Folate and DNA Methylation

18 Pages·2011·1.07 MB·English

REVIEW
Folate and DNA Methylation: A Review of
Molecular Mechanisms and the Evidence
for Folate’s Role1,2
Krista S. Crider,3* Thomas P. Yang,4 Robert J Berry,3 and Lynn B. Bailey5
3NationalCenteronBirthDefectsandDevelopmentalDisabilities,DivisionofBirthDefectsandDevelopmentalDisabilities,Atlanta,GA,
4DepartmentofBiochemistryandMolecularBiology,CenterforEpigenetics,DivisionofPediatricGenetics,UniversityFloridaCollegeofMedicine,
Gainesville,FL,UniversityofFlorida,Gainesville,FL,and5FoodsandNutritionDepartment,UniversityofGeorgia,Athens,GA
D
o
w
ABSTRACT nlo
a
d
DNAmethylationisanepigeneticmodificationcriticaltonormalgenomeregulationanddevelopment.Thevitaminfolateisakeysourceofthe ed
onecarbongroupusedtomethylateDNA.BecausenormalmammaliandevelopmentisdependentonDNAmethylation,thereisenormous fro
m
interestinassessingthepotentialforchangesinfolateintaketomodulateDNAmethylationbothasabiomarkerforfolatestatusandasa a
d
mechanisticlinktodevelopmentaldisordersandchronicdiseasesincludingcancer.ThisreviewhighlightstheroleofDNAmethylationinnormal v
a
genomefunction,howitcanbealtered,andtheevidenceoftheroleoffolate/folicacidintheseprocesses. Adv.Nutr.3:21–38,2012. nc
e
s
.n
u
tritio
n
Introduction is a covalent modification of genomic DNA that modifies .org
Folateisanessentialwater-solublevitaminoccurringnatu- geneexpressionandprovidesamechanismfortransmitting b
y
rally in select foods as well as in the synthetic form (folic andperpetuatingepigeneticinformationthroughDNArep- g
u
acid) used in supplements and in food fortification pro- lication and cell division. The role of DNA methylation in es
grams (1–3). There are many critical cellular pathways de- cellular regulation has also provided the potential for a t on
A
pendent on folate as a 1-carbon source including DNA, new paradigm of disease intervention and treatment. The p
RNA, and protein methylation as well as DNA synthesis development of various inhibitors of DNA methylation ril 1
5
andmaintenance.Folatecanbealimitingfactorinallthese that alter methylation patterns within intact mammalian , 2
0
reactions (Fig.1). cells has led to the clinical use of some inhibitorsin exper- 1
5
Epigeneticsisthestudyofheritablechangesinphenotype imentaltherapiesforhumandiseasessuchashematological
or gene expression that do not result from changes in the malignancies (8)and myelodysplasticdisorders(9).
primary DNA sequence (4,5). Recognized mechanisms of DNAmethylationpatternsarestableandareretainedin
epigenetic regulation in mammals include DNA methyla- purifiedgenomicDNA;therefore,studiesofDNAmethyl-
tion,post-translationalmodificationofhistones,chromatin ationareamenabletoawidevarietyofcell-freeassaysand
remodeling, microRNAs, and long noncoding RNAs (6,7). technologiesincludingDNAmethylationanalysisatsingle-
Theseepigeneticregulatorymechanismsmodulatechroma- nucleotide resolution, next-generation sequencing, and
tinstructureandcontributetoregulationofthemajormo- genome-wide methylation profiling (10) (Table 1). Cur-
lecular processes in the nucleus including transcription, rently, new DNA sequencing technologies are beginning
replication, repair, and RNA processing. DNA methylation to provide novel insight into genome-wide patterns of
DNAmethylation(10–13).Althoughhigh-resolutiontruly
genome-wide studies have been limited to a very small
1Authordisclosures:Theauthorsreportnoconflictsofinterest.K.S.CriderandR.J.Berry samplesize,arecentstudydescribedtheDNAmethylation
performedthisworkasemployeesoftheUSgovernment.Thefindingsandconclusionsin levelat1505individualsites(loci)in808genesin1628hu-
thisreportarethoseoftheauthorsanddonotnecessarilyrepresenttheofficialpositionof
man genomic DNA samples (14). Even this tremendous
theCentersforDiseaseControlandPrevention.
2SupplementalTable1andSupplementalFigure1areavailablefromthe“OnlineSupporting datasetonlycoversatiny portionofthegenomeinalim-
Material"linkintheonlinepostingofthearticleandfromthesamelinkintheonlinetable ited number of samples. Despite these recent advances, a
ofcontentsathttp://advances.nutrition.org.
basic understanding of normal variation in genomic
*Towhomcorrespondenceshouldbeaddressed.E-mail:[email protected].
ã2012AmericanSocietyforNutrition.Adv.Nutr.3:21–38,2012;doi:10.3945/an.111.000992. 21
Figure1 Folic acid metabolism. This
schematic shows the process by which
folate/folicacidisusedforDNAmethylation.
TheMTHFR677C/Tvariantreducesenzyme
activity (175) and may help to divert the
available methyl groups from the DNA
methylation pathway toward the DNA
synthesis pathway (176–178). The pathway
is complex and highly regulated. with
feedback loops and interactions not shown
in the schematic. Gene names for enzymes
areinitalicsandcofactorsareinparentheses.
DHF, dihydrofolate; DHFR, dihydrofolate
reductase; DNMT, DNA methyltransferase;
dTMP, thymidylate; dUMP, deoxyuridine
monophosphate; MS, methionine synthase;
MTHFR, methylenetetrahydrofolate
reductase; SAH, S-adenosylhomocysteine; SAM, S-adenosylmethionine; SHMT, serine hydroxymethyltransferase; THF,
tetrahydrofolate; TS, thymidylate synthase.
D
o
w
patternsofDNAmethylationinhumansacrosstissues,age, Current status of knowledge n
lo
populations,disease,orenvironmentalconditions(includ- Part I: overview of DNA methylation a
d
e
ing dietary intakes) have not been well described. The Introduction to DNA methylation. Methylation of cyto- d
methods needed to undertake these types of studies and sineiscommonthroughoutthehumangenome.Thiscova- fro
m
theinfrastructuretodotheseanalysesarerapidlyemerging lentmodificationmostcommonlyoccursatcytosineswithin a
d
(10,15,16). a 59-CpG-39 dinucleotide when a methyl group from S- va
n
Thisreviewwaswrittentoprovidethenutritionalscien- adenosylmethionine (SAM)6 is enzymatically transferred to ce
s
tistwithasynopsisofthemechanisticaspectsofDNAmeth- the 5 position of cytosine to generate 5-methylcytosine .n
u
ythlaetinountraisenatbfaoclaktgerotounadffefcotrtuhnesdeerssatmanedminoglethcuelaprotpernotcieaslsfeosr. (p5r-oMduCc)tionfgtehneofmreiqcuDenNcAy.oDfNcyAtomsientehyDlaNtiAonmpeatthteyrlnatsioanreaat trition
Specifically,highlightsofthefollowingarecovered:theroles specific sites along a strand of DNA. Recent genome-wide .org
and mechanisms of DNA methylation and demethylation; high-resolutionDNAmethylationanalysisofaprimaryhu- by
the mechanisms of gene silencing by DNA methylation; manfibroblastcelllinedemonstratedthat4.25%oftotalcy- gu
e
the function of DNA methlyation; reprogramming of tosinesingenomicDNAaremethylated,67.7%ofCpGsare st o
DNA methylation patterns during development and differ- methylated,and99.98%ofDNAmethylationoccursinCpG n
A
entiation;andtheimportanceofchangesinDNAmethyla- dinucleotides (10). Similar analysis of a human embryonic p
tion (hyper- and hypomethylation). Following this, the stem cell line showed that 5.83% of cytosines are methyl- ril 1
5
potential role of folate in the DNA methylation process ated, 82.7% of CpG dinucleotides are methylated, and , 2
0
and its assessment as a folate status biomarker and link to 25% of all cytosine methylation occurs at non-CpG sites 15
disease outcomes are covered: folate’s role in 1-carbon me- (10); an earlier study also reported high levels of non-CpG
tabolism related to DNA methylation; low folate status methylation in mouse embryonic stem cells in contrast to
and DNA methylation; background of DNA methylation, methylationpatternsinmousesomaticcells(17).Although
cancer,andfolate;studiesofcancerpatientsandfolatestatus the importance of CpG methylation is established, the role
and global DNA methylation; studies of healthy adults and of non-CpGmethylation is a newactiveareaof research.
folate status; changes in DNA methylation in response to Interestingly,thefrequencyofCpGdinucleotidesislower
theenvironment and diet—the importanceof the develop- than would be expected throughout most mammalian ge-
mentaltimingofexposure;studiesoffolateinthefetus,in- nomes (18). This is believed to be due to the spontaneous
fants, cord blood; and high folate and folic acid intake and deamination of 5-MC to yield thymine and a cytosine-to-
DNAmethylation.Thereviewconcludeswithafocusonre- thymidinetransitioninDNA(19).However,certainregions
search issues including methodological considerations that of the genome have approximately a 10-fold higher
arekeyinplanningandinterpretingresearchinvestigations
related to assessing differences in DNA methylation in re-
sponse to changes in folate status. Clearly there are many
6Abbreviationsused:BER,baseexcisionrepair;BPA,bisphenolA;DMR,differentially
challenging research issues that need to be addressed to fill methylatedregion;DNMT,DNAmethyltransferase;5-HMC,5-hydroxymethylcytosine;IAP,
inthegapsinourknowledgerelatedtothepotentialforfo- intracisternalAparticle;5-MC,5-methylcytosine;5-methylTHF,5-methyltetrahydrofolate;
LINE,longinterspersedelement;MTHFR,methylenetetrahydrofolatereductase;NTD,neural
late to modulate DNA methylation and potentially have an
tubedefect;RBC,redbloodcell;SAH,S-adenosylhomocysteine;SAM,
impact on development and health maintenance. S-adenosylmethionine;SINE,shortinterspersedelement;THF,tetrahydrofolate.
22 Crideretal.
ConsReferences Requirescostlyinstrumentation;re-Shellnutetal.(137)wμquires1gamountsofDNA Requirescostlyinstrumentation;re-Armstrongetal.(186)μquiresgamountsofDNALackssensitivity;imprecise;weakBalaghiandWagner(187)reproducibility Limitedtoanalysisofindividualre-WolfandMigeon(188)strictionsites;notstrictlyquantitativeNotquantitative;limitedsequenceHermanetal.(189)coverage Requirescostlyinstrumentation;re-TostandGut(190)strictedsequencecoverageRequirescostlyinstrumentation;re-Bibikovaetal.(191),wμquires0.5gDNAChristensenetal.(131),Fernandezetal.(14)Laborintensive;time-consuming;re-Clarketal.(192)strictedsequencecoverage RequiressequencingoffullgenomeListeretal.(45)wat15-foldcoverage;cost Lowresolution,notfullgenomecov-Weberetal.(193),Nairetfierage;efciencydependentonal.(194),Rauchetal.CpGdensity(195)BiasedtowardanalyzingCpGislands;Guetal.(196)lacksfullgenomecoverage LacksfullgenomecoverageIrizarryetal.(197) LacksfullgenomecoverageKhulanetal.(198) Notstrictlyquantitative;indirectas-Balletal.(199)sessmentofmethylationbyde-tectingunmethylatedCpGsites;lacksfullgenomecoverage encing,RRBS,etc.,allcytosinesexceptmethylatedcytosinesingenomicDNAarecomprehensivelistofmethodsusedtoassayDNAmethylation.)CHARM,com-performanceliquidchromatography;LC-MS,liquidchromatography-massspec-cutcounting;RRBS,reducedrepresentationbisulfitesequencing.
Table1.1RepresentativelistofcommonlyusedmethodologiesforanalyzingDNAmethylation DescriptionPros Globalchromatog-Digestion/hydrolysisofgenomicDNAtonucleosidesandHighlysensitive,quantitative,accurate,andLiquidraphytandemmassfractionationbyLC-MSreproduciblespectrometryHPLCDigestion/hydrolysisofgenomicDNAtonucleotidesorba-Highlyquantitative,accurate,andreproduciblesesandfractionationonHPLCMethylacceptanceGlobalmethylationdeterminedbylevelofinvitromethyla-TechnicalsimplicitytionofgenomicDNAbySssImethylaseGenespecificfiMethyl-sensitivere-AssessesmethylationatspecicrestrictionsitesusingStandardmolecularbiologytechniquesstrictionenzymeSouthernblottingdigestionfifiMethylation-specificSpecicallyampliesmethylatedgenomicDNAsequencesSensitivity;technicalsimplicityfipolymerasechainaftersodiumbisultetreatmentofgenomicDNAreactionfiPyrosequencingDetermineslevelofmethylationatspecicCpGsitesHighlyquantitative;single-nucleotideresolution;technicallystraightforwardfifiMicroarrayLargenumbersofindividualCpGsitesfromspecicgenesQuantitative;specic;technicallystraightforward;areassayedusingstandardmicroarrayequipmentafterlargenumberofCpGsatatimefibisultetreatment;1500>450KsitespersampleSodiumbisulfiteDeterminesmethylationstatusofeachcytosineinaregionofSingle-nucleotideresolution;revealsDNAmeth-genomicinterestwithinasingleDNAmoleculeylationpatternsofindividualDNAmoleculessequencingGenome-wideprofilingMethylC-seqDeterminesmethylationstatusofeachcytosineinthege-Single-nucleotideresolution;fullgenomecover-nomeafterchemicalconversionofunmethylatedcyto-age;themostcomprehensivemethodforge-fisinestouracilsingenomicDNAnome-wideprolingofDNAmethylationmeDIP/meDNAUsesantimethylatedcytosineantibodyormethylatedDNA-Successfullyusedinawidevarietyofpublishedpull-down/MIRAbindingproteintoenrichformethylatedDNAstudies;technicallyaccessible RRBSAsize-selectedportionofthegenomeisisolatedafterre-Single-nucleotideresolutionstrictionenzymedigestionandsequencedaftersodiumfibisultetreatmentfiCHARMUsesrestrictionenzymethatspecicallycleavesmethylatedQuantitativeDNAHELPComparestherelativerepresentationofMspIandHpaIIAssaysbothCpGislandsandnon-CpGislandfragmentsatindividuallocisites;quantitativeMSCCAssessesmethylationstatusatallrecognitionsitesofaHighresolution;notrestrictedtoCpGislandsmethylation-sensitiverestrictionenzyme 1Forsodiumbisulfiteconversionusedinmethylation-specificpolymerasechainreaction,pyrosequencing,sodiumbisulfitegenomicsequencing,fullgenomesequchemicallyconverteduracilbysodiumbisulfite;methylatedcytosinesareresistanttoconversionandcanbedetectedbyvariousmethods.(Thistableisnotaprehensivehigh-throughputarraysforrelativemethylation;HELP,HpaIItinyfragmentenrichmentbyligation-mediatedpolymerasechainreaction;HPLC,high-trometry;meDIP,methylatedDNAimmunoprecipitation;meDNA,methylatedDNA;MIRA,methylatedCpGislandrecoveryassay;MSCC,methylation-sensitive by guest on April 15, 2015advances.nutrition.orgDownloaded from
FolateDNAmethylation 23
frequency of the CpG dinucleotide than the rest of the ge- important region for DNA methylation are CpG island
nome. These regions are referred to as CpG islands and “shores,”whichareCpG-containingregionsthatarewithin
are often correlated with the location of genes, particularly 2kbofaCpGisland,containlowdensitiesofCpGsites,and
promoter regions and other regulatory regions (20,21). have been shown in recent work to be a location of tissue-
Theseislandscompose<1%ofthegenomeandaretypically specific differential methylation in normal tissues (13).
unmethylated(Fig.2).MappingCpGislandshasbeenused Methylation of CpG island shores has been correlated with
asatoolforgenediscoverybecause>50%ofallmammalian gene expression, and certain shores have shown altered
genes are associated with CpG islands (19,22–24). CpG is- methylation patterns incoloncancer(13).
lands were originally defined as genomic regions w200
base pairs in size with a C+G content of 50% and an ob- DNA methylation and demethylation: roles and
served CpG/expected CpG >0.6 (20), although this defini- mechanisms. Methylation at CpG dinucleotides provides
tion is somewhat arbitrary and has been refined more amechanismfortransmittingDNAmethylationpatternsaf-
recently to correlate more closely with promoter regions terDNAreplicationandperpetuatingpatternsofepigenetic
and regulatory regions (21) (Fig. 2). Another functionally regulationthroughsubsequentcellgenerations.Methylation
D
o
w
n
lo
a
d
e
d
fro
m
a
d
v
a
n
c
e
s
.n
u
tritio
n
.o
rg
b
y
g
u
e
s
t o
n
A
p
ril 1
5
, 2
0
1
5
Figure2 DNA methylationpatterns atdifferent types ofgeneticloci andhow thepatternschangein cancer andin responseto
maternaldiet.This isa cartoon ofa hypothetical singlesection ofdouble-stranded DNA. Eachbarrepresents a 59-CpG -39site.
MethylatedCpGareindicatedbybarswithballsattheend.Thegreenarrowindicatestranscriptionstartandactivetranscription.The
redXequals gene silencing. Thenumbered bluerectangles areexons. (A)DNA methylationpatternsconducive totranscription:–
patternssetduringdevelopment. CpGsthroughout thegenomeare targeted formethylationby DNAmethyltransferases (DNMT1,
DNMT3a,DNMT3b).CpGislandsareregionswithaconcentrationofCpGsoften(butnotalways)associatedwiththepromoterregions
ofgenes.CpG islandsgenerally are thoughttobe actively protectedfrom DNAmethylation toallow forappropriate regulation of
transcription(179). (B)Dysregulation ofDNAmethylation seen incancer post-transformation. Cancer cellshaveaberrant DNA
methylationpatterns.OverallDNAmethylationlevelsarereduced(58,180–182).Asubsetofgeneticsitesishypermethylated,including
sometumor suppressor genes; thespecific loci hypermethylated varyamong types oftumors (114,131).(C)Maternal diet changes
methylationlevels atthemetastable epiallele, whichcan changetranscription ofaneighboring region. Metastableepialleles are
regionsofthegenomethathaveavariablemethylationpatternsthataresemiheritableacrossgenerations,illustratedbytheagouti
mouse model(147). In mice,a maternaldiet thatincludes additional methyl donorscan lead toincreased methylationat the
metastable epiallele(the intracisternal Aparticle retrotransposon), achangein expression atnearbygenetic loci,andpermanent
changesinthephenotypedistributionoftheoffspring(147).Thesequenceoftheretrotransposonisacandidateformethylationinthe
earlyembryo;however,theextenttowhichitbecomesmethylatedvariesingeneticallyidenticalindividualsbasedonmaternaldiet
(methyl donorsaswellasother dietary factors)(146,183). Arecentstudyfound candidate loci in thehumangenome thatshow
differentialmethylation dependent onseason ofbirth(148).
24 Crideretal.
of CpG dinucleotides results in symmetrically methylated the existence of mechanisms of active DNA demethylation
CpG sites (i.e., methylation of the CpG sequence on both (33).However,studiesofDNAdemethylationhavebeencon-
strands of DNA) (as seen in Fig. 2). However, after DNA troversialandmechanismsofdemethylationinmammalsstill
replication,the2resultingdaughterDNAmoleculesbecome are not well understood (33). One mechanism of passive
hemimethylated with the parental template strand main- DNA demethylation involves inhibition or loss of mainte-
taining its methylation and the newly synthesized strand nance methylation at CpG sites so that multiple rounds of
lacking methylation. Continued DNA replication of these DNA replication and cell division eventually result in the
hemimethylateddaughtermoleculeswouldeventuallyresult lossofDNAmethylationatthesesitesinsubsequentcellgen-
inthelossofDNAmethylationpatternsinsubsequentgen- erations. Several mechanisms of active DNA demethylation
erations. In 1975, Riggs (25) and Holliday and Pugh (26) alsohavebeenpostulatedandreported,butnonehavebeen
proposed a scheme by which DNA methylation patterns at conclusivelydemonstrated. One potential mechanism of ac-
CpG dinucleotides could be restored in newly synthesized tive demethylation involves direct removal of the methyl
DNAafterDNAreplication.They postulatedthathemime- groupviacleavageofahighlystableC-Cbond;however,re-
thylated CpG sites generated by DNA replication could be ports demonstrating this mechanism have not been subse-
specificallyrecognizedbyamaintenanceDNAmethyltrans- quently confirmed or reproduced (34–36). Other potential
ferasethatwouldmethylateCpGsitesonthenewlysynthe- mechanismsofactivedemethylationinvolvebaseexcisionre-
sized DNA strand immediately after DNA replication. This pair(BER)ofDNAaftereventssuchasremovalof5-MCba-
would regenerate the patterns of symmetrically methylated sesinDNAbyglycosylaseactivityorenzymaticdeamination
CpGdinucleotidesindaughtercellsandmaintaintheepige- of5-MCtothymidineandformationofaT:Gmismatch;in
D
netic states and patterns of gene regulation that existed in eithercase,theexcisedbasewouldbereplacedwithanunme- o
w
parental cells. In 1983, a DNA methyltransferase (DNMT) thylatedcytosinebyBER(37).Nucleotideexcisionrepairhas n
lo
(now referred to as DNMT1) was identified that demon- alsobeenreportedtoplayaroleinactiveDNAdemethylation, a
d
e
strated the predicted preference for hemimethylated DNA although other studies seem to contradict this mechanism d
(27,28). Thus, the general mechanism is now widely ac- (38–40). Most recently, the appreciation that notable levels fro
m
cepted as a mechanism for maintaining DNA methylation of 5-hydroxymethylcytosine (5-HMC) are present in mam- a
d
patterns after DNA replication and cell division, although malian genomic DNA and that conversion of 5-MC to va
n
more recent evidence suggests that maintenance methyla- 5-HMC in DNA can be catalyzed by mammalian TET c
e
s
tion alone may not be sufficient to fully perpetuate DNA (ten-eleven translocation) proteins has generated interest .n
u
mmeatinhtyelantaionnce pDaNttAernmsetgheynlaotmioen-woifdehe(m2i9m).ethNyolanteetdhesluesbs-, idnemtheethpyolasstiibolne (ro4l1e–4o3f)5.-OHMneCpiontemnteicahlamnisemchsanoifsmDNoAf trition
stratesprovidesaclearexampleofamechanismthatallows 5-HMC–mediateddemethylationwouldinvolverecognition .org
the transmission and retention of epigenetic information andremovalof5-HMCandreplacementwithanunmethy- b
y
through multiple cell generations. It has been postulated lated cytosine by BER. Thus, what was once regarded as a gu
e
that the availability of folate as a source of methyl groups highly stable and terminal epigenetic state, DNA methyla- s
t o
mayaffecttheabilitytomaintainDNAmethylationpatterns tion is becoming increasingly viewed as a more dynamic n
A
in replicatingcells(30,31). process. As more genome-wide methylation studies are re- p
Mammals have 3 types of DNMT: DNMT1, DNMT3a, ported, there is evidence of both increases and decreases ril 1
5
and DNMT3b (32). DNMT1 is the most abundant DNMT in DNA methylation on differentiation (44–46). Further- , 2
0
incellsand,aspreviouslydescribed,isbelievedtoact asthe more,thefindingofvariationinDNAmethylationbetween 1
5
primary maintenance methyltransferase to methylate hemi- thesexesandchangesinmethylationduringagingsuggests
methylatedDNAafterDNAreplicationandpreserveparental thatthereismoreinterindividualvariationandvariationin
DNA methylation patterns in daughter cells. In contrast, DNA methylation patterns than was previously expected
DNMT3aandDNMT3bfunctionasdenovomethyltransfer- (47–50).
asestomethylatefullyunmethylatedCpGsites.Theyfunction
primarilyduringmammaliandevelopmenttoestablishDNA Mechanisms of gene silencing by DNA methylation.
methylation patterns as epigenetic remodeling and reprog- MethylationofCpGislands,especiallythoseislandscolocal-
ramming proceed during differentiation. DNMT3a and izedwithpromotersorotherregulatoryregions,isgenerally
DNMT3b methylate different subsets of DNA sequences in associatedwithgenerepression.Themechanismsbywhich
the genome as evidenced by the different phenotypes of DNAmethylationsilencestranscriptionarenotfullyunder-
mice carrying a knockout of either methyltransferase (32). stood, although several mechanisms have been proposed.
These de novo methyltransferases also appear to play a role First, methylation of a specific regulatory DNA sequence
in maintenance methylation to fully preserve parental DNA andtheattendantinsertionofthemethylgroupintothema-
methylationpatternsindaughtercells(29). jor groove of DNA may prevent stable binding of a regula-
AlthoughDNAmethylationhasusuallybeenregardedasa tory protein (e.g., transcription factor) to that sequence,
relatively stable epigenetic mark, there are clear examples thereby directly preventing gene activation by a transcrip-
where DNA demethylation in mammalian cells occurs in tion factor. In vitro binding studies have identified se-
theabsence ofDNA replication and cell division, suggesting quence-specific DNA-binding transcription factors whose
FolateDNAmethylation 25
interactionwiththeircognateDNArecognitionsequenceis correlated with transcriptional silencing of the associated
negatively affected by methylation of a CpG dinucleotide gene, particularly for CpG islands associated with a gene
withinthebindingsequence(51).However,manytranscrip- promoterregion(55).Oneclearexampleofthisistheaber-
tionfactorsarenotsensitivetoDNAmethylationandnotall rant stable transcriptional silencing of certain tumor sup-
DNAsequencesboundbyasequence-specifictranscription pressor genes in cancer cells by the hypermethylation of
factor will contain a CpG dinucleotide. Therefore, other their promoter CpG islands (64). However, methylation of
moreindirectmechanismsmustalsobeinvolvedinrepres- a CpG island does not necessarily lead to gene silencing.
sing transcription by DNA methylation. One such indirect For example, the gene for telomerase has been shown to
mechanismisviaDNA-bindingproteinscontainingmethyl- beactivatedbymethylation(65).DNAmethylationalsooc-
ated DNA-binding domains (52). These methylated DNA- curswithinthebodyofgenesandbetweengenes;thefunc-
binding proteins, such as methyl CpG–binding protein 2 tion of methylation at CpG sites in these regions still is
and methylated DNA–binding domain 2, recognize and poorlyunderstood.Also,itisnotclearwhethermethylation
bindto1ormoremethylatedCpGdinucleotides.Thesepro- of critical CpG sites within CpG islands triggers hyperme-
teinsinturninteractwithorrecruittranscriptionalsilencing thylationoftheislandandgenesilencing(66),whetherover-
complexes containing chromatin remodeling complexes all levels of hypermethylation within a CpG island govern
and/orhistonemodificationenzymes(e.g.,histonedeacety- silencing,orwhetherbothmechanismscontributetosilenc-
lases) that act to form a transcriptionally repressive con- ing.Recently,incontrasttoCpGislandmethylation,reports
densed chromatin structureatthe associated gene (53–55). indicatethatmethylationinthebodyofgeneshasaroleasa
DNMTshavealsobeenfoundascomponentsofmultisubu- positiveregulatoroftranscription(56)andmayalsohavea
D
nit chromatin modifying and remodeling complexes in- role in transcription elongation (67), alternative promoter o
w
volved in transcriptional repression, suggesting that the use (68), and nucleosome positioning (69). Generally, n
lo
DNA methylation machinery also may be recruited to spe- DNAmethylationisthoughttomaintaingenesilencingver- a
d
e
cific genes by other mechanisms (similar to those that re- susinitiatinggenesilencing(70)suchthatageneissilenced d
cruit histone modification and remodeling activities) and initially by another mechanism and that silencing is stabi- fro
m
thatchromatinmodification/remodelingandDNAmethyl- lized by methylation of associated regulatory regions (e.g., a
d
ation act cooperatively to silence transcription. In contrast, CpG island, promoter region). va
n
intragenic DNA methylation of active genes is often higher X-chromosome inactivation is the process by which c
e
s
than in promoter regions, suggesting that methylation in female mammals compensate for the unequal dose of .n
u
tahreebceondtysotufdgyenseusgigsesatspothsiattivtehrisegmulaaytoorcocufrgebnyeseuxppprreessssioinng; X(7-1li)n.kFeedmgaelnesesrabnedtwomeelnyminaalceti(vXatYe)1anodf tfheme a2leX(XchXr)omceoll-s trition
therepressiveeffectsofthepolycombcomplex,whichisas- somes present in each and every somatic cell in a process .org
sociated with chromatinremodeling (56). that occurs in early female embryogenesis, resulting in b
y
monoallelic expression of genes subject to X-chromosome gu
e
Function of DNA methylation. Cytosine methylation has inactivation.Onceafemalecellinearlyembryogenesisran- s
t o
been hypothesized to be an ancient component of the im- domlychooseswhichXchromosometoinactivate(i.e.,the n
A
mune system designed to recognize and inactivateparasitic maternally or paternally inherited X chromosome), all mi- p
viralDNAsequencesthatinfiltratethegenome(57).Mam- totic progeny of that cell maintain inactivation of the same ril 1
5
malian genomes contain a large proportion of repetitive X chromosome throughout the remainder of development , 2
0
DNA sequences (e.g., nearly 50% of human genomic into adult tissues. One of the hallmarks of the inactive X 1
5
DNA)thatincluderetroviralsequences,transposons,retro- chromosome is the hypermethylation of promoter CpG is-
transposons long interspersed elements (LINEs), short in- lands along the length of the inactive X, which contributes
terspersed elements (SINEs), extended blocks of tandemly to the stable maintenance of transcriptional silencing of
repeated DNA sequences (e.g., satellite DNA), etc. Most of the associated genes on the inactive X chromosome (72);
theDNAmethylationpresentinmammaliangenomesisas- these same CpG islands are typically unmethylated on the
sociated with these repetitiveDNA sequencesand servesto activeX chromosome.
suppress the potential activity and deleterious effects of re- Genomic imprinting is another mammalian epigenetic
petitive DNA (58–60). This includes suppressing mobility systemofmonoallelicgeneexpressioninwhichthe2alleles
andpropagationofrepetitiveelementscapableoftransposi- of a particular gene within the same cell are differentially
tionandsuppressingillegitimaterecombinationbetweenre- regulated, in this case, in a parent-of-origin fashion (73).
lated repetitive elements (58–60). These functions serve to Thus, for some imprinted genes, only the paternally inher-
prevent widespread genome instability and rearrangement itedalleleisexpressed,whereasthematernallyinheritedal-
mediated byrepetitiveDNA sequences (61). leleisstablysilenced.Conversely,forotherimprintedgenes,
It is hypothesized that over time the function of DNA thepaternallyinheritedalleleisstablysilencedandonlythe
methylationhasevolvedandnowplaysadditionalroles,in- maternally inherited allele is transcriptionally active. This
cludingtranscriptionalregulation,X-chromosomeinactiva- differentialmonoallelicexpressionofthematernalorpater-
tion, imprinting, and tissue-specific gene expression nalalleleisusuallyassociatedwithdifferentialDNAmethyl-
(19,62,63). Hypermethylation of CpG islands is highly ationof 1ormoreregulatory regionsof these genes. These
26 Crideretal.
so-called differentially methylated regions (DMRs) are exceptiontothisdynamicreprogrammingofDNAmethyl-
generally (but not always) hypermethylated on the tran- ationpatternsduringembryogenesisoccurswithimprinted
scriptionally silenced allele and hypomethylated on the ex- genes that escape this process of erasure and resetting,
pressed allele. Imprinting control regions that regulate thereby maintaining throughout embryogenesis the differ-
imprinting across an entire domain of imprinted genes are entialDNAmethylationofthepaternallyandmaternallyin-
typically DMRs (74,75). The differential parent-of-origin herited alleles in somatic cells characteristic of most
DNA methylation patterns at imprinted genes are estab- imprinted loci(90).
lishedduringgametogenesiswhereDNAofthemaleandfe- TheseperiodsofdynamicreprogrammingofDNAmeth-
male gametes acquire either maternal- or paternal-specific ylation patterns during gametogenesis and embryogenesis
methylation patterns before fertilization (although some mayalsopresentwindowsofopportunityforenvironmental
parent-of-origin methylation patterns may also continue influences (e.g., diet and nutrition) on the developing em-
to develop post-zygotically) (Supplemental Fig. 1A) (76). bryotoalterthenormalprocessofestablishingDNAmeth-
These differential gametic methylation imprints are then ylation (and other epigenetic) patterns (91–93). Such
maintained and propagated during subsequent somaticde- changes in DNA methylation patterns could affect normal
velopment to facilitate the parent-of-origin patterns of patternsofgeneexpressionoraltergenomestability,thereby
monoallelic transcription associated with imprinted genes. leading to an increased risk of diseases later in life, even if
Imprintedgenesarethoughttocompriseasmallproportion thesechangesinmethylationoccurinonlyasubsetofcells
of mammalian genomes with <100 currently confirmed in of a giventissue.
humansandmice(77).Inhumansthereisalimitednumber It is also possible to alter DNA methylation levels and
D
of known imprinted loci (<80), and only a subset of these patterns within intact mammalian cells by treatment with o
w
aresharedwiththemouse(78,79).However,arecentstudy various chemical inhibitors, most commonly cytidine ana- n
lo
of imprinting in embryonic and adult mouse brains found logs such 5-azadeoxycytidine. 5-Azadeoxycytidine (and its a
d
e
parent-of-origin bias in the expression of 1300 genes and ribonucleotide analog 5-azacytidine) functions as a deme- d
transcript isoforms and, if confirmed, would suggest that thylating agent(9), resulting in hypomethylation of the ge- fro
m
imprinting(orparent-of-origineffects)maybemorewide- nome in treated cells. These cytidine analogs and other a
d
spread and complex than originallybelieved(80). DNA-demethylating drugs have recently been introduced va
n
as potential therapeutic agents for the treatment of human c
e
s
Reprogramming of DNA methylation patterns during diseases, particularly myelodysplastic syndromes (9). .n
u
dteervnesloupnmdeerngtoarenmdoddieflfienrgendtuiaritniognt.heDcNoAursmeeotfhmylaatmiomnapliaatn- What does a change in DNA methylation reflect? trition
differentiation and development to eventually generate the Hypermethylation versus hypomethylation. Any CpG .org
cell type–specific methylation patterns found in adult so- site in a single DNA molecule can either be methylated or b
y
maticcells.Dynamicchangesinmethylationareparticularly unmethylated,sothelevelofmethylationatthatsitewillei- gu
e
pronouncedduringgametogenesisandearlyembryogenesis ther be 0% or 100%. However, within a single diploid cell, s
t o
(SupplementalFig.1).Duringthecourseofgametogenesis, the methylation level of a given CpG site on 1 strand of n
A
DNA methylation patterns are erased (i.e., demethylated) DNAcanbe0%(symmetricallyunmethylatedonbothchro- p
within each developing germ cell (primordial germ cells) mosomes), 50% (symmetrically methylated on 1 chromo- ril 1
5
andthenremethylatedandresettothegenomicDNAmeth- some and unmethylated on the other chromosome, as , 2
0
ylationpatternsspecifictoeitherthespermortheeggasga- oftenseeninDMRsofimprintedgenes),or100%(symmet- 1
5
metogenesis proceeds (81,82). This process is particularly rically methylated on both chromosomes). DNA methyla-
important for establishing the gamete-specific methylation tion assays are usually performed on a population of cells;
patternsofimprintedgenesinspermandeggbeforefertili- therefore, the level of methylation at a given CpG site in
zation. Current data suggest that other nonimprinted loci these assays reflects the collective methylation at this site
mayalso be subject to at least partial erasure, and resetting inmultiplecells(andmultipleDNAmolecules)andthevar-
of DNA methylation patterns during gametogenesis (83– iation in methylation seen at individual CpG sites. Thus, a
85). In early embryogenesis, DNA methylation patterns finding that a particular CpG site is 50% methylated in a
also undergo a genome-wide process of erasure, that in populationofcellscouldmeanthatthissiteis50%methyl-
mouse embryogenesis begins shortly after fertilization and atedineverycell(i.e.,1allelemethylatedandtheotherallele
continuestotheblastocyststage(86).Thiswaveofdemeth- unmethylated),thatthissiteisunmethylatedin50%ofthe
ylationerasesthemethylationpatternsofthegametesandis cellsinthepopulationandfullymethylatedintheother50%
followed by DNA remethylation during the remaining of cells, or some other combination of methylation levels
course of embryogenesis to establish the genome-wide and amongcellsinthepopulationthatyieldsacompositemeth-
cell-type specific DNA methylation patterns found in fully ylationlevelof50%.Thedifferentfunctionalconsequences
differentiatedadultsomaticcells(44,87,88).Insmallstudies of these 3different scenarios areimportant.
ofhumanembryos,ithasbeenshownthatthereisawaveof Whenconsideringtheimplicationsofchangesintheper-
demethylation atthe 4-cellstage followedbyremethylation centage of methylation at a specific CpG site (e.g., in re-
beginning as early as the late morula (89). An important sponse to dietary conditions), it is important to consider
FolateDNAmethylation 27
thata10%changeinDNAmethylationlevelatasitecould homocysteine remethylation, SAM formation, and DNA
be distributed across an entire population of cells (and not methylation. These nutrients include vitamin B-6 (serine
appreciably affect transcription), reflect a 100% change in hydroxymethyltransferaseactivity),riboflavin(MTHFRsta-
asubpopulationofcells(andrepresentgenesilencingorac- bility), vitamin B-12 (methionine synthase function), and
tivation in these cells), or something in between. Each of choline(betaineprecursorasahepaticmethylsourceviabe-
these potential scenarios has different implications for the taine:homocysteinemethyltransferase) (98).
impact of thesame10% change in methylation level. The1-carbonpathway,andthusDNAmethylation,func-
InthecaseofCpGislands,genesilencingisgenerallyas- tionsundertightregulatorycontrols.SAMisthemajorreg-
sociatedwithmultiplemethylatedCpGsitesacrossastrand ulator of folate-dependent homocysteine remethylation
ofDNA(Fig.2).However,site-specificmethylationcanre- because itis apotentinhibitorof MTHFR. Under thecon-
sultingenesilencingifthemethylatedCpGblocksthebind- dition of high SAM concentration, MTHFR is inhibited,
ing of a DNA methylation-sensitive transcription factor. At which reducesthe synthesisof 5-methylTHF and hencere-
most genetic loci, it is unknown which CpG sites or how methylation of homocysteine. Conversely, when SAM con-
many along the strand of DNA must be methylated to si- centrations are low, remethylation of homocysteine is
lenceanygivengene.Whenevaluatingtheimpactofchanges favored.MTHFRactivityandthus5-methylTHFformation
inDNAmethylation,itisimportanttoconsidertheexactlo- may also be modified by the common genetic variant,
cationassayedandthepotentialofmethylationatthatsiteor 677C/T,whichreducesenzymeactivity(99).Itisalsorec-
region to change gene expression or other biological pro- ognized thatS-adenosylhomocysteine(SAH)functionsasa
cessesof interest. potentproductinhibitorofSAM-dependentmethyltransfer-
D
ases (100). For this reason, continual hydrolysis of SAH to o
w
Part II: Folate and DNA methylation homocysteineisessentialtomaintainnormalDNAmethyl- n
lo
Folate’s role in 1-carbon metabolism related to DNA ation (101). Moderate elevations in plasma homocysteine a
d
e
methylation. Under normal dietary conditions, absorbed concentration have been reported to be associated with in- d
folate is metabolized to 5-methyltetrahydrofolate (5-meth- creasedconcentrationofSAH,butnotSAM,andincreased fro
m
ylTHF, monoglutamyl form) in the intestine and/or liver. SAHconcentrationhasbeenassociatedwithglobalDNAhy- a
d
5-MethylTHF is the primary folate constituent taken up pomethylation (102). va
n
by nonhepatic tissues, which then must be polyglutamated c
e
s
forcellularretentionand1carboncyclecoenzymefunction. LowfolatestatusandDNAmethylation.Lowfolatestatus .n
u
Tpeotlryaghluydtarmofaotleatseyn(tThHetFa)sei;sththeeremforoes,t5e-fmfecettihvyelTsHubFstmrautestfboer (caesntdreafitinoends,bfyolvaatreioinutsakmee,aasnudr/eosrinfoclliucdaincigdbilnotoadkef)olhaatsebcoeenn- trition
converted to THF via the methionine synthase reaction associated with an increased risk of cardiovascular disease, .org
(Fig. 1). When folic acid is consumed in fortified foods or multiple cancers, and neural tube defects (103–105). The by
supplements, it is metabolized primarily to 5-methylTHF mechanismsbywhichlowfolatestatuscontributestothese gu
e
duringintestinalabsorptionand/orfirstpassintheliver,af- disorders remain unclear. During DNA replication, folate st o
terwhichitbehavesidenticallytonaturaldietaryfolate.Folic depletion can have destabilizing consequences. Inadequate n
A
acidisnormally firstreducedtodihydrofolatebydihydrofo- folateavailabilityduringcelldivisioncanresultinthecom- p
late reductase and subsequently to THF to enter the folate promisedproductionofthymidine,suchthaturacilmaybe ril 1
5
pool(Fig.1).Insomecases,thecapacityofdihydrofolatere- substituted in the DNA sequence (Fig. 1). This mutagenic , 2
0
ductaseisexceededandfolicacidmayappearinthecircula- eventmaytriggerattemptstorepairthedefectandincreases 15
tion in the oxidized form (94). Once the THF coenzyme is the frequency of chromosomal breaks (104). In tissue cul-
formedfromeitherfolicacidordietaryfolate,itisfirstcon- ture,ithasbeenshownthatlowfolicacidresultsinthefor-
vertedto5,10-methyleneTHFbythevitaminB-6–dependent mationofmicronuclei(indicativeofchromosomebreakage)
enzyme serine hydroxymethyltransferase and subsequently andthattheMTHFRTTgenotypeleadstoincreasedmicro-
irreversibly reduced to 5-methylTHF by methylenetetrahy- nuclei formationunder low folateconditions (106).
drofolatereductase(MTHFR).Thisreactioniskey tomain-
taining the flux of methyl groups for the remethylation of DNAmethylation,cancer,andfolate.Globalhypomethy-
homocysteinetomethionineviathevitaminB-12–dependent lation and targeted hypermethylation are considered defin-
methioninesynthasereaction.Methionineisthesubstratefor ing characteristics of human tumors (14,23,58,107,108)
SAMorAdoMet,acofactorandmethylgroupdonorfornu- (Fig.2B).DNAmethylationpatternsarewidelydysregulated
merous methylation reactions including the methylation of inhumancancer(Fig.2B).EarlyinthestudyofDNAmeth-
DNA, RNA, neurotransmitters, and other small molecules, ylation,genome-widehypomethylationwasfoundintumor
phospholipids,andproteins,includinghistones(95).Anum- tissues compared with matched healthy tissues (58). Addi-
ber of SAM-dependent reactions have regulatory roles by tionally, w5% of genes were found to be hypermethylated
affectingbothgenomestabilityandgenetranscription(55),lo- in nonrandom patterns specific to the type of tumor
calizationofprotein(96),andsmallmoleculedegradation(97). (109). Itwas determinedthat thesechangesin DNA meth-
Inadditiontofolate,anumberofotherdietarynutrients ylationintumorsresultedinsilencingoftumorsuppressor
are required to maintain 1 carbon flux, ensuring normal genes and chromosome instability (Fig. 2B) (110,111).
28 Crideretal.
Hypomethylationofrepetitiveelementsmaybepredictiveof The association of global hypomethylation with folate
increasedriskofcancerandincreasedmortalityfromcancer status(e.g.,intake,bloodfolateconcentration)hasbeenin-
(112). In recent large-scale DNA methylation studies of consistent among studies of cancer patients (Supplemental
multipletypesofprimarytumors(thousandsoflociinhun- Table 1). This is not unexpected given the variety of study
dredsoftumors),DNAmethylationchangesintumorswere designs,differentglobalDNAmethylationassays,andtypes
associatedwithconcomitantchangesingeneexpressionand oftissuesexamined.Therearemanytechniquesavailableto
tumor characteristics (113,114). The DNA methylation assess “global” DNA methylation level including those that
changes in cancerdo not appear to be random because tu- look at repetitive element methylation level LINE, SINE,
morsofspecificcelllineagesshowsimilarDNAmethylation and Alu and those that examine various restriction sites
changesthataredistinctfromothertumortypesandnormal throughout the genome. However, the correlation between
tissues.Astumorsincreaseinseverity,aprogressivenumber thesemeasuresisactuallyquitelow(127).Thisisnotneces-
of loci show increased DNA methylation primarily at CpG sarilyduetotheaccuracyofthemeasurebutduetothefact
islandsanddecreasedDNAmethylationinnon-CpGislands thatalthoughtheycanbethoughtofasglobal,theyareac-
sites(14).ThemagnitudeofDNAmethylationchangesob- tually measuring different types and subsets of DNA se-
servedincancertissuesissubstantiallygreaterthananydif- quences in disparate regions of the genome that may be
ferencesinDNAmethylationbetweentypesofnormaltissue regulateddifferently.Indeed,recentworksuggeststhatthere
(14). aredifferentchangesinDNAmethylationbetweentypesof
Folicacidintakehasbeenreportedtopreventlossofhet- tumorsandtypesofrepetitiveDNAelementsoftenusedas
erozygosity of a tumor suppressor gene in a small supple- proxies for global methylation (128). Additionally, each of
D
mentation trial in humans (115). Folate depletion of the early trials of global DNA methylation in colon cancer o
w
humantissueculturecellscanresultnotonlyintheexpected patients used the methyl acceptance assay (30,118,119). n
lo
global hypomethylation (116,117) but also targeted hyper- Thisassay providesanestimateofglobalDNAmethylation a
d
e
methylation of the H-cadherin locus (117). It is unclear by using an enzyme to methylate purified genomic DNA d
whether increases in dietary folate and/or folic acid result with radiolabeled tritium. Methyl acceptance is limited in fro
m
inchangesinhealthytissuesthatcanpredisposetocarcino- precision and necessitates intact DNA (129) so that the a
d
genesis.Manystudiesoftheeffectsoffolicacidsupplemen- use of damaged DNA (which may occur as a result of the va
n
tation have focused on either preventing or promoting outcome of interest or during preparation or handling) c
e
s
cancer,especiallycolon cancer(Supplemental Table 1). willresultininaccurateresults.Carefulconsiderationofas- .n
u
Studies of cancer patients and folate status and global sparyiamteeitnhtoedrporloetgaytiiosncraitnicdarlefporrosdtuudcyibdileitsyigonfttoheendsuatrae.appro- trition
DNA methylation. Therehasbeengreatinterestinrevers- .org
ing epigenetic changes seen in early tumorigenesis (e.g., Studies of cancer patients and folate status and site- b
y
global hypomethylation) as rapidly dividing tissue tumors specific DNA methylation. For patients with a previous gu
e
may be susceptible to low folate availability, resulting in adenomaofthecolon,todate,only1trialexaminedtheef- s
t o
global hypomethylation. Three clinical trials examined the fect of folic acid supplementation (1 mg/d for 3 y) on site- n
impactoffolicacidsupplementation(0.4–10mg/d)onglobal specific DNA methylation, and it was limited to 2 loci Ap
DNAmethylationincoloncancerpatients(30,118,119)(Sup- (49). In tumors, methylation of tumor suppressor CpG is- ril 1
5
plemental Table 1). These studies have shown that DNA lands is generally either fully methylated and silenced , 2
0
methylation might be a biomarker forcolorectal cancer and (>80% methylated), or unmethylated (<20% methylated) 1
5
that methylation of DNA in the colon and leukocytes can depending on the exact location of the CpG sites (130).
be increased by supplementation with folic acid in human The changes in DNA methylation observed in association
patients (118,120). Among the observational studies (e.g., withvariation in red blood cell (RBC) folate concentration
cohort and case-control) of the association of folate status inhumansaremuchsmaller thanthosepresentintumors.
(measuredasfolate/folicacidintakeandorbloodfolatecon- Wallace et al. (49) evaluated DNA methylation levels at 2
centration) and cancer, 5 studies examined the global DNA CpG island–containing genes in colon tissue of persons
methylationlevel(121–125)(SupplementalTable1).Inthese with a previous adenoma after a placebo-controlled folic
studies,anassociationwasfoundbetweencancerandglobal acid supplementation trial (1 mg/d for 3 y). They found a
DNA hypomethylation in circulating blood and/or tumor trend of greater methylation at each genetic locus as the
tissue; however, there was not a consistent association be- quintile of RBC folate increased [ERa: lowest RBC folate
tween global DNA methylation and folate status (intake or quintile(Q1 43.3–520.9mg/L)vs.highestRBCfolatequin-
bloodfolate) acrossstudies(121–125).Thislackofadefini- tile (Q5 1081.2–2620.8); 10.3% vs. 11.3% methylated;
tiveassociationbetweenlowfolatestatusandglobalDNAhy- P-trend = 0.03 and SFRP1: Q1 vs. Q5; 21.2% methylated
pomethylation could be due to a limited sample size or vs.23.0%methylatedP-trend=0.01].Althoughstatistically
imprecisionoftheassays.AcrossstudiesoutlinedinSupple- significant, the functional impact of these small changes in
mentalTable1,globalhypomethylationwasfoundintumors DNA methylation (<2%) on gene transcription or longer
and even normal colon cells and blood of cancer patients term health outcomes is unclear at this time. There was
(121,122,124,126). no association of DNA methylation at either of these loci
FolateDNAmethylation 29
withtreatmentwith1mg/dfolicacidfor3y,plasmafolate group),differencesinagegroups,differentinitialfolatesta-
concentration, or dietary folate intake. Although age, pro- tus, and varying lengths ofintervention.
tein intake, and race were strongly associated with DNA Two additional observational studies examined global
methylationattheseloci.Inthisstudy,therewerenostatis- DNAmethylationandfolatestatusinpatientswithoutcan-
tically significant associations of methylation at either of cer (126,139) (Supplemental Table 1). Friso et al. (139)
these loci with hyperplasic polyps or adenomas after 3 yof found an MTHFR genotype-dependent association be-
treatment. tweenlowerglobalDNAmethylationandlowerplasmafo-
Twoadditionalobservationalstudiesofcancerexamined late concentration. Pufulete et al. (126) observed a trend
site-specific DNA methylation and folate status (131,132). toward lower DNA methylation with lower serum folate
Kim et al. (132) examined DNA methylation at 3 genetic concentration. Taken together, these findings suggest that
lociandfoundthathighplasmafolateconcentrationamong there may be a differential global DNA methylation re-
cases($6.7mg/Lvs.<4.1mg/L)wasassociatedwithanin- sponsetofolatedepletionandrepletiondependentonage,
creased methylation at p73 but not p16in4A or hMLH1. genotypes, duration, and magnitude of exposure. None
Christensen et al. (131) recently examined a cohort of of these studies of healthy adults examined site-specific
women with breast cancer and the DNA methylation level DNA methylation changes; however, in a study of men
at1413sitesin733genes.Theyfoundthatdietaryfolatein- with hyperhomocystinemia, global hypomethylation was
take was associated with DNA methylation class member- reversedandmonoallelicexpressionofimprintedgenesre-
ship in primary breast tumors (131). This type of study stored in patients given folic acid therapy (140) (Supple-
examiningthousandsoflocisimultaneouslywilllikely pro- mentalTable1).
D
duceabetterunderstandingoftheinteractionbetweenDNA o
w
methylationandfolate,althoughsubstantialanalyticandin- ChangesinDNAmethylationinresponsetotheenvironment n
lo
formaticschallenges exist. anddiet:theimportanceofthedevelopmentaltimingof a
d
e
Itisimportanttonotethatlowfolatestatuscanleadto exposure. There is increasing evidence of more interindi- d
both hypo- and hypermethylation, resulting in misregula- vidual variation and variation in DNA methylation pat- fro
m
tionofthiscomplexsystem.Thechangesobservedincan- terns across the life span than was previously expected, a
d
cer do not appear to be random such that not all CpG andthesechangesmaybemodulatedbyenvironmentalex- va
n
islands are susceptible to hypermethylation (133,134). posures. DNA methylation varies between the sexes and c
e
Some of these changes appear to precede transformation changes during aging (47–50). In addition, an increasing s.n
u
(oifntirtiaantsiofonrmofatciaonnc.eAr)t,thanisdtiomthe,eritsims uayncbleearawcohnetsheeqruesunbce- naiurmpboellrutoiofnen(1vi4r1o)n,mbeennztaelneexepxopsousruersei(n14h2u)m, aanndspsuarcthicua-s trition
sets of genetic loci are susceptible to DNA methylation late pollution (141) have been linked to changes in DNA .org
changesinresponsetofolate/folicacidintakeincancerpa- methylation (e.g., repetitive element hypomethylation) in b
y
tients; the studies highlighted in Supplemental Table adults.Furthermore,Frisoetal.(143)foundthatestrogen gu
e
1showanumberoflocithatwarrantexaminationinfuture replacement resulted in lowering of plasma homocysteine s
t o
studies. Large-scale genome-wide analysis may provide concentration and increases in global methylation in a n
A
additional candidates. small (n = 13 postmenopausal women) double-blind, p
placebo-controlled, double-crossover study, whereas no ril 1
5
Studies of healthy adults and folate status. In healthy changes in methylation were found at the ERa, ERb, and , 2
0
adults,dividingtissuessuchasthebloodandthegutmucosa p16 genes. Thus, certain environmental exposures have 1
5
arelikelytobemostsusceptibletolowfolateinthediet.In been associated with changes in DNA methylation and
humans, 3 studies examined the effects of controlled folate changes in 1-carbon metabolites.
depletionandrepletiononglobalDNAmethylationinleu- There has been increasing interest in in utero effects of
kocytes(135–137)(SupplementalTable1).Folatedepletion environmental exposures (from toxins to nutrition) on
resulted in reductions in global DNA methylation in older long-term health outcomes, and it has been hypothesized
women in controlled feeding studies (135,136), a finding that disease risk may in part be determined by maternal
thatwasnotobservedbyShelnuttetal.(137)inyoungerfe- and paternal diet (87,144,145). Studies of cancers have
maleadults.BothJacobetal.(136),andShelnuttetal.(137) found that there are changes in DNA methylation in the
observed considerable increases in methylation with folate blood that reflect the cancer risk in the tissue of interest
intake during the repletion period. In the Shelnutt et al. (119,124).Thismaybebecausedividingtissuesaresuscep-
(137) study, the increases were limited to those with the tible to the same insults or that the change that resulted in
MTHFR TT genotype (see Fig. 1 for location in the folic thepredispositiontocancerhappenedintheembryonicpe-
acid cycle). Axume et al. (138) did not find DNA methyla- riodandisreflectedinmanytissues(SupplementalFig.1C).
tionchangesinresponsetodepletionorrepletion;however, In mice it is clear that maternal dietary changes in 1-
thestudywasrestrictedtothosewiththeMTHFRCCgen- carbonavailabilitycanaffecttheDNAmethylationpatterns
otype.Althoughtheresultsfromthecontrolledfeedingtrials andphenotypeofoffspring(87).Thebest-studiedexample
of healthy adults are somewhat inconsistent, these studies istheagoutimouse.Theagoutimousestrainhasaninsertion
are limited by small sample size (n = 8–33 individuals per ofanintracisternalAparticle(IAP)sequenceinanupstream
30 Crideretal.