Mitochondrial antioxidative capacity regulates muscle glucose uptake in the conscious mouse: effect of exercise and diet

Abstract

The objective of this study was to test the hypothesis that exercise-stimulated muscle glucose uptake (MGU) is augmented by increasing mitochondrial reactive oxygen species (mtROS) scavenging capacity. This hypothesis was tested in genetically altered mice fed chow or a high-fat (HF) diet that accelerates mtROS formation. Mice overexpressing SOD2 (sod2^Tg), mitochondria-targeted catalase (mcat^Tg), and combined SOD2 and mCAT (mtAO) were used to increase mtROS scavenging. mtROS was assessed by the H₂O₂ emitting potential (JH₂O₂) in muscle fibers. sod2^Tg did not decrease JH₂O₂ in chow-fed mice, but decreased JH₂O₂ in HF-fed mice. mcat^Tg and mtAO decreased JH₂O₂ in both chow- and HF-fed mice. In parallel, the ratio of reduced to oxidized glutathione (GSH/GSSG) was unaltered in sod2^Tg in chow-fed mice, but was increased in HF-fed sod2^Tg and both chow- and HF-fed mcat^Tg and mtAO. Nitrotyrosine, a marker of NO-dependent, reactive nitrogen species (RNS)-induced nitrative stress, was decreased in both chow- and HF-fed sod2^Tg, mcat^Tg, and mtAO mice. This effect was not changed with exercise. Kg, an index of MGU was assessed using 2-[¹⁴C]-deoxyglucose during exercise. In chow-fed mice, sod2^Tg, mcat^Tg, and mtAO increased exercise Kg compared with wild types. Exercise Kg was also augmented in HF-fed sod2^Tg and mcat^Tg mice but unchanged in HF-fed mtAO mice. In conclusion, mtROS scavenging is a key regulator of exercise-mediated MGU and this regulation depends on nutritional state.

reactive oxygen species (ROS) are generated inside and outside the mitochondria. In the mitochondria they are formed when electrons escape electron transport chain and react with ground state oxygen to form the superoxide anion (O₂·⁻). Superoxide dismutase (SOD) converts O₂˙⁻ to hydrogen peroxide (H₂O₂), which is reduced to water (H₂O) by the action of glutathione peroxidase or peroxiredoxins. Catalase, an antioxidant enzyme located in peroxisomes also decomposes H₂O₂ to H₂O and oxygen. In the presence of reduced metal atoms H₂O₂ may be oxidized to the highly reactive hydroxyl (˙OH) or peroxyl (˙OOH) radicals (27, 35). Skeletal muscle generates ROS, and this generation is increased with insulin resistance (1, 2, 24). Paradoxically ROS production is also increased by muscle contraction (35), a condition associated with increased muscle glucose uptake (MGU) and increased insulin action (40). One distinction between insulin resistance and muscle contraction is that in the insulin resistant state increased ROS is due to formation of mitochondrial O₂˙⁻ (mtO₂˙⁻) (2, 24), whereas recent work in isolated fibers suggest that the increment in ROS during contraction is extra-mitochondrial (26). In addition to site of origin, the specific ROS species may be significant in metabolic regulation. mtO₂˙⁻ has been implicated in insulin resistance (2, 24). On the other hand, H₂O₂-activated signaling pathways have been proposed to contribute to the stimulation of MGU during contraction (35, 48, 57). The significance of ROS both in insulin resistance and with contraction is controversial and largely remains unresolved.

One important reason the role of oxidative status in MGU remains unclear and conflicting results exist is that the oxidative status of a cell is highly sensitive and subject to the experimental model and design. The role of ROS in contraction-stimulated MGU has been studied ex vivo (48), in situ (34), and in humans (37). These different models have yielded different conclusions (35). Studies of human subjects are obviously of most direct relevance, but can be limited by practical considerations regarding the use of human subjects. Ex vivo and in situ models in rodents are powerful experimental systems, but have limitations. Ex vivo models lack the intact physiological interactions between muscle fibers and surrounding matrix and capillaries, whereas in situ models are usually performed in anesthetized animals using a perfusion protocol. In both cases muscle tissue is reliant on extra-physiological factors to maintain tissue and cell integrity.

Strategies for increasing antioxidant capacity that specifically focus on the removal of mitochondrial O₂˙⁻ or H₂O₂ have rarely been evaluated in the context of contracting muscle. This is important in defining potential mechanisms involved in stimulating MGU. One study showed that overexpression of SOD2, the mitochondrial form of SOD, stimulates MGU in isolated contracting muscle (48). SOD2 overexpression has been shown to enhance mitochondrial antioxidant capacity and provide resistance against factors that induce mitochondrial permeability (39, 51). It has also been shown that high-fat (HF)-fed insulin resistant obese mice, which have accelerated mtROS production (2), have decreased exercise capacity and exercise-induced MGU (15). The aim of the present studies was to test the hypotheses that an increase in mitochondrial ROS (mtROS) scavenging would increase exercise-stimulated MGU in vivo. Mice overexpressing SOD2 (sod2^Tg) were used to increase mtO₂˙⁻ scavenging (41). Mice expressing a catalase transgene (mcat^Tg) with a mitochondrial localization sequence (50) were used to increase mitochondrial H₂O₂ (mtH₂O₂) scavenging. Mice overexpressing both SOD2 and mCAT (mtAO) were also studied, representing an extremely high mitochondrial antioxidative capacity state. Exercise-stimulated MGU was measured in unstressed chronically catheterized mice, which allows for study of the whole organism and preservation of the physiological milieu. Mice fed chow or a HF diet that accelerates ROS production were combined with the transgenic mouse models to dissect the importance of pathways for ROS scavenging and to test the hypothesis that MGU is improved by increased mtROS scavenging capacity.

MATERIALS AND METHODS

Mouse Models

All mice were housed in cages under conditions of controlled temperature and humidity with a 12:12-h light/dark cycle. Four lines of sex- and littermate-matched C57BL/6J mice, wild type (WT), mice overexpressing SOD2 (sod2^Tg), the mitochondrial isoform of SOD (41), mice overexpressing catalase targeted to mitochondria with a mitochondrial localization signal in the human catalase transgene (mcat^Tg) (50), and mice with overexpression of both SOD2 and mCAT (mtAO) were used. Mice used in experiment 1 were fed chow and mice used in experiment 2 were fed a HF diet (F3282, BioServ), which contains 60% calories as fat, for 16 wk. mcat^Tg mice were generously provided by Dr. Peter Rabinovitch of the University of Washington (50). The Vanderbilt and East Carolina Animal Care and Use Committees approved procedures involving animals.

Exercise Stress Test

Two days prior to the exercise stress test, mice performed a 10 min bout of exercise at 10 m/min for acclimation purposes. V̇o₂ was measured on an enclosed treadmill using an Oxymax System (Columbus Instruments). Following a 45-min sedentary period, mice commenced running at 10 m/min. Speed was increased 4 m/min every 3 min until exhaustion (30, 31). Sedentary V̇o₂ was measured for the last 15 min of the sedentary period, and peak V̇o₂ during exercise was achieved when V̇o₂ no longer increased despite an increase in work rate.

Exercise Metabolism in Vivo

Metabolic studies were conducted on 149 WT, sod2^Tg, mcat^Tg, and mtAO mice. Experiment 1 used 73 mice fed chow, and experiment 2 used 76 mice fed a HF diet. Catheters were implanted in the left carotid artery and the right jugular vein for sampling and infusions, respectively, >5 days before the study (5, 6). The day before surgery, all mice were acclimated to the treadmill. On the day of the study, tubing was connected to the exteriorized catheters and mice were put in the enclosed treadmill to acclimate (16). At time 0, a baseline blood sample was taken to measure arterial glucose, insulin, non-esterified fatty acid (NEFA), and lactate. Mice remained sedentary or performed a single bout of exercise at 12 m/min for 30 min. This particular speed reflects 30% of the maximal running speed in the chow-fed mice (experiment 1) and 50% of the maximal running speed in the HF-fed mice (experiment 2). The same absolute amount of work and not the relative amount of work was used regardless of diet. This made direct comparison between different diets difficult. On the other hand, mice regardless of diet performed comparable absolute work, creating similar demands on energy producing processes. All mice from the exercise group completed the 30 min treadmill running.

An index of MGU was determined using 13 μCi of 2-[¹⁴C]-deoxyglucose ([¹⁴C]2-DG) administered as an intravenous bolus at t = 5 min. Blood samples were then taken at 7, 10, 15, 20, and 30 min to measure arterial glucose and plasma [¹⁴C]2-DG. Plasma insulin, NEFA, and lactate were also measured in blood sample collected at 30 min as the postexperimental parameters. After the last sample, mice were anesthetized and tissues were excised and frozen in liquid nitrogen.

Analyses

[¹⁴C]2-DG and [¹⁴C]2-DG-phosphate radioactivity.

Radioactivity of [¹⁴C]2-DG and [¹⁴C]2-DG-phosphate ([¹⁴C]2-DG-P) was determined as described (4). The fractional uptake of [¹⁴C]2-DG (Kg) was calculated by the equation: Kg = [¹⁴C]2-DG-P_tissue/AUC[¹⁴C]2-DG_plasma, where [¹⁴C]2-DG-P_tissue is the [¹⁴C]2-DG-P radioactivity in tissues (dpm/g), and AUC[¹⁴C]2-DG_plasma is the area under the plasma [¹⁴C]2-DG disappearance curve (dpm·min⁻¹·ml⁻¹).

Plasma insulin and substrates.

Plasma insulin and NEFA were determined as described (28). Plasma lactate was determined enzymatically (33). Blood glucose was determined from ∼5 μl of arterial blood using an ACCU-CHEK monitor.

Muscle glycogen.

Glycogen was measured in neutralized gastrocnemius extracts. Glucose units were determined enzymatically (33).

Muscle nitrotyrosine.

Nitrotyrosine was assessed in the fasted state and following exercise by immunohistochemistry in paraffin-embedded tissue sections (5 μm) with anti-nitrotyrosine antibody (Millipore). Images were captured using a Q-Imaging Micropublisher camera mounted on an Olympus upright microscope. Immunostaining was quantified using ImageJ software.

AMPK activity.

AMPK activity was determined as previously described (59). AMPK activity was calculated as picomoles phosphate incorporated into the SAMS peptide per minute per milligram of protein.

Mitochondrial H₂O₂ Emitting Potential

Mitochondrial H₂O₂ emitting potential (JH₂O₂) was measured in permeabilized muscle fiber bundles prepared as described (2). Muscle fiber bundles in a buffer containing 105 mM K-MES, 30 mM KCl, 1 mM EGTA, 10 mM KH₂PO4, 5 mM MgCl₂, 0.5 mg/ml BSA, pH 7.4, 5 μM glutamate, and 2 μM malate were treated with 20 mM pyrophosphate to deplete adenine nucleotides and to inhibit contraction of the fibers. JH₂O₂ was measured at 30°C in response to increasing succinate concentrations or 25 μM palmitoyl-carnitine and 2 mM malate during state 4 respiration (10 μg/ml oligomycin) by continuous spectrofluorometric monitoring oxidation of Amplex Red (10 μM). Exogenous Cu/Zn SOD (for the detection of non-matrix sources of superoxide production) was not included in the assay to isolate the effects of the SOD2 transgenic mouse model. JH₂O₂ was expressed as picomoles per minute per milligram dry weight.

Western Blotting

Isolated mitochondria samples were used for the detection of SOD2 and catalase proteins. Muscle mitochondria were isolated from frozen tissues using the Mitochondria Extraction Kit (IMGENEX). Protein concentration was then measured using a Bio-Rad assay with BSA as a standard. The absence of protein GAPDH was used to document the purity of mitochondria fractions. Protein (30 μg) was applied to 4–12% SDS-PAGE gel using the NuPAGE system (Invitrogen). SOD2 protein was probed using the SOD2 polyclonal antibody (Assay Designs). Catalase was probed using the monoclonal anti-catalase antibody (Sigma). V-DAC was used as a loading control. For the detection of hexokinase II (HKII) and GLUT4, gastrocnemius muscle was homogenized in homogenization buffer containing 50 mM Tris, pH 7.5, 1 mM EDTA, 1 mM EGTA, 10% glycerol, 1% Triton X-100, 1 mM DTT, 1 mM PMSF, 5 μg/ml protease inhibitor cocktail, 50 mM NaF, 5 mM sodium pyrophosphate, and centrifuged at 13,000 g for 20 min at 4°C. The protein concentration was then determined, and homogenates were run on SDS-PAGE gels. HKII and GLUT4 were probed using HKII and GLUT4 antibodies (Abcam).

Reduced and Oxidized Glutathione Measurements

Gastrocnemius was homogenized in buffer containing 10 mM Tris, pH 7.4, 1 mM EDTA, 1 mM EGTA, 5 μg/ml protease inhibitor cocktail, 1 mM sodium orthovanadate, 5 mM sodium pyrophosphate, 50 mM sodium fluoride. For oxidized glutathione (GSSG) measurement, 5 mM 1-methyl-2-vinylpyridinium trifluoromethanesulfonate was used to scavenge all reduced thiols in the sample. Total reduced glutathione (GSH) and GSSG concentrations were then measured using the BIOXYTECH GSH/GSSG-412 assay kit (Oxis Research).

Statistical Analysis

Data are expressed as means ± SE. Statistical analyses were performed using either student's t-test or two-way ANOVA followed by Tukey's post hoc tests as appropriate. The significance level was P < 0.05.

RESULTS

Protein Expression of SOD2 and Catalase in the Transgenic Mice

SOD2 protein expression was increased by approximately threefold in muscle mitochondria of sod2^Tg and mtAO mice relative to WT mice (Fig. 1A). SOD2 protein was not affected in the mitochondria of muscle in mcat^Tg mice. mcat^Tg and mtAO mice had a ∼15-fold increase in catalase expression in muscle mitochondria compared with WT mice (Fig. 1B). Catalase protein was equal in sod2^Tg and WT mice. HF feeding in mice had no effect on SOD2 or catalase expression in muscle mitochondria (Fig. 1, A and B).

Fig. 1.

Protein expression of SOD2 and catalase. Western blotting of SOD2 (A) and catalase (B) in mitochondria isolated from gastrocnemius muscle in mice after chow or high-fat diet (HF) feeding for 16 wk. Integrated intensities of bands are listed below each band and are represented as means ± SE. Data are normalized to chow-fed wild type (WT). V-DAC expression was not affected by genotype or HF feeding and was used as a loading control. N is equal to 3–8, and *P < 0.05 compared with chow-fed WT.

Experiment 1: Increased Mitochondrial Antioxidant Capacity in Chow-Fed Mice

Baseline metabolic parameters.

Body weight was unaffected by genotype (Table 1). Basal arterial glucose, plasma insulin, and NEFA were similar between all genotypes. Basal lactate was similar between genotypes except that mtAO had lower lactate, possibly reflecting a reduced redox state or a more efficient oxidation of pyruvate.

Table 1.

Characteristics of chow-fed sod2^Tg, mcat^Tg, and mtAO mice in sedentary and exercise states

	Sedentary				Exercise
	WT	sod2Tg	mcatTg	mtAO	WT	sod2Tg	mcatTg	mtAO
N (female/male)	8 (4/4)	7 (4/3)	10 (5/5)	7 (4/3)	11 (5/6)	10 (5/5)	12 (7/5)	8 (4/4)
Body weight, g	26 ± 1	24 ± 1	25 ± 1	25 ± 1	26 ± 1	24 ± 1	25 ± 1	26 ± 2
Glucose, mg/dl
Basal	9.3 ± 0.4	8.8 ± 0.3	10.1 ± 0.5	9.1 ± 0.5	9.4 ± 0.3	10.0 ± 0.9	9.8 ± 0.4	9.1 ± 0.7
Postexperimental	10.0 ± 0.6	10.6 ± 0.6	10.6 ± 0.4	9.9 ± 0.4	12.2 ± 0.6†	13.6 ± 0.6†	13.1 ± 0.7†	12.0 ± 1.0†
Insulin, ng/m
Basal	0.58 ± 0.06	0.61 ± 0.07	0.48 ± 0.06	0.45 ± 0.02	0.61 ± 0.09	0.60 ± 0.11	0.59 ± 0.05	0.62 ± 0.09
Postexperimental	0.58 ± 0.06	0.70 ± 0.13	0.47 ± 0.06	0.51 ± 0.03	0.38 ± 0.04†	0.45 ± 0.06†	0.42 ± 0.05†	0.38 ± 0.04†
NEFA, mM
Basal	0.95 ± 0.04	1.03 ± 0.25	1.46 ± 0.17	1.29 ± 0.27	1.43 ± 0.41	1.36 ± 0.28	1.42 ± 0.12	1.39 ± 0.23
Postexperimental	0.92 ± 0.03	1.00 ± 0.11	1.11 ± 0.11	0.89 ± 0.15	0.75 ± 0.13†	0.73 ± 0.11†	0.71 ± 0.09†	1.02 ± 0.10
Lactate, mM
Basal	1.01 ± 0.16	0.86 ± 0.17	0.86 ± 0.04	0.48 ± 0.07*	0.76 ± 0.11	1.01 ± 0.22	0.76 ± 0.07	0.50 ± 0.03*
Postexperimental	1.02 ± 0.20	0.92 ± 0.31	0.75 ± 0.06	0.55 ± 0.07*	2.72 ± 0.38†	2.33 ± 1.24†	2.35 ± 0.07†	1.18 ± 0.1†*

Values are mean ± SE. Basal refers to data from samples collected at 0 min. Postexperimental refers to data from samples collected at 30 min.

^*P < 0.05 compared with wild type (WT);

^†P < 0.05 compared with basal.

Antioxidant capacity.

Succinate-supported JH₂O₂ was increased in sod2^Tg but decreased to baseline in mcat^Tg and mtAO mice (Fig. 2A). Palmitoyl-carnitine/malate-supported JH₂O₂ was unaffected in sod2^Tg but remarkably decreased in mcat^Tg and mtAO mice (Fig. 2B). Results in mcat^Tg and mtAO mice demonstrate that expression of catalase in mitochondria is effective in scavenging mtH₂O₂.

Fig. 2.

Antioxidant capacity in chow-fed mice. A and B: muscle mitochondrial H₂O₂ emitting potential (JH₂O₂) was measured in permeabilized gastrocnemius fiber bundles prepared from sedentary chow-fed WT, sod2^Tg, mcat^Tg, and mtAO mice. JH₂O₂ was measured either in response to increasing succinate concentrations (A) or 25 μM palmitoyl-carnitine and 2 mM malate (B) during state 4 respiration (10 μg/ml oligomycin). C: GSH/GSSG ratio was determined in muscle homogenates. D: protein expression of nitrotyrosine was detected by immunohistochemistry in gastrocnemius. Representative images were displayed and the magnification of images was ×20. E: nitrotyrosine protein was quantified by the integrated intensity of staining and normalized to WT chow. Data are expressed as means ± SE and n = 3–9 for each group. ξP < 0.05 for genotype effect compared with WT. *P < 0.05 compared with chow-fed WT.

The ratio of reduced to oxidized glutathione (GSH/GSSG) was not changed in sod2^Tg but was increased in mcat^Tg and mtAO mice (Fig. 2C). These results are parallel to the JH₂O₂ data as GSH participates in the detoxification of H₂O₂ by glutathione peroxidases. In contrast to some findings (20) but consistent with others (34), GSSG concentration was undetectable in mice after exercise.

Nitrotyrosine is a product of tyrosine nitration of reactive nitrogen species (RNS) such as peroxynitrite anion and nitrogen dioxide, both of which are formed in vivo by the reaction of the free radical O₂˙⁻ and the free radical nitric oxide (NO). Nitrotyrosine was decreased in sod2^Tg, mcat^Tg, and mtAO mice in the fasted stated (Fig. 2, D and E). This relationship was not affected by exercise (Fig. 2, D and E). Together these data show that overexpression of SOD2 and catalase decreases intramyocellular oxidative stress.

Exercise capacity.

Although sod2^Tg, mcat^Tg, and mtAO altered mitochondrial antioxidant capacity, running time to exhaustion during an exercise stress test was unaltered (20 ± 2 min in WT, 15 ± 3 min in sod2^Tg, 19 ± 1 min in mcat^Tg, and 16 ± 1 min in mtAO). Sedentary V̇o₂ was similar between genotypes (79 ± 3 in WT, 84 ± 7 in sod2^Tg, 73 ± 2 in mcat^Tg, and 83 ± 3 ml·kg⁻¹·min⁻¹in mtAO). During the exercise stress test, peak V̇o₂ was similar in WT, mcat^Tg, and mtAO mice but was lower in the sod2^Tg mice [131 ± 3 in WT, 117 ± 3 (P < 0.05 vs. WT) in sod2^Tg, 124 ± 3 in mcat^Tg, and 128 ± 5 ml·kg⁻¹·min⁻¹in mtAO].

Metabolic responses to exercise.

Mice in the sedentary groups maintained arterial glucose, plasma insulin, NEFA, and lactate constant (Table 1 and Fig. 3A). Thirty minutes of exercise increased arterial glucose and lactate and decreased plasma insulin in all genotypes (Table 1 and Fig. 3B). Exercise arterial glucose and plasma insulin did not differ between genotypes, and exercise plasma lactate in mtAO mice was lower than that of WT mice. Plasma NEFA was decreased to a similar extent in WT, sod2^Tg, and mcat^Tg mice, but was unchanged in chow-fed mtAO mice (Table 1).

Fig. 3.

Metabolic responses to treadmill exercise in chow-fed mice. A and B: arterial glucose concentrations either at sedentary or during exercise were measured by a glucose meter in chow-fed mice. C: Kg was determined in gastrocnemius using the nonmetabolizable glucose analog 2-[¹⁴C]-deoxyglucose during the sedentary (Sed) or treadmill exercise (Ex) study in mice fed chow diet. D: muscle glycogen content was measured by an enzymatic assay in gastrocnemius collected immediately after the sedentary or exercise study from mice fed chow diet. E: AMPK activity was determined in gastrocnemius muscle collected immediately after the sedentary or exercise study. F: HKII and GLUT4 expression in muscle homogenates. Integrated intensities of bands are listed below each band. Data are expressed as means ± SE and n = 7–12 for blood glucose and Kg measurements, 4–5 for muscle glycogen and AMPK activity measurements, n = 3 for HKII and GLUT4 expression. *P < 0.05 compared with WT within Sed or Ex. +P < 0.05 compared with Sed within a genotype.

Sedentary Kg was the same among WT, sod2^Tg, and mtAO mice, but was higher in mcat^Tg mice (Fig. 3C). Thirty minutes of exercise increased muscle Kg in all mice. Compared with WT mice, sod2^Tg, mcat^Tg, and mtAO mice had enhanced exercise Kg in gastrocnemius. Muscle glycogen was similar at rest between all genotypes, and 30 min of treadmill exercise equivalently decreased glycogen in all genotypes (Fig. 3D).

Increased AMPK activity is a well-characterized response to exercise (18, 44, 63). Sedentary muscle AMPK activity was equivalent in all genotypes (Fig. 3E). Thirty minutes of exercise increased AMPK activity similarly in chow-fed WT, sod2^Tg, mcat^Tg, and mtAO mice.

MGU is regulated by the delivery of glucose from the blood to the interstitial space, glucose transport across the membrane by GLUT4, and intracellular phosphorylation of glucose to glucose 6-phosphate by hexokinase II (HKII) (60, 61). Despite increased exercise Kg in the transgenic mice, muscle HKII and GLUT4 expression was similar in chow-fed WT, sod2^Tg, mcat^Tg, and mtAO mice (Fig. 3F).

Experiment 2: Increased Mitochondrial Antioxidant Capacity in HF-Fed Mice

Baseline metabolic parameters.

Body weight was unaffected by genotype (Table 2). Arterial glucose, plasma insulin, NEFA, and lactate were similar in all genotypes (Table 2).

Table 2.

Characteristics of HF-fed sod2^Tg, mcat^Tg, and mtAO mice in sedentary and exercise states

	Sedentary				Exercise
	WT	sod2Tg	mcatTg	mtAO	WT	sod2Tg	mcatTg	mtAO
N (female/male)	9 (5/4)	8 (4/4)	7 (4/3)	7 (4/3)	11 (5/6)	11 (6/5)	12 (7/5)	11 (6/5)
Body weight, g	37 ± 1	37 ± 3	39 ± 3	41 ± 4	39 ± 2	37 ± 2	36 ± 2	39 ± 3
Glucose, mg/dl
Basal	10.9 ± 0.5	12.1 ± 1.3	10.9 ± 0.8	10.5 ± 1.1	11.9 ± 0.6	11.5 ± 0.4	11.4 ± 0.7	10.2 ± 0.4
Postexperimental	12.1 ± 0.5	12.6 ± 1.2	12.5 ± 0.7	12.0 ± 1.2	14.8 ± 0.7†	14.3 ± 0.3†	15.2 ± 0.9†	13.5 ± 0.5†
Insulin, ng/m
Basal	1.90 ± 0.04	1.92 ± 0.54	2.40 ± 0.74	2.13 ± 0.93	2.29 ± 0.22	1.82 ± 0.38	1.75 ± 0.34	1.88 ± 0.42
Postexperimental	1.96 ± 0.23	1.79 ± 0.64	2.88 ± 0.88	1.98 ± 1.02	1.04 ± 0.15†	0.64 ± 0.1†*	0.58 ± 0.1†*	1.00 ± 0.16†
NEFA, mM
Basal	1.10 ± 0.20	1.01 ± 0.04	0.89 ± 0.05	0.79 ± 0.12	1.09 ± 0.19	0.81 ± 0.06	0.81 ± 0.09	0.96 ± 0.10
Postexperimental	0.74 ± 0.06	0.94 ± 0.06	0.86 ± 0.12	0.75 ± 0.08	0.72 ± 0.10	0.58 ± 0.07	0.61 ± 0.09	0.73 ± 0.09
Lactate, mM
Basal	0.84 ± 0.14	0.78 ± 0.03	1.10 ± 0.28	0.81 ± 0.32	0.95 ± 0.03	1.05 ± 0.10	0.96 ± 0.07	0.81 ± 0.04
Postexperimental	0.81 ± 0.12	1.07 ± 0.17	1.25 ± 0.34	0.98 ± 0.07	2.05 ± 0.22†	2.85 ± 0.46†	2.44 ± 0.26†	3.07 ± 0.2†*

Values are mean ± SE. Basal refers to data from samples collected at 0 min. Postexperimental refers to data from samples collected at 30 min.

^*P < 0.05 compared with WT;

^†P < 0.05 compared with basal.

Antioxidant capacity.

HF diet elevated both succinate- and palmitoyl-carnitine/malate-supported JH₂O₂ in the muscle of WT mice (succinate: 221 ± 28 vs. 136 ± 15 pmol·min⁻¹·mg⁻¹ at 3 mM succinate, P < 0.05; palmitoyl-carnitine/malate: 192 ± 11 vs. 146 ± 13 pmol·min⁻¹·mg⁻¹, P < 0.05). In contrast to increased succinate-supported JH₂O₂ and unaltered palmitoyl-carnitine/malate-supported JH₂O₂ in chow-fed sod2^Tg mice, both of these were decreased by ∼40% in HF-fed sod2^Tg mice compared with HF-fed WT mice (Fig. 4, A and B). Succinate- and palmitoyl-carnitine/malate-supported JH₂O₂ was decreased by ∼60% in mcat^Tg mice and was decreased to baseline rates in HF-fed mtAO mice (Fig. 4, A and B).

Fig. 4.

Antioxidant capacity in HF-fed mice. A and B: muscle mitochondrial H₂O₂ emitting potential (JH₂O₂) was measured in permeabilized gastrocnemius fiber bundles prepared from sedentary HF-fed WT, sod2^Tg, mcat^Tg, and mtAO mice. JH₂O₂ was measured either in response to increasing succinate concentrations (A) or 25 μM palmitoyl-carnitine and 2 mM malate (B) during state 4 respiration (10 μg/ml oligomycin). C: GSH/GSSG ratio was determined in muscle homogenates. D: protein expression of nitrotyrosine was detected by immunohistochemistry in gastrocnemius. Representative images were displayed and the magnification of images was ×20. E: nitrotyrosine protein was quantified by the integrated intensity of staining and normalized to WT chow. Data are expressed as means ± SE and n = 3–9 for each group. ξP < 0.05 for genotype effect compared with WT. *P < 0.05 compared with HF-fed WT.

HF diet feeding decreased GSH/GSSG ratio in muscle of WT mice (13 ± 3 vs. 27 ± 5, P < 0.05). GSH/GSSG ratio was increased in HF-fed sod2^Tg, mcat^Tg and mtAO mice relative to HF-fed WT mice (Fig. 4C). GSSG concentration was undetectable in mice after exercise.

HF diet increased muscle nitrotyrosine in WT mice (1.42 ± 0.19-fold increase, P < 0.05). Nitrotyrosine was reduced in HF-fed sod2^Tg, mcat^Tg, and mtAO mice compared with HF-fed WT mice in the fasted state (Fig. 4, D and E). This effect was not changed with exercise (Fig. 4, D and E). These data show that mitochondria-targeted increase in the antioxidant capacity effectively decreases the oxidative state of muscle in HF-fed mice.

Exercise capacity.

HF feeding equivalently impaired exercise tolerance in all genotypes (12 ± 1 in WT, 13 ± 2 in sod2^Tg, 12 ± 1 in mcat^Tg, and 11 ± 1 min in mtAO). V̇o₂ was similar between genotypes in both sedentary (76 ± 5 in WT, 98 ± 7 in sod2^Tg, 81 ± 5 in mcat^Tg, and 91 ± 6 ml·kg⁻¹·min⁻¹ in mtAO) and peak exercise states (101 ± 6 in WT, 117 ± 11 in sod2^Tg, 109 ± 5 in mcat^Tg, and 113 ± 6 ml·kg⁻¹·min⁻¹ in mtAO).

Metabolic responses to exercise.

The running speed of 12 m/min corresponded to 30% of maximal speed in the chow-fed mice and 50% in HF-fed mice. Because HF diet feeding decreased exercise capacity, HF-fed mice ran at a higher relative work rate although they ran at the same absolute speed as the chow-fed mice. Arterial glucose, plasma insulin, NEFA, and lactate remained constant in the sedentary groups (Table 2 and Fig. 5A). Exercise arterial glucose and plasma NEFA did not differ between genotypes (Table 2 and Fig. 5B). Exercise plasma insulin in sod2^Tg and mcat^Tg mice was significantly lower than that of WT mice (Table 2). In contrast to chow-fed mice, exercise lactate was higher in mtAO mice relative to WT on a HF diet (Table 2).

Fig. 5.

Metabolic responses to treadmill exercise in HF-fed mice. A and B: arterial glucose concentrations either at sedentary or during exercise were measured by a glucose meter in HF-fed mice. C: Kg was determined in gastrocnemius using the nonmetabolizable glucose analog 2-[¹⁴C]-deoxyglucose during the sedentary (Sed) or treadmill exercise (Ex) study in mice fed HF diet. D: muscle glycogen content was measured by an enzymatic assay in gastrocnemius muscle collected immediately after the sedentary or exercise study from mice fed HF diet. E: AMPK activity was determined in gastrocnemius muscle collected immediately after the sedentary or exercise study. F: HKII and GLUT4 expression in muscle homogenates. Integrated intensities of bands are listed below each band. Data are expressed as means ± SE and n = 7–12 for blood glucose and Kg measurements, 4–5 for muscle glycogen and AMPK activity measurements, n = 3 for HKII and GLUT4 expression. *P < 0.05 compared with WT within Sed or Ex. +P < 0.05 compared with Sed within a genotype.

Sedentary Kg was similar and exercise increased Kg in HF-fed mice of all genotypes (Fig. 5C). Exercise Kg in HF-fed mice was higher than that in the chow-fed mice with the exception of the mtAO mice (Fig. 5C vs. Fig. 3C). This is likely due to increased relative work load in the HF-fed mice. Exercise Kg, however, was enhanced in HF-fed sod2^Tg and mcat^Tg mice compared with HF-fed WT mice but unchanged in HF-fed mtAO mice. Muscle glycogen was similar at rest and with exercise between all genotypes (Fig. 5D).

In WT mice, HF feeding abolished the exercise-induced increase in muscle AMPK activity present in chow-fed mice (Fig. 5E). This increase in AMPK activity was restored in sod2^Tg and mcat^Tg mice. In contrast, muscle AMPK activity in mtAO mice was not restored but remained unchanged as in WT mice. The restoration of AMPK responses to exercise in HF-fed sod2^Tg and mcat^Tg mice paralleled the increases in exercise-stimulated Kg in these mice. The failure to see an increase in AMPK activity in HF-fed mtAO was consistent with the absence of an enhanced Kg in HF-fed mtAO mice.

Consistent with chow-fed mice, muscle HKII and GLUT4 expression was not different between WT mice and the transgenic mice on a HF diet (Fig. 5F).

DISCUSSION

These studies show that increased antioxidant capacity of muscle mitochondria by genetic means can increase MGU during exercise. There are two distinct aspects of these studies that make it particularly novel. First, we used serial genetic manipulations to distinguish mtO₂˙⁻ scavenging by SOD2 from changes in mtH₂O₂ formation. This is significant because H₂O₂ is an important signaling molecule. Second, we determined the effects of mitochondrial antioxidant capacity during exercise in an in vivo system. This is important because it permitted the study of mitochondrial oxidative status under physiological conditions free of the constraints imposed by in vitro experimental systems. Exercise has been widely used to assess cellular and molecular adaptations in muscle (7). The novel catheterized mouse model we used permits blood sampling during exercise. Exercise intensity and whether stress is induced are extremely important in examining physiological responses to exercise. In the current study we used a treadmill running protocol at 12 m/min at 0% grade for 30 min (30% maximal running speed in chow-fed mice and 50% maximal running speed in HF-fed mice). ROS signaling is highly sensitive to experimental conditions. The use of physiological exercise rather than ex vivo (48) or in situ (34) contractions may contribute to the experimental outcome and explain differing results (35).

Experiment 1: Increased Mitochondrial Antioxidative Capacity on Muscle Glucose Metabolism

Increased mitochondrial antioxidant capacity in mcat^Tg and mtAO mice was evident by decreased JH₂O₂ from muscle fibers of sedentary mice. It is important to emphasize that JH₂O₂ was measured in situ under conditions in which the redox pressure on the electron transport system was progressively increased (succinate titration/palmitoyl-carnitine) and thus only represents a measure of the potential for JH₂O₂. There are no methods available to measure mtO₂˙⁻ or mtH₂O₂ production in vivo in real time. Increased mitochondrial antioxidant capacity in mcat^Tg and mtAO mice resulted in increased intracellular antioxidant capacity as shown by increased GSH/GSSG, the major intracellular redox buffer. In contrast to mcat^Tg and mtAO mice, sod2^Tg mice did not increase mtH₂O₂ scavenging. This is not surprising because SOD2 converts mtO₂˙⁻ to mtH₂O₂ production. Despite unchanged GSH/GSSG in sedentary sod2^Tg mice, the expression of nitrotyrosine, the reaction product of mtO₂˙⁻ and NO, was decreased. Taken together, these data suggest that overexpression of the antioxidant enzymes SOD2 and catalase in mitochondria induces chronic elevation in muscle antioxidant capacity, which may not uniformly exhibit changes in acute indices of oxidative status.

Despite this increase in the antioxidant capacity, the transgenic mice have unaltered peak exercise capacity when assessed by an incremental exercise test. Results in both rodents (23, 56) and humans (64) have shown that increased antioxidative capacity by antioxidant vitamin C or E supplementation does not alter endurance training adaptation. However, studies by Ristow et al. (46) in humans and Gomez-Cabrera et al. (21) in both humans and rats reported that the same supplementation hampers training-induced adaptive responses of mitochondrial biogenesis, insulin action, and endurance capacity. This is at odds with a rodent study indicating a beneficial effect of vitamin E supplementation on endurance capacity in aging rats (3). These disparities could be due to differences in the dosage of the supplements used or other technical differences. The role of antioxidant capacity in exercise capacity remains to be clearly defined.

An increase in AMPK activity during exercise is well established (18, 44, 63). However, the role of AMPK in the coupling of ROS to MGU is controversial. While one study suggests that ROS accelerates contraction-mediated MGU by activating AMPK (48), another elegantly showed that N-acetyl-cysteine (NAC), a non-specific antioxidant, inhibits contraction-stimulated MGU independent of AMPKα2 (36). Exercise-stimulated MGU was enhanced in mice with increased mtROS scavenging although the exercise-mediated AMPK activation in chow-fed mice was not. This disassociation of AMPK signaling and MGU is consistent with the finding of Merry et al. (36). The discrepancy between our results and those of Sandstrom et al. (48) could be attributed to differences in the experimental paradigm. NAC was used by Sandstrom et al. to simultaneously reduce both mtROS and extra-mtROS. This approach contrasts with that of the present study where we targeted the mitochondria specifically with increased endogenous antioxidant capacity by overexpressing the antioxidant enzymes SOD2 and catalase. In addition, in vivo treadmill exercise was used in the present study, whereas ex vivo contraction was used by Sandstrom et al. (48).

MGU is regulated by the delivery of glucose from the blood to the interstitial space, glucose transport across the membrane by GLUT4, and phosphorylation of glucose to glucose-6-phosphate by HKII (60, 61). Membrane transport is the primary barrier to MGU in the fasted, sedentary state, as membrane GLUT4 content is low and the membrane is relatively impermeable to glucose. GLUT4 translocation is greatly increased during exercise, resulting in a shift in the barrier of MGU from membrane transport to phosphorylation (13). Indeed, we found no differences in GLUT4 expression between WT mice and the transgenic mice. We did not measure GLUT4 in the plasma membrane; however, it is unlikely based on data obtained in mice (12–14, 17), rats (22), and humans (29, 43) that increased GLUT4 translocation would further increase MGU during exercise. HK exerts greater control of MGU during exercise (13). Although we did not find differences in total HKII expression between genotypes, it has been suggested that HK compartmentation and spatial barriers also determine the ability of HKII to phosphorylate glucose (61). We speculate that these events may be affected in the transgenic mice contributing to increased exercise Kg. It is also possible that overexpression of SOD2 and/or catalase in mitochondria changes the expression of other genes that could conceivably impact our results through undefined mechanisms.

Experiment 2: Effect of Chronic Increase in Antioxidative Capacity in Muscle Mitochondria of HF-Fed Insulin Resistant Mice

The absence of the enhanced exercise Kg in HF-fed mtAO mice contrasts with the enhanced Kg in chow-fed mtAO mice and in other HF-fed transgenic mice. This remarkable finding adds important insight into the mechanisms by which oxidative status influences metabolic regulation. HF diet feeding accelerates JH₂O₂ and impairs insulin sensitivity (2). The fact that a ∼50% reduction of JH₂O₂ in sod2^Tg or mcat^Tg mice was associated with enhanced exercise-mediated MGU whereas a more complete reduction of mtH₂O₂ in mtAO mice fully abolished this effect, suggests that the metabolic surplus of excess dietary fat intake shifts control of MGU by mtROS so that a critical mtROS level is necessary for augmentation of MGU during exercise. This suggests that a critical mitochondrial oxidative status must be present for its role in cell signaling (45, 58).

Insulin resistance and obesity impairs AMPK activation during exercise in skeletal muscle (30, 53). Exercise-induced AMPK activation was absent in HF-fed WT mice, but restored in HF-fed sod2^Tg and mcat^Tg mice to levels in chow-fed WT mice. In contrast, AMPK activity was unaffected in HF-fed mtAO mice. This coincided with the lack of an enhanced Kg in HF-fed mtAO mice. These data show that restoration of the exercise-induced increase in AMPK signaling associates with enhanced exercise-mediated Kg in insulin-resistant mice. Whether there is a causal linkage remains to be defined. The role of mtROS on regulating AMPK activation has not yet been demonstrated. One possibility is that increased mtROS production associated with HF feeding could cause an oxidative modification of the enzyme itself, therefore impairing its activation in response to exercise. It is also possible that increased mtROS associated with HF feeding decreases the upstream kinase of AMPK. Studies have shown that obese insulin-resistant rodents have decreased LKB1-AMPK-PGC-1α pathway (54). PGC-1α is an AMPK-regulated transcription factor that regulates mitochondria biogenesis (25). All these data suggest that mtROS production is associated with the regulation of AMPK activation in the HF condition. However, the effects of AMPK activation are diverse, and increased mtROS scavenging could conceivably act on a range of pathways when it restores AMPK activation. Furthermore, increased exercise Kg in HF-fed sod2^Tg and mcat^Tg mice was not associated with alterations in the expression of GLUT4 or HKII. The contribution of HKII compartmentation and spatial barriers under these conditions remains to be studied.

Nutritional State Determines the Route of O₂˙⁻ Removal

Effects of increased SOD expression on production of H₂O₂ are controversial (11, 42, 47, 49, 55, 62). Different paradigms and mathematical models have been used to explain inconsistencies (8, 9, 19, 32). The reason for these seemingly paradoxical findings is that H₂O₂ production depends on the balance of various processes that react with O₂˙⁻ (Fig. 6). These are the 1) dismutation reaction catalyzed by SOD, which produces one-half molecule of H₂O₂ per molecule of O₂˙⁻; 2) reactions that consume O₂˙⁻ to produce one or more molecules of H₂O₂ (e.g., reactions with [4Fe-4S]-containing dehydratases, hydroquinols, and thiols); and 3) reactions that consume O₂˙⁻ without producing H₂O₂ (e.g., reactions with nitric oxide, ubiquinone, and cytochrome c³⁺). The finding that chow-fed sod2^Tg mice have either increased or unchanged JH₂O₂ suggests that the rate of O₂˙⁻ removal in WT mice through reaction 3 (no H₂O₂ formation) equals to or exceeds the rate of removal through reaction 2 (≥1 H₂O₂ formation). Therefore, when SOD2 is overexpressed and the reactions for O₂˙⁻ removal are shifted to reaction 1 from reactions 2 and 3, the net balance of H₂O₂ will be either increased or remain the same. This is because the decrease in H₂O₂ produced from reactions 2 and 3 is less than or equal to the increase in H₂O₂ produced by reaction 1. Our results are consistent with the mathematic model of Gardner et al. (19). Gardner et al. (19) analyzed a minimal mathematic model where O₂˙⁻ dismutation by SOD is treated as pseudo-first-order processes, and both O₂˙⁻ production and the activity of enzymes other than SOD are considered constant. They concluded that at steady state when O₂˙⁻ production equals the rate of O₂˙⁻ consumption, the overall rate of H₂O₂ production is dependent on the rate constants of reactions 2 and 3. When rate constant for reaction 2 (k₂) is less than rate constant for reaction 3 (k₃), H₂O₂ production increases when SOD concentration increases. When k₂ = k₃, H₂O₂ production stays constant when SOD concentration increases. When k₂ > k₃, H₂O₂ goes down when SOD concentration increases (see Ref. 19 for detailed model). Therefore, our data support the existing mathematic model for this complex system.

Fig. 6.

Proposed models for how SOD2 overexpression influences mitochondrial H₂O₂ production in muscle of sedentary chow- and HF-fed mice. A: superoxide anion (O₂˙⁻), predominantly generated from complex I and III of the electron transport chain (ETC) in mitochondria is removed by 3 processes: 1) reaction catalyzed by SOD2 that produces one half of H₂O₂ per O₂˙⁻ molecule consumed; 2) reactions that produce one or more H₂O₂ per O₂˙⁻; and 3) reactions that do not produce any H₂O₂ molecules. H₂O₂ is then broken down to H₂O and O₂ by antioxidant enzymes, such as catalase targeted to mitochondria (mCAT) and glutathione peroxidase (GPx). A thicker line represents a greater reactivity. k₂ and k₃ are rate constants for reactions 2 and 3. B: as SOD2 concentration increases, the steady-state concentration of O₂˙⁻ decreases and the flux of O₂˙⁻ consumption shifts from reactions 2 and 3 to reaction 1. Dashed lines represent reduced reactions and bold lines represent increased reactions. C: on a HF diet, both steady-state levels of O₂˙⁻ and H₂O₂ are increased. We propose that O₂˙⁻ removal is reliant on reaction 2 in HF-fed WT mice. Enlarged fonts represent increased concentrations and reduced fonts represent decreased concentrations. D: in HF-fed sod2^Tg mice, mitochondrial H₂O₂ production is reduced compared with HF-fed WT mice, attributable to the shift from reaction 2 that produces one or more H₂O₂ to reaction 1 that produces only one-half of H₂O₂ per molecule of O₂˙⁻.

SOD2 activity is normally sufficient to capture >90% of the O₂˙⁻ generated in mitochondria attributable to its high activity and abundance, relative to other O₂˙⁻ removal processes (32). Hence JH₂O₂ is only increased by ∼40% in sod2^Tg mice that have a ∼300% increase in SOD2 activity (Fig. 2A). The balance of reactions described in Fig. 6 undergoes a transition with a HF diet. In contrast to an increased or constant JH₂O₂ evident in muscle of chow-fed sod2^Tg mice, sod2^Tg mice have decreased muscle JH₂O₂ in HF-fed mice. This reflects a transition so that, compared with chow-fed mice, HF-fed mice consume more mtO₂˙⁻ by the reactions that comprise 2 rather than those reactions that comprise 3 (Fig. 6). Furthermore, these data suggest that the main routes of mtO₂˙⁻ removal shift with HF feeding from reaction 1 to the reactions that comprise 2. This is supported by the fact that HF diet feeding causes a net increase in JH₂O₂ (2) with a potential decrease in SOD2 gene expression (52). One could speculate that a redox shift in sod2^Tg mice alters the susceptibility of complexes I and III to a HF diet so that excessive oxidation of these complexes could repress their ability to generate mtO₂˙⁻. This is unlikely because V̇o₂ in HF-fed sod2^Tg mice was not impaired compared with HF-fed WT mice.

Mitochondria as a Source of ROS During Exercise

It has long been assumed that ROS formation by working muscle is a by-product of increased mitochondrial oxygen consumption (10). Recent work, however, showed that mitochondria are not the major source of ROS during contraction in isolated muscle (38). The mitochondrial ROS production rate is apparently lower than original estimates, and the increase in intracellular ROS is less than the increase in extracellular ROS during exercise (26). Results from the present study, however, show that in the whole organism the mitochondrial oxidative status is critical, as an increase in the antioxidant capacity specifically targeted to mitochondria enhanced MGU during exercise. It is impossible to directly measure the predominant organelle responsible for the increase in ROS in vivo. The approach we used is advantageous in that it does not rely on direct measurements, but rather on functional outcomes of genetic manipulations of mitochondrial antioxidant enzymes. This paper shows a functional role for high chronic mitochondrial antioxidant scavenging capacity in MGU during exercise, despite the probability that extramitochondrial scavenging capacity is unaffected.

Summary

These data show for the first time that a genetic increase in mitochondrial antioxidative capacity increases an index of exercise-stimulated MGU in vivo in mice fed chow or a HF diet that accelerates ROS formation. These results link mitochondrial function to the exercise-induced increase in glucose metabolism. The present studies also emphasize that the regulation of mtROS on MGU is dependent on metabolic state because HF-fed mice are uniquely affected by muscle mitochondrial antioxidative capacity. It is noteworthy that an increase in mitochondrial antioxidative capacity alone reduces nitrotyrosine accumulation by 25–50% regardless of diet, supporting the concept that in vivo the mitochondria are important determinants of total muscle oxidative status. The present studies shed new light on the important role of mitochondrial antioxidative capacity in the physiological response to exercise in healthy and metabolic disease states.

GRANTS

This work was supported by National Institutes of Health Grants DK-073488 (to P. D. Neufer), DK-054902 (to D. H. Wasserman), and DK-059637 (Mouse Metabolic Phenotyping Center; to D. H. Wasserman).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

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