The importance of cell–matrix interactions during early odontogenesis is well established and these interactions may be mirrored during tissue repair in mature dental tissues. The cells and extracellular matrix of the dentine-pulp complex are in close relation throughout dentinogenesis and this may be important in maintaining the physiological function of these tissues. Dynamic approaches to the study of dentinogenesis during development and tissue repair have adopted both in vivo and in vitro techniques. Whilst in vitro culture models are used extensively in studies of tooth development, they have been little applied to the study of the more mature dentine–pulp complex. Culture of pulpal fibroblasts has allowed investigation of their synthetic and secretory behaviour, but there are few studies on the culture of odontoblasts from more mature teeth.

Attempts to culture odontoblasts in vitro have shown that contact between them and the dentine matrix is required to maintain their phenotypic morphology and secretory activity. Culture of the odontoblasts in isolation after removal from the dentine matrix leads to loss of phenotype and the appearance of a fibroblastic morphology. Components within the dentine matrix exert an influence on odontoblast secretory activity both in vivo and in vitro, emphasizing the role of cell–matrix interactions during dentinogenesis. Culture of odontoblasts on the surface of dentine matrix from which the pulpal soft tissues had been removed, has allowed study of dentinogenic events over several hours. However, the presence of the pulpal soft tissues during culture of the intact dentine–pulp complex may provide support and nutrition to the odontoblasts for longer periods. The use of tooth slices may help to overcome limitations to nutrient diffusion and gaseous exchange in such organ cultures. A tooth-slice system for the culture of the dentine–pulp complex in new-born rat tissue has been reported, but has been used only for short-term culture.

A model allowing longer-term culture of the dentine–pulp complex would be valuable for the study of molecular and cellular processes involved in normal dentinogenesis and tissue repair. A tooth-slice culture model for the human dentine–pulp complex was recently reported in which cell viability was maintained for longer periods. In this model, slices were cultured on the base of a dish, which limited experimental manipulation of growth conditions. An approach that overcomes some of these problems has been used for culture of the embryonic dental papilla, which was embedded in a semisolid agar medium and grown at the gas–medium interface on filters in Trowel-type cultures. That technique might be adapted to the successful, longer-term culture of the dentine–pulp complex from mature teeth. Our aim now was to develop that technique for the culture of slices from the mature rat incisor tooth for up to 2 weeks.

Upper and lower incisor teeth were dissected from 28-day-old male Wistar rats killed by cervical dislocation. They were placed in sterile washing medium consisting of DMEM (Sigma, U.K.), and subsequently embedded in dental impression compound (KEMCO precision; Kemdent) so that transverse sections of approx. 2-mm thickness could be cut with a segmented, diamond-edged rotary saw (Taab, Aldermaston, Berks, U.K.) cooled with water. The tooth slices were immediately placed in sterile washing medium before culture.

The sections of incisors were washed several times in washing medium at 37°C immediately after cutting, and transferred to individual wells of a plastic, 96-well dish. Culture medium (100 ?l) containing DMEM, vitamin C (0.15 mg/ml), 10% heat inactivated fetal calf serum, -glutamine (200 mM), 1% penicillin/streptomycin solution and 1% low melting-point agar maintained at 37°C was added to each well and allowed to cool. When semisolid, the embedded tooth slices were removed from the multiwell dish and transferred to a sterile Millipore (mixed esters of cellulose acetate and nitrate) filter. With the aid of a plastic support, the filter was floated on the surface of 4 ml DMEM in Trowel-type cultures in 35×10 mm Petri dishes. The tooth slices were cultured at 37°C in an atmosphere of 5% CO₂ in air, in a humidified incubator for 2–14 days, changing the medium every 2 days. After culture, the slices were fixed in 10% neutral-buffered formalin for 24 h, demineralized in 10% formic acid for 48 h and then processed and embedded in paraffin wax for histological examination. Sections were cut at 7 ?m and routinely stained with haematoxylin and eosin.

At the end of a culture period, tooth slices were removed and placed in 4 ml DMEM containing 0.0001% acridine orange, together with heat-inactivated fetal calf serum, vitamin C, glutamine and antibiotics. The tissues were then incubated for 2 h at 37°C in 5% CO₂ in air before mounting the slice on a microscope slide in phosphate-buffered saline and examining it in a Leitz Fluorart fluorescence microscope.

Secretory activity of cultured tissues was assessed using radiolabelled proline as a precursor for the biosynthesis of extracellular matrix. In cultures to be labelled with radioactive marker, 100 ?l of semisolid culture medium (see organ culture of tooth slices above) containing 10 ?Ci of [³H]proline was added to each well containing a tooth slice during the initial embedding stage and slices were then cultured as previously described in 4 ml of culture medium containing 20 ?Ci (5 ?Ci/ml of medium) of [³H]proline. Slices were cultured at 37°C in 5% CO₂ in air for 48 hr. The culture medium containing [³H]proline was then removed, fresh medium without radiolabel added, and culture continued for a further 5 days. Tooth slices cultured in the absence of [³H]proline in both semisolid and liquid culture media served as controls.

Autoradiography was done on 7-?m thick sections of cultured tissue in a darkroom. Nuclear emulsion, LM-1 (Amersham, Little Chalfont, U.K.), was melted at 43°C and mixed; sections mounted on slides were then dipped into the molten emulsion and removed. The slides were left to gel on a cold plate for 10 min, then air-dried for 2–3 h in a light-tight box. Once dry, they were placed in a light-tight box and left to expose at 4°C for 2–21 days.

Slides were processed in Phenisol (Ilford) diluted 1:4 with distilled water for 3–5 min and gently agitated. They were then placed in a stop solution of 0.5% (w/v) acetic acid, transferred to a fixing solution of 30% (w/v) sodium thiosulphate for 4 min, soaked in that solution for a further 4 min and finally washed in distilled water before staining.

After culture, tooth slices were fixed in 2.5% glutaraldehyde–phosphate buffer (0.1 M, pH 7.3) for 24 h. Tissues were demineralized in 5% EDTA in 2.5% glutaraldehyde–phosphate buffer with 6.5% sucrose (0.1 M, pH 7.3). Slices were washed in 0.2 M sucrose–phosphate buffer (0.1 M, pH 7.3) overnight, postfixed in 1% osmium tetroxide for 1 hr, washed in buffer and dehydrated through a series of graded ethanols. They were then placed in a 50:50 Spurr resin/ethanol mix overnight, and embedded in Spurr resin for 3–4 h before polymerization at 70°C for 24 h. Thin sections were cut on a Reichart ultracut microtome E (Reichart Jung, Germany); some semithin sections (1 ?m thickness) were collected on a miroscope slide and stained with toluidine blue to assess the tissue orientation. Ultrathin sections (90 nm thickness) were mounted on copper grids, counterstained with uranyl acetate and lead citrate, and examined in a Phillips EM 208 transmission electron microscope.

The dentine–pulp complex in transverse sections of rodent incisor teeth after the initial preparation of the thick tooth slice, but before culture, showed maintenance of normal morphology throughout. The odontoblasts were attached to the predentine and appeared as tall, columnar cells. They were larger and more columnar in the radicular region, which was probably the most metabolically active area. Fibroblasts were observed within the pulpal extracellular matrix in slices taken from all areas of the tooth.