Main

Oral squamous cell carcinoma is the most common cancer of the head and neck region and it accounts for 274 000 new cancer cases and almost 145 000 deaths annually.1 World wide, the incidence of oral cancer varies greatly and high incidence is seen in India and other parts of Asia, France and Brazil.2 The differing social customs, genetic and environmental factors are likely responsible for this geographical variation in incidence.2 The aetiological difference is also reflected in genetic aberrations between oral cancers from the Western and South East Asian countries. In Indian oral cancers, a high prevalence of Ras and a lower incidence of p53 mutations were reported than in Western cases.3 Despite radical changes in the therapeutic management of oral cancer, the overall survival rates have not improved over 30 years and remain unsatisfactory.4

The Ras–Raf–MEK–ERK–MAP kinase signalling transduction pathway regulates cell cycle progression and apoptosis in diverse types of cells.5 The Ras–Raf–MEK–ERK signalling cascade is now established as an important target in cancer therapy.6 At least 30% of human tumours contain mutation in one of the three Ras genes; K-, N- and H-Ras.7 The mutation at codons 12, 13 and 61 leads to constitutive activation of the downstream signalling driven by Ras. In oral carcinomas, the frequency of Ras mutation showed great demographic variation. Ras mutation has been reported in 20–35% of the cases from India,8, 9, 10 whereas lower frequencies (4%) were reported from the United Kingdom11 and no mutation was reported from the Unites States.12 Davies et al13 revealed activating mutations in the B-Raf kinase gene in greater than 60% of melanomas and a broad range of other cancers. Even though only less than 1% of the cancer samples have concurrent B-Raf and Ras mutations, an apparent trend was observed in which the type of tumours with B-Raf mutations are similar to those with Ras mutation.13 In head and neck carcinoma, B-Raf mutation was reported to be a rare event and was found only in pharyngeal region.14 Even though Ras mutations are highly prevalent in Indian oral carcinoma, B-Raf mutation has not been investigated. Hence, we hypothesize that B-Raf mutations can also play a role in oral carcinogenesis in the Indian scenario.

Ras proteins have essential roles in controlling the activity of multiple downstream effector pathways that regulate normal cellular proliferation.6 Ras-controlled pathways modulate several transcription factors that eventually link Ras activity to various phases of the cell cycle.5 Oncogenic Ras efficiently transform immortal rodent and human cell line but fails to transform primary cells.15, 16 In primary rat Schawnn cells, oncogenic Ras causes premature G1 arrest and senescence mediated by the induction of p15, p16 and p21.17 This indicates that additional genetic events are required in order for Ras to be transforming. Loss of the Rb family or p16INK4A together with p53 loss, enables Ras signals to transform normal cells.18 It has been shown that activation of Ras results in increased levels of cyclin D1.19 The consequential increase in cyclin D1/CDK4 or CDK6 complex in turn inactivates Rb function and thus abrogates G1 restriction point. Thus, Rb and p53 pathways play a critical role in the Ras-mediated oncogenesis.20 Despite the large body of in vitro evidence, the multidimensional interaction between mutant Ras pathway and G1 cell cycle regulatory proteins is poorly determined in tumours in vivo. In the present study, we have analysed cooperative interaction between Ras and B-Raf genes in oral carcinogenesis in Indian patients. Moreover, the biological consequence of the activating mutations in these genes on key G1 cell cycle regulatory proteins and their prognostic significance were characterized.

Materials and methods

A total of 152 oral squamous cell carcinoma patients were studied from our original cohort of 348 patients described previously.21 Availability of tumour biopsy for DNA isolation was the criteria for selecting the present cohort from the original one. The clinical findings, treatment and follow-up have been recorded prospectively. The clinical and pathological features of the patients are summarized in Table 1. An incision biopsy of tumour tissue was collected from each eligible patient before treatment. Criteria for histological diagnosis were based upon WHO guidelines for the histological classification of oral lesions. There were 51 female (33.6%) and 101 male (66.4%) patients. Clinical follow-up of all these patients after treatment were carried out until death or up to a maximum of 50 months with a median follow-up of 22 months. The Institute's Research Review Board and the Human Ethics Committee approved the present study.

Table 1 Clinicopathological characteristics of the patients
Full size table

Mutation Analysis of B-Raf and Ras Genes

DNA from tissue samples and blood was extracted using standard phenol–chloroform method. B-Raf mutation was reported to confine to the exons 11 and 15.13 DNA samples were screened for mutations within these regions of B-Raf22 and hot-spot codons 12, 13 and 61 of all three Ras genes by PCR-SSCP.23, 24, 25, 26, 27 The positive controls used for the B-Raf mutations analyses were, exon 15–NPA cell lines, 1205 Lu and a mutation confirmed papillary thyroid tumour DNA and for exon 11 was NCI-1755 cell line. For Ras mutation analyses T24 (H-Ras exon 1), A431, SW480 (both for K-Ras exon 1), MOLT4 (N-Ras exon 1) and HL60 (N-Ras exon 2) cell lines were used. Normal blood DNA as well as placental DNA was included as wild-type control for SSCP. Following amplification, 9 μl of the PCR product was mixed with an equal volume of loading buffer containing 95% formamide, 10 mM NaOH, 0.25% bromophenol blue and 0.25% xylene cyanol. The mixture was denatured at 95°C for 10 min and then quickly cooled on ice. The whole mixture was then loaded on 5–8% non-denaturing polyacrylamide gel (bisacrylamide:acrylamide ratio—1:49) with or without 5% glycerol and respectively run at room temperature or 4°C (300 V) for 12–14 h. After electrophoresis, the bands were visualized by silver staining.

Direct DNA Sequencing

All samples exhibiting mobility shifts in SSCP were re-amplified from genomic DNA using the same sets of primers. The PCR products were gel purified and sequenced with ABI Big Dye Terminator Cycle Sequencing Kit (Applied Biosystems, Rotkreuz, Switzerland) and analysed in the ABI prism 377 Automated DNA sequencer. The sequencing was carried out from both sense and antisense orientation for confirmation.

Immunohistochemistry

Immunohistochemical expression of cell cycle regulatory proteins—cyclin D1, CDK4, Rb, p16, p53, p21 and proliferation marker PCNA in these samples were reported previously.21 Standard avidin–biotin complex (ABC) method was employed for immunohistochemical studies.21 Briefly, 5 μm thick tissue sections were cut from buffered formalin-fixed, paraffin-embedded tissues. After blocking the nonspecific sites, the sections were incubated overnight at 4°C with optimally diluted respective primary antibodies: Rb (clone IF8), cyclin D1 (clone DCS6), CDK4 (clone DCS35), p16 (clone ZJ14), p53 (clones DO7 and PAb240), p21 (clone DCS60.2) and PCNA (clone PC10). All the antisera were purchased from Neomarkers, CA, USA. The antibody-bound sites were visualized using an avidin–biotin–immunoperoxidase system (VectaElite, Vector Laboratories Inc., Burlingame, CA, USA) with diaminobenzidine and nickel intensification (Vector Laboratories Inc.) and counterstained with haematoxylin. Negative controls (with diluted non-immune mouse sera instead of primary antibody) were also run with each sample. The surrounding stromal cells and adjacent normal epithelia present in the sections served as internal controls.

The intensity of staining was evaluated using a four-point semiquantitative scale: 0=negative, 2=mild, 4=moderate and 6=intense. Besides the staining intensity, the percentage of positive cells was also noted. For assessing the positivity, the cells seen at the invading tumour front or the deep malignant islands were considered. For analysis, the results were expressed as expression index, that is, the product of percentage of positivity and score of the staining intensity seen in the majority of the cells. For analysis, median of expression index was taken as cutoff point to dichotomize the data. The expression index values above the median were categorized as overexpression and median value or values below the median were considered as low expression.21

Statistical Analysis

Association between mutation and clinical parameters as well as the expression of cell cycle regulatory proteins was analysed by χ2 tests. The univariate survival analyses were carried out using log-rank test. Recurrence-free survival curve was generated by the Kaplan–Meier method. The significance of the difference between the curve was tested by log-rank test. A two-tailed P-value ≤0.05 was considered ‘statistically significant’ in all statistical analysis.

Results

B-Raf Mutation

A representative SSCP gel of exon 15 of B-Raf is shown in Figure 1. Despite clear resolution of the positive controls, all oral cancer samples studied, showed only a normal SSCP pattern. Thus, indicating the absence of activating mutations in exons 11 and 15 of B-Raf gene in oral cancer.

Figure 1
figure 1

A representative PCR-SSCP gel of mutation analysis of the B-Raf gene exon 15. P1–3 are positive controls; N, normal samples; 1–4, tumour samples; arrows indicate mobility shift.

Full size image

Ras Mutation: Clinical and Prognostic Significance

Among the three Ras genes screened for mutation, only H-Ras showed mutation. In all, 19 samples (12.5%) were detected carrying mutation in H-Ras gene (Figure 2). The nature of mutation and amino-acid changes are detailed in Table 2. Majority of mutations were at codon 12 (63%) followed by codon 13 (32%). While only one sample showed codon 61 mutation (5%). One sample had double mutation at codon 13 (Figure 2b1). Thus, a total of 20 mutations were found in the 19 samples. The clinical significance of these mutations was statistically determined. Significant gender difference was found in the H-Ras mutation (Table 3). Females had comparatively higher percentage of H-Ras mutation. Intraoral site-specific difference in the distribution of H-Ras mutation was also seen. Most of the Ras mutations were at oral sites excluding the tongue (18/19) although only one tongue carcinoma showed Ras mutation (P=0.056; Table 3). H-Ras mutation did not show any significant association with nodal status, tumour size, grade of differentiation and stage of the disease. In survival analysis, a significant relation was noticed for Ras mutation with recurrence-free survival. H-Ras mutation carriers had significantly high recurrence-free survival (P=0.033). Even though not significant, the overall survival was also high in H-Ras mutation carriers. The Kaplan–Meier recurrence-free survival curve for patients according to H-Ras mutation is given in Figure 3.

Figure 2
figure 2

Representative electropherograms of H-Ras mutation analysis. (a1 and a2)—Sense and antisense strands of a sample showing H-Ras codon 12 mutation (GGC-AGC), respectively; (b1)—antisense strand of a sample showing double mutation at codon 13; (b2)—antisense strand of a sample showing a mutation at codon 13.

Full size image
Table 2 H-Ras mutations, the amino-acid change, cyclin D1 expression and survival status of the patients are shown
Full size table
Table 3 Variables showed significant association with H-Ras mutation
Full size table
Figure 3
figure 3

Kaplan–Meier recurrence-free survival curve of patients with oral carcinoma dichotomized according to H-Ras mutation. Ras-Mut+—patients with H-Ras mutation and Ras-Mut−—patients without H-Ras mutation. Log-rank test P=0.033.

Full size image

Association between H-Ras Mutation and G1 Cell Cycle Regulators

The general pattern of immunohistochemical expression of cell cycle regulatory proteins, cyclin D1, CDK4, Rb, p16, p53, p21 and proliferation marker PCNA, in these samples was previously reported.21 Briefly, all the examined proteins (except p16) were overexpressed in tumour cells, while p16 exhibited downregulation in the majority of malignant cases. Localization of the studied proteins was primarily nuclear, but, cyclin D1, CDK and p16 also showed some degree of cytoplasmic expression. Expression pattern of p53 was assessed with two clones of monoclonal antibodies, DO7 and Pab240 (mutant specific), which showed some prominent differences in expression pattern. The staining intensity and positivity were comparatively more with DO7 than with PAb240. A total of 5.3% of cases were negative and 72.0% showed intense staining with DO7 clone. However, when PAb240 clone was used, 40.4% of cases showed negative staining and only 29.8% of cases showed intense staining. This could be due to DO7 detecting both mutant and wild type p53, whereas, PAb240 detects only mutant protein under non-denaturing condition. The expression of PCNA was not only confined to the basaloid cells, but was also seen in differentiating squamous cells. Majority of the samples showed intense (92.4%) expression of PCNA. We have analysed the association between the expressions of cell cycle regulatory proteins and H-Ras mutation after dichotomizing the expression data with median of expression index. H-Ras mutation showed significant association with cyclin D1 (P=0.027), CDK4 (P=0.046), Rb (P=0.011) and p16 (P=0.026) expression (Table 3). The expression of cyclin D1 and CDK4 was significantly low and that of p16 and Rb was high in H-Ras mutant cases. H-Ras mutation did not show any association with the expression of p53 (with both clones), p21 and PCNA.

Discussion

Ras genes are crucial regulators of several aspects of normal growth and transformation and have been reported to be mutated in 20–35% of Indian oral carcinomas.5, 6, 8, 9, 10 The H-Ras found to be mutated in 12.5% of the cases in the present study. In spite of 12.5% of H-Ras mutation, the present study demonstrated lack of activating B-Raf mutation in the Indian oral carcinoma. H-Ras mutation showed significant association with low expression of G1 cell cycle regulatory proteins cyclin D1 and CDK4 and overexpression of p16 and Rb proteins. Moreover, the survival analysis revealed that H-Ras mutation associated with high recurrence-free survival in these patients.

The Ras–Raf–MEK–ERK–MAP kinase pathway is a well-characterized signalling pathway that couples cellular response to growth signals. Activating mutations of the Ras gene is a relatively frequent event in oral carcinogenesis in Indian patients. Previous studies from three different parts of India showed variation in Ras mutation both in terms of percentage and types of Ras genes involved.8, 9, 10 A study from southern India showed mutation in H-Ras gene (17%) and N-Ras (2%).9 In the present study (Southern most Indian population) 12.5% of the samples showed H-Ras mutation. There is also divergence in the afflicted codon in all these studies. These differences might be due to demographic variation. In the present study, despite 12.5% of the samples showing H-Ras mutation, we could not find mutations in B-Raf gene. The present study constitutes the first study of B-Raf mutations in a large series of HNSCC of a single anatomic site—oral cavity. It is also noteworthy that the present study included oral cancer patients from an area that has relatively high Ras mutation. A study, which investigated B-Raf mutation in head and neck carcinoma also, did not find mutation in oral cancer samples.14 Lack of B-Raf mutation could be due to specificity of carcinogens, which might target Ras gene rather than Ras pathway in oral carcinogenesis.

The Ras–Raf–MEK–ERK signal transduction pathway regulates cell-cycle progression in diverse cell types by modulating transcription factors.5 Ras acts at different phases of the cell cycle, including early G1, the G1/S boundary and at G2/M.28 Among the major G1 cell cycle regulatory proteins analysed, cyclin D1, CDK4, p16 and Rb showed significant association with H-Ras mutation. Several studies found that Ras activation results in increased levels of cyclin D1.19 Ras signalling also plays an important role in the assembly of cyclin D1 and CDK4 or 6.29 In this context the observed low expression of cyclin D1 in tumours with H-Ras mutation is indeed surprising. In vitro experiments suggest that oncogenic Ras inhibits cell cycle by the induction of the cyclin-dependent kinase inhibitors, p16 and p21 in primary cells.17 Cells expressing oncogenic Ras and dominant negative p53 are able to proliferate in the absence of mitogen, but are unable to proliferate in the absence of anchorage and are contact inhibited. Either loss of p16 or inactivation of the Rb family makes them anchorage independent and avoid contact inhibition.18 These studies, thus, demonstrate the likely importance of p16 and Rb in the Ras-mediated transformation.18 Recently, Williams et al30 reported that loss of the Rb tumour suppressor leads to a proliferative disadvantage in tumour cell harbouring activating mutation in the Ras gene. The observed overexpression of Rb in the Ras-mutated cases may provide a favourable environment for Ras-mediated oncogenesis. At the same time, Ras-mediated upregulation of p16 and downregulation of cyclin D1 and CDK4 may guard against the oncogenic potential of Ras by maintaining an antiproliferative response to continuous strong activation of Ras.18 In a previous study, we have reported that high expression of p16 and downregulation of cyclin D1 associated with favourable outcome in oral carcinoma.21 The molecular phenotype associated with H-Ras mutation, thus, might reduce the oncogenic potential of Ras genes and provide a survival advantage for this subset of patients. Consistent with the above-mentioned possibilities, we demonstrated a survival advantage in patients with H-Ras mutation. The present study, thus, points to a direct interaction between H-Ras mutation and G1 cell cycle regulatory protein in tumours in vivo and provides the impetus for additional studies in this direction not only in oral carcinoma but also in other types of tumours.

Analysis between genders showed significant difference in the distribution of H-Ras mutation. Females showed a higher rate of H-Ras mutations (Table 3). The odds ratio (OR) for paan chewing was more elevated among women oral cancer patients (OR 42; 95% CI 24–76) than among men in one study from southern India.31 Majority of the observed mutations (G>A transitions (13/20–65%) followed by G>T transversion (4/20–20%)) in the present study was reported to be attributed to the exposure to tobacco-derived carcinogens.32 Therefore, the high prevalence of tobacco-chewing habits among women might be the reason for the observed high rate of H-Ras mutation among women. The tongue and buccal mucosal carcinoma is reported to be biologically distinct.33 Similarly, here we also found a difference in the distribution of H-Ras mutation between tongue and other oral sites, reaffirming our previous results.33 This suggests an aetiological and biological difference in the carcinogenesis of the different intra-oral sites. Even though Ras mutation is moderately high in Indian oral cancer scenario; none of the studies investigated its prognostic potential. The present study demonstrated that oral carcinoma with H-Ras mutation represent a unique molecular subtype with favourable prognosis. Similar to the present study, Ras overexpression was reported to be associated with favourable prognosis in HNSCC,34 although, conflicting reports are also available.35

In summary, the present study demonstrated that H-Ras mutation associated with expression of key cell cycle regulatory proteins such as cyclin D1, CDK4, Rb and p16 in vivo. It seems that molecular phenotype associated with mutant Ras guard against the oncogenic potential of H-Ras mutation in oral carcinoma and thus provide favourable prognosis to the patients. Similar to this proposed theory, surprisingly, H-Ras mutation predicts favourable prognosis in this patient population. Such an interaction between H-Ras mutation and cell cycle regulatory proteins should be reconfirmed in larger number of samples. Thus, H-Ras mutation defines a molecular subtype of oral carcinoma with favourable outcome and unique biology.