Subject characteristics

Sixteen sedentary males and females (mean±sd; age 50±8; range 37–64 yr) volunteered for the study (Table 1). Because of novel findings regarding DNA turnover, a second group of participants (3 female and 1 male; age 21±2; range 19–23 yr) who did not perform exercise training was added to provide a young sedentary comparison group (Table 1). Participants in the exercise group were matched for age, sex, body mass index (BMI), and maximal aerobic capacity (Vo_2max) and then randomized to consume either a carbohydrate (CHO) or protein + carbohydrate (PRO) drink following each exercise session. There were no baseline differences (P>0.05) between PRO and CHO in age, BMI, or Vo_2max (Table 1). The participants were healthy based on medical history questionnaire and results from a Bruce protocol graded exercise test with 12-lead electrocardiogram. Body composition was determined using dual-energy X-ray absorptiometery (DEXA; GE Medical Systems, Madison, WI, USA). All subjects were free of any medications that could impair mitochondrial adaptations, including nonselective β-adrenergic antagonists, HMG-CoA reductase inhibitors (statins), and nonsteroidal anti-inflammatory drugs (26). Medication classes that were allowed included selective serotonin reuptake inhibitors, proton pump inhibitors, and multivitamins.

Study overview

Participants completed a progressive aerobic exercise protocol (3 sessions/wk for 6 wk) and consumed a postexercise drink of either CHO or PRO. The drinks were isocaloric (300 kcal), and PRO included 20 g of mixed milk proteins (drinks were kindly provided by the Gatorade Sports Science Institute, Barrington IL, USA). D₂O was consumed daily to isotopically label newly synthesized skeletal muscle proteins, DNA, and membrane phospholipid glycerol. Vo_2max, DEXA, muscle biopsy, and blood sampling were performed before and after the 6-wk training protocol (Fig. 1). The participants were instructed to maintain their normal dietary habits. Dietary habits were not changed throughout the study, as shown from 3-d records collected every 2 wk (data not shown).

Open in a separate window

Figure 1.

Sixteen participants completed 6 wk of progressive aerobic training while consuming either carbohydrates or carbohydrates with 20 g of protein following each exercise session. D₂O was consumed 3 times daily (tid) or 2 times daily (bid) to isotopically label newly synthesized products obtained in muscle biopsy samples. Body composition and maximal aerobic fitness (Vo_2max) were determined before and after training. Saliva and blood samples were used to determine steady-state D₂O content in body water and precursor enrichments for synthesis rate calculations.

Vo_2max

Vo_2max was measured using a cycle ergometer (Lode Excalibur; Medical Graphics Corp., St. Paul, MN, USA) and indirect calorimeter (ParvoMedics, Sandy, UT, USA), as described previously (15). Heart rate was continuously recorded during the test and used to determine the intensity of exercise sessions. The heart rates recorded at various percentages of Vo_2max (60, 65, 75, and 85%) were used to determine exercise intensity during the 6-wk exercise protocol.

Exercise protocol

The 6-wk training program was chosen because this period has been identified as a the period necessary to induce a new steady state of mitochondrial biosynthesis in humans (27). Exercise was performed on a treadmill and included warm-up (15 min at 60%) and workout stages (30 min, progressing from 65 to 85% by wk 5). The participants wore heart rate monitors (Polar Electro, Lake Success, NY, USA), and the speed or grade of the treadmill was changed to maintain heart rate at the specified percentage of Vo_2max. A skilled technician supervised each exercise session and monitored all exercise intensities and times. The drinks were consumed immediately after exercise under supervision. We had 100% compliance with all exercise sessions and drink consumption.

Deuterium labeling

Deuterium labeling of newly synthesized products was achieved using oral consumption of D₂O (70%; Cambridge Isotope Laboratories, Andover MA, USA) throughout the entire 6-wk exercise protocol (or 4 wk for sedentary group). A target of 1–2% enrichment was achieved during a 1-wk priming stage (50 ml 3×/d = 150 ml/d) and maintained for 5 wk (50 ml 2×d = 100 ml/d). Body water enrichment was determined from saliva swabs collected at wk 2, 4, and 6 as described previously (22, 28). Participants were instructed to not eat or drink anything for 30 min before saliva sampling. Saliva swabs were stored at −80°C until analysis.

Tissue sampling

Participants arrived at the laboratory following an overnight fast for blood and muscle sampling before and after the 6 wk exercise protocol. Venous blood was collected; 1 ml of whole blood was removed and stored at −80°C for peripheral blood mononuclear cell (PMBC) isolation as described below. The remaining whole blood was centrifuged (1200 g, 4°C, 15 min) to separate plasma and buffy coat layers that were removed separately and stored at −80°C. Muscle biopsy samples (~100–150 mg) of the vastus lateralis were removed while the subjects were under local anesthesia (1% lidocaine) using a 5-mm Bergstrom needle with manual suction and then immediately frozen in liquid nitrogen and stored at −80°C.

Mixed muscle protein synthesis rate

Mixed muscle protein synthesis (MPS) rate was determined from ~20 mg of the postmuscle using gas-chromatography mass-spectrometry (GC-MS) analysis of deuterium-labeled alanine, as described previously (20). Approximately 20 mg of mixed skeletal muscle was hydrolyzed by incubation in 6 N HCl at 120°C for 24 h. The hydrolysates were ion exchanged, dried under vacuum, and then suspended in 1 ml of 50% acetonitrile and 50 mM K₂HPO₄, pH 11. Pentafluorobenzyl bromide (20 μl; Pierce Scientific, Rockford, IL, USA) was added, and the sealed mixture was incubated at 100°C for 1 h. Derivatives were extracted into ethyl acetate, and the top, organic layer was removed and dried by addition of solid Na₂SO₄ followed by vacuum centrifugation. With the use of negative chemical ionization, derivatized amino acids were analyzed on a DB225 gas chromatograph column. The starting temperature was 100°C, increasing 10°C/min to 220°C. The mass spectrometry used negative chemical ionization with helium as the carrier gas and methane as the reagent gas. The mass-to-charge ratios of 448, 449, and 450 were monitored for the pentafluorobenzyl-N,N-di(pentafluorobenzyl)alaninate derivative. In all cases, these mass-to-charge ratios represented the primary daughter ions that included all of the original hydrocarbon bonds from the given amino acid. The newly synthesized fraction (f) of muscle proteins was calculated from the true precursor enrichment (p) using mass isotopomer distribution analysis (MIDA; refs. 20, 29). MPS was calculated as the change in enrichment of deuterium-labeled alanine (25) bound in muscle proteins over the entire labeling period.

Muscle DNA extraction

Muscle DNA synthesis rate (DNA %F) was determined from DNA extracted from muscle biopsy samples. Total muscle DNA (~8 μg) was extracted from 50 mg tissue using the MiniDNA kit (Qiagen, Valencia, CA, USA), following manufacturer's instructions, and eluted in 200 μl TE buffer (10 mM Tris and 1 mM EDTA, pH 8.0). A 20-μl aliquot of total DNA was used for PCR procedures (below). The remaining DNA was precipitated with cold ethanol, suspended into 200 μl nuclease-free H₂O, and hydrolyzed to free deoxyribonucleic acids. Briefly, isolated DNA was hydrolyzed overnight at 37°C with nuclease S1 and potato acid phosphatase. Hydrolyzates were reacted with pentafluorobenzyl hydroxylamine and acetic acid and then acetylated with acetic anhydride and 1-methylimidazole. Dichloromethane extracts were dried, resuspended in ethyl acetate, and analyzed by GC-MS on a DB-17 column with negative chemical ionization, using He as carrier and CH₄ as reagent gas. The fractional molar isotope abundances at m/z 435 (M0 mass isotopomer) and 436 (M1) of the pentafluorobenzyl triacetyl derivative of purine dR were quantified using ChemStation software (Agilent Technologies, Santa Clara, CA, USA). The deoxyadenosine fraction was separated and analyzed for deuterium content by GC-MS, as described previously (25).

Blood processing

PMBCs were purified from frozen whole-blood samples using magnetic beads (Miltenyi Biotech, Auburn, CA, USA) following the manufacturer's protocol. The buffy coat layer was used for one participant because the whole-blood sample yielded insufficient DNA. Briefly, anti-CD¹⁴⁺ beads were added to whole blood (1 ml) or buffy coat (~500 μl) and collected using whole-blood columns and MiniMacs separator (Miltenyi Biotech). The PMBC fraction was suspended in 200 μl PBS (3.2 mM Na₂HPO₄, 0.5 mM KH₂PO₄, 1.3 mM KCl, and 135 mM NaCl, pH 7.5). DNA was extracted from PBMCs using a DNA Mini kit (Qiagen), eluted using 200 μl nuclease-free H₂O, and processed for GC-MS analysis as described for DNA %F (25).

Calculation of DNA %F

Deuterium labeling of the deoxyribose moiety of DNA occurs exclusively through de novo nucleotide synthesis and allows calculation of the rate of newly synthesized DNA over extended periods (21, 25). The enrichment of a synthesized product cannot exceed the enrichment of the true precursor; therefore, the enrichment of DNA in a cell that is fully replaced during the labeling period (e.g., PMBC) can be used as an estimate of the enrichment of the precursor in the individual (EM1*). We determined EM1* using 2 methods: PMBC DNA enrichment, and the relationship between measured body water deuterium enrichment and MIDA calculations of EM1* at different body water D₂O enrichments, as described previously (25, 29). A generalized form of the relationship between EM1* and body water D₂O is EM1* = 3.193 * BW + 0.0013, where BW is determined from the body water D₂O enrichment measured in saliva samples. There was no difference between the DNA %F calculated from body water and PMBCs, and the reported values are from body water.

Muscle membrane phospholipid synthesis

Total phospholipids were extracted from muscle biopsy samples, and the glycerol fraction was separated by thin-layer chromatography and analyzed by mass spectrometry (22). Briefly, a lipid-containing fraction was collected during the DNA isolation as described above. The digested muscle sample was applied to a spin column and centrifuged, and then a lipid-enriched fraction was collected from the flowthrough. Lipids were extracted using chloroform, and then fatty acid methyl esters were separated from glycerol by a modified Folch technique. The aqueous phase containing glycerol was lyophilized, derivatized to glycerol triacetate, and resolved by thin-layer chromatography. The labeling of phospholipid glycerol represents the formation of new phospholipid molecules in the same manner as for triglyceride molecules (22); glycerol hydrogens equilibrate rapidly with body water during the formation of triosephosphate, and these hydrogens are then permanently incorporated into the glycerol backbone of monoacylglycerides, which in turn form phospholipids. The synthesis rate of membrane phospholipids (PL %F) is an index of membrane lipid synthesis.

Real-time PCR

The ratio of mitochondrial DNA (mtDNA) to nuclear DNA (nDNA) copy number was determined using real-time PCR from an aliquot of total muscle DNA. Briefly, 10 ng of DNA was amplified in a 20-μl reaction with a premade master mix (Thermo Fisher, Rockford, IL, USA) and TaqMan primers and probes (Applied Biosystems, Carlsbad, CA, USA). Samples were amplified in triplicate for each primer/probe sequence (singleplex), and all plates included blank and internal controls. PCR conditions were a hot start (15 min at 95°C) followed by 40 cycles of denaturing (15 s at 95°C) and annealing (60 s at 60°C) in 96-well clear reaction plates using a 7300 Real-Time PCR System (Applied Biosystems). One nuclear and one mitochondrial primer/probe set was custom designed using Primer Express 3.0 software (Applied Biosystems), and the other primers and probes were inventoried TaqMan gene expression assays for genomic DNA (Applied Biosystems; Table 2). The inventoried assays include proprietary sequences that are not specified by Applied Biosystems.

mtDNA content

The relative copy (RC) number of mtDNA to nDNA (30) adjusted for qPCR efficiency (31) was calculated as RC = E^ΔCt where

The slope of the linear regression line of diluted DNA samples covering the expected threshold cycle C_t for gene targets revealed equal efficiency E between gene targets (E=75%). C_t values were determined automatically using the SDS 1.4 software (Applied Biosystems), and then it was manually verified that the threshold line was located early in the linear amplification phase. C_t values were consistent between the control samples run on separate plates.

Muscle protein and DNA synthesis

Body water enrichment reached ~1.5–2.5% by wk 2 and was maintained through wk 6 (Fig. 3). The increased body water enrichments had no adverse effects, and only one participant, with the lowest body weight, reported a slight and transient dizziness during the initial loading phase of heavy water consumption. Raw enrichment values from skeletal muscle fractions are displayed in Table 3. Average MPS (Fig. 4A) over 6 wk of endurance exercise did not differ between PRO and CHO groups (PRO, 39.6±5.9; CHO, 43.5±5.9% new in 6 wk; P = 0.22). Because MPS did not differ between CHO and PRO groups, we combined the two groups into a single older exercise group and compared it with the younger sedentary group. The older exercise group had a higher MPS than the younger sedentary group (7.4±1.0 vs. 5.54±1.0%/wk; P = 0.0043; Fig. 4B).

Open in a separate window

Figure 3.

Body water D₂O enrichment measured from saliva swabs was raised to 1.5–2.5% during the 6-wk labeling period.

Table 3.

Precursor and product enrichment of skeletal muscle protein, DNA, and membrane phospholipids following ²H₂O labeling

Fraction	Older exercise		Young sedentary
	CHO	PRO
DNA
Precursor (%)	6.74 ± 2.50	5.69 ± 1.85	3.03 ± 1.10
Product (%)	0.42 ± 0.22	0.28 ± 0.14	0.05 ± 0.04
Protein
Precursor (%)	4.34 ± 1.56	3.97 ± 2.75	2.83 ± 0.81
Product (%)	1.91 ± 0.60	1.58 ± 1.00	1.20 ± 0.33
Phospholipids
Precursor (%)	6.96 ± 2.42	6.39 ± 1.81	7.53 ± 2.85
Product (%)	2.79 ± 0.82	2.36 ± 0.82	0.41 ± 0.08

Open in a separate window

Data are means ± sd. Older exercise groups were labeled for 6 wk; young sedentary group was labeled for 4 wk.

Open in a separate window

Figure 4.

A) There were no differences in MPS between CHO or PRO nutrition with aerobic exercise training. B) Younger sedentary group (Yg-SED) had lower MPS than the older exercise group (Older-EX). *P < 0.005.

The average DNA %F (Fig. 5A) over 6 wk of endurance exercise did not differ between the PRO and CHO groups. Combining the two groups resulted in an average DNA %F of 4.2 ± 2.8% (95% CI: 2.7–5.8% newly synthesized DNA) in 6 wk or ~0.5–1%/wk. In contrast, the calculated DNA %F was nearly 0 in the young sedentary group; even considering the shorter labeling time, there was significantly less ²H incorporation into DNA compared with the older exercise group (Fig. 5B). The final enrichment of the young sedentary group was similar to background samples and resulted in DNA %F that did not differ from 0 (P>0.05). DNA %F was weakly correlated with MPS in the older exercise group (Pearson r²=0.25; P=0.054; Fig. 5C).

Open in a separate window

Figure 5.

A)There were no differences in DNA %F between CHO or PRO nutrition with aerobic training. B) DNA %F was lower in the younger sedentary group (Yg-SED) than in the older exercise group (Older-EX). Low enrichment of Yg-SED resulted in a DNA synthesis rate that did not differ from 0. Data are expressed as percentage new per week because the Yg-SED group was labeled for 4 wk and the Older-EX group for 6 wk. C) MPS and DNA %F were correlated over 6 wk of aerobic training but not for the Yg-SED (not shown). *P = 0.013.

The average PL %F over 6 wk of endurance exercise did not differ between the PRO and CHO groups (37.6±11.9 vs. 42.2±11.2% new in 6 wk, respectively; Fig. 6A). The combined CHO and PRO groups (older exercise group) had higher average PL %F than the young sedentary group (7.13±2.1 vs. 1.36±0.55%/wk, respectively; P<0.0001; Fig 6B). PL %F was not correlated to MPS or DNA %F (data not shown).

Open in a separate window

Figure 6.

A) PL %F, an indication of membrane turnover, did not differ between CHO or PRO nutrition groups. B) PL %F was lower in the younger sedentary (Yg-SED) than older exercise (Older-EX) group. *P < 0.0001.

DNA synthesis rate

The D₂O technique labels DNA through de novo nucleoside synthesis pathways and represents almost exclusively S-phase DNA synthesis rather than DNA repair (21). In principle, 3 sources could contribute to the measured DNA synthesis: mitochondrial biogenesis, cells that are continuously replicating in muscle tissue, or satellite cell recruitment.

The measured DNA synthesis can not be accounted for by mtDNA turnover. First, the lack of mtDNA:nDNA change with training suggests that even if mtDNA changed, then nDNA would have changed in the same direction with equal magnitude. Second, the small size of mtDNA (~16.5 kb) and mtDNA copy number (~2500 per nuclear genome) is <1% of total DNA (~3 Gb). Therefore, even a fully turned over mtDNA pool could not account for the ~4% newly synthesized DNA reported in our study.

Second, it is not likely that cells other than myocytes are contributing to our measured DNA synthesis rate. A cell type or collection of multiple types that contribute ~4% of total DNA could be fully turned over and could theoretically account for the newly synthesized DNA over 6 wk. We recruited the sedentary group to determine the rate of DNA synthesis under sedentary and free-living conditions to account for basal turnover of other cell types. DNA in the sedentary group had enrichments that did not differ from background following 4 wk of labeling, indicating that the cells represented in the biopsy samples have basal DNA synthesis rates near zero. Thus, any contribution of cells other than myofibrils is small and does not adequately account for our measured DNA synthesis rate during aerobic training.

The third potential source of DNA synthesis is satellite cell replication. Satellite cells are resident stem cells within skeletal muscle that are usually quiescent but can be activated to divide into 2 daughter cells. A portion of activated satellite cells is used to maintain the population of satellite cells, and the others fuse to existing myofibrils or form nascent myofibrils (32, 33). Early work has shown that isotopic labeling of DNA in skeletal muscle can be attributed to satellite cell replication (34). In addition, ~80% of satellite cells can be labeled within 5 d in growing rats (33). Satellite cells make up ~2–5% of total nuclei in muscle samples from adult humans (35, 36); therefore, it is possible that replication of satellite cells could account for ~4% newly synthesized DNA.

Few studies have considered satellite cell activation with aerobic exercise; however, a recent report in humans showed a single bout of aerobic exercise increased markers of satellite cell activation (37). If satellite cells are activated and differentiated, then an increase in MPS would be stimulated to maintain the posited myonuclear domain (38), which is maintained with aging (39). Consistent with this hypothesis, a positive correlation (r²~25%) was documented between muscle protein and DNA synthesis over the 6 wk of aerobic training.

Muscle protein synthesis

Average muscle protein synthesis did not differ between protein and carbohydrate groups, in contrast to short-term studies. Postexercise protein consumption has been repeatedly shown to increase muscle protein synthesis in the several hours following aerobic (40, 41) and resistance exercise (6, 42). In addition, the importance of timed protein consumption after exercise is supported by studies over several days and weeks that reported increased nitrogen retention (15) and strength gains (43). Our results suggest that any short-term increases in protein synthesis following exercise and protein consumption may not augment cumulative protein synthesis over several weeks compared with carbohydrates only. Instead, they are consistent with results showing that the timing of protein intake can vary around exercise and still allow increased amino acid uptake over several hours (44, 45) or muscle cross-sectional area with training (46).

There are some important considerations when comparing our results to previous findings. First, acute tracer studies are conducted in inpatient settings and commonly have strict dietary controls. Our long-term measures were made on free-living people with fluctuations in daily energy balance, which can alter protein balance (47). Second, we used heavy water to directly measure skeletal muscle protein synthesis, while others have used nonspecific methods that measure net changes, such as nitrogen balance (15) or muscle cross-sectional area (43). Perhaps the timing of protein intake is less critical in people who are consuming adequate calories and protein outside of a tightly controlled inpatient setting.

Muscle protein synthesis was higher in our older exercise group compared with the younger sedentary group and provides strong, though indirect, support for the hypothesis that aerobic training increases protein synthesis in older subjects. The younger untrained group represents an upper bound on the rate of long-term basal synthesis because previous reports show either similar (48) or lower skeletal muscle protein synthesis with aging (10, 49). Therefore, the higher protein synthesis during aerobic training over resting may be greater if compared with an older nonexercising group.

Membrane phospholipid synthesis

The synthesis of membrane phospholipids did not differ between protein and carbohydrate groups and was not correlated to synthesis of skeletal muscle DNA or protein. Interestingly, the synthesis of membrane phospholipids was greater in the aerobic training group compared with sedentary controls and suggests cellular proliferation or remodeling within myofibrils. By area, cellular organelles contain most of the phospholipid content and likely represent the majority of membrane remodeling. In particular, the mitochondrial reticulum is a large portion (20–40%) of intracellular membranes with continuous remodeling (50). Application of this method to isolated mitochondria could facilitate investigation of long-term adaptive changes to the mitochondrial reticulum. Deuterium labeling combined with subfractional analysis can provide unique insight into long-term adaptive changes to cellular membranes.

Mitochondrial adaptations

We did not detect any differences in mtDNA:nDNA between groups at baseline or following training. mtDNA copy number per nDNA is an index of mitochondrial content (11, 51). Interestingly, the increase in Vo_2max in the protein group suggests that consuming protein postexercise during endurance training can promote long-term aerobic adaptations independent to changes in mtDNA:nDNA. Each participant in the protein group increased both absolute and relative Vo_2max by ~10%, and the average value for the protein group was greater than for the carbohydrate group. These findings indicate that postexercise protein consumption increases aerobic adaptations to endurance exercise. Mixed muscle protein did not differ between the groups, but that does not rule out that mitochondrial specific protein synthesis increased to a greater degree in the protein group. Previous work has shown that aerobic exercise and feeding increases mitochondrial protein synthesis in the untrained and trained state (12). It is likely that the aerobic exercise increases mRNA of mitochondrial related proteins and the consumption of protein increases the translation of those mRNA (52). The combination of exercise and protein consumption is an easily achievable lifestyle modification that may help promote aerobic fitness during aging.

The authors appreciate the enthusiasm of all the study participants during the project and thank everyone who helped with exercise training and heavy water preparation: Jon Land, Kyle Barnes, Karen Warnersdorfer, Rebecca Peterson, Carrie Sousek, Shannon Wells, and the Adult Fitness Program at Colorado State University. The authors also thank Christopher Bell, Rebecca Scalzo, and Garrett Peltonen for providing the sedentary control data and Wyatt Voyles for providing medical oversight. Claire Emson and Kelvin Li (KineMed Inc.) provided technical expertise for deuterium analysis.

Funding for the project was provided to B.F.M. by U.S. National Institutes of Health/National Institute of Diabetes and Digestive and Kidney Diseases pilot grant P30 DK048520 and Colorado State University.