Knockout of endothelial cell-derived endothelin-1 attenuates skin fibrosis but accelerates cutaneous wound healing

Abstract

Endothelin (ET)-1 is known for the most potent vasoconstrictive peptide that is released mainly from endothelial cells. Several studies have reported ET-1 signaling is involved in the process of wound healing or fibrosis as well as vasodilation. However, little is known about the role of ET-1 in these processes. To clarify its mechanism, we compared skin fibrogenesis and wound repair between vascular endothelial cell-specific ET-1 knockout mice and their wild-type littermates. Bleomycin-injected fibrotic skin of the knockout mice showed significantly decreased skin thickness and collagen content compared to that of wild-type mice, indicating that bleomycin-induced skin fibrosis is attenuated in the knockout mice. The mRNA levels of transforming growth factor (TGF)-β were decreased in the bleomycin-treated skin of ET-1 knockout mice. On the other hand, skin wound healing was accelerated in ET-1 knockout mice, which was indicated by earlier granulation tissue reduction and re-epithelialization in these mice. The mRNA levels of TGF-β, tumor necrosis factor (TNF)-α and connective tissue growth factor (CTGF) were reduced in the wound of ET-1 knockout mice. In endothelial ET-1 knockout mouse, the expression of TNF-α, CTGF and TGF-β was down-regulated. Bosentan, an antagonist of dual ET receptors, is known to attenuate skin fibrosis and accelerate wound healing in systemic sclerosis, and such contradictory effect may be mediated by above molecules. The endothelial cell-derived ET-1 is the potent therapeutic target in fibrosis or wound healing, and investigations of the overall regulatory mechanisms of these pathological conditions by ET-1 may lead to a new therapeutic approach.

Figure 1. Comparison of the ET-1 expression between wild-type and ET-1^f/f; Tie-2-Cre (+) mice skins.

(A) Gross comparison of a wild-type (WT) and an ET-1^f/f; Tie-2-Cre (+) (KO) mouse at 8 weeks. (B) Representative haematoxylin and eosin (HE) staining of skin section from a wild-type (WT) and an ET-1^f/f; Tie-2-Cre (+) (KO) mouse. (C) Representative ET-1 staining of subcutaneous blood vessels in a wild-type (WT) and an ET-1^f/f; Tie-2-Cre (+) (KO) mouse. Scale bar = 20 µm. (D) Dermal vessels of wild-type (WT) and ET-1^f/f; Tie-2-Cre (+) (KO) mice were stained with antibodies against isolectin IB4 (green) and ET-1 (red). Scale bar = 10 µm.

Figure 2. Bleomycin-induced skin fibrosis in wild-type and ET-1^f/f; Tie-2-Cre (+) mice.

(A) The protocol for Figure 2B and 2C is shown. Bleomycin or PBS was locally injected in the back of the wild-type (WT) and ET-1^f/f; Tie-2-Cre (+) (KO) mice daily for 4 weeks. The back skin was obtained on day 29. (B) Hematoxylin and eosin (HE) staining of PBS-treated mice skin. WT; wild-type, KO; ET-1^f/f; Tie-2-Cre (+). Scale bar = 100 µm. (C) HE (upper panels) and Masson's trichrome staining (lower panels) of bleomycin-treated mice skin. WT; wild-type, KO; ET-1^f/f; Tie-2-Cre (+). Scale bar = 100 µm. (D) Dermal thickness of the wild-type (WT) and ET-1^f/f; Tie-2-Cre (+) (KO) mice was evaluated by measuring the distance between the epidermal-dermal junction and the dermal-fat junction in HE sections under 100-folds magnification. Data are shown on the ordinate (n = 6). Bars show means. *P<0.05. (E) Relative collagen contents of paraffin-embedded sections from the wild-type (WT) and ET-1^f/f; Tie-2-Cre (+) (KO) mice were determined as described in “Materials and Methods” (n = 6). *P<0.05.

Figure 3. Infiltrating cells in the bleomycin-treated skin of the wild-type (WT) and ET-1^f/f; Tie-2-Cre (+) (KO) mice.

Myeloperoxidase (A), F4/80 (B) and CD3 (C) were stained. Positive cells were counted in five random high-power fields (0.06 mm², magnification, ×400). Data were expressed as the mean ± SD of six independent counts (left panel). The representative results of immunostaining for myeloperoxidase, F4/80 and CD3 are shown (right panel).

Figure 4. Cytokine expression in the bleomycin-treated skin from the wild-type (WT) and ET-1^f/f; Tie-2-Cre (+) (KO) mice.

Total RNA was extracted from the skin, and the mRNA expression levels of indicated cytokines were determined by real-time PCR. Data are expressed as the mean ± SD of six independent experiments. *P<0.05 as compared with the value in WT mice (1.0).

Figure 5. Cutaneous wound healing in WT and ET-1^f/f; Tie-2-Cre (+) mice.

(A) Representative wound closure in wild-type (WT) and ET-1^f/f; Tie-2-Cre (+) (KO) mice at days 0, 5, 7, 10 post-wounding. (B) Hematoxylin-Eosin (HE) staining of wound tissues derived from wild-type (WT) and ET-1^f/f; Tie-2-Cre (+) (KO) mice at days 3, 7, 12 post-wounding. Arrow heads indicated bilateral edges of wound granulation tissue. Double-headed arrows indicated the distance between the leading edges of wounded epidermis. The one representative result is shown. Scale bar = 1000 µm. (C) Measurements of granulation tissue size (the distance between the arrow heads) in the wild-type (WT) and ET-1^f/f; Tie-2-Cre (+) (KO) mice at days 3, 7, 12 post-wounding. Data are expressed as the mean ± SD of six independent experiments. *P<0.05 as compared with the value in WT mice. (D) Measurements of epithelial gap (the distance of double-headed arrows) in the wild-type (WT) and ET-1^f/f; Tie-2-Cre (+) (KO) mice at days 3, 7, 12 post-wounding. Data are expressed as the mean ± SD of six independent experiments. *P<0.05 as compared with the value in WT mice.

Figure 6. Infiltrating cells in the wounding bed at day 3, 7 and 12 of the wild-type (WT) and ET-1^f/f; Tie-2-Cre (+) (KO) mice.

Myeloperoxidase (A), F4/80 (B) and CD3 (C) were stained. Positive cells were counted in five random high-power fields (0.06 mm², magnification, ×400). Data were expressed as the mean ± SD of six independent counts (left panel). The representative results of immunostaining for myeloperoxidase, F4/80 and CD3 are shown (right panel).

Figure 7. Cytokine expression in the wounding bed at day 3, 7 and 12 of the wild-type (WT) and ET-1^f/f; Tie-2-Cre (+) (KO) mice.

Total RNA was extracted from the skin, and the mRNA expression levels of indicated cytokines were determined by real-time PCR. Data are expressed as the mean ± SD of six independent experiments. *P<0.05 as compared with the value in WT mice (1.0).

Figure 8. Schematic model of attenuated bleomycin-induced skin fibrosis and accelerated wound healing in endothelial cell-specific endothelin 1 knockout mice.