Summary

45S5 bioactive glass combined with hMSC did not permit de novo ectopic bone formation. Such absence of osteogenicity was most likely due to the alkalinization of the 45S5 microenvironment that affects adversely the osteogenic differentiation of stem/precursor cells.

Bone marrow stromal cells (BMSCs) are capable of bone formation and can promote the repair of osseous defects when implanted in appropriate scaffolds. The most promising biomaterials for application in bone tissue engineering (TE) are hydroxyapatite (HA), tricalcium phosphate (TCP), calcium carbonate (coral) ceramics or bioactive glasses (BG) because of their osteoconductive properties and ability to enhance bone formation. However, information regarding the osteogenic potential of hBMSCs in combination with BG scaffolds is strikingly lacking in the TE field. The present study focused on evaluating the osteogenicity of bone constructs prepared from particles of 45S5 BG combined with hBMSCs in comparison with biphasic HA/TCP or coral particles, in a mouse ectopic model.

The in vivo osteogenicity was then correlated with various aspects of the effects of the scaffold materials tested on hBMSCs functions pertinent to bone tissue formation. Particular attention was given to the pH in the microenvironment where the cells reside in TE constructs and its effect on the osteoblastic differentiation of hBMSCs. In vivo experiments evidenced that 45S5 BG constructs with hBMSCs failed to form ectopic bone. In contrast, the cell constructs prepared with either HA/TCP or coral ceramics displayed great and consistent capacity for the ectopic bone formation. The cytocompatibility of hBMSCs on BG material was addressed and no differences were evidenced between HA/TCP and coral substrates related to the adhesion of hBMSCs and their proliferation in vitro. The hBMSCs viability was even higher within the 45S5 BG-containing constructs compared to the other two types of material constructs tested both in vitro and in vivo. These findings indicated that the absence of de novo bone formation in the hBMSCs-containing 45S5 BG constructs was not the result of cytotoxic effects of the BG material.

The potential of osteogenic differentiation of hBMSCs cultured on material substrates was next addressed and the ALP activity of hBMSCs was significantly diminished when these cells were cultured on 45S5 BG as compared to either HA/TCP or coral substrates. Because BG materials are well-known for causing external alkalinisation, the pH was specifically measured in TE constructs. The pH inside the cell-containing BG constructs, measured ex vivo, was 8.0 (i.e. 0.4–0.5 units more alkaline than that measured in the coral- or HA/TCP-constructs). The impact of such external alkalinisation on the osteogenic differentiation of hBMSCs was assessed by culturing the cells over a wide range of alkaline pH. The hBMSCs expression of osteogenic markers, ALP activity and mineralization were not significantly affected at moderate external alkaline pH (≤ 7.90) but were dramatically inhibited at higher pH.

Altogether, these findings provided evidence that despite 45S5 BG are reported to be good osteoconductive materials, they are not necessarily good scaffolds for TE, most likely due to the alkalinization of the 45S5 microenvironment that affects adversely the osteogenic differentiation of precursor cells. Controlling the shifts of pH in the local engineered extracellular environment is a critical issue for the development of bioactive TE scaffolds.