Proteoglycan 4 (PRG4) is a mucinous proteoglycan with an apparent molecular weight of -345 kDa. It has been detected in a variety of tissues including cartilage, tendon, bone heart and liver, and is also known as cartilage superficial zone protein (SZP), lubricin, megakaryocyte stimulating factor (MSF) precursor protein and camptodactyly-arthropathy-coxa vara pericarditis (CACP) protein. PRG4 has been shown to reduce friction in joints, and immunological studies have located PRG4 on the surface of articular cartilage, in synovial fluid, on the surface of meniscus and on the surface of mature compressed tendon. PRG4 is involved in the congenital joint pathology known as CACP. Analyses of the PRG4 sequence reveal a propensity for a diverse number of functions including lubrication, matrix-binding, self-aggregation, cytoprotection and cell proliferation. This study set out to investigate the potential functions of the N- (exons 2-5) and C-terminal (exons 7-12) structural domains of human and bovine PRG4, facilitated by recombinant protein expression. Preliminary results from solid-phase binding assays showed an interaction between the N- terminal domain of PRG4 and plasminogen activator inhibitor-1 (PAI-1), type II collagen and a 70kDa fibronectin fragment. Immunoprecipitaion experiments demonstrated a potential interaction between the PRG4 C-terminal domain and fibronectin fragments. Both the N- and C- terminal contain functional heparin binding sites. The C-terminal domain was shown to interact with superficial zone chondrocytes, an interaction that was perturbed by heparin. The study also used purified full-length PRG4 from three sources: human recombinant, bovine articular cartilage and bovine tendon. These PRG4 species bound to heparin and displayed similar glycosylation profiles. This study also shows that PRG4 is susceptible to degradation by a number of matrix proteinases including elastase, plasmin, matrix metalloproteinase-7 and cathepsin B. Collectively, the data presented in this thesis provides novel information concerning the biochemistry, susceptibility to digestion and functional capabilities of PRG4.

Les lésions de cartilage ne cicatrisent pas spontanément et la réparation de ce tissu est un challenge. Les techniques chirurgicales restant insatisfaisantes, la thérapie cellulaire et l'ingénierie tissulaire sont maintenant envisagées. La transplantation de chondrocytes autologues (TCA) existe déjà mais cette procédure nécessite l'amplification des chondrocytes qui s'accompagne d'une perte du phénotype différencié (dont l'indicateur est le collagène de type II), au profit d'un phénotype fibroblastique (dont l'indicateur est le collagène de type I, retrouvé dans les tissus fibreux). La TCA conduit donc à une greffe de chondrocytes dédifférenciés produisant un fibrocartilage, dont les propriétés mécaniques sont différentes du cartilage hyalin natif. L'objectif de mes travaux était de développer un nouveau kit d'ingénierie tissulaire du cartilage par association de chondrocytes humains, de biomatériaux et d'une sélection de facteurs solubles. Nous avons utilisé le cocktail FGF-2/insuline (FI) pour l'amplification cellulaire et le cocktail BMP-2/insuline/T3 (BIT) pour redifférencier les chondrocytes dans des éponges de collagène. Nos résultats ont montré que cette combinaison permet la synthèse d'une matrice cartilagineuse dans les supports collagène. Cependant, cette synthèse s'est trouvée favorisée en périphérie des éponges cultivées en conditions statiques. Nous avons ensuite utilisé un bioréacteur pour perfuser les éponges et nos résultats ont révélé alors un dépôt plus homogène de cartilage dans ces supports. De manière très intéressante, nous avons aussi observé l'arrêt de l'expression du collagène de type I. Ainsi, notre approche multifactorielle combinant des chondrocytes humains, des biomatériaux collagène, une combinaison FI-BIT et une culture en perfusion permet la reconstruction d'un cartilage non fibrotique / Cartilage lesions are irreversible and cartilage repair is challenging. Actual surgical techniques remain unsatisfactory and therefore, cell therapy and tissue engineering approaches are now considered. The Autologous Chondrocytes Transplantation (ACT) already exists but this procedure requires chondrocytes amplification. During this amplification, a dedifferentiation process occurs: chondrocytes lose their differentiated phenotype (characterized by type II collagen) towards a fibroblastic phenotype (characterized by type I collagen, a component of fibrous tissues). ACT leads to the graft of dedifferentiated chondrocytes, hence provoking the production of a fibrocartilage that presents different mechanical properties than native hyaline cartilage. The aim of my work was to develop a new kit of tissue engineering for cartilage repair using human chondrocytes, biomaterials and a selection of soluble factors. We used a cocktail of FGF-2 and insulin (FI) for cell amplification and a cocktail of BMP-2, insulin and T3 (BIT) for chondrocyte redifferentiation in collagen sponges. Our results showed that the combination allows the synthesis of a cartilaginous matrix in collagen scaffolds. However, matrix production is favored in periphery of the sponges cultivated in static conditions. A perfusion bioreactor was then used to perfuse the sponges and our results revealed a more homogeneous deposition of cartilage in the scaffolds. Very interestingly, we also noticed a stop of type I collagen expression. Thus, our multifactorial approach combining human chondrocytes, collagen scaffold, the combination FI-BIT and culture under perfusion allows the reconstruction of a non-fibrotic cartilage