Mutations in the CDKN2A (p16INK4a) gene in microdissected sporadic primary melanomas

Recognizing Pattern and Rule of Mutation Signatures Corresponding to Cancer Types

Cancer has been generally defined as a cluster of systematic malignant pathogenesis involving abnormal cell growth. Genetic mutations derived from environmental factors and inherited genetics trigger the initiation and progression of cancers. Although several well-known factors affect cancer, mutation features and rules that affect cancers are relatively unknown due to limited related studies. In this study, a computational investigation on mutation profiles of cancer samples in 27 types was given. These profiles were first analyzed by the Monte Carlo Feature Selection (MCFS) method. A feature list was thus obtained. Then, the incremental feature selection (IFS) method adopted such list to extract essential mutation features related to 27 cancer types, find out 207 mutation rules and construct efficient classifiers. The top 37 mutation features corresponding to different cancer types were discussed. All the qualitatively analyzed gene mutation features contribute to the distinction of different types of cancers, and most of such mutation rules are supported by recent literature. Therefore, our computational investigation could identify potential biomarkers and prediction rules for cancers in the mutation signature level.

Do CDKN2 p16 540 C>G, CDKN2 p16 580 C>T, and MDM2 SNP309 T>G gene variants act on colorectal cancer development or progression?

CDNK2 p16 plays a pivotal role in G1/S transition by regulating the p53 pathway, which was regulated by a nuclear oncoprotein, mouse double minute 2 (MDM2). Overexpression of the MDM2 gene has been shown in a number of tumor types, its gene amplification is found to associate with accelerated tumor development and failure to treatment in both hereditary and sporadic cancers. Although genetic association studies have revealed the relationship between certain genetic polymorphisms and genes that play important roles in the development and progression of colorectal cancer (CRC), it is still unknown. Therefore, the polymorphisms of p16 540 C>G, 580 C>T, and MDM2 SNP309 T>G designed to investigate the risk of CRC development and progression in a Turkish population. We enrolled 87 patients with CRC and 75 healthy controls into the study. Genotypings were determined using polymerase chain reaction-restriction fragment length polymorphism techniques. Genotype distributions of p16 540 C>G and 580 C>T were found in agreement with the Hardy-Weinberg equilibrium in patients and controls. MDM2 SNP309 T>G was found in agreement with the Hardy-Weinberg equilibrium in controls, but not in patients. The results of our study, the G allele of p16 540 C>G and GG genotype of MDM2 SNP309 T>G were found significantly lower in patients compared with controls (p<0.001, p<0.05, respectively). Haplotype analyses have shown that the C allele of both the CDKN2 p16 540 C>G and 580 C>T variants together indicate a risk haplotype for the patient group; besides, carrying the G allele of p16 540 and G allele of MDM2 also seems a risk haplotype for the patient group. Our study is the first study that investigates the relationship among variants of CDKN2 p16 540 C>G, 580 C>T, and MDM2 SNP309 T>G risk of CRC and the development and progression in the Turkish population.

Rapid repair of UVA-induced oxidized purines and persistence of UVB-induced dipyrimidine lesions determine the mutagenicity of sunlight in mouse cells

Despite the predominance of ultraviolet A (UVA) relative to UVB in terrestrial sunlight, solar mutagenesis in humans and rodents is characterized by mutations specific for UVB. We have investigated the kinetics of repair of UVA- and UVB-induced DNA lesions in relation to mutagenicity in transgenic mouse fibroblasts irradiated with equilethal doses of UVA and UVB in comparison to simulated-sunlight UV (SSL). We have also analyzed mutagenesis-derived carcinogenesis in sunlight-associated human skin cancers by compiling the published data on mutation types found in crucial genes in nonmelanoma and melanoma skin cancers. Here, we demonstrate a resistance to repair of UVB-induced cis-syn cyclobutane pyrimidine-dimers (CPDs) together with rapid removal of UVA-induced oxidized purines in the genome overall and in the cII transgene of SSL-irradiated cells. The spectra of mutation induced by both UVB and SSL irradiation in this experimental system are characterized by significant increases in relative frequency of C-->T transitions at dipyrimidines, which are the established signature mutation of CPDs. This type of mutation is also the predominant mutation found in human nonmelanoma and melanoma tumor samples in the TP53, CDKN2, PTCH, and protein kinase genes. The prevailing role of UVB over UVA in solar mutagenesis in our test system can be ascribed to different kinetics of repair for lesions induced by the respective UV irradiation.

Genetic mutations involved in melanoma: a summary of our current understanding

The biomolecular understanding of melanoma is in flux. The importance of high-penetrance genes involved in familial melanoma includes a significant number of mutations that directly lead to impairment of the checkpoints of the normal cell cycle. Furthermore, a greater understanding of the interaction between genetic factors and environmental factors, such as MC1R, CDKN2A, BRAF, and ultraviolet light, is emerging from landmark research. Although currently and with rare exception most clinicians still confine genetic testing to the realm of research, even in familial melanoma, continued and major advances in this arena may lead to development of new and revolutionary means of diagnosis and treatment, patterned on improved understanding of melanoma-related genetic mutations and resultant aberrations in cellular pathways.

The role of cell cycle regulatory proteins in the pathogenesis of melanoma

The transformation of melanocytes to melanoma cells is characterised by abnormal proliferation resulting from alterations in cell cycle regulatory mechanisms. This occurs through alterations in the two major cell cycle regulatory pathways, the retinoblastoma (Rb) and p53 tumour suppressor pathways. This review summarises the current knowledge of alterations in these two pathways at G1/S transition and specifically the role of the key cell cycle regulatory proteins pRb, p16INK4a (p16), cyclin D1, p27Kip1 (p27), p53 and p21Waf1/Cip1 (p21) in the pathogenesis of melanoma. It also considers their prognostic significance. Current data indicate that alterations of cyclin kinase inhibitor (cdki) levels are implicated in the pathogenesis of melanoma and may be useful prognostic markers. However, large validation studies linked to comprehensive clinical follow up data are necessary to clarify the prognostic significance of cell cycle regulatory proteins in individual patients.

Differences in global gene expression in melanoma cell lines with and without homozygous deletion of the CDKN2A locus genes

We studied differential global gene expression in four melanoma cell lines with three cell lines without homozygous deletion of the CDKN2A locus using HG-U133A microarrays with 22 277 transcripts. None of the cell lines carried mutations in the B-RAF and N-RAS genes. Data analysis using stringent criteria showed specific upregulation of 70 genes and downregulation of 86 genes in cell lines with homozygous deletion of the CDKN2A gene. A comparison with previous expression data showed overlapping of upregulation and downregulation of seven and 23 genes, respectively, in melanoma cell lines with homozygous deletion of the CDKN2A locus or mutations in the B-RAF and N-RAS genes. Microarray data for eight selected genes were validated with an extended number of cell lines using quantitative real-time polymerase chain reaction. The upregulated genes in cell lines with the deletion besides others included MAGE A2 [fold change 128, 95% confidence interval (CI) 82.8-172.2; t-test P=0.004], MAGE A6 (fold change 623, 95% CI 473.4-772.1; t-test P=0.001), MAGE A12 (fold change 90, 95% CI 65.1-115.5; t-test P=0.001) and dopachrome tautomerase (fold change 42, 95% CI 32.5-51.8; t-test P=0.001). Downregulated genes included interleukin 18 (fold change 489, 95% CI 146.4-831.2; t-test P=0.04), ID2 (fold change 3, 95% CI 2.2-4.9; t-test P=0.001), KLF4 (fold change 9, 95% CI 4.3-14.7; P=0.01) and CD24 antigen (fold change 1308, 95% CI 766.0-1850.8; t-test P=0.01). The upregulated genes common to cell lines with homozygous deletion of the CDKN2A gene and mutations in B-RAF and N-RAS gene included those that are involved in RAS/RAF/MEK/ERK pathways. Our results highlight effects of homozygous deletion of the CDKN2A locus on global gene expression.

The genetics of melanoma

Melanoma is an increasingly common cancer and in order to direct preventative advice at those at risk, an understanding of susceptibility is crucial. This review summarizes what is known about common low-risk genes (such as those controlling red hair) and rare high-risk genes.

Re-expression of the retinoblastoma-binding protein 2-homolog 1 reveals tumor-suppressive functions in highly metastatic melanoma cells

The loss of cell cycle control in malignant melanomas is thought to be due to a lack of retinoblastoma protein (pRb) activity. We have recently reported a progressive deficiency of the retinoblastoma-binding protein 2-homolog 1 (RBP2-H1) in advanced and metastatic melanomas in vivo, suggesting a role of RBP2-H1 in loss of pRb-mediated control. Therefore, in this study, we re-established the pRb-modulating function of RBP2-H1 in highly metastatic A375-SM melanoma cells by re-expressing its C-term (cRBP2-H1). As previously shown, the corresponding domains comprise the pRb-binding region of the RBP2-H1 protein (non-T/E1A-pRb-binding domain (NTE1A)). As a result, we detected pRb-hypophosphorylation selectively at Ser795, but not at Ser780 and Ser807/811 throughout the G1 phase of the cell cycle. As a further consequence, a block in G1/S transition was observed accompanied by a significant decrease of DNA replication and cellular proliferation. As demonstrated by cDNA microarrays of cRBP2-H1-transduced cells and confirmed by quantitative TaqMan reverse transcriptase-PCR, differential expression of melanoma-progression-related genes was observed, among them bone morphogenetic protein 2, follistatin, transforming growth factor alpha, hepatocyte growth factor, transcription factor 4 and microphthalmia-associated transcription factor. Conclusively, these data suggest that RBP2-H1 exerts a broad tumor-suppressive function partially mediated by pRb modulation. Therefore, re-establishing of RBP2-H1 could evolve as an interesting novel approach in developing experimental treatments for metastatic melanomas.

Recurrent Patterns of Dual RB and p53 Pathway Inactivation in Melanoma

Two growth inhibitory hurdles that must be overcome by the evolving cancer cell include pathways regulated by RB and p53. In human melanoma cells, inactivation of a single locus, CDKN2A, can lead to abrogation of both RB and p53 functionality through loss of the two CDKN2A cognate transcripts-p16 and p14ARF, respectively. We thus set out to assess how recurrent patterns of genetic injury at three critical human melanoma loci-CDKN2A, TP53, and CDK4-cooperate to disrupt both RB and p53 pathways. Overall, 77.8% of the melanoma cell lines analyzed showed genetic evidence of dual RB and p53 pathway compromise; this percentage is even higher if protein expression loss is considered. Although homozygous deletion of all three critical CDKN2A exons (exons 1 beta, 1 alpha, and 2) represent the most common mechanism, concurrent loss of CDKN2A(Exon1 alpha) and CDKN2A(Exon1 beta) and simultaneous point mutagenesis of CDK4 and TP53 reflect alternative cassettes of dual inactivation. In cell lines with isolated CDKN2A(Exon2) mutations, coincident alterations in TP53 or deletion of CDKN2A(Exon1 beta) suggest that p16 transcript may be preferentially targeted over the p14ARF transcript as additional p53 pathway lesions are recruited. Moreover, predictive modeling of CDKN2A(Exon2) missense mutations further suggests that the amino acid substitutions in this region negatively impact p16 to a greater extent than p14ARF. Taken together, our data point to a clear need in human melanoma cell lines, as in its murine counterpart, to disrupt both RB and p53 pathways and recurrent mechanisms may play into the unique genetic vulnerabilities of this tumor type.

Polymorphism of p16 INK4A and cyclin D1 in adenocarcinomas of the upper gastrointestinal tract

PURPOSE

We investigated the prevalence of single nucleotide polymorphisms in the p16 gene (C540G) and the cyclin D1 gene (G870A), both known to regulate function in G1 arrest and therefore, may play an important role in carcinogenesis.

METHODS

Using PCR based restriction fragment length polymorphism and single strand conformational polymorphism, we determined single nucleotide exchanges in the p16 and cyclin D1 genes among 56 esophageal adenocarcinomas (ADC) arising in Barrett's esophagus, 95 cardiac gastric ADC, and in 191 distal gastric ADC. The allelic frequencies were compared to a control group of 253 healthy blood donors.

RESULTS

The C/G genotype of p16 was identified in 10.4% of esophageal carcinomas, 13.3% of cardiac carcinomas, and in 14.1% of gastric carcinomas, compared to 17.4% in the healthy control group. All other cases showed the C/C wildtype, as no homozygous G/G nucleotide exchange was detected in the group of cancer patients or in the control group. In esophageal cancer, cyclin D1 G/G genotype was found 28.6%, A/G in 46.4%, and A/A in 25.0%. In cardiac carcinoma, frequency of cyclin D1 genotype was 27.4% for G/G, 57.9% for A/G, and 14.7% for AA. In distal gastric carcinoma, both homozygous genotypes (G/G and A/A) had a frequency of 15.2% each, while the heterozygous A/G genotype occurred in 69.6% of patients. The control group displayed 24.9% G/G, 53.8% A/G, and 21.3% A/A genotype.

CONCLUSIONS

Our results show that frequencies of p16 or cyclin D1 polymorphisms in gastric and esophageal ADC do not differ significantly from the healthy control group. Therefore, these polymorphisms are unlikely to be associated with risk of ADC of the upper gastrointestinal tract.

BRAF mutations are common somatic events in melanocytic nevi

We determined mutations in the BRAF, N-ras, and CDKN2A genes in 27 histologically diverse melanocytic nevi and corresponding surrounding tissues from 17 individuals. Mutations in the BRAF and N-ras gene were found in 22 nevi (81%) from 16 individuals (94%). The predominant BRAF mutation T1799A (V600E) was detected in 18 nevi; 1 nevus had a novel A1781G (D594V) mutation in the same gene and 3 nevi had mutations in codon 61 of the N-ras gene. In 4 individuals both nevi carried a BRAF mutation, whereas in 2 other individuals 1 nevus showed a BRAF mutation and the second nevus had an N-ras mutation. In 2 individuals normal skin distant from nevi showed a BRAF mutation. No mutations were detected in the CDKN2A gene. The mutations in the BRAF and N-ras genes, in this study, were not associated with histologic type, location, skin type, size, or numbers of nevi. Our results suggest that mutations in the BRAF gene and to some extent in the N-ras gene represent early somatic events that occur in melanocytic nevi. We hypothesize the dual effect of solar ultraviolet irradiation on melanoma, through mutagenesis and by increasing the number of melanocytic nevi, many of which carry a BRAF or N-ras mutation.

High frequency of CDKN2A alterations in esophageal squamous cell carcinoma from a high-risk Chinese population

Because previous studies have shown that loss of heterozygosity (LOH) is common on chromosome arm 9p in esophageal squamous cell carcinoma (ESCC) and that genetic alterations in CDKN2A and CDKN2B on 9p are also common, we sought to determine whether LOH and these genetic alterations are related. We performed LOH studies on chromosome bands 9p21-p22 and searched for genetic alterations of CDKN2A and CDKN2B in 56 ESCCs from a high-risk Chinese population. Seventy-three percent of patients were found to have LOH at one or more loci on chromosome bands 9p21-p22, and LOH occurred more frequently in patients with a family history of upper gastrointestinal cancer than in those with a negative family history (P = 0.01, global permutation test). CDKN2A mutations (point mutations, deletions, insertions) were observed in 25% (14 of 56) of cases, and the LOH pattern was significantly different for individuals with and without a CDKN2A mutation (P = 0.01, global test). Three new single nucleotide polymorphisms (SNPs) and 2 previously reported SNPs were identified in this group of patients. Intragenic allelic loss at polymorphic sites in CDKN2A was detected in 32% (18 of 56) of patients. Seven of the 56 (13%) cases exhibited what is considered classic evidence (n = 4) or showed potential evidence (n = 3) of biallelic inactivation. Only one alteration was observed in CDKN2B, G171A in the 5' untranslated region. Both mutation and intragenic allelic loss in CDKN2A appear to play a role in the development of ESCC.

Activating BRAF and N-Ras mutations in sporadic primary melanomas: an inverse association with allelic loss on chromosome 9

We searched and report mutations in the BRAF and N-ras genes in 22 out of 35 (63 percent) primary sporadic melanomas. In three melanomas, mutations were concomitantly present in both genes. In all, 10 out of 12 mutations in the BRAF gene involved the 'hot spot' codon 600 (In all communications on mutations in the BRAF gene, the nucleotide and codon numbers have been based on the NCBI gene bank nucleotide sequence NM_004333. However, according to NCBI gene bank sequence with accession number NT_007914, there is a discrepancy of one codon (three nucleotides) in exon 1 in the sequence with accession number NM_004333. The sequence analysis of exon 1 of the BRAF gene in our laboratory has shown that the sequence derived from NT_007914 is correct (Kumar et al., 2003). Due to the correctness of the latter, sequence numbering of codons and nucleotides after exon 1 are changed by +1 and +3, respectively.), one tandem CT1789-90TC base change represented a novel mutation and another mutation caused a G466R amino-acid change within the glycine-rich loop in the kinase domain. Mutations in the N-ras gene in 11 melanomas were at codon 61 whereas two melanomas carried mutations in codon 12 including a tandem mutation GG>AA. We observed an inverse association between BRAF/N-ras mutations and the frequency of loss of heterozygosity (LOH) on chromosome 9 at 10 different loci. Melanomas with BRAF/N-ras mutations showed a statistically significant decreased frequency of LOH on chromosome 9 compared with cases without mutations (mean fractional allelic loss (FAL)=0.29+/-0.23 vs 0.72+/-0.33; t-test, P=0.0001). Difference in the FAL value between tumours with and without BRAF/N-ras mutations on 33 loci on five other chromosomes was not statistically significant (mean FAL 0.17+/-0.19 vs 0.25+/-0.22; t-test, P=0.24). Melanoma cases with BRAF/N-ras mutations were also associated with lower age at diagnosis than cases without mutations (mean age 80.38+/-7.24 vs 65.77+/-19.79 years; t-test, P=0.02). Our data suggest that the occurrence of BRAF/N-ras mutations compensate the requirement for the allelic loss at chromosome 9, which is one of the key events in melanoma.

Frequency of UV-inducible NRAS mutations in melanomas of patients with germline CDKN2A mutations

BACKGROUND

Germline alterations in cyclin-dependent kinase inhibitor 2A (CDKN2A) are important genetic factors in familial predisposition to melanoma. Activating mutations of the NRAS proto-oncogene are among the most common somatic genetic alterations in cutaneous malignant melanomas. We investigated the occurrence of NRAS mutations in melanomas and dysplastic nevi in individuals with germline CDKN2A mutations.

METHODS

Genomic DNA was extracted from 39 biopsy samples (including primary melanomas, metastatic melanomas, and dysplastic nevi) from 25 patients in six Swedish families with a hereditary predisposition to melanoma who carried germline CDKN2A mutations. DNA was also extracted from 10 biopsy samples from patients with sporadic melanomas. NRAS was analyzed using polymerase chain reaction, single-strand conformation polymorphism analysis, and nucleotide sequence analysis. Differences in NRAS mutation frequency between hereditary and sporadic melanomas were analyzed by the chi-square test. All statistical tests were two-sided.

RESULTS

Activating mutations in NRAS codon 61, all of which were either CAA(Gln)-AAA(Lys) or CAA(Gln)-CGA(Arg) mutations, were found in 95% (20/21) of primary hereditary melanomas but in only 10% (1/10) of sporadic melanomas (P<.001). Multiple activating NRAS mutations were detected in tumor cells from different regions of individual primary melanomas in nine patients. Activating mutations that were detected in the primary melanomas of these patients were also retained in their metastases. NRAS mutations at sites other than codon 61 were also present in the primary melanomas, indicating genetic instability of this locus. NRAS codon 61 mutations were also detected in dysplastic nevi and in an in situ melanoma, suggesting a role for such mutations during early melanoma development.

CONCLUSIONS

The high frequency of NRAS codon 61 mutations detected in these hereditary melanomas may be the result of a hypermutability phenotype associated with a hereditary predisposition for melanoma development in patients with germline CDKN2A mutations.

Clinical course of bladder neoplasms and single nucleotide polymorphisms in the CDKN2A gene

Point mutations and single nucleotide polymorphisms (SNPs) in the CDKN2A gene in bladder cancer patients have been resolved only to a limited extent. The exact frequency of mutations remains uncertain and reports on SNPs are lacking. In this population-based study we investigated mutations and polymorphisms in the CDKN2A gene in bladder cancer patients from all hospitals within the Stockholm County. Mutations were determined in 4 exons of the CDKN2A gene in tumor-tissues from 172 bladder cancer patients and 2 single nucleotide polymorphisms in the 3' UTR of the CDKN2A gene were studied in 309 cases. Missense mutations were identified in only 4 of 172 (2.3%) cases, including 1 in the germ-line. Frequencies of the 500 C-->G and 540 C-->T polymorphisms in the 3' UTR of the CDKN2A in bladder cancer cases were not statistically significantly different compared to an ethnically matched control population. The tumor-specific survival was significantly shorter in patients with either the 500 C-->G or 540 C-->T polymorphism than those with wild-type CDKN2A gene (P = 0.02). Our results corroborate the earlier findings that single base mutation is not the prime mode of inactivation of the CDKN2A gene in bladder cancer. Further, the results indicate, a role for the 3' UTR polymorphisms in the CDKN2A gene in tumor invasiveness.

Phosphorylation of p16INK4A correlates with Cdk4 association

Progression through the eukaryotic cell cycle is driven by the activity of cyclin-dependent kinases. The cyclin D-dependent kinase Cdk4 promotes progression through the G(1) phase of the cell cycle and is deregulated in many human tumors. The tumor suppressor protein p16(INK4A) (p16) forms a complex with Cdk4 and inhibits kinase activity. Here we report that p16 is phosphorylated, and the phosphorylated form of p16 is preferentially associated with Cdk4 in normal human fibroblasts. We mapped phosphorylation sites on exogenously overexpressed p16 to serines 7, 8, 140, and 152 and found that endogenous p16 associated with Cdk4 is phosphorylated at serine 152. All mapped phosphorylation sites lie outside of the conserved kinase-binding domain of p16 but in regions of the protein affected by mutations in familial and sporadic cancer. Our results suggest a novel regulation of p16 activity.

Repair of UV light-induced DNA damage and risk of cutaneous malignant melanoma

BACKGROUND

The mechanism underlying the role of UV light exposure from sunlight in the etiology of cutaneous malignant melanoma (CMM) is unclear. Patients with xeroderma pigmentosum, a disease characterized by severe sensitivity to UV radiation and a defect in nucleotide excision repair, have a high incidence of CMM, which suggests that DNA repair capacity (DRC) plays a role in sunlight-induced CMM in the general population as well.

METHODS

We conducted a hospital-based case-control study of DRC and CMM among 312 non-Hispanic white CMM patients who had no prior chemotherapy or radiation therapy, and 324 cancer-free control subjects who were frequency-matched to case patients on age, sex, and ethnicity. Information on demographic variables, risk factors, and tumor characteristics was obtained from questionnaires and medical records. We used the host-cell reactivation assay to measure the DRC in study subjects' lymphocytes. All statistical tests were two sided.

RESULTS

Case patients had a 19% lower mean (+/- standard deviation [SD]) DRC (8.5 +/- 3.4%) than control subjects (10.5 +/- 5.1%), a statistically significant difference (P<.001). DRC that was at or below the median value (i.e., 9.4%) in control subjects was associated with increased risk for CMM after adjustment for age, sex, and other covariates (odds ratio [OR] = 2.02, 95% confidence interval [CI] = 1.45 to 2.82). We observed a dose-response relationship between decreased DRC and increased risk of CMM (P(trend)<.001). Patients with tumors on sun-exposed skin had statistically significantly lower DRC than patients with tumors on unexposed skin (8.2 +/- 3.3% versus 9.5 +/- 3.5%; P =.004).

CONCLUSIONS

Reduced DRC is an independent risk factor for CMM and may contribute to susceptibility to sunlight-induced CMM among the general population.

Significant impact of promoter hypermethylation and the 540 C>T polymorphism of CDKN2A in cutaneous melanoma of the vertical growth phase

Promoter hypermethylation, mutations, and loss of heterozygosity in the CDKN2A gene as well as polymorphisms at the 3'-untranslated region were determined in vertical growth phase melanomas. Methylation-specific polymerase chain reaction in soluti and in situ showed that 19% of the cases were hypermethylated at the CDKN2A promoter region, and some of these cases were heterogeneous with both methylated and unmethylated tumor cells. Methylation was associated with increased tumor cell proliferation by Ki-67 expression (P = 0.01) and decreased patient survival (P = 0.025). Point mutations in CDKN2A were found in 4% of the cases, whereas 90% had loss of heterozygosity at one or more of 4 markers studied. Furthermore, presence of the 540 C>T polymorphism at the 3'-untranslated region of CDKN2A (23%) was associated with improved survival in multivariate analysis (hazard ratio, 2.6; P = 0.02). Our results suggest that promoter methylation of the CDKN2A gene is present in a subgroup of the tumors and associated with increased tumor cell proliferation and reduced survival. Further, the 540 C>T polymorphism might define a distinct subgroup of low-grade vertical growth phase melanomas. These findings support a significant role of the CDKN2A gene in melanoma progression.

Multiple primary melanoma revisited

BACKGROUND

Incidence of cutaneous melanoma continues to increase in the Caucasian population worldwide. Approximately 5% of melanoma patients develop additional primary melanoma. This rate is significantly higher than the estimated lifetime risk of an individual for developing the disease (1.4%). These features suggest that a genetic predisposition may underlie multiple primary melanomas (MPMs). Prior studies had identified CDKN2A mutations in a few MPM individuals. The objectives of this study were to determine the frequency of family history of melanoma in MPM cases, to characterize other clinical features including history of other cancer, and to determine the association with functional CDKN2A mutations.

METHODS

This study used a case series design. All living patients who had been seen in the Pigmented Lesion Clinic at the University of Pennsylvania and who had more than one primary invasive malignant melanoma or an invasive primary followed by an in situ melanoma were eligible for participation.

RESULTS

Individuals with MPM frequently had a family history of melanoma, dysplastic nevi (DN), and/or another cancer including basal cell carcinoma (BCC), and squamous cell carcinoma breast cancer, and a personal history of DN, and basal cell carcinoma. Germline mutations in CDKN2A gene were identified in 8 of 96 MPM cases (8.3%, 95% confidence interval, 6.7-9.9%).

CONCLUSIONS

These data indicate that the presence of MPM is associated with a modest incidence of a family history of melanoma, DN, or BCC and a small association with CDKN2A mutations. Therefore, in addition to the MPM index case, other family members can benefit from screening and regular surveillance for melanoma, DN, and BCC.

A single nucleotide polymorphism in the 3'untranslated region of the CDKN2A gene is common in sporadic primary melanomas but mutations in the CDKN2B, CDKN2C, CDK4 and p53 genes are rare

In this report we present the results of mutational analysis of the CDKN2B, CDKN2C, CDK4, p53 genes and 5'UTR of the CDKN2A gene in a set of 44 sporadic primary melanomas, which had been earlier analysed for mutations in the CDKN2A (p16/p14(ARF)) gene. No tumour-associated mutations were detected except in 1 melanoma where we found a CC>T* deletion-mutation in the codon 151-152 (exon 5) of the p53 gene. On the basis of our preliminary results, we did extended genotyping of the 500 C>G and 540 C>T polymorphisms in the 3'UTR of the CDKN2A gene in 229 melanoma cases and 235 controls. The T-allele frequency (for 540 C>T polymorphism) in melanomas was significantly higher than in controls (0.14 vs. 0.08; chi(2) = 5.95, p = 0.01; OR = 1.71, 95%CI = 1.11-2.66). The heterozygote frequency for this polymorphism was 0.26 (59/229) in melanomas compared to 0.13 (30/235) in healthy controls (chi(2) = 11.4; p = 0.0007; OR = 2.34, 95% CI = 1.40-3.92). The frequency of the 500 C>G polymorphism in the 3'UTR in the CDKN2A gene was not significantly higher in melanomas compared to healthy controls. The 500 C>G polymorphism, however, was in linkage disequilibrium with approximately 50 kb apart the C>A intronic polymorphism in the CDKN2B gene (determined in 44 melanomas and 90 controls; Fisher exact test, p<0.0001). Finally, the sequence analysis of genomic DNA isolated from T cell lymphocytes of healthy individuals exhibited that the codon reported as last of exon 2 of the CDKN2C gene is rather the first codon of exon 3.

Mutations in the INK4a/ARF melanoma susceptibility locus functionally impair p14ARF

The INK4a/ARF locus encodes two cell cycle regulatory proteins, the cyclin-dependent kinase inhibitor, p16(INK4a), and the p53 activator, p14(ARF). Germline mutations in this locus are associated with melanoma susceptibility in 20-40% of multiple case melanoma families. Many of these mutations specifically impair p16(INK4a), whereas mutations uniquely targeting p14(ARF) are rare. Nevertheless, the importance of p14(ARF) has not been excluded because more than 40% of INK4a/ARF alterations affect p16(INK4a) and p14(ARF). We now report that p14(ARF) is functionally impaired in melanoma kindreds carrying INK4a/ARF mutations. Of the seven INK4a/ARF mutations tested, three altered the subcellular distribution of p14(ARF) and diminished the ability of p14(ARF) to activate the p53 pathway. This work establishes the importance of p14(ARF) in melanoma predisposition.

Sporadic multiple primary melanoma cases: CDKN2A germline mutations with a founder effect

Multiple primary cancers are one of the hallmarks of inherited predisposition. Outside the familial context, multiple primary tumors could be related either to germline de novo mutations or to low-penetrance mutations, in predisposing genes. We selected 100 patients who displayed multiple primary melanoma (MPM) without any known melanoma cases recorded within their families and looked for germline mutations in the two melanoma-predisposing genes identified to date, CDKN2A and CDK4 exon 2. Nine patients (9%) had germline mutations in CDKN2A, whereas none carried germline mutations in exon 2 of CDK4. Seven cases displayed a recurrent missense mutation, G101W, already described in more than 20 melanoma-prone families; one case carried a missense mutation never reported to date (P114S), and the last case was a carrier of a 6 bp insertion at nucleotide 57 resulting in a duplication of codons 18 and 19. To ascertain whether the G101W was a mutational hot spot for de novo mutations or a common founder mutation, we genotyped eight microsatellite markers flanking the CDKN2A gene. After allowing for recombination over time, haplotype sharing provided evidence for an original G101W mutation common to 6 out of 7 sporadic MPM cases. Therefore, it can be concluded that de novo germline CDKN2A mutations associated with MPM are rare.

Detection of three novel translocations and specific common chromosomal break sites in malignant melanoma by spectral karyotyping

Chromosomal aberrations in malignant melanoma cells have been reported using standard chromosome banding analysis and comparative genomic hybridization. To identify marker chromosomes and translocations that are difficult to characterize by standard banding analysis, 15 early passage malignant melanoma cell lines were examined using spectral karyotyping. All 15 tumor cell lines had lost all or part of 1p and 10q. Losses of material on chromosome arms 4p (12/15), 6q (12/15), 9p (15/15), 12p (13/15), 12q (13/15), 13q (11/15), and 19q (14/15) were the next most frequent events. Gain of chromosome arms 1q (11/15), 6p (13/15), and 20q11 (14/15) was also observed. Interestingly, we identified translocations der(12)t(12;20)(q15;q11), der(19)t(10;19)(q23;q13), and der(12)t(12;19)(q13;q13) in 4/15 tumors. Three recurring translocations involving four of the most frequent break points were detected. The identification of recurring translocations and unique chromosome break points in melanoma will aid in the identification of the genes that are important in the neoplastic process.

Expresson of vascular endothelial growth factor, its receptors (FLT-1, KDR) and TSP-1 related to microvessel density and patient outcome in vertical growth phase melanomas

Microvessel density (MVD) was estimated in a series of 202 vertical growth phase (VPG) melanomas and 68 corresponding metastases, using a marker for angiogenic endothelial cells (CD105) and Factor-VIII. The expression pattern of vascular endothelial growth factor (VEGF), FLT-1, KDR and thrombospondin-1 (TSP-1) was studied by immunohistochemistry, in situ hybridization and reverse-transcriptase polymerase chain reaction. CD105 stained significantly less vessels, but gave only limited additional prognostic information compared with Factor-VIII, and MVD was an independent prognostic factor for both markers. Ninety-eight percent of all cases showed expression of VEGF, and higher expression was found significantly more frequent in thinner and less vascularized tumors. Possible autocrine loops were suggested by co-expression of VEGF and its two receptors in tumor cells, and by a significant correlation between KDR and tumor cell proliferation (Ki-67) in the subgroup of thicker tumors. Staining of VEGF receptors in endothelium was not correlated with MVD. Strong expression of TSP-1 in tumor stroma was found in 43% of the primary tumors, and was significantly correlated with increased thickness, proliferation and MVD, as well as decreased survival. These data suggest that MVD is associated with prognosis in cutaneous melanomas, and that the VEGF system and particularly TSP-1 seem to be involved in the regulation of angiogenesis and progression of these tumors.

p16(MTS-1/CDKN2/INK4a) in cancer progression

Since its discovery as an inhibitor of cyclin-dependent kinases 4 and 6, the tumor suppressor p16 has continued to gain widespread importance in cancer. The high frequency of deletions of p16 in tumor cell lines first suggested an important role for p16 in carcinogenesis. This initial genetic evidence was subsequently strengthened by numerous studies documenting p16 inactivation in kindreds with familial melanoma. Moreover, a high frequency of p16 gene alterations was found in primary tumors, while recent studies have identified p16 promoter methylation as a major mechanism of tumor-suppressor-gene silencing. Additional insight into p16's role in cancer has come from the genetic analysis of precancerous lesions and various tissue culture models. It is now believed that loss of p16 is an early and often critical event in tumor progression. Consequently, p16 is a major tumor-suppressor gene whose frequent loss occurs early in many human cancers.

CDKNA2A mutation analysis, protein expression, and deletion mapping of chromosome 9p in conventional clear-cell renal carcinomas: evidence for a second tumor suppressor gene proximal to CDKN2A

Inactivation of tumor suppressor genes on chromosome 9p is considered a critical event in renal cell carcinoma pathogenesis. Alterations of CDKN2A on 9p21 have been reported in renal cancer cell lines, but their relevance for primary renal carcinomas is unclear. Loss of heterozygosity (LOH) was analyzed by using four polymorphic microsatellites at D9S970 (9p12-9p13), D9S171 (9p13), D9S1748 (9p21), and D9S156 (9p21) in 113 primary conventional clear-cell renal cell carcinomas (CRCCs). Allelic deletion was detected in 21 of 88 informative CRCCs (24%) with the highest rate of LOH being observed at D9S171 on 9p13 (20%). Chromosome 9p LOH was associated with short tumor-specific survival in stage pT3 RCC (P = 0.01). Fluorescence in situ hybridization analysis of 54 CRCCs revealed no homozygous CDKN2A deletions indicating that this mechanism of CDKN2A inactivation is rare in CRCC. Sequencing of 113 CRCCs showed that 13 tumors (12%) had a 24-bp deletion abrogating codons 4 through 11 of CDKN2A. Immunohistochemical CDKN2A expression was absent in normal renal tissue and was only detected in six of 382 CRCCs (1.5%) on a renal tumor microarray. These data suggest that CDKN2A alterations are present in a small subset of CRCCs and a second, yet unknown tumor suppressor gene proximal to the CDKN2A locus, may play a role in CRCC development.

Single nucleotide polymorphism analyses of the human proliferating cell nuclear antigen (pCNA) and flap endonuclease (FEN1) genes

The products of proliferating cell nuclear antigen (PCNA) and flap endonuclease (FEN1) genes are multifunctional proteins that are involved in DNA replication and damage repair. Yeast models suggest association of mutant forms of PCNA and FEN1 with genomic instability. In our study, we have determined the single nucleotide polymorphisms in human PCNA and FEN1 genes. We sequenced the coding region and adjacent noncoding region of both the PCNA and FEN1 genes in 120 alleles (60 individuals). In the PCNA gene, we detected 9 sequence variants with Hardy-Weinberg allele frequency ranging from 0.008 to 0.088. No polymorphism was detected in the FEN1 gene. The sequence variants in the PCNA gene included 7 intronic single nucleotide polymorphisms (SNP) and 2 synonymous exonic SNPs. All the intronic SNPs were located in introns 1 and 4, which contain several regulatory elements involved in the control of PCNA gene expression. Six of the 7 intronic SNPs showed complete linkage disequilibrium. We confirmed this allelic linkage disequilibrium by allele-specific PCR sequencing. We genotyped 117 additional individuals belonging to 3 population subgroups using the PCR-RFLP method. Finally, to see if the detected polymorphisms are associated with any cancer type, we genotyped cases with melanomas (37 cases), breast cancers (118 cases) and lung cancers (100 cases). We did not find statistical difference in frequency of polymorphism in any cancer type compared with healthy controls, although in breast cancer the frequency was low. Our results suggest that the coding regions of the PCNA and FEN1 genes are highly conserved when compared with other DNA repair genes. The potential of some of the detected intronic polymorphisms to effect regulation of the PCNA gene expression remains to be determined.

Analysis of G(1)/S checkpoint regulators in metastatic melanoma

We have analyzed the expression of the CDKN1A (p21(CIP1)), CDKN1B (p27(Kip1)), TP53, RB1 and MDM2 proteins and tumor cell proliferation by immunohistochemical staining in 59 cases of metastatic melanoma. The genomic status of the CDKN2A (INK4-ARF, p16/p14(ARF)), CDKN2B (p15) and CDKN2C (p18) genes was determined by PCR-SSCP (single-strand conformation polymorphism) in 46 of these cases. These results were correlated with various clinico-pathological parameters, including the outcome of combined chemoimmunotherapy. We found positive correlations between the expression of CDKN1A and MDM2 (r = 0.5063, P = 0.001), between the expression of CDKN1B and RB1 (r = 0.5026, P = 0.001), and between RB1 expression and tumor cell proliferation (0.5564, P<0.001). Two mutations in the CDKN2A (p16) gene were detected, including a novel base change AAC-->ATC (Asn to Ile) at codon 71, that also changes the codon 85 of the alternative reading frame gene p14(ARF) from CAA to CAT (Gln to His). Homozygous deletion at exon 2 of the CDKN2A (INK4-ARF) gene was detected in six cases. In seven cases, the 540C-->G polymorphism in the 3'UTR of the CDKN2A (p16) gene was found in linkage disequilibrium with the 74C-->A polymorphism in intron 1 of the CDKN2B gene (P < 0.0001). These cases had significantly lower expression of the TP53 protein (P = 0.0032). Both 540C-->G and 580C-->T polymorphisms in the 3'UTR of the CDKN2A (p16) gene were associated with significantly shorter progression time from primary to metastatic disease (P = 0.0071). We conclude, that although none of the analyzed cell cycle regulators could be singled out as a major prognostic factor, G(1)/S checkpoint abnormalities remain one of the most significant factors in the development of malignant melanoma.

INK4a/ARF locus alterations in human non-Hodgkin's lymphomas mainly occur in tumors with wild-type p53 gene

INK4a/ARF locus codes for two different proteins, p16(INK4a) and p14(ARF), involved in cell cycle regulation. p14(ARF) is considered an upstream regulator of p53 function. To determine the role of these genes in the pathogenesis of human non-Hodgkin's lymphomas we have analyzed exon 1beta, 1alpha, and 2 of the INK4a/ARF locus and p53 gene aberrations in 97 tumors previously characterized for p16(INK4a) alterations. p53 alterations were detected in four of 51 (8%) indolent lymphomas but in 15 of 46 (33%) aggressive tumors. Inactivation of p14(ARF) was always associated with p16(INK4a) alterations. Exon 1beta was concomitantly deleted with exon 1alpha and 2 in eight tumors. One additional lymphoblastic lymphoma showed deletion of exon 1alpha and 2 but retained exon 1beta. No mutations were detected in exon 1alpha and 1beta in any case. Two of the three mutations detected in exon 2 caused a nonsense mutation in the p16(INK4a) reading frame and a missense mutation in the ARF reading frame involving the nucleolar transport domain of the protein. The third mutation was a missense mutation in the p16(INK4a) reading frame, but it was outside the coding region of p14(ARF). Aggressive lymphomas with p14(ARF) inactivation and p53 wild type showed a significantly lower p53 protein expression than tumors with no alteration in any of these genes. In this series of tumors, inactivation of the INK4a/ARF locus mainly occurred in tumors with a wild-type p53 gene because only two lymphomas showed simultaneous aberrations in these genes. Tumors with concomitant alterations of p16(INK4a) and p14(ARF)/p53 genes seem to exhibit a worse clinical behavior than lymphomas with no alterations or isolated inactivation of any of these genes. These findings indicate that p14(ARF) genetic alterations occur in a subset of aggressive NHLs, but they are always associated with p16(INK4a) aberrations. Concomitant disruption of p16(INK4a) and p14(ARF)/p53 regulatory pathways may have a cooperative effect in the progression of these tumors.

Melanoma genetics: an update with focus on the CDKN2A(p16)/ARF tumor suppressors

Investigative interest in atypical nevi and familial melanoma has contributed to the identification of several candidate melanoma loci within the human genome. Molecular defects in both tumor suppressor genes and oncogenes have been pathogenically linked to melanoma in recent studies. Of the loci currently characterized, the major gene resides on chromosome 9p and encodes a tumor suppressor designated p16. This gene, which is also known as CDKN2A, is either mutated or deleted in a large majority of melanoma cell lines, as well as in many uncultured melanoma cells and in the germline of melanoma kindreds. A novel aspect of the p16 locus is that it encodes not just one but two separate gene products that are transcribed in alternative reading frames. Both products function as negative regulators of cell cycle progression. The p16 protein itself executes its effects by competitively inhibiting cyclin-dependent kinase 4, which is a factor necessary for cellular progression through a major regulatory transition of the cell division cycle. Inherited and acquired deletions or point mutations in the p16 gene increase the likelihood that potentially mutagenic DNA damage will escape repair before cell division. Notably, the second product of the locus, ARF (for alternative reading frame), regulates cell growth through independent effects on the p53 pathway. Although there is little evidence that ARF by itself is involved in the pathogenesis of melanoma, deletions at the p16 locus disable two separate pathways that control cell growth. These recent advances open up the possibility of genetic testing for melanoma susceptibility in the setting of familial melanoma and suggest novel therapeutic strategies for melanoma based on gene therapy or small molecule mimicry targeted to the correction of defects in the p16 regulatory pathway. (J Am Acad Dermatol 2000;42:705-22.)

LEARNING OBJECTIVE

At the conclusion of this learning activity, participants should be familiar with the historical aspects of melanoma genetics and should have a greater understanding of the CDKN2A(p16)/ARF tumor suppressor genes.

Functional reassessment of P16 variants using a transfection-based assay

CDKN2A appears to be the major melanoma susceptibility gene, and is also mutated/deleted in sporadic tumours of various types including melanoma. Thus far most approaches to assessing the functionality of mutations in this gene have used in vitro methods such as CDK4 binding and kinase inhibition assays, with sometimes disparate conclusions about functional significance of some variants between studies. We have used a melanoma cell line (MM96L) with no functional p16, as the basis for a "semi-in vivo" transfection-based assay for exogenous p16 functionality based on the growth parameters of the cells and the behaviour of variant proteins after transfection of different CDKN2A cDNAs. Colony counts performed on these transfectants revealed that all but the wild type, + 24 bp ad A148T variants have a diminished ability to inhibit cell growth. All other variants detected either constitutionally in familial melanoma patients (I49T, R87P, G101W and V126D) or somatically in melanomas (N71S, and P81L), appeared functionally impaired in this assay. This diminution of function was independent of CDK4 and CDK6 binding ability. Furthermore, the predominant localization of these variants within the cell was different from that of wt p16. This mislocalization may provide an explanation for their lack of function, or alternatively, it may also be an indicator that the cells are processing unstable, misfolded p16 proteins. This novel assay for assessment of functionality of p16 variants may better reflect the role of some of these mutations in vivo, and as such is a useful adjunct to other in vitro assays.

Selective deletion of exon 1 beta of the p19ARF gene in metastatic melanoma cell lines

The INK4A locus on 9p21 is deleted or rearranged in a large number of human cancers. The locus encodes two unrelated and independently acting negative cell-cycle regulators, p16 and p19ARF, arising in alternate reading frames from a partly shared sequence. We analyzed five human melanoma cell lines for deletions at the INK4 loci and flanking microsatellite markers on 9p21. All the cell lines displayed deletions of varying sizes. The metastatic cell line IGR-1 showed a large deletion between the markers D9S736 and D9S171. In the cell lines WM-115 and WM-266-4, the deletion included exon 1alpha of p16, exon 1beta of p19ARF, and exon 2 of the INK4B (p15) gene. Two cell lines, SK-MEL-5 and A2058, had deletions confined to exon 1beta and the microsatellite marker D9S942. RT-PCR experiments showed the presence of the p16 and p15 transcripts and absence of p19ARF expression in both SK-MEL-5 and A2058 cell lines. The selective loss of the exon 1beta of p19ARF and retention of the p16 and p15 genes and their expressions in these two cell lines support the putative tumor suppressor role for the alternate reading frame p19ARF gene.

Cloning and characterization of the CDKN2A and p19ARF genes from Monodelphis domestica

The tumor suppressor gene, CDKN2A (p16), encodes a cyclin-dependent kinase inhibitor and functions as a negative regulator in the retinoblastoma pathway that blocks cell cycle progression from the G1 phase. The gene has been found to be deleted, truncated, mutated, or silenced by promoter methylation in a wide range of tumor types. Where melanoma CDKN2A mutations have been characterized, C --> T and CC --> TT transitions were found, indicating a direct role for ultraviolet radiation (UVR)-induced pyrimidine dimers in the formation of some tumors. The South American opossum, Monodelphis domestica, has been shown by our group and others to be susceptible to the induction of melanoma on chronic exposure to UVR alone. The CDKN2A gene and its exon 1beta alternate transcript p19ARF were cloned and sequenced from M. domestica to investigate the role of these genes in the development of UVR-induced melanoma and non-melanoma tumors. Both genes were first amplified by polymerase chain reaction (PCR) using cDNA from an opossum corneal-tumor cell-line library and degenerate primers based on human, mouse, and rat CDKN2A gene sequences. To verify these as normal sequences, both genes were then RT-PCR amplified from cultured normal opossum melanocyte mRNA. When comparing the tumor and melanocyte sequences, we found a UVR signature point mutation, a C --> T transition, within exon 2 in the corneal tumor cell line. The same mutation at this site in other tumors has been shown to alter the CDKN2A protein's ability to bind CDK4 kinase, which may lead to uncontrolled cell cycling. A comparison of the amino acid sequence of opossum CDKN2A showed identities relative to human, mouse, and rat between 57% and 63%, and when conserved amino acid substitutions are considered (similarity), the range is 63% to 67%. The amino acid identity and similarity for p19ARF ranged from 39% to 49%.