Skip to main content
NCBI home page
Translational Oncology logoLink to Translational Oncology
. 2012 Oct 1;5(5):321–326. doi: 10.1593/tlo.12232

Tumor-suppressor Gene Promoter Hypermethylation in Saliva of Head and Neck Cancer Patients1

Dmitry A Ovchinnikov *, Matthew A Cooper †, Pratibala Pandit *, William B Coman ‡,§, Justin J Cooper-White *,, Patricia Keith *, Ernst J Wolvetang *, Paul D Slowey #, Chamindie Punyadeera *
PMCID: PMC3468923  PMID: 23066440

Abstract

Head and neck squamous cell carcinoma (HNSCC) accounts for a bulk of the oral and laryngeal cancers, the majority (70%) of which are associated with smoking and excessive drinking, major known risk factors for the development of HNSCC. In contrast to reports that suggest an inverse relationship between smoking and global DNA CpG methylation, hypermethylation of promoters of a number of genes was detected in saliva collected from patients with HNSCC. Using a sensitive methylation-specific polymerase chain reaction (MSP) assay to determine specific methylation events in the promoters of RASSF1A, DAPK1, and p16 genes, we demonstrate that we can detect tumor presence with an overall accuracy of 81% in the DNA isolated from saliva of patients with HNSCC (n = 143) when compared with the DNA isolated from the saliva of healthy nonsmoker controls (n = 31). The specificity for this MSP panel was 87% and the sensitivity was 80% (with a Fisher exact test P < .0001). In addition, the test panel performed extremely well in the detection of the early stages of HNSCCs, with a sensitivity of 94% and a specificity of 87%, and a high κ concordance value of 0.8, indicating an excellent overall agreement between the presence of HNSCC and a positive MSP panel result. In conclusion, we demonstrate that the promoter methylation of RASSF1A, DAPK1, and p16 MSP panel is useful in detecting hypermethylation events in a noninvasive manner in patients with HNSCC.

Introduction

There are 780,000 new cases of head and neck squamous cell carcinoma (HNSCC) worldwide each year, which result in more than 300,000 deaths annually. HNSCC is often associated with a poor prognosis and has a significant impact on morbidity, mortality, and health-care expenditure. Squamous cell carcinomas, in particular their oral form, represents a significant proportion of HNSCC, and accounts for 90% of all tumors with a relatively low 5-year survival rate [1]. This poor prognosis is largely caused by the late diagnosis of the disease [1]. About 30% to 50% of HNSCC are caused as a direct result of human papilloma viral infections [2]. The absence of definitive early warning signs for oral HNSCCs highlights the need for sensitive and specific biomarkers that would have a clear utility and be of significant importance for screening, particularly in high-risk individuals [3]. Detection of HNSCC is currently based on an expert clinical examination of the upper aerodigestive tract and histologic analysis of suspicious areas, but lesions can remain undetected in obscured sites such as the crypts at the base of the tongue and tonsils. It is therefore imperative to develop tools to diagnose HNSCC at an early stage of the disease enabling effective and timely interventions whereby reducing mortality and morbidity associated with this type of cancer.

Tobacco use is a major risk factor for HNSCC and smoking kills more than 1,000,000 people a year worldwide, accounting for 30% of all cancer-related deaths. According to the World Health Organization report [4], one in three people worldwide is addicted to nicotine. Smoking is thought to be responsible for ∼45% of HNSCC cases in men and 75% of cases in women [5]. Smoking has been associated with significant changes in the methylation status of the genome, causing hypomethylation and potential destabilization of many repetitive, including retro-transposon-derived, sequences. Smoking can, however, also promote hypermethylation of CpG islands within promoters of tumor-suppressor genes, thus repressing their expression [6,7], and this is a well-recognized early epigenetic event in cancer [8,9]. Several genes are frequently methylated in HNSCCs, such as the cell cycle regulator p16INK4a, DAPK1, and RASSF1A, all three of which are known to act as modulators of apoptosis [10,11].

Human saliva is an ideal diagnostic medium for investigating smoking-related cancers because of the close proximity between the oral cavity and the sites of occurrence of the relevant types of HNSCC. Saliva further represents an extremely attractive diagnostic biologic fluid, in particular for DNA-based diagnostics, because of both the ease of collection in a clinical setting [12] and its slightly basic nature (in general, a healthy human saliva has a pH of 7.4) [13] that appears to facilitate the preservation of intact high molecular weight DNA [14–16]. Indeed, saliva represents a unique biologic fluid that may offer a diagnostic window on the status of the whole organism as it carries a compendium of biomarkers derived from a wide range of organs and systems [17]. It is important to discern a sampling method that will be least affected by the disease status and, more importantly, will enable sensitive detection of the potential malignant transformation at a particular site. Our study objectives were three-fold: first, to compare two saliva collection methods, which provide distinct ways of sampling within the oral cavity [drool method yielding mainly unstimulated saliva [7] vs the oral scrapes using DNA-SAL (Oasis Diagnostics Corporation)], as there are many other methods of saliva collection, including swabs, spitting, etc., that were not used in this study (Mohamed et al., accepted to the Journal of Clinical and Translational Medicine, August, 2012); second, to determine whether DNA methylation of CpG islands in the promoters of RASSF1A, DAPK1, and p16INK4a in saliva-derived DNA is a surrogate noninvasive biomarker panel to discriminate healthy controls from patients with HNSCC; and third, to investigate whether the methylation status of CpG islands in the three target promoter genes in the DNA isolated from saliva collected from healthy control nonsmokers versus smokers can potentially serve as noninvasive screening method for early detection of HNSCC. Our study highlights the utility of saliva for methylation-specific polymerase chain reaction (MSP)-based diagnostic detection of smoking-associated DNA promoter hypermethylation of DAPK1, RASSF1A, and p16INK4a during various stages of the HNSCC development and progression. We have therefore demonstrated that saliva can be used as an analytical matrix for the development of a noninvasive and cost-effective MSP-based test, which allows for the early detection of promoter hypermethylation associated with gene cellular transformation events in patients with HNSCC.

Materials and Methods

Study Design

This study was approved by the University of Queensland Medical Ethical Institutional Board and by the Princess Alexandra Hospital Ethics Review Board. All participants gave informed consent before sample collection. Healthy control subjects (n = 46 both smokers and nonsmokers, n = 31 healthy nonsmokers) without any clinical signs of cancer as well as patients with HNSCC (n = 143) at various clinical stages of cancer (stages I–IV) were recruited to our study (Table 1).

Table 1.

Demographic Characteristics of Study Participants.

Healthy Controls Patients with HNSCC P
Age (mean ± SD) 48 ± 9 62 ± 13 <.0001
Gender
Females 21 (46%) 31 (22%) .012
Males 25 (54%) 112 (78%)
Smoking status
Smokers 15 (33%) 119 (83%) >.001
Nonsmokers 31 (67%) 24 (20%) κ = 0.3
Open in a new tab

Saliva Sample Collection

DNA promoter methylation of the three-gene panel was assayed in DNA isolated from whole mouth saliva (drool, unstimulated) and from buccal cell scrape samples (DNA-SAL) from both patients with HNSCC and healthy controls, both smokers (including recent quitters) and nonsmokers. The volunteers were asked to sit in a comfortable upright position and were asked to rinse their mouth with water (to remove food debris) and were asked to tilt their heads down and to pool saliva in the mouth for about 2 to 5 minutes. Saliva samples were collected in sterile urine containers (Sarstedt, Australia) and were transported on dry ice to the laboratory. Samples were then thawed at room temperature and centrifuged at 500g at 4°C for 10 minutes. The supernatant was discarded and the cellular pellet was frozen at -80°C until further analysis. Collection of buccal cell scrapings was performed using DNA-SAL Salivary DNA Collection Device (Oasis Diagnostics Corporation) as per the manufacturer's protocol with the exception that TE buffer was used in place of the stabilization buffer provided by the manufacturer. Cell scrapes were rinsed with TE buffer and were centrifuged (500g at 4°C) to collect the cell pellets and these were stored at -80°C.

DNA Extraction and Bisulfite Conversion of Saliva Samples

DNA extraction and subsequent bisulfite conversions were carried out using the EpiTect Plus Kit (Qiagen, Valencia, CA) according to the manufacturer's instructions with the exception of a longer incubation time (10 minutes instead of 1 minute) and the use of a larger elution volume (17 µl instead of 15 µl). Bisulfite-converted DNA was eluted from the column in elution buffer (10 mM Tris-HCl, pH 8.0) and immediately used for the first-stage nested MSP or stored at -80°C. All the converted DNA samples were assessed for their DNA purity and quantified on a ND-1000 Spectrophotometer (Thermo Scientific, Wilmington, DE).

Nested MSP Analysis of the Biomarker Panel

The selection of the three-gene panel (RASSF1A, p16INK4a, and DAPK1) was based on existing data showing high concordance between hypermethylation of the promoters of these genes and oral cancers [18,19]. MSP is the most commonly used method for detecting methylated or unmethylated alleles in human genomic DNA [10]. In this study, we used a modification of a previously reported approach [6] to detect methylation at specific sites located at (or near) the 3′ ends of the genes [20] of RASSF1A, p16INK4a, and DAPK1 (see Table 2 and Figure 1A). Briefly, first-stage methylation-independent primers were used at 0.5 µM in a standard polymerase chain reaction (PCR) (25 µl of reaction volume), using Taq DNA polymerase (TDP-500 Scientifix) with 2 to 5 µl of converted DNA template (diluted either three or five times depending on the DNA concentration) with supplied buffers/dNTPs using the following cycling conditions: 30 cycles of 15 seconds at 94°C, 15 seconds at 60°C, and 15 seconds at 72°C. To detect unmethylated or methylated alleles for DAPK1, separate second-stage reactions were carried out using similar cycling conditions and the corresponding primer set. For RASSF1A and p16INK4a, second-stage unmethylated and methylated PCRs were touchdown gradient PCRs, with annealing temperature decreasing from 64°C to 58°C in 2°/5-cycle steps. PCR products were visualized on 2% agarose gel electrophoresis in 1x TAE buffer. As a positive control for methylation-specific PCR, we used bisulfite-converted methylated HeLa cell line (NEB, Ipswich, MA, Catalogue No. 4007s). Only MSP primer sets amplifying methylated alleles not present in the majority of normal cellular DNA as well as transformed cell lines of unrelated etiologies were selected for these assays. Thus, the results of the nested MSP reactions were interpreted in a binary manner, such that presence of any detectable amplification product using a methylated allele-specific primer set was interpreted as a positive result. Each reaction of the first round of nested MSPs was carried out with one of the three gene-specific primer pairs. As an internal control for the quality of each bisulfite-converted gDNA sample, a second round of unmethylated allele-specific MSPs was carried out for all the three genes. Only samples that gave a strong consistent band using the unmethylated allele-specific reference PCR were used for promoter methylation analysis of RASSF1A, p16INK4a, and DAPK1.

Table 2.

The MSP Primers Used in Our Study.

Gene Nucleotide Sequence PCR Product Size (bp)
Primer sequences for stage 1 PCR (methylation-independent)
p16 Forward: 5′-GAGGAAGAAAGAGGAGGGGTTG-3′ 274
Reverse: 5′-ACAAACCCTCTACCCACCTAAATC-3′
RASSF1A Forward: 5′-GGAGGGAAGGAAGGGTAAGG-3′ 260
Reverse: 5′-CAACTCAATAAACTCAAACTCCC-3′
DAPK1 Forward: 5′-GGTTGTTTYGGAGTGTGAGGAGG-3′ 236
Reverse: 5′-CTAAAAACTCCCCCRATCCCT-3′
Stage 2 primer sequences.unmethylated allele-specific
p16 Forward: 5′-TTATTAGAGGGTGGGGTGGATTGT-3′ 145
Reverse: 5′-CAACCCCAAACCACAACCATAA-3′
RASSF1A Forward: 5′-GGTTTTGTGAGAGTGTGTTTAG-3′ 72
Reverse: 5′-ACACTAACAAACACAAACCAAAC-3′
DAPK1 Forward: 5′-GGAGGATAGTTGGATTGAGTTAATGTTT-3′ 80
Reverse: 5′-CAAATCCCTCCCAAACACCAA-3′
Stage 2 primer sequences-methylated allele-specific
p16 Forward: 5′-GAGGGTGGGGCGGATCGC-3′ 43
Reverse: 5′-GACCCCGAACCGCGACCG-3′
RASSF1A Forward: 5′-GGGGGTTTTGCGAGAGCGC-3′ 203
Reverse: 5′-CCCGATTAAACCCGTACTTCG-3′
DAPK1 Forward: 5′-ATAGTCGGATCGAGTTAACGTC-3′ 152
Reverse 5′-AAAACTAACCGAAACGACGACG-3′
Open in a new tab

Figure 1.

Figure 1

Open in a new tab

(A) The detection of the promoter hypermethylation events for p16 across different sample groups and tumor grades. U = unmethylated PCR and M = methylated MSPs; samples 1 to 4 are four different patient samples, samples 5 and 6 are HeLa positives, and sample 7 is a negative PCR control. (B) The percentages of samples positive for either whole three-gene MSP panel (dark gray) or p16 (light gray columns) in different categories.

Statistical Analysis of the Results

Determination of the clinical performance (sensitivity, specificity, and accuracy, as well as positive and negative predictive values) was performed using standard algorithms based on the 2 x 2 contingency tables (specific for each comparison) and using an available online resource hosted by the Johns Hopkins Medical Center (http://www.rad.jhmi.edu/jeng/javarad/roc/JROCFITi.html). Standard error and confidence intervals (CIs) for sensitivity and specificity were calculated using standard formulas and a multiplier of 1.96 for 95% CI [21]. Fisher exact test and other parameters used for data evaluation were also calculated using online statistical resources (e.g., http://statpages.org/ctab2x2.html). Agreement quality (κ) quantification using the same 2 x 2 contingency tables was performed using the GraphPad online resource (http://graphpad.com/quickcalcs/kappa1.cfm).

Results

Validation of the Performance of the Three-Marker Panel in Diagnosing HNSCC

Of a cohort of 143 patients with HNSCC, only 24 had never smoked according to the World Health Organization criteria (http://www.who.int/whosis/indicators/compendium/2008/2ptu/en/index.html). The analysis of DNA methylation at the three promoters of DAPK1, p16INK4a, and RASSF1A genes in the saliva collected either as a drool sample or with the use of a DNA cell scrape allowed us to identify individuals with HNSCC with an overall accuracy of 81% (P =.05, CI = 75–84%) in patients with HNSCC when compared with healthy nonsmoker controls (n = 31). The test also yielded high specificity of 87% (71–96%) and sensitivity of 82% (78–84%) for the three-marker panel, with a Fisher exact test (P < .0001). The test is also characterized by a very high positive predictive value (0.97, P = .05, CI = 0.93–0.99) and a good negative predictive value (NPV) (0.5). As expected, the overall positiveness for the panel appeared to correlate well with the presence of the HNSCC, with κ = 0.51.

The test panel performed particularly well in the detection of early stage HNSCCs, with detection of well to poorly differentiated SCCs (grade II) with a high sensitivity of 94% (P = .05, CI = 85–98), a specificity of 87%, an accuracy of 91%( P = .05, CI = 81–96%), and a very high concordance between positiveness for the MSP panel and the presence of the tumor(s), with κ value of 0.8, indicating a fair overall agreement. For patients with grade III tumors, a similar specificity of 87% was observed, with an overall accuracy of 86%. The test also performed well in the detection of the metastatic HNSCC (grade IV), at an accuracy of 87%, a sensitivity of 88%, and a specificity of 87%, with P < .001. Determination of the κ value (0.72) indicated a good agreement between gene promoter hypermethylation and the overall incidence of HNSCC, with κ at 0.8 for grade II and good agreement at κ = 0.7 for grade III (see also Figure 2). Similarly, the NPV rose from 0.53 for grade I/II to 0.60 for grade III and 0.72 for grade IV. It is clear from Figure 1, depicting a fraction of samples in categories that tested positive for the marker set and that the panel as a whole acts as a better indicator of HNSCC status than when a single gene (p16) is used in a stand-alone fashion (Figure 1B).

Figure 2.

Figure 2

Open in a new tab

Detection of the promoter hypermethylation using MSPs and in patients with different tumor grades. (A) Sensitivity and specificity (expressed as percentages) for each of the MSP markers used singly and in the three-gene panel combination. Notably, DAPK1 hypermethylation detection is very highly specific but found only in a small fraction of patients (see text for more details). Sample sizes were 31 for controls and 143 for patients with HNSCC. (B) Grade-specific breakdown of the detection of hypermethylation in patient saliva samples using the three-gene panel. Notice that detection sensitivity increases with the advancement of the tumor stage. In A and C bars, P = .05 CIs.

Comparison of DNA Methylation of the Three Target Genes in Saliva Collected Using the Drool Method versus DNA-SAL

To perform a comparative evaluation of the drool and DNA-SAL collection methods, we used both sampling methods to obtain bisulfite-converted DNA samples from 53 patients with HNSCC. After the quality of all 106 (53 x 2) samples was confirmed using MSP primer sets specific for unmethylated variants of the gene panels' promoters, detection of the hypermethylated variants using the three-gene panel was carried out. Twelve patients were negative in the drool- and the DNA-SAL-derived DNA samples. In 19 cases, positive results were obtained from both the drool and DNA-SAL samples. Fourteen cases were positive in the drool-derived DNA samples only, and the remaining eight cases were positive in DNA-SAL- but not saliva-derived samples. Estimation of the κ value suggests not only a lack of statistically significant agreement (i.e., that measurements are indeed likely to survey correlating parameters) but also the fact that with high likelihood the measurements from the two tests are totally unrelated, as could be expected because of an essentially distinct nature of sampled cellular populations (see further discussion below), with the predicted value of κ < -0.3.

Effects of Smoking on DNA Hypermethylation within the Control Group

When comparing the smoker healthy controls with nonsmoker healthy controls, the methylation of the three-marker panel correlated moderately well with the smoking status (for DAPK1 and p16INK4a, with an accuracy of 72%, a specificity of 87%, and a concordance coefficient of ∼0.3). In contrast, when these markers were individually tested, they gave a better agreement with smoking status. For instance, in the patient group, DAPK1 was predictive for the smoking status with a high degree of specificity (88%, high positive predictive value). A high specificity of 84% was also observed for p16INK4a.

Discussion

Our study clearly demonstrates the clinical utility of using saliva as a biologic matrix to detect the molecular changes that occur in a tumor sample. DNA methylation of DAPK1, p16INK4a, and RASSF1A can be used to predict the risk of incidence of HNSCC in individuals, highlighting the potential as a screening biomarker panel in high-risk groups. In addition, the test panel performed very well (specificity of 87% and sensitivity of 80%) in the detection of early stage HNSCCs, with detection of well to poorly differentiated HNSCCs further illustrating the screening power of this DNA methylation marker panel. DNA promoter hypermethylation of DAPK1 and p16INK4a was significantly associated with smoking status. Our data are consistent with recent animal and in vitro work, demonstrating that the prevalence of p16INK4a methylation increases with the duration of smoking in a dose-dependent fashion [22–24]. In addition, methylation of p16INK4a was significantly associated with the pack years smoked [25]. Tobacco smoke has been shown to affect the methylation of the DAPK1 and p16INK4a promoters, although the precise mechanism still remains to be elucidated.

Saliva composition can constitute a representative “sampling matrix” for many molecular and physiological processes in the human body and can therefore be an extremely useful medium for detecting unwanted changes in many organ systems. At the same time, when the highest sensitivity is sought, it is more advantageous to analyze tissue at a specific locale as provided by the DNA-SAL tool. This latter method provides a scrape from the selected locale on the inner cheek and also has the advantage of providing larger amounts of the cellular material. The drawbacks of this method include its slightly more invasive nature and potentially the fact that in these particular patients scrapings are normally taken from outside of the area of the lesion, resulting in an increase in the noise from the normal buccal tissue and a dilution of signal from HNSCC cells. Saliva also contains cellular populations of mobilized cells that are likely to be enriched with tumor-derived cells (from our own observations, the abundance of the cellular component of saliva strongly and positively correlated with severity of oral HNSCC lesions), incidentally including those from tumors outside of the oral cavity, which could still be of significant relevance for diagnostic purposes.

One of the many factors contributing to the poor 5-year survival rate for HNSCCs is the delay in tumor diagnosis. To ameliorate this, a simple, noninvasive, and economical test would provide an ideal screening tool for use with large cohorts of high-risk groups. Saliva collection by the unstimulated drool method is possibly the most noninvasive test available for harvesting bodily fluids, with the exception perhaps of urine collection. The work by Esteller et al. has demonstrated aberrant DNA promoter methylation in the sputum samples of patients with lung cancer up to 3 years before the clinical manifestation of the disease [26]. Similarly, they were able to identify patients at high risk of developing lung cancer by detecting DNA hypermethylation in sputum samples in a prospective study. We have demonstrated that DNA methylation of DAPK1, p16INK4a, and RASSF1A in saliva provides a highly sensitive method for the detection of promoter hypermethylation not only in patients with HNSCC but also in discriminating healthy smokers from nonsmokers with a high specificity of 73%, suggesting likely causative correlation between monitored hypermethylation events and smoking.

The panel of genes we used for the MSP is fairly compact and consists only of three elements. This simplicity together with the relative low cost of MSP analysis renders the assay amenable to point-of-care test development. The qualitative nature of the MSP assays used in our study, in conjunction with significant advances in the development of small-scale DNA analytical devices (such as molecular beacons), is likely to lead to the future development of highly portable detection devices. It is also important to note the high specificity of the MSP test for the hypermethylation of the promoters of the tumor-suppressor genes in individuals who smoke, suggesting the possibility of use of similar testing protocols as an “early warning” of molecular changes in cells in the oral cavity of smokers. It is also important to highlight the significance of the higher sensitivity of salivary DNA MSP testing compared to the buccal cell scrape, it is likely to be because of a higher contribution by the normal buccal cellular material, and it is likely to be augmented by the wider representation of cells from various locales (and possibly shed more motile tumor cells) in the saliva.

Our study investigates three genes with high relevance to neoplastic transformation, which have been previously reported to be hypermethylated in HNSCC saliva [2]. Interestingly, the percentage of positive samples in our study appears significantly higher for the RASSF1A gene (∼50% vs 17% reported by Righini et al. [18]) and very similar for the other two genes, p16 and DAPK1, at around 25% and 15%, respectively [2]. This is possibly because of the fact that Righini et al. surveyed methylation at a different CpG site(s) using RASSF1A MSP when compared with our study, in which we probed the site that appears to be more frequently methylated in HNSCC and thus a significantly more sensitive readout in detection of the potentially undesirable hypermethylation of RASSF1A. The increase in frequency of methylation events detected using our panel correlates well with the tumor's histologic grade (Figure 2B), suggesting disease relevance of the assessed hypermethylation sites within the panel of selected genes.

In summary, using patients with HNSCC and healthy control samples, we demonstrate the relevance of a compact three-gene panel MSP-based screening tool for the identification of HNSCC and associated hypermethylation events in saliva. We also illustrate the suitability of the saliva test for assessing the hypermethylation of tumor-suppressor genes and demonstrate that a subset of the CpG islands in the RASSF1A promoter represents a better target for potential risk prediction tests.

Acknowledgments

The authors acknowledge all of the staff at the Head and Neck Cancer Clinic at the Princess Alexandra Hospital in Woolangabba and thank Associate Professor Chris Perry, Associate Professor Ben Panizza, and Dr Warren Joubert. The authors also acknowledge Professor Ian Frazer and Dr Deanne Whitworth for providing valuable comments and suggestions on the manuscript.

Footnotes

1

The authors acknowledge the financial support of the Queensland Government Smart Futures Fellowship Program (QGSFF), Collaborative Industry Engagement Fund University of Queensland (UQCIEF), and the University of Queensland New Staff Research Funds (UQNSRSF 601252). The authors declare no conflict of interest.

References

  • 1.Brinkmann O, Kastratovic DA, Dimitrijevic MV, Konstantinovic VS, Jelovac DB, Antic J, Nesic VS, Markovic SZ, Martinovic ZR, Akin D, et al. Oral squamous cell carcinoma detection by salivary biomarkers in a Serbian population. Oral Oncol. 2011;47:51–55. doi: 10.1016/j.oraloncology.2010.10.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Wu JY, Yi C, Chung HR, Wang DJ, Chang WC, Lee SY, Lin CT, Yang YC, Yang WC. Potential biomarkers in saliva for oral squamous cell carcinoma. Oral Oncol. 2010;46:226–231. doi: 10.1016/j.oraloncology.2010.01.007. [DOI] [PubMed] [Google Scholar]
  • 3.Freedman ND, Abnet CC, Leitzmann MF, Hollenbeck AR, Schatzkin A. Prospective investigation of the cigarette smoking-head and neck cancer association by sex. Cancer. 2007;110:1593–1601. doi: 10.1002/cncr.22957. [DOI] [PubMed] [Google Scholar]
  • 4.WHO, author. WHO Report. 2008. http://www.who.int/whosis/indicators/compendium/2008/2ptu/en/index.html.
  • 5.Viet CT, Schmidt BL. Methylation array analysis of preoperative and postoperative saliva DNA in oral cancer patients. Cancer Epidemiol Biomarkers Prev. 2008;17:3603–3611. doi: 10.1158/1055-9965.EPI-08-0507. [DOI] [PubMed] [Google Scholar]
  • 6.Richards KL, Zhang B, Baggerly KA, Colella S, Lang JC, Schuller DE, Krahe R. Genome-wide hypomethylation in head and neck cancer is more pronounced in HPV-negative tumors and is associated with genomic instability. PLoS One. 2009;4:e4941. doi: 10.1371/journal.pone.0004941. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Smith IM, Mydlarz WK, Mithani SK, Califano JA. DNA global hypomethylation in squamous cell head and neck cancer associated with smoking, alcohol consumption and stage. Int J Cancer. 2007;121:1724–1728. doi: 10.1002/ijc.22889. [DOI] [PubMed] [Google Scholar]
  • 8.Herman JG, Baylin SB. Gene silencing in cancer in association with promoter hypermethylation. N Engl J Med. 2003;349:2042–2054. doi: 10.1056/NEJMra023075. [DOI] [PubMed] [Google Scholar]
  • 9.Jones PA, Baylin SB. The fundamental role of epigenetic events in cancer. Nat Rev Genet. 2002;3:415–428. doi: 10.1038/nrg816. [DOI] [PubMed] [Google Scholar]
  • 10.Fan CY. Epigenetic alterations in head and neck cancer: prevalence, clinical significance, and implications. Curr Oncol Rep. 2004;6:152–161. doi: 10.1007/s11912-004-0027-0. [DOI] [PubMed] [Google Scholar]
  • 11.Shaw R. The epigenetics of oral cancer. Int J Oral Maxillofac Surg. 2006;35:101–108. doi: 10.1016/j.ijom.2005.06.014. [DOI] [PubMed] [Google Scholar]
  • 12.Punyadeera C. Human saliva as a tool to investigate intimate partner violence. Brain Behav Immun. 2012;26:541–542. doi: 10.1016/j.bbi.2012.02.006. [DOI] [PubMed] [Google Scholar]
  • 13.Bahar G, Feinmesser R, Shpitzer T, Popovtzer A, Nagler RM. Salivary analysis in oral cancer patients: DNA and protein oxidation, reactive nitrogen species, and antioxidant profile. Cancer. 2007;109:54–59. doi: 10.1002/cncr.22386. [DOI] [PubMed] [Google Scholar]
  • 14.Punyadeera C, Dimeski G, Kostner K, Beyerlein P, Cooper-White J. One-step homogeneous C-reactive protein assay for saliva. J Immunol Methods. 2011;373:19–25. doi: 10.1016/j.jim.2011.07.013. [DOI] [PubMed] [Google Scholar]
  • 15.Pfaffe T, Cooper-White J, Beyerlein P, Kostner K, Punyadeera C. Diagnostic potential of saliva: current state and future applications. Clin Chem. 2011;57:675–687. doi: 10.1373/clinchem.2010.153767. [DOI] [PubMed] [Google Scholar]
  • 16.Topkas E, Keith P, Dimeski G, Cooper-White J, Punyadeera C. Evaluation of saliva collection devices for the analysis of proteins. Clin Chim Acta. 2012;413:1066–1070. doi: 10.1016/j.cca.2012.02.020. [DOI] [PubMed] [Google Scholar]
  • 17.Schulz BL, Cooper-White J, Punyadeera CK. Saliva proteome research: current status and future outlook. Crit Rev Biotechnol. 2012 doi: 10.3109/07388551.2012.687361. [E-pub ahead of print 21 May] [DOI] [PubMed] [Google Scholar]
  • 18.Righini CA, de Fraipont F, Timsit JF, Faure C, Brambilla E, Reyt E, Favrot MC. Tumor-specific methylation in saliva: a promising biomarker for early detection of head and neck cancer recurrence. Clin Cancer Res. 2007;13:1179–1185. doi: 10.1158/1078-0432.CCR-06-2027. [DOI] [PubMed] [Google Scholar]
  • 19.Righini CA, de Fraipont F, Reyt E, Favrot MC. Aberrant methylation of tumor suppressor genes in head and neck squamous cell carcinoma: is it clinically relevant? Bull Cancer. 2007;94:191–197. [PubMed] [Google Scholar]
  • 20.Swafford DS, Middleton SK, Palmisano WA, Nikula KJ, Tesfaigzi J, Baylin SB, Herman JG, Belinsky SA. Frequent aberrant methylation of p16INK4a in primary rat lung tumors. Mol Cell Biol. 1997;17:1366–1374. doi: 10.1128/mcb.17.3.1366. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Belinsky SA. Role of the cytosine DNA-methyltransferase and p16INK4a genes in the development of mouse lung tumors. Exp Lung Res. 1998;24:463–479. doi: 10.3109/01902149809087381. [DOI] [PubMed] [Google Scholar]
  • 22.Lee YW, Klein CB, Kargacin B, Salnikow K, Kitahara J, Dowjat K, Zhitkovich A, Christie NT, Costa M. Carcinogenic nickel silences gene expression by chromatin condensation and DNA methylation: a new model for epigenetic carcinogens. Mol Cell Biol. 1995;15:2547–2557. doi: 10.1128/mcb.15.5.2547. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Kim DH, Nelson HH, Wiencke JK, Zheng S, Christiani DC, Wain JC, Mark EJ, Kelsey KT. p16INK4a and histology-specific methylation of CpG islands by exposure to tobacco smoke in non-small cell lung cancer. Cancer Res. 2001;61:3419–3424. [PubMed] [Google Scholar]
  • 24.Wild D. The Immunoassay Handbook. New York, NY: Nature Publishing Group; 2001. p. 906. [Google Scholar]
  • 25.Palmisano WA, Divine KK, Saccomanno G, Gilliland FD, Baylin SB, Herman JG, Belinsky SA. Predicting lung cancer by detecting aberrant promoter methylation in sputum. Cancer Res. 2000;60:5954–5958. [PubMed] [Google Scholar]
  • 26.Esteller M, Corn PG, Baylin SB, Herman JG. A gene hypermethylation profile of human cancer. Cancer Res. 2001;61:3225–3229. [PubMed] [Google Scholar]

Articles from Translational Oncology are provided here courtesy of Neoplasia Press

RESOURCES