La dysplasie géléophysique (DG) est une maladie rare qui appartient à la famille des dysplasies acroméliques. Cette pathologie est caractérisée par un retard statural, une brachydactylie, une raideur articulaire, une dysmorphie faciale, une peau épaisse, une atteinte bronchopulmonaire et une surcharge valvulaire cardiaque conduisant le plus souvent à une mort précoce dans les premières années de la vie. Deux modes de transmissions ont été identifiés. Le premier autosomique récessif est dû à des mutations dans le gène ADAMTSL2. Le second, autosomique dominant est dû à un hot-spot de mutations dans les exons 41 et 42 qui codent pour le domaine Transforming Growth Factor (TGF) β-binding protein-like domain 5 (TB5) du gène FBN1. FBN1 et ADAMTSL2 codent pour des protéines sécrétées de la matrice extracellulaire (MEC). FBN1 code pour la fibrilline-1, une composante des microfibrilles qui jouent un rôle dans la biodisponibilité du TGFβ. La protéine ADAMTSL2 fait partie de la famille des ADAMTS mais n’a pas d’activité enzymatique dû à l’absence de domaine catalytique. Sa fonction est encore inconnue. Cependant des partenaires d’ADAMTSL2 ont été identifiés par notre équipe : latent-transforming growth factor beta-binding protein 1 (LTBP1) et FBN1 qui sont directement impliqués dans le stockage de TGFβ. Récemment une autre protéine, FBN2, a aussi été découverte comme partenaire d’ADAMTSL2 (Hubmacher D et. al.). L’objectif de ma thèse était de comprendre le mécanisme physiopathologique de la DG, grâce à l’analyse de modèles murins. Un premier modèle murin pour la forme récessive de la DG appelé CreCMV; Adamtsl2f/f (ou KO) a été généré. L’analyse phénotypique de ces souris a montré un retard statural, des os longs courts, des extrémités courtes. Dans les plaques de croissance des os longs des souris mutantes, nous avons observé une désorganisation des colonnes chondrocytaires associée à une diminution de l’expression du collagène de type 10, marqueur de la différentiation des chondrocytes. L’analyse de la matrice extracellulaire des plaques de croissance a révélé une désorganisation structurale importante. Une diminution de la fibrilline-1 et de LTBP-1 a été observée ainsi qu’une augmentation de l’activation de la voie de signalisation TGFβ au niveau de la plaque de croissance des souris mutantes. Nous avons observé une désorganisation du réseau microfibrillaire sur des cultures de chondrocytes de souris mutantes. Ces résultats nous ont permis de suggérer que la protéine ADAMTSL2 est impliquée dans la structure du réseau microfibrillaire, lieu de stockage du TGFβ et de démontrer un rôle majeur d’ADAMTSL2 dans la régulation de la chondrogenèse. Afin d’étudier la forme dominante de la DG, le modèle FBN1TB5+/- a été généré. Il est issu d’un système knock-in avec une mutation dans l’exon 42 du gène fbn1 qui correspond au domaine TB5 de la fibrilline-1. Nos résultats ont montré une réduction de la taille des souris hétérozygotes et homozygotes en comparaison aux souris sauvages au stade P1 et P30. Au stade P1, nous avons observé des chondrocytes plus larges et une dérégulation des marqueurs de la chondrogenèse au niveau de la plaque de croissance des fémurs des souris hétérozygotes, ainsi que chez les souris homozygotes. De plus, nous avons observé une très forte mortalité des souris homozygotes vers l’âge de 2 ou 3 mois. Nous en avons conclu que des mutations domaine TB5 de la fibrilline étaient liées à un retard statural et donc que FBN1 avait un rôle majeur dans la chondrogenèse. / Geleophysic dysplasia (GD) is a rare disease, which belong to acromelic group. This pathology is characterized by short stature, brachydactyly, joint stiffness, thick skin, facial dimorphism, broncho-pulmonary insufficiency and cardiac disease which lead to an early death in the first years of life. Two mode of inheritance have been identified. The first one, autosomal recessive, is due to mutations in ADAMTSL2 gene. The second, autosomal dominant, is due to hot-spot mutations in exon 41-42 of FBN1 gene, which encode the Transforming Growth Factor (TGF) β-binding protein-like domain 5 (TB5) of the protein. FBN1 and ADAMTSL2 encode secreted proteins of the extracellular matrix (ECM). FBN1 encodes fibrilline-1, component of microfibrillar network, playing a role in the bioavailability of TGF- β. ADAMTSL2 protein belongs to ADAMTS family, but does not have enzymatic activity due to lack of catalytic domain. Its function remains unknown. However, ADAMTSL2 partners have been identified by our team: latent-transforming growth factor beta-binding protein 1 (LTBP1) and FBN1, which are directly implied in storage of TGF-β. Recently, another protein, FBN2, have been identified as an ADAMTSL2 partner (Hubmacher D et. al.). The aim of my study was to understand the physiopathological mechanism of Geleophysic dysplasia by analysing murine models. A first murine model for the GD recessive form, CreCMV; Adamtsl2f/f (KO), have been generated. Phenotypic analysis of these mice showed short stature and shorter long bones and extremities. In long bone growth plate of mutant mice, we observed disorganization of chondrocyte columns, associated with decrease of collagen 10 expression, marker of chondrocyte differentiation. Analysis of ECM in growth plate revealed strong structural disorganization. Decrease of FBN1 and LTBP1 and were observed with an overactivation of TGF-β pathway in growth plate of mutant mice. We observed disorganization of microfibrillar network in chondrocyte cultures of mutant mice. These results suggest that ADAMTSL2 protein is implied in structure of microfibrillar network, where is stored TGF-β, and demonstrate major role of ADAMTSL2 in chondrogenesis. In order to study dominant form of GD, mouse model FBN1TB5+/-, have been generated. The mice were obtained by knock-in system, with mutation in exon 42 of FBN1 gene. Our results showed short stature of heterozygous (HT) and homozygous (Ho) mice compared to wild)type mice, at stage P1 and P30. At stage P1, we observed larger chondrocytes and deregulation of chondrogenesis markers in growth plate of HT and Ho mice. Furthermore, we observed high mortality of Ho mice at 2-3 months. We concluded that mutations in TB5 domain of FBN1 were linked to short stature and thus FBN1 have major role in chondrogenesis.

Abstract not available

<p>Migration of T-lymphocytes on a surface coated with extracellular matrix (ECM) components (two-dimensional (2-D) migration) and migration (infiltration) into a matrix (Three-dimesional (3-D) migration) are complex events and the underlying mechanisms are not yet fully understood. Here 2-D and 3-D migration were studied by use of seven leukemic T-cell lines representing discrete differentiation stages, a non-leukemic T-cell clone, and normal peripheral blood T cells. peripheral blood lymphocytes and the T-cell clone produced nanogram quantities of various chemokines, as compared to a production of ≤ 0.05 ng/ml by the T leukemia cell lines. In a Boyden chamber system, the leukemic T-cell lines showed haptotactic migration on fibronectin. The migration was augmented bu exposure to chemokines, including RANTES, MIP-1α, MIP-1β, and IL-8. The T-cell lines showed a peak response at a chemokine concentration of 10-50 ng/ml, whereas the T-cell clone responded optimally at 100 ng/ml. In contrast to a general capability of T-cells to migrate on 2-D ECM, only some of the T-cell lines were capable of 3-D migration into Matrigel or a collagen matrix. The infiltrative capacity was unrelated to the capacity to migrate on or adhere to the substrata. T-cell lines with a capacity to infiltrate produced matrix metalloproteinase-9 (MMP-9) and tissue inhibitor of matrix metalloproteinases-1 (TIMP-1), whereas non-infiltrating cell lines did not produce MMP-9. T-cell lines capable of infiltrating Matrigel or collagen responded to chemokines exposure with increased infiltration, but the chemokines did not render non-infiltrative cell lines infiltrative. Stimulation of infiltration of T-cell lines into collagen by the chemokine SDF-1α was inhibited by somatostatin, a neuropeptide with immunosuppressive properties. In conclusion, the ability to migrate on 2-D substrata and to infiltrate into 3.D substrata was found to be distinguishable properties of T cells. failure of some T-cell lines to infiltrate correlated with the lack of expression of MMP-9. Chemokines stimulated infiltration of infiltrative T-cell lines into collagen and Matrigel but did not render non-infiltrative T-cell lines infiltrative. Finally, a possible physiological mechanism for modulation of the chemokine-stimulated 3-D migration was demonstrated.</p>

<p>Bikunin, a chondroitin sulfate-containing protein of 25 kDa, has protease inhibitory activity and occurs in the plasma in free and complexed form. In inter-α-inhibitor (IαI) and pre-a-inhibitor (PαI) it is covalently linked through its chondroitin sulfate (CS) chain to two or one other polypeptide of about 80 kDa – heavy chains 1 and 2 (H1, H2) and heavy chain 3 (H3) – respectively. Bikunin and the heavy chains are synthesized as precursors, which are proteolytically cleaved and assembled into IαI and PαI in the secretory pathway. The C-terminal extension (CTX) of the heavy chains seems to mediate its own cleavage and theassembly of the complexes. The heavy chains of the IαI family become transferred to hyaluronan during ovulation and inflammation.</p><p>In this thesis, the biosynthesis of PαI, the plasma clearance of bikunin and the binding of IαI to collagen were studied. We found that in H3, a short segment on the N-terminal side of the CTX cleavage site is required for cleavage. Furthermore, the H3 could become linked to free CS chains primed by a xyloside, showing that the bikunin protein core is not needed for coupling. We also identified His649 as a residue essential for coupling, but not for cleavage. </p><p>Bikunin labelled with a residualizing agent, 125I-tyramine cellobiose, was injected into mice to identify tissues involved in its uptake. Half of the radioactivity was recovered in the kidneys, 10% in the liver, and the rest distributed in other tissues. We determined the half-life of bikunin in rat plasma using two independent methods: injection of 125I-bikunin, or hepatectomy followed by assessing the rate of disappearance of endogenous bikunin. Both methods yielded half-time values of 5-7 minutes. Removal of the CS chain did not affect the clearance rate of bikunin.</p><p>IαI and its heavy chains were found to bind to collagen with dissociation constants greater than 2 μM and 0.4-0.6 μM, respectively and this binding was independent of divalent metal ions. We suggest that the interaction of IαI with collagen may play a modulatory role in cell migration or in remodelling of the extracellular matrix.</p>

Bikunin, a chondroitin sulfate-containing protein of 25 kDa, has protease inhibitory activity and occurs in the plasma in free and complexed form. In inter-α-inhibitor (IαI) and pre-a-inhibitor (PαI) it is covalently linked through its chondroitin sulfate (CS) chain to two or one other polypeptide of about 80 kDa – heavy chains 1 and 2 (H1, H2) and heavy chain 3 (H3) – respectively. Bikunin and the heavy chains are synthesized as precursors, which are proteolytically cleaved and assembled into IαI and PαI in the secretory pathway. The C-terminal extension (CTX) of the heavy chains seems to mediate its own cleavage and theassembly of the complexes. The heavy chains of the IαI family become transferred to hyaluronan during ovulation and inflammation. In this thesis, the biosynthesis of PαI, the plasma clearance of bikunin and the binding of IαI to collagen were studied. We found that in H3, a short segment on the N-terminal side of the CTX cleavage site is required for cleavage. Furthermore, the H3 could become linked to free CS chains primed by a xyloside, showing that the bikunin protein core is not needed for coupling. We also identified His649 as a residue essential for coupling, but not for cleavage. Bikunin labelled with a residualizing agent, 125I-tyramine cellobiose, was injected into mice to identify tissues involved in its uptake. Half of the radioactivity was recovered in the kidneys, 10% in the liver, and the rest distributed in other tissues. We determined the half-life of bikunin in rat plasma using two independent methods: injection of 125I-bikunin, or hepatectomy followed by assessing the rate of disappearance of endogenous bikunin. Both methods yielded half-time values of 5-7 minutes. Removal of the CS chain did not affect the clearance rate of bikunin. IαI and its heavy chains were found to bind to collagen with dissociation constants greater than 2 μM and 0.4-0.6 μM, respectively and this binding was independent of divalent metal ions. We suggest that the interaction of IαI with collagen may play a modulatory role in cell migration or in remodelling of the extracellular matrix.

Migration of T-lymphocytes on a surface coated with extracellular matrix (ECM) components (two-dimensional (2-D) migration) and migration (infiltration) into a matrix (Three-dimesional (3-D) migration) are complex events and the underlying mechanisms are not yet fully understood. Here 2-D and 3-D migration were studied by use of seven leukemic T-cell lines representing discrete differentiation stages, a non-leukemic T-cell clone, and normal peripheral blood T cells. peripheral blood lymphocytes and the T-cell clone produced nanogram quantities of various chemokines, as compared to a production of ≤ 0.05 ng/ml by the T leukemia cell lines. In a Boyden chamber system, the leukemic T-cell lines showed haptotactic migration on fibronectin. The migration was augmented bu exposure to chemokines, including RANTES, MIP-1α, MIP-1β, and IL-8. The T-cell lines showed a peak response at a chemokine concentration of 10-50 ng/ml, whereas the T-cell clone responded optimally at 100 ng/ml. In contrast to a general capability of T-cells to migrate on 2-D ECM, only some of the T-cell lines were capable of 3-D migration into Matrigel or a collagen matrix. The infiltrative capacity was unrelated to the capacity to migrate on or adhere to the substrata. T-cell lines with a capacity to infiltrate produced matrix metalloproteinase-9 (MMP-9) and tissue inhibitor of matrix metalloproteinases-1 (TIMP-1), whereas non-infiltrating cell lines did not produce MMP-9. T-cell lines capable of infiltrating Matrigel or collagen responded to chemokines exposure with increased infiltration, but the chemokines did not render non-infiltrative cell lines infiltrative. Stimulation of infiltration of T-cell lines into collagen by the chemokine SDF-1α was inhibited by somatostatin, a neuropeptide with immunosuppressive properties. In conclusion, the ability to migrate on 2-D substrata and to infiltrate into 3.D substrata was found to be distinguishable properties of T cells. failure of some T-cell lines to infiltrate correlated with the lack of expression of MMP-9. Chemokines stimulated infiltration of infiltrative T-cell lines into collagen and Matrigel but did not render non-infiltrative T-cell lines infiltrative. Finally, a possible physiological mechanism for modulation of the chemokine-stimulated 3-D migration was demonstrated.

Background. Myocardial hypertrophy is a risk factor for cardiovascular morbidity and mortality. Independent of underlying disease, the cardiac muscle strives in different ways to compensate for an increased workload. This remodelling of the heart includes changes in the extracellular matrix which will affect systolic and diastolic cardiac function. Furthermore, signal transduction, molecular diffusion and microcirculation will be affected in the hypertrophic process. One important extracellular component is the glycosaminoglycan hyaluronan. It has been shown to play a major role in other conditions that feature cellular growth and proliferation, such as wound healing and malignancies. The aim of this thesis was to investigate hyaluronan and its role in both an experimental rat model of cardiac hypertrophy as well as in cultured mouse cardiomyocytes and fibroblasts. Methods. Cardiac hypertrophy was induced in rats by aortic ligation. Hyaluronan concentration was measured and expression of genes coding for hyaluronan synthases were quantified after 1, 6 and 42 days after operation, in cardiac tissue from the left ventricular wall. Localization of hyaluronan and its receptor CD44 was studied histochemically. Hyaluronan synthesis was correlated to gene transcription using microarray gene expression analysis. Cultures of cardiomyocytes and fibroblasts were stimulated with growth factors. Hyaluronan concentration was measured and expression of genes coding for hyaluronan synthases were detected. Hyaluronan size was measured and crosstalk between cardiomyocytes and fibroblasts was investigated. Results. Increased concentration of hyaluronan in hypertrophied cardiac tissue was observed together with an up-regulation of two hyaluronan synthase genes. Hyaluronan was detected in the myocardium and in the adventitia of cardiac arteries whereas CD44 staining was mainly found in and around the adventitia. Hyaluronan synthesis correlated to the expression of genes, regulated by transcription factors known to initiate cardiac hypertrophy. Stimulation of cardiomyocytes by PDGF-BB induced synthesis of hyaluronan. Cardiomyocytes also secreted a factor into culture media that after transfer to fibroblasts initiated an increased synthesis of hyaluronan. When stimulated with hyaluronan of different sizes, a change in cardiomyocyte gene expression was observed. Different growth factors induced production of different sizes of hyaluronan in fibroblasts. The main synthase detected was hyaluronan synthase-2. Cardiomyocytes were also shown to secrete microvesicles containing both DNA and RNA. Isolated microvesicles incubated with fibroblasts were observed by confocal microscopy to be internalized into fibroblasts. Altered gene expression was observed in microvesicle stimulated fibroblasts. Conclusion. This study shows that increased hyaluronan synthesis in cardiac tissue during hypertrophic development is a part of the extracellular matrix remodelling. Cell cultures revealed the ability of cardiomyocytes to both synthesize hyaluronan and to convey signals to fibroblasts, causing them to increase hyaluronan synthesis. Cardiomyocytes are likely to express receptors for hyaluronan, which mediate intracellular signalling causing the observed altered gene expression in cardiomyocytes stimulated with hyaluronan. This demonstrates the extensive involvement of hyaluronan in cardiac hypertrophy.

Expression studies have implicated Fibronectin Leucine Rich Transmembrane protein 2 (FLRT2) in cranial neural crest cell migration and pre-chondrogenic cell condensation during craniofacial skeletogenesis. This aim of this study was to characterize the expression of FLRT2 and its relationship to the extracellular matrix (ECM) in ATDC5 chondroprogenitor cells. Immunofluorescence studies localized FLRT2 to the cell membrane as well as exracellularly, where it colocalized with fibronectin. FLRT2 was identified in the ATDC5-derived ECM after cell extraction. Further to its colocalization with fibronectin, FLRT2 associated with fibronectin-coated beads in cell cultures. Co-immunoprecipitation confirmed that FLRT2 and fibronectin interact, either directly or indirectly. Blocking fibronectin fibril formation in ATDC5 cell cultures demonstrated a concomitant decrease in extracellular FLRT2 accumulation. It appears that FLRT2 may exist in both a membrane-bound and a shed form. Either or both of these forms may participate in cell-ECM interactions in cooperation with fibronectin or other ECM proteins.

Expression studies have implicated Fibronectin Leucine Rich Transmembrane protein 2 (FLRT2) in cranial neural crest cell migration and pre-chondrogenic cell condensation during craniofacial skeletogenesis. This aim of this study was to characterize the expression of FLRT2 and its relationship to the extracellular matrix (ECM) in ATDC5 chondroprogenitor cells. Immunofluorescence studies localized FLRT2 to the cell membrane as well as exracellularly, where it colocalized with fibronectin. FLRT2 was identified in the ATDC5-derived ECM after cell extraction. Further to its colocalization with fibronectin, FLRT2 associated with fibronectin-coated beads in cell cultures. Co-immunoprecipitation confirmed that FLRT2 and fibronectin interact, either directly or indirectly. Blocking fibronectin fibril formation in ATDC5 cell cultures demonstrated a concomitant decrease in extracellular FLRT2 accumulation. It appears that FLRT2 may exist in both a membrane-bound and a shed form. Either or both of these forms may participate in cell-ECM interactions in cooperation with fibronectin or other ECM proteins.

El conocimiento de las interacciones entre célula/proteína/biomaterial es fundamental para la ingeniería de superficies debido a las numerosas aplicaciones biomédicas y biotecnológicas que se están desarrollando así como al éxito clínico que han alcanzado muchos implantes. La respuesta biológica final inducida por los implantes está fuertemente influenciada por las interacciones superficiales entre los componentes biológicos y el material sintético. Las propiedades físicas y químicas de la superficie de un biomaterial, en lugar de las propiedades en su masa, influyen directamente en la capa de proteínas que se adsorben sobre el biomaterial y, como consecuencia de ello, en la respuesta celular a la misma, tanto in vitro como in vivo.El objetivo de esta tesis doctoral es profundizar en el conocimiento de las interacciones material-biosistema, con el énfasis en el descubrimiento de relaciones entre las propiedades superficiales de las superficies de titanio y su respuesta biológica in vitro.El titanio comercialmente puro (Ti c.p.) está siendo ampliamente utilizado con éxito durante muchos años como biomaterial para implantes en cirugía ósea. Su excelente biocompatibilidad se basa en sus adecuadas propiedades mecánicas y, con mayor importancia, en su excelente resistencia a la corrosión. Esta última se debe principalmente a la formación espontanea de una fina película de óxido de titanio que le confiere protección natural contra los ataques degradativos. La modificación de la topografía de la superficie del titanio ha sido objeto de investigación en el pasado con el fin de mejorar la osteointegración. El granallado de partículas es una de las tecnologías más utilizadas para conferir rugosidad a las superficies del titanio. La rugosidad óptima y el tipo de partículas abrasivas del granallado para una respuesta óptima in vitro e in vivo fue previamente determinada en nuestro laboratorio. Sin embargo, todavía están por determinar cuáles son las causas últimas que llevan al biomaterial a su exitosa respuesta biológica.En este trabajo se han estudiado superficies pulidas y rugosas de Ti c.p. obtenidas mediante el granallado con partículas abrasivas de diferente composición química(Al2O3 y SiC) y diferentes tamaños (212-300&#956;m; 425-600&#956;m; 1000-1400&#956;m). La completa caracterización de las propiedades física y química de la superficie, incluyendo la rugosidad, la composición química, la mojabilidad/energía libre y la carga eléctrica de las superficies ensayadas ha llevado a una serie de relevantes conclusiones. Entre ellas, cabe destacar que a) la composición química de las partículas de granallado, así como el método de esterilización fueron los principales factores que influyeron en la mojabilidad y la energía libre superficial de las superficies de titanio estudiadas, b) el método de esterilización cambió en la energía superficial el carácter de donante de electrones de las superficies mediante el cambio de la cantidad y la naturaleza de las sustancias adsorbidas, y c) la composición química de las partículas de granallado no influyó en la carga eléctrica a pH fisiológico ni en el punto isoeléctrico de las superficies.Un segundo paso consistió en el uso de una microbalanza de cristal de cuarzo con monitorización de la energía de disipación, para el estudio de la cinética de adsorción (cantidad y conformación) y de los procesos de adsorción competitiva de tres proteínas de especial interés en los procesos de curación del hueso - la albúmina de suero bovino (BSA), el fibrinógeno (Fbg), y la fibronectina (Fn)- en sensores lisos recubiertos de TiO2. Se determinaron diferentes modelos de procesos de adsorción con una, dos o múltiples pasos distinguibles en función de las proteínas en solución. La capa adsorbida de BSA mostró los cambios más significativos en sus propiedades mecánicas, de conformación y de incorporación de agua hasta que se alcanzaron las condiciones estables de adsorción de proteínas. La BSA, la más pequeña de las proteínas ensayadas, desplazó la Fn y el Fbg cuando se ensayó en condiciones de la competencia por la adsorción, indicando su mayor afinidad por las superficies de TiO2. También se emplearon técnicas de marcaje fluorescente para el estudio de la adsorción proteica en superficies rugosas granalladas. En este estudio, por un parte, se pudo determinar que la cantidad de Fn y BSA adsorbidas en las superficies granalladas está directamente correlacionada con su energía superficial. Por otra parte, se visualizó la adsorción de fibronectina en solución sobre muestras granalladas rugosas de Ti. La Fn formó un patrón irregular de adsorción con una mayor cantidad de proteína adsorbida en los picos que en los valles de la topografía.También se evaluó la organización espacial de la matriz extracelular de los osteoblastos, ECM, sobre superficies de Ti lisas y rugosas por medio de la visualización de las fibrillas de Fn teñidas con marcador fluorescente. Las células osteoblásticas depositaron las fibrillas de Fn con un determinado patrón organizado dentro de la matriz total secretada. Aparecen como una película que cubre la parte superior de las diferentes superficies rugosas de titanio. Un resultado relevante es que el espesor de esta capa aumentó con la rugosidad de la topografía subyacente. Sin embargo no más de la mitad de la máxima distancia pico-valle se cubrió con la proteína secretada y/o reorganizada.Por último, teniendo en cuenta las diferencias en la organización de la ECM y laadsorción de Fn en las superficies ensayadas de Ti, se realizó un estudio de qRT-PCR para determinar la influencia de las propiedades superficiales del titanio, con y sin preadsorción de Fn, en la respuesta osteoblástica. La expresión génica de la subunidad &#61537;5 de la integrina celular, como marcador de la adhesión celular, se incrementó en las superficies granalladas con SiC en comparación con las granalladas con alúmina. Este resultado fue correlacionado con la mayor cantidad de Fn adsorbida debido a la mayor energía superficial de las superficies granalladas con SiC. El aumento de la rugosidad, así como la presencia de partículas de alúmina en las superficies rugosas incrementó la actividad de ALP y la expresión génica de ALP mRNA por los osteoblastos, y por lo tanto su diferenciación. / The understanding of cell/protein/biomaterial interactions is critical to the engineering of substrates for numerous biomedical and biotechnological applications and to the clinical success of implants. The final biological response induced by implants is strongly influenced by the biological-components/synthetic-material surface interactions. It is well accepted that the physical and chemical surface properties of a biomaterial rather than its bulk properties will influence the protein adlayer and then the cell response to it, both in vitro and in vivo.The aim of this PhD thesis is to gain an increased understanding of the materialbiosystem interactions, with an emphasis on establishing correlations between surface properties of titanium surfaces and its in vitro biological response.Commercially pure titanium (c.p. Ti) is being widely and successfully used implant biomaterial in bone surgery over many years. Its excellent biocompatibility is based in its appropriate mechanical properties and, more importantly, in its excellent corrosion resistance, which is mainly due to the presence of a naturally-occurring thin protective titanium oxide film. Modification of titanium surface topography has been a subject of research in the past with the purpose of improving its osseointegration. Grit blasting is one of the most used technologies to roughen titanium surfaces for this purpose. The optimal roughness and type of abrasive blasting-particles for a better in vitro and in vivo response was previously determined in our lab. However, which and how different relevant surface properties of the blasted titanium surfaces induce that optimal biological behavior is still poorly understood.Smooth/polished and rough c.p. Ti surfaces obtained by blasting with abrasiveparticles of different chemical composition (Al2O3 and SiC) and different sizes (212-300&#956;m; 425-600&#956;m; 1000-1400&#956;m) were studied. The comprehensive characterization of physical and chemical surface properties, including roughness, chemical composition, wettability/free energy and electrical charge of the tested surfaces led to a series of relevant conclusions. Among them, it is worth noting that a) the chemical composition of the grit-blasting particles as well as the method of sterilization were found the main factors influencing wettability and surface free energy of the titanium surfaces; b) the sterilization method changed the electron donor character of the surfaces by changing the amount/nature of physisorbed substances on the surfaces, and c) the chemical composition of the blasting particles did not influence on the electrical charge at physiological pH and the isoelectric point of the surfaces.A second step consisted in the use of a quartz crystal microbalance with monitoring of the energy dissipation to study the adsorption kinetics (amount and conformation) and adsorption competition processes of three proteins of special interest in the healing processes of bone -bovine serum albumin (BSA), fibrinogen (Fbg), and fibronectin (Fn)-on smooth TiO2-coated sensors. Different patterns of adsorption with processes in one, two or multiple distinguishable steps were determined depending of the protein in solution. The BSA adlayers showed the most significant changes in their mechanical properties/conformation/incorporation of water until steady protein-adsorption conditions were reached. BSA, the smallest of the tested proteins, displaced Fn and Fbg when in competition for adsorption, which is an indication of its higher affinity for TiO2 surfaces. Fluorescent labelling techniques where used to study protein adsorption on blasted rough surfaces. Most significantly, the amount of Fn and BSA adsorbed on blasted surfaces was positively correlated with their surface energy. The adsorption of fibronectin from solution on shot-blasted rough titanium surfaces resulted in an irregular pattern of adsorption with a higher amount of protein adsorbed on peaks than on valleys of the topography.Further, the spatial organization of the osteoblast extracellular matrix, ECM, on smooth and rough Ti surfaces was evaluated by visualizing fluorescently-stained Fn-fibrils. Osteoblast-like cells deposited Fn- fibrils in a specific facet-like pattern that was organized within the secreted total matrix. It appeared as a film overlying the top of the different rough titanium surfaces. Interestingly, the thickness of this layer increased with the roughness of the underlying topography, but no more than half of the total maximum peak-to-alley distance was covered.Finally, taking into consideration the differences in ECM organization and Fn adsorption on the tested Ti surfaces a qRT-PCR study was carried out to elucidate the influence of titanium surface properties with and without Fn-precoatings on the osteoblast response. The expression of &#61537;5 integrin subunit gene, as a marker for cell adhesion, was increased in SiC-blasted surfaces compared to alumina-blasted surfaces. This was related to the higher amount of adhesive-protein Fn adsorbed caused by the higher surface energy of SiC-blasted surfaces. The increase of roughness as well as the presence of alumina particles on blasted surfaces increased ALP activity and ALP mRNA gene expression by osteoblasts, and so their differentiation.This research work contribute to increase our knowledge on the interactions taking place at the bio/non-bio interface between different biological components -water, proteins, cells- and materials of clinical relevance, such as rough titanium. Theintertwined effects of the different properties of the synthetic surfaces appear as a challenge to unravel the ultimate causes that determine the fate of cells on synthetic biomaterials.