A melanoma-associated germline mutation in exon 1b ... - Nature

Oncogene (2001) 20, 5543 ± 5547 ã 2001 Nature Publishing Group All rights reserved 0950 ± 9232/01 $15.00 www.nature.com/onc

A melanoma-associated germline mutation in exon 1b inactivates p14ARF Helen Rizos*,1,3, Susana Puig*,2,3, CeÁlia Badenas2, Josep Malvehy2, Artur P Darmanian1, Loli JimeÂnez2, Montserrat MilaÁ2 and Richard F Keord1 1

Westmead Institute for Cancer Research, University of Sydney at Westmead Millennium Institute, Westmead Hospital, Westmead NSW 2145, Australia; 2Hospital Clinic, IDIBAPS, Universitat de Barcelona, 08036 Barcelona, Spain

The INK4a/ARF locus encodes the cyclin dependent kinase inhibitor, p16INK4a and the p53 activator, p14ARF. These two proteins have an independent ®rst exon (exon 1a and exon 1b, respectively) but share exons 2 and 3 and are translated in dierent reading frames. Germline mutations in this locus are associated with melanoma susceptibility in 20 ± 40% of multiple case melanoma families. Although most of these mutations speci®cally inactivate p16INK4a, more than 40% of the INK4a/ARF alterations located in exon 2, aect both p16INK4a and p14ARF. We now report a 16 base pair exon 1b germline insertion speci®cally altering p14ARF, but not p16INK4a, in an individual with multiple primary melanomas. This mutant p14ARF, 60ins16, was restricted to the cytoplasm, did not stabilize p53 and was unable to arrest the growth of a p53 expressing melanoma cell line. This is the ®rst example of an exon 1b mutation that inactivates p14ARF, and thus implicates a role for this tumour suppressor in melanoma predisposition. Oncogene (2001) 20, 5543 ± 5547. Keywords: p14ARF; mutation

p16INK4a;

melanoma;

germline

Familial melanoma is a genetically heterogeneous disease, with approximately half of the families showing evidence of linkage to chromosome band 9p21. Although this locus encodes several cell cycle regulators, including p16INK4a, p15INK4b and p14ARF, it is widely accepted that the CDK inhibitor, p16INK4a, is the melanoma predisposition gene within 9p21. Inactivating germline p16INK4a mutations have been identi®ed in approximately 50% of families in which melanoma is linked to 9p21 (Dracopoli and Fountain, 1996). In contrast, the CDK inhibitor, p15INK4b, does not appear to be important in melanomagenesis, essentially because this gene is frequently retained in melanomas and no germline p15INK4b mutations have been identi®ed in familial melanoma kindreds (Flores et al., 1997; Platz et al., 1997; Stone et al., 1995).

*Correspondence: H Rizos; E-mail: [email protected] or S Puig; E-mail: [email protected] 3 Both authors contributed equally to this work Received 15 May 2001; revised 13 June 2001; accepted 14 June 2001

Linkage analysis in melanoma prone families (Ruiz et al., 1999) and loss of heterozygosity studies in melanoma tumours (Puig et al., 2000; Ruiz et al., 1998) also suggest the existence of other melanoma genes located in 9p21. The role of p14ARF in melanoma susceptibility remains controversial and has been dicult to resolve because of its unusual genomic structure; p14ARF shares signi®cant coding sequence with p16INK4a. The p14ARF and p16INK4a genes have unique ®rst exons (exon 1b and exon 1a, respectively) but share exons 2 and 3 (Quelle et al., 1995). As a result, when tumour-associated mutations dually aect p16INK4a and p14ARF, a situation occurring in approximately 40% of human tumours (Pollock et al., 1996), the relative contribution of ARF in the genesis of the cancer is not known. The frequent alteration of both p14ARF and p16INK4a in tumours is consistent with a model in which the inactivation of the p53 and retinoblastoma cell cycle pathways cooperates in the process of immortalization. p16INK4a regulates G1-phase exit by interacting with CDK4 and CDK6 to inhibit the phosphorylation of the retinoblastoma protein (pRb) (Serrano et al., 1993). p14ARF regulates both the p53 and pRb pathways, by binding to and inhibiting the function of the proto-oncogene hdm2 (Honda and Yasuda, 1999; Pomerantz et al., 1998; Stott et al., 1998; Weber et al., 1999; Zhang et al., 1998) and targeting the transcription factors, E2F-1, -2 and -3, for degradation (Eymin et al., 2001; Martelli et al., 2001). Hdm2 promotes the nuclear export and degradation of p53 (Fuchs et al., 1998; Roth et al., 1998) and functionally inactivates pRb. The inactivation of pRb stimulates the activity of the E2F-1 and -3 transcription factors, which are positive regulators of cell cycle progression (Sun et al., 1998; Xiao et al., 1995). Consequently, the growth inhibitory activity of ARF is suppressed by inactivation of both the pRb and p53 pathways, but remains eective when only one of these pathways has been inactivated (Carnero et al., 2000). It is not surprising, therefore, that speci®c knock out mice for the murine homologue of p14ARF, p19ARF, are highly tumour prone and their mouse embryo ®broblasts are immortal (Kamijo et al., 1997). The identi®cation of alterations speci®cally targeting p14ARF, while not altering p16INK4a or p15INK4b, in human tumours has provided evidence that p14ARF is speci®cally targeted during cancer development. The

Melanoma-associated mutation impairs p14ARF function H Rizos et al

5544

Oncogene

p14ARF-speci®c exon 1b is deleted in the majority of T cell acute lymphoblastic leukemias (Gardie et al., 1998), and in some astrocytomas and metastatic melanoma cell lines (Kumar et al., 1998; Newcomb et al., 2000). Mutations in exon 1b have been identi®ed in the HCT116 colon cancer cell line and in a primary colon carcinoma (Burri et al., 2001). Furthermore, a germline deletion speci®c to exon 1b has been detected in a single kindred predisposed to melanoma and neural system tumours (Randerson-Moor et al., 2001) and a germline missense mutation aecting exon 1b has been recently identi®ed in a French melanoma kindred (Bressac-de Paillerets, personal communication). Although the functional consequence of this mutation is not known, Zhang and Xiong (1999) provided evidence that two cancer-associated missense mutations (p14ARFR98L and p14ARFR98Q) impaired the function of p14ARF. We now present an exon 1b mutation identi®ed in a female with multiple primary melanomas. This 16 base pair (bp) insertion does not aect the coding sequence of p16INK4a but causes a frameshift in p14ARF after codon 21, and inactivates the cell cycle inhibitory function of this tumour suppressor. The 16 bp insertion (60ins16) in exon 1b of p14ARF was detected in a single individual who developed primary melanomas at 37 and 43 years of age (Figure 1). This Spanish woman did not exhibit dysplastic naevi or a high number of moles, and was identi®ed during an analysis of individuals with multiple primary melanomas, in a study to be reported separately. Subsequent investigation of the family revealed that two of the three proband's daughters (30 and 36 years old) also carry the 60ins16 mutation, but have not developed melanoma or clinically atypical naevi (Figure 1). The 60ins16 mutation was not detected in 266 normal chromosomes derived from a group of unrelated control individuals. This alteration places the p14ARF sequence beyond codon 21 in a dierent reading frame and introduces a stop codon at codon 67. The insertion of 16 bp appears to result from a duplication of a CG rich region within exon 1b (Figure 1), and in this sense is similar to the 24 bp p16INK4a duplication which has been observed in several melanoma prone kindreds (Walker et al., 1995). Such duplications are thought to arise during DNA replication, through polymerase slippage or through homologous recombination errors (Pollock et al., 1998). The functional analysis of the 60ins16 mutant protein was of interest because this mutant ARF retains the N-terminal nucleolar localization sequence and the hdm2-binding domain of ARF (Figure 1). In fact, the amino terminal 15 amino acids of p14ARF are capable of translocating heterologous proteins into the nucleus and nucleolus (Lohrum et al., 2000; Midgley et al., 2000; Rizos et al., 2000; Weber et al., 2000), stabilizing p53 in vivo and inhibiting the ubiquitination of p53 in vitro (Midgley et al., 2000). To compare the subcellular distribution of wild type p14ARF and the 60ins16 mutant, the cDNA of each was cloned in frame with the FLAG-epitope in the

Figure 1 Analysis of the 16 bp exon 1b insertion. (a) Pedigree of a patient with germ-line exon 1b mutation and multiple primary melanomas. dx, age of diagnosis; +, mutation carrier; 7, noncarrier; ®lled in symbol, aected; open symbol, unaected; numbers, age (of onset for the aected); arrow indicates proband. (b) The wild type p14ARF and mutant (p14ARF60ins16) sequences are shown. The exon 1b insertion was identi®ed using SSCP analysis and con®rmed by sequencing PCR products in both orientations and sequencing the cloned exon 1b mutant. The 16 bp repeat is highlighted in the mutant sequence and the Nterminal nucleolar localization domain (Rizos et al., 2000) is underlined

pFLAG-CMV-5b vector. The subcellular distribution of each construct was evaluated in transiently transfected NM39 melanoma cells. These cells lack endogenous p16INK4a and p14ARF, but express p53 and pRb (Rizos et al., 1999). As expected p14ARF-FLAG localized predominantly in the nucleoli of most transfected NM39 melanoma cells (Figure 2). In contrast, the p14ARF60ins16 mutant was retained exclusively in the cytoplasm of almost all transiently transfected NM39 cells (Figure 2) and Saos-2 cells (data not shown). Considering that p14ARF60ins16 retains the N-terminal nucleolar localization sequence, it was surprising that


5545

Figure 2 Cellular distribution of FLAG-epitope tagged wild type p14ARF and the p14ARF60ins16 mutant in NM39 melanoma cells. Wild type p14ARF-FLAG was synthesized as previously described (Rizos et al., 2000). The p14ARF60ins16 mutant was engineered by PCR-mediated mutagenesis, ligated to the Nterminus of the FLAG-epitope encoded by the pFLAG-CMV-5b vector (Sigma) and completely sequenced. Each p14ARF-FLAGtagged construct was transiently introduced using Lipofectamine 2000 reagent (GIBCO ± BRL), into the NM39 melanoma cells grown on coverslips in 6-well plates. Approximately 40 h post transfection cells were ®xed in 3.7% formaldehyde, permeabilized and immunostained for 50 min with a monoclonal mouse aFLAG M2 antibody (Sigma) followed by a 50 min exposure to a FITC-conjugated a-mouse IgG antibody (Roche). Nuclei were visualised by Hoechst 33258 staining (2 mg/ml). LM, light microscopy. The proportion of transfected cells showing predominantly nucleolar (Nlo), nuclear and cytoplasmic (N+C) or cytoplasmic only (C) staining is shown in the histograms. Subcellular distribution was determined from two independent experiments. The number (n) of ¯uorescent cells counted is indicated

ectopically expressed p14ARF60ins16 was retained in the cytoplasm. Only a small proportion of transiently transfected NM39 and Saos-2 cells (2.5% and 6%, respectively) accumulated p14ARF60ins16 in the cytoplasm and the nucleolus. It is possible that the structure of the chimeric 60ins16 protein masks the ARF-derived nucleolar localization domain resulting in the cytoplasmic sequestration of this mutant. It is well established that nuclear ARF is required for p53 stabilization and cell cycle inhibition. The p53 stabilizing activity of the p14ARF60ins16 was evaluated by transiently transfecting NM39 cells with the p14ARF60ins16-FLAG or wild type ARF-FLAG construct. Approximately 40 h post transfection cells were immunostained for p53. Of the transfected NM39 cells expressing wild type ARF, 93% accumulated high levels of p53 (Figure 3). Similarly, 88% of the transfected NM39 cells expressing wild type ARF also accumulated increased levels of the p53 transcriptional target, hdm2 (data not shown). In contrast, only 4% of NM39 cells expressing p14ARF60ins16-FLAG accumulated high levels of the p53 protein (Figure 3) and there was no increase in the levels of hdm2 in transfected

Figure 3 Transient expression of wild type p14ARF but not p14ARF60ins16 stabilizes p53. (a) NM39 cells transfected with the wild type p14ARF-FLAG or the p14ARF60ins16-FLAG plasmid were immunostained, 40 h post transfection, for the FLAG and p53 antigens by dual sequential immuno¯uorescence. Proteins were detected with mouse a-FLAG M2 antibody (Sigma) and rabbit a-p53 (Santa Cruz) followed by a 50 min exposure to a FITC-conjugated a-mouse secondary IgG (Roche) and texas redconjugated a-rabbit secondary IgG (Jackson ImmunoResearch). Nuclei were visualized by Hoechst 33258 staining. LM, light microscopy. (b) The proportion of NM39 cells expressing either p14ARF or p14ARF60ins16 and accumulating increased levels of p53 is expressed as a percentage of all transfected cells. The number (n) of transfected cells counted is indicated

NM39 cells expressing the 60ins16 variant (data not shown). The eect of wild type p14ARF or p14ARF60ins16 on cell cycle progression was investigated by transiently transfecting FLAG-tagged constructs into NM39 melanoma cells. Forty hours after transfection, DNA content was assayed by ¯ow cytometry to determine cell cycle distribution. At this time point, the expression of wild type p14ARF in NM39 cells induced potent G1 cell cycle arrest. In contrast, the p14ARF60ins16 variant reproducibly failed to arrest transiently transfected NM39 cells (Figure 4). Consequently, the predominantly cytoplasmic p14ARF60ins16 mutant is unable to stabilize p53 and fails to induce cell cycle arrest in p53 expressing cells. The contribution of p14ARF to melanoma development remains unresolved. To date, three alterations speci®c to ARF have been reported in the germline of individuals with melanoma. The functionally impaired Oncogene


5546

nomas, but this individual's daughters are also carriers of the mutant ARF gene and are so far unaected at ages under 40 years. A p14ARFGly16Asp mutant was recently identi®ed in a French melanoma kindred, but there is no reported functional data on this mutant (Bressac-de Paillerets, personal communication). A germline deletion of the p14ARF-speci®c exon 1b was associated with melanoma-astrocytoma syndrome in a four-case melanoma family (Randerson-Moor et al., 2001). In addition, approximately 50% of kindreds with melanoma linked to 9p21 do not have detectable alterations in p16INK4a. In two such families we found that the allele-speci®c expression of p14ARF from the chromosome segregating with disease was signi®cantly reduced (Rizos et al., 1997). If we also consider that 40% of melanoma-prone kindreds carry germline alterations that dually aect p16INK4a and p14ARF, it would appear that the role of p14ARF in melanoma predisposition might have been underestimated. Further studies, including functional analysis of p14ARF mutants and genotype-phenotype correlations are needed to resolve the relative roles of p14ARF and p16INK4a in melanoma susceptibility and development. Figure 4 Cell cycle inhibitory activity of melanoma-associated p14ARF60ins16. NM39 melanoma cells (36105) were transfected with the indicated p14ARF-FLAG plasmid (3 mg) and pCMVEGFP-spectrin (1 mg). The cell cycle distribution of green ¯uorescent cells was determined, 40 h post transfection, using propidium iodide staining. (a) DNA content from at least 3000 EGFP-spectrin positive cells was analysed using ModFit software. (b) The G1-phase arrest induced by wild type p14ARF was set at 100% and the arrest induced by the p14ARF60ins16 mutant was expressed relative to the wild type protein. Each result is derived from at least three independent transfection experiments

p14ARF60ins16 mutant described in this study, occurs in an individual diagnosed with multiple primary mela-

Acknowledgments The identi®cation of melanoma-associated mutations was performed within the multidisciplinary malignant melanoma group integrated by: Tessa Castel, Carles Conill, Francisco Cuellar, Loli Jimenez, Josep Malvehy, Rosa Marti, BegonÄa Mellado, Josep Palou, Susana Puig, RamoÂn Rull, Jordi Segura, Jose Soler, Mauricio Vera, Sergi Vidal, Antoni Vilalta, RamoÂn Vilella and Eva Yachi. We thank Dr Tom Shenk for supplying the pCMVEGFP-spectrin plasmid. This work was supported by the National Health and Medical Research Council, the NSW Cancer Council, the NSW Health Department, the Fondo de Investigaciones Sanitarias and Generalitat de Catalunya.

References Burri N, Shaw P, Bouzourene H, Sordat I, Sordat B, Gillet M, Schorderet D, Bosman FT and Chaubert P. (2001). Lab. Invest., 81, 217 ± 229. Carnero A, Hudson JD, Price CM and Beach DH. (2000). Nature Cell. Biol., 2, 148 ± 155. Dracopoli NC and Fountain JW. (1996). Cancer Surv., 26, 115 ± 132. Eymin B, Karayan L, Seite P, Brambilla C, Brambilla E, Larson CJ and Gazzeri S. (2001). Oncogene, 20, 1033 ± 1041. Flores JF, Pollock PM, Walker GJ, Glendening JM, Lin AHT, Palmer JM, Walters MK, Hayward NK and Fountain JW. (1997). Oncogene, 15, 2999 ± 3005. Fuchs SY, Adler V, Buschmann T, Wu X and Ronai Z. (1998). Oncogene, 17, 2543 ± 2547. Gardie B, Cayuela JM, Martini S and Sigaux F. (1998). Blood, 91, 1016 ± 1020. Honda R and Yasuda H. (1999). EMBO J., 18, 22 ± 27. Kamijo T, Zindy F, Roussel MF, Quelle DE, Downing JR, Ashmun RA, Grosveld G and Sherr CJ. (1997). Cell, 91, 649 ± 659. Oncogene

Kumar R, Sauroja I, Punnonen K, Jansen C and Hemminki K. (1998). Genes Chrom. Cancer, 23, 273 ± 277. Lohrum MA, Ashcroft M, Kubbutat MH and Vousden KH. (2000). Current Biol., 10, 539 ± 542. Martelli F, Hamilton T, Silver DP, Sharpless NE, Bardeesy N, Rokas M, DePinho RA, Liningston DM and Grossman SR. (2001). Proc. Natl. Acad. Sci. USA., 98, 4455 ± 4460. Midgley CA, Desterro JMP, Saville MK, Howard S, Sparks A, Hay RT and Lane DP. (2000). Oncogene, 19, 2312 ± 2323. Newcomb EW, Alonso M, Sung T and Miller DC. (2000). Hum. Pathol., 31, 115 ± 119. Platz A, Hansson J, Mansson-Brahme E, Lagerlof B, Linder S, Lundqvist E, Sevigny P, Inganas M and Ringborg U. (1997). J. Natl. Cancer Inst., 89, 697 ± 702. Pollock PM, Pearson JV and Hayward NK. (1996). Genes Chrom. Cancer, 15, 77 ± 88.


Pollock PM, Spurr N, Bishop T, Newton-Bishop J, Gruis N, van der Velden PA, Goldstein AM, Tucker MA, Foulkes WD, Barnhill R, Haber D, Fountain J and Hayward NK. (1998). Human Mutation, 11, 424 ± 431. Pomerantz J, Schreiber-Agus N, LieÂgeois NJ, Silverman A, Alland L, Chin L, Potes J, Chen K, Orlow I, Lee H-W, Cordon-Cardo C and DePinho RA. (1998). Cell, 92, 713 ± 723. Puig S, Castro J, Ventura PJ, Ascaso C, Malvehy J, Estivill X, Vilalta A, Lecha M and TC. (2000). Melanoma Res., 10, 231 ± 236. Quelle DE, Zindy F, Ashmun RA and Sherr CJ. (1995). Cell, 83, 993 ± 1000. Randerson-Moor JA, Harland M, Williams S, CuthbertHeavens D, Sheridan E, Aveyard J, Sibley K, Whitaker L, Knowles M, Newton Bishop J and Bishop DT. (2001). Hum. Mol. Genet., 10, 55 ± 62. Rizos H, Becker TM, Holland EA, Keord RF and Mann GJ. (1997). Oncogene, 15, 515 ± 523. Rizos H, Darmanian AP, Indsto JO, Shannon JA, Keord RF and Mann GJ. (1999). Melanoma Res., 9, 10 ± 19. Rizos H, Darmanian AP, Mann GJ and Keord RF. (2000). Oncogene, 19, 2978 ± 2985. Roth J, Dobbelstein M, Freedman DA, Shenk T and Levine AJ. (1998). EMBO J., 17, 554 ± 564. Ruiz A, Puig S, Lynch M, Castel T and Estivill X. (1998). Int. J. Cancer, 76, 312 ± 316.

Ruiz A, Puig S, Malvehy J, Lazaro C, Lynch M, M G-AA, Puig L1, Sanchez-Conejo J, Estivill X and Castel T. (1999). J. Med. Genet., 36, 490 ± 494. Serrano M, Hannon GJ and Beach D. (1993). Nature, 366, 704 ± 707. Stone S, Jiang P, Dayananth P, Tavtigian SV, Katcher H, Parry D, Peters G and Kamb A. (1995). Cancer Res., 55, 2988 ± 2994. Stott FJ, Bates S, James MC, McConnell BB, Starborg M, Brookes S, Palmero I, Ryan K, Hara E, Vousden KH and Peters G. (1998). EMBO J, 17, 5001 ± 5014. Sun P, Dong P, dai K, hannon GJ and Beach D. (1998). Science, 282, 2270 ± 2272. Walker GJ, Hussussian CJ, Flores JF, Glendening JM, Haluska FG, Dracopoli NC, Hayward NK and Fountain JW. (1995). Hum. Mol. Gen., 4, 1845 ± 1852. Weber JD, Kuo ML, Bothner B, DiGiammarino EL, Kriwacki RW, Roussel MF and Sherr CJ. (2000). Mol. Cell Biol., 20, 2517 ± 2528. Weber JD, Taylor LJ, Roussel MF, Sherr CJ and Bar-Sagi D. (1999). Nature Cell. Biol., 1, 20 ± 26. Xiao Z-X, Chen J, Levine AJ, Modjtahedl N, Xing J, Sellers WR and Livingston DM. (1995). Nature, 375, 694 ± 697. Zhang Y and Xiong Y. (1999). Mol. Cell, 3, 579 ± 591. Zhang Y, Xiong Y and Yarbrough WG. (1998). Cell, 92, 725 ± 734.

5547

Oncogene