To examine transcriptional alterations associated with aging in skeletal muscle and the heart, we and others have used DNA microarrays to compare the gene expression profile of young and old animals. Aging results in a differential gene expression pattern specific to each tissue, and most alterations can be completely or partially prevented by caloric restriction (CR) in both heart and skeletal muscle. Transcriptional patterns of tissues from calorie-restricted animals suggests that CR retards the aging process by reducing endogenous damage and by inducing metabolic shifts associated with specific transcriptional profiles. These studies demonstrate that DNA microarrays can be used in cardiovascular aging research to generate panels of hundreds of transcriptional biomarkers, providing a new tool to measure biological age of cardiac and skeletal muscles and to test interventions designed to retard aging in these tissues.
The increasing probability of death with age is due to a number of factors, including neoplasia, sepsis, and organ-specific failure. Given the long mammalian lifespan, measuring the aging process on an organ-specific basis in mammals, through a panel of molecular markers, is likely to be useful for evaluating interventions. In order to gain a better understanding of general aspects of aging and to identify aging biomarkers, we are currently using DNA microarray analysis to study caloric restriction (CR) as a model system of aging retardation in mammals. We have characterized the gene expression profile associated with the aging process and CR in several organs, including skeletal muscle [1,2], brain , heart , and adipose tissue . These studies have shown that DNA microarrays can be used to identify potential mechanisms of aging and its retardation by CR. In this review we highlight our findings in skeletal and cardiac muscles, two tissues of critical importance for health maintenance in the aged.
2. The gene expression profile of aging and CR in mouse skeletal muscle
Aging in muscle is associated with a decrease in mass, strength, and rate of contraction. One of the most significant effects of aging is loss of muscle mass, known as sarcopenia, a process that may be triggered by reactive oxygen species (ROS) that are generated throughout an organism's lifespan [6,7]. Aging of muscle is also characterized by muscle cell atrophy, presence of lipofuscin deposits, mitochondrial abnormalities, and increase in oxidative damage to proteins, lipid, and DNA [8–10].
Our previous analysis of gene expression profile of aging in gastrocnemius muscle from 5- and 30-month-old male C57BL/6 mice revealed that aging resulted in a differential gene expression pattern . Of the genes that increased in expression with age, 16% were mediators of stress responses, including heat shock response genes, oxidative stress inducible genes, and DNA damage inducible genes (Table 1). Genes involved in energy metabolism were down-regulated with aging, including genes associated with mitochondrial function and turnover. This suppression of metabolic activity was accompanied by a concerted a decline in expression of genes involved in glycolysis, glycogen metabolism, and the glycerophosphate shunt (Table 1). Aging was also characterized by inductions of genes involved in neuronal growth and large reductions in the expression of biosynthetic enzymes. A comparison of 30-month-old control and CR mice revealed that aging-related changes in gene expression profiles were remarkably attenuated by CR. Most age-related alterations were either completely or partially prevented by CR, suggesting that CR mice appear to be biologically younger than animals receiving the control diet. Examination of global changes in gene expression induced by CR suggested a metabolic reprogramming in CR mice. CR resulted in a transcriptional shift toward energy metabolism, increased biosynthesis, and macromolecular turnover (Table 1). The CR-mediated reduction of mRNA encoding inducible genes involved in metabolic detoxification, oxidative stress response, and DNA repair is in agreement with the observed reduction in oxidative damage in calorie-restricted rodents .
Global view of transcriptional changes induced by aging and caloric restriction in skeletal muscle and heart of mice
↑ Stress response
↑ Protein metabolism
Induction of: Heat shock response
Oxidative-stress inducible genes
DNA damage inducible genes
↑ Energy metabolism
Induction of: Glycolysis
↓ Energy metabolism
Pentose phosphate shunt
↑ Neuronal injury
Induction of: Fatty acid synthesis
Neurite extension and sprouting
↓ Macromolecular damage
Suppression of: Inducible heat shock factors
Inducible detoxification systems
Inducible DNA repair systems
↑ Structural proteins
↓ Structural proteins
Induction of: Extracellular matrix components
Suppression of: Extracellular matrix components
Actin/tubulin structural modulation
↓ Fatty acid metabolism
↓ Immune response
Suppression of: MHC molecules
↓ Protein synthesis
Suppression of initiation factors
↓ DNA damage
Suppression of DNA-damage inducible transcripts
Induction of DNA repair genes
Suppression of pro-apoptosis factors
Induction of Inhibitors of apoptosis
We have also investigated the effects of aging and adult-onset CR on the gene expression profile in the vastus lateralis muscle from rhesus monkeys . We observed the concerted induction of genes encoding proteins linked to an inflammatory/immune function. Additionally, genes involved in oxidative stress responses and neuronal death, remodeling, and repair were induced with aging. In contrast, genes involved in energy metabolism, such as mitochondrial electron transport and oxidative phosphorylation, were down-regulated in aged muscles of rhesus monkeys.
To determine the influences of CR on gene expression profiles, the transcriptional patterns of middle-aged monkeys (19–20 years old) that have been subjected to CR since early adulthood were compared to those of age-matched monkeys receiving the control diet . The major classes of genes induced by CR were composed of genes encoding structural proteins and genes involved in cellular growth. CR resulted in a down-regulation of genes involved in energy metabolism, suggesting that CR monkeys may be in a hypometabolic state associated with reduced activity of the mitochondrial electron transport system. Surprisingly, we did not observe beneficial effects of adult-onset CR on the progression of age-related changes in gene expression. One possible explanation for these findings is that CR effects in aging retardation as reflected in gene expression can only be observed if CR is initiated early in life, or if tissues are profiled late in life. Alternatively, the impact of CR on the aging process at the transcriptional level may be species-specific, being greater in shorter-lived organisms.
Recently, the gene expression profile of aging in human skeletal muscle has been conducted with vastus lateralis biopsies obtained from healthy young (21–27 years old) and older men (67–75 years old) . A consequence of muscle aging in men was an induction of genes encoding metallothioneins, high-mobility-group proteins, heterogeneous nuclear ribonucleoproteins, and other RNA binding/processing proteins, and components of the ubiquitin–proteasome proteolytic pathway. Muscle aging in men was also characterized by decreased expression of genes involved in energy metabolism and mitochondrial protein synthesis . In a separate study with skeletal muscle of young (20–29 years old) and old (65–71 years old) women, genes encoding pre-mRNA/mRNA binding proteins were up-regulated, whereas genes involved in energy metabolism were down-regulated with aging. . The largest differential expression between young and old humans, in both men and women, was observed for the cell cycle regulator p21 (cyclin-dependent kinase inhibitor 1A, cdkn1a). A recent study from our laboratory has revealed that the age-related up-regulation of p21 is a component of an age-related transcriptional program induced in skeletal muscle of multiple species, including mice  and rhesus monkeys . In the mouse, this transcriptional program involves an induction of p53 tumor suppressor gene and several of its transcriptional targets, including p21 and GADD45α [M. Edwards and T.A Prolla, unpublished results]. Interestingly, the human DNA microarray studies have not found any consistent pattern of activation of oxidative stress response genes. This result is in contrast to the observed age-dependent increases in oxidative damage of DNA [8,10] and lipids [8,10] in human skeletal muscle, both of which support a role for oxidative stress in muscle aging. Taken as a whole, recent gene expression profile studies using DNA microarrays suggest that skeletal muscle aging is associated with decreased energy metabolism and induction of a stress response.
3. The gene expression profile of aging and CR in the mouse heart
Physiological investigations have shown that the aging heart in human and rodents is characterized by a loss of myocytes (cardiomyopathy) [14,15], enlargement of the remaining myocytes (hypertrophy) [15,16], and cardiac fibrosis . At the molecular level, age-related changes in hearts include lipofuscin pigment accumulation  and a reduction in calcium transport across the sarcoplasmic reticulum membrane . Other alterations include the induction of apoptotic factors  and loss of stem cells . In rats, moderate CR reduces the progression of cardiomyopathy  and lifelong CR prevents the age-related impairments of late diastolic function in B6D2F1 mice . CR also attenuates the age-related increases of molecular damage, including DNA oxidation , protein cross-linking , and somatic mitochondrial genomic rearrangements .
To understand the underlying mechanisms of aging, transcriptional alterations were measured in 5- and 30-month-old male B6C3F1 mice . Transcriptional profiles associated with heart aging are characterized by a metabolic shift toward carbohydrate metabolism, induction of genes having a structural role, and reduced protein synthesis (Table 1). Aging in the heart was also associated with increased expression of genes encoding cellular structural proteins, including extracellular matrix (ECM) components, junctional proteins, and smooth muscle contraction modulators. This evidence at the transcriptional level is supported by known age-related alterations in heart structure, such as ECM protein deposition , fibrosis , and cardiomyocyte hypertrophy . We also observed a metabolic shift from fatty acid to carbohydrate metabolism (Fig. 1). Key regulators of fuel selection were down-regulated and several genes involved in fatty acid β-oxidation in mitochondria were decreased in expression with aging. CR initiated in middle age markedly suppresses most age-related alterations in gene expression. In addition to preventing age-related transcriptional changes, CR also resulted in alterations in gene expression consistent with a marked cytoskeletal reorganization, decreased immune response, reduced DNA damage, and modulation of apoptosis (Table 1). Transcriptional profile of CR mice suggests that CR protects cardiomyocytes from age-dependent apoptosis by enhancing DNA repair and reducing endogenous DNA damage .
Genes involved in β-oxidation are reduced in expression as a result of aging. The transcriptional patterns also suggest that glucose utilization is increased as a result of heart aging. Genes encoding enzymes involved in each metabolic step are italicized. Increases (+) or decreases (−) in mRNA levels are indicated next to gene names. All genes shown were altered in expression as a result of aging, and the alteration in expression was either completely or partially prevented by CR. Reproduced from .
In rats, old hearts subjected ischemia/reperfusion showed an injury-related response, associated with up-regulation of genes associated with hypertrophy or apoptosis , and gene expression profiling demonstrated that cardiac aging is associated with transcriptional alterations that may contribute to reduced β-adrenergic signaling . Large-scale transcriptome analysis of failing and nonfailing human myocardium has identified genes associated with heart failure, including mitogen-activated protein kinases (MAPKs) that are activated in response to stress stimulation and have been implicated in cardiac hypertrophy and failure . These gene expression profiles of heart aging display similarities with known heart disorders and provide support for the hypothesis that age-related transcriptional alterations in cardiomyocytes play an important and perhaps underlying role in functional losses and reduced stress resistance in aged hearts. Interestingly, cardiac chamber-specific gene expression has demonstrated a different pattern of gene expression among the four heart chambers and the interventricular septum , suggesting that transcriptional profiling of whole hearts may represent a heterogeneous transcriptome with important age-related changes potentially masked.
4. Age-related impairment of the transcriptional responses to oxidative stress in the mouse heart
Heart aging is associated with impaired resistance to stressors, such as hypoxic injury . At the gene expression level, heart aging has been shown to be associated with an augmentation of cardiac IL-6 induction following endotoxin challenge  and reduced expression of immediate early genes (IEGs) c-fos, c-myc, and c-jun following hemodynamic stress  and pressure overload . In order to learn more about the specific pathways that are induced by oxidative stress, and how this response is affected by aging, we treated young (5 months old), middle-aged (15 months old), and aged (30 months old) mice with the ROS generator Paraquat (50 mg/kg of body weight) and sacrificed animals at 1, 3, 5, and 7 h after injection . To determine if paraquat induced oxidative stress, we assayed cardiac tissue for the presence of F(2)-isoprostanes. Basal levels of isoprostanes were 1.5 ± 0.1 and 1.1 ± 0.1 ng/g in control old (30 months old) and young (5 months old) animals, and were raised to 2.3 ± 0.4 and 2.2 ± 0.1 ng/g in paraquat injected old and young aged animals, respectively. Of the 9977 genes probed on the oligonucleotide microarray we identified 249 genes altered in expression (ANOVA, P<0.01) in the young mice, 298 in the middle-aged mice and 256 in the old mice. Interestingly, only 55 transcripts were determined to be paraquat responsive for all age groups. Paraquat induced several genes previously known to mediate stress responses, including Metallothioneins 1 and 2, GADD45, p21, and Sestrin. The induction of several genes appears to be associated with a protective metabolic stress response in the heart, including Bcl-XL, an anti-apoptototic protein that allows cells to maintain oxidative metabolism during cellular stress by allowing continued transport of metabolites across the outer mitochondrial membrane  and 5′nucleotidase, an enzyme that controls the production of adenosine in the heart through dephosphorylation of AMP .
A major class of genes displaying increased expression after paraquat injection was the IEGs, including zfp36, btg2, ptpn16, cyr61, nr4a1, atf3, junb, krox-24, and ptgs2. As the first transcriptional response to damage, IEG protein products initiate a coordinated cascade of adaptive gene expression in response to a given stimulus . We found that young mice displayed a greater number of IEGs induced compared with middle-aged and old animals (12 genes in the young, 7 genes in the middle-aged, and 5 genes in the old). Interestingly, several IEGs that show age-related alterations in expression have been shown to be dependent on mitogen-activated protein kinase (MAPK) signaling for expression. Aging was also associated with an impairment in the induction of several stress response genes, including the GADD45 isoforms α, β, and γ, MAP3K6, and JunB. Interestingly, we observed lower constitutive levels of several antioxidant genes in the aged hearts, including Gpx4, Peroxiredoxin 1, Peroxiredoxin 2, Peroxiredoxin 5, Sod1, and Sod2. These observations suggest that aged hearts may be less resistant to oxidative stress, although the mechanism of reduction in expression of antioxidant genes is not obvious from the DNA microarray data. Our findings are consistent with a previous study based on cDNA libraries from mouse ventricular cardiac muscle cells, which indicates reduced protection against stress-induced injury and impairment of mitochondrial electron transport system associated with the development of contractile dysfunction in aged mouse .
5. The impact of α-lipoic acid, coenzyme, Q10 and caloric restriction on gene expression patterns in the aging heart
We evaluated the efficacy of three dietary interventions started at middle age (14 months) to retard the aging process in mice . These were supplemental alpha-lipoic acid (LA) or coenzyme Q10 (CQ) and caloric restriction (CR, a positive control). LA is a potent antioxidant that acts as a cofactor of key mitochondrial enzymes, such as pyruvate dehydrogenase and α-ketoglutarate dehydrogenase . The aging rat myocardium exhibits increased oxidant production, significantly lower ascorbic acid, and a marked increase of 8-oxo-2′-deoxyguyanosine (8-oxo-dG) . LA supplementation significantly reverses the age-related decline in myocardial ascorbic acid content, and lowers the rate of oxidant production and the steady-state level of 8-oxo-dG . LA also ameliorates the age-related increase in lipid peroxidation . Other benefits of LA supplementation include a lowering of blood pressure in spontaneously hypertensive rats , an increase of both oxidative and nonoxidative glucose metabolism associated with the enhancement of insulin sensitivity in skeletal muscle of insulin-resistant fatty Zucker rats , increased tissue levels of glutathione in mice , and an increase of insulin-stimulated glucose metabolism in type 2 diabetics patients .
Coenzyme Q serves the dual function of an electron carrier/proton translocator in the mitochondrial electron transport chain  and also of an antioxidant that directly scavenges radicals [51,52] and regenerates α-tocopherol [53–55]. Dietary supplementation of mice with CQ results in increases in both CQ and α-tocopherol in mitochondria of several tissues . Pretreatment with CQ improves the functional recovery of senescent rat hearts after aerobic stress  and the contractile function of elderly patients after cardiac surgery , and also attenuates cytochrome c release from isolated mitochondria treated with the complex III inhibitor antimycin . In humans, coenzyme Q10 deficiency is associated with a genetic syndrome of multiple respiratory-chain dysfunction that is responsive to dietary CQ . Interestingly, lack of CQ in the diet increases the lifespan of Caenorhabditis elegans, perhaps by lowering metabolism and inducing an adaptive response . These studies suggest that CQ plays central roles as an electron carrier, antioxidant, and as a lifespan determinant.
In our study, LA and CQ had no impact on longevity or tumor patterns compared with control mice fed the same number of calories, whereas CR increased maximum life span by 13% (p<.0001) and reduced tumor incidence . To evaluate these interventions at the molecular level, we used microarrays to monitor the expression of 9977 genes in hearts from young (5 months) and old (30 months) mice. Of the 9977 genes studied, 5325 (53%) and 5231 (52%) genes were expressed in the heart of 30-month-old LA and CQ treated groups, respectively, as determined by the Affymetrix algorithm. LA, CQ, and CR inhibited age-related alterations in the expression of genes involved in the extracellular matrix, cellular structure, and protein turnover. However, unlike CR, LA, and CQ did not prevent age-related transcriptional alterations associated with energy metabolism. LA supplementation lowered the expression of genes encoding major histocompatibility complex (MHC) components and of genes involved in protein turnover and folding. We have previously observed that CR is associated with a reduction in complement cascade encoding genes and MHC genes , and we believe that these gene expression shifts reflect reduced endogenous ROS production.
CQ increased expression of genes involved in oxidative phosphorylation and reduced expression of genes involved in the complement pathway and several aspects of protein function. These included Cytochrome C oxidase subunits and ATP synthase. CQ also reduced the expression of the stress response genes Peroxiredoxin 5, GADD45γ, Hsp27, and Stip1. These observations strongly suggest that CQ supplementation reduces oxidative stress in the aging heart. We divided age-related transcriptional alterations in four broad categories, extracellular matrix genes, structural component genes, protein turnover, and energy metabolism. CR had a broad impact on aging retardation in this model, and the antioxidants offered partial protection in some gene categories (Table 2). The observations suggest that supplementation with LA or CQ results in transcriptional alterations consistent with a state of reduced oxidative stress in the heart, but that these dietary interventions are not as effective as CR in inhibiting the aging process in this tissue. In particular, age-associated transcriptional patterns of genes involved in energy metabolism are strikingly similar to the metabolic consequences of cardiac hypertrophy and atrophy [62,63]. CR, but not antioxidants, is effective in preventing these alterations. Our observations suggest that oxidative stress contributes to heart aging, but that it is unlikely to be the major causal event. Possibly, alterations in mitochondrial bioenergetics play a causative role, since CR, as opposed to the dietary interventions, largely opposed such alterations. These observations are in agreement with the recent finding that mice lacking expression of antioxidant genes do not appear to display accelerated aging phenotypes .
Comparative impact of CR, LA, and CQ on opposing the development of age-related transcriptional alterations
% CR effect
% LA effect
% CQ effect
Extracellular matrix (6 genes)
Type XV collagen alpha 1
Procollagen C-proteinase enhancer protein
Type VI collagen alpha 3
Type IV collagen alpha 1
Hyaluronan synthase 1
Cellular/cytosolic structure (9 genes)
Integrin alpha 6
Intercellular adhesion molecule 2
Troponin T1, skeletal, slow
Actin, alpha 2, smooth muscle, aorta
Ankyrin 1, erythroid
Protein turnover (4 genes)
Proteasome 28 subunit, 3 (Psme3)
Peptidylprolyl isomerase C
Ubiquitin specific protease 23
Ubiquitin-like 1 activating enzyme E1A
Energy metabolism (14 genes)
Fatty acid transporter
Fatty acid translocase (Cd36)
Carnitine O-palmitoyl transferase I
Acyl-CoA thioesterase 1
Lipase, hormone sensitive
Acyl-CoA oxidase 1
3, 2 Enoyl CoA isomerse (Peci)
Enoyl CoA hydratase 1
3, 2 trans-Enoyl CoA isomerse (Dci)
Pyruvate dehydrogenase kinase 4 (Pdk4)
Uncoupling protein 3 (Ucp3)
The percent effect of each treatment was computed using ((O−Trt)/(O−Y)) × 100 for each gene, where O, Trt, and Y indicate the average signals of 30-month-old control group, 30-month-old treatment group (either caloric restriction (CR), alpha-lipoic acid (LA), or coenzyme Q10 (CQ)), and the 5-month-old control group, respectively. Average effects for the whole classes, as well as individual genes are shown. Negative percentages correspond to exacerbation of age-related alterations.
Heart and skeletal muscle undergo major age-related alterations in gene expression. Gene expression patterns suggest induction of stress response pathways and reduced expression of genes involved in energy metabolism, both of which are broadly prevented by long-term CR in rodents. Antioxidants such as lipoic acid and CoQ10 can partially prevent age-related alterations in gene expression, but are not as effective as caloric restriction. Possibly, combinations of antioxidants and compounds that optimize mitochondrial function will be more effective. Extension of these DNA microarray studies to other organs should result in the identification of hundreds of tissue-specific biomarkers of aging, facilitating the elucidation of aging mechanisms and the testing of interventions. Because the pathophysiology of aging is complex, it is likely that a detailed analysis of gene expression profiles in multiple tissues will reveal both tissue-specific and general patterns. However, certain limitations remain associated with data generated by microarrays. First, the observed collection of gene expression alterations in whole tissues such as muscle and heart is complex, reflecting the presence of diverse cell types. Secondly, changes in mRNA levels may not always result in a parallel alteration in protein levels. Moreover, changes in mRNA levels may be due to age-dependent changes in mRNA decay processes, such as the reported decline of mRNA stabilizing protein ARE-HuR with aging . However, the complete or partial prevention of the majority of the observed aging alterations by CR in heart and skeletal muscle suggests that gene expression patterns can be used to assess the biological age of the tissue under study, and to infer mechanisms of action of CR. Gene expression profiling may also prove valuable to a new area in CR research, the discovery of nutrient or drug interventions that mimic the calorie-restricted state .
Work in the laboratory of Dr. Prolla in the area of skeletal muscle and heart aging has been supported by National Institute of Health Grant R01AG18922.
(1997) Multi-organ characterization of mitochondrial genomic rearrangements in ad libitum and caloric restricted mice show striking somatic mitochondrial DNA rearrangements with age. Nucleic Acids Res 25:974–982.
(1980) NADH- and NADPH-dependent lipid peroxidation in bovine heart submitochondrial particles. Dependence on the rate of electron flow in the respiratory chain and an antioxidant role of ubiquinol. Biochem J 192:853–860.