Summary Although matrix metalloproteinases (MMPs) are among the potential key mediators of cancer invasion, their involvement in premalignant lesions and conditions is not clarified. Therefore, we studied, using in situ hybridization, immunohistochemistry and zymography the expression and distribution of MMP-1 and -2, and their tissue inhibitors (TIMPs -1, -2 and -3) in oral squamous cell carcinomas (SCC) and lymph node metastases as well as in oral lichen planus, epithelial dysplasias and normal buccal mucosa. In oral SCC and lymph node metastasis, MMP-1 mRNA was detected in fibroblastic cells of tumoral stroma. In two out of ten carcinomas studied, the peripheral cells of neoplastic islands were also positive. MMP-2 mRNA expression was noted in fibroblasts surrounding the carcinoma cells, and no signal in carcinoma cells was detected. A clear TIMP-3 mRNA expression was seen in stromal cells surrounding the neoplastic islands in all SCCs and lymph node metastases studied. TIMP-1 mRNA was detected in some stromal cells surrounding the neoplastic islands, whereas the mRNA expression for TIMP-2 was negligible. On the other hand, expression of MMPs and TIMPs was consistently low in oral epithelial dysplasias, lichen planus and normal mucosa. In certain epithelial dysplasias and lichen planus, MMP-1 and -2 mRNA expressions were detected in few fibroblasts under the basement membrane zone, but normal mucosa was completely negative. In SCC and lymph node metastasis, a detectable immunostaining for MMP-1 in stromal cells and in some carcinoma cells was observed. MMP-2 immunoreactivity was detected in the peripheral cell layer in neoplastic islands and in some fibroblast-like cells of tumoral stroma. Immunostaining for TIMP-3 was detected in stromal cells surrounding the neoplastic islands. A weak positive staining for TIMP-1 was located in tumoral stroma, whereas the immunostaining for TIMP-2 was negative. Using zymography, elevated levels of MMP-2 and MMP-9 were observed in carcinoma samples in comparison with lichen planus or normal oral mucosa. Our results indicate that the studied MMPs and TIMPs are clearly up-regulated during invasion in oral SCC. However, there was also a clear, although weak, up-regulation of the expression of the MMPs but not TIMPs in some of the lichen planus and dysplastic lesions.Keywords: oral squamous cell carcinoma; matrix metalloproteinase; tissue inhibitor of metalloproteinase; mouth neoplasm Oral squamous cell carcinoma (SCC) has a high potential for invasiveness associated with a high rate of fatality. Distant-organ metastasis and regional lymph node metastasis are the major cause of mortality from the oral SCC. Oral lichen planus is regarded as a potential condition for malignant transformation, and thus dysplasia or carcinoma could arise from oral lichen planus (WHO, 1997).Matrix metalloproteinases (MMPs) are a highly regulated superfamily of enzymes that degrade almost all extracellular matrix and basement membrane components, processes which are essential for invasion and subseque...

The aim of the present study was to characterize the eventual presence and molecular forms of gelatinase/type IV collagenase activities in gingival crevicular fluid (GCF) and saliva in different forms of periodontitis; patients with clinically healthy periodontium served as controls. Enzyme activities were monitored electrophoretically by zymography using gelatin and type IV collagen as substrates and analyzed visually and/or densitometrically. Both saliva and GCF collected from adult periodontitis, localized juvenile periodontitis and type II diabetes mellitus periodontitis patients contained species moving identically with gelatinase isolated from human neutrophils or MMP-9 (mean 98 kD), and species with mobility similar to gelatinase in fibroblast cell culture supernatants or MMP-2 (mean 76 kD). Hitherto, undescribed high molecular weight forms (mean 128 kD), were found, possibly representing polymerized or complexed enzyme active/activated in situ in the gel matrix. Small molecular forms of gelatinases (mean 51 kD and 46 kD), unable to cleave type IV collagen, were also found, most likely representing in vivo proteolytically activated, truncated enzymes. Although multiple forms of gelatinases/type IV collagenases in saliva and GCF may take part in the tissue destruction in periodontitis, their profile judged according to molecular weights does not differentiate between different forms of periodontitis.

Accelerated periodontal tissue destruction in patients with labile insulin-dependent diabetes mellitus (DM) and localized juvenile periodontitis (LJP) has been suggested to be related to functional abnormalities of neutrophils. We have recently found that collagenase in gingival crevicular fluid (GCF) of adult periodontitis patients is primarily derived from neutrophils and that neutrophil collagenase activity is more sensitive to inhibition by tetracyclines than collagenase produced by fibroblasts. This study is to characterize the cellular sources, activation and inhibition of collagenase in GCF of DM patients and to compare it with collagenase in LJP GCF. We found differences which may have therapeutic implications. Specific doxycycline inhibition tests revealed that GCF collagenase in DM is derived from neutrophils, whereas the enzyme in LJP originates primarily from fibroblasts. Oxidant, sodium hypochlorite, activated efficiently GCF collagenase of DM but not LJP patients. In contrast, plasmin activated LJP GCF collagenase but not that of DM patients. In GCF of DM patients 50-60% of collagenase existed in an active form, whereas in LJP GCF, the enzyme was almost completely in a latent form. The results suggest that collagenase in GCF of periodontitis patients with labile DM is primarily derived from neutrophils and that tetracycline therapy may be an effective adjunct in treatment aimed at controlling the periodontal breakdown in these patients. On the other hand, in LJP the anti-collagenase property of tetracyclines may be less important for control of periodontal tissue destruction because of the tetracycline-resistance of fibroblast collagenase.

This study examined cellular and vascular changes in gingival connective tissue samples by stereologic point-counting procedures and interactive digital analyzing systems in long-term insulin-dependent diabetes mellitus patients. Gingival connective tissue capillaries representing a clinically healthy sulcus with no evidence of periodontal disease at the site of biopsy were studied in 29 patients with diabetes. Based upon their long-term medical records, 19 were identified as having poorly controlled (PIDD) and 10 as controlled insulin-dependent diabetes mellitus (CIDD). Ten nondiabetic, age- and gender-matched individuals served as controls. Thirty-nine biopsies were processed for light microscopy, and the blood vessel area was analyzed using an interactive digital analyzing system; 9 gingival biopsies, 5 diabetic and 4 controls, were processed for morphometric electron microscopic analysis. For each individual, site-specific recordings were made for the plaque index, bleeding index, probing depth, loss of attachment, and radiographic loss of interproximal alveolar bone. No evident signs of periodontitis occurred at the biopsy sites. For each PIDD patient, respective volumetric and numeric densities of cellular components including fibroblasts, neutrophilic granulocytes, monocyte/macrophages, mast cells, lymphocytes, blast cells, and plasma cells were recorded in the inflamed connective tissue (ICT). Non-cellular components such as collagen fibers and blood vessels were also recorded. PIDD patients had elevated plasma cell levels relative to controls and they appeared also to have a decreased collagen fiber density. In addition, fibroblasts occupied less volume in the ICT of PIDD patients than in controls. PIDD patients had the largest mean area of cross-section of the blood vessels, but this difference was not statistically significant (P > or = 0.211; t-test). No specific characteristics of ICT or vascular changes were detectable in adult well-controlled long-term diabetics under similar plaque conditions. Swollen and proliferated endothelial cells were frequently found in PIDD patients and the mean distance from the lumen to the outer border of basement membrane was greater in the PIDD than in the controls (P < 0.001; t-test). Overall, our findings that cellular, vascular, and connective tissue changes indicative of increased catabolism rather than anabolism detected in gingiva are especially associated with poorly controlled long-term insulin-dependent diabetes.

Laminin-332 degradation is pH-dependent and differences exist between studied Candida strains. Lm-332 degradation can exert functional disturbances on basement membrane integrity, possibly aiding Candida cell invasion into tissues. Certain synthetic matrix metalloproteinase inhibitors (CMTs, CTT) can inhibit Candida proteinases and may be therapeutically useful in future.