We evaluated promoter hypermethylation of a panel of tumor suppressor genes as a means to detect epigenetic alterations in oral squamous cell carcinomas (OSCC) of Indian-origin and compare with North-American head and neck squamous cell carcinomas (HNSCC). Quantitative-methylation-specific PCR was used to investigate the promoter methylation status of DCC, EDNRB, p16INK4a and KIF1A in 92 OSCC, and compared to 48 paired normal tissues and 30 saliva and sera samples from healthy control subjects. Aberrant methylation of at-least one of these genes was detected in 74/92 (80.4%) OSCC; 72.8% at EDNRB, 71.7% at KIF1A, 47.8% at p16INK4a and 58.7% at DCC; and in 5 of 48 (10.4%) normal oral tissues. None of the saliva and sera samples from controls exhibited DNA methylation in these four target genes. Thirty-two of 72 node positive cases harbored p16INK4a and DCC hypermethylation (p = 0.005). Thus, promoter hypermethylation in genes analyzed herein is a common event in Indian OSCC and may represent promising markers for the molecular staging of OSCC patients. We found higher frequency of p16INK4a methylation (47.8%) in this Indian cohort in comparison with a North-American cohort (37.5%). In conclusion, aberrant methylation of EDNRB, KIF1A, DCC and p16INK4a genes is a common event in Indian OSCC, suggesting that epigenetic alterations of these genes warrant validation in larger studies for their potential use as biomarkers.
Head and neck squamous-cell carcinoma (HNSCC) is the sixth most common cancer in United States and the fourth most prevalent cancer in men worldwide.1 Rapid advances in treatment modalities and improvements in the early detection of HNSCC have not significantly impacted the overall survival rates of cancer patients (about 50% at 5 years). Development of novel biomarkers offer the potential to transform clinical practice by improving the efficacy of cancer screening and diagnosis based on molecular markers as a complement to routine clinical screening and diagnostic strategies. An epigenetic pathway of transcriptional inactivation for many tumor suppressor genes includes CpG island hypermethylation within promoter regions.2–5 This pathway has been identified in several human cancers including HNSCC.5–8 Promoter hypermethylation is a powerful and ubiquitous mechanism of gene silencing which can be detected in tissue samples using quantitative methylation-specific PCR (Q-MSP); this real-time PCR methodology enables an objective, robust, and rapid assessment of promoter methylation status. The ability to quantify the methylation provides the potential for determination of a threshold value of methylation to improve sensitivity and specificity in detection of tumor-specific signal.5, 9
Some of us (JC, SD) have recently demonstrated the utility of analysis of promoter methylation of a panel of genes in HNSCC. The ongoing discovery efforts to define specific and sensitive promoter methylation based markers for HNSCC and other tumors have resulted in an ever expanding roster of genes with tumor specific hypermethylation.10 However, there have often been differences in molecular alterations in HNSCC, related to geographic location. In particular, oral squamous cell carcinoma (OSCC) the commonest form of HNSCC in the Indian population have been noted to have a different spectrum of molecular alterations when compared to the North American and European populations,11 perhaps related to differences in risk factors. Defining population based differences in molecular alterations in OSCC would significantly impact understanding of the biology underlying oral carcinogenesis, and indicate a need for population specific approaches to development of novel therapies as well as molecular detection strategies. Even though the gene-promoter methylation data in the oral cavity is meager, measurement of such patterns has shown promise in cancer detection schemes.8 Because of the sensitivity and specificity of PCR-based assays (especially quantitative methylation-specific PCR) and the potential for use in high-throughput assays, determination of methylation status in oral cancers may have great potential for early detection, monitoring, and treatment.
The objective of our study was to determine the frequency of promoter hypermethylation in a panel of genes (found to be methylated in HNSCC in a North American population by SD and JC), in an Indian cohort. We specifically chose genes that were altered with high specificity, as these genes would be candidates for broad use in molecular detection strategies, if they maintained their specificity across diverse populations. During our study, we were able to show specific differential promoter hypermethylation in Indian OSCC patients compared to normal individuals for four genes, tumor suppressor gene deleted in colorectal carcinoma (DCC), Endothelin receptor type B (EDNRB), p16INK4a and Kinesin chain member 1A (KIF1A). In addition, associations of promoter hypermethylation in these genes with clinicopathological parameters were also assessed.
CDK: cyclin dependent kinase; DCC: deleted in colorectal carcinoma; EDNRB: endothelin receptor type B; ET: endothelin; HNSCC: head and neck squamous-cell carcinoma; KIF1A: kinesin chain member 1A; OSCC: oral squamous cell carcinoma; Q-MSP: quantitative methylation-specific PCR
Material and Methods
Tissue samples from 92 OSCC patients and 48 normal tissues from a distant site were obtained from patients undergoing surgical treatment in the Institute of Rotary Cancer Hospital (IRCH) at All India Institute of Medical Sciences, using appropriate informed consent obtained after institutional research ethics committee approval. Control saliva rinses and sera from 30 healthy subjects with no evidence of disease were collected as per the protocol described by Carvalho et al.5 Tumors were snap frozen and microdissected on a cryostat to at least 80% purity. Hematoxylin and eosin stained sections were evaluated by the experienced pathologist. Histologically proven cancer and normal tissues were used for DNA extraction as per the method described below.
DNA from tissues, saliva and sera were extracted by digestion with 50 μg/mL proteinase K (Boehringer, Mannheim, Germany) in the presence of 1% SDS at 48°C overnight, followed by phenol/chloroform extraction and ethanol precipitation.
Quantitative methylation-specific PCR
DNA from primary tumors, adjacent normal tissues from OSCC patients, and saliva and sera from controls were modified with bisulfite and cleaned using EpiTect® bisulfite conversion kit (Cat. No. 59110) purchased from Qiagen (Chatsworth, CA). The bisulfite-modified DNA was used as a template for fluorescence-based real-time PCR, as previously described.6 In brief, primers and probes were designed to specifically amplify the bisulfite-converted DNA for the ß-actin gene (sense, 5′-TGG TGA TGG AGG AGG TTT AGT AAG T-3′; antisense, 5′-AAC CAA TAA AAC CTA CTC CTC CCT TAA-3′; and probe, 6-FAM-5′-ACC ACC ACC CAA CAC ACA ATA ACA AAC ACA-3′-6-TAMRA); DCC gene (sense, 5′-TTG TTC GCG ATT TTT GGT TTC-3′; antisense, 5′-AAA AAC CAC TTA CCG ATT ACT TAA AAA-3′ and probe, 6-FAM-5′-GCG CTA AAC AAA AAA ACT CCG AAA A-3′-6-TAMRA)6; p16 gene (sense, 5′-TTA TTA GAG GGT GGG GCG GAT CGC-3′; antisense, 5′-AGT AGT ATG GAG TCG GCG GCG GG-3′; and probe, 6-FAM-5′-GAC CCC GAA CCG CGAC CGT AA-3′-6-TAMRA)12; EDNRB gene (sense, 5′-GGG AGT TGT AGT TTA GTT AGT TAG GGA GTA G-3′; antisense, 5′-CCC GCG ATT AAA CTC GAA AA-3′; and probe, 6-FAM-5′-TTT TTA TTC GTC GGG AGG AG-3′-6-TAMARA) (Demokan et al., submitted) and KIF1A gene (sense, 5′-GCG CGA TAA ATT AGT TGG CGA TT-3′; antisense, 5′-CTC GAC GAC TAC TCT ACG CTAT-3′; and probe, 6-FAM-5′-CCT CCC GAA ACG CTA ATT AAC TAC GCG-3′-6-TAMRA) (Demokan et al., submitted). The ratios between the values of the gene of interest and the internal reference gene, ß-actin, which was obtained by Taqman analysis, were used as a measure for representing the relative level of methylation in a particular sample (gene of interest/reference gene × 100). Fluorogenic PCR reactions were carried out in a reaction volume of 20 μL consisting of 600 nmol/L of each primer; 200 nmol/L probe; 0.75 units platinum Taq polymerase (Invitrogen); 200 μmol/L of each dATP, dCTP, dGTP, and dTTP; 200 nmol/L ROX dye reference (Invitrogen); 16.6 mmol/L ammonium sulfate; 67 mmol/L Trizma (Sigma, St. Louis, MO); 6.7 mmol/L magnesium chloride (2.5 mmol/L for p16 gene amplification); 10 mmol/L mercaptoethanol; and 0.1% DMSO. Three microliters of treated DNA solution were used in each real-time methylation-specific PCR (MSP) reaction. Amplifications were carried out in 384-well plates in a 7900 Sequence Detector System (Perkin-Elmer Applied Biosystems, Norwalk, CT). Thermal cycling was initiated with a first denaturation step at 95°C for 2 min, followed by 50 cycles of 95°C for 15 sec and 60°C for 1 min. Leukocytes from a healthy individuals were methylated in vitro with excess SssI methyltransferase (New England Biolabs Inc., Beverly, MA) to generate completely methylated DNA, and serial dilutions of this DNA were used for constructing the calibration curves on each plate. Each reaction was carried out in triplicate. Positive and negative controls were used for every plate in our study. All the samples were run in triplicates in individual runs. The QMSP runs for individual samples were performed once. However, the test was repeated for samples which did not give consistent results in all the three wells. In addition, about 10% of the samples were repeated to verify the test results.
Hypermethylation of each gene was treated as a binary variable (methylation versus no methylation) by dichotomizing each gene. Sensitivity and specificity were calculated and quantified using receiver operating characteristic (ROC) analyses for each gene. Cut-off value was taken at specificity value of ∼90% for individual genes. Proportions of gene methylation were compared between tumor samples (from cases) and normal tissues (adjacent to tumor tissues) and salivary rinses or sera samples (from controls) using Fisher's exact test. The potential of the confounding effect of the covariates, including age, gender, tobacco and alcohol consumption, was assessed using regression analysis.
OSCC patients and control characteristics
OSCC patients were mainly males (79%) with age ranging from 23 to 72 years (median, 46.5 years; Table 1). The control population was also mainly males (77% with age ranging from 21 to 71 years; median, 37.5 years). Tobacco consumption was observed in 87% of OSCC patients and 60% control subjects.
Table 1. Demographics of study patients
Promoter methylation in OSCCs
QMSP was used to analyze the promoter methylation status of four genes in 92 primary OSCC, 48 matched normal tissues, and sera and saliva from 30 healthy individuals (Table 2). Aberrant methylation was observed in at least one of these genes in 74 of 92 (80.4%) OSCC. Two genes were methylated in 70 of 92 (76.1%) tumors; three genes were methylated in 55 of 92 (59.8%) tumors. EDNRB was methylated in 67 of 92 (72.8%) tumors, KIF1A was methylated in 66 of 92 (71.7%) tumors, p16 was methylated in 44 of 92 (47.8%) tumors and DCC was methylated in 54 of 92 (58.7%) tumors.
Table 2. Methylation analysis of genes in normal and OSCC tissues
No methylation was observed in any of the four genes in DNA isolated from sera and saliva of healthy individuals; however three of the histologically proven normal tissues obtained from cancer patients showed methylation in the four genes analyzed in this cohort (Figs. 1a–1d). Significant association was observed between methylation status of the four genes analyzed in this cohort of OSCC patients (Table 3).
Table 3. Significant association between methylation of genes studied in patients with OSCC
Relationships between promoter methylation and clinicopathological parameters of OSCC patients
We performed an analysis of clinical and pathologic variables associated with patients with primary OSCC, including age, gender, stage, grade, site, and tobacco and alcohol exposure with methylation status of DCC, EDNRB, p16INK4a and KIF1A. No associations with tumor-related factors were noted with methylation status of the genes analyzed except p16. p16 promoter methylation was significantly associated with nodal involvement [p = 0.04, OR = 3.3, 95% C.I. = 1.1–10.2]. Thirty-nine of 72 (54.2%) of the node positive OSCC patients showed methylated p16 promoter. Even when tumor size, differentiation and tobacco consumption were taken into consideration, p16 methylation was significantly associated with nodal involvement in multivariate analysis (p = 0.007, OR = 9.1, 95% C.I. = 1.8–45.6; Table 4). Further, when different combinations of genes were evaluated, OSCC harboring both p16 and DCC methylation emerged to be more significantly associated with nodal involvement than p16 alone (p = 0.005, OR = 19.2, 95% C.I. = 2.4–152.8; Table 5). Importantly, 31 of the 56 (55.4%) OSCC patients with tobacco chewing habit were found to have ≥3 genes methylated (MI ≥ 0.75, p = 0.056, OR = 2.4, 95% C.I. = 1.0–5.7).
Table 4. Association between nodal involvement and methylation
Table 5. Association between nodal involvement and methylation
Using QMSP-PCR, we observed promoter hypermethylation in at least one of the four genes analyzed in 80.4% of primary OSCC. Among this panel of genes, EDNRB and KIF1A were methylated in more than 70% of the OSCC patients. Importantly, EDNRB, DCC; Microtubule-dependent motor protein: KIF1A and cell cycle regulatory gene: p16INK4a have been investigated in head and neck cancer in North American population by our group (Demokan et al., submitted).6 In our study, we determined the methylation status of these genes in OSCC and adjacent normal tissues of Indian origin. We have recently shown that DCC and p16INK4a are epigenetically inactivated by promoter hypermethylation in HNSCC.5, 6DCC is a putative conditional tumor-suppressor gene located at chromosome 18q21. It encodes a transmembrane protein with structural similarity to neural cell adhesion molecule.13 DCC promoter region has been found to be hypermethylated in multiple tumor types.6, 14, 15 p16 is a recognized tumor suppressor which induces G1 cell cycle arrest by inhibiting the phosphorylation of pRb by the cyclin-dependent kinases CDK4 and CDK6.16 EDNRB is a non-selective G protein-coupled receptor which activates phosphatidylinositol-calcium second messenger system and is proposed to negatively regulate the activity of ET-1 which includes ET-1 clearance, inhibition of ET-1 secretion, and activation of signaling transduction pathways that counter-regulate ET-1.17 ETAR-mediated mitogenic and additive proliferative effect of ET-1 may be enhanced by methylation inactivation of the EDNRB gene and the loss of its unique negative regulatory functions.18 It has been demonstrated to be hypermethylated in some types of tumors viz., nasopharyngeal, prostate, bladder, hepatocellular, lung, esophageal.19–24 Kinesin superfamily proteins (KIFs) constitute a large superfamily of microtubule-dependent proteins that transport membranous organelles and macromolecules fundamental for cellular functions along microtubules.25 KIF1A is an anterograde motor protein that transports membranous organelles along axonal microtubules. Several studies have investigated the effects of p16INK4a methylation in primary OSCC26–28 and reported a range of 23 to 87% methylation, making methylation of this gene a very important component of marker panels for tumor screening and detection.29–32 Our study showed methylated p16INK4a in 47.8% OSCCs. p16INK4a methylation was observed in 36.4% (32/88) of head and neck cancer patients (oral cavity, oropharynx, larynx, neck) in North American population. The methylation level was 35.1% (13/37) in tumor samples of patients with cancer in the oropharynx region (n = 37). The methylation level was 37.5% (9/24) in tumor samples of patients with cancer in the oral cavity (n = 24) of the North American population (Demokan et al., unpublished results) as compared to 47.8% in the present study in OSCC of Indian population. DCC was observed to be methylated in 59% OSCCs. Ogi et al.33 showed promoter methylation of DCC gene in 16 of 96 primary OSCC. DCC methylation was correlated with bone invasion, and deep invasion of the tongue. Primary oral cancer patients with methylated DCC had a significantly reduced survival than those who did not have DCC methylation.33 The presence of p16INK4a or DCC methylation in the primary tumor has been shown to be associated with poor outcome.31, 33, 34 Our study supports these findings. p16INK4a methylation was found to be significantly associated with nodal involvement. Although the presence of DCC methylation alone was not a predictor of nodal involvement, combination with p16INK4a emerged to be more significantly associated than p16INK4a alone. The impact of p16INK4a and DCC methylation in the primary tumor on nodal involvement was not related to histology or the tumor size. Thirty two of the 34 cases harboring promoter hypermethylation of DCC and p16INK4a showed nodal involvement. Thus, aberrant methylation of EDNRB, KIF1A, DCC and p16INK4a genes is highly prevalent in OSCC and may play an important role in the pathogenesis of OSCC.
OSCC in India often has a different etiology, in that these cancers are usually associated with betel quid use and/or tobacco exposure. Our study examines a different ethnic cohort from a different geographic region with different exposures than those found in often reported North American cohorts. However, there are remarkable similarities in findings regarding a panel of differentially hypermethylated tumor suppressor gene candidates. These data show that promoter hypermethylation of specific genes in an Indian cohort, including EDNRB, KIF1A, DCC and p16INK4a are fairly comparable to alterations in the North American cohort. However, more populations, geographic locations, and exposures, need to be evaluated before concluding that OSCC carcinogenetic pathway is fairly constant, and that development of prognostic and detection markers may be applied across geographic populations with success. Further, our data underscore the need for more complementary and confirmatory works to be done in this field before these epigenetic biomarkers can be routinely used in the screening and early detection of HNSCC. Thus, a larger study cohort involving different populations and exposures needs to be evaluated in support of our findings.