Purpose. The data regarding the effects of noggin on bone morphogenetic protein (BMP)-induced osteogenesis of mesenchymal stem cells (MSCs) are controversial. Most studies performed in rodent cells/models indicated that noggin was a negative regulator of BMP-2-induced osteogenesis; however, one study conducted with human MSCs in culture showed that the addition of noggin induced osteogenesis in vitro. To clear the controversy, we designed this study to evaluate the effects of knocking down noggin gene expression on BMP-2-induced osteogenesis of human bone marrow-derived primary MSCs in vitro. Method. MSCs were isolated from human tibial bone marrow by density gradient centrifugation. Two noggin small interfering RNAs (siRNAs) were used in this study to knockdown noggin gene expression. There were four study groups: MSCs with no transfection of siRNA (named as NT group), MSCs transfected with non-targeting negative control siRNA (named as control group), MSCs transfected with noggin siRNA1 (named as NOGsi1 group), and MSCs transfected with noggin siRNA2 (named as NOGsi2 group). After transfection, MSCs were induced to undergo osteogenic differentiation by incubating in basal medium containing 0.1 μg/ml BMP-2 for 35 days. The expression levels of osteoblastic marker genes were measured by real-time quantitative PCR on day 14. Also assessed was alkaline phosphatase (ALP) activity by a colorimetric kinetic assay and Fast Blue B staining on day 14. Calcium deposition was determined by the calcium assay on day 35. Results. The expression levels of integrin binding sialoprotein (IBSP) and osteocalcin (OC) were significantly decreased in both NOGsi1 and NOGsi2 groups compared with NT and control groups (all p<0.038). Although the expression level of runt-related transcription factor 2 (RUNX2) was also reduced in NOGsi1 and NOGsi2 groups compared with NT and control groups, it did not reach statistical significance. ALP activity was significantly lower in NOGsi1 and NOGsi2 groups than that of NT group (both p<0.024). The same pattern was also observed in ALP Fast Blue B staining. Calcium deposition was also significantly decreased in both NOGsi1 and NOGsi2 groups compared with NT group (both p<=0.048). Conclusion. Noggin suppression by siRNA inhibits BMP-2-induced osteogenesis of human bone marrow-derived MSCs. Our results, contrary to the extensive studies conducted in rodent cells/models, corroborated with the previous study that the addition of noggin in the cell culture increased osteogenesis of human MSCs. This suggests that the effects of noggin on BMP-2-induced osteogenesis of MSCs might be species-specific

Biomechanical stimuli have fundamental roles in the maintenance and remodeling of ligaments including collagen gene expressions. Mechanical stretching signals are mainly transduced by cell adhesion molecules such as integrins. However, the relationships between stress-induced collagen expressions and integrin-mediated cellular behaviors are still unclear in anterior cruciate ligament cells. Human ACL cells were harvested from ligament samples donated by patients who underwent total knee arthroplasties with informed consents. Interface cells were isolated from the 5-mm-end of ACL. Midsubstance cells were cultured from the middle part of ACL. The cells were seeded onto stretch chambers (2Ä−2 cm, 50,000 cells/chamber) and uni-axial cyclic mechanical stretch (0.5 Hz, 7%) was applied for 2 h using a ST140. RNA samples were reverse-transcripted and quantitative real-time RT-PCR analysis were performed. To inhibit the function of integrin alphaVbeta3 subunit or alpha5 in stretching experiments, anti-human integrin alphaVbeta3 and alpha5 functional blocking antibodies (alphaVbeta3: 20 mg/ml, alpha5: 4 mg/ml) were used. To investigate the cellular attachments responding to mechanical stretch, we observed the distribution of integrins and stress fibers in both ACL cells. The shape of midsubstance cells showed spindle and fibroblastic cellular morphologies. On the other hand, the interface cells displayed chondroblastic appearances such as small and triangular morphologies. The expressions of COL1A1, COL2A1, and COL3A1 genes were detected in the tissue RNAs of interface zones. However, these expressions were decreased in cultured interface cells. In midsubstance cells, the expression of COL1A1 gene was equally detected in both tissues and cultured cells. COL3A1 gene expression was maintained in cultured midsubstance cells. These results indicated that the phenotypes of both ACL cells were changed by cultured conditions, especially in the interface cells. After mechanical stretch, the COL1A1 expression of midsubstance and interface cells were stimulated up to 6 and 14-fold levels of each non-stretched control, respectively. The COL3A1 expressions were also enhanced up to 1.8-fold level of controls by stretching treatment in both cells. Integrin alphaVbeta3 was shifted to the peripheral edge of cells by stretching treatment. In addition, mechanical stretch changed the integrin alphaVbeta3-dependent stress fiber formation in both ACL cells. The functional blocking of integrin alphaVbeta3 inhibited stretch-activated COL1A1 and COL3A1 expressions. However, the functional blocking of integrin alpha5 did not suppress the stretch-induced COL1A1 and COL3A1 expressions in both ACL cells. Cultured interface cells loose their phenotypes in collagen gene expressions. However, mechanical stretch reproduces the expression of COL1A1 and COL3A1 genes in cultured ACL cells. The present study demonstrated that stretch-activated collagen gene expressions depend on the integrin alphaVbeta3-mediated cellular adhesions