INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Alterations in the growth environment of a bacterium, such as changes in carbon, carbohydrate, and nitrogen source, growth temperature, and iron concentration, have been shown to alter the expression of a variety of virulence factors (13, 15, 32, 48). These environmental alterations, or stress responses, also result from placing the bacterium in adverse growth environments, such as conditions of starvation, nutrient limitation, pH shift, heat shock, or change in oxygen tension. With few exceptions, all bacteria, whether living free or associated with a host, such as the periodontopathogenic bacteria, must be able to respond to these signals or cues. Adjustment to these environmental changes permits them to survive and multiply in this changing environment (11, 13, 34-36). Oral pathogens are an excellent example of bacteria that survive, grow, and multiply in a complex and rapidly changing environment. With potentially hundreds of different bacterial species occupying a periodontal niche, producing many metabolic end products and associated physicochemical changes, these bacteria are highly adapted for this niche. We have noted that hemin limitation increases the expression of new outer membrane proteins (2, 27) by Porphyromonas gingivalis. Additionally, various studies have shown that hemin limitation has a significant effect on P. gingivalis virulence in mice (22, 31, 33). Moreover, we have noted that environmental growth conditions alter the virulence potential (24) and in vitro S-layer expression of Campylobacter rectus (38).

Ultimately, to understand the disease caused by any pathogen, its virulence potential in an in vivo environment must be examined (20, 45). In vivo studies provide a means to address the role of the host and selected host-associated environmental factors on specific manifestations of disease progression. Recent work in our laboratory has shown that the murine abscess model is capable of establishing a role for the trypsin-like protease activity in P. gingivalis pathogenesis (25). Although we have some knowledge of the in vitro elaboration of putative treponeme virulence factors (i.e., hemolysin, trypsin-like enzyme, major outer sheath proteins, and lipopolysaccharide [LPS]-like molecules), we still have very little understanding of their functional role in vivo or regulation by environmental factors such as iron (28, 41, 52). The oral treponemes Treponema denticola, Treponema socranskii, Treponema pectinovorum, and Treponema vincentii have been implicated as etiological agents of severe periodontal disease in adults (19, 30, 37, 44, 46). Among the oral treponemes, T. denticola is the predominant one identified within the gingival crevice and subgingival ecology of the developing periodontal pocket and has been the most extensively studied. In vitro studies have revealed that oral treponemes elaborate a variety of proteolytic enzymes, hemolysins, esterases, collagenase, fibrinolytic enzymes, iminopeptidases, phospholipase C, hyaluronic acid, and chondroitin sulfate-degrading enzymes. The functional significance of these virulence factors has not been clearly elucidated in vivo. There are no data documenting the role of iron as a modulator of oral treponeme virulence. The murine abscess model provides the wherewithal to study the role of putative virulence factors of the oral treponemes in tissue destruction. Previously, we employed this model to explore variations in virulence among the oral treponemes, as well as the contribution of a trypsin-like protease to the virulence of T. denticola and T. pectinovorum strains (26).

The objectives of this study were to determine the role of iron, cysteine (as a sulfhydryl reducing agent), local inflammation, and endotoxin in the pathogenesis of T. denticola and T. pectinovorum in a murine abscess model. The hypotheses tested included the following. (i) Alterations in iron availability, in vitro or in vivo, modify the virulence of T. denticola and T. pectinovorum. (ii) Cysteine, which significantly enhances in vitro growth of the treponemes, increases the virulence of the species. (iii) Existing localized inflammation enhances treponeme virulence. (iv) Both T. denticola and T. pectinovorum demonstrate toxic activities in mice which had been sensitized for evaluation of endotoxin-induced lethality.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Bacterial strains. The Treponema spp. used and their sources are given in Table 1. T. denticola strains were grown in GM-1 broth (1) in a Coy anaerobic chamber in an atmosphere of 85% N₂, 5% CO₂, and 10% H₂ for 72 h at 37°C. T. pectinovorum strains were grown in GM-1 broth supplemented with 0.3% pectin (Sigma) for 24 to 72 h (49, 50). All manipulations were carried out under anaerobic conditions to ensure maximum cell viability. Culture purity was determined by dark-field and phase-contrast microscopy. Viability was estimated by the degree of motility and the presence or absence of spherical bodies of the treponemes. Growth in liquid media was estimated at an optical density of 660 nm. Log-phase cultures were harvested by centrifugation (9,000 × g for 10 min), and pellets were resuspended in the same medium under anaerobic conditions. An aliquot of the culture was removed from the chamber, and 10-fold dilutions were made for estimating total counts with a Petroff-Hausser bacterial counting chamber. After allowing sufficient time for treponemes to settle in the counting chamber, the cells were enumerated, with spherical bodies excluded.

Mice. ICR (Harlan Sprague-Dawley Inc., Indianapolis, Ind.) female mice, aged 8 to 12 weeks old, were utilized. The animals were housed in isolator cages in an American Association for Accreditation of Laboratory Animal Care-accredited animal facility at the University of Texas Health Science Center at San Antonio and were provided with autoclaved TEKLAD chow (Harlan Sprague-Dawley Inc., Madison, Wis.) and water ad libitum. BALB/c mice were used for testing the lethality of the T. denticola GM-1 strain, and ICR mice were used in all other experiments.

Murine virulence model. The murine abscess model (23, 24, 26) was used to examine the virulence capacity of the oral treponeme strains. For determination of virulence and abscess-forming capability, appropriate dilutions of the treponemes were made under anaerobic conditions and mixed with the virulence-modulating agents. Mice were injected subcutaneously (s.c.) or intraperitoneally (i.p.) within 15 to 30 min of sample preparation. After challenge, the animals were monitored daily for symptoms of infection. Progression of lesion formation for up to 3 weeks and virulence were scored on the basis of (i) onset and characteristic of lesion, (ii) size of localized abscess and/or necrotic skin lesion, (iii) size of spreading exudative hemorrhagic lesion, and (iv) death. Subcutaneous abscess and necrotic skin lesion sizes (length and width) were measured daily for 15 days with a caliper gauge, and the area was determined and expressed in square millimeters. The maximum abscess size was noted for each animal, and the mean per group was calculated. Three experimental models were used.

(i) Bacterial growth modulation. The cysteine concentration was altered such that T. denticola and T. pectinovorum were grown in GM-1 medium (1, 49, 50) supplemented with 6, 18, or 36 mM cysteine. Logarithmic-phase broth cultures were harvested by centrifugation and resuspended in GM-1 medium (with or without cysteine) to a final concentration of 2.5 × 10¹¹ treponemes per ml. GM-1 medium supplemented with 6, 18, or 36 mM cysteine alone was injected into groups of animals to serve as controls.

(ii) Bacterial environment modulation in vivo. Alterations in T. denticola virulence were assessed by the extent of abscess formation following coinoculation of mice with treponemes and Cytodex 1 dextran microcarrier beads essentially as described by Ford et al. (16). The beads induce a localized inflammatory response and can increase bacterial pathogenicity. Cytodex 1 microcarrier beads (Sigma) were suspended in phosphate-buffered saline (50 ml/g) overnight and autoclaved at 120°C for 10 min. The sterile supernatant fluid was decanted from the sedimented microcarrier beads, equal volumes of packed microcarrier beads and an appropriate concentration of T. denticola were mixed, and 0.2 ml of the mixture was injected s.c. into the posterior dorsolateral surface of each mouse for determination of abscess formation. Sterile packed Cytodex 1 beads alone were injected s.c. into a group of five animals as a negative control.

(iii) Host environment modulation. To evaluate the effect of iron pretreatment of mice on the virulence of T. denticola and T. pectinovorum, mice were administered iron dextran (ferric hydroxide dextran complex; Sigma) i.p. at 5 mg per animal 3 h prior to (41) treponeme challenge. Alternatively, administration of phenylhydrazine (Sigma) to mice elicits a rapid lysis of circulating erythrocytes with a resulting substantial increase in the levels of free hemoglobin in plasma (21). This host alteration lasts for approximately 4 days, at which time the blood levels of hemoglobin return to normal. Phenylhydrazine (2 mg/animal) was administered i.p. 3 h prior to T. denticola or T. pectinovorum challenge. Groups of animals without iron or phenylhydrazine pretreatment were also infected with the treponemes to serve as untreated controls; mice injected with iron dextran or phenylhydrazine alone served as treatment controls.

Statistical analyses. Statistical differences in lesion size between various groups were determined by use of a Wilcoxon-Mann-Whitney U test (Minitab, State College, Pa.).

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

In vitro environmental modulation of virulence. The ability of cysteine to enhance the virulence of T. denticola and T. pectinovorum, as determined by the production of soft tissue lesions in mice following s.c. challenge, was evaluated (Fig. 1). We have previously described the characteristics of the induction, progression, and gross and histologic characteristics of these lesions (26). T. denticola and T. pectinovorum grown in 18 mM cysteine induced significantly (P < 0.01) larger local abscesses than bacteria grown in 6 mM cysteine when a challenge dose of 10¹⁰ was used. Also, T. denticola grown in 36 mM cysteine induced significantly (P < 0.01) larger abscesses than bacteria grown in 6 mM cysteine, although the lesion size was similar to that of bacteria grown in 18 mM cysteine (data not shown). Similar onsets and similar lesion characteristics were elicited by both treponemes, irrespective of the availability of cysteine during in vitro cultivation. As we had previously noted, there was no mortality induced by s.c. challenge infection with these treponemes under any of the conditions.

View larger version (39K): [in this window] [in a new window]   FIG. 1.   Effect of growth of T. denticola ATCC 35404 and T. pectinovorum ATCC 33768 in cysteine-containing GM-1 medium on virulence in a murine abscess model. The bars denote the mean abscess size in square millimeters, and the error bars represent 1 standard deviation derived from groups of five mice. NT, not tested. *, significantly different at P < 0.01 when results for treponemes grown in 18 and 6 mM cysteine are compared. The treponeme-free negative control injected with cysteine in GM-1 medium demonstrated neither toxicity nor any lesion at the site of injection (data not shown).

In vivo bacterial environment modulation. Injection of the inflammation-stimulating Cytodex 1 beads into the dorsal surface of mice induced a closed sterile abscess of approximately 80 mm² (Fig. 2). The abscess persisted for approximately 14 days and then resolved. A similar s.c. injection of T. denticola alone at a dose of 10⁹ bacteria/site produced a small abscess of approximately 28 mm². In contrast to the response induced by the Cytodex 1 beads (i.e., a closed sterile abscess), the abscess induced by T. denticola was expressed through the skin and healed more slowly (i.e., 2 to 3 weeks). When 10⁸ or 10⁹ T. denticola bacteria were combined with the Cytodex 1 beads, the abscess formed was significantly larger (P < 0.05) than that resulting from T. denticola injection alone (Fig. 2). Interestingly, while the abscess was larger (110 to 150 mm²) in the T. denticola-Cytodex 1 bead-injected animals than in the animals injected with T. denticola or Cytodex 1 beads alone, the lesion from the combined injection appeared to be the result of an additive virulence outcome rather than a synergistic effect. Nevertheless, this local abscess persisted for more than 3 weeks and did not break through the skin, as was noted with T. denticola alone.

View larger version (46K): [in this window] [in a new window]   FIG. 2.   Effect of sterile Cytodex 1 microcarrier bead-induced abscess on the virulence of T. denticola ATCC 35404. Mice were challenged with 108 or 109 bacteria as described in Materials and Methods. NT, not tested. The bars represent the mean abscess size in square millimeters, and the error bars represent 1 standard deviation derived from groups of five mice.

Host environment modulation. The in vivo administration of iron-containing compounds (e.g., iron dextran, hemin, hemoglobin, or transferrin) has been shown to enhance the virulence of many bacterial pathogens (28, 42, 51, 52), although these studies have generally emphasized microorganisms which cause systemic infections. Mice pretreated with iron dextran and infected with T. denticola (grown in 18 mM cysteine) induced at least 50% larger lesions (P < 0.025) than untreated T. denticola-challenged mice (Fig. 3). No abscess or symptoms of toxicity were observed in control mice that received iron dextran alone. In contrast, iron dextran had no effect on the virulence of infection with T. pectinovorum (grown in 18 mM cysteine) in comparison to that in untreated control animals (Fig. 3).

View larger version (36K): [in this window] [in a new window]   FIG. 3.   Effect of administration of iron dextran (FeDex) to mice on the virulence of T. denticola ATCC 35404 and T. pectinovorum ATCC 33768 cultivated in medium with cysteine at various concentrations (6 or 18 mM). Animals were challenged with 1 × 1010 or 5 × 1010 bacteria as described in Materials and Methods. The bars represent the maximum mean lesion size in square millimeters, and the error bars represent 1 standard deviation derived from groups of five mice. NT, not tested. *, significantly different (P < 0.01) when results of growth in 18 mM cysteine with host iron pretreatment and growth in 18 mM cysteine alone are compared. Mice injected with iron dextran alone demonstrated neither toxicity nor any lesion at the site of injection (data not shown).

View larger version (34K): [in this window] [in a new window]   FIG. 4.   Effect of administration of phenylhydrazine (PHZ) to mice on the virulence of T. denticola ATCC 35404 and T. pectinovorum ATCC 33768 cultivated in medium with cysteine at various concentrations (6 or 18 mM). Animals were challenged with 1 × 1010 or 5 × 1010 bacteria as described in Materials and Methods. The bars represent the maximum mean lesion size in square millimeters, and the error bars represent 1 standard deviation derived from groups of five mice. The stacked bars for PHZ treatment and T. denticola challenge represent the localized abscess (solid bar) and total (spreading necrotic lesion plus local abscess) (horizontally lined bar) lesion sizes in the mice. NT, not tested. *, significantly different (P < 0.01) from results for PHZ treatment of control infected mice. Mice injected with PHZ alone demonstrated neither toxicity nor any lesion at the site of injection except for a brief (<24-h) period of malaise with ruffled hair (data not shown).

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

In vitro studies of environmental regulation of bacterial virulence have provided a number of important observations regarding strategies used by pathogens to respond and survive under changing environmental conditions in vivo. However, recent evidence, including sophisticated in vivo expression technologies (20, 45), have demonstrated the importance of the in vivo environment in delineating significant virulence determinants which contribute to the host-parasite interactions resulting in manifestations of disease. An in vivo expression technology vector (pPGIVET) has been constructed and used to determine the expression of hemagglutinin genes hagB and hagC of P. gingivalis during an infectious process in mice (30). Thus, these genetic approaches to evaluate the virulence mechanisms of oral microorganisms have been initiated, but the capabilities of these approaches are still quite limited. The experiments described in this report examine the oral treponemes T. denticola and T. pectinovorum as members of the putative periodontopathic microbiota in the subgingival sulcus (30, 44, 46). In particular, we used more classic approaches to specifically evaluate the ability of in vitro and in vivo environmental conditions to modify the virulence capacity of oral treponemes by using a murine model (14, 23, 24).

Our initial investigation expanded previous in vitro observations from studies using cysteine (9) or iron levels (3, 4, 7) to modulate the growth characteristics of oral microorganisms. These have been suggested to result in the up- and down-regulation of selected surface and cytoplasmic molecules that aid in the ability of the bacteria to utilize alternative mechanisms for acquisition of critical nutrients and cofactors. We have noted that cysteine can stimulate or enhance enzyme activities of P. gingivalis, although high levels of this sulfhydryl-containing amino acid can actually abrogate the virulence of P. gingivalis (unpublished observation). Cysteine has also been identified as a protease activator, although it had been shown previously to exhibit a minimal effect on a number of T. denticola enzyme functions in vitro (26). However, previous studies in our laboratory have clearly shown that cell-associated hemolytic and hemoxidative activities of T. denticola are cysteine dependent (6). Further studies of these biologic activities described the isolation and characterization of a 46-kDa protein from T. denticola, referred to as cystalysin, with hemolytic activity different from that of complement and classical hemolysin (8). Its preferred substrate compound was cysteine with a sulfur atom adjacent to both a -methylene and an -amino group (9). The cystalysin exhibited its enzymatic activity by degrading cysteine, yielding H₂S and pyruvate, which increased the growth of the bacteria and caused the hemolytic activity (9). In contrast, T. pectinovorum does not express a cystalysin activity. We hypothesized that cysteine would significantly increase the in vivo virulence of T. denticola, via these cystalysin activities. The results showed that, in fact, the virulence of T. denticola was significantly increased by growth in media with elevated levels of cysteine. However, this amino acid also appeared to modify the characteristics of T. pectinovorum and increased its lesion-inducing capacity. Thus, while the T. denticola cystalysin may be a virulence factor affected by the cysteine, it appears that the oral treponemes may have multiple mechanisms for interacting with this molecule in the environment.

Previous studies have clearly shown that environmental iron has a significant regulatory effect on the growth (3) and virulence expression (22) of P. gingivalis. We have noted that cultivation of oral treponemes under iron-limited conditions alters outer membrane protein profiles (7), enzymatic activities (9), and hemolytic activities (8). Recently, we have observed that T. denticola and T. pectinovorum have a minimal capacity to survive under continued iron-depleted conditions in vitro (unpublished observations), in contrast to the ability of P. gingivalis to survive and grow for three to seven passages in media devoid of iron (unpublished data). Thus, we hypothesized that depleting the environment of iron would have a profound effect on virulence expression by the oral treponemes. Although iron is essential for the in vitro growth of oral treponemes and appears to alter putative virulence characteristics, including toxins, cytotoxins, and hemolysins (6), depletion of this molecule during a single in vitro passage had a negligible impact on in vivo virulence. These results suggest that the treponemes may be less affected by the dramatic fluctuations in environmental iron in the gingival sulcus and, as such, are less dependent upon the local availability of iron for maintenance of their virulence capability.

Iron is a necessary factor for the survival and growth of prokaryotes, which is clearly evident in the numerous evolutionary strategies that have been developed by bacteria to acquire iron from the natural environment (51). Various studies have begun to contribute to a clearer understanding of mechanisms by which oral anaerobes acquire iron from the environment (3, 4, 18), particularly with respect to P. gingivalis. We have recently reported preliminary observations concerning the accession of iron in vivo by P. gingivalis (22). Additionally, we have determined that P. gingivalis virulence is enhanced by pretreatment of mice with iron dextran, hemin, and transferrin (unpublished data). This has been suggested to result from increased iron availability and altered regulation of virulence determinants. However, very little is known concerning the characteristics of iron acquisition and utilization by oral treponemes. To this end, we assessed iron utilization by T. denticola and T. pectinovorum in vivo in mice. Specifically, numerous studies have demonstrated that iron loading of mice significantly increases the virulence of systemic pathogens, such as Vibrio vulnificus (21), Escherichia coli (48), and Yersinia enterocolitica (41), although this model has never been used to examine localized, tissue-destructive pathogens. In our study, iron dextran administration significantly enhanced the virulence (abscess size) of T. denticola. Similarly, phenylhydrazine pretreatment of mice, which has been reported to induce hemoglobinemia in mice, significantly enhanced lesion formation by T. denticola. The phenylhydrazine pretreatment also substantially altered the virulence characteristics of T. denticola either by a direct effect of the increased iron on bacterial virulence mechanisms or by altering host resistance and allowing the T. denticola to spread from the infection site. Both iron dextran and phenylhydrazine pretreatments appeared to have less of an effect on the virulence of T. pectinovorum in vivo. One interpretation of these findings is that the alterations were at the level of the microorganism rather than the host, although further studies will be required to validate this concept. Additionally, the minimal effect of host iron changes on T. pectinovorum suggested that this microorganism either requires less iron for in vivo survival and virulence expression or does not express the capacity to utilize either of these host iron sources but acquires its iron by other mechanisms. Again, these options need to be evaluated in future studies.

Extensive inflammation, which enhances delivery of soluble and cellular serum components to the sulcular environment, results in dramatic changes in the milieu of the sulcus, from health to periodontitis. Extensive data have shown that the oral treponemes emerge as dominant members of the oral ecology at gingival sites with increased inflammation (19, 37). This change in the inflammatory environment may be expected to result in alterations in a variety of local cues which require bacterial responsiveness to the milieu. Thus, we tested the synergistic effect of cysteine and iron, based upon the findings that cysteine and in vivo iron individually enhance T. denticola virulence. We noted that cysteine and iron contributed additively to the virulence of T. denticola. In contrast, these two environmental cues appeared to act independently with respect to the virulence of T. pectinovorum. These findings suggest that the two signals may be acting upon different aspects of each microorganism's survival and virulence strategies as expressed in this murine model.

It has also not been determined whether the increase in treponemes associated with gingivitis and periodontitis (19, 30, 37, 44) represents the treponemes contributing to the induction of an inflammatory environment or if their increase in the plaque results from the creation of a more favorable ecological niche at the inflammation sites. Thus, we hypothesized that locally induced inflammation would create a more favorable environment and increase the ability of T. denticola to induce tissue-destructive lesions, which could be evaluated in a murine model. We had previously reported that Cytodex 1 microcarrier beads, which create a localized inflammatory milieu in vivo, enhance the infectivity and pathogenicity of C. rectus as assessed by the murine abscess model (24). We evaluated the ability to affect treponeme virulence by altering the local in vivo environment of the bacteria. The abscess-promoting agent Cytodex microcarrier beads was evaluated for increasing local inflammation and for potentially providing a more favorable environment for the treponemes to establish an infectious focus. The results indicate that within the limitations of this model, local inflammation appears to have a minimal effect on virulence expression by treponemes. Of potential importance is that this model may be used to explore other modifications of the local in vivo environment and to determine their effect on treponeme virulence.

The LPS from gram-negative bacteria is capable of eliciting a wide range of biological responses and is implicated in the pathogenesis of a variety of diseases. Members of the genus Treponema have an outer envelope resembling that of gram-negative bacteria. Several reports have described an LPS-like component in T. denticola (10, 29, 53, 54), Serpulina (Treponema) hyodysenteriae (39, 40), and the nonpathogenic Reiter treponeme (12). Injection of GalN makes mice increasingly susceptible to the lethal effects of endotoxin (17). This model has been suggested to be a quantitative measure of the endotoxicity of LPS, particularly as related to the toxic activity of the lipid A portion of the molecule (47). We have recently identified an LPS-like macromolecule isolated from the outer sheath of T. pectinovorum (unpublished data). Thus, we hypothesized that LPS-like macromolecules could contribute to the pathogenicity of oral treponemes and evaluated this biological activity in a galactosamine sensitization model in mice. The results demonstrated a general lack of toxicity of both treponemes when injected i.p. into normal mice. However, both treponemes were lethal for mice that were concomitantly treated with GalN and elicited systemic changes normally associated with endotoxic activities. The majority of GalN-sensitized animals died within 8 h of bacterial challenge, consistent with an LPS-mediated effect (17). The systemic manifestations of endotoxemia have been identified as resulting from the ability of LPS to activate host cells to produce circulating levels of cytokines, such as interleukin 1, interleukin 6, and tumor necrosis factor alpha, which are the direct effectors of endotoxic shock and death. Therefore, while this study did not identify cytokines in the murine serum, nor specifically identify the capacity of the LPS to affect host responses, the results suggest that examination of the factor(s) that elicited the lethal outcome would be important in understanding treponeme-mediated pathogenicity.

The findings in this study clearly demonstrate the ability to modify the virulence of oral treponemes by altering the in vitro and in vivo environments to which these bacteria are subjected. This model can be manipulated to evaluate the contribution of putative virulence components of oral treponemes within the host.