580 and ATCC 700013, T. pectinovorum ATCC 33768, T. socranskii subsp. buccale (ATCC 35534), T. socranskii subsp. paredis (ATCC 35535), T. socranskii subsp. socranskii (ATCC 35536), T. maltophilum ATCC 51939 (gift from Chris Wyss, University of Zurich, Zurich, Switzerland), T. medium G7201 (gift from Toshihiko Umemoto, Asahi University, Asahi, Japan), T. pallidum (D. Thomas), and cultivable PROS isolates OMZ-802, OMZ-804, and OMZ-805 (C. Wyss).

All treponemas were maintained in OMIZ-P4 broth (unpublished formula from Chris Wyss), which was a modification of OMIZ-WI (17). OMIZ-P4 differs from OMIZ-WI as follows: deletion of lecithin, 1,4-dihydroxy-2-naphthoic acid, n-acetyl muramic acid, and l-valyl-l-lysine HCl and addition of 1,536 mg of glutathione, 50 mg of pyridoxal HCl, 250 mg of thiamine pyrophosphate, 5 mg of spermidine, 1.4 mg of adenine, 1 mg of rifampin, and 100 mg of fosfomycin per liter, as well as 0.05% (vol/vol) yeast extract solution (Gibco BRL, Grand Island, N.Y.), and 1% (vol/vol) heat-inactivated human serum (Sigma Chemical Co., St. Louis, Mo.). We further modified the OMIZ-P4 medium by eliminating the following trace metals and carbohydrates: MgSO₄, CuSO₄, MnSO₄, ZnSO₄, NiSO₄, SnCl₂, NaCO₃, d-glucose, d-fructose, d-maltose, d-mannitol, glucuronic acid, and galacturonic acid. The modified OMIZ-P4 medium is referred to as mP4. Maintenance cultures were incubated at 35°C in anaerobic GasPak jars.

Isolation of cultivable PROS. OHSU 242-9 and 242-10 Treponema strains were isolated as follows. Subgingival plaque taken from sites of periodontitis was suspended in normal saline and combined with an equal volume of 2× mP4 broth. Suspensions were enriched for spirochetes by overnight incubation at 35°C in anaerobic GasPak jars. Enrichment cultures were viewed by ×400 dark-field microscopy to estimate spirochete numbers, and 10-fold dilutions were made to produce spirochete concentrations between 10 and 1,000 cells per ml. Pour plates were made by combining 1 ml of each dilution with 30 ml of molten mP4 agar supplemented with 1.5% SeaPlaque agar (FMC Bioproducts, Rockland, Maine) in disposable petri dishes. After 7 days of incubation at 35°C in anaerobic GasPak jars, discrete colonies were picked and suspended in 0.5 ml of mP4. PROS were defined by reactivity with H9-2 according to an established protocol (10). This process was repeated with H9-2-positive suspensions through second and third rounds of pour plates to isolate pure cultures.

Western blots. Treponemas were washed twice in normal saline. Washed cell pellets were resuspended in 2× treatment buffer (0.125 M Tris-HCl [pH 6.8], 4% sodium dodecyl sulfate, 20% glycerol, 10% 2-mercaptoethanol). Protein concentrations were determined by the bicinchoninic acid protein assay (Pierce, Rockford, Ill.) and adjusted to 2 μg/μl. Protein was extracted by boiling at 70°C for 5 min followed by sonication on ice for 4 min. Crude extracts were then centrifuged at 10,000 × g for 10 min, and supernatants were electrophoresed on 4 to 20% Tris-glycine gradient gels (Novex Experimental Technology, San Diego, Calif.) at 150 V for 70 min. Experimental samples were routinely placed in alternate lanes, leaving one lane blank between each sample. Prestained molecular mass markers (6.5 to 200 kDa; Bio-Rad Laboratories, Hercules, Calif.) were introduced into lanes 1, 11, 12, and 22 in order to facilitate the interpretation of experimental and control bands.

Proteins were then electroblotted onto nitrocellulose membranes, and membranes were blocked overnight with 3% nonfat milk in Tris-buffered saline (TBS). Separate immunoblots were incubated with one of three T. pallidum-specific MAbs, H9-1, H9-2, or F-5 (Shiela Lukehart, Seattle, Wash.), diluted 1:10 in 1% nonfat milk–TBS, for 2.5 h at 4°C. Blots were washed in TBS and then incubated for 2.5 h at 4°C with alkaline phosphatase-labeled goat anti-mouse immunoglobulin G (IgG) (Sigma) at a dilution of 1:1,000 in blocking solution. Alkaline phosphatase color development reagents (5-bromo-4-chloro-3-indolylphosphate p-toluidine salt and p-nitroblue tetrazolium chloride; Bio-Rad) were used for detection.

Media modulation and cell staining. Log-phase cultures of T. vincentii, OMZ-802, and OMZ-804 were established in both modified new oral spirochete (NOS) broth (ATCC medium 1494) and mP4. Cells were subcultured three times at 48-h intervals. In another series of experiments, cells were subcultured from NOS to mP4 or from mP4 to NOS and after 48 h were subcultured back to the original medium. Cells from each subculture were washed twice in normal saline, spotted onto glass slides, and air dried. Whole-cell reactivities with H9-1, H9-2, and F-5 were determined by a biotin-amplified microscopic immunocytochemical assay (10). A pool of two MAbs, specific for T. socranskii subsp. buccale and T. socranskii subsp. socranskii, was used as the control in each experiment.

DNA isolation. One milliliter of the log-phase culture was sedimented, and cells were washed three times in sterile normal saline. Each pellet was resuspended in 25 to 50 μl of distilled water and then lysed by three cycles of exposure to dry ice for 10 min followed by 95°C for 10 min. DNA was isolated by centrifugation at 10,000 × g for 5 min, and supernatant was transferred and stored at −20°C until analyzed.

16S RFLP. PCR with a 100-μl reaction mixture was carried out in a buffer containing 2.5 mM MgCl₂, 50 mM KCl, 10 mM Tris-HCl (pH 8.3; Boehringer Mannheim, Indianapolis, Ind.), 50 μM (each) deoxynucleoside triphosphates (Sigma), 10 pM (each) 1492R and 8F primers (4, 15), and 2.5 U of Taq polymerase (Boehringer Mannheim), and the mixture was overlaid with 1 drop of mineral oil (Sigma). Four microliters of DNA was added as the template. After initial denaturation at 97°C for 1 min, 26 cycles were performed as follows: denaturation at 97°C for 45 s, annealing at 55°C for 45 s, and extension at 72°C for 1 min. A final extension was conducted at 72°C for 4 min. Amplified product was stored at 4°C until analyzed on a 1.0% agarose gel in 0.5% TBE buffer (89 mM Tris, 89 mM borate, and 2 mM EDTA). Three reaction mixtures for each template were combined, precipitated with 0.5 volume of ammonium acetate and 2 volumes of isopropanol, centrifuged for 20 min at 10,000 × g, washed twice in 70% ethanol, dried, and resuspended in 30 μl of 0.1× TE buffer (10 mM Tris and 1 mM EDTA). Ten microliters of each PCR product was digested overnight at 37°C with 5 U of HinfI, HaeIII, or RsaI (Gibco BRL); dried down completely; resuspended in 4 μl of loading dye; and electrophoresed in 2% agarose and 0.5× TBE at 4 V/cm until the leading dye front ran to the 15-cm mark. Gels were stained with ethidium bromide, destained, and photographed.

Species-specific PCR. PCR was performed in 25-μl reaction volumes as previously described (16). In brief, PCR was performed as described above for RFLP except that the concentration of MgCl₂ was reduced to 1.5 mM and the 8F universal primer was used with species-specific reverse primers.

AP-PCR analysis. AP-PCR was performed in 25-μl reaction volumes as described previously (6), by using 10 pM random primer OPA-2 (AGTCAGCCAC; Operon Technologies, Alameda, Calif.), 7 mM MgCl₂, 2 μl of target DNA, and 50 μM (each) deoxynucleoside triphosphates in a buffer containing 50 mM KCl, 10 mM Tris-HCl (pH 8.3; Boehringer Mannheim), and 1.25 U of Taq polymerase. Amplifications were performed with the following parameters: 1 min of initial denaturation at 96°C followed by 45 cycles of 30 s at 95°C, 30 s at 36°C, and 100 s at 72°C and a final 3-min elongation at 72°C. Twenty-four microliters was loaded on a 1.5% agarose gel, stained, and photographed as described above.

MAb reactivity with whole cells grown in NOS and mP4. Table 1 shows that cultivable PROS strains OMZ-802 and OMZ-804 and T. vincentii ATCC 35580 varied in the expression of determinants that cross-reacted with T. pallidum MAb H9-2. T. vincentii grown in NOS did not cross-react with H9-2, and a 48-h exposure to mP4 was not sufficient to promote a cross-reaction. On the other hand, T. vincentii grown in mP4 did cross-react with H9-2, and cells retained their reactivity after subculture in NOS for 48 h. OMZ-802 performed in a manner similar to that of T. vincentii, except that it appeared to be more responsive to brief changes in medium. OMZ-802 grown in NOS acquired cross-reactivity after 48 h in mP4, and cells grown in mP4 lost cross-reactivity after 48 h in NOS. In contrast to both T. vincentii and OMZ-802, OMZ-804 expressed cross-reactivity with H9-2 when grown in either NOS or mP4.

TABLE 1 Reactivity of T. pallidum MAb H9-2 with PROS and T. vincentii whole cells during sequential subculture in mP4, NOS, or combinations of mP4 andNOS

Table 2 shows that reactivity of oral treponemas with T. pallidum MAbs H9-1 and F5 also varied according to growth conditions and that strain differences were similar to those described for H9-2. Cultivable PROS and T. vincentii grown in mP4 demonstrated strong cross-reactivity with all three MAbs, while only OMZ-804 retained cross-reactivity with H9-2 and F5 when grown in NOS. Strong cross-reactions were characterized by confluent staining of all, or nearly all, cells grown in mP4. Weak cross-reactions were observed in some experiments with cells grown in NOS. Weak reactions were characterized by partial staining of some, but not the majority, of the cells in representative microscopic fields.

TABLE 2 Reactivity of whole treponema cells with T. pallidum MAbs and control T. socranskii MAbs following continuous culture in either mP4 or NOSmedium

Table 2 also shows that, in addition to T. vincentii, T. maltophilum, T. pectinovorum, and two strains of T. denticola, grown in mP4, also cross-reacted with H9-1, H9-2, and F5. However, with the exception of the weak reactivity of H9-2 with T. maltophilum, none of these treponemas grown in NOS cross-reacted with T. pallidum MAbs. In almost every instance, Treponema cells that cross-reacted with H9-2 also reacted with H9-1 and F5. The one exception was T. medium. When grown in mP4 it displayed weak cross-reactivity with H9-2 and no cross-reactivity with either H9-1 or F5. T. medium grown in NOS did not cross-react with any T. pallidum MAb.

Other treponemas were consistently nonreactive in both NOS and mP4 media. T. phagedenis, T. socranskii, and T. denticola strains GM-1, ATCC 35404, ATCC 35405, and ATCC 33521 never reacted with H9-1, H9-2, or F5.

The T. socranskii MAb pool control never reacted with treponemas other than appropriate T. socranskii target cells.

Cross-reactivity of treponema proteins with T. pallidum MAbs in immunoblots. All cultivable treponemas were grown in mP4. Figure 1 shows that H9-2 reacted with reduced proteins from T. pallidum, T. vincentii ATCC 35580 and ATCC 700013, and PROS strains OHSU 242-9, OHSU 242-10, OMZ-802, OMZ-804, and OMZ-805. Cross-reactive proteins from oral treponemas were slightly larger than the 37-kDa protein of T. pallidum that reacts with H9-2. H9-2 did not react with proteins from any other cultivable treponemas. H9-1 and F5 reacted with T. pallidum but not with proteins from any other treponema, including those strains displaying cross-reactivity between whole cells and MAbs.

FIG. 1 Representative immunoblot for treponema following incubation with T. pallidum MAb H9-2. Lanes 1, 11, 12, and 22, molecular mass markers (200, 116, 97, 66, 45, 31, 21.5, and 14.4 kDa; Bio-Rad); lanes 2 and 13, T. pallidum band at 37 kDa (all other bands (more ...)

Species-specific PCR. PCR studies are summarized in Table 3. T. vincentii-specific PCR generated a product of approximately 200 bp with DNA templates from OHSU 242-9, OHSU 242-10, and OMZ-802. T. vincentii-specific PCR did not amplify DNA from OMZ-804 or OMZ-805. OMZ-805 was amplified by T. medium-specific PCR, producing an amplicon of approximately 200 bp, identical in size to the product obtained with the T. medium template. OMZ-804 was not amplified by T. vincentii, T. medium, or T. pallidum species-specific primers. T. pallidum-specific PCR did not amplify DNA from any oral treponema, including cultivable PROS.

TABLE 3 Amplification of cultivable PROS isolates by species-specificPCRa

RFLP. Figure 2 shows RFLP patterns obtained with RsaI enzyme digests of 16S rDNA PCR products. OHSU 242-9, OHSU 242-10, and OMZ-802 had RFLP patterns identical to that of T. vincentii. The pattern of OMZ-804 was identical to that of T. medium, and the pattern of OMZ-805 was identical to that of T. vincentii.

FIG. 2 RFLP patterns obtained with RsaI enzyme digests of 16S rDNA PCR products. Lanes: 1 and 12, 1-kb DNA marker (Gibco BRL); 2, DNA from T. vincentii ATCC 35580; 3, DNA from T. vincentii ATCC 700013; 4, DNA from OHSU 242-9; 5, DNA from OHSU 242-10; 6, DNA (more ...)

OHSU 242-9 and 242-10, OMZ-802, -804, and -805, T. medium, and T. vincentii were found to have identical RFLP patterns when HinfI and HaeIII were used.

AP-PCR. Figure 3 shows that T. vincentii ATCC 35580 and ATCC 700013 and PROS isolates OHSU 242, OMZ-802, OMZ-804, and OMZ-805 all have unique AP-PCR profiles with random primer OPA-2. OHSU 242-9 and OHSU 242-10 had identical AP-PCR profiles.

FIG. 3 AP-PCR patterns obtained from whole genomic DNA with random primer OPA-2. Lanes: 1 and 10, 1-kb DNA marker (Gibco BRL); 2, DNA from T. vincentii ATCC 35580; 3, DNA from T. vincentii ATCC 700013; 4, DNA from OHSU 242-9; 5, DNA from OHSU 242-10; 6, DNA (more ...)

The present work helps to reconcile previous contradictory reports (1, 3, 8–10, 12) by confirming that oral treponemas grown in NOS medium did not react with T. pallidum MAbs while cells grown in mP4 media did cross-react with H9-1, H9-2, and F5 MAbs. Furthermore, cross-reactivity was modulated when treponemas were subcultured from one medium to another. An unexpected finding was that phenotypic modulation and T. pallidum MAb cross-reactivity extended to oral treponemas other than cultivable PROS. T. denticola, T. maltophilum, T. pectinovorum, and T. vincentii also acquired cross-reactivity with T. pallidum MAbs after growth in mP4. It is unlikely that cross-reactivity was due to an artifact created by mP4 because T. denticola GM-1, ATCC 35404, ATCC 35405, and ATCC 33521, three subspecies of T. socranskii, and T. phagedenis failed to react with T. pallidum MAbs after growth in mP4.

While whole treponema cells cross-reacted with three T. pallidum MAbs, only H9-2 reacted in Western blots with reduced proteins from cultivable PROS and T. vincentii. We confirmed that the cross-reactive protein identified by H9-2 is slightly larger than the 37-kDa protein from T. pallidum that bears the H9-2 determinant (3). The nature of cross-reactive determinants identified on whole cells by H9-1 and F5 MAbs has not been determined, but unpublished observations indicate that they are preserved by extraction under nonreducing conditions (7). Cross-reactive determinants expressed by PROS in subjects with necrotizing ulcerative gingivitis may have been responsible for serum antibodies that reacted with proteins from T. pallidum (12). However, in a recent study (14) subjects with no detectable oral PROS were just as likely to have IgA, IgG, or IgM to 15-, 37-, or 47-kDa proteins from T. pallidum as were subjects with PROS. The presence of cross-reactive serum antibodies from subjects with no detectable oral spirochetes (14) suggests that cross-reactive determinants may be expressed by moieties other than oral treponemas.

Choi et al. (3) suggested that PROS were related to T. vincentii based upon DNA sequence homology, and the present investigation supports this close relationship for three of five PROS strains. OHSU 242-9, OHSU 242-10, and OMZ-802 were amplified by T. vincentii-specific PCR, and their 16S rDNA RFLP patterns were identical to that of T. vincentii. However, not all PROS isolates, defined by H9-2 cross-reactivity, were so easily aligned with T. vincentii. For example, OMZ-805 displayed MAb phenotypes like those of T. vincentii but was amplified by T. medium-specific PCR, not by T. vincentii-specific PCR. Furthermore, the RFLP pattern for OMZ-805 appeared to contain bands common to both T. medium and T. vincentii. OMZ-804 was also different. It was the only strain that reacted with T. pallidum MAbs in both NOS and mP4 media; it was not amplified by species-specific PCR for T. medium or T. vincentii, but the RFLP profile for OMZ-804 was identical to that for T. medium. Thus, the available evidence suggests that OHSU 242-9, OHSU 242-10, and OMZ-802 may be most like T. vincentii; that OMZ-805 is related to both T. medium and T. vincentii; and that OMZ-804 is the most different, sharing RFLP patterns only with T. medium. Based upon these provisional relationships, it may be that PROS and T. medium belong to the group I treponemas described by Choi et al. (2).

PCR amplification and RFLP patterns for OHSU 242-9, OHSU 242-10, and OMZ-802 suggest that they may be closely related to T. vincentii. However, AP-PCR revealed not only that each PROS isolate was unique (OHSU 242-9 and OHSU 242-10 had identical profiles but differed from the others) but also that T. vincentii ATCC 35580 and ATCC 700013 were distinct. Thus, these three PROS isolates are not clonotypes of T. vincentii. OHSU 242-9 and OHSU 242-10 were not identical because they differed in morphology and invasive potential (13a).

Finally, these experiments cast aside any thoughts that PROS are related to T. pallidum. No PROS strain was amplified by T. pallidum-specific PCR, and their RFLP profiles were different (data not shown). Cross-reactivity with T. pallidum MAbs H9-1 and F5 appeared to be a consequence of some undefined effect of growth conditions, which created changes in secondary or tertiary protein structure, a process that was reversed when NOS was used for growth. OMZ-804 was the exception to this rule because it cross-reacted with T. pallidum MAbs in both mP4 and NOS media. Reactivity with H9-2 was preserved after reduction in buffer for immunoblots, but the protein bearing the determinant was larger than the 37-kDa protein from T. pallidum. Perhaps most telling was the observation that four oral Treponema species, T. denticola, T. maltophilum, T. pectinovorum, and T. vincentii, cross-reacted with T. pallidum MAbs if they were grown in mP4 medium. Thus, reactivity with H9-1 and F5 should be considered an artifact related to growth conditions, and H9-2 should not be regarded as specific for T. pallidum unless close attention is paid to the size of reactive proteins in Western blots.

In conclusion, immunologic and genetic techniques disclosed considerable diversity among five strains of cultivable PROS, including those that are very similar to T. vincentii, some that appear to be intermediate between T. medium and T. vincentii, and those that are somewhat more closely related to T. medium. The implied relatedness between T. pallidum MAb-reactive oral spirochetes and T. pallidum is not supported by molecular data. It is proposed that the PROS nomenclature be dropped.

This work was supported in part by the Oregon Health Sciences Foundation, Portland, Oreg.