Diabetes is one of the risk factors

Transcription

1 J Periodontol February 2006 Collagenases in Gingival Crevicular Fluid in Type 1 Diabetes Mellitus Bedia Safkan-Seppälä,* Timo Sorsa,* Taina Tervahartiala,* Arzu Beklen, and Yrjö T. Konttinen i Background: Studies have demonstrated that high levels of collagenase activity in gingival crevicular fluid (GCF) are associated with degradation of periodontal tissues in progressive periodontitis compared to periodontally healthy tissues. Because the activation of collagenases is an important issue in periodontitis, we have studied the activation of collagenase in gingival crevicular fluid samples of diabetic patients. Methods: Collagenase activity was studied in human gingival crevicular fluids. Twenty-two poorly controlled diabetic patients (e.g., blood glucose: mmol/l; hemoglobin A 1c [HbA 1c ]: 9.6% 0.3%) and five well-controlled diabetic patients were compared to six chronic periodontitis subjects and five healthy controls. Collagenase activity against type I collagen was measured using sodium dodecyl sulfatepolyacrylamide gel electrophoresis analysis quantitated by laser densitometry. Results: The poorly controlled diabetic patients had more alveolar bone loss than the well-controlled diabetic subjects and controls (P <0.001; t test). The activity of collagenases in GCF in poorly controlled diabetic patients was similar to that seen in chronic periodontitis subjects (P >0.05) but higher than in healthy controls (P <0.01; t test), whereas there was no difference between the well-controlled diabetic subjects and systemically healthy controls (P >0.05; t test). Conclusion: Poorly controlled diabetes is strongly related to periodontal tissue destruction, and collagenases in GCF may mediate and reflect this effect. J Periodontol 2006;77: KEY WORDS Collagenase; gingival crevicular fluid; insulin-dependent diabetes mellitus; periodontitis. * Institute of Dentistry, University of Helsinki, Helsinki, Finland. Department of Medicine/Invärtes Medicin, Helsinki University Hospital, Helsinki, Finland. Institute of Biomedicine/Anatomy, University of Helsinki. Medico-Social Centre, Dental Clinic, Bogazici University, Istanbul, Turkey. i ORTON Orthopaedic Hospital of the Invalid Foundation, Helsinki, Finland. COXA Joint Replacement Hospital, Tampere, Finland. Diabetes is one of the risk factors for chronic periodontitis. 1 Crosssectional and longitudinal studies have demonstrated that poor control is associated with more loss of attachment and alveolar bone than good disease control in aggressive type 1 diabetes mellitus. 2-4 To our knowledge, no previous studies have been published on collagenases in gingival crevicular fluid (GCF) of patients with long-term poorly controlled type 1 diabetes and periodontitis. Collagenases are considered key mediators and indicators of inflammatory tissue destruction in periodontal diseases. This makes collagenases amenable targets in strategies aiming to prevent attachment and bone loss. Indeed, long-term low-dose doxycycline therapy with anticollagenolytic effects may be an effective adjunct in the treatment of chronic periodontitis. 5 Periodontal tissue destruction is reflected in GCF as elevated levels of collagenases. 6 The aim of the present study was to analyze gingival crevicular fluid collagenases of poorly and well-controlled type 1 diabetic and chronic periodontitis subjects compared to non-diabetic with chronic periodontitis and periodontally healthy control subjects. MATERIALS AND METHODS Patients From January 1, 1990 through January 31, 1993, human subjects participated in the study after providing informed consent to a protocol that was reviewed and doi: /jop

2 Diabetes Mellitus and Periodontal Disease Volume 77 Number 2 approved by an appropriate Institutional Review Board. At the time of this study, 204 diabetic individuals were treated at the Department of Medicine, University Central Hospital of Helsinki. Of these subjects, 83 individuals had had type 1 diabetes for at least 10 years. Twenty-seven of these subjects (14 males and 13 females) were selected. All long-term diabetes patients (35 to 65 years) had severe chronic periodontitis. They were accepted as patients to the Department of Periodontology and included in this study. Out of these 27, five individuals were, based upon their medical records, classified by diabetes specialists as having well-controlled type 1 diabetes at the initiation of the study. Their hemoglobin A1c (HbA 1c ) value (8.4% 0.3%) was slightly higher than the currently recommended cut-off point, which requires HbA 1c <7.5%, the gold standard for good control. All diabetic type 1 patients were outpatients at the University Central Hospital of Helsinki, and, therefore, the diabetic subjects were the most medically demanding due to their long-standing complications of diabetes. Twenty-two of the diabetic patients were identified as having poorly controlled type 1 diabetes (Table 1) with a history of problems related to the control of their diabetes, such as hypo- or hyperglycemias, recurrent infections, ketoacidosis, diabetic coma, glucosuria, and different stages of retinopathies, nephropathies, and neuropathies. The age range of the poorly controlled diabetic subjects (N = 22) was 40 to 65 years, and the mean duration of their diabetes was 21.2 years. They were compared to subjects with wellcontrolled diabetes (N = 5; age range: 42 to 51 years; mean duration of diabetes: years) (Table 1), chronic periodontitis (N = 6; age range: 36 to 61 years; median: 50 years), and periodontally healthy non-diabetic dental students (N = 5; age range: 21 to 23 years). Clinical Data The dental examinations were performed by a blinded specialist in periodontology during 1990 to 1993 at the University of Helsinki. Recordings were made for the buccal, lingual, and two interproximal surfaces of each tooth with a World Health Organization (WHO) ball-tip probe (tip diameter: 0.5 mm). The interproximal sites were examined by probing. Orthopantomograms were done at baseline. After GCF sampling, clinical recordings were made as follows. Plaque index. Recordings were made for the gingival third of each tooth surface. Individual scores for percentages of tooth surfaces with plaque index 1 were recorded. Bleeding index. Bleeding from the bottom of the clinical periodontal pocket was determined at four sites of each tooth 10 to 15 seconds after probing with the blunt WHO probe and standardized pressure. Bleeding after probing at sample site. After gingival crevicular fluid sampling, bleeding after probing at the sampling site was evaluated 10 to 15 seconds after probingwiththewhoprobeandstandardizedpressure. Loss of attachment. The distance from the cemento-enamel junction to the bottom of the periodontal pocket was measured with the WHO probe to the nearest millimeter. The loss of attachment measurements were made at four sites of each tooth except the molars. The molars were examined at mesial and distal sites and, in addition, at two buccal and two lingual sites. Probing depth and probing depth of sample sites. The distance from the gingival margin to the bottom of the periodontal pocket was measured. Only pockets with probing depths >3 mm were recorded. Bone loss. Approximal loss of marginal alveolar bone was measured from orthopantomograms. # The distance from the cemento-enamel junction to the alveolar bone crest was measured mesially and distally on all teeth to the nearest millimeter using a transparent scale. A mean score based on these measurements was calculated for each individual. Gingival Crevicular Fluid Gingival crevicular fluid samples were collected from periodontally diseased sites with a mean probing depth 5 mm ( mm; Table 2) after the double-blind roentgenographic examination. At the time of examination, none of the examiners knew the grouping of the diabetic subjects, and, therefore, the two diabetic groups were not statistically equal, as in our earlier studies. 3 The corresponding interproximal sites from control subjects were matched to the wellcontrolled diabetic group. These sites (two to five sites per subject) were cleaned and dried gently and supragingivally with sterile curets and kept dry with cotton wool rolls. Visible plaque was carefully removed. One to two filter paper strips were placed at the opening of the gingival margin 0 to 1 mm into the sulcus for 3 minutes for gingival crevicular fluid collection. Strips were placed into Eppendorf tubes. Measurement of Collagenase Activity Against Type I Collagen Samples were studied blindly without the examiner knowing the grouping of the poorly controlled insulindependent diabetes (PIDD) and well-controlled insulin-dependent diabetes (CIDD) subjects. Gingival crevicular fluid samples were assayed for interstitial collagenase activity by quantitative sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS- PAGE) and laser densitometry. 6 Native human type I collagen was extracted from human skin and purified # PM2002CC, Planmeca Oy, Helsinki, Finland. 190

3 J Periodontol February 2006 Safkan-Seppälä, Sorsa, Tervahartiala, Beklen, Konttinen Table 1. Characteristics of Diabetic Study Groups* Subjects Age (years) Duration of Diabetes (years) Insulin Dosage (IU/day) Glycosylated Hemoglobin (HbA 1c ) (%) Blood Glucose (mmol/l) Poorly controlled (N = 22) x SEM Median Well controlled (N = 5) x SEM Median P value * t test used for comparison. Mean. Table 2. Comparison of Clinical Parameters of Sample Groups (t or x 2 tests) Clinical Parameters Poorly Controlled (N = 22) Well Controlled (N = 5) Controls (N = 6) Poor Versus Good Poor Versus Controls Good Versus Controls Plaque index* x NS SEM Bleeding index x NS i SEM Bleeding after probing at sample x NS i NS sites SEM Loss of attachment x NS SEM Probing depth x NS SEM Probing depth of sample sites x NS SEM Bone loss x i SEM * Mean plaque index scores (% scores 1). Mean. P < P <0.01. i P <0.05. NS = not statistically significant. by selective salt precipitation at acid and neutral ph. The purity of type I collagen substrate was examined by cyanogen bromide cleavage. Gingival crevicular fluid samples were incubated with 1.5 mm collagen in the presence or absence of 1 mm aminophenylmercuric acetate (APMA),** an optimal organomercurial activator of interstitial collagenases. The enzyme reaction was terminated, and the collagen was denatured by boiling for 3 minutes after adding 0.1 M Tris-phosphate buffer, ph 6.8, containing 20% glycerol, 6% sodium dodecyl sulfate, and 0.04% bromophenol blue. Intact collagen, 3/4 (a 1A and a 2A ), and 1/4 (a 1B and a 2B ) degradation products were separated by using 11% (T%) SDS-PAGE. The gels were stained with Coomassie brilliant blue and destained in 5% acetic acid and 10% methanol in water. 6 ** Sigma, St. Louis, MO. 191

4 Diabetes Mellitus and Periodontal Disease Volume 77 Number 2 Degradation products were quantified densitometrically. The values representing the degraded a 1A and a 2A collagen chains were multiplied by 4/3. 6 Statistics Results are expressed as mean standard error of the mean (SEM) and medians. The t test was used to compare means for normally distributed variables provided that there were no statistically significant differences in the variances. The x 2 test was used to compare frequencies between different groups. Pearson s correlation matrix was used to calculate correlations between parameters. The correlation matrix was tested for homogeneity using the Bartlett x 2 test. Computations of probabilities were done using a matrix of Bonferroni probabilities. A statistical program was used for statistical analysis of the clinical data. RESULTS Diabetes Poorly controlled diabetic patients (N = 22) were characterized by higher mean blood glucose levels than well-controlled diabetic subjects (N = 5; mmol/l versus mmol/l; P <0.05; t test). The difference in glycosylated hemoglobin A 1c, the gold standard of diabetic control, was 9.6% 0.3% and 8.4% 0.3%, respectively, and did not reach statistical significance due to the low number of CIDD subjects who were randomly chosen and blindly studied without knowing the grouping of the subjects according to the study plan (P = 0.101; t test). There were no statistical differences in the mean duration of diabetes (median: years versus years; P = 0.625; t test) or the daily dosage of insulin units (median: 50 versus 38 IU/day; P = 0.101; t test) between the poorly and well-controlled diabetic subjects (Table 1). Periodontitis Poorly controlled diabetic patients (N = 22) had more alveolar bone loss than well-controlled diabetic subjects (N = 5) ( mm versus mm, respectively; P <0.05l; x 2 test). There was a difference between the diabetic groups and the control group regarding alveolar bone loss (P <0.001; t test). Patients with poorly and well-controlled diabetes had more loss of attachment than the controls (P <0.01 and P <0.001, respectively; t test). There was no difference between poorly and well-controlled diabetic patients in the mean probing depth or the mean probing depth of the sites chosen for sampling (Table 2). The differences of bleeding indices in general and after probing at the sample sites were not significant between poorly and well-controlled diabetic patients, but the differences were significant when comparing PIDD subjects to controls (P <0.01 and P <0.05, respectively; t test; Table 2). At the baseline investigation, the poorly and well-controlled diabetic subjects had higher mean plaque indices than the control subjects (P <0.001 and P <0.01; t test; Table 2). Collagenase Diabetic subjects with periodontal disease as a group had higher GCF collagenase activities than the healthy control subjects (P <0.001; x 2 test; Fig. 1). Poorly controlled diabetic patients with periodontitis were characterized by higher GCF collagenase activities than well-controlled diabetic subjects ( versus IU/ml; P <0.05; x 2 test; Fig. 2, lanes 2 and 3). Total collagenase activity was higher in poorly controlled diabetic patients than in well-controlled diabetic subjects (P <0.05; t test) or controls (P <0.01; t test). There was no significant correlation between the age of the patients and the collagenase values in the present study. Therefore, age does not explain the difference observed between poorly and well-controlled type 1 diabetes. There were no significant correlations between collagenase activity values and clinical parameters when diabetic patients were analyzed as one group (Bartlett x 2 test). When poorly controlled diabetic patients (N = 22) and well-controlled diabetic subjects (N = 5) were tested separately, significant correlations between vertical bone loss and loss of attachment (r = 0.846; P <0.001), bleeding on probing and loss of attachment (r = 0.726; P <0.01), and vertical bone loss and bleeding on probing (r = 0.648; P <0.05) were observed in poorly controlled diabetic patients. In our set of patients, the total collagenase activity in chronic periodontitis was intermediate between poorly and wellcontrolled type 1 diabetes complicated with chronic periodontitis but significantly higher (P <0.001; t test) than in healthy controls (Fig. 1). DISCUSSION There is a relationship between periodontal disease and long-term type 1 diabetes with poor metabolic control. 2,4 Decreased collagen and glycosaminoglycan synthesis by fibroblasts and elevated collagenases 5 lead to loss of tooth support in periodontitis. The present study extends these findings to human type 1 diabetes by demonstrating high collagenase activity, both latent and proteolytically activated, in gingival crevicular fluid. It is likely that collagenases, as downstream effector molecules, are directly responsible for tissue destruction. However, it is still unclear what molecules act as upstream effectors upregulating collagenase activity in such subjects. We believe that proinflammatory cytokines such as interleukin-1 and tumor necrosis factor-a play a role, as we have previously shown that the number of tumor Ultrascan Laser Densitometer 2202, LKB, Bromma, Sweden. Systat, Systat Software, Richmond, CA. 192

5 J Periodontol February 2006 Safkan-Seppälä, Sorsa, Tervahartiala, Beklen, Konttinen Figure 1. Collagenase activity (10-8 IU/ml; x SEM) without or with in vitro activation with aminophenylmercuric acetate (APMA), respectively, in poorly controlled type I diabetes: / ; well-controlled type 1 diabetes: / ; chronic periodontitis: / ; and healthy controls: / in gingival crevicular fluid samples from sites with severe periodontal disease (probing depth 6 mm) (*P <0.05; P <0.001; P <0.01; NS = not significant [x 2 test]). Figure 2. Degradation of native human skin type I collagen by GCF collected from poorly controlled diabetic patients (lanes 2 and 3: +1 and -1 mm APMA pretreatment, respectively), controlled diabetic subjects (lanes 4 and 5: +1 and -1 mm APMA pretreatment, respectively), and healthy controls (lanes 6 and 7: +1 and -1 mm APMA pretreatment, respectively). Native type I collagen was incubated with buffer (lane 1). GCF samples and buffer were incubated with 1.5 mm native soluble type I collagen. Incubations were terminated by boiling (5 minutes) in Laemmli s sample buffer and analyzed by 11% SDS-PAGE. a indicates intact human skin type I collagen monomers, and 4/3a indicates classical 75%-degradation products resulting from collagenase action on native type I collagen. necrosis factor-a containing cells is clearly increased in chronic periodontitis compared to healthy controls (1, cells/mm 2 versus cells/mm 2, respectively; P <0.001). 7 Diabetes leads to production of advanced glycosylation end products. Formation of advanced glycosylation end products is correlated to the level and duration of hyperglycemia. Advanced glycosylation end products modulate the formation of cross-links, which change collagen solubility and turnover. 8 Poorly controlled type 2 diabetes mellitus patients with advanced periodontitis, a complication of diabetes, have recently been shown to express high levels of active matrix metalloproteinase (MMP)-8/collagenase-2 in gingival crevicular fluid pathologically. 9 The pathogen microflora in hyperglycemic poorly controlled type 1 diabetes may also contribute to periodontal breakdown through exaggerated and/or prolonged host responses. 4,10,11 This study demonstrates that type 1 diabetes, particularly when poorly controlled, is characterized by high collagenase activity levels in gingival crevicular fluid. These findings suggest that poor metabolic control of type 1 diabetes is related to an increased risk of periodontitis. These results are supported by observations in rat models. 12,13 Furthermore, even well-controlled type 1 diabetes is characterized by relatively high endogenously activated collagenase compared to controls. This is in accordance with the hypothesis that type 1 diabetes is characterized by periodontal tissue destruction, particularly in subjects having poor metabolic control. Further studies are necessary to confirm or exclude the apparent role on the relationship between high collagenase activity and the production of advanced glycosylation end products. In this study, the classification of the state of type 1 diabetic patients to poor or good control of their systemic disease was not based upon any single blood glucose or HbA 1c evaluation but rather on the longterm medical history of each patient as assessed by diabetes specialists. Due to our study plan with blinded researchers, the lack of more subjects with good control of their type 1 diabetes mellitus provided some statistical problems, but they were solved technically and mathematically. Poorly controlled type 1 and 2 diabetic patients with elevated blood glucose and HbA 1c levels exhibit more attachment and alveolar bone loss than subjects with controlled blood glucose and HbA 1c levels. 4,10,14,15 Early diagnosis and strict control of blood glucose levels have been demonstrated to be important in the control of periodontal disease in diabetic individuals. 12,16 The present study suggests that the periodontal tissue destruction in type 1 diabetes is probably at least partially mediated by functionally active collagenases. ACKNOWLEDGMENTS The authors acknowledge the help of Mika Seppälä, professor of mathematics at Florida State University and at the University of Helsinki, for performing the statistical calculations. The authors also thank Prof. Jukka Ainamo, Institute of Dentistry, University of Helsinki, for his continuous support and encouragement. This work was supported in part by the Center for International Mobility (CIMO) Finland, evo grant HUS TYH 2235, Finska Läkaresällskapet, the Finnish 193

6 Diabetes Mellitus and Periodontal Disease Volume 77 Number 2 Dental Society Apollonia, the Invalid Foundation, Helsinki University, and the National Center of Excellence program supported by the Academy of Finland, Finnish Funding Agency for Technology and Innovation (TEKES), and Ministry of Education. REFERENCES 1. Genco RJ, Löe H. The role of systemic conditions and disorders in periodontal disease. Periodontol ;2: Safkan-Seppälä B, Ainamo J. Periodontal conditions in insulin-dependent diabetes mellitus. J Clin Periodontol 1992;19: Safkan-Seppälä B. Periodontal disease in insulindependent diabetics. [Thesis]. Helsinki: University of Helsinki; p. 4. Genco RJ. Current view of risk factors for periodontal diseases. J Periodontol 1996;67(Suppl.): Golub LM, Lee HM, Ryan ME, Giannobile WV, Payne J, Sorsa T. Tetracyclines inhibit connective tissue breakdown by multiple non-antimicrobial mechanisms. Adv Dent Res 1998;12: Uitto VJ, Suomalainen K, Sorsa T. Salivary collagenase. Origin, characteristics and relationship to periodontal health. J Periodontal Res 1990;25: Tervahartiala T, Koski H, Xu J-W, et al. Tumor necrosis factor-a and its receptors, p55 and p75, in gingiva of adult periodontitis. J Dent Res 2001;80: Monnier VM, Glomb M, Elgawish A, Sell DR. The mechanism of collagen cross-linking in diabetes: A puzzle nearing resolution. Diabetes 1996;45(Suppl. 3): S67-S Sorsa T, Ingman T, Suomalainen K, et al. Cellular source and tetracycline-inhibition of gingival crevicular fluid collagenase of patients with labile diabetes mellitus. J Clin Periodontol 1992;19: Seppälä B, Ainamo J. Dark field microscopy of the subgingival microflora in insulin-dependent diabetics. J Clin Periodontol 1996;23: Seppälä B, Sorsa T, Ainamo J. Morphometric analysis of cellular and vascular changes in gingival connective tissue in long-term insulin-dependent diabetes. J Periodontol 1997;68: Golub LM, Garant PR, Ramamurthy NS. Inflammatory changes in gingival collagen in the alloxan-diabetic rat. J Periodontal Res 1977;12: Golub LM, Lee HM, Lehrer G, et al. Minocycline reduces gingival collagenolytic activity during diabetes. Preliminary observations and a proposed new mechanism of action. J Periodontal Res 1983;18: Ainamo J, Lahtinen A, Uitto VJ. Rapid periodontal destruction in adult humans with poorly controlled diabetes: A report of 2 cases. J Clin Periodontol 1990; 17: Kinane DF, Marshall GJ. Periodontal manifestations of systemic disease. Aust Dent J 2001;46: Westfelt E, Rylander H, Blohme G, Jonasson P, Lindhe J. The effect of periodontal therapy in diabetics. Results after 5 years. J Clin Periodontol 1996;23: Correspondence: Prof. Yrjö T. Konttinen, Biomedicum, P.O. Box 700 (Haartmaninkatu 8), Helsinki, FIN HUS Finland. Fax: ; yrjo.konttinen@ helsinki.fi. Accepted for publication June 25,