Association Between Endothelin Receptor B Nonsynonymous Variants and Melanoma Risk

Nadem Soufir 0 Roubila Meziani 0 Jean- Jacques Lacapre 0 Guylene Bertrand 0 Frederic Fumeron 0 Agnes Bourillon 0 Bndicte Grard 0 Vincent Descamps 0 Batrice Crickx 0 Laurence Ollivaud 0 Alain Archimbaud 0 Cleste Lebbe 0 Nicole Basset-Seguin 0 Philippe Saiag 0 Bernard Grandchamp for the Investigators of the Melan-Cohort 0 0 Journal of the National Cancer Institute , Vol. 97, No. 17, September 7, 2005 - The endothelin signaling pathway plays a crucial role in melanocyte differentiation and migration. In this study, we investigated whether germline mutations of endothelin receptor B (EDNRB), a gene involved in Hirschsprung disease (HSCR), could also predispose for malignant melanoma (MM). The coding region of EDNRB was sequenced in 137 MM patients and in 130 ethnically matched Caucasian control subjects. Six nonsynonymous EDNRB variants were found in 15 patients (11%), but only two were found in four control subjects (3%, odds ratio [OR] = 3.87, 95% confidence interval [CI] = 1.25 to 12; P = .012). Overall, 14 out of 15 MM patients carried EDNRB mutations reported in HSCR, some of which had previously been shown to lead to loss of function. In multivariable logistic regression analysis including skin type, eye and hair color, number of nevi, and dorsal lentigines (freckles), the association between EDNRB mutations and MM risk remained statistically significant (OR = 19.9, 95% CI = 1.34 to 296.2; P = .03). Our data strongly suggest that EDNRB is involved in predisposition for two different multigenic disorders, HSCR and melanoma. [J Natl Cancer Inst 2005;97:1297301] The incidence of malignant melanoma (MM) has doubled every 10 years in most Caucasian populations (1). Multiple phenotypic traits, including the number of melanocytic nevi, dysplastic nevi, freckling and sunburn tendencies, and severe sunburn episodes during childhood, have been shown to be MM risk factors (2). In addition to phenotypic traits, two highly penetrating melanomapredisposing genes, CDKN2A and CDK4, are involved in predisposition to familial cutaneous melanoma, which accounts for 510% of all melanoma cases (36). Furthermore, the inheritance of melanoma can also be polygenic, involving variant alleles in several different genes (7,8), including the melanocortin 1 receptor MC1R (911). The endothelin receptor B (EDNRB) is a G-protein-coupled, seven-transmembrane receptor that interacts with endothelins (EDN1, EDN2, and EDN3), multifunctional peptides involved in many cellular and physiologic processes (12). EDNRB signaling is required during embryogenesis for the migration of melanoblasts from the neural crest to the interfollicular epidermidis, retinal epithelium, and follicular hair (13) and plays a crucial role in regulating the number of progenitor melanocytes and melanocyte differentiation (14). Loss of EDNRB function is associated with several genetic diseases: Hirschsprung disease (HSCR, 600155) (15,16), Waardenburg syndrome 4 (WS4; OMIM 277580) (17), and Waardenburg-Shah syndrome (18). These diseases are characterized by major defects in the neural crest cell lineages. In addition to its role in melanocyte development, activation of the endothelin system also plays a crucial role in the skins response to ultraviolet irradiation. EDN1, a paracrine peptide that is secreted by keratinocytes in response to UVB radiation (1921), is a ligand for EDNRB and induces proliferation of mature melanocytes and stimulates melanogenesis, producing a photoprotective effect (22). We therefore tested the hypothesis that a germline defect of the EDNRB pathway could be associated with a genetic predisposition to melanoma. The coding sequence and intronic flanking regions of the EDNRB gene were entirely sequenced in this case control study. Melanoma patients (n = 137) and control subjects (n = 130) were prospectively recruited between 1999 and 2004 by the Dermatology Departments of the Bichat Claude-Bernard, the Percy, the Ambroise Par, and the Saint-Louis Hospitals in Paris. The study population consisted of patients aged 2080 years with histologically confirmed MM. Twenty-nine of the patients had at least one relative with melanoma, but none harbored a CDKN2A or a CDK4 germline mutation. The control group was extended from the one previously described (23) and was composed of individuals who had no personal or family history of skin cancer, were in the same age range, and were referred by the same departments and hospitals as the MM patients. The birthplaces of the parents and grandparents were recorded to ensure that all patients and control subjects were of Caucasian origin. The Hospital Medical Ethics Committee (CCPPRB) approved the study protocol. Informed consent was obtained from all patients and control subjects enrolled in the study. The clinical characteristics of the MM patients and control subjects are summarized in Table 1. The strongest risk factors identified for MM were a mole count of >50 (P<.001), the presence of an atypical mole syndrome (P<.001), fair skin color (P<.001), and having dorsal lentigines (P<.001). Other pigmentation characteristics (light-colored eyes and hair, as well as skin type I or II) were not associated with MM risk. In addition, the presence of MC1R functional Affiliations of authors: Laboratoire de Biochimie Hormonale et Gntique, Hpital Bichat-Claude Bernard, AP-HP, Facult de Mdecine Paris VII, Paris, France (NS, RM, GB, AB, BG); Inserm U 410 Facult de Mdecine Paris VII Bichat-Claude Bernard, Paris, France (J-JL); EA 3516, Universit Paris 7, Facult de Mdecine Paris VII BichatClaude Bernard, Paris, France (FF); Service de Dermatologie, Hpital Bichat-Claude Bernard, Paris, AP-HP, Facult de Mdecine ParisVII, Paris, France (VD, BC); Service de Dermatologie, Hpital Saint-Louis, AP-HP, Facult de Mdecine Paris VII, Paris, France (LO, AA, CL, NB-S); Service de Dermatologie, Hpital Ambroise Par, AP-HP, Facult de Mdecine Paris-Ile de France Ouest, Boulogne Billancourt, France (PS). Correspondence to: Nadem Soufir, MD, PhD, Laboratoire de Biochimie Hormonale et Gntique, IFR02; Hopital Bichat-Claude Bernard, 46 rue Henri Huchard, 75018 Paris, France (e-mail: ). See Notes following References. Melanoma patients (n = 137) Control subjects (n = 130) 53.4 (63) 46.6 (55) 61.2 (79) 38.8 (50) 70.5 (91) 29.5 (38) 65.6 (84) 34.4 (44) OR (95% CI) (Referent) 1.6 (0.95 to 2.6) (Referent) 5.8 (2.5 to 13.7) (Referent) 2.4 (1.4 to 4.1) (Referent) 2.5 (1.6 to 4.4) (Referent) 13.2 (5.9 to 29.4) (Referent) 7.3 (2.4 to 1.9) (Referent) 5.23 (2.96 to 9.22) (Referent) 2.94 (1.73 to 5.02) 8.88 (3.19 to 24.67) variants was strongly associated with MM risk (for the presence of one variant, odds ratio [OR] = 2.94, 95% confidence interval [CI] = 1.73 to 5.02, and for two variants, OR = 8.88, 95% CI = 3.19 to 24.67). A higher frequency of nonsynonymous EDNRB variants was observed in the MM patients than in the control subjects. Six nonsynonymous EDNRB variants were characterized in 15 of t (...truncated)