Previously, we demonstrated that high-volume resistance exercise stimulates mitochondrial protein synthesis (a measure of mitochondrial biogenesis) in lean but not obese Zucker rats. Here, we examined factors involved in regulating mitochondrial biogenesis in the same animals. PGC-1α was 45% higher following exercise in obese but not lean animals compared with sedentary counterparts. Interestingly, exercised animals demonstrated greater PPARδ protein in both lean (47%) and obese (>200%) animals. AMPK phosphorylation (300%) and CPT-I protein (30%) were elevated by exercise in lean animals only, indicating improved substrate availability/flux. These findings suggest that, despite PGC-1α induction, obese animals were resistant to exercise-induced synthesis of new mitochondrial and oxidative protein. Previously, we reported that most anabolic processes are upregulated in these same obese animals regardless of exercise, so the purpose of this study was to assess specific factors associated with the mitochondrial genome as possible culprits for impaired mitochondrial biogenesis. Exercise resulted in higher mRNA contents of mitochondrial transcription factor A (∼50% in each phenotype) and mitochondrial translation initiation factor 2 (31 and 47% in lean and obese, respectively). However, mitochondrial translation elongation factor-Tu mRNA was higher following exercise in lean animals only (40%), suggesting aberrant regulation of mitochondrial translation elongation as a possible culprit in impaired mitochondrial biogenesis following exercise with obesity.
- peroxisome proliferator-activated receptor-γ coactivator-1α
- mitochondrial protein synthesis
- mitochondrial transcription and translation
- oxidative metabolism
the prevalence of obesity and diabetes has risen to epidemic proportions in the US (3). These disease states are often associated with dysregulated metabolism, including reduced glucose uptake, altered protein turnover, and dysregulated lipid metabolism. In particular, obesity is associated with increased rates of incomplete mitochondrial β-oxidation, an inability to switch to carbohydrate fuel supplies, and the accumulation of intramuscular acylcarnitines (12, 18). These disruptions are often associated with poor mitochondrial quality and impaired ability to build new mitochondria (mitochondrial biogenesis). Furthermore, although some debate remains, mitochondrial dysfunction has been implicated in the development of insulin resistance (4, 17, 29, 35). Kelley et al. (16) have reported reduced mitochondrial function concurrent with the appearance of lipid vacuoles identified as being of mitochondrial origin in skeletal muscles of diabetic patients. In fact, mitochondrial dysfunction is proposed to be a primary defect leading to insulin resistance (4). One of the primary means by which mitochondrial function may be enhanced is through the biogenesis of new mitochondrial components. The assessment of mitochondrial biogenesis has been made by several techniques. Miller and Hamilton (23) proposed that the use of mitochondrial protein synthesis rates (mtFSR) may be the closest true measurement of mitochondrial biogenesis, as these measurements describe novel addition of protein components to the mitochondria in “real time.” Using this technique, we reported in a previous study by Nilsson et al. (26) an impaired ability to stimulate mtFSR following high-volume resistance exercise (RE) in obese/insulin resistant Zucker rats despite an observed RE-induced mtFSR in lean counterparts. Therefore, utilizing the same animals and tissues, the present investigation was performed to elucidate potential new mechanisms that may contribute to impaired stimulation of mitochondrial biogenesis in the obese Zucker rat, a widely accepted model of obesity and insulin resistance.
Peroxisome proliferator-activated receptor (PPAR)-γ coactivator-1α (PGC-1α) is considered to be the “master regulator of mitochondrial biogenesis” (21). PGC-1α functions through transcriptional coactivation, stimulating many transcription factors, including the PPAR family of transcription factors, nuclear respiratory factors, transcription factor of activated mitochondria (TFAM), and others (24, 27). The commonality of these transcription factors is that they all serve to stimulate the transcriptional activation of genes necessary for mitochondrial biogenesis. In fact, data from muscle-specific PGC-1α-knockout mice demonstrate that in the absence of PGC-1α, exercise training-induced mitochondrial biogenesis is severely attenuated (11). In normal healthy conditions, exercise training, including both endurance and resistance exercise, enhances PGC-1α expression due to transcriptional regulation (1, 6, 10). Furthermore, Sriwijitkamol et al. (33) have demonstrated previously that 7 wk of aerobic exercise training induced elevated PGC-1α protein expression in obese Zucker rats.
However, as independent organelles with their own genomes, including 13 protein-encoding genes, all of which are critical to the function of the electron transport system, the mitochondria are necessarily equipped with their own equipment for transcription and translation. Whereas PGC-1α and exercise are known to induce the activity of TFAM and thereby mitochondrial DNA transcription (15, 30), the impact of exercise on mRNA translation of the mitochondrial genome is relatively unknown. In fact, a review of the literature shows no previous report on the exercise induction of the mitochondrial translation machinery. However, these steps may be of critical importance, as the mitochondrial genome includes 13 protein-encoding genes, each of which are necessary for the electron transport system and thereby mitochondrial function. Of note to the present investigation, a recent report by Mercader et al. (22) implicates two essential proteins for the mitochondrial translation process as novel candidate genes in human type 2 diabetes. Those proteins, mitochondrial translation initiation factor 2 (mtIF2) and mitochondrial translation elongation factor-Tu (TUFM; also termed mtEF-TU, COXPD4, P43), exhibit primary responsibility for the translation initiation and elongation processes, respectively. mtIF2 and TUFM are highly conserved nuclear-encoded proteins translated within the cytosol by the canonical translation machinery and imported to the mitochondria. There, they are necessary components of the mitochondrial translation process within the mitochondrial matrix. Currently, it is believed that the initiation of translation occurs upon mtIF2 binding to the mitoribosomal small subunit (SSU), which carries fMet-tRNA to the SSU, which is enhanced by the presence of GTP. GTP hydrolysis appears to allow the release of mtIF2 from the mitoribosomal complex, leading to elongation. GTP-TUFM then carries tRNA to the SSU at the aminoacyl acceptor site for translation of the mRNA; following proofreading, TUFM is then similarly released from the unit through GTP hydrolysis, allowing another tRNA-TUFM to come in. This process continues to the end of the mRNA via a similar stop codon mechanism to nuclear translation. Smits et al. (32) have previously published a more detailed review of this process. The precise regulation of each step and protein involved is still undetermined. However, it appears that a number of transcription factors exhibit potential binding sites to the mtIF2 and TUFM promoters, and additionally, we have noted 27 predicted binding sites for posttranslational modification of TUFM (14). Therefore, it stands to reason that complete mitochondrial biogenesis requires the upregulation of the mitochondrial translation machinery to add these essential components.
The current study was performed to examine potential mechanisms behind impaired mitochondrial biogenesis in the obese/insulin resistant condition. We initially hypothesized that PGC-1α induction would be impaired in these animals. However, these assessments revealed an induction of PGC-1α mRNA and protein content in the obese Zucker rat following RE, thereby uncoupling PGC-1α expression from measured mitochondrial biogenesis. This is similar to a previous report by Holloway et al. (13), who showed in lean and obese women that PGC-1α was uncoupled from fatty acid oxidation in obesity. Considering this finding and given our previous reports indicating most anabolic processes to be upregulated in obese animals (25, 26) regardless of exercise, we elected to focus on specific anabolic factors associated with the mitochondrial genome as possible culprits for the lack of mitochondrial biogenesis in these animals. Therefore, we amended our hypothesis that exercise would induce transcription and translation of the mitochondrial genome but that said induction would be impaired in the obese/insulin resistant animal. In fact, in the same animals in which we previously reported an impaired exercise-induced increase in mtFSR, we have now observed an aberrant regulation of the mitochondrial elongation factor TUFM in the obese Zucker rat following RE, suggesting a novel target to combat mitochondrial dysfunction in obesity and diabetes.
MATERIALS AND METHODS
Lean (fa/−; n = 16) and obese (fa/fa; n = 14) Zucker rats were purchased (Charles River Laboratories, Wilmington, MA) at 16 wk of age, and all methods were approved by the Institutional Animal Care and Use Committee of Texas A & M University. All animals were singly housed in a secure temperature- and humidity-controlled environment and maintained on a constant 12:12-h light-dark cycle with food and water provided ad libitum. Animals were assigned to one of four conditions: lean-sedentary (SED; n = 8), lean-resistance exercised (RE; n = 8), obese-SED (n = 6), and obese-RE (n = 8). Prior to the study, animals were weighed and analyzed for body composition by dual-energy X-ray absorptiometry scans. To perform scans, animals were anesthetized using a ketamine (Ketaset; 37.5 mg/kg body wt) and medetomidine (Domitor; 0.25 mg/kg body wt) cocktail administered by intraperitoneal (ip) injection and immediately awakened postscan by administration of antisedan. Using these data, animals were matched such that body weight and composition were similar within lean and obese phenotypes prior to exercise training (26).
Operant conditioning and RE.
All animals were operantly conditioned to the RE protocol over a period of 2 wk consisting of six total sessions. The exercise protocol has been described previously (9) and resembles that of a traditional “squat” performed by humans in a weight room. Briefly, rats were taught to press an illuminated lever in a specially designed cage to avoid a brief foot shock stimulus (<3 mA, 60 Hz; 1–5 V). This movement facilitates a full extension and flexion of the hindlimb. During the final two conditioning sessions, a nonweighted Velcro vest was placed over the scapulae, and the animals were required to perform the same movement. Once operantly conditioned, animals performed the exercise with little or no requirement for shock.
The experimental protocol consisted of four RE sessions (RE1–RE4) over an 8-day period and was progressive in nature, with weight and repetitions increased with each session such that during RE3 and RE4 animals performed an average of 88 repetitions per session. During the RE sessions the Velcro vest was placed over the scapulae, and additional weight was added to the vest. All RE animals received the same absolute amount of additional overload such that during RE4 animals performed 80 (18 repetitions), 130 (16 repetitions), 180 (14 repetitions), 230 (30 repetitions), and 280 g (6 repetitions). SED animals experienced only normal cage activity during the experiment. The total number of shocks administered to each RE animal was recorded, and SED animals were given the average number of shocks received by the RE animals. Operant conditioning and experimental RE sessions were all separated by 48–72 h and conducted during the animals' dark cycle. Muscles were harvested and animals euthanized on the 9th day.
Tissue and blood collection.
Consistent with previous research (8), all tissue samples were collected 16 h following exercise. Fasting blood samples (12-h fast) were collected from the lateral saphenous vein 16 h following RE3 for analysis of serum lipids and cholesterols. To allow for a better comparison of the fasting blood samples to the post RE4 condition, the total volume of work (weight × repetitions) completed during RE3 and RE4 was matched approximately. Blood samples were again collected by cardiac puncture at the time of euthanization (post-RE4). All blood samples were collected in vacutainers containing either EDTA for plasma separation or a gel with a clot activator for serum separation and frozen as plasma and serum at −80°C until analysis.
On the morning of muscle harvest, rat chow was withdrawn 4 h prior to muscle excision, and the animals were transported to the core laboratory. Approximately 16 h following RE4, animals were anesthetized using ketamine (Ketaset; 37.5 mg/kg body wt) and medetomidine (Domitor; 0.25 mg/kg body wt) by ip injection. Two milliliters of whole blood was collected by cardiac puncture; the hindlimb muscles were quickly harvested and the animal euthanized. Fat, blood, and connective tissue were removed from muscles before snap-freezing in liquid nitrogen and were subsequently pulverized before storing at −80°C until further analyses. Each group (lean RE, lean SED, obese RE, and obese SED) was represented on any given experimental day and selected in a random order to eliminate potential bias.
Assessment of blood lipids and lipoproteins.
All lipid and lipoprotein analyses were performed using serum collected after RE3 and RE4. Analysis of serum nonesterified fatty acids (NEFA), total cholesterol, LDL cholesterol, and HDL cholesterol was performed colorimetrically using commercially available kits (Wako Diagnostics, Richmond, VA). All lipid and lipoprotein concentrations are expressed in millimolars per liter.
Isolation of protein and Western blot analysis.
All muscle tissue analyses were performed using mixed gastrocnemius. Western blot analysis of proteins was performed as described previously, with minor modifications (7). Briefly, whole tissue was weighed and powdered at the temperature of liquid nitrogen, and then 20 mg was homogenized in cold buffer (25 mM HEPES, 4 mM EDTA, 25 mM benzamidine, 1 μM leupeptin, 1 μM pepstatin, 0.15 μM aprotinin, and 2 mM phenylmethylsulfonyl fluoride, pH 7.4). The homogenate was then centrifuged (10,000 g for 30 min at 4°C). Protein concentration of the supernatant was determined according to the bicinchoninic acid method described by Smith et al. (31). Samples were stored at −80°C until sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis.
Before gel electrophoresis, an aliquot of the supernatant was diluted in an equal volume of buffer (125 mM Tris, pH 6.8, 4% SDS, 20% glycerol, 200 mM DTT, and 0.002% bromophenol blue). Protein was then separated by electrophoresis across an 8% polyacrylamide gel and transferred to nitrocellulose membrane (Amersham Biosciences, Piscataway, NJ). Membranes were then incubated in blocking solution (containing 5% nonfat dried milk in Tris-buffered saline) at room temperature. Following blocking, membranes were incubated with rabbit anti-PPARδ, PGC-1α, carnitine palmitoyltranserase (CPT)-Iβ, FAT/CD36, F1 ATPase, mtIF2 (Santa Cruz Biotechnology, Santa Cruz, CA), AMPKα, phospho-AMPKα (Thr172), cytochrome c oxidase IV (COX-IV), TFAM (Cell Signaling Technology, Danvers, MA), and TUFM (Santa Cruz Biotechnology and Abcam, Cambridge, MA). The membranes were then washed and incubated again with anti-rabbit IgG coupled to horseradish peroxidase (Cell Signaling Technology) and developed using chemiluminescence and imaged using an Alpha Innotech [PPARδ, PGC-1α, CPT-Iβ, FAT/CD36, F1 ATPase, AMPKα, phospho-AMPKα (Thr172), COX-IV; Alpha Innotech, FluorChem SP, San Leandro, CA] or LiCor C-DiGiT Blot Scanner (TFAM, mtIF2, TUFM; Li-Cor Biosciences, Lincoln, NE). Because of technical issues, blots imaged with Alpha Innotech were normalized to Ponceau S stain, and blots imaged with LiCor were normalized to GAPDH. Absorbance was normalized to an internal control protein standard (obtained from rat quadriceps; the same sample was loaded to each gel and used to control for gel to gel variability) loaded on each gel and expressed as arbitrary units. Each experimental group was equally represented on each gel.
RNA isolation, cDNA synthesis, and quantitative real-time PCR.
RNA isolation, cDNA synthesis, and quantitative real-time PCR were performed as reported previously (36), with minor modifications. Briefly, RNA was extracted with Trizol reagent (Life Technologies, Grand Island, NY), as suggested by the manufacturer. Total RNA was isolated and DNase treated, and concentration and purity was determined by fluorometry using the Qubit 2.0 protocol (Life Technologies). cDNA was reverse transcribed from 1 μg of total RNA using Quanta qScript cDNA Supermix (Quanta BioSceinces, Gaithersburg, MD) according to the manufacturer's instructions. Real Time Polymerase Chain Reaction (PCR) was performed using the StepOne Real-Time PCR system (Life Technologies and Applied Biosystems, Grand Island, NY), and results were analyzed using StepOne Software. cDNA was amplified in a 25-μl reaction containing appropriate primer pairs and TaqMan Universal Mastermix (Applied Biosystems). Samples were incubated at 95°C for 4 min, followed by 40 cycles of denaturation, annealing, and extension at 95, 60, and 72°C. TaqMan fluorescence was measured at the end of the extension step for each cycle. Fluorescence-labeled probes for PGC-1α, TFAM, mtIF2, TUFM (FAM dyes), and 18s (VIC dye) were purchased from Applied Biosystems and quantified with TaqMan Universal mastermix. Cycle threshold (CT) was determined, and the ΔCT value was calculated as the difference between CT value and 18s CT value. 18S CT values were unaffected by either obesity or exercise. Final quantification of gene expression was calculated using the ΔΔCT method. Relative quantification was calculated as 2−ΔΔCT.
The independent factors in this study were exercise (RE vs. SED) and phenotype (lean vs. obese). Dependent variables of interest included protein content of PGC-1α, PPARδ, AMPKα, select oxidative and mitochondrial target proteins (CD36/FAT, CPT-I, COX-IV, and F1 ATPase) TFAM, mtIF2, and TUFM, AMPKα relative phosphorylation, mRNA content of PGC-1α, TFAM, mtIF2, and TUFM, and serum lipid and lipoprotein concentrations. An exercise by phenotype (2 × 2) ANOVA was employed as the global analysis for each dependent variable of interest. The comparison-wise error rate α was set at 0.05 for all statistical tests. When significant F ratios were found, a Fisher's least significant difference post hoc analysis was used to distinguish differences among means. Correlations between PPARδ content and serum NEFA were assessed by Pearson's product-moment correlations. All data were analyzed using the Statistical Analysis System (version 9.3; SAS, Cary, NC) and expressed as means ± SE.
Serum lipid and lipoprotein concentrations are greater in obese compared with lean Zucker rats.
RE did not alter serum lipids and lipoproteins in either the fasted (RE3) or fed (RE4) state; therefore, data were collapsed across exercise treatments for each phenotype (Table 1). Serum NEFA concentrations were greater in obese compared with lean rats (+0.6 mmol/l post-RE3, +0.2 mmol/l post-RE4). Total cholesterol was also greater in obese rats (+6.2 mmol/l post-RE3, +1.89 mmol/l post-RE4). No differences were observed in LDL cholesterol concentrations between phenotypes. Additionally, obese animals exhibited elevated serum insulin (∼5- to 6-fold) and glucose (∼25% greater) (Ref. 26 and data not shown).
RE enhances protein expression of PGC-1α and PPARδ in the obese Zucker rat but fails to induce AMPK phosphorylation.
RE obese Zucker rats demonstrated an ∼45% greater PGC-1α protein content than obese sedentary animals (Fig. 1A), with no effect of RE on PGC-1α protein in lean animals. Similarly, Pgc-1α mRNA was induced 65% with exercise training in obese animals only (Fig. 1B). PPARδ protein content was greater in both lean (47%) and obese (>200%) RE animals compared with respective sedentary controls (Fig. 1C). No differences were observed for total AMPKα protein content; however, relative AMPKα phosphorylation at Thr172 was ∼300% greater in lean RE than all other groups (Fig. 1D). Of interest, because PPARδ is a proposed fatty acid sensor (28), we assessed the correlation between PPARδ protein content and serum NEFA concentration and found a significant correlation in lean animals (r = 0.741, P = 0.001) but not in obese animals (Fig. 1, E and F). Sample immunoblots are presented in Fig. 4.
RE fails to induce expression of proteins in oxidative function and fat metabolism in the obese Zucker rat.
COX-IV protein content, a marker for mitochondrial volume, was unaffected by RE in lean animals (Fig. 2A). However, COX-IV content was ∼60% greater in sedentary obese animals compared with lean animals. These data further showed reduced COX-IV protein content in obese animals following RE, indicating normalized COX-IV as a result of RE. (Fig. 2A). Mean CD36/FAT content trended toward an increase in lean RE animals compared with sedentary animals (13%), although this difference was not significant (Fig. 2B). CPT-I content was 30% greater in lean RE animals compared with lean sedentary animals (Fig. 2C). However, CPT-I content was not significantly different between obese sedentary and obese RE animals. Conversely, F1 ATPase, a subunit of the ATP synthase, was greater in obese RE than obese sedentary animals (37%), with no difference seen between lean groups. Sample immunoblots are presented in Fig. 4.
RE induces expression of mitochondrial transcription and translation factors.
In our previous work, Nilsson et al. (26) demonstrated that RE induces enhanced mitochondrial protein synthesis in the lean but not obese Zucker rat. In the same animals and tissues, we now report that this occurred despite a profound increase in PGC-1α expression in muscle from obese rats following exercise. The present investigation sought to understand this obesity-related mitochondrial anabolic resistance to exercise by assessing potential defects between PGC-1α signaling and mitochondrial protein synthesis, including key components of the transcription and translation machinery of the mitochondrial genome itself. These analyses revealed greater TFAM mRNA and protein content (∼50% greater) in RE compared with sedentary in both lean and obese phenotypes (Fig. 3, A and B). Similarly, mtIF2 was ∼31% greater in RE compared with sedentary in lean rats and 47% greater following RE in obese rats (Fig. 3, C and D). Conversely, TUFM mRNA content was ∼40% greater in lean RE compared with lean sedentary animals, with no difference between obese sedentary and obese RE groups (Fig. 3E). In contrast, TUFM protein content was reduced 71% following exercise in lean animals, although it was not significantly affected by exercise in obese animals (data confirmed using two antibodies to TUFM protein; Fig. 3F). Sample immunoblots are presented in Fig. 4.
Previously, we reported that high-volume RE induces mtFSR, a measure of mitochondrial biogenesis in lean but not obese Zucker rats (26). From these same animals and tissues (mixed gastrocnemius muscle), we have now shown that despite this apparent induction in mitochondrial biogenesis in lean rats only, we now report that PGC-1α mRNA and protein expression are induced in obese but not lean rats following high-volume RE. RE led to enhanced expression of the transcription factor PPARδ in both lean and obese phenotypes. Interestingly, similarly to mtFSR, CPT-I content and AMPK phosphorylation were enhanced following RE in lean animals only, and CD36/FAT content also appeared greater following RE in lean rats only (P > 0.05). This apparent disconnect between PGC-1α expression and mitochondrial biogenesis and oxidative markers led us to examine potential mechanisms by which mitochondrial biogenesis may be impaired downstream of PGC-1α in obese/insulin-resistant animals. Although we now know that the regulation of mitochondrial quality and function is a multifaceted process (37), the addition of new and healthy mitochondrial units and proteins is vital to the ability to enhance total cell mitochondrial function. Therefore, the mitochondrial anabolic resistance to exercise we observed previously in these obese Zucker rats could severely limit mitochondrial quality. Furthermore, given that our previous report also indicated that most anabolic processes are upregulated in obese animals (25, 26) with or without exercise, we elected to focus our attention on specific anabolic factors associated with the mitochondrial genome as possible culprits contributing to the lack of mitochondrial biogenesis in these animals, specifically the regulation of the transcription and translation processes of the mitochondrial genome. Similar to previous reports (15, 30), these analyses revealed that TFAM mRNA and protein are elevated following exercise. To our knowledge, we are the first to report that exercise induces an increase in the expression of mtIF2. Interestingly, our analysis of TUFM mRNA specifically shows exercise induction in lean animals only, similar to mtFSR, whereas TUFM protein content is reduced significantly following exercise in the lean but not obese Zucker rat. It is notable that TUFM mRNA content appears to already be elevated in obese compared with lean animals in the sedentary condition, which could represent a ceiling on TUFM mRNA expression. To our knowledge, TUFM is the primary protein regulating translation elongation of the mitochondrial genome, and therefore, it is critical to the completion of mitochondrial translation. Therefore, these data suggest that the lack of an exercise-induced proliferation of mitochondrial proteins in the obese Zucker rat is due to an impaired response of translation elongation, limiting the development of a new mitochondrial protein.
As stated previously, PGC-1α is considered the primary regulator of mitochondrial biogenesis. Therefore, it stands to reason that any intervention targeted to induce the expression or activity of PGC-1α would greatly benefit the mitochondrial phenotype of the given tissue. However, in the current study we observed a rise in PGC-1α only in obese animals, where induction of other markers of mitochondrial biogenesis (mtFSR, COX-IV) and enhanced oxidative metabolism (protein expression of CPT-I, CD36/FAT, and AMPK phosphorylation) were not present. We should note that our finding that CD36/FAT content was not significantly different at baseline between lean and obese animals is consistent with a previous report by Stefanyk et al. (34). We are not the first to suggest that induction of PGC-1α may not always be associated directly with enhancement in oxidative phenotype. In fact, Holloway et al. (13) previously reported the loss of the association between PGC-1α and fatty acid oxidation in obese humans. Further enhancing this is that despite the induction of PPARδ in lean and obese animals, the relationship between PPARδ, thought to be a fatty acid sensor (28), and serum NEFA was lost in the obese animals, indicating an impaired sensory function of PPARδ. Although we did not directly assess mitochondrial function in the current study, these findings demonstrate an ability for exercise-induced upregulation of upstream regulators PGC-1α and PPARδ, with no measured induction of downstream targets involved in the oxidative adaptation to exercise.
In the current study, we have elected to focus our interpretation of the stimulation of mitochondrial biogenesis on the assessment of mtFSR that was reported previously in these same animals (26). In agreement with Miller and Hamilton (23), we have taken the stance that conventional measurements of mitochondrial density (by microscopy or mitochondrial protein content) or mitochondrial enzyme activity fail to directly assess the actual synthesis of new mitochondrial components. A review by Yan et al. (37) described the potential for the accumulation of damaged mitochondria. Such an accumulation would lead to an elevation in the content of mitochondrial components, which could be mistakenly identified as signs of mitochondrial biogenesis. This speculation is based on our previous report that mitochondrial protein is about 31% greater per unit wet muscle mass in obese than in lean animals (26). Furthermore, a specific marker of mitochondrial density, COX-IV content (a nuclear-encoded mitochondrial protein closely tied to mitochondrial volume), was highest in obese sedentary rats, but “normalized” (reduced) by RE. We hypothesize that elevated COX-IV content in obese SED was reflective of increased mitochondrial content as a result of impaired mitochondrial autophagy rather than de novo mitochondrial biogenesis. Unfortunately, attempts to assess autophagy markers in these animals were unsuccessful. However, we should note that in this study we did not directly assess mitochondrial function and quality. Nevertheless, Kelley et al. (16) reported reduced mitochondrial function concurrent with the appearance of lipid vacuoles identified by immunogold electron microscopy as being of mitochondrial origin in skeletal muscles of diabetic patients. Such an activated clearance of mitochondria, likely through mitophagy, would agree with the recent report by Lira et al. (20), which indicated that exercise training stimulates enhanced basal autophagy, perhaps leading to an enhanced clearance of damaged organelles and proteins.
Considering the implications that the obese Zucker rat is able to upregulate primary factors regulating mitochondrial biogenesis (PGC-1α and PPARδ) without a corresponding exercise-induced enhancement of mitochondrial biogenesis (mtFSR, COX-IV) or oxidative phenotype (CD36/FAT, CPT-I, and AMPK phosphorylation), we sought to determine potential mechanisms by which this signal is uncoupled from its downstream adaptations. Since most anabolic processes are upregulated in the obese animals, we focused on specific anabolic factors associated with the mitochondrial genome that were unresponsive or perceived as rate limiting following exercise. These analyses revealed that TFAM and mtIF2, regulators of mitochondrial transcription and translation initiation, respectively, were upregulated following RE in both lean and obese animals compared with their phenotypic sedentary controls. This finding suggests that obese animals are able to properly stimulate mitochondrial transcription and translation initiation following high-volume RE. However, in moving further downstream, we noted that TUFM mRNA, the primary factor in mitochondrial translation elongation, was enhanced only in lean animals in response to RE. Notably, we did observe a greater mean TUFM mRNA content in obese compared with lean animals prior to exercise, and although not significant, this observation may indicate a ceiling on TUFM gene expression. In contrast to mRNA content of TUFM, we observed a 71% reduction in TUFM protein following exercise in the lean animal with a blunted, nonsignificant (P = 0.27) decline in TUFM protein following exercise in the obese animals. We should note that TUFM protein following RE was not significantly different between lean and obese phenotypes; however, the diminished reduction in TUFM protein following RE in the obese compared with lean animals is notable. The disparity between TUFM mRNA and protein content suggests that TUFM is highly regulated both transcriptionally and posttranslationally. Utilizing resources available through the DECODE database (SA Biosciences, Valencia, CA), we found that this gene is the likely target of a number of transcription factors. Notably, the findings by Mercader et al. (22) demonstrate that mitochondrial translation factors are linked to insulin signaling through the NF-κB1/IκBKB pathway. Furthermore, TUFM associates specifically with insulin signaling through the canonical translation pathway via RPS6, eIF4E, and eEF2 (22). Therefore, it appears likely that similar mechanisms that govern the expression of the canonical nuclear genome translation pathway, such as YY1 (2, 5), may be involved in the regulation of TUFM; however, such an interpretation requires further investigation to elucidate. Finally, although these mechanisms may lead to better insight into the regulation of TUFM gene expression, they fail to explain why TUFM protein content is decreased following exercise in the lean animal.
The TUFM protein contains nine potential ubiquitination sites, making the protein highly susceptible to ubiquitin-mediated degradation (14). Additionally, Lei et al. (19) demonstrated recently that TUFM complexes with the Atg5/Atg12 autophagy complex. Together, we speculate that this complex may help to specifically carry TUFM to the autophagosome for selective degradation. Although further studies are necessary to best elucidate the mechanisms by which TUFM is regulated, we suggest that TUFM protein is activated at a high level following exercise, which leads to eventual damage and a high requirement for turnover of the protein, leading to elevated transcription of the gene concurrent with high rates of TUFM protein degradation. Interestingly, these processes, specifically as they relate to turnover of TUFM, appear blunted in the obese/insulin-resistant condition. As noted, to our knowledge, TUFM is the primary protein regulating the translation elongation process for the mitochondrial genome, and therefore, it would appear critical to the mitochondrial translation process. This effect is likely reflective of a bottleneck point at which the obese animal is incapable of inducing mitochondrial translation elongation and thereby fails to induce a measurable mitochondrial protein synthesis. Thus, it appears that obese animals are able to integrate the signals from high-volume RE to initiate an adaptive oxidative response (PGC-1α, PPARδ, TFAM, and mtIF2), but the failed mitochondrial biogenesis (mtFSR and COX-IV) and lack of enhancement in oxidative proteins (CPT-I, CD36/FAT, and AMPK phosphorylation) may be due to an inadequate mitochondrial translation elongation response. Our future work will investigate specifically whether targeting the mitochondrial translation process in an effort to overcome mitochondrial anabolic resistance will allow for improved mitochondrial quality and function in obesity and diabetes.
In summary, high-volume RE leads to elevated mtFSR in lean but not obese Zucker rats (26). The obese Zucker rat shows signs of attempted mitochondrial biogenesis and oxidative adaptations, as evidenced by the enhanced expression of PGC-1α and PPARδ. However, any predictable outcomes leading to mitochondrial biogenesis, such as elevations of mtFSR or oxidative/mitochondrial protein content, were not observed in this phenotype, strongly suggesting that culprits leading to impaired biogenesis in these animals are downstream of these important initiators. In support of this conclusion, we noted that lean and obese Zucker rats are equally able to stimulate expression of factors involved in mitochondrial genome transcription (TFAM) and translation initiation (mtIF2) in response to exercise. However, the regulation of mitochondrial translation elongation at the TUFM protein, the primary factor regulating mitochondrial translation elongation, appears to be aberrant in the obese rat, suggesting a failure to elicit a complete response of mitochondrial protein synthesis/anabolism following exercise in this condition. Therefore, we propose that stimulation of mitochondrial biogenesis with obesity/diabetes is impaired downstream of the activation of PGC-1α and PPARδ and that this limitation occurs at the level of mitochondrial translation elongation.
These studies were funded in part by the Sydney and J. L. Huffines Institute for Sports Medicine and Human Performance and the Texas Chapter of the American College of Sports Medicine.
The authors declare no conflicts of interest, financial or otherwise.
N.P.G., M.I.N., S.F.C., and J.D.F. contributed to the conception and design of the research; N.P.G., M.I.N., D.E.L., L.A.B., A.M.P., K.L.S., E.S.G., S.F.C., and J.D.F. performed the experiments; N.P.G., T.A.W., D.E.L., L.A.B., A.M.P., K.L.S., E.S.G., and J.D.F. analyzed the data; N.P.G., T.A.W., D.E.L., E.S.G., S.F.C., and J.D.F. interpreted the results of the experiments; N.P.G. and D.E.L. prepared the figures; N.P.G. drafted the manuscript; N.P.G., M.I.N., T.A.W., D.E.L., L.A.B., A.M.P., K.L.S., E.S.G., S.F.C., and J.D.F. edited and revised the manuscript; N.P.G., T.A.W., S.F.C., and J.D.F. approved the final version of the manuscript.
We thank the dedicated faculty, staff, and students of the Applied Exercise Science and Muscle Biology Laboratories of Texas A & M University and the Human Performance Laboratory of the University of Arkansas, without whom these studies would not have been possible.
- Copyright © 2014 the American Physiological Society