Topographically induced bone formation in vitro: Implications for bone implants and bone grafts

We have investigated the influence of substrate topography on the timing and location of bone formation by rat osteoblasts. 250 μm thick slabs of dental tissues were used intact or had a rectangular grid of grooves ( 350 μm wide and of variable depth) cut with a diamond wheel. They were then seeded with rat calvarial osteoblasts and cultured in MEM with 10% FCS at 37°C in 5% CO2. Ascorbic acid 50 μg/mL and (β-glycero-phosphate 2 mmol/L were added at confluence. Cultures were observed daily from 2 to 4 weeks, until fixation (and storage) in 70% ethanol. Most were stained with alizarin red S to visualize the newly formed bone. The presence of gap junctions in the bone nodules was determined using connexin-43 immunolabeling and confocal microscopy. Two specimens were embedded in polymethylmethacrylate (PMMA): micromilled blockfaces were coated with carbon and examined by digital backscattered electron (BSE) microscopy. Bone formation began in the second week, preferentially wherever cellular condensation was favored: these locations were (a) within the grooves; (b) at the junction between the slab and the bottom of the culture dish; (c) at the periphery of the dish; and (d) in cracks where dissimilar tissues had separated. In the grooves, a grid of aligned bone developed, the deeper trenches showing bone formation earlier than shallower ones, with bone formation tapering off as a groove became shallower. BSE images showed that the bone formed was well mineralized and contained a high volume proportion of osteocytes. Mean and median values for the mean BSE coefficients were: in vitro bone in grooves 0.138528, 0.141484; in vivo aged bone (2 year old rat mandible) 0.143431, 0.144206; and in vivo young bone (neonate rat cranium) 0.129011, 0.132696. Connexin-43 gap junctions were immunolocalized on osteocytes fully enclosed within bone and on osteoblasts overlying it. We conclude first that local topography is an important factor in the location and timing of bone formation in vitro, and that it is likely to be equally important in vivo in normal bone turnover, fracture repair and the incorporation of bone grafts. Second, the mineral density of the bone formed in vitro is consistent with its being true bone. (Bone 18:115–123; 1996)