Endothelin-1 is a transcriptional target of p53 in epidermal keratinocytes and regulates ultraviolet-induced melanocyte homeostasis

Abstract

Keratinocytes contribute to melanocyte activity by influencing their microenvironment, in part, through secretion of paracrine factors. Here, we discovered that p53 directly regulates Edn1 expression in epidermal keratinocytes and controls UV-induced melanocyte homeostasis. Selective ablation of endothelin-1 (EDN1) in murine epidermis (EDN1(ep-/-) ) does not alter melanocyte homeostasis in newborn skin but decreases dermal melanocytes in adult skin. Results showed that keratinocytic EDN1 in a non-cell autonomous manner controls melanocyte proliferation, migration, DNA damage, and apoptosis after ultraviolet B (UVB) irradiation. Expression of other keratinocyte-derived paracrine factors did not compensate for the loss of EDN1. Topical treatment with EDN1 receptor (EDNRB) antagonist BQ788 abrogated UV-induced melanocyte activation and recapitulated the phenotype seen in EDN1(ep-/-) mice. Altogether, the present studies establish an essential role of EDN1 in epidermal keratinocytes to mediate UV-induced melanocyte homeostasis in vivo.

Figure 1

Positive regulation of Edn1 expression after UV exposure by p53 in murine keratinocytes. (A) Schematic of predicted p53 binding locations on murine and human Edn1/EDN1 promoters. Arrows indicate primers designed for chromatin immunoprecipitation (ChIP). (B) ChIP assay on primary murine keratinocytes using anti-p53 antibody following presence or absence of UV exposure. Results were analyzed by qPCR using primers specific to proximal, mid and distal regions (indicated in A). Primers directed against the 3’ UTR region of EDN1 and non-specific IgG antibody were used as negative controls. (C) Relative gene expression of Edn1 in epidermis from adult wildtype C57BL/6 and p53^−/− mice at designated time points post-UV exposure. All experiments were done using a minimum of three biological replicates from each group and in all cases are expressed as mean +/− SEM. Statistical analysis was performed using Graphpad Prism, * = p < 0.05.

Figure 2

Characterization of EDN1^ep−/− mice showed reduced epidermal and dermal melanocytes in adult skin. (A) Schematic diagram of Cre-targeted loxP sites flanking exon 2 of Edn1. Genotyping results using primers directed against both Cre and Edn1 from epidermal genomic DNA. (B) qPCR analysis of mRNA expression levels for Edn1 and Edn3 in untreated P2 mouse whole skin are shown. (C,D) Fontana-Masson (black) and TRP1 (green) staining of untreated P2 and P42 mouse dorsal skin sections, arrows indicate dermal melanocytes; scale bar=20µm. IHC sections counterstained with DAPI (blue) and white dashed lines indicate epidermal-dermal junction. (E) Bar graph comparing melanocytes per field between untreated P2 and P42 EDN1^L2/L2 and EDN1^ep−/− skin. All experiments were done using a minimum of three biological replicates from each group and in all cases are expressed as mean +/− SEM. Statistical analysis was performed by Student’s t-test using Graphpad Prism, ** = p < 0.01.

Figure 3

Decreased epidermal and dermal melanocytes in EDN1^ep−/− and BQ788-treated wildtype neonatal skin post-UVR. (A) Relative gene expression of Edn1 and Edn3 over time following UVR of EDN1^L2/L2 and EDN1^ep−/− P2 skin. (B) Fontana Masson (FM) stained images at 72 and 96hr post-UVR of P2 skin, arrows indicate melanocytes. Yellow dashed line represents epidermal-dermal junction, scale bar=20µm. (C) Epidermal melanocyte counts per field in EDN1^L2/L2 and EDN1^ep−/− skin over time post- UVR of P2 skin. (D) Dermal melanocyte counts per field in EDN1^L2/L2 and EDN1^ep−/− skin over time post- UVR of P2 skin. (E) Epidermal melanocyte counts per field in wildtype B6 mice treated with topical BQ788 or vehicle at 72 and 96hrs post-UVR of P3 skin for both FM and IHC (TRP1+). All experiments were done using a minimum of three biological replicates from each group of mice and in all cases are expressed as mean +/− SEM. Statistical analysis was performed using Graphpad Prism, * = p < 0.05, ** = p < 0.01, *** = p < 0.001.

Figure 4

Immunohistochemical characterization for proliferation, DNA damage and apoptosis of melanocytes post-UVR. (A) IHC analysis of EDN1^L2/L2 and EDN1^ep−/− skin at 72 or 24hrs post-UVR of P2 mice stained with anti-PCNA (red), anti-CPD (red), anti-TRP1 (green and red) primary antibodies and TUNEL assay (green). Yellow scale bar=62µm, white scale bar=20µm. E=epidermis, D=dermis, HF=hair follicle. PCNA and CPD sections counterstained with DAPI (blue) White dashed line indicates dermalepidermal junction. (B) Percentage of PCNA, CPD or TUNEL-positive melanocytes out of total DAPIstained cells between EDN1^L2/L2 and EDN1^ep−/− skin post-UVR of P2 mice. All experiments were done using a minimum of three biological replicates from each group of mice and in all cases are expressed as mean +/− SEM. Statistical analysis was performed using Graphpad Prism, * = p < 0.05.

Figure 5

Decreased epidermal and dermal melanocytes and altered melanocyte proliferation and DNA damage in p53^−/− neonatal skin post-UVR. (A) Fontana Masson (FM) and IHC stained images of p53^+/+ and p53^−/− mouse skin at 72hr post-UV treatment of P3 mice, arrows indicate melanocytes. Anti-PCNA (red), anti-CPD (red), anti-TRP1 (green) primary antibodies were used, PCNA and CPD sections counterstained with DAPI (blue). White dashed line represents epidermal-dermal junction, E=epidermis, D=dermis, scale bar=62µm. (B) Epidermal melanocyte counts per field using both FM+ and TRP1+ cells 72 hours post-UVR of P3 skin in p53^+/+ and p53^−/− mice. (C) Percentage of PCNA or CPD positive melanocytes out of total DAPI-stained cells between p53^+/+ and p53^−/− neonatal mice 72 hours post-UVR of P3 mice. All experiments were done using a minimum of three biological replicates from each group of mice and in all cases are expressed as mean +/− SEM. Statistical analysis was performed using Graphpad Prism, ** = p < 0.01, *** = p < 0.001.

Figure 6

Activation of MAPK and PKC signaling by EDN1 is specific to EDNRB receptor. (A) Immunoblot analysis of EDNRB expression in lysates from primary murine keratinocytes [KC] and melanocytes [MC], β-actin levels were used as controls. (B) Immunoblot analysis for ERK phosphorylation after addition of exogenous EDN1 with or without the EDNRB antagonist BQ788. Total ERK and β-actin levels were used as controls, as well as minimal and complete mediums. (C) Activation of PKC after addition of exogenous EDN1 with or without the presence of EDNRB antagonist BQ788. All experiments were performed in triplicate and PKC activation results are expressed as mean +/− SEM. (D) Real-time transwell migration assay comparing slopes for exponential migration phase of wildtype melanocytes. (E) IHC analysis of ERK activation in EDN1^L2/L2 and EDN1^ep−/− skin at 0, 24, 48 and 72hrs post-UVR treatment of P2 mice with anti-pERK (red) and anti-TRP1 (green) primary antibodies. All sections are counterstained with DAPI (blue), white dashed line represents epidermal-dermal junction. E=epidermis, D=dermis, white scale bar=62 µm, arrows and white boxes indicate pERK activated melanocytes. (F) Percentage of pERK+ cells out of total DAPI-stained epidermal cells between EDN1^L2/L2 and EDN1^ep−/− mice post-UVR treatment of P2 mice. (G) Percentage of pERK+ melanocytes out of total TRP1+ stained cells in EDN1^L2/L2 and EDN1^ep−/− after UVR treatment of P2 mice. All experiments were done using a minimum of three biological replicates from each group of mice and in all cases are expressed as mean +/− SEM. Statistical analysis was performed using Graphpad Prism, ** = p < 0.01, *** = p < 0.001.