Exercise training reverses cardiac aging phenotypes associated with heart failure with preserved ejection fraction in male mice

Abstract

Heart failure with preserved ejection fraction (HFpEF) is the most common type of HF in older adults. Although no pharmacological therapy has yet improved survival in HFpEF, exercise training (ExT) has emerged as the most effective intervention to improving functional outcomes in this age‐related disease. The molecular mechanisms by which ExT induces its beneficial effects in HFpEF, however, remain largely unknown. Given the strong association between aging and HFpEF, we hypothesized that ExT might reverse cardiac aging phenotypes that contribute to HFpEF pathophysiology and additionally provide a platform for novel mechanistic and therapeutic discovery. Here, we show that aged (24–30 months) C57BL/6 male mice recapitulate many of the hallmark features of HFpEF, including preserved left ventricular ejection fraction, subclinical systolic dysfunction, diastolic dysfunction, impaired cardiac reserves, exercise intolerance, and pathologic cardiac hypertrophy. Similar to older humans, ExT in old mice improved exercise capacity, diastolic function, and contractile reserves, while reducing pulmonary congestion. Interestingly, RNAseq of explanted hearts showed that ExT did not significantly modulate biological pathways targeted by conventional HF medications. However, it reversed multiple age‐related pathways, including the global downregulation of cell cycle pathways seen in aged hearts, which was associated with increased capillary density, but no effects on cardiac mass or fibrosis. Taken together, these data demonstrate that the aged C57BL/6 male mouse is a valuable model for studying the role of aging biology in HFpEF pathophysiology, and provide a molecular framework for how ExT potentially reverses cardiac aging phenotypes in HFpEF.

Full text

Aging Cell. 2020;19:e13159.  
|
1 of 12
https://doi.org/10.1111/acel.13159
wileyonlinelibrary.com/journal/acel
Received:5December2019
|
Revised:26February2020
|
Accepted:12April2020
DOI: 10.1111/acel.13159
ORIGINAL ARTICLE
Exercise training reverses cardiac aging phenotypes associated
with heart failure with preserved ejection fraction in male mice
Jason D. Roh1 | Nicholas Houstis1 | Andy Yu1 | Bliss Chang1 | Ashish Yeri1 |
Haobo Li1 | Ryan Hobson1 | Carolin Lerchenmüller2 | Ana Vujic3 |
Vinita Chaudhari1 | Federico Damilano1 | Colin Platt1 | Daniel Zlotoff1 |
Richard T. Lee3 | Ravi Shah1 | Michael Jerosch-Herold4 | Anthony Rosenzweig1
Thisisanop enaccessarti cleundertheter msoftheCreativeCommonsAttributionL icense,whichpe rmitsuse,dis tribu tionandreprod uctioninanymed ium,
provide d the original wor k is properly cited.
©2020TheAuthors.Aging Cellpublis hedbyAnatomicalSocietyandJohnWiley&SonsLtd
1CorriganMinehanHear tCenter,
Massachuset tsGen eralHospita l,Har vard
MedicalSchoo l,Boston,MA ,USA
2Depar tmentofCardiology,Angiolog y,
andPulmonolog y,UniversityHo spital
Heidelberg,Heidelberg,Germany
3Depar tmentofStemCellandRegenerative
Biolog y,Harva rdStemCellInstitute,
HarvardUnive rsity,Cambridge,MA ,USA
4Depar tmentofRadiol ogy,Brighamand
Women’sHospital,HarvardMedic alScho ol,
Boston ,MA,USA
Correspondence
JasonD.Roh ,Massachuset tsGe neral
Hospit al,SimchesResearchCe nter,Room
3.186,Bos ton,MA02114,USA.
Email: jroh@mgh.harvard.edu
Funding information
AmericanHeartAssociation,Gr ant/
AwardNumber:16F TF29630016and
16SFRN3172000;NationalI nstituteon
Aging ,Grant /AwardNumb er:AG0 47131,
AG061034andAG064328;National
Heart,Lung ,andBlo odInst itute,G rant/
AwardNumber:HL119230 ,HL122987
andHL135886;Else-Kroner-Fresenius-
StiftungFound ation;Deutsc he
Forschungsgemeinschaft,Grant/Award
Number :DGFLE32571-1;FredandInes
YeattsFundf orInnovativeResearch
Abstract
Heart failurewithpreservedejectionfraction(HFpEF)isthe mostcommontypeof
HFin olderadults. Althoughnopharmacological therapyhasyetimprovedsurvival
inHFpEF,exercisetraining(ExT)hasemergedasthemosteffectiveinterventionto
improving fu nctional outcomes in t his age-related disease. T he molecular mecha-
nismsbywhichExTinduces itsbeneficialeffectsin HFpEF,however,remainlargely
unknown.GiventhestrongassociationbetweenagingandHFpEF,wehypothesized
that ExTmight reverse cardiac agingphenotypes that contribute to HFpEF patho-
physiology and additionally provide a platform for novel mechanistic and therapeutic
discovery.Here,weshowthataged(24–30months)C57BL/6malemicerecapitulate
manyofthehallmarkfeaturesofHFpEF,includingpreservedleftventricularejection
fraction,subclinicalsystolicdysfunction,diastolicdysfunction,impairedcardiacre-
serves,exerciseintolerance,andpathologiccardiachypertrophy.Similartoolderhu-
mans,ExTinoldmiceimprovedexercisecapacity,diastolicfunction,andcontractile
reserves,whilereducingpulmonarycongestion.Interestingly,RNAseqof explanted
heartsshowedthatExTdid notsignificantlymodulatebiologicalpathwaystargeted
byconventionalHFmedications.However,itreversedmultipleage-relatedpathways,
includingtheglobaldownregulationofcellcyclepathwaysseeninagedhearts,which
was associated with increasedcapillary density, but no effects on cardiac mass or
fibrosis.Takentogether,thesedatademonstratethattheagedC57BL/6malemouse
isavaluablemodelforstudyingtheroleofagingbiologyinHFpEFpathophysiology,
andprovide a molecular framework forhowExTpotentially reverses cardiac aging
phenotypesinHFpEF.
KEYWORDS
aging,cardiac,cardiovascular,exercise,heartfailure,RNAsequencing

2 of 12
|
ROH et al .
1 | INTRODUCTION
Heartfailurewithpreservedejectionfraction(HFpEF)isacomplex,
heterogenous clinical syndrome strongly associated with advanced
age (Upadhya, Taffet,Cheng, & Kit zman, 2015). It nowrepresents
themost commonform of HF in olderadults withagrowing prev-
alence largelyattributed to globalpopulation aging (Dunlay,Roger,
& Redfiel d, 2017). Unfortunately, prognosis for older adults wit h
HFpEFremainspoor.HFremainstheleadingcauseofhospitalization
amongstpersonsover65yearsold,andnearly1/3ofallolderadults
hospit alized for HF are eith er readmitte d or dead within 9 0 days
of discharge (Upadhya et al., 2015). Not ably, no pharmacological
agent, includingneurohormonalant agonistsandnitrate derivatives
(Borlauget al.,2018;Massieetal.,2008;Pitt etal., 2014;Redfield
etal., 2015), hasimproved survival in HFpEF,making itone of the
largest unmet needs in geriatric and cardiovascular medicine (Parikh
etal.,2018).
The reasons we lack effective pharmacologic al inter ventions
in HFpEF are multifold, but largely stem from an incomplete
unders tanding of the u nderlying me chanisms that dr ive HFpEF
pathophysiology (Roh, Houstis, & Rosenzweig, 2017). In older
adults, molecular,struc tural,an dfunc tionalchanges associated
withcardiacaginghavelongbeenhypothesizedtobemajorcon-
tributorstoHFpEF(Roh, Rhee,Chaudhari, &Rosenzweig, 2016;
Strait&Lakatt a,2013;Upadhyaetal.,2015). However,whether
the biology of cardiac agingcan be targeted forHFpEF therapy
is unclear.
Despite limitedsuccess of currentpharmacologicalagents,aer-
obic exercisetraining(ExT) has emerged as one of themost effec-
tive strategies for improving functional outcomes in older adults
with HFpEF (Edelmann etal., 2011; Kit zman et al., 2016; Kit zman,
Brubaker,Morgan, Stewart,&Little, 2010; O’Connor etal., 2009).
Whether ExT can alter cardiac aging phenot ypes that contribute
to HFpEF patho physiology, however, is contr oversial. Whil e some
studies have suggested that ExT improves diastolic func tion and
cardiac r eserves i n older HFpEF pati ents, othe rs have shown that
ExThasminimalef fect sonthesecardiacagingphenotypes(Angadi
etal.,2015;Edelmannetal.,2011;Haykowskyetal.,2012;Kitzman
etal.,2010,2016 ;N ol teetal.,2014).Mo re ov er,themo le cul ar mech-
anisms by whi ch ExTp otentially imp roves cardiac pe rformance i n
HFpEFareunknown.
This study addresses these critical issues by first demonstrat-
ingthattheagedC57BL/6malemouseisparticularlywellsuited
forstudyingtheroleofagingbiologyinHFpEFpathophysiology.
Usingthisage-relatedHFpEFmodel,wethenshowthatExTpar-
tiallyreversesmany,butnotall,ofthecardiacagingphenotypes
associatedwith HFpEF.Finally, combining RNAseq profiling and
ExT in an integrate d platform for therapeut ic target discover y
provides a scientific rationale for why previous drug targets may
havefailedinclinical HFpEF trials and implicatesalternative bi-
ological pathways as candidates for therapeutic inter vention in
age-relatedHFpEF.
2 | RESULTS
2.1 | Cardiac functional HFpEF phenotyping
Since advanced age represent s one of the dominant risk factors
for HFpEF, we hypothesized that old mice might share hallmark
phenot ypes prese nt in human HFpEF. Toev aluate the aged m ouse
as a HFpEF model, we firstset criteria based on the most common
pathophysiologic features seen in clinical HFpEF (Borlaug, 2014;
Mohammedetal.,2015),whichincludedthefollowing:(a)preserved
left ventricular(LV)systolic function,measured by ejec tion fraction
(EF) or fr actional sh ortening (FS ); (b) impaired exer cise capacit y; (c)
impairedcontractile orchronotropicreserves; (d)increasedintracar-
diac filling pressures, B-type natriuretic peptide (BNP) expression,
orpulmonary congestion;and (e)histologicfeaturesconsistentwith
pathologiccardiachypertrophy.Usingtheseprespecifiedcriteria,we
performed comprehensivephenoty pingin young (3–4months), old
(24–26mo nths), and very ol d (28–30 months) C57BL/6 male m ice.
TwostagesofadvancedagewereusedtodeterminewhetherHFpEF
phenot ypes progressed in the late stages of th e murine lifespan to fur-
therinvestigatetheroleofagingbiologyinHFpEFpathophysiology.
Restingcardiacfunctionalphenotypeswereassessedusinganex-
tensive multimodality approach. Transthoracic echocardiography per-
formed in a large cohort of animals (n=43)foundthatLVfractional
shor ten in gw asge ner al lypre ser ve dinoldand ve r yo ld mi ce ,c ompar ed
toyoung mice(Figure1a).Importantly,this definingfeatureof HFpEF
was further validated in smaller subgroups using cardiac magnetic res-
onance imaging (Figure S1) and invasive intracardiac hemodynamic
tes ting(Fig ureS2).SimilartohumanHFpEF,botholda ndver yoldmice
displayed evidence of subclinicalLVsystolic dysfunctionas reflected
inreducedsystolicstrain(Figure1b).Additionally,impairedmyocardial
relaxation,amarkerofdiastolicdysfunctioncommonlyseeninHFpEF,
wasseeninagedmice(Figure1c,FigureS2).
2.2 | Exercise HFpEF phenotyping
The most c onsistent fun ctional im pairment se en in clinical H FpEF
is exercise intolerance (Borlaug, 2014). Using stress echocardi-
ography, we found that exercise capacity was markedly reduced
in old and ver y old mice, even when adjusted for body weight
(Figure2a,FigureS3).SimilartohumanagingandHFpEF(DelBuono
etal.,2019;Fleg et al., 20 05;Strait & Lakatta,2013), therewasan
age-associateddecline in exercisecapacity that fur ther progressed
from 24 to 30 months (Figure 2a). Bothchronotropic andcontrac-
tilecardiacreservesdecreasedwithage,althoughonlychronotropic
reser ves continued to de cline from 24 to 30 months (F igure 2b).
Bothchronotropicandcontractilereservescorrelatedwithexercise
capacity(Figure2c), suggestingthattheimpairments in cardiac re-
servesseen in oldermice likelycontributetotheir age-relatedde-
clineinexercisecapacity.Notably,thesedataarenotonlythefirstto
demonstrateamarkedage-relateddecrementincardiacreservesin

|
3 of 12
ROH et al .
C57BL/6miceinthecontextofexercise,butalsosupporttheuseof
thismodelforstudyingthepathophysiologyofage-relatedHFpEF.
2.3 | Histopathologic cardiac HFpEF phenotyping
Inadditiontothecardiacfunctionalimpairmentsandexerciseintoler-
anceseeninHFpEF,postmortemhistopathologiccharacterizationof
heartsfrom HFpEFpatientshasrevealedacommonpathologiccar-
diac hypertrophy phenotype that includes increased cardiomyocyte
size,fibrosis,andmicrovascularrarefaction(Mohammedetal.,2015).
Todetermine whether aged C57BL/6 male mice exhibit these his-
topathologic HFpEF phenotypes, we first performed gravimetric
analyseson young,old,andvery old mice. Indexedlungweights,an
indicator of pulmonary congestion, increased with age (Figure S4),
sugges ting that as the se mice age, a progr essive HF syn drome oc-
curs that parallels the age-related decline in functional exercise ca-
pacity(Figure2).Althoughasignificantincreaseincardiacmasswas
seenbetween4and26months,therewasnofur ther increaseaf ter
26months(FigureS4).Thus,wefocusedsubsequentanalysesonthe
young and old age groups. Old mice fully recapitulated the patho-
logic cardiac hypertrophy seen in human HFpEF,demonstrating in-
creased cardiomyocyte size, myocardial fibrosis, and microvascular
rarefaction(Figure 3a),whichwasassociatedwithincreasedcardiac
BNPexpression,abiomarker ofincreasedmyocardialstress and HF
(Figure 3b). N otably, blood pres sures were similar b etween young
and old mice (Figure 3c), suggesting that the pathologic cardiac re-
modeling seen in aged mice is not driven by overt hypertension but
more likely related to processes intrinsic to cardiovascular aging.
2.4 | Reversal of HFpEF phenotypes in aged
C57BL/6 male mice with exercise training
Since aged C 57BL/6 male mice recap itulate many of the H FpEF
phenotypesobservedinolderhumans,wenextexaminedwhether
FIGURE 1 Age-relatedchangesinrestingcardiacfunctioninC57BL/6malemicearesimilartocardiacfunctionalphenotypesinhuman
HFpEF.NonsedatedtransthoracicechocardiographyinC57BL/6malemiceat3–4months(young( Y),n=12),24–26months(old(O),n=17),
and28–30months(veryold(VO),n=14).(a)Fractionalshortening,(b)systolicstrain,(c)earlydiastolicstrainrate(SR).Datashownas
mean ± SEM,withallindividualdatapointsplotted.One-wayANOVAwithposthocTukey'stestusedforanalyses.*p<.05,**p<.01,***p < .001
FIGURE 2 Progressiveage-relateddeclineinexercisecapacit yandcardiacreservesinC57BL/6malemicerecapitulatesexercise
intolerancephenotypesinhumanHFpEF.StressechocardiographytestinginC57BL /6malemiceat3–4months(Y,n=12),24–26months
(O,n=17),and28–30months(VO,n=13).(a)Exercisecapacitymeasuredbytotaldistancerunandtotalworkachieved(adjustedfor
bodyweight).(b)Chronotropicandcontractilereservesmeasuredatpeakexercise.(c)Pearsoncorrelationofexercisecapacit y(work)with
chronotropicorcontractilereserves.Inpanelsaandb,datashownasmean±SEM,withallindividualdatapointsplotted,andone-way
ANOVAwithposthocTukey'stestusedforanalyses.*p<.05,**p<.01,***p < .001

4 of 12
|
ROH et al .
ExT could effectively reverse these age-related HFpEF phe-
notyp es. Mice were match ed based on body w eight and HFpEF
phenotypes (Table 1) and then divided into either an 8-week
modera te-intensity trea dmill running prot ocol (45 min at 10 m/
minat10° incline)versus no intervention(normalsedentary life-
style)(Figure 4a). ExTinducedmultiplefunctional improvements,
someofwhichhavebeenreportedinolderHFpEFpatients(Angadi
et a l.,2 015 ; Ede l m a nne t al., 2 011;H a y ko w s k yet a l .,2 0 12; K i t z man 
etal., 2010, 2016; Nolteet al., 2014). Specifically,improvements
inexercise capacity,systolic strain, diastolic function,contractile
reserves, and pulmonary congestion were seenaf tereight weeks
ofExT(Figure 4b–e, FigureS5).Although cardiacmass and fibro-
siswerenotsignificantlychanged,capillarydensityincreasedwith
ExT(Figure4f–h).
2.5 | Exercise-mediated transcriptome changes
in the aged heart
Toexplorethe molecular pathwaysthroughwhich ExTmodulatescar-
diacaging phenotypesassociatedwithHFpEF,we per formedRNAseq
ona subsetof cardiacsamples fromtheExTandsedentary aged mice.
Fourteengenes (11of whichareknown) were differentially expressed
after adjusting for multiple hypothesis testing (Figure 5b, Table S1).
Although RNAseq revealed only a small number of genes in the aged
heart that were differentially regulated by ExT, gene setenrichment
analysis did identify 479 significantly upregulated and 77 downregu-
lated biological pa thways (Tables S2 and S3). Inte restingly, pathways
associated with drug targets previously tested in clinical HFpEF trials,
that is, adrenergic, renin–angiotensin–aldosterone (RAAS) and nitric
oxide-cGMP-phosphodiesterase signaling pathways, were generally
not significantly altered by ExT, although positive regulation of nitric
oxide synt hase biosynthe sis reached our sign ificance thres hold (NES
1.67;FDR0.24)andcellularresponsestonitricoxide(NO)hadapositive
trend(NES1.64,FDR 0.26)(Table2).Ratherthepathwaysmosthighly
upregulated by ExTwere predominantly cell cycle-related processes,
while the most highly downregulated were related to cellular respiration
(TablesS2and S3). Takentogether,thesedatasuggest that the cardiac
benefitsofExTseeninolderanimalswithHFpEFpathophysiologymay
bedriven more bychangesin thesealternativebiologicalpathways, as
opposed to the neurohormonal pathways that are targeted with current
HFmedications.
FIGURE 3 Pathologic cardiac
hypertrophyinagedC57BL/6malemice
recapitulates histopathologic cardiac
phenotypesinhumanHFpEF.(a)Cardiac
histopathologicassessmentsof4-month
(Y,n=6–12)and25-to26-month(O,
n=6–9)mice,includingrepresentative
photomicrographsandquantificationof
cardiomyocytecross-sectionalarea(CM
CSA),myocardialfibrosis,andcapillary
density.Scalebar=100µm.(b)Relative
cardiacmRNAexpressionofBNP.n = 5/
group.(c)Systemicmeanar terialpressure
(MAP)inY(n=11)versusO(n=8)mice.
Forallpanels,datashownasmean±SEM,
with all individual data points plot ted.
UnpairedStudent'st test used for
analyses.*p<.05,**p<.01,***p < .0 01

|
5 of 12
ROH et al .
2.6 | Reversal of cardiac aging pathways by
exercise training
TodeterminewhetherExTpotentiallyreversescardiacagingbiol-
ogyassociatedwithHFpEF,wenextassessedhowthecardiactran-
scriptome changes with normal aging to provide a comparison to the
ExT-inducedtranscriptomeseenintheagedheart.Notsurprisingly,
RN A seq ana lys e si d ent ifi eda lar gen umb ero fg ene sth atw ere dif f er-
entiallyexpressedin30-month-oldcomparedto4-month-oldhearts
(Figure 5a,TableS4).Gene set enrichmentanalysisconfirmedthat
biological processes implicated in cardiac aging were changing in the
expec ted directi ons in our aged mice (B ergmann et al ., 2009; Dai
etal.,2009;Eisenbergetal.,2016;Hulsmansetal.,2018;Lee,Alison,
Brand,Weindruch,&Prolla,2002; Roh etal., 2016). Inflammation,
cytokineproduction,complement activation,andextracellularma-
trixproductionpathwayswereupregulatedintheagedhear t,while
cellcycle, DNA repair,mitochondrial function, oxidativephospho-
rylation,fatt yacidmetabolism,cardiaccontractilityand relaxation,
vasculogenesis,autophagy,andproteasomepathwaysweredown-
regulate d (Tables S5 and S6). Notab ly,t he generalized i ncrease in
chronic inflammatory processes of the innate immune system along
with downregulation of pathways relevant to cardiac muscle me-
chanics and vascular growth highlights the parallels between cardiac
agingbiology andleadingHFpEFhypotheses,whichhaveproposed
a central role for microvascular inflammation inducing the cardio-
myocytedysfunctionseeninHFpEF(Paulus&Tschope,2013).
Comparativeanalysisbetweentheaging(4moversus30mo)and
ExT(30 mo sedentary vs. 30moExT)cohorts identified 216 path-
ways,whichwereregulatedinopposingdirectionsbyagingandExT
(TablesS2,S3,S5 and S6).Ofthese 216pathways, the most highly
significant changes predominantly occurred in cell cycle or cell di-
vision pathways (Figure5c), suggesting thatreversingimpairments
in this hallmark of aging (Lopez-Otin, Blasco, Partridge, Serrano,
& Kroeme r,2013) may b e an import ant contribut or by which ExT
improves the performance of the aged heart. The main drivers of
thethreemosthighlyupregulatedpathways(cellcycleprocess,cell
cycle, mitoticcellcycle)wereHAUS8, GADD45A, MAPRES2, MCM5,
and FANCI. In creased exp ression tre nds of these gen es were vali-
dated by QPCR in an independent cohort of old male mice that
underwent eight weeks of voluntary wheel running (Figure S6a),
sug gestingthatdifferentmo de sofaerob icExTcaninducesim ilarbi-
ologicaleffectsintheagedheart.Althoughthechangeswerenotas
robustasthecellcyclechangesnoted above,ExTalsoreversedthe
downregulation of ubiquitin–proteasome,cellular st ressre sponse,
heatshockproteinbinding,andfat tyacidmetabolismpathwaysas-
sociatedwithaging( TablesS2,S3,S5andS6).
3 | DISCUSSION
HFp EF is ac li ni c al sy ndr om ew ithhigh mo rbidi tyand mo r talit y,mo st 
commonlyseenin olderadults (Dunlayet al., 2017). Given the lack
of effec tive pharm acologic al therapi es for this dis ease, along w ith
its projected future growth with ongoingpopulation aging, HFpEF
has been labeled as one of the largest unmet needs in cardiovascular
medicine(Parikh et al.,2018). The reason for the lack of effective
therapiesinHFpEFismultifold ,butlargelyduetoanincompleteun-
derstandingofitscomplexpathophysiolog y.
Thisstudyaddressesoneofthemajorshortcomingsinthisfield,
which is the limited number of animal models to identify and study
causal molecular mechanisms in HFpEF (Roh et al., 2017). Previous
studieshave suggested that although aged mice exhibit some fea-
turesincommonwithHFpEF,theydonotnecessarilymodelHFpEF
(Daiet al.,2009;Eisenberget al., 2016).Themerepresenceofcar-
diacHFpEFphenotypes inmicedoesnotnecessarily equateto the
clinicalsyndromeofHF,whichisinherentlydifficult toascertainin
animals . However, our comprehe nsive funct ional, histol ogical, an d
molecular phenot yping provides strong evidence that the aged
C57BL/6 male mo use capture s many of the major c ardiac pheno -
types that have been implicated as core pathophysiologic mediators
ofHFpEF(Borlaug,2014).Importantly, the HFpEF phenotypes ob-
serve d in aged C57BL/6 male mic e occur in the abse nce of overt
hypertension, which has been a common adjunct intervention
used to generateHFpEF phenotypes in animal models(Eisenberg
etal.,2016;Hulsmans et al., 2018; Schiattarellaetal., 2019).Thus,
weproposethattheagedC57BL/6malemousenotonlyrepresents
TABLE 1 Baselinecharacteristicsof28-month-oldC57BL/6
miceincludedintheexercisetrainingsubstudy
Characteristic
Sedentary
(n = 7)
Exercise
(n = 7)
p
value
Age(month) 28 28 n.a.
Sex male male n.a.
Strain C 57 BL /6 C57B L/6 n.a.
Weight(g) 31.6±2.9 34.8 ± 3.5 .09
Resting cardiac function
Fractional
shortening(%)
52.9 ± 3.2 50.2 ± 2. 8 .12
Exercisecapacity
Distance(m) 131.5±26.3 136.2±19.4 .71
Work(Joules) 7.0±1.5 8.0±0.7 .16
Lact ateatpeak
exercise(mM)
8.2±1.6 9.0 ± 1.1 .28
Cardiac reserves
Chronotropic,HR
atpeakexercise
(bpm)
665.5±17.5 664.3±46.8 .95
Contractile,ΔFS
atpeakexercise
(%)
0.2 ± 8.4 2.2 ± 5.0 .59
Note: Priortoinitiationoftheexercisetrainingprotocol,nosignificant
baselinedifferencesinbodyweight,res tingcardiacfunction,exercise
capacity,orcardiacreservesweredetec tedbet weenthesedentary
andexercise-trainedgroups.n=7/group.Datashownasmean±SEM.
UnpairedStudent'st test used for analyses. p < .05 considered
significant.

6 of 12
|
ROH et al .
a valuabl e and comple mentar y model of HFpEF bu t is partic ularly
well suitedfor studying the role ofaging biology inHFpEF patho-
physiolog y.This supportsanemergingparadigm inthisfield, which
suggeststhatgiventheheterogeneityinHFpEF,effectivetherapeu-
tic discover y and implementation will likely need to focus on specific
subgroups(e.g.,olderadults)inwhichtheprimarydriversofHFpEF
maydiffer(Shahetal.,2016).
The second major aim of this study was to begin to elucidate
biologicalmechanismsbywhichExTpotentiallyalterscardiacaging
phenotypes thatcontributetoHFpEF inolderadults.AerobicExT
and caloric restriction have been the only interventions to improve
functionalcapacity inolder HFpEFpatient sinrandomizedclinical
trials(Kitzmanetal.,2016).However,duetolimitedaccesstocar-
diac tissue, the molecular mechanisms by which these interven-
tionsinducethesebenefitshaveremainedlargelyunknown.Here,
wefocused specifically on ExTwith the hypothesisthat it would
inducesimilarfunctionalben efitsint heagedC57BL/6malemouse
and that RN Aseq analy ses of cardiac ti ssue would not onl y pro-
videmechanisticinsightsintotheroleof ExTin cardiacaging,but
alsopotentiallyidentifymuch-needednoveltherapeutictargetsfor
HFpEF.Indeed,ourfindingsshowthatalthoughExTonlyimproves
somecardiacHFpEFphenotypesinthis model,itimproves overall
cardiac p erform ance and exerc ise capacit y.We pr opose that t he
functional parallels seenwithExTin thesemiceand humanswith
HFpEF not only p rovide fur ther suppor t for the use of t he aged
C57BL/6 male mo use as a model of age- related HFpEF, but also
for the use of ExT as a platform for therapeuticdiscover y in this
disease.
Importantly,thesearethefirstdatatoassesshowExTmodulates
thetra ns cr ipto meofth ea ge dh ea rtinthecontex toffunct ionalphe-
notyping. Our analyses provide two major mechanistic insight s into
thetherapeutic roleof ExTinage-related HFpEF.First, in the aged
male hear t, ExTdoes not significantly regulatebiologicalpathways
associated with conventional HF drugs, including β-blockers and
RAASinhibitors.Thisisconsistentwiththefailureofpharmacologi-
caltherapiestargetingthesepathwaystoimproveHFpEFoutcomes
inolderadults.Borderlinesignificancetowardenhancedproduction
orcellularresponses to NO was detectedin our pathway analysis,
whichcould suggestthatNObiologymaystillbea promisingther-
apeuti c target for age-re lated HFpEF. However, given the neutr al
result s in randomi zed controlle d trials wit h organic nit rate and in-
organic nitritetherapies(Borlaugetal., 2018;Redfield etal.,2015),
alternat ive approac hes to target ing this biolo gy need to ex plored.
Second,incontrast to its lack of effect on thetranscriptionalpro-
file of neurohormonal pathways, ExT induced a marked reversal in
theglobaldownregulationofcellcycle-relatedpathwaysseeninthe
aged hear t (Figure 5c). T his study was no t designed to det ermine
whichcellpopulationsthissignalpertainsto.However,theincrease
in capill ary densit y and upregu lation of angioge nesis pathways , in
the absen ce of cardiac mas s or fibrosis cha nges, sugges t that en-
dothelialproliferation is likelyinvolved, whichwould be consistent
withpreviousreportsonExT-inducedcardiacangiogenesis(Lemitsu,
FIGURE 4 AerobicexercisetrainingreversessomeHFpEFphenot ypesinagedmalemice.(a)Experimentaldesignofexercisetraining
substudy.Priortostudycompletion,3micediedinthesedentarygroup(SED)and2micediedintheexercise-trainedgroup(ExT).(b)Resting
cardiacfunctionbystrainechocardiography.(c)Exercisecapacity.(d)Chronotropicandcontractilecardiacreserves.(e)Lungweight(LW)
normalizedtotibiallength( TL).(f)Heartweight(HW)normalizedtoTL.(g)%Fibrosis.(h)Capillar ydensity.Forfibrosisandcapillarydensity
measurements,histologicsectionsfromaSEDanimalthatdiedonedaypriortothefinaltimepointfunctionaltestingwereincludedin
the final analyses. Data shown as mean ± SEM,withallindividualdatapointsplotted.UnpairedStudent'sttestusedforanalyses.*p<.05,
**p<.01,***p < .001

|
7 of 12
ROH et al .
Maeda,Jesmin,Otsuki,&Miyauchi,2006)andthetrendsseeninNO
pathways.ItispossiblethatExT-inducedcardiomyogenesismayalso
be contributing to this signal. Recent work from our group has shown
thataerobicExTgeneratesarobustincreaseincardiomyogenesisin
youngmice(Vujicetal.,2018).Althoughthishasyettobestudiedin
oldanimals,wehavefoundthatexercise-mediatedmoleculardrivers
ofcardiomyogenesisnotonlymodulateproliferationprocesses,but
often promote otherpro-survivaland protective growth pathways
incardiomyocy tes(Bostrometal.,2010;Liuetal.,2015).Giventhe
profound cardiomyocyte loss and reduced regenerative capacity
in the aged h eart (Ber gmann et al., 2 009), even smal l increases i n
cardiomyogenesis would likely have substantial impacts on cardiac
function.
Atanindividualgene level,after adjusting formultiple hypoth-
esis test ing, our RNAs eq analyses di d identify 11 cand idates that
were dif ferentially reg ulated by ExT in the aged he art (Figure 5a ,
TableS1).Ofthese11candidates,theincreasesinSORL1 and AC TA 1
expression were fully validated by QPCR in an independent ExT
cohortof old mice(FigureS6b).SORL1encodesforthesortilin-like
receptor 1, a low-density lipid receptor, whose downregulation
has been implicated in age-related Alzheimer disease (Rogaeva
etal., 20 07). Although SORL1 hasyetto be studied in the context
ofcardiacaging,HF,orexercise,givenitsroleinendosomalprotein
recycli ng, it is possib le that its u pregulati on by ExTc ould mitiga te
someoftheimpairedproteostasisseenincardiacagingandHF.The
ExT-inducedupregulationofcardiacAC TA1 expressioninagedmice
wasunexpected.ACTA 1 is a member of the “fetal" gene profile typ-
ically increased in pathological cardiac hypertrophy and downregu-
latedinexercise-inducedphysiologicalhypertrophy(Vega,Konhilas,
Kelly,&Leinwand,2017). However,high-intensityExTcan increase
AC TA1  expre ssion in the he art (Cast ro et al., 2013). It is pla usible
thateventhoughourExTprotocolwas initiallygradedasmoderate
intensity,itbecameprogressivelymorestrenuousfortheoldanimals
astheyagedover8weeks.Althoughwedidnotdetectasignificant
differenceinc ardiac massin our ExToldmice,average cardiomyo-
cyte size i ncreased by ~1.4-fold, whic h would be consistent w ith
the increased cardiac AC TA1 expressionobservedwithExT.Further
workisneededtodeterminewhetherexerciseintensityhasdiffer-
ential effects on the “fetal gene” profile associated with pathologic
cardiachypertrophy.Importantly,thesedataalsoraisethequestion
FIGURE 5 Reversalofcardiacagingpathwayswithexercisetraining.RNAseqanalysesoncardiactissuefromyoung(4months)
sedentary,veryold(30months)sedent ary,andveryold(30months)exercise-trained(ExT)C57BL/6malemice.ExTmicecompletedan
8-weektreadmillrunningprotocolfrom28to3 0monthspriortosacrifice.n=3/group.(a)Volcanoplotofdifferentiallyexpressedgenes
betweenyoungandveryoldsedentarymice.(b)VolcanoplotofdifferentiallyexpressedgenesbetweensedentaryandExTveryoldmice.(c)
Mosthighlysignificantpathwaysthataredifferentiallyregulatedinopposingdirectionsbyage(red,youngsedentaryvs.veryoldsedentary)
andExT(blue,veryoldsedentaryvs.veryoldE xT).OnlypathwayswithanFDR<0.05inboththeagingandexercise-trainedcohor tswith
normalizedenrichmentscores(NES)inopposingdirectionsaredisplayed.Forvolcanoplots,black=nonsignificant;orange=log2(FC)≥1;
red = padj < .05; green = log2(FC)≥1+padj < .05

8 of 12
|
ROH et al .
ofwhetherthefetal geneexpressionprofilecanreliablydistinguish
between physiologic and pathologic hypertrophy in older animals
and humans. Evidence in humans has suggested that moderate in-
tensity distance running specifically increases circulating BNP,an-
other member of the pathologic cardiac hypertrophy fetal gene
profile,in older,butnot youngerhumans (Kim et al., 2017). In our
ExToldmice,theimprovementsincardiacfunctionandabsenceof
fibrotic changes suggest that despite an overall upregulation in the
fetal gene expression profile, exercise appears to induce a benefi-
cial effect in the aged murine heart. Lastly,it is important to note
thatourRNAseqanalysesdidnotidentifysignificanttranscriptional
changes in targets thathavebeen previously reportedinExTaged
rodents,suchasSERCA2a,VEGF,andSIRT1(Laietal.,2014;Lemit su
etal.,2006;Tateetal.,1996).However,itislikelythatsomeofthese
TABLE 2 EffectsofexercisetraininginagedheartsonbiologicalpathwaysassociatedwithpreviouslytesteddrugtargetsinHFpEF
Pathway NES FDR
Adrenergic system
Adrenergicreceptoractivity 0.89 0.81
Adrenergicreceptorbinding 0.69 0.94
Adrenergicreceptorsignalingpathway 0.88 0.82
Catecholamine binding 1.01 0.73
Catecholamine biosynthetic process 0.95 0.76
Catecholamine metabolic process 1.33 0.49
Catecholamine transport 1.18 0.60
Negativeregulationofcatecholaminesecretion 1.01 0.73
Positive regulation of catecholamine metabolic process −1. 13 0.82
Regulation of norepinephrine secretion −1. 0 0 0.88
Response to epinephrine −1. 3 8 0.71
Renin–angiotensin–aldosterone system
Angiotensinreceptorbinding −0.85 0.93
Regulationofbloodvolumebyrenin–angiotensin 0.98 0.75
Regulationofsystemicarterialbloodpressurebycirculatingrenin–angiotensin 1.48 0.38
Regulationofsystemicarterialbloodpressurebyrenin–angiotensin 1.45 0.40
Response to mineralocorticoid −0.86 0.93
Nitric oxide-cGMP-phosphodiesterase system
3’5’cGMPphosphodiesteraseactivity −1.56 0.57
Cellularresponsetonitricoxide 1.64 0.26
cGMP binding −1. 13 0.82
cGMP biosynthetic process 1.10 0.66
cGMP met abolic process 1.08 0.68
Negativeregulationofnitricoxidemetabolicprocess 0.95 0 .76
Nitricoxidemediatedsignaltransduction −1.4 5 0. 67
Nitricoxidemetabolicprocess 1. 29 0.52
Nitricoxidesynthasebinding −1.08 0.84
Positiveregulationofnitricoxidesynthaseactivity −0.96 0.90
Positiveregulationofnitricoxidesynthasebiosyntheticprocess 1.67 0. 24
Regulation of cGMP biosynthetic process −1. 0 3 0.86
Regulation of cGMP metabolic process 1.08 0.68
Regulationofnitricoxidebiosyntheticprocess −0.86 0.93
Regulationofnitricoxidesynthaseactivity −0.77 0.97
Regulationofnitricoxidesynthasebiosyntheticprocess 1.42 0.42
Responsetonitricoxide 1.50 0.36
Note: GenesetenrichmentanalysisofRNAseqprofilesfromcardiactissueofsedentaryver susexercise-trainedagedmalemiceusingtheGene
Ontolog y pathway dataset. N=3/group.NES=normalizedenrichmentscore.Falsediscover yrate(FDR )<0.25consideredsignificant.

|
9 of 12
ROH et al .
targets,suchasSERCA2a,arelargelyregulatedatapost-transcrip-
tionallevelintheagedheart(Rohetal.,2019).
Some limitations of the study warrant emphasis. First, this
study wa s done exclusive ly in male mice and , thus, does no t ad-
dresssex-relateddifferences inage-relatedHFpEF.Evidencesug-
gests that there are likely molecular differences in how male and
femaleheartsage,andmoreover,howtheyremodelinresponseto
physiologicandpathologicstress(Konhilaset al., 200 4;Piro,Della
Bona,Abbate,Biasu cci,&Crea,2010;Weinbergetal.,1999).While
our findings strongly suggestthat theaged C57BL /6male mouse
recapitulatesmanyoftheclinicalHFpEFphenotypes,furtherwork
needstobedonetodeterminewhetherHFpEFphenotypesarealso
presentinagedfemalemice,andifcellcyclepathwaysaresimilarly
modulatedbyageandexerciseinfemales.Second,althoughwein-
cluded ma ny of the core path ophysiologic f eatures in ou r HFpEF
phenotyping,thisdid not include assessmentofotherpotentially
causal comorbidities, such as obesity, or the role of peripheral
mechanisms in HFpEF pathophysiology (Borlaug, 2014; Kitzman
etal ., 20 16).Nonc a rd iacph enotyp eslikel yc on tri butetot he age-re-
latedexerciseintoleranceseeninthismodelandneedtobefurther
investigated.Third, while nodifferences insystemicarterialpres-
sureweredetectedbetweenyoungandoldmice,wedidnotassess
forchangesinaortic stiffness,whichincreases with ageandcould
also be contributing to the pathologic cardiac hypertrophy pheno-
type seen in this aged mouse model (Fleenor et al., 2014). Lastly,
alteredpathwayswereidentifiedusingRNAseqonwholeheartex-
tracts.Thus,whilethisisthefirstExT-relatedtranscriptomeanaly-
se si na ge dhe arts ,fu tu res tud ie swill ne e dt od ef i ne thesp eci ficce ll 
populations driving the functional and molecular changes induced
byExTintheagedheart.Moreover,RNAseqdoesnotidentifyreg-
ulationthatoccursatthelevelofproteinexpressionorpost-trans-
lational modifications of protein.
Takento gether, this stu dy address es some of the ma jor short-
comings in HFpEF research, particularly in the context of aging.
It esta blishes the age d C57BL/6 male mouse as a v aluable mod el
for stud ying the role of agin g biology in HFpEF pat hophysiology.
Moreover, by using E xTas a pl atform for t herapeu tic discover y,it
identifiesmultiplepathwaysimplicatedincardiacaging,mostnota-
blyimpairedcellcycle-relatedpathways,thatmaypotentiallyrepre-
sentpromising targets fortherapeuticdevelopmentin age-related
HF pEF.
4 | EXPERIMENTAL PROCEDURES
4.1 | Mice
All animal studies were approved by the Beth Israel Deaconess
Medical Center and Massachusetts General Hospital Institutional
Animal CareandUse Committees.Aged C57BL/6males were gen-
erouslyprovidedbytheNationalInstituteonA ging.AgedC57BL/6
females were unavailable atthetime ofthisstudy.YoungC57BL/6
maleswerepurchasedfromJacksonLaboratory.
4.2 | Echocardiography
Echocardiography was performed on unanesthetized mice with
Vivid 7andE90systems (GEHealthcare). Systolicfunctionwas as-
sessed byfractional shortening and radial systolic strain, while di-
astolic function was assessed by early diastolic strain rate. Refer to
supplemental methods for details on echocardiographic image ac-
quisitionandanalysis.
4.3 | Cardiac magnetic resonance imaging
Mice were ane sthetized wit h isoflurane an d imaged using a 9.8-T
MRIsys tem(BrukerBiospi n).Refertosuppl ement almet ho dsf ord e-
tails of the cardiac MRI protocol.
4.4 | Invasive intracardiac hemodynamics
Mice were ane sthetized w ith isoflura ne and mechan ically venti-
latedthroughouttheprocedure.TheLVwasenteredvia theright
carotid arterywithaScisence1.2Fhigh-fidelitymicromanometer
catheter(TransonicSystems Inc.)to record pressure-volume(PV)
loops. PV loopswereanalyzedoff-linewithLabScribe2 software
(iWorx).
4.5 | Stress echocardiography exercise testing
Tome asure exercis e capacit y and cardia c reser ves, a stres s echo-
cardiography protocolwasdesignedinwhichmice were run to ex-
haustion and then immediately imaged via echocardiography. Refer
to the supplemental methods for details of the protocol.
4.6 | Exercise training protocols
Twoaerobicexercisetrainingprotocolswere used in this study.In
the initial discovery cohort, moderate-intensity treadmill running
was performed five days per week for eight consecutive weeks.
Treadmill running was done on an automated treadmill (Columbus
Instruments) at a constant speed of 10 m/min at 10° incline. To
ensure that the biologicaleffectsof ExTwerenotlimited to aspe-
cifictypeof aerobicexercise, wealsoperformed eight consecutive
weeks of vol untary wh eel running (S TARR Life Science s)i n an in-
dependent validation cohort, usingpreviously published methods
(Vujicetal.,2018).
4.7 | Histologic and immunohistochemical analyses
Formalin-fixed, paraffin-embedded mid-ventricular sections were
stained with periodic acid–Schiff for cardiomyocyte cross-sectional

10 of 12
|
ROH et al .
area (CSA), Masson's trichrome for fibrosis, and rabbit-anti-mouse
CD31(1:50,CellSignalingTechnologies,#77699)forcapillar ydensity.
CardiomyocyteCSAsweremeasuredinthreetofiverandomsections
(40–60 cells/section,~200cells/hear t),which wereaveragedto rep-
resentasingledat apointfore achheart.C apillarydensityw asquanti-
fied by dividing the number of CD31+cellsbythearea of randomly
selected sections. Three to five sections were measured per heart and
averagedtorepresentasingledatapoint.Giventheextensivevariabil-
ity in fibrosis distribution throughoutthe hear t, BZ-XAnalyzer soft-
ware (Keyence)was used to quantify fibrosis in full mid-ventricular
sections. Percent fibrosis was calculated as the ratio of fibrotic area
to total tissue area. Measurements from two sections were averaged
to represent a single data point for each heart. Quantitative histologic
analyses were done in a blinded fashion.
4.8 | Quantitative real-time PCR
Real-time PCR productswere carried out using SYBR-green and
standard amplification protocols. Expression levels were cal-
culated using the ΔΔCt method. Prim er sequences ar e listed in
TableS7.
4.9 | RNA sequencing
RNA sequencing was performed by the MGH Sequencing Core.
LibrarieswereconstructedfrompolyA-selectedRNAusingaNEBNext
Ultra DirectionalRNALibrary Prep Kit(NewEngland Biolabs)andse-
quenced o n Illumina HiSe q2500 inst rument. Th e R package DES eq2
was used fo r differentia l gene expressio n analysis. Ge nes were con-
sidered differentially expressed if upregulated by log2FC>+1 or
downregulated by log2FC<−1 with an adjusted p value < .05 (using
Benjamini-Hochbergcorrection).Pathwayanalysiswasperformedwith
GeneSetEnrichmentAnalysis(GSEA,BroadInstitute)usingtheGene
Ontology database. For all differentially expressed genes, a metric
wascomputedasthe productof logFCand−log10(p-value).A“running
sum” statistic was calculated for each gene set in the pathway dat a-
base based on the ranks of the members of the set relative to those of
thenonmembers.Enrichmentscore(ES)wasdefinedasthemaximum
sum of the ru nning sum wit h the genes ma king up this ma ximum ES
contributingtothe coreenrichmentinthat pathway.Anormalizeden-
richmentscore(NES)wasgeneratedbasedonthegenesetenrichment
scores for all dataset permutations. Pathways with a false discovery
rate(FDR)<0.25wereconsideredsignificant.
4.10 | Statistical analyses
RNAseq and GSE A data analyses were performed with R and
DESeq2software,asdescribedinSection4.9.GraphPadPrism(ver-
sion7.0)was usedforallotherdataanalyses.Inallgraphs,dataare
shown as means ± SEMwithallindividualdatapointsdisplayed.For
comparisons of two groups,unpaired Student's t test s were per-
formed.For comparisons of≥3 groups, one-wayANOVA followed
byposthocTukey'smultiplecomparisontestingwasdone.Pearson
methodwasusedforcorrelationstudiesinFigure2.p value <.05 was
considered statistically significant.
ACKNOWLEDGMENTS
This work was suppor ted by the NIH (AG0 47131, AG061034,
AG064328, HL119230, HL122987, HL135886), AHA
(16SFRN31720000, 16FTF29630016), German Research
Foundation (DFG, LE 3257 1-1), Else-Kroner-Fresenius-Stif tung
Foundation, and the Fred and Ines Yeatts Fund for Innovative
Research.
CONFLICT OF INTEREST
Nonedeclared.
AUTHORS’ CONTRIBUTIONS
JDR,NH,andARdesignedthestudy.JDR,NH,AYu,HL,CL,AV,FD,
DZ,RS,andMJHperformedtheinvivoexperiments.JDR,NH,AYu,
BC,AYe,HL,RH,VC ,CL,AV,CP,FD,DZ,RS,MJH,andRTLassisted
with tissueanalysesand/ordat ainterpretation.JDR and AR wrote
the manuscript with contributions from all authors.
DATA AVAIL AB ILI T Y STAT E MEN T
The data that support the findings of this study are all present
in the paper or the Supplemental Materials. The raw RNA se-
quencingdatausedinthisstudywillbeavailableintheNCBISR A
repository.
ORCID
Jason D. Roh https://orcid.org/0000-0002-6999-6868
Haobo Li https://orcid.org/0000-0002-5660-7835
Ryan Hobson https://orcid.org/0000-0001-8370-8570
REFERENCES
Angadi , S. A., Mook adam, F., Lee, C. D., Tucker, W. J., Haykows ky,
M.J.,& Gaesser,G.A .(2015).High-intensity inter val training vs.
moder ate-intensity co ntinuous exerc ise training in h eart failur e
withpreserved ejec tion fraction:A pilot study.Journal of Applied
Physiolog y, 119, 753–758. ht tps ://doi.org/10.1152/japplphysi
ol.00518.2014
Bergmann,O., Bhardwaj,R .D.,Bernard,S., Zdunek,S., Barnabé-Heider,
F.,Walsh,S.,…Frisén,J.(2009).Evidenceforcardiomyocy terenewal
in humans. Science, 324, 98–102. https://doi.org/10.1126/scien
ce.116 468 0
Borlau g, B. A. (2014). The pat hophysiolog y of heart fail ure with pre-
served ejection fraction. Nature Reviews Cardiology, 11, 507–515.
https://doi.org/10.1038/nrcar dio.2014.83
Borlaug,B.A.,Anstrom,K.J.,Lewis,G.D.,Shah,S.J.,Levine,J.A.,Koepp,
G.A.,…Redfield,M.M.(2018).Effectofinorganicnitritevsplacebo
on exercise capacity among patients with heart failure with pre -
servedejectionfraction:TheINDIE-HFpEFrandomizedclinicaltrial.
JAMA,320,1764–1773.https://doi.org/10.1001/jama.2018.14852
Bostro m, P.,Ma nn, N., Wu, J. , Quintero, P. A., Pl ovie, E. R ., Panakova ,
D., … Spiegel man, B. M. (2 010). C/EBPβ contro ls exercise-in duced

|
11 of 12
ROH et al .
cardiac growth and protects ag ainst pathologic cardiac remodeling.
Cell,143,1072–1083.https://doi.org/10.1016/j.cell.2010.11.036
Castro,V.,Grisdale-Helland, B., Helland, S. J.,Torgersen,J., Kristensen,
T., Claireau x, G., … Takle, H. (2 013). Cardiac m olecular-accli mation
mechanisms in response tosw imming-inducedexercise in Atlantic
salmon. PLoS ONE, 8, e55056. https://doi.org/10.1371/journ
al.pone.0055056
Dai, D.-F., Santana, L . F.,Vermulst, M., Tomazela, D. M., Emond, M. J.,
MacCoss,M.J., …R abinovitch, P.S. (20 09). Overexpressionofc at-
alase targeted to mitochondria attenuates murine cardiac aging.
Circulation, 119, 2789–2797. https://doi.org/10.1161/CIRCULATIO
NAHA.108.822403
Del Buono, M. G., Arena, R., Borlaug, B . A., C arbone, S., Canada, J.
M., Kirkman, D. L ., … Abbate, A. (2019). Exercise intolerance in
patient s with heart failure: JACC state-of-the-art review. Journal
of the American College of Cardiology, 73, 2209–2225. https://doi.
org/10.1016/j.jacc.2019.01.072
Dunlay, S. M. , Roger, V.L ., & Redfield , M. M. (2017). Epi demiolog y of
heart failure with preserved ejection fraction. Nature Reviews
Cardiology,14,591–602.https://doi.org/10.1038/nrcardio. 2017.65
Edelmann, F., Gelbrich, G., Düngen, H .-D., Fröhling, S., Wachter, R.,
Stahrenberg,R ., … Pieske, B.(2011).E xercise trainingimproves ex-
ercise capacit y and diastolic function in patients with heart failure
with pre served eje ction frac tion: Result s of the Ex-DHF (E xercise
traininginDiastolicHeartFailure)pilotstudy.Journal of the American
College of Cardiology, 58, 1780–1791. htt ps://doi.org/10.1016/j.
jacc.2011.06.054
Eisenberg, T., Abdellatif,M., Schroeder,S., Primessnig, U., Stekovic , S.,
Pendl, T.,…Madeo, F.(2016).Cardioprotectionand lifespanexten-
sion by the natural polyamine spermidine. Nature Medicine,22,1428–
1438. https://doi.org/10.1038/nm.4222
Fleenor,B. S., Eng,J. S.,Sindler,A. L.,Pham, B. T.,Kloor,J. D., & Seals ,
D.R.(2014).Superoxidesignalinginperivascularadiposetissuepro-
motes age-related artery stiffness. Aging Cell,13,567–578.https://
doi.org/10.1111/acel.12196
Fleg,J.L.,Morrell,C.H.,Bos,A.G.,Brant ,L.J.,Talbot,L.A.,Wright,J.G.,
&L akatta, E. G.(2005).Accelerated longitudinaldecline of aerobic
capacity in healthy older adults. Circulation, 112 ,674–682.https://
doi.org/10.1161/CIRCU L ATIONA HA .105. 545 459
Haykowsk y, M. J., Brubaker, P. H., Stewart, K. P., Morgan, T. M.,
Eggebeen,J.,& Kit zman, D.W.(2012). Effec tof endurance training
onthedeterminantsofpeakexerciseox ygenconsumptioninelderly
patient s with stable compensated heart failure and preser ved ejec-
tion fraction. Journal of the American College of Cardiology,60, 120–
128.https://doi.org/10.1016/j.jacc.2012.02.055
Hulsmans,M.,Sager,H.B.,Roh,J.D.,Valero-Muñoz,M., Houstis,N.E.,
Iwamoto, Y., … Nah rendorf, M. (2018). C ardiac macro phages pro-
mote diastolic dysfunction. Journal of Experimental Medicine, 215,
423–440.https://doi.org/10.1084/jem.20171274
Kim, J.H., Ko,Y.-A.,Hedley,J.,MacNamara,J.,Awad, M., Taylor,W., …
Baggish,A .L .(2017).Theimpact ofmoderatedistance recreational
running and ageing on cardiac physiology. Heart, 103, 219–226.
htt ps://doi.org/10.113 6/hear tjnl-2016 -310 059
Kitzm an, D. W., Bruba ker,P., Morga n, T., Haykowsk y,M. , Hundley, G.,
Kraus,W.E., … Nicklas, B. J. (2016). Effect of c aloricrestrictionor
aerobic exercise training on peak oxygen consumption and qual-
ity of life in obese older patients with hear t failure with preser ved
ejection fraction. JAMA, 315, 36–46. https://doi.org/10.1001/
jama. 2015.173 46
Kitzman,D.W.,Brubaker,P.H.,Morgan,T.M.,Stewar t,K.P.,&Little,W.
C. (2010). E xercisetraining in older patients with hear t failure and
preser ved ejec tion frac tion: A rand omized, cont rolled, sing le-blind
trial. Circulation: Heart Failure,3,659–667.https://doi.org/10.1161/
CIRCHEARTFAILURE.110.958785
Konhilas,J.P., Maass,A . H., Luckey,S. W., Stauffer,B. L.,Olson,E .N.,
& Leinwan d, L. A . (2004) . Sex modif ies exerci se and card iac adap-
tation in mice. Ame rican Journal of Physiology. Heart and Circulatory
Physiolog y, 287, H2768–H2776. https://doi.org/10.1152/ajphe
art.00292.2004
Lai, C .-H., Ho, T.-J., Kuo, W.-W., Day, C.-H., Pai , P.-Y., Chung , L.-C., …
Huang, C .-Y. (2014). Exercise training enhanced SIRT1 longevity
signalingreplacesthe IGF1 survival pathway to attenuate aging-in-
duced rat heart apoptosis. Age, 36, 9706 . htt ps://doi.o rg/10.10 07/
s 1 1 3 5 7 - 0 1 4 - 9 7 0 6 - 4
Lee, C. K., Alison, D.B.,Brand,J., Weindruch, R.,& Prolla,T.A. (2002).
Transcriptional profiles associated with aging and middle age-on-
set caloric restriction in mouse hearts. Proceedings of the National
Academy of S ciences of the United Sta tes of Am erica,99,14988–14993.
https://doi.org/10.1073/pnas.232308999
Lemitsu, M., Ma eda, S., Jesmin , S., Otsuki, T., & Miy auchi, T. (2006).
Exercis e training i mproves agin g-induced dow nregulatio n of VEGF
angiogen ic signaling cas cade in heart s. American Jo urnal of Physiolog y.
Heart and Circulatory Physiology,291(3), H1290–H1298. ht tps://doi.
org/10.1152/ajphe art.00820.2005
Liu, X.,X iao, J., Zhu,H., Wei, X.,Platt,C., Damilano, F.,…Rosenzweig,
A.(2015).miR-222isnecessaryforexercise-inducedcardiacgrowth
and protects against pathologic cardiac remodeling. Cell Metabolism,
21,584–595.https://doi.org/10.1016/j.cmet.215.02.014
Lopez-Otin, C.,B lasco, M. A ., Partridge,L ., Serr ano, M., & Kroemer, G.
(2013). The hallmarks of aging. Cell, 153, 1194–1217. https://doi.
org/10.1016/j.cell.2013.05.039
Massie, B . M., Carson , P. E., McMur ray, J. J., Komajda, M., M cKelvie,
R., Zile, M. R.,…Ptaszynska, A. (2008).Irbesartan in patients with
heart failure and preserved ejection fraction. New England Journal
of Medicine, 359, 2456–2467. https://doi.org/10.1056/NEJMo
a0805450
Mohammed, S. F., Hussain, S., Mirzoyev, S. A., Edwards, W. D.,
Maleszewski,J.J.,&Redfield,M.M.(2015).Coronarymicrovascular
rarefaction and myocardial fibrosis in heart failure with preserved
ejection fraction. Circulation,131,550–559.https://doi.org/10.1161/
C I R C U L A T I O N A H A . 1 1 4 . 0 0 9 6 2 5
Nolte,K.,Schwarz,S.,Gelbrich,G.,Mensching,S.,Siegmund,F.,Wachter,
R., … Ede lmann, F. (2014). Effect s of long-term e ndurance a nd re-
sistancetrainingondiastolicfunction,exercisecapacit y,andquality
of life in asymptomatic diastolic dysfunction vs. heart failure with
preserved ejection fraction. ESC Heart Failure, 1, 59–74. ht t ps ://doi .
org /10.10 02/ehf2.120 07
O’Connor,C. M., Whellan,D.J., Lee,K.L.,Keteyian, S.J.,Cooper,L.S.,
Ellis, S. J.,…Piña,I. L. (2009). Efficac y and safety of exercise train-
ing in pati ents with c hronic hea rt failu re: HF-ACTIO N randomize d
controlled trial. JAMA, 3 01, 1439–1450. https://doi.org/10.1001/
ja m a.20 0 9. 45 4
Pa ri k h, K .S .,S ha r ma ,K . ,F i uz at, M.,Su rks ,H .K.,Ge o rg e,J.T.,H on arp our, 
N., … Felker, G. M . (2018). Heart failure with preser ved ejection
fraction expert panel report: Current controversies and implica-
tions for clinical trials. JACC: Hear t Failure, 6, 619–632. https://doi.
org/10.1016/j.jchf.2018.06.008
Paulus,W.J., & Tschope, C . (2013). A novel paradigm for heart f ailure
with preserved ejection fraction: Comorbidities drive myocardial
dysfunction and remodeling through coronary microvascular endo-
thelial inflammation. Journal of the American College of Cardiolog y,62,
263–271.https://doi.org/10.1016/j.jacc.2013.02.092
Pir o,M.,De ll aB on a,R. ,Abbate ,A .,Bi as ucci,L.M.,&Crea,F.(2 010 ).Sex-
related differences in myocardial remodeling. Journal of the American
College of Cardiology, 55, 1056–1065. ht tps://doi.o rg/10.1016/j.
jacc.2009.09.065
Pitt,B.,Pfeffer,M.A.,Assmann,S.F.,Boineau,R.,Anand,I.S.,Claggett,
B., … McKi nlay, S. M. (2014). Spiro nolact one for hear t failur e with

12 of 12
|
ROH et al .
preserved ejection fraction. New England Journal of Medicine, 370,
1383–1392.https://doi.org /10.1056/NEJMoa1313731
Redfield, M. M.,Anstrom, K.J., Levine, J. A.,Koepp, G.A., Borlaug,B.
A., Chen, H. H ., … Braunwald, E. (2015).Isosorbide mononitrate in
heart failure with preserved ejection fraction. New England Journal
of Medicine, 373, 2314–3234. https://doi.org/10.1056/NEJMo
r15107 74
Rogaeva,E .,Me ng,Y.,Lee,J.H.,G u,Y.,Kawar ai,T.,Zou,F.,…StGeor ge-
Hyslop, P. (2007 ). The neurona l sortilin -related rece ptor SORL1 is
genetic ally assoc iated with Al zheimer dis ease. Nature Genetics, 39,
168–177.https://doi.org/10.1038/ng1943
Roh, J.D.,Hobson,R .,Chaudhari, V.,Quintero, P.,Yeri, A.,Benson,M.,
…Rosenzweig,A .(2019).ActivintypeIIreceptorsignalingincardiac
aging and heart failure. Science Translational Medicine,11,eaau868 0.
https://doi.org/10.1126/scitranslmed.aau8680
Roh,J.,Houstis,N.,&Rosenz weig,A .(2017).Whydon’twehaveproven
treatmentsforHFpEF?Circulation Research,120,1243–1245.https://
doi .org/10.1161/CIRCRESAHA .116.310119
Roh,J.,Rhee,J., Chaudhari, V.,&Rosenzweig,A .(2016).Theroleof ex-
ercise in cardiac aging: From physiolog y to molecular mechanisms.
Circulation Research, 118, 279–295. https://doi.org/10.1161/CIRCR
ESAHA.115.305250
Schiattarella,G.G .,Alt amirano,F.,Tong,D.,French,K.M.,Villalobos,E.,
Kim, S. Y.,… Hill, J. A. (2019).Nitrosative stre ssdrives heart failure
with preserved ejection fraction. Nature,56 8,351–356.https://doi.
org /10.103 8/s41586 -019-1100-z
Shah, S . J., Kitzm an, D. W., Borla ug, B. A. , van Heere beek, L ., Zile, M .
R.,Kass,D.A.,&Paulus,W.J.(2016).Phenotype-specifictreatment
ofheartfailurewithpreservedejectionfraction:Amultiorganroad-
map. Circulation, 134, 73–90.https://doi.org/10.1161/CIRCULATIO
NAHA.116.021884
Strait, J. B., & Lakatta, E. G. (2013). Aging-associated cardiovascular
changes and their relationship to heart failure. Hear t Failure Clinics,8,
143–164.https://doi.org/10.1016/j.hfc.2011.08.011
Tate,C.A.,Helgason,T.,Hyek,M.F.,McBride,R.P.,Chen,M.,Richardson,
M.A.,&Taffet,G.E.(1996).SERCA2aandmitochondrialcytochrome
oxidase e xpression ar e increased in h earts of exe rcise-traine d old
rats. American Journal of Physiology, 271, H68–H72. https://doi.
org /10.1152/ajpheart .1996. 271.1.H68
Upadhya,B.,Taffet ,G .E .,Cheng,C.P.,& Kit zman, D.W.(2015).Heart
failurewit hpreser ved ejection frac tion in the elderly: Scopeof the
problem. Journal of Molecular and Cellular Cardiology, 83, 73–87.
https://doi.org/10.1016/j.yjmcc.2015.02.025
Vega, R. B., Konhilas, J. P., Kelly, D. P., & Leinwand, L. A. (2017).
Molecular mechanisms underlying cardiac adaptation to exer-
cise. Cell Metabolism, 25, 1012–1026. h ttps://doi.org/10.1016/j.
cmet.2017.04.025
Vujic, A., Lerchenmüller, C., Wu, T.-D., Guillermier, C., Rabolli, C. P.,
Gonzal ez, E., … Rosenz weig, A. (2018). E xercise indu ces new car-
diomyocyte generation in the adult mammalian heart. Nature
Communications, 9, 1659. https ://doi.org/10.1038/s41467-018-
0 4 0 8 3 - 1
Weinberg , E. O., Thienelt , C. D., Katz, S . E., Bartune k, J., Tajima, M.,
Rohrbach,S.,…Lorell,B.H.(1999).Genderdifferencesinmolecular
remodeling in pressure overload hypertrophy. Journal of the Ame rican
College of Cardiology, 34, 264–273. https://doi.org/10.1016/s0735
- 1 0 9 7 ( 9 9 ) 0 0 1 6 5 - 5
SUPPORTING INFORMATION
Additional supporting information may be found online in the
SupportingInformationsection.
How to cite this article:RohJD,HoustisN,YuA,etal.Exercise
training reverses cardiac aging phenotypes associated with
heart failure with preserved ejection fraction in male mice.
Aging Cell. 2020;19:e13159. htt ps://doi.or g/10.1111/
acel.13159