|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Journal of Clinical Oncology, Vol 22, No 2 (January 15), 2004: pp. 286-292 © 2004 American Society of Clinical Oncology. DOI: 10.1200/JCO.2004.07.112 BRAF Gene Is Somatically Mutated but Does Not Make a Major Contribution to Malignant Melanoma Susceptibility: The Italian Melanoma Intergroup StudyFrom the Istituto di Chimica Biomolecolare-Sezione di Sassari, CNR, Alghero; Istituto di Anatomia Patologica, Università di Sassari, Sassari; Istituto Nazionale Tumori Fondazione G. Pascale, Napoli; Dipartimento di Oncologia, Biologia e Genetica, Università di Genova, Genova; Biologia Molecolare e Cellulare, Istituto Dermopatico dell'Immacolata, IRCCS, Rome; Centro di Osservazione Epidemiologica Multizonale, Azienda USL1, Sassari, Italy; and the Cancer Genome Project, Wellcome Trust Sanger Institute, Hinxton, United Kingdom Address reprint requests to Giuseppe Palmieri, MD, Istituto di Chimica Biomolecolare-Sezione di Sassari, CNR, Località Tramariglio-Alghero, 07040 Santa Maria La Palma (Sassari), Italy; e-mail:gpalmieri{at}yahoo.com
PURPOSE: Oncogenic activation of the BRAF gene has been demonstrated to be involved in the pathogenesis of malignant melanoma (MM). In this study, we investigated the contribution of BRAF to melanoma susceptibility, also making a comparison with frequency of CDKN2A germline mutations in MM patients from different areas in Italy. PATIENTS AND METHODS: Using a combination of denaturing high-performance liquid chromatography analysis and automated sequencing on genomic DNA from peripheral blood or tumor tissue samples, 569 MM patients (211 from northern Italy and 358 from southern Italy) were screened for BRAF mutations. RESULTS: Three BRAF germline sequence variants (M116R, V599E, and G608H) were identified in four (0.7%) of 569 MM patients. The most common BRAF mutation, V599E, was detected in one germline DNA sample only; M116R and G608H were newly described mutations. A high frequency (59%) of BRAF mutations was instead observed in tumor samples from patients also undergoing germline DNA analysis; at the somatic level, substitution of valine 599 was found to account for the majority (88%) of BRAF mutations. We then estimated the germline mutation rates in BRAF and CDKN2A among 358 consecutively collected patient samples originating in southern Italy; a low (2.5%) or very low (0.29%) prevalence of CDKN2A and BRAF mutations, respectively, was detected. CONCLUSION: Mutation analysis of either blood DNA from a large collection of MM patients or matched MM tissues from a subset of such patients revealed that BRAF is somatically mutated and does not play a major role in melanoma susceptibility. The present study further suggests that patient origin may account for different mutation rates in candidate genes.
Incidence and mortality rates of malignant melanoma (MM) in fair-skinned populations are increasing worldwide [1]. Environmental and genetic factors have been indicated as responsible for neoplastic transformation of the melanocytic cells [2]. The role of ultraviolet radiation in MM pathogenesis is not well understood yet, and epidemiologic data on this issue are controversial [2,3]. Melanomas arise at different body sites, sun-exposed or not, presenting different associations with melanocytic nevi and solar keratoses [3,4]. In terms of genetic factors, germline mutations of the CDKN2A gene have been widely reported as the most common cause of inherited susceptibility to MM [2,5]. Moreover, the prevalence of CDKN2A mutations within different geographic areas has been demonstrated to be tightly associated with the variation of the population incidence rates for MM [6-9]. Although the role of genetic factors is clearly evident, predisposition to MM seems indeed to be genetically heterogeneous, and several low-penetrance candidate genes have been implicated [5, 10]. Recently, our group has reported somatic mutations of the BRAF gene in the majority (approximately two thirds) of MMs [11]. The BRAF gene encodes a serine/threonine kinase that acts in the mitogen-activated protein kinase (MAPK) pathway to transduce regulatory signals from RAS to MEK1/2. Activation of the MAPK pathway, through both receptor tyrosine kinases and G-protein–coupled receptors, has a central role in melanocyte proliferation [11]. To investigate MM susceptibility that is attributable to mutations in BRAF, we carried out mutation analysis in families with MM, individuals with melanoma negative for family history (sporadic cases), and controls (also including tumor tissues from subsets of MM cases undergoing investigations for germline DNA mutations). Patients originated from northern and southern Italy, where different incidence rates of MM have been reported (standardized rates per 100,000 inhabitants: 9.5 to 11.7 cases in northern Italy v 4.0 to 4.5 cases in southern Italy for males [mean, 8.5] and 11.0 to 13.3 cases in northern Italy v 3.5 to 4.3 cases in southern Italy for females [mean, 9.6] [1]). In addition, we here provide the first opportunity to compare the prevalence of germline mutations in both BRAF and CDKN2A genes among a large series of MM patients with the same geographic origin.
Cases and Controls Five hundred sixty-nine patients with histologically proven diagnosis of MM were included in the study. For 358 patients originating from southern Italy, clinicopathologic features, such as histologic classification (including Breslow thickness and Clark level of invasion), disease stage at diagnosis, disease-free survival, and overall survival, were abstracted from hospital records. The main clinical characteristics of southern Italian MM patients are reported in Table 1. Disease stage was defined according to the recent American Joint Committee on Cancer guidelines [12].
Familial recurrence of melanoma was ascertained in the consecutively collected southern Italian patients through specific cancer-evaluation programs at both the National Tumor Institute of Naples (which represents the principal institution, accounting for a large fraction of cancer patients from the southern part of Italy) and the University of Sassari (accounting for majority of cancer patients from Sardinia). Melanoma families were selected according to the standardized criteria [13]: at least three affected members, or two affected members and presence of at least one of the following subcriteria: at least one affected member younger than 50 years at onset or a case of pancreatic cancer in a first- or second-degree relative; or one affected member with multiple primary melanomas. After patients were informed about the aims and limits of the study, blood samples were obtained with their written consent. Normal and neoplastic tissue samples were obtained from publicly available banks, in the case of cell lines, and from individual investigators using appropriate local informed consent protocols, in the case of short-term cultures (< passage 15) or tumor tissue specimens from melanoma patients. The study was reviewed and approved by the ethical review boards of the National Tumor Institute of Naples and both the University of Sassari and the University of Genova.
Mutation Screening of BRAF and CDKN2A The full coding sequence and splice junctions of BRAF (except exons 1 and 18, which failed to amplify, despite the use of different primer sets) were screened for mutations by denaturing high-performance liquid chromatography (DHPLC), the WAVE Nucleic Acid Fragment Analysis System (Transgenomic, Omaha, NE), or direct sequencing, using an automated fluorescence-based cycle sequencer (ABIPRISM3100; Applied Biosystems, Foster City, CA). Primer sets and protocols for polymerase chain reaction (PCR) assays were as previously described [11]. The complete coding sequences and intron-exon boundaries of the CDKN2A gene were amplified in 358 patients whose information on both clinicopathologic features and family history for cancer were available, as previously reported [14]. Each MM case with absence of a disease-causing mutation (such as a truncating mutation or one of the previously described pathogenic missense variants) was defined as noncarrier of CDKN2A. To confirm that each gene variant detected by sequencing was a real mutation or a polymorphism, 55 unrelated healthy individuals (corresponding to 110 control chromosomes), originating from the same geographic area and with no family history for cancer, were used as controls and screened for each sequence variation identified.
Genomic DNA from 569 melanoma patients (211 from northern Italy and 358 from southern Italy) was screened for germline mutations of the BRAF gene using a two-step approach. First, PCR products corresponding to the coding exons and intron-exon junctions were analyzed by DHPLC, using previously reported protocols [15]. All PCR products presenting an abnormal denaturing profile in comparison with the normal controls were sequenced using an automated approach. Three germline variants, at exons 3 and 15 of BRAF, were detected in four MM patients (0.7%). Only one case presented the substitution of valine by a glutamic acid at position 599 (V599E), which has been demonstrated to account for majority (more than 80%) of the BRAF mutations identified [11, 16-18]. The remaining two variants, M116R and G608H, have not been reported before in databases; negative screening of germline DNA from 55 unrelated normal individuals (corresponding to 110 control chromosomes) allowed us to define the M116R at exon 3 and the G608H at exon 15 as new mutations. To compare frequencies of somatic and germline mutations of the BRAF gene in our series, the pattern of BRAF mutations was then assessed in tumor DNA from a subset of melanoma patients who also underwent mutation analysis on germline DNA. Therefore, we screened the full coding sequence and intron-exon junctions of BRAF in genomic DNA from different melanoma samples: short-term cultures and tumor tissues (in both cases, from either primary melanomas or metastatic MM lesions). In Table 2, the list of all variants identified in the present study is reported. Except for the germline polymorphisms and the M116R variant, all mutations were grouped in two clusters involving either the kinase domain at exon 15 or the ATP-binding pocket at exon 11 of the BRAF gene. All identified mutations were single- or dinucleotide substitutions with missense effects and correspondent amino acid changes (Table 2). Figure 1 shows the nucleotide sequences for the five new BRAF mutations.
The high prevalence of somatic mutations previously reported [11] in melanoma samples (as summarized in Table 3) was also found in this study among tumor samples from a subset of melanoma patients undergoing germline DNA screening (Table 4). Altogether, mutations were detected in 15 (57.7%) of 26 melanoma tissues and 11 (61.1%) of 18 short-term MM cultures. All seven (58.3%) of 12 melanoma immortalized cell lines (obtained from public banks and used as positive controls) carried the most prominent V599E mutation (Table 4), confirming the findings previously described [11] for such MM samples (Table 3). None of the sequence changes identified was present in normal adjacent tissue from the same MM patients (Table 4), further indicating that these variants are tumor-specific and somatically acquired mutations.
Prevalence of germline mutations in CDKN2A (the major gene involved in melanoma susceptibility) was simultaneously evaluated in 358 consecutively collected MM patient samples originating from southern Italy. (The remaining 211 patients of our series were previously classified as sporadic cases and noncarriers of CDKN2A originating from northern Italy.) Mutation analysis for all coding regions and splice boundaries of the CDKN2A gene was again performed by DHPLC and automated sequencing. Six CDKN2A mutations were found in nine MM patients (2.5%); two of these variants (S12Term and E26Term) have not been previously reported in the Human Gene Mutation Database at http://archive.uwcm.ac.uk/uwcm/mg/hgmd0.html (Table 5). According to the selection criteria, 17 individuals (4.7%) with familial recurrence of melanoma were identified; although none of them was found positive for germline BRAF mutations, three familial cases (17.6%) were found to carry CDKN2A mutations (Table 4). Although less striking, an evident difference between prevalence rates of CDKN2A germline mutations (six mutations [1.8%] in 341 patients) and BRAF germline mutation (one mutation [0.29%] in 341 patients) was observed among sporadic MM patients from the same geographic area (Table 5).
No BRAF variant (either mutation or polymorphism) was found in individuals with CDKN2A mutations. Among noncarriers of CDKN2A, we detected BRAF germline mutations in three (1.42%) of 211 patients from northern Italy and one (0.29%) of 349 patients from southern Italy. Altogether, BRAF mutations were present at high rates in tumor DNAs (26 mutations [59.1%] in 44 tumor DNA samples) and nearly absent in blood DNA samples (one mutation [0.28%] in 358 blood DNA samples) from southern Italian melanoma patients (Table 6). Although with a population prevalence lower than expected, CDKN2A mutations presented a frequency about nine-fold higher than that of BRAF mutations in germline DNA samples from the same series of patients (Table 6).
Finally, univariate analysis using the Pearson's  2 test revealed no significant statistical correlation between the presence of either BRAF variants or CDKN2A mutations and any clinicopathologic parameter (Breslow thickness, Clark level of invasion, age and disease stage at diagnosis, disease-free survival, and overall survival) among the 358 southern Italian patients whose hospital records were available.
A mutation analysis of the BRAF gene, which we previously demonstrated to be somatically mutated in two thirds of malignant melanomas and at lower frequency in a wide range of human cancers [11], was performed in germline DNA from 569 melanoma patients originating from the northern and southern parts of Italy. A subset of patients was also screened for germline mutations in the CDKN2A gene, which has been demonstrated to be mostly involved in melanoma susceptibility [2,5,13]. Thus, the present study represents the first in which prevalence of germline mutations in these two candidate genes, BRAF and CDKN2A, has been extensively evaluated in MM patients with the same geographic origin. After a molecular screening based on a combination of DHPLC and sequence analysis, three BRAF germline mutations were found in four melanoma patients (0.7%). Conversely, the distribution of BRAF somatic mutations in the short-term culture or tumor tissue samples from the same series of melanoma patients undergoing screening for BRAF germline mutations was remarkably similar to that seen in the melanoma cell lines (26 somatic mutations in 44 samples [59.1%] v 7 somatic mutations in 12 cell lines [58.3%]). Our findings confirmed that changes at V599 (V599R, V599K, V599E, or V599D) account for majority of BRAF somatic mutations (overall, 54 [91.5%] of 59 mutations; Table 2). In addition to our previous data [11], we here demonstrated that similar rates of BRAF mutations are present in primary and metastatic melanomas as well as in cultured MM cells, indicating that BRAF mutations occur before tumor dissemination, and their incidence does not vary during tumor progression. All sequence variants found in melanoma tissues for which a matched normal sample (either tissue or peripheral blood) from the same individual was available were shown to be somatic mutations, strongly suggesting that BRAF does not play a major role in melanoma predisposition. Germline CDKN2A mutations, which have been instead widely demonstrated to confer susceptibility to melanoma [2, 5, 13], were detected in nine (2.5%) of 358 consecutively collected MM patient samples from the same geographic area ( Tables 5 and 6). In this series, presence of familial recurrence of melanoma was ascertained through well defined clinical criteria and without prior linkage analysis. Taking also into consideration some lack of sensitivity of our screening assay to reduce the number of positive cases (although the percentages of the BRAF somatic mutations detected in this study are similar to those previously reported [11]), a very low frequency of germline mutations in these two candidate genes was observed among southern Italian melanoma patients. In contrast to a higher frequency (approximately 9% [19]) of CDKN2A germline mutations observed in nonfamilial cases from northern Italy, our data among sporadic patients from southern Italy (six [1.8%] of 341 patients) strongly suggest that discrepancy in CDKN2A mutation frequency may be due to patient origin and/or to the different genetic backgrounds of populations. Data from previous studies on familial melanoma [20, 21] also indicating that the penetrance varies with melanoma population incidence rates [9] are consistent with this hypothesis. In fact, either the standardized incidence rates of melanoma (roughly, 12 cases per 100, 000 inhabitants in northern Italy versus four cases per 100.000 inhabitants in southern Italy [1]) or the prevalence of familial MM cases (9.2% in patients from northern Italy [19, 21] v 4.7% in those from southern Italy in this study) are significantly different in these two geographic areas. Thus, the same factors that affect population incidence of MM may also mediate CDKN2A penetrance. Considering the frequency of BRAF germline mutations in noncarriers of CDKN2A with sporadic melanoma (1.42% in MM cases from northern Italy v 0.29% in those from southern Italy), one could further sustain the hypothesis that patient origin may account for different mutation prevalences in several candidate genes or, more generally, for differences in genetic factors involved in tumorigenesis. Interestingly, frequency of germline mutations in both CDKN2A and BRAF genes was found to be about five-fold higher in MM cases from northern Italy than in those from the southern part of Italy. (One could speculate that genetic factors involved in melanoma predisposition may be more heterogeneous within this latter geographic area.) Our study provides clear indications that mutational activation of the BRAF gene is a pathogenetic event contributing to melanoma tumorigenesis at the somatic level with minor participation to the melanoma susceptibility at the germline level. BRAF somatic mutations were also found in majority of histologically different nevi [22], implying that, although a crucial step in the initiation of melanocytic transformation, mutation of this gene alone is insufficient for the development of melanoma. Additional molecular defects are thus required to transform a nevus containing a BRAF mutation into a melanoma, further suggesting that genes acting into the RAS/BRAF/MAPK pathway may not be able to contribute to MM susceptibility.
The authors indicated no potential conflicts of interest.
The authors are grateful to the patients and their families for their important contribution to this study. We thank Dr Maria Napolitano for technical assistance and Dr Assunta Criscuolo for data management.
Supported by Associazione Italiana Ricerca sul Cancro, Ricerca Finalizzata Ministero della Salute (F.S.N.), and Regione Autonoma della Sardegna Progetto Genetica e Tumori nel Nord Sardegna. Authors' disclosures of potential conflicts of interest are found at the end of this article.
1. Parkin DM, Whelan SL, Ferlay J, et al: Cancer incidence in five continents, Vol VIII. Lyon, France, IARC Press, 2003 2. Bataille V: Genetic epidemiology of melanoma. Eur J Cancer 39:1341-1347, 2003
3. Gilchrest BA, Eller MS, Geller AC, et al: The pathogenesis of melanoma induced by ultraviolet radiation. N Engl J Med 340:1341-1348, 1999 4. Jhappan C, Noonan FP, Merlino G: Ultraviolet radiation and cutaneous malignant melanoma. Oncogene 22:3099-3112, 2003[CrossRef][Medline] 5. Hayward NK: Genetics of melanoma predisposition. Oncogene 22:3053-3062, 2003[CrossRef][Medline]
6. Soufir N, Avril M-F, Chompret A, et al: Prevalence of p16 and CDK4 germline mutations in 48 melanoma-prone families in France. Hum Mol Genet 7:209-216, 1998
7. Aitken J, Welch J, Duffy D, et al: CDKN2A variants in a population-based sample of Queensland families with melanoma. J Natl Cancer Inst 91:446-452, 1999 8. Harland M, Holland EA, Ghiorzo P, et al: Mutation screening of the CDKN2A promoter in melanoma families. Genes Chromosom Cancer 28:45-57, 2000[CrossRef][Medline]
9. Bishop DT, Demenais F, Goldstein AM, et al: Geographical variation in the penetrance of CDKN2A mutations for melanoma. J Natl Cancer Inst 94:894-903, 2002 10. Goldstein AM, Chidambaram A, Halpern A, et al: Rarity of CDK4 germline mutations in familial melanoma. Melanoma Res 12:51-55, 2002[CrossRef][Medline] 11. Davies H, Bignell GR, Cox C, et al: Mutations of the BRAF gene in human cancer. Nature 417:949-954, 2002[CrossRef][Medline]
12. Balch CM, Buzaid AC, Soong SJ, et al: Final version of the American Joint Committee on Cancer Staging System for Cutaneous Melanoma. J Clin Oncol 19:3635-3648, 2001 13. Haluska FG, Hodi FS: Molecular genetics of familial cutaneous melanoma. J Clin Oncol 16:670-682, 1998[Abstract] 14. Smith-Sorensen B, Hovig E: CDKN2A (p 16INK4A) somatic and germline mutations. Hum Mutat 7:294-303, 1996[CrossRef][Medline]
15. Palmieri G, Palomba G, Cossu A, et al: BRCA1 and BRCA2 germline mutations in Sardinian breast cancer families and their implications for genetic counseling. Ann Oncol 13:1899-1907, 2002
16. Cohen Y, Xing M, Mambo E, et al: BRAF mutation in papillary thyroid carcinoma. J Natl Cancer Inst 95:625-627, 2003
17. Brose MS, Volpe P, Feldman M, et al: BRAF and RAS mutations in human lung cancer and melanoma. Cancer Res 62:6997-7000, 2002
18. Yuen ST, Davies H, Chan TL, et al: Similarity of the phenotypic patterns associated with BRAF and KRAS mutations in colorectal neoplasia. Cancer Res 62:6451-6455, 2002 19. Mantelli M, Pastorino L, Ghiorzo P, et al: Frequency of CDKN2A mutation in Ligurian non-familial melanoma patients. Presented at third Research Meeting on Melanoma, Milan, Italy, May 26-28, 2003 20. Della Torre G, Pasini B, Frigerio S, et al: CDKN2A and CDK4 mutation analysis in Italian melanoma-prone families: Functional characterization of a novel CDKN2A germ line mutation. Br J Cancer 85:836-844, 2001[CrossRef][Medline] 21. Mantelli M, Barile M, Ciotti P, et al: High prevalence of the G101W germline mutation in the CDKN2A (P16(ink4a) gene in 62 Italian malignant melanoma families. Am J Med Genet 107:214-221, 2002[CrossRef][Medline] 22. Pollock PM, Harper UL, Hansen KS, et al: High frequency of BRAF mutations in nevi. Nat Genet 33:19-20, 2003[CrossRef][Medline] Submitted July 15, 2003; accepted November 13, 2003.
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2004 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
