| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1Research Center for Genetic Medicine, Children's National Medical Center, Washington, District of Columbia; 2Division of Cardiology, Department of Medicine and Cell Biology, Duke University Medical Center, Durham; and 3Department of Exercise and Sport Science, and the Human Performance Laboratory and Diabetes/Obesity Center, East Carolina University, Greenville, North Carolina
Submitted 25 March 2004 ; accepted in final form 31 August 2004
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
1). All were also quickly downregulated by detraining, although the induction was not an acute response to activity. Protein and enzymatic assays were used to validate mRNA induction with aerobic training and loss with detraining (96 h to 2 wk) in 10 male and 10 female STRRIDE subjects. We propose that training coordinately increases the levels of aspartate aminotransferase 1, lactate dehydrogenase B, and pyruvate dehydrogenase-1 subunit, increasing glucose metabolism in muscle by liberating pyruvate for oxidative metabolism and, therefore, limiting lactate efflux. Serial measurement of fasting plasma lactate from 62 subjects from the same exercise group demonstrated a significant decrease of circulating lactate with training. We also found evidence for sex-specific molecular remodeling of muscle with ubiquinol-cytochrome c reductase core protein II, a component of mitochondrial respiratory complex III, which showed an increase after training that was specific to women. These biochemical adaptations complement existing molecular models for improved glucose tolerance with exercise intervention in prediabetic individuals. microarray; aspartate aminotransferase; lactate dehydrogenase
-oxidation of fatty acids occurs at rates similar to those at rest or lower (8, 49). Glycogenolysis and glycolysis can generate ATP more rapidly than oxidative metabolism but do so at the expense of metabolic economy, generating only 2 or 3 ATP molecules per plasma glucose or glycogen-derived glucosyl unit, respectively, compared with the 36–38 produced by oxidative metabolism (5). Skeletal muscle from untrained subjects is proportionately more reliant on glycogenolysis and glycolysis to generate ATP because it is less perfused and the mitochondrial reticulum is less elaborated (5, 24). This also limits the ability of muscle to utilize both blood-borne free fatty acids and intramuscular triglycerides as substrates for energy metabolism (5). However, when adapted to habitual aerobic exercise, muscle is able to utilize the oxidative catabolism of both substrates to generate ATP at higher absolute workloads than in the untrained state (5, 16). This is particularly relevant to the prediabetic metabolic syndrome, which is characterized by impaired glucose tolerance, insulin sensitivity, and mild to severe lipid abnormalities (26), and highlights the importance of exercise in limiting or reversing its progression to Type 2 diabetes.
Traditionally, studies of the metabolic remodeling of muscle have relied on biochemical and physiological measurements. Although these have been highly informative, it is difficult to parallelize data generation and thereby begin to develop a broader picture of muscle remodeling after training. Newly emerging mRNA profiling and proteomic methods can impart a "genomewide" assessment of muscle plasticity. The reproducibility and redundancy of the Affymetrix GeneChip platform has been particularly effective in producing high-quality gene expression data from skeletal muscle and in generating new hypotheses about its regenerative capacity and phenotypic plasticity (12, 13, 20, 21, 53). In a previous microarray-based pilot study using three male subjects from the studies of a targeted risk reduction intervention through defined exercise (STRRIDE) exercise intervention study, our laboratory (20) found that trained muscle produces large amounts of clot-destroying fibrinolytic proteins. This study showed that the systemic fibrinolytic state was favorably modulated by chronic exercise. This finding was particularly significant because the metabolic syndrome is an independent risk factor for cardiovascular disease (22, 51). To explore this model further, we applied similar methods toward a systematic evaluation of skeletal muscle transcriptional remodeling in response to chronic exercise training of overweight individuals with characteristics of the metabolic syndrome. Toward this end, we studied biopsies from three female and three male STRRIDE subjects each in the "high" exercise group (2,200 kcal/wk, 3 mo ramp up, 6 mo training) (26). We analyzed four muscle biopsies from each subject: one at entry, and three after 9 mo of aerobic training [24 h after the last bout, 96 h after the last bout, and 14 days after the last bout (detraining)]. We then used a validation set of seven men and seven women from the same group, as well as plasma samples (for fasting lactate) from larger numbers of subjects. Our rationale for studying this population was to increase understanding of aspects of the metabolic syndrome in participants in STRRIDE where serum lipid profiles and insulin resistance were normalized after the exercise intervention (22, 25).
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
O2)] of the STRRIDE study; three of each sex were used for mRNA profiling and seven for protein validation of mRNA results.
Women and men ranged from 40 to 64 yr of age, were sedentary (exercise < once weekly; peak O2 25.8–34.6 ml·kg–1·min–1) and overweight (body mass index 27.9–36.1), and exhibited signs of fasting hyperinsulinemia (fasting insulin 12.3 to 18.8 IU/ml) with mild to moderate lipid abnormalities (low-density lipoprotein 132–163 mg/dl and high-density lipoprotein 27–40 mg/dl). Skeletal muscle biopsies were obtained by the percutaneous needle biopsy technique using 1% lidocaine without epinephrine (26). Four serial vastus lateralis muscle biopsies (100–200 mg) were obtained with a triple pass of the bioptome needle: one on entry into the study, one after 9 mo (3-mo ramp up, 6-mo training) of exercise training (within 24 h after the last bout of exercise), and two detraining time points 96 h (4 days) and 14 days (2 wk) after the last exercise bout. Biopsy samples were flash frozen for histological analysis, and an average of 25 mg/biopsy were shipped on dry ice for gene expression profiling and protein and enzymatic analysis.
RNA isolation, two-round amplification, and expression profiling.
For the isolation of total RNA, twenty 10-µm sections (
5 mg) from the belly of each muscle biopsy material were homogenized by using the TRIzol reagent (Invitrogen, Carlsbad, CA), according to the manufacturer's protocol. Total RNAs were extracted concurrently from the muscle biopsies obtained at entry, after 9-mo exercise training, and after 96-h and 14-day detraining time points (26). A two-round cRNA amplification protocol (Affymetrix, Santa Clara, CA) was used to produce sufficient biotinylated cRNA target for Affymetrix analysis of the small-muscle biopsy material. This protocol has been shown to be highly reproducible when starting with as little as 10 ng of total RNA. Despite a small loss in sensitivity by microarray analysis, two-round amplified cRNA has been shown to correlate quantitatively with RT-PCR analysis of one round of linear cRNA amplification (3, 18, 20). Briefly, 100 ng of total RNA were converted to double-stranded cDNA by use of a T7-promoter primer. This cDNA was then used as a template to drive the production of cRNA using the Megascript T7 in vitro transcription kit (Ambion, Austin, TX). A small amount of the first round amplified cRNA (<300 ng) was used to synthesize double-stranded cDNA by using a random primer to drive the first strand synthesis and then the T7 promoter/primer to drive the second strand synthesis. This amplified cDNA was used to drive the synthesis of biotinylated cRNA from the T7 promoter using the standard Affymetrix protocol described previously (12). Biotinylated cRNAs for each exercise time point were hybridized to four separate Affymetrix Human U133A GeneChips such that each individual had his or her own four-chip temporal series. Procedures for cRNA preparation and GeneChip processing were performed as previously described (12), following quality control procedures that our laboratory has published (21).
Gene expression data analysis. Absolute analysis of Affymetrix "raw" data was conducted using Affymetrix Microarray Suite 5.0 (MAS). Scanned chip images were first scaled to a target intensity of 150 to account for variations in cRNA target hybridization efficiencies. Background and scaled noise were similarly averaged for all chips before analysis. Of the 24 chips analyzed, the average scaling factor was 2.4, the number of genes called "present" by MAS ranged from 30 to 39%, and the average 3'-to-5' ratio for GAPDH was 2.2. A full list of our in-house quality control standards developed for one- and two-round amplified cRNA is available at the following URL: http://microarray.cnmcresearch.org/pgaoutline-qcofsamples.asp.
The 18,400 transcripts (referred to hereafter as genes) represented on the Affymetrix Human U133A oligonucleotide array describe 14,500 well-characterized genes and 400 expressed sequence tags. In all, 22,000 probe sets were analyzed by using 22 oligonucleotide probes (11 distributed probe pairs) for each probe set, producing multiple independent measurements for each gene. Comparison of the hybridization signal from the perfect match and mismatch probes allows for a specificity measure of signal intensity and elimination of most nonspecific cross-hybridization signals. Values of intensity differences as well as ratios of each probe pair were used (statistically) to determine whether a gene is called present or absent within the analysis. Data analysis required >30% of profiles to show a present for that gene to be carried into the next analysis. GeneSpring 5.1 (Silicon Genetics, Redwood City, CA) was then used to filter potential exercise-responsive genes for statistical significance (P < 0.05; paired t-test and ANOVA for pairwise and time-series analyses, respectively) relative to the study entry. All statistical analyses were done using a longitudinal design, in which each subject served as his or her own control (time 0 at entry vs. three subsequent training and detraining time points).
In addition to the MAS 5.0 probe set analyses, we also reanalyzed the data set using the dChip 1.2 difference model (41). We have found this complementary analysis of the Affymetrix data files to provide superior signal-to-noise ratios in human muscle experiments (43a). dCHIP differs from MAS in two ways: the probe set algorithm used and the normalization method. dCHIP uses a project-based global normalization, rather than the chip-based normalization of MAS 5.0. We did not use a "present call filter" in this second analysis, allowing greater sensitivity for genes potentially excluded because of poor probe set performance in the MAS 5.0 analysis.
Western blot analysis. Skeletal muscle protein extracts were obtained from 10 serial 14-µm cryostat sections of frozen biopsy material that was then homogenized in chilled extraction buffer (50 mM Tris·HCl, pH 7.5, 250 mM sucrose, 1 mM EDTA, 0.1% Nonidet P-40) containing an anti-protease mixture (Roche, Indianapolis, IN). Protein concentration was determined (in triplicate) via a microplate-based Bio-Rad (Hercules, CA) Bradford assay using BSA to construct a relative standard curve of protein concentration. Protein samples were mixed 1:1 with 2 x SDS-PAGE loading buffer (80 mM Tris·HCl pH 6.8; 10% mercaptoethanol; 2% SDS; 10% glycerol; bromphenol blue) and heat denatured, and 10 µg of each sample were loaded onto a 4–20% SDS-PAGE gradient gel. After electrophoretic size fractionation, proteins were transferred to a nitrocellulose membrane (Amersham, Piscataway, NJ). Membranes were blocked for 1 h at room temperature in 5% nonfat powdered milk in Tris-buffered saline plus 0.1% Tween 20 (TBST). Membranes were incubated in TBST for 2 h at room temperature or at 4°C overnight with primary antibodies [GOT1 (Research Diagnostics, Flanders, NJ), PDHA1 and UQCRC2 (Molecular Probes, Eugene, OR), and LDHB (Sigma, St. Louis, MO)], washed, and incubated for 1 h at room temperature with peroxidase-conjugated secondary antibodies (Upstate Biotechnology, Lake Placid, NY). Signal was detected by using an enhanced chemiluminescence detection kit (Amersham) and Kodak X-Omat AR scientific imaging film (Fisher, Palo Alto, CA). Each nitrocellulose membrane and gel was subsequently stained with Coomassie brilliant blue to confirm equal sample loading and transfer on the basis of the density of the predominant myosin heavy chain band. Expression levels were quantified from scanned X-ray film images by UNSCANIT automated digitizing system (Silk Scientific, Orem, UT).
Lactate dehydrogenase isoform analysis.
A nondenaturing gel electrophoresis method was adapted for skeletal muscle lactate dehydrogenase (LDH) isoform analysis (40, 47). Briefly, an 8% resolving gel and a 3% stacking gel were polymerized in the absence of SDS. Twenty micrograms of protein suspended in native loading buffer (3 ml glycerol, 0.6 ml 50x running buffer, 6.4 ml H2O, bromphenol blue) were run in a cold room at 30 mA (stack) and then 40 mA (resolving) until the gel tracking dye was 1 cm from the bottom. Gels were washed with running buffer (50x diluted, 7.5 g Tris base, 36 g glycine, water to 250 ml) and then incubated in the dark at 37°C in an 80 mM potassium phosphate buffer (pH 7.4) with 0.1 M lithium lactate, 1 mM nitroblue tetrazolium, and 25 µM phenazine methosulfate. Gels were then scanned and dried for storage. Expression levels were quantified from scanned X-ray film images by use of UNSCANIT automated digitizing system (Silk Scientific).
Aspartate aminotransferase assay.
The kinetic assay for aspartate aminotransferase (GOT) couples the products of endogenous GOT with exogenous malate dehydrogenase (MDH) and the concomitant oxidation of NADH (2). GOT catalyzes the transfer of an amino group between L-aspartate and 2-oxoglutarate, forming oxaloacetate (OAA), which is subsequently reduced with NADH (in a 1:1 ratio) in the presence of MDH, forming oxidized NAD+, which can be monitored at 340 nm. GOT assay reagent (all reagents from Sigma) was composed of 12 mM 2-oxoglutarate, 200 mM L-aspartic acid, 0.19 mM NADH, and 600 U/l MDH in a 100 mM sodium phosphate buffer, pH 7.8. Reagent was prewarmed to 37°C for 1 min, and then the absorbance was blanked against distilled water. In a 96-well Labsystems Multiskan MCC/340 microplate reader (Fisher), 200 µl of reagent and 10 µg of protein were combined, and the initial absorbance at 340 nm was taken before and after 1-min incubation (linear portion of curve) for a total of 2 min. L-Aspartate was omitted in a parallel experiment to determine background levels of activity produced by endogenous -keto acids. For the purpose of this experiment, exogenous GOT was purchased (Sigma) to determine the linear range of the assay by a microplate procedure.
Plasma analysis. A fasting venous blood sample was obtained, and plasma was frozen at –80°C for the subsequent determination of lactate. Plasma was analyzed for lactate with an oxidation reaction (YSI 2300 STAT Plus glucose and lactate analyzer, YSI, Yellow Springs, OH) (6, 22).
Statistics and data analysis. Unless otherwise stated, all statistical analysis were conducted by using a paired Student's t-test. Significance for all comparisons was set at P < 0.05. Data are represented as means ± SE.
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Of the 18,400 genes probed using the Affymetrix human U133A arrays after scaling, correcting for background, and filtering using flags (present calls) and statistical analysis (P < 0.05 for paired biological replicates), 429 genes were differentially expressed in skeletal muscle of women after 9 mo of aerobic exercise training on the basis of P value with no adjustment for multiple testing. We first focused on the gene expression changes that showed an average 1.5-fold change or greater (136 of 492; 28%). Using the same filtering criteria for the male skeletal muscle samples, we found significantly fewer expression changes (182 genes with paired t-test <0.05 in men compared with 492 in women). When we filtered for fold change, 73 genes showed upregulation and 23 were downregulated with a relative fold change 
1.5 (about one-half the changes found in women). These data suggest that women exhibit a more pronounced transcriptional response to resistance training than men. A full list of statistically significant aerobic exercise-responsive genes in the men and three women (STRRIDE U133A Genes.xls) is available for download as supplementary material at: http://microarray.cnmcresearch.org/pgadatatable.asp under the project name Wkraus STRRIDE Study, or via our PEPR database: http://pepr.cnmcresearch.org/browse.do?action=list_prj_exp&projectId=192. The latter includes gene-based queries of either longitudinal or cross-sectional data via the "chart" function.. In addition to our "in house" database, a full data set was submitted to Gene Expression Omnibus under the accession number series GSE1295.
Metabolic remodeling. We focused on mRNAs that encoded proteins involved in substrate and energy metabolism with the goal of providing insight into the pathophysiology of the metabolic syndrome and the benefits of exercise intervention. Genes discovered using dCHIP or MAS5.0 included UQCRC2, glycogen synthase 1, ALAS1 (heme synthesis), GOT2 [malate-aspartate (MA) shuttle], IGFBP7, the mitochondrial isoforms of creatine kinase and aconitase and others (supplementary data) (23, 34). GOT1 and LDHB were detected by using both dCHIP and GeneSpring. Some genes, such as UQCRC2, ALAS1, and IGFBP7, were differentially expressed in a single sex, whereas GOT1, LDHB, PDHA1, aconitase, creatine kinase, and glycogen synthase 1 were upregulated in both sexes. Several of these genes, particularly those exhibiting gender and/or fiber-type specificity, are currently being investigated.
We focused on a subset of metabolic proteins of known function that were strongly induced by training in both men and women, or showed sexual dimorphism in this response [paired t-test P < 0.05 between time point 1 (entry) and time point 2 (9 mo training)]. Each was validated in a set of muscle biopsies from the same exercise group by immunoblot and/or enzyme activity. The validation set included the three individuals of each sex used for microarray analysis (24 biopsies) and an additional set of 7 male and 7 female subjects (56 biopsies).
Aspartate aminotransferase 1 expression and total activity. Cytosolic aspartate aminotransferase, also known as glutamate-oxoglutarate transaminase (GOT1), is the rate-limiting enzyme in the MA shuttle (2). The three known NADH shuttles function in part to regenerate the NAD+, which in turn is needed to drive the glycolytic production of pyruvate destined for the TCA cycle. It is worth noting that the inhibition of GOT activity significantly decreases the oxidative catabolism of glucose and increases lactate production (7). In addition, the activities of MA shuttle enzymes have previously been shown to increase with exercise training in young healthy men, but never in the context of the metabolic syndrome, and women have not been studied (42).
GOT1 mRNA showed strong induction with training in women and to a lesser extent in men; this was most evident when normalizing to mean expression levels across all time points in the three women (Fig. 1A) and three men (Fig. 1B) studied by expression profiling. During the 96-h and 14-day detraining time points, steady-state GOT1 mRNA levels dropped nearly to baseline, pretraining levels in both men and women. These data suggest that the expression of this gene in untrained muscle is sexually dimorphic, with men showing considerably higher basal levels pretraining than women.
|
50% in both men and women relative to the 9-mo training point (Fig. 1, A and B). The drop in protein levels was not as dramatic as the decrease in mRNA levels; however, this is commonly observed, as most proteins have longer half-lives than mRNAs. Steady-state protein levels detected by immunoblotting do not necessarily reflect enzymatic activity. We, therefore, studied GOT1 enzymatic activity in muscle biopsy lysates from the validation muscle biopsy set (7 men, 7 women). We studied total GOT activity (cytosolic plus mitochondrial), rather than cytosolic alone, because pure cytosolic preparation was not possible given the biopsy size and flash-frozen preservation technique used. The mRNAs of GOT2 and other proteins involved in the MA shuttle were also increased with training but less significantly than GOT1 (supplementary data). Total skeletal muscle GOT activity increased significantly in women (0.107–0.223 U/mg protein, P = 0.045) and to a lesser magnitude in men (0.26–0.390 U/mg protein, P = 0.056) after exercise training (Fig. 1C). Enzymatic activity data mirrored GOT1 protein abundance in (male and female) muscle before and after training.
Lactate dehydrogenase B expression and LDH isozyme analysis. The lactate dehydrogenase complex (LDH) catalyzes the near-equilibrium interconversion of pyruvate to lactate with concomitant conversion of NADH to NAD+. Lactate dehydrogenase B (LDHB) or heart-type isoform exhibits a higher affinity for lactate than lactate dehydrogenase A (LDHA) or muscle-type isoform and is preferentially expressed in highly oxidative tissues (6). In the context of an enhanced capacity for NAD+ regeneration (via the MA shuttle), increased LDHB expression would increase the economy of glucose metabolism in trained muscle, making it better able to respond to insulin-stimulated glucose uptake. Although changes in LDH isozyme distribution have been reported previously in highly trained athletes, it has not been investigated in the context of exercise intervention and the metabolic syndrome and the coordinated upregulation of pathway members LDHB and GOT1 (5).
In the 10 women studied (Fig. 2A), the amount of immunoreactive LDHB protein increased significantly (6.7 ± 2.1-fold) after exercise training and then decreased incrementally after 4 days and 14 days of posttraining to near untrained levels. The LDHB protein level changes in women were greater in magnitude yet closely mirrored the trend of Affymetrix LDHB transcript levels (Fig. 2A). The amount of immunologically reactive LDHB protein also increased significantly (8.1 ± 0.5-fold) after exercise training in the 10 men studied. These levels persisted or dropped slightly after 4 days of detraining and then returned to near untrained levels 14 days posttraining (Fig. 2B).
|
Serum lactate analysis. The increased levels of GOT1, LDHB, and PDHA1 suggested an increased ability of the muscle to metabolize glucose more efficiently, producing less lactate as a result. This suggested that serum lactate might decrease as a function of aerobic training, in that muscle could better utilize endogenous lactate and thus release less into the serum. Fasting serum lactate levels, before and after exercise training, were quantified in 36 men and 26 women from the same high-dose, high-intensity STRRIDE exercise group. By use of a paired statistical analysis, changes in fasting lactate were decreased in both men and women but reached statistical significance only in women (Fig. 3A). When data are combined irrespective of gender, changes in fasting lactate were significant and rapidly reversible with cessation of regular exercise in the detraining time points (Fig. 3B).
|
1-subunit expression.
Increased glucose and lactate metabolism via LDHB and GOT1 should lead to greater flux of intracellular pyruvate, and thus the potential for increased activity of pyruvate dehydrogenase (PDH) to shuttle carbon chains into the mitochondria to drive the oxidative phosphorylation of ATP. In fact and very recently, the pyruvate dehydrogenase 1-subunit (PDHA1) has been show to increase in muscle with training (28); however, this previous study was in young healthy volunteers, and the extent of transcriptional changes previously observed was considerably less than what we observed here in individuals with the metabolic syndrome. In the 10 women studied, the amount of immunologically reactive PDHA1 protein increased significantly (2.1 ± 0.9-fold) after exercise training and then decreased rapidly to pretraining levels after 14 days of detraining (Fig. 4A). As well, in the 10 men, the amount of immunologically reactive PDHA1 protein increased significantly (4.1 ± 1.5-fold) after exercise training, persisted or dropped slightly after 4 days of detraining, and then also decreased rapidly after 14 days of detraining (Fig. 4B). Western blot data indicated that untrained women have relatively high concentrations of PDHA1 compared with similarly untrained men, opposite the trend observed for GOT1. As well, the protein levels of PDHA1 closely followed the Affymetrix transcriptional profile of the corresponding gene in both men and women over the exercise training and detraining cycle.
|
The protein level of UQCRC2 in the 10 women studied showed a strong increase in immunologically reactive UQCRC2 protein with training (7.1 ± 1.1-fold increase). However, unlike GOT1, LDHB, and PDHA1 proteins, the relative expression of UQCRC2 remained high even after 14 days of detraining (Fig. 5A). In men, there were no significant changes in UQCRC2 protein levels after training, indicating a gender difference in exercise responsiveness and in baseline protein levels in trained and detrained skeletal muscle (Fig. 5B).
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Also, when compared against previous microarray studies on the acute effects of exercise on rat muscle (12), no significant changes were noted in the metabolic genes described herein. This suggests that these changes in gene expression are the result of exercise training rather than an acute bout effect. Integrating previous literature with our data elaborates on existing models of glucose and lactate metabolism by proposing biochemical cooperation and pretranslational coordination between the MA shuttle, LDH, and PDH complexes and links this with training-associated changes in fasting serum lactate.
Our model of the metabolic remodeling of skeletal muscle in metabolic syndrome patients (Fig. 8) describes an adaptive increase in the integration of glucose oxidation in response to the rapid changes in cellular energy demand with habitual aerobic exercise. Specifically, these adaptations are mediated by increased protein and activity levels of GOT1, LDHB, and PDHA1 and, in women only, UQCRC2.
|
-ketoglutarate), producing OAA and glutamate. At this point, the OAA may be used as a gluconeogenic precursor (via PEPCK) or as an electron shuttle across the inner mitochondrial membrane when reduced by MDH (2, 31). Collectively, the three known cellular NADH shuttles provide a conduit for the transfer of reducing equivalents across the inner mitochondrial membrane. The other two NADH shuttles include the glycerol phosphate shuttle, which is less efficient than the MA shuttle (17), and the recently proposed intracellular lactate shuttle (9). Regardless of the specific shuttle, the regeneration of cytosolic NAD+ is critical for sustaining a high rate of glycolytic ATP production. However, when glycolytic NADH generation exceeds the electron shuttle capacity (i.e., during intense exercise), pyruvate is increasingly favored as an electron acceptor and is rapidly reduced to lactate and temporarily (if it is reoxidized in the muscle) or permanently (if it is exported into circulation) sequestered from the high-ATP-yielding oxidative metabolism within the mitochondria (9, 19). We propose that, in untrained metabolic syndrome patients, low levels of GOT1 protein and activity in skeletal muscle limit its capacity for glycolytic flux in the absence of significant lactate production. This is particularly relevant during higher intensity exercise, such as that used in this study, when muscle favors glucose as a substrate, and the lack of sufficient NADH shuttle capacity would drive the production of lactate. For this reason, we believe that untrained skeletal muscle becomes a net producer of lactate, which likely contributes to elevated fasting lactate levels in untrained individuals and impaired glucose tolerance (Fig. 3). Conversely, increased GOT1 protein and activity in trained muscle should explain, in part, training-associated decreases in fasting lactate levels in women and men, although ideally this should be determined by a lactate threshold or stress test. This may be significant because consistently elevated levels of circulating lactate stimulate hepatic glucose production, which, when combined with peripheral insulin resistance and decreased metabolic clearance of glucose (both characteristics of the metabolic syndrome), leads to increasingly severe bouts of hyperglycemia typical of Type 2 diabetes (4, 27, 29).
Increased levels of GOT1 activity have been associated with increased maximal O2 of exercise-trained humans and animals (33, 38); however, the data presented here show that basal levels of GOT1 are very low in subjects with features of metabolic syndrome. At the exercise intensities associated with the ascertainment of peak or submaximal O2 measurements, muscle is highly dependent on glycolysis and glycogenolysis (8). The production of lactate rapidly limits substrate entry into the (oxidative) TCA cycle, particularly when muscle glycogen supply is limiting because of lack of training and insulin resistance. Our model then describes how the enhanced capacity of the MA shuttle (via GOT1) liberates pyruvate for oxidative ATP production in the TCA cycle. This may account also for the up to 39% increase in relative peak O2 observed in both men and women within the "high" STRRIDE cohort used in this study and for increased glucose tolerance with training (22a, 25).
Further evidence for the enhanced ability of muscle from trained metabolic syndrome patients to metabolize glucose is provided by the LDHB data shown here (Fig. 8). The LDHB protein favors the binding and oxidation of lactate relative to LDHB (39). In untrained muscle, lactate is typically exported via the monocarboxylate transporter 4, which is preferentially expressed in white glycolytic muscle fibers (32), whereas the monocarboxylate transporter 1 is associated with the import of lactate into oxidative red fibers from adjacent muscle fibers, with serum, and more recently with the import of lactate into the mitochondrion via the intracellular lactate shuttle (9, 10). In accordance with this model, the upregulation of LDHB and the resulting LDH isoform shift in exercise-trained muscle (Fig. 2C) reflects an enhanced capacity for glycolytic flux in the absence of net lactate production. Because inhibition of GOT activity increases lactate production, we propose that the coordinate induction of LDHB and GOT1 increases the economy of muscle glucose oxidation by reducing lactate production during changes in glycolytic flux and driving instead the production of pyruvate, which is channeled to the oxidative production of ATP via the TCA cycle and mitochondrial electron transport chain.
The PDH complex has a molecular weight of 7 million and is composed of multiple copies of three enzymes: pyruvate decarboxylase (E1), hydrolipoyl transacetylase, and dihydrolipoyl dehydrogenase (46). The E1 enzyme is itself a complex structure, a heterotetramer of two - and two -subunits. The E1- subunit (PDHA1) plays a key role in the function of the PDH complex because it contains the E1 active site and is thought to be the rate-limiting step in the conversion of pyruvate to acetyl-CoA (30). Recently, PDHA1 expression in muscle has been shown to increase in response to 8 wk of aerobic exercise training in healthy young men (28). This responsiveness of PDHA1 to exercise training, coupled with our observations of its increase with training and decrease with detraining (Fig. 4), suggests that its expression levels and perhaps the activity of the PDH complex itself are impaired in obese and/or glucose-intolerant individuals.
Taken together, the increased abundance and, in some cases, the increased activity of PDHA1, LDHB, and GOT1 likely reflect the enhanced integration and economy of glycolysis and the TCA cycle within exercise-trained skeletal muscle of obese individuals with characteristics of the metabolic syndrome.
The data presented here also provide new insights into gender disparity with regard to baseline muscle metabolism and subsequent remodeling with training. Specifically, UQCRC2 was discovered to be upregulated in women but not men according to dCHIP analysis. UQCRC2 is one of the core proteins of the mitochondrial electron transport chain complex III, which transfers electrons from ubiquinone to cytochrome c by using the change in electron energy state to translocate protons across the inner mitochondrial membrane (17, 35). The striking differences in exercise-responsive protein levels of UQCRC2 between women and men may simply reflect higher basal UQCRC2 protein content in men or possibly significant gender differences in mitochondrial adaptation to exercise (16). In addition to UQCRC2, many genes associated with the mitochondrial electron transport chain, mitochondrial structure and function, and metabolism were upregulated in skeletal muscle from exercise-trained muscle of women and men (supplementary data), indicating a general increase in mitochondrial density and volume with exercise training (24). Realizing this, we integrated UQCRC2 into our metabolic model to reflect this overall increase in mitochondrial density and how this would effectively increase the capacity for oxidative metabolic flux by providing more surface area for the translocated NADH and pyruvate to power the oxidative phosphorylation of ATP. A recent microarray survey of skeletal muscle from insulin-resistant and diabetic individuals described the coordinated downregulation of mitochondrial genes of oxidative and energy metabolism (35, 52). Not only are these findings in agreement with our model, but they may also explain, in part, some of the beneficial effects of habitual exercise in ameliorating these conditions in metabolic syndrome patients.
A true coordinated cluster of genes must be regulated by one or several common transcriptional response or promoter elements. LDHB, GOT1, and PDHA1 have all been reported to be glucocorticoid responsive, and all show GREs in their respective gene promoters (31, 43, 46). Specifically, it has been shown that both endogenous and exogenous glucocorticoids induce the expression of LDHB and GOT1 in liver of various animal models and that this effect is modulated by multiple promoter elements that are responsive to a variety of physiological stimuli. We believe that the metabolic adaptations to exercise training described in this report are mediated, at least in part, by glucocorticoid receptor activators and coactivators that induce a training-responsive transcriptional response. Interestingly, well-characterized genetic variations in the glucocorticoid receptor have been associated with both visceral adiposity and the metabolic syndrome (14, 48, 36). It is also worth noting that mitochondrial biogenesis is stimulated by glucocorticoids specifically in active skeletal muscle (50), a fact that strengthens our assertion that cortisol may play an important role in muscle metabolic adaptation to exercise. The surprising lack of glucocorticoid-induced expression of GOT1 and LDHB in rat skeletal muscle in response to an exogenous methylprednisolone bolus may reflect tissue-specific or species-specific response to glucocorticoids. It would be worthwhile, therefore, to investigate effects of cortisol on myotubes or in the context of an animal exercise model, because it is known that exercise limits the glucocorticoid-induced muscle atrophy (50). Most studies of glucocorticoid-induced expression of LDHB and GOT1 have been investigated in the liver of nonhuman animal models, and none has yet been conducted in skeletal muscle (41, 42, 46). These studies revealed the presence of NF1 and SP1 as well as negative glucose and insulin response elements in close proximity to these GREs, which may function in a combinatorial and tissue-specific manner (7, 37). Chronically high circulating levels of insulin and glucose may, therefore, explain the relatively low expression of GOT1 seen in untrained muscle (Fig. 1), whereas decreased fasting insulin (Fig. 7), and improved glucose tolerance with exercise training may contribute to increased GOT1 expression.
In summary, we propose that changes in glucose metabolism economy that occur in aerobic exercise-trained skeletal muscle are also responsible for the cardiovascular benefits of habitual exercise in individuals with impaired glucose tolerance. We propose also a plausible but unconfirmed model and a plausible mechanism for coordinated regulation of these genes and that these changes are regulated, at least in part, by the (co)activation of the glucocorticoid receptor, which has also been implicated in obesity. Recently, we described significant improvements in muscle-associated changes in circulating fibrinolytic proteins with exercise training (20). Coupled with increased lactate clearance and improved glucose metabolism (22a), microarrays have helped us to describe the molecular basis for the improvement in two cardiovascular risk factors associated with the metabolic syndrome (15) and the importance of skeletal muscle in governing these changes.
|
GRANTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
