Lifelong voluntary exercise in the mouse prevents age-related alterations in gene expression in the heart
A. M. Bronikowski1,7,
P. A. Carter2,
T. J. Morgan2,
T. Garland, Jr3,
N. Ung1,
T. D. Pugh4,
R. Weindruch4,6 and
T. A. Prolla5
1 Department of Zoology and Genetics, Iowa State University, Ames, Iowa 50011
2 School of Biological Sciences, Washington State University, Pullman, Washington 99164
3 Department of Biology, University of California, Riverside, California 92521
4 Geriatric Research, Education and Clinical Center, W. S. Middleton Memorial Veterans Administration Hospital, Madison 53706
5 Department of Genetics and Medical Genetics, University of Wisconsin, Madison, Wisconsin 53706
6 Department of Medicine, University of Wisconsin, Madison, Wisconsin 53706
7 Department of Zoology, University of Wisconsin, Madison, Wisconsin 53706
 |
ABSTRACT
|
---|
We present the first quantitative gene expression analysis of cardiac aging under conditions of sedentary and active lifestyles using high-density oligonucleotide arrays representing 11,904 cDNAs and expressed sequence tags (ESTs). With these data, we test the hypothesis that exercise attenuates the gene expression changes that normally occur in the aging heart. Male mice (Mus domesticus) were sampled from the 16th generation of selective breeding for high voluntary exercise. For the selective breeding protocol, breeders were chosen based on the maximum number of wheel revolutions run on days 5 and 6 of a test at 8 wk of age. For the colony sampled herein, mice were housed individually over their entire lifetimes (from weaning) either with or without access to running wheels. The hearts of these two treatment groups (active and sedentary) were assayed at middle age (20 mo) and old age (33 mo). Genes significantly affected by age in the hearts of the sedentary population by at least a 50% expression change (n = 137) were distributed across several major categories, including inflammatory response, stress response, signal transduction, and energy metabolism. Genes significantly affected by age in the active population were fewer (n = 62). Of the 42 changes in gene expression that were common to both treatment groups, 32 (72%) displayed smaller fold changes as a result of exercise. Thus exercise offset many age-related gene expression changes observed in the hearts of the sedentary animals. These results suggest that adaptive physiological mechanisms that are induced by exercise can retard many effects of aging on heart muscle at the transcriptional level.
aging; artificial selection; exercise; microarray; Mus domesticus; stress/inflammation response
 |
INTRODUCTION
|
---|
EXERCISE HAS BEEN SHOWN to improve overall health (e.g., increased cardiovascular performance; Ref. 5) and to offset the risk of age-related disease (2). However, in rodents, the impact of exercise on lifespan is primarily on median, not maximum lifespan (15, reviewed in Ref. 27). This observation suggests that although exercise is beneficial, it may not retard aging at the molecular level. In fact, there is a potential cost of exercise in terms of cellular oxidative damage, particularly in the heart (17). Recently developed techniques for high-throughput microarray analyses of transcripts have advanced the mechanistic study of aging and are promising tools for evolutionary studies of aging and rate-of-senescence interventions. Experiments have begun to reveal global patterns of age-related gene expression changes in thousands of genes in multiple tissues and the effects of genetic and nutritional interventions (e.g., 23). Exercise has been unequivocally associated with a slowing of age-specific mortality increases in rats, and thus with increased median lifespan (14, 15). However, because of increased oxygen consumption and metabolism during exercise, exercise is associated with an increase in the production of reactive oxygen species (ROS) and upregulation of antioxidant defenses during bouts of exercise in rodent skeletal muscle, heart, and liver (e.g., 17, 18). If this results in an increase in oxidative damage to cellular components, then exercise would be detrimental to health and lifespan. Alternatively, exercise may retard the aging process through assorted beneficial physiological consequences or by inducing an adaptive response that results in a net decrease in oxidative damage. In any case, the lack of molecular markers of the aging process has prevented an evaluation of the impact of this intervention at the molecular level of target tissues.
Here we present heart gene expression profiles from both active and sedentary middle- and old-aged house mice that have been selectively bred for 16 generations for high, early age voluntary exercise on running wheels. Because these mice run at relatively high levels for most of their lifespan, they represent a novel opportunity to study the effects of voluntary exercise. We ask: 1) What changes in gene expression occur late in life relative to mid-life in a sedentary lifestyle? 2) How does lifelong exercise affect these expression changes? and 3) What are the genes affected by age in the population of mice that have voluntarily exercised their entire lifetimes? We discuss our findings in relation to recent reports implicating insulin signaling and the stress and inflammation responses in the aging process and in relation to the survival patterns and exercise levels in these unique lines of mice.
 |
MATERIALS AND METHODS
|
---|
Mouse colony.
Mice in this report were members of a colony derived from a laboratory selective-breeding population designed to study the evolutionary correlates of exercise behavior (41). The selection colony was begun from outbred genetically heterogeneous Institute for Cancer Research mice (Mus domesticus) purchased from Harlan Sprague Dawley (Indianapolis, IN; Building 202, Barrier A). To study the health effects of voluntary wheel-running on lifetime measures of exercise, health, and survival, the population in this report was produced from second litters of the generation 15 breeders for generation 16 of the selection colony. At generation 15, the divergence between selection and control males in running behavior was 150% at 2 mo of age (i.e., the selection test age); control males averaged 4.5 km/day, whereas artificial selection males averaged 11.3 km/day (7). The artificial selection experiment maintains a replicated design. There are four closed selection lines and four closed control lines (i.e., randomly bred with respect to wheel running). Ten families are used to propagate each line. Five breeding pairs from each of the four selectively bred lines and four randomly bred lines were mated to produce 20 families from each selection history; details of aging colony design may be found in Refs. 8 and 32. Four experimental groups were initiated with four males and four females from each of these 40 families: selectively bred active (with wheels) and sedentary (without wheels), and control active and sedentary. Thus each of these four treatment groups contained 40 males and 40 females. In this report, we sampled male (nonsibling) mice from the selectively bred treatment groups (active and sedentary) at two ages.
In the colony sampled herein, housing with and without access to running wheels was begun when mice were 28 ± 3 days old in the mouse facility at Washington State University (WSU); mice were housed under the following conditions from weaning until natural death. Mice in the active group were placed individually (at weaning) in cages with a 10-cm radius running wheel and electronic wheel-revolution counter built into the cage top. The mouse thus had the option of voluntarily getting into the wheel and running. On the same day, mice designated to be in the sedentary group were placed individually in standard rodent cages of the same dimensions. Thus each "sedentary" mouse was free to move about its cage, but did not have access to a running wheel, whereas each "active" mouse was housed from weaning through death with access to a running wheel. Body mass and number of running revolutions were recorded weekly, the latter by an automated magnetized counter. Mice were checked daily; water and food (Harland Teklad rodent diet W 8604) were available ad libitum. Apparent food consumption was determined weekly by subtracting the amount eaten from the amount offered. Extra siblings from all 40 families were maintained as sentinel mice for specific pathogen screens. All screens were negative until an exposure to mouse hepatitis virus (MHV) at 24 mo post start date (through a barrier breakdown). No treatment was initiated for the virus. However, no necropsies indicated either MHV or liver damage as cause of death (performed by the veterinarians of Laboratory Animal Resource Center at WSU). Necropsies also included the examination of all other organs. Cage bottoms were cleaned once every 2 wk, wheels were cleaned once every 4 wk, and clean wheels were randomly assigned to active mice; this ensured no consistent effect on individual mice of any differences in mechanical resistance among wheels. Mean and maximum lifespan are reported in RESULTS.
Animals and tissues.
At 20 and 33 mo of age, eight male selectively bred mice were randomly (conditional on surviving to the sample age) chosen for death by decapitation; four sedentary males and four active males, one from each of the four replicate lines. Mice were decapitated, exsanguinated, and dissected; tissues and organs were weighed and placed immediately into a -80°C freezer (all protocols approved by WSU Animal Care and Use Committee). All organs were examined for overt disease; if an animal expressed tumors or other abnormalities, another animal was chosen. For the active old group, only three mice were used in this study based on this disease-free criterion. Both ventricles (left and right) were weighed and subjected to total RNA extraction using the guanidinium isothiocyanate method (TRIzol Reagent; Life Technologies, Grand Island, NY).
High-density oligonucleotide array hybridization.
Each mouse heart transcriptome was assayed individually, one array per mouse heart. Ten micrograms of total RNA was assayed from each mouse heart; thus the amount of heart tissue used per mouse varied among individuals (see Table 1). All mRNAs present in 10 µg of total RNA per individual were converted to double-stranded cDNA (Superscript Choice System, Life Technologies) and used as templates to synthesize biotin-labeled cRNA (T7 Megascript kit; Ambion, Austin, TX). Biotin-labeled cRNA was purified using RNeasy affinity columns (Qiagen, Valencia, CA). We hybridized cRNA to high-density mouse oligonucleotide arrays (Mouse Genome Array MU74Av2; Affymetrix, La Jolla, CA) as described (23). After hybridization, the gene arrays were washed and stained in a fluidic station (model 800101, Affymetrix) and scanned at a resolution of 6 µm with a Hewlett-Packard GeneArray scanner (model 900154, Affymetrix).
Preliminary data analysis by Affymetrix algorithms.
The Affymetrix Mouse Genome Array MU74Av2 was based on the cDNA sequences from the UniGene (8/96 and Build 4.0) and TIGR (Build 1.0 beta) databases. This mouse array contains 12,422 probe sets representing 11,904 known or putative genes. Approximately 16 probe pairs of oligonucleotide probes in a probe set (16 perfect match and 16 mismatch probes) are used to measure the transcript level of a gene. Each probe pair consists of a perfect match (PM) probe and a mismatch probe (MM), which allows direct subtraction of cross-hybridization signals after background subtraction. GeneChip Analysis Suite 3.3 was used to quantify the image data. Affymetrix software determines the presence of mRNA in samples and computes the signals of probe sets. The software calculates differences and ratios between perfect match and mismatch signals, which are representative of the hybridization levels of their targets in each probe set. The average of the differences between perfect match and mismatch signals (after removing the outliers beyond 3 standard deviations) is used to estimate the relative mRNA level of each transcript. Signals in each image are normalized by global scaling, in which the average signals of all probe sets in an image are scaled to the target average intensity by multiplying a scaling factor. We used the resulting variable, an individuals probe set signal intensity ("signal intensity") for each gene, as the dependent variable in all statistical analyses. Additionally, we report the fold change (FC) for genes determined to change significantly with age. As the probe set signal intensity is directly related to its expression level, the ratio of intensities between two images becomes the FC. To calculate FC between young and old mice, the following formula is used by the software
where SignalOld1 is the signal of a probe set for gene 1 from a mouse that was either an old animal in the sedentary or exercised groups, and SignalYoung1 is the signal of the same probe set, for the same gene 1 from a mouse in the corresponding middle-aged group, and Qfactor is a measure of the nonspecific fluorescent intensity background. Sixteen possible pairwise comparisons were performed (i.e., four middle-aged compared with four old-aged individuals) and their average was computed as a final FC for each treatment group.
Statistical analysis.
In this report, we focus on two comparisons: "sedentary aging," which comprises the gene expression changes with age in selection males housed without running wheels; and "exercise aging," which is the same but for selection males housed with access to running wheels. Statistical analyses were performed for each gene on each individuals calculated signal intensity. We used two approaches to ascertain probable signal in our data while attending to concerns about experiment-wise error rates for the analysis of thousands of expression variables. First, we based all of our analyses on an adjusted Students t statistic calculated for each gene. Adjusted posterior probabilities (in the Bayesian tradition) adjust the significance level based on regularized expressions for the variance of each gene (3). These post hoc t statistic adjustments of the log-transformed data were implemented in the software Cyber-T (25); the Bayesian prior was calculated with the 101 genes whose signal intensities were most similar to the gene in question. For example, if a genes signal intensity was 1,000, then the nearest 101 genes were used to obtain an expected variance for that gene. Genes that met the criterion of P < 0.01 were considered significant. However, to warrant consideration, a gene had to be expressed in all individuals in at least one treatment group. This limited the number of genes considered to n = 8,300.
Second, because some of these genes may have represented false-positive results, we queried our data for the false discovery rate (FDR) for different numbers of significant genes. This rate is experiment-specific and depends on the observed variability of each gene. FDR was calculated using the SAM Microsoft Excel add-in (45). By allowing a 10% FDR, an almost complete overlap was observed between this method and the conservative Bayesian method in the genes that were identified. Thus we report those genes implicated by the Bayesian t-test at P < 0.01 and note that based on the separate FDR analysis, the global FDR of these lists is 10%.
For the genes significantly affected by age in the sedentary population, we report the proportion by which exercise prevented this change in expression
where FC sedentary is the average fold change in old relative to young sedentary mice, and FC active is the same value in active mice. Values greater than one occurred when the FC in the active population was of opposite sign than in the sedentary population (e.g., induced in sedentary animals with age and downregulated in active animals with age). Values less than one occurred when exercise exacerbated the gene expression change (e.g., induced in sedentary animals with age, greater induction in active animals with age).
 |
RESULTS
|
---|
Survival and running phenotypes.
We present summary results for male selection mice: median (50% survival) and maximum (10% survival) lifespan, kilometers per day at ages 2 mo (i.e., the selective-breeding test age), 20 mo, and 33 mo. Selection males housed with running wheels (n = 40) had greater median lifespan than did those not housed with running wheels (n = 40) (50% survival, SD, SE = 698 days, 4.0 days, 2.2 days; and 599 days, 7 days, 2.1 days for active and sedentary selection males, respectively), but little difference in maximum lifespan (10% survival, SD, SE = 921 days, 21 days, 2 days; and 902 days, 11 days, 1.6 days for active and sedentary selection males, respectively) (Fig. 1). For selected male mice housed with running wheels, mean kilometers per day decreased over the lifespan from an average 6.2 to 4.8 to 0.6 km/day at 2 mo, 20 mo, and 33 mo of age. We note that selection males ran more than control males across the lifespan (control-to-selection differentials: 2 mo, 77%; 20 mo, 92%; 33 mo, 100%).

View larger version (14K):
[in this window]
[in a new window]
|
Fig. 1. Age-specific survival (lx) for males of the selected genetic background in two exercise treatments; active mice (n = 40) are indicated with a black line, sedentary mice (n = 40) are indicated with a gray line. The lx values were computed with the actuarial method in Program SAS, Procedure Lifetest. Median lifespan (50% survival) and maximum lifespan (10%) survival are indicated.
|
|
In an analysis of variance of body mass testing for the effects of age, exercise, and their interaction, only age was significant; body mass was smaller in older than in younger animals (F1,11 = 6.29, P = 0.03). When heart mass was tested for the same effects and included body mass as a covariate, only body mass was a significant predictor of heart mass (larger animals had smaller hearts, F1,11 = 6.16, P = 0.03). Finally, in an analysis of variance testing the same effects and including heart mass as a covariate, only heart mass was a significant predictor of amount of total RNA; larger hearts yielded less RNA than smaller hearts (F1,11 = 7.11, P = 0.02). It therefore follows that younger animals yielded more total RNA than older animals. Note however, that regardless of total yield, 10 µg total RNA was used from each animal.
General patterns of gene expression in the aging hearts of sedentary mice.
Of the 8,300 genes that were expressed in all members of at least one of the four experimental groups, 137 (1.6%) were significantly changed in the old relative to young mice in the sedentary population (P < 0.01) and were up- or downregulated by at least 50% (i.e., ± |1.5|-fold change). Most of these 137 genes could be classified into broad functional groups, although 36 code putative proteins and are as yet unnamed and of unknown function. Genes classified under the inflammatory, signal transduction, stress, and energy metabolism categories are listed in Tables 2 (increased expression with age) and 3 (decreased expression with age). For the complete list of significantly altered genes by at least 50% with age, see Supplemental Tables 7 and 8 (Supplemental Tables 710 are available online at the Physiological Genomics web site).1
View this table:
[in this window]
[in a new window]
|
Table 2. Genes significantly increased (P < 0.01) in expression with age in the sedentary population (inflammatory response, stress response, signal transduction, and energy metabolism)
|
|
View this table:
[in this window]
[in a new window]
|
Table 3. Genes significantly decreased (P < 0.01) in expression with age in the sedentary population (inflammatory response, stress response, signal transduction, and energy metabolism)
|
|
Lifelong voluntary exercise prevents many age-related changes in gene expression.
Of the 137 genes whose expression was significantly different in old vs. young sedentary mice, 70 were attenuated by exercise by at least 50%, i.e., had exercise prevention proportions of at least 0.50. In addition, six expression changes were exacerbated by exercise, and 28 were reversed in the active population (see Supplemental Tables 7 and 8). Furthermore, in the active population, fewer genes were significantly affected by age by at least a 50% change in expression level (n = 62) than in the sedentary population. The effect of exercise on large alterations in gene expression was particularly striking given that we observed only one gene, atrial natriuretic factor (ANF), displaying a greater than 10-fold increase in expression in the exercised animals compared with six genes displaying such changes in the sedentary animals. Most of the 62 genes could be classified into several functional groups, but 13 were of unknown function. Genes classified under inflammatory, stress, signal transduction, and energy metabolism categories are listed in Tables 4 and 5. For the complete list of all genes significantly up- or downregulated by at least 50% with age in the active population, see Supplemental Tables 9 and 10.
View this table:
[in this window]
[in a new window]
|
Table 4. Genes significantly increased (P < 0.01) in expression with age in the exercised population (inflammatory response, stress response, signal transduction, and energy metabolism)
|
|
View this table:
[in this window]
[in a new window]
|
Table 5. Genes significantly decreased (P < 0.01) in expression with age in the exercised population [inflammatory response, stress response (none), signal transduction, and energy metabolism]
|
|
As another measure of the effect of exercise, we investigated genes that were either upregulated or downregulated as a result of aging in both the exercised and sedentary populations. We included in this analysis all genes that had an age change in expression that was statistically significant at P < 0.05 (Bayesian adjusted) in both the exercised and sedentary populations and which also displayed a similar overall trend with respect to induction or repression. These genes represent the most statistically robust alterations in gene expression as a function of age because they were identified as significant in two independent gene expression analyses. Therefore, they represent a useful set to evaluate the differential impact of aging on the sedentary and exercised populations. We identified 42 genes that met these criteria, 27 representing upregulations in mRNA level and 15 representing downregulation in mRNA levels (Table 6). We then conducted a slope heterogeneity test to determine whether the interaction between age and exercise was significant. In practice, this test measures differences in the signal-intensity changes in the two activity treatments (active vs. sedentary). Of the 42 genes examined, only three showed a statistical difference at P < 0.05. However, three-factor analysis of variance, which tests for the interaction of age-by-gene-by-exercise, suggested that gene expression changes in relation to aging are different in the exercised and sedentary groups when the population of genes is analyzed as a whole (F41,462 = 1.36, P < 0.07). This is supported by the fact that the overall effect of exercise on age-related gene expression changes was 60%, i.e., a global exercise prevention proportion of 0.60. Additionally, of the 42 changes in gene expression that were common to both activity groups, 32 (72%) displayed smaller FCs as a result of exercise. When taken as a whole, our data support the conclusion that exercise retards the transcriptional alterations associated with age at two levels: a reduction in the number of genes displaying statistically significant changes and a reduction in fold changes for those genes.
 |
DISCUSSION
|
---|
We undertook this study to ask: 1) How does aging affect gene expression profiles in sedentary and exercising populations?; and ultimately to ask 2) Can exercise attenuate age-related phenotypes at the transcriptional level? Using lines of genetically heterogeneous mice, selectively bred for high voluntary exercise, we found the majority of genes affected by age in our sedentary population to belong to the general categories of inflammation and stress response, and cell signaling (with 10% FDR). We further found that many of these transcription-level changes were offset by exercise in our active population. Additionally, the active population was characterized by fewer age-related expression changes overall. These results suggest that lifelong exercise can retard cardiac aging at the transcriptional level. We note that posttranscriptional events may negate subsequent effects on protein levels, and the reactions they catalyze, and caution is required in interpreting the biological implications of changes in gene expression. We first discuss our results relative to the survival and running phenotypes of these mice, and second, relative to recent reports implicating each of inflammatory response, stress response, and cell signaling in the aging process.
Median lifespan was lengthened as a result of lifelong exercise.Our data provide further support to the previous observations that exercise increases mean, but not maximum, lifespan in rodents. Median lifespan was increased by 17% (
100 days) in our wheel-access selectively bred male mice, but maximum lifespan was unaffected (Fig. 1). These results are in agreement with work by Holloszy (14, 15; see also Ref. 31) and McCarter and colleagues (27, 28) who have also demonstrated extension of median, but not maximum, lifespan in rodents with access to running wheels.
Why does exercise not extend maximum lifespan if it appears to retard aging at the molecular level as indicated by the gene expression analysis? Most likely, aging retardation at the molecular level by exercise is not observed in all tissues, including some that may limit lifespan. For example, if exercise does not reduce aging rates in replicative tissues, then it will not retard age-related tumor onset, which tends to limit maximum lifespan. Another possibility relates to the observation that wheel running decreased to an average 680 m/day at 33 mo of age vs. >6,000 m/day at 2 mo of age. This minimal level of late-age running may explain the lack of maximum lifespan extension if the benefits of exercise are short-term and thus require habitual high exercise. That more genes were affected with age in the sedentary population relative to the active population suggests that exercise affords protection in the heart muscle against normal age-related expression changes. Sedentary mice, while able to move about and conduct normal home-cage activity, presumably were unable to obtain cardiovascular conditioning, which has been shown to prolong average lifespan and lead to physiological adaptations in antioxidant defenses (18) and to decreased oxidative stress in heart and liver (22).
Aging is associated with inflammatory and stress responses in the heart. The immune/inflammation response has been shown to become less effective in fighting infectious disease and injury with increasing age (reviewed in Ref. 30). Furthermore, long-term exercise training may counteract age-related declines in immune function (e.g., 33, 46). The major transcriptional class induced as a result of the aging process in sedentary mice was the inflammatory response (21 genes). This class included a concerted induction of complement genes, which are involved in innate immunity, including the genes encoding the component 1, q subunits, C1qb and C1qc, and Complement C4. The low level of induction of this system in our exercising population suggests less inflammation to heart muscle with exercise. Lee et al. (23) reported an induction of the complement cascade in the aging brain of mice, which, similar to exercise in our experiment, was offset by caloric restriction, an intervention that retards the aging process and extends lifespan in rodents (49). Previously, McGeer and colleagues (50) demonstrated that myocardial tissue locally expresses complement and that this expression significantly increases in response to ischemia and reperfusion. Other inflammatory response genes induced in the sedentary population, but not in the active population, include genes involved in B and T cell functioning and genes involved in the modulation of inflammation (Table 2). Although these same categories were induced in the active population with age, the greatest induction was by 180% in active mice (a total of six inflammatory response genes were upregulated 100% or more) vs. an almost 3,800% increase in the sedentary population (where 18 inflammatory response genes were upregulated by 100% or more).
Empirical evidence has been rapidly accumulating in several model systems that specifically implicates stress resistance as a factor in aging either through decreased efficiency of the stress response with age or from altered stress resistance at particular ages [e.g., rhesus monkey (20), mouse (23, 29), nematode (19), and fruit fly (24, 34)]. In our study, aging was associated with the induction of a stress response, including the expression of the antioxidant enzymes glutathione peroxidase, extracellular superoxide dismutase (EC-SOD), and metallothionein I. Also induced were two genes previously associated with oxidative stress responses, the apolipoproteins ApoD and ApoE. ApoE induction is associated with Alzheimers disease, and ApoE-deficient animals display increased tissue levels of isoprostanes (38) and 3-nitrotyrosine (26), both of which are markers of oxidative stress. ApoD is constitutively expressed in multiple tissues (39), whereas both ApoE and ApoD are dramatically induced in response to denervation injury (6). The induction of these genes in the heart has not been previously reported, but our results suggest that they may be involved in the cardiac response to aging, perhaps in association with damage to the sympathetic neurons that innervate the heart or because of oxidative stress to cardiomyocytes. Taken as a whole, our observations suggest that the senescent heart is under a proinflammatory state associated with oxidative stress.
Aging in the heart is associated with alterations in expression of genes involved in hypertrophy and bioenergetics. In general, the aged heart displays increases in extracellular matrix (ECM) protein deposition (9), fibrosis (48), and cardiomyocyte hypertrophy (35). Although heart mass did not differ between active and sedentary mice, sedentary aged mice exhibited a 3.8-fold increase in the expression of the gene encoding atrial natriuretic factor (ANF), a peptide hormone of cardiac origin that participates in the homeostatic control of intravascular volume and vascular tone, and that is elevated in congestive heart failure (CHF) (37). Aging also resulted in the induction of myosin light chain type 2 (MLC2V), a gene previously identified as induced in idiopathic dilated cardiomyopathy (IDCM) (13). These observations suggest commonality of pathways involved in CHF, IDCM, and the aging process. Possibly, these alterations result from age-associated alterations in cardiomyocyte bioenergetics, as suggested by upregulation in the mitochondrial electron transport system (ETS) genes cytochrome oxidase (COX) Va and COX VIa, and lactate dehydrogenase. Eukaryotic COX is a multicomponent enzyme consisting of 13 polypeptides in mammals and it is believed that subunits VIII and VIa may function to modulate enzyme activity in response to changes in metabolic conditions (47). We also observed decreases in expression of genes involved in fatty acid metabolism, such as
-methylacyl-CoA racemase, involved in peroxisomal fatty acid
-oxidation, and heart fatty-acid binding protein (HFABP), which functions as a vehicle of cytosolic fatty acid transport. Interestingly, we also observed a decrease in expression of the F1F0-ATP synthase E chain, which is regulated in response to fatty acid intake (42). These gene expression results are in agreement with the previous observation of impaired cardiac mitochondrial fatty acid oxidation with aging (36).
A reduction in levels of homologues of insulin, insulin-like growth factors (IGFs), and receptors in the insulin-signaling pathway has been shown to confer greater longevity in yeast (12, 16), nematodes (21, 44), fruit flies (10, 43), mutant long-lived mice (4, 11), and caloric-restricted mice (40). Therefore, the as-yet unidentified mechanism of insulin signaling on lifespan may be evolutionarily conserved. A primary mammalian insulin pathway ligand, IGF-I, as well as insulin-related receptor (IRR) and insulin-activated AAAT (IaAAAT) were upregulated in sedentary old mice relative to young mice. IGF-I and IRR inductions were almost completely offset in the active population (96% and 95%, respectively), whereas IaAAAT was reversed, which is consistent with the observation of increased median lifespan in active mice. That genes involved in insulin signaling were not significantly reduced in expression in old active mice correlates with the observation that exercise training eliminates age-related differences in skeletal muscle insulin receptor and insulin receptor substrate 1 (IRS-1) abundance in rats (1).
We have presented the first large-scale assay of transcriptional changes in aging mouse hearts and their modulation by exercise. We found more genes to be significantly affected by age in our sedentary population of mice than in our active population of mice. Moreover, many alterations in gene expression affected by age in the sedentary population were attenuated by exercise. We also observed a striking effect of exercise on large alterations in gene expression, as we observed only one gene displaying greater than a 10-fold increase in expression in the exercised animals compared with six genes displaying such changes in the sedentary animals. Our results indicate a global, overall decrease in the induction of inflammatory and stress responses in the presence of a lifetime of active exercise, despite decreased exercise levels in the very old exercising mice. Further studies specifically targeting the genes implicated in this study as well as genes involved in insulin signaling will further our understanding of how habitual physical activity promotes extended life expectancy at the transcriptional level. Additionally, investigations of other tissues in the exercising and sedentary populations should reveal whether the benefits of exercise in aging retardation affect multiple tissues, a finding that would be consistent with the increase in median lifespan induced by exercise.
 |
ACKNOWLEDGMENTS
|
---|
We are grateful to S. Hall, S. Kane, S. Thomson, L. Jenkins, M. Baze, B. Irwin, M. Schmit, A. Poopatanapong, J. Robertson, F. Muller, and D. Baker for the many hours of assistance with colony care and data collection; and to the veterinarians of the Laboratory Animal Resources Committee at WSU for technical assistance. We thank Rich Miller for statistical advice and critical comments on an earlier version of this manuscript.
This research was supported by grants from the WSU College of Sciences (to P. A. Carter); by the National Science Foundation Grants DEB-0083638 (to P. A. Carter), DEB-0105079 (to P. A. Carter and T. J. Morgan), and IBN-0212567 (to T. Garland, Jr.); and by National Institutes of Health Grants AG-18922 to (to T. A. Prolla and R. Weindruch) and AG-05784 (to A. M. Bronikowski).
 |
FOOTNOTES
|
---|
Article published online before print. See web site for date of publication (http://physiolgenomics.physiology.org).
Address for reprint requests and other correspondence: A. M. Bronikowski, Dept. of Zoology and Genetics, Iowa State Univ., Ames, IA 50011 (E-mail: abroniko{at}iastate.edu).
10.1152/physiolgenomics.00082.2002.
1 The Supplemental Material (Tables 710) to this article is available online at http://physiolgenomics.physiology.org/cgi/content/full/12/2/129/DC1. 
 |
References
|
---|
- Arias EB, Gosselin LE, and Cartee GD. Exercise training eliminates age-related differences in skeletal muscle insulin receptor and IRS-1 abundance in rats. J Gerontol A Biol Sci Med Sci 56: B444B450, 2001.
- Astrand PO. Physical activity and fitness. Am J Clin Nutr 50: 1231S1236S, 1992.
- Baldi P and Long AD. A Bayesian framework for the analysis of microarray expression data: regularized t-test and statistical inferences of gene changes. Bioinformatics 17: 509519, 2001.[Abstract/Free Full Text]
- Bartke A, Brown-Borg H, Mattison J, Kinney B, Hauck S, and Wright C. Prolonged longevity of hypopituitary dwarf mice. Exp Gerontol 36: 2128, 2001.[ISI][Medline]
- Blair SN, Kohl HW III, Barlow CE, Paffenbarger RS Jr, Gibbons LW, and Macera CA. Changes in physical fitness and all-cause mortality. A prospective study of healthy and unhealthy men. JAMA 273: 10931098, 1995.[Abstract]
- Boyles JK, Notterpek LM, and Anderson LJ. Accumulation of apolipoproteins in the regenerating and remyelinating mammalian peripheral nerve. J Biol Chem 265: 1780517815, 1990.[Abstract/Free Full Text]
- Bronikowski AM, Carter PA, Swallow JG, Girard IA, Rhodes JS, and Garland T Jr. Open-field behavior of house mice selectively bred for high voluntary wheel running. Behav Genet 31: 309316, 2001.[ISI][Medline]
- Bronikowski AM, Morgan TJ, Garland T Jr, and Carter PA. Antioxidant gene expression in active and sedentary house mice (Mus domesticus) selected for high voluntary wheel-running behavior. Genetics 161: 17631769, 2002.[Abstract/Free Full Text]
- Burgess ML, McCrea JD, and Hedrick HL. Age-associated changes in cardiac matrix and integrins. Mech Ageing Dev 122: 17391756, 2001.[ISI][Medline]
- Clancy DJ, Gems D, Harshman LG, Oldham S, Stocker H, Hafen E, Leevers SJ, and Partridge L. Extension of life-span by loss of CHICO, a Drosophila insulin receptor substrate protein. Nature 292: 104106, 2001.
- Dozmorov I, Bartke A, and Miller RA. Array-based expression analysis of mouse liver genes: effect of age and of the longevity mutant Prop1df. J Gerontol A Biol Sci Med Sci 56: B72B80, 2001.[Abstract/Free Full Text]
- Fabrizio P, Pozza F, Pletcher SD, Gendron CM, and Longo VD. Regulation of longevity and stress resistance by Sch9 in Yeast. Science 292: 288290, 2001.[Abstract/Free Full Text]
- Haase D, Lehmann MH, Korner MM, Korfer R, Sigusch HH, and Figulla HR. Identification and validation of selective upregulation of ventricular myosin light chain type 2 mRNA in idiopathic dilated cardiomyopathy. Eur J Heart Fail 4: 2331, 2002.[ISI][Medline]
- Holloszy JO. Exercise and longevity: studies on rats. J Gerontol 43: B149B151, 1988.[ISI][Medline]
- Holloszy JO. Exercise increases average longevity of female rats despite increased food intake and no growth retardation. J Gerontol 48: B97B100, 1993.[ISI][Medline]
- Imai SI, Armstrong CM, Kaeberlein M, and Guarente L. Transcriptional silencing and longevity protein SIR2 is an NAD-dependent histone deacetylase. Nature 403: 795799, 2000.[ISI][Medline]
- Ji LL. Exercise-induced oxidative stress in the heart. In: Handbook of Oxidants and Antioxidants in Exercise, edited by Sen CK, Packer L, and Hänninen OOP. Amsterdam: Elsevier, 2000, p. 689712.
- Ji LL, Leeuwenburgh C, Leichtweis S, Fiebig R, Hollander J, Gore M, and Bejma J. Oxidative stress and aging: role of exercise and its influences on antioxidant systems. Ann NY Acad Sci 845: 102117, 1998.
- Johnson TE, Cypser J, de Castro E, de Castro S, Henderson S, Murakami S, Rikke B, Tedesco P, and Link C. Gerontogenes mediate health and longevity in nematodes through increasing resistance to environmental toxins and stressors. Exp Gerontol 35: 687694, 2000.[ISI][Medline]
- Kayo T, Allison DB, Weindruch R, and Prolla TA. Influences of aging and caloric restriction on the transcriptional profile of skeletal muscle from rhesus monkeys. Proc Natl Acad Sci USA 98: 50935098, 2001.[Abstract/Free Full Text]
- Kenyon C. A conserved regulatory system for aging. Cell 105: 165168, 2001.[ISI][Medline]
- Kim JD, Yu BP, McCarter RJ, Lee SY, and Herlihy JT. Exercise and diet modulate cardiac lipid peroxidation and antioxidant defenses. Free Radic Biol Med 20: 8388, 1996.[ISI][Medline]
- Lee CK, Weindruch R, and Prolla TA. Gene-expression profile of the ageing brain in mice. Nat Genet 25: 294297, 2000.[ISI][Medline]
- Lin YJ, Seroude L, and Benzer S. Extended life-span and stress resistance in the Drosophila mutant methuselah. Science 282: 943946, 1998.[Abstract/Free Full Text]
- Long AD. Cyber-T software [Online]. Functional Genomics. Univ. of California, Irvine, 2001. http://www.genomics.uci.edu/software.html (updated 6 Dec. 2000).
- Matthews RT and Beal MF. Increased 3-nitrotyrosine in brains of Apo E-deficient mice. Brain Res 718: 181184, 1996.[ISI][Medline]
- McCarter RJM. Caloric restriction, exercise and aging. In: Handbook of Oxidants and Antioxidants in Exercise, edited by Sen CK, Packer L, and Hänninen OOP. Amsterdam: Elsevier, 2000, p. 797829..
- McCarter R, Shimokawa I, Ikeno Y, Higami Y, Hubbard G, Yu BP, and McMahan CA. Physical activity as a factor in the action of dietary restriction on aging: effects in Fischer 344 rats. Aging (Milano) 9: 7379, 1997.
- Migliaccio E, Giorgio M, Mele S, Pelicci G, Reboldi P, Pandolfi PP, Lanfrancone L, and Pelicci PG. The p66shc adaptor protein controls oxidative stress response and life span in mammals. Nature 402: 309313, 1999.[ISI][Medline]
- Miller RA. Aging and the immune response. In: Handbook of the Biology of Aging, edited by Schneider EL and Rowe JW. San Diego, CA: Academic, 1990, p. 157180.
- Mlekusch W, Tillian H, Lamprecht M, Trutnovsky H, Horejsi R, and Reibnegger G. The effect of reduced physical activity on longevity of mice. Mech Ageing Dev 88: 159168, 1996.[ISI][Medline]
- Morgan TJ, Garland T Jr, and Carter PA. Ontogenies in mice selected for high voluntary wheel-running activity. I. Mean ontogenies. Evolution In press.
- Nash MS. Exercise and immunology. Med Sci Sports Exerc 26: 125127, 1994.[ISI][Medline]
- Novoseltsev VN, Novoseltseva J, and Yashin AI. A homeostatic model of oxidative damage explains paradoxes observed in earlier aging experiments: a fusion and extension of older theories of aging. Biogerontology 2: 127138, 2001.[ISI][Medline]
- Olivetti G, Melissari M, Capasso JM, and Anversa P. Cardiomyopathy of the aging human heart. Myocyte loss and reactive cellular hypertrophy. Circ Res 68: 15601568, 1991.[Abstract]
- Paradies G, Ruggiero FM, Petrosillo G, Gadaleta MN, and Quagliariello E. Carnitine-acylcarnitine translocase activity in cardiac mitochondria from aged rats: the effect of acetyl-L-carnitine. Mech Ageing Dev 84: 103112, 1995.[ISI][Medline]
- Perrella MA, Schwab TR, OMurchu B, Redfield MM, Wei CM, Edwards BS, and Burnett JC Jr. Cardiac atrial natriuretic factor during evolution of congestive heart failure. Am J Physiol Heart Circ Physiol 262: H1248H1250, 1992.[Abstract/Free Full Text]
- Pratico D, Tangirala RK, Rader DJ, Rokach J, and Fitzgerald GA. Vitamin E suppresses isoprostane generation in vivo and reduces atherosclerosis in ApoE-deficient mice. Nat Med 4: 11891192, 1998.[ISI][Medline]
- Seguin D, Desforges M, and Rassart E. Molecular characterization and differential mRNA tissue distribution of mouse apolipoprotein D. Brain Res Mol Brain Res 30: 242250, 1995.[ISI][Medline]
- Shelley CX, Dhahbi JM, Mote PL, and Spindler SR. Genomic profiling of short- and long-term caloric restriction effects in the liver of aging mice. Proc Natl Acad Sci USA 98: 1063010635, 2001.[Abstract/Free Full Text]
- Swallow JG, Carter PA, and Garland T Jr. Artificial selection for increased wheel-running behavior in house mice. Behav Genet 28: 227237, 1998.[ISI][Medline]
- Swartz DA, Park EI, Visek WJ, and Kaput J. The e subunit gene of murine F1F0-ATP synthase. Genomic sequence, chromosomal mapping, and diet regulation. J Biol Chem 271: 2094220948, 1996.[Abstract/Free Full Text]
- Tatar M, Kopelman A, Epstein D, Tu MP, Yin CM, and Garofalo RS. A mutant Drosophila insulin receptor homolog that extends life-span and impairs neuroendocrine function. Science 292: 107110, 2001.[Abstract/Free Full Text]
- Tissenbaum HA and Guarente L. Increased dosage of a sir-2 gene extends lifespan in Caenorhabditis elegans. Nature 410: 227230, 2001.[ISI][Medline]
- Tusher VG, Tibshirani R, and Chu G. Significance analysis of microarrays applied to the ionizing radiation response. Proc Natl Acad Sci USA 98: 51165121, 2001.[Abstract/Free Full Text]
- Utsuyama M, Ichikawa M, Konno-Shiradawa A, Fujita Y, and Hirokawa K. Retardation of the age-associated decline of immune functions in aging rats under dietary restriction and daily physical exercise. Mech Ageing Dev 91: 219228, 1996.[ISI][Medline]
- Wan B and Moreadith RW. Structural characterization and regulatory element analysis of the heart isoform of cytochrome c oxidase VIa. J Biol Chem 270: 2643326440, 1995.[Abstract/Free Full Text]
- Wanagat J, Wolff MR, and Aiken JM. Age-associated changes in function, structure and mitochondrial genetic and enzymatic abnormalities in the Fischer 344 x Brown Norway F1 hybrid rat heart. J Mol Cell Cardiol 34: 1728, 2002.[ISI][Medline]
- Weindruch R and Walford RL. The Retardation of Aging and Disease by Dietary Restriction. Springfield, IL: Charles C. Thomas, 1988.
- Yasojima K, Kilgore KS, Washington RA, Lucchesi BR, and McGeer PL. Complement gene expression by rabbit heart: upregulation by ischemia and reperfusion. Circ Res 82: 12241230, 1998.[Abstract/Free Full Text]