1、Cell MetabolismPerspectiveEpigenetic Mechanisms of Transmissionof Metabolic Disease across GenerationsVicencia Micheline Sales,1Anne C.Ferguson-Smith,2and Mary-Elizabeth Patti1,*1Integrative Physiology and Metabolism,Research Division,Joslin Diabetes Center and Harvard Medical School,One Joslin Plac
2、e,Sixth Floor,Boston,MA 02215,USA2Department of Genetics,University of Cambridge,Downing Street,Cambridge CB2 3EH,UK*Correspondence:mary.elizabeth.pattijoslin.harvard.eduhttp:/dx.doi.org/10.1016/j.cmet.2017.02.016Both human and animal studies indicate that environmental exposures experienced during
3、early life canrobustly influence risk for adult disease.Moreover,environmental exposures experienced by parents duringeither intrauterine or postnatal life can also influence the health of their offspring,thus initiating a cycle of dis-easeriskacrossgenerations.InthisPerspective,wefocusonepigeneticm
4、echanismsingermcells,includingDNA methylation,histone modification,and non-coding RNAs,which collectively may provide a non-geneticmolecular legacy of prior environmental exposures and influence transcriptional regulation,developmentaltrajectories,and adult disease risk in offspring.IntroductionMeta
5、bolic disease arises at the confluence of genetics and envi-ronmental factors,such as nutrition,exercise,chemical expo-sure,behavior,and the microbiome(Baccarelli and Bollati,2009;Barre s and Zierath,2016;Gallou-Kabani and Junien,2005;McGowan et al.,2009;Miska and Ferguson-Smith,2016;Skinner et al.,
6、2010;Song et al.,2013).Interestingly,studies in both humans and animal models indicate that environ-mental exposures experienced by parents during either intra-uterine or postnatal life can also influence the health of theiroffspring,thus initiating a cycle of disease risk across genera-tions.In thi
7、s review,we briefly discuss potential epigeneticmechanisms that may contribute to environmentally inducedmetabolic disease after direct exposure and in subsequent gen-erations,witha particular focus onpaternal lineage effects medi-ated by germ cells.Evidence for Non-genetic Impact on Metabolic Disea
8、seRisk in Subsequent GenerationsMetabolic disease,including obesity and type 2 diabetes(T2D),arises at the interface of genetic and environmental factors.While family history has long been recognized as an importantdeterminant,only?5%10%of T2D risk can be attributed togenetic factors(Voight et al.,2
9、010).This so-called missing her-itability(Maher,2008)may result fromenvironmental exposuresshared by family members,including those experienced duringearly life.The concept of developmental origins of adult disease pro-poses that during intrauterine life,the fetus adapts to theenvironment to which i
10、t was exposed,potentially conferringresistance to similar exposures experienced postnatally(Barker,2000).Unfortunately,if the postnatal environment differs fromthe prenatal exposures,such responses may be maladaptive.For example,exposure to undernutrition during pregnancymay program the fetus to hav
11、e more efficient energy meta-bolism,thus reducing weight loss and enhancing survival duringfuture periods of undernutrition(Beauchamp et al.,2015).How-ever,if the postnatal environment is instead marked by nutrientexcess,offspring will be more susceptible to weight gain andobesity.Unfortunate times
12、of humanitarian catastrophes,such as warsand political/economic crises,have provided the possibility tostudy the influence of poor nutrition on human populations.Retrospective studies of human cohorts exposed to undernutri-tion during prenatal and early postnatal life provide evidence tosupport inte
13、rgenerational effects.Studies of the Dutch hungerwinter,aperiodofabruptonsetofseverefamineduringGermanblockade of the Netherlands at the end of World War II(19441945),have been very informative.Poor nutrition experiencedby pregnant mothers during the famine was associated withincreased fat mass,hype
14、rtension,glucose tolerance,and psy-chiatric disorders emerging in their children during adult life(Lumey et al.,2007;Ravelli et al.,1998).Similar findings havebeen demonstrated for populations around the world.TheUkraine famine(19321933)also revealed associations betweenearly gestational nutrient re
15、striction and development of T2D inadulthood(Lumey et al.,2015).Similarly,individuals born imme-diately after periods of famine in Austria(19181919,1938,19461947)(Thurner et al.,2013)or China(19591961)havehigher rates of T2D or hyperglycemia(Li et al.,2010).Scarcityof food during the Biafran famine
16、during the Nigerian Civil War(19671970)and resulting undernutrition during gestation andchildhood has also been associated with impaired glucose toler-ance in adulthood(Hult et al.,2010).Low birth weight,which can be independent of nutritional fac-tors,increases the risk of metabolic disease in adul
17、t life(Barker,2000;Mericq et al.,2017).David Barker was the first to identifyrelationships between birth weight and the development of car-diovascularandothermetabolicdiseaseduringadultlife(Barker,2000).Subsequent studies in distinct populations worldwidehave confirmed these relationships,including
18、the Helsinki birthcohort study(Eriksson,2006)and the China birth weight study(Li et al.,2010).While birth weight is a phenotype easily dis-cerned in epidemiologic studies,birth weight itself is unlikely toplay a causal role in adult disease risk;rather,birth weight is abiomarker arising from reduced
19、 fetal growth,in turn resultingCell Metabolism 25,March 7,2017 2017 Elsevier Inc.559from placental dysfunction,hypoxia,poor nutrition,or otherpregnancy stressors.In a recent review,Mericq summarizedthe extensive body of literature supporting the association be-tween small for gestational age or prem
20、aturity and the subse-quent development of metabolic disease(i.e.,higher glucoselevels,T2D,cardiovascular disease)(Mericq et al.,2017).Theseassociationsarefurthermagnifiedwhenlow-birth-weightindividuals experience postnatal catch-up growth(Dulloo,2009).Conversely,overnutrition experienced during ges
21、tationalso predisposes the fetus to develop disease later in life(Boneyet al.,2005;Dabelea and Crume,2011;Yu et al.,2013)Exposures experienced by parents either prior to or existing atthe time of conception can also impact the health of future gen-erations(Alfaradhi and Ozanne,2011;Patti,2013).For e
22、xample,either maternal or paternal prenatal exposure to famine(Chinesefamine of 19591961)has been associated with increased hy-perglycemia in offspring;these effects are greater when bothparentshadprenatalexposure(Lietal.,2017).Moreover,studiesfrom a population born in the Overkalix region of northe
23、rnSweden from1890to 1920analyzed foodavailability duringspe-cific phases of childhood.Interestingly,overnutrition during thechildhood slow-growth period in males was associated withincreased risk of cardiovascular disease and diabetes in hisgrandsonsa transgenerationally inherited phenotype(Kaatiet
24、al.,2002).Experimental Models of Multigenerational Disease RiskMultiple animal models have demonstrated that early-lifeexposures experienced by the parent can potently impactphenotypes in the developing offspring and in subsequentgenerations,evenwithout furtherenvironmentalstressors(Figure 1).For ex
25、ample,undernutrition experienced by themotherduringgestationcancausenotonlyincreasedadiposity and glucose intolerance/diabetes in her offspring(F1)Figure 1.The Vicious Cycle ofIntergenerational Paternal Disease RiskAdverse intrauterine exposure(1)can impact thedevelopment and implantation of the emb
26、ryo,ultimately increasing disease risk in the offspring.In addition,postnatal environmental factors(2)andmetabolic disease in adult life(3)also contribute toepigenetic changes,including alterations in thegerm cells,both sperm and oocytes.Note thatgerm cells can be affected by both the originaldirect
27、 exposure to the embryo,including exposureto his developing germ cells(*),and metabolicdisease in the adult male,thus impacting subse-quent generations.(Isganaitis et al.,2009;Jimenez-Chillaronet al.,2005)but also increased adiposityand glucose intolerance in the next(F2)generation(Hanafiet al.,2016
28、;Jimenez-Chillaron et al.,2009).We and othershave demonstrated that prenatal andearly postnatal exposures can reducethe number and function of pancreatic is-lets(Jimenez-Chillaron et al.,2005;Weiet al.,2014)and nephrons(Benz andAmann,2010)and can alter reproductiveorgans(Chan et al.,2015)and even st
29、em cells(Woo et al.,2011).Moreover,if an undernutrition insult is sustained,therecan be further propagation of metabolic phenotypes acrossmany generations.This was demonstrated in a recent study inWistar rats subjected to continuous 50%caloric restrictionover 50 generations.Offspring animals had fas
30、ting hyperinsuli-nemia,glucose intolerance,and increased adiposity with aging(Hardikar et al.,2015).Neither metabolic nor epigenetic pheno-types were reversed by restoration of nutrition for two genera-tions,indicatingstrikingdurabilityofancestraleffects.Conversely,overnutrition and obesity in the F
31、0 dam can alsoyield phenotypes in the F2 and F3 generation(Dunn and Bale,2009,2011;Gniuli et al.,2008).Environmental exposure tochemical agents also leads to increased metabolic disease sus-ceptibility in subsequent generations(Baccarelli and Bollati,2009;Holliday,1998;Skinneretal.,2010).Interesting
32、ly,acquiredbehaviors can also be transmitted across multiple generationsthrough non-genetic factors,as recently reviewed(Bohacekand Mansuy,2015).In summary,many influential factors may contribute to non-genetic inheritance of disease risk.Hence,consideration ofpotential mechanisms contributing to su
33、ch phenotypes acrossgenerations requires analysis of several key components:1.Do phenotypes in offspring result from alterations inparental health or behavior existing at the time of concep-tion?Alternatively,can these offspring phenotypes resultfrom parental germ cell exposures,either during prenat
34、alor postnatal life of the parent?Parental health pre-conception,at conception,and post-conception have all been shown to impact offspringhealth.Specifically,nutritional status(both over-andundernutrition),insulin resistance,exercise,chemicalexposure,and behavior in the parent before and during560Ce
35、ll Metabolism 25,March 7,2017Cell MetabolismPerspectivepregnancy can all affect both paternal and offspring health(Anderson et al.,2006;Anway et al.,2006;Chamberset al.,2016;Hanafiet al.,2016;Isganaitis et al.,2014;Jimenez-Chillaron et al.,2009;Murashov et al.,2016;Rodgers et al.,2013,2015;Skinner e
36、t al.,2010;Stanfordet al.,2015).These patterns of parent-to-child transmission of diseaserisk could result from the impact of parental health per seon the pregnancy environment(particularly prominent inthe case of maternal exposures)or on either maternal orpaternal germ cells.Evidence that the isola
37、ted germ cellcan mediate offspring disease was recently described byHuypens and collaborators,who utilized in vitro fertiliza-tion to demonstrate that germ cells harvested from miceexposed to nutritional factors(low-fat diet,normal diet,andhigh-fatdietHFD)areabletotransmitmetabolicphe-notypes to off
38、spring.Moreover,both offspring andparental sex influence the severity of offspring adiposityand glucose intolerance(Huypens et al.,2016).Thus,both parental health as well as direct exposuresexperienced by germ cells may contribute to offspringdevelopmental programming and disease risk.2.Are phenotyp
39、es transmitted via maternal or paternal line-age,or both?Mechanisms contributing to transmission of diseaserisk likely depend on whether phenotypes are trans-mitted through the maternal or paternal lineage,or both(Table 1).Both parents contribute to offspring geneticsas well as shared postnatal envi
40、ronments(e.g.,familydiet,activity,behavior,socioeconomic factors).Uniquematernal factors include the well-recognized effects ofmaternal diabetes or obesity,altered structure/functionof reproductive organs,alterations in vaginal or gut mi-crobiome,mitochondrialDNAinheritance,placentalfunction,or epig
41、enetic phenotypes in female germ cells.By contrast,unique paternally mediated effects onoffspring implicate indirect effects on the fetal compo-nent of the placenta,seminal fluid proteins,or spermepigenetic mechanisms(refer to Table 1 for detailedcitations).Due to space restriction,we will only brie
42、flyreview maternal effects on intergenerational diseasetransmission,and focus on paternal lineage mediatedphenotypes in greater detail;for deeper informationon the maternal lineage,please refer to other reviewsTable 1.Parental Contribution for Metabolic Disease ProgrammingMaternal LineagePaternal Li
43、neageNuclear DNA inheritanceYesaYesbMitochondrial DNA inheritanceYesc,dNoeEpigenetic modificationGerm cellsYesf,gYesh,i,j,k,l,m,n,o,p,q,r,s,t,u,v,w,x,y,z,aa,ab,acMaternal environment during pregnancyHormonesYesadNot applicableNutrientsYesae,af,ag,ah,ai,aj,ak,al,am,an,aoMetabolismYesap,aq,ar,asUterin
44、e structure/functionYesat,au,av,awBehaviorYesay,aoChemical exposureYesba,bbMilk compositionYesbcPlacental structure/functionYesbd,ay,be,bfYes(paternal contribution to expression of placentaldevelopment genes)bg,bhShared postnatal environmentDietYesbi,bjYesbjBehaviorYesbkYesbkEnvironmental chemicalsY
45、esblYesblActivityYesbmYesbmMicrobiomeYesbnYesbnaFerguson-Smith et al.,2006;bSchagdarsurengin and Steger,2016;cSaben et al.,2016;dWu et al.,2015;eWang et al.,2016;fBranco et al.,2016;gHuypens et al.,2016;hChen et al.,2016;iDaxinger et al.,2016;jde Castro Barbosa et al.,2015;kDenham et al.,2015;lDonki
46、n et al.,2016;mHammoudet al.,2010;nHuypens et al.,2016;oJodar et al.,2013;pLane et al.,2014a;qMcPherson et al.,2015;rOkada et al.,2010;sOst et al.,2014;tPalmer et al.,2012;uPeng et al.,2012;vRadford et al.,2014;wRodgers et al.,2015;xSchuster et al.,2016;ySharma et al.,2016;zShea et al.,2015;aaSiklen
47、ka et al.,2015;abTerashima et al.,2015;acvan de Werken et al.,2014;adSeckl and Meaney,2004;aeAlfaradhi and Ozanne,2011;afBenz and Amann,2010;agCarone et al.,2010;ahChambers et al.,2016;aiDunn and Bale,2009;ajDunn and Bale,2011;akJimenez-Chillaron et al.,2005;alJimenez-Chillaronet al.,2009;amMcCurdy
48、et al.,2009;anONeill et al.,2014;aoStanford et al.,2015;apDabelea and Crume,2011;aqGue nard et al.,2013;arIsganaitiset al.,2014;asYu et al.,2013;atBromfield et al.,2014;auChan et al.,2015;avIba n ez et al.,2011;awvan Dijk et al.,2015;baHolliday,1998;bbSkinneretal,2010;bcSilvermanet al.,1992;bdGatfor
49、d etal.,2010;ayHowerton et al.,2013;beMacLennanet al.,2004;bfNelissenet al.,2011;bgFerguson-Smithetal.,2006;bhWang etal.,2013;biJimenez-Chillaron etal.,2006;bjSavage etal.,2007;bkMcGowan etal.,2009;blBaccarelli and Bollati,2009;bmBarre sand Zierath,2016;bnSong et al.,2013.Cell Metabolism 25,March 7,
50、2017561Cell MetabolismPerspective(Alfaradhi and Ozanne,2011;Clarke and Vieux,2015;Ferguson-SmithandPatti,2011;RandoandSim-mons,2015).3.Is the inheritance pattern intergenerational or transgenera-tional?Understanding the distinction between these two terms isimportant,as they reflect differences in p