Poster BB13 - Microchannel directionality drives vascularization of a cartilaginous scaffold template for bone engineering

Introduction: Despite bone’s innate regenerative potential, up to 2 million bone-grafting procedures are performed worldwide annually. Within these, critical-sized bone defects are particularly difficult to heal due to poor tissue perfusion, as angiogenesis and revascularization are necessary for normal bone healing. While commercial bone allograft is a clinical gold standard, they may cause nonunion at the allograft-host junction and potential tissue necrosis. Thus, alternative scaffold approaches that support vascular integration in addition to bone formation are necessary. We have previously developed a novel scaffold based on porcine meniscus decellularization (MEND) for cartilage repair. The selective digestion of elastin fibers from meniscal cartilage preserves the bulk of its ECM structures and mechanical properties whilst creating a number of channels for recellularization. Notably, a cartilaginous template may be remodeled through endochondral ossification to form bone, much as it happens during development and fracture healing. Thus, MEND may serve as an ideal platform for bone repair if it can be easily infiltrated by blood vessels. In this study, we aimed to (i) show that MEND can be infiltrated by host vasculature upon implantation and (ii) investigate whether structural morphology within scaffolds can alter the extent of vascular infiltration. To achieve this goal, we implanted MEND with different channel directions and assessed vascularization using a chicken embryo choriolallantoic membrane (CAM) model. The CAM model serves as an ex vivo living system with rapid vasculature development that recapitulates physiological vascular infiltration of material scaffolds.

Materials and

Methods: MEND fabrication: MEND is prepared as previously described by sectioning the medial region of porcine meniscus parallel or perpendicular to the circumferential elastin fibers to obtain scaffolds (0.5 -1 mm) with channels in the transverse plane (tMEND) or sagittal plane (sMEND). Sections are subjected to 4 freeze/thaw cycles (1hr) followed by pepsin and elastase treatment for decellularization and selective removal of elastin. GELFOAM (positive control) and GelMA (negative control) are used in 3mm cylindrical plugs obtained with a biopsy punch.

CAM Assay: Eggs are placed vertically in an incubator at 37.5 C that rotates hourly. On incubation day 4, an incision is made on the blunt point of the egg and the eggshell window fragment is removed to reveal the CAM underneath. On incubation day 8, scaffolds are first treated with bovine serum albumin (BSA, 15min) before second treatment with either BSA or human fibroblast growth factor-2 (hFGF-2, 15min) and individually placed onto the CAM of the chicken embryo. Scaffolds are harvested on incubation day 12 and 14 for fixation, and CAM integration was assessed macroscopically by microscope images on both sides of the scaffolds.

Histology and Immunostaining: Samples are processed through paraffin embedding and sectioning (8um). Then, sections are assessed by histology (Hematoxylin and Eosin (H&E) for structure, Safranin-O and Alcian Blue for GAGs, and Van Gieson’s for elastin, and immunofluorescence (collagen I and alpha-smooth muscle actin (α-SMA) for vascular infiltration).

Results, Conclusions, and Discussions: MEND was successfully decellularized post-enzymatic treatment. Histology revealed complete removal of circumferential elastin fibers. The resulting channels were evenly distributed within the scaffold with diameter (average 8 um) unaffected by channel directions, making it ideal for capillary infiltration (2-12 um).

For in ovo implantations, scaffolds were treated with hFGF-2 to stimulate angiogenesis. The gross structure of the scaffolds did not drastically alter during the 5-7 days in ovo. GELFOAM, sMEND, and tMEND adhered tightly to CAM, while GelMA adhered weakly and often detached. Macroscopically, GELFOAM was nearly encased in CAM tissues with high vascular density within the scaffold. Both sMEND and tMEND were similarly surrounded by CAM tissue but showed different extent of vascular integration into the scaffold. GelMA showed close to no vasculature infiltration.

Histological examination revealed close interactions between CAM and porous scaffolds (Fig. 1). GELFOAM exhibited the highest vascular infiltration, confirming that high scaffold porosity allows for more angiogenesis. For MEND scaffolds, sagittal channels better supported vascular invasion while transverse channels posed an obstacle to infiltrating vessel depth and restricted their localization to the scaffold periphery. α-SMA immunostaining was used to mark perivascular cells to further examine vascular network distribution within each scaffold (Fig. 2). sMEND showed clear infiltration of α-SMA+ cells along the vertically aligned channels, whereas tMEND showed weak α-SMA+ along the scaffold periphery. This suggests that reorienting scaffold microstructures can be leveraged to control vascular infiltration without significantly altering structural and mechanical properties of MEND.

Conclusion:
We here demonstrated that MEND can be extensively infiltrated with vasculature in an in ovo model. Varying channel directions within MEND alters accessible microstructures that serve as entry points for vascular infiltration. This shows promise for MEND’s application in bone tissue engineering as a cartilaginous template that can be vascularized and subsequently remodeled via endochondral ossification.

Acknowledgements (Optional):