Diabetes mellitus is associated with bone loss. Patients with type 2 diabetes are frequently treated with oral antidiabetic drugs such as sulfonylureas, biguanides, and thiazolidinediones. Rosiglitazone treatment has been shown to increase adipogenesis in bone marrow and to induce bone loss. In this study we evaluated the effect of in vivo and in vitro treatment with metformin on bone marrow progenitor cells (BMPCs), as well as the involvement of AMPK pathway in its effects. The in vitro effect of coincubation with metformin and rosiglitazone on the adipogenic differentiation of BMPCs also was studied. In addition, we evaluated the effect of in vivo metformin treatment on bone regeneration in a model of parietal lesions in nondiabetic and streptozotocin-induced diabetic rats. We found that metformin administration both in vivo and in vitro caused an increase in alkaline phosphatase activity, type I collagen synthesis, osteocalcin expression, and extracellular calcium deposition of BMPCs. Moreover, metformin significantly activated AMPK in undifferentiated BMPCs. In vivo, metformin administration enhanced the expression of osteoblast-specific transcription factor Runx2/Cbfa1 and activation of AMPK in a time-dependent manner. Metformin treatment also stimulated bone lesion regeneration in control and diabetic rats. In vitro, metformin partially inhibited the adipogenic actions of rosiglitazone on BMPCs. In conclusion, our results indicate that metformin causes an osteogenic effect both in vivo and in vitro, possibly mediated by Runx2/Cbfa1 and AMPK activation, suggesting a possible action of metformin in a shift toward the osteoblastic differentiation of BMPCs. ß

An association has been previously established between uncompensated diabetes mellitus and the loss of bone mineral density and/or quality. In this study, we evaluated the effects of metformin on the growth and differentiation of osteoblasts in culture. Treatment of two osteoblast-like cells (UMR106 and MC3T3E1) with metformin (25-500 microM) for 24 h led to a dose-dependent increase of cell proliferation. Metformin also promoted osteoblastic differentiation: it increased type-I collagen production in both cell lines and stimulated alkaline phosphatase activity in MC3T3E1 osteoblasts. In addition, metformin markedly increased the formation of nodules of mineralization in 3-week MC3T3E1 cultures. Metformin induced activation and redistribution of phosphorylated extracellular signal-regulated kinase (P-ERK) in a transient manner, and dose-dependently stimulated the expression of endothelial and inducible nitric oxide synthases (e/iNOS). These results show for the first time a direct osteogenic effect of metformin on osteoblasts in culture, which could be mediated by activation/redistribution of ERK-1/2 and induction of e/iNOS.

The adhesion of osteoblasts to bone extracellular matrix, of which type-I collagen constitutes >85%, can modulate diverse aspects of their physiology such as growth, differentiation and mineralisation. In this study we examined the adhesion of UMR106 rat osteoblast-like cells either to a control (Col) or advanced-glycation-endproduct-modified (AGEs-Col) type I collagen matrix. We investigated the possible role of different integrin receptors in osteoblastic adhesion, by co-incubating these cells either with beta-peptide (conserved sequence 113-125 of the beta subunit of integrins) or with two other peptides, RGD (Arg-Gly-Asp) and DGEA (Asp-Gly-Glu-Ala), which are recognition sequences for the alpha-subunits of alpha(1,5)beta(1) and alpha(2)beta(1) integrins. Collagen glycation inhibited the adhesion of UMR106 osteoblasts to the matrix (40% reduction versus Col, P > 0.001). beta-Peptide showed a dose- and glycation-dependent inhibitory effect on adhesion, and at a concentration of 100 microM decreased the attachment of UMR106 cells to both matrices (42% to Col, P<0.001and 25% to AGEs-Col, P<0.01). The synthetic peptides RGD (1mM) and DGEA (5mM) inhibited the attachment of UMR106 cells to Col (30 and 20%, P > 0.01 and P< 0.001, respectively), but not to AGEs-Col. beta-Peptide induced an increase in UMR106 cell clumping and a decrease in cellular spreading, while DGEA increased spreading with cellular extensions in multiple directions. These results indicate that both alpha and beta integrin subunits participate in osteoblastic attachment to type-I collagen, probably through the alpha(1,5)beta(1) and alpha(2)beta(1) integrins. AGEs-modification of type-I collagen impairs the integrin-mediated adhesion of osteoblastic cells to the matrix, and could thus contribute to the pathogenesis of diabetic osteopenia.

Patients with long-standing Diabetes mellitus can develop osteopenia and osteoporosis. We have previously shown that advanced glycation endproducts reduce the bone-forming activity of osteoblasts. Bisphosphonates are used for the treatment of various bone disorders, since they reduce osteoclastic function and survival, and stimulate osteoblastic bone-forming capacity. In this work we have investigated whether bisphosphonates are able to revert advanced glycation endproducts-induced deleterious effects in osteoblasts. MC3T3E1 and UMR106 osteoblastic cells were incubated with control or advanced glycation endproducts-modified bovine serum albumin, in the presence or absence of different doses of the bisphosphonates Alendronate, Pamidronate or Zoledronate. After 24-72 h of culture, we evaluated their effects on cell proliferation and apoptosis, type-1 collagen production, alkaline and neutral phosphatase activity, and intracellular reactive oxygen species production. Advanced glycation endproducts significantly decreased osteoblast proliferation, alkaline phosphatase activity and type 1 collagen production, while increasing osteoblastic apoptosis and reactive oxygen species production. These effects were completely reverted by low doses (10 − 8 M) of bisphosphonates. High doses of bisphosphonates (10-10 − 5 M) were toxic for osteoblasts. Nifedipine (L-type calcium channel blocker) did not affect the advanced glycation endproducts-induced decrease in osteoblastic proliferation, although it blocked the reversion of this effect by 10 − 8 M Alendronate. Both advanced glycation endproducts and Alendronate inhibited the activity of intracellular neutral phosphatases. In conclusion, we show that bisphosphonates revert the deleterious actions of advanced glycation endproducts on osteoblastic cells, and that these effects of bisphosphonates depend on: (a) Ca 2+ influx through L-type voltage-sensitive channels, and (b) blockage of advanced glycation endproducts-induced reactive oxygen species generation.

Fructose-induced MS is associated with deleterious effects on bone microarchitecture and with a decrease in bone repair. These alterations could be due to a deviation in the adipogenic/osteogenic commitment of MSC, probably by modulation of the Runx2/PPARγ ratio.