Skip to main content
NCBI home page
Journal of Cancer Research and Clinical Oncology logoLink to Journal of Cancer Research and Clinical Oncology
. 2003 Sep 3;129(11):642–650. doi: 10.1007/s00432-003-0485-z

Differential alterations of the genes in the CDKN2A-CCND1-CDK4-RB1 pathway are associated with the development of head and neck squamous cell carcinoma in Indian patients

Anusri Tripathi (Bhar) 1, Soma Banerjee 2, Neelanjana Chunder 1, Anup Roy 3, Arunava Sengupta 4, Bidyut Roy 5, Susanta Roychowdhury 2, Chinmay Kumar Panda 1,✉
PMCID: PMC12161921  PMID: 14586645

Abstract

Purpose

The aim of this study was to analyse the alterations of the genes in the CDKN2A/CCND1/CDK4/RB1 pathway in the G1-S phase of the cell cycle during development of head and neck squamous cell carcinoma (HNSCC).

Methods

The alterations of these genes were analysed in 22 dysplastic lesions, 26 stage-I/II and 33 stage-III/IV HNSCC tumours of Indian patients.

Results

The alterations [mutation, hypermethylation, homozygous deletion and loss of heterozygosity/microsatellite size alteration (LOH/MA)] in the CDKN2A were found to be highest in 57% of the samples, followed by CCND1 amplification and LOH/MA at the RB1 locus in 14% and 8.5% of the samples, respectively. No dominant CDK4 Arg24Cys mutation was seen in our samples. Comparatively high frequency of CDKN2A alterations (except homozygous deletion) was found in dysplastic head and neck lesions and remained almost constant or increased during progression of the tumour, whereas the homozygous deletion of CDKN2A and the alterations in CCND1 and RB1 genes were seen mainly in the later stages of the tumour.

Conclusions

Our study suggested that mutation/hypermethylation/allelic alterations (LOH/MA) of CDKN2A were associated with the development of dysplastic head and neck lesions. All the other alterations might provide some cumulative effect during progression of later stages of the tumour to have selective growth advantages.

Keywords: Head and neck squamous cell carcinoma, CDKN2A-CCND1-CDK4-RB1, Mutation, Hypermethylation, Homozygous deletion, Loss of heterozygosity, Microsatellite size alteration

Introduction

Cell cycle progression is regulated by complexes formed between cyclins and cyclin-dependent kinases (cdks). The kinase activity of cdk4 and cdk6 is dependent on their association with cyclin D1 at the G1-S phase of the cell cycle (Lunderberg and Weinberg 1999). The cyclin D1–cdk4/6 complexes phosphorylate the retinoblastoma protein (pRb), resulting in the release of ubiquitous transcription factor E2F for cell cycle progression. The activity of cyclin D1-cdk4/6 complexes is blocked by cdk inhibitors, including p15, p16, p21, p27. The p16 binds with cdk4/6 and blocks their interaction with cyclin D1. Thus, any deregulation in the cell cycle regulatory pathway involving p16/cdk4/cyclin D1/pRb may result in imbalance in the cell cycle and genomic stability, eventually leading to tumorigenesis.

In India and South East Asia, HNSCC is one of the leading cancers, accounting for approximately 40% of all malignancies (Saranath et al. 1993). Accumulating molecular studies have shown that there is a great variability in the frequency of alterations of the CDKN2A (encoding p16), CCND1 (encoding cyclin D1) and RB1 (encoding pRb) genes in HNSCC tumours. The mutation, promoter hypermethylation, and homozygous deletion of CDKN2A gene, have been found in 6–16% (Lin et al. 2000; Zhang et al. 1994), 23–50% (Wu et al. 1999; Shintani et al. 2001) and 0–78% (Lin et al. 2000; Cerilli et al. 1999; Shahnavaz et al. 2001) of HNSCC tumours, respectively. Microsatellite markers from the CDKN2A locus at chromosome (chr.) 9p21 have shown loss of heterozygosity (LOH) in 31–72% of the HNSCC tumours (Wu et al. 1999; van der Riet et al. 1994). Also, in this region, high frequencies of microsatellite size alteration (MA) were seen in HNSCC, indicating the genetic instability at this locus (Shahnavaz et al. 2001). This variability in the frequencies of CDKN2A alterations among different studies might be due to differences in the methodologies used, sample preparation, or histological differences in the HNSCC tumours analysed. In pre-malignant head and neck lesions the frequencies of CDKN2A mutation and homozygous deletion have been found to be very low (Matsuda et al. 1996; Papadimitrakopoulou et al. 1997), but the presence of LOH at chr.9p21 and the absence of p16 protein (as detected by immunohistochemistry) in hyperplastic head and neck lesions led Mao and El-Naggar (1999) to suggest that CDKN2A inactivation might be associated with the development of hyperplastic head and neck lesions from normal mucosa. On the other hand, Nagai (1999) proposed that CDKN2A inactivation might be involved in the development of carcinoma-in-situ from dysplastic head and neck lesions. Thus, it seems that CDKN2A would function as either an initiating tumour suppressor gene (TSG) or a progression TSG in HNSCC, along with other genetic alterations. The frequency of amplification of CCND1 in the HNSCCs varied from 16–37%, with copy number ranging between twofold and tenfold (Callender et al. 1994). However, there is ambiguity as to whether CCND1 amplification occurs at early or advanced stages of the tumour (Mao and El-Naggar 1999; Kyomoto et al. 1997). Previous studies have revealed LOH at the RB1 locus in 27–48% of HNSCC samples (Li et al. 1994; Yoo et al. 1994), but its involvement in the development of pre-malignant head and neck lesions has not yet been explored in detail. The dominant CDK4 Arg24Cys mutation seen in melanoma, which specifically inhibits the interaction between p16 and cdk4 (Zuo et al. 1996), has not yet been studied in HNSCC samples; thus, the alterations in any one of these genes, i.e. CCND1, CDK4 and RB1, might have the same effect as that of the inactivation of CDKN2A. On the other hand, it might also be possible that the alterations of these genes could have an additive effect, along with the inactivation of CDKN2A. However, to rule out the ambiguities about the timing of alteration of these genes, observed among the different investigators, no one has made any comparative analysis of the alterations of these genes, i.e. CDKN2A, CCND1, CDK4 and RB1, at different stages of HNSCC development in the same set of samples.

In the present study we have analysed the prevalence of CDKN2A alterations, CCND1 amplification, and CDK4-Arg24Cys mutation and deletion at the RB1 locus in 22 dysplastic head and neck lesions, 26 stage-I/II, and 33 stage-III/IV HNSCC tumours from Indian patients to find out the involvement of these genes during the development of HNSCC.

Materials and methods

Sample collection and clinical data

Specimens from the head and neck region of 81 patients undergoing operation were collected, together with the corresponding normal tissue or peripheral blood leukocytes (PBLs), at Chittaranjan National Cancer Institute and the Cancer Centre and Welfare Home, Calcutta. All samples were collected after consent had been obtained from the patients and hospital authorities. The samples were frozen immediately after collection and stored at −80°C until required. The detailed clinical history of the patients is presented in Table 1. Among the 12 orofacial samples, four were from the maxilla, three were from the mandible and nasal cavity each and two were collected from the cheek. The tumours were graded and staged according to the UICC TNM classification (4th edition, 1987). Approximately 81% (66/81) of the patients were male, with mean age of 49 years, and 19% (15/81) were female, with mean age of 46 years. The patients were considered as tobacco habituated if they had consumed at least 10–15 cigarettes or the equivalent amount of chewable tobacco per day for at least 10 years.

Table 1.

Clinicopathological features of head and neck lesions

Clinical features No. of patient Mean age Age range
(years) (years)
Primary site
Oral cavity 61 47 8–70
Larynx 08 61 48–75
Orofacial 12 47 37–70
Clinical stage
Dysplasia 22 47 32–70
TNM stage I 04 55 45–65
TNM stage II 22 46 8–75
TNM stage III 20 52 37–73
TNM stage IV 13 49 26–70
Gender
Male 66 49 8–75
Female 15 46 26–60
Tumour differentiation
Well 29 51 30–73
Moderate 24 47 8–70
Poor 06 51 32–75
Lymph node
Positive 28 46 26–70
Negative 31a 53 8–75
Tobacco
Tobacco + 47 50 26–73
Tobacco − 34 47 8–70
Open in a new tab

aExcluding dysplasia

DNA extraction

We removed the normal cells present as contaminants in the dysplastic head and neck lesions and HNSCC tumours by micro-dissection (Dasgupta et al. 2002), using a cryostat (Leica CM 1800, Germany). We took extreme care to avoid cross-contamination among the samples. DNA was extracted from the micro-dissected tissue sections and their corresponding normal tissue or PBL by proteinase-K digestion followed by phenol: chloroform extraction (Dasgupta et al. 2002).

Mutation analysis of CDKN2A gene

DNA samples from all the head and neck lesions were screened for mutation in exon-1 and exon-2 of the CDKN2A gene by single-strand conformation polymorphism (SSCP) analysis, according to the method described by Heyman et al. (1996). All the PCR reactions were performed in 10 μl reaction volume containing 1× PCR buffer [20 mmol/l Tris-HCl (pH 8.4), 50 mmol/l KCl], 1 mmol/l MgCl2, 0.2 mmol/l of each dNTP, 5 pmol of each primer, 0.5 units of Taq DNA polymerase (Gibco BRL, USA), 50–100 ng of template DNA, and 5% and 10% DMSO for CDKN2A exon-1 and exon-2 analysis, respectively, at 57°C annealing temperature. The sequences of the primers used in this analysis were as follows: CDKN2A exon-1 (2F: 5′-GAAGAAAGAGGAGGGGCTG-3′ and 1108R: 5′-GCGCTACCTGATTCCAATTC-3′) and CDKN2A exon-2 (5′-GGAAATTGGAAACTGGAAGC-3′ and 5′-TCTGAGCTTTGGAAGCTCT-3′). We obtained radiolabelled PCR products by substituting 0.2 μl of water with [α32P] dCTP (specific activity 3,000 Ci/mmol; NEN, USA) in 10 μl reaction volume. The PCR products were diluted to a final volume of 15 μl, including 1.5 μl of appropriate buffer, and digested with SmaI (Heyman et al. 1996). Three microlitres of this SmaI-digested sample was added to 17 μl of stop solution [95% vol/vol formamide, 20 mmol/l EDTA (pH 8), 0.05% bromophenol blue and 0.05% xylene cyanol], heated to 95°C for 10 min and quickly chilled in ice. Of this mixture, 2 μl was electrophoresed overnight in SSCP gels containing 6% acrylamide (acrylamide:bisacrylamide = 49:1), 1×TBE with 10% glycerol or without glycerol at 2 W. The glycerol-containing gels were run at 15–20°C and the non-glycerol gels were run at 4°C. After electrophoresis, the gels were transferred to Whatman 3MM paper, dried, and autoradiographed on X-ray film (Kodak). DNA samples showing bands of altered mobility were subjected to direct nucleotide sequencing in an automated DNA sequencer (ABI 377). Samples were determined to be mutation positive if both forward and reverse strands exhibited the same mutation.

Analysis of methylation status of the CDKN2A promoter region

Approximately 100 ng of genomic DNA was digested overnight with Hpa II and Msp I (Gibco BRL), separately, using 30 units of enzyme per reaction (Klangby et al. 1998). Another mock digestion was performed with each sample, where water was added instead of the enzymes. The digested DNA was alcohol precipitated, dried, and dissolved in 10 μl water for PCR amplification in 20 μl reaction volume, as described above, using the same CDKN2A exon-1 primers (i.e. 2F and 1108R). PCR was performed for 30 cycles at 57°C annealing temperature, and the PCR product was then run on 2% agarose gel, stained with ethidium bromide, visualized under a UV-transilluminator and photographed. The gel was transferred to nylon membrane (Gene Screen, NEN, USA) and hybridized with random 32P-labelled CDKN2A cDNA probe (Klangby et al. 1998) for final confirmation of the PCR product. All the samples showing CDKN2A methylation were reproduced twice.

Homozygous deletion analysis of the CDKN2A gene

Homozygous deletion of CDKN2A gene was confirmed by comparative multiplex PCR involving amplification with two different sets of primer pairs in the same reaction mixture (Liew et al. 1999). A D9S169 marker set (https://www.gdb.org/), located approximately 11 Mbp centromeric to CDKN2A, was used as a control to test for homozygous deletion of CDKN2A exon-2. The primers for exon-2 were the same as used in our SSCP analysis. Forward primers of D9S169 marker and CDKN2A exon-2 were end labelled with [γ32P] ATP (specific activity 3,000 Ci/mmol; NEN) using T4 polynucleotide kinase (Gibco BRL). PCR was performed for 30 cycles at 57°C annealing temperature in 20 μl reaction volume containing 1× PCR buffer [20 mmol/l Tris-HCl (pH 8.4), 50 mmol/l KCl], 1 mmol/l MgCl2, 0.2 mmol/l of each dNTP, 5 pmol of each primer, 0.5 units of Taq DNA polymerase (Gibco BRL), 50–100 ng of template DNA and 10% DMSO. Then, 1 μl of the radiolabelled PCR product was mixed with 4 μl of the stop solution (described above), heated at 95°C for 5 min, and chilled on ice. Three microlitres of the mixture was loaded on 7% polyacrylamide denaturing sequencing gel containing 8 mol/l urea for electrophoresis at 50 W for 3 h. After electrophoresis, the gels were exposed to X-ray film (Kodak) and kept at −80°C for 24 h. All the comparative multiplex PCRs for the tumours with homozygous deletion were reproduced at least twice. Results were obtained by comparison of the allele intensities in matched normal/tumour DNA by densitometric scanning (Shimadzu, CS-9000). Homozygous deletion was scored if the signal intensity of the CDKN2A in tumour tissue was at least ten-times less than the signal from the normal tissue, whereas the intensity of the D9S169 control allele was approximately equal in both the tumour and normal tissues (Liew et al. 1999).

Analysis of D9S942 status

We analysed the D9S942 microsatellite marker (https://www.gdb.org/), located 5 kb upstream of CDKN2A exon-1, to determine any LOH or MA of the CDKN2A locus. PCR was performed in 20 μl reaction volume, as described by Dasgupta et al. (2002), which contained 1× PCR buffer [67 mmol/l Tris (pH 8.7), 16.6 mmol/l (NH4)2SO4, 0.01% Tween-20], 1 mmol/l MgCl2, 0.2 mmol/l of each dNTP, 5 pmol of each primer, 0.5 units of Taq DNA polymerase (Gibco BRL) and 50–100 ng of template DNA. The forward primer of D9S942 was end labelled with [γ32P] ATP (specific activity 3,000 Ci/mmol; NEN) using T4 polynucleotide kinase (Gibco BRL). The PCR was carried out at 58°C annealing temperature for 30 cycles. The amplified products were separated by 7% polyacrylamide denaturing gels and autoradiographed as described in the homozygous deletion analysis.

The LOH and MA were detected by densitometric scanning as described by Dasgupta et al. (2002) and MacGrogan et al. (1994). The allele loss was recorded if there was a complete absence of one allele or if the relative band intensity of one allele was reduced by at least 50% in the tumour in comparison with the homologous allele in the corresponding normal DNA. The MAs were detected by a shift in the mobility of one (MA1) or both (MA2) alleles in comparison with the same allele in the corresponding normal DNA. MA in one allele and loss of the other allele was regarded as LMA.

Analysis of CCND1 amplification

We measured the amplification of CCND1 (located at 11q13) by differential PCR, using dopamine-D2-receptor gene (DRD2) as reference sequence located at 11q22–23 (Kyomoto et al. 1997). We used the DRD2 gene to avoid the aneuploidy problem of chromosome 11. The PCR reaction was performed in a 20-μl reaction volume containing 1× buffer (Dasgupta et al. 2002), 1 mmol/l MgCl2, 0.2 mmol/l of each dNTP, 4 pmol of each of CCND1 and DRD2 primers (Kyomoto et al. 1997), 0.5 units of Taq DNA polymerase (Gibco BRL), 0.1 μl of [α32P] dCTP (specific activity 3,000 Ci/mmol; NEN) and 50–100 ng of template DNA. The PCR reaction was carried out for 30 cycles at 64°C annealing temperature. The CCND1 primer sequences were as follows- 5′-ACCAGCTCCTGTGCTGCGAAGTG-3′ and 5′-GACGGCAGGACCTCCTTCTGCACA-3′. The DRD2 primer sequences were as follows: 5′-TGATGATGATCTGGAGAGGCAGAAC-3′ and 5′-TGCCGAAGACGATGACAGCGATGAG-3′. The radiolabelled PCR products were electrophoresed in denaturing gel, autoradiographed, and densitometrically scanned as described above. CCND1 was considered to be amplified if the CCND1/DRD2 index of tumour was more than twice that of the corresponding normal sample (Kyomoto et al. 1997).

CDK4 mutation analysis

The p16 binding to cdk4 is specifically affected by a C-to-T transition at nucleotide 297 of CDK4 exon-2 (Gene Bank accession number M14505). The C-to-T transition creates a novel StuI site by changing the sequence from AGGCCC to AGGCCT (Zuo et al. 1996). Hence, this transition was checked by PCR followed by StuI restriction digestion. The PCR was performed as described in the CCND1 amplification analysis in 20 μl reaction volume containing 1 mmol/l MgCl2 and 5 pmol of each CDK4 primer (Zuo et al. 1996) for 30 cycles at 60°C annealing temperature. The CDK4 primer sequences were as follows: 5′-GCTGCAGGCTCATACCATCCT-3′ and 5′-CTCTCACACTCTTGAGGGCC-3′. The PCR product was digested with StuI in a final reaction volume of 25 μl. The digested PCR product was electrophoresed in 2% agarose gel and stained with ethidium bromide for visualization under UV light. In this primer set, the C-to-T transition produces 149 bp and 49 bp fragments instead of 198 bp PCR product after StuI digestion.

LOH analysis of RB1 locus

The LOH at the RB1 locus was determined by PCR analysis using the intron-20 VNTR marker, Rb1 (Li et al. 1994). The PCR was performed in 20 μl reaction volume containing 1× PCR buffer [50 mmol/l Tris-HCl, 10 mmol/l KCl, 5 mmol/l (NH4)2SO4, pH 8.3], 2 mmol/l MgCl2, 0.2 mmol/l of each dNTP, 5 pmol of each primer, 2 units of FastStart Taq DNA polymerase (Roche, Germany) and 50–100 ng of template DNA. PCR was performed in two-step annealing temperatures, firstly at 61°C for 5 cycles and then at 57°C for 25 cycles.

Statistical analysis of the clinical data

To determine the association between the alterations with different clinicopathological features of the tumours, we performed chi-square analysis. Probability values of P<0.05 were regarded as statistically significant.

Results and discussion

To understand the role of the CDKN2A/CCND1/CDK4/RB1 pathway during HNSCC development, we analysed some alterations of these genes in 22 dysplastic head and neck lesions, 26 stage-I/II and 33 stage-III/IV HNSCC samples (Table 1). The mutation, promoter hypermethylation, homozygous deletion and LOH/MA in the CDKN2A gene were seen in 57% (46/81) of the samples (Table 2 and Figs. 1 and 2). In SSCP analysis, abnormal band shifts were detected in only exon-2 of 15 samples (Fig. 1). To determine the types of mutations, we carried out direct sequencing in these samples. Two samples (nos. 149 and 326) showed Ala148Thr polymorphism, which we confirmed by sequencing normal DNA of the corresponding patients. This polymorphism was also reported by other investigators, in melanoma and ovarian carcinoma (Pollock et al. 1996; Schuyer et al. 1996). Sample no. 326 also harboured a point mutation at codon 114. The G→C transversion at the –4 position of the intron-1/exon-2 splice junction was detected in four samples. In the remainder of the samples only transition mutations were scattered between codons 58 to 122 (Table 3). The frequencies of transition mutations were in the following order: C→T>G→A>T→C. Similar to the CDKN2A mutations reported in HNSCC (Kannan et al. 2000), melanoma and other tumours (Pollock et al. 1996), some common mutations, i.e. G→C transversion at the –4 position of the intron-1/exon-2-splice junction, Pro114Leu, and Arg58Term, were also identified in our samples. Other mutations detected in our samples affected the same codons that were frequently altered in melanoma, bladder carcinoma and HNSCC, but with different nucleotide changes (Pollock et al. 1996). This type of phenomenon has also been seen in other types of tumours (Pollock et al. 1996; Park et al. 1999). Similar to our result of 17% (10/58) CDKN2A mutation in HNSCC tumours, Zhang et al. (1994) detected 16% mutation in both exon-1 and exon-2 of the CDKN2A gene in primary HNSCC tumours. Also, 9–11% P16 mutation has been reported by other investigators in primary HNSCC samples including some samples of Indian origin (Kannan et al. 2000). However, no statistical correlation has been found between CDKN2A mutation and tobacco intake, nodal involvement, and progression of the disease (Table 4) in our analysis. Interestingly, the frequency of mutation remained almost constant (approximately 18%) during progression of the tumour, i.e. from dysplasia to stage-III/IV (Fig. 3). This is the first report demonstrating CDKN2A mutation to be one of the early events necessary for the development of dysplastic head and neck lesions. On the other hand, Matsuda et al. (1996) found no CDKN2A mutation in pre-malignant lesions of the head and neck. Papadimitrakopoulou et al. (1997) identified a non-sense mutation in only one of the 74 oral pre-malignant lesions analysed.

Table 2.

Alterations of CDKN2A/CCND1/CDK4/RB1genes in head and neck lesions (+ Positive for the particular alteration, − negative for the particular alteration, P CDKN2A polymorphism, ND not done, L loss of heterozygosity, MA1 microsatellite size alteration of one allele, MA2 microsatellite size alterations of both alleles, LMA loss of one allele and size alteration of the other allele, RH retention of heterozygosity, NI non-informative, AMP amplification, D dysplastic lesions)

Parameter L15 L25 4248 2024 L13 L14 L26 L28 L33 L47 L38 L39 4188 292 222 7428 1047 558 6290 1295 7216 5959 410 939 3216 825 149 2592 1367 326 7059 5325 308 4332 2086 7783 693 5364 1068 311 5011 2618 802 2030 5184 1445 4904 1491 7286
CDKN2A mutation Exon-1 − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − −
Exon-2 + − + + + + + + + P P,+ + + + + +
CDKN2A methylation − − + − − + − + − − − + + − − + − + + − − + + − − − − + + − − − + − + − − − − + + − − + − − ND − −
CDKN2A homozygous deletion − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − + − − − + + − − − − − −
D9S942 status MA2 L RH NI LMA RH MA2 L RH RH MA2 L RH L RH RH RH RH L L RH RH RH RH MA2 MA2 LMA RH RH RH L L RH MA2 LMA MA2 MA1 MA1 ND RH NI RH RH L MA2 MA2 MA1 MA2 L
CCND1 amplification AMP AMP AMP AMP ND AMP AMP AMP AMP AMP AMP AMP
CDK4 Arg24Cys mutation − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − −
RB1 status NI RH RH NI RH RH NI ND RH ND NI NI L NI NI NI RH RH ND NI NI ND NI NI RH L NI NI RH NI RH NI NI NI NI RH NI MA2 NI NI NI L NI RH L NI NI NI L
Stage D D D D D D D D D D D D I I I II II II II II II II II II II II II III III III III III III III III III III III III IV IV IV IV IV IV IV IV IV IV
Open in a new tab

Fig. 1.

Fig. 1.

Open in a new tab

SSCP analysis of CDKN2A exon-2. Lane 1: normal DNA; lanes 2–9: tumour DNAs of samples 939, L15, 7286, 4904, L47, 7216, 326, 1047 respectively. Arrows indicate the band shift

Fig. 2A–E.

Fig. 2A–E.

Open in a new tab

Representative photograph of the alterations of CDKN2A, CCND1 and RB1 locus in head and neck lesions. A Analysis of CDKN2A methylation. U mock digestion, M Msp I digestion, H Hpa II digestion. B Allelotyping of the D9S942 locus. LOH loss of heterozygosity, MA1 microsatellite size alteration of one allele, MA2 microsatellite size alteration of both alleles, LOH+MA loss of one allele and size alteration of the other allele (LMA). Arrows indicate the loss of the corresponding allele. Asterisks indicate the size alteration of the corresponding allele. C Analysis of homozygous deletion of the CDKN2A locus. Asterisk indicates homozygous deletion of the CDKN2A locus. D Analysis of CCND1 amplification. Asterisk indicates the amplification of the CCND1 locus. E Allelotyping of RB1 locus. Arrow indicates the loss of the corresponding allele, Asterisks indicate the size alteration of the allele. T tumour DNA of the corresponding patient, N normal DNA of the corresponding patient

Table 3.

Compilation of the mutations and polymorphisms at CDKN2A exon-2 in head and neck lesions

Sample Codon CDKN2A
Nucleotide change Amino acid altered
L26, 1047, 7216, 4904 −4 Position of intron-1/exon-2 splice junction G→C –
5325, 222 58 CGA→TGA Arg to Term
939, 7286 108 GAT→AAT Asp to Asn
L47 113 CTG→CCG Leu to Pro
326, 693 114 CCC→CTC Pro to Leu
308 118 GCT→GTT Ala to Val
L13, L25 122 GGC→GAC Gly to Asp
149, 326 148 GCG→ACG Ala to Thr
Open in a new tab

Table 4.

Clinicopathological correlation of the alterations of CDKN2A, CCND1 and RB1 in head and neck lesions

Parameter CDKN2A mutation P CDKN2A methylation P CDKN2A homo-del P D9S942 alterations P CCND1 amplification P RB1 amplification P
+ − + − + − + − + − + −
Tobacco+ 9 38 0.601 8 40 0.219 2 46 0.739 20 22 0.009 8 37 0.447 3 36 0.860
Tobacco− 5 29 9 23 1 32 6 26 3 30 3 29
Node+ 3 25 0.407 6 21 0.879 3 25 0.070 10 14 0.418 6 20 0.107 3 24 0.847
Node− 11 42 11 42 0 53 16 34 5 47 3 41
Dysplasia 4 18 0.971 4 18 0.686 0 22 0.062 7 15 0.299 1 21 0.0191 0 15 0.179
Stage I/II 4 22 6 20 0 26 6 17 2 24 2 21
Stage III/IV 6 27 7 25 3 30 13 16 8 22 4 29
Open in a new tab

Fig. 3.

Fig. 3.

Open in a new tab

Stage-wise alterations of CDKN2A, CCND1 and RB1 in the head and neck lesions. Diagonal hatching percentage of dysplastic head and neck lesions, stipples percentage of stage-I/II HNSCC tumours, cross-hatching percentage of stage-III/IV HNSCC tumours

In the analysis of the methylation status of CDKN2A promoter, 21% (17/81) of the samples were found to be hypermethylated (Table 2 and Fig. 2), but, no significant correlation has been found between CDKN2A methylation and different clinicopathological features (Table 4). However, similar to the CDKN2A mutation, the frequency of CDKN2A promoter hypermethylation was seen to be more or less constant (18–23%) throughout the progression of the tumour (Fig. 3). This indicated that CDKN2A promoter hypermethylation was necessary for the development of dysplastic head and neck lesions. Only one sample (no. 308) showed both mutation and methylation. On the other hand, homozygous deletion of the CDKN2A locus was seen in only 4% (3/81) of the samples (Table 2 and Fig. 2), and a trend towards association (P=0.062–0.07) was found between this phenomenon and nodal involvement and progression of the tumour (Table 4). This indicated that homozygous deletion of the CDKN2A locus might be associated with the progression of some tumours at later stages (Table 2 and Figs. 2 and 3). Using similar methods, other investigators have reported 23–27% hypermethylation and 0–14.6% homozygous deletion in HNSCC tumours (Lin et al. 2000; Wu et al. 1999; Tsai et al. 2001). On the other hand, approximately 50% hypermethylation and 44−78% homozygous deletion has been reported in HNSCC by different investigators (Shintani et al. 2001; Shahnavaz et al. 2001; Nakahara et al. 2001). This variation in the frequencies of hypermethylation and homozygous deletion in the CDKN2A gene, found by the different investigators, might be due to the use of different methodologies, controls, and different histological status of the HNSCC tumours analysed.

In our study, 35% (26/74) of the samples showed allelic alterations (LOH and/or MA) at the CDKN2A locus. Like us, Shahnavaz et al. (2001) also detected both LOH and MA in some of the primary pre-malignant lesions of head and neck and HNSCC samples. The MA at the CDKN2A locus represents some form of genomic instability in this region. Similarly, using a dense set of markers from this CDKN2A locus, Wu et al. (1999) detected 31% LOH in the primary HNSCC samples. On the other hand, using D9S171 marker (located approximately 5 Mbp centromeric to the D9S942 marker used in our study), van der Riet et al. (1994) observed a high frequency of LOH in this tumour. In our study, significant correlation (P=0.009) was found between the alteration at the D9S942 locus and the tobacco habit of the patients (Table 4). The LOH/MA at the CDKN2A locus remained more or less constant (26–31%) in the dysplastic lesions and stage-I/II tumours and considerably increased in stage-III/IV tumours (Fig. 3). Thus, similar to mutation and methylation in the CDKN2A gene, the LOH/MA at this locus were also necessary for the development of dysplastic lesions, and its increase in frequency in the later stages of the tumour might provide some growth advantage. These alterations in the CDKN2A gene might affect p16 protein expression in the pre-malignant lesions of the head and neck, as detected by Mao and El-Naggar (1999) by immunohistochemistry. Among the 46 samples showing the CDKN2A alterations, only 15 showed any of the two types of alterations (Table 2). Thus, it seemed that in HNSCC, CDKN2A predominantly acted as a haplo-insufficient TSG, in comparison with the classical two-hit model requiring inactivation of both the alleles. This type of phenomenon has also been suggested in other tumours (Knudson 1995; Zabarovsky et al. 2002). Thus, the acquiring of other genetic modification in different genes was also associated with this tumour development.

The potentially dominant Arg24Cys mutation of CDK4 was not detected in any of the samples (Table 2). CCND1 was amplified in only 14% (11/78) of the samples analysed, and its amplification was significantly associated (P=0.0191) with progression of the tumour (Tables 2 and 4 and Figs. 2 and 3). The amplification range of the CCND1 locus in our samples has been found to be between 2.0-fold and 5.0-fold. Most of the samples with CCND1 amplification (except sample nos. L33 and 1068) had at least one of the alterations in the CDKN2A locus, indicating additive effect of CCND1 amplification in this tumour development. However, using our same method, Kyomoto et al. (1997) detected CCND1 amplification in 25% of dysplastic lesions of the head and neck and 22% of HNSCC samples. On the other hand, CCND1 amplification has been seen in 16–38% of HNSCC tumours by other groups (Callender et al. 1994). In our study, the LOH/MA at the RB1 locus was seen in only six HNSCC samples (Table 2 and Figs. 2 and 3), and no significant correlation has been found between this phenomenon and different clinicopathological features (Table 4). The tumours with changes at the RB1 locus had at least one of the alterations in the CDKN2A locus, and some of these samples (3/6) also showed CCND1 amplification (Table 2). Thus, it seemed that the allelic alteration at the RB1 locus was not necessary for initiation of the HNSCC and might have some cumulative role in tumour progression along with the alterations in the CDKN2A and CCND1 loci.

Thus, considering all alterations in the CDKN2A, CCND1 and RB1 genes together in the 81 samples analysed, we can see that 50% (11/22) of the dysplastic lesions, 58% (15/26) of the stage-I/II tumours and 67% (22/33) of the stage-III/IV tumours showed at least one of the alterations in these genes (Table 2). It could be concluded from our study that mutation/hypermethylation/LOH/MA at the CDKN2A locus were necessary for the development of pre-malignant lesions of the head and neck, whereas CDKN2A homozygous deletion along with CCND1 amplification and loss of RB1 were associated with progression of the tumour.

Acknowledgements

We are grateful to the Director of Chittaranjan National Cancer Institute (CNCI), Calcutta, Dr. S. Gupta, Director of the Cancer Centre and Welfare Home, Calcutta, and Drs. E. Zabarovsky, A Dam and S Mondal for their active support during this work. Financial support for this work was provided by grant no. BT/MB/05/002/94 from the Department of Biotechnology of the Government of India.

References

  1. Callender T, El-Naggar AK, Lee MS, Frankenthaler R, Luna MA, Batsakis JG (1994) Prad-1 (CCND1)/Cyclin D1 oncogene amplification in primary head and neck squamous cell carcinoma. Cancer 74:152–158 [DOI] [PubMed] [Google Scholar]
  2. Cerilli LA, Swartzbaugh JR, Saadut R, Marshall CE, Rumpel CA, Moskaluk CA Jr, Frierson H (1999) Analysis of chromosome 9p21 deletion and p16 mutation in salivary gland carcinomas. Hum Pathol 30:1242–1246 [DOI] [PubMed] [Google Scholar]
  3. Dasgupta S, Mukherjee N, Roy S, Roy A, Sengupta A, Roychowdhury S, Panda CK (2002) Mapping of the candidate tumor suppressor genes' loci on human chromosome 3 in head and neck squamous cell carcinoma of Indian patient population. Oral Oncol 38:6–15 [DOI] [PubMed] [Google Scholar]
  4. Heyman M, Rasool O, Brandter LB, Liu Y, Grander D, Soderhall S, Gustavsson G, Einhorn S (1996) Prognostic importance of p15INK4B and p16INK4A gene inactivation in childhood acute lymphocytic leukemia. J Clin Oncol 14:1512–1520 [DOI] [PubMed] [Google Scholar]
  5. Kannan K, Munirajan AK, Krishnamurthy J, Bhuvarahamurthy V, Mohanprasad BK, Panishankar KH, Tsuchida N, Sanmugam G (2000) The p16 INK4 alpha/p19ARF gene mutations are infrequent and are mutually exclusive to p53 mutations in Indian oral squamous cell carcinomas. Int J Oncol 16:585–590 [DOI] [PubMed] [Google Scholar]
  6. Klangby BU, Okan I, Magnusson KP, Wendland M, Lind P, Wiman KG (1998) P16/INK4a and p15/INK4b gene methylation and absence of p16/INK4a mRNA and protein expression in Burkitt's lymphoma. Blood 91:1680–1687 [PubMed] [Google Scholar]
  7. Knudson AG Jr (1995) Mutation and cancer: a personal odyssey. Adv Cancer Res 67:1–23 [DOI] [PubMed] [Google Scholar]
  8. Kyomoto R, Kumazawa H, Toda Y, Sakaida N, Okamura A, Inawaga M, Shintaku M, Yamashita T, Hiai H, Fukumoto M (1997) Cyclin-D1-gene amplification is a more potent prognostic factor than its protein overexpression in human head-and-neck squamous-cell carcinoma. Int J Cancer 74:576–581 [DOI] [PubMed] [Google Scholar]
  9. Li X, Lee NK, Ye YW, Waber PG, Schweitzer C, Cheng QC, Nisen PD (1994) Allelic loss at chromosome 3p, 8p, 13q and 17p associated with poor prognosis in head and neck cancer. J. Natl. Cancer Inst 86:1524–1529 [DOI] [PubMed] [Google Scholar]
  10. Liew CL, Li HM, Lo KW, Leow CK, Chan JY, Hin LY, Lau WY, Lai PBS, Lim BK, Huang J, Leung WT, Wu S, Lee JCK (1999) High frequency of p16INK4A gene alterations in hepatocellular carcinoma. Oncogene 18:789–795 [DOI] [PubMed] [Google Scholar]
  11. Lin SC, Chang KW, Chang CS, Liu TY, Tzeng YS, Yang FS, Wong YK (2000) Alterations of p16/MTS1 gene in oral squamous cell carcinomas from Taiwanese. J Oral Pathol Med 29:159–166 [DOI] [PubMed] [Google Scholar]
  12. Lunderberg AS, Weinberg RA (1999) Control of the cell cycle and apoptosis. Eur J Cancer 35:531–539 [PubMed] [Google Scholar]
  13. MacGrogan D, Levy A, Bostwick D, Wagner M, Wells D, Bookstein R (1994) Loss of chromosome arm 8p loci in prostate cancer: mapping by quantitative allelic imbalance. Genes Chromosomes Cancer 10:151–159 [DOI] [PubMed] [Google Scholar]
  14. Mao L, El-Naggar AK (1999) Molecular changes in the multistage pathogenesis of head and neck cancer. In: Srivastava S, et al. (eds) Molecular pathology of early cancer. IOS Press, pp 189–205
  15. Matsuda H, Konishi N, Hiasa Y, Hayashi Y, Tsuzuki T, Tao M, Kitahori Y, Yoshioka N, Kirita T, Sugimura M (1996) Alterations of p16/CDKN2, p53 and ras genes in oral squamous cell carcinoma and premalignant lesions. J Oral Pathol Med 25:232–238 [DOI] [PubMed] [Google Scholar]
  16. Nagai MA (1999) Genetic alterations in head and neck squamous cell carcinomas. Braz J Med Biol Res 32:897–904 [DOI] [PubMed] [Google Scholar]
  17. Nakahara Y, Shintani S, Mihara M, Ueyama Y, Matsumura T (2001) High frequency of homozygous deletion and methylation of p16(INK4A) gene in oral squamous cell carcinomas. Cancer Lett 163:221–228 [DOI] [PubMed] [Google Scholar]
  18. Papadimitrakopoulou V, Izzo J, Lippman SM, Lee JS, Fan YH, Clayman G, Ro JY, Hittelman WN, Lotan R, Hong WK, Mao L (1997) Frequent inactivation of p16INK4a in oral premalignant lesions. Oncogene 14:1799–1803 [DOI] [PubMed] [Google Scholar]
  19. Park JS, Dong SM, Kim HS, Lee JY, Um SJ, Park IS, Kim SJ, Namkoong SE (1999) Detection of p16 gene alteration in cervical cancer using tissue microdissection and LOH study. Cancer Lett 136:101–108 [DOI] [PubMed] [Google Scholar]
  20. Pollock PK, Pearson JV, Hayward NK (1996) Compilation of somatic mutations of the CDKN2 gene in human cancers: non-random distribution of base substitutions. Genes Chromosomes Cancer 15:77–88 [DOI] [PubMed] [Google Scholar]
  21. Saranath D, Bhoite LT, Deo MG (1993) Molecular lesions in human oral cancer: the Indian scene. Oral Oncol 29B:107–112 [DOI] [PubMed] [Google Scholar]
  22. Schuyer M, Staveren I, Klijn JGM, Burg MEL, Stoter G, Henzen-Logmans SC, Foekens JA, Berns EMJJ (1996) Sporadic CDKN2A gene alterations in human ovarian tumors. Br J Cancer 74:1069–1073 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Shahnavaz SA, Bradley G, Regezi JA, Thakker N, Gao L, Hogg D, Jordan RC (2001) Patterns of CDKN2A gene loss in sequential oral epithelial dysplasias and carcinomas. Cancer Res 61:2371–2375 [PubMed] [Google Scholar]
  24. Shintani S, Nakahara Y, Mihara M, Ueyama Y, Matsumura T (2001) Inactivation of the p14(ARF), p15(INK4B) and p16(ink4a) genes is a frequent event in human oral squamous cell carcinomas. Oral Oncol 37:498–504 [DOI] [PubMed] [Google Scholar]
  25. Tsai CH, Yang CC, Chou LS, Chou MY (2001) The correlation between alteration of p16 gene and clinical status in oral squamous cell carcinoma. J Oral Pathol Med 30:527–531 [DOI] [PubMed] [Google Scholar]
  26. van der Riet P, Nawroz H, Hruban RH, Corio R, Tokino K, Koch W, Sidransky D (1994) Frequent loss of chromosome 9p21–22 early in head and neck cancer progression. Cancer Res 54:1156–1158 [PubMed] [Google Scholar]
  27. Wu CL, Roz L, Mckown S, Sloan P, Read AP, Holland S, Porter S, Scully C, Paterson I, Tavasoli M, Thakker N (1999) DNA studies underestimate the major role of CDKN2A inactivation in oral and oropharyngeal squamous cell carcinomas. Genes Chromosomes Cancer 25:16–25 [PubMed] [Google Scholar]
  28. Yoo GH, Xu HJ, Brennan JA, Westra W, Hruban RH, Koch W, Benedict F, Sidransky D (1994) Infrequent inactivation of the retinoblastoma gene despite frequent loss of chromosome 13q in head and neck squamous cell carcinoma. Cancer Res 54:4603–4606 [PubMed] [Google Scholar]
  29. Zabarovsky ER, Lerman MI, Minna JD (2002) Tumor suppressor genes on chromosome 3p involved in the pathogenesis of lung and other cancers. Oncogene 21:6915–6935 [DOI] [PubMed] [Google Scholar]
  30. Zhang SY, Klein-Szanto AJP, Sauter ER, Safarenko M, Mitsunaga S, Nobori T, Carson DA, Ridge JA, Goodrow TL (1994) Higher frequency of alterations in the p16/CDKN2 gene in squamous cell carcinoma cell lines than in primary tumors of head and neck. Cancer Res 54:5050–5053 [PubMed] [Google Scholar]
  31. Zuo L, Weger J, Yang Q, Goldstein AM, Tucker MA, Waker GJ, Hayward N, Dracopoli NC (1996) Germline mutations in the p16INK4a binding domain of CDK4 in familial melanoma. Nat Genet 12:97–99 [DOI] [PubMed] [Google Scholar]

Articles from Journal of Cancer Research and Clinical Oncology are provided here courtesy of Springer

RESOURCES