Involvement of PAR-2 in the Induction of Cell-Specific Matrix Metalloproteinase-2 by Activated Protein C in Cutaneous Wound Healing

Abstract

We previously reported that human keratinocytes express protease-activated receptor (PAR)-2 and play an important role in activated protein C (APC)-induced cutaneous wound healing. This study investigated the involvement of PAR-2 in the production of gelatinolytic matrix metalloproteinases (MMP)-2 and -9 by APC during cutaneous wound healing. Full-thickness excisional wounds were made on the dorsum of male C57BL/6 mice. Wounds were treated with APC on days 1, 2, and 3 post-wounding. Cultured neonatal foreskin keratinocytes were treated with APC with or without intact PAR-2 signalling to examine the effects on MMP-2 and MMP-9 production. Murine dermal fibroblasts from PAR-2 knock-out (KO) mice were also assessed. MMP-2 and -9 were measured via gelatin zymography, fluorometric assay, and immunohistochemistry. APC accelerated wound healing in WT mice, but had a negligible effect in PAR-2 KO mice. APC-stimulated murine cutaneous wound healing was associated with the differential and temporal production of MMP-2 and MMP-9, with the latter peaking on day 1 and the former on day 6. Inhibition of PAR-2 in human keratinocytes reduced APC-induced MMP-2 activity by 25~50%, but had little effect on MMP-9. Similarly, APC-induced MMP-2 activation was reduced by 40% in cultured dermal fibroblasts derived from PAR-2 KO mice. This study shows for the first time that PAR-2 is essential for APC-induced MMP-2 production. Considering the important role of MMP-2 in wound healing, this work helps explain the underlying mechanisms of action of APC to promote wound healing through PAR-2.

Keywords: activated protein C; cutaneous wound healing; matrix metalloproteinase-2, -9; murine; protease activator receptor-2.

Figure 1

APC accelerates wound healing in C57/Bl6 mice. Full-thickness 6 mm diameter wounds were made and treated topically with 20 μL of PBS or APC (10 μg) once a day for 3 consecutive days. Haematoxylin and Eosin (H&E)-stained paraffin sections from day 1, 3, and 6 post-wounding. Sections represent 6 separate experiments in each individual group. Arrows denote the leading edge of the migrating epithelial tongue. e, eschar; he, hyper-proliferative epithelium. Scale bar = 1 mm. Abbreviations: APC, activated protein C; PBS, phosphate-buffered saline.

Figure 2

Temporal profile of MMP-2 and MMP-9 in healing wounds. Tissue homogenates were obtained from PBS- and APC-treated wounds on day 1, day 3, day 6, and day 13. (A) MMP-2 and -9 production was assessed using gelatin zymography. (B) Band densitometry was performed using ImageJ software V 1.8.0. Bars represent mean ± SD, n = 3. * p < 0.05, ** p < 0.01 vs. PBS-treated wounds by two-way ANOVA with Tukey’s post-hoc test. Abbreviations: APC, activated protein C; MMP, matrix metalloproteinase, PBS, phosphate-buffered saline.

Figure 3

Immunohistochemical analysis of MMP-2 and MMP-9 in healing wounds. (A) Representative MMP-9 staining of PBS- and APC-treated wounds on day 1, day 3, and day 6 showed strong epithelial localization in APC-treated wounds on day 3 and on day 6 in PBS-treated wounds. (B) Representative MMP-2 staining of PBS- and APC-treated wounds on day 1, day 3, and day 6 showed an increased MMP-2 expression in both the epidermis and dermis in APC-treated wounds on day 6. (C) Negative control staining—rabbit IgG at the same dilution of MMP-2 and MMP-9. Scale bar = 1000 µm. Abbreviations: APC, activated protein C; MMP, matrix metalloproteinase, PBS, phosphate-buffered saline.

Figure 4

PAR-2 mediates APC-induced MMP-2 production by primary human keratinocytes. Cell supernatant was obtained from cultured primary human keratinocytes and MMP-2 and -9 production was assessed using gelatin zymography. (A) A dose-dependent increase in MMP-2 activation by 0–5 μg/mL of APC over 48 h. (B) Effect of 10 μg/mL of the PAR2-blocking antibody, SAM11, on APC-induced MMP-2 activation. (C) PAR-2 siRNA-transfected keratinocytes were treated with 1 μg/mL of APC for 48 h. MMP-2 within the cell supernatant was activated by the addition of 1 mM amino-phenyl mercuric acetate (APMA). The fluorescent intensity was measured at indicated time points. Bars represent mean ± SD, n = 4; * p < 0.05, by one-way ANOVA. Abbreviations: APC, activated protein C; MMP, matrix metalloproteinase, PAR, protease-activated receptor; RFU, relative fluorescent units; siRNA, small interfering ribonucleic acid.

Figure 5

PAR-2 mediates APC-induced MMP-2 activation by primary murine dermal fibroblasts. Cell supernatant was obtained from cultured PAR-2 KO- and WT-mice dermal fibroblasts and treated with 0–10 μg/mL of APC for 48 h. (A) Representative gel of gelatin zymography, showing MMP-2 and -9 production. (B) Densitometric analysis of pro-MMP2. (C) The proportion (%) of active MMP-2 was analysed via band densitometry using ImageJ software. Bars represent mean ± SD, n = 4. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001 by 2-way ANOVA with Tukey’s post-hoc test. Abbreviations: APC, activated protein C; a.u., arbitrary unit; MMP, matrix metalloproteinase, PAR, protease-activated receptor.

Figure 6

Proposed mechanism of PAR-2-induced gelatinolytic MMP activation in APC-stimulated wound healing. Based on data presented in this study, our previous studies, and other studies, we propose that APC-mediated PAR2 activation causes intracellular activation of AKT via PI3K, which phosphorylates IĸBα, leading to its proteasomal degradation, leaving NF-ĸB free to translocate into the nucleus to promote MMP-2 and MMP-9 expression. Released pro-active MMP-2 and MMP-9 are cleaved by APC through unknown mechanisms and promote cutaneous wound healing. ? indicates unknown mechanism (s).