The purpose of this study was to assess the effect of endotoxin adsorbed to dental surfaces and to collagen type I on the migration, attachment, and orientation of human gingival fibroblasts (HGF). Transversely cut porcine tooth root slices (RS), 200 micron thick, were prepared. Half of the RS obtained were partially demineralized in EDTA. Half of the demineralized and non-demineralized RS were incubated with 400 micrograms/mL of endotoxin for 24 hr, whereas the other half were maintained in PBS and served as controls. Experimental and control RS were placed on confluent layers of HFG and cultured for six days. Cell migration toward and cell attachment to the periphery of the RS and the formation of oriented cell sheets were assessed by means of photographic techniques. Additionally, six-day-old cultures were fixed and processed for SEM observation. In separate experiments, the effect of endotoxin on cell attachment to collagen type I and on contraction of three-dimensional collagen gels was assessed. It was found that: (i) bacterial endotoxin inhibited migration and attachment of HGF to both demineralized and non-demineralized cementum and interfered with the development of oriented cellular structure: (ii) the inhibitory effect was significantly more pronounced for non-demineralized than for demineralized cementum: (iii) the morphology of HGF attached to endotoxin-treated dental surfaces was altered compared with that of their controls: and (iv) bacterial endotoxin inhibited cell attachment to collagen type I and delayed the contraction of collagen gel.

Bacterial endotoxin inhibits the attachment of human gingival fibroblasts to collagen. The present study attempted to elucidate the possible mechanism of this inhibition. Two mechanisms were considered: direct toxicity to the cells and steric interference. Collagen substrates were prepared by rat type I collagen being air-dried in the wells of 24 multi-well plates. Experimental collagen substrates were treated with 50 micrograms of endotoxin/well, while untreated collagen substrates served as controls. Two mL of cell suspension (10(4) cells/mL) was added to each well, and these were incubated at 37 degrees C for two h. The average cell number/mm2 attached to experimental and control substrates was determined. Cell attachment to endotoxin-treated collagen was inhibited by 78%, compared with that to untreated collagen. The washing of the endotoxin-treated collagen for two h did not affect the inhibition of cell attachment, whereas after 24 h of washing, cell attachment was inhibited by 54%, compared with that to untreated collagen. Pre-incubation of the cells in endotoxin for two h did not affect their attachment to collagen. The addition of fetal calf serum (15%) to the experimental system completely reversed the inhibition of fibroblast attachment to endotoxin-treated collagen. These findings suggest that endotoxin interferes with fibroblast attachment to collagen through a steric phenomenon, possibly by blocking the binding sites on the collagen molecule recognized by the membrane receptor for collagen.