R M Locklin

Effects of TGFβ and BFGF on the differentiation of human bone marrow stromal fibroblasts

Adipocytes and osteoblasts have common origins from fibroblastic stem cells. Consequently, modulation of the processes of adipogenesis and osteogenesis has implications for the possible treatment of metabolic bone diseases, such as osteoporosis, in which medullary fat accumulates and trabecular bone volume decreases. It is likely that the balance between these two systems is affected by particular endogenous growth factors which are known to affect bone metabolism. We have therefore investigated the effects of transforming growth factor beta (TGFβ), basic fibroblast growth factor (bFGF) and dexamethasone (Dex) on cultured human bone marrow (HBM) fibroblastic cells to observe the effects on adipogenesis and osteogenesis. In the absence of fetal calf serum (FCS), TGFβ caused a dose‐dependent increase in cell growth and alkaline phosphatase activity (AP); however, in the presence of FCS growth was inhibited at high concentrations and AP unaffected. TGFβ increased matrix proteoglycan and collagen synthesis. bFGF inhibited AP and increased colony number and size, while Dex treatment increased AP activity and colony number, and both factors in combination resulted in an additive increase in growth. Dex‐induced adipocyte formation was accelerated but not increased by bFGF. A significant inhibition of adipogenesis by TGFβ was observed within 7 days. These results demonstrate the importance of biological factors known to be involved in bone remodelling in the regulation of osteogenesis and adipogenesis.

Differences between bisphosphonates in binding affinities for hydroxyapatite

Bisphosphonates (BPs) inhibit bone resorption and are widely used for the treatment of bone diseases, including osteoporosis. BPs are also being studied for their effects on hydroxyapatite (HAP)-containing biomaterials. There is a growing appreciation that there are hitherto unexpected differences among BPs in their mineral binding affinities that affect their pharmacological and biological properties. To study these differences, we have developed a method based on fast performance liquid chromatography using columns of HAP to which BPs and other phosphate-containing compounds can adsorb and be eluted by using phosphate buffer gradients at pH 6.8. The individual compounds emerge as discrete and reproducible peaks for a range of compounds with different affinities. For example, the peak retention times (min; mean +/- SEM) were 22.0 +/- 0.3 for zoledronate, 16.16 +/- 0.44 for risedronate, and 9.0 +/- 0.28 for its phosphonocarboxylate analog, NE10790. These results suggest that there are substantial differences among BPs in their binding to HAP. These differences may be exploited in the development of biomaterials and may also partly explain the extent of their relative skeletal retention and persistence of biological effects observed in both animal and clinical studies.

MODULATION OF OSTEOGENIC DIFFERENTIATION IN HUMAN SKELETAL CELLS IN VITRO BY 5‐AZACYTIDINE

Cellular differentiation is controlled by a variety of factors including gene methylation, which represses particular genes as cell fate is determined. The incorporation of 5‐azacytidine (5azaC) into DNA in vitro prevents methylation and thus can alter cellular differentiation pathways. Human bone marrow fibroblasts and MG63 cells treated with 5azaC were used as models of osteogenic progenitors and of a more mature osteoblast phenotype, respectively. The capacity for differentiation of these cells following treatment with glucocorticoids was investigated. 5azaC treatment led to significant expression of the osteoblastic marker alkaline phosphatase in MG63 osteosarcoma cells, which was further augmented by glucocorticoids; however, in human marrow fibroblasts alkaline phosphatase activity was only observed in glucocorticoid‐treated cultures. MG63 cells represent a phenotype late in the osteogenic lineage in which demethylation is sufficient to induce alkaline phosphatase activity. Marrow fibroblasts are at an earlier stage of differentiation and require stimulation with glucocorticoids. In contrast, the expression of osteocalcin, an osteoblastic marker, was unaffected by 5azaC treatment, suggesting that regulation of expression of the osteocalcin gene does not involve methylation. These models provide novel approaches to the study of the control of differentiation in the marrow fibroblastic system.

Selective targeting of death receptor 5 circumvents resistance of MG-63 osteosarcoma cells to TRAIL-induced apoptosis

Tumor necrosis factor–related apoptosis-inducing ligand (TRAIL), a tumor necrosis factor superfamily member, targets death receptors and selectively kills malignant cells while leaving normal cells unaffected. However, unlike most cancers, many osteosarcomas are resistant to TRAIL. To investigate this resistance, we characterized the response of MG-63 osteosarcoma cells and hPOB-tert osteoblast-like cells to TRAIL and agonist antibodies to death receptor 4 (DR4) and death receptor 5 (DR5). We found that MG-63 osteosarcoma cells and hPOB-tert osteoblast-like cells show no or very little response to TRAIL or a DR4 agonist, but MG-63 cells undergo apoptosis in response to a DR5 agonist. Analysis of TRAIL receptor expression showed that normal osteoblastic and osteosarcoma cells express a variety of TRAIL receptors but this does not correlate to TRAIL responsiveness. Production of the soluble decoy receptor osteoprotegerin also could not explain TRAIL resistance. We show that TRAIL activates the canonical caspase-dependent pathway, whereas treatment with cycloheximide increases the sensitivity of MG-63 cells to TRAIL and anti-DR5 and can also sensitize hPOB-tert cells to both agents. Proapoptotic and antiapoptotic protein expression does not significantly differ between MG-63 and hPOB-tert cells or change following treatment with TRAIL or anti-DR5. However, sequencing the death domain of DR4 in several osteoblast-like cells showed that MG-63 osteosarcoma cells are heterozygous for a dominant-negative mutation, which can confer TRAIL resistance. These results suggest that although the dominant-negative form of the receptor may block TRAIL-induced death, an agonist antibody to the active death receptor can override cellular defenses and thus provide a tailored approach to treat resistant osteosarcomas. [Mol Cancer Ther 2007;6(12):3219–28]

Innate immune response to human bone marrow fibroblastic cell implantation in CB17 scid/beige mice.

Immunocompromised mouse models have been extensively used to assess human cell implantation for evaluation of cytotherapy, gene therapy and tissue engineering strategies, as these mice are deficient in T and B lymphoid cells. However, the innate immune response and its effect on human cell xenotransplantation in these mouse models are mainly unknown. The aim of this study is to characterise the myeloid populations in the spleen and blood of CB17 scid beige (CB17 sb) mice, and to study the inflammatory cell responses to xenogeneic implantation of enhanced green fluorescent protein (GFP)‐labelled human bone marrow fibroblastic (HBMF) cells into CB17 sb mice. The results indicate that even though CB17 sb mice are deficient in B‐ and T‐cells, they exhibit some increases in their monocyte (Mo), macrophage (MΦ) and neutrophil (Neu) populations. NK cell and eosinophil populations show no differences compared with wild‐type Balb/C mice. An innate immune response, identified by CR3 (CD11b/CD18)‐positive myeloid inflammatory cells and F4/80‐positive macrophages, was evident in the tissues where HBMF cells were implanted. As a consequence, the majority of implanted HBMF cells were eliminated by 4 weeks after implantation. Interestingly, the mineralised matrix formed by osteogenic HBMF cells was also eroded by multinuclear MΦ‐like giant cells. We conclude that CB17 sb mice retain active innate immune cells, which respond to HBMF cell xenotransplantation. This study highlights the importance of the innate immune cells in the anti‐xenograft response and suggests that strategies to block the activities of these cells may ameliorate the progressive long‐term elimination of xenotransplants. J. Cell. Biochem. 98: 966–980, 2006.

Stimulation of osteoclast formation by inflammatory synovial fluid

Peri-articular bone resorption is a feature of arthritis due to crystal deposition and rheumatoid disease. Under these conditions, the synovial fluid contains numerous inflammatory cells that produce cytokines and growth factors which promote osteoclast formation. The aim of this study was to determine whether inflammatory synovial fluid stimulates the formation of osteoclasts. Synovial fluid from rheumatoid arthritis (RA), pyrophosphate arthropathy (PPA) and osteoarthritis (OA) patients was added to cultures (n=8) of human peripheral blood mononuclear cells (PBMCs) in the presence and absence of macrophage colony-stimulating factor (M-CSF) and the receptor activator of NF-κB ligand (RANKL). Osteoclast formation was assessed by the formation of cells positive for tartrate-resistant acid phosphatase (TRAP) and vitronectin receptor (VNR) and the extent of lacunar resorption. The addition of 10% OA, RA and PPA synovial fluid to PBMC cultures resulted in the formation of numerous multinucleated or mononuclear TRAP+ and VNR+ cells which were capable of lacunar resorption. In contrast to PBMC cultures incubated with OA synovial fluid, there was marked stimulation of osteoclast formation and resorption in cultures containing inflammatory RA and PPA synovial fluid which contained high levels of tumour necrosis factor alpha, a factor which is known to stimulate RANKL-induced osteoclast formation.

Fates and osteogenic differentiation potential of human mesenchymal stem cells in immunocompromised mice

Human mesenchymal stem cells (hMSCs) from bone marrow were genetically marked by using a murine leukaemia virus construct encoding enhanced green fluorescent protein (eGFP). The marked cells were either directly implanted into the tibialis anterior muscle or introduced into a variety of other tissue sites in immunocompromised mice (NOD/SCID and C.B-17 SCID/beige) to investigate their fates and differentiation potentials. It was observed that the hMSCs survived for up to 12 weeks and showed site-specific morphological phenotypes. hMSCs delivered by intravenous injection were found mainly in the lungs and were detected rarely in other organs. Histomorphometry showed that, after implantation of hMSCs into the tibialis anterior muscle juxtaskeletally, the areas of reactive host callus formation at 1 and 2 weeks and of ectopic human bone formation at 1 week were significantly increased compared with the control group. Expression of eGFP and human RUNX2, alkaline phosphatase, osteocalcin, osteopontin, and collagen type I mRNAs were detected in mice implanted with the labelled hMSCs but not in sham-treated samples. Active clearance of the reactive callus and ectopic calcified tissue by osteoclast-like tartrate-resistant acid phosphatase-positive cells was observed. We conclude that the eGFP-labelled hMSCs can survive and retain the potential to differentiate morphologically into a variety of apparent mesenchymal phenotypes in vivo. Absolute confirmation of differentiation capacity requires further study and is complicated by known possibilities of fusion of donor and host cells or limited transfer of genetic material. Nevertheless, the genetically marked hMSCs are shown to participate extensively in bone formation and turnover. Control of the host osteoclast/macrophage responses resulting in clearance of formed osteogenic tissue warrants further investigation to promote prolonged human osteogenesis in immunocompromised mice. Furthermore, any proposed general cytotherapeutic strategy for enhanced osteogenesis is likely to require supplementation of local bone-forming biological signals.