Cartilage development relies on the coordinated presentation of biological signals to direct chondrocyte morphology and function. This is largely controlled by perlecan, a heparan sulfate proteoglycan (HSPG). Understanding the role of perlecan and its pendant glycosaminoglycan chains (GAG) in cartilage development is essential for advances in tissue engineered cartilage replacement strategies. Perlecan was immunolocalised to the pericellular matrix of prehypertrophic and hypertrophic chondrocytes in human fetal feet. Human fetal chondrocytes were isolated and cultured in 3-dimensional (3D) scaffolds for a period of 4 weeks. Their chondrogenic phenotype, based on extracellular matrix (ECM) components, was assessed and compared to 2D cultures. Chondrocyte perlecan was immunopurified from human fetal chondrocytes grown in vitro and fetal cartilage tissue and characterised using a combination of antibody-based techniques (ELISA, Western blotting) and gel electrophoresis. The biological function of chondrocyte perlecan was determined by its ability to form ternary complexes with fibroblast growth factors (FGF) and their receptors (FGFR) using an antibody-based technique as well as a cell proliferation assay using cells expressing FGFR isotypes. Perelcan was restricted to the prehypertrophic and hypertrophic zones of cartilage. This zonal organisation of chondrocytes and chondrogenic properties, determined by their morphology and PG deposition, was recapitulated in the 3D constructs while 2D cultures displayed dedifferentiated chondrocytes. Exogenous FGF2 promoted chondrocyte proliferation, while FGF18 stimulated the synthesis of perlecan, reflecting chondrocyte hypertrophy. Chondrocyte perlecan (630kDa) contained HS, chondroitin sulfate (CS) and keratan sulfate (KS) chains. Chondrocyte perlecan formed HS dependent ternary complexes with FGF2-FGFR1c and FGF18-FGFR3c, while FGF18-FGFR3c binding to perlecan protein core was also observed. Binding of FGF18-FGFR3c to chondrocyte perlecan HS was more promiscuous than FGF2-FGFR1c. Furthermore, chondrocyte perlecan HS mediated biological activity with FGF18 via FGFR3c, which was modulated by mammalian heparanase, while no biological activity was elicited by FGF2-FGFR1c. The findings underline how perlecan and its GAGs interact with FGF and FGFR in a spatio-temporal manner to promote signalling, effecting chondrocyte behaviour and morphology in cartilage development. This insight can be utilised in tissue engineering to improve the development of biologically functional cartilage replacements.

The main question that this thesis addresses is what is the optimal way of producing tissue grown in the peritoneal cavity around a foreign body for its use as a vascular graft? It is known that a foreign body implanted into the peritoneal cavity induces an inflammatory response with cells recruited from within the peritoneal cavity to encapsulate the foreign body. Over the course of two to three weeks these cells produce an organised matrix and differentiate to become myofibroblasts. Tubes of these ‗tissue capsules‘ have been transplanted into the arterial vasculature in several animal models where the tissue capsule differentiates into an arterial structure. This structure consists of a layer of smooth muscle-like cells, adventitia of dense connective tissue including vasa-vasorum and an endothelial layer of flattened mesothelial cells. In order to determine whether the tissue would further differentiate ex vivo in response to mechanical stimulus an in-vitro bioreactor system was built to house tissue capsules produced in a variety of animal models. This bioreactor system could house 4 tissue capsules under physiological conditions including standard pulse rates, pressures and temperatures experienced by an artery. Boiled blood clot (BBC) scaffolds were implanted into the peritoneal cavity of rats to produce tissue capsules. After two weeks of development in the peritoneal cavity, tissue capsules were harvested and implanted into the bioreactor. Tissue capsules grafted into the bioreactor were subjected to mechanical force for a range of time-points, pressure, pulse and flow rates. When analysing tissue immunohistochemically, elastin, myosin, αSMA and desmin were detected. This staining was not consistent across all samples and only present in small parts of some tissue tested. Western analysis did not show any expression of αSMA or myosin. Finally the morphology of the tissue also resembled that of tissue previously implanted into the arterial circulation, but development of mechanical properties were not to the extent that would make the tissue useful as a vascular graft. The bioreactor system was thus modified to be able to house tissue for a period of 3 weeks. This system successfully housed tissue capsules under mechanical force in physiological ranges. Next, a range of materials were tested for their ability to be included into the peritoneal implant device used for the large animal model. Elasteon 80A did not produce any cellular growth or peritoneal pathology in all implanted samples (n = 4). Cloisite, a pro-inflammatory material produced large tissue capsule development over a 2 week implant period in 25% of samples however this tissue was heavily adhered to the greater omentum and dependent on its vascular supply. This data suggested that Elasteon could be used to coat the outer surface of a peritoneal implant device to decrease the rate of peritoneal adhesions. Three devices were designed and fabricated for their use in generating tissue for the modified Mitrofanoff procedure which requires a length of tissue to be implanted between the umbilicus and the bladder as a fistula. In all three cases no implantable material was produced that could be used for this procedure. To modify the device that could be used to produce tissue for any surgical application, a range of devices was produced and the animal model was changed to pigs. Materials incorporated into these devices include Dexon mesh and polyethylene. These devices also did not produce any tissue that could possibly be used as a vascular graft. A novel material, polymer BD347 was then produced for use in developing tissue within the interior of the device to provide greater growth and mechanical properties for developing a vascular graft. In toxicological studies, the replacement rate of cells was unaffected after seven days of incubation of fibroblasts at confluence with the polymer. A range of mechanical properties from pig vasculature was gained so that a sheet of polymer with similar properties to that of a vascular graft could be made. This polymer was fabricated as a tube and implanted into the peritoneal cavity of rats. The implanted polymer remained free-floating with a capsule of tissue in 78% of cases. A device was designed that has the ability to impart a physiological pulsation force on the developing tissue capsule in the peritoneal cavity using a sheep model. When two devices were implanted for a period of 10 days in each animal these devices produced no complications for the animal. Upon harvest all devices were free of adhesion and did not cause any peritoneal or dermal infection. In 100% of cases this device produced tissue that was thick and consistent along the length of the implant. The quality of tissue differed greatly macroscopically between tissue produced around pulsing and non-pulsing scaffolds, but microscopically the structure of both tissues was not significantly different. Approximately 90% of cells in this tissue stained positively for CD45. Tissue in pulsing devices produced a higher amount of vimentin expression in CD45 positive staining cells than tissue in non-pulsing devices. Mechanical properties of tissue in pulsed devices were also much greater than tissue in non-pulsed devices. Two of the pulsed tissues were grafted into the carotid artery of sheep as arterial patches. In one animal tissue lasted a period of 1 week before it ruptured. In the second animal tissue lasted a period of 2 weeks at which time the animal was sacrificed. In this sheep a layer of endothelial cells had migrated to populate areas of the tissue patch. Pulsation of the implant device enhanced the development of tissue capsule in the peritoneal cavity towards arterial properties. These studies provide information on the materials and designs required to produce peritoneal-derived tissue capsules that can be used in a range of surgical applications. These studies also provide information on how this tissue responds to mechanical force and provides an in vitro system for testing this tissue. This work in this thesis has produced a device that is in the stage of pre-clinical development to be used as a potential therapy for cardiovascular disease. This device is a novel development from previous devices used for generating tissue capsules for engraftment and is a significant contribution to work in developing a replacement artery.

Cartilage development relies on the coordinated presentation of biological signals to direct chondrocyte morphology and function. This is largely controlled by perlecan, a heparan sulfate proteoglycan (HSPG). Understanding the role of perlecan and its pendant glycosaminoglycan chains (GAG) in cartilage development is essential for advances in tissue engineered cartilage replacement strategies. Perlecan was immunolocalised to the pericellular matrix of prehypertrophic and hypertrophic chondrocytes in human fetal feet. Human fetal chondrocytes were isolated and cultured in 3-dimensional (3D) scaffolds for a period of 4 weeks. Their chondrogenic phenotype, based on extracellular matrix (ECM) components, was assessed and compared to 2D cultures. Chondrocyte perlecan was immunopurified from human fetal chondrocytes grown in vitro and fetal cartilage tissue and characterised using a combination of antibody-based techniques (ELISA, Western blotting) and gel electrophoresis. The biological function of chondrocyte perlecan was determined by its ability to form ternary complexes with fibroblast growth factors (FGF) and their receptors (FGFR) using an antibody-based technique as well as a cell proliferation assay using cells expressing FGFR isotypes. Perelcan was restricted to the prehypertrophic and hypertrophic zones of cartilage. This zonal organisation of chondrocytes and chondrogenic properties, determined by their morphology and PG deposition, was recapitulated in the 3D constructs while 2D cultures displayed dedifferentiated chondrocytes. Exogenous FGF2 promoted chondrocyte proliferation, while FGF18 stimulated the synthesis of perlecan, reflecting chondrocyte hypertrophy. Chondrocyte perlecan (630kDa) contained HS, chondroitin sulfate (CS) and keratan sulfate (KS) chains. Chondrocyte perlecan formed HS dependent ternary complexes with FGF2-FGFR1c and FGF18-FGFR3c, while FGF18-FGFR3c binding to perlecan protein core was also observed. Binding of FGF18-FGFR3c to chondrocyte perlecan HS was more promiscuous than FGF2-FGFR1c. Furthermore, chondrocyte perlecan HS mediated biological activity with FGF18 via FGFR3c, which was modulated by mammalian heparanase, while no biological activity was elicited by FGF2-FGFR1c. The findings underline how perlecan and its GAGs interact with FGF and FGFR in a spatio-temporal manner to promote signalling, effecting chondrocyte behaviour and morphology in cartilage development. This insight can be utilised in tissue engineering to improve the development of biologically functional cartilage replacements.

Articular cartilage disorders are a leading cause of human disability in many countries around the world. In this work, new techniques and strategies were developed to improve the quality of cartilage produced in vitro by methods of tissue engineering. Chondrocytes were isolated from the hip and knee joints of aborted human foetuses. The cells were expanded and seeded into scaffolds and the seeded scaffolds were cultured in perfusion bioreactors. The quality of the final cartilage constructs was assessed biochemically by measuring their content of glycosaminoglycan (GAG), total collagen and collagen type II and histologically by staining cross-sections of the constructs for GAG, collagen type I and collagen type II. The amount of proteoglycan released in the culture medium was also measured at regular intervals. Proteoglycans from tissue-engineered cartilage and spent culture medium were compared and analysed for degradation and capability of aggregation. During monolayer expansion, the chondrocyte differentiation indices decreased, the cell size increased and the percentage of cells present in G2/S??M phase decreased with the greatest changes occurring during the first passage. Expanding chondrocytes in PGA or PGA??alginate scaffolds produced cells with a higher level of differentiation than monolayer-expanded cells. However, PGA and PGA??alginate could not be justified as suitable systems for the routine expansion of chondrocytes mainly because of the relatively low cell proliferation obtained. Two new methods for seeding of cells into scaffolds were investigated using PGA and PGA??alginate as scaffold materials. Both methods produced high seeding efficiencies and homogeneous distribution of cells. When seeded PGA??alginate scaffolds were cultured in perfusion bioreactors, they produced good quality constructs with higher concentrations of extracellular matrix (ECM) components compared with previously described methods. However, when seeded PGA scaffolds were cultured in perfusion bioreactors, they produced small constructs of poor quality. Investigation of the effect of medium flow rate on the PGA scaffolds showed that a low flow rate was needed at the beginning of the culture to enable the cells to form a framework onto which other synthesised elements could deposit. Applying a gradual increase in medium flow rate to PGA scaffolds cultured in perfusion bioreactors solved the shrinkage problem and produced constructs with quality similar to those produced using PGA??alginate scaffolds. A novel compression bioreactor that mimicked the physiological stimulation of cartilage by joint movement was constructed. Using this bioreactor, compressed constructs showed significantly higher wet weight and higher concentrations of GAG, total collagen and collagen type II compared with non-compressed constructs.

Thesis (Ph. D.)--Tufts University, 2004. / Adviser: David L. Kaplan. Submitted to the Dept. of Chemical and Biological Engineering. Includes bibliographical references. Access restricted to members of the Tufts University community. Also available via the World Wide Web;

Thesis (Ph.D.)--Tufts University, 2004. / Adviser: David L. Kaplan. Submitted to the Dept. of Biotechnology Engineering. Includes bibliographical references (leaves 162-222). Access restricted to members of the Tufts University community. Also available via the World Wide Web;

Thesis (Ph. D.)--University of Hong Kong, 2007. / Also available in print.

Thesis (Ph.D.)--University of Delaware, 2007. / Principal faculty advisors: John F. Rabolt and D. Bruce Chase, Dept. of Materials Science & Engineering. Includes bibliographical references.

Diss. Stockholm : Karolinska institutet, 2006.

Zugl.: München, Techn. Univ., Diss., 2007