Receptor-Interacting Protein Kinase 3 Deficiency Delays Cutaneous Wound Healing

Abstract

Wound healing consists of a complex, dynamic and overlapping process involving inflammation, proliferation and tissue remodeling. A better understanding of wound healing process at the molecular level is needed for the development of novel therapeutic strategies. Receptor-interacting protein kinase 3 (RIPK3) controls programmed necrosis in response to TNF-α during inflammation and has been shown to be highly induced during cutaneous wound repair. However, its role in wound healing remains to be demonstrated. To study this, we created dorsal cutaneous wounds on male wild-type (WT) and RIPK3-deficient (Ripk3-/-) mice. Wound area was measured daily until day 14 post-wound and skin tissues were collected from wound sites at various days for analysis. The wound healing rate in Ripk3-/- mice was slower than the WT mice over the 14-day course; especially, at day 7, the wound size in Ripk3-/- mice was 53% larger than that of WT mice. H&E and Masson-Trichrome staining analysis showed impaired quality of wound closure in Ripk3-/- wounds with delayed re-epithelialization and angiogenesis and defected granulation tissue formation and collagen deposition compared to WT. The neutrophil infiltration pattern was altered in Ripk3-/- wounds with less neutrophils at day 1 and more neutrophils at day 3. This altered pattern was also reflected in the differential expression of IL-6, KC, IL-1β and TNF-α between WT and Ripk3-/- wounds. MMP-9 protein expression was decreased with increased Timp-1 mRNA in the Ripk3-/- wounds compared to WT. The microvascular density along with the intensity and timing of induction of proangiogenic growth factors VEGF and TGF-β1 were also decreased or delayed in the Ripk3-/- wounds. Furthermore, mouse embryonic fibroblasts (MEFs) from Ripk3-/- mice migrated less towards chemoattractants TGF-β1 and PDGF than MEFs from WT mice. These results clearly demonstrate that RIPK3 is an essential molecule to maintain the temporal manner of the normal progression of wound closure.

Fig 1. Ripk3 ^-/- mice show delayed wound closure.

WT (n = 7) and Ripk3 ^-/- mice (n = 7) were subjected to dorsal cutaneous wounds as described in Materials and Methods for a 14-day wound healing study. Wound area percentage was determined based on measuring wound area daily using NIHImage J analysis as described in Materials and Methods. Data are represented as mean ± SEM and compared by one-way ANOVA using SNK test. *P < 0.05 versus WT at the indicated day.

Fig 2. Ripk3 ^-/- mice show impaired quality of wound closure.

Skin samples from the wound sites of WT and Ripk3 ^-/- mice were collected at days 7 and 14 for histological evaluation by H&E staining (n = 5 mice per group). Representative images of H&E stained skin sections (40×magnification) from day 7 (A, B) and day 14 (C, D) wounds from WT (A, C) and Ripk3 ^-/- (B, D) mice are shown. Angiogenic sites are depicted by arrows; ep, new epithelium; wb, wound bed. Scale bar = 500 μm.

Fig 3. Ripk3 ^-/- mice show impaired collagen deposition.

Skin samples from the wound sites of WT and Ripk3 ^-/- mice were collected at the indicated days for evaluation of collagen deposition by Masson-Trichrome staining and qPCR (n = 5 mice per group). Representative images of Masson-Trichrome stained skin sections (40× magnification) from day 7 (A, B) and day 14 (C, D) wounds from WT (A, C) and Ripk3 ^-/- (B, D) mice are shown. Collagen is stained blue. Angiogenic sites are depicted by arrows; ep, new epithelium; wb, wound bed. Scale bar = 500 μm. (E) The expression analysis of collagen type 1 α 1 (Col1a1) mRNA by qPCR. Data expressed as mean ± SEM (n = 5 mice per group) and compared by one-way ANOVA using SNK test. *P < 0.05 versus WT at the indicated day.

Fig 4. Ripk3 ^-/- mice show altered early neutrophil infiltration and inflammation.

(A-E) Skin samples from the wound sites of WT and Ripk3 ^-/- mice were collected at days 1 and 3 for immunohistochemical staining with Gr-1 antibody. Representative images of Gr-1 stained skin sections (200× magnification) from day 1 (A, B) and day 3 (C, D) wounds from WT (A, C) and Ripk3 ^-/- (B, D) mice and quantitative analysis of neutrophil infiltration expressed as number of neutrophils/field (E) are shown. Neutrophils are stained brown. Scale bar = 100 μm. (F-I) Skin samples from the wound sites of WT and Ripk3 ^-/- mice were collected at the indicated days. RNA was extracted and cDNA was prepared from these samples for gene expression analysis of IL-6 (F), KC (G), IL-1β (H) and TNF-α (I) by qPCR. Data expressed as mean ± SEM (n = 3 (E) and 5 (F-I) mice per group for each time point) and compared by one-way ANOVA using SNK test. *P < 0.05 versus WT at the indicated day.

Fig 5. Ripk3 ^-/- mice show reduced MMP-9 levels.

Skin samples from the wound sites of WT and Ripk3 ^-/- mice were collected at days 1, 7 and 14 for immunohistochemical staining with MMP-9 antibody (n = 5 mice per group). Representative images of MMP-9 stained skin sections (100× magnification) from day 1 (A, B), day 7 (C, D) and day 14 (E, F) wounds from WT (A, C, E) and Ripk3 ^-/- (B, D, F) mice are shown. MMP-9 is stained brown; ep, new epithelium; wb, wound bed. Scale bar = 100 μm. (G-H) RNA was extracted and cDNA was prepared from wound skin samples collected at the indicated days. MMP-9 (G) and Timp-1 (H) gene expression was analyzed by qPCR. Data expressed as mean ± SEM (n = 5 mice per group) and compared by one-way ANOVA using SNK test. *P < 0.05 versus WT at the indicated day.

Fig 6. Ripk3 ^-/- mice show delayed angiogenesis and altered pattern of growth factor induction.

Skin samples from the wound sites of WT and Ripk3 ^-/- mice were collected at the indicated days. (A-D) Skin samples were sectioned and immunostained with antibody against the endothelial cell junction molecule CD31. Representative images of CD31 stained skin sections (100× magnification) from day 7 (A, B), and day 14 (C, D) wounds from WT (A, C) and Ripk3 ^-/- (B, D) mice are shown. CD31 is stained brown. Blood vessels were identified by CD31 positive endothelial cells lining the lumen and are depicted by arrows; ep, new epithelium; wb, wound bed. Scale bar = 100 μm. (E) Statistical evaluation of the density of CD31-positive blood vessels (expressed as number of vessels/field) in wound sections at day 7 and 14 post-wound. (F-G) Total RNA extracted from skin tissues were analyzed for mRNA expression of VEGF (F) and TGF-β1 (G) by qPCR. Data expressed as mean ± SEM (n = 3 (E) and 5 (F-G) mice per group for each time point) and compared by one-way ANOVA using SNK test. *P < 0.05 versus WT at the indicated day.

Fig 7. Ripk3 ^-/- mice show reduced chemotaxis of fibroblasts in response to growth factors.

Mouse embryonic fibroblasts (MEFs) were prepared from WT and Ripk3 ^-/- mice and subjected to transwell migration assay as described in Materials and Methods with chemoattractants PDGF (A), and TGF-β1 (B). (C) The numbers of MEFs from WT and Ripk3 ^-/- mice migrated toward the indicated chemoattractant. The experiment was repeated three times and data expressed as mean ± SEM (n = 3–4 per group per experiment) and compared by Student’s t-test. *P < 0.05 versus WT at the indicated day.