Endurance exercise rescues progeroid aging andinduces systemic mitochondrial rejuvenation inmtDNA mutator miceAdeel Safdara,b,c, Jacqueline M. Bourgeoisd, Daniel I. Ogborne, Jonathan P. Littlea, Bart P. Hettingab,Mahmood Akhtarb, James E. Thompsonf, Simon Melovg, Nicholas J. Mocellinb, Gregory C. Kujothh,Tomas A. Prollah, and Mark A. Tarnopolskyb,c,1

Departments of aKinesiology, bPediatrics, cMedicine, dPathology and Molecular Medicine, and eMedical Sciences, McMaster University, Hamilton, ON, CanadaL8N 3Z5; fDepartments of Medicine and Immunology, Roswell Park Cancer Institute, Buffalo, NY 14263; gBuck Institute for Age Research, Novato, CA 94945;and hDepartments of Genetics and Medical Genetics, University of Wisconsin, Madison, WI 53706

Edited* by Bruce M. Spiegelman, Dana-Farber Cancer Institute/Harvard Medical School, Boston, MA 02115, and approved January 28, 2011 (received forreview December 30, 2010)

A causal role for mitochondrial DNA (mtDNA) mutagenesis inmammalian aging is supported by recent studies demonstratingthat the mtDNA mutator mouse, harboring a defect in theproofreading-exonuclease activity of mitochondrial polymerasegamma, exhibits accelerated aging phenotypes characteristic ofhuman aging, systemic mitochondrial dysfunction, multisystempathology, and reduced lifespan. Epidemiologic studies in humanshave demonstrated that endurance training reduces the risk ofchronic diseases and extends life expectancy. Whether enduranceexercise can attenuate the cumulative systemic decline observed inaging remains elusive. Here we show that 5 mo of enduranceexercise induced systemic mitochondrial biogenesis, preventedmtDNA depletion and mutations, increased mitochondrial oxida-tive capacity and respiratory chain assembly, restored mitochon-drial morphology, and blunted pathological levels of apoptosisin multiple tissues of mtDNA mutator mice. These adaptationsconferred complete phenotypic protection, reduced multisystempathology, and prevented premature mortality in these mice. Thesystemic mitochondrial rejuvenation through endurance exercisepromises to be an effective therapeutic approach to mitigatingmitochondrial dysfunction in aging and related comorbidities.

proliferator-activated receptor gamma coactivator-1α | sarcopenia |cardiac hypertrophy

The mitochondrial theory of aging postulates that the lifelongaccumulation of somaticmitochondrial DNA (mtDNA)muta-

tions leads to mitochondrial abnormalities resulting in a pro-gressive decline in tissue function (1, 2). Mitochondrial abnor-malities and mtDNA mutagenesis are well-established intrinsicinstigators that drive multisystem degeneration, stress intoler-ance, and energy deficits during aging in humans (3), monkeys (4),and rodents (5). Reduced mitochondrial quality and content inmultiple tissues is also implicated in several aging-associatedconditions, including cancer, obesity, cardiovascular diseases, hy-pertension, type 2 diabetes, osteoporosis, and dementia, as well asin the pathogenesis of neurometabolic syndromes, psychiatricdisorders, end-stage renal disease, and mitochondrial cytopathies(6–10). Current treatment strategies for conditions associatedwith mitochondrial dysfunction address the secondary symptomsbut not the deficiency itself (11). One possible approach to miti-gating the primary deficiency is to boost the residual mitochon-drial oxidative capacity by increasing functional mitochondrialmass in the affected tissues.The epidemic emergence of modern chronic diseases largely

stems from the adoption of a sedentary lifestyle and excess en-ergy intake (12). There is incontrovertible evidence from epide-miologic studies that endurance exercise extends life expectancyand reduces the risk of chronic diseases (7–10, 13–21). Endur-ance exercise is the most potent physiological inducer of mito-

chondrial biogenesis in skeletal muscle (12) and also has pro-found effects on metabolism in various other tissues, includingheart, brain, adipose tissue, and liver (22, 23). These adaptationsresult in improved healthspan, reduced risk of morbidity andmortality, and enhanced quality of life (12, 24). In this work, weused the mtDNA mutator mouse (designated the PolG mouse),a model of progeroid aging that exhibits elevated mtDNA pointmutations and systemic mitochondrial dysfunction and pheno-copies human aging (25, 26), to investigate whether enduranceexercise can effectively counteract the entrenched multisystemdegenerationandmitochondrial dysfunction tomitigateprematureaging in these mice.

Results and DiscussionEndurance Exercise Conferred Complete Phenotypic Protection andPrevented Early Mortality in PolG Mice. As early as 6 mo of age,sedentary PolG mice (PolG-SED) displayed symptoms of ac-celerated aging, as described previously (25, 26), including alope-cia, graying hair, weight loss, poor body condition, and impairedmobility (Fig. S1A and Movie S1). At 8 mo of age, PolG micethat had undergone 5 mo of forced endurance exercise (PolG-END; 15 m/min for 45 min, 3 times/wk) lacked visible features ofthe accelerated aging phenotype (alopecia and graying hair) andwere visually indistinguishable from age-matched WT littermates(Fig. S1A and Movie S1). Endurance exercise also attenuated thedecline in body weight and body condition in PolG mice (Fig. S1D and E). In addition, the PolG-END mice exhibited similarlevels of physical activity and motor performance as the WT ice(Movies S2 and S3), indicating improved muscle function. Whensubjected to a progressive exhaustive exercise test, PolG-ENDmice had significantly greater functional endurance capacitycompared with PolG-SED mice (Fig. 1A and Movie S4). Re-markably, the endurance capacity of PolG-END mice surpassedthat of WT mice in trials 2–4 (Fig. 1A). We also observed a sig-nificant and complete prevention of early mortality in PolG-ENDmice (P < 0.01; Fig. S1 B and C). Given the large effect size, thesefindings are highly significant despite the relatively small numberof animals per group; however, future studies with larger groupsof animals subjected to lifelong endurance exercise are needed todefine the full extent of this life-prolonging effect.

Author contributions: A.S. and M.A.T. designed research; A.S., J.M.B., D.I.O., J.P.L., B.P.H.,M.A., J.E.T., S.M., and N.J.M. performed research; J.E.T., S.M., G.C.K., T.A.P., and M.A.T.contributed new reagents/analytic tools; A.S., J.M.B., and B.P.H. analyzed data; and A.S.wrote the paper.

The authors declare no conflict of interest.

*This Direct Submission article had a prearranged editor.1To whom correspondence should be addressed. E-mail: tarnopol@mcmaster.ca.

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1019581108/-/DCSupplemental.

www.pnas.org/cgi/doi/10.1073/pnas.1019581108 PNAS | March 8, 2011 | vol. 108 | no. 10 | 4135–4140

MED

ICALSC

IENCE

S

Dow

nloa

ded

by g

uest

on

June

2, 2

020

Dow

nloa

ded

by g

uest

on

June

2, 2

020

Dow

nloa

ded

by g

uest

on

June

2, 2

020

Endurance Exercise Mitigated Multisystem Degeneration in PolGMice. Aging is characterized by a loss of muscle mass (sarcope-nia) and brain atrophy (9, 10, 27, 28). Consistent with the causalrole of mtDNA mutations in multiorgan pathology in PolG mice,we found significant reductions in skeletal muscle (25, 26, 29) andbrain mass in PolG-SED mice compared with WT mice (Fig. 1 Band C and Fig. S2A). PolG-SED mice also displayed significant

cardiac hypertrophy (Fig. 1D and Fig. S2 B and C) and heartpathology (Fig. S2D) compared with WT mice. Endurance exer-cise mitigated age-associated sarcopenia, brain atrophy, and car-diomyopathy in PolG-END mice (Fig. 1 B–D and Fig. S2 A–D).The PolG-SED mice had decreased body fat (lipodystrophy),

consistent with their overall cachectic appearance and reducedweight of abdominal and retroperitoneal fat pads (Fig. 1E).

Fig. 1. Endurance exercise mitigates the exercise intolerance and systemic pathology observed in PolG-SED mice. (A) The time to exhaustion (in min) of PolG-END mice in monthly endurance stress test trials was significantly greater compared with both PolG-SED and WT mice (n = 10/group). (B and C) Weights ofskeletal muscle (quadriceps femoris and gastrocnemius) (B) and brain (C) fromWT, PolG-SED, and PolG-END mice at 8 mo of age (n = 18/group). (D) Trichrome-stained cross-sections of heart from WT, PolG-SED, and PolG-END mice (n = 6/group). Representative images of heart from each group are displayed. (Scalebar: 50 mm.) (E) Weight of abdominal (ABDO) and retroperitoneal (RETRO) fat pads from WT, PolG-SED, and PolG-END mice (n = 10/group). (F) H&E-stainedsections of dorsal skin from WT, PolG-SED, and PolG-END mice (n = 3/group). Open arrowheads indicate dermal skin; closed arrowheads, subcutaneous fat;diamond arrowheads, subcutaneous muscle. (G) Weights of gonads (ovaries and testes) from WT, PolG-SED, and PolG-END mice (n = 10/group). (H) Hemo-globin values in from WT, PolG-SED, and PolG-END mice (n = 10/group). *P < 0.05, **P < 0.01, and ***P < 0.001, PolG-SED versus both WT and PolG-END;†P < 0.05, ‡P < 0.01, PolG-END versus WT. Error bars represent SEM.

4136 | www.pnas.org/cgi/doi/10.1073/pnas.1019581108 Safdar et al.

Dow

nloa

ded

by g

uest

on

June

2, 2

020

Endurance exercise induced significant gains in the mass of ab-dominal and retroperitoneal fat pads, thereby attenuating lip-odystrophy in PolG mice (Fig. 1E). In addition, histologicalexamination of dorsal skin and subcutis showed that exercisesuppressed the depletion of subcutaneous adipose tissue andthinning of the dermis in PolG mice (Fig. 1F). PolG-END micealso maintained a normal melanin content comparable to that ofWT mice (Fig. S2E). Profoundly reduced fertility has beenreported in PolG mice (26), congruent with reduced sperm countand decreased female fecundity with advancing age in these mice(26). We noted gonadal atrophy in 8-mo-old PolG-SED micecompared with WT mice (Fig. 1G). Endurance exercise attenu-ated the gonadal atrophy present in PolG mice of both sexes(Fig. 1G).Anemia is a frequent clinical problem seen in elderly humans

(30) and patients with acute leukemias (31). Consistent withprevious reports (26, 31), we found signs of macrocytic anemiawith abnormal erythroid maturation and megaloblastic changes,characterized by significantly higher mean corpuscular volumeand lower hemoglobin, erythrocyte and leukocyte concentrationsin peripheral blood of PolG-SED mice compared with WT mice(Fig. 1H and Fig. S3 A–C). We also noted splenic enlargement inPolG-SED mice, indicative of stress erythropoiesis (Fig. S3D).Endurance exercise prevented the development of anemia in thePolG mice, and hemoglobin, mean corpuscular volume, erythro-cyte, and leukocyte levels and spleen size were indistinguishablebetween WT and PolG-END mice (Fig. 1H and Fig. S3 A–D).

Endurance Exercise Attenuated the Decline in mtDNA Copy Numberand Reduced the Frequency of mtDNA Point Mutations in PolG Mice.Development of the progeroid phenotype in PolG mice iscausally associated with reduced mtDNA copy number and accu-mulation of mtDNA point mutations (25, 26). We found a sig-nificant depletion in full-length mtDNA content (skeletal mus-cle, heart, and liver) and an increase in mtDNA point mutations(skeletal muscle) in PolG-SED mice compared with WT mice(Fig. 2 A and B and Fig. S3 E–G). Endurance exercise com-pletely rescued mtDNA depletion in skeletal muscle, heart, andliver of the PolG mice (Fig. 2 A and B and Fig. S3 E and F). Acausative role for mtDNA point mutations via impaired assemblyof respiratory chain complexes in driving the premature aging inPolG mice has been reported (32). This phenomenon is exem-plified by mitochondrial cytochrome c oxidase (COX) complexassembly, wherein the mtDNA-encoded subunits are highly con-served and form the catalytic core of the complex. The assemblyof the mtDNA-encoded subunits of COX complex provides aplatform for the subsequent incorporation of the nuclear sub-units (33). Amino acid substitutions in the mtDNA-encoded sub-units of the COX complex, a consequence of mtDNA pointmutations, are deleterious to the function and stability of thiscomplex. Thus, such amino acid substitutions are strongly se-lected against in the germ line (34). Endurance exercise reducedthe frequency of point mutations in the PolG mice, resulting ina concomitant increase in mitochondrial COX complex assembly(Fig. 2B andC and Fig. S3G). Our results indicate that enduranceexercise-mediated normalization of the systemic degenerativepathology in PolG mice is closely associated with maintenance ofhigh levels of mtDNA copy number and a reduction in mtDNApoint mutation load. Clearly, exercise in PolG mice preventsmtDNAmutations from reaching a critical threshold above whichpathology manifests, and it represents a viable presymptomatictherapy for patients carrying polymerase gamma mutations knownto cause pathology (35).

Endurance Exercise Promoted Systemic Mitochondrial OxidativeCapacity and Restored Mitochondrial Morphology in PolG Mice.Endurance exercise is known to induce metabolic adaptationsvia activation of the transcriptional coactivator peroxisome

proliferator-activated receptor γ coactivator-1 α (PGC-1α), themaster regulator of mitochondrial metabolism and biogenesisthat has been touted as a potential therapeutic target for aging-associated diseases (36). Studies in primary cells from patientswith mitochondrial disorders and in skeletal muscle of COX10conditional knockout mice have indicated that inducing mito-chondrial biogenesis via ectopic expression of PGC-1α (37) orendurance training (38) has beneficial effects on the mitochon-drial pathology of genetic origin. Whether PGC-1α is the centralregulator of mitochondrial biogenesis in other tissues remainsunknown (36, 39). Interestingly, mild overexpression of PGC-1αin skeletal muscle alone is known to be protective against sar-copenia, to attenuate inactivity-induced fiber atrophy, to ame-liorate Duchenne muscular dystrophy and Huntington’s pathol-ogy, to reduce systemic chronic inflammation, and to maintain

Fig. 2. Endurance exercise rescues mtDNA depletion, mitigates mtDNArandom point mutations, and enhances COX assembly in PolG mice. (A)Quantification of mtDNA levels relative to diploid nuclear genome inskeletal muscle (soleus) from WT, PolG-SED, and PolG-END mice at 8 mo ofage (n = 10/group). (B) Frequency of mtDNA point mutations in 3,325, 2,473,and 3,842 reads of mtDNA sequences from multiple mtDNA populations,yielding 1.46 × 106, 1.25 × 106, and 9.21 × 105 bp of mtDNA sequences fromthe PolG-SED, PolG-END, and WT mice, respectively. (C) COX assembly inskeletal muscle (quadriceps femoris) from WT, PolG-SED, and PolG-ENDmice (n = 4–6/group) using 2D Blue-Native PAGE. Blots were probed withCOX subunits I, IV, Vb, and VIc. *P < 0.05, **P < 0.01, and ***P < 0.001,PolG-SED versus both WT and PolG-END; †P < 0.05, PolG-END versus WT.Error bars represent SEM.

Safdar et al. PNAS | March 8, 2011 | vol. 108 | no. 10 | 4137

MED

ICALSC

IENCE

S

Dow

nloa

ded

by g

uest

on

June

2, 2

020

systemic glucose and insulin homeostasis in aged mice (40–44).We found lower basal PGC-1α mRNA expression and nuclearabundance in skeletal muscle of PolG-SED mice compared withWT mice (Fig. 3A). The reduced nuclear PGC-1α content in

PolG-SED mice was associated with no change in the nuclearabundance of receptor-interacting protein (RIP) 140, a negativeregulator of PGC-1α (45). We also detected significant decreasesin the mRNA and protein content of mitochondrial transcriptionfactor A (Tfam), amediator ofmtDNA integrity and transcription(46), in concert with lower levels of other downstream PGC-1αtargets (Fig. 3A and Fig. S4 A–C). This is consistent with a recentstudy reporting profound down-regulation of gene sets associatedwith mitochondrial metabolism in PolG mice (29). Enduranceexercise abrogated the nuclear accumulation of RIP140, whileincreasing the content of nuclear PGC-1α and Tfam, as well asmRNA expression of several downstream targets in skeletalmuscle of PolG mice, thus shifting the cellular dynamics towardmitochondrial biogenesis (Fig. 3A and Fig. S4 A–C). PGC-1αmRNA expression remained unchanged in other organs (Fig.S4C), further supporting the association between maintenance ofmuscle-specific PGC-1α expression and exercise-mediated multi-system rejuvenation of PolG mice. Because PGC-1α lacks func-tional DNA- and ligand-binding domains, pharmacologicinterventions that directly promote PGC-1α function without af-fecting the activity of its upstream metabolic regulators remainelusive. Endurance exercise is the only practical way to “selec-tively” modulate PGC-1α function within a therapeutically bene-ficial window, thereby circumventing the unknown and unwantedside effects of the drugs and of nonspecific activation of PGC-1α.Enhanced mitochondrial biogenesis in response to endurance

exercise is supported by an increase in mitochondrial electrontransport chain (ETC) subunits and COX activity in skeletalmuscle of PolG-END mice compared with PolG-SED mice (Fig.3 B and C). Electron microscopy studies of skeletal muscle andheart also demonstrated increased mitochondrial abundance inPolG-END mice versus PolG-SED mice (Fig. 4 A–C and Figs. S5A–F and S6D). Electron microscopy revealed an accumulation ofswollen, pleomorphic, oversized mitochondria in skeletal muscleand heart of PolG-SED mice (Fig. 4 B and D–G and Figs. S5 Cand D and S6C). The abnormal mitochondria in the skeletalmuscle of PolG-SED mice exhibited cristae fragmentation (Fig.4D), vacuolization (Fig. 4E), disrupted membranes (Fig. 4F), andlarge myelin-like structures (Fig. 4G). Similar alterations in mi-tochondrial morphology have been documented to occur withage and in humans with mitochondrial myopathy (47, 48). En-durance exercise abrogated these morphological irregularities inmitochondrial morphology in PolG-END mice (Fig. 4 A, C, Hand I and Figs. S5 A, B, E, and F and S6 C and D).Although numerous epidemiologic studies have clearly shown

that exercise reduces morbidity and mortality (19, 21), data de-scribing the systemic effects of endurance exercise are scarce.Given our findings of dramatic suppression of the acceleratedaging phenotype and rescue of multisystem degenerate pathol-ogy, we sought to determine whether endurance exercise sys-temically promoted mitochondrial biogenesis in PolG mice. Wefound a significant reduction in the protein levels of mitochon-drial ETC subunits and mitochondrial COX activity in heart,liver, brain, pancreas, and gonadal tissue in PolG-SED micecompared with WT mice (Fig. 3 B and C and Fig. S6 A and B).This decline in mitochondrial ETC subunit content and COXactivity was prevented with endurance exercise in all tissuesstudied, indicating multisystem mitochondrial restoration (Fig. 3B and C and Fig. S6 A and B).

Endurance Exercise Mitigated Systemic Apoptosis in PolG Mice. Sus-tained mitochondrial dysfunction leads to activation of the cas-pase cascade culminating in DNA fragmentation, a hallmark ofapoptosis that has been observed in aged tissues, acute leukemia,and neurometabolic disorders (29, 49, 50). The PolG-SED micedisplayed increased DNA fragmentation in skeletal muscle,heart, liver, spleen, intestine, kidney, and gonads (Fig. 3D andFig. S6E), suggesting that dysregulated systemic apoptosis is in-

Fig. 3. Endurance exercise induces systemic mitochondrial biogenesis,enhances systemic COX activity, and mitigates dysregulated systemic apo-ptosis in PolG mice. (A) Gene expression and protein content of nuclear PGC-1α and Tfam in skeletal muscle (tibialis anterior) of PolG-SED and PolG-ENDmice versus WT mice (n = 10/group) at 8 mo of age. (B) RepresentativeWestern blot of PGC-1α–mediated downstream proteins—(i) complex V ATPsynthase subunit α, (ii) complex III subunit core 2, (iii) complex IV subunit I,(iv) complex II subunit 30 kDa, and (v) complex I subunit NADH-ubiquinoneoxidoreductase 1β subcomplex 8—in skeletal muscle (quadriceps femoris),lung, and heart extracts from WT, PolG-SED (PS), and PolG-END (PE) mice(n = 8/group). (vi) Actin was used as a housekeeping loading control. (C) COXactivity in skeletal muscle (quadriceps femoris), heart, and ovary of PolG-SEDand PolG-END mice versus WT mice (n = 10/group). (D) Nuclear DNA frag-mentation (apoptotic index) in the cytosolic fractions prepared from skeletalmuscle (quadriceps femoris), heart, and liver from WT, PolG-SED, and PolG-END mice (n = 4–6/group). *P < 0.05, **P < 0.01, and ***P < 0.001, PolG-SEDversus both WT and PolG-END; †P < 0.05, PolG-END versus WT. Error barsrepresent SEM.

4138 | www.pnas.org/cgi/doi/10.1073/pnas.1019581108 Safdar et al.

Dow

nloa

ded

by g

uest

on

June

2, 2

020

tegral to driving multisystem degenerative pathologies (25, 29).Indeed, suppression of apoptosis successfully prevented cardio-myopathy in the heart-specific mitochondrial mutator mousemodel (51). A recently proposed, intriguing mechanism postu-lates that somatic mtDNA mutation-driven generation of mis-folded mitochondrial proteins may bind to proapoptotic proteinsand activate systemic apoptosis (52). This might explain themitochondrial ETC complex assembly defect and the dispro-portionate increase in apoptosis seen in the PolG-SED mice.Endurance exercise systemically abrogated this apoptosis indexin the PolG mice, indicating that induction of prosurvivalmechanism(s) is crucial for multisystem maintenance (Fig. 3Dand Fig. S6E).

Summary and Perspectives. Although a plethora of previous studiesfound strong correlations among mtDNA mutations, mosaic re-spiratory chain dysfunction, and mammalian aging (2, 53–56),PolG mice provided the first direct cause-and-effect evidence thatmtDNA mutagenesis and mitochondrial dysfunction results inprogeroid aging phenotypes and associated multisystem patholo-gies (25, 26). Here we report that an increased burden of somaticmtDNA point mutations in PolGmice results in profound declinesin mitochondrial biogenesis and systemic oxidative metabolism,reductions in mtDNA copy number, defects in the assembly ofETC functional complexes, accumulation of degenerate mito-chondria, and a pathological increase in systemic apoptosis (25, 26,32). Strikingly, 5 mo of endurance exercise promoted systemicmitochondrial biogenesis and increased multiorgan oxidative ca-pacity, contributing to the complete phenotypic protection of thePolG mice. Whether the central mechanism driving mammalianaging and associated pathologies is mtDNA mutagenesis and de-pletion, enhanced systemic apoptosis, or some other form of mi-tochondrial dysfunction remains unknown (25, 26). Clearly, thetherapeutic effects of endurance exercise are unprecedented andmultifactorial in nature. The obvious question is how exercise canalleviate the mutational load despite the continued presence ofdefective mitochondrial polymerase γ. We hypothesize that en-durance exercise-mediated regulation of muscle-specific PGC-1αmay impose selective mitochondrial biogenesis of healthy mito-chondria via modulation of mitochondrial dynamics (fusion andfission) and targeted autophagy of mitochondria carrying patho-logical levels of mutatedmtDNA. This, together with the inductionof secondary polymerase γ-independent mtDNA repair pathwayswith exercise, may maintain a pool of bioenergetically functionalmitochondria. In addition, we speculate that endurance exercisemodulates the release of systemic factors (e.g., chemokines, cyto-kines, metabolites) that may promote organ cross-talk, resulting insystemic mitochondrial biogenesis and multisystem rejuvenation.These adaptations may mitigate systemic mitochondrial dysfunc-tion and accelerated cell death by diluting the pathological effectsof mtDNA point mutations incurred systemically in PolG mice.Our data clearly support endurance exercise as a medicine and

a lifestyle approach to improving systemic mitochondrial func-tion, which is critical for reducing morbidity and mortality acrossthe lifespan (7, 19, 21). Our findings also have substantialimplications for exercise therapy in young asymptomatic orpaucisymptomatic patients harboring known pathogenic muta-tions in mtDNA regulatory proteins, such as POLG1, twinklehelicase, and others. Understanding the multiple molecular cuesthat lead to endurance exercise-mediated systemic mitochondrialrejuvenation in the mtDNA mutator mouse could also lead to thedevelopment of novel nutritional, pharmacologic, and exercise-based therapeutic interventions designed to ameliorate the struc-tural and functional mitochondrial alterations associated withaging and metabolic diseases.

Materials and MethodsSee SI Materials and Methods for details regarding animal breeding, exerciseprotocol, anthropometric measurements, endurance stress testing, and sur-vival analyses. Molecular analyses including electron microscopy, melaninassays, mRNA expression, mtDNA copy number and point mutation analyses,subcellular fractionation, 2D BN-PAGE, immublotting, COX activity assays,apoptosis cell death detection ELISA, and statistical analyses are described inSI Materials and Methods. mtDNA mutator mice used in this study weredescribed in ref. 25.

ACKNOWLEDGMENTS. This work was supported by the Canadian Institutesof Health Research (Grant MOP97805), a kind donation from Mr. WarrenLammert and family (to M.A.T.), and Nathan Shock Award P30AG025708 (toS.M.). A.S. is funded by a Canadian Institutes of Health Research Instituteof Aging doctoral research scholarship and a Canada PRESTIGE fellowship.D.I.O. and J.P.L. are funded by National Sciences and Engineering ResearchCouncil scholarships. G.C.K. and T.A.P. were awarded US Patent 7,126,040for the PolgD257A mouse.

Fig. 4. Endurance exercise restores mitochondrial abundance and mor-phology in PolG mice. (A–C) Electron micrographs of myofibers (quadricepsfemoris) from WT (A), PolG-SED (B), and PolG-END (C) mice (n = 6/group) at 8mo of age. (Scale bar: 1 μm.) (D–I) Myofibers of PolG-SED mice (D–G) arepopulated with enlarged, abnormally shaped mitochondria containingvacuoles, fragmented cristae, disrupted external membranes, and largemyelin-like figures compared with mitochondria observed in WT (H) andPolG-END (I) mice. (Scale bar: 100 nm.)

Safdar et al. PNAS | March 8, 2011 | vol. 108 | no. 10 | 4139

MED

ICALSC

IENCE

S

Dow

nloa

ded

by g

uest

on

June

2, 2

020

1. Harman D (1972) The biologic clock: The mitochondria? J Am Geriatr Soc 20:145–147.2. Larsson NG (2010) Somatic mitochondrial DNA mutations in mammalian aging. Annu

Rev Biochem 79:683–706.3. Corral-Debrinski M, et al. (1992) Mitochondrial DNA deletions in human brain:

Regional variability and increase with advanced age. Nat Genet 2:324–329.4. Schwarze SR, et al. (1995) High levels of mitochondrial DNA deletions in skeletal

muscle of old rhesus monkeys. Mech Ageing Dev 83:91–101.5. Khaidakov M, Heflich RH, Manjanatha MG, Myers MB, Aidoo A (2003) Accumulation

of point mutations in mitochondrial DNA of aging mice. Mutat Res 526:1–7.6. Wallace DC (2005) A mitochondrial paradigm of metabolic and degenerative diseases,

aging, and cancer: A dawn for evolutionary medicine. Annu Rev Genet 39:359–407.7. Johannsen DL, Ravussin E (2009) The role of mitochondria in health and disease. Curr

Opin Pharmacol 9:780–786.8. Warburton DE, Charlesworth S, Ivey A, Nettlefold L, Bredin SS (2010) A systematic

review of the evidence for Canada’s Physical Activity Guidelines for Adults. Int J BehavNutr Phys Act 7:39.

9. Safdar A, et al. (2010) Aberrant mitochondrial homeostasis in the skeletal muscle ofsedentary older adults. PLoS ONE 5:e10778.

10. Safdar A, Debeer J, Tarnopolsky MA (2010) Dysfunctional Nrf2-Keap1 redox signallingin skeletal muscle of sedentary old. Free Radic Biol Med 49:1487–1493.

11. DiMauro S, Mancuso M (2007) Mitochondrial diseases: Therapeutic approaches. BiosciRep 27:125–137.

12. Holloszy JO, Booth FW (1976) Biochemical adaptations to endurance exercise inmuscle. Annu Rev Physiol 38:273–291.

13. Herring MP, O’Connor PJ, Dishman RK (2010) The effect of exercise training onanxiety symptoms among patients: A systematic review. Arch Intern Med 170:321–331.

14. Nader GA, Lundberg IE (2009) Exercise as an anti-inflammatory intervention tocombat inflammatory diseases of muscle. Curr Opin Rheumatol 21:599–603.

15. Brenner I (2009) Exercise performance by hemodialysis patients: A review of theliterature. Phys Sportsmed 37:84–96.

16. Beal MF (2009) Therapeutic approaches to mitochondrial dysfunction in Parkinson’sdisease. Parkinsonism Relat Disord 15(Suppl 3):S189–S194.

17. Balbuena L, Casson AG (2009) Physical activity, obesity and risk for esophagealadenocarcinoma. Future Oncol 5:1051–1063.

18. Buchner DM (2009) Physical activity and prevention of cardiovascular disease in olderadults. Clin Geriatr Med 25:661–675, viii.

19. Stessman J, Hammerman-Rozenberg R, Cohen A, Ein-Mor E, Jacobs JM (2009) Physicalactivity, function, and longevity among the very old. Arch Intern Med 169:1476–1483.

20. ten Hacken NH (2009) Physical inactivity and obesity: Relation to asthma and chronicobstructive pulmonary disease? Proc Am Thorac Soc 6:663–667.

21. Chakravarty EF, Hubert HB, Lingala VB, Fries JF (2008) Reduced disability andmortality among aging runners: A 21-year longitudinal study. Arch Intern Med 168:1638–1646.

22. Sutherland LN, Bomhof MR, Capozzi LC, Basaraba SA, Wright DC (2009) Exercise andadrenaline increase PGC-1alpha mRNA expression in rat adipose tissue. J Physiol 587:1607–1617.

23. Boveris A, Navarro A (2008) Systemic and mitochondrial adaptive responses tomoderate exercise in rodents. Free Radic Biol Med 44:224–229.

24. Hawley JA, Holloszy JO (2009) Exercise: It’s the real thing! Nutr Rev 67:172–178.25. Kujoth GC, et al. (2005) Mitochondrial DNA mutations, oxidative stress, and apoptosis

in mammalian aging. Science 309:481–484.26. Trifunovic A, et al. (2004) Premature ageing in mice expressing defective

mitochondrial DNA polymerase. Nature 429:417–423.27. Lexell J, Taylor CC, Sjöström M (1988) What is the cause of the ageing atrophy? Total

number, size and proportion of different fiber types studied in whole vastus lateralismuscle from 15- to 83-year-old men. J Neurol Sci 84:275–294.

28. Ito M, Hatazawa J, Yamaura H, Matsuzawa T (1981) Age-related brain atrophy andmental deterioration—a study with computed tomography. Br J Radiol 54:384–390.

29. Hiona A, et al. (2010) Mitochondrial DNA mutations induce mitochondrialdysfunction, apoptosis and sarcopenia in skeletal muscle of mitochondrial DNAmutator mice. PLoS ONE 5:e11468.

30. Balducci L (2003) Epidemiology of anemia in the elderly: Information on diagnosticevaluation. J Am Geriatr Soc 51 (3, Suppl)S2–S9.

31. Chen ML, et al. (2009) Erythroid dysplasia, megaloblastic anemia, and impairedlymphopoiesis arising from mitochondrial dysfunction. Blood 114:4045–4053.

32. Edgar D, et al. (2009) Random point mutations with major effects on protein-codinggenes are the driving force behind premature aging in mtDNA mutator mice. CellMetab 10:131–138.

33. Dell’Agnello C, et al. (2007) Increased longevity and refractoriness to Ca(2+)-dependent neurodegeneration in Surf1 knockout mice. Hum Mol Genet 16:431–444.

34. Stewart JB, et al. (2008) Strong purifying selection in transmission of mammalianmitochondrial DNA. PLoS Biol 6:e10.

35. Ashley N, et al. (2008) Depletion of mitochondrial DNA in fibroblast cultures frompatients with POLG1 mutations is a consequence of catalytic mutations. Hum MolGenet 17:2496–2506.

36. Handschin C, Spiegelman BM (2008) The role of exercise and PGC1α in inflammationand chronic disease. Nature 454:463–469.

37. Srivastava S, et al. (2009) PGC-1α/β-induced expression partially compensates forrespiratory chain defects in cells from patients with mitochondrial disorders. HumMolGenet 18:1805–1812.

38. Wenz T, Diaz F, Hernandez D, Moraes CT (2009) Endurance exercise is protective formice with mitochondrial myopathy. J Appl Physiol 106:1712–1719.

39. Handschin C (2009) The biology of PGC-1α and its therapeutic potential. TrendsPharmacol Sci 30:322–329.

40. Handschin C, et al. (2007) Abnormal glucose homeostasis in skeletal muscle-specificPGC-1α knockout mice reveals skeletal muscle-pancreatic β cell crosstalk. J Clin Invest117:3463–3474.

41. Handschin C, et al. (2007) PGC-1α regulates the neuromuscular junction program andameliorates Duchenne muscular dystrophy. Genes Dev 21:770–783.

42. Wenz T, Rossi SG, Rotundo RL, Spiegelman BM, Moraes CT (2009) Increased musclePGC-1α expression protects from sarcopenia and metabolic disease during aging. ProcNatl Acad Sci USA 106:20405–20410.

43. Chaturvedi RK, et al. (2009) Impaired PGC-1α function in muscle in Huntington’sdisease. Hum Mol Genet 18:3048–3065.

44. Sandri M, et al. (2006) PGC-1α protects skeletal muscle from atrophy by suppressingFoxO3 action and atrophy-specific gene transcription. Proc Natl Acad Sci USA 103:16260–16265.

45. Seth A, et al. (2007) The transcriptional corepressor RIP140 regulates oxidativemetabolism in skeletal muscle. Cell Metab 6:236–245.

46. Larsson NG, et al. (1998) Mitochondrial transcription factor A is necessary for mtDNAmaintenance and embryogenesis in mice. Nat Genet 18:231–236.

47. Tate EL, Herbener GH (1976) A morphometric study of the density of mitochondrialcristae in heart and liver of aging mice. J Gerontol 31:129–134.

48. Lindal S, et al. (1992) Mitochondrial diseases and myopathies: A series of musclebiopsy specimens with ultrastructural changes in the mitochondria. Ultrastruct Pathol16:263–275.

49. Wohlgemuth SE, Seo AY, Marzetti E, Lees HA, Leeuwenburgh C (2010) Skeletalmuscle autophagy and apoptosis during aging: Effects of calorie restriction and life-long exercise. Exp Gerontol 45:138–148.

50. Wasilewski M, Scorrano L (2009) The changing shape of mitochondrial apoptosis.Trends Endocrinol Metab 20:287–294.

51. Mott JL, et al. (2004) Cardiac disease due to random mitochondrial DNA mutations isprevented by cyclosporin A. Biochem Biophys Res Commun 319:1210–1215.

52. Mott JL, Zhang D, Zassenhaus HP (2005) Mitochondrial DNA mutations, apoptosis,and the misfolded protein response. Rejuvenation Res 8:216–226.

53. Cantuti-Castelvetri I, et al. (2005) Somatic mitochondrial DNA mutations in singleneurons and glia. Neurobiol Aging 26:1343–1355.

54. Fayet G, et al. (2002) Ageing muscle: Clonal expansions of mitochondrial DNA pointmutations and deletions cause focal impairment of mitochondrial function.Neuromuscul Disord 12:484–493.

55. Michikawa Y, Mazzucchelli F, Bresolin N, Scarlato G, Attardi G (1999) Aging-dependent large accumulation of point mutations in the human mtDNA controlregion for replication. Science 286:774–779.

56. Nekhaeva E, et al. (2002) Clonally expanded mtDNA point mutations are abundant inindividual cells of human tissues. Proc Natl Acad Sci USA 99:5521–5526.

4140 | www.pnas.org/cgi/doi/10.1073/pnas.1019581108 Safdar et al.

Dow

nloa

ded

by g

uest

on

June

2, 2

020

Editorial Expression of Concern

MEDICAL SCIENCESPNAS is publishing an Editorial Expression of Concern regardingthe following article: “Endurance exercise rescues progeroid agingand induces systemic mitochondrial rejuvenation in mtDNA mu-tator mice,” by Adeel Safdar, Jacqueline M. Bourgeois, Daniel I.Ogborn, Jonathan P. Little, Bart P. Hettinga, Mahmood Akhtar,James E. Thompson, Simon Melov, Nicholas J. Mocellin, GregoryC. Kujoth, Tomas A. Prolla, and Mark A. Tarnopolsky, which wasfirst published February 22, 2011; 10.1073/pnas.1019581108 (ProcNatl Acad Sci USA 108:4135–4140).The editors wish to note that the following concerns were

brought to our attention by Mark Tarnopolsky: “The loadingcontrols in Fig. 3B, vi, and Fig. S6A, vi, appear to be identical inboth figures.” The editors are therefore issuing an expression ofconcern and will notify readers once more information is available.

Published under the PNAS license.

Published online June 11, 2018.

www.pnas.org/cgi/doi/10.1073/pnas.1808586115

E5834 | PNAS | June 19, 2018 | vol. 115 | no. 25 www.pnas.org