Bioactive glasses and electrospun composites that release cobalt to stimulate the HIF pathway for wound healing applications

Abstract

Background: Bioactive glasses are traditionally associated with bonding to bone through a hydroxycarbonate apatite (HCA) surface layer but the release of active ions is more important for bone regeneration. They are now being used to deliver ions for soft tissue applications, particularly wound healing. Cobalt is known to simulate hypoxia and provoke angiogenesis. The aim here was to develop new bioactive glass compositions designed to be scaffold materials to locally deliver pro-angiogenic cobalt ions, at a controlled rate, without forming an HCA layer, for wound healing applications.

Methods: New melt-derived bioactive glass compositions were designed that had the same network connectivity (mean number of bridging covalent bonds between silica tetrahedra), and therefore similar biodegradation rate, as the original 45S5 Bioglass. The amount of magnesium and cobalt in the glass was varied, with the aim of reducing or removing calcium and phosphate from the compositions. Electrospun poly(ε-caprolactone)/bioactive glass composites were also produced. Glasses were tested for ion release in dissolution studies and their influence on Hypoxia-Inducible Factor 1-alpha (HIF-1α) and expression of Vascular Endothelial Growth Factor (VEGF) from fibroblast cells was investigated.

Results: Dissolution tests showed the silica rich layer differed depending on the amount of MgO in the glass, which influenced the delivery of cobalt. The electrospun composites delivered a more sustained ion release relative to glass particles alone. Exposing fibroblasts to conditioned media from these composites did not cause a detrimental effect on metabolic activity but glasses containing cobalt did stabilise HIF-1α and provoked a significantly higher expression of VEGF (not seen in Co-free controls).

Conclusions: The composite fibres containing new bioactive glass compositions delivered cobalt ions at a sustained rate, which could be mediated by the magnesium content of the glass. The dissolution products stabilised HIF-1α and provoked a significantly higher expression of VEGF, suggesting the composites activated the HIF pathway to stimulate angiogenesis.

Keywords: Bioactive composites; Bioactive glass; Cobalt; HIF pathway; Wound healing.

Fig. 1

Elemental concentration following dissolution of glasses into SBF over time: a Co for 5Co-A and 5Co-B (inset: Co release over the first four days); b Co for 2Co-A and 2Co-B (inset: Co release over the first four days); c Ca; d P; e Si; and f pH of SBF buffer. Data points are mean ± SD of one independent experiment performed in triplicate

Fig. 2

Change in glass structure after incubation in SBF for 21 d: (a) XRD and (b) FTIR. Filled circles (●) represent peaks relating to apatite and hollow squares (□) represent peaks relating to calcium carbonate. s = stretch, b = bend

Fig. 3

SEM images showing changes in glass surface morphology of glass particles (a) before and (b) after incubation in SBF for 21 d. Scale bar is 1 μm

Fig. 4

TEM-EDS analysis of the surface cross-sections of (a) 5Co-A and (b) 2Co-A particles after incubation in SBF for 21 d, and with individual spectra taken at multiple locations marked by a ★. Sections were cut perpendicular to the glass surface by FIB. Scale bar is 250 nm

Fig. 5

a SEM image of a typical electrospun composite. Scale bar is 20 μm. b Co ion release from glass particles and composite in DMEM cell culture media at 0.5 h, 1 h, 2 h, 4 h, 8 h, 24 h, 3 d and 7 d. Data shown is mean ± SD of one independent experiment performed in triplicate

Fig. 6

AlamarBlue reading of cells in contact with control (DMEM) media, conditioned media from glass composites, and DMEM containing 100 μM CoCl₂. Data shown is mean ± SD of three independent experiments. Two way ANOVA conducted with Tukey test with * p < 0.05 compared to Day 1 and # p < 0.05 compared to Day 3

Fig. 7

a A typical western blot showing increased expression of HIF-1α with conditioned media and representative western blot quantification (representative images shown from a total of three independent experiments); (b) VEGF expression of fibroblasts relative to protein: data shown are mean ± SD of three independent experiments. One way ANOVA performed with * p < 0.05