Les affections ostéo-articulaires concernent des millions de personnes. L’ingénierietissulaire osseuse, associant cellules souches mésenchymateuses humaines (CSM) etmatériaux synthétiques, pourrait répondre aux besoins cliniques. Pour cela, les différentescomposantes de cette approche et leur association doivent être mieux étudiées pour la rendreutile cliniquement. Durant cette thèse, une première étude animale proche du cas cliniquenous a permis de définir les points à améliorer pour le traitement des pertes osseuses. Nousavons ainsi pu développer un milieu spécifique induisant une différenciation rapide etterminale des CSM en ostéoblastes. Par la suite, l’utilisation de particules de céramiquescomme support cellulaire nous a permis d’obtenir des hybrides riches en matriceextracellulaire. Cet environnement 3D biomimétique permet l’engagement spontané des CSMvers un phénotype ostéoblastique et l’obtention d’une quantité osseuse importante in vivo.L’ensemble de ces résultats met en évidence l’importance de l’environnement et du stade dedifférenciation cellulaire pour la formation osseuse par ingénierie tissulaire osseuse. / Osteo-articular disorders affect millions of people over the world. Bone tissueengineering, an approach combining human mesenchymal stem cells (MSC) and syntheticmaterials, could potentially fulfill clinical needs. However, the different components of thisapproach and their association should be investigated further to make it clinically useful. Inthis thesis, an initial animal study close to clinical situation allowed us to identify areas thatneed improvement for regenerating bone defect. We were then able to develop a specificmedium which induces a rapid and terminal osteoblastic differentiation of MSC.Subsequently, the use of ceramic particles as cell support has allowed us to obtain hybridmainly composed of extracellular matrix. This biomimetic 3D environment allowsspontaneous osteoblastic commitment of MSC and induces a large bone quantity in vivo.Overall, these results highlight the importance of the environment and the cell differentiationstate for bone formation using bone tissue engineering.

The field of tissue engineering aims to develop viable substitutes with the ability to repair and regenerate the functions of damaged tissue. Common practices to supplement bone regeneration in larger defects include bone graft biomaterials such as autografts, allografts, xenografts, and synthetic biomaterials. Autologous bone grafting is the current gold-standard procedure used to replace missing or damaged bone. However, these grafts have disadvantages such as donor site morbidity, limited availability, and the need for a secondary surgery. The focus of this study is to tissue engineer a lyophilized gelatin composite sponge composed of hydroxyapatite (HA), chitin whiskers (CW), and preparations rich in growth factors (PRGF) to provide sufficient structural support to the defect site while enhancing the body’s own reparative capacity, ultimately eliminating the need for autologous tissue harvesting or repeat operations. The present study investigates several in vitro evaluations on multiple compositions of modified gelatin sponge scaffolds for use in bone graft applications. Gelatin sponges were fabricated via freeze-drying, enhanced with PRGF, HA, and/or CW, and cross-linked with 50 mM 1-Ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride (EDC) either during or post gelation. Initial evaluation of all scaffold combinations indicated that scaffolds released contents up to 90 days, EDC cross-linking during gelation allowed for more protein release, and had the ability to swell. Since the incorporation of PRGF, HA, and CW increased cell infiltration, and production of cell-created mineral matrix over 90 days in culture, these scaffolds were further characterized. Freeze-dried gelatin sponges enhanced with PRGF, HA, and CW and cross-linked during gelation with EDC (PHCE) were mineralized (M) in 5x revised simulated body fluid (r-SBF) for 1 hour to create a bone-like mineral surface. Gelatin EDC scaffold controls (GE), GE-M, PHCE, and PHCE-M scaffolds were characterized for their ability to swell, mineralizing potential, surface morphology, growth factor incorporation and release, uniaxial compression properties, and cell attachment, proliferation, infiltration, and protein/cytokine secretion.. After mineralization, scanning electron microscopy showed sparse clusters of mineral deposition for GE-M scaffolds while PHCE-M scaffolds exhibited a more uniform mineral deposition. Both GE and PHCE scaffolds were porous structures that swelled up to 50% of their original volume upon hydration. Over 21 days incubation, PHCE-M scaffolds cumulatively released about 30% of their original protein content, significantly more than all other scaffolds. Multiplex Luminex assays confirmed the successful incorporation of PRGF growth factors within PRGF sponges. For acellular uniaxial compression testing, PHCE-M scaffolds reported lower Young’s modulus values (1.3 - 1.6 MPa) when compared to GE and GE-M scaffolds (1.6 – 3.2 MPa). These low modulus values were comparable to values of tissue found in early stages of bone healing. DAPI (4',6-diamidino-2-phenylindole) staining and imaging showed an increase in initial cell attachment and infiltration of PHCE and PHCE-M scaffolds on day 1. GE-M scaffolds also appeared to attach more cells than the GE control. MTS cell proliferation assay results indicated that on days 4 and 7, PHCE scaffolds increased cell proliferation (compared to GE controls). MTS also illustrated that the addition of a mineral coating increases and decreases cell proliferation on GE-M and PHCE-M scaffolds, respectively. Multiplexer analysis of MG-63 protein/cytokine secretion suggests that cells are responding in a bone regenerative fashion on all scaffolds, as evidence of osteocalcin secretion. Little to no secretion of osteopontin, IL-1β, and TNF-α demonstrates that scaffolds are not influencing cells to secrete factors associated with bone resorption. The compressive mechanical properties of cellularized scaffolds did not differ much from acellular scaffolds. The collective results indicated increased cellular attachment, infiltration, and bone regenerative protein/cytokine secretion by cells on GE-M scaffolds, which support the addition of a bone-like mineral surface on GE scaffolds. Cellularized PHCE and PHCE-M scaffolds report similar advantages as well as Young’s modulus values in the range of native tissues present in the early stages of bone healing. The results of this study propose that the developed PHCE and PHCE-M scaffolds exhibit good cellular responses and mechanical properties for use in early bone healing applications.

A l’heure actuelle, les alliages à bases de titane sont les matériaux les plus utilisés en implantologie osseuse. Les procédés émergents de fabrication additive, tel que la fusion par faisceau d’électrons (EBM), permettent de fabriquer des structures architecturées sur-mesure en titane. Dans les cas cliniques difficiles, il est nécessaire de stimuler activement les cellules souches osseuses pour qu’elles produisent de l’os. Les protéines osseuses morphogénétiques (BMP-2, BMP-7) ont cette capacité d’ostéo-induction et sont utilisées en clinique. Cependant, leur délivrance par matrice de collagène est très mal contrôlée. Des revêtements de surface à base de polymères naturels, tels que la poly(L-lysine) et l’acide hyaluronique (PLL/HA), peuvent former des films biomimétiques servant de nanoréservoir pour ces protéines. L’objectif de cette thèse était de développer un implant innovant constitué de structures 3D en titane à la fois architecturées et ostéo-inductrices. Pour cela, des structures 3D poreuses en alliage de titane (Ti-6Al-4V) constituées de cellules cubiques ont été construites par EBM. La porosité a été bien contrôlée avec une différence par rapport aux modèles CAO de moins de 1%. La BMP-7 a été chargée et quantifiée dans les films biomimétiques. La capacité d’ostéo-induction des films a été évaluée avec des cellules souches mésenchymateuses de souris par leur expression de la phosphatase alcaline. L’expression de cette enzyme a augmenté de façon dose-dépendante avec la dose de BMP-7 initialement chargée. Le dépôt du film ostéo-inducteur sur les structures 3D architecturées a été caractérisé par microscopies optique et électronique. Les cellules souches cultivées au sein des structures 3D bioactives se différencient en cellules osseuses démontrant ainsi leur capacité ostéo-inductrice sur le court terme in vitro. Des tests préliminaires in vivo sont actuellement réalisés pour tester ces structures 3D bioactives dans un modèle fémoral de défaut osseux chez le rat. / To date, titanium-based alloys (Ti) remain the most used implantable materials for load-bearing applications. Emerging additive manufacturing techniques such as electron beam melting (EBM) enable to custom-build architectured scaffolds of controlled macroporosity. In very difficult clinical situations, potent bioactive signals are needed to boost stem cells: osteoinductive molecules such as bone morphogenetic proteins (BMP-2) are currently used for this purpose. However, one of their limitations is their inappropriate delivery with collagen sponges. Biomimetic surface coatings made of the biopolymers poly(L-lysine) and hyaluronic acid, (PLL/HA) polyelectrolyte films, have recently been engineered as nanoreservoirs for BMP proteins. The aim of this PhD thesis was to develop architectured and osteoinductive 3D titanium-based scaffolds as innovative synthetic bone grafts. To this end, we used the EBM additive manufacturing technique to engineer porous scaffolds with cubit unit-cells. Their surface was coated with biomimetic films containing the bone morphogenetic protein 7 (BMP-7). The porosity was well controlled with a difference from CAD models of less than 1%. The osteoinductive capacity of BMP-7 loaded films was assessed using murine mesenchymal stem cells (MSCs) by quantifying their alkaline phosphatase (ALP) expression, which increased in a dose-dependent manner. The coating of the 3D architectured scaffolds by the bioactive film was characterized using optical and electron microscopy techniques. Finally, the 3D architectured scaffolds coated with BMP-7-loaded films were proved to be osteoinductive at the early stage in vitro. Preliminary experiments are currently done to assess their performance in an in vivo model of a critical size femoral bone defect in rat.

Bone undergoes constant turnover throughout life and has the capacity to regenerate itself. However, the repair of critical size defects, caused by bone diseases such as osteoporosis, can be more problematic. Therefore, there is a clinical need for a bone graft substitute that can be used at sites of surgical intervention to enhance bone regeneration. Poly(vinylphosphonic acid-co-acrylic acid) (PVPA-co-AA) has recently been identified as a potential candidate for use in bone tissue scaffolds. It is hypothesised that PVPA-co-AA can mimic the action of bisphosphonates – a class of drugs used in the treatment of osteoporosis – by binding to calcium ions from bone mineral surfaces. In this way, bisphosphonates can affect bone turnover by increasing the activity of osteoblasts and reducing osteoclast activity. Although PVPA-co-AA has been shown to improve bone formation, the mechanism of action has so far not been fully elucidated. Therefore, this work aims to understand the effect of copolymer composition on the properties of PVPA-co-AA, and thus to determine its effect on osteoblast adhesion and proliferation. PVPA-co-AA copolymers have been synthesised with a range of monomer feed ratios. It was found that a VPA content of 30 mol % led to the greatest calcium binding affinity of the copolymer and is thus expected to lead to enhanced bone formation and mineralisation of the matrix produced by osteoblast cells. The release profile of PVPA-co-AA from electrospun PCL scaffolds was investigated. It was shown that all of the PVPA-co-AA was released into aqueous media within 8 h of immersion. It was also found that the calcium chelation from osteogenic differentiation media significantly increased within the first 8 h. Therefore, it was concluded that PVPA-co-AA is released from the scaffolds, where it can then bind to calcium ions from the bone mineral surface to promote mineralisation, thus acting as a mimic of non-collagenous proteins, which are present in the extracellular matrix (ECM) of bone. Hydrogels of PVPA-co-AA have been produced and the effect of monomer feed ratio (0-50 mol % VPA) on the properties of the gels was explored. It was found that an increase in VPA content led to greater hydrogel swelling and increased porosities. Hydrogels that contained 30 and 50 mol % VPA were shown to have similar morphologies to the native ECM of bone. Rheological testing showed that hydrogels with higher VPA contents were more flexible and could be deformed to a large extent without permanent deformation of their structure. An increase in osteoblast adhesion and proliferation was observed for hydrogels with 30 and 50 mol % VPA content as well as superior cell spreading. Osteoblast cell metabolic activity also increased as a function of VPA content in the hydrogels. This work indicates that hydrogels of PVPA-co-AA, with VPA contents of 30 or 50 mol %, are ideal for use as bone tissue scaffolds. Furthermore, the mechanical and cell adhesion properties of the gels can be tuned by altering the copolymer composition. Finally, composite hydrogels of PVPA-co-AA and hydroxyapatite (HA) have been produced and investigated for their ability to remove fluoride ions from groundwater. It was found that the fluoride uptake ability of PVPA-HA hydrogels was significantly enhanced when compared with HA powder alone. Furthermore, the fluoride uptake was dependent on many factors, including pH, contact time and the presence of competing ions. It was possible to regenerate the hydrogel to remove the fluoride ions, and thus it was shown that the material can be used a number of times with only a slight reduction in its fluoride uptake capacity.

Tissue engineering is a multidisciplinary field that is rapidly emerging as a promising new approach in the restoration and reconstruction of tissues. In this approach, three dimensional (3D) scaffolds are of great importance. Scaffolds function both as supports for cell growth and depot for sustained release of required active agents (e.g. enzymes, genes, antibiotics, growth factors). Scaffolds should possess certain properties in accordance with usage conditions. Wet-spinning is a simple technique that has been widely used for the fabrication of porous scaffolds for tissue engineering applications. Natural polymers can effectively be used in scaffold fabrication due to their biocharacteristics. Among natural polymers, chitosan and alginate are two of the most studied ones in tissue engineering and drug delivery fields because of being biologically renewable, biodegradable, biocompatible, non-antigenic, non-toxic and biofunctional. In this study, two kinds of porous scaffolds were produced as chitosan and alginate coated chitosan fibrous scaffolds by wet-spinning technique In order to investigate the delivery characteristics of the scaffolds, loading of gentamicin as a model antibiotic and bovine serum albumin (BSA) as a model protein was carried out in different loading models. Resultant scaffolds were characterized in terms of their structural formation, biodegradation, biomineralization, water uptake and retention ability and mechanical properties. Additionally, release kinetics of gentamicin and BSA were examined. Efficiency of gentamicin on Escherichia coli (E.coli) was examined. Characterization of scaffolds revealed their adequacy to be used in bone tissue engineering applications and capability to be employed as bioactive agent delivery systems.

This study was aimed at designing a cell carrier from an intelligent polymer to achieve loading of mechanical stress for the purpose of improving the tissue engineering capability in vitro. Ethyleneglycoldimethacrylate (EGDMA) crosslinked poly(Nisopropylacrylamide) (pNIPAM) films were prepared by radical polymerization with ultraviolet light (UV) in the presence of photoinitiator 2,2&#039 / -azoisobutyronitrile (AIBN) in isopropanol/water (1:1). Patterns were formed on the surface of the polymers by using silicon wafers with microridges (2 &amp / #956 / m) and grooves (10 &amp / #956 / m) that were prepared by photolithography technique as the template. The surfaces of the films were also modified by adsorption of ELP-RGD6 polypeptide. Bone marrow stem cells (BMSCs) isolated from 6 week old Sprague-Dawley rats were seeded onto the pNIPAM films with different surface topography and chemistry and cultured under static and dynamic conditions. Dynamic conditions were generated by cyclic temperature changes (15 min at 29&deg / C, 30 min at 37&deg / C) for 10 times a day during 5 days starting on the second day post-cell seeding. ELP-RGD6 on the films enhanced initial cell attachment but had no effect on proliferation in long term culturing. However, for the dynamic culturing, ELP was crucial for both retaining cells attached on the surface when the surface became hydrophilic and resulted in weakened cell attachment, and for better communication between cell and material which enhanced the ability of pNIPAM films to transfer mechanical stress on the cells. Dynamic conditions improved cell proliferation but decreased differentiation. Presence of the patterns also influenced the differentiation but did not affected proliferation.

During last decade, polyurethanes (PUs) which are able to degrade into harmless molecules upon implantation have received a significant level of attention as a biomaterial in tissue engineering applications. Many studies are focused especially on development of PUs based on amino acid derivatives / however, there are only few applications of amino acid based PUs in tissue engineering. In this study, a biocompatible and biodegradable thermoplastic poly(ester-urethane) (PEU) based on L-lysine diisocyanate (LDI) and polycaprolactone diol (PCL) was synthesized and used for the preparation of two dimensional (2D) films and three dimensional (3D) scaffolds. The resulting polymer was casted as 2D films for full characterization purpose and it was found that it is highly elastic with modulus of elasticity ~12 MPa. Surfaces of 2Ds were modified via micropatterning and fibrinogen coating to check the material-cell interaction. The 3D scaffolds were obtained by salt leaching and rapid prototyping (bioplotting) techniques. The 3D scaffolds had various pore size and porosity with different mechanical strength. The bioplotted scaffolds had uniform pore size of ~450 &micro / m and exhibited higher compressive modulus (~4.7 MPa) compared to those obtained by salt leaching (~147 kPa). Salt leached 3D scaffolds had inhomogenous pore size distribution in the range of 5 &micro / m - 350 &micro / m and demonstrated greatest degradation profile compared to 2D films and 3D bioplotted samples under enzymatic condition. Rat bone marrow stem cells (BMSCs) were used to investigate the biocompatibility of the polymer and suitability of fabricated scaffolds for osteogenesis. Presence of micropatterns on 2D matrices did not show any influence on osteoblastic function, but presence of fibrinogen enhanced cell attachment and proliferation. All of the fabricated 3D PEU matrices supported proliferation, osteoblastic differentiation and extracellular matrix (ECM) deposition with highest osteoblastic activity on bioplotted scaffolds which confirmed by von Kossa staining and EDX analysis. The results indicated that the synthesized PEU based scaffolds were able to induce osteoblastic differentiation and mineralization of BMSC and therefore these scaffolds can be good candidates to be used in bone tissue engineering

Large bone defects pose a significant clinical challenge currently lacking an adequate therapeutic solution. Bone tissue engineering (BTE) therapies aim to provide that solution by combining structural scaffolds, bioactive factors, and/or osteogenic cells. Cellular therapies are likely vital to repair severe defects in patients lacking sufficient endogenous cells. Stem cells are attractive cell choices due to their osteogenic differentiation and extensive proliferation abilities, but their therapeutic potential is still uncertain, as studies comparing stem cell sources and delivery methods have produced inconsistent results. In this thesis, we developed a challenging in vivo large bone defect model for quantitative comparison of human stem cell-based therapies and then evaluated the abilities of adult or fetal stem cell-seeded constructs to enhance defect repair, with or without added osteogenic cues. First, we showed that cellular construct treatment enhanced defect healing over acellular construct treatment, although there were no differences between adult or fetal cell sources. We next labeled stem cells with a fluorescent tracking agent, the quantum dot, to determine biodistribution of implanted cells during the repair process. While quantum dots effectively labeled cells in vitro, they were ineffective in vivo tracking agents due to false positive signals and detrimental effects on stem cell-mediated repair. Finally, we developed a novel gene therapy technique using virus-coated scaffolds to deliver the osteogenic factor bone morphogenetic protein 2 (BMP2) to defect sites, either by in vitro (BMP2 transduction of seeded stem cells pre-implantation) or in vivo (BMP2 transduction of defect-site host cells) means. While defect-site BMP2 delivery through gene therapy methods improved repair, in vivo therapy enhanced healing more than stem cell-based in vitro therapy. This finding does not rule out the potential of stem cell-based in vitro gene therapy treatment for functional bone repair, as increases in viral dose may improve stem cell-mediated healing, but it does present evidence of a novel acellular BTE therapy with potential off-the-shelf clinical application in large bone defect repair, as scaffolds could be virally coated with the gene for BMP2 expression and frozen until implantation.

This work examines the osteogenic potential and immune-privileged properties of human umbilical cord perivascular cells (HUCPVCs) in normal Wistar rats and athymic rnu/rnu rats for up to 60 days. HUCPVCs demonstrated a mesenchymal stromal cell phenotype, assayed through flow cytometry, and RT-PCR analysis detected their expression of osteogenic genes. A bone tissue engineering construct was developed through centrifugal seeding of HUCPVCs onto calcium phosphate-coated PLGA scaffolds. These cell-scaffold constructs were transplanted into bilateral femoral defects. HUCPVCs did not induce any systemic biological response in normal rats; however, they did not engraft and impaired bone healing up to 60 days. When transplanted into athymic rats, HUCPVCs were detected up to 30 days in the femoral defects, improved bone regeneration at 15 and 30 days, as measured by micro computed tomography, and expressed osteogenic proteins. These findings demonstrate that HUCPVCs are suitable for bone tissue engineering studies in larger animals.

This work examines the osteogenic potential and immune-privileged properties of human umbilical cord perivascular cells (HUCPVCs) in normal Wistar rats and athymic rnu/rnu rats for up to 60 days. HUCPVCs demonstrated a mesenchymal stromal cell phenotype, assayed through flow cytometry, and RT-PCR analysis detected their expression of osteogenic genes. A bone tissue engineering construct was developed through centrifugal seeding of HUCPVCs onto calcium phosphate-coated PLGA scaffolds. These cell-scaffold constructs were transplanted into bilateral femoral defects. HUCPVCs did not induce any systemic biological response in normal rats; however, they did not engraft and impaired bone healing up to 60 days. When transplanted into athymic rats, HUCPVCs were detected up to 30 days in the femoral defects, improved bone regeneration at 15 and 30 days, as measured by micro computed tomography, and expressed osteogenic proteins. These findings demonstrate that HUCPVCs are suitable for bone tissue engineering studies in larger animals.