Osteoblasts respond to microarchitectural features of their substrate. On smooth surfaces (tissue culture plastic, tissue culture glass, and titanium), the cells attach and proliferate but they exhibit relatively low expression of differentiation markers in monolayer cultures, even when confluent. When grown on microrough Ti surfaces with an average roughness (Ra) of 4-7 µm, proliferation is reduced but differentiation is enhanced and in some cases, is synergistic with the effects of surface microtopography. In addition, cells on microrough Ti substrates form hydroxyapatite in a manner that is more typical of bone than do cells cultured on smooth surfaces. Osteoblasts also respond to growth factors and cytokines in a surface-dependent manner. On rougher surfaces, the effects of regulatory factors like 1α,25(OH) 2 D 3 or 17β-estradiol are enhanced. The response to the surface is mediated by integrins, which signal to the cell through many of the same mechanisms used by growth factors and hormones. Studies using PEGmodified surfaces indicate that increased differentiation may be related to altered attachment to the surface. When osteoblasts are grown on surfaces with chemistries or microarchitectures that reduce cell attachment and proliferation, and enhance differentiation, the cells tend to increase production of factors like TGF-β1 that promote osteogenesis while decreasing osteoclastic activity. Thus, on microrough Ti surface, osteoblasts create a microenvironment conducive to new bone formation.

Titanium implant surfaces with rough microtopographies exhibit increased pullout strength in vivo suggesting increased bone-to-implant contact. This is supported by in vitro studies showing that as surface microroughness increases, osteoblast proliferation decreases whereas differentiation increases. Differentiation is further enhanced on microrough surfaces by factors stimulating osteogenesis including 1alpha,25(OH)2D3. Levels of PGE2 and TGF-beta1 are increased in cultures grown on rough microtopographies; this surface effect is enhanced synergistically by 1alpha,25(OH)2D3-treatment. PGE2 and TGF-beta1 regulate osteoclasts as well as osteoblasts, suggesting that surface microtopography may modulate release of other factors from osteoblasts that regulate osteoclasts. To test this hypothesis, we examined the effects of substrate microarchitecture on production of osteoprotegerin (OPG) and receptor activator of nuclear factor kappa B ligand (RANKL), which have been identified as a key regulatory system of bone remodeling. We also examined the production of 1alpha,25(OH)2D3, which regulates osteoblast differentiation and osteoclastogenesis. MG63 osteoblast-like cells were grown on either tissue culture plastic or titanium disks of different surface microtopographies: PT (Ra < 0.2 microm), SLA (Ra = 4 microm), and TPS (Ra = 5 microm). At confluence, cultures were treated for 24 h with 0, 10(-8) M or 10(-7) M 1alpha,25(OH)2D3. RANKL and OPG were determined at the transcriptional level by RT-PCR and real time PCR and soluble RANKL, OPG and 1alpha,25(OH)2D3 in the conditioned media were measured using immunoassay kits. Cell number was reduced on SLA and TPS surfaces and 1alpha,25(OH)2D3 caused further decreases. OPG mRNA levels increased on rougher surfaces and 1alpha,25(OH)2D3 treatment caused a further synergistic increase. While the cells expressed RANKL mRNA, levels were low and independent of surface microtopography. OPG protein was greater when cells were grown on SLA and TPS. 1alpha,25(OH)2D3 increased OPG by 50% on the smooth Ti surface but on SLA, 10(-8) M 1alpha,25(OH)2D3 caused a 100% increase and 10(-7) M 1alpha,25(OH)2D3 increased OPG by 200%. On TPS 10(-7) M 1alpha,25(OH)2D3 increased OPG 350%. Soluble RANKL was not detected in the conditioned media of any of the cultures. 1alpha,25(OH)2D3 was produced endogenously and levels were positively correlated with surface roughness. Thus, on surfaces with rough microtopographies, osteoblasts secrete factors that enhance osteoblast differentiation while decreasing osteoclast formation and activity.

Osteoblasts are attachment-dependent cells that interact with their surface through integrin-mediated mechanisms. Their differentiation is regulated by 1,25-dihydroxyvitamin D3 [1alpha,25(OH)(2)D(3)] and is affected by substrate chemistry and microtopography, suggesting that 1alpha,25(OH)(2)D(3) may regulate integrin expression in a surface-specific manner. To test this hypothesis, osteoblast-like human MG63 cells were grown on tissue culture plastic and on grit-blasted and acid-etched titanium disks with a complex microtopography to induce osteoblast differentiation. Expression of alpha(2), alpha(5), alpha(v), beta(1), and beta(3) integrins were quantified by real-time polymerase chain reaction (PCR) as a function of time in culture and treatment with 1alpha,25(OH)(2)D(3). Results were correlated with expression of osteocalcin, a marker of a differentiated osteoblast. Osteocalcin mRNA increased with time and 1alpha,25(OH)(2)D(3) treatment and these changes were greater in cultures on the titanium disks. Integrin expression varied with time in culture and this was also surface dependent. At each time point, beta(1) and alpha(2) mRNAs were greater on titanium than on plastic, whereas alpha(5) expression was reduced and alpha(v),beta(3) expression was unaffected. 1alpha,25(OH)(2)D(3) increased beta(1) mRNA on both surfaces at all time points, but it increased alpha(2) expression only in 8-d cultures. 1alpha,25(OH)(2)D(3) caused reduced alpha(5) expression only in cultures grown on plastic for 8 d, and had no effect on either alpha(v) or beta(3) expression regardless of surface. These results show that integrin expression in human osteoblast-like cells is differentially modulated by 1alpha,25(OH)(2)D(3) in a time-dependent manner that is sensitive to the surface on which the cells are grown.

Phospholipase A2 (PLA2) is pivotal in the rapid membrane-mediated actions of 1,25-dihydroxyvitamin D3 [1alpha,25(OH)2D3]. Microarray analysis indicated that PLA2 activating protein (PLAA) mRNA is upregulated 6-fold before rat growth plate cells exhibit 1alpha,25(OH)2D3-dependent protein kinase C (PKC) increases, suggesting that it plays an important role in 1alpha,25(OH)2D3's mechanism of action. PLAA mRNA was confirmed in 1alpha,25(OH)2D3-responsive growth zone (prehypertrophic and upper hypertrophic cell zones) chondrocytes by RT-PCR and Northern blot in vitro and by in situ hybridization in vivo. PLAA protein was shown by Western blot and immunohistochemistry. PLAAs role in 1alpha,25(OH)2D3 signaling was evaluated in growth zone cell cultures using PLAA peptide. Arachidonic acid release was increased as was PLA2-specific activity in plasma membranes and matrix vesicles. PKCalpha, but not PKCbeta, PKCepsilon, or PKCzeta, was increased. PLAAs effect was comparable to that of 1alpha,25(OH)2D3 and was additive with 1alpha,25(OH)2D3. PLA2 inhibitors quinacrine and AACOCF3, and cyclooxygenase inhibitor indomethacin blocked the effect of PLAA peptide on PKC, indicating arachidonic acid and its metabolites were involved. This was confirmed using exogenous arachidonic acid. Prostaglandin acted via EP1 based on inhibition by SC19220 and not via EP2 since AH6809 had no effect. Like 1alpha,25(OH)2D3, PLAA peptide also increased activity of phospholipase C-specific activity via beta-1 and beta-3 isoforms, but not delta-1 or gamma-1; the effect of PLAA was via lysophospholipid but not via arachidonic acid. PLAA peptide decreased [3H]-thymidine incorporation to 50% of the decrease caused by 1alpha,25(OH)2D3. In contrast, PLAA peptide increased alkaline phosphatase-specific activity and proteoglycan production in a manner similar to 1alpha,25(OH)2D3. This indicates that PLAA is a specific activator of PLA2 in growth plate chondrocytes, and suggests that it mediates the membrane effect of 1alpha,25(OH)2D3, thereby modulating physiological response.