INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

MITOCHONDRIAL DNA (mtDNA) encodes 13 proteins essential for electron transport and ATP synthesis, as well as tRNAs and rRNAs needed for their translation. In several species, aging is associated with reduced concentrations of the mRNAs encoded by mtDNA, both in muscle and in other tissues (1, 3, 4, 6, 8, 20, 26). Diminished levels of mitochondrial-encoded mRNAs could be the result of a reduced number of mtDNA molecules, reduced transcription of each mtDNA molecule, or increased mitochondrial mRNA degradation. In the aged human brain, mitochondrial mRNA levels are reduced even though mtDNA levels are increased (3). The mitochondrial mRNA/mtDNA ratios also decline with senescence in rat brain and in Drosophila (4, 8). In contrast, the reduced mitochondrial mRNA expression in muscle of old rats appears to be explained by a lower mtDNA concentration (1). Studies of patients undergoing orthopedic surgery indicated that muscle of older patients had more mtDNA than muscle of young patients (2, 15). However, in one of these studies, the concentrations of mitochondrial-encoded mRNAs were not affected by aging, and in the other they were not determined. Examination of a few of these samples revealed increased expression in older muscle of mitochondrial transcription factor A (Tfam; protein and mRNA), which is required for mtDNA transcription and replication (replication is linked to transcription because RNA primers are required), and of nuclear respiratory factor-1 (NRF-1) mRNA (11). NRF-1 regulates transcription of the Tfam gene and other genes involved in mitochondrial biogenesis and energy production (19). In our previous studies, in which concentrations of mitochondrial-encoded mRNAs were reduced in older muscle, we did not observe an increase in expression of Tfam mRNA (26, 27) or protein (unpublished data). However, we did not examine the mtDNA concentration or NRF-1 expression. We explored this issue further in the present study by comparing levels of mtDNA, Tfam protein and mRNA, and NRF-1 mRNA in young and old muscle. We also examined expression of several mRNAs encoding other proteins required for mtDNA replication. Because a muscle's mtDNA concentration is determined in part by its contractile activity (9), we examined the relation between mtDNA levels and both self-reported physical activity and aerobic fitness [maximal oxygen consumption (O_2 max)].

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects were 24 healthy volunteers, 12 of them 21-27 yr old and 12 of them 65-75 yr old (Table 1). All signed a consent form approved by the University of Rochester Research Subjects Review Board, after risks and procedures had been explained.

The subjects were students, clerical workers, professional/technical workers, or retirees. We did not enroll any subjects who had been engaged in vigorous exercise for more than 2 h/wk during the month before the initial interview. Physical activity during the 1-yr period before the study was assessed by the monitoring trends and determinants of cardiovascular disease optional study of physical activity questionnaire (14). The questionnaire evaluated the amount of time spent walking and the amount of time spent doing sports and exercises (only those done at least 12 times during the preceding year, and only the two sports or exercises done most frequently). Activities done <12 times in the preceding year, and activities not considered to be a sport or exercise by the subjects, would not be included in responses to these questions. For the present study, we evaluated only sports or exercises that caused a "moderate increase" or a "large increase" in the rate or depth of breathing, excluding those that caused "no change" or a "small increase" in the rate or depth of breathing. The activity questionnaire also instructed subjects to assign their present level of physical activity to one of four categories: 1) no weekly activities; 2) only light activity most weeks; 3) vigorous activity for at least 20 min once or twice each week; 4) vigorous activity for at least 20 min three or more times per week. Vigorous activity was defined as activity causing shortness of breath, rapid heart rate, and sweating.

Body composition was determined by whole-body dual energy X-ray absorptiometry (GE-Lunar Prodigy). O_2 max during stationary bicycling was determined with a standard protocol for increasing the load until exhaustion or until there was no further increase in oxygen consumption. Oxygen consumption was assessed with a MedGraphics CPX/D gas analyzer (Medical Graphics St. Paul, MN). The O_2 max test was separated from the muscle biopsy procedure by at least 3 days.

Needle biopsies of the left vastus lateralis muscle were taken in the morning after subjects had rested for at least 90 min. Samples were frozen in liquid nitrogen within 30 s after removal, then stored at 70°C until analysis. Subjects were asked to refrain from activities more strenuous than walking for 3 days before the muscle biopsy procedure. They stayed at the University of Rochester General Clinical Research Center the night before the biopsy procedure, so that subject-to-subject variability in physical activity was minimized.

DNA was extracted from ~10-20 mg tissue, after degradation of proteins and RNA with proteinase K and RNase, by phenol-chloroform extraction and ethanol precipitation. The total amount of DNA recovered was determined by ultraviolet absorbance. Relative amounts of nuclear DNA (nDNA) and mtDNA were determined by quantitative PCR (26). Because the total amount of nDNA per milligram muscle tissue is unaffected by aging (25), the ratio of mtDNA to nDNA reflects the tissue concentration of mtDNA. For assessment of nDNA, we amplified from 300 ng total DNA per tube a 308-bp segment of the IGF-1 gene. A 241-bp competitive internal standard (8 fg) was amplified with the same primers in the same tube. The sequences of the primers have been published (27). For assessment of mtDNA, we amplified from 3 ng total DNA per tube a 239-bp segment of the COX-2 gene. A 213-bp competitive internal standard (0.75 pg) was amplified with the same primers in the same tube. COX-2 primers were 5'-acctgcgactccttgacgttg and 5'-taggacgatgggcatgaaactg. PCR products were separated by PAGE. The gel was soaked in SYBR Green dye (Molecular Probes, Eugene, OR), and fluorescence of the bands was quantified with a FluorImager (Amersham Biosciences, Piscataway, NJ). The amounts of competitive standards were determined by preliminary experiments, to produce similar amounts of the gene segments and the internal standards after PCR amplification. Standard curves with fixed amounts of competitive standard, and varying amounts of DNA indicated that the ratio of amplified gene to amplified standard was proportional to the initial ratios over a wide range. All samples produced ratios within the ranges of the standard curves.

RNA was extracted from ~40 mg of tissue with TriReagent (Molecular Research Center, Cincinnati, OH) and quantified by ultraviolet absorbance. Quality was confirmed by the presence of ribosomal bands in an agarose gel soaked in ethidium bromide. The gel was scanned with a FluorImager, and the intensity of the 28S band was quantified. An aliquot of the RNA was analyzed with high-density oligonucleotide arrays as described below. Another aliquot was used for reverse transcription of the RNA after treatment with DNase I to remove any genomic DNA present in the RNA samples. The COX-2 cDNA was then quantified by PCR with a competitive internal standard, as described above for COX-2 mtDNA. No COX-2 band was detected when the reverse transcription step was omitted. Because the RNA concentration of muscle is not affected by aging (25, 26), and most of the RNA is ribosomal, the ratio of COX-2 mRNA to 28S rRNA is an index of the tissue concentration of COX-2 mRNA.

Tfam protein expression was determined by Western blotting. Proteins were extracted from ~10-25 mg of tissue with 6 M urea, 1% SDS, and 1% 2-mercaptoethanol (50°C for 30 min). Insoluble proteins were removed by centrifugation. Protein concentrations of the supernatants were determined with the RC DC assay kit (Bio-Rad, Hercules CA). After proteins were separated by SDS-PAGE (100 µg of total protein per lane), they were transferred to nylon membranes. Samples from a younger and an older subject were placed in adjacent lanes, and an equal number of samples from younger and older subjects was analyzed with each membrane. Staining of the blots with SYPRO Ruby (Molecular Probes) confirmed that there was no significant difference in transfer efficiency between samples from younger and older subjects. Polyclonal anti-human Tfam rabbit immunoglobulin, generously donated by Dr. David Clayton (Howard Hughes Medical Institute, Stanford University), was the primary antibody. The secondary antibody (anti-rabbit IgG conjugated to horseradish peroxidase) was detected by chemiluminescence on X-ray film. Quantitation of relative band intensities was done by laser scanning densitometry. Each blot, with six to eight samples per blot, was analyzed as a separate entity to minimize variability introduced by day-to-day variations in incubation and washing times, film exposure times, etc. Thus data were normalized separately for each blot, with the mean background-subtracted band intensity of samples from younger subjects arbitrarily set to a value of 100. Tfam expression could not be determined for six subjects (four younger and two older) because of an insufficient amount of tissue.

Levels of mRNAs encoding Tfam, NRF-1, RNase MRP, DNA polymerase  (catalytic and accessory subunits), single-stranded DNA binding protein 1, and mitochondrial RNA polymerase were analyzed with Affymetrix (Santa Clara, CA) HG-U133A microarrays by the University of Rochester Microarray Core facility with methods that have been described elsewhere (27). These arrays have ~22,000 probe sets; other results will be presented in a separate report. Some of the genes are represented by more than one probe set; in this case, we report the results from the probe set that generated the highest signal. Arrays were normalized so that the average signal across all probe sets, excluding the highest and lowest 2% of the signals, was 500 for each array. Microarray data are not available for 6 of the 24 subjects (4 young men, 2 older women).

Data are presented as means and SD. Comparisons between younger and older subjects were made with two-tailed t-tests. Pearson's correlation coefficient was used to determine the relation between two variables. Statistics were computed with Microsoft Excel 97, except partial correlation coefficients were computed with SPSS 11.5 for Windows.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The lean tissue mass of the legs (excluding bone), relative to body weight, was significantly less in older subjects (Table 1). O_2 max per kilogram body mass was 37% lower in older subjects (Table 1). The lean mass of the legs was a slightly better predictor of O_2 max (r = 0.79) than was total lean body mass (r = 0.76) and was a much better predictor than was total body weight (r = 0.35). When O_2 max was expressed per kilogram of lean tissue mass of the legs, the age-related decline was smaller (20%) but still statistically significant.

The mean COX-2 mRNA level, relative to 28S rRNA, was 22% lower in older muscle than in younger muscle (Fig. 1). There was an even larger decline (38%) in the mtDNA/nDNA ratio with aging of muscle (Fig. 1). Tfam protein and mRNA levels were similar in younger and older muscle (Table 2). There was no effect of age on expression of NRF-1 mRNA or mRNAs encoding proteins essential for mtDNA replication (Table 2).

View larger version (22K): [in this window] [in a new window]   Fig. 1.   Mean mitochondrial DNA [mtDNA, relative to nuclear DNA (nDNA)] and cytochrome-c oxidase-2 (COX-2) mRNA (relative to 28S rRNA) in 12 younger (21-27 yr old) and 12 older (65-75 yr old) subjects. Error bars are 1 SD. P determined by 2-sided t-test.

There was a correlation between mtDNA/nDNA ratios and O_2 max (r = 0.58, P = 0.003, Fig. 2). The relation between O_2 max and mtDNA/nDNA ratios appeared to be explained for the most part by the fact that both measures declined with aging, because O_2 max was not a significant predictor of the mtDNA/nDNA ratio after the influence of age was accounted for (partial r = 0.29, P = 0.185), whereas age was a significant predictor of the mtDNA/nDNA ratio after accounting for the influence of O_2 max (partial r = 0.52, P = 0.011).

View larger version (17K): [in this window] [in a new window]   Fig. 2.   Relation between maximal oxygen consumption (O2 max) during stationary bicycling, per kilogram lean tissue mass of legs, and mtDNA/nDNA ratio in vastus lateralis. Solid line is the linear regression line for all subjects regardless of age. , Younger subjects; upper dashed line, regression line for younger subjects; , older subjects; lower dashed line, regression line for older subjects.

Seven young subjects who had spent 68-234 h engaged in sports or exercises (done at least 12 times) during the preceding year according to the activity questionnaire had 43% more mtDNA (P = 0.005) than five young subjects who spent 0-26 h in such activities. Only one older subject regularly performed a sport or exercise that significantly increased his rate and depth of breathing (78 h during preceding year), and he had the lowest mtDNA/nDNA ratio. Older subjects reported spending more time walking (mean 1.8 h/day) than did younger subjects (mean 1.3 h/day), but the difference was not statistically significant and there was not a significant correlation between time spent walking and the mtDNA/nDNA ratio within either age group or overall. There also was no apparent relation between self-reported scores for current level of vigorous activity (1-4 scale) and mtDNA/nDNA ratios.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

A decline with aging in the concentration of mtDNA transcripts has been observed in tissues of several species, including human muscle (26). The age-related decrease in muscle levels of COX-2 mRNA, a representative mtDNA transcript, was confirmed in the present study. In principle, this effect could be mediated by a reduced mtDNA concentration, by diminished transcription of mtDNA, by increased mtRNA degradation, or by a combination of these factors. The present study suggests that in human muscle, a reduced number of copies of mtDNA can explain the lower abundance of mitochondrial mRNAs in older subjects. The decline in mtDNA was greater than the decline in COX-2 mRNA, suggesting that there could have been compensatory increases in mtDNA transcription or reduced mtRNA degradation in older muscle.

The reduced mtDNA concentration in older muscle could reflect either fewer mitochondria or fewer copies of mtDNA per mitochondrion. We did not have enough muscle tissue to determine the number of copies of mtDNA per mitochondrion. Previous studies indicated that the total mitochondrial volume within muscle declines with aging because of a reduced volume of each mitochondrion rather than a reduced number of mitochondria (Table 3). On the basis of these data, it seems likely that there are fewer copies of mtDNA per mitochondrion in older muscle.

The present results contrast sharply with studies of orthopedic patients (2, 11, 15), which indicated that aging was associated with increased levels of mtDNA (>1.5-fold), Tfam protein (2.6-fold), Tfam mRNA (11-fold), and NRF-1 mRNA (6-fold). The reason for such discrepant findings is unclear. The orthopedic patients were studied during surgery, but the specific conditions requiring surgery were not listed. In one of the previous studies (15), muscle samples from both the vastus lateralis and rectus abdominalis were used to examine mtDNA content, whereas we examined only vastus lateralis. It is not clear whether the different muscles had different mtDNA levels or whether the type of muscle examined was evenly distributed across age groups. Several orthopedic patients were in their ninth decade of life, whereas older subjects in the present study were all in their seventh and eighth decades. However, the amount of mtDNA (or Tfam and NRF-1 expression) did not change between the eighth and ninth decades in the orthopedic patients.

The present study did not reveal a molecular explanation for the reduced mtDNA concentration in older muscle. According to microarrays, there was no age-related difference in the expression of several genes involved in mtDNA replication. The lack of change in Tfam and mitochondrial RNA polymerase mRNA expression with aging confirms our quantitative RT-PCR data from a separate group of subjects (26). However, we cannot exclude the possibility that aging affected translation of the mRNAs (except for Tfam mRNA, because the Tfam protein concentration also was not affected by aging) or activity of the proteins. It must be emphasized that expression of these genes was determined after subjects had refrained from vigorous physical activity for 3 days. It certainly is possible that intermittent activity-induced increases in expression or activity of these genes for several weeks before the study could have contributed to a greater mtDNA content in young muscle. Another potential explanation is that the mtDNA from older muscle is a poor template for replication. Somatic mutations accumulate with aging in the control region of muscle mtDNA (24). However, if such mutations were to impair replication of mtDNA, they would be expected to be diluted to low levels by the preferential replication of normal mtDNA (unless very little normal mtDNA survives to old age).

The mtDNA concentration depends on both mtDNA replication and mtDNA degradation. One route of mtDNA degradation is lysosomal digestion of the whole organelle. Human mitochondrial bulk protein turnover rate declines with aging (18). Because the mitochondrial inner membrane and matrix appear to turn over as a unit (12), the slower protein turnover suggests slower macroautophagy of mitochondria. If this is the case, there should be a reduced rate of mtDNA loss by organellar degradation in older muscle. The extent to which mtDNA is degraded within intact mitochondria is unclear. There is evidence that oxidatively damaged mtDNA fragments are present in mitochondria (21). It is not known whether aging affects the rate of degradation of mtDNA independently of bulk mitochondrial degradation.

If younger subjects were more active than older subjects during the weeks before the study, this might contribute to their higher mtDNA concentration. The fact that Tfam, NRF-1, and other activity-dependent genes were not differentially expressed in younger and older muscle does not address this issue, because muscle samples were taken after subjects had refrained from vigorous activities for at least 3 days. The fact that young subjects had higher maximal rates of oxygen consumption does not prove that their level of activity was greater, because O_2 max declines with aging even among the most active individuals (7, 17, 22, 23). The lack of a significant correlation between O_2 max and mtDNA content within each age group (Fig. 2) suggests that aerobic fitness was not the primary determinant of mtDNA content. Moreover, responses to a question about the usual weekly volume of vigorous activity at the time of the muscle biopsy suggested that mtDNA levels were lower in older muscle at every activity level. However, responses to other questions suggested that the decline in mtDNA content of older muscle might be explained in part by less participation in vigorous sports or exercises in the months before the muscle biopsy. The questionnaire might not have captured all activities that influence mtDNA levels. The subjective nature of self-assessment of activity, together with the fact that not all "vigorous" activities require a high rate of respiration by the vastus lateralis, makes it difficult to draw conclusions on the basis of questionnaire data. A definitive assessment of the effect of aging independent of altered physical activity would require a more complex study involving long-term control of the amount and type of physical activity. Even if the effect of aging on the muscle mtDNA concentration is secondary to a change in the amount of physical activity, the prevalence of the decline in vigorous activity with aging makes it likely that the present results are applicable to a large proportion of the population.