Hindawi Publishing Corporation
Journal of Oncology
Volume 2012, Article ID 170172, 12 pages
Methylation-Mediated MolecularDysregulation in
Rebecca Towle1andCathie Garnis1,2
1Department of Integrative Oncology, British Columbia Cancer Research Centre, 675 West 10th Avenue, Vancouver,
BC, Canada V5Z 1L3
2Division of Otolaryngology, Department of Surgery, Faculty of Medicine, University of British Columbia, 910 West 10th Avenue,
Vancouver, BC, Canada V5Z 4E3
Correspondence should be addressed to Cathie Garnis, email@example.com
Received 24 December 2011; Revised 18 February 2012; Accepted 19 February 2012
Academic Editor: Frederick E. Domann
Copyright © 2012 R. Towle and C. Garnis. This is an open access article distributed under the Creative Commons Attribution
License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly
Herein we provide a concise review of the state of methylation research as it pertains to clinical oral cancerous and precancerous
tissues. We provide context for ongoing research efforts in this field and describe technologies that are presently being applied to
analyze clinical specimens. We also discuss the various recurrent methylation changes that have been reported for oral malignancy
(including those genes frequently silenced by promoter methylation and the small RNAs with activity modulated by methylation
changes) and describesurrogatedisease markers identified viaepigenetic analysis ofsaliva and bloodspecimens frompatients with
Oral cancer remains a major global killer . Its entrenched
poor survival rates make essential the ongoing molecular
evaluation of oral cancers and precancers for the purpose of
uncovering new tools for detecting and treating this disease.
As with most solid epithelial tumors, oral cancers develop
through a series of histopathological stages; hyperplasia
leads to various degrees of dysplasia, which are followed by
carcinoma in situ and finally invasive disease stages. The
accumulation of various genetic and epigenetic alterations
is understood to drive this progression paradigm. Herein
we will discuss the role of DNA methylation in clinical oral
tumorigenesis, focusing specifically on oral squamous cell
DNA methylation occurs most frequently at cytosine
residues of CpG dinucleotides in gene promoter regions, and
much less frequently, within a gene . CpG islands (CpG-
rich regions spanning >500bp with >55% GC content)
exist in approximately 60–70% of promoters in the human
genome [2, 3]. Methylation in the promoter region of a given
gene can serve to decrease expression of that gene. This is
thought to occur by either physically inhibiting the binding
of proteins essential for transcription, or by recruiting pro-
teins that have transcription repressive properties . This
reversible process helps govern gene expression activity in
individual cells and is commonly disrupted in cancer, where
gene silencing via methylation in particular can contribute
Further, global hypomethylation of genes is understood
to serve as a mechanism of oncogene activation, providing
another avenue for methylation changes to contribute to
tumorigenesis. In addition to being identified as an early
event in tumorigenesis for many epithelial cancers, aber-
rant methylation has also been identified in dysplasia—
complexity at this level of epigenetic dysregulation.
Myriad technologies exist for detecting DNA methylation
alterations and most of these have already been applied
in an oral cancer context. The most common technique
2 Journal of Oncology
is methylation-specific PCR (MSP). This approach involves
bisulfate conversion of unmethylated cytosines within CpG
islands of the genome, with methylated cytosines unchanged
by this conversion. In MSP, following design of primers for
both methylated and unmethylated sequences of a specific
locus, converted DNA is amplified and then separated by
gel electrophoresis. Differential analysis of resultant MSP
products—say for tumor versus patient-matched normal
tissues—reveals changes in methylation status. MSP is par-
ticularly useful because it has sufficient sensitivity to detect
one methylated cytosine in 1000 and the primers used are
also highly specific (i.e., have a low false positive rate). MSP
is also a relatively quick and affordable technique, though its
limitations include the reality that it is more qualitative than
quantitative and the fact that it is generally unhelpful for the
assessment of genome-wide methylation changes.
Other techniques for evaluating methylation changes are
built on the principles underpinning MSP. MethyLight—
which integrates sodium bisulfite conversion and quanti-
tative fluorescence PCR—offers a sensitive, highly specific,
and rapid means for assessing methylation status for a
particular locus . This technique involves use of primers
with a fluorescent 5?reporter dye (typically FAM) and a 3?
quencher dye. During amplification, Taq DNA polymerase
cleaves the probe and releases the reporter dye. The resultant
fluorescence is quantified by laser in associated equipment
and will be proportional to the number of methylated
cytosines at a given locus. This approach allows a more
precise quantitation of methylation status compared to MSP,
which can only be measured qualitatively. Additionally,
MethyLight has been shown to be more sensitive than MSP
by a factor of ten (detecting even one methylated cytosine in
Sequence alterations resulting from bisulfite conversion
can remove cut sites recognized by specific restriction
enzymes. For example, if methylation has been lost, the cut
to TGTG. COBRA (combined bisulfite restriction analysis)
leverages this fact to allow quantification of methylation
in PCR products generated following bisulfite conversion
. Ultimately, the amount of total methylation per locus
corresponds to the degree to which the strands are cut by
the restriction enzyme. The strength of this technique lies in
its quantitative results, ease of use, and compatibility with
paraffin embedded samples. However, COBRA does have
its limitations, as it is a locus specific technique and thus
unhelpful for a genome wide analysis.
Pyrosequencing is a technique that is quantitative, easy
to use, and accurate . It utilizes a sequence-by-synthesis
method that monitors the bioluminometric signal that
results when a pyrophosphate group is released during DNA
synthesis. This technique can assess the methylation status
of genes by bisulphate conversion, PCR, and subsequent
comparison of the ratio of T and Cs in the samples.
A recurring disadvantage to the above techniques is
the fact that they do not easily facilitate global analyses
of methylation changes; they are all locus-specific. Several
have been developed to address this issue. Currently, the
most common approach for genome-wide analysis involves
methylation-specific microarrays. A range of arrays are avail-
able, but most are based on one of three general principles to
sample processing: (MSRE) methylation-sensitive restriction
enzyme digests, methylation-specific immunoprecipitation,
or sodium bisulfite conversion .
Regarding enzyme digestion approaches, sample pro-
cessing with endonucleases specific to methylation sites
(e.g., HpA1I) was one of the first methods of array-based
methylation profiling . There are several different types
of arrays that use this approach. For example, in differen-
tial methylation hybridization arrays, DNA is digested by
MSREs. This pool of DNA fragments is labeled with one
fluorescent dye, and a pool of the same sample with no
digestion is labeled with second dye. They are concurrently
hybridized on an array and relative signal intensities are
measured to determine methylation status at a given locus
. This method of methylation profiling can interrogate
everything from hundreds of different CpG islands to the
entire genome. These approaches all have relatively high
sensitivity; however, their utility for analyzing disease tissues
can be variable as DNA sample input requirements can
be quite high, requiring approximately 2μg of tissue, and
clinical lesions—particularly premalignant ones—can be
quite small (thus limiting the amount of specimen available
for analysis). Additionally, these methylation arrays can be
samples can be processed in parallel, and analysis is limited
to those sequences that possess recognition sites for the
restriction enzyme associated with a given platform .
Immunoprecipitation of either methylated DNA or
chromatin, known as MeDIP and ChIP, respectively, has
also been used to facilitate global analyses of methylation
alterations [11, 12]. These methods use antibodies to pull
down material of interest and subsequently subject it to
microarray analysis. This method has the ability to detect
methylation status at a genome-wide level, does not require
specific primers, and avoids the sequence bias introduced
by restriction enzyme methods . However, this technique
can also have lower throughput than other approaches and
has low sensitivity to methylated areas that occur in CpG
poor regions. Additionally, instead of microarray analysis of
pulled down material, ChIP-Seq combines immunoprecipi-
tation with sequencing technology to elucidate protein-DNA
interactions on a genome-wide scale .
A third common approach to methylation-specific mi-
croarrays involves use of the sodium bisulfite conversion
methods described above. Illumina arrays offer one platform
based on this approach . These arrays are automated in
use and allow analysis of hundreds of samples simultane-
ously. They are also capable of delineating methylation status
for thousands of CpG sites (if not the entire genome), allow
and have relatively low sample input requirements (thus
more readily facilitating analysis of low yield clinical tissues).
That said, this technique can introduce bias via incomplete
bisulfite conversion reactions and PCR.
Lastly, a technique that is rapidly becoming increasingly
common is whole genome bisulfite sequencing. Due to
Journal of Oncology3
improvements in protocols, increase of throughput, and
the rapid decline of sequencing cost, this method is robust
and highly sensitive. In brief, this protocol involves bisulfite
conversion, PCR amplification, and sequencing of products
to determine methylation status to a single base pair
resolution . Historically, this method was problematic
due to processing errors of bisulfite conversion, degradation
of DNA due to long processing times, and the need
for a relatively large amount of sample . However,
improvements on protocols have greatly increased efficiency
of this technique, which has great promise for the profiling
of methylation status.
3.Genes Evincing Methylation Changes in Oral
As specific genes are implicated as critical drivers of malig-
nant phenotypes, they become attractive candidates for
evaluation as targets of novel therapeutics or as biomark-
ers for guiding patient management decisions. Recurrent
methylation-mediated alterations of several genes have been
reported for all invasive epithelial cancer types, includ-
ing oral tumors. It is understood that in general cancer
cells exhibit increased global hypomethylation with specific
regions of hypermethylation. However, it is currently not
known how specific genes are targeted. The major etio-
logical factors for oral squamous cell carcinoma (OSCC)
are smoking, alcohol use, and (to a lesser extent) Human
Papillomavirus (HPV) infection . How these factors
influence DNA methylation in oral tumorigenesis has not
been evaluated. However, smoking has been shown in other
cancer types to activate DNA methyltransferase 1 (DNMT1)
which catalyzes DNA methylation [18, 19]. A summary of
specific genes implicated in oral cancers and understood to
be governed by methylation changes follows below.
CDKN2A, mapping to chromosome 9p21.3, produces
two major proteins: p16(INK4), which is a cyclin-dependent
kinase inhibitor, and p14(ARF), which binds the p53-
stabilizing protein MDM2 and is involved in cell cycle
control. Deletion at this locus is regularly reported to be one
of the earliest events in oral cancer initiation and progression
. Hypermethylation of the CDKN2A promoter region
has been extensively evaluated in oral cancers with the
frequency of hypermethylation being reported as anywhere
from 28% to 86% [21, 22]. Aberrant methylation of this
panel of cell lines was investigated for homozygous deletion,
hypermethylation, and point mutations at p16, with results
indicating that the first two alteration types were the more
common modes of p16 disruption in OSCC .
In a specific cohort of betel chewing individuals with oral
cancer, methylation of p16 was detected in 63% of OSCCs
and 67% of verrucous carcinomas . In a panel of indi-
viduals of Indian descent, methylation of p16 was detected
in 23% of OSCC cases . In general, data do suggest
that differences in patient ethnicity, etiological factors, and
tissue type (since the OSCC category actually spans different
tissues) can influence the molecular alterations detected for
p16INK4A/p14ARF with various clinical features for
oral cancers, though results have varied. In one study, it was
observed that people with p16 promoter methylation had
a lower mean age, a higher risk of lymph node invasion in
young patients, a higher risk of distant metastasis in older
patients, and shortened disease free survival in older patients
. Other work showed that p16INK4A methylation was
associated with increased likelihood of disease recurrence,
whereas p14ARF is associated with lower recurrence rates
. Concurrent promoter hypermethylation of p16 and
p14 correlated significantly with tumor size and lymph node
metastasis and with later stage of OSCC in one study ,
while a separate study found methylation of p14ARF alone
correlated with a good prognosis for patients . Larger
scale trials are needed to fine tune how methylation status
of p16 and p14 promoters may best be applied to manage
P16 promoter methylation has been assessed for squa-
mous cell carcinomas of the tongue as well as in margin
tissues that remain in patients following surgical resec-
tions. As expected, tumors showed a high frequency of
p16 promoter hypermethylation (86.8%) . Regarding
tissues at surgical margins—which were all histologically-
characterized as disease-free—43.3% exhibited p16 pro-
moter hypermethylation . Significantly, those cases with
margin tissues harboring p16 hypermethylation had a 6.3-
fold increased risk for local recurrence.
Similarly, a separate group assessed p16 promoter hyper-
methylation status in OSCC tumors, associated normal
tissues, and a panel of healthy controls . They found no
methylation of p16 in the healthy control group whereas p16
for 27.3% of cases (and in all of those cases, concurrent p16
hypermethylation was also detected for matched tumors).
In this study, clinical features and habitual factors did not
correlate with methylation status. Another study did not
report prognostic significance for p16 methylation, again
showing that there is lingering ambiguity regarding whether
E-cadherin (CDH1) plays a critical role in cell adhesion
processes and is known to significantly influence epithelial
tissue architecture . With respect to malignancy, it
is known to function as a suppressor of invasion and
metastasis formation and has previously been reported as
undergoing hypermethylation-mediated silencing in several
cancer types . Previous studies of tongue squamous cell
via promoter hypermethylation was significantly associated
with poorer rates of disease-free survival . Independent
reports have confirmed association of epigenetically-silenced
CDH1 expression and poorer overall survival for oral cancer
(also demonstrating that CDH1 promoter hypermethylation
is associated specifically with poorer survival for node-
positive cases and individuals with stage III disease) .
4 Journal of Oncology
Other groups, on the other hand, while reporting associa-
tions between CDH1 promoter hypermethylation and oral
cancers (when compared to normal oral mucosa), have
failed to detect significant associations between these same
clinical parameters and this epigenetic event . Indepen-
dent evaluation based on a large OSCC patient cohort is
needed to more accurately determine the significance of
the methylation status of CDH1 vis-` a-vis clinical outcomes
(something that can also be said for other genes reported as
epigenetically dysregulated in oral tumors).
O6-methylguanine-DNA methyltransferase (MGMT) is a
DNA repair gene that protects from toxicity and mutations
that occur by alkylating agents through the removal of O6-
guanine DNA adducts. CpG island hypermethylation of the
MGMT promoter region results in gene silencing, with loss
of MGMT repair capacity thought to drive cancer progres-
sion via the emergence of genomic instability. Decreased
expression of MGMT via epigenetic silencing has been
reported for many cancer types and loss of its expression can
be tied to greater sensitivity to alkylating chemotherapeutic
agents . Epigenetic silencing of MGMT has been asso-
ciated with OSCCs where tobacco exposure and betel quid
chewing are suspected etiological factors [38–40]. MGMT
promoter hypermethylation has also been associated with
poorer outcomes for oral cancer, including a greater likeli-
hood of nodal metastases, tumor recurrence, and decreased
survival [41, 42]. MGMT promoter hypermethylation has
also been associated with poorer outcomes for oral cancer,
including a greater likelihood of nodal metastases, tumor
expression has also been associated with these parameters
in head and neck squamous cell carcinomas generally, and
OSCCs specifically [41, 43]. Ongoing or elevated MGMT
expression has been associated with resistance to alkylating
agents in multiple cancer types including gliomas, astro-
cytomas, and melanomas [44, 45]. While alkylating agents
such as ifosfamide and cyclophosphamide have been applied
to manage various stages of oral and other head and neck
malignancy, we have not found any reports to date regarding
the role of MGMT silencing in modulating response to these
other cancer types provide a strong rationale for pursuing
Death-associated protein kinase (DAPK) encodes a ser-
ine/threonine kinase that is required for apoptosis induced
by IFN-γ . Loss of its expression via promoter hyperme-
thylation has been associated with formation of metastases
and advanced disease stages in multiple cancer types, includ-
ing head and neck cancers [38, 48, 49]. Regarding OSCCs,
DAPK hypermethylation has been reported as associated
with increased likelihood of lymph node involvement,
though these results have not always attained statistical
significance [34, 50]. Interestingly, other groups have not
reported associations between these clinical features and
hypermethylation of DAPK (or other genes discussed here)
[42, 51]. Again, this may be a function of the tissue hetero-
geneity that exists within the oral squamous cell carcinoma
category. Detection of DAPK promoter hypermethylation
at resection margins of oral tumors has been significantly
associated with decreased overall survival, suggesting that it
may have utility as a biomarker for guiding patient follow-
up strategies . As with MGMT, DAPK hypermethylation
has also been associated with oral tumors where tobacco
consumption is a suspected etiological factor .
The TGFβ superfamily transcription factor, runt-related
transcription factor 3 (RUNX3), functions as a tumor
suppressor gene and is involved in mediating apoptotic
ing of RUNX3 has been reported for many cancer types
. Recently, hypermethylation of the RUNX3 promoter
region was found to be significantly associated with the
presence of lymph node metastases and tumor stage in
tongue carcinomas . Other groups have reported that
reduced RUNX3 expression or promoter hypermethylation
is associated not only with progression in oral cancers, but
also with disease recurrence and poorer prognoses . The
fact that other groups have not found significant associations
between RUNX3 promoter hypermethylation and patient
outcomes suggests that the role of epigenetic silencing of
this gene in oral cancers bears further scrutiny . The
emerging role of RUNX3-mediated perturbation of the
canonical Wnt signaling pathway in oral cancers also needs
to be further evaluated.
3.1. Wnt Pathway Genes. Dysregulation of the canonical
Wnt signaling pathway—via disrupted function of genes
such as adenomatous polyposis coli (APC), AXIN1, β-catenin
(CTNNB1), and secreted frizzled-related proteins (SFRPs)—
has been noted for a variety of cancer types, including
oral malignancies [55–58]. Briefly, canonical Wnt signaling
way is inactive, CTNNB1 exists in a phosphorylated form
and is marked for degradation, with this phosphorylation
mediated by a protein complex that incorporates Glycogen
Synthase Kinase 3β (GSK3β), APC, and AXIN proteins.
Dishevelled (Dsh) is activated and, consequently, GSK3β is
inhibited. This in turn causes CTNNB1 dephosphorylation,
which stabilizes the molecule and allows it to accumulate
in the cell nucleus, where it can induce TCF/LEF-mediated
transcription of several target genes. Regarding oncogenic
processes, the downstream effects of CTNNB1 activation
via Wnt signaling include enhanced cell proliferation and
Given their role as Wnt antagonists, SFRPs function as
tumor suppressors. Marsit et al., reporting on head and
neck cancers in general, described methylation differences
for SFRP genes based on alcohol consumption, smoking
reports involving analysis of both oral cancer cell lines
and clinical OSCC cases have reported that promoter
hypermethylation for SFRPs is associated with disease [60,
61]. Interestingly, the findings in these studies for SFRP1
have been conflicting, with both hypermethylation and
demethylation of this gene reported [60–62].
APC also functions as a tumor suppressor gene and
has also been reported as downregulated in oral tumors.
Journal of Oncology5
Disruption of APC function in OSCC has been attributed
to loss of heterozygosity (LOH alterations), mutations, and
epigenetic alterations . Regarding the latter, in vitro
studies demonstrated that treatment with demethylating
agent 5-aza-2?-deoxycytidine resulted in restoration of APC
expression in oral cancer cells, supporting the functional
importance of epigenetic silencing for this gene . The
drug received FDA approval for the treatment of myelodys-
plastic syndrome and is currently in clinical trials testing its
utility in the treatment of several solid tumors. Increased
promoter methylation for APC has been reported in clinical
oral cancer tissues and some data do suggest a relationship
between APC methylation status and development of lymph
node metastases when it is analyzed in concert with the
promoter methylation status of CDH1 .
As described above, activation of CTNNB1 is a critical
consequence of canonical Wnt pathway signaling—and
a variety of oncogenic processes can be turned on by
this activity . Elevated expression and greater nuclear
localization of CTNNB1 have been reported for oral cancer
and a multitude of other cancer types [56, 65]. While
recent work has shown CTNNB1 immunostaining levels to
be significantly associated with lymph node status, survival
outcomes, and different invasive stages for oral cancers, data
regarding the association of CTNNB1 methylation status for
this disease have not . Data from other cancer types
suggest that activating mutations of CTNNB1 that prevent
its downregulation may be a more common event than
WIF1 functions as a tumor suppressor that inhibits
Wnt signaling through direct interaction with Wnt proteins,
its activation leading to cell cycle arrest . It has been
described as downregulated in several cancer types .
Promoter hypermethylation-mediated downregulation of
WIF1 in oral cancers has been reported by multiple groups,
so there is evidence to suggest that epigenetic alteration
of this gene can contribute to invasive disease phenotypes
[52, 61]. More recently, a group studying tongue cancers
found no significant associations between WIF1 promoter
hypermethylation and lymph node metastasis formation,
tumor stage, or overall survival . In a separate study,
this same group found no significant association between
WIF1 promoter hypermethylation status in tissues from
histologically negative resection margins and oral cancer
recurrence . Results so far suggest that WIF1 alterations
used for diagnosis of invasive disease.
Figure 1 places the genes discussed in this section in their
proper context in the canonical Wnt signaling cascade.
4.Methylation Changes Governing Behavior of
Noncoding RNAs inOral Cancers
Noncoding RNAs, particularly microRNAs (miRNAs), have
been shown to play a role in many biological processes and
cellular pathways—and their behavior has been shown to be
24 nucleotide transcripts that regulate mRNA expression by
binding to and subsequently silencing the mRNA target [70–
72]. MiRNAs can act as oncogenes or tumor suppressors and
have been identified as dysregulated in several cancer types,
including oral cancer [73–77]. As is the case with protein
coding genes, there are several mechanisms that can lead
of CpG island promoters [78–82]. Using a functional-based
screen in two OSCC cell lines, Uesugi et al. profiled tumor
suppressive miRNAs that are silenced by hypermethylation
. They identified 110 miRNAs that exhibited inhibitory
properties and, when compared to additional cell lines and
tumor tissue, they found that miR-218 and miR-585 were
frequently silenced by DNA hypermethylation .
Separate work has also shown that silencing of four
tumor suppressive miRNAs (miR-34b, miR-137, miR-193a,
and miR-203) can be mediated by aberrant methylation in
OSCC cells . This same work reported downregulation
of miRNA expression through tumor-specific hypermethy-
lation as being more frequent for miR-137 and miR-193a
than for miR-34b and miR-203 when analyzed in primary
tumors and paired normal oral mucosa. A separate group
investigated the association between promoter methylation
of miR-137 and both overall survival and disease-free
survival, as well as with various prognostic factors .
This study found an association between methylation at the
miR-137 promoter and poorer overall survival, though no
associations were observed with disease-free survival or any
of the other evaluated prognostic factors. MiR-137 has been
reported as a negative regulator of CDK6/CCND1-mediated
cell cycle progression , hence promoter methylation
might serve as a means of inactivating tumor suppressive
miR-137 function. Langevin et al. suggest that their inability
to detect associations between miR-137 and other prognostic
features in their study may have been a product of factors
such as limited sample sizes and insufficiently long followup,
proposing that stronger associations between miR-137 and
oral cancer outcomes may exist . Very recent work has
suggested that epigenetically-mediated miR-137 expression
may differ in oral cancer cells depending on whether they
are part of a stem-like subpopulation . The argument for
further evaluation of miR-137 in the context of oral cancer
outcomes is bolstered by evidence from other cancer types
that also suggest a tumor suppressive biological role for this
miRNA and an association with poorer disease outcomes
[86–88]. Though only a few reports have evaluated the
impact of methylation changes on small RNA behaviors in
oral cancers and precancers, these early data do suggest a role
for these molecules in oral malignant processes and provide
a strong rationale for further studies in this area.
Premalignant lesions in the oral cavity are readily detectable
owing to the accessibility of the organ site. That said, one
factor that contributes to the poor prognosis of OSCC
is the current inability to determine which premalignant
6 Journal of Oncology
interactions are shown as activating (arrows) or inhibiting (blocked arrows). Boxes indicate processes that will ultimately be influenced by
WNT signaling. Solid lines represent direct interactions and dashed lines indicate indirect effects. Lines connecting genes represent binding
or association interactions. Complexes are represented by ovals that overlap.
Journal of Oncology7
standard of care is “watchful waiting” since histopathological
review is presently incapable of delineating progression
risk and intervention with all early lesions would lead to
overtreatment that is costly both in terms of dollars and
patient quality of life. Since earlier stage lesions are smaller
progression likelihood for a given dysplasia case represent a
key means for improving patient outcomes.
Several studies have indicated that deletion and silencing
of loci mapping to chromosome arms 9p and 3p are
common events in the early development of OSCC [89, 90].
Thus, studies investigating the role methylation changes
in oral premalignant lesion (OPL) behavior have generally
investigated genes mapping within these regions.
More specifically, several studies have investigated the
impact of methylation changes p14 and p16—both of which,
as described above, are encoded by CDKN2A and map to
promoter hypermethylation in patients with histologically
confirmed severe dysplasias (57.5% and 3.8% resp.) .
Another group reported that p16 hypermethylation in OPLs
was associated with a greater risk for progression to invasive
disease; patients with confirmed OPLs (as determined by
WHO criteria and independent review by two pathologists)
that harbored hypermethylated p16 promoters were approx-
imately two and a half times more likely to develop OSCC
than patients with OPLs that did not exhibit p16 promoter
hypermethylation . Interestingly, another study reported
higher rates of promoter methylation for neighboring p15
relative to p16 across multiple stages of histopathologically
confirmed oral dysplasia (50% versus 18%) . This
same study also reported a much higher frequency of p14
promoter methylation; though the fact that analyzed tissues
came from a population of betel quid chewers—where the
suspected disease etiology was therefore different than other
populations (where tobacco is the predominant etiological
factor)—may have impacted these results.
Another study assessed whether hypermethylated genes
previously implicated in oral dysplasias could be used as
markers for progression likelihood . In this instance,
longitudinal follow-up data made this evaluation possible
since progression status for each patient with clinically
confirmed disease was known. Interestingly, the results of
this work suggested a much lower incidence of promoter
methylation at previously identified candidate genes than
in the studies discussed above. The authors in this instance
posited that use of pyrosequencing instead of MSP may
explain the discrepancy, the former technique being less
prone to false positives. This study did conclude that p16
could be a predictor of progression, though its specificity
came in at a fairly low value (57%).
Though wealth of clinical data associated with canonical
genomic alterations at chromosome 9p and 3p makes these
regions attractive for locus-specific analysis of methylation
alterations, other loci are also worthy of further scrutiny
for the utility in predicting disease progression. 14-3-3-σ
has also found to be recurrently methylated in histologically
confirmed oral dysplasias and has been associated with
coincident methylation at p16, making it an attractive can-
didate for further evaluation in an oral cancer progression
functions to inhibit angiogenesis, invasion, and metastasis,
has a potential to be a biomarker as it is recurrently found
to be hypermethylated in the normal mucosa adjacent to
the tumor . Ultimately, whole methylome analyses of a
large panel of well-annotated OPLs with extensive followup
are needed to uncover disease-relevant biomarkers that will
impact disease management and oral cancer survival rates.
6.Global Methylome Changes and the Role
of HypomethylationinOral Malignancy
In addition to investigating the methylation status of specific
genes, some studies have begun analyzing the oral cancer
methylome as a whole. Early evidence suggests that the
distribution of methylation at various gene promoters across
a series of lesions of progressing grade from a single
individual can follow a specific pattern. Further, the CpG
island methylation phenotype (CIMP) has been observed in
several cancer types and is based on the observation that
those tumors exhibiting aberrant methylation of one gene
are more likely to have other sites of aberrant methylation
[97, 98]. To determine whether the CIMP phenomenon was
detectable in OSCC, the methylation status of ten genes
was evaluated in a large panel of oral tumors . This
analysis revealed a cluster of tumors with a greater degree
of promoter methylation than would be predicted by chance
alone. These cases were identified as “CIMP +ve” and results
suggest (1) that these tumors exhibited a less aggressive
tumor biology than other oral cancers and (2) that these
cases were characterized by a greater host inflammatory
response (a finding which the authors note is consistent with
findings in other cancer types). Another group has reported
observing CIMP in head and neck cancers, though it is clear
additional studies are needed to elucidate the mechanisms
driving this phenotype .
While the majority of literature has focused on pro-
moter hypermethylation silencing of tumor suppressor
genes as a critical mechanism driving oral tumorigenesis,
DNA hypomethylation is also understood to contribute
to development of many epithelial cancers by facilitating
activation of candidate oncogenes . More specifically, it
has been proposed that DNA hypomethylation contributes
to tumorigenesis by two potential mechanisms. The first
represents another global level of methylation dysregulation,
where highly methylated repetitive elements (such as long
interspersed nuclear element-1 LINE-1 and Alu sequences)
are demethylated, resulting in increased chromosomal insta-
bility that leads to irregular mitoses that, in turn, drive the
emergence of further mutations that can activate oncogenes
[97, 102, 103]. The second is more specific, where inadver-
are normally methylation-silenced in the human genome
[97, 102, 103].
8 Journal of Oncology
With respect to oral cancer specifically, little work has
been done in clinical tissues. Analysis of in vitro OSCC
models of interleukin-mediated chronic inflammation have
indicated that chronic inflammation can drive both global
hypomethylation of LINE-1 sequences as well as specific
CpG methylation changes . More precisely, this model
showed that IL-6 specifically was inducing these changes,
with methylation alterations also found to be associated with
downstream changes in gene expression. The suggestion that
inflammatory responses can play a critical role in mediating
cancer-causing methylation changes is of particular interest
to carcinogens, is frequently under inflammatory stress.
Another group investigated the role of hypomethylation in
a murine model of oral cancer which used DNMT1 hypo-
esophageal carcinogenesis was induced by 4-nitroquinoline
1-oxide . This group reported that reduction of DNA
methylation levels led to the suppression of tumor formation
in certain cell types.
A separate group detected global LINE-1 hypomethyla-
tion in oral rinses obtained from OSCC patients (relative
to healthy controls) . This is touched on further
in the following section. The relative scarcity of studies
addressing either global trends in the oral cancer methylome
or the role of specific instances of hypomethylation in oral
tumorigenesis points to a need to develop these research
streams more fully.
7.MethylationChanges as SurrogateMarkers
for Oral Malignancy
7.1. Saliva. The use of saliva for the diagnosis or establishing
prognosis of OSCC patients is a promising screen, as it is
both noninvasive and inexpensive. The potential of this tool
has been studied by utilizing a genome-wide methylation
array to profile the methylation status of 13 OSCC patients
before and after surgery, as well as ten normal samples
. This work identified 34 genes of interest, including
p16, and proposed panels ranging from 4–7 genes that are
likely to have the optimal sensitivity and specificity for
clinical applications (the range of specificity being 62–77%
and sensitivity being 83–100%). This group also used saliva
samples taken from the same patients at both the pre- and
postoperative stages, eliminating issues arising from using
saliva samples from unmatched normal subjects as a control.
Another group analyzed the methylation status of 11
genes by MSP for a panel of primary tumors and saliva sam-
tissue samples including previously-implicated candidates
such as p16, MGMT, DAPK, and RASSF1. Additionally, they
were able to accurately detect malignant cells in saliva several
months before tumor recurrence was otherwise detected in
patients, indicating that saliva-based assays may be a very
useful tool for effective followup of oral cancer patients.
As referenced above, a separate study of oral rinses
identified LINE-1 hypomethylation as being more prevalent
in rinses obtained from OSCC patients when compared to
healthy controls . While the authors did not establish
any statistically significant associations between their rinse
results and disease stage, histological grade, lesions site,
or carcinogen exposures (e.g., tobacco smoke), the utility
of this noninvasive approach for differential diagnosis of
oral malignancy does suggest it is worth evaluating in
independent patient cohorts.
To assess the potential of saliva as a screen for pre-
malignant lesions, another group collected saliva samples
from patients that had leukoplakia and were at risk for the
development of OSCC . They assessed the methylation
status of p16, p14, and MGMT from these specimens using
MSP and observed a relatively high frequency of hyperme-
thylation of MGMT and p16 relative to levels typically found
in OSCC patients. However, followup had not been done on
the patients, so whether or not the hypermethylation of p16
and MGMT was unique to patients at risk for progressing
into OSCC is not known.
7.2. Serum. In addition to saliva rinses, methylation changes
detected in DNA isolated from serum from peripheral blood
also hold potential as a diagnostic or prognostic tools for
detection and management of disease. This approach has
previously been used in a variety of other cancers such
as lung, liver, and colorectal [110–112]. As with studies
involving saliva, the methylation status of p16 has been the
predominant focus of this work in oral cancer.
One group found that parallel evaluation of pro-
moter methylation status of multiple genes (including p16)
improved the sensitivity and specificity of disease discrimi-
methylation status of multiple genes in serum samples
obtained from OSCC cases and found that in a majority of
cases, at least one of the four assayed genes was detectable
as hypermethylated . It has been reported that when a
gene is found to be hypermethylated within a tumor, DNA
from a matched peripheral blood sample will harbor similar
methylation changes at a frequency ranging from 31 and
54.5% [49, 114, 115].
Evaluation of methylation alterations in samples ob-
tained by minimally invasive procedures such as collection
of saliva or blood remains a very attractive avenue for the
discovery of biomarkers for management of clinical oral
malignancy. Expanded collection efforts and evaluation of
these biological materials should be a component of all
molecular investigations of oral cancers and precancers.
Interrogation of methylation changes in clinical oral cancers
and precancers has revealed multiple recurrent alterations
at genes and chromosomal loci that are associated with
oncogenic processes in this and other cancer types. Some
analyses have indicated strong associations between these
changes and specific patient outcomes, including the likeli-
hood of progressing from premalignant to invasive disease
Journal of Oncology9
samples, that are easily accessible, has also generated data
that methylation changes detected in these specimens might
have utility as surrogate biomarkers for managing disease.
global health challenge, these more affordably processed
specimen types should be of particular research interest
going forward if the field is serious about developing tools
with utility worldwide. We conclude that changes in DNA
methylation from clinical tissues do show early promise as
markers to improve management for this cancer type, which
has had survival rates that have remained stubbornly low
over recent decades. We ultimately feel that the true utility
of methylation markers for management of oral malignancy
will only be realized when analyses are expanded to much
larger patient cohorts. This reasoning holds whether the
methylation status is being evaluated for an individual locus
or by the most current and robust platform for whole
methylome analysis. We also feel that a much stronger
research emphasis on developing new methylation markers
for managing smaller, more treatable premalignant lesions
would address a currently underserved path for improving
oral cancer outcomes.
Conflicts of Interest
The authors acknowledge financial support from Pacific
Otolaryngology Foundation, Rotary Hearing Foundation,
and CIHR (201003MOP-221786-CPT-CAAA-136662).
 J. Ferlay, H. R. Shin, F. Bray, D. Forman, C. Mathers, and D.
GLOBOCAN 2008,” International Journal of Cancer, vol. 127,
no. 12, pp. 2893–2917, 2010.
 D. Takai and P. A. Jones, “Comprehensive analysis of CpG
National Academy of Sciences of the United States of America,
vol. 99, no. 6, pp. 3740–3745, 2002.
 R. S. Illingworth and A. P. Bird, “CpG islands—“a rough
guide”,” FEBS Letters, vol. 583, no. 11, pp. 1713–1720, 2009.
 A. Bird, “DNA methylation patterns and epigenetic mem-
ory,” Genes and Development, vol. 16, no. 1, pp. 6–21, 2002.
 A. G. Knudson Jr., “Mutation and cancer: statistical study
of retinoblastoma,” Proceedings of the National Academy of
 C. A. Eads, K. D. Danenberg, K. Kawakami et al., “Methy-
Light: a high-throughput assay to measure DNA methyla-
tion,” Nucleic Acids Research, vol. 28, no. 8, p. E32, 2000.
 Z. Xiong and P. W. Laird, “COBRA: a sensitive and quantita-
tive DNA methylation assay,” Nucleic Acids Research, vol. 25,
no. 12, pp. 2532–2534, 1997.
 J. Tost and I. G. Gut, “DNA methylation analysis by
pyrosequencing,” Nature Protocols, vol. 2, no. 9, pp. 2265–
 M. Bibikova and J.-B. Fan, “Genome-wide DNA methylation
profiling,” Wiley Interdisciplinary Reviews: Systems Biology
and Medicine, vol. 2, no. 2, pp. 210–223, 2010.
 T. H. M. Huang, M. R. Perry, and D. E. Laux, “Methylation
profiling of CpG islands in human breast cancer cells,”
Human Molecular Genetics, vol. 8, no. 3, pp. 459–470, 1999.
 M. Weber, J. J. Davies, D. Wittig et al., “Chromosome-wide
and promoter-specific analyses identify sites of differential
DNA methylation in normal and transformed human cells,”
Nature Genetics, vol. 37, no. 8, pp. 853–862, 2005.
 J. Wu, L. T. Smith, C. Plass, and T. H. M. Huang, “ChIP-chip
comes of age for genome-wide functional analysis,” Cancer
Research, vol. 66, no. 14, pp. 6899–6902, 2006.
 M. Bibikova, J. Le, B. Barnes et al., “Genome-wide DNA
1, no. 1, pp. 177–200, 2009.
 M. Frommer, L. E. McDonald, D. S. Millar et al., “A
genomic sequencing protocol that yields a positive display
of 5-methylcytosine residues in individual DNA strands,”
Proceedings of the National Academy of Sciences of the United
States of America, vol. 89, no. 5, pp. 1827–1831, 1992.
 M. Ehrich, S. Zoll, S. Sur, and D. van den Boom, “A new
treatment,” Nucleic Acids Research, vol. 35, no. 5, article e29,
 S. D. D. Silva, A. Ferlito, R. P. Takes et al., “Advances and
applications of oral cancer basic research,” Oral Oncology,
vol. 47, no. 9, pp. 783–791, 2011.
 R. K. Lin, Y. S. Hsieh, P. Lin et al., “The tobacco-specific
carcinogen NNK induces DNA methyltransferase 1 accumu-
lation and tumor suppressor gene hypermethylation in mice
and lung cancer patients,” Journal of Clinical Investigation,
vol. 120, no. 2, pp. 521–532, 2010.
 S. A. Belinsky, D. M. Klinge, C. A. Stidley et al., “Inhibition
of DNA Methylation and Histone Deacetylation Prevents
Murine Lung Cancer,” Cancer Research, vol. 63, no. 21, pp.
 V. Papadimitrakopoulou, J. Izzo, S. M. Lippman et al.,
“Frequent inactivation of p16(INK4α) in oral premalignant
lesions,” Oncogene, vol. 14, no. 15, pp. 1799–1803, 1997.
 Y. Nakahara, S. Shintani, M. Mihara, Y. Ueyama, and
T. Matsumura, “High frequency of homozygous deletion
and methylation of p16INK4A gene in oral squamous cell
carcinomas,” Cancer Letters, vol. 163, no. 2, pp. 221–228,
 D. Akanuma, N. Uzawa, M. A. Yoshida, A. Negishi, T.
Amagasa, and T. Ikeuchi, “Inactivation patterns of the p16
(INK4a) gene in oral squamous cell carcinoma cell lines,”
Oral Oncology, vol. 35, no. 5, pp. 476–483, 1999.
 D. T. Cody II, Y. Huang, C. J. Darby, G. K. Johnson, and
F. E. Domann, “Differential DNA methylation of the p16
INK4A/CDKN2A promoter in human oral cancer cells and
normal human oral keratinocytes,” Oral Oncology, vol. 35,
no. 5, pp. 516–522, 1999.
 M. Viswanathan, N. Tsuchida, and G. Shanmugam, “Pro-
moter hypermethylation profile of tumor-associated genes
p16, p15, hMLH1, MGMT and E-cadherin in oral squamous
cell carcinoma,” International Journal of Cancer, vol. 105, no.
1, pp. 41–46, 2003.
 R. J. Shaw, T. Liloglou, S. N. Rogers et al., “Promoter methy-
lation of P16, RARβ, E-cadherin, cyclin A1 and cytoglobin in
10Journal of Oncology
oral cancer: quantitative evaluation using pyrosequencing,”
British Journal of Cancer, vol. 94, no. 4, pp. 561–568, 2006.
 J. K. Lee, M. J. Kim, S. P. Hong, and S. D. Hong, “Inactivation
patterns of p16/INK4A in oral squamous cell carcinomas,”
Experimental and Molecular Medicine, vol. 36, no. 2, pp. 165–
 T. N. Tran, Y. Liu, M. Takagi, A. Yamaguchi, and H.
Fujii, “Frequent promoter hypermethylation of RASSF1A
and p16INK4a and infrequent allelic loss other than 9p21
in betel-associated oral carcinoma in a Vietnamese non-
smoking/non-drinking female population,” Journal of Oral
Pathology and Medicine, vol. 34, no. 3, pp. 150–156, 2005.
 P. F. Su, W. L. Huang, H. T. Wu, C. H. Wu, T. Y. Liu,
and S. Y. Kao, “P16INK4A promoter hypermethylation is
associated with invasiveness and prognosis of oral squamous
cell carcinoma in an age-dependent manner,” Oral Oncology,
vol. 46, no. 10, pp. 734–739, 2010.
 R. Sailasree, A. Abhilash, K. M. Sathyan, K. R. Nalinakumari,
S. Thomas, and S. Kannan, “Differential roles of p16INK4A
and p14ARF genes in prognosis of oral carcinoma,” Cancer
Epidemiology, Biomarkers & Prevention, vol. 17, no. 2, pp.
 E. Ishida, M. Nakamura, M. Ikuta et al., “Promotor hyper-
methylation of p14ARF is a key alteration for progression of
oral squamous cell carcinoma,” Oral Oncology, vol. 41, no. 6,
pp. 614–622, 2005.
 K. Ogi, M. Toyota, M. Ohe-Toyota et al., “Aberrant methy-
lation of multiple genes and clinicopathological features in
oral squamous cell carcinoma,” Clinical Cancer Research, vol.
8, no. 10, pp. 3164–3171, 2002.
 P. Sinha, S. Bahadur, A. Thakar et al., “Significance of pro-
moter hypermethylation of p16 gene for margin assessment
 K. Kato, A. Hara, T. Kuno et al., “Aberrant promoter hyper-
methylation of p16 and MGMT genes in oral squamous cell
carcinomas and the surrounding normal mucosa,” Journal of
Cancer Research and Clinical Oncology, vol. 132, no. 11, pp.
 G.ˇSupi´ c,R.Kozomara,M.Brankovi´ c-Magi´ c,N.Jovi´ c,andZ.
Magi´ c, “Gene hypermethylation in tumor tissue of advanced
oral squamous cell carcinoma patients,” Oral Oncology, vol.
45, no. 12, pp. 1051–1057, 2009.
 G. Berx and F. van Roy, “Involvement of members of
the cadherin superfamily in cancer,” Cold Spring Harbor
Perspectives in Biology, vol. 1, no. 6, p. a003129, 2009.
 H. W. Chang, V. Chow, K. Y. Lam, W. I. Wei, and A. P.
WingYuen, “Loss of E-cadherin expression resulting from
prognostic significance,” Cancer, vol. 94, no. 2, pp. 386–392,
mechanism,” Annual Review of Pharmacology and Toxicology,
vol. 45, pp. 629–656, 2005.
 S. L. B. Rosas, W. Koch, M. D. G. Da Costa Car-
valho et al., “Promoter hypermethylation patterns of
p16, O6-methylguanine-DNA-methyltransferase, and death-
associated protein kinase in tumors and saliva of Head &
Neck cancer patients,” Cancer Research, vol. 61, no. 3, pp.
 V. Kulkarni and D. Saranath, “Concurrent hypermethylation
of multiple regulatory genes in chewing tobacco associated
oral squamous cell carcinomas and adjacent normal tissues,”
Oral Oncology, vol. 40, no. 2, pp. 145–153, 2004.
 S.-H. Huang, H.-S. Lee, K. Mar, D.-D. Ji, M.-S. Huang,
and K.-T. Hsia, “Loss expression of O6-methylguanine DNA
methyltransferase by promoter hypermethylation and its
relationship to betel quid chewing in oral squamous cell
Radiology and Endodontology, vol. 109, no. 6, pp. 883–889,
 C. Zuo, L. Ai, P. Ratliff et al., “O6-methylguanine-DNA
methyltransferase gene: epigenetic silencing and prognostic
value in Head & Neck squamous cell carcinoma,” Cancer
Epidemiology Biomarkers and Prevention, vol. 13, no. 6, pp.
 J. Paluszczak, P. Misiak, M. Wierzbicka, A. Wo´ zniak, and
W. Baer-Dubowska, “Frequent hypermethylation of DAPK,
RARbeta, MGMT, RASSF1A and FHIT in laryngeal squa-
mous cell carcinomas and adjacent normal mucosa,” Oral
Oncology, vol. 47, no. 2, pp. 104–107, 2011.
 M. Sawhney, N. Rohatgi, J. Kaur et al., “MGMT expression
in oral precancerous and cancerous lesions: correlation with
progression, nodal metastasis and poor prognosis,” Oral
Oncology, vol. 43, no. 5, pp. 515–522, 2007.
 G. Jiang, Z.-P. Wei, D.-S. Pei, Y. Xin, Y.-Q. Liu, and J.-
N. Zheng, “A novel approach to overcome temozolomide
resistance in glioma and melanoma: inactivation of MGMT
by gene therapy,” Biochemical and Biophysical Research
Communications, vol. 406, no. 3, pp. 311–314, 2011.
 S. Sharma, F. Salehi, B. W. Scheithauer, F. Rotondo, L. V.
progression, diagnosis, treatment and prognosis,” Anticancer
Research, vol. 29, no. 10, pp. 3759–3768, 2009.
 M. Airoldi, G. Cortesina, C. Giordano, F. Pedani, and C.
Bumma, “Ifosfamide in the treatment of Head & Neck
cancer,” Oncology, vol. 65, supplement 2, no. 2, pp. 37–43,
 M. S. Kies, D. H. Boatright, G. Li et al., “Phase II trial of
inductionchemotherapy followed by surgery for squamous
cell carcinoma of the oral tongue in young adults,” Head &
Neck. In press.
 A. M. Michie, A. M. McCaig, R. Nakagawa, and M. Vukovic,
“Death-associated protein kinase (DAPK) and signal trans-
duction: regulation in cancer,” FEBS Journal, vol. 277, no. 1,
pp. 74–80, 2010.
 M. Sanchez-Cespedes, M. Esteller, L. Wu et al., “Gene
promoter hypermethylation in tumors and serum of Head
& Neck cancer patients,” Cancer Research, vol. 60, no. 4, pp.
 W. J. Kong, S. Zhang, C. Guo, S. Zhang, Y. Wang,
and D. Zhang, “Methylation-associated silencing of death-
associated protein kinase gene in laryngeal squamous cell
cancer,” Laryngoscope, vol. 115, no. 8, pp. 1395–1401, 2005.
 R. P. Dikshit, A. Gillio-Tos, P. Brennan et al., “Hypermethyla-
tion, risk factors, clinical characteristics, and survival in 235
vol. 110, no. 8, pp. 1745–1751, 2007.
 G. Supic, R. Kozomara, N. Jovic, K. Zeljic, and Z. Magic,
“Prognostic significance of tumor-related genes hyperme-
thylation detected in cancer-free surgical margins of oral
squamous cell carcinomas,” Oral Oncology, vol. 47, no. 8, pp.
 Y. Kudo, T. Tsunematsu, and T. Takata, “Oncogenic role
of RUNX3 in Head & Neck cancer,” Journal of Cellular
Biochemistry, vol. 112, no. 2, pp. 387–393, 2011.
 F. Gao, C. Huang, M. Lin et al., “Frequent inactivation
of RUNX3 by promoter hypermethylation and protein
Journal of Oncology 11
mislocalization in oral squamous cell carcinomas,” Journal of
Cancer Research and Clinical Oncology, vol. 135, no. 5, pp.
 T. Reya and H. Clevers, “Wnt signalling in stem cells and
cancer,” Nature, vol. 434, no. 7035, pp. 843–850, 2005.
 N. Barker and H. Clevers, “Mining the Wnt pathway for
cancer therapeutics,” Nature Reviews Drug Discovery, vol. 5,
no. 12, pp. 997–1014, 2006.
 R. Fodde and T. Brabletz, “Wnt/β-catenin signaling in cancer
stemness and malignant behavior,” Current Opinion in Cell
Biology, vol. 19, no. 2, pp. 150–158, 2007.
 A. A. Molinolo, P. Amornphimoltham, C. H. Squarize, R. M.
Castilho,V.Patel, andJ.S.Gutkind,“Dysregulated molecular
networks in Head & Neck carcinogenesis,” Oral Oncology,
vol. 45, no. 4-5, pp. 324–334, 2009.
 C. J. Marsit, M. D. McClean, C. S. Furniss, and K. T. Kelsey,
“Epigenetic inactivation of the SFRP genes is associated with
drinking, smoking and HPV in Head & Neck squamous cell
carcinoma,” International Journal of Cancer, vol. 119, no. 8,
pp. 1761–1766, 2006.
 Y. Sogabe, H. Suzuki, M. Toyota et al., “Epigenetic inac-
tivation of SFRP genes in oral squamous cell carcinoma,”
International Journal of Oncology, vol. 32, no. 6, pp. 1253–
 G. Pannone, P. Bufo, A. Santoro et al., “WNT pathway in oral
Reports, vol. 24, no. 4, pp. 1035–1041, 2010.
 C. H. Lee, Y. J. Hung, C. Y. Lin, P. H. Hung, H. W. Hung,
and Y. S. Shieh, “Loss of SFRP1 expression is associated with
aberrant β-catenin distribution and tumor progression in
mucoepidermoid carcinoma of salivary glands,” Annals of
Surgical Oncology, vol. 17, no. 8, pp. 2237–2246, 2010.
 M. P´ erez-Say´ ans, J. M. Su´ arez-Pe˜ naranda, M. Herranz-
Carnero et al., “The role of the adenomatous polyposis coli
(APC) in oral squamous cell carcinoma,” Oral Oncology, vol.
48, no. 1, pp. 56–60, 2012.
 H. Uesugi, K. Uzawa, K. Kawasaki et al., “Status of reduced
expression and hypermethylation of the APC tumor sup-
pressor gene in human oral squamous cell carcinoma,”
International Journal of Molecular Medicine, vol. 15, no. 4, pp.
 M. Fujii, N. Katase, M. Lefeuvre et al., “Dickkopf (Dkk)-3
and β-catenin expressions increased in the transition from
normal oral mucosal to oral squamous cell carcinoma,”
Journal of Molecular Histology, vol. 42, no. 6, pp. 499–504,
 G. Ravindran and H. Devaraj, “Aberrant expression of
β-catenin and its association with ΔNp63, Notch-1, and
clinicopathological factors in oral squamous cell carcinoma,”
Clinical Oral Investigations. In press.
 Y. Ying and Q. Tao, “Epigenetic disruption of the WNT/β-
catenin signaling pathway in human cancers,” Epigenetics,
vol. 4, no. 5, pp. 34–39, 2009.
silencing of tumor-suppressor microRNAs in cancer,” Onco-
gene, vol. 31, no. 13, pp. 1609–1622, 2012.
 A. Lujambio, A. Portela, J. Liz et al., “CpG island
hypermethylation-associated silencing of non-coding RNAs
transcribed from ultraconserved regions in human cancer,”
Oncogene, vol. 29, no. 48, pp. 6390–6401, 2010.
 K. C. Miranda, T. Huynh, Y. Tay et al., “A pattern-based
method for the identification of microRNA binding sites and
their corresponding heteroduplexes,” Cell, vol. 126, no. 6, pp.
 H. Iwama, T. Masaki, and S. Kuriyama, “Abundance of
microRNA target motifs in the 3?-UTRs of 20 527 human
genes,” FEBS Letters, vol. 581, no. 9, pp. 1805–1810, 2007.
 A. M. Duursma, M. Kedde, M. Schrier, C. Le Sage, and R.
Agami, “miR-148 targets human DNMT3b protein coding
region,” RNA, vol. 14, no. 5, pp. 872–877, 2008.
 A. Esquela-Kerscher and F. J. Slack, “Oncomirs—microRNAs
with a role in cancer,” Nature Reviews Cancer, vol. 6, no. 4,
pp. 259–269, 2006.
 S. M. Hammond, “MicroRNAs as tumor suppressors,”
Nature Genetics, vol. 39, no. 5, pp. 582–583, 2007.
 A. Lujambio and M. Esteller, “How epigenetics can explain
8, no. 3, pp. 377–382, 2009.
 A. Uesugi, K.-I. Kozaki, T. Tsuruta et al., “The tumor sup-
pressive microRNA miR-218 targets the mTOR component
rictor and inhibits AKT phosphorylation in oral cancer,”
Cancer Research, vol. 71, no. 17, pp. 5765–5778, 2011.
 J. Zheng, H. Xue, T. Wang et al., “miR-21 downregulates
the tumor suppressor P12 CDK2AP1and stimulates cell
proliferation and invasion,” Journal of Cellular Biochemistry,
vol. 112, no. 3, pp. 872–880, 2011.
 Y. Saito, G. Liang, G. Egger et al., “Specific activation of
microRNA-127 with downregulation of the proto-oncogene
BCL6 by chromatin-modifying drugs in human cancer cells,”
Cancer Cell, vol. 9, no. 6, pp. 435–443, 2006.
 L. Han, P. D. Witmer, E. Casey, D. Valle, and S. Sukumar,
“DNA methylation regulates microRNA expression,” Cancer
Biology and Therapy, vol. 6, no. 8, pp. 1284–1288, 2007.
 A. Lujambio, S. Ropero, E. Ballestar et al., “Genetic unmask-
ing of an epigenetically silenced microRNA in human cancer
cells,” Cancer Research, vol. 67, no. 4, pp. 1424–1429, 2007.
 M. Toyota, H. Suzuki, Y. Sasaki et al., “Epigenetic silencing
of microRNA-34b/c and B-cell translocation gene 4 is
associated with CpG island methylation in colorectal cancer,”
Cancer Research, vol. 68, no. 11, pp. 4123–4132, 2008.
 A. Lujambio, G. A. Calin, A. Villanueva et al., “A microRNA
DNA methylation signature for human cancer metastasis,”
Proceedings of the National Academy of Sciences of the United
States of America, vol. 105, no. 36, pp. 13556–13561, 2008.
 S. M. Langevin, R. A. Stone, C. H. Bunker et al., “MicroRNA-
137 promoter methylation is associated with poorer overall
and neck,” Cancer, vol. 117, no. 7, pp. 1454–1462, 2011.
 K. I. Kozaki, I. Imoto, S. Mogi, K. Omura, and J. Inazawa,
“Exploration of tumor-suppressive microRNAs silenced by
DNA hypermethylation in oral cancer,” Cancer Research, vol.
68, no. 7, pp. 2094–2105, 2008.
 E. D. Wiklund, S. Gao, T. Hulf et al., “MicroRNA alterations
and associated aberrant DNA methylation patterns across
multiple sample types in oral squamous cell carcinoma,”
PLoS ONE, vol. 6, no. 11, Article ID e27840, 2011.
 F. Zhi, X. Chen, S. Wang et al., “The use of hsa-miR-21,
hsa-miR-181b and hsa-miR-106a as prognostic indicators of
astrocytoma,” European Journal of Cancer, vol. 46, no. 9, pp.
 J. Silber, D. A. Lim, C. Petritsch et al., “miR-124 and miR-
137 inhibit proliferation of glioblastoma multiforme cells
and induce differentiation of brain tumor stem cells,” BMC
Medicine, vol. 6, article 14, 2008.
 L. T. Bemis, R. Chen, C. M. Amato et al., “MicroRNA-
137 targets microphthalmia-associated transcription factor
in melanoma cell lines,” Cancer Research, vol. 68, no. 5, pp.
12Journal of Oncology Download full-text
 M. P. Rosin, W. L. Lam, C. Poh et al., “3p14 and 9p21 loss
is a simple tool for predicting second oral malignancy at
previously treated oral cancer sites,” Cancer Research, vol. 62,
no. 22, pp. 6447–6450, 2002.
 L. Mao, J. S. Lee, Y. H. Fan et al., “Frequent microsatellite
alterations at chromosomes 9p21 and 3p14 in oral prema-
lignant lesions and their value in cancer risk assessment,”
Nature Medicine, vol. 2, no. 6, pp. 682–685, 1996.
p16INK4a and p14ARF in patients with severe oral epithelial
dysplasia,” Cancer Research, vol. 62, no. 18, pp. 5295–5300,
 J. Cao, J. Zhou, Y. Gao et al., “Methylation of p16 CpG island
associated with malignant progression of oral epithelial dys-
plasia: a prospective cohort study,” Clinical Cancer Research,
vol. 15, no. 16, pp. 5178–5183, 2009.
 M. Takeshima, M. Saitoh, K. Kusano et al., “High frequency
of hypermethylation of p14, p15 and p16 in oral pre-
cancerous lesions associated with betel-quid chewing in Sri
Lanka,” Journal of Oral Pathology and Medicine, vol. 37, no. 8,
pp. 475–479, 2008.
 G. L. Hall, R. J. Shaw, E. A. Field et al., “p16 promoter methy-
lation is a potential predictor of malignant transformation
in oral epithelial dysplasia,” Cancer Epidemiology Biomarkers
and Prevention, vol. 17, no. 8, pp. 2174–2179, 2008.
 M. Gasco, A. K. Bell, V. Heath et al., “Epigenetic inactivation
of 14-3-3 σ in oral carcinoma: association with p16INK4a
silencing and human papillomavirus negativity,” Cancer
Research, vol. 62, no. 7, pp. 2072–2076, 2002.
 N. K. Long, K. Kato, T. Yamashita et al., “Hypermethylation
of the RECK gene predicts poor prognosis in oral squamous
cell carcinomas,” Oral Oncology, vol. 44, no. 11, pp. 1052–
 M. Toyota, N. Ahuja, M. Ohe-Toyota, J. G. Herman, S. B.
Baylin, and J. P. J. Issa, “CpG island methylator phenotype
in colorectal cancer,” Proceedings of the National Academy of
Sciences of the United States of America, vol. 96, no. 15, pp.
 M. Toyota, N. Ahuja, H. Suzuki et al., “Aberrant methylation
in gastric cancer associated with the CpG island methylator
phenotype,” Cancer Research, vol. 59, no. 21, pp. 5438–5442,
phenotype (CIMP) in oral cancer: associated with a marked
inflammatory response and less aggressive tumour biology,”
Oral Oncology, vol. 43, no. 9, pp. 878–886, 2007.
 C. J. Marsit, E. A. Houseman, B. C. Christensen et al.,
“Examination of a CpG island methylator phenotype and
implications of methylation profiles in solid tumors,” Cancer
Research, vol. 66, no. 21, pp. 10621–10629, 2006.
 W. Sun, Y. Liu, C. A. Glazer et al., “TKTL1 is activated by
promoter hypomethylation and contributes to Head & Neck
squamous cell carcinoma carcinogenesis through increased
aerobic glycolysis and HIF1α stabilization,” Clinical Cancer
Research, vol. 16, no. 3, pp. 857–866, 2010.
 Y. Yamada, L. Jackson-Grusby, H. Linhart et al., “Opposing
effects of DNA hypomethylation on intestinal and liver
carcinogenesis,” Proceedings of the National Academy of
Sciences of the United States of America, vol. 102, no. 38, pp.
 P. W. Laird, L. Jackson-Grusby, A. Fazeli et al., “Suppression
of intestinal neoplasia by DNA hypomethylation,” Cell, vol.
81, no. 2, pp. 197–205, 1995.
 J. A. Gasche, J. Hoffmann, C. R. Boland, and A. Goel,
“Interleukin-6 promotes tumorigenesis by altering DNA
methylation in oral cancer cells,” International Journal of
Cancer, vol. 129, no. 5, pp. 1053–1063, 2011.
 S. Baba, Y. Yamada, Y. Hatano et al., “Global DNA
hypomethylation suppresses squamous carcinogenesis in the
tongue and esophagus,” Cancer Science, vol. 100, no. 7, pp.
 K. Subbalekha, A. Pimkhaokham, P. Pavasant et al., “Detec-
tion of LINE-1s hypomethylation in oral rinses of oral
squamous cell carcinoma patients,” Oral Oncology, vol. 45,
no. 2, pp. 184–191, 2009.
 C. T. Viet and B. L. Schmidt, “Methylation array analysis
of preoperative and postoperative saliva DNA in oral cancer
patients,” Cancer Epidemiology Biomarkers and Prevention,
vol. 17, no. 12, pp. 3603–3611, 2008.
 C. A. Righini, F. De Fraipont, J. F. Timsit et al., “Tumor-
specific methylation in saliva: a promising biomarker for
early detection of Head & Neck cancer recurrence,” Clinical
Cancer Research, vol. 13, no. 4, pp. 1179–1185, 2007.
 M. L´ opez, J. M. Aguirre, N. Cuevas et al., “Gene promoter
hypermethylation in oral rinses of leukoplakia patients—
a diagnostic and/or prognostic tool?” European Journal of
Cancer, vol. 39, no. 16, pp. 2306–2309, 2003.
 M. Esteller, M. Sanchez-Cespedes, R. Resell, D. Sidransky, S.
hypermethylation of tumor suppressor genes in serum DNA
from non-small cell lung cancer patients,” Cancer Research,
vol. 59, no. 1, pp. 67–70, 1999.
 I. H. N. Wong, Y. M. D. Lo, J. Zhang et al., “Detection of
aberrant p16 methylation in the plasma and serum of liver
cancer patients,” Cancer Research, vol. 59, no. 1, pp. 71–73,
 K. Hibi, C. R. Robinson, S. Booker et al., “Molecular
detection of genetic alterations in the serum of colorectal
cancer patients,” Cancer Research, vol. 58, no. 7, pp. 1405–
 A. L. Carvalho, C. Jeronimo, M. M. Kim et al., “Evaluation
of promoter hypermethylation detection in body fluids as
a screening/diagnosis tool for Head & Neck squamous cell
carcinoma,” Clinical Cancer Research, vol. 14, no. 1, pp. 97–
 Y. Nakahara, S. Shintani, M. Mihara, S. Hino, and H.
Hamakawa, “Detection of p16 promoter methylation in the
serum of oral cancer patients,” International Journal of Oral
and Maxillofacial Surgery, vol. 35, no. 4, pp. 362–365, 2006.
 K. Hibi, M. Taguchi, H. Nakayama et al., “Molecular
detection of p16 promoter methylation in the serum of
patients with esophageal squamous cell carcinoma,” Clinical
Cancer Research, vol. 7, no. 10, pp. 3135–3138, 2001.