Effect of Free and Vesicle-Bound Cysteine Proteinases of

Download PDF
0 downloads 0 Views 245KB Size Report
Comment
because of the inhibitory effect of albumin. These results imply that ... gesting a disorder of the blood coagulation system at such lesions. Since coagulation ...
INFECTION AND IMMUNITY, Dec. 1995, p. 4877–4882 0019-9567/95/$04.0010 Copyright q 1995, American Society for Microbiology

Vol. 63, No. 12

Effect of Free and Vesicle-Bound Cysteine Proteinases of Porphyromonas gingivalis on Plasma Clot Formation: Implications for Bleeding Tendency at Periodontitis Sites TAKAHISA IMAMURA,1 JAN POTEMPA,2 ROBERT N. PIKE,3 JAMES N. MOORE,4 MICHELLE H. BARTON,4 AND JAMES TRAVIS3* Department of Biochemistry3 and Departments of Physiology & Pharmacology and Large Animal Medicine,4 College of Veterinary Medicine, University of Georgia, Athens, Georgia 30602; Division of Molecular Pathology, Department of Neuroscience and Immunology, Kumamoto University Graduate School of Medical Sciences, Kumamoto 860, Japan1; and Department of Microbiology and Immunology, Institute of Molecular Biology, Jagiellonian University, 31-120 Krako ´w, Poland2 Received 8 June 1995/Returned for modification 17 July 1995/Accepted 18 September 1995

Infection by Porphyromonas gingivalis is strongly associated with adult periodontitis, with proteinases from this bacterium now considered to be important virulence factors. In order to investigate possible pathological functions of these enzymes, we examined the effect of both free and vesicle-bound forms of the two major cysteine proteinases (gingipains) of P. gingivalis on plasma clot formation by using thrombin time (TT) measurements. Both Lys-gingipain (gingipain-K) and Arg-gingipain (gingipain-R) prolonged plasma TT in a dose- and time-dependent manner, and this was also found with vesicles which are the biological carriers of P. gingivalis proteinases. The increase in plasma TT by vesicles could be completely reversed by treatment with nonspecific cysteine proteinase inhibitors but only partially by compounds selective for either gingipain-K or gingipain-R. Preincubation of vesicles with a gingipain-K-specific inhibitor (z-FK-ck) reduced plasma TT more than a gingipain-R-specific inhibitor (leupeptin), suggesting that under physiological conditions gingipain-K was more effective in fibrinogen destruction. Each purified enzyme also markedly increased fibrinogen TT, gingipain-R being fourfold more potent than gingipain-K. However, in plasma, gingipain-R was ineffective because of the inhibitory effect of albumin. These results imply that cysteine proteinases, especially gingipain-K, abrogate the clotting potential of fibrinogen and, therefore, may contribute to the bleeding tendency and to persistent inflammation in periodontitis sites infected with P. gingivalis. binding and degrading fibrinogen (19–21). Such data corroborate and extend the findings of earlier reports which showed that this oral microorganism has fibrinogenolytic and fibrinolytic activities (27, 39). Proteolytic enzymes, which are produced in large quantity by P. gingivalis, are important factors in periodontitis (10, 23, 33). However, the individual proteinase(s) potentially responsible for the fibrin(ogen)olytic activity of this microorganism has not been clearly described. Recently, we purified three major cysteine proteinases from P. gingivalis, referred to as gingipains: two molecular mass variants of an arginine-X peptide bondspecific gingipain, of 50 and 95 kDa (3, 28), and a lysine-X bond-specific gingipain of 105 kDa (28). It has now been proposed by the International Union of Biochemistry that the naming of these enzymes be changed to account for their individual specificities. Thus, the arginine-specific cysteine proteinase from P. gingivalis will be referred to here as gingipain-R, and similarly, the lysine-specific proteinase will be referred to here as gingipain-K. Scott and coworkers had already demonstrated that a lysine-specific proteinase from P. gingivalis (probably gingipain-K) had potent fibrinogenolytic activity (31). However, their proteinase solution apparently contained significant amounts of at least one of the forms of gingipain-R. Thus, the fibrinogenolytic activity of their preparation could have been a cooperative effect of gingipain-K and gingipain-R, as we have recently shown in the release of bradykinin from

Periodontitis is characterized by gingival inflammation and loss of connective tissue and bone from around the roots of the teeth (9), resulting in formation of a chronically inflamed softtissue pocket between the gingiva and the root surface. Porphyromonas gingivalis has been implicated as an etiologic agent of human adult periodontitis (24, 32); however, the mechanism(s) by which this organism contributes to the pathogenic changes seen in this disease is not completely clear. Gingival bleeding is a typical sign of inflammation in periodontitis, suggesting a disorder of the blood coagulation system at such lesions. Since coagulation involves conversion of fibrinogen to fibrin by thrombin (5, 7), degradation of this clotting factor would clearly result in impaired clot formation and consequent bleeding. A positive correlation between the presence of P. gingivalis in periodontal pockets and a bleeding tendency has previously been shown (25, 34). Furthermore, a negative correlation between the proportion of intact fibrinogen in gingival crevicular fluid and clinical periodontitis parameters, including the Papilla Bleeding Index, has also been found (37). These data suggest the involvement of P. gingivalis as a causative agent for the tendency to bleed at periodontitis sites, presumably through degradation of fibrinogen (29). Indeed, studies by Lantz and coworkers indicate that P. gingivalis is capable of * Corresponding author. Phone: (706) 542-1711. Fax: (706) 5423719. 4877

4878

IMAMURA ET AL.

INFECT. IMMUN.

high-molecular-weight kininogen (15). They also did not examine the fibrinogenolytic activity of their gingipain-K preparation in plasma, where this activity could possibly be negated by competition from other proteins present in this fluid, e.g., albumin, high-molecular-weight kininogen, prekallikrein, and factor X, all of which are known to be cleaved by proteinases of P. gingivalis (1, 14, 26, 31). Therefore, from a biological point of view it seemed important to examine the fibrinogenolytic function of these enzymes directly in plasma. In this report, we describe the results of investigations of the effect of free and vesicle-bound gingipain-K and gingipain-R on the clotting potential of fibrinogen, each by itself and in plasma. MATERIALS AND METHODS Materials. Human fibrinogen; bovine thrombin; 1,10-phenanthroline; tosyl-Llysine chloromethyl ketone (TLCK); iodoacetamide; benzoyl-L-arginine-p-nitroanilide (BAPNA); Cibacron blue 3GA-Sepharose; and plasmas deficient in Hageman factor (factor XII), factor X, prekallikrein, and high-molecular-weight kininogen were purchased from Sigma Chemicals, St. Louis, Mo. Plasmin was from Calbiochem, San Diego, Calif., and its molar concentration was calculated by assuming that 1 U of plasmin activity corresponded to 0.061 mg of plasmin in the TAME assay (2). Benzoxycarbonyl-L-lysine-p-nitroanilide (z-KPNA) was from Novabiochem, La Jolla, Calif. Benzyloxycarbonyl-Phe-Lys-CH2OCO(2,4,6-Me3) phenyl z HCl (z-FK-ck) was kindly donated by A. Krantz, Syntex, Palo Alto, Calif. H-D-Phe-Pro-Arg-chloromethylketone (FPRck) was from Bachem Bioscience, Inc., Philadelphia, Pa. Human plasma was obtained from a single healthy volunteer by the addition of 9 volumes of freshly drawn blood to 1 volume of 3.8% (wt/vol) sodium citrate followed by centrifugation. Bacterial cultivation and preparation of vesicles. P. gingivalis ATCC 53978 (W50) was purchased from the American Type Culture Collection, and the H66 strain, a laboratory strain, was a gift from Roland Arnold, University of North Carolina, Chapel Hill. The total gingipain-R and gingipain-K activities were similar for these strains, but most of the activities of H66 were released into the culture media while those of P. gingivalis ATCC 53978 were localized in vesicles and the cell membrane (30). The cells were grown in 100 ml of broth containing 15.0 g of Trypticase soy broth (Difco), 2.5 g of yeast extract, 2.5 mg of hemin, 0.25 g of cysteine, 0.05 g of dithiothreitol, and 0.5 mg of menadione (all from Sigma), anaerobically, at 378C for 24 to 30 h in an atmosphere of 85% N2, 10% CO2, and 5% H2. The entire culture was then used to inoculate 2 liters of the same broth which was incubated, anaerobically, at 378C for about 48 h until the late stationary phase of bacterial growth in order to obtain maximum vesicle production. To obtain vesicles, cultures of P. gingivalis ATCC 53978 (250 ml) were centrifuged (6,000 3 g, 30 min, 48C) to remove cells. The culture fluid obtained was then subjected to ultracentrifugation (100,000 3 g, 60 min, 48C). The precipitate, resuspended in 20 mM Bis-Tris–150 mM NaCl–5 mM CaCl2 (pH 7.8), was regarded as the vesicle fraction. The amounts of gingipain-K and gingipain-R activities associated with vesicles were measured by titration with specific inhibitors as described below. Proteinase purification. Gingipain-K, 50-gingipain-R, and 95-gingipain-R were purified according to the method of Pike et al. (28). In all experiments performed in the present study it was found that the two gingipain-R proteinases behaved identically. Therefore, for convenience, we have utilized the generic term gingipain-R. The amount of active gingipain-K or gingipain-R in each batch of purified proteinase was determined by active-site titration with z-FK-ck and FPRck, respectively (30). Briefly, activated gingipain-K or gingipain-R was preincubated with increasing amounts of either z-FK-ck or FPRck; this was followed by measurement of residual enzyme activity with z-KPNA or BAPNA. The concentration of gingipain-K or gingipain-R was calculated from the amount of inhibitor needed for complete inactivation of each enzyme. Enzyme activity assay. The amidolytic activity of purified gingipain-R, gingipain-K, and vesicles was determined with either BAPNA or z-KPNA. Samples were preincubated in 0.1 M Tris-HCl–5 mM CaCl2–10 mM cysteine (pH 7.6) for 5 min at 378C and then assayed for amidase activity with 1 mM substrate. The formation of p-nitroaniline was monitored spectrophotometrically at 405 nm. Activation of proteinases. Both purified proteinases and those associated with vesicles were activated with 10 mM cysteine at pH 8.0 at 378C for 10 min; this was followed by dilution with Tris-buffered saline (TBS) (10 mM Tris-HCl–150 mM NaCl [pH 7.3]) containing 10 mM cysteine and 1 mM CaCl2. Thrombin time (TT) measurement. Fibrinogenolytic activity of the various proteinases tested was assayed by changes in TT, measured with a coagulometer (COAG-A-MATE XC; General Diagnostics, Morris Plains, N.J.) according to the manufacturer’s instructions. Briefly, 180 ml of citrated human plasma or fibrinogen (3 mg/ml) was incubated with 20 ml of proteinase or vesicles diluted in TBS, at 378C for 180 s in a plastic cell. Bovine thrombin (5 U/ml; 200 ml utilized for plasma and 100 ml utilized for fibrinogen) was added, and the clotting time was measured spectrophotometrically. TBS, in the absence of proteinase, was used as a control. Each assay was performed in triplicate, and the value was expressed as the mean 6 standard deviation.

FIG. 1. (A) TTs of human plasma incubated with various concentrations of cysteine proteinases were assayed. Final concentrations of proteinases in plasma are shown. The dotted zone indicates the range of TT of plasma incubated with TBS. E, gingipain-K; Ç, gingipain-R; F, gingipain-K treated with TLCK; å, gingipain-R treated with TLCK; h, human plasmin. (B [Inset]) TT of plasma diluted with TBS. Dashed lines indicate TTs of gingipain-K (7 nM) and gingipain-R (100 nM) and plasma concentrations corresponding to each TT.

Purification of human albumin. Human albumin was purified according to the method of Travis and Pannell (38). Preparation of albumin-depleted plasma. Six milliliters of normal plasma, diluted threefold with TBS, was applied to 40 ml of Blue-Sepharose, previously swollen with TBS. After being stirred mildly for 30 min, the solution was centrifuged at 1,300 3 g for 10 min and the supernatant was used as albumindepleted plasma. The albumin content in this plasma was less than 0.1% of the initial plasma albumin content when measured by a single immunodiffusion assay with goat anti-human albumin antibody.

RESULTS Effect of cysteine proteinases from P. gingivalis on TT. In order to examine the effect of gingipain-K and gingipain-R on fibrinogen under semiphysiological conditions, we measured TT after incubating plasma with each proteinase. Each enzyme prolonged TT at concentrations of 2 nM (gingipain-K) and 10 nM (gingipain-R) in a dose-dependent manner. However, prolongation by gingipain-K (7 nM) was longer than that by gingipain-R (100 nM), indicating that gingipain-K was apparently 13-fold more potent in this assay (Fig. 1A). When TBS-diluted plasma was used as the source of fibrinogen, TT was prolonged in a linear relationship with the logarithmic decrease of percent plasma concentration (Fig. 1B). This was similar to the relationship previously shown with fibrinogen (31). From this standard curve, gingipain-K was estimated to have reduced the clot-forming ability of the fibrinogen in plasma by about 70% at 7 nM, while gingipain-R eliminated about 60% of the fibrinogen clotting activity at 100 nM. Since proteinases inactivated with TLCK did not increase TT (Fig. 1A and 2), prolongation of clotting by gingipain-K or gingipain-R is obviously dependent on enzymatic activity. The TT of plasma in the presence of gingipain-K or gingipain-R could be increased by prolonged incubation (Fig. 2). Since the TTs of plasma incubated for 3 min with either gingipain-K (70 nM) or gingipain-R (100 nM) were similar to those of plasma incubated for 30 min with gingipain-K (7 nM) and gingipain-R (10 nM), respectively (Fig. 1A and 2), it appears that none of the plasma proteinase inhibitors significantly inhibits either proteinase. On the other hand, plasmin, a representative fibrino(geno)lytic proteinase,

VOL. 63, 1995

P. GINGIVALIS PROTEINASES AND GINGIVAL BLEEDING

FIG. 2. TTs of human plasma assayed after incubation with cysteine proteinases for various periods at 378C. Final concentrations of proteinases in plasma are shown. The values at time zero show TT of plasma incubated with TBS for 3 min. E, 7 nM gingipain-K; Ç, 10 nM gingipain-R; F, 7 nM gingipain-K treated with 10 mM TLCK and dialyzed against TBS; å, 10 nM gingipain-R treated with 10 mM TLCK and dialyzed against TBS.

exhibited a negligible TT prolongation even at 100 nM (Fig. 1A), probably because of the presence of plasmin inhibitors in plasma, particularly a2-antiplasmin. In order to confirm the effect of each of the P. gingivalis proteinases on fibrinogen, the TT of this clotting factor was measured after incubation with gingipain-K or gingipain-R. In both cases, the fibrinogen TT was prolonged in a dose-dependent manner, with the effect of gingipain-K being comparable to that of plasmin (Fig. 3). Curiously, gingipain-R abolished fibrinogen clot formation at much lower concentrations of enzyme than those obtained with plasma, while gingipain-K exhibited a similar dose-dependent behavior with either fibrinogen or plasma (Fig. 3). As the TT of fibrinogen incubated with gingipain-R (10 nM) was significantly longer than that after incubation with gingipain-K (40 nM), gingipain-R is at least

FIG. 3. TTs of human fibrinogen (3 mg/ml) assayed after incubation with cysteine proteinases. Final concentrations of proteinases in fibrinogen solution are shown. The dotted zone indicates the range of TT of fibrinogen incubated with TBS. E, gingipain-K; Ç, gingipain-R; h, human plasmin.

4879

FIG. 4. TTs of plasma (A) or fibrinogen (3 mg/ml) (B) incubated with cysteine proteinases for 5 or 180 s. u, TBS; ■, gingipain-K; 3, gingipain-R. (A) Gingipain-K (20 nM) and gingipain-R (100 nM) were used. (B) Gingipain-K (40 nM) and gingipain-R (10 nM) were used.

fourfold more potent in reducing fibrinogen clot-forming ability in this system (Fig. 3). In order to exclude the possibility that the TT prolongation effect of gingipain-K and gingipain-R is due to inactivation of thrombin by these proteinases, the TTs of plasma and fibrinogen, to which thrombin was added immediately after the addition of each proteinase, were measured. As shown in Fig. 4, the TTs of plasma and fibrinogen were not prolonged after 5 s of incubation with each proteinase, whereas they were markedly prolonged after 180 s of incubation. The results indicate that gingipain-K and gingipain-R prolong TT through abrogation of fibrinogen clotforming ability but not through inactivation of thrombin activity. Effect of albumin on the prolongation of fibrinogen TT by bacterial cysteine proteinases. The fact that gingipain-R prolonged TT more efficiently in fibrinogen solution than in plasma suggested that some plasma proteins interfere with this activity. We investigated the TTs of plasmas deficient in prekallikrein, Hageman factor, high-molecular-weight kininogen, or factor X after incubation with gingipain-R or gingipain-K. Gingipain-K was more efficient than gingipain-R in TT prolongation for both normal plasma and all of the deficient plasmas, indicating that none of these proteins had interfered significantly in the plasma TT assay (Table 1). However, albumin-depleted plasma increased gingipain-R TT prolongation to the gingipain-K level but had no effect on gingipain-K. Significantly, the increased gingipain-R TT prolongation was normalized by reconstitution of the depleted plasma with physiological concentrations of albumin (Table 1). These results indicate that albumin inhibits fibrinogen degradation by gingipain-R in plasma. Effect of vesicles of P. gingivalis on TT. The fact that gingipain-K is more potent than gingipain-R in the plasma TT assay suggests that most of the fibrinogen-degrading activity of P. gingivalis is dependent on gingipain-K. In order to investigate this further, we assayed the TT of plasma incubated with P. gingivalis vesicles which contain both gingipain-K and gingipain-R. Vesicles prolonged the TT in a dose- and incubation time-dependent manner, but the effect was only one-half of that obtained with a mixture of gingipain-K and gingipain-R with equivalent activity (16). We therefore determined which proteinases were associated with the fibrinogen-degrading ac-

4880

IMAMURA ET AL.

INFECT. IMMUN.

TABLE 1. TTs of various plasmas incubated with gingipain-K or gingipain-Ra Plasmab

Normal PK deficient HF deficient Factor X deficient HMWK deficient Albumin depletede Albumin depleted 1 albumin (10 mg/ml)f

TT (s)c 2d

Gingipain-K

Gingipain-R

15.5 6 0.5 15.6 6 0.2 15.0 6 0.5 14.6 6 0.1 17.1 6 1.5 14.8 6 0.3 14.5 6 0.5

22.1 6 1.2 25.1 6 1.1 21.0 6 1.1 22.3 6 0.9 30.3 6 1.5 19.5 6 0.7 19.1 6 0.5

17.1 6 0.8 18.3 6 0.8 17.3 6 0.7 17.5 6 0.9 21.2 6 0.8 19.2 6 0.7 16.7 6 0.7

a Deficient plasma (180 ml), diluted threefold with TBS, was incubated with 20 ml of a given proteinase (5 nM, final concentration) at 378C for 180 s, and the clotting time was measured with a coagulometer after adding 100 ml of thrombin (5 U/ml). b PK, prekallikrein; HF, Hageman factor (coagulation factor XII); factor X, coagulation factor X; HMWK, high-molecular-weight kininogen. c Data are means 6 standard deviation. d 2, TBS was added to plasma instead of proteinase. e Normal plasma diluted threefold with TBS, followed by Blue-Sepharose treatment. f Reconstituted, albumin-depleted plasma (final concentration, 10 mg/ml).

tivity by examining the effect of proteinase inhibitors on the extension of plasma TT induced by vesicles. The fibrinogenolytic activity was found to be very low in the absence of cysteine, and in its presence it was almost completely inactivated by iodoacetamide or TLCK (Fig. 5). These results suggest that cysteine proteinases associated with vesicles, presumably gingipain-K and gingipain-R, are likely to be responsible for the protraction of plasma TT. By examining the effect of two proteinase inhibitors, one specific for gingipain-K (z-FK-ck) and the other specific for gingipain-R (leupeptin), we were able to determine which of the two proteinases was involved in vesicleinduced loss of fibrinogen clot-forming ability. Both inhibitors reduced activity, but z-FK-ck was more effective than leupeptin (Fig. 6), and neither had any effect on the TT of control

FIG. 5. Effect of inhibitors of gingipain-K and gingipain-R in vesicles on the TT of plasma. The final concentrations of gingipain-K and gingipain-R in the vesicles were 20 and 33 nM, respectively. (2), TBS instead of vesicles; non-act., vesicles which were not activated with cysteine; non-tr., vesicles which were activated with cysteine but not treated with any inhibitors; IA-tr., vesicles which were activated, treated with 10 mM iodoacetamide, and dialyzed against TBS; TLCK-tr., vesicles which were activated, treated with 10 mM TLCK, and dialyzed against TBS.

FIG. 6. Effect of inhibitors on TT of plasma treated with gingipain-K (KGP), gingipain-R (RGP), or vesicles. Samples were treated with z-FK-ck or leupeptin (final concentrations, 3 and 5 mM, respectively), and TTs of plasma incubated with each of these proteinases or vesicles were measured. Final concentrations of gingipain-K and gingipain-R were 10 and 20 nM, respectively, and the vesicles contained equivalent activities. Dashed lines indicate the range of control TTs of plasma, which was incubated with TBS containing 1 mM CaCl2 and 10 mM cysteine. u, non-treated; ■, treated with z-FK-ck; 3, treated with leupeptin.

plasma. These results suggest that gingipain-K is the primary factor in vesicles which is responsible for the interruption in fibrinogen clotting. DISCUSSION Previous findings showing the interaction of either P. gingivalis or purified bacterial proteinases with fibrinogen were based on their binding and subsequent digestion of the purified protein (19–21, 27, 39). However, from a biological point of view it would seem to be more important to focus on the effect of the bacterial proteinases in plasma and examine any effect on clotting of fibrinogen. The results presented here clearly indicate that the two cysteine proteinases gingipain-K and gingipain-R are essential fibrinogenolytic factors for P. gingivalis and support data indicating the association of this organism with a tendency to bleed at periodontitis sites. The data obtained which show that gingipain-K prolonged both plasma TT and fibrinogen TT in a parallel manner (Fig. 1 and 3) are consistent with our recent observation that gingipain-K binds fibrinogen with considerable specificity and fragments this protein as rapidly in plasma as it does with purified protein (29). The preference of gingipain-K for fibrinogen parallels the fact that it is also the major P. gingivalis proteinase which reduces fibrinogen clotting in plasma (Fig. 6). In addition to the P. gingivalis proteinases, plasmin is a well-known fibrinogenolytic proteinase produced through activation of plasminogen (12) and, therefore, also possibly involved in fibrinogen degradation at periodontitis sites. As addition of plasminogen does not, apparently, enhance the prolongation of fibrinogen TT by gingipain-K or gingipain-R (16), it is unlikely that these proteinases activate this zymogen. However, tissue plasminogen activator is also present in gingival crevicular fluid and gingival tissue (4, 18), and it is possible that plasmin production could occur in this manner at periodontal sites. However, since both a2-antiplasmin and plasminogen activator inhibitors are each present in gingival crevicular fluid and gingival tissue (4, 18), the production of plasmin activity would be restricted specifically to the gingiva and periodon-

VOL. 63, 1995

P. GINGIVALIS PROTEINASES AND GINGIVAL BLEEDING

tium. In contrast, gingipain-K, which is as active as plasmin in fibrinogen TT prolongation (Fig. 3), is not at all inactivated in plasma (Fig. 2), suggesting that it would be the more effective fibrinogenolytic enzyme in vivo. The fact that gingipain-R-induced TT prolongation was augmented only in albumin-depleted plasma (Table 1) clearly indicates that albumin is a specific inhibitor of fibrinogen degradation by this enzyme. It has been recently shown that human fibrinogen and albumin form a stable complex which is able to be clotted by thrombin but resistant to plasmin degradation (22). Therefore, it may well be that gingipain-R is unable to cleave fibrinogen in plasma because it cannot attack such a complex, in comparison to gingipain-K. However, this may not be the only mechanism for regulating degradation of this important clotting factor by P. gingivalis proteinases, since it is now known that gingipain-R can be slowly inhibited by the plasma proteinase inhibitor, a2-macroglobulin (11), while gingipain-K is unaffected. All of these data, taken together with the fact that gingipain-K prolonged plasma TT in a manner similar to the case for fibrinogen TT (compare Fig. 1 and 3) and was unaffected by the presence or absence of albumin (Table 1), strongly suggest that this latter proteinase must be the enzyme of choice which is involved in the degradation of this important coagulation factor. Gingipain-K and gingipain-R are associated with vesicles in most P. gingivalis strains (30). These proteinases are highly concentrated on the surface of the vesicles and, unlike their soluble forms, neither diffused nor diluted; hence, the vesicle proteinases are capable of being present at high local concentrations. In spite of these advantages, we found vesicles to be less active than equivalent concentrations of gingipain-K in the plasma TT assay (Fig. 6), indicating that this enzyme is probably less active in vesicles than soluble gingipain-K. This is also likely to be true for gingipain-R, since vesicles pretreated with z-FK-ck to eliminate gingipain-K activity were still less active in extending plasma TT than an equivalent concentration of soluble gingipain-R (Fig. 6). This is also in accord with the fact that vascular permeability enhancement activity produced by gingipain-R on vesicles is less than that of an equivalent concentration of soluble gingipain-R (15). In comparison to soluble forms, it is likely that either proteinase when anchored within vesicles is less active on large molecules such as fibrinogen, prekallikrein, and high-molecular-weight kininogen, presumably because of the inaccessability of these substrates. In summary, the present study indicates that P. gingivalis interferes with plasma clot formation by digesting fibrinogen with either vesicle-bound or secreted cysteine proteinases, especially gingipain-K. This result demonstrates a pathogenic correlation for fibrinogen digestion by the bacterial cysteine proteinases with a tendency to bleed at periodontitis sites (25, 34). Since the essential role of fibrin in wound repair is well established (6), fibrin digestion by P. gingivalis enzymes might result in retarded repair of infected and damaged sites, contributing to the development of periodontitis as a chronic disease. Both plasmin and gingipain-K cleave fibrin after lysine residues (12). Thus, it is possible that gingipain-K could produce fibrin(ogen) degradation products D, E or D-dimer, each of which enhances vascular permeability, induces leukocyte migration (8, 35, 36), and modulates leukocyte microbicidal activity and O2 release (17). The D-dimer also stimulates cells of the monocyte/macrophage lineage to secrete interleukin-1 (13). Accordingly, fibrin(ogen) digestion by gingipain-K and/or the products formed is likely to enhance the progression of periodontitis associated with the presence of P. gingivalis.

4881

REFERENCES 1. Carlsson, J., J. F. Ho ¨fling, and G. K. Sundqvist. 1984. Degradation of albumin, haemopexin, haptoglobin and transferrin, by black-pigmented Bacteroides species. J. Med. Microbiol. 18:39–46. 2. Castellino, F. J., and J. R. Powell. 1981. Human plasminogen. Methods Enzymol. 80:365–378. 3. Chen, Z., J. Potempa, A. Polanowski, M. Wikstro ¨m, and J. Travis. 1992. Purification and characterization of a 50-kDa cysteine proteinase (gingipain) from Porphyromonas gingivalis. J. Biol. Chem. 267:18896–18901. 4. Cohen, R. L., K. Watanabe, J. Schmid, J. M. Brown, and D. A. Chambers. 1992. Plasminogen activator in periodontal health and disease. Ann. N.Y. Acad. Sci. 667:183–185. 5. Doolittle, R. F. 1984. Fibrinogen and fibrin. Annu. Rev. Biochem. 53:195– 229. 6. Dvorak, H. F., A. P. Kaplan, and R. A. F. Clark. 1988. Potential functions of the clotting system in wound repair, p. 57–85. In R. A. F. Clark and P. M. Henson (ed.), The molecular and cellular biology of wound repair. Plenum Press, New York. 7. Fenton, J. W., II. 1986. Thrombin. Ann. N. Y. Acad. Sci. 485:5–15. 8. Ge, M., T. J. Ryan, H. Lum, and A. B. Malik. 1991. Fibrinogen degradation product fragment D increases endothelial monolayer permeability. Am. J. Physiol. 261(Lung Cell. Mol. Physiol. 5):L283–L289. 9. Grant, D. A., I. B. Stern, and M. A. Listgarten. 1988. Periodontics, 6th ed., p. 348–375. The C.V. Mosby Co., St. Louis. 10. Grenier, D., and D. Mayrand. 1987. Selected characteristics of pathogenic and nonpathogenic strains of Bacteroides gingivalis. J. Clin. Microbiol. 25: 738–740. 11. Grøn, H., R. Pike, J. Potempa, J. Travis, I. B. Thøgersen, J. J. Enghild, and S. Pizzo. The potential role of a2-macroglobulin in the control of trypsin-like cysteine proteases from Porphyromonas gingivalis. Submitted for publication. 12. Halkier, T. 1991. Mechanism in blood coagulation, fibrinolysis and the complement system, p. 131–160. Cambridge University Press, Cambridge. 13. Hamaguchi, M., Y. Morishita, I. Takahashi, M. Ogura, J. Takamatsu, and H. Saito. 1991. FDP D-dimer induced the secretion of interleukin-1, urokinase-type plasminogen activator, and plasmin activator inhibitor-2 in a human promonocytic leukemia cell line. Blood 77:94–100. 14. Imamura, T., R. N. Pike, J. Potempa, and J. Travis. 1994. Pathogenesis of periodontitis: a major arginine-specific cysteine proteinase from Porphyromonas gingivalis induces vascular permeability enhancement through activation of the kallikrein/kinin pathway. J. Clin. Invest. 94:361–367. 15. Imamura, T., J. Potempa, R. N. Pike, and J. Travis. 1995. Dependence of vascular permeability enhancement on cysteine proteinases in vesicles of Porphyromonas gingivalis. Infect. Immun. 63:1999–2003. 16. Imamura, T., J. Potempa, and J. Travis. Unpublished data. 17. Kazura, J. W., J. D. Wenger, R. A. Salata, A. Z. Budzynski, and G. H. Goldsmith. 1989. Modulation of polymorphonuclear leukocyte microbicidal activity and oxidative metabolism by fibrinogen degradation products D and E. J. Clin. Invest. 83:1916–1924. 18. Kinnby, B., I. Lecander, G. Martinsson, and B. Åstedt. 1991. Tissue plasminogen activator and placental plasminogen activator inhibitor in human gingival fluid. Fibrinolysis 5:239–242. 19. Lantz, M. S., R. D. Allen, P. Bounelis, L. M. Switalski, and M. Hook. 1990. Bacteroides gingivalis and Bacteroides intermedius recognize different sites on human fibrinogen. J. Bacteriol. 172:716–726. 20. Lantz, M. S., R. D. Allen, T. A. Vail, L. M. Switalski, and M. Hook. 1991. Specific cell components of Bacteroides gingivalis mediate binding and degradation of human fibrinogen. J. Bacteriol. 173:495–504. ´ witalski, and M. Ho 21. Lantz, M. S., R. W. Rowland, L. M. S ¨o ¨k. 1986. Interaction of Bacteroides gingivalis with fibrinogen. Infect. Immun. 54:654–658. 22. Lipinski, B. 1995. In vitro preparation and partial characterization of a novel complex of human fibrinogen with albumin. J. Protein Chem. 14:259–262. 23. Marsh, P. D., A. S. McKee, A. S. McDermid, and A. B. Dowsett. 1989. Ultrastructure and enzyme activities of a virulent and an avirulent variant of Bacteroides gingivalis W50. FEMS Microbiol. Lett. 59:181–186. 24. Mayrand, D., and S. C. Holt. 1988. Biology of asaccharolytic black-pigmented Bacteroides species. Microbiol. Rev. 52:134–152. 25. Nakagawa, T., S. Yamada, M. Tsunoda, T. Sato, Y. Naito, I. Takazoe, and K. Okuda. 1990. Clinical, microbiological, and immunological studies following initial preparation in adult periodontitis. Bull. Tokyo Dent. Coll. 31:321–331. 26. Nilsson, T., J. Carlsson, and G. Sundqvist. 1985. Inactivation of key factors of the plasma proteinase cascade systems by Bacteroides gingivalis. Infect. Immun. 50:467–471. 27. Nitzan, D., J. F. Sperry, and T. D. Wilkins. 1978. Fibrinolytic activity of oral anaerobic bacteria. Arch. Oral Biol. 23:465–470. 28. Pike, R., W. McGraw, J. Potempa, and J. Travis. 1994. Lysine- and argininespecific proteinases from Porphyromonas gingivalis: isolation, characterization, and evidence for the existence of complexes with hemagglutinins. J. Biol. Chem. 269:406–411. 29. Pike, R. N., J. Potempa, W. McGraw, T. H. T. Coetzer, and J. Travis. Characterization of the binding activities of proteinase-adhesions from Porphyromonas gingivalis. Submitted for publication.

4882

IMAMURA ET AL.

30. Potempa, J., R. Pike, and J. Travis. 1995. The multiple forms of trypsin-like activity present in various strains of Porphyromonas gingivalis are due to the presence of either Arg-gingipain or Lys-gingipain. Infect. Immun. 63:1176– 1182. 31. Scott, C. F., E. J. Whitaker, B. F. Hammond, and R. W. Colman. 1992. Purification and characterization of a potent 70-kDa thiol lysyl-proteinase (Lys-gingivain) from Porphyromonas gingivalis that cleaves kininogens and fibrinogen. J. Biol. Chem. 268:7935–7942. 32. Slots, J., and R. J. Genco. 1984. Microbial pathogenicity. Black-pigmented Bacteroides species, Capnocytophaga species, and Actinobacillus actinomycetemcomitans in human periodontal disease: virulence factors in colonization, survival, and tissue destruction. J. Dent. Res. 63:412–421. 33. Smalley, J. W., A. J. Birss, H. M. Kay, A. S. McKee, and P. D. Marsh. 1989. The distribution of trypsin-like enzyme activity in cultures of a virulent and an avirulent strain of Bacteroides gingivalis W50. Oral Microbiol. Immunol. 4:178–179.

Editor: J. R. McGhee

INFECT. IMMUN. 34. Southard, S. R., C. L. Drisko, W. J. Killoy, C. M. Cobb, and D. E. Tira. 1989. The effect of 2% chlorhexidine digluconate irrigation on clinical parameters and the level of Bacteroides gingivalis in periodontal pockets. J. Periodontol. 60:302–309. 35. Stecher, V. J., and E. Sorkin. 1972. The chemotactic activity of fibrinolysis products. Int. Arch. Allergy Appl. Immunol. 43:879–886. 36. Sueishi, K., S. Nanno, and K. Tanaka. 1981. Permeability enhancing and chemotactic activities of lower molecular weight degradation products of human fibrinogen. Thromb. Haemostasis 45:90–94. 37. Talonpoika, J. T., E. So ¨derling, and K. Paunio. 1993. Characterization of fibronectin and fibrin(ogen) fragments in gingival crevicular fluid. Scand. J. Dent. Res. 101:26–32. 38. Travis, J., and R. Pannell. 1973. Selective removal of albumin from plasma by affinity chromatography. Clin. Chim. Acta 49:49–52. 39. Wiksto ¨m, M. B., G. Dahle´n, and A. Linde. 1983. Fibrinogenolytic and fibrinolytic activity in oral microorganisms. J. Clin. Microbiol. 17:759–767.