Tube and Tooth Fragment Skinfold Models for Pulp Neovascularization Studies: A Qualitative Assessment

Previously, our in vitro tube model, simulating the root canal space, demonstrated that human dental pulp stem cell (DPSC)-derived endothelial-like cells (ECs) in fibrin gel can form vascular-like structures. In the present study, we extended to an in vivo tube model and developed a tooth fragment dorsal skinfold window chamber (t-SWC) in vivo model to investigate vascularization and blood perfusion from the host vasculature. In the tube model, we observed significant variation in the extension of red blood cell (RBC)-present vasculature from the host tissue into the tube. There was a demarcation between the RBC-present vessels and RBC-absent vessels in the tissue within the tube. Cells were labeled with green fluorescent protein (GFP) and traced with anti-GFP antibodies via immunohistochemistry. RBC-present vessels were made of both GFP⁺ and GFP^- cells, indicating the ingrowth of the mouse vascular system via angiogenesis. Cells within the RBC-absent vessel region exhibited robust growth, and morphologically odontoblast-like cells were observed at the coronal end in some samples. DPSC-ECs appeared to contribute to RBC-containing vessel formation, although they were insufficient to independently generate fully mature, functional vasculature capable of providing blood supply. Freshly prepared mineral trioxide aggregate exhibited cytotoxic effects on implanted cells within the tube, and this toxicity was not mitigated by the in vivo environment. The t-SWC model allowed real-time and direct observation of the in vivo neovascularization. Under this testing model, blood perfusion within the canal space was not detected during up to 50 d of observation. These qualitative observations suggest that, in the evaluable specimens, DPSC-EC-derived vascular-like structures did not show detectable functional maturation into perfused vessels. The limitations of the ectopic models may play a role in the findings, as subcutaneous tissue is different from the periapical environment. Regardless, the 2 models provided experimental platforms for evaluating host-construct vascular interactions and neovascularization in the canal space.