Abstract:
INTRODUCTION: While a critical role of angiogenesis in bone tissue engineering has been well established, the spatiotemporal regulation of angiogenesis and the interaction of engineered construct with host vascular microenvironment during repair have not been well studied. Our understanding of the intricate relationship between osteogenesis and angiogenesis at a repair site has been hampered by the lack of an effective approach that allows tracking of bone healing and neovascularization simultaneously at a high spatiotemporal resolution. To overcome this barrier, we have recently established a cranial bone defect window chamber model in mice that permits high resolution, four-dimensional imaging and analyses of bone defect healing over a period of months using multiphoton laser scanning microscopy (MPLSM). The goal of our current study is to utilize this intravital imaging approach to delineate the spatiotemporal regulation of angiogenesis and the interaction of tissue engineered constructs with host microenvironment during cranial bone defect repair. To enable successful repair of a 2mm cranial defect, nanofibrous matrices prepared from electrospinning were used. Our data revealed an intricate balance between osteogenesis and angiogenesis during nanofiber-mediated repair, underscoring the importance of development smart biomaterials capable of modifying host vascular microenvironment at multiple time scales for effective repair and regeneration. METHODS: Composite nanofibrous matrices were prepared from polycaprolactone (PCL), collagen type I, and hydroxyapatite nanoparticles (nHA) via electrospinning. Bone marrow stromal cells (BMSCs) derived from global green fluorescent protein (GFP) transgenic mice were seeded onto the scaffold. About 20 layers of BMSC-seeded fiber sheets were stacked layer-by-layer to form a 3D tissue construct. Circular grafts (2mm in diameter) were punched out of the assembled 3D construct and used to repair 2 mm defects created in the parietal bone of mouse calvarium in immunodeficient mice. A glass window was mounted on top of the cranial wound for imaging as previously described. An Olympus FV1000-AOM multiphoton imaging system equipped with a Titanium:Sapphire laser was used for imaging of bone healing. Second harmonic generation (SHG), green or red fluorescence were visualized simultaneously using the excitation wavelength 780 nm to generate multiphoton excitation signals from collagen matrix, cells, and blood vessels, which were perfused with a fluorescent dye. Three-dimensional reconstruction and analyses of multichannel z-series image stack were performed using Amira image analysis software (Visage Imaging). Bone healing was further imagined by in vivo longitudinal MicroCT scanning at multiple time points. RESULTS: The assembled 3D construct was examined prior to implantation by Scanning Electron Microscope (SEM), histology and MPLSM. The SEM showed mean fiber diameter of ~600nm. MPLSM demonstrated multi-layered membranous scaffold with uneven distribution of collagen deposit. When inserted into the 2-mm cranial defect created in immunodeficient mice, the tissue grafts induced significant bone formation within defect as indicated by live MicroCT scanning (Fig.1A&B). Although uneven bone formation was observed both in the implants seeded with or without BMSC, quantification demonstrated 3.2-fold higher mineralization in defects treated with BMSC-seeded constructs over a period of 8 weeks (n=7, P<0.05). Histologic analyses at the end time point confirmed bone formation around the scaffolds and within layered fibrous matrices (Fig. 1D-F). MPLSM scanning showed significant amount of SHG+ collagen matrix deposition between fibrous layers (Fig. 1F) with scattered GFP+ osteocytes embedded in the newly formed bone. Histomorphometric analyses demonstrated nearly 3-fold more bone formation in defects treated with BMSC-seeded grafts than those treated with noncellularized bone grafts (Fig. 1I, n=7, p<0.05). Control defect without treatment showed no bone formation within the defect region over 12 week period. MPLSM was performed longitudinally at multiple time points following transplantation in the cranial defect window chamber model to evaluate neovascularization (Fig. 2). Angiogenesis peaked at week 3 prior to significant bone formation and mineralization. By 8-10 weeks, overall angiogenesis was reduced in all three groups examined. Despite marked difference in bone formation, the overall vascularity in terms of vascular volume, diameters, length at the defect region was variable and not significantly different among groups treated with or without BMSCs. However, when analyses were performed to exclude the surface vessels, marked difference was observed with BMSC-seeded grafts showing significantly higher vascularity (e.g. vascular volume, diameters and length) than non-cellular matrices at all depths examined. The differences were further confirmed by immunohistochemical staining for CD31using histologic sections. In attempt to understand the uneven bone formation among constructs implanted into the defects, we further conducted analyses to compare vascularity in areas associated with or without bone formation. We found that in nearly all animals exhibiting poor or partial bone healing, a large number of irregularly-shaped, wavy and entangled blood vessels were present at the areas devoid of bone formation (Fig. 2 panel B&C). Red blood cell flow analyses using MPLSM-enabled line scan showed that these vessels had slow, stalled or disorganized flow rates compared to those vessels associated with bone formation. In addition to disorganized vessels, we also found persistent large diameter arterioles in areas devoid of bone formation. These data suggest that vessels associated with bone formation or fibrotic inflammatory tissues are distinct. Further analyses are underway to define the composition of these different vessel networks associated with bone or fibrotic tissue formation. DISCUSSION: Using layer-by-layer approach we succeeded in constructing a cellularized 3D bone tissue construct for repair of bone defect. To the end of understanding the intricate relationship of osteogenesis and angiogenesis in nanofibrous scaffold-mediated repair, we further employed an intravital imaging approach combining both MicroCT and MPLSM to examine scaffold interaction with the host vascular microenvironment at the repair site. Our study demonstrates that the spatiotemporal regulation of angiogenesis is critical for effective scaffold-mediated defect repair. Dysregulation of angiogenesis could disrupt bone formation and mineralization. By further characterization of the distinct vascular networks associated with healing and non-healing bone defect, we are working towards delineating necessary vascular microenvironmental cues for effective bone defect repair. SIGNIFICANCE. Our current study could contribute to the important knowledge of vascular interaction with bone tissue scaffolds, further aiding in development of therapeutic strategies and smart implants for effective repair and regeneration of bone tissue.