The effect of periodontal proteolytic Bacteroides species on proteins of the human complement system

Expression of Porphyromonas gingivalis proteolytic activity in Escherichia coli

Porphyromonas gingivalis (formerly Bacteroides gingivalis) degrades numerous protein substrates including collagen, fibrinogen, fibronectin, gelatin, casein, immunoglobulins and complement components. In order to clone one or more of these protease genes, a genomic library was constructed with Sau3A1 restriction fragments of chromosomal DNA from P. gingivalis ATCC 33277 ligated into the temperature-regulated vector pCQV2, and expressed in Escherichia coli DH5 alpha mcr. The electro-transformants (3 x 10(4)) were screened for general protease activity on Luria broth agar containing ampicillin (50 mg/l) and sodium caseinate (2%). One casein-hydrolyzing clone was detected and subcultured, and the activity of the cell extracts was characterized. We were able to show that the protease-positive clone, (pTEM1), had broad substrate specificity. Colorimetric assays indicated the hydrolysis of azocoll, azocasein, collagen, elastin-congo red and artificial substrates. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis was used to confirm that collagen, casein, fibrinogen and fibronectin were degraded by the clone.

Cloning, expression, and sequencing of a protease gene (tpr) from Porphyromonas gingivalis W83 in Escherichia coli

Porphyromonas gingivalis is a highly proteolytic organism which metabolizes small peptides and amino acids. Indirect evidence suggests that the proteases produced by this microorganism constitute an important virulence factor. In this study, a gene bank of P. gingivalis W83 DNA was constructed by cloning 0.5- to 20-kb HindIII-cut DNA fragments into Escherichia coli DH5 alpha by using the plasmid vector pUC19. A clone expressing a protease from P. gingivalis was isolated on LB agar containing 1% skim milk. The clone contained a 3.0-kb insert that coded for a protease with an apparent molecular mass of 64 kDa. Sequencing part of the 3.0-kb DNA fragment revealed an open reading frame encoding a protein of 482 amino acids with a molecular mass of 62.5 kDa. Putative promoter and termination elements flanking the open reading frame were identified. The activity expressed in E. coli was extensively characterized by using various substrates and protease inhibitors, and the results suggest that it is possibly a thiol protease.

Inactivation of human serum bactericidal activity by a trypsinlike protease isolated from Porphyromonas gingivalis

A protease was isolated from an outer membrane vesicle preparation of Porphyromonas gingivalis ATCC 33277 and assessed for its ability to inactivate the bactericidal activity of normal human serum. The enzyme, which was activated by reducing agents, was found to be a trypsinlike protease with a molecular mass of approximately 80 kDa. Prior to being tested in the bactericidal assay, pooled human serum was preincubated with the partially purified enzyme. Under conditions in which the trypsinlike protease was activated, a strong reduction of the serum bactericidal activity against Capnocytophaga ochracea was noted. On the other hand, no reduction of the bactericidal action of serum was observed when the serum-protease mixture was preincubated in the presence of an inhibitor of the enzyme. As determined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, the protease was shown to degrade immunoglobulins G and M as well as complement factor C3. This study confirms the previous hypothesis that the proteases of P. gingivalis can interfere with the protective action of human serum.

Influence of immunization on Porphyromonas gingivalis colonization and invasion in the mouse chamber model

The effects of immunization with invasive or noninvasive Porphyromonas (Bacteroides) gingivalis strains on the pathogenesis of infection in a mouse chamber model were examined. BALB/c mice were immunized by a single injection of heat-killed P. gingivalis invasive strain A7436 or W83 or noninvasive strain 33277, HG405, or 381 directly into subcutaneous chambers. P. gingivalis-specific antibody was detected in chamber fluid 21 days postimmunization, and mice were subsequently challenged by injection of exponential-phase P. gingivalis into chambers. Immunization with A7436 or W83 followed by challenge with A7436 protected mice against secondary abscess formation and death; however, P. gingivalis persisted in chambers for up to 14 days postchallenge. Immunization with noninvasive strain 33277, HG405, or 381 followed by challenge with invasive strain A7436 or W83 protected mice against secondary lesion formation and death. P. gingivalis was cultured from 33277- or HG405-immunized and nonimmunized animals to day 14. All P. gingivalis strains induced an immunoglobulin G response, as measured by an enzyme-linked immunosorbent assay and Western immunoblotting of P. gingivalis whole-cell and outer membrane protein preparations. Western blot analyses indicated that sera from mice immunized with different invasive and noninvasive strains recognized common P. gingivalis antigens. In summary, immunization with invasive P. gingivalis A7436 and W83 or noninvasive P. gingivalis 33277, HG405, and 381 protected mice from secondary lesion formation and death after challenge with invasive P. gingivalis A7436 or W83. P. gingivalis-specific antibody did not, however, inhibit the colonization of P. gingivalis within chambers.

Host responses in periodontal diseases: current concepts

In periodontal diseases, bacteria trigger inflammatory host responses which, along with the direct destructive effects of the bacteria, cause most of the tissue destruction. Periodontal inflammatory responses are, by and large, immunologic, and our understanding of these reactions has been advanced by the explosion of knowledge in immunobiology, some of which is discussed in this review. Understanding the role of immune cells and their regulatory cell surface molecules such as the MHC, CD antigens, and receptors, as well as knowledge of effector systems set into motion such as phagocytes and cytotoxic T-cells, and the effector molecules such as antibodies, complement, and cytokines, have led to better understanding of the complex pathogenesis of periodontal disease. The role of mediators including the matrix metalloproteinases, proteoglycans, the kinins and anaphylatoxins, and low molecular weight mediators including products of arachidonic metabolism is beginning to be elucidated in periodontal disease. Important avenues of research for development of diagnostic tests based upon host response are apparent. For example, tissue products released during periodontal inflammation including the metalloproteinases, elastase, cytokines, prostaglandins, antibodies, and complement components may provide the basis for future diagnostic indicator tests. The recognition that the neutrophil/antibody/complement axis is critical for protection against periodontal bacteria and that abnormalities in this system often lead to increased periodontal susceptibility provide approaches for the development of diagnostic tests assessing risk. A group of factors which are negative regulators of inflammation including TGF-beta, gamma-interferon, and IL-1 receptor antagonist provide potential for assessment of periodontal disease in remission or in the healing phase. Finally, factors such as HLA associations and the molecular basis for neutrophil abnormalities may provide genetic markers for periodontal disease susceptibility. Diagnostic factors based upon host response measures offer great potential for predicting host susceptibility and will likely be used in combination with microbial diagnostics which identify specific infecting organisms.

Role of superoxide dismutase in resistance of Porphyromonas gingivalis to killing by polymorphonuclear leukocytes

Porphyromonas gingivalis in which the synthesis of superoxide dismutase (SOD) was induced by nitrate or by aeration was rendered resistant to killing by polymorphonuclear leukocytes. SOD purified from either anaerobically maintained or aerated cells also inhibited bacterial killing when added exogenously, and no difference between the effects of the two SODs was observed. These results suggest that SOD may form part of a defense mechanism that helps protect P. gingivalis against killing by polymorphonuclear leukocytes.

Protective effect of Porphyromonas gingivalis outer membrane vesicles against bactericidal activity of human serum

The present study was undertaken to evaluate the effect of Porphyromonas gingivalis outer membrane vesicles on the bactericidal activity of human serum. Human serum was pretreated with extracellular vesicles and then incubated with a cell suspension of Capnocytophaga ochracea. After 2 h at 37 degrees C, the percent viability of C. ochracea was determined by cultivation on blood agar plates. At a final concentration of 0.3 mg/ml, outer membrane vesicles completely inhibited the serum bactericidal activity against C. ochracea. Boiling the vesicles prevented this inhibition. However, partial inhibition of the serum lethal action was obtained when a higher concentration (1.5 mg/ml) of boiled vesicles was used, which indicates the involvement of both heat-labile and heat-stable components associated with vesicles. Combining vesicles at a suboptimal concentration (0.1 mg/ml) with a reducing agent brought back inhibition of the bactericidal activity, whereas combining vesicles at an optimal concentration (0.3 mg/ml) with a thiol-blocking reagent caused a restoration of the bactericidal activity. When a purified preparation of P. gingivalis lipopolysaccharides was used instead of vesicles, inhibition of the bactericidal activity was also observed. These results indicate that the lipopolysaccharides and the proteolytic enzyme(s) associated with P. gingivalis outer membrane vesicles are likely to represent the heat-stable and the heat-labile components, respectively. It is possible that outer membrane vesicles released by P. gingivalis protect other bacterial species from complement action, thus favoring the pathogenic process of periodontal disease.

Complement factor D-like activity of Porphyromonas gingivalis W83

Porphyromonas gingivalis is a proteolytic gram-negative anaerobic bacterium that is frequently isolated from lesions of human periodontal disease. Previous studies have shown that P. gingivalis strain W83 inactivates C3 in pooled normal human serum (NHS) by a mechanism that is inhibitable by EDTA, yet it degrades purified complement proteins by a mechanism that is not EDTA-inhibitable. Furthermore, during complement activation, only a small number of C3 molecules accumulate on the surface of this organism unless the bacteria are treated with the protease inhibitor TLCK prior to complement activation. The hypothesis was tested that P. gingivalis W83 contains protease activity mimicking that of complement factor D, thus enabling it to activate C3 in serum without significant C3 accumulation on the cell surface. It was first noted that incubation of P. gingivalis W83 in absorbed human serum that was depleted of factor D resulted in C3 consumption that was reversed in the presence of the protease inhibitor TLCK. To directly demonstrate that factor B-dependent C3 consumption occurs in the absence of factor D, P. gingivalis W83 was incubated with purified C3 or a mixture of C3 and B. Although some proteolysis of C3 was noted, increased C3 consumption was noted in mixtures containing both C3 and B. This increment in C3 consumption was inhibited by both EDTA and TLCK. Furthermore, the addition of purified factor H to this mixture inhibited the increment in C3 consumption, indicating that a C3 convertase was probably formed.(ABSTRACT TRUNCATED AT 250 WORDS)

Stimulation of proteinase and amidase activities in Porphyromonas (Bacteroides) gingivalis by amino acids and dipeptides

Proteolytic enzymes from the organism Porphyromonas gingivalis are believed to be involved in the development of periodontitis. Studies on both crude extracts and purified trypsinlike enzymes from this organism indicate that substantial stimulation of both amidase and proteinase activities can be obtained during incubation with glycine-containing compounds. We postulate that P. gingivalis may have developed this unusual property to take advantage of the glycine-rich environment which occurs during the periodontitis-associated degradation of gingival collagen. The finding of such a stimulation in crevicular fluids from discrete periodontal sites has been correlated with the presence of P. gingivalis and could be utilized for the early detection of infection by this organism during the onset of periodontitis.

Humoral immune response to an antigen from Porphyromonas gingivalis 381 in periodontal disease

The humoral immune responses of patients with periodontitis were evaluated to characterize the host response to Porphyromonas gingivalis. A sonic extract of P. gingivalis 381 from whole cells was fractionated by gel chromatography and ion-exchange chromatography. The fractionated extracts were evaluated by Western blot (immunoblot) analyses with patient sera. A dominant antigen was identified from the sonic extract with an apparent molecular mass of 53 kDa. The 53-kDa protein antigen (Ag53) was purified by affinity chromatography by using a monoclonal antibody. Ag53 was detected on the vesicle surface of P. gingivalis 381 by immunoelectron microscopy by using the monoclonal antibody and was detected as a major protein in the outer membrane and in vesicles by Western blot analysis. Monoclonal antibody cross-reactivity to Ag53 in the sonic extracts of P. gingivalis ATCC 33277, P. gingivalis 1021, and Porphyromonas endodontalis ATCC 35406 was revealed. Seventy-seven patients with periodontitis were examined for their responses to Ag53. Serum immunoglobulin G (IgG) from 54 patients reacted strongly to Ag53; however, serum IgG from the remaining 23 patients did not exhibit detectable reactivity at all to Ag53, even though the patients had high serum IgG titers to the sonic extract. Ag53 is a new marker that represents an interesting aspect of the humoral immune response to P. gingivalis in patients with periodontitis.

Antibody-dependent alternate pathway of complement activation in opsonophagocytosis of Porphyromonas gingivalis

It has been suggested that the ability of Porphyromonas gingivalis to proteolyse complement, as well as its production of a capsule, contributes to resistance to phagocytosis by polymorphonuclear leukocytes. In this report, the opsonic role of serum complement and its activation pathways were investigated, using individual sera heat treated or depleted of factors B, C2, and C1q and the divalent cations Mg2+ and Ca2+. A fluorochrome microassay was used to quantitate phagocytosis of P. gingivalis A7436 by human polymorphonuclear leukocytes. Heat treatment of rabbit antiserum to P. gingivalis (RaPg) (56 degrees C, 30 min) resulted in a reduction in phagocytosis from 100% to 55% +/- 5%, while heat treatment of chronic adult periodontal disease serum abrogated phagocytosis. The heat-labile activity of RaPg was fully restored with MgEGTA-chelated rabbit serum but not EDTA- or EGTA-chelated rabbit serum. The addition of serum depleted of factor B but not C2 or C1q restored most of the heat-labile activity; however, the factor B-depleted serum was suspect, due to low-level opsonization of zymosan (inhibitable by EDTA but not MgEGTA). Adding C1q at 80 micrograms/ml to serum depleted of C1q restored much but not all of the activity lost through heat treatment or through depletion of C1q. A large part of opsonic activity with C2- and C1q-depleted sera was enhanced by the addition of 4 x 10(-3) M Mg2+. The data indicate that although opsonophagocytosis of P. gingivalis A7436 is dependent on the classical complement pathway, a significant contribution is made by an antibody-dependent alternate pathway.

Phagocytosis of virulent Porphyromonas gingivalis by human polymorphonuclear leukocytes requires specific immunoglobulin G

No studies to date clearly define the interactions between Porphyromonas gingivalis and human peripheral blood polymorphonuclear leukocytes (PMN), nor has a protective role for antibody to P. gingivalis been defined. Using a fluorochrome phagocytosis microassay, we investigated PMN phagocytosis and killing of P. gingivalis as a function of P. gingivalis-specific antibody. Sera from a nonimmune rabbit and a healthy human subject were not opsonic for virulent P. gingivalis A7436, W83, and HG405; phagocytosis of these strains (but not 33277) required opsonization with hyperimmune antiserum (RaPg). Diluting RaPg with a constant complement source decreased proportionally the number of P. gingivalis A7436 cells phagocytosed per phagocytic PMN. Enriching for the immunoglobulin G fraction of RAPg A7436 enriched for opsonic activity toward A7436. An opsonic evaluation of 18 serum samples from adult periodontitis patients revealed that only 3 adult periodontitis sera of 17 with elevated immunoglobulin G to P. gingivalis A7436 were opsonic for A7436 and, moreover, that the serum sample with the highest enzyme-linked immunosorbent assay titer was most opsonic (patient 1). However, the opsonic activity of serum from patient 1 was qualitatively and not just quantitatively different from that of the nonopsonic human sera (but was less effective opsonin than RaPg). Strain variability was observed in resistance of P. gingivalis to phagocytosis, and opsonization was strain specific for some, but not all, strains tested. An evaluation of killing of A7436 revealed that serum killing and extracellular killing of P. gingivalis were less effective alone when compared with intracellular PMN killing alone.

A novel mouse model to study the virulence of and host response to Porphyromonas (Bacteroides) gingivalis

We describe here the development of a mouse subcutaneous chamber model that allows for the examination of host-parasite interactions as well as the determination of gross pathology with Porphyromonas (Bacteroides) gingivalis challenge. When inoculated into stainless-steel chambers implanted subcutaneously in female BALB/c mice, P. gingivalis W83, W50, and A7436 (10(8) to 10(10) CFU) caused cachexia, ruffling, general erythema and phlegmonous, ulcerated, necrotic lesions, and death. P. gingivalis W50/BEI, HG405, and 33277 (10(10) CFU) produced localized abscesses in the mouse chamber model with rejection of chambers at the injection site. Analysis of chamber fluid from 33277-, HG405-, and W50/BEI-infected mice by cytocentrifugation revealed inflammatory cell debris, polymorphonuclear leukocytes, and high numbers of dead bacteria. In contrast, fluid from A7436-, W50-, and W83-infected mice revealed infiltration predominantly of polymorphonuclear leukocytes and live bacteria. Bacteria were found primarily associated with polymorphonuclear leukocytes in the fluid from W50-, HG405-, and W83-infected mice but not from A7436-infected mice. Viable isolates were recoverable from the chamber fluid through day 3 for W50/BEI, day 5 for 33277, day 6 for HG405, day 7 for W50, day 14 for W83, and day 26 for A7436. All strains induced a systemic immunoglobulin G response in serum and chamber fluid samples. The use of this model will allow us to examine the virulence of P. gingivalis as defined by the interaction of host response to localized infection with P. gingivalis.

Factors in virulence expression and their role in periodontal disease pathogenesis

The classic progression of the development of periodontitis with its associated formation of an inflammatory lesion is characterized by a highly reproducible microbiological progression of a Gram-positive microbiota to a highly pathogenic Gram-negative one. While this Gram-negative microbiota is estimated to consist of at least 300 different microbial species, it appears to consist of a very limited number of microbial species that are involved in the destruction of periodontal diseases. Among these "putative periodontopathic species" are members of the genera Porphyromonas, Bacteroides, Fusobacterium, Wolinella, Actinobacillus, Capnocytophaga, and Eikenella. While members of the genera Actinomyces and Streptococcus may not be directly involved in the microbial progression, these species do appear to be essential to the construction of the network of microbial species that comprise both the subgingival plaque matrix. The temporal fluctuation (emergence/disappearance) of members of this microbiota from the developing lesion appears to depend upon the physical interaction of the periodontal pocket inhabitants, as well as the utilization of the metabolic end-products of the respective species intimately involved in the disease progression. A concerted action of the end-products of prokaryotic metabolism and the destruction of host tissues through the action of a large number of excreted proteolytic enzymes from several of these periodontopathogens contribute directly to the periodontal disease process.(ABSTRACT TRUNCATED AT 250 WORDS)

Effects of anaerobiosis and aerobiosis on interactions of human polymorphonuclear leukocytes with the dental plaque bacteria Streptococcus mutans, Capnocytophaga ochracea, and Bacteroides gingivalis

Human polymorphonuclear leukocytes (PMN) were able to generate and release superoxide anions upon stimulation of Streptococcus mutans, Bacteroides gingivalis, and Capnocytophaga ochracea when incubated aerobically but not when incubated anaerobically. Lysozyme release and phagocytosis by PMN were independent of oxygen, and no difference between PMN incubated aerobically or anaerobically was observed (PMN stimulated by B. gingivalis released 7.6% total lysozyme when aerobic and 6.9% when anaerobic). There were variations in lysozyme release and phagocytosis for the three organisms, particularly for phagocytosis. B. gingivalis and C. ochracea yielded lower phagocytosis values than those for S. mutans, e.g., at 1 h 67% of the initial inoculum of S. mutans was phagocytosed (versus only 40% for B. gingivalis). Transmission electron microscopy showed that both S. mutans and B. gingivalis were internalized into classical phagolysosomes. In contrast, C. ochracea showed two forms of internalization; C. ochracea either formed a classical phagolysosome or was tightly bound in the cytoplasm with no surrounding cell membrane. Intracellular killing of S. mutans and C. ochracea was unaffected by anaerobiosis, but killing of C. ochracea was much lower than that of S. mutans (1 x 10(7) to 2 x 10(7) bacteria killed compared with 5.1 x 10(7) bacteria killed at 6 h). In contrast, a greater number of B. gingivalis was killed in the presence of oxygen (5.3 x 10(7) bacteria were killed when aerobically incubated and 1.9 x 10(7) bacteria were killed when anaerobically incubated). These results suggest that the ability to survive anaerobically may enable some bacteria to evade PMN killing; however, abnormal phagocytosis may represent a more efficient way to evade both oxygen-dependent and -independent killing mechanisms, leading to enhanced virulence of the organism.

Activation of complement by Treponema denticola

Oral spirochetes have been shown to be associated with periodontal diseases and are present in increased numbers in lesions of greater severity. In this study, the interaction of Treponema denticola with human complement, a major antibacterial defense system, was examined. For each of two strains of T. denticola, it was found that both the classical and alternative pathways of human complement were activated in human serum upon incubation at 37 degrees C. C3 fragments were deposited on the surface of this organism following complement activation; the fragments bound included both of the major C3-derived opsonic fragments C3b and iC3b. Under incubation conditions identical to those carried out for complement activation in serum, T. denticola failed to degrade purified, hemolytically-active C3, although it readily degraded inactivated C3. Thus, despite the documented proteolytic activity of this organism, complement activation and deposition of complement-derived opsonins may be important defense mechanisms in the control of infections with T. denticola.

Cell-associated and extracellular proteolytic activity of an oral flagellate, Trichomonas tenax

Proteolytic activities in crude extracts and culture filtrates from Trichomonas tenax were determined using hide powder azure as substrate and the proteinase profiles in both samples were analysed in SDS-polyacrylamide gels containing copolymerized gelatin. The enzyme activity in the crude extract was detected over a broad pH range and was strongly activated by dithiothreitol, mainly in the pH range 5-8, and inhibited by cysteine proteinase inhibitors. Extracellular enzyme activity in culture filtrates was SH-dependent and increased continuously during incubation of the cell suspension, suggesting proteinase release. A total of seven distinct proteolytic bands could be detected in crude preparations. Three of these, with apparent Mr values 35,000, 45,000 and 56,000 and a pH optimum of 4-7, were SH-dependent and their inhibitory sensitivities were characteristic for cysteine proteinases. The 45,000 and 56,000 proteinases probably corresponded to those found in the culture filtrates. Proteolytic bands with apparent Mr 76,000, 87,000, 102,000 and 270,000 and pH optima in the alkaline region, pH 8-9, were independent of SH groups and were inhibited by a chelating agent EDTA, suggesting that they belong to the metalloproteinase family.

The role of complement in periodontal diseases

The complement system has been implicated as both a pathogenic mechanism and a means of protection in periodontal diseases. It is well known that bacteria activate complement; such activation can initiate a number of events, including bacterial opsonization and killing, release of inflammatory agents, and modulation of other immune reactions. Cleavage of complement proteins has been observed in gingival fluids from individuals with periodontal disease and some investigators have observed complement deposition in diseased gingival tissues. Furthermore, a number of bacterial from individuals with periodontal diseases have been found to activate complement in vitro; some of these organisms appear to have the capacity to evade opsonization due to their proteolytic capacity. However, concrete evidence is not yet available that indicates that complement activation occurs in human periodontal disease and is important in either its pathogenesis or in protection against bacterial virulence factors.

A proteolytic enzyme secreted by Proteus mirabilis degrades immunoglobulins of the immunoglobulin A1 (IgA1), IgA2, and IgG isotypes

Proteus mirabilis strains associated with human urinary tract infections have previously been shown to secrete an extracellular metalloproteinase which cleaves serum immunoglobulin A (IgA). The enzyme has now been purified to apparent homogeneity from culture supernatants of P. mirabilis 64676. The protease activity is associated with a 50-kilodalton (kDa) protein. Unlike that of the classic IgA1 proteases, the substrate specificity of the P. mirabilis protease has been found to extend to both sublcasses of IgA, IgG, and the nonimmunoglobulin substrates, secretory component and casein. Cleavage of IgA1 and IgA2 by the P. mirabilis protease yielded fragments on sodium dodecyl sulfate-polyacrylamide gel electrophoresis whose sizes were consistent with a cleavage site outside the hinge region. Both secretory IgA1 and IgA2 were also cleaved by P. mirabilis protease, although the secretory IgA2 molecule was less readily cleaved than secretory IgA1. Free and IgA-bound secretory components were degraded to some extent by P. mirabilis protease. Cleavage of IgG, however, occurred at the hinge region as a two-stage process. The first stage was pepsinlike and generated an F(ab')2 fragment of 120 kDa and a small pFc fragment detected on nonreduced polyacrylamide gels. In the second stage, the F(ab')2 product was cleaved to yield papainlike Fab and Fc fragments, visualized as a diffuse band of 40 to 50 kDa.

Failure of Bacteroides gingivalis W83 to accumulate bound C3 following opsonization with serum

Our previous studies have demonstrated that strains of Bacteroides gingivalis are capable of proteolytic degradation and inactivation of complement proteins including the third component of complement C3. Since a crucial step in the ability of complement to control bacterial infections is the binding of C3 fragments to the bacterial surface with subsequent enhancement of phagocytosis, further examination of the importance of the proteolytic capacity of Bacteroides in interactions with complement proteins was carried out by quantitating the amount of C3 bound to two proteolytic Bacteroides gingivalis strains. Pooled normal human serum (NHS) containing 125I-C3 was incubated with strains of B. gingivalis (W83 and ATCC 33277) and the non-proteolytic pathogen A. actinomycetemcomitans strain Y4, and samples of the reaction mixtures were removed at various time intervals for determination of bound C3. B. gingivalis 33277 bound only half the number of C3 molecules as did A. actinomycetemcomitans, while B. gingivalis W83 bound very little C3. A large increase in the number of C3 molecules bound to B. gingivalis W83 was noted in assays carried out in the presence of the protease inhibitor TLCK, indicating that bacterial proteases may be responsible for the lack of binding of C3 to strain W83. TLCK treatment modestly increased the accumulation of C3 on strain 33277, but had no effect on A. actinomycetemcomitans. Analysis of 125I-C3 in supernatants from reaction mixtures of strain 33277, W83, or a proteolytic strain of B. intermedius demonstrated no qualitative differences in the C3 fragments amongst the tested strains or in the presence or absence of TLCK.(ABSTRACT TRUNCATED AT 250 WORDS)