New susceptibility loci for cutaneous melanoma risk and progression revealed using a porcine model

Emmanuelle Bourneuf,Hélène Gilbert,Silvia Vincent-Naulleau,Amandine Blin,Maria Del Pilar Schneider,Florence Demenais,Amaury Vaysse,Jordi Estellé,Myriam Brossard,Mark Lathrop,Françoise Créchet

doi:10.18632/oncotarget.25455

Abstract

Despite major advances, it is estimated that a large part of melanoma predisposing genes remains to be discovered. Animal models of spontaneous diseases are valuable tools and experimental crosses can be used to identify and fine-map new susceptibility loci associated with melanoma. We performed a Genome-Wide Association Study (GWAS) of melanoma occurrence and progression (clinical ulceration and presence of metastasis) in a porcine model of spontaneous melanoma, the MeLiM pig. Five loci on chromosomes 2, 5, 7, 8 and 16 showed genome-wide significant associations (p < 5 × 10–6) with either one of these phenotypes. Suggestive associations (p < 5 × 10–5) were also found at 16 additional loci. Moreover, comparison of the porcine results to those reported by human melanoma GWAS indicated shared association signals notably at CDKAL1 and TERT loci but also nearby CCND1, FTO, PLA2G6 and TMEM38B-RAD23B loci. Extensive search of the literature revealed a potential key role of genes at the identified porcine loci in tumor invasion (DST, PLEKHA5, CBY1, LIMK2 and ETV5) and immune response modulation (ETV5, HERC3 and DICER1) of the progression phenotypes. These biological processes are consistent with the clinico-pathological features of MeLiM tumors and can open new routes for future melanoma research in humans.

Highlights

Cutaneous melanoma is the most deadly skin cancer and has shown a growing incidence worldwide in the last decades [1]
The associated regions obtained with the mixed model approach are summarized in Tables 1–3 for melanoma occurrence, clinical tumor ulceration and metastasis, respectively
CDK4 alleles did not segregate with melanoma in a reference population [19], and lastly an association analysis showed that MITF could not be related to melanoma occurrence in the Melanoblastoma-bearing Libechov Minipig (MeLiM) model [23]

Summary

Introduction

Cutaneous melanoma is the most deadly skin cancer and has shown a growing incidence worldwide in the last decades [1]. Two high-penetrance genes, CDKN2A and CDK4, both regulators of the cell cycle, have been first evidenced in around 20% of the families predisposed to melanoma [4,5,6,7]. Rare mutations in other genes, including POT1, TERF2IP and ACD, have been recently identified by whole-exome sequencing in melanoma families [8,9,10]. Low frequency variants conferring moderate risk of melanoma were identified in two genes, MC1R and MITF, which play a key role in melanocyte biology and pigment synthesis regulation [13, 14]. Genome-wide association studies have highlighted common genetic variants associated with melanoma risk, at loci containing genes involved in pigmentation and naevus count (e.g. TYR, SLC45A2, ASIP, PLA2G6), and DNA repair genes (PARP1, ATM see [15, 16] for reviews)

Methods

Results

Discussion

Conclusion