Received June 11, 2008; Last revision October 3, 2008;
Accepted October 21, 2008
S.P. Barros* and S. Offenbacher
Center for Oral and Systemic Diseases, Department of
Periodontology, School of Dentistry, University of North
Carolina at Chapel Hill, Room 222, CB 7455, Chapel Hill, NC,
USA 27599; *corresponding author, firstname.lastname@example.org
J Dent Res 88(5):400-408, 2009
Genetic information is encoded not only by the linear
sequence of DNA, but also by epigenetic modifica-
tions of chromatin structure that include DNA methy-
lation and covalent modifications of the proteins that
bind DNA. These “epigenetic marks” alter the struc-
ture of chromatin to influence gene expression.
Methylation occurs naturally on cytosine bases at CpG
sequences and is involved in controlling the correct
expression of genes. DNA methylation is usually asso-
ciated with triggering histone deacetylation, chromatin
condensation, and gene silencing. Differentially meth-
ylated cytosines give rise to distinct patterns specific
for each tissue type and disease state. Such methyla-
tion-variable positions (MVPs) are not uniformly dis-
tributed throughout our genome, but are concentrated
among genes that regulate transcription, growth,
metabolism, differentiation, and oncogenesis. Altera-
tions in MVP methylation status create epigenetic
patterns that appear to regulate gene expression pro-
files during cell differentiation, growth, and develop-
ment, as well as in cancer. Environmental stressors
including toxins, as well as microbial and viral expo-
sures, can change epigenetic patterns and thereby
effect changes in gene activation and cell phenotype.
Since DNA methylation is often retained following
cell division, altered MVP patterns in tissues can accu-
mulate over time and can lead to persistent alterations
in steady-state cellular metabolism, responses to stim-
uli, or the retention of an abnormal phenotype, reflect-
ing a molecular consequence of gene-environment
interaction. Hence, DNA epigenetics constitutes the
main and previously missing link among genetics,
disease, and the environment. The challenge in oral
biology will be to understand the mechanisms that
modify MVPs in oral tissues and to identify those
epigenetic patterns that modify disease pathogenesis
or responses to therapy.
Key words: epigenetics, DNA methylation,
gene regulation, infection, inflammation, field
environment and Genotype to
Phenotype and disease
as the “breakthrough of the year” (Kennedy, 2007). This report emphasized the
number of studies in that year that led to a new view of human genetic diversity,
with appreciation of the extent to which our genomic sequences differ from per-
son to person, and the implications of these variations for potentially decipher-
ing the complexity of the biological systems in the human body. Undoubtedly,
variations in the linear sequence of the genetic code, like single nucleotide poly-
morphisms (SNPs), may play a key role in explaining inter-individual differ-
ences in structure and function, as well as insight into disease susceptibility and
resistance. However, the function of our genome is also dependent upon epige-
netic mechanisms—which are by definition “beyond the genome”, and include
alterations of chromatin structure, involving covalent modification of the central
DNA molecule itself, as well as the complex macromolecules that form chro-
matin. The rapidly evolving field of epigenetics is contributing to our under-
standing of gene-environment interactions, as epigenetic mechanisms exert an
additional layer of transcriptional control that regulates gene expression.
our years after the unveiling of the complete sequence of the human genome
in 2003, “Human Genetic Variation” was recognized by the journal Science
Prior to the middle of the twentieth century, before DNA was given a special
status in biology, the developmental biologist and evolutionist Conrad H.
Waddington (1905-1975) emphasized that genetics and developmental biol-
ogy were related (Pennisi, 2001), hypothesizing that patterns of gene expres-
sion, turning genes on and off, and not the genes themselves, define each cell
type, thus linking genes and gene action to development. To denote the
dynamic actions leading from the genotype to the phenotype, Waddington
coined the term ‘epigenetics’ from the Greek word epigenesis, referring to
embryology and genetics as “a gradual coming into being of newly formed
organs and tissues out of an initially undifferentiated mass”. In this way,
Waddington indicated that an epigenetic landscape underlies each developing
organism, referring to the existence of a complex network in which genetic
interactions, the feedback and “feedforward” relationships among DNA, pro-
teins, and other internal and external biochemical compounds are highly
intermingled (Van de Vijver et al., 2002). In 1975, two papers were published
(Holliday and Pugh, 1975; Riggs, 1975) outlining a molecular model for the
switching of gene activities, and also the heritability of gene activity or inac-
tivity. This model was based on the enzymatic methylation of cytosine in
DNA. The suggestion was that DNA methylation could have strong effects on
gene expression, and changes in DNA methylation may therefore explain the
switching on and off of genes during development, and that the pattern of
methylation could be heritable, persisting through cell divisions. That is, dur-
ing DNA replication, the methylation patterns in cytosine bases would also be
conserved during strand duplication. More recently, the scope of epigenetics
crItIcAL reVIews IN orAL bIoLoGy & MedIcINe
J Dent Res 88(5) 2009 Epigenetics 401
extends to heritable modifications
of genes, leading to alteration in
the expression of specific DNA
sequences that vary among differ-
ent tissues within the same orga-
nism and cannot be explained by
changes in DNA sequence.
Epigenetics, as the term sug-
gests, can be seen as a major turn
away from molecular biology’s
Central Dogma, recognizing that
there are epigenetic inheritance
sequence-dependent DNA varia-
tions can be transmitted in cell,
tissue, and organismal lineages.
Thus, current epigenetics not only
offers new insights into gene regu-
lation and heredity, but it also pro-
foundly challenges the way we
think about evolution, genetics,
and development. Most interest-
ingly, it suggests testable mecha-
factors (ranging from stress to
infection) can influence genetic
expression. Furthermore, these
potential epigenetic modifications
can occur throughout the lifetime
of the organism, beginning as early
as the intra-uterine environment,
and can accumulate in tissues and
cells over time to modify gene
expression patterns and cellular
ePIGeNoMe ANd dNA MetHyLAtIoN
The most-studied epigenetic modification of DNA in mam-
mals is methylation of cytosine in CpG dinucleotides (Bird,
2002). The other main group of epigenetic modifications
includes post-translational modification of histones, princi-
pally changes in phosphorylation, acetylation, and ubiquitiny-
lation status. Often, these epigenetic mechanisms are coupled
and interact to modify chromatin structure and function. DNA
methylation is a covalent biochemical modification that, in the
mammalian genome, takes place predominantly at cytosine
bases that are located 5´ to a guanosine (Fig. 1). CpGs are
vastly under-represented in the genome, as compared with
what would be expected by chance (0.23 in the human
genome). Furthermore, CpG-rich regions are not evenly dis-
tributed throughout the genome, but appear most often among
the promoter regions and first exons of specific genes (Larsen
et al., 1992). Within the structure of chromatin, DNA is
complexed with histone proteins to form octamers around
which DNA loops to form the nucleosome, the individual
packaging unit of genomic DNA. The enzymes involved in
DNA methylation include the DNA methyltransferases
(DNMTs), which establish and maintain DNA methylation pat-
terns using S-adenosyl methionine (SAM) as the methyl donor.
When DNA becomes methylated, those methyl groups pro-
trude from the cytosine nucleotides into the major groove of
the DNA to displace transcription factors that normally bind to
the DNA (Hark et al., 2000). The exposed methylation sites
attract methyl-binding proteins, the methyl-CpG-binding
domain proteins (MBDs), which are involved in ‘reading’
methylation marks (Loenen, 2006) to affect chromatin conden-
sation by recruiting histone deacetylases that covalently mod-
ify the tails of histone proteins (Vucic et al., 2008). Histone
modification results in gene silencing and chromatin compac-
tion (Bird and Wolffe, 1999) (Fig. 2).
It is important that methylation patterns be generally con-
served following DNA replication carrying epigenetic patterns
to cellular progeny, thereby creating a link or communication
that conveys the developmental or environmental pressures of
preceding cellular generations. In 2004, Fazzari and Greally
pointed out that the methylation of CpG-rich promoters is used
by mammals to prevent transcriptional initiation and to ensure
Figure 1. Methylation of DNA occurs at cytosine residues when present as CG dinucleotides. Methylation
occurs by the addition of a methyl group at the 5’ site of cytosine (depicted as shaded sphere).
402 Barros & Offenbacher J Dent Res 88(5) 2009
the silencing of genes on the inactive X chromosome, imprinted
genes, and parasitic DNAs. The potential role of DNA methyla-
tion in tissue-specific gene expression has been explored more
recently, since it was realized that CpG methylation can regulate
tissue-specific gene expression and can be influenced by exter-
nal stressors, environmental toxins, and aging (Dolinoy et al.,
2007; Hanson and Gluckman, 2008), potentially increasing or
decreasing the level of transcription, depending on whether
the methylation inactivates a positive or negative regulatory
The ontological role of DNA methylation, through transcrip-
tional silencing, contributes to an epigenetic regulation of the
embryonic and morphogenetic developmental gene expression
program (Holliday and Pugh, 1975). DNA methylation is also
recognized as an ancient host defense system designed to protect
against exogenous parasitic nucleic acid sequence elements or
deleterious endogenous sequences, which have been evolution-
arily incorporated and retained vestigially within our genome
(Doerfler, 1991; Yoder et al., 1997). In normal cells, DNA methy-
lation occurs predominantly in repetitive genomic regions, includ-
ing satellite DNA and parasitic elements (such as long interspersed
transposable elements [LINES], short interspersed transposable
elements [SINES], and endogenous retroviruses [Yoder et al.,
1997]), offering a mechanism by
which the environment can stably
change gene expression. Changes in
methylation status can also regulate
which, in turn, modulates post-tran-
scriptional gene expression and plays
important roles in essential processes,
such as differentiation, cell growth,
and cell death (Miska, 2005; Zamore
and Haley, 2005). The key concept is
that the epigenome, consisting of
chromatin and its modifications, func-
tions as an interface between the
inherited genome and the dynamism
imposed by the environment, and such
interaction promotes epigenetic modi-
fications that are specific DNA meth-
ylation patterns, which then result in a
relatively stable or homeostatic pro-
file of gene expression. This has been
referred to (Feinberg, 2008) as a meta-
stable condition that represents an epi-
genetic modification, resulting in a
new cellular or tissue homeostatic
“set-point” with a new range of gene
expression patterns that differ from
those of the original state. Chromatin
modifications, including CpG methy-
lations, allow for sculpting of the
epigenome during development, mod-
ified by individual environmental
unique identity, even for monozygotic
twins (Fraga et al., 2005).
In a landmark publication, Fraga and colleagues (2005)
reported, in a large cohort of monozygotic twins, that DNA
methylation increased over time within different tissues and cell
types, including oral epithelial, lymphocytic, muscle, and fat.
Most importantly, tissues from identical twins who were 3 years
of age displayed low levels of DNA methylation and virtually
identical gene expression profiles for the different tissues obtained
from each twin, reflecting the high concordance of gene expres-
sion patterns associated with monozygosity, i.e., identical DNA
sequences. However, when they compared DNA methylation
patterns in identical twins at the age of 50, there were higher
DNA methylation levels in general. The gene expression patterns
in the twins were significantly different, and those differences
were mapped to the alterations in DNA methylation status. Thus,
the authors suggested that the “distinct profiles of DNA methyla-
tion and histone acetylation patterns that among different tissues
(that) arise during the lifetime of monozygotic twins may con-
tribute to the explanation of some of their phenotypic discor-
dances and underlie their differential frequency/onset of common
diseases”, dependent upon the cumulative history of external
exposures (i.e., environmental, nutritional, toxins, and infec-
tious) that reprogram epigenetic status.
Figure 2. The shaded sphere depicts the octameric histone complex, which forms the nucleosome with
the acetylated tails of histones and the cytosines of the CpG sites in an unmethylated state, shown as
open white circles. In this conformation, the chromatin is loosely packed and available for the binding
of transcriptional activating proteins, which, by the action of RNA polymerase II, synthesize mRNA.
The action of DNA methyl transferase (DNMT) methylates the cytosine residues, depicted as red
circles, which provide a docking site for the methyl binding domain proteins (MBD), which aggregate
in conjunction with the action of the histone deacetylase, which cleaves the histone acetyl group. Both
of these serve to alter the structure of the chromatin by causing a condensation that impedes the access
of the transcriptional activating proteins and thereby blocks mRNA synthesis. Alternatively, the normal
active structure of chromatin can become inaccessible for the binding of transcriptional activating pro-
teins by the action of CpG methylation at sites that sterically hinder the binding of activating proteins,
independent of MBD aggregation.
J Dent Res 88(5) 2009 Epigenetics 403
Diploid organisms carry two copies of every autosomal gene, one
from each parent. In the great majority of cases, the two copies are
either repressed or transcribed identically, but this is not the case
for genes that exhibit the phenomenon of parental imprinting.
Observations through the centuries have suggested that the genes
passed on by each parent had somehow been permanently
marked—or “imprinted”, as it eventually came to be known—so
that the expression patterns of the maternal and paternal genes
differ in their progeny. Genomic imprinting in mammals repre-
sents a situation where there is non-equivalence in the expression
of alleles at certain gene loci, dependent on the parent of origin
(Reik and Walter, 1998).The expression of either the paternally or
maternally inherited allele is consistently repressed, resulting in
mono-allelic expression of a particular gene. Thus, imprinted
genes show markedly different behavior, depending on their
parental origin. The same pattern of mono-allelic expression is
faithfully transmitted to daughter cells following cell division.
Imprinting is not a phenomenon entirely unique to mammals,
since it also happens in plants, where most commonly the paternal
genes are imprinted (Grant-Downton and Dickinson, 2005).
However, it is only since 1991 that researchers have begun to
isolate a variety of genes whose expression depended upon their
parents of origin. That year, researchers identified two genes,
Igf2r [insulin-like growth factor-2 receptor] and H19, that are
active only when inherited from the mother; a third, called Igf2
[insulin-like growth factor-2], is turned on only when inherited
from the father. Those findings raised essential questions on
how genomes become marked differentially during gametogen-
esis, and how this marking is maintained on the gene throughout
development, as well as prompting a broader search for other
imprinted genes (Bartolomei et al., 1991; DeChiara et al., 1991;
Constancia et al., 1998). DNA methylation appears to be the key
mechanism by which one copy of a gene is preferentially
silenced according to parental origin and maintained during cell
division by 5-cytosine DNA methyltransferase-1 (DNMT1)
(Okano et al., 1999; Miranda and Jones, 2007). DNMT1 prefer-
entially methylates hemi-methylated CpG sites, thus copying
established methylation patterns to the newly synthesized DNA
strands. Reik and Walter (1998) proposed that at least a subset
of ~ 100 genes of the ~20,000-25,000 genes in the mammalian
genome are thought to be imprinted, and a list of known
imprinted genes can be accessed at http://www.geneimprint.
com/site/genes-by-status. Although it has not been demon-
strated, it appears likely that many small RNA sequences that
arise from non-coding regions, and are also involved in gene
regulation and metabolism, might be candidates for epigenetic
marks and possibly demonstrate parent-of-origin imprinting
properties. This possibility is suggested by the presence of CpG-
rich regions that are occasionally present in non-coding regions,
including miRNAs. The concept of imprinting is an interesting
example of interactions between heterologous genomes (mater-
nal and paternal) that seek to establish some sort of mutually
acceptable state of equilibrium. The role of imprinting in IGF2
regulation in placental and fetal growth is a fascinating example.
This insulin-like growth hormone is the major somatic growth
factor for the fetus, since it enhances placental nutrient exchange
of glucose for fetal growth, and it serves as a pluripotent tissue
growth hormone for the fetus. IGF2 impairment restricts fetal
growth. It is perhaps no surprise that the paternal genome serves
to enhance IGF-II secretion to stimulate the transfer of nutrients
from mother to fetus to make the baby as large and healthy as
possible, whereas the maternal genome seeks to attenuate this
response by controlling the expression level of the receptor for
this hormone. Thus, one can appreciate that the epigenetic
response that mediates IGF2 gene expression strikes a balance
between competing genetic programs. It now appears that epi-
genetic mechanisms serve to mediate genomic conflict that can
occur between host and exogenous genomes such as those pro-
vided by viral infections or even the commensal microbiome.
Differences in programmed gene expression that result in the
development of organs, tissues, and cell lineages generally occur
without changes to the sequence of our DNA (with one or two
exceptions, e.g., immunoglobulin synthesis), indicating that
development is, by definition, epigenetic (Reik, 2007). Certain
cell lineages are dependent upon epigenetic programs that
sequentially regulate gene expression patterns to direct differen-
tiation, maturation, and function effectively. The helper T-cell
population of the immune system is an example of this epige-
netic programming (Ansel et al., 2003). During the differentia-
tion of CD4+ T-cells, there is an epigenetic activation of the
interferon gamma gene (IFNG) and a silencing of the interleukin
4 (IL4) gene. This results in a progressive polarization of T-cell
responsiveness as the epigenetic modifications are further modi-
fied by antigenic and cytokine actions via sequential divisions
within the lineage. Thus, different T-helper cells emerge and
maintain a polarized phenotype based upon epigenetic modifica-
tions that are retained following cell divisions, but additionally
polarized by antigenic and environmental cytokines, which add
to the epigenetic modulation. In mice, naïve T-cells have hyper-
methylated CpG islands within the Il4 locus. Extended demethy-
lation within this region results in Il4 gene activation and is
coupled to Ifng silencing, leading to a TH2 phenotype commit-
ment. Retention of Il4 methylation and Ifng activation results in
a TH1 phenotype. Thus, epigenetic memory of the T-cells and the
lineage modulates immune responses via TH2 (which up-regu-
lates IL4, IL5, and IL13) or TH1 (with enhanced IFN-γ and IL-2)
cytokine responses (Fields et al., 2002).
During development, imprinting of fetal tissues includes
the placenta, principally targeting trophoblast cells. Abnormali-
ties in placental imprinting have been recently implicated as a
cause of preeclampsia (Van Dijk et al., 2005) and fetal growth
restriction (McMinn et al., 2006). Imprinting patterns have
been associated with congenital disorders affecting growth and
neuro development that persist into adulthood, including Prader-
Willi and Angelman syndromes, which are two clinically
distinct diseases associated with abnormal imprinting on chro-
mosome 15q11-q13. Loss of maternal imprinting is responsible
for the Angelman syndrome, which is characterized by mental
404 Barros & Offenbacher J Dent Res 88(5) 2009
retardation, ataxia, and social disposition. In Prader-Willi syn-
drome, loss of paternal imprinting in the same region is charac-
terized by learning difficulties, hypogonadism, short stature, and
small hands and feet. Beckwith-Weidemann syndrome, another
imprinting disorder characterized by macrosomia, hemihyper-
trophy, abdominal wall defects, organomegaly, and susceptibil-
ity to Wilm’s tumor, is the result of loss of imprinting of
insulin-like growth factor 2 (IGF2) on chromosome 11p15
(Reik and Maher, 1997; Tycko and Morison, 2002). We recently
demonstrated, in a pregnant murine model, that infection with
Campylobacter rectus could induce an alteration in placental
Igf2 methylation patterns that resulted in reduced insulin-like
growth factor II mRNA expression with an associated fetal
growth restriction (Bobetsis et al., 2007). This suggests that
external stimuli or stressors like infection can modify imprinting
patterns in utero.
eNVIroNMeNtAL stressors As
Alterations in DNA methylation status as a result of environmen-
tal stressors have been documented to begin before birth. For
example, the methylation of fetal DNA that occurs in utero as a
result of low dietary levels of folate, methionine, or selenium can
change epigenetic programming that can persist into adulthood
(Post et al., 1999; Lund et al., 2004; Zaina et al., 2005). Although
many epigenetic marks are potentially reversible, the mecha-
nisms for reversal remain to be clearly elucidated, and many
epigenetic changes appear to persist throughout the cell lineage
and life of the organism. This provides perhaps an explanation
for the Barker hypothesis (Barker et al., 2002), which posits that
intra-uterine exposures can result in fetal programming that per-
sists into adulthood and may contribute to the risk for adult-onset
diseases such as cardiovascular disease and type 2 diabetes.
Intra-uterine nutrition can determine epigenetic program-
ming of the fetus. For example, methyltetrahydrofolate (folate)
is a critical methyl donor for S-Adenosyl Methionine (SAM),
which is used by the enzyme DNA methyltransferase (DNMT)
to methylate CpG residues selectively during embryonic devel-
opment (Razin and Shemer, 1995; Carlone and Skalnik, 2001;
Hershko et al., 2003). Maternal folate deficiency during preg-
nancy leads to inadequate levels of SAM, a critical substrate for
DNMT-dependent methylation (Okano et al., 1999). Folate defi-
ciencies can thereby result in DNA hypomethylation, which can
contribute to improperly elevated expression of certain genes, as
well as genetic instability facilitating abnormal chromosomal
re-arrangements (Zaina et al., 2005). These folate deficiencies
often result in abnormalities in placental development and func-
tion and alter patterns in fetal DNA methylation that can result
in growth defects and birth anomalies, including neural tube
defects (Blom et al., 2006). Thus, gross nutritional deficiencies
can lead to impairment of normal epigenetic programming,
resulting in abnormal ontological outcomes.
Such dietary or other exogenous environmental factors also
appear to modulate chromatin epigenetic marks throughout life.
For example, hyperhomocysteinemia (a marker for low levels of
methyl donors) is associated with global hypomethylation and
has been reported in atherosclerosis models and in humans,
which supports an emerging view that alterations in global
methylation patterns are characteristic of early stages of cardio-
vascular disease. In advanced stages of atherosclerosis, smooth-
cell proliferation or monocytic clonal expansion within the
atheroma may be associated with altered DNA methylation
patterns (Castro et al., 2006). The first findings linking
DNA methylation of CpG islands to cardiovascular disease
identified increased methylation of the CpG region of the estro-
gen receptor-α (ERα) gene seen in coronary atherectomy or
carotid endarterectomy samples (Post et al. 1999).
It has also been proposed that sensitivity to diet or to envi-
ronmental toxins may vary among individuals, due to pre-
existing genetic variants that can challenge methyl metabolism
and predispose individuals to epigenetic changes (Lund and
Zaina, 2007). Other environmental stimuli that may potentially
function as epigenetic modifiers are exposures to metals and
aromatic hydrocarbons (e.g., benzopyrene), found in occupa-
tional chemicals, fossil fuel emissions, contaminated drinking
water, cigarette smoke, and infection (Risch and Plass, 2008). A
recent study on the influence of smoking on global DNA methy-
lation indicated that smoking induces generalized alterations in
DNA methylation across multiple tissues and organ systems,
also showing an association of the offspring’s DNA methylation
with paternal DNA methylation that was strongest if both had
never smoked (p = 0.02) (Hillemacher et al., 2008). The authors
also found that the association completely vanished if descen-
dants smoked themselves or had ever smoked, suggesting an
association between smoking behavior and global DNA methy-
lation, which may be of importance for a wide range of diseases.
Smoking may also be linked to oncogenesis by inducing specific
epigenetic modifications (Tessema et al., 2008). For example,
promoter methylation of several tumor suppressor genes has
frequently been reported in a high percentage (20-100%) of
human lung cancers (Zochbauer-Muller et al., 2002). Methylation
of the tumor suppressor p16 gene has been suggested to play a
critical role in lung cancer survival rate pathogenesis (Kim
et al., 2001; von Zeidler et al., 2004; Dammann et al., 2005).
ePIGeNetIcs, cANcer, ANd INFLAMMAtIoN
The best-studied epigenetic alteration in cancer is DNA methyla-
tion. During tumorigenesis, methylation is usually decreased
genome-wide, with selective hypermethylation of CpG sites within
promoters of tumor-suppressor genes, leading to their silencing
and subsequent tumor progression (Breivik and Gaudernack,
1999). This suggests that oncogenesis may also occur through
epigenetic dysregulation. Feinberg (2007) has recently reviewed
the epigenetic mechanisms involved with oncogene activation or
tumor suppressor gene silencing in cancer initiation and progres-
sion, discussing the new idea that epigenetic modifications may
play a role in cancer predisposition, and that such changes should
be considered as targets for preventive oncology.
The role of host inflammation on modification of epigenetic
patterns is still unknown, but the activation of the immune
response involving potential epigenetic changes has been
suggested (Adcock and Lee, 2006). Inflammatory signals that
J Dent Res 88(5) 2009 Epigenetics 405
activate NF-κ B have been shown to alter histone methylation
patterns and activate gene expression (Ito, 2007). Thus, inflam-
mation has some potential to modify chromatin structure via
histone structure; however, the role of inflammation in modulat-
ing CpG methylation patterns, which are more likely to be
conserved following cell replication, remains unclear. Recent
reports suggest that loss of epigenetic control over this complex
process contributes to autoimmune disease (Yung and Julius,
2008). Logically, this may involve abnormal function and matu-
ration in T-cell lineages as a consequence of aberrant epigenetic
Epigenetic mechanisms may also explain, in part, the linkage
between inflammation and oncogenesis, and the relationship
between CpG island methylation phenotype in tumors and
inflammation has been discussed (Shaw et al., 2007). For
example, gastric inflammation due to bacterial infection with H.
pylori has been linked to alteration in DNA methylation patterns
of tumor suppressor genes (Tsuji et al., 2006; Ushijima, 2007).
H. pylori is an etiologic gastric carcinogen, with about 80% of
gastric cancers being H. pylori-related (Forman et al., 1991).
However, the cancer risks are different among H. pylori-infected
individuals, which probably reflects the diversity of H. pylori
strains, and differences in host susceptibility or other environ-
mental factors (Uemura et al., 2001). There exists, however, a
close anatomical relationship in cancerous lesions, whereby H.
pylori is in direct contact with gastric cells that display altered
MVP methylation patterns. One specific MVP target in gastric
cancer appears to be cyclo-oxygenase 2 (COX-2). Low levels of
gastric secretion of prostaglandin E2 (a product of cyclo-oxyge-
nase 2) and COX-2 suppression have long been known to be
associated with gastric cancer. Recent studies (Huang et al.,
2006; Perri et al., 2007) have demonstrated that cancerous gas-
tric cells exhibit abnormal PTGS2 (Prostaglandin G/H
Synthase-2) promoter hyper methylation patterns. Non-cancerous
regions of the gastric lesion do not demonstrate the presence of
H. pylori or the abnormal PTGS2 promoter hypermethylation
and COX-2 suppression. Thus, the evidence supports the con-
cept that the H. pylori infection promotes oncogenesis by epige-
netic modification that includes the PTGS2 promoter.
In 2007, we were the first group to report alteration in DNA
methylation status of the Igf2 gene in murine placental tissues
due to maternal infection with the periodontopathogenic bacte-
ria Campylobacter rectus (Bobetsis et al., 2007). By analogy, H.
pylori and C. rectus are close phylogenetic neighbors that share,
for example, GroEL protein (HSP60 family) expression, which
can stimulate IL-6 production (Tanabe et al., 2003). We were
able to demonstrate, in placentas from growth-restricted fetuses,
that the hypermethylation found in the promoter region (P0) of
the Igf2 gene was related to the C. rectus placental exposure in
pregnant mice. This appears causal, since Igf2 gene function
involves growth and development. Population studies have
shown that, among humans, prematurity and impaired fetal
growth have been associated with adult-onset diabetes and car-
diovascular disease (Barker et al., 2002). If abnormalities in
IGF2 methylation were to occur in utero among humans, it is
possible that these epigenetic marks could persist into adulthood
and may be associated with the observed abnormalities in IGF2
metabolism in adults with cardiovascular disease and diabetes.
Thus, the role of infection as an intra-uterine modifier of epige-
netic programming should be considered as a possible link to
adult disease. This may be particularly relevant in humans, since
C. rectus exposure of the fetus has been found to be associated
with preterm delivery (Madianos et al., 2001).
Following the observation and establishment of a connection
between oral bacteria altering placental DNA methylation, we
asked the question, Could the oral biofilm epigenetically modify
the local adjacent periodontal tissues? We have conducted a pilot
survey using CpG Island Microarray analysis (data not shown)
comparing genomic-wide MVP methylation status of periodon-
tally diseased gingival tissues with healthy gingival tissue. We
could preliminarily identify a list of genes that were differentially
methylated in gingival tissues from individuals with periodontal
disease, and these results will be forthcoming shortly. Thus, the
role of bacterial infection and chronic inflammation as a potential
stimulus for altering local periodontal tissue DNA methylation
patterns provides a fertile area for further investigation. Furthermore,
the link between inflammation and oral cancer is well-established,
and a connection between bacterial infection and inflammation is
evident. Thus, epigenetic influences may serve as a plausible
potential mechanism that connects all three pathways and should
be further explored, especially as it relates to mucosal cancers,
which emerge in the presence of high microbial burdens.
ePIGeNetIc MArKs IN cArcINoGeNesIs
Epigenetic changes set the stage for alterations in gene expres-
sion and have been identified as important components of
carcinogenesis. As previously mentioned, global DNA hypo-
methylation is a general feature of genomic DNA derived from
solid and hematologic tumors as isolated from animal models
and human tumors (Gaudet et al., 2003; Fraga et al., 2005; Holm
et al., 2005). Hypomethylation is consistent with the overall
increased transcriptional activity seen in most tumors. However,
loss of DNA methylation, which often occurs at sequences which
are unstable, is likely related to increased tumor frequency due to
chromosomal instability, and it has been considered as the earli-
est epigenetic change from a normal to a pre-malignant cell.
However, the expression of certain oncogenes appears to be
directly activated by hypomethylation, whereas hypermethyla-
tion of tumor suppressor genes can also be seen. Examples of
promoter DNA hypermethylation and chromatin hypoacetyla-
tion, which result in the silencing of tumor suppressor genes,
include p16 (also known as cyclin-dependent kinase inhibitor
2A) and MutL protein homologue 1 (MLH1) (Herman and
Baylin, 2003; Feinberg and Tycko, 2004).
In a recent review, Choi and Myers (2008) emphasized the
genetic and epigenetic alterations in the molecular pathogenesis
of oral squamous cell carcinoma (OSCC), discussing the role of
oncogenes like Ras oncogene, Cyclin D1, AP-1 complex, and
tumor suppressor genes like p53, p16, and p21. These investiga-
tors reported aberrant hypermethylation patterns in the promoter
region of p16 and E-cadherin which influence cell division and
cell-cell adhesion, respectively. p16 was one of the first genes to
be found associated with aberrant DNA methylation patterns in
406 Barros & Offenbacher J Dent Res 88(5) 2009
head and neck cancer (Reed et al., 1996). p16 inhibits G1 to S
phase passage by binding cyclin-dependent kinase, preventing
formation of its complex with cyclin D. Methylation of p16
promoter has been considered as a predictive marker for malig-
nant transformation, since the methylation depicts uncontrolled
cell division (Hall et al., 2008). In OSCC, p16 methylation has
been reported to vary between 31% (Maruya et al., 2004) and
67% (Kulkarni and Saranath, 2004). E-cadherin plays a role in
cell-cell adhesion, and, when underexpressed, may affect tumor
invasion by leading to a greater probability of tumor invasion or
metastasis. E-cadherin was found silenced by hypermethylation
in other studies of oral cancer (Hasegawa et al., 2002; Kudo
et al., 2004; Maruya et al., 2004).
Other genes have also been investigated for aberrant DNA
methylation in oral squamous cell carcinomas. The epigenetic
silencing of the MGMT (O6-methylguanine–DNA methyltrans-
ferase) DNA-repair gene which, by promoter methylation, com-
promises DNA repair, has been considered an early event in the
development of oral cancer and is associated with 25-52% of
primary oral squamous-cell carcinomas (Viswanathan et al.,
2003; Kulkarni and Saranath, 2004; Maruya et al., 2004). The
death-associated protein kinase 1(DAPK1) gene, a tumor-
suppressor gene involved in apoptosis, is also found methylated
between 7% and 68% in oral cancers (Li, 2002; Ogi et al., 2002;
Maruya et al., 2004).
It has also been shown that miRNAs, which modulate post-
transcriptional gene expression, can be aberrantly expressed or
mutated in cancers, suggesting that they may also function as
oncogenes or tumor suppressor genes. More recently, studies have
shown that miRNA genes may be regulated also by epigenetic
mechanisms and have even been identified embedded within CpG
islands (Lujambio and Esteller, 2007). The promoter of an
miRNA methylated in normal tissues can be maintained in cancer,
as in the case of two putative tumor suppressor miRNAs—
miR-127 and miR-124a—which are transcriptionally inactivated
by CpG island hypermethylation (Saito et al., 2006; Lujambio
et al., 2007), whereas in lung cancer, the overexpression of let-
7a-3 seems to be due to DNA hypomethylation (Brueckner et al.,
2007; Weber et al., 2007). In both cases, the functional signifi-
cance is the opposite. miR-127 and miR-124a seem to act as
tumor suppressors, and so methylation is maintained in cancer,
while let-7a-3 is thought to act as an oncogene, so its demethyla-
tion would contribute to the tumoral phenotype. Clearly, the role
of epigenetic influence on miRNA function is a new area of inves-
tigation and represents just one area in which the role of epigenet-
ics and oral disease needs further exploration.
Future research will help us understand, for example, how
systemic exposures, like smoking, may alter global epigenetic
patterns to affect the expression of oral conditions such as oral
cancer or advanced periodontitis. We need to understand how
the oral microbiome and local biofilm may create an epigenetic
“footprint” in the adjacent mucosa and periodontal tissues, and
potentially modify the local inflammatory response and onco-
genic potential. Biofilm-induced epigenetic patterns may influ-
ence local tissue metabolism to alter the microbial ecology and
alter local healing responses of the periodontal tissues. Recently,
Park and colleagues (2008) were able to reprogram somatic
human cells to a pluripotent state, which is in essence a reversal
of differentiation to a more embryonic state, by inducing the
ectopic expression of four transcriptional regulatory factors
(Oct4, Sox2, Klf4, and Myc). This resulted in significant epige-
netic remodeling, which was sufficient to result in cellular
reprogramming to a pluripotent state. This suggests that epige-
netic reprogramming might prove to be a mechanism to create
new wound-healing or tissue-regenerative potential, and agents
which modify epigenetic patterns are a fertile area for new drug
development strategies. These are just a few important questions
and opportunities that will await further studies that explore the
role of epigenetics in oral biology. Thus, epigenetic codes,
which are just becoming revealed, can help us better understand
the biological phenotype that arises from the interaction of the
human genome with the environment in health and in disease.
This work is supported by grant DE-01243 from NIDCR and
grant RR-00046 from NCRR to SO.
Adcock IM, Lee KY (2006). Abnormal histone acetylase and deacetylase
expression and function in lung inflammation. Inflamm Res 55:311-
321, erratum in Inflamm Res 55:572, 2006.
Ansel KM, Lee DU, Rao A (2003). An epigenetic view of helper T cell dif-
ferentiation. Nat Immunol 4:616-623.
Barker DJ, Eriksson JG, Forsen T, Osmond C (2002). Fetal origins of adult
disease: strength of effects and biological basis. Int J Epidemiol
Bartolomei MS, Zemel S, Tilghman MS (1991). Parental imprinting of the
mouse H19 gene. Nature 351:153-155.
Bird A (2002). DNA methylation patterns and epigenetic memory. Genes
Bird AP, Wolffe AP (1999). Methylation-induced repression: belts, braces,
and chromatin. Cell 99:451-454.
Blom HJ, Shaw GM, den Heijer M, Finnell RH (2006). Neural tube defects
and folate: case far from closed. Nat Rev Neurosci 9:724-731.
Bobetsis YA, Barros SP, Lin DM, Weidman JR, Dolinoy D, Jirtle RL, et al.
(2007). Bacterial infection promotes DNA hypermethylation. J Dent
Breivik J, Gaudernack G (1999). Genomic instability, DNA methylation,
and natural selection in colorectal carcinogenesis. Semin Cancer Biol
Brueckner B, Stresemann C, Kuner R, Mund C, Musch T, Meister M,
et al. (2007). The human let-7a-3 locus contains an epigenetically
regulated microRNA gene with oncogenic function. Cancer Res 67:
Carlone DL, Skalnik DG (2001). CpG binding protein is crucial for early
embryonic development. Mol Cell Biol 22:7601-7606.
Castro R, Rivera I, Blom HJ, Jakobs C, Tavares de Almeida I (2006).
Homocysteine metabolism, hyperhomocysteinaemia and vascular dis-
ease: an overview. J Inherit Metab Dis 29:3-20.
Choi S, Myers JN (2008). Molecular pathogenesis of oral squamous cell
carcinoma: implications for therapy. J Dent Res 87:14-32.
Constancia M, Pickard B, Kelsey G, Reik W (1998). Imprinting mecha-
nisms. Genome Res 8: 881-900.
Dammann R, Strunnikova M, Schagdarsurengin U, Rastetter M, Papritz M,
Hattenhorst UE, et al. (2005). CpG island methylation and expression
of tumour-associated genes in lung carcinoma. Eur J Cancer 41:
DeChiara M, Robertson EJ, Efstratiadis A (1991). Parental imprinting of the
mouse insulin-like growth factor II gene. Cell 64:849–859.
J Dent Res 88(5) 2009 Epigenetics 407
Doerfler W (1991). Patterns of DNA methylation—evolutionary vestiges of
foreign DNA inactivation as a host defense mechanism. A proposal.
Biol Chem Hoppe Seyler 372:557-564.
Dolinoy DC, Huang D, Jirtle RL (2007). Maternal nutrient supplementation
counteracts bisphenol A-induced DNA hypomethylation in early devel-
opment. Proc Natl Acad Sci USA 104:13056-13061.
Fazzari MJ, Greally JM (2004). Epigenomics: beyond CpG islands. Nat Rev
Feinberg AP (2007). Phenotypic plasticity and the epigenetics of human
disease. Nature 447:433-440.
Feinberg AP (2008). Epigenetics at the epicenter of modern medicine. J Am
Med Assoc 299:1345-1350.
Feinberg AP, Tycko B (2004). The history of cancer epigenetics. Nat Rev
Fields PE, Kim ST, Flavell RA (2002). Cutting edge: changes in histone
acetylation at the IL-4 and IFN-gamma loci accompany Th1/Th2 dif-
ferentiation. J Immunol 169:647-650.
Forman D, Newell DG, Fullerton F, Yarnell JW, Stacey AR, Wald N, et al.
(1991). Association between infection with Helicobacter pylori and
risk of gastric cancer: evidence from prospective investigation. BMJ
Fraga MF, Ballestar E, Villar-Garea A, Boix-Chornet M, Espada J, Schotta G,
et al. (2005). Loss of acetylation at Lys16 and trimethylation at Lys20 of
histone H4 is a common hallmark of human cancer. Nat Genet 37:391-400.
Gaudet, F, Hodgson JG, Eden A, Jackson-Grusby L, Dausman J, Gray JW,
et al. (2003). Induction of tumors in mice by genomic hypomethylation.
Grant-Downton RT, Dickinson HG (2005). Epigenetics and its implications
for plant biology. The epigenetic network in plants. Ann Bot (Lond)
Hall GL, Shaw RJ, Field EA, Rogers SN, Sutton DN, Woolgar JA, et al.
(2008). p16 promoter methylation is a potential predictor of malignant
transformation in oral epithelial dysplasia. Cancer Epidemiol
Biomarkers Prev 17:2174-2179.
Hanson MA, Gluckman PD (2008). Developmental origins of health and
disease: new insights. Basic Clin Pharmacol Toxicol 102:90-93.
Hark AT, Schoenherr CJ, Katz DJ, Ingram RS, Levorse JM, Tilghman SM
(2000). CTCF mediates methylation-sensitive enhancer-blocking activ-
ity at the H19/Igf2 locus. Nature 405:486-489.
Hasegawa M, Nelson HH, Peters E, Ringstrom E, Posner M, Kelsey KT
(2002). Patterns of gene promoter methylation in squamous cell cancer
of the head and neck. Oncogene 21:4231-4236.
Herman JG, Baylin SB (2003). Gene silencing in cancer in association with
promoter hypermethylation. N Engl J Med 349:2042-2054.
Hershko AY, Kafri T, Fainsod A, Razin A (2003). Methylation of HoxA5 and
HoxB5 and its relevance to expression during mouse development.
Hillemacher T, Frieling H, Moskau S, Muschler MA, Semmler A, Kornhuber
J, et al. (2008). Global DNA methylation is influenced by smoking
behaviour. Eur Neuropsychopharmacol 18:295-298.
Holliday R, Pugh JE (1975). DNA modification mechanisms and gene activ-
ity during development. Science 187:226-232.
Holm TM, Jackson-Grusby L, Brambrink T, Yamada Y, Rideout WM 3rd,
Jaenisch R (2005). Global loss of imprinting leads to widespread tumor-
igenesis in adult mice. Cancer Cell 8:275-285; erratum in Cancer Cell
8:433, 2005, and Cancer Cell 9:69, 2006.
Huang L, Zhang KL, Li H, Chen XY, Kong QY, Sun Y, et al. (2006).
Infrequent COX-2 expression due to promoter hypermethylation in
gastric cancers in Dalian, China. Hum Pathol 37:1557-1567.
Ito K (2007). Impact of post-translational modifications of proteins on the
inflammatory process. Biochem Soc Trans 35:281-283.
Kennedy D (2007). Breakthrough of the year. Science 318:1833.
Kim DH, Nelson HH, Wiencke JK, Zheng S, Christiani DC, Wain JC, et al.
(2001). p16 (INK4a) and histology-specific methylation of CpG islands
by exposure to tobacco smoke in non-small cell lung cancer. Cancer
Kudo Y, Kitajima S, Ogawa I, Hiraoka M, Sargolzaei S, Keikhaee MR, et al.
(2004). Invasion and metastasis of oral cancer cells require methylation
of E-cadherin and/or degradation of membranous beta-catenin. Clin
Cancer Res 10:5455-5463.
Kulkarni V, Saranath D (2004). Concurrent hypermethylation of multiple
regulatory genes in chewing tobacco associated oral squamous cell
carcinomas and adjacent normal tissues. Oral Oncol 40:145-153.
Larsen F, Gundersen G, Lopez R, Prydz H (1992). CpG islands as gene
markers in the human genome. Genomics 13:1095-1107.
Li E (2002). Chromatin modification and epigenetic reprogramming in
mammalian development. Nature Rev Genet 3:662-673.
Loenen WA (2006). S-adenosylmethionine: jack of all trades and master of
everything? Biochem Soc Trans 34(Pt 2):330-333.
Lujambio A, Esteller M (2007). CpG island hypermethylation of tumor sup-
pressor microRNAs in human cancer. Cell Cycle 6:1455-1459.
Lujambio A, Ropero S, Ballestar E, Fraga MF, Cerrato C, Setién F, et al.
(2007). Genetic unmasking of an epigenetically silenced microRNA in
human cancer cells. Cancer Res 67:1424-1429.
Lund G, Zaina S (2007). Atherosclerosis, lipids, inflammation and epigenet-
ics. Curr Opin Lipidol 18:699-701.
Lund G, Andersson L, Lauria M, Lindholm M, Fraga MF, Villar-Garea A,
et al. (2004). DNA methylation polymorphisms precede any histologi-
cal sign of atherosclerosis in mice lacking apolipoprotein E. J Biol
Madianos PN, Lieff S, Murtha AP, Boggess KA, Auten RL Jr, Beck JD,
et al. (2001). Maternal periodontitis and prematurity. Part II: Maternal
infection and fetal exposure. Ann Periodontol 6:175-182.
Maruya S, Issa JP, Weber RS, Rosenthal DI, Haviland JC, Lotan R, et al.
(2004). Differential methylation status of tumor-associated genes in
head and neck squamous carcinoma: incidence and potential implica-
tions. Clin Cancer Res 10:3825-3830.
McMinn J, Wei M, Schupf N, Cusmai J, Johnson EB, Smith AC, et al.
(2006). Unbalanced placental expression of imprinted genes in human
intrauterine growth restriction. Placenta 27:540-549.
Miranda TB, Jones PA (2007). DNA methylation: the nuts and bolts of
repression. J Cell Physiol 213:384-390.
Miska EA (2005). How microRNAs control cell division, differentiation,
and death. Curr Opin Genet Dev 5:563-568.
Ogi K, Toyota M, Ohe-Toyota M, Tanaka N, Noguchi M, Sonoda T, et al.
(2002). Aberrant methylation of multiple genes and clinicopathological
features in oral squamous cell carcinoma. Clin Cancer Res 8:
Okano M, Bell DW, Haber DA, Li E (1999). DNA methyltransferases
DNMT3a and DNMT3b are essential for de novo methylation and
mammalian development. Cell 99:247-257.
Park IH, Zhao R, West JA, Yabuuchi A, Huo H, Ince TA, et al. (2008).
Reprogramming of human somatic cells to pluripotency with defined
factors. Nature 451:141-146.
Pennisi E (2001). Behind the scenes of gene expression. Science 293:
Perri F, Cotugno R, Piepoli A, Merla A, Quitadamo M, Gentile A, et al.
(2007). Aberrant DNA methylation in non-neoplastic gastric mucosa of
H. pylori infected patients and effect of eradication. Am J Gastroenterol
Post WS, Goldschmidt-Clermont PJ, Wilhide CC, Heldman AW, Sussman
MS, Ouyang P, et al. (1999). Methylation of the estrogen receptor gene
is associated with aging and atherosclerosis in the cardiovascular sys-
tem. Cardiovasc Res 43:985-991.
Razin A, Shemer R (1995). DNA methylation in early development. Hum
Mol Genet 4:1751-1755.
Reed AL, Califano J, Cairns P, Westra WH, Jones RM, Koch W, et al.
(1996). High frequency of p16 (CDKN2/MTS-1/INK4A) inactivation
in head and neck squamous cell carcinoma. Cancer Res 56:3630-
Reik W (2007). Stability and flexibility of epigenetic gene regulation in
mammalian development. Nature 447:425-432.
Reik W, Maher ER (1997). Imprinting in clusters: lessons from Beckwith-
Wiedemann syndrome. Trends Genet 13:330-334.
Reik W, Walter J (1998). Imprinting mechanisms in mammals. Curr Opin
Genet Dev 8:154-164.
408 Barros & Offenbacher J Dent Res 88(5) 2009 Download full-text
Riggs AD (1975). X inactivation, differentiation, and DNA methylation.
Cytogenet Cell Genet 14:9-25.
Risch A, Plass C (2008). Lung cancer epigenetics and genetics. Int J Cancer
Saito Y, Liang G, Egger G, Friedman JM, Chuang JC, Coetzee GA, et al.
(2006). Specific activation of microRNA-127 with downregulation of
the proto-oncogene BCL6 by chromatin-modifying drugs in human
cancer cells. Cancer Cell 9:435-443.
Shaw RJ, Hall GL, Lowe D, Bowers NL, Liloglou T, Field JK, et al. (2007).
CpG island methylation phenotype (CIMP) in oral cancer: associated
with a marked inflammatory response and less aggressive tumour biol-
ogy. Oral Oncol 43:878-886
Tanabe S, Hinode D, Yokoyama M, Fukui M, Nakamura R, Yoshioka M,
et al. (2003). Helicobacter pylori and Campylobacter rectus share a
common antigen. Oral Microbiol Immunol 18:79-87.
Tessema M, Willink R, Do K, Yu YY, Yu W, Machida EO, et al. (2008).
Promoter methylation of genes in and around the candidate lung cancer
susceptibility locus 6q23-25. Cancer Res 68:1707-1714.
Tsuji S, Tsujii M, Murata H, Nishida T, Komori M, Yasumaru M, et al.
(2006). Helicobacter pylori eradication to prevent gastric cancer:
underlying molecular and cellular mechanisms. World J Gastroenterol
Tycko B, Morison IM (2002). Physiological functions of imprinted genes
(review). J Cell Physiol 192:245-258.
Uemura N, Okamoto S, Yamamoto S, Matsumura N, Yamaguchi S, Yamakido
M, et al. (2001). Helicobacter pylori infection and the development of
gastric cancer. N Engl J Med 345:784-789.
Ushijima T (2007). Epigenetic field for cancerization. J Biochem Mol Biol
Van de Vijver G, Van Speybroeck L, De Waele D (2002). Epigenetics: a
challenge for genetics, evolution, and development? Ann NY Acad Sci
Van Dijk M, Mulders J, Poutsma A, Könst AA, Lachmeijer AM, Dekker
GA, et al. (2005). Maternal segregation of the Dutch preeclampsia
locus at 10q22 with a new member of the winged helix gene family. Nat
Viswanathan M, Tsuchida N, Shanmugam G (2003). Promoter hypermethy-
lation profile of tumor-associated genes p16, p15, hMLH1, MGMT and
E-cadherin in oral squamous cell carcinoma. Int J Cancer 105:41-46.
von Zeidler SV, Miracca EC, Nagai MA, Birman EG (2004). Hypermethylation
of the p16 gene in normal oral mucosa of smokers. Int J Mol Med
Vucic EA, Brown CJ, Lam WL (2008). Epigenetics of cancer progression.
Weber B, Streseman C, Brueckner B, Lyko F (2007). Methylation of human
microRNA genes in normal and neoplastic cells. Cell Cycle 6:1001-1005.
Yoder JA, Walsh CP, Bestor TH (1997). Cytosine methylation and the ecol-
ogy of intragenomic parasites. Trends Genet 13:335-340.
Yung RL, Julius A (2008). Epigenetics, aging, and autoimmunity.
Zaina S, Lindholm MW, Lund G (2005). Nutrition and aberrant DNA
methylation patterns in atherosclerosis: more than just hyperhomo-
cysteinemia? J Nutr 135:5-8.
Zamore PD, Haley B (2005). Ribo-gnome: the big world of small RNAs.
Zochbauer-Muller S, Minna JD, Gazdar AF (2002). Aberrant DNA methyla-
tion in lung cancer: biological and clinical implications. Oncologist