Granulocyte/macrophage colony-stimulating factor influences angiogenesis by regulating the coordinated expression of VEGF and the Ang/Tie system

Abstract

Granulocyte/macrophage colony-stimulating factor (GM-CSF) can accelerate wound healing by promoting angiogenesis. The biological effects of GM-CSF in angiogenesis and the corresponding underlying molecular mechanisms, including in the early stages of primitive endothelial tubule formation and the later stages of new vessel maturation, have only been partially clarified. This study aimed to investigate the effects of GM-CSF on angiogenesis and its regulatory mechanisms. Employing a self-controlled model (Sprague-Dawley rats with deep partial-thickness burn wounds), we determined that GM-CSF can increase VEGF expression and decrease the expression ratio of Ang-1/Ang-2 and the phosphorylation of Tie-2 in the early stages of the wound healing process, which promotes the degradation of the basement membrane and the proliferation of endothelial cells. At later stages of wound healing, GM-CSF can increase the expression ratio of Ang-1/Ang-2 and the phosphorylation of Tie-2 and maintain a high VEGF expression level. Consequently, pericyte coverages were higher, and the basement membrane became more integrated in new blood vessels, which enhanced the barrier function of blood vessels. In summary, we report here that increased angiogenesis is associated with GM-CSF treatment, and we indicate that VEGF and the Ang/Tie system may act as angiogenic mediators of the healing effect of GM-CSF on burn wounds.

Figure 1. The speed and quality of burn wound healing.

(A) Photographs showing macroscopic wound healing in both the control and GM-CSF treatment groups at different time points following the burn event. (B) Wound healing in the control and GM-CSF treatment groups in self-controlled rat models at days 7 and 14 after the burn event. (C) Graph showing the wound healing rate in the two groups (mean±SD) (*P<0.05). (D) Histological changes in burn wounds in the two groups at days 14 and 21 after the burn event. Scale bar = 50 mm.

Figure 2. Immunofluorescence assay of proliferating endothelial cells in the wound.

(A) Representative examples of double staining of Ki67/CD31 (green, CD31; red, Ki67; nucleus, blue) in skin sections from the control group and the GM-CSF treatment group at day 3 after the burn event. Original magnification, ×200. (B) Quantitative comparison of the proliferating capillary index (PCI) in the two groups. The PCI was used to assess the percentage of microvessels with proliferating endothelial cells. All data are expressed as the mean±SD (*P<0.05).

Figure 3. Immunofluorescence analysis of microvascular density (MVD) and pericyte coverage of microvessels in the wound.

(A) A double-labeling technique was used to stain endothelial cells for CD31 expression (green) and pericytes for α-SMA expression (red). Representative images of the 2 groups at days 7 and 14 after the burn event are shown. Original magnification, ×200. (B) The MVD in the two groups at different time intervals. (C) The number of pericytes in the two groups at different time intervals. (D) The microvessel pericyte coverage index (MPI) was quantified by assessing the percentage of microvessels that were associated with α-SMA-positive pericytes. All data are expressed as the mean±SD (*P<0.05).

Figure 4. Expression of Ang-1 and Ang-2 in the skin during wound healing.

(A) Ang-1 mRNA levels in the control and GM-CSF treatment groups were assessed by qRT-PCR. (B) Immunohistochemical staining of Ang-1 protein in the two groups at day 7 after the burn event. Ang-1 protein signals were present mainly in pericyte-like perivascular mural cells. Scale bar = 100 mm. (C) Representative Western blots showing Ang-1 protein levels in the two groups during the healing process. (D) Statistical analysis of Ang-1 protein levels. (E) qRT-PCR analysis of Ang-2 mRNA expression levels in the two groups. (F) Immunohistochemical staining of Ang-2 protein at day 7 after the burn event. Ang-2 protein signals were present in capillary luminal and perivascular mural cells. Scale bar = 100 mm. (G) Representative Western blots showing Ang-2 protein levels during the healing process. (H) Statistical analysis of Ang-2 protein levels. (I) The protein expression ratio of Ang-1/Ang-2. All data above are expressed as the mean±SD (*P<0.05).

Figure 5. Tie2 protein levels, tyrosine phosphorylation, and VEGF expression in the skin wounds.

(A) Immunohistochemical staining of Tie-2 protein in the two groups at day 7 after the burn event. Tie-2 protein signals were present mainly in endothelial cells. Scale bar = 100 mm. (B) Representative Western blots showing Tie2 tyrosine phosphorylation levels and Tie2 protein levels in the two groups. (C) Quantification of the relative intensity of phosphotyrosine-Tie-2 compared with total Tie-2. (D) qRT-PCR analysis of VEGF mRNA expression levels in the two groups. (E) Immunohistochemical staining of VEGF protein at day 7 after the burn event. VEGF protein signals were present in pericyte-like perivascular mural cells. Scale bar = 100 mm. (F) Representative Western blots showing VEGF protein levels during the healing process. (G) Statistical analysis of VEGF protein levels. All of the above data are expressed as the mean±SD (*P<0.05).

Figure 6. Time course of changes in MMP-2 and MMP-9 expression in the wounds.

(A) MMP-2 mRNA levels in the control and GM-CSF treatment groups were assessed by qRT-PCR. (B) Immunohistochemical staining of MMP-2 protein in the two groups at day 3 after the burn event. Scale bar = 100 mm. (C) Representative Western blots showing MMP-2 protein levels in the two groups during the healing process. (D) Statistical analysis of MMP-2 protein levels. (E) qRT-PCR analysis of MMP-9 mRNA expression levels in the two groups. (F) Immunohistochemical staining for MMP-9 protein at day 3 after the burn event. Scale bar = 100 mm. (G) Representative Western blots showing MMP-9 protein levels during the healing process. (H) Statistical analysis of MMP-9 protein levels. All data above are expressed as the mean±SD (*P<0.05).

Figure 7. Electron micrographs of the newly formed microvessels in the wounds.

(A) Ultrastructural characterization of microvessels in the control group at day 14 after the burn event. The microvessels have a poorly organized basement membrane and swelled endothelial cells. Few pericytes coated the endothelium. Original magnification, ×5800. (B) In the GM-CSF treatment group, the endothelial cells formed tight junctions, several elongated pericytes were embedded within an integrated basement membrane, and their longitudinal processes completely surrounded the endothelial tube. E, endothelial cell; P, pericyte; BM, basement membrane. Original magnification, ×5800. (C) Enlargement of the image in the black box. A pericyte can be observed making direct contact with an endothelial cell. A high-power view of an EPI is shown, which is composed of a pericyte cytoplasmic projection and corresponding endothelial indentation. (D) Time course of vascular permeability changes after the burn event in the two groups. Data are expressed as the mean±SD (*P<0.05).