Purpose

Bioengineering autogenous tissue for repair of craniofacial defects avoids the limitations associated with traditional autografts or allogeneic transplantation. Successful engineering of thick 3D tissue constructs large enough to solve actual clinical problems requires novel tissue-engineering strategies that address chemotransportative requirements in their design and implementation. Furthermore, since nature ensures adequate bony chemotransportation by establishing vascular networks, if an in vitro tissue engineered construct is to survive in vivo, it must have a vascular network. Engineering an in vitro microvasculature network de novo requires simultaneous co-culture of endothelial cells and osteogenic cells. We hypothesise that de novo vascularized osseous tissue can be constructed by co-culturing multiple cell types on 3D scaffolds in a flow perfusion bioreactor.

Materials and Methods

Adipose-derived mesenchymal stem cells (MSCs) were isolated and expanded from human lipoaspirate. MSCs were differentiated separately in osteogenic and vasculogenic media or differentiated together in combine osteogenic-vasculogenic (dual-differentiation) media. Characterisation of osteoprogenitor-rich (OPC), endothelioprogenitor-rich (EPC), or dually differentiated progenitor-rich (ddPC) cell populations was confirmed by quantitative real-time PCR. Normal human osteoblasts (NHOst) and human umbilical vein endothelial cells (HUVEC) served as terminally differentiated cell lines. The effects of 2D co-culture of various cell combinations (e.g OPC+HUVEC, OPC+EPC etc) on capacity for bone nodule formation was evaluated by von Kossa assay. Optimal culture conditions were tested with cell-seeded custom thick (>6mm) 3D HA-TCP scaffolds in a novel flow-perfusion bioreactor. FEM was used to predict oxygen delivery and fluid shear stress of fluid flow through offset or onset scaffolds. 1.5 x10⁶ MSCs were seeded onto onset or offset 60:40 HA/TCP scaffolds and cultured in static conditions or in flow perfusion. Scaffolds were harvested at 7 and 10 days and cell distribution was assessed by scanning electron microscopy (SEM). Cellularity was quantified by DNA fluorospectrophotometry. Cellular function was assessed by alkaline phosphatase (AP) colorimetric assay.

Results

In 2D, proliferation and function were greatest when more-differentiated vasculogenic cells (i.e. HUVECs) were co-cultured with less-differentiated osteogenic cells (i.e. OPCs). For example, co-cultured HUVEC/OPC formed more bone nodules (263,945µm²) than HUVEC/NHOst (179,840µm²) or NHOst alone (89,608µm²). In 3D, cellular function was enhanced in flow perfusion. FEM analysis revealed areas of low fluid shear stress and relative hypoxia due to laminar flow through onset scaffolds, corrected in offset scaffolds. At days 7 and 10, SEM demonstrated that MSC were only viable on the periphery of the scaffolds in static culture. In contrast, MSC were viable throughout the scaffolds cultured in flow perfusion. DNA fluorospectrophotometry showed greater cellular viability in offset scaffolds than in onset scaffolds at day 7 (578±20 vs. 407±7 U/g scaffold p=0.064) and at day 10 (662±13 vs. 443±6 U/g p=0.072). By 10 days in flow perfusion culture, AP activity was significantly greater (3.86±0.320mM/g) than in static culture (0.909±0.460mM/g, p=0.002).

Conclusions

This is the first report to describe optimizing co-culture combinations for engineering of vascularized osseous tissue. By expanding chemotransportation boundaries using 3D flow-perfusion co-culture, we can develop composite tissue constructs for replacement and repair.