Cementum is a mineralized dental tissue, present on the roots of teeth, which, together with the gingiva, periodontal ligament, and alveolar bone, forms the periodontium. It has long been thought that cementum is an ‘inert’ tissue only serving for the attachment of teeth to the alveolar bone by the incorporation of periodontal ligament (Sharpey’s) fibres. It is now recognized that cementum has a variety of biological functions among which is its contribution to the integrity of the dentogingival junction through the interaction of laminin of the extracellular matrix of cementum with its receptor on gingival epithelial cells . Cementum is also believed to participate actively in repair and regeneration following periodontal disease by virtue of its content of proteoglycans/glycosaminoglycans , growth factors , and attachment proteins . The chemotactic effect of healthy cementum extracts on gingival and periodontal ligament fibroblasts , and the inhibition of fibroblast migration by diseased cementum extracts , further support the role of cementum in the restoration of periodontal health.In contrast to the detailed knowledge available on the structure of the other dental and periodontal tissues, cementum has only recently attracted attention and there is little definitive information on its extracellular matrix. A variety of collagenous and non-collagenous proteins have been identified in cementum , of which type I collagen is the main constituent , in addition to minor amounts of types III , V and VI (. Among the non-collagenous proteins that have been demonstrated in the extracellular matrix of cementum are the proteoglycans with their glycosaminoglycan components .

The presence of glycosaminoglycans in human cementum was first demonstrated using histochemical staining with alcian blue. demonstrated the presence of hyaluronan, chondroitin sulphate and dermatan sulphate in healthy human cementum using cellulose acetate electrophoresis. Later, biochemically characterized chondroitin sulphate as the predominant glycosaminoglycan species in addition to dermatan sulphate and heparan sulphate. They used a monoclonal antibody (9-A-2) to localize chondroitin-4-sulphate immunohistochemically in bovine cementum. demonstrated chondroitin sulphate in sheep cementum, also using cellulose acetate electrophoresis. More recently, immunolocalized keratan sulphate around cementocytes of bovine cementum.

Few such studies have been directed at human cementum. Our aim now was, therefore, to investigate immunohistochemically the distribution of a range of glycosaminoglycan epitopes in human cementum using a broader panel of antibodies than previously available. In contrast to previous studies, we have also investigated their presence in healthy and periodontally involved human teeth, in addition to age-related changes in both groups.

Forty-eight permanent human teeth were collected from the Department of Oral Surgery, Cardiff Dental School immediately after extraction and fixed in 90% ethanol for 24 hr at 4°C. The teeth were separated into eight different age groups (from 12 to 90 years), and within each group the teeth were further subdivided into two clinical categories, teeth with periodontal diseases and periodontal disease-free teeth. The age groups included were: 12–20, 21–30, 31–40, 41–50, 51–60, 61–70, 71–80 and 81–90 years.

The teeth were then demineralized in 10% EDTA, pH 7.45, and demineralization was confirmed radiographically. The extraction of proteoglycans during demineralization in EDTA cannot be ruled out , although those workers, who examined proteoglycan loss upon demineralization, used powdered tissue, which has a much greater surface area. have shown that the use of aqueous EDTA provides the best compromise in terms of speed of demineralization and epitope preservation.

The demineralized teeth were sliced and embedded in paraffin wax. Sections of 5–10 ?m were cut and mounted on slides precoated with 0.1% poly-

-lysine (Sigma). Haematoxylin- and -eosin sections were used to confirm the histological appearances.

A panel of well-characterized monoclonal antibodies was obtained from Professor B. Caterson (Cardiff). The antibody 2B6 recognizes neo-epitopes in chondroitin-4-sulphate and dermatan sulphate, generated by chondroitinase ABC digestion, which consist of 4,5 unsaturated non-reducing terminal hexuronate (glucuronic acid), linked to a 4-sulphated N-acetyl galactosamine. This antibody also recognizes unsaturated epitopes in chondroitin-4-sulphate generated by digestion with chondroitinase AC-II. Monoclonal antibody 3B3 recognizes enzyme-generated neo-epitopes in chondroitin-6-sulphate following predigestion with chondroitinase ABC, which consist of non-reducing terminal unsaturated hexuronic acid residues adjacent to 6-sulphated N-acetylgalactosamine . In addition, 3B3 recognizes native epitopes in osteoarthritic cartilage . Monoclonal antibody 5D4 recognizes native epitopes in skeletal and corneal keratan sulphate. Monoclonal antibody 7D4 recognizes native chondroitin sulphate epitopes that are thought to be oversulphated and are present within the native chondroitin sulphate chain.

Before staining, the sections were deparaffinized in xylene and rehydrated in descending concentrations of alcohol, water and PBS. To expose the chondroitin-4-sulphate/dermatan sulphate neo-epitopes, sections were pretreated with chondroitinase ABC 0.1 IU/ml (Sigma) in 0.1 M Tris buffer containing 0.1% BSA, pH 7.2, at 37°C for 1 hr. Digestion with 0.1 IU/ml chondroitinase AC-II (Sigma) in 0.1 M Tris buffer containing 0.1% BSA, pH 6.4, at 37°C before incubation with the antibody 2B6 was also used to exclusively detect chondroitin-4-sulphate. To expose the neo-epitope recognized by antibody 3B3, the sections were pretreated with chondroitinase ABC as described above. After enzymatic pretreatment the sections were rinsed thoroughly in PBS and Tween for 15 min, and incubated for 15 min with 3% H₂O₂ in methanol to eliminate endogenous peroxidase activity. The sections were then blocked with 10% normal goat serum (Histostain-SP^TM kit; Zymed LAB-SA System, San Francisco, CA) for 30 min at room temperature. The primary antibodies were diluted as: 2B6, 1:1000; 3B3, 7D4 and 5D4 as 1:1500 with 0.1% BSA and 0.1% Tween 20 in PBS. The sections were incubated with the primary antibodies for 60 min at room temperature in a humidity chamber, then rinsed. Secondary biotinylated antibody (goat antimouse rabbit, guinea-pig and rat IgG; Histostain-SP^TM kit) was applied for 30 min. The sections were then incubated with the enzyme conjugate (horseradish peroxidase-labelled streptavidin) for 30 min and, finally, the chromogen diaminobenzydine solution was added for up to 60 sec. The sections were rinsed well with distilled water, counterstained with haematoxylin, and washed in water and PBS. All sections were then dehydrated in ascending concentrations of alcohol, cleared in xylene and mounted with Histomount (Histostain-SP^TM kit) for light-microscopic examination, and photomicrographs taken with a Leica microscope. The number of cementocytes was counted in three random fields of view on five separate slides for each antibody used, and the number of positive cells expressed as a percentage of the total number of cells.

Sections used as negative controls were treated in three main ways. First, by substitution of the primary antibody with PBS or mouse immunoglobulin (MIG, Sigma; concentration 1:500), which do not react with the secondary antibody. Secondly, for staining with 2B6 and 3B3, where enzymatic predigestion is required to generate the neo-epitope recognized by the antibody, enzymatic pretreatment with chondroitinase ABC and AC-II was omitted. Thirdly, before incubation with 5D4, some control sections were treated with keratanase (Sigma) 0.1 IU/ml (in 0.1 M Tris–HCl buffer, pH 7.4), which digests keratan sulphate chains, and hence removes the epitope. The pulp, predentine and dentine of the teeth were used for reference, based upon previous findings .

Localization of chondroitin-4-sulphate/dermatan sulphate, chondroitin-6-sulphate, keratan sulphate and novel sulphated chondroitin sulphate epitopes was observed in cellular cementum and at the cementum–periodontal ligament interface. A number of randomly selected fields were assessed to determine the proportion of positive lacunae, as described in Methods. With monoclonal antibody 2B6 after pretreatment with chondroitinase ABC, as well as after pretreatment with chondroitinase AC-II there was intense positive staining of the margins and sometimes the lumina of the majority of the lacunae and canaliculi in cellular cementum (99% of cementocytes demonstrated positive staining with 2B6 and chondroitinase ABC predigestion; 87% of cells demonstrated positive immunoreactivity with 2B6 and chondroitinase AC-II predigestion), and within a short distance at the periphery of cellular and acellular cementum where periodontal ligament fibres insert. No staining was observed in any other parts of the cementum matrix. With 3B3 antibody, after pretreatment with chondroitinase ABC, there was staining of a proportion (48%) of the lacunae and canaliculi and at the cementum–periodontal ligament interface .