Assessment of protein synthesis in highly aerobic canine species at the onset and during exercise training

Abstract

Canis lupus familiaris, the domesticated dog, is capable of extreme endurance performance. The ability to perform sustained aerobic exercise is dependent on a well-developed mitochondrial reticulum. In this study we examined the cumulative muscle protein and DNA synthesis in groups of athletic dogs at the onset of an exercise training program and following a strenuous exercise training program. We hypothesized that both at the onset and during an exercise training program there would be greater mitochondrial protein synthesis rates compared with sedentary control with no difference in mixed or cytoplasmic protein synthesis rates. Protein synthetic rates of three protein fractions and DNA synthesis were determined over 1 wk using ²H₂O in competitive Alaskan Huskies and Labrador Retrievers trained for explosive device detection. Both groups of dogs had very high rates of skeletal muscle protein synthesis in the sedentary state [Alaskan Huskies: Mixed = 2.28 ± 0.12, cytoplasmic (Cyto) = 2.91 ± 0.10, and mitochondrial (Mito) = 2.62 ± 0.07; Labrador Retrievers: Mixed = 3.88 ± 0.37, Cyto = 3.85 ± 0.06, and Mito = 2.92 ± 0.20%/day]. Mitochondrial (Mito) protein synthesis rates did not increase at the onset of an exercise training program. Exercise-trained dogs maintained Mito protein synthesis during exercise training when mixed (Mixed) and cytosolic (Cyto) fractions decreased, and this coincided with a decrease in p-RpS6 but also a decrease in p-ACC signaling. Contrary to our hypothesis, canines did not have large increases in mitochondrial protein synthesis at the onset or during an exercise training program. However, dogs have a high rate of protein synthesis compared with humans that perhaps does not necessitate an extra increase in protein synthesis at the onset of aerobic exercise training.

canis lupus familiaris, the domesticated dog, is capable of extreme endurance exercise performance. At the annual Iditarod Race, winning sled dog teams can cover 1,600 km in less than 9 days. When the energetic demands of this extreme activity are combined with environmental stress, total energy expenditure approximates 50,000 kJ/d (7), which far exceeds, on a relative and absolute basis, the highest values recorded in humans [Tour de France cyclists (25)]. Although the ability to sustain these high energy expenditures may be unique to Alaskan Huskies, other canine breeds also have a large aerobic capacity as is evident by a maximal oxygen consumption (V̇o_2max) of 145.8 ml·kg⁻¹·min⁻¹ recorded in Labrador Retrievers (18) and 240 ml·kg⁻¹·min⁻¹ in one greyhound (17).

The ability to perform sustained aerobic exercise is dependent on a well-developed mitochondrial reticulum. The mitochondrial surface area in the skeletal muscle of Labrador Retrievers has been shown to be twice as large as a similarly sized goat, which is a less aerobic species (24). Mitochondrial biogenesis is the making of new mitochondria and is defined by the making of new proteins with subsequent import into the existing organelle (21). Mitochondrial biogenesis requires a large investment of energy given the energetic cost of protein synthesis. Thus although it is thought that mitochondrial biogenesis is primarily transcriptionally mediated through peroxisome proliferator-activated receptor gamma coactivator 1-alpha (PGC-1α), the energetically costly process of translation is likely regulated as well (12). Mitochondrial biogenesis is, therefore, beneficial to sustain high aerobic energy production, but is also reliant on adequate energy availability. In accordance with this idea, we have previously demonstrated in rodent models that during energetic stress, associated with an increase in AMP-activated protein kinase (AMPK) activity or a decrease in the activity of the mechanistic target of rapamycin (mTOR), mitochondrial proteins are often selectively translated (5, 10). This selective translation of mitochondrial proteins could be a countermeasure to energetic restriction.

Mitochondrial protein synthesis can be assessed with stable isotopes. Traditional stable isotope techniques assess protein synthesis over a period of 20 min (flooding technique) and up to 6 h (constant infusion) (23). Therefore, conclusions are impacted by the specific 20-min to 6-h period that is captured (e.g., after an overnight fast vs. postprandial vs. postexercise, etc.). Longer periods of assessment are limited by the need to maintain a constant infusion of isotopes, thus restricting free-living conditions. To address these limitations, we (4, 5, 13, 14, 19, 22) and others (2, 6, 16, 26) have employed stable isotope techniques that use deuterium oxide (²H₂O). By providing ²H₂O in the drinking water of humans (19, 22) or rodents (4, 5, 13, 14), we have assessed cumulative protein synthesis over the period of weeks, rather than hours. By doing so, the contributions of all key periods (e.g., overnight fast vs. postprandial vs. postexercise, etc.) are incorporated into one value as well as the contributions of both rapidly and slowly turning over proteins. Recently we proposed that rates of protein synthesis should be considered in conjunction with changes in DNA synthesis (4, 10). By performing analyses this way, we gain a greater appreciation for differences in growth vs. proteostatic mechanisms, as protein synthesis is considered in proportion to cell proliferation.

To our knowledge, mitochondrial biogenesis has not been previously assessed in the highly aerobic canine species. In this study we examined the cumulative skeletal muscle protein and DNA synthesis in two groups of athletic dogs that were previously engaged in a strenuous exercise training program, and Alaskan Huskies at the onset of an exercise training program. Because of the highly aerobic nature of canines, we hypothesized that both at the onset and during an exercise training program there would be an increase in mitochondrial protein synthesis compared with sedentary control with no difference in myofibrillar or cytoplasmic protein synthesis rates.

METHODS

Overall study design.

All procedures were approved by the Oklahoma State University IACUC prior to the start of the experiments. Alaskan Huskies and Labrador Retrievers were tested by slightly different protocols at two different locations. The Alaskan Huskies were studied at a competitive racing kennel in Alaska. The kennel had previously won the Yukon Quest 1,000 mile sled dog race, had multiple top 10 finishes at the Iditarod sled dog race, and had been elite endurance competitors for over 10 years. The dogs were both males (n = 8) and females (n = 8) and were 7 ± 2 years old (Table 1). Testing took place in June 2013, the “offseason” from racing, when the dogs had not been undergoing regular exercise training for the previous 3 mo. The sedentary dogs (n = 4) maintained their normal housing conditions. The exercised dogs (n = 12) began exercise training similar to the onset of their normal training program at the conclusion of the summer. The experiment took place over 7 days with the dogs weighed each day of exercise. During the study, the dogs were taken out for off-leash runs every other day. During the 3 exercise days the dogs covered an average of 15.5 ± 1.5 km/day, as determined by GPS collars (Garmin Astro 320 receiver and DC-50 collars). At the onset of the study, the exercised dogs were divided into two groups that were fed two different quantities of protein (2.8 or 3.5 gm/kg BW), but equal total calories. Subsequent analysis determined that there were not differences in protein synthesis between the groups, thus the dogs were combined into one group. The sedentary dogs ate 64 ± 6 kcal·kg^0.75·day⁻¹, while the exercised dogs ate 113 ± 18 kcal·kg^0.75·day⁻¹. Total energy intake was measured by recording the mass of food provided to each dog, as well as the mass of any food refusals. A commercial laboratory measured the nutritional content of a sample of food obtained from the study diet.

Table 1.

Characteristics of Alaskan Huskies studied at the onset of exercise and Labrador Retrievers studied during an exercise training program

		Weight, Kg
	Age, yr	Pre	Post	Change
Alaskan Husky
Sedentary	8 ± 1	20.0 ± 2.6	20.2 ± 2.8	0.3 ± 0.2
Exercise	6 ± 2	21.1 ± 3.7	21.5 ± 3.8	0.4 ± 0.5
Labrador
Sedentary	4 ± 1	34.5 ± 3.2	35.0 ± 3.2	0.5 ± 0.6
Exercise	4 ± 1	29.9 ± 3.8	29.9 ± 3.8	0.0 ± 0.4

Pre and Post refer to values at the beginning and end of the 7-day testing period.

In May 2012 in North Carolina, we studied Labrador Retrievers that were part of the improvised explosive device detection (IDDs) used by dismounted Marine Corps patrols to facilitate standoff detection of explosive devices. The dogs were studied over a period of 7 days with both male (n = 4) and female (n = 8) dogs that were 4 ± 1 years old (Table 1). These dogs were housed in sheltered outdoor runs measuring 3′ × 12′ and were fed twice daily with water provided ad libitum. Sedentary dogs (n = 4) and half of the exercised dogs (n = 4) had recently returned from military deployment and maintained fitness up to the time of study with walks and sprints according to standard Marine Corps training. The other half of the exercised dogs (n = 4) had completed an 8-wk exercise training program consisting of progressive increases in off-leash exercise duration three times weekly designed to enhance fitness for prolonged endurance exercise. Both groups of exercised dogs were considered trained, and since subsequent analysis determined that there were no differences in protein synthesis outcomes in response to the exercise period, they were studied as one group.

Exercised dogs were tested during a deployment simulation consisting of 5 days of simulated deployment activity lasting ∼9 h/day. A separate group of dogs in the same treatment covered an average of 31.2 ± 5.8 km/day according to GPS collars. Body weight was recorded at the beginning of each exercise day. The sedentary dogs ate 207 ± 23 kcal·kg^0.75·day⁻¹, while the exercised dogs ate 248 ± 61 kcal·kg^0.75·day⁻¹. Total energy intake was measured by recording the mass of food provided to each dog, as well as the mass of any food refusals. A commercial laboratory measured the nutritional content of a sample of food obtained from the study diet.

Experimental procedures.

Protein synthesis was determined the same way in both studies. On the day prior to study, the dogs received 200 ml of 99% enriched ²H₂O mixed into kibble or drinking water. Thereafter, the dogs received 8% enriched ²H₂O in all drinking water for the duration of the study. Importantly, at no time were the dogs allowed access to other drinking water or other nondeuterium-enriched water sources (puddles, streams, etc.). Blood was drawn by jugular venipuncture every day (Labrador Retrievers) or every other day (Alaskan Huskies) during the study. For the Labrador Retrievers the last blood draw occurred the day before the final biopsy, whereas for the Alaskan Huskies the last blood draw occurred on the morning of the biopsy. On the morning after the last exercise bout, the dogs received a skeletal muscle biopsy of the semimembranosus muscle while under general anesthesia (propofol, 7 mg/kg IV bolus), and the biopsies were immediately frozen in liquid nitrogen.

Protein isolation and DNA isolation.

Skeletal muscle tissue was fractionated according to our previously published procedures (4, 5, 13, 14). Muscle (15–50 mg) was homogenized 1:10 with a bead homogenizer (Next Advance, Averill Park, NY) in isolation buffer (100 mM KCl, 40 mM Tris HCl, 10 mM Tris Base, 5 mM MgCl₂, 1 mM EDTA, 1 mM ATP, pH=7.5) with phosphatase and protease inhibitors (HALT, Thermo Scientific, Rockford, IL). After homogenization, subcellular fractions of mixed (Mixed, containing primarily contractile elements and nuclei), cytosolic (Cyto), and mitochondrial (Mito) proteins were isolated via differential centrifugation as previously described (4, 5, 13, 14). After muscle fractions were isolated and purified, 250 μl 1 M NaOH was added, and pellets were incubated for 15 min at 50°C and 900 rpm. For DNA, ∼100 ng/μl of total DNA was extracted from ∼20 mg tissue (QiAamp DNA mini kit Qiagen, Valencia, CA).

Sample preparation and analysis of analytes via GC/MS.

Protein was hydrolyzed by incubation for 24 h at 120°C in 6 N HCl. The hydrolysates were ion exchanged, dried under vacuum, and resuspended in 1 ml molecular biology grade H₂O. Five hundred microliters of suspended samples were derivatized [500 μl acetonitrile, 50 μl 1 M K₂HPO₄ (pH = 11), and 20 μl of pentafluorobenzyl bromide (Pierce Scientific, Rockford, IL)], sealed, and incubated at 100°C for 1 h. Derivatives were extracted into ethyl acetate. The organic layer was removed and dried by N₂ followed by vacuum centrifugation. Samples were reconstituted in 1 ml ethyl acetate then analyzed.

The pentafluorobenzyl-N,N-di(pentafluorobenzyl) derivative of alanine was analyzed on an Agilent 7890A GC coupled to an Agilent 5975C MS as previously described (4, 5, 13, 14). The newly synthesized fraction (f) of proteins was calculated from the true precursor enrichment (p) with plasma analyzed for ²H₂O enrichment and adjusted by mass isotopomer distribution analysis (MIDA) (2). Protein synthesis was calculated as the ratio of deuterium-labeled to unlabeled alanine (2) in proteins over the entire labeling period.

To determine body water enrichment, 125 μl of plasma was placed into the inner well of o-ring screw cap and inverted on heating block overnight. Two microliters of 10 M NaOH and 20 μl of acetone was added to all samples and to 20 μl 0–20% D₂O standards then capped immediately. Samples were vortexed at low speed and left at room temperature overnight. Extraction was performed by the addition of 200 μl hexane. The organic layer was transferred through anhydrous Na₂SO₄ into GC vials and analyzed via EI mode with a DB-17MS column.

Determination of ²H incorporation into purine deoxyribose (dR) of DNA was performed as previously described (4, 5, 14, 15). Briefly, DNA isolated from whole tissue was hydrolyzed overnight at 37°C with nuclease S1 and potato acid phosphatase. Hydrolysates 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 as previously described (4, 5, 14, 15).

Western blotting.

To assess AMPK activity and mTOR activity, we used the downstream effectors Acetyl-CoA carboxylase (ACC) and ribosomal protein S6 (RpS6), respectively. In addition we examined the “master regulator” of mitochondrial biogenesis, peroxisome proliferator-activated receptor gamma coactivator 1-alpha (PGC-1α). Western blots were completed on a portion of the Cyto fraction, or (for mitochondrial proteins) the supernatant remaining from a low speed spin (800 g × 10 min). Protein concentration was determined with a bicinchoninic acid assay (Thermo Fisher, Rockford, IL). Samples were diluted to equal concentrations, boiled with Laemmli buffer, and then 20–40 μg of protein was separated with 6–12% SDS-PAGE at 100 V. Small molecular weight proteins (15–90 kDa) were transferred at 4°C (100 V for 75 min in 20% wt/vol methanol, 0.02% wt/vol SDS, 25 mM Tris Base, 192 mM glycine, pH=8.3) to nitrocellulose paper. Large molecular weight proteins (200+ kDa) were transferred overnight at 4°C (10 V in 10% wt/vol methanol, 0.1% wt/vol SDS, 25 mM Tris Base, 192 mM glycine, pH=8.3) to nitrocellulose paper. Blots incubated in Superblock (Thermo Scientific, Rockford, IL) for 1 h after transfer. Antibodies were purchased from Cell Signaling Technologies (rpS6 phospho-Ser[235/236] No. 4858, rpS6 total No. 2217, ACC phospho No. 3661, ACC total No. 3662; Boston, MA) or Santa Cruz Biotechnology (β-tubulin No. sc-5274, PGC-1α No. sc-3067; Santa Cruz, CA). Blots were incubated overnight at 4°C with primary antibodies diluted 1:2,000 (rpS6-phospho), 1:1,000 (rpS6-total), 1:200 (PGC-1α), 1:500 (ACC phospho and ACC total), or 1:1,000 (β-tubulin). Blots were washed in 1 × TBST and incubated with HRP-conjugated secondary antibody antirabbit (rpS6, PGC-1α, ACC) diluted 1:5,000 or antimouse (β-tubulin) diluted 1:2,000 in Superblock with subsequent chemiluminescence detection (West Dura, Pierce, Rockford, IL). Images were captured and densitometry analyzed by UVP Bioimaging system (Upland, CA). Blots were probed for phosphorylated proteins first, placed in stripping buffer (GM Biosciences, Rockville, MD), and then reprobed for total protein. Equal loading was verified by ponceau-s staining and β-tubulin.

Statistics.

Data are presented as means ± SE. For protein synthetic outcomes the data were compared by a two-way (fraction and activity) ANOVA with LSD post hoc analysis when differences were determined (P < 0.05). A comparison of synthetic rates between dog breeds used a one-way ANOVA with LSD post hoc analysis. For Western blot analysis, pre- and poststudy biopsies were compared within a treatment group and analyzed by a two-tailed t-test and considered significant when P < 0.05.

RESULTS

Both the Alaskan Huskies and Labrador Retrievers maintained body weight throughout the week of study (Fig. 1A); however, there was a significant effect of treatment in that the sedentary Labrador Retrievers were heavier than the exercised group. Body water enrichment of ²H₂O was within the 1–4% target range (Fig. 1B). However, the enrichment increased over time from the beginning to the end of the study. Therefore, the calculation of synthesis rates used the weighted average enrichment over time as the precursor enrichment. In the dogs that were initiating exercise training, exercise did not have a significant effect on rates of protein synthesis, in that Mixed, Cyto, and Mito fractions were not different from the sedentary condition (Fig. 2A). However, synthesis rates in the Mixed fractions were significantly slower than the Cyto fractions in both sedentary and exercise-trained dogs. Although the exercise-trained Labrador Retrievers participated in two different training programs prior to study, these two groups had synthesis rates during the period of study that were not significantly different from each other (data not shown), thus justifying studying the two separate groups as one exercise-trained group. In the exercise-trained dogs, there was a significant effect of exercise where Mixed and Cyto, but not Mito, fractions had slower protein synthesis rates compared with the sedentary condition (Fig. 2B). In addition, the sedentary Mito protein synthesis rate was lower than that of Mixed and Cyto. In a separate analysis we compared synthesis rates of protein fractions between the two dog breeds (Table 2). There were significant differences in the protein synthetic rates of Mixed and Cyto, but not Mito, fractions in the sedentary dogs, and no differences between fractions in the exercised dogs. To our surprise, the DNA synthesis rates were too low to detect in both groups of dogs (data not shown).

Fig. 1.

Body weight over time (A) and water enrichment over time (B) in both Alaskan Huskies and Labrador Retrievers. There was a significant difference between sedentary and exercised Labrador Retrievers, but no change over time. Water enrichment rose over time in both dog breeds, thus calculations of synthesis were made using average enrichment for the labeling period. Values are means ± SE, and n = 4–12/group. *Significantly different between sedentary and exercised.

Fig. 2.

Protein fractional synthesis rates (FSR) of three different skeletal muscle protein fractions in Alaskan Huskies during sedentary conditions or at the onset of exercise (A), and Labrador Retrievers during sedentary conditions or during an exercise training program (B). Values are means ± SE, and n = 4–12/group *Significantly different between sedentary and exercised. ^#Significantly different between protein fractions.

Table 2.

Comparison of the synthesis rates of three protein fractions in Alaskan Huskies and Labrador Retrievers in the sedentary and exercised state presented in percent per day

	Rest			Exercise
	Alaskan Husky	Labrador Retriever	Human*	Alaskan Husky	Labrador Retriever	Human*
Mixed	2.28 ± 0.12	3.88 ± 0.37†	1.02 ± 0.13	2.20 ± 0.12	2.26 ± 0.19	1.22 ± 0.07
Cyto	2.91 ± 0.10	3.85 ± 0.60†	1.02 ± 0.14	2.90 ± 0.16	2.65 ± 0.19	1.22 ± 0.07
Mito	2.62 ± 0.07	2.92 ± 0.20	0.087 ± 0.10	2.66 ± 0.11	2.34 ± 0.15	1.04 ± 0.07

^*Human values are provided for reference and are adapted from our previous study in young human subjects undergoing an exercise training program (22). Values are means ± SE.

^†Significantly different between Alaskan Huskies and Labrador Retrievers.

Poststudy Western blot analysis of skeletal muscle of both Alaskan Huskies and Labrador Retrievers showed no differences in PGC-1α content (Fig. 3, A and B). Measures of p-RpS6/total RpS6 and p-ACC/total ACC were not different between sedentary and exercised Alaskan Huskies (Fig. 3, C and E), however, phospho-protein ratios were significantly lower in exercise-trained compared with sedentary Labrador Retrievers (Fig. 3, D and F).

Fig. 3.

Western blot analysis of skeletal muscle PGC1α (A and B), p-RpS6/total RpS6 (C and D), and p-ACC/total ACC (E and F) in sedentary and exercised Alaskan Huskies and Labrador Retrievers. Analysis was performed on biopsies taken at the conclusion of the testing period. Values are presented as percent of sedentary control and are means ± SE, and n = 4–12/group. *Significantly different between sedentary and exercised.

DISCUSSION

Canines are capable of extreme endurance performance, thus we hypothesized that mitochondrial proteins would be preferentially synthesized over other protein fractions at the initiation and during an exercise training program. Our main findings were that 1) Mito protein synthesis did not increase at the onset of an exercise training program when dogs were kept in energy balance; 2) dogs that were apparently in negative energy balance maintained Mito protein synthesis during exercise training, whereas Mixed and Cyto fractions were decreased; and 3) Alaskan Huskies and Labrador Retrievers, two active dog breeds, have very high rates of skeletal muscle protein synthesis in the sedentary state compared with humans.

Methodological approach.

For several reasons, we have advocated the use of ²H₂O for the measurement protein synthesis (10) and mitochondrial biogenesis (12). First, since protein translation is energetically expensive there is extensive regulation posttranscriptionally that determines whether or not a protein is synthesized. This regulation is especially important during periods in which energetic signaling is changed (5, 13). Therefore, markers of mitochondrial biogenesis such as cellular signaling or mRNA content have limitations when determining mitochondrial biogenesis (12). Second, using ²H₂O to measure synthetic responses over a period of days to weeks integrates all periods of the day and does not restrict one to measuring the response in one physiological state (e.g., postexercise, postfeeding, overnight fasted), which is common with techniques using the continuous infusion of labeled amino acids. Last, because of the ability to simultaneously assess DNA synthesis with ²H₂O, important tissue-specific insight into the role of protein synthesis to maintain existing cells vs. to accommodate cellular expansion is gained (4, 10).

This study is the first to apply the use of ²H₂O to dogs undergoing exercise training. One practical issue was that dogs were kept away from all water sources (e.g., puddles, streams, water bowls, etc.) that were not mixed with ²H₂O by the study investigators. In both studies, the amount of ²H₂O needed was slightly overestimated and there was a rising enrichment over the course of the study. However, through frequent blood sampling to account for the rising enrichment, we were able to use the average enrichment for the labeling period to calculate precursor enrichment. Because of the volume of ²H₂O required and the challenge associated with controlling the exercise training of two separate groups of dogs, we restricted the measurement period to 1 wk. The 1-wk period of measurement was shorter than the typical 2 to 6-wk measurement we have employed in the past with humans and rodents (4, 5, 13, 14, 19, 22). However, during this period of measurement 15–25% of the proteins were newly synthesized, giving us confidence that an integrated response of varying protein pools was measured.

Protein and DNA synthesis at the onset of exercise training.

Our first goal was to investigate protein synthesis and DNA synthesis at the onset of exercise training in the dogs. The training and racing season for Alaskan Huskies typically begins in September/October and lasts until late March/early April. Training distances used in this study were modest (15.5 ± 1.5 km/day) to simulate the initiation of training. Although the Alaskan Huskies were not undergoing any exercise training prior to study, the exercise stimulus used in this study did not seem to change protein synthetic rates. In addition, there was no detectable DNA synthesis, although an increase in DNA synthesis is something we have observed at the onset of exercise training in human subjects (19). In both sedentary and exercise-trained dogs, Mixed protein synthesis, which is largely reflective of myofibrillar protein synthesis, was slower than the Cyto fraction. These data indicate that metabolic priorities did not shift at the onset of exercise. From these data, we conclude that three bouts of 15 km were not enough to increase protein synthesis over the sedentary condition.

Protein and DNA synthesis during exercise training.

Our second goal was to investigate skeletal muscle protein and DNA synthesis during an exercise training period. For this study the Labrador Retrievers were engaged in an exercise training program for 8 wk prior to testing. Entering the study the sedentary dogs were significantly heavier than the exercise-trained group and this difference was maintained during the experimental time period. During the testing period, the Labradors covered ∼30 km/day and we attempted to match the expected energy expenditure with energy intake based on data from the lead-in period. Despite these efforts, p-RpS6, a downstream indicator of mTOR activity, was decreased in the exercised animals compared with the sedentary animals indicating that the exercised dogs were potentially in negative energy balance because of the inhibitory effect of AMPK activation on mTOR activity. However, AMPK activation is also known to increase p-ACC, whereas the exercised dogs had decreased p-ACC/total, thus clouding the interpretation of energy balance.

Whether or not there was a true change in energy balance throughout the exercise period, there were clear compensatory changes in the exercising dogs compared with sedentary dogs as indicated by protein synthetic rates. Interestingly, the protein synthetic rate of the Mixed and Cyto fractions of these exercising dogs were slower than the sedentary dogs, but the Mito fraction was maintained. The maintenance of mitochondrial protein synthesis (mitochondrial biogenesis) when other protein fractions decrease synthesis rates is something we have observed previously in a rodent model treated with rapamycin (5, 10). Our overall interpretation is that the selective translation of mitochondrial proteins helps to maintain efficient energy production without the energetic cost associated with increases in global protein synthesis. Importantly, the synthesis of new proteins, even in the absence of a net gain in mitochondrial proteins, could improve mitochondrial function to improve energy production. Further studies on mitochondrial function are needed to confirm this speculation. A second possibility for decreased Mixed and Cyto fractions in the exercised dogs is that amino acids from muscle protein could have been used as an additional energy source in the exercising dogs, as we have previously documented high rates of gluconeogenesis in dogs (11). By this rationale, there could be some loss of labeled protein, thus underestimating protein synthesis rates. Although this scenario is possible, in studies of rodents with breakdown rates that approximate the synthesis rates in the Labradors (8), the loss of label would not account for the difference in synthesis rates between the sedentary and exercised groups. A more likely explanation is that some of the dietary intake of protein is directed immediately to energy production, thus decreasing overall synthesis rates, in which case the measured synthetic rates represent a real difference between sedentary and exercised dogs.

Comparison between breeds and to humans.

We previously demonstrated that Alaskan Huskies are capable of extreme metabolic changes for prolonged exercise (11). It is not known if this plasticity is unique to Alaskan Huskies or if other breeds share this capacity. To address this question, we made comparisons of protein synthesis between the two breeds in our study (Table 2). In the rested dogs, the rates of protein synthesis of the Mixed and Cyto fractions were higher in the Labrador Retrievers compared with the Alaskan Huskies. In the exercised condition, there were no differences between the breeds. Since we did not obtain a final blood draw on the day of the biopsy in the Labrador Retrievers (but did in the Alaskan Huskies), it is possible that a precursor enrichment that continued to rise slightly would result in a slightly lower synthesis rate in the Labrador Retrievers. One interpretation of the data is that the rested Alaskan Huskies keep a proportionally faster Mito protein synthesis, relative to the other two fractions, during the untrained state. This interpretation implies that the Alaskan Huskies maintain a relatively greater (as compared with Laboratory Retrievers) aerobic phenotype, which would be commensurate with the known highly aerobic nature of the breed (17), even in the untrained state (1). A second interpretation is that rather than the exercised Labrador Retrievers being in a negative energy balance, the sedentary Labrador Retrievers were actually in a positive energy balance and the greater synthesis rates of Mixed and Cyto fractions were from weight gain. In support of this idea is the greater energy intake of the sedentary Labrador Retrievers (207 ± 23 kcal·kg^0.75·day⁻¹) compared with the Alaskan Huskies (64 ± 6 kcal·kg^0.75·day⁻¹). However, in the absence of biopsies taken prior to initiation of the exercise training in the Labrador Retrievers, it is hard to make this determination. Although there is evidence from rodents that weight gain does not selectively change the synthesis rates of protein fractions (16), it is unknown whether this is true in dogs.

Finally, it is very informative to compare the values from the dogs to our previously published values in humans (22) (Table 2). For this purpose, reference values are presented, but statistical comparisons not performed because of methodological differences in studies. At rest, the dogs had two-to-four times greater protein synthetic rates in the three fractions of proteins, while during exercise this decreases to roughly two times that of humans. Although the smaller difference in protein synthesis during the exercise periods are the result of a nonchange in the Alaskan Huskies and a decrease from rest to exercise in the Labrador Retrievers, the data illustrate the extremely high protein synthetic rates in dogs during sedentary conditions. The tradeoff is interesting in that the maintenance of high protein synthetic rates is energetically very costly (20). However, the ability to adjust to extreme environmental changes, which dogs do both quickly and comprehensively (3, 9, 11), could be facilitated by this excess capacity. This very high resting protein synthesis could also help explain why the onset of exercise training did not result in increases in Mito protein synthesis. Future studies are needed to determine how much further the dogs can adjust to physiological stresses and the mechanisms of further acclimatization.

Limitations and future studies.

We acknowledge several limitations of the current work. The study was conducted as a field research project and was thus limited by the complications of that style of research. However, that it is a field study is also its strength. Second, we were working with some of the most elite terrestrial aerobic athletes known, thus we were somewhat restricted in our access to the animals as is the case with most studies of elite athletes. Since the Labrador Retrievers were engaged in an 8-wk training program prior to study, we were not able to obtain a pretraining muscle biopsy, thus limiting comparisons to the sedentary dogs rather than pre-and/or posttesting. The differences in weights maintained during the trial and the difference between AMPK related outcomes therefore lead to ambiguity as to when changes in negative balance occurred. Second, the two dog breeds were tested in two different locations at two different times. However, both studies were performed in June, and because of unusually warm conditions in Alaska, both were performed in ambient temperatures of 27–34°C. Last, labeling for 1 wk may have been an insufficient period of time to have DNA incorporation that was within the limits of detection of the GC-MS, thus limiting our ability to interpret proteostatic mechanisms (4, 10). Current studies are in progress to determine mitochondrial function, to determine whether changes occur in the absence of changes in mitochondrial protein synthesis, and to further elucidate mechanisms of physiological remodeling under various environmental stresses.

Conclusions.

The canine species has an incredible capacity to adjust to environmental challenges, including extreme exercise. Here we demonstrate that contrary to our hypothesis, canines do not have large increases in mitochondrial protein synthesis at the onset or during an exercise training program. However, sedentary dogs have a high rate of protein synthesis that perhaps does not necessitate an extra increase in protein synthesis at the onset of aerobic exercise training. Further, when an additional stress is added, potentially a negative energy balance, mitochondrial protein synthesis is maintained while the synthesis of other protein fractions are decreased. Future studies are needed in the dogs to explore whether skeletal muscle remodeling and mitochondrial function improvements occur during exercise training without a net increase in protein synthesis over sedentary values.

GRANTS

This work was supported by the United States of America Army Research Office Division of Life Sciences Awards W911NF0910549 and W911NF-13-1-0091 and the Office of Naval Research Award N00014-11-C-0493.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

B.F.M., S.P.-P., M.D., and K.L.H. conception and design of research; B.F.M., S.E.E., J.C.D., L.M.B., S.P.-P., M.D., and K.L.H. performed experiments; B.F.M., S.E.E., J.C.D., F.F.P., L.M.B., M.D., and K.L.H. analyzed data; B.F.M., S.E.E., M.D., and K.L.H. interpreted results of experiments; B.F.M. and S.E.E. prepared figures; B.F.M. drafted manuscript; B.F.M., S.E.E., J.C.D., S.P.-P., M.D., and K.L.H. edited and revised manuscript; B.F.M., S.E.E., J.C.D., F.F.P., L.M.B., S.P.-P., M.D., and K.L.H. approved final version of manuscript.