WO1994005806A1

WO1994005806A1 - Method and product for diagnosis of collagen tissue destructive diseases such as periodontitis

Info

Publication number: WO1994005806A1
Application number: PCT/CA1993/000361
Authority: WO
Inventors: Jaroslav Sodek; Christopher A. G. Mcculloch
Original assignee: Jaroslav Sodek; Mcculloch Christopher A G
Priority date: 1992-09-02
Filing date: 1993-09-02
Publication date: 1994-03-17
Also published as: CA2143634A1; AU4938993A

Abstract

A method and product for the diagnosis of collagen tissue destructive diseases, such as periodontitis. A method includes collecting a sample of extracellular fluid from a potentially diseased site of a mammal. It is ascertained whether an amount of collagen-destructive enzyme in an active form is present in the sample in excess of a pre-determined threshold amount, such presence indicating the likely diseased condition of the site. The enzyme may be active gelatinase or active collagenase. Examples involving periodontitis and the use of a mouth rinse for collecting the sample are given. A capsule, being of collagen (or gelatin) is described. The capsule contains a colored substance and a wall of such thickness that when the capsule is exposed to a sample containing active collagenase (or active gelatinase) exceeding the threshold amount, degradation of the wall occurs to relase the dye from the capsule to indicate the presence of the disease at the site from which the sample was collected.

Description

METHOD AND PRODUCT FOR DIAGNOSIS OF

COLLAGEN TISSUE DESTRUCTIVE DISEASES

SUCH AS PERIODONTITIS

FIELD OF THE INVENTION

This invention relates to methods for diagnosing the presence of collagen destructive diseases, such as periodontitis, in connective tissues of mammals. Particularly, this invention relates to comparing the level of an enzyme in its active form, such as collagenase or gelatinase, in a sample containing extracellular fluid collected from a potentially diseased site relative to a pre-determined level of the enzyme in order to ascertain whether the site is diseased or not.

GENERAL INTRODUCTION AND BACKGROUND

Collagen fibers provide structural support of a tooth as part of connective tissue between the tooth and alveolar bone of the jaw. Periodontal diseases comprise a group of infections that exhibit loss of collagen in different disease types. Such loss weakens tooth attachment and can eventually lead to tooth loss. Destructive periodontal diseases are thought to affect up to 7% of the adult North American population (1). Early diagnosis of the presence of disease is desirable so that therapy can be instituted prior to the occurrence of substantial damage to the connective tissue. Diagnosis is also desirable during treatment of the disease to monitor effectiveness of therapy as such therapies are often expensive and at times painful (2).

Various approaches have been taken in recent years in addressing the need for improved diagnostic methods. These approaches measure: physical changes in the attachment of the gum tissue to the teeth after infection; etiological agents; or the response of the host to infective agents. Physical measurements are reasonably accurate in detecting evidence of at least the past presence of disease in demonstrating that destruction has occurred. Such measurements, taken over a period of a few months indicate the presence of ongoing disease. Physical measurements necessarily require the further destruction of connective tissue to indicate the presence of ongoing disease, and in this way are considered inadequate as a diagnostic method. Further, such measurements do not necessarily indicate if tissue destruction is likely to occur in the future and may thus lead to unecessary treatment.

While it is possible to assay bacterial pathogens, their presence only weakly correlates with progression of tissue destruction.

A number of tests that measure host response have been reported. For example, enzymes such as alkaline phosphatase (3), elastase (4), cathepsins (5) and beta-glucoronidase (6) have been measured in the exudate from inflamed gum tissues. Although the measured levels of these enzymes reflect increased inflammation in the gum tissues, they do not discriminate clearly between inflamed tissues that are stable (i.e. not undergoing net destruction) and tissues that are undergoing active destruction of connective tissues. Thus, it would appear that the release of these enzymes occurs as a consequence of inflammation instead of being causally linked to destruction.

Since the major structural proteins of periodontal tissues are the collagens, it would seem logical that there may be a connection between enzymes that participate in collagenolysis and disease progression. Enzymes that exhibit collagenolytic activity have thus been measured as potential diagnostic agents for monitoring progression of destructive periodontal diseases. In earlier studies, collagenolytic activity in crevicular fluid (7, 8) and extracts of gingiva (9, 10) has been related to tissue inflammation. It has also been shown that the major enzymes responsible for collagen destruction in periodontal disease are host-derived collagenase and gelatinase, which are detectable in the inflammatory exudate of diseased tissues in both an active form and as a latent, pro-enzyme form (11). Previous reports indicate that there may be an association between the amount of active collagenase and loss of attachment in dogs (8, 12) and that the active enzyme is found in higher concentrations in sites affected by localized juvenile periodontitis (11, 13, 14) and recurrent periodontitis. However, because of a lack of measurement of reproducibility and assay precision, and because of lack of appropriate methods to control assay errors, these previous reports did not provide strong support for use of collagenase and gelatinase in diagnostic testing. Further, they could be not be used to detect or predict active destruction of tissues. Consequently, their clinical utility as diagnostic tools was limited.

DETAILED BACKGROUND TO THE INVENTION

Extracellular collagen degradation is a central feature of inflammatory connective tissue lesions, although the degradative mechanisms are not well understood. Destructive lesions of rheumatoid arthritis, periodontitis and cornea (15) are associated with secretion of neutral matrix metalloproteinases (MMP), enzymes that can specifically cleave and degrade collagens at physiological pH and temperature. For example, several in vivo studies have demonstrated mRNA expression of interstitial collagenase (MMP-l) by fibroblasts and macrophages in synovial tissues (16, 17, 18) and active collagenase has been identified in synovial fluid from rheumatoid arthritis joints (19). In inflamed connective tissues, collagenolytic enzymes are also secreted by infiltrating polymorphonuclear leukocytes (MMP-8; MMP-9) and these enzymes, along with myeloperoxidase and elastase are found in the tissue or in inflammatory fluids of pulmonary fibrosis (20), interstitial lung disease (21), synovial fluid (22, 23) and periodontitis (24, 25). Currenty available evidence to demonstrate the role of neutrophil MMP's in the degradation of extracellular collagen is largely indirect and has been derived mainly from in vitro studies or from animal studies (26). However the lack of homology between the properties of human neutrophils and the neutrophils of other animals commonly used in animal models underlines the need for caution in extending insights in animals to human disease. There is little direct evidence to demonstrate simulatenously, connective tissue degradation in humans with the activity of neutrophil MMP's, in part because of the lack of readily available model systems. This has prevented precise definition of the role of neutrophil MMP's in the pathogenic mechanisms of connective tissue degradation in human diseases. In this context, periodontal diseases may provide a useful model to relate temporally, the rate of connective tissue destruction with the concentration and activity of degradative enzymes.

Periodontal diseases comprise a group of infections that exhibit well-defined and readily measurable loss of collagen in different disease types (27). In subjects with progressive lesions, the destruction of collagen can be monitored over time by repeated measurement of the level of the gingival attachment to the tooth (28). Previous studies have demonstrated that the inflammatory exudate (gingival crevicular fluid) draining from infected periodontal tissues can be non-invasively collected (11) and that the collagenolytic activity can be measured with a high degree of reproducibility (29). The concentration of total enzyme is positively associated with the volume of exudate (7) and with the amount of cumulative destruction (14). After pharmacological reduction of infection (e.g. antibiotics) or after mechanical debridement, the rate of connective tissue destruction is reduced and the collagenolytic activity decreases (13).

The destruction of collagen fibers in the extracellular matrix involves the concerted activities of neutral metalloproteinases (MMPs) which may be derived from connective tissue cells or PMNs. There is good evidence that the collagenase detected in gingival crevicular fluid is derived from emigrating polymorphonuclear leukocytes (25). The PMN collagenase and gelatinase enzymes are derived from genes distinct from those of fibroblasts and the individual enzymes can be distinguished by differences in molecular mass. The PMN procoUagenase enzyme (MMP-8) exhibits a higher molecular mass than the fibroblast enzyme (MMP-1; 75 kDa versus 57 KDa; (30)), has different substrate kinetics (31), different sequence specificity (32) and is also antigenically distinct (33). The fibroblast progelatinase (MMP-2) migrates at 72 kDa compared to the PMN progelatinase (MMP-9) which migrates at 92 kDa. In each case, the proteolytically activated forms of these enzymes migrate slightly faster than the proenzyme forms, and the latent and active forms of gelatinase can be identified by enzymography (11). Previous studies (8, 13, 14) have shown that collagenase activity can be detected in GCF of inflamed periodontal tissues and that the collagenase activity increases with the severity of inflammation (13). Moreover, the concentration of active enzyme is positively associated with the rate of GCF flow and attachment loss (8, 14), and both active and total collagenases decrease following clinical treatment of periodontal disease (13, 34). Leukocyte collagenase is released in a latent form from secretory granules (35) and both the latent and active forms are detectable in the gingival crevicular fluid (13). Activation of the proenzyme may occur by a variety of mechanisms that include catalytic cleavage and conformational change (36, 37) but the regulation of these processes is not well understood in situ (26). Although the activation of the latent to the active enzyme is theoretically an important rate-limiting step in collagenolytic degradation of connective tissues (38, 39), the demonstration of the importance of the active neutrophil enzyme in the destruction of inflamed connective tissue has not been definitively demonstrated in vivo.

Recently, a mouthrinse sampling procedure has been developed which permits rapid and non-invasive collection of GCF and facilitates accurate assessment of whole mouth collagenase and gelatinase activities (11). The mouthrinse protocol also reduces salivary contamination of GCF samples by matrix metalloproteinase inhibitors such as the tissue inhibitors of metalloproteinases (TIMPs).

SUMMARY OF RESULTS

COLLAGENASE

To assess directly the relationship between connective tissue degradation and the concentration of active and latent collagenase expression in the inflammatory exudates of human periodontal lesions, a study involving three diagnostically defined groups was undertaken. Each group consisted of subjects which exhibited either: 1) inflammation and progressive loss of bone and connective tissue detachment; 2) inflammation and bone loss without progressive connective tissue detachment; and 3) inflammation without connective tissue detachment from the tooth and without bone loss. Examination of these groups permitted discrimination between the enzymes found in lesions with only inflammation and in lesions with both inflammation and net connective tissue destruction. The study found there to be significantly increased amounts of active, but not latent, collagenase in subjects with progressive loss of connective tissue and bone over time. A temporal relationship between loss of connective tissue and level of active collagenase was found.

GELATINASE

Active and latent forms of gelatinase were measured in gingival crevicular fluid collected in mouthrinses from three diagnostically different groups of human subjects. Each group consisted of subjects which either: 1) exhibited gingivitis (G); 2) were well-treated and maintained periodontitis patients (TP) without detectable loss of attachment; or 3) exhibited recurrent loss of periodontal attachment and/or abscess formation (RP). The study found there to be significantly elevated amounts of active, but not latent, gelatinase in the TP and RP groups. A temporal relationship between loss of connective tissue and level of active gelatinase was found.

The invention thus provides a method for diagnosing the presence of collagen tissue destructive dieseases at a collagen connective tissue site of a mammal. A sample of extracellular fluid is collected from the area of the site. It is then ascertained whether the sample contains an amount of collagen-destructive enzyme in an active form exceeding a pre-determined threshold amount. An indication of the presence of disease is given if the ascertained amount exceeds the threshold amount.

The enzyme may be active collagenase or active gelatinase derived from neutrophils.

Typically the sample is obtained from the oral cavity, or mouth, of the mammal and may be collected directly from the juncture of a tooth and jaw bone or may be collected in a mouth rinse.

In the examples described below, the amount of enzyme is ascertained by determination of the level of proteolytic activity of the active enzyme.

In a particular embodiment, a visual marker, such as an intensely colored dye, encapsulated by a gelatin film is exposed to the sample. The film has a wall having a thickness such that if the amount of active gelatinase exceeds the threshold amount then degradation of the film occurs causing the dye to be released and thereby indicating the presence of disease. Alternatively, the capsule wall may be of collagen which may be degraded by active collagenase.

In another embodiment, the invention may include such a gelatin or collagen capsule for use in diagnosis of collagen tissue destructive diseases.

DESCRIPTION OF THE DRAWINGS

In the drawings,

FIGURE 1 shows collagenase activity in periodontal lesions. Active and latent neutrophil enzymes from periodontal pocket exudate were measured by functional assays in subjects with progressive net loss of connective tissue (Progressive), previous loss of connective tissue but currently clinically stable (Stable), or no net loss of connective tissue but with inflammation (Gingivitis). Mean active collagenase in progressive lesions was five fold higher (p<0.05) than stable or gingivitis subjects. Data shown as Mean ± SEM;

FIGURE 2 shows the ratio of active to total collagenase as a function of disease type;

FIGURE 3 shows active collagenase over time. Linear regression with 95 % confidence limits of active collagenase over time in subjects with progressive (A), stable (B) and gingivitis (C) lesions. Subjects with progressive loss of connective tissue exhibited steady increases of collagenase activity over time (904 CAU/day) while stable and gingivitis groups had no statistically significant increase of activity over time; FIGURE 4 shows latent collagenase over time. Linear regression with 95 % confidence limits of latent collagenase as in Figure 3. There was no net increase of latent collagenase activity over time in any of the subject groups;

FIGURE 5 shows an enzymogram analysis for active and latent gelatinase activities. An aliquot (10 μl) containing a mouthrinse sample diluted 5x in sample buffer was loaded in individual wells and the proteins electrophoresed on 12% crosslinked SDS-PAGE gels containing 40 μg/ml gelatin as described in "Material and Methods". A supernatant obtained from lysed human neutrophils (PMN) and GCF from a subject (CP.) with known gelatinase activity served as internal controls. Two aliquots of human gingival fibroblast culture media, without and with Con-A treatment (GF-1; GF-2; respectively) provided fibroblast-derived gelatinase. Gelatinase activity in three representative mouthrinse samples (1,2,3) selected from gingivitis (G), treated and well-maintained periodontitis (TP) and recurrent periodontitis (RP) groups are shown. Molecular mass markers are shown on the left. SA, serum albumin (62 kDa). ** indicates lower molecular mass (43 kDa) gelatinase activity produced by further degradation.

FIGURE 6A shows a histogram showing percentage of tests demonstrating presence of active, total and latent gelatinase activities for gingivitis (G), treated and well-maintained periodontitis (TP) and recurrent periodontitis (RP) groups. The differences of active gelatinase activity between the three groups were statistically significant (p < 0.0001).

FIGURE 6B shows a histogram of active, total and latent gelatinase activities (mean + S.E.M., standard error of the mean; in gelatinase units-U) for gingivitis (G), treated and well-maintained periodontitis (TP) and recurrent periodontitis (RP) groups obtained by enzymography. The differences between the three groups were statistically significant (p < 0.00 1) except for the latent gelatinase activity of the TP and RP; and

FIGURE 7 shows a histogram of active, total and latent gelatinase activities (mean ± S.E.M., standard error of the mean, in gelatinase units-U) before, during (trigger session) and after metronidazole treatment of recurrent periodontitis (RP) group. The differences between trigger session and after medication of active, total and latent gelatinase activities in RP group were statistically significant (p < 0.002; 0.0001; 0.0001; respectively).

MATERIALS AND METHODS

COLLAGENASE

Reagents

Type I collagen was metabolically labeled with L-[l-¹⁴C]glycine (New England Nuclear Corp.; Boston, MA) in cultures of a rat calvarial cell line (RC III 3.2; kindly provided by Dr. J. E. Aubin; University of Toronto).

Leech collagenase (CALONASE™; Calbiochem, La Jolla, CA) which produces a collagen cleavage pattern indistinguishable from mammalian collagenases was used as a laboratory reference standard (29). All other materials and reagents used were as described (29).

Study Populations

Subjects were recruited through newspaper advertising and were enrolled in the study after clinical screening. Prospective calculation of sample size (40) was used (alpha =0.05, beta=0.1, expected difference between groups =25, 000 units; standard deviation =20,000 units) to estimate a minimum of 14 subjects per group. Subjects were classified into one of three different groups according to the following inclusion criteria:

Group 1 - Progressive Periodontitis (PP)

Previous surgical treatment for periodontitis within the last 5 years and on the basis of clinic charts exhibited continued periodontal abscesses, or loss of periodontal attachment in excess of 2 mm around one or more tooth surfaces, or tooth loss because of periodontitis. After enrollment, subjects were exited from the study if antibiotics were ingested or if there were any periodontal treatment outside the Faculty of Dentistry during the study period.

Group 2 - Stable Periodontitis (SP)

This group of subjects exhibited the same inclusion and exit criteria as Group 1 except there was no detectable loss of attachment > 2mm within the last 3 years and no tough loss because of periodontitis in the last 4 years.

Group 3 - Gingival Inflammation with No Attachment Loss (G)

Subjects in this group exhibited the same inclusion and exit criteria as Group 2 patients except there was no history of connective tissue attachment loss and no periodontal treatment had been provided within the last year. However there was generalized gingival inflammation with bleeding on probing of the periodontal tissues. Exclusion criteria for all groups included confinement to a hospital or institution, history of rheumatic fever or congenital heart disease, history of renal or liver disease, blood dyscrasia or anticoagulant therapy and history of antibiotic usage within the previous 6 months. All subjects were monitored for up to 12 months. Group 1 subjects with no detectable progressive disease after 12 months of monitoring were dismissed from the study and their clinical and laboratory data were not analyzed.

Sample Collection and Processing

After obtaining informed consent, subjects were assigned a study number to facilitate masked evaluations of laboratory samples. Prior to longitudinal monitoring, all subjects in groups 1 and 2 received thorough periodontal treatment including debridement at the outset of the study. This procedure reduced inflammation and the likelihood of connective tissue attachment loss initially and also provided an opportunity for baseline measurements to be made. Subsequently it was expected that the Group 1 subjects would continue to exhibit loss of connective tissue on the basis of their past histories. Group 3 subjects were examined but were not treated. All subjects were then enrolled in a three-month disease-monitoring phase designed to verify that subjects conformed to the selection criteria, and in the date reported here, all patients conformed to the inclusion criteria throughout the study.

To assess connective tissue degradation over time, estimates of the level of the gingival attachment to the teeth were made monthly (41) and a relatively high threshold (2 mm) of attachment loss was used to ensure that type I error was minimized (p < 0.001). To assess inflammation, bleeding on probing of the gingival tissues was recorded and crevicular fluid flow rate was measured electronically (Periotron 6000; (42)). These baseline measurements were repeated every month thereafter and the first pair of baseline measurements were used to establish the magnitude of intra-examiner error.

At the time of each examination, gingival crevicular fluid was collected from all subjects at 6 specified sample teeth (43) and from teeth that exhibited loss of connective tissue attachment > 2mm. The samples of exudate were collected by micropipettes and analyzed as described (29). Aliquots (10 μl) of samples were incubated at 22°C with [¹⁴C]collagen (10 μl; 2000 dpm) and either 10 μl of distilled water to assess active collagenase activity or 3 mM p-amino phenyl mercuric acetate (APMA) to assess total collagenase activity. After incubation for 18 hours at 22 °C, reactions were terminated by the addition of 10 μl of sample buffer (0.5 M Tris-HCl, pH 6.8, 8 M Urea, 0.3 M SDS, 8% (v/v) Bromophenol Blue) and heated at 60°C for 20 min. An aliquot (15 μl) of each heated reaction mixture was resolved by 7.5% SDS-PAGE and fluorography. The fluorographs were scanned by a laser densitometer at 620 nm (UltroScan XL™, Pharmacia LKB, Baie d'Urfe, QU) and density measurements of the collagen degradation products were analyzed by the GelScan XL"" program (Pharmacia LKB) as previously described (29).

Positive controls were obtained by incubating labeled collagen substrate with 1 unit of CALONASE™ in an identical assay volume as the active colagenase assays. One collagenase activity unit (CAU) was defined as equivalent to one CALONASE™ unit per ml, and one CALONASE™ unit was defined as the amount of enzyme that produced an increase in absorbance at 520 nm of 0.0042 after incubation with lOmg AZOCOLL™ (Calbiochem) substrate for 24 h at 37°C in 25 mM Tris-HCl, pH 7.5, and 5 mM CaCl₂.

Negative controls were obtained by incubating distilled water, APMA and collagenase assay buffer in identical assay volumes as the total collagenase assays. Each batch of CALONASE™ was reconstituted for each run of assays and a standard curve of enzyme activity was constructed as previously described (29). Each assay was performed immediately after sample collection or the samples were frozen at -20 °C and assayed within 6 months of collection. Previous studies in our laboratories have demonstrated no loss of collagenase activity after storage under these conditions. All samples were anlayzed by one of us (W.L.) in a separate laboratory without knowledge of clinical measurements and patient histories.

Data Management and Statistical Analysis

Collagenase activities were expressed as CAU and were calculated from densitometry data in terms of the percentage degradation of [¹⁴C]collagen substrate alpha chains into 3/4 alpha chains. These data were converted to CAU using the volume of gingival crevicular fluid collected for each specific sample and from interpolation of the CALONASE™ standard curve. The calculation of collagenase activity took into account the collagenase activity of the positive control for each assay date and the data were normalized to the equivalent CALONASE™ enzyme activity in the positive control sample. Total enzyme activity was estimated from assays performed in the presence of APMA and active enzyme activity was estimated from samples without APMA. Latent enzyme activity was estimated by subtraction of active from total values.

All clinical and laboratory data were analyzed statistically with

SAS™ (SAS Institute Inc., Cary, NC). Means and standard errors of untransformed data were computed and linear regression was performed with the general linear models program in SAS. Comparisons of CAU values within each patient group were analyzed after logarithmic transformation (44). Preliminary analyses demonstrated that the data were normalized most effectively by this transformation. Analysis of variance of the log-transformed data set was used to compare groups and individual comparisons were performed with Duncan's Multiple Range Test or Tukey's test. GELATINASE

Patient Population

Three groups of patients were enrolled in this study: 10 patients (4F, 6M, range: 20-32 years; mean age = 26.4 years) with gingivitis (G); 10 well-treated and maintained periodontitis patients (5F, 5M, range: 32-58 years; mean age = 43.2 years) who had received surgical treatment within the previous 5 years and without detectable loss of attachment within the previous 3 years (TP); and 9 patients (4F, 5M, range: 35-67 years; mean age = 51 years) with recurrent loss of periodontal attachment (> 2mm) and/or abscess formation (RP). No pre-study sample size was calculated and the number of patients was chosen by convenience. The G group patients exhibited generalized gingival inflammation with bleeding on probing and a Gingival Index (GI; (45)) > 1. Exclusion criteria for the G group included history of periodontal abscess, periodontal attachment loss in excess of 2 mm, tooth loss due to periodontitis, and periodontal scaling, prophylaxis or surgery within the last year. Patients included in the TP and RP groups must have been surgically treated for periodontitis within the previous 5 years. Patients in the RP group must have exhibited: i) a periodontal abscess; or ii) periodontal attachment loss in excess of 2 mm around one or more tooth surfaces; or iii) tooth loss due to periodontitis within the last year. Patients in the TP group exhibited a clinically healthy periodontium with no periodontal pockets greater than 4 mm. All patients were enrolled in a three-month disease-monitoring phase which was designed to verify that all patients conformed to the above inclusion criteria.

Exclusion criteria for all patients during the study period included: confinement to a hospital or institution, history of rheumatic fever or congenital heart disease, history of renal or liver disease, hypersensitivity to metronidazole, blood dyscrasia or anticoagulant therapy. Criteria for exiting patients from the study after enrollment included development of severe or superinfection, loss of periodontal attachment greater than 2 mm in the G and TP groups, and any periodontal treatment outside the Faculty of Dentistry during the study period. The protocol used was approved by the

Human Subjects Review Committee, University of Toronto. Informed consent was obtained from patients before the study was started in September, 1989. All patients were considered to be in good health without previous history of antibiotic usage within the previous 6 months.

Clinical Protocol and Sampling Methods

All patients were examined at base-line for the following periodontal measurements: pocket depth (PD, in mm), gingival attachment level (GAL, in mm), number of sites with bleeding on probing (BLSIT), gingival index (GI; (45)), plaque index (PI; (46)), tooth mobility (MOB; by Periotest, range -5 to 50 mobility units, Siemens, FRG), and gingival crevicular fluid flow (CFF; by Periotron, range 0 to 200 units; IDE-Interstate, Amity ville, New York). CFF measurements were made on six sample teeth (16, 21, 24, 36, 41, 46) as described by Ramfjord (41). PD, GAL, and BLSIT were measured at 6 sites (MB, mid-B, DB, ML, mid-L, DL) per tooth. PI, GI and MOB were measured on each tooth. GAL was measured by a pressure-sensitive probe (Vine Valley Research, Middlesex, New York) precalibrated to provide 30 g force. Custom-fabricated, heat-polymerized acrylic splints with steering grooves were used to facilitate reproducible probe placement (41). All of the above measurements were collected monthly for a maximum of 10 months. However, not all patients provided samples for the entire 10 month period. All clinical measurements and procedures were performed by an experienced dental hygienist who had been previously calibrated during a pre-study period. The clinical investigator had no knowledge of laboratory results. In the RP group, patients were administered a course of metronidazole (250 mg, tid, for 7 days) when loss of GAL > 2mm or/ and abscess formation were detected (trigger session, only one trigger session was observed in each patient). Then, patients were re-examined at 2 and 8 weeks after the trigger session. All the TP and RP patients received periodontal maintenance therapy that included a thorough peridontal scaling, root planing and prophylaxis requiring 1 to 1 1/2 in hours, every 3 months.

GCF samples were obtained at each appointment. Patients rinsed twice with 5 ml of distilled water for 10 sec each and expectorated to remove as much saliva and debris as possible from the oral cavity. After 30 sec, patients rinsed vigorously for 30 sec with 5 ml of distilled water and the expectorate was collected in a 15 ml centrifuge tube. Tubes were frozen immediately and stored at -20 °C for enzymography.

Assay of Gelatinase

Latent and active gelatinase were assayed by gelatin-substrate enzymography as described by Heussen and Dowdle (47), and modified by Overall and Limeback (48). Briefly, discontinuous 12% (w/v) cross-linked SDS-polyacrylamide mini-slab (2.5 mm) gels containing 40 μg/ml gelatin were used. Ten μl/ml sample buffer (50 mM Tris-HCl, 0.2 M NaCl, 5 mM CaCl₂, 0.5 μg/ml Brij 35, 0.2 μg/ml NaN₃, pH 7.4) were added to 20 μl of a 5x diluted mouthrinse sample. Aliquots (10 μl) of the diluted mouthrinse were electrophoresed at 150 V for one hour, gels were rinsed twice for 10 min in 2.5% (w/v) Triton X-100 to remove SDS, and incubated in assay buffer at 37° C for 3 hours. The gels were fixed for 5 min in 10% (w/v) acetic acid and stained with Coomassie brilliant blue G 250 (49). To determine whether bacterial gelatinases were present in GCF samples and thereby identity the potential sources of enzyme, in some experiments gelatinase activity was assayed following a 20 min activation of the diluted sample with 10 μl, of 1 mM p-aminophenylmercuric acetate (APMA) at pH 7.5. The APMA activates the latent gelatinase derived from fibroblasts and PMNs but inhibits the microbial gelatinases.

Gelatinase activity was detected as a clear band against a blue-stained background and was quantitated by laser densitometry at 633 nm (Ultrascan II, Pharmacia) within the linear range of the densitometric response. The area under the inverted peaks was integrated using customized software (Curves; written by Dr. P.N. Lewis; Department of Biochemistry, University of Toronto) run on a Macintosh computer. Gelatinase activity was measured in arbitrary units set automatically by "Curves" . The activity obtained from the activation of the latent progelatinase by the SDS, was obtained by subtracting the values obtained from the active gelatinases from the total MMP gelatinase activity. To standardize the analyses, a fixed aliquot of PMN gelatinase was run on each gel and the densitometric values were corrected for any variation.

To evaluate the cellular sources of the gelatinase bands, a GCF sample from a subject (CP.) with known gelatinase activity, a sample of lysed human neutrophil (PMN) supernatant (50), and aliquots of culture medium from human gingival fibroblast cultures, with and without Con- A (51) treatment (GF-2; GF-1; respectively) were analyzed by enzymography. The gelatinase in aliquots of the supernatant from lysed human neutrophils also served as an internal standard of enzyme activity. The major gelatinase activities were characterized further using various proteolytic enzyme inhibitors as described in detail previously. Thus, the major band at 92 kDa was identified as the progelatinase MMP-9, whereas the faster migrating band at 84 kDa was identified as the activated form. Quantitation of the latent and active gelatinases was performed by one investigator (Y.T.) without knowledge of the clinical data. Assay Reproducibility Test

To assess the reproducibility of the assay procedures, 10 randomly selected mouthrinse samples from all three groups of patients were analyzed. To assess intergel (SDS-PAGE) variation, the same assay sample was loaded onto two different gels. To assess inter-assay variation, the same sample was analyzed in different preparations of sample buffer loaded onto the same gel. To assess inter-scan variation, 10 randomly selected assays with marked active gelatinase activity were scanned for a second time without knowledge of the previous results. To assess diurnal variations, replicate mouthrinse samples were collected from 5 subjects 3 minutes apart and repeat samples were collected 2 hours and 4 hours later for a total of 30 samples. Each sample was assayed and scanned twice.

Statistical Analysis

Parametric analyses of the untransformed data set were performed using SAS™ (SAS Institute Inc., Cary, NC). The individual mean values for each clinical parameter were calculated per patient. Group means were then computed from individual patient means collected for the total number of appointments (equal to the number of mouthrinse samples). To obtain estimates of the most severely involved site within each patient, the highest single measurement for PD, PI, GI, MOB, CFF, were recorded at each session and then group means of these peak values were computed. To analyze presence/absence data of active, latent and total gelatinase activity, the chi-square test was employed. The paired t test and the paired sign test were employed for significance testing of assay reproducibility and the effectiveness of metronidazole in reducing gelatinase activity. Repeated measures analysis of variance was employed to examine diurnal variation. Both whole mouth patient means and peak values of clinical measurements were analyzed in testing statistical significance. For each group, the mean values of enzyme activities were computed and included all samples over the 10 month study period. ANOVA was employed to test for differences between groups. Multiple step-wise regression analysis was used to assess the association between the dependent variable (gelatinase activity) and the independent variables (clinical parameters). Data from the RP group of patients were separated into sessions before, during (trigger) and after loss of GAL or abscess formation to evaluate temporal variations of gelatinase activity within patients.

RESULTS

COLLAGENASE

Clinical Measurements

A total of 58 subjects were enrolled in the study (Group 1-N = 14, Mean age ± s.e.m. +49.3 ±3.0 years; Group 2-N=27, age=51.3±2.1 years; Group 3-N = 17, age=28.5±1.4 years). Only subjects in Group 1 with progressive lesions exhibited loss of connective tissue attachment (4.3 ±0.5 mm) and the average time from the start of monitoring to the time of >2mm connective tissue loss was 204 ± 35 days. Crevicular fluid flow, a measure of the extent of inflammation, was high in all groups and was significantly (p<0.G01) lower in subjects with inflammation and no loss of connective tissue than in subjects with stable or progressive lesions

(Group 1 = 15.7± 1.3; Group 2=25.2±0.8; Group 3=30.5 ± 1.5).

Enzyme Activity

Untransformed data of collagenase activity from all 3 groups of subjects exhibited a markedly skewed distribution with a long and flat tail towards higher CAU values. After logarithmic transformation, analysis by the Shapiro-Wilk test (52) demonstrated a good fit to a normal distribution (W=0.8). Previous analyses of assay reproducibility have shown that only 8% of the variation of collagenase activity can be attributed to operator and assay errors using the methods employed here (29) while the remaining 92% is due to real biological variation.

Comparisons of the active and latent collagenase levels in the three groups of subjects demonstrated that for all sampling sessions taken together, subjects with progressive loss of connective tissue exhibited about 5 fold higher levels (p <0.05; Fig. 1) of active collagenase than subjects without loss of connective tissue attachment. In individual subjects who were monitored over time, active collagenase levels exhibited apparently random fluctuations. Correlation analysis of replicate measurements of active collagenase made one month apart demonstrated wide variations (Group 1 : r=0.32; Group 2: r=0.50; Group 3: r=0.02). At the time of detection of connective tissue attachment loss, individual subjects with progressive lesions exhibited large and discrete increases of active collagenase levels. Overall, a 40% increase in active collagenase activity (1.11 ±0.29X10⁵ collagenase units-pre-breakdown; l.Sό±O.όβxlC^-breakdown) was observed in the Group 1 patients. No similar trends were observed in any of the subjects in Groups 2 and 3.

In contrast to the variations of active enzyme between groups, latent enzyme levels were increased between the progressive to stable to gingivitis groups (Fig. 1). This relationship between connective tissue loss and the relative abundance of latent and active enzymes was also examined by computing the ratio of active to total enzyme activity (total = latent + active). As some samples exhibited no latent enzyme, it was not possible to compute the active to latent enzyme ratio. The ratio of active to total enzyme increased significantly (p< 0.001, Figure 2) with the extent of connective tissue degradation (Group 1 : 0.243 ±0.017; Group 2 : 0.11 + 0.005 ; Group 3 : 0.124 ±0.13) and was more than double in subjects with progressive disease.

To assess temporal variation of collagenase activity between groups, linear regression of collagenase activity was performed for each group (Fig. 3). Subjects with progressive disease exhibited a linear increase of active collagenase with time and collagenase activity increased at a rate of 904 units per day. In contrast, individuals with stable periodontal lesions exhibited increases of activity of only 32 units per day. To compare the relative importance of latent and active enzymes, linear regression of latent enzyme activity was also plotted. In comparison to active enzyme, latent enzyme exhibited no statistically significant increase of activity with time in any of the study groups although the absolute levels of the latent enzyme were 2-3 fold higher in subjects with inflammation but no loss of connective tissue (Fig. 4).

GELATINASE

In this study, every patient had at least 23 teeth. No patients were exited from the G and TP groups and none of these patients exhibited attachment loss or abscess formation. All patients in the RP group exhibited attachment loss or abscess formation at one or more measurement sessions (trigger session). A total of 184 mouthrinse samples and 184 sets of clinical replicate measurements (35 from G, 90 from TP, 59 from RP) from 29 patients over a 10 months period were analyzed. The mean values of all clinical measurements obtained over 10 months demonstrated that the mean and peak PI and GI were higher in the G group than the TP and RP groups (p < 0.0001) whereas the mean and peak PD, MOB, and CFF were highest (p < 0.0001) in the RP group (Table 1). Loss of GAL occurred only in the RP group (mean = 2.4 ±0.4 mm). BLSIT was higher in the G and RP groups (Table 1) than the TP group and was not simply the result of unequal numbers of teeth per patient between groups. Thus, the clinical periodontal parameters obtained from the individual patients were consistent with the original assignation of these patients into the G, TP and RP groups respectively.

Based on the migration patterns of GCF gelatinase on SDS-PAGE and the effects of APMA activation, the major GCF enzymes were identified as the latent 92 kDa progelatinase (MMP-9) that co-migrated with progelatinase from lysed neutrophils, and the faster migrating activated forms (p-gelatinase and gelatinase; (53, 54)). In these samples there was no evidence of fibroblast-derived gelatinase (Fig. 5) which migrated at 68 kDa in these gels. Lower Mr bands (43 kDa or lower) of gelatinase activity were also observed in some samples (Fig. 5). However, Chi-square analysis demonstrated no significant difference (p > 0.87) between the frequency of occurrence of the lower M_r gelatinase activity in the three patient groups.

Reproducibility testing demonstrated no significant difference (paired t test; p > 0.60; p > 0.08; p > 0.36, respectively; paired sign test, p > 0.025, one tailed) in the inter-gel, inter-assay, and inter-scan results.

Repeated measures analysis of variance for the diurnal variation demonstrated that there was no significant difference (p > 0.11) between time, test and within-subject effects. In the test of diurnal variation, we observed a slight increase of the active enzyme activity at the second test period (2 hour sampling) for most of the test samples. However, this difference was not statistically significant.

The mean level of active gelatinase activity for all measurement sessions was significantly higher (p < 0.001; ANOVA) in the RP group (71,000 U) compared to the TP group (43,814 U). Both of these groups were higher (p< 0.001) than the G group (2,824 U; Fig. 6). The mean values of the total and latent gelatinase activities were also significantly higher (p < 0.001) in the TP and RP groups compared to the G group (Fig.6). During the trigger sessions, the mean value of active gelatinase was 129,414 U (Fig.6) which was about 2 fold greater than the mean values of the TP group and the "before trigger" session of the RP group.

Analysis of replicate samples (35 from G, 90 from TP, 59 from RP) demonstrated the presence of active gelatinase in 11.4% of G, 97.8% of TP and 86.4% of RP samples (Fig.6). Parallel with the parametric analysis, chi-square analysis demonstrated a highly significant difference (p < 0.0001) between the G and TP/RP groups based on the presence or absence of active gelatinase. Thus, presence/absence analysis provided discrimination between gingivitis and periodontitis but not between inactive and active periodontitis.

Loss of GAL was observed only in the RP group and was more strongly associated (r = 0.52) with active gelatinase than either total gelatinase (r = 0.41) or with latent gelatinase (r = 0.27) levels (Table 2). Active, total and latent gelatinase levels were weakly associated with clinical peak PD measurement (r=0.34, r=0.28, r=0.20, respectively; Table 2). The peak PD (r=0.46, p < 0.0001) was associated more strongly with loss of GAL than the mean MOB (r=0.36, p < 0.0001). The number of sites with bleeding on probing (BLSIT) was associated more strongly with mean PD (r = 0.59, p < 0.0001; data not shown) than with any other measurements. Multiple regression analysis also demonstrated a significant association between active gelatinase and loss of GAL (r=0.52, p < 0.0001), and mean MOB (r=0.31, p < 0.03). Other clinical measurements (PD, PI, GI, CFF, BLSIT) were not significantly (p > 0.05) associated with active gelatinase.

Patients in the RP group that exhibited loss of GAL in excess of 2 mm and/or abscess formation were treated with metronidazole and demonstrated 4-to 6-fold reduction of enzyme activities two to eight weeks after the trigger session (active: p < 0.002; total: p < 0.0001; latent: p < 0.0001; Fig. 7). DISCUSSION

COLLAGENASE

The human periodontal disease model system permits direct in vivo evaluation of the role of MMP-8 (25) in connective tissue degradation at isolated sites and at frequent time intervals. The approach facilitates temporal study of the relationship between collagen degradation and neutrophil collagenase activity. As the collagenase is present in large quantities in exudate draining from affected sites, it is possible to collect samples containing the collagenase non-invasively and thereby allow frequent sampling that does not perturb local degradative processes. Thus, it has been possible to identify precisely those time intervals during which connective tissue degradation has occurred and to relate these same time intervals to periods of increased enzyme activity. In subjects with progressive loss of connective tissue, there were large increases of active enzyme that were contemporaneous with loss of connective tissue at affected sites. Further, subjects that exhibited progressive loss of connective tissue exhibited several-fold higher levels of active enzyme than did subjects with no loss of connective tissue. These differences of active enzyme levels were found in spite of the presence of inflammation in both groups of subjects. As collagenase activity was computed in terms of activity per unit volume, the variations of exudate flow rate between groups would not explain the 5-fold higher levels of active collagenase nor the very rapid loss of connective tissue in the subjects with progressive lesions. Our findings that relate connective tissue degradation and enzyme activity are consistent with previous human studies (29) in which administration of anti-microbial drugs in combination with debridement reduced active collagenase levels 3-fold and also blocked further loss of connective tissue. These data indicate that reduction of microbial colonization of the tooth surface blunts inflammatory mediators of connective tissue destruction, including those affecting the activation of neutrophil collagenase.

The utlization of a functional assay (55) allows discrimination between the latent (proenzyme) and active collagenases in gingival crevicular fluid. Unlike earlier reports (30) in which total enzyme activity was measured after in vitro activation, the approach used here assessed both the latent and the active enzyme. This analysis has demonstrated the relative importance of the active compared to the latent neutrophil enzyme in connective tissue degradation. Indeed, the relatively high levels of latent enzyme were much more strongly associated with inflamed lesions without connective tissue loss than with progressive lesions, indicating that release of the latent enzyme alone is not associated with destruction. Further, the finding that inflammation is a necessary but not a sufficient condition for connective tissue degradation in the subjects examined here, points to the presence of enzyme inhibitors, activators and other regulatory mechanisms that control the activity of degradative enzymes and that are possibly independent of the inflammatory process.

A large body of data has indicated that neutrophils secrete MMP's in a latent form (56) and that activation mechanisms including chlorinated oxidants released by neutrophils (26) are then responsible for conversion of the latent to the active form. Activation appears to occur by a stepwise mechanism through which sequential processing events occur in the propeptide region (57, 58). Availability of activator enzymes (59) or of HOCl (56) to cleave the N-terminal propeptide of MMP's and to thereby displace Cys⁷³ from the zinc atom in the catalytic site is likely a rate-limiting factor (60). Once the latent collagenase is activated, a wide variety of inhibitors including alpha₂-macroglobulin and the tissue inhibitors of the metalloproteinases (TIMPs) may also regulate enzyme activity. However, recent evidence implicating neutrophil-inactivation of alpha_:-macroglobulin (61) and of neutrophil elastase inactivation of TIMP (62) indicate that the neutrophil in inflamed connective tissues possesses a number of mediators that can overcome the anti-proteinase screen that normally regulates collagenase activity (26).

As enzyme activity is dependent on levels of synthesis, inhibition and activation (63), it is evident that in the periodontal disease model system, the level of active neutrophil enzyme, as reflected in the amount of active enzyme detected in a sample collected from a subject, is an important factor in the rate of connective tissue degradation. Our data show that only in individuals with rapid increases of enzyme activity over time was there demonstrable loss of connective tissue, suggesting a type of threshold, which when exceeded leads to connective tissue loss.

To determine the level of threshold, we computed the active collagenase level that would be expected at day 204, the mean time period from treatment to the time when connective tissue destruction was detected. Assuming linear increases in the rate of collagenase production per day in all subjects with progressive lesions, we calculated tiiat sites with greater than 184,500 units of collagenase activity per day (calculated from the increase of 904 units/day over the 204 days shown in Figure 3) will exhibit loss of connective tissue. To test whether this calculated value was consistent with data from independent studies, we calculated the mean plus the 95% confidence limit of the enzyme activity of progressive sites from an earlier separate study that used a completely different group of subjects and who exhibited very similar destructive lesions of periodontal connective tissue (N=30; (29)). This value (198,030 collagenase units) was recorded at the time of active destruction and is reasonably close to the value computed in the present study. The concordance of these data supports the existence of a threshold level of about 150,000 to 200,000 collagenase units, above which progressive loss of connective tissue occurs. A second threshold level of about 75,000 collagenase units, can be estimated, below which the presence of disease is very unlikely.

As the cleavage of the native collagen molecule by neutrophil collagenase is the initial step in collagen degradation, the activity of MMP-8 is of considerable significance in inflammatory diseases of connective tissues.

Our results show that active collagenase and not latent collagenase in exudates is temporally linked to destruction. If the degradative mechanisms in periodontal diseases are common in other inflammatory connective tissue diseases, then elucidation of the activation mechanisms of the proenzyme could provide novel approaches to therapeutic control of these diseases.

GELATINASE

The major gelatinase activity in GCF collected by the mouthrinse procedure used in this study appears to be derived from neutrophils and not from fibroblasts, as described in greater detail elsewhere. This finding is in agreement with recent work demonstrating that collagenolytic activity obtained from inflamed human gingival extracts (24), gingival crevicular fluid (11, 24) and whole saliva (64) represents the neutrophil collagenase. The fibroblast-derived collagenase (MMP-1) can be differentiated from neutrophil-derived collagenase (MMP-8) by molecular mass (24, 35, 54, 55, 64, 65), mechanisms of activation (24, 35, 64, 66), antigenic properties and substrate specificity (67). Discrete differences have also been demonstrated between the 72 kDa fibroblast gelatinase (MMP-2) and the 92 kDa neutrophil gelatinase (MMP-9) (53). Although the 92 kDa gelatinase can be expressed by other cell types including endothelial and epithelial cells, the MMP-8 collagenase is characteristic of PMN's and is not known to be expressed by these other cells. Notably, the PMN gelatinase is believed to be released from a cytoplasmic, peroxidase-negative, third granule compartment separate from the intracellular source of collagenase (68). Although the activation mechanisms of the neutrophil-derived collagenase and gelatinase in vivo are still unclear, it is conceivable that the release of both enzymes upon degranulation and their activation may be crucial regulatory steps at different phases of periodontal inflammation and destruction.

Collagenolytic enzymes are also produced by microbial strains such as Porphyromonas gingivalis (69) that have been associated with periodontitis, and microbial collagenolytic enzymes may play a role in the migration of bacteria through the dense periodontal connective tissues (64, 69). Thus it is conceivable that some of the gelatinases observed on the enzymograms could be derived from bacterial and not neutrophil enzymes. However, compared to host-derived MMP's, the microbial proteases: i) are not inhibited by alpha₂-macroglobulin; ii) are inhibited by APMA (93%) and NEM (N-ethylmaleimide, 60%); iii) exhibit little ability to cleave N-terminal nonhelical telopeptides; and iv) demonstrate dual dependence on free thiol groups (-SH) and metal ions for catalysis (69). In contrast, our results showed activation of the latent gelatinase with APMA. Moreover, since the bacterial enzymes are present at low concentrations even in pure cultures it is likely that microbial gelatinase activities were not measured in this assay.

The 92 kDa progelatinase was activated by SDS and was detected as a band of gelatinase activity. However, the specific activity of the progelatinase on the enzymograms may be lower than that of the activated forms, as observed with the 72 kDa latent gelatinases (65). Therefore, the amount of latent gelatinase relative to active gelatinase may be much higher than measured by the functional assays. Further the high levels of latent gelatinase may result in part from lysis of PMNs in the water rinse. This might be avoided in future studies by using isotonic rinses. Activation of the latent collagenolytic enzymes that occurs in association with periodontal disease may result from the action of chlorinated oxidants (26) generated by PMN myeloperoxidase activity, or through the activity of bacterial metabolites such as methyl mercaptan and hydrogen sulphide (70, 71), which can activate latent collagenase. In the gingivitis group, the majority of the gelatinase activity was in the latent gelatinase form (92 kDa progelatinase, MMP-9), a finding consistent with previous studies demonstrating that latent collagenolytic enzymes are generally present at mildly inflamed or clinically healthy sites (8, 11).

Based on multiple regression analysis, active gelatinase was most strongly associated with loss of GAL and to a lesser extent with mean tooth mobility, whereas total and latent gelatinase activities were not associated closely with any clinical parameters of periodontitis. Previous analyses of active collagenase have demonstrated that the amount of active enzyme is positively associated with loss of attachment in dogs (8, 12) and is also found in higher concentrations in sites affected by localized juvenile periodontitis (11, 13, 14), adult periodontitis (11, 14) and recurrent periodontitis (29). The statistical relation between active gelatinase activity and mean tooth mobility in this study indicates that inflamed periodontal tissues with attachment loss are most likely associated with loose teeth, consistent with previous work by Hakkarainen et al. (34). Further, the statistical associations between active gelatinase and loss of GAL at a single site indicate that the whole mouth enzyme test may be capable of detecting periodontal disease destruction at a single site. Previously, a relationship between active periodontal disease and collagenase activity at single sites has been observed (8, 14).

The increased mean gelatinase activities reflecting active and latent forms of MMP-9 observed in me G, TP and RP groups in association with the increased severity of disease, together with the close temporal relationship between loss of gingival attachment level (GAL) and increased amount of active gelatinase during the trigger session, indicate that fluctuations of enzyme levels may reflect the loss of homeostasis of host defense mechanisms. In conjunction with previous reports (8, 11, 29), our findings support the existence of a host-threshold for maintenance of gingival attachment level in the presence of inflammation. This relationship can be modelled by constructing a 2 x 2 table using the mean active gelatinase level of the RP group as a threshold to define a positive test result. Sequential monitoring of RP patients demonstrated that when the threshold cut-off point was exceeded in mouthrinse samples, loss of GAL occurred in 75% of these patients, 25% greater than that expected by chance alone. This model is summarized in Table 3 and indicates that when a post hoc estimate of a positive test result is computed from the mean active gelatinase level, there is a relatively high positive predictive value (75%). Thus, this model indicates that detection of active gelatinase above a certain threshold provides reasonably good prediction of progressive periodontitis in high risk patients. However, there were significant age differences between the three study groups (youngest in G group, oldest in RP group) and the higher levels of active gelatinase and increased frequency of loss of gingival attachment in the RP group could be in part age-related. Previous cross-sectional and longitudinal studies (72, 73, 74) of the natural history of periodontitis have demonstrated the dependence of disease severity on aging and this phenomenon could have contributed to increased gelatinase activity in the RP group.

Several previous reports (75) have demonstrated the effectiveness of metronidazole treatment in reducing probing depth and increasing attachment level in deep pockets with anaerobic infections. Metronidazole appears to reduce the prevalence of Porphyromonas gingivalis and spirochetes in active sites of recurrent periodontitis subjects (75). The results presented here indicate that metronidazole treatment also results in the reduction of both active and latent forms of gelatinase. The reduction of gelatinase activity is best explained by the antimicrobial effect of metronidazole which appears to blunt destructive host immune and inflammatory responses. Therefore, the gelatinase assay could be used to monitor the efficacy of therapeutic procedures.

The results of this study indicate tiiat the mouthrinse protocol for collection of GCF combined with the gelatinase (MMP-9) assay can provide simple and noninvasive methods for studying potentially useful biochemical markers of gingival crevicular fluid. Since gelatinase activity was detected in highly diluted moud rinse samples of GCF, it is apparent that enzymography provides an extremely sensitive assay for gelatinase. Although the test in its present form is not ideal for roμtine clinical use, it avoids the use of radioactive substrates that are used in other assays for detection of collagenase activity. Moreover, the analyses of the intergel, interassay and interscan variations indicate that the reproducibility of the gelatinase assay is high and that diurnal variation is small. Thus, analysis of active gelatinase in mouthrinse samples may provide a sensitive indicator of periodontal tissue destruction that could facilitate detection of active periodontal lesions.

In a particular embodiment the results presented herein may be used for an easy-to-use kit for diagnosing the presence of periodontitis in a human subject by use of a sample obtained using a mouth rinse. The sample would be incubated with either a native collagen (for active collagenase activity) or a gelatin (for active gelatinase activity) substrate prepared as a film of defined thickness, encapsulating an intensely colored dye, for example. The capsule thickness would be such that if the amount of active enzyme in the sample exceeds a threshold amount, then degradation of the substrate by the enzyme contained in the sample would cause the dye to be released into the incubating solution to give a positive determination of the presence of a destructive collagen tissue disease. In such a test the degradation of periodontal tissues is mimicked in a test-tube, utilizing relatively inexpensive reagents. The procedure for collecting samples is non- invasive (i.e. by a mouthrinse with a physiological buffer solution, for example) and entails no contact of chemicals other those used in the rinse solution with the patient. The simplicity of the approach, the current high cost of dental care and the potential for self-diagnostic use in the home, suggest that this diagnostic procedure, in the form of a home test kit, would be desirable.

As alternatives to a colored dye, other techniques for enzyme activity assay may be used. For example, a film may encapsulate a substance which is itself a substrate of the enzyme being assayed, action of the enzyme upon the substrate producing a detectable indicator. For example, an amide or polypeptide or other hydrolyzable substrate which produces a colored solution, or photometrically detectable product, upon hydrolysis catalyzed by the enzyme may be used. It may be possible that such a substrate may be used without the need for a film, if the enzyme whose activity is being assayed is selective enough for the substrate.

According to an alternative approach, a substrate has a core coated with collagen (or gelatin) in which the coating is of such a thickness that when the substrate is exposed to a sample collected from a diseased patient for a period of time, the coating is degraded to expose the core to indiciate the presence of disease.

An alternative embodiment kit may include a core coated with gelatin or collagen, the particular coating being of varying thickness. The degree of degradation of the gelatin (or collagen) coating due to the presence of active gelatinase (or collagenase) would correlate with the amount of active gelatinase in the sample. The core could be colored or marked such that the areas of degradation would be visible. Areas of thicker coating would be degraded in a given period of time only if sufficient gelatinase (or collagenase) were present, thereby giving an indication of the amount of gelatinase (or collagenase) present in the sample.

An alternative to assays which depend upon detection of enzyme action, such as the above embodiments, is assay of the enzyme itself. A competitive protein binding assay may be used.

Radioimmunoassays (RIAs) or immunoradiometric assays (IRMAs) could be used, but these of course would be limited to use in a clinical setting, radioisotopes and their detection equipment being required for such procedures.

An enzyme-linked immunosorbent assay (ELISA) would have in common with RIAs and IRMAs a relatively high degree of sensitivity, but would generally not rely upon the use of radioisotopes. A visually detectable substance may be produced or at least one detectable in a spectrophotometer.

An assay relying upon fluroescence of a substance bound by the enzyme being assayed could be used.

Of course, it would be possible to combine the approaches of determination of enzyme action and assay of the enzyme itself. For example, a binding component of an ELISA assay may be contained within a collagen (or gelatin) film, which component would be released only after the film had been degraded by active collagenase (or gelatinase).

It will be appreciated that there are a number of reporter systems which may be used, according to the present invention, to detect active gelatinase or collagenase, or either or both of their respective activities, in determining whether one or the other or both of the active enzymes is present above or below a threshold amount to diagnose collagen destructive diseases.

Given the results described herein, a skilled person would be capable of developing other assays and methods for determining whether a sample collected from a potentially diseased tissue site contains an active enzyme exceeding a threshold amount to thereby determine if a diseased condition exists in the tissue.

Table 1: Mean values of all clinical parameters of three study groups

G TP RP

N = 35 N = 90 N = 59

MEAN S.E. MEAN S.E. MEAN S.E.

GAL 2.4 + 0.4

BLSΓΓ 14.4 ± 2.2 8.2 ± 0.9 19.6 + 2.5

PD in mm, MOB in Periotest units, CFF in Periotron units

GAL in mm, BLSIT⁰ = No. of sites with bleeding upon probing

S.E. = standard error

N = No. of mouthrinse samples

Note that not all patients provided samples for the entire ten months 37 -

Table 2: Multiple correlation analysis of gelatinase activity and clinical measurements

Boxes marked with asterisk (*) are correlations with P > 0.05 and are not reported. Table 3: Post hoc estimates of diagnostic test values of the RP group

Reference Standard

(Loss of GAL in RP)

> 2mm <_ 2mm total Sen = 78.9% Spe = 87.5%

Ppv = 75%

Test > mean 15 5 20 Npv = 89.7%

(active gelatinase

≥_ mean) < mean 4 35 39

total 19 40 59

Sen = Sensitivity, Spe = specificity, Ppv = positive predictive value, Npv = Negative predictive value, mean = mean of active gelatinase activity of RP group (=71,006 U)

LIST OF ABBREVIATIONS^*

APMA p-aminophenylmercuric acetate

BLSIT number of sites with bleeding on probing

CFF gingival crevicular fluid flow G patients exhibiting gingivitis with no attachment loss

GAL gingival attachment level

GCF gingival crevicular fluid

GI gingival index

MMP neutral matrix metalloproteinase MMP-1 procoUagenase from fibroblasts

MMP-2 progelatinase from fibroblasts

MMP-8 procoUagenase from connective tissue cells

MMP-9 progelatinase from connective tissue cells

MOB tooth mobility PD pocket depth

PI plaque index

PMN connective tissue cell

PP patients with progressive periodontitis

RP patients who experienced recurrent loss of periodontal attachment SP patients who exhibited stable periodontitis

TP maintained periodontitis patients

A more detailed definition may be given in the text at the point of first usage. REFERENCES

1. Miller et al. , N.I.H. Survey of Oral Health in U.S. , 1987.

2. Listgarten, J Clin Periodontol 1980; 7: 165.

3. Ishikawa and Cimasoni, Arch Oral Biol 1970; 15: 1401. 4. Palcanis, et al., J Periont 1992; 63: 237.

5. Ishikawa et al. , Arch Oral Biol 1972; 17: 111.

6. Lamster et al. , J Periodontol 1985; 56: 139.

7. Golub LM, Siegel K, Ramamurthy NS, Mandel ID, J Dent Res 1976; 55: 1049-1057. 8. Kryshtalskyj E, Sodek J, Ferrier JM, Archs Oral Biol 1986; 31:

21-31.

9. Uitto et al., J Periodont Res 1981; 16: 417.

10. Overall et al., J Dent Res 1983; 62: 667.

11. Gangbar S, Overall CM, McCulloch CAG, Sodek J, J Periodont Res 1990; 25: 257-267.

12. Kryshtalskyj E, Sodek J, J Periodont Res 1987; 22: 264-269.

13. Larivee J, Sodek J, Ferrier JM, J Periodont Res 1986; 21: 702-715.

14. Villela B, Cogen RB, Bartolucci AA, Birkedal-Hansen H, J Periodont Res 1987; 22: 209-211, 381-389. 15. Harris ED, Krane SM, N Engl J Med 1974; 291: 557-661.

16. Firestein GS, Paine MM, Liftman BH, Arth Rheum 1991; 34: 1094-1105.

17. Gravellase EM, Darling JM, Ladd AL, Katz JN, Glimcher LH, Arth Rheum 1991; 34: 1076-1084. 18. McCachren SS, Arth Rheum 1991; 34: 1085-1093.

19. Harris Jr ED, Cohen GL, Krane SM, Arth Rheum 1969; 12: 92-102. 20. Cantin AM, North SL, Fells GA, Hubbard RC, Crystal RG, J Clin Invest 1987; 79: 1665-1673.

21. Weiland JE, Garcia JG, Davis WB, Gadek JE, J Appl Physiol 1987; 62: 628-633. 22. Edwards SW, Hughes V, Barlow J, Bucknall R, Biochem J 1988; 250:

81-85.

23. Konttinen YT, Otso L, Kemppinen P, Saari H, Suomalainen K, Vauhkonen M, Lindy S, Sorsa T, Matrix 1991; 11: 296-301.

24. Sorsa T, Uitto V-J, Suomalainen K, Vauhkonen M, Lindy S, J Periodont Res 1988; 23: 386-393.

25. Overall CM, Sodek J, McCulloch AG, Birek P, Infect and Immun 1991; 59: 4687-4692.

26. Weiss SJ. Tissue destruction by neutrophils, N Engl J Med 1989; 320: 365-376. 27. Page RC, Schroeder HE, A summary of current work. Lab Invest 1976; 33: 235-249.

28. Goodson J, Tanner ACR, Haffajee AD, Sornberger GC, Socransky SS, . J Clin Periodont 1982; 9: 472-481.

29. Lee WS, Aitken S, Kulkarni G, Birek P, Overall CM, Sodek J, McCulloch CAG, J Periodont Res 1991; 26: 479-485.

30. Murphy G, Reynolds JJ, Bretz U, Baggiolini M, Biochem J 1982; 203: 209-221.

31. Hasty KA, Jeffrey JJ, Hibbs MS, Welgus HG, J Biol Chem 1987; 262: 10048-10052.

32. Netzel-Arnett S, Fields GB, Birkedal-Hansen H, Van Wart HE, J Biol

Chem 1991; 266: 6747-6755.

33. Hasty KA, Hibbs MS, Kang AH, Mainardi CL, J Exp Med 1984; 159: 1455-1463.

34. Hakkarainen K, Uitto V-J, Ainamo J, J Periodont Res 1988; 23: 204-210. 35. Hasty KA, Hibbs MS, Kang AH, Mainardi CL, J Biol Chem 1986; 261: 5645-5650.

36. Grant GA, Eisen AZ, Marmer BL, Roswit WT, Goldberg GI, J Biol Chem 1987; 262: 5886-5889.

37. Witham SE, Murphy G, Angel P, Rahmsdorf H-J, Smith BJ, Lyons TJR, Harris JJR, Herrlich P, Docherty AJP, Biochem J 1986; 240: 913-916.

38. Weiss SJ, Peppin G, Ortiz X. Ragsdale C, Test ST, Science 1985; 227: 747-749. 39. Unemori EN, Bair MJ, Bauer EA, Amento EP, J Biol Chem 1991; 266: 23477-23482.

40. Pocock SJ, Clinical Trials: A Practical Approach. John Wiley & Sons Ltd, London 1983; 266pp.

41. McCulloch CAG, Birek P, Overall C, Aitken S, Lee W, Kulkarni G, J Clin Periodontol 1990; 17: 616-622.

42. Hinrichs JE, Bandt CL, Smith JA, J Periodont 1984; 55: 294-298.

43. Ramfjord SP, J Periodont Res 1974; 9 (suppl 14): 78-93.

44. Zar JH, Biostatistical Analysis 1974; Prentice-Hall, Inc., Englewood Cliffs, NJ, 620pp.

45. Lδe H, Silness J, Ada Odont Scand 1963; 21: 533-551.

46. Silness J, Lδe H, Acta Odont Scand 1964; 24: 747-756.

47. Heussen C, Dowdle EB, Anal Biochem 1980; 102: 196-202.

48. Overall CM, Limeback H, Biochem J 1988; 256: 965-972.

49. Neuhoff V, Stamm R, Eibl H, Electrophoresis 1985; 6: 427-448.

50. Weiss SJ, Klein A, Slivka A, Wei M, J Clin Invest 1982; 70:

598-607.

51. Wang HM, Hurum S, Sodek J, J Periodont Res 1983; 18: 149-155.

52. Shapiro SS, Wilk MB, Chen HJ, J Amer Statist Assoc 1968; 63: 1343-1372. 53. Hibbs MS, Hasty K. Seyer JM, Kang AH, Mainardi CL, J Biol Chem 1985; 260: 2493-2500.

54. Murphy GM, Bretz U, Baggiolini M, Reynolds JJ, Biochem J 1980; 192: 517-525. 55. Overall CM, Sodek J, J Dent Res 1987; 66: 1271-1282.

56. Weiss SJ, Pippin GJ, Biochem Pharmacol 1986; 35: 3189-3197.

57. Suzuki K, Enghild JJ, Morodomi T, Salvesen G, Nagase H, Biochem 1990; 29: 10261-10270.

58. Nagase H, Enghild JJ, Suzuki K, Salvesen G, Biochem 1990; 29: 5783-5789.

59. Nagase H, Ogata Y, Suzuki K, Enghild JJ, Salvesen G, Biochem Soc Transact 1991; 19: 715-718.

60. Springman EB, Angleton EL, Birkedal-Hansen H, Van Wart HE, Proc Natl Acad Sci USA 1990; 87: 364-368. 61. Reddy VY, Pizzo SV, Weiss SJ, J Biol Chem 1989; 264: 13801-13809.

62. Okada Y, Watanabe S, Nakanishi I, Kishi J, Hayakawa T, Watorek W, Travis J, Nagase H, FEBS Lett 1988; 229 157-160.

63. Murphy G, Hembry RM, J Rheumatol (suppl 32) 1992; 19: 61-64. 64. Uitto V-J, Suomalainen K, Sorsa T, J Periodont Res 1990; 25: 135-142.

65. Overall CM, Wrana JL, Sodek J, J Biol Chem 1989; 264: 1860-1869.

66. Stricklin GP, Jeffrey JJ, Roswit WT, Eisen AZ, Biochem 1983; 22: 61-68. 67. Welgus GH, Jeffrey JJ, Eisen AZ, J Biol Chem 1981; 256:

9511-9518.

68. Hibbs MS, Bainton DF, J Clin Invest 1989; 84: 1395-1402.

69. Birkedal-Hansen H, Taylor RE, Zambon JJ, Barwa PK, Neiders ME, J Periodont Res 1988; 23: 258-264. 70. Uitto V-J, Turto H, Huttunen A, Lindy S, Uitto J, Biochem Biophys Acta 1980; 613: 168-177.

71. Perrson S, Claesson R, Carlsson J, Oral Microbiol & Immunol 1989; 3: 169-172.

72. Lδe H, Anerud A, Boysen H, Smith, M, J Periodont Res 1978; 13:

550-562.

73. Lόe H, Anerud A, Boysen H, Morrison E, J Clin Periodontol 1986; 13: 431-440.

74. Lindhe J, Okamoto H, Yoneyama T, Haffajee A, Socransky SS, J Clin Periodontol 1989; 16: 671-678.

75. Kulkarni G, Lee W, Aitken S, Birek P, McCulloch CAG, J Periodontol 1991; 62: 197-202.

Claims

1. A method for diagnosing the presence of collagen tissue destructive disease at a collagen connective tissue site of a mammal, comprising the steps of:

collecting a sample of extracellular fluid from the area of the site; and

ascertaining whether the sample contains an amount of a collagen-destructive enzyme in an active form exceeding a pre-determined first threshold amount, wherein a said amount of the collagen-destructive enzyme which exceeds the threshold amount indicates the presence of collagen destructive tissue disease.

2. The method of claim 1 wherein the enzyme is collagenase.

3. The method of claim 1 wherein the enzyme is gelatinase.

4. The method of claim 2 wherein the collagenase is derived from neutrophils.

5. The method of claim 3 wherein the gelatinase is derived from neutrophils.

6. The method of claim 1 wherein the site is a juncture of a tooth and jaw bone.

7. The metiiod of claim 1 wherein the sample is obtained from an oral cavity of the mammal.

8. The method of claim 7 wherein the sample is obtained in a mouth rinse from the oral cavity.

9. The method of claim 1 wherein the step of ascertaining includes the steps of determining a level of activity of the enzyme in the sample and comparing the determined level with a pre-determined level of activity wherein a said determined level exceeding the pre-determined level indicates the presence of collagen destructive tissue disease.

10. The method of claim 1 wherein the ascertaining step further comprises the step of determining whether the amount of the collagen-destructive enzyme in said active form is below a pre-determined second threshold amount, wherein a said amount of the collagen-destructive enzyme below the second threshold amount indicates the absence of collagen-destructive tissue disease.

11. A method for diagnosing the presence of periodontal diseases in a mammalian subject comprising the steps of:

obtaining a sample of gingival crevicular fluid from the oral cavity of the subject; and

ascertaining whether the sample contains an amount of a collagen-destructive enzyme in an active form exceeding a pre-determined threshold amount, wherein a said amount of the collagen-destructive enzyme which exceeds the threshold amount indicates the presence of periodontal diseases.

12. The method of claim 1 wherein the subject is a human and obtaining the sample includes the step of having the mouth of the subject rinsed with a pre-determined amount of water for a pre-determined length of time and recovering the sample-containing rinse from the mouth of the subject.

13. The method of claim 2 wherein the ascertaining step includes determining whether the rinse contains an amount of active collagenase which exceeds a pre-determined amount thereof.

14. The method of claim 2 wherein the ascertaining step includes determining whether the rinse contains an amount of active gelatinase which exceeds a pre-determined amount thereof.

15. The method of claim 2 wherein the ascertaining step includes exposing a collagen capsule containing a colored dye to the sample, wherein the capsule has a wall having a thickness such that if the amount of active collagenase in the sample exceeds the threshold amount then degradation of the capsule wall occurs to cause the dye to be released to indicate the presence of the disease.

16. The method of claim 3 wherein the ascertaining step includes exposing a gelatin capsule containing a colored dye to the sample, wherein the capsule has a wall having a thickness such that if the amount of active gelatinase in the sample exceeds the threshold amount then degradation of the capsule wall occurs to cause the dye to be released to indicate the presence of the disease.

17. A gelatin capsule for use in diagnosis of collagen tissue destructive disease in a mammalian subject, comprising a wall encapsulating a colored dye wherein the wall has a thickness such that when the capsule is exposed to a sample collected from a potentially diseased site of the subject containing active collagenase exceeding a pre-determined level, degradation of the wall occurs to release the dye from the capsule to indicate the presence of collagen destsructive tissue disease in the subject.

18. A collagen capsule for use in diagnosis of collagen tissue destructive disease in a mammalian subject, comprising a wall encapsulating a colored dye wherein the wall has a thickness such that when the capsule is exposed to

SUBSTITUTE SHEET a sample collected from a potentially diseased site of the subject containing active collagenase exceeding a pre-determined level, degradation of the wall occurs to release the dye from the capsule to indicate the presence of collagen destsructive tissue disease in the subject.

19. A substrate comprising a solid core having a collagen coating in which the coating is of such a thickness that when the substrate is exposed to a sample collected from a diseased patient for a period of time, the coating is degraded to expose the core to indiciate the presence of disease.

20. A substrate comprising a solid core having a collagen coating in which the coating is of such a thickness that when the substrate is exposed to a sample collected from a diseased patient for a period of time, the coating is degraded to expose the core to indiciate the presence of disease.

21. An indicator for diagnosis of periodontal diseases in a mouthrinse sample of a subject, the indicator comprising, a solid core having markers at various locations, the core being at least partially coated with collagen of a predetermined thickness at each location to cover the marker at the location, such that when the indicator is exposed to a said sample which contains an amount of active collagenase for a period of time, if the amount of active collagenase exceeds a threshold amount the collagen at a particular location is degraded to visually expose the indicator at that location, the amount of active collagenase required to expose a said marker at at least one of the locations corresponding to an amount of active collagenase indicating presence of periodontal diseases in the subject.

22. An indicator for diagnosis of periodontal diseases in a mouthrinse sample of a subject, the indicator comprising, a solid core having markers at various locations, the core being at least partially coated with gelatin of a predetermined thickness at each location to cover the marker at the location,

SUBSTITUTE SHEET such that when the indicator is exposed to a said sample which contains an amount of active gelatinase for a period of time, if the amount of active gelatinase exceeds a threshold amount the gelatin at a particular location is degraded to visually expose the indicator at that location, the amount of active gelatinase required to expose a said marker at at least one of the locations corresponding to an amount of active gelatinase indicating presence of periodontal diseases in the subject.

SUBSTITUTE SHEET