The vascularisation of collagen-glycosaminoglycan scaffolds for bone tissue engineering.

A major problem in the field of tissue engineering (TE) is graft failure due to avascular necrosis of TE constructs. To overcome this problem, one approach might be to first vascularise the TE construct in vitro prior to implantation in vivo. Using this method, the TE construct does not have to rely on host vessel invasion; rather it relies on the anastomosis between preengineered and host vessels. With this in mind, the overall aim of this thesis was to first vascularise a collagen-glycosaminoglycan (CG) scaffold in vitro prior to implantation and to investigate whether the presence of an inherent vascular network within the CG scaffold has the ability to enhance bone tissue repair. The study demonstrated that optimal vessel formation occurred in a highly spatial and temporal manner in vitro. It was shown that the timely addition of mesenchymal stem cells (MSCs), in the role of perivascular cells, to preformed human umbilical vein endothelial cell (EC) networks at day 3, followed by a subsequent 3 days in culture, produced well-developed vessels within the scaffold. This was shown to be regulated by two key growth factors: platelet derived growth factor (PDGF) and vascular endothelial growth factor (VEGF). Functional assessment of these pre-vascularised CG scaffolds, using a subcutaneous implant model in athymic rats, demonstrated that in vitro engineered vessels functionally anastomosed with the host vasculature. Further to this, it was shown that vascularisation was increased in the co-culture group compared to ECs alone and this was attributed to the presence of MSCs as perivascular cells which were observed to be actively involved in vessel formation.

Further assessment was carried out to determine whether the presence of a functional vascular network within the CG scaffold could promote healing in an immunocompetent rat cranial defect model. The co-culture group demonstrated the ability to promote healing at the earlier time point of 4 weeks compared to the empty defect, CG scaffold and ECs alone. Investigation into the host immune response toward xenogenic cells demonstrated that in the coculture group, the presence of MSCs had the ability to exert some immunomodulatory effects and reduce the pro-inflammatory M1 macrophage response at 8 weeks. This was in direct contrast to ECs which observed a further increase in the M1 response at 8 weeks. Finally, this study has demonstrated that the presence of an in vitro engineered vascular network in the CG scaffold led to increased osteogenesis in vitro compared to a CG scaffold containing osteogenically primed MSCs alone. This highlighted the relationship between vasculogenesis and osteogenesis and identified some factors thought to modulate these processes. Furthermore, new information was provided on the ability of a CHA scaffold to support in vitro vessel formation and to further enhance osteogenesis compared to CG scaffolds.

Collectively, this thesis has demonstrated an optimal method with which to vascularise a CG scaffold in vitro that can functionally anastomose with the host vasculature in vivo and promote early bone formation in a critical-sized defect. These processes were shown to occur in a highly regulated manner and were dependent on cellular crosstalk whereby ECs formed the initial vascular structures which were enhanced by delayed addition of MSCs in the role of the perivascular cell. Furthermore, the presence of the EC/MSC co-cultures was shown to interact with osteogenically-primed MSCs to increase osteogenesis in vitro compared to MSCs alone. Key factors such as PDGF and most importantly VEGF were shown to modulate these processes.