Bone grafts are utilized clinically to guide tissue regeneration. Autologous bone and allogeneic bone are the current clinical standards. However, there are significant limitations to their use. To address the need for alternatives to autograft and allograft, researchers have worked to develop synthetic grafts, also referred to as scaffolds. Despite extensive efforts in this area, a gap persists between basic research and clinical application. In particular, solutions for repairing critical size and/or load-bearing defects are lacking. The aim of this thesis work was to address two critical barriers preventing design of successful tissue engineering constructs for bone regeneration within critical size and /or load-bearing defects. Those barriers are insufficient osteointegration and slow neovascularization. In this work, the effects of scaffold microporosity, recombinant human bone morphogenetic protein-2 delivery and endothelial colony forming cell vasculogenesis were evaluated in the context of bone formation in vivo. This was accomplished to better understand the role of these factors in bone regeneration, which may translate to improvements in tissue engineering construct design.

Biphasic calcium phosphate (BCP) scaffolds with controlled macro- and microporosity were implanted in porcine mandibular defects. Evaluation of the BCP scaffolds after in vivo implantation showed, for the first time, osteocytes embedded in bone within scaffold micropores (< 10 μm) as well as the most extensive bone growth into micropores to date with bone penetration throughout rods 394 um in diameter. The result is the first truly osteointegrated bone scaffolds with integration occurring at both the macro and micro length scales, leaving no "dead space" or discontinuities of bone in the defect site. The scaffold forms a living composite upon integration with regenerating bone and this has significant implications with regard to improved scaffold mechanical properties.

The presence of osteocytes within scaffold micropores is an indication of scaffold osteoinductivity because a chemotactic factor must be present to induce cell migration into pores on the order of the cell diameter. It is likely that the scaffold undergoes in vivo modifications involving formation of a biological apatite layer within scaffold micropores and possibly co-precipitation of endogenous osteoinductive proteins. To further investigate the effects of scaffold osteoinductivity, BCP scaffolds were implanted in porcine mandibular defects with rhBMP-2, which was partially sequestered in the micropores. Cell migration into osteoinductive scaffold micropores can be enhanced through the delivery of exogenous rhBMP-2 further promoting multi-scale osteointegration.

Finally, endothelial colony forming cells (ECFCs) isolated from human umbilical cord blood (UCB) were evaluated in terms of their in vivo vasculogenic potential in the context of bone formation. This work was completed to determine if ECFCs could be utilized in a bone tissue engineering construct to promote neovascularization. ECFCs were combined with a BCP scaffold and rhBMP-2 and implanted subcutaneously on the abdominal wall of NOD/SCID mice. The result was formation of perfused human vessels within BCP scaffold macropores that were present at 4 weeks. The high density and persistence of human vessels at four weeks indicates that human UCB ECFCs exceed their reported in vivo vasculogenic potential when combined with rhBMP-2 and a BCP scaffold. This shows a dual role for BMP-2 in the context of bone regeneration.

Collectively, the thesis demonstrates that 1) the design of synthetic bone scaffolds should include controlled multi-scale porosity to promote multi-scale osteointegration, which may significantly improve scaffold mechanical properties and 2) human umbilical cord blood-derived endothelial colony forming cells have potential for promoting neovascularization in a bone defect when combined with rhBMP-2.