epigenetics and its role in periodontal diseases: a state

Review

Epigenetics and Its Role in Periodontal Diseases:A State-of-the-Art Review

Lena Larsson,* Rogerio M. Castilho,† and William V. Giannobile‡§

The immune response to oral bacteria and the subsequentactivation of inflammatory signaling is not only dependent ongenetic factors. The importance of so-called epigeneticmechanisms presents additional regulatory pathways ofgenes involved in maintaining chronic inflammation, includ-ing gingivitis and periodontitis. The term epigenetics relatesto changes in gene expression that are not encoded in theDNA sequence itself and include chemical alterations ofDNA and its associated proteins. These changes lead toremodeling of the chromatin and subsequent activation or in-activation of a gene. Epigenetic mechanisms have been foundto contribute to disease, including cancer and autoimmune orinflammatory diseases. In this state-of-the art review, the au-thors provide the latest findings on the involvement of epigeneticmodifications in the development of periodontal disease andpresent emerging therapeutic strategies aimed at epigenetictargets (epidrugs) associated with the disruption of tissue ho-meostasis and the development of periodontitis. J Periodon-tol 2015;86:556-568.

KEY WORDS

Acetylation; DNA methylation; epigenomics; genetics;histones; periodontitis.

Periodontitis is an infectious dis-ease characterized by chronic in-flammation in the gingival tissues

in response to bacteria colonizing thetooth surface, ultimately leading to tissuebreakdown with loss of alveolar boneand connective tissue.1 The epithelial cellsin the oral cavity are constantly being ex-posed to high levels of various bacteria,which leads to activation of multiplesignaling pathways that induce variousinflammatory response mechanisms.1

The host inflammatory response to bac-teria is in turn influenced by both geneticand environmental factors.1,2

REGULATION OF IMMUNERESPONSE

The central dogma of gene expression isthat DNA is transcribed into messengerRNA (mRNA), which serves as a tem-plate for the protein synthesis.3 Thebasic model for the transcriptional in-duction in the immune response is thatexternal stimuli, such as bacterial prod-ucts, start a signaling cascade within thecells, activating the binding of transcrip-tion factors to regulatory sequences ofthe DNA, so-called promoters and en-hancers. This assembles the transcriptioninitiation machinery and induces the ex-pression of certain gene sets.4 At thislevel, genetic changes such asmutations,polymorphisms, or alterations in the DNAsequence alter the binding sites fortranscription factors in the DNA regula-tory region (promoter), thereby affecting

* Currently, Department of Periodontology, Institute of Odontology, University ofGothenburg, Sweden; previously, Department of Periodontics and Oral Medicine, Schoolof Dentistry, University of Michigan, Ann Arbor, MI.

† Laboratory of Epithelial Biology, Department of Periodontics andOral Medicine, School ofDentistry, University of Michigan.

‡ Department of Biomedical Engineering, College of Engineering, University of Michigan.§ Department of Periodontics andOralMedicine,School ofDentistry,University ofMichigan.

doi: 10.1902/jop.2014.140559

Volume 86 • Number 4

556

not only the initiation speed but also the magnitude ofthe gene expression. In addition, the form of theprotein, should these changes occur within the codingregions of the gene (exons), results in changes in theamino acid sequence of the protein.5

A recent study has revealed additional regulatorylayers superimposed on this basic model.4 Specificfunctional programs able to be rapidly turned on andoff in response to external stimuli have been identi-fied.4,6 These processes have been proposed asa mechanism in creating a selectivity of the hostresponse to stimuli, and also to regulate the mag-nitude of the inflammatory response to clear the

pathogens or drive chronic inflammation.7 In addi-tion, this presents a potential mechanism of howdifferent bacterial species lead to activation of sig-naling pathways, inducing inflammatory responsemechanisms.8

The layers of gene expression regulation of animmune response include structures of the chromatin,transcription factors, and posttranscriptional modifi-cations, e.g., RNA splicing, mRNA polyadenylation,and regulatory molecules, which include microRNA(miRNA) and the recently identified long non-codingRNAs (lncRNAs)9 (Fig. 1). The first two mechanismsare: 1) modifications of the mRNA transcript and

2) regulation of the stability,storage, and amount of mRNAtranscripts present in the cyto-plasm. When DNA is transcribedinto mRNA, a process calledsplicing removes the non-codingsequences of mRNA (introns).Bacterial products can alter thesplicing pattern, resulting in dif-ferent isoforms of the proteinproduced. After the transcriptionand splicing, the mRNA tran-script acquires a sequence ofadenosine bases at the terminalend of the transcript called a poly(A)tail. The length of this poly(A)tail determines and contributes toboth mRNA stability and trans-port to the cytoplasm. It also actsas a mechanism for the cell tostore silenced mRNA within thecytoplasm, which in response tobacterial stimuli can be used torapidly induce gene expressionwithout having to start the DNAtranscription process in the nu-cleus. This was also suggestedas a way to limit immune re-sponse to avoid inducing chronicinflammation.10

Not all RNA is protein coding.miRNAs are small, evolution-arily conserved ncRNAs that arederived from longer transcriptsencoded within the genome.They inhibit gene expression bybinding tomRNA by a sequence-pairing homology to regulateand fine-tune gene expression.miRNA expression is altered inresponse to bacterial products,but inflammatory cytokines canalso influence the regulation of

Figure 1.Epigenetic studies have identified multiple layers of regulatory factors and mechanisms that add tothe basic model for transcriptional induction in the immune response. Bacterial products and theinflammatory process influence the transcriptional cascade at all levels.

J Periodontol • April 2015 Larsson, Castilho, Giannobile

557

miRNA production.7 Recently, another class ofncRNAs was discovered, the lncRNAs.7 Differentbiologic functions have been attributed to lncRNAs.A potential mechanism for their ability to regulategene expression is that they bind to chromatinregulatory proteins, thereby controlling access orinhibition of proteins binding to enhancer regions inthe DNA.3,11,12 One function of these RNAs in theimmune response has been suggested to involve theregulation of the host response, including innateimmunity, since virus infection alters the expressionof lncRNAs.13 Interestingly, a recent report identifieda new ncRNA that is linked to epigenetic mecha-nisms.14

Chronic inflammatory diseases, such as peri-odontitis, have specific target tissues in which theinflammation results in tissue breakdown.1 A specificgene may possess different epigenetic patterns de-

pending on the cell type, resulting in local and sys-temic expression of the gene. This indicates a local,site-specific change in the immune response to ex-ternal stimuli and/or the inflammatory response it-self, which may differ among individuals and add tothe susceptibility to disease. This review focuses onthe epigenetic alterations of the chromatin and DNA,how they can contribute to periodontal disease, andthe recent literature on their potential use as newtreatment targets.

EPIGENETICS

The term epigenetics relates to changes in geneexpression that are not encoded in the DNA se-quence.15 Epigenetic modifications include chemicalalterations of DNA and associated proteins, leadingto remodeling of the chromatin and activation orinactivation of a gene. These changes can contribute

Figure 2.Schematic illustration of the influence of histone acetylation on gene transcription. Met = methylated; DNMTs = DNA methyltransferases; 5mC =5-methylcytosine, K9HMT = Lysine 9 histone methyltransferase; AC = acetylation.

Epigenetics and Periodontal Diseases Volume 86 • Number 4

558

to the development and maintenance of cancer andautoimmune or inflammatory diseases, including peri-odontitis. Interestingly, some epigenetic modificationsare reversible and can be induced and/or altered byenvironmental factors, therefore presenting a link be-tween the inherited genome and the environment.16,17

The two major epigenetic modifications are DNAmethylation and histone acetylation andmethylation.17

The building blocks of chromatin are the nucleo-somes.17 One nucleosome consists of 146 base pair(bp) DNA and a core histone complex, which includestwo copies each of histones H2A, H2B, H3, and H4and the linker histone H1 that connects the nucleo-somes, forming the primary chromatin structure, oftenreferred to as ‘‘beads-on-a-string.’’ Histones can beacetylated or methylated at histone amino acid tailsthat protrude from the nucleosome.18,19 Histoneacetylation is regulated by histone acetyltransferases(HATs) and histone deacetylases (HDACs).20 Re-moval of acetyl groups by HDACs leads to alterationsin the packing of DNA around histones and a sub-sequent inactivation of genes (Fig. 2). In contrast,hyperacetylation is associated with transcription-ally active chromatin.20,21 Methylation occurs atdifferent lysine molecules on histones H3 and H4,with H3 lysines K4 and K9 being most commonlymethylated.19 Although high levels of methylated K4have been associated with transcriptionally activeregions, methylated K9 was found in silent chro-

matin regions.19 Importantly, a lysine can be mono-,di-, or tri-methylated, further adding to the variousfunctions of methylated histones.19 However, recentresearch in the area of histone methylation revealsevidence for an even more complicated regulationand function of histone methylation, with specificmethylation patterns in promoters, introns, andexons resulting in a fine-tuning of gene expression.22

In the classic DNAmethylation model, the additionof methyl groups to cytosine bases situated next toa guanine base occurs at specific sites in the DNAsequence, so-called CpG islands or CpG sites by thede novo DNA methyltransferases (DNMTs) DNMT3aand DNMT3b.15,23 This results in both methylated aswell as unmethylated CpG sites in the DNA strands.The maintenance enzyme DNMT1 then remethylateshemimethylated CpG sites.24 This modification altersthe configuration of the DNA and the binding oftranscription factors, resulting in changes in geneexpression. Transcriptionally active genes are as-sociated with low levels of DNA methylation. How-ever, in 2014, Jeltsch and Jurkowska24 presenteda modified version of this classic model based onrecent research findings. It was suggested that theaverage DNA methylation (global DNA methylation)level of DNA is inherited rather than the methylationstate of individual CpG sites (site-specific DNAmethylation). To add another level to the conceptof DNA methylation, it was discovered that 5-methylcytosine (5mC) could be further oxidized into5-hydroxymethylcytosine (5hmC) by the ten-eleventranslocation (TET) family of enzymes.25 The TETenzymes can then convert 5hmC into unmethylatedcytosine, resulting in DNA hypomethylation.26 Thebiologic function of 5hmC is not clear, but it has beensuggested to be an intermediate leading to deme-thylation of 5mC and thereby re-expression of genessilenced by DNA methylation (Fig. 3).27

Histonemodifications and DNAmethylation are notseparate events but are linked, and the resultingcombinations promote unique cellular responses.21,23

In addition, hypo-acetylated histones and hyper-methylated CpGs in chronic inflammatory diseasesmay result in persistent inflammation.20 In a recentstudy, treatment of mice with a compound capableof inducing a immune response (antigen stimula-tion) resulted in a decrease of DNMT1 followed by anincrease in HDAC3.28

These changes may provide evidence of the in-dividual differences in local expression of genes as-sociated with inflammation. Specific factors havebeen reported as risk factors for periodontitis, e.g.,smoking, age, bacteria, and genetic polymorphisms.Association between age and risk of developingperiodontitis has been reported.29,30 Comparison ofepigenetic patterns over time in twins at different

Figure 3.Illustration of the different forms of DNA methylation and the proteinsregulating this process. 5caC = 5-carboxydeoxycytidine, 5fc =5-formyldeoxycytidine, c = cytosine.


559

Table

1.

Ove

rview

ofStudiesonEpigeneticMechanismsin

Relationto

Periodontitis

Study

Mechanism

Marker

Material

Metho

dResults

Andiaet

al.201035

DNA

methylatio

nIL-8

Epithelialcells

from

37patients

with

generalized

AgP

and37

controls

MSP

The

IL-8

promoterwas

foundto

be

hypomethylatedin

individuals

with

generalized

AgP.

Baptista

etal.201444

DNA

methylatio

nSO

CS1,LINE-1

Oralcellsfrom

30patientswith

generalized

AgP

and30

healthycontrols

COBRA

Ahigher

percentageof

dem

ethylatio

nofSO

CS1

was

foundforthehealthygroup.A

similarresultwas

reported

for

theLINE-1gene.

BarrosandOffenbacher

201433

DNA

methylatio

nGenome-wide

screening

Gingivaltissuesamplesfrom

healthysitesanddiseased

sitesfrom

patientswith

periodontitisandtissue

samplesfrom

healthy

individuals

Microarray,LC

MSeveralthousandgeneswere

analyzed,and

differencesinDNA

methylatio

nwerefoundbetween

healthyandinflamed

tissue.

Cantleyet

al.201146

Histone

analysis

Periodontalbone

loss

Experimentalperiodontitis

mousemodel

Microcomputed

tomography,

histologicanalysis

HDACiwas

reported

tosuppress

bone

loss.Differenceswere

foundforthedifferentHDACis

used.

deFariaAmorm

inoet

al.

201343

DNA

methylatio

nTLR

2Gingivalbiopsies

from

20

patientswith

periodontitis

and20controls

Methylprofiler

DNA

methylatio

nqPCR

assay

Hypermethylatio

nofTLR

2was

reported

fortheperiodontitis

group.

deOliveira

etal.201142

DNA

methylatio

nTLR

2,TLR

4Gingivaltissuefrom

11

controls,11sm

okers

with

periodontitis,and12no

n-sm

okers

with

periodontitis

Methylatio

nanalysis

usingspecific

methylatio

n-sensitive

restrictionenzymes

andPCR

TLR

4was

unmethylatedin

most

samplesin

allgroups,whereas

theTLR

2gene

was

foundto

be

amix

ofboth

methylatedand

unmethylatedsitesin

allgroups.

Ishidaet

al.201237

DNA

methylatio

nIL-6

Peripheralbloodsamplesfrom

30patientswith

CP,30

patientswith

RA,and30

controls

Bisulfitegeno

mic

sequencing

The

methylatio

nleveloftheCpG

motif

-74bpwas

lower

inpatientswith

RA

andCP

compared

with

controls.


560

Table

1.(continued)

Ove

rview


Relationto

Periodontitis

Study

Mechanism

Marker

Material

Metho

dResults

Larssonet

al.201240

DNA

methylatio

n,histone

analysis

IL-10

Gingivalbiopsies

andperipheral

bloodsamplesfrom

six

individualswith

periodontitis;

lympho

blastoid

celllines

Bisulfate

sequencing,

HRM,ChIP

LPSstimulationofBcells

resulted

inan

increase

inacetylationof

H4andmethylatio

nofH3.

Treatm

entwith

HDACiand5-

azaresultedin

increase

inIL-10

mRNA

expression.

The

three

CpG

sitesin

thedistalIL-10

promoterregionwere

methylatedin

both

bloodand

gingivaltissue.

Looet

al.201045

DNA

methylatio

nE-Cadherin,

COX-2

Bloodsamplesfrom

108

periodontallyhealthy

controls,110patientswith

CP,and106patientswith

breastcancer

MSP

Hypermethylatio

nofE-Cadherin

andCOX-2

was

reported

for

theCPgroup

andthecancer

group.None

was

foundin

the

healthygroup.

Oliveira

etal.200934

DNA

methylatio

nIL-18

Epithelialcellsfrom

41controls,

30sm

okers

with

CP,and40

non-sm

okers

with

CP;blood

leukocytesandgingivalcells

from

13individualsfrom

each

group

MSP

Inflammationin

theoralmucosa

influencedtheDNAmethylatio

npattern

ofIL-8.Nodifferences

werefoundin

methylatio

nbetweenbloodleukocytesand

gingivalcells.A

higher

percentage

ofmethylatio

nfrequencywas

foundincontrolscompared

with

patientswith

periodontitis,in

which

mostsampleshadahigh

levelofunmethylatio

n.

Stefaniet

al.201336

DNA

methylatio

nIL-6

Gingivalbiopsies

from

21

patientswith

CPand21

controls

MSP

Nodifference

inmethylatio

nlevel

ofIL-6

was

foundwithin

the

regions

analyzed.

Viana

etal.201138

DNA

methylatio

nIFN-g,IL-10

Gingivalbiopsies

from

18

patientswith

periodontitis

and16controls

MSP,bisulfite

sequencing

The

methylatio

nlevelofIFN-g

and

IL-10was

similarin

both

groups.

Mostsampleswerepositivefor

IFN-g

methylatio

n.Ofsamples

intheperiodontitisgroup,11%

wereunmethylated,whereas

nounmethylatedsampleswere

foundforIL-10.


561

ages showed large differences in geneexpression patterns, indicating thatepigenetic changes increase during life.However, age itself contrasts with thegenetic changes seen in twins, sinceage itself is a risk factor for epigeneticchanges.31 In addition, it has beensuggested that the epigenome of olderpatients with periodontitis may differfrom that of younger patients.32 Epi-genetic mechanisms may also explainwhy some patients may not respond totreatment and why some individualsare more susceptible to periodontitis.32

Studies on epigenetic mechanisms inrelation to periodontitis33-47 are presentedin Table 1.

DNA Methylation and PeriodontitisIn periodontitis, epigenetic modifica-tions during inflammation occur locallyat the biofilm–gingival interface aroundthe teeth. It has been reported that theepigenome differs between inflamedperiodontal sites and non-inflamed sitesin the same individual.33 In addition,different methylation patterns asso-ciated with pathways regulating celldifferentiation, apoptosis, lipopolysaccha-ride (LPS)-mediated signaling, oncogen-esis, and cell adhesion were found ininflamed tissue from patients with peri-odontitis compared with tissue fromhealthy individuals.33

Several studies have investigated theDNA methylation of inflammatory cyto-kines in chronic periodontitis (CP) andaggressive periodontitis (AgP).34-41

Analysis of the interleukin (IL)-8promoter indicates a tissue-specificpattern in DNA methylation.34 Inepithelial cells, the methylation fre-quency of IL-8 was higher in controlsthan in patients with periodontitis. Nodifference was seen in gingival cells orblood leukocytes.34 These findings arein line with data that found that theIL-8 promoter was hypomethylated inoral epithelial cells from individualswith generalized AgP compared withhealthy controls.35 The IL-6 promoterwas found to be partially methylated inboth healthy individuals and patientswith periodontitis,36 even though thelevel of expression of IL-6 was higherin patients with periodontitis. A similarT

able

1.(continued)

Ove

rview


Relationto

Periodontitis

Study

Mechanism

Marker

Material

Metho

dResults

Zhang

etal.201039

DNA

methylatio

nIFN-g

Gingivalbiopsies

from

23

periodontallyhealthysites,

12experimentally

induced

gingivitissites,and12CP

sites

Pyrosequencing

Anoveralllower

DNAmethylatio

nlevelwas

foundin

the

periodontitisgroup

compared

with

periodontallyhealthyand

gingivitisgroups.

Zhang

etal.201047

DNA

methylatio

nPTGS2

10biopsies

from

inflamed

gingivaltissueandsixno

n-inflamed

gingivaltissue

biopsies

Bisulfite-specificPCR,

cloning,and

sequencing

Ininflamed

gingivaltissue,

hypermethylatio

nofthePTGS2

promoterwas

associated

with

lower

levelsofPTGS2

transcriptio

n.

Zhang

etal.201341

DNA

methylatio

nTNF-a

Gingivalbiopsies

from

17

patientswith

CP,18

periodontallyhealthy

individuals,and11with

experimentalgingivitis;T

HP1

andRAW

264.7

celllines

Bisulfite-specificPCR

andpyrosequencing

TwoCpG

sitesin

theTNF-a

promoterhadahigher

levelof

DNA

methylatio

nin

biopsies

from

patientswith

periodontitis

compared

with

those

from

controlpatients.Treatm

entwith

5-aza

resultedin

anincrease

inTNF-amRNAexpression.

IL=interleu

kin;AgP=aggressiveperiodontitis;

MSP=methylation-spec

ificpolymerase

chain

reaction;SOCS1=su

ppressorofcy

tokinesignalin

g1;LIN

E-1

=longintersperse

delem

ent-1;COBRA=co

mbined

bisulfite

restrictionanalysis;

LCM

=lase

rca

pture

microdisse

ction;HDACi=

histonedea

cetylase

inhibitor;TLR2=Toll-lik

erece

ptor2;qPCR=quantitative

polymerase

chain

reaction;CP=ch

ronic

periodontitis;

RA

=rheu

matoid

arthritis;

HRM

=high-res

olutionmeltingcu

rveanalysis;

ChIP

=ch

romatinim

munoprecipitationassay;5-aza

=5-aza-deo

xycy

tidine;

COX-2

=cy

clooxy

gen

ase

2;IFN

=interferon;PTGS2=

prostaglandin-endoperoxidesy

nthase

2;TNF-a

=tumornec

rosisfactor-alpha.


562

increase in IL-6 serum level in patients with CPcompared with healthy individuals was associatedwith hypomethylation of the -74 bp CpG site inthe IL-6 promoter. This suggests that DNA meth-ylation of a single CpG site may influence cytokineproduction and hence disease.37

The methylation level of interferon (IFN)-g hasbeen investigated in two studies.38,39 Viana et al.38

reported a similar methylation level of IFN-g and IL-10in gingival biopsies from patients with CP comparedwith biopsies from healthy controls. In contrast, Zhanget al.39 reported a lower methylation level of CpG sitesin periodontitis biopsies compared with periodontallyhealthy tissue biopsies. This was related to an increasein IFN-g transcription. The increase in IFN-g in ex-perimental gingivitis was found to be independent ofpromoter methylation. Regarding the IL-10 gene,no unmethylated samples were found in either peri-odontitis or healthy tissues.38 Three CpG sites close tothe IL-10 promoter region around the -1,087 poly-morphism were methylated in DNA from peripheralblood cells and from gingival tissue. Treatmentof B cells with the DNA methylation inhibitor5-aza-deoxycytidine resulted in an increase in IL-10mRNA regardless of genotype for the -1,087 poly-morphism.40 TwoCpG sites in the tumor necrosis factor(TNF)-a promoter were reported to be hyper-methylated in CP. In addition, transcription ofTNF-a was shown to be regulated by the meth-ylation level, since treatment of ThP1 cells with 5-aza-deoxycytidine increased TNF-a expression.41

In addition to methylation level of cytokines,a number of other genes related to inflammation havebeen analyzed in periodontitis.42-45 The DNA meth-ylation pattern in the Toll-like receptor 2 (TLR2) andTLR4 genes was evaluated in gingival tissue sam-ples.42 The TLR4 gene promoter was unmethylatedin most samples in all groups,42 whereas the TLR2promoter was both methylated and unmethylatedin most samples. A hypermethylation and a subsequentlow transcription of TLR2 in periodontitis tissues wasreported.43 Interestingly, a positive correlation wasfound between TLR2 methylation and periodontalprobing depth. In a study on AgP, an overall deme-thylation pattern of the suppressor of cytokine sig-naling 1 (SOCS1) gene was found.44 In healthyindividuals, this demethylated level was higher, andalso total demethylated samples were found for thisgroup compared to periodontitis. A significant differencewas also reported regarding the methylation pattern forthe long interspersed element-1 (LINE-1) gene.44

The methylation status of the E-Cadherin and cy-clooxygenase 2 (COX-2) genes was investigated inpatients with CP and patients with breast cancer.45

The results showed a hypermethylation of these twogenes in both patients with periodontitis and those

with cancer compared with healthy individuals. Thesefindings confirm the present view that chronic in-flammation and cancer may have a similar epigeneticpattern and that DNA methylation may be a linkbetween inflammation and cancer.48

A mechanism of how DNA methylation of a genepromoter may influence gene expression was illus-trated in a study by Zhang et al.47 in which hyper-methylation of the prostaglandin-endoperoxidesynthase 2 (PTGS2) promoter in inflamed periodontaltissue was associated with lower levels of PTGS2transcription. Interestingly, a CpG site close to a nu-clear factor (NF)-kB transcription factor-binding sitewas highly methylated, suggesting that methylationstatusmay exclude binding of this transcription factor.This is in line with other studies on methylation ofcytokine promoter regions in which the CpG sitesanalyzed were in close proximity to binding sites fortranscription factors NF-kB and specificity protein 1(Sp1).40,41 These two factors are known to be strongregulators of gene expression of the immune re-sponse. They are activated by bacteria binding toTLRs and the subsequent activation of mitogen-activated protein kinase (MAPK) pathways.49,50 Theinfluence of Sp1 and NF-kB on the chromatin struc-ture was demonstrated byWidlak et al. in 1997,51 whoshowed that the binding of Sp1 and NF-kB to adjacentsites induced a repositioning of the nucleosomesupstream and downstream of the factors, therebyinfluencing gene transcription.51 It has been reportedthat enzymes that cause methylation can be regulateddirectly or indirectly by the extracellular milieuthrough signal transduction pathways such as theMAPK extracellular signal-regulated kinase (ERK)kinase pathway.52

The inflammatory processes may also result in ab-errant DNA methylation in tissues.53 This mechanismhas been suggested as a link between inflammationand alterations in methylation level in cancers. In in-flamed tissue, neutrophils and eosinophils producesubstances that further form halogenated cytosineproducts.53 The DNMT1 enzyme cannot distinguishbetween halogenated andmethylated DNA, resulting inan increase in de novo methylation in cells of the in-flamed tissue.53 In addition, antigen stimulation canresult in downregulation of DNMT1 and HDAC2, withan increase in HDAC3, together with induction of Sp1and NF-kB p65 binding to the HDAC3 promoter. Thedownregulation of DNMT1 results in induction of ad-hesion molecules and cellular interactions.28

The process of 5mC oxidation into 5hmC has beenwell studied in cancer, but understanding of the role ofthese markers in inflammation is limited. In a studyon the autoimmune disease systemic lupus eryth-ematosus (SLE),54 the levels of TET1, TET2, and 5hmCin cluster of differentiation 4–positive (CD4+) T cellswere


563

compared to those from healthy individuals. A high levelof these markers in patients with SLE was found. It wassuggested that the DNA hydroxymethylation may becontributing to the pathogenesis of SLE.54 Because SLEand periodontitis share common features regarding theimmune response, it may also be that DNA hydrox-ymethylation plays a role in periodontitis similar to thatin SLE.55 Bacterial influence of TET2 expression hasbeen previously shown, as Escherichia coli stimulationof THP1 cells for 24 hours resulted in more than doublethe amount of TET2 mRNA.56

Histone Modifications and PeriodontitisFew studies are available on histone modifications inperiodontitis.40,46 Maintaining histone acetylation ofgenes related to osteoclastogenesis was found to beimportant for preventing bone loss in experimentalperiodontitis. Treatment with histone deacetylaseinhibitor (HDACi) resulted in improved bone levels.53

In a study from the authors’ laboratory,40 it wasdemonstrated that different epigenetic modificationsof DNA and histones influence IL-10 gene expression.The authors also found an association between LPSstimulation and histone modifications. Furthermore,these alterations differed among IL-10 genotypes forthe -1,087 G/A polymorphism in the IL-10 genepromoter.40 Administration of HDACi trichostatin Adid not alter the IL-10 production; however, treatmentusing HDACi butyrate and valproic acid resulted in anincrease in IL-10 gene expression.

Influence of Periodontal Pathogens on ChromatinModificationsAn early report on the bacterial influence on DNAmethylation showed that bacterial infection withCampylobacter rectus promoted DNA hyper-methylation of the insulin-like growth factor 2(IGF2) gene.57 Stimulation with the periodontalpathogen Treponema denticola on periodontal liga-ment cells showed a decrease in methylation of thematrix metalloproteinase (MMP)-2 promoter.58

However, the authors could not confirm these findings using the more sensitive bisulfite DNA se-quencing method.58 In addition, the MMP-2 promoterwas found to be hypomethylated regardless of thepresence of T. denticola. It was suggested that theadherence and/or internalization of T. denticola maycontribute to regulation of matrix degradation and boneresorption.58 Studies have shown that the culture super-natant from the periodontal pathogen Porphyromonasgingivalis contains high levels of butyric acid.59,60 Thisculture supernatant was reported to inhibit HDACs,resulting in increased histone acetylation.59,60

Periodontal pathogens can induce cell-specificchanges as shown in a study61 comparing the ex-pression of epigenetic markers in gingival fibroblastsand keratinocytes. P. gingivalis LPS stimulation re-

sulted in a decrease inmRNA transcription of DNMT1,DNMT3, and histone demethylases Jumonji domaincontaining 3 (JMJD3) in keratinocytes but not ingingival fibroblasts. In contrast, RNA analysis fromgingival biopsies from periodontitis and healthy sitesshowed no difference in gene expression for thesetarget genes.61 JMJD3 binds to genes and regulatesthe H3K27me3 level and hence transcription activity.It was suggested that a decrease in JMJD3 couldlead to changes in transcription levels of inflammatorygenes activated by LPS. Interestingly, JMJD3 is re-gulated by NF-kB.61 Consistent with these findingsare the results presented by Yin and Chung,8 whoinvestigated the influence of P. gingivalis and Fuso-bacterium nucleatum on epigenetic modifications ingingival epithelial cells. P. gingivalis stimulation re-sulted in a decrease in HDAC1, HDAC2, and DNMT1,whereas no difference was found after stimulationwith F. nucleatum for those markers compared withuntreated cells. These results indicate that oralbacterial infection is associated with changes in thehistone H3 lysine 4 (H3K4) methylation level, andthat the bacteria investigated induced different levelsof trimethylated H3K4. In addition, these bacteriadrove different epigenetic changes for the anti-microbial proteins human b-defensin 2 and CCchemokine ligand 20.

Specific cells within the oral mucosa may responddifferently to bacterial infection or to the periodontalinflammatory processes.34 This results in differencesin the methylation pattern of a certain gene amongcell types. Analysis of epigenetic modifications fromgingival biopsies shows the combined results of allcells present in the sample, making it difficult toestimate the role of each cell type in the epigeneticpattern at the site of inflammation.61 Barros andOffenbacher used laser capture microdissection(LCM) to analyze epithelial cells from the sulcularepithelium.33 A large number of genes were analyzed,but the gamma-amino-n-butyrate transaminase(GATA) gene better illustrated differences in geneexpression between the gingival biopsies and LCMtissue samples. The results from gingival samplesshowed a 2.5-fold increase in GATA gene methyla-tion, comparing healthy individuals versus patientswith periodontitis. This was confirmed using LCM,with which a 1.2-fold increase was found.

However, not only do bacteria influence themethylation level of a host’s genes, the bacterial DNAcan also differ in methylation level, which influencestheir ability to induce the production of inflammatorycytokines.62 The production of TNF-a in macrophageswas induced only when bacterial DNA from Ag-gregatibacter actinomycetemcomitans, P. gingi-valis, Peptostreptococcus micros, and E. coli wasunmethylated.62


564

The studies on epigenetics and periodontitis de-scribe how the immune response influences theepigenetic pattern of genes. However, it should benoted that epigenetic changes may also affect theimmune response itself, thereby providing a potentialexplanation for how individuals are more prone todevelop chronic inflammation.

CLINICAL APPLICATIONS TARGETINGEPIGENETIC MODIFICATIONS

Limited research has been done using HDAC or DNAmethylation inhibitors as treatments for periodontitisor oral inflammation. In contrast, there are numerousreports on the use of these epigenetic inhibitors incancer research. In a recent review,63 different areasfor use of epigenetic factors were identified to im-prove individualized drug therapy, in which pharmaco-epigenetic biomarkers can be used to predict drugresponse. One example is the O6-methylguanineDNA methyltransferase molecule that can be de-tected in serum and saliva at levels corresponding tothat of the primary tumor. In this manner, it would bepossible to monitor epigenetic changes in a specifictissue by using more easily accessible body fluids.Epidrugs are defined as ‘‘drugs that inhibit or activatedisease-associated epigenetic proteins for amelio-rating, curing, or preventing the disease.’’63 Indeed,the present authors have found that administration ofepidrugs (HDACi) can prevent the accumulation of

head and neck cancer stem cells and increase thesensitivity of tumor cells to chemotherapy.64,65 Atpresent, two DNMT inhibitors and two HDAC in-hibitors have been approved by the U.S. Food andDrug Administration. Although these developmentsare exciting, the complexity of solid tumors are farbeyond current understanding, and the influence ofgene expressionmediated by histonemodifications isin part impacted by the latest identification of severalmutations in histones and histone modifiers.66

Recently, HDACis have been found not only toregulate histone acetylation but also to influenceDNA methylation.52 The HDACi suberoylanilide hy-droxamic acid inhibits histone acetylation but alsoregulates DNMT1 protein expression.52 Treatmentusing another HDAC inhibitor, 1,179.4b (Classes Iand II HDACi), showed reduced alveolar bone lossand less osteoclasts per square millimeter in micewith P. gingivalis–induced periodontitis comparedwith untreated mice.46 However, no influence wasfound on the level of inflammation, which was sug-gested to indicate a direct effect of this HDACi onbone regeneration rather than an indirect effect bytargeting the inflammation.46 Moderate daily exer-cise and diet were found to reduce P. gingivalis–induced periodontitis in diet-induced obese mice byrestoring the level of histone 3 acetylation to normal,as well as NF-kB recruitment to TNF-a and IL-10promoter.67 It should be noted that P. gingivalis

produces butyric acid that in-hibits HDACs and increaseshistone acetylation, which wasshown to reactivate Epstein-Barrvirus (EBV) as well as human im-munodeficiency virus 1, indicatingthat periodontal disease may con-tribute toEBV-relateddiseases.59,60

A group of epigenetic regula-tory proteins are the bromodo-main and extraterminal domain(BET) proteins, which are able toread acetylated histone tails andrecruit transcription complexesto regulate gene transcription. Ina recent study, the BET inhibitorJQ1 was found to inhibit both aninflammatory response and alve-olar bone loss in experimentalperiodontitis. In vitro analysis ofthe function of JQ1 revealed thatbinding of the BET protein BRD4to the NF-kB promoters was de-creased, as was NF-kB phos-phorylation, and treatment withJQ1 also attenuated nuclear NF-kB. Based on these findings, it

Figure 4.Based on the studies presented in the present review, a potential hypothesis of the epigenetic eventsoccurring in a CP lesion. ROS = reactive oxygen species. AC = acetylase; SP1 = transcription factor.


565

was suggested that JQ1 may be a potential treat-ment model for periodontitis.68 Interestingly, aspirinhas been found to have a negative effect on allergicinflammation by regulating DNMT1 expression. Inthis model, aspirin was shown to prevent antigenfrom influencing the expression of HAT Tip60, HDAC3,and DNMT1.28

CONCLUSIONS

Recent studies have reported on the ability and im-portance of environmental factors to induce changesin epigenetic regulatory mechanisms of the immuneresponse (Fig. 4). Knowledge of the role of epige-netics in development of periodontal diseases is stilllimited, but the studies presented in this reviewhighlight that both DNA methylation and histonemodifications occur in the oral mucosa in response tobacteria and the inflammatory processes. To betterunderstand the influence of environmental cuesduring the development of periodontal disease, themolecular events driven by epigenetic modificationsneed to be identified. This information may alsoprovide some explanations as to why patients withsimilar clinical characteristics respond differently totreatment. Identifying epigenetic patterns associatedwith the development of periodontitis may also im-prove personalized approaches that can be used toindividualize treatment plans for patients with peri-odontitis.

ACKNOWLEDGMENTS

This work was supported by IRIS stipend (Gothen-burg, Sweden) to LL, Colgate Palmolive Company(Piscataway, New Jersey) to WVG, and the Universityof Michigan School of Dentistry (Ann Arbor, Michi-gan) to RMC. The authors report no conflicts of inter-est related to this study.

REFERENCES1. Kornman KS. Mapping the pathogenesis of peri-

odontitis: A new look. J Periodontol 2008;79(Suppl.8):1560-1568.

2. Offenbacher S, Barros SP, Beck JD. Rethinking peri-odontal inflammation. J Periodontol 2008;79(Suppl.8):1577-1584.

3. Rinn JL, Chang HY. Genome regulation by longnoncoding RNAs. Annu Rev Biochem 2012;81:145-166.

4. Smale ST. Transcriptional regulation in the innateimmune system. Curr Opin Immunol 2012;24:51-57.

5. Shastry BS. SNPs: Impact on gene function andphenotype. Methods Mol Biol 2009;578:3-22.

6. Ivanov P, Anderson P. Post-transcriptional regulatorynetworks in immunity. Immunol Rev 2013;253:253-272.

7. O’Connell RM, Rao DS, Baltimore D. microRNA regu-lation of inflammatory responses. Annu Rev Immunol2012;30:295-312.

8. Yin L, Chung WO. Epigenetic regulation of humanb-defensin 2 and CC chemokine ligand 20 expressionin gingival epithelial cells in response to oral bacteria.Mucosal Immunol 2011;4:409-419.

9. Carpenter S, Ricci EP, Mercier BC, Moore MJ,Fitzgerald KA. Post-transcriptional regulation of geneexpression in innate immunity. Nat Rev Immunol 2014;14:361-376.

10. Weill L, Belloc E, Bava F-A, Mendez R. Translationalcontrol by changes in poly(A) tail length: RecyclingmRNAs. Nat Struct Mol Biol 2012;19:577-585.

11. Bonasio R, Shiekhattar R. Regulation of transcriptionby long noncoding RNAs. Annu Rev Genet 2014;48:433-455.

12. Zhang X, Li H, Burnett JC, Rossi JJ. The role ofantisense long noncoding RNA in small RNA-triggeredgene activation. RNA 2014;20:1916-1928.

13. Peng X, Gralinski L, Armour CD, et al. Unique signa-tures of long noncoding RNA expression in response tovirus infection and altered innate immune signaling.MBio 2010;1:e00206.

14. Di Ruscio A, Ebralidze AK, Benoukraf T, et al. DNMT1-interacting RNAs block gene-specific DNA methyla-tion. Nature 2013;503:371-376.

15. Bird A. DNA methylation patterns and epigeneticmemory. Genes Dev 2002;16:6-21.

16. Barros SP, Offenbacher S. Epigenetics: Connectingenvironment and genotype to phenotype and disease.J Dent Res 2009;88:400-408.

17. Wilson AG. Epigenetic regulation of gene expression inthe inflammatory response and relevance to commondiseases. J Periodontol 2008;79(Suppl. 8):1514-1519.

18. Szerlong HJ, Hansen JC. Nucleosome distribution andlinker DNA: Connecting nuclear function to dynamicchromatin structure.BiochemCell Biol 2011;89:24-34.

19. Turner BM. Cellular memory and the histone code. Cell2002;111:285-291.

20. Backdahl L, Bushell A, Beck S. Inflammatory signalling asmediatorofepigeneticmodulation in tissue-specificchronicinflammation. Int J Biochem Cell Biol 2009;41:176-184.

21. Jenuwein T, Allis CD. Translating the histone code.Science 2001;293:1074-1080.

22. Black JC, Van Rechem C, Whetstine JR. Histone lysinemethylation dynamics: Establishment, regulation andbiological impact. Mol Cell Rev 2012;48:491-507.

23. Robertson KD, Wolffe AP. DNA methylation in healthand disease. Nat Rev Genet 2000;1:11-19.

24. Jeltsch A, Jurkowska RZ. New concepts in DNAmethylation. Trends Biochem Sci 2014;39:310-318.

25. Tahiliani M, Koh KP, Shen Y, et al. Conversion of5-methylcytosine to 5-hydroxymethylcytosine in mam-malian DNA by MLL partner TET1. Science 2009;324:930-935.

26. Ito S, Shen L, Dai Q, et al. Tet proteins can convert5-methylcytosine to 5-formylcytosine and 5-carboxylcyto-sine. Science 2011;333:1300-1303.

27. Kraus TF, GlobischD,Wagner M, et al. Low values of 5-hydroxymethylcytosine (5hmC), the ‘‘sixth base,’’ areassociated with anaplasia in human brain tumors. Int JCancer 2012;131:1577-1590.

28. Kim Y, Kim K, Park D, et al. DNA methyl transferase Iacts as a negative regulator of allergic skin inflamma-tion. Mol Immunol 2013;53:1-14.

29. Eke PI, Dye BA, Wei L, Thornton-Evans GO, GencoRJ; CDC Periodontal Disease Surveillance Work-group. Prevalence of periodontitis in adults in the


566

United States: 2009 and 2010. J Dent Res 2012;91:914-920.

30. Papapanou PN, Lindhe J. Epidemiology of periodontaldisease. In: Lindhe J, Lang NP, Karring T, eds. ClinicalPeriodontology and Implant Dentistry, 5th ed. Oxford:Blackwell Munksgaard; 2008:129-179.

31. Fraga MF, Ballestar E, Paz MF, et al. Epigeneticdifferences arise during the lifetime of monozygotictwins. Proc Natl Acad Sci USA 2005;102:10604-10609.

32. Lod S, Johansson T, Abrahamsson KH, Larsson L. Theinfluence of epigenetics in relation to oral health. Int JDent Hyg 2014;12:48-54.

33. Barros SP, Offenbacher S. Modifiable risk factors inperiodontal disease: Epigenetic regulation of geneexpression in the inflammatory response. Periodontol2000 2014;64:95-110.

34. Oliveira NF, Damm GR, Andia DC, et al. DNA methyl-ation status of the IL8 gene promoter in oral cells ofsmokers and non-smokers with chronic periodontitis. JClin Periodontol 2009;36:719-725.

35. Andia DC, de Oliveira NF, Casarin RC, Casati MZ, LineSR, de Souza AP. DNA methylation status of the IL8gene promoter in aggressive periodontitis. J Periodon-tol 2010;81:1336-1341.

36. Stefani FA, Viana MB, Dupim AC, et al. Expression,polymorphism andmethylation pattern of interleukin-6in periodontal tissues. Immunobiology 2013;218:1012-1017.

37. Ishida K, Kobayashi T, Ito S, et al. Interleukin-6 genepromoter methylation in rheumatoid arthritis andchronic periodontitis. J Periodontol 2012;83:917-925.

38. Viana MB, Cardoso FP, Diniz MG, et al. Methylationpattern of IFN-g and IL-10 genes in periodontal tissues.Immunobiology 2011;216:936-941.

39. Zhang S, Crivello A, Offenbacher S, Moretti A, PaquetteDW, Barros SP. Interferon-gamma promoter hypome-thylation and increased expression in chronic peri-odontitis. J Clin Periodontol 2010;37:953-961.

40. Larsson L, Thorbert-Mros S, Rymo L, Berglundh T.Influence of epigenetic modifications of the interleukin-10 promoter on IL10 gene expression. Eur J Oral Sci2012;120:14-20.

41. Zhang S, Barros SP, Moretti AJ, et al. Epigeneticregulation of TNFA expression in periodontal disease.J Periodontol 2013;84:1606-1616.

42. De Oliveira NF, Andia DC, Planello AC, et al. TLR2 andTLR4 gene promoter methylation status during chronicperiodontitis. J Clin Periodontol 2011;38:975-983.

43. de Faria Amormino SA, Arao TC, Saraiva AM, et al.Hypermethylation and low transcription of TLR2 genein chronic periodontitis. Human Immunobiol 2013;74:1231-1236.

44. Baptista NB, Portinho D, Casarin RC, et al. DNAmethylation levels of SOCS1 and LINE-1 in oralepithelial cells from aggressive periodontitis patients.Arch Oral Biol 2014;59:670-678.

45. Loo WT, Jin L, Cheung MN, Wang M, Chow LW.Epigenetic change in E-cadherin and COX-2 to predictchronic periodontitis. J Transl Med 2010;8:110-115.

46. Cantley MD, Bartold PM, Marino V, et al. Histonedeacetylase inhibitors and periodontal bone loss. JPeriodontal Res 2011;46:697-703.

47. Zhang S, Barros SP, Niculescu MD, Moretti AJ, PreisserJS, Offenbacher S. Alteration of PTGS2 promotermethylation in chronic periodontitis. J Dent Res 2010;89:133-137.

48. Kundu JK, Surh YJ. Inflammation: Gearing the journeyto cancer. Mutat Res 2008;659:15-30.

49. Saccani S, Pantano S, Natoli G. p38-Dependent mark-ing of inflammatory genes for increased NF-kappa Brecruitment. Nat Immunol 2002;3:69-75.

50. Liu Y-W, Chen C-C, Tseng H-P, Chang W-C.Lipopolysaccharide-induced transcriptional activationof interleukin-10 is mediated by MAPK- and NF-kappaB-induced CCAAT/enhancer-binding proteind in mouse macrophages. Cell Signal 2006;18:1492-1500.

51. Widlak P, Gaynor RB, Garrard WT. In vitro chromatinassembly of the HIV-1 promoter. ATP-dependent polarrepositioning of nucleosomes by Sp1 and NFkappaB.J Biol Chem 1997;272:17654-17661.

52. Sarkar S, Abujamra AL, Loew JE, Forman LW, PerrineSP, Faller DV. Histone deacetylase inhibitors reverseCpG methylation by regulating DNMT1 through ERKsignaling. Anticancer Res 2011;31:2723-2732.

53. Valinluck V, Sowers LC. Endogenous cytosine damageproducts alter the site selectivity of human DNAmaintenance methyltransferase DNMT1. Cancer Res2007;67:946-950.

54. Zhang Y, Zhao M, Sawalha AH, Richardson B, Lu Q.ImpairedDNAmethylation and itsmechanisms inCD4(+)T cells of systemic lupus erythematosus. J Autoimmun2013;41:92-99.

55. Berglundh T, Donati M, Zitzmann N. B cells in peri-odontitis: Friends or enemies? Periodontol 2000 2007;45:51-66.

56. Neves-Costa A, Moita LF. TET1 is a negative tran-scriptional regulator of IL-1b in the THP-1 cell line.MolImmunol 2013;54:264-270.

57. Bobetsis YA, Barros SP, Lin DM, et al. Bacterial in-fection promotes DNA hypermethylation. J Dent Res2007;86:169-174.

58. Miao D, Godovikova V, Qian X, Seshadrinathan S,Kapila YL, Fenno JC. Treponema denticola upregulatesMMP-2 activation in periodontal ligament cells: Inter-play between epigenetics and periodontal infection.Arch Oral Biol 2014;59:1056-1064.

59. Imai K, Ochiai K, Okamoto T. Reactivation of latentHIV-1 infection by the periodontopathic bacteriumPorphyromonas gingivalis involves histone modifica-tion. J Immunol 2009;182:3688-3695.

60. Imai K, Inoue H, Tamura M, et al. The periodontalpathogen Porphyromonas gingivalis induces theEpstein-Barr virus lytic switch transactivator ZEBRAby histone modification. Biochimie 2012;94:839-846.

61. de Camargo Pereira G, Guimaraes GN, Planello AC,et al. Porphyromonas gingivalis LPS stimulation down-regulates DNMT1, DNMT3a, and JMJD3 gene expres-sion levels in human HaCaT keratinocytes. Clin OralInvestig 2013;17:1279-1285.

62. Nonnenmacher C, Dalpke A, Zimmermann S, Flores-De-Jacoby L, Mutters R, Heeg K. DNA from periodon-topathogenic bacteria is immunostimulatory for mouseand human immune cells. Infect Immun 2003;71:850-856.

63. Ivanov M, Barragan I, Ingelman-Sundberg M. Epige-netic mechanisms of importance for drug treatment.Trends Pharmacol Sci 2014;35:384-396.

64. Almeida LO, Abrahao AC, Rosselli-Murai LK, et al.NFkB mediates cisplatin resistance through histonemodifications in head and neck squamous cell carci-noma (HNSCC). FEBS Open Bio 2014;4:96-104.


567

65. Giudice FS, Pinto DS Jr., Nor JE, Squarize CH, CastilhoRM. Inhibition of histone deacetylase impacts cancerstem cells and induces epithelial-mesenchyme transi-tion of head and neck cancer. PLoS ONE 2013;8:e58672.

66. Martins MD, Castilho RM. Histones: Controlling tumorsignaling circuitry. J Carcinog Mutagen 2013;29(Suppl. 5):1-12.

67. ZhouQ, Leeman SE, Amar S. Signalingmechanisms inthe restoration of impaired immune function due todiet-induced obesity. Proc Natl Acad Sci USA 2011;108:2867-2872.

68. Meng S, Zhang L, Tang Y, et al. BET inhibitor JQ1blocks inflammation and bone destruction. J Dent Res2014;93:657-662.

Correspondence: Dr. William V. Giannobile, Departmentof Periodontics and Oral Medicine, University of Michi-gan School of Dentistry, 1011 North University, AnnArbor, MI 48109-1078. Fax: 734/763-5503; e-mail:[email protected].

Submitted October 1, 2014; accepted for publicationNovember 12, 2014.


568

mailto:[email protected]

epigenetics and its role in periodontal diseases: a state

Documents