During endochondral bone formation, as occurs in fracture healing, chondrocytes are one of the first cells to see an implant surface. We tested the hypothesis that chemical composition and surface roughness affect chondrocyte differentiation, matrix synthesis, and local factor production and that the nature of the response is dependent on the state of maturation of the cells. To do this, we harvested rat growth zone and resting zone chondrocytes and examined their response to smooth and rough disk surfaces manufactured from either commercially pure titanium or titanium alloy. Profilometry, scanning electron microscopy, Auger spectroscopy, and Fourier transform infrared spectroscopy were used to characterize the surfaces. Average roughness values were 0.22 microm for smooth titanium surfaces, 0.23 microm for smooth titanium alloy surfaces, 4.24 microm for rough titanium surfaces, and 3.20 microm for rough titanium alloy surfaces. Cells were grown on the different disk surfaces until the cultures had reached confluence on plastic. The effect of the surfaces was determined by assaying cell number and [3H]thymidine incorporation as measures of cell proliferation, cell layer and cell alkaline phosphatase specific activity as markers of differentiation, and collagen production and [35S]sulfate incorporation as indicators of extracellular matrix production. In addition, the synthesis of prostaglandin E2 and transforming growth factor-beta were examined to measure changes in local factor synthesis. In growth zone and resting zone cultures, cell number and [3H]thymidine incorporation were decreased on rough surfaces; however, this effect was greater on commercially pure titanium surfaces. Cell layer and cell alkaline phosphatase specific activity were decreased in resting zone cells grown on rough surfaces. Cell alkaline phosphatase specific activity in growth zone cells was decreased on rough surfaces, whereas cell layer alkaline phosphatase specific activity was increased only in growth zone cells grown on rough commercially pure titanium surfaces. Resting zone cell collagen production was decreased only on rough commercially pure titanium, whereas in growth zone cells, collagen production was increased. Increased prostaglandin E2 release into the media was found for growth zone and resting zone cell cultures on the disks with rough surfaces. The observed effect was greater on rough commercially pure titanium. Production of transforming growth factor-beta by resting zones was similarly affected, whereas an increase in its production by growth zone cells was measured only on rough commercially pure titanium. These results indicate that surface roughness affects chondrocyte proliferation, differentiation, matrix synthesis, and local factor production and that these parameters are also affected by chemical composition. Furthermore, the nature and extent of the cell response is dependent on cell maturation. The overriding variable in response to an implant material, however, appears to be roughness of the surface.
