Author + information
- Received February 7, 2008
- Revision received March 21, 2008
- Accepted April 14, 2008
- Published online August 5, 2008.
- Christian Werner, MD⁎,
- Milad Hanhoun, MD⁎,
- Thomas Widmann, MD†,
- Andrey Kazakov, MD⁎,
- Alexander Semenov, MD⁎,
- Janine Pöss, MD⁎,
- Johann Bauersachs, MD‡,
- Thomas Thum, MD‡,
- Michael Pfreundschuh, MD†,
- Patrick Müller, MD⁎,
- Judith Haendeler, MD§,
- Michael Böhm, MD⁎ and
- Ulrich Laufs, MD⁎,⁎ ()
- ↵⁎Reprint requests and correspondence:
Dr. Ulrich Laufs, Klinik für Innere Medizin III, Kardiologie, Angiologie und Internistische Intensivmedizin, Universitätsklinikum des Saarlandes, 66424 Homburg/Saar, Germany.
Objectives The purpose of this study was to study the underlying molecular mechanisms of the protective cardiac effects of physical exercise.
Background Telomere-regulating proteins affect cellular senescence, survival, and regeneration.
Methods C57/Bl6 wild-type, endothelial nitric oxide synthase (eNOS)–deficient and telomerase reverse transcriptase (TERT)–deficient mice were randomized to voluntary running or no running wheel conditions (n = 8 to 12 per group).
Results Short-term running (21 days) up-regulated cardiac telomerase activity to >2-fold of sedentary controls, increased protein expression of TERT and telomere repeat binding factor (TRF) 2, and reduced expression of the proapoptotic mediators cell-cycle–checkpoint kinase 2 (Chk2), p53, and p16. Myocardial and leukocyte telomere length did not differ between 3-week- and 6-month-old sedentary or running mice, but telomerase activity, TRF2 and TERT expression were persistently increased after 6 months and the expression of Chk2, p53, and p16 remained down-regulated. The exercise-induced changes were absent in both TERT−/− and eNOS−/− mice. Running increased cardiac expression of insulin-like growth factor (IGF)-1. Treatment with IGF-1 up-regulated myocardial telomerase activity >14-fold and increased the expression of phosphorylated Akt protein kinase and phosphorylated eNOS. To test the physiologic relevance of these exercise-mediated prosurvival pathways, apoptotic cardiomyopathy was induced by treatment with doxorubicin. Up-regulation of telomere-stabilizing proteins by physical exercise in mice reduced doxorubicin-induced p53 expression and potently prevented cardiomyocyte apoptosis in wild-type, but not in TERT−/− mice.
Conclusions Long- and short-term voluntary physical exercise up-regulates cardiac telomere-stabilizing proteins and thereby induces antisenescent and protective effects, for example, to prevent doxorubicin-induced cardiomyopathy. These beneficial cardiac effects are mediated by TERT, eNOS, and IGF-1.
Regular physical activity is associated with a decrease of cardiovascular events (1,2). Physical training improves exercise capacity, endothelial function, and collateralization in patients with coronary artery disease and chronic heart failure (3–5). Physical activity is associated with improved blood pressure, insulin sensitivity, mood, body weight, and hemostatic and inflammatory variables (6). However, despite the wealth of evidence derived from epidemiological and interventional trials, there is limited understanding of the underlying molecular mechanisms, especially in the heart (7).
The prevalence, incidence, and morbidity of cardiac diseases increase with the aging human population. On the cellular level, senescence, chromosome stability, and cell viability are regulated by the telomeres and their associated proteins, deoxyribonucleic acid-protein complexes located at both ends of eukaryotic chromosomes (8). Shortening of the telomeres has been shown to be associated with increased mortality rate from heart disease (9). In men and in mice, a ribonuclear complex, called telomerase, is the central component of the telomere complex. The activity of telomerase decreases in aging mouse myocytes (10). The best-characterized function of the telomeric complex is to protect the chromosome ends from degradation. The enzyme telomerase contains a catalytic subunit, the telomerase reverse transcriptase (TERT). In isolated cells, TERT gene transfer reduces replicative senescence and extends the life span of numerous cell types, including cardiac myocytes (11). Important proteins that form the telomeric complex include the telomere repeat binding factors (TRFs) 1 and 2, which can directly bind to the TTAGGG repeats of telomeres. The TRF2 protein is crucial for maintaining a normal chromosomal end structure because it collaborates in the processes leading to T-loops (12). In addition, TRF2 interacts with other factors, serving as binding platforms for a number of additional telomere-associated proteins. The function of these protein complexes is not completely known but includes signaling to deoxyribonucleic acid damage checkpoint controls (13,14). In addition to the protection of chromosome ends, the components of the telomeric complex enhance cell survival. Suppression of telomerase enzyme activity promotes apoptosis, whereas overexpression of TERT prevents programmed death by interfering with a pre-mitochondrial step in the cell death cascade (11,15). Gene products implicated in growth arrest and apoptosis, such as p53, p16, and cell-cycle–checkpoint kinase 2 (Chk2), are potential downstream effectors of the telomeric complex and increase with age in cardiac myocytes (10,13,14,16–19).
Cardiomyocyte apoptosis has been shown to contribute significantly to the pathogenesis of heart failure. In end-stage human heart failure, myocyte apoptosis increases (20). Similarly, acute toxic cardiomyopathy, such as that induced by doxorubicin (a frequently used anticancer drug), is characterized by cellular senescence and apoptosis (21). However, our understanding of potential strategies to prevent cardiomyocyte senescence and apoptosis is limited (20). Here we identify and characterize voluntary exercise as a potent measure to increase telomere-stabilizing proteins in the myocardium and to prevent doxorubicin-induced cardiac myocyte apoptosis.
Animals and exercising
Animal experiments were approved by the animal ethics committee of the Universität des Saarlandes, conformed with U.S. National Institutes of Health's Guide for the Use and Care of Laboratory Animals (NIH Pub. No. 85-23, revised 1996), and were conducted in accordance to institutional guidelines. Eight-week-old male C57/Bl6 (Charles River Laboratories, Wilmington, Massachusetts), endothelial nitric oxide synthase (eNOS)−/− (B6.129/P2-Nos3, Charles River) mice, TERT−/− (B6.129S-Terttm1Yjc/J, mutant generation 2, Jackson Laboratory, Bar Harbor, Maine) mice, and strain-matched controls to the mutants were kept under usual care at 1 to 6 mice per cage. Each exercising mouse was kept in an individual cage supplied with a running wheel (12.8-cm diameter) equipped with a tachometer (BC 500, Sigma Sport Europe, Neustadt, Germany) recording the daily running distance. Mice ran voluntarily. The mean running distance was 5,100 ± 800 m/24 h and did not differ in mice after doxorubicin treatment as well as in eNOS−/− and TERT−/− mice. Transthoracic echocardiography was performed using a GE Systems Vivid 5 scanner (GE Healthcare, Munich, Germany), 13-MHz broadband transducer; fractional shortening, end-diastolic thickness of interventricular septum and the left posterior wall, as well as the left ventricular end-diastolic diameter was taken. Indicated mice were treated with doxorubicin 22.5 mg/kg (Medac, Wedel, Germany), administered intraperitoneally (i.p.) for 24 h; with growth hormone (GH) (Sigma-Aldrich, Munich, Germany), 2.5 μg GH/g/day solved in 150 μl phosphate-buffered saline (PBS) once per day i.p. for 7 days; with mouse insulin-like growth factor (IGF)-1 (Sigma-Aldrich), 1.5 μg mouse IGF-1/g body weight solved in 150 μl PBS for 3 times per day i.p. for 2 days; or with 150 μl PBS as control as described (22).
Telomere length analysis by flow-fluorescence in-situ hybridization (FISH) assays
Telomere length was determined by the FISH method (23). In brief, 600,000 peripheral blood leukocytes were washed once (5% dextrose, 0.1% bovine serum albumin [BSA], 10 mmol/l 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid) and resuspended in hybridization buffer (75% deionized formamide, 20 mmol/l tris(hydroxymethyl)aminomethane [Tris] [pH 7.1], 20 mmol/l NaCl, 1% BSA) for 10 min. After denaturing cells, 90 min incubation with either no probe (unstained control) or 0.18 μg fluorescein isothiocyanate (FITC)–labeled, telomere-specific (C3TA2)3 deoxyribonucleic acid probe (Applied Biosystems, Langen, Germany) was performed. After 3 rounds of washing (75% deionized formamide, 0.1% BSA, 10 mmol/l Tris, 0.1% Tween 20), cells were resuspended in PBS, 0.1% BSA, RNAse A at 10 μg/ml and 0.1 μg/ml LDS 751 (Exciton, Munich, Germany) for deoxyribonucleic acid (DNA) counterstain. Flow cytometric analysis was carried out on a FACSCanto (Becton Dickinson, Heidelberg, Germany). Telomere length was expressed as mean fluorescent signal intensity. To control for interday variation, FITC-labeled beads (Quantum 24, Caltag, Hamburg, Germany) with defined amounts of fluorescence (molecules of equivalent soluble fluorochrome) were run in parallel each day. Telomere length was converted into base pairs according to a previously established standard curve (23).
Myocardial telomere length analysis by FISH
Telomere length is directly related to its integrated fluorescence intensity when marked with a specific telomere peptide nucleic acid probe using quantitative FISH as described (24). We fixed 5-μm cryosections of the hearts in 4% paraformaldehyde for 30 min at room temperature followed by treatment with 0.1% Pepsine solution (Sigma-Aldrich) for 10 min at 37°C. After washing and dehydration, 25 μl of the hybridization mix (70% formamide, 0.13 μg Cy3-conjugated telomere-specific peptide nucleic acid probe (C2TA3)3 (Applied Biosystems), blocking reagent (Roche, Mannheim, Germany), MgCl2 buffer, and 1 mmol/l Tris (pH 7.2) was added to the sections. Denaturation was carried out at 80°C for 3 min and preceded incubation for 2 h in a humid chamber and washing with wash buffer (70% formamide, 10 mmol/l Tris [pH 7.2], 0.1% BSA), Tris-buffered saline (+1% Tween) and PBS for a total of 1 h. Nuclei were then counterstained with 4,′6-Diamidino-2-phenylindoldihydrochloride (DAPI). All sections were analyzed using a Nikon E600 epifluorescence microscope (Nikon, Düsseldorf, Germany) and 1,000× magnification by a blinded investigator. At least 3 representative high power fields of the Cy3 and DAPI channel from each slide were recorded with Lucia G software, version 4.81 (Laboratory Imaging for Nikon, Prague, Czech Republic) with identical exposure time. Telomere fluorescence intensity was calculated with TFL-Telo freeware version 2.2.07.0418—2002 (25,26).
Telomerase activity was quantified using a telomerase repeat amplification protocol (23,27,28). Protein extracts (1 μg) of mouse left ventricle, 0.1 μg of primer TS (5′-ATCGCTTCTCGGCCTTTT-3′) (template), 0.05 μg primer ACX (5′-GCGCGG [CTTACC]3CTAACC-3′) in 20 μl LightCycler FastStart SYBR Green PCR Master Mix (Roche) containing 1.5 mmol/l MgCl2 were incubated at 30°C for 30 min to allow template elongation by telomerase activity. After immediate transfer to the LightCycler instrument (Roche), telomerase activity was terminated and hot start DNA polymerase activated by incubation at 95°C for 10 min. Forty cycles of amplification were carried out with 20 s at 95°C, 30 s at 60°C, and 50 s at 72°C. Telomerase activity was displayed as relative telomerase activity to 1,000 human embryonic kidney (HEK 293) cells (Gibco, Karlsruhe, Germany). A standard titration curve of HEK 293 cells was established from 0 to 1,000 cells to ensure linearity of the assay (R2 = 0.99). A positive result of the telomerase assay was considered if the quantity of telomerase activity was 3 times above the standard deviation of the mean negative control's background level (heat-inactivated lysate from 1,000 HEK 293 cells). A positive control (protein extracts from 1,000 HEK 293 cells) was run in every experiment.
Immunofluorescence analysis of Ki-67 in cardiomyocytes
Coimmunostaining for the proliferation marker Ki-67 (Novocastra, Leica Biosystems, Newcastle, United Kingdom) and alpha-sarcomeric actin (clone5c5, Sigma-Aldrich) was performed in 3-μm thick myocardial sections as previously described (29). The percentage of Ki-67 expression was calculated from the total number of Ki-67 positive cardiomyocytes and the total number of cardiomyocytes per area.
Western blot analysis
Left ventricular tissue was homogenized with 500 μl lysis buffer (100 mmol/l Tris [pH 6.8], 4% sodium dodecyl sulfate [SDS], 20% glycerol) containing the protease inhibitor M phenylmethanesulfonyl fluoride 0.1 mmol/l, leupeptin 0.5 μl, and aprotinin 0.5 μl. We separated 50 μg of proteins on SDS–polyacrylamide gel electrophoresis 10%. Proteins were transferred to nitrocellulose membrane (162-0112, Bio-Rad Laboratories, Hercules, California) blocked with 5% dry milk or blocking solution for Western blot (Roche) and exposed to rabbit polyclonal immunoglobin G (IgG) TRF2 (H-300: sc-9143, Santa Cruz Biotechnology, Santa Cruz, California; dilution 1:200), mouse monoclonal IgG p16 (F-12: sc-1661, Santa Cruz Biotechnology; dilution 1:250), rabbit polyclonal IgG anti-p53 (FL-393: sc-6243, Santa Cruz Biotechnology; dilution 1:500 in dry milk 1%), mouse monoclonal IgG Chk2 (A-11: sc-17747, Santa Cruz Biotechnology; dilution 1:1,000 in dry milk 1%), rabbit polyclonal IgG anti-p-Akt (Ser473) (#9271, Cell Signaling Technology, Danvers, Massachusetts; dilution 1:1,500), mouse IgG anti-p-eNOS (pS1177) (#612392, Becton Dickinson; dilution 1:1,000), and mouse monoclonal IgG glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (6C5: sc-32233, Santa Cruz Biotechnology; dilution 1:1,000). For analysis of TERT, immunoprecipitation was performed using polyclonal IgG anti-TERT (H-231: sc-7212, Santa Cruz Biotechnology) and agarose-A protein goat antirabbit IgG (Sigma-Aldrich). Immunodetection was accomplished using goat anti-rabbit IgG (Sigma-Aldrich) and goat anti-mouse (170-6516, Bio-Rad) secondary antibodies (1:4,000 dilution), and an enhanced chemiluminescence kit (Amersham Biosciences) (30).
Reverse transcription–polymerase chain reaction
Reverse transcription—polymerase chain reaction standardized to GAPDH was performed using the following primers:
TRF1-rev 5′-ATCTGGCCTATCCTTAGACG-3′ 55°C, 27 cycles
TRF2-for 5′- TGTCTGTCGCGCATTGAAGA-3′
TRF2-rev 5′-GCTGGAAGACCTCATAGGAA-3′ 55°C, 26 cycles
Ku70-rev 5′-CAGCATGATCCTCTTGTGAC-3′ 55°C, 26 cycles
Ku80-rev 5′-AACTGCAGAGAGATGCCAGA-3′ 56°C, 26 cycles
IGF-1-rev 5′-ATTCTGTAGGTCTTGTTTCC-3′ 54°C, 27 cycles
P16-rev 5′-TACACAAAGACCACCCAGCG-3′ 53°C, 40 cycles
P53-for 5′-GGGACAGCCAACTCTGTTATG TGC-3′
P53-rev 5′-CTGTCTTCCAGATACTCGGGA TAC-3′ 62°C, 30 cycles
Chk2-rev 5′-GAAGTAGAGCTTACAGGTGG-3′ 53°C, 40 cycles
GADPH-rev 5′-TCCACCACCCTGTTGCTGTA-3′ 60°C, 27 cycles
Quantification of apoptosis by hairpin oligonucleotide assay
To detect apoptosis, 3-μm thick paraffin sections of formalin-fixed mice heart tissue were examined using the ApopTag Peroxidase In Situ Oligo Ligation Kit (Chemicon International, Millipore, Billerica, Massachusetts) (31). The in situ oligo ligation assay specifically detects apoptosis by staining only cells that contain double-stranded breaks that are blunt-ended or have a 1 base 3′ overhang (cells containing nicked, gapped, 3′-recessed, 3′-overhanging ends longer than 1 base and single-stranded ends are not detected). Unlike conventional terminal transferase-based labeling (terminal deoxynucleotidyl transferase mediated 2′-deoxyuridine 5′-triphosphate nick end labeling [TUNEL]), the assay stains apoptotic but not necrotic or transiently damaged cells (31). To distinguish between cardiomyocytes and other cardiac cells, eosin staining was performed after the in situ oligo ligation assay.
Band intensities were analyzed by densitometry. All values are expressed as mean ± standard error of the mean. Unpaired Student t tests and analysis of variance for multiple comparisons were applied. Post-hoc comparisons were performed with the Bonferroni adjustment test. The Mann-Whitney U test was used for the analysis of the Ki-67 positive cardiomyocytes because the data were not normally distributed. SPSS software, version 12.0 (SPSS Inc., Chicago, Illinois), was used. Differences were considered significant at p < 0.05.
Voluntary physical exercise increases telomere-stabilizing proteins
The mean voluntary running distance was 5,100 ± 800 m/24 h. Voluntary exercise for 21 days, n = 8 to 12 per group, did not change heart or body weight or the ratio of heart weight to tibia length (Fig. 1A). Echocardiography of mouse hearts showed no change of fractional shortening, end-diastolic and end-systolic thickness of the interventricular septum and the left posterior wall, as well as the left ventricular diameters. Flow-FISH assays showed equal mean length of telomeres compared with sedentary controls (not shown). At the same time, telomerase repeat amplification protocol assays revealed that 3 weeks of exercising up-regulated cardiac telomerase activity to 230.7 ± 21%, p < 0.01 (Fig. 1B). Exercising increased the expression of TERT in the heart to 165 ± 4%, p < 0.05 (Fig. 1C). Next, the effects of voluntary running on cardiac T-loop stabilizing proteins were assessed. Three weeks of running had no effect on the messenger ribonucleic acid (mRNA) expression of TRF1, but up-regulated TRF2 mRNA to 145 ± 12% and TRF2 protein expression to 168.7 ± 8.4%, p < 0.05. Exercise up-regulated mRNA expression of the 80-kDa subunit but not the 70-kDa subunit of the repair protein Ku (Figs. 1D to 1F).
Exercise decreases markers of cellular aging in the heart
Compared with sedentary controls, the left ventricles of the mice that were supplied with a running wheel for 21 days were characterized by a decreased expression of the aging marker protein p16 (52.6 ± 3.3% of control, p < 0.01). Similarly, expression of Chk2, which mediates cell-cycle arrest and apoptosis, was reduced to 78.7 ± 2.7%, p < 0.05. The cardiac expression of the proapoptotic transcription factor p53 was reduced by one-half (56.4 ± 2.7%, p < 0.01). Data are shown in Figure 2. Reverse transcription–polymerase chain reaction analysis showed that p16 and Chk2 mRNA expression were significantly down-regulated (53.3 ± 7.2% of control, p < 0.01 and 75.7 ± 10.8% of control, p < 0.01, respectively), but p53 mRNA levels were not altered after running (101 ± 12% of control).
Effects of long-term voluntary exercise on the telomere complex and cellular aging
To test if the regulation observed after 21 days was a transient effect, mice were equipped with individual running wheels or no running wheel for the duration of 6 months (n = 8 per group). As expected, long-term running induced a mild myocardial hypertrophy that did not impair left ventricular fractional shortening (Fig. 3A). Compared with the 21-day training, the long-term training exerted qualitatively and quantitatively similar effects and significantly increased the activity of telomerase and the expression TRF2. The expression of the markers of reduced cellular life span—p16, Chk2, and p53—was markedly inhibited by 29 ± 3%, 34.8 ± 5%, and 43 ± 4.3%, respectively (p < 0.05) (data depicted in Fig. 3). The experiments show that exercising exerts a continuous and persistent effect on regulators of cellular survival.
Flow-FISH assays detected no shortening of telomeres in leukocytes from 6-month-old mice compared with 3-week-old mice (Fig. 4A). Consequently, mean telomere length did not differ between mice running for 6 months (21.6 kb) and sedentary animals (21.4 kb). A group of 18-month-old C57/Bl6 animals was studied as a positive control that showed reduction of telomere length (17.5 kb in 18-month-old animals compared with 21.8 kb in 3-week-old animals, p = 0.001).
Telomere length in the hearts of corresponding left ventricular myocardial sections was directly examined by FISH using a Cy3-conjugated telomere-specific PNA probe (Figs. 4B and 4C). There was no significant difference in the myocardial telomere length between mice after 3 weeks and 6 months of exercise or sedentary condition (3 weeks sedentary: 4,495.0 ± 159.7 individual telomeric fluorescence units [TFU]; 6 months sedentary: 5,172.0 ± 361.1 TFU; and 6 months running: 4,766.0 ± 468.3 TFU). The 18-month-old mice exhibited a reduced telomere length (2,687.3 ± 146.7 TFU, p < 0.001 vs. all other groups).
Influence of exercise on Ki-67 expression in cardiomyocytes
Coexpression of alpha-sarcomeric actin and Ki-67 was not detected in myocardial sections (n = 8) of control mice. However, there was a marked increase to 0.023 ± 0.01% Ki-67 positive cardiomyocytes in the hearts of mice after 3 weeks of running (p < 0.02) (Fig. 4D).
Effect of voluntary exercise in TERT−/− mice
To test if the effects of physical training on the expression of Chk2, p16, and p53 are coincidence or mediated by the telomere complex, second-generation TERT−/− mice and their strain-matched wild-type (WT) were subjected to 21 days of voluntary running. Wild-types exhibited a significant up-regulation of TRF2 protein expression to 466 ± 121% and a down-regulation of p16, p53, and Chk2 expression to 68 ± 8.7%, 50.6 ± 9.0%, and 56.4 ± 10.6%, respectively, n = 8 per group, p < 0.05. However, these exercise-induced changes were absent in the TERT−/− mice (Fig. 5).
IGF-1 and eNOS mediate the effects of exercising on survival proteins
Physical exercise is associated with increased serum concentrations of IGF-1 (32). IGF-1 is decreased in elderly patients and has been shown to enhance telomerase delaying cellular aging and death (10). The effect of exercising on cardiac IGF-1 is not known. Voluntary running induced up-regulation of myocardial IGF-1 expression to 169.7 ± 13% (Fig. 6A). To provide evidence for a potential role of IGF-1 as mediator of exercise-induced cardiac telomerase activity, mice were treated with IGF-1 and GH, which increases endogenous IGF-1 levels (22). Treatment with GH led to an 8-fold increase of telomerase activity and IGF-1 up-regulated telomerase activity by 14-fold, p < 0.001 (Fig. 6B). The IGF-1 treatment of animals resulted in 2-fold increased expression of p-Akt (p < 0.01) and p-eNOS (p < 0.05) in the myocardium (Figs. 6C to 6E). Running is known to up-regulate the expression and function of eNOS, which has been confirmed in our model of voluntary running at several time points (33–35). To test a potential role of eNOS as a mediator of the observed effects, the experiments were repeated in eNOS−/− mice as a head-to-head comparison with strain-matched WT animals (n = 8 per group, 21 days). The running distance was not different between groups. The results are summarized in Figures 6F to 6J. Sedentary eNOS−/− mice showed a decrease of TRF2 expression to 70.4 ± 2.4% and an increase of p16 expression to 130 ± 5.1% compared with WT sedentary mice (p < 0.05); telomerase activity and p53 expression were not significantly altered. The eNOS-deficient mice—in contrast to the WT mice—did not show that exercise modified the cardiac telomerase, TRF2, p16, p53, or Chk2.
Exercise prevents doxorubicin-induced cardiac apoptosis
Doxorubicin is a commonly used antineoplastic drug with frequent cardiotoxicity. The acute doxorubicin-induced cardiomyopathy is characterized by cellular senescence and apoptosis (21). To test the functional relevance of the observed exercise-induced up-regulation of telomere stabilizing and down-regulation of proapoptotic genes, we exposed C57/Bl6 and TERT−/− mice with and without running wheels to a single injection with 22.5 mg/kg i.p. of doxorubicin (n = 8 per group). We measured TRF2 expression 24 h after injection and found that it was moderately decreased and telomerase activity was not changed in sedentary control mice (Figs. 7A and 7B). Voluntary running was able to increase TRF2 as well as telomerase activity in the presence of doxorubicin. Doxorubicin-induced p53 expression in WT mice by 2-fold to 213 ± 7%, p < 0.01. The effect was significantly blunted in mice that had been running voluntarily for 21 days. Exercise prevented the doxorubicin-induced down-regulation of TRF2 and markedly reduced up-regulation of p53 (163 ± 25% of control, p < 0.05 vs. doxorubicin group) (Figs. 7C and 7D).
Apoptosis was quantitated by hairpin oligonucleotide assays. Interestingly, 21 days of voluntary exercising decreased apoptotic cardiomyocytes by 5-fold (0.03 ± 0.01 vs. 0.17 ± 0.03%, p < 0.01). As expected, doxorubicin increased cardiomyocyte apoptosis in the inactive animals (0.29 ± 0.08%, p < 0.05). However, mice supplied with running wheels were completely protected from doxorubicin-induced programmed cell death of cardiomyocytes (Fig. 7D). Cardiomyocyte apoptosis rate of sedentary versus running TERT+/+ B6.129S mice was in the same range of the corresponding C57/Bl6 strain as discussed previously (0.51 ± 0.21% and 0.21 ± 0.03%, p = 0.03 for control vs. control + doxorubicin). In contrast, sedentary TERT−/− mice of the same strain showed an increased susceptibility to doxorubicin-induced apoptosis (0.73 ± 0.14%, p < 0.05 vs. WT doxorubicin control) (Fig. 7F). In these mice, exercise was not able to render a significant protection (p = NS for TERT−/− doxorubicin control vs. runner; p < 0.001 vs. WT doxorubicin runner) demonstrating the importance of TERT for the exercise-mediated effects on apoptosis.
The study identifies a novel effect of physical exercise on the heart. Compared with mice kept under the regular conditions of laboratory animals, voluntary running exercise up-regulated telomere-stabilizing proteins, reduced cellular senescence, and prevented doxorubicin-induced apoptotic cell death. Further characterization shows that the effect depends on TERT expression and is mediated by IGF-1 and eNOS. The effect was observed after 3 weeks of voluntary running and persisted for at least 6 months.
Physical activity is associated with cardiovascular protection and longevity (1–3,5,6,8,9,36). However, the underlying molecular mechanisms remain only partially understood (7). Here we observe that exercising increases the activity of the telomerase, the reverse transcriptase responsible for the extension of telomeric repeat sequences, as well as the expression of its catalytic subunit, TERT. Both telomerase and TERT have been shown to regulate cardiac muscle cell growth and survival (10,11,16,37), and telomerase ribonucleic acid knockout mice (Terc−/−) develop cardiac dysfunction, increased expression of p53, and increased apoptosis (16). Overexpression of TERT causes hypertrophy in cultured cardiac myocytes and protects from apoptosis in vitro and in vivo (11). In addition to the effects on telomerase and TERT, exercising mice showed up-regulation of TRF2 mRNA and protein expression. Telomere repeat binding factor 2 can bind to the TTAGGG repeats of telomeres and contributes to the formation of the chromosome-protecting T-loops (12). Importantly, TRF2 serves as the binding platform for additional telomere-associated proteins and mediates signaling to DNA damage checkpoint controls (13,14). Interestingly, cardiac apoptosis in human heart failure was associated specifically with defective expression of TRF2 and activation of the DNA damage checkpoint kinase, Chk2 (17). In addition, exogenous TRF2 was shown to confer protection from oxidative stress (17). In circulating progenitor cells, TRF2 was identified as a regulator of clonogenic potential and migratory capacity that can be modified by pharmacological treatment (38,39). Here, voluntary exercise resulted in up-regulation of telomerase, TERT, and TRF2 after only 3 weeks and the regulation persisted for at least 6 months. Therefore, physical exercise represents a potent regulator of the telomere complex.
Relatively little is known about the physiologic development of telomere length in healthy untreated WT C57/Bl6 mice over time (8,14,40). After 6 months, we observed no shortening of telomere length, and telomere length did not differ between running and sedentary animals both in blood leucocytes as well as in the myocardium. A control group of 18-month-old C57/Bl6 mice exhibited significantly shorter telomeres. This finding may be interpreted in several ways. Six months may be too short a time, and significantly longer observation, such as 18 months, may be necessary to potentially observe a protective effect of exercising on telomere length. On the other hand, it is possible that the enhanced telomerase activity may counteract a putative telomerase-independent negative effect of exercising. However, the interesting aspect of these data is that the regulation of the telomere-regulating proteins by exercise-mediated survival signaling seems to be independent of telomere length. Indeed, recent evidence suggests a direct telomerase-dependent transcriptional regulation of genes involved in cell growth, which has been recently suggested as an additional mechanism by which telomerase promotes cell proliferation. The regulation of telomerase activity and subsequent replicative potential is reported to occur rapidly and independently of telomere length in several cell types (8,22,41). Defects in mice lacking the ribonucleic acid component of telomerase involve apoptosis, not just proliferation defects, and telomerase directly protects cells against programmed cell death (8,16,42,43). Similarly, TRF2 was suggested to mediate proapoptotic signaling in post-mitotic, noncycling cardiomyocytes (17) and was shown to signal independently of telomere length in endothelial progenitor cells (39). It therefore seems likely that the telomeric complex serves as a regulator of cellular aging and function beyond and potentially independently of protecting telomere length.
Several gene products implicated in senescence and cell death, such as transformation-related protein p53, p16, and Chk2, have been identified downstream of the telomeric complex (10,13,14,16–19). The tumor suppressor protein p53 has been shown to mediate telomere dysfunction (16,19). Protein p53 modulates apoptosis and senescence by increasing the expression of specific proteins, including Bax, Bad, and p21. Importantly, p53 was recently identified as a key mediator of maladaptive cardiac remodeling essential for the transition from cardiac hypertrophy to heart failure (18). Here, the data identify voluntary running as a novel and potent inhibitor of cardiac Chk2, p16, and p53.
To test whether the exercise-induced regulation of the telomere complex and the regulation of p53 represent a coincidence or may be causally related, the experiments were repeated in TERT-deficient mice. Running induced very similar regulation of the transformation-related proteins in TERT−/− WT and C57/Bl6 mice, but the exercise-mediated effects were absent in the TERT−/− mice. These data support the concept that the regulation of the telomeric complex by exercising mediates the downstream effects on apoptosis.
Physical exercise is associated with increased serum concentrations of IGF-1 (32,44). IGF-1 enhances telomerase, delaying cellular aging and death (10,22). Cardiac overexpression of IGF-1 in transgenic mice increases the heart weight (45,46). In IGF-1 transgenic mice, cardiac stem cell division is increased, which is accompanied by enhanced telomerase activity and delayed senescence (10). Similarly, age-dependent impairment of endothelial progenitor cells is corrected by GH-mediated increase of IGF-1 (22). IGF-1 and its ligand may have the potential to regulate the re-entry of adult ventricular myocytes into the cell cycle (46,47). In contrast to the systemic effects of exercising on IGF-1, the effect of exercise on local cardiac IGF-1 is not known. Here, the data show that voluntary running increases the expression of IGF-1 in the heart. To test the role of IGF-1 in exercise-induced telomerase regulation, mice were treated with recombinant GH and IGF-1, which resulted, respectively, in powerful 8- and 14-fold increases of telomerase activity. The effects of IGF-1 on telomerase have been shown to be associated with the regulation of endothelial nitric oxide and phosphatidylinositol 3-kinase–Akt-p70S6K signaling, a pathway that plays an important role in regulating cardiac hypertrophy, viability, and homeostasis (22,48). In line with these data, treatment with recombinant IGF-1 led to significant increases in myocardial expression of p-Akt and p-eNOS. Indeed, one of the best characterized molecular effects of exercising in the model of voluntary running is the up-regulation of eNOS (33,35). Therefore, the experiments were repeated in eNOS−/−mice showing that the effects of exercising on telomere-regulating proteins and subsequent survival signaling are mediated via eNOS. Taken together, the data identify IGF-1, Akt, and eNOS as important mediators of the exercise-induced up-regulation of telomerase activity.
Apoptosis has been implicated in both acute and chronic heart diseases causing progressive loss of cardiac myocytes. End-stage human heart failure is characterized by increased myocyte apoptosis (20). Similarly, several forms of acute toxic cardiomyopathy are characterized by cellular senescence and apoptosis; however, our understanding of potential strategies to prevent cardiomyocyte senescence and apoptosis is limited (20). Doxorubicin is a potent, widely used antineoplastic agent in cancer chemotherapy. However, its clinical application is compromised by its dose-dependent cardiotoxicity mediated by cardiomyocyte apoptosis (21). Here, we applied the well-characterized model of acute doxorubicin-induced cardiac apoptosis to test if the observed effects of exercising on telomere-regulating proteins and survival markers are able to confer protection from cardiomyocytes apoptosis. After acute exposure to doxorubicin, mice showed a small reduction of TRF2 expression. However, in mice that had access to a running wheel during the 3 weeks prior to doxorubicin exposure, TRF2 was significantly up-regulated. Similarly, the doxorubicin-induced increase of cardiac p53, a key mediator of cardiac cell death (16,18,19), was markedly blunted in mice of the running group. Cardiomyocyte apoptosis was assessed by the hairpin oligonucleotide assay that specifically detects apoptosis by selectively staining cells that contain double-stranded DNA breaks that are blunt-ended or have a 1 base 3′ overhang whereas cells containing nicked, gapped, 3′-recessed, 3′-overhanging ends longer than 1 base, and single-stranded ends are not detected. Unlike conventional terminal transferase-based labeling, the assay stains apoptotic but not necrotic or transiently damaged cells (31). The hairpin oligonucleotide experiments revealed that providing mice with a running wheel to allow voluntary running for the relatively short period of 3 weeks represents a powerful intervention to protect from doxorubicin-induced cardiomyocyte apoptosis.
Somewhat unexpectedly, voluntary running decreased the basal rate of cardiomyocyte apoptosis compared with sedentary animals. Ascensão et al. (49) observed prevention of doxorubicin-induced increase in Bax, Bax-to-Bcl-2 ratio, and tissue caspase-3 activity by a 14-week training protocol in rats but no effect of exercise on the basal rate of these markers in left ventricular homogenates. Chicco et al. (50) show a decrease of caspase-3 activity in the rats' left ventricles induced by exercising both in the presence and in the absence of doxorubicin; however, the latter effect did not reach statistical significance. None of these studies has directly assessed cardiomyocyte specific apoptosis; in addition, there were significant differences in the protocols, species, and modalities of training. In agreement with previous studies, long-term voluntary running (e.g., 6 months) is associated with increased heart weight, an adaptive change without impairment of cardiac function. In the light of these data, it is interesting to speculate that the opportunity to exercise resembles the natural habitat of mice more closely than cages without running wheels. Consequently, the inactivity of regular laboratory mice could be considered the experimental intervention in this study. The provocative hypothesis would be that the observed cardiac morphology in voluntary exercise is “normal” and that sedentary animals exhibit an adaptive cardiac hypotrophy. The notion of an “anti-aging” effect of exercising on the heart is supported by earlier reports in the literature (51) that survival rates decrease in sedentary as opposed to exercising rodents.
The effects of physical activity and inactivity are not limited to the regulation of telomere-regulating proteins. Improved cardiac antioxidant capacity and beneficial effects on both circulating progenitor cells and cardiac resident stem cells are likely to contribute to the molecular and cellular actions of exercise (7,10,22,35,52). In agreement with this hypothesis, Ki-67–positive cardiomyocytes were detected in the hearts of running but not in sedentary mice. In our opinion, future research is needed to further characterize both the quantitative and the cell-type specific contribution of the exercise-induced mechanisms with the aim to develop more specific therapeutic interventions. Specifically, the effects of exercise on telomere biology in cardiac stem cells may be of functional significance (10).
A significant proportion of elderly heart failure patients show no other confounding variables suggesting that age as such may be a primary cause of cardiac decompensation and diastolic dysfunction (53,54). Shortening of the telomeres has recently been shown to predict cardiac morbidity and mortality, and telomere length of circulating leukocytes is decreased in patients with chronic heart failure (9,14,17,55). In addition, recent evidence suggests that telomere biology represents an indicator for the effect of a pharmacologic intervention (56). Here, we identify physical exercise as a potent antisenescent intervention to up-regulate telomere-stabilizing proteins and to reduce cardiomyocyte apoptosis, improving the molecular understanding of the beneficial cardiovascular effects of exercise. Furthermore, the data set the stage to prospectively investigate these novel cardioprotective effects in specific clinical situations, for example, for the prevention of doxorubicin-induced cardiomyopathy.
The authors thank Sascha Jakob, Ellen Becker, and Simone Jäger for excellent technical assistance.
Supported by grants from the Deutsche Forschungsgemeinschaft (KFO 196) and the Universität des Saarlandes (HOMFOR).
- Abbreviations and Acronyms
- bovine serum albumin
- cell-cycle–checkpoint kinase 2
- endothelial nitric oxide synthase
- fluorescence in-situ hybridization
- glyceraldehyde-3-phosphate dehydrogenase
- growth hormone
- human embryonic kidney
- insulin-like growth factor
- messenger ribonucleic acid
- phosphate buffered saline
- telomerase reverse transcriptase
- telomeric fluorescence units
- telomere repeat binding factor
- tris(hydroxymethyl) aminomethane
- Received February 7, 2008.
- Revision received March 21, 2008.
- Accepted April 14, 2008.
- American College of Cardiology Foundation
- Hambrecht R.,
- Fiehn E.,
- Weigl C.,
- et al.
- Hornig B.,
- Maier V.,
- Drexler H.
- Torella D.,
- Rota M.,
- Nurzynska D.,
- et al.
- Oh H.,
- Taffet G.E.,
- Youker K.A.,
- et al.
- Fuster J.J.,
- Andres V.
- Haendeler J.,
- Hoffmann J.,
- Brandes R.P.,
- Zeiher A.M.,
- Dimmeler S.
- Leri A.,
- Franco S.,
- Zacheo A.,
- et al.
- Oh H.,
- Wang S.C.,
- Prahash A.,
- et al.
- Kang P.M.,
- Izumo S.
- Arola O.J.,
- Saraste A.,
- Pulkki K.,
- Kallajoki M.,
- Parvinen M.,
- Voipio-Pulkki L.M.
- Thum T.,
- Hoeber S.,
- Froese S.,
- et al.
- Lansdorp P.M.,
- Verwoerd N.P.,
- van de Rijke F.M.,
- et al.
- Poon S.,
- Lansdorp P.
- TFL-Telo. Available at: http://www.bccrc.ca/tfl/research_lansdorp/Applications.htm. Accessed January 14, 2008.
- Kim N.W.,
- Wu F.
- Müller P.,
- Kazakov A.,
- Semenov A.,
- Böhm M.,
- Laufs U.
- Maack C.,
- Kartes T.,
- Kilter H.,
- et al.
- Gertz K.,
- Priller J.,
- Kronenberg G.,
- et al.
- Laufs U.,
- Werner N.,
- Link A.,
- et al.
- Hornig B.,
- Maier V.,
- Drexler H.
- Leri A.,
- Barlucchi L.,
- Limana F.,
- et al.
- Spyridopoulos I.,
- Haendeler J.,
- Urbich C.,
- et al.
- Haendeler J.,
- Hoffmann J.,
- Diehl J.F.,
- et al.
- Herbert B.,
- Pitts A.E.,
- Baker S.I.,
- et al.
- Reiss K.,
- Cheng W.,
- Ferber A.,
- et al.
- Delaughter M.C.,
- Taffet G.E.,
- Fiorotto M.L.,
- Entman M.L.,
- Schwartz R.J.
- McMullen J.R.,
- Shioi T.,
- Zhang L.,
- et al.
- Ascensão A.,
- Magalhaes J.,
- Soares J.M.C.,
- et al.
- Chicco A.J.,
- Hydock D.S.,
- Schneider C.M.,
- Hayward R.
- Laufs U.,
- Wassmann S.,
- Czech T.,
- et al.
- van der Harst P.,
- van der Steege G.,
- de Boer R.A.,
- et al.