The work presented in this thesis focuses on the development of strategies and smart bioactive materials for the treatment of osteoporosis. High and low molecular weight soluble hyaluronic acid-bisphosphonate (HA-BP) derivatives were investigated for their ability to inhibit osteoclasts. Low molecular weight HA-BP (L-HA-BP) was most effective in inhibiting active resorption of both murine and human osteoclasts (without affecting osteoblasts) compared to free bisphosphonate (BP). Precursor monocytes were unaffected, suggesting the specificity of HA-BP towards osteoclasts. This new class of functionalized hyaluronic acid could lead to rapid development of tailor-made pro-drugs for targeted treatment of osteoporosis. Polyphosphoesters (PEP) have been widely studied for their pro-osteoblast effects, primarily due to their involvement in cellular energy production pathway leading to the formation of inorganic phosphates that contribute to mineralized bone. Given that the effect of PEP on human osteoclasts is little studied, this work on poly(ethylene sodium phosphate) (PEP.Na) explores the potential to use PEP.Na as an inhibitor of osteoclast activity for the first time. PEP.Na exposure led to a dose-dependent toxicity of osteoclasts with reduction in their capacity to form resorption pits over 24h. Currently, there is a dearth of in vitro cell-culture systems that can study osteoclast-related resorption and osteoblast-related mineralization in a single co-culture system, and to simultaneously quantify the effects of soluble factors on these processes. Described here, is the development of a novel and simple two-sided co-culture system that can overcome these limitations with reliable and quantifiable readouts. In comparison with traditional one-sided co-culture systems, the two-sided co-culture was able to generate similar readouts for alkaline phosphatase (ALP) and tartrate-resistant acid phosphatase (TRAP) markers. There is also the advantage of distinctly separate and quantifiable readouts for mineralization and resorption, which has been demonstrated using Pamidronate. Finally, HA-BP was synthesized with pre-determined amounts of BP groups. The BP groups attached to HA allowed the tunable incorporation of BMP-2 in hydrogels. The charge-based affinity of BMP-2 and BP allowed stable incorporation of specific amounts of BMP-2, which could be tuned by the ratio of BP groups. 125I-labelled BMP-2 was loaded into hydrogels and their release was studied. Radioactive measurements revealed the tunable sequestration and controlled release of protein over time. This result was corroborated by ALP measurements of cells exposed to released BMP-2. ALP production was found to be almost 5-fold higher in HA-BP hydrogels loaded with BMP-2 which suggested that the sequestered BMP-2 is not only available to cells but also remains highly potent, even in entrapped form, The release of BMP-2 is dependent upon the rate of diffusion, swelling in hydrogels and degradation pattern of the gels and may assist in the long-term and rapid regeneration of osteoblasts in vitro.

PHEX est une protéine importante dans le processus de minéralisation osseuse. Des mutations ou la délétion d’une partie de ce gène causent l’hypophosphatémie liée au chromosome X (XLH). Cette maladie est caractérisée par une hypophosphatémie, accompagnée de défauts de minéralisation, de rachitisme et de lésions ostéomalaciques. Avec l’hypophosphatémie, les taux circulants de vitamine D devraient être augmentés, ce qui n’est pas le cas d’où une régulation anormale de la production de vitamine D a lieu. Cependant, malgré le fait que cette protéine soit une peptidase, aucun substrat physiologique n’a encore été répertorié pour PHEX. PHEX est une protéine membranaire de type II de la famille M13 des métalloendopeptidases à zinc possédant un court domaine N-terminal cytosolique, un segment transmembrannaire d’environ 20 acides aminés et une large portion C-terminale extracellulaire où se trouve le site actif de l’enzyme. PHEX est exprimée de façon majoritaire dans les os et dans les dents et elle apparaît à l’initiation de la minéralisation. Les patients souffrant de XLH et la souris Hyp, qui est un modèle animal de la maladie humaine, montrent des quantités importantes de la protéine FGF23. De plus, FGF23 est impliqué dans une autre maladie reliée au métabolisme du phosphate, l’hypophosphatémie rachitique autosomale dominante (ADHR) où des mutations de FGF23 causent sensiblement les mêmes symptômes que XLH. FGF23 est produit principalement par les ostéoblastes et les ostéocytes. FGF23 cause une hypophosphatémie par la diminution de l’expression du cotransporteur NaPi de type II, responsable de la réabsorption du phosphate rénal. L’hypothèse proposée dans la littérature serait que PHEX activerait ou inactiverait des peptides importants pour la minéralisation osseuse. Plus spécifiquement, l’activation ou l’inactivation de ces peptides aurait pour rôle de réguler les quantités de FGF23. Selon l’hypothèse mentionnée précédemment, la régulation de PHEX pourrait donc avoir un effet sur la minéralisation. Une quantité croissante de données sur la régulation de PHEX sont maintenant disponibles. Par exemple, la vitamine D diminue l’expression de PHEX tandis que les glucocorticoïdes et l’hormone de croissance augmentent son expression. Dans une première étude, nous avons voulu déterminer si un peptide relié à la minéralisation osseuse, le PTHrP1-34, pouvait réguler l’expression de PHEX. Nous avons déterminé que le PTHrP1-34 peut réguler de façon négative l’expression de PHEX dans les cellules UMR-106, une lignée cellulaire ostéoblastique. Cette régulation passe par la voie de l’AMPc/protéine kinase A. De plus, cette diminution d’expression est également observée au jour 7 dans des cultures primaires d’ostéoblastes de rat en minéralisation. Par la suite, nous avons étudié un mutant de PHEX, le mutant E4Q retrouvé chez un patient souffrant de XLH, où la mutation se retrouve dans le domaine cytosolique de PHEX. Cette mutation n’interfère pas avec le site catalytique de l’enzyme puisque ce mutant de PHEX peut tout aussi bien cliver un substrat synthétique que la protéine sauvage. Il a été déterminé que cette mutation annule un motif di-acide. Nous avons démontré que ce motif di-acide est responsable de la liaison de PHEX à COPII, responsable de la formation de vésicules de sécrétion. De plus, il semblerait que ce motif soit important, probablement par son interaction avec COPII, à l’incorporation de PHEX dans des vésicules de calcification, lesdites vésicules étant importantes dans le processus de minéralisation. Finalement, des essais de compétitions ont démontré que la minéralisation pouvait être perturbée lorsque l’on surexprimait la queue cytosolique sauvage de PHEX, contrairement à la queue mutée. Ceci suggère possiblement que l’interaction avec COPII menant à l’incorporation de PHEX dans les vésicules de calcification ou d’autres protéines comprenant de tels motifs pourrait être importante pour la minéralisation. Finalement, la dernière étude porte sur la protéine FGF23. Nous avons démontré, par la surexpression de FGF23 dans la lignée MC3T3 d’ostéoblastes de souris, que cette surexpression a un effet sur la sénescence de ces cellules. En effet, des essais de sénescence ont montré l’augmentation de celle-ci lorsque FGF23 est surexprimé. Par contre, la prolifération n’est pas altérée. De plus, il semblerait que la différenciation soit plus rapide, tel qu’observé par une minéralisation survenant plus tôt, mais n’étant pas plus importante. Bref, la surexpression de FGF23 semblerait faire en sorte que les ostéoblastes se différencient plus rapidement et passent donc à un état de sénescence prématuré comparativement aux cellules sauvages. Ceci est en accord avec la littérature où KLOTHO, un cofacteur de FGF23 permettant sa liaison avec une plus grande affinité sur son récepteur, lorsqu’inactivé démontre un phénotype similaire au vieillissement incluant un phénotype de sénescence. / PHEX is an important protein in the process of osseous mineralisation. Mutations or deletions of a part of the PHEX gene cause X-linked hypophosphatemia (XLH). This disease is characterized by hypophosphatemia, accompanied by defects of bone mineralisation, rickets and osteomalacia. With the hypophosphatemia, the circulating levels of vitamin D should be increased, which is not the case where an abnormal regulation of the production of vitamin D takes place. However, in spite of the fact that this protein is a peptidase, no physiological substrate has been identified. PHEX is a membrane type II integral protein member of the M13 family of zinc metalloendopeptidasee. These proteins have a short N-terminal cytosolic domain, a transmembrane domain of approximately 20 amino acids and a large extracellular C-terminal portion where the active site of the enzyme is located. PHEX is expressed predominantly in bones and teeth in osteoblasts and odontoblasts, respectively. PHEX is expressed at initiation of mineralization. Patients suffering from XLH and the Hyp mouse, which has been used widely as an animal model of the human disease, show large quantities of the FGF23 protein. Moreover, FGF23 is implicated in another disease connected to phosphate metabolism, the autosomal dominant hypophosphatemic rickets (ADHR) where activating mutations in FGF23 cause roughly the same symptoms as XLH. FGF23 is produced mainly by osteoblasts and osteocytes. FGF23 causes hypophosphatemia by decreasing the expression of the type II sodium phosphate cotransportor, partly responsible for renal phosphate reabsorption. The hypothesis suggested in the literature would be that PHEX would activate or inactivate important peptides for osseous mineralisation. More specifically, the activation or the inactivation of these peptides would have a role in the control of FGF23 expression. According to the assumption mentioned previously, the regulation of PHEX could thus have an effect on mineralization. An increasing quantity of data on the regulation of PHEX is now available. For example, vitamin D decreases the expression of PHEX while glucocorticoids and growth hormone increase its expression. In a first study, we examined the possibility that a peptide connected to osseous mineralization could control the expression of PHEX. We determined that PTHrP1-34 can control in a negative way the expression of PHEX in UMR-106 cells, a cell line of osteoblastic origin. This regulation involves the cAMP/protein kinase A pathway. Moreover, this decrease in PHEX expression is also observed at day 7 in primary cultures of mineralizing rat osteoblasts. Next, we looked more closely at PHEX cellular localization. We used a mutant of PHEX, mutant E4Q identified in an XLH patient, where the mutated amino acid is found in the cytosolic domain of PHEX. This change does not interfere with the catalytic site of the enzyme since this PHEX mutant can still cleave a synthetic substrate as well as wildtype protein. This mutation disrupts a di-acidic motif present in the cytosolic domain of PHEX. We showed that this di-acidic motif is reponsible for the interaction of PHEX with COPII, a protein complex involved in the formation of secretion vesicles. Moreover, it would seem that this di-acidic motif is important, probably by its interaction with COPII, to the incorporation of PHEX in matrix vesicles, which are important in the mineralization process. Finally, competition assays showed that mineralization could be disturbed when the wildtype PHEX cytosolic tail is overexpressed, as opposed with the mutated cytosolic tail. This suggests that the interaction with COPII and the subsequent incorporation of PHEX in matrix vesicles or other proteins that possesses this motif could be important for mineralization. Finally, the last study examined the role of FGF23 on mineralization. We showed, by the overexpression of FGF23 in the mouse MC3T3 osteoblast cell line, that FGF23 can cause senescence of these cells. On the other hand, proliferation is not affected. Moreover, differentiation seems to occur at a faster rate, as indicated by earlier mineralization. Overexpression of FGF23 would accelerate differentiation and induce senescence. This is in agreement with the literature where KLOTHO, a FGF23 cofactor that increase the affinity of FGF23 for its receptor, when inactivated, shows a similar phenotype that includes senescence and aging.

The objective of the present study was to synthesize porous, biodegradable poly (D, l- lactide-co-glycolide) PLGA-B-HA (Bovine hydroxyapatite) composite and evaluate the effect of ceramic content on bone marrow cell differentiation in vitro. A macroporous biodegradable PLGA-B-HA composite with the pore size varying from 0.1 to 1000?? and a highly interconnected structure was fabricated using the freeze-drying/lyophilization technique. A pilot study was done to determine the effects of B-HA on to the osteoblast function. The main study was done to determine the effect of the increase in B-HA concentration on to the mesenchymal stem cell differentiation. Morphological characteristics of the composites were analyzed using FTIR and SEM/EDX analysis. The composites were seeded with neonatal rat calvarial osteoblasts (NRCO). The polymer: ceramic ratio in this study was 35%:65%. For comparison parallel experiments involving pure HA-200 discs were performed. SEM results indicated a higher proliferation and mineralization on PLGA-B-HA composites than pure HA discs. In addition, we evaluated the in vitro characteristics of PLGA-B-HA composites with varying ratios, i.e., 1:1, 1:2 and 1:3, seeded with rat marrow cells. FTIR indicated an increase in the area under the ceramic peak as ceramic concentration was increased. In addition, the average roughness values increased in the order of 1:3 andgt; 1:2 andgt; 1:1. Both compressive strength and modulus of 1:1 were significantly higher than 1:2 and 1:3 PLGA-B-HA composites. No significant difference in compressive modulli and strengths could be observed for 1:2 and 1:3 PLGA-B-HA composites. Cellular activity was determined by measuring AP activity, total protein analysis and osteocalcin concentration. Evaluation of alkaline phosphatase activity showed bone cells attached to 1:3 (PLGA-B-HA) expressed significantly higher alkaline phosphatase as compared to 1:1 and 1:2 PLGA-B-HA composites. In addition, cells seeded on to 1:3 composites secreted significantly higher osteocalcin and at a relatively short time period as compared to the other samples. Corrosion studies (ICP) and pH values indicate minimal difference in the concentration of Ca and P and pH in tissue culture media for all the samples at the end of all time periods. Hence we conclude that an increase in the ceramic concentration stimulated mesenchymal stem cell differentiation thereby promoting osteogenesis.

The neural crest (NC) derives from the dorsal borders of the vertebrate neural tube. During development, the NC cells migrate and contribute to the formation of different tissues and organs. Along the anteroposterior axis, the NC gives rise to neurons and glia of the peripheral nervous system and to melanocytes. Furthermore, the cephalic NC yields mesenchymal tissues, which form all facial cartilages and bones, the large part of skull, facial dermis, fat cells and smooth muscle cells in the head. In the trunk of amniotes Vertebrates, these tissues are derived from the mesoderm, not from the NC. In lower Vertebrates, however, the trunk NC generates some mesenchymal tissues, such as in the dorsal fins of zebrafish. The question therefore is raised whether the ability of the NC to produce mesenchymal cells was totally lost in the trunk of amniote Vertebrates during evolution, or if it can still be achieved under specific conditions. This work is interested in uncovering the mesenchymal potential of the avian trunk NC, with special interest in the differentiation into osteoblasts and adipocytes.Our experimental approach was to examine the skeletogenic and adipogenic differentiation potentials of quail trunk NC cells after in vitro culture. Cell differentiation was evidenced by the analysis of lineage-specific genes and markers using in situ hybridization (ISH), immunocytochemistry and RT-PCR. The established culture conditions allowed observation of both skeletogenesis and adipogenesis. Osteogenesis was initially characterized by expression of Runx2, the first transcription factor specific of the osteoprogenitors, which was detected by ISH from 5 days of culture. Later, we observed osteoblast maturation, with the expression of collagen1 protein, osteopontin mRNA and alkaline phosphatase mRNA, until the bone matrix mineralization stage. The trunk NC cells also underwent chondrogenesis, as demonstrated by Sox9, aggrecan and collagen10 mRNA expression, and Alcian blue staining. The observation of the mineralized areas and chondrogenesis suggested that the trunk NC cells in vitro are able to perform endochondral and membranous ossifications. In same culture conditions, the cells differentiated also into adipocytes, identified from 10 days of culture by Oil Red O staining. The mRNAs of the CEBP, PPAR and FABP4 adipogenic markers were detected by RT-PCR from 3 days of culture. For the characterization of bone and adipocyte progenitors, we evaluated the differentiation potential of individual trunk NC cells. The phenotypic analysis of these clonal cultures showed that 76% of the cells generated Runx2-positive osteoblasts. Moreover, most of the clone-forming trunk NC cells were multipotent progenitors endowed with both neural and osteogenic potentials. Furthermore, in another clonal culture condition, adipocytes were found in 35.3% of the clones, and approximately half of them also contained glial and/or melanogenic cells.These results show that the trunk NC cells in vitro are able to differentiate not only in their classical derivatives found in vivo (melanocytes, neurons and glial cells), but also in mesenchymal phenotypes, including adipocytes and osteoblasts. Importantly, as in cephalic NC cells, mesenchymal phenotypes differentiated from multipotent progenitor cells, suggesting that, during evolution, the NC stem cells intended for both mesenchymal and neural fates, had the expression of their mesenchymal potential inhibited in the trunk. Thus, although at the dormant state and not expressed in vivo, a significant mesenchymal potential is present in the trunk NC cells of amniotes Vertebrates and can be disclosed in vitro

Advanced glycation end products (AGEs) irreversibly cross-link proteins with sugars and accumulate at a higher age and in diabetes, processes which can interfere with the integration of implants into the tissue. Glyoxal is a highly reactive glycating agent involved in the formation of AGEs and is known to induce apoptosis, as revealed by the upregulation of caspase-3 and fractin (caspase-3 being a key enzyme activated during the late stage of apoptosis and fractin being a caspase-cleaved actin fragment). In this study, we investigated the influence of collagen type I coating on the cytotoxic effect of glyoxal on rat calvarial osteoblastic cells and on human osteosarcoma cells (Saos-2) grown on titanium alloy, Ti6Al4V. Activation of caspase-3 and fractin was measured by counting immunohistochemically stained cells and by flow cytometry with propidium iodide (detection of the apoptosis indicating a sub-G1 peak). Our results showed an increased number of apoptotic osteoblasts after incubation with glyoxal on Ti6Al4V discs. However, the number of apoptotic cells on collagen-coated titanium was significantly smaller than on uncoated titanium after the same treatment. The present findings demonstrate that osteoblasts treated with glyoxal undergo apoptosis, whereas collagen type I coating of titanium alloys (used for implants) has an antiapoptotic function. / Dieser Beitrag ist mit Zustimmung des Rechteinhabers aufgrund einer (DFG-geförderten) Allianz- bzw. Nationallizenz frei zugänglich.

1a,25-Dihydroxyvitamin D3 (1a,25(OH)2D3) is a major functional metabolic form of vitamin D. 1a,25(OH)2D3 has drawn increasing attention due to its functions in addition to maintaining calcium phosphate homeostasis. It directly regulates mineralization by osteoblasts, matrix production and remodeling by chondrocytes, and contraction of cardiomyocytes. 1a,25(OH)2D3 and its analogues have shown beneficial effects in treating multiple sclerosis, diabetes and various types of cancer. In order to maximize the pharmaceutical potential of 1a,25(OH)2D3, a better understanding its cell signaling pathway is necessary. 1a,25(OH)2D3 regulates osteoblasts through both classical nuclear vitamin D receptor (nVDR) mediated genomic effects and plasma membrane receptor-mediated rapid responses. The identity of the plasma membrane receptor for 1a,25(OH)2D3 is controversial. Protein disulfide isomerase associated 3 (Pdia3) has been hypothesized as one of the putative plasma membrane receptors for 1a,25(OH)2D3. The overall goal of this thesis was to understand the general role and the molecular mechanism of Pdia3 in 1a,25(OH)2D3-initiated rapid responses, and to determine the role of Pdia3 and its dependent signaling in osteoblast biology. The results show that Pdia3 is required for membrane-mediated responses of 1a,25(OH)2D3. Moreover, both Pdia3 and nVDR are critical components of the plasma membrane receptor complex for 1a,25(OH)2D3. Finally, Pdia3 and signaling via Pdia3 regulate osteoblast differentiation and mineralization. Taken together, this study demonstrates the role of Pdia3 in rapid responses to 1a,25(OH)2D3 and osteoblast biology, reveals the unexpected complexity of the 1a,25(OH)2D3 plasma receptor complex and opens the new target, Pdia3, for pharmaceutical application and tissue engineering.

Musculoskeletal pathologies associated with decreased bone mass, including osteoporosis and disuse-induced bone loss, affect millions of Americans annually. Many pharmaceutical treatments have slowed osteoporosis, but there is still no countermeasure for bone loss observed in astronauts. Additionally, high magnitude and low frequency impact has been recognized to increase bone and muscle mass under normal but not microgravity conditions. However, a low magnitude and high frequency (LMHF) mechanical load experienced in activities such as postural control has also been shown to be anabolic to bone. While several clinical trials have demonstrated that the LMHF mechanical loading normalizes bone loss in vivo, the target tissues and cells of the mechanical load and underlying mechanisms mediating the responses are unknown. As such, the objectives of this project are to analyze cellular and molecular changes induced in osteoblasts by LMHF loading and to investigate the utility of a LMHF mechanical load in mitigating microgravity-induced bone loss. The central hypothesis of the project is that simulated microgravity or disuse conditions induce bone loss by inhibiting expression of genes critical in regulating bone formation, osteoblast differentiation, and subsequent mineralization while a LMHF mechanical load prevents these effects. To test this hypothesis, we developed an in vitro disuse system using the Random Positioning Machine (RPM). For the first time, we reported systemic gene expression studies in 2T3 preosteoblasts using the RPM disuse system showing that 140 genes were altered by RPM exposure with over two-fold statistically significant changes. Moreover, we also utilized an independent simulator called the Rotating Wall Vessel (RWV) to partially validate the in vitro disuse systems and to confine the list of genes to those most critical in regulating bone formation. After comparative studies, we constricted the list to 15 commonly changed genes, three of which were not only decreased with disuse but also increased with mechanical loading in vivo. Furthermore, we employed the RPM disuse system to evaluate the mechanism by which a LMHF load mitigates bone loss. Exposure of osteoblasts to the RPM decreased both ALP activity and mineralization even in the presence of bone morphogenic protein 4 (BMP4), and the LMHF mechanical loading prevented the RPM-induced decrease in both markers. Mineralization induced by LMHF mechanical loading was enhanced by treatment with BMP4 and blocked by the BMP antagonist noggin, suggesting a role for BMPs in this response. In addition, LMHF mechanical loading rescued the RPM-induced decrease in gene expression of ALP, runx2, osteomodulin, parathyroid hormone receptor 1, and osteoglycin. These findings show that osteoblasts directly respond to LMHF mechanical loading, potentially leading to normalization or prevention of bone loss caused by disuse or microgravity conditions. The mechanosensitive genes identified here provide potential targets for pharmaceutical treatments that may be used in combination with LMHF mechanical loading to better treat osteoporosis, disuse-induced bone loss, or microgravity-induced bone loss.

Biomaterial surface properties such as microtopography and energy can change cellular responses at the cell-implant interface. Phospholipase D (PLD) is required for differentiation of osteoblast-like MG63 cells on machined and grit-blasted titanium surfaces. Here, we determined if PLD is also required on microstructured/high-energy substrates and the mechanism involved. shRNAs for human PLD1 and PLD2 were used to silence MG63 cells. Wild-type and PLD1 or PLD1/2 silenced cells were cultured on smooth-pretreatment surfaces (PT); grit-blasted, acid-etched surfaces (SLA); and SLA surfaces modified to have higher surface energy (modSLA). PLD was inhibited with ethanol or activated with 24,25-dihydroxyvitamin-D₃ [24R,25(OH)₂D₃]. As surface roughness/energy increased, PLD mRNA and activity increased, cell number decreased, osteocalcin and osteoprotegerin increased, and protein kinase C (PKC) and alkaline phosphatase specific activities increased. Ethanol inhibited PLD and reduced surface effects on these parameters. There was no effect on these parameters after knockdown of PLD1, but PLD1/2 double knockdown had effects comparable to PLD inhibition. 24R,25(OH)₂D₃increased PLD activity and production of osteocalcin and osteoprotegerin, but decreased cell number on the rough/high-energy surfaces. These results confirm that surface roughness/energy-induced PLD activity is required for osteoblast differentiation and that PLD2 is the main isoform involved in this pathway. Here we showed that PLD is activated by 24R,25(OH)₂D₃ in a surface-dependent manner and inhibition of PLD reduced the effects of surface microstructure/energy on PKC, suggesting that PLD mediates the stimulatory effect of microstructured/high-energy surfaces via PKC-dependent signaling.

[Truncated abstract] Bone is a living tissue and is maintained by the coordinate action of osteoblasts and osteoclasts. The intercellular communication between these two cells is the quintessential mechanism in bone remodelling. Unfortunately, the importance of this interaction is often neglected and its significance is only realised when disruption of this “cross-talk” results in debilitating bone diseases. Additionally, the number of known proteins that are involved in this “cross-talk”, especially those that are osteoclast-derived, and act specifically on osteoblasts, is limited. This discrepancy leads to the question: Can osteoclasts directly control the growth and function of osteoblastic cells by expressing specific proteins that bind directly to osteoblasts? If so, is it possible to use these proteins to control and, possibly, treat bone disease? The objective of this thesis is to identify and characterise osteoclast-derived factors that can modulate bone homeostasis, as well as contribute to the intercellular communication between osteoblasts and osteoclasts ... Collectively, the data in this thesis culminates in one important conclusion: the identification of a novel paracrine secretory factor that has the potential to directly induce the formation of bone. These findings represent the first ever characterisation of a protein that allows the osteoclasts to directly control the growth and function of osteoblasts. Due to the potential function of ODOF to induce bone formation, this protein may be used therapeutically to treat bone disease.

Diss. (sammanfattning) Stockholm : Karol. inst., 2002. / Härtill 4 uppsatser.