ARTICLES Epigenetic Regulation Identifies RASEF as a Tumor-Suppressor Gene in Uveal Melanoma Willem Maat,1 Sigrid H. W. Beiboer,2 Martine J. Jager,1 Gre´ P. M. Luyten,1 Nelleke A. Gruis,3 and Pieter A. van der Velden1,3 5 PURPOSE. Recently, a segregation study in families with uveal Recently, Jo¨nsson et al. revealed a genetic component in three and cutaneous melanoma identified 9q21 as a potential locus such families, in which members are affected by either uveal or harboring a tumor-suppressor gene (TSG). One of the genes in cutaneous melanoma. Linkage analysis in these families identi- this area, RASEF, was then analyzed as a candidate TSG, but fied a potential uveal melanoma susceptibility locus on chro- lack of point mutations and copy number changes could not mosome 9, area q21. confirm this. In this study, the RASEF gene was investigated for This locus has a long history in melanoma that started with potential mutations and gene silencing by promoter methyl- detection of isochromosome 9q with cytogenetic analysis.6,7 ation in uveal melanoma. Loss of heterozygosity (LOH) of markers at 9q22 was subse- ETHODS quently frequently reported and was shown to be associated M . Eleven uveal melanoma cell lines and 35 primary 8,9 uveal melanoma samples were screened for mutations in the with proliferation and tumor progression. Recently, single RASEF gene by high-resolution melting-curve and digestion nucleotide polymorphism (SNP) analysis has confirmed the analysis. Expression of RASEF was determined by real-time LOH of this locus in melanoma, while genome-wide analysis in dizygotic twins for nevi numbers also showed linkage with this RT-PCR in all cell lines and 16 primary uveal melanoma sam- 10,11 ples, and the methylation status of the promoter of the RASEF 9q region. In addition, a gene slightly distal to RASEF, gene was analyzed and confirmed by direct sequencing. RMI1, has recently been shown to be a risk factor for cutane- ous melanoma, whereas the locus for familial melanoma sus- RESULTS. Mutation screening revealed a known polymorphism 12,13 3 ceptibility is located on the short arm of chromosome 9. (R262C; C T) in exon 5 of the RASEF gene that displayed a Cutaneous melanomas are often characterized by loss of the normal frequency (54%). Of the primary uveal melanomas, 46% cell-cycle regulator p16 and/or activation of the RAS/RAF/ERK presented a heterozygous genotype, and 10 (91%) of 11 cell pathway.14,15 These hallmarks of melanoma are also recog- lines showed a homozygous genotype. Melting-curve analysis nized in uveal melanoma, although the underlying mechanisms indicated loss of heterozygosity in at least two primary tumors. differ.16,17 Whereas in cutaneous melanoma, p16 is commonly Low RASEF expression in the cell lines and primary tumors lost by chromosomal deletion of the CDKN2A gene, the pref- correlated with methylation of the RASEF promoter region. erential mechanism in uveal melanoma appears to be silencing Homozygosity and methylation of the RASEF gene in primary 18 ϭ of the p16-encoding CDKN2A promoter by methylation. Mu- tumors were associated with decreased survival (P 0.019). tations in BRAF, NRAS,orc-kit lead to constitutive ERK acti- CONCLUSIONS. Homozygosity, in combination with methylation, vation in most cutaneous melanomas.19,20 However, mutations appears to be the mechanism targeting RASEF in uveal mela- in BRAF have only rarely been reported in uveal melanoma, noma, and allelic imbalance at this locus supports a TSG role whereas activating NRAS and c-kit mutations have never been for RASEF.(Invest Ophthalmol Vis Sci. 2008;49:1291–1298) reported.21 Still, ERK activation is also present in uveal mela- DOI:10.1167/iovs.07-1135 noma, and this knowledge leads to the question of what causes ERK activation in the absence of activating mutations in BRAF, veal melanoma is the most common primary intraocular NRAS,orc-kit.16,17,21 Uneoplasm in adults, with an annual incidence of six to The RASEF (RAS and EF hand domain containing) gene is eight per million in Caucasian populations.1 In contrast to located on chromosome 9, area q21, and encodes a protein cutaneous melanoma, clustering of uveal melanoma in families with calcium-binding EF-hand and Ras GTPase (Rab family) is extremely rare.2–4 Occurrence of both uveal melanoma and motifs (http://www.genome.ucsc.edu/ provided in the public cutaneous melanoma in a single family has been observed.4 domain by the Genome Bioinformatics Group, University of Santa Cruz, CA); it is also known as RAB45 or FLJ31614.22 Based on the functional domains in RASEF, the gene product may be engaged in the RAS pathway and in combination with 1 3 From the Departments of Ophthalmology and Dermatology, evidence for linkage of the RASEF region with cutaneous and Leiden University Medical Center (LUMC), Leiden, The Netherlands; uveal melanoma, molecular analysis of this gene is warranted. and 2Hogeschool Leiden, Leiden, The Netherlands. In line with the analysis of cutaneous melanoma reported Supported by Dutch Cancer Society (KWF) Grant RUL 2001-2472. 5 Submitted for publication August 31, 2007; revised November 9, by Jo¨nsson et al., we therefore set out to analyze RASEF for 2007; accepted February 18, 2008. mutations and for expression of the gene in uveal melanoma. Disclosure: W. Maat, None; S.H.W. Beiboer, None; M.J. Jager, None; G.P.M. Luyten, None; N.A. Gruis, None; P.A. van der Velden, None MATERIALS AND METHODS The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertise- Cell Lines and Primary Uveal ment” in accordance with 18 U.S.C. §1734 solely to indicate this fact. Melanoma Specimens Corresponding author: Pieter A. van der Velden, Skin Research Lab, Department of Dermatology, Leiden University Medical Center, In total, 11 cell lines derived from primary uveal melanomas (92.1; PO Box 9600, 2300 RC Leiden, The Netherlands; [email protected]. OCM-1, -3, and -8; and Mel-202, -270, -285, and -290) and uveal mela- Investigative Ophthalmology & Visual Science, April 2008, Vol. 49, No. 4 Copyright © Association for Research in Vision and Ophthalmology 1291 Downloaded from iovs.arvojournals.org on 10/02/2021 1292 Maat et al. IOVS, April 2008, Vol. 49, No. 4 TABLE 1. Tumor Characteristics and Survival Data of 35 Uveal Melanoma Patients Sorted by Methylation Status and RASEF Genotype Survival Hypermethylated Tumor ID Cell Type (mo) Present Status RASEF Genotype UM1 Spindle 210 Alive Not present Hom C UM16 Epithelioid 115 Alive Not present Hom C UM29 Spindle 24 Died, due to metastases Not present Hom C UM3 Mixed 95 Died, due to metastases Not present Hom C UM33 Mixed 24 Died, due to metastases Not present Hom C UM12 Epithelioid 137 Died, due to metastases Not present Hom T UM19 Spindle 15 Died, due to metastases Not present Hom T UM26 Spindle 122 Alive Not present Hom T UM31 Mixed 34 Died, due to metastases Not present Hom T UM32 Epithelioid 63 Died, due to metastases Not present Hom T UM4 Spindle 31 Died, due to metastases Not present Hom T UM6 Epithelioid 57 Died, due to metastases Not present Hom T UM27 Mixed 23 Died, due to metastases Present Hom C UM28 Epithelioid 33 Died, due to metastases Present Hom C UM30 Spindle 113 Alive Present Hom C UM11 Mixed 13 Died, due to metastases Present Hom T UM18 Spindle 12 Died, due to metastases Present Hom T UM5 Spindle 50 Died, due to metastases Present Hom T UM35 Mixed 94 Alive Present Het/homT* UM21 Epithelioid 167 Died, due to metastases Present Het/homT* UM13 Epithelioid 30 Died, due to metastases Present Het/homT* UM15 Mixed 23 Died, due to metastases Present Het UM17 Epithelioid 33 Died, due to metastases Present Het UM2 Mixed 29 Died, due to metastases Not present Het UM14 Mixed 42 Died, due to metastases Not present Het UM25 Mixed 29 Died, other cause Not present Het UM7 Mixed 63 Died, other cause Not present Het UM23 Spindle 152 Died, unknown cause Not present Het UM22 Spindle 2 Lost to follow up Not present Het UM8 Spindle 191 Alive Not present Het UM9 Mixed 131 Alive Not present Het UM10 Mixed 136 Alive Not present Het UM20 Mixed 187 Alive Not present Het UM24 Mixed 143 Alive Not present Het UM34 Mixed 106 Alive Not present Het * Loss of heterozygosity/allelic imbalance. noma metastases (OMM-1, -2.3, and -2.5) were analyzed. All melanoma Zymo Research Corp., Orange, CA). Enzymatically methylated human cell lines were cultured in RPMI 1640 medium (Invitrogen-Gibco, DNA (Chemicon Europe Ltd., Hampshire, UK) was used as the positive Paisley, Scotland, UK) supplemented with 3 mM L-glutamine (Invitro- control in all experiments. DNA and RNA concentrations were deter- gen-Gibco), 2% penicillin-streptomycin, and 10% FBS (Hyclone, Logan, mined by spectrophotometer (model ND-1000; NanoDrop Technolo- UT). All cell cultures were incubated at 37°C in a humidified 5% CO2 gies Inc., Wilmington, DE). atmosphere. Archival frozen tumor specimens of primary uveal mela- noma came from 35 patients who attended the Leiden University Mutation Screening and Genotyping Medical Center between 1988 and 1996. All tumors were primary A 96-well light scanner (Idaho Technologies Inc., Salt Lake City, UT) for lesions with a tumor diameter greater than 12 mm, a prominence high-resolution melting-curve analysis was used to scan all amplicons greater than 6 mm, and no treatment before enucleation. The validity of the RASEF gene. The primers are shown in Table 2. DNA samples of the diagnosis of uveal melanoma was confirmed histologically in all were amplified with a double-stranded DNA-binding dye (LC Green cases, and clinical and survival data were listed for use in the study Plus; Idaho Technologies).