One of the greatest obstacles to clinical translation of bone tissue engineering is the inability to effectively and efficiently vascularise scaffolds. This limits the size of defects that can be repaired, as blood perfusion is necessary to provide nutrient and waste exchange to tissue at the core of scaffolds. The goal of this work was to systematically explore whether architecture, at a scale of hundreds of microns, can be used to direct the growth of microvessels into the core of scaffolds. A pipeline was developed for the production of hydroxyapatite surfaces with controlled architecture. Three batches of hydroxyapatite were used with two different particle morphologies and size distributions. On sintering, one batch remained phase pure and the other two batches were biphasic mixtures of α-tricalcium phosphate (α-TCP) and hydroxyapatite. Sample production methods based on slip casting of a hydroxyapatite-gelatin slurry were explored. The most successful of these involved the use of curable silicone to produce moulds of high-resolution, three dimensional (3D) printed parts with the desired design. Parts were dried and sintered to produce patterned surfaces with higher resolution than obtainable through conventional 3D printing techniques. Given the difficulties associated with the structural reproducibility of concave pores architectures in 3D reported in the literature, in this work, a 2.5D model has been developed that varies architectural parameters in a controlled manner. Six contrasting architectures consisting of semi-circular ridges and grooves were produced. Grooves and ridges were designed to have widths of 330 μm and 660 μm, with periodicities, respectively, of 1240 μm and 630 μm. Groove depth was varied between 150 μm and 585 μm. Co-cultures of endothelial cells and osteoblasts were optimised and used to grow microcapillary-like structures (referred to as "microvessels") on substrates. Literature shows that these precursors to microcapillaries contain lumina and can produce functional vasculature, demonstrating their clinical promise. The effects of the composition and surface texture of grooved samples on microvessel formation were studied. It was found that surface microtopography and phase purity (α-TCP content) did not affect microvessel formation. However, hydroxyapatite architecture was found to significantly affect microvessel location and orientation. Microvessels were found to form predominantly in grooves or between convexities. Two metrics - the degree of alignment (DOA) and the degree of containment (DOC) - were developed to measure the alignment of endothelial cell structures and their localisation in grooves. For all patterned samples, the CD31 (an endothelial cell marker) signal was at least 2.5 times higher along grooves versus perpendicular to grooves. In addition, the average signal was at least two times higher within grooves than outside grooves for all samples. Small deep grooves had the highest DOA and DOC (6.13 and 4.05 respectively), and individual, highly aligned microvessels were formed. An image analysis method that compares sample X-ray microtomography sections to original designs to quantify architectural distortion was developed. This method will serve as a useful tool for improvements to architectural control for future studies. This body of work shows the crucial influence of architecture on microvessel self-assembly at the hundreds of micron scale. It also highlights that microvessel formation has a relatively low sensitivity to phase composition and microtopography. These findings have important implications for the design of porous scaffolds and the refinement of fabrication technologies. While important results were shown for six preliminary architectures, this work represents a toolkit that can be applied to screen any 2.5D architecture for its angiogenic potential. This work has laid the foundations that will allow elucidating the precise correspondence between architecture and microvessel organisation, ultimately enabling the "engineering" of microvasculature by tuning local scaffold design to achieve desirable microvessel properties.

Cette thèse transdisciplinaire réalisée en collaboration avec le laboratoire SPCTS (Sciences des Procédés Céramiques et Traitement de Surface) et l’EA 3842 (Homéostasie cellulaire et pathologies) de l’université de Limoges est un projet de recherche à l’interface entre la biologie et la chimie et a été consacrée à l’étude de l’influence des propriétés physico-chimiques de biocéramiques de phosphate de calcium sur leur comportement biologique in vitro.L’exploration des processus d’interaction entre matériaux et cellules reste une problématique scientifique de premier plan tant d’un point de vue fondamental qu’appliqué pour la mise au point de biomatériaux performants. L’objectif final est d’optimiser l’efficacité thérapeutique des céramiques phosphocalciques comme matériaux de substitution pour la régénération osseuse. La première partie de la thèse est une revue bibliographique générale présentant la problématique actuelle abordée en lien avec les besoins cliniques et les limitations des études actuelles. Les connaissances sur la biologie du tissu osseux sain ainsi que les aspects de régulation du processus de remodelage osseux ont également été abordés dans ce chapitre. Ce chapitre se termine par une synthèse bibliographique sur les biomatériaux et la régénération osseuse. Le chapitre 2 est relatif à la synthèse puis à la caractérisation physico-chimique des matériaux céramiques. Des céramiques de trois compositions chimiques : HA (hydroxyapatite : Ca10(PO4)6(OH)2 , SiHA (hydroxyapatite silicatée : Ca10(PO4)5,6(SiO4)0,42(OH)1,6 et CHA (hydroxyapatite carbonatée : Ca9,5(PO4)5,5(CO3)0,48(OH)1,08(CO3)0,23 , chacune avec deux microstructures différentes : dense ou poreuse, ont été élaborées et rigoureusement caractérisées (porosité, topographie de surface, mouillabilité, potentiel zêta, taille des grains, distribution et taille des pores, surface spécifique). Le chapitre 3 décrit l’approche expérimentale employée pour l’évaluation biologique des interactions matériaux/cellules explorées dans ce travail. Les analyses biologiques ont été réalisées avec deux lignées cellulaires différentes. La lignée cellulaire pré-ostéoblastique MC3T3-E1 et la lignée cellulaire de monocytes/macrophages, précurseurs des ostéoclastes RAW 264.7, (très importantes pour les aspects osseux, mais moins souvent explorées que les lignées ostéoblastiques dans la littérature). Enfin, le chapitre 4 reporte et commente les résultats biologiques obtenus dans ce travail. Tous les biomatériaux évalués dans cette étude sont biocompatibles, néanmoins, le biomatériau poreux CHA s’est avéré le plus prometteur des six variantes de biomatériaux testés. / This transdisciplinary thesis, carried out in collaboration with the SPCTS laboratory (sciences of ceramic processes and surface treatment) and EA 3842 (Cellular homoeostasis and pathologies) of the University of Limoges, is a research project at the interface between biology and chemistry and was devoted to the study of the influence of the physico-chemical properties of calcium phosphate bioceramics on their biological behavior in vitro.The exploration of the processes of interaction between materials and cells remains a major scientific issue, both from a fundamental and applied point of view for the development of highperformance biomaterials. The ultimate objective is to optimize the therapeutic efficiency of phosphocalcic ceramics as substitute materials for bone regeneration.The first part of the thesis is a general bibliographic review presenting the current issues tackled with the clinical needs and limitations of current studies. Knowledge of the biology of healthy bone tissue as well as the regulatory aspects of the bone remodeling process was also discussed in this chapter. It includes also a bibliographic overview of biomaterials and bone regeneration.Chapter 2 relates to the synthesis and the physico-chemical characterization of ceramic materials. HA (hydroxyapatite: Ca10 (PO4) 6 (OH) 2, SiHA (silicated hydroxyapatite: Ca10 (PO4) 5.6 (SiO4) 0.42 (OH) 1.6 and CHA (carbonated hydroxyapatite: Ca9.5 (PO4) 5.5 (CO3) 0.48 (OH) 1.08 (CO3) 0.23, ceramics each with two different microstructures : dense or porous, have been elaborated and thoroughly characterized (porosity, surface topography, wettability, zeta potential, grain size, pore size and distribution, specific surface area). Chapter 3 describes the experimental approach used for the biological evaluation of the interactions between materials and cells. Biological analyzes were performed with two different cell lines. The pre-osteoblastic MC3T3-E1 cell line and the RAW 264.7cell line of monocytes / macrophages, precursors of the steoclasts, (very important for the bone aspects, but less often explored than the osteoblastic lines in the literature). Finally, Chapter 4 reports and comments on the biological results obtained in this work. All biomaterials evaluated are biocompatible, nevertheless, the porous CHA biomaterial was the most promising of the six variants of biomaterials tested.