Evaluation by blue native polyacrylamide electrophoresis colorimetric staining of the effects of physical exercise on the activities of mitochondrial complexes in rat muscle

Blue native polyacrylamide electrophoresis (BN-PAGE) is a technique developed for the analysis of membrane complexes. Combined with histochemical staining, it permits the analysis and quantification of the activities of mitochondrial oxidative phosphorylation enzymes using whole muscle homogenates, without the need to isolate muscle mitochondria. Mitochondrial complex activities were measured by emerging gels in a solution containing all specific substrates for NADH dehydrogenase and cytochrome c oxidase enzymes (complexes I and IV, respectively) and the colored bands obtained were measured by optique densitometry. The objective of the present study was the application of BN-PAGE colorimetric staining for enzymatic characterization of mitochondrial complexes I and IV in rat muscles with different morphological and biochemical properties. We also investigated these activities at different times after acute exercise of rat soleus muscle. Although having fewer mitochondria than oxidative muscles, white gastrocnemius muscle presented a significantly higher activity (26.7 ± 9.5) in terms of complex I/V ratio compared to the red gastrocnemius (3.8 ± 0.65, P < 0.05) and soleus (9.8 ± 0.9, P < 0.001) muscles. Furthermore, the complex IV/V ratio of white gastrocnemius muscle was always significantly higher when compared to the other muscles. Ninety-five minutes of exhaustive physical exercise induced a decrease in complex I/V and complex IV/V ratios after all resting times (0, 3 and 6 h) compared to control (P < 0.05), probably reflecting the oxidative damage due to increasing free radical production in mitochondria. These results demonstrate the possible and useful application of BN-PAGE-histochemical staining to physical exercise studies. Correspondence


Introduction
Skeletal muscles have a highly variable requirement for energy, depending on the demand placed on them.Skeletal muscle has four major fiber types, which can be biochemically and physiologically distinguished and are classified according to the velocity of ATP hydrolysis by the myosin heavy chain as follows: fibers I, IIa, IIb and IIx/d.The myosin heavy chain composition of a muscle fiber is strongly correlated with the maximum velocity of the fiber contraction (1)(2)(3).In general, the fibers are identified by techniques such as histochemical staining and SDS-electrophoresis.These techniques have been well established for classifying fiber differences although there are still discrepancies among the different classifications and the metabolic properties of the fibers, which must have a consistently specialized machinery for their specific action (4).
Skeletal muscles possess quite a remarkable capacity of adaptation to changes in metabolic demands and physical exercise induces changes in many structural and biochemical components of this tissue (5)(6)(7).Different muscle fiber types respond differently depending on the intensity, type and duration of the exercise.Sprint training seems to be associated with an increase of fasttwitch fibers (8).Other studies report adaptations induced by endurance training in variables responsible for the improvement of oxidative metabolism such as a coordinated increase in the capacity for fatty acid, carbohydrate, ketone and amino acid oxidation and in the enzymatic pathways required for handling the reducing equivalents (9)(10)(11).Despite the beneficial adaptations induced by physical exercise, which leads to qualitative changes in muscle metabolism, an important side effect should be noted, such as the increased generation of reactive oxygen species (ROS), especially during exhaustive and unaccustomed exercise (12)(13)(14)(15).
Mitochondrial electron transport has long been recognized as a major intracellular source of oxygen radicals and hydrogen peroxide during exercise (16)(17)(18).Elevated ROS production exceeding the cellular antioxidant defense system might result in an increased oxidative stress level.Several investigators have shown greater amounts of lipid peroxidation and protein breakdown products in rodent and human muscles after an acute bout of exhaustive exercise than at rest (13,16,17,19).It is also known that antioxidant enzymes have fiber-type dependent activities, and that oxidative damage is one of the determinants of mitochondrial turnover (16,(20)(21)(22).Finally, recent reports have associated ROS production with apoptosis after physical effort, a situation in which the apoptotic mitochondrial pathways may play a major role by releasing cytochrome c and activating initiators such as caspases (23)(24)(25).Within this context, with mitochondria playing a major role in several mechanisms of physiological and biochemical muscle adaptations, our objective was to adapt the recently developed blue native polyacrylamide gel electrophoresis (BN-PAGE) method to the analysis of mitochondrial oxidative phosphorylation (OXPHOS) enzymes in different muscles.For this purpose we used the quantification of skeletal muscle OXPHOS enzymes by colorimetric enzymatic staining of BN-PAGE.This is a method recently developed for analyzing membrane complexes (26)(27)(28), which has the important advantage of using small quantities of muscle for analysis and which does not require the isolation of mitochondria.Moreover, it specifically measures the OXPHOS enzymes whereas most spectrophotometric assays also detect other cellular activities (27,28).
We standardized BN-PAGE and colorimetric enzymatic staining to directly quantify complex I and complex IV mitochondrial activities in tissue extract of rat muscles with different morphological and biochemical properties such as soleus, red and white gastrocnemius and semitendinosus muscles.We also analyzed the activities of mitochondrial complexes I and IV at different times after an acute bout of exhaustive exercise to determine if this technique could be applied to physical exercise studies.

Animals and muscle sampling
The experiments were performed using 2-month-old male Wistar rats maintained on an inverted 12-h dark-light cycle, at 22ºC, with food and water ad libitum.The animals were anesthetized with an intraperitoneal injection of 10% chloral hydrate (w/v), 0.3 ml/100 g body weight.The red gastrocnemius, white gastrocnemius, soleus, and semitendinosus muscles were carefully dissected, immediately frozen in liquid nitrogen and subsequently stored at -70ºC until analysis.The experimental procedures were approved by the Committee for Animal Use in Research of the Biological Institute, UNICAMP, Brazil.

Sample preparation and electrophoresis technique
The method of sample preparation for BN-PAGE was that originally described by Zerbetto et al. (27), with the following modifications: muscle samples (30 mg) were minced and homogenized in 1 ml of a cooled solution containing 20 mM MOPS, 440 mM sucrose, 1 mM EDTA and 5 mM PMSF, pH 7.2, in a Kinematica Polytron Homogenizer (Westbury, NY, USA) with a generator diameter of 7 mm at the highest speed for 10 s.After centrifugation at 20,000 g for 20 min the supernatants were submitted to the enzyme assay (citrate synthase), as described by Srere (29).The pellet was re-suspended in 80 µl of 1 M 6-aminocaproic acid and 50 mM Bis-Tris, pH 7.0, containing 5 mM PMSF.The membranes were solubilized by the addition of 30 µl of freshly prepared 10% n-dodecylmaltoside.After 25 min of centrifugation at 100,000 g, the supernatants were collected and stored in small aliquots of 20 µl each at -70ºC until analysis.These aliquots were charged with 1 µl of 5% serva blue G diluted in 1 M 6-aminocaproic acid.Then, 5 µl of each sample was loaded per slot of a 5-11% polyacrylamide gradient gel.BN-PAGE was performed as described elsewhere (26).Protein concentration was determined by the method of Bradford (30).These procedures were performed at 4ºC.
All chemicals were obtained from Sigma, except for 6-aminocaproic acid and ndodecylmaltoside, which were obtained from Fluka and Boehringer Mannheim, respectively.

Colorimetric enzymatic staining of BN-PAGE gels
BN-PAGE gel slices were submitted to colorimetric enzymatic staining immediately after the electrophoretic run and complex I (NADH dehydrogenase) and complex IV (cytochrome c oxidase, COX) activities were evaluated.Complex I activity was analyzed by incubating the gel slices in a solution containing 2 mM Tris-HCl, pH 7.4, 0.1 mg/ ml NADH and 2.5 mg/ml NBT, diluted 1:10 (w:v) at room temperature (27).The gel was incubated in 50% methanol and 10% acetic acid for 15 min to fix the color of complex I reacting bands and then preserved in 10% acetic acid.Complex IV COX activity was estimated by incubating the BN-PAGE gels with 5 mg of DAB dissolved in 9 ml 50 mM phosphate buffer, pH 7.4, 10 mg cytochrome c, 0.05 ml catalase (30 mg/ml), and 0.76 mg sucrose as described by Zerbetto et al. (27).The gel was then fixed and preserved as described above.The colorimetric reactions were allowed to continue until maximum color was developed.The color intensity of the stained bands was assessed by scanning the still wet gel with the "Sharp-Scanner" for the gel image and the area was quantified by the Image Master 1D-program (Amersham Pharmacia).The areas of the bands are reported either in arbitrary units or relative to the area of Coomassie blue-stained complex V, thus permitting the BN-PAGE results to be reported quantitatively.

Exercise protocol
Twenty rats that showed willingness to run on a motorized treadmill were randomly assigned to a sedentary control group (5 animals) or to an exhaustion exercise group (15 animals).All the animals were habituated to a low intensity treadmill running daily session of 5 min at a speed of 5 m/min for one week.This regimen was used to ensure that the rats could run as outlined in Table 1, but would be minimally influenced by a training effect.The results showed that 95 min was the maximum time the sedentary rats could run before exhaustion, when they refused to continue running.Soleus muscles were removed either 0, 3 or 6 h after the end of the exercise protocol (5 animals for each time) and the same procedure was performed with the control group.

Statistical analysis
Before applying the statistical tests, a pretest was performed to determine whether a parametric or nonparametric test should be used.Citrate synthase activity, protein determination, evaluation of colorimetric enzymatic staining, and kinetics experiments after exercise are reported as means ± SEM and statistical comparison was performed by one-way ANOVA, with the level of significance set at P ≤ 0.05.The graphs for the correlation between complex V area and citrate synthase activity had a 95% confidence interval, assuming Gaussian distributions, and P ≤ 0.05 was considered to be significant.

Mitochondrial complexes I and IV from different types of skeletal muscle fibers
Figure 1A shows the Coomassie blue staining of the gels for all OXPHOS enzymes and the colorimetric enzymatic staining of NADH and COX complexes of four rat muscles (soleus, red gastrocnemius, white gastrocnemius, and semitendinosus) after   detergent extraction and BN-PAGE.The five major bands represent the oxidative phosphorylation enzymes although the best resolutions were consistently achieved for complexes I, III and V. Complexes IV and I were well identified using the enzymatic colorimetric reactions set up for histochemical analysis.The colorimetrically stained enzymatic activities of the complexes were localized specifically in a single or double band, while after Coomassie blue staining the complexes were generally identified as a single band, less strong than that obtained by colorimetric staining.Only the COX activity appeared in a single or double staining band due to its monomeric and dimeric forms.There were apparent differences in the width of the gel bands representing OXPHOS enzymes from different muscles, mainly by histochemical staining.Broader bands were observed for the more oxidative muscles (soleus and red gastrocnemius) and narrower bands were observed for more glycolytic muscles (white gastrocnemius and semitendinosus).This result may reflect the higher and lower oxidative potentials of these muscles, respectively.This observation is confirmed and quantified in Table 2 which shows the activity of citrate synthase, used as a mitochondrial content marker.Table 2 also shows the total protein content of the same samples.It is evident that this measurement could not be used as a good marker of mitochondrial content because it does not reflect the oxidative capacity of different muscles.
As shown in Figure 2, a linear correlation between complex V area and citrate synthase activity was found for all four muscles.This indicates that the Coomassie bluestained complex V area is a better indicator of mitochondrial protein content than the total protein content shown in Table 2, and that the complex V area is a useful internal standard of the OXPHOS proteins loaded onto the gel, as shown by Zerbetto et al. (27).Table 3 shows the stained area of complexes I and IV reported as a function of the stained area of the Coomassie bluestained complex V (internal standard).The ratio between the colorimetric and Coomassie blue staining of oxidative muscles, soleus and red gastrocnemius, fell within a narrow range of values for both complexes and also showed a linear correlation with the internal standard, complex V (Table 3).The same ratios were more variable in glycolytic muscles, white gastrocnemius and semitendinosus, and presented higher values when compared to those for the oxidative muscles.There were significant differences in the complex I/V ratio only when white gastrocnemius muscle was compared to red gastrocnemius and semitendinosus muscles (Table 3) while there was a significant difference in the complex IV/V ratio between white gastrocnemius and all the other muscles analyzed (Table 3).Citrate synthase (nmol min -1 mg tissue -1 ) was determined in the 20,000 g supernatant of a 30-mg homogenate.Protein content (mg/ml) was determined in the 20,000 g supernatant by the method of Bradford.CS = citrate synthase; RG = red gastrocnemius; SO = soleus; ST = semitendinosus; WG = white gastrocnemius.*P < 0.001 compared to all muscles; # P < 0.001 compared to WG and ST (one-way ANOVA).**P < 0.01 compared to WG and ST; + P = 0.01 compared to WG (one-way ANOVA).Quantitative data were obtained from the ratio of the activity and the internal standard of each muscle.RG = red gastrocnemius; SO = soleus; ST = semitendinosus; WG = white gastrocnemius.I/V: *P < 0.001 compared to RG; # P < 0.05 compared to SO. IV/V: *P < 0.01 compared to RG and ST; # P < 0.05 compared to SO (one-way ANOVA).

Soleus muscle mitochondrial complex I and IV activities after exhaustive exercise
The colorimetric enzymatic staining of OXPHOS enzymes was applied to the analysis of the effect of exhaustive physical exercise on the mitochondrial complexes IV and I of rat soleus muscle.Samples were collected at different resting periods after exercise and the results are presented in Figure 1B and Table 4.There was a significant reduction of the complex I/V and complex IV/V ratios, detected immediately after the end of exhaustive exercise (0 h).These de-creased ratios persisted for 3 and 6 h after the end of exercise (P < 0.01 and P < 0.05) for both complexes.

Discussion
BN-PAGE is an electrophoresis technique developed for analyzing membrane complexes and is particularly useful for characterizing the mitochondrial OXPHOS enzymes.As shown by Schagger and Jagow (26), complexes I and V run on the gel in the monomeric state with molecular masses of about 750 and 600 kDa, respectively, while complex III runs on the gel as a dimer with a molecular mass of 500 kDa.Complex IV, 200 kDa, runs in its monomeric form, but occasionally a second band for this complex is detected, possibly representing the dimeric form of the same complex.Complex II runs at the bottom of the gel, with a molecular mass of 130 kDa.Studies combining BN-PAGE with colorimetric enzymatic staining have been recently performed in an attempt to simultaneously identify and quantify the activity of the OXPHOS complexes on the gel.These methods have been successfully applied to samples of skeletal muscles from humans, cows and rats (27).However, the results obtained in the analysis of rat skeletal muscles were not totally satisfactory.In an attempt to improve this characterization, the morphological (myosin heavy chain composition of a muscle fiber) and functional (oxidative and glycolytic metabolism) differences between fiber types were considered.The present results showed that BN-PAGE in combination with the colorimetric enzymatic staining permits the identification and quantification of the activities of mitochondrial complexes I and IV from rat skeletal muscles with different oxidative capacities, as already shown for human skeletal muscles.
According to Delp and Duan (31), the white gastrocnemius and semitendinosus muscles showed the higher percentage of type IIb fibers, 95 and 74% respectively, whereas the soleus and red gastrocnemius muscles presented a higher percentage of type I fibers, 86 and 55%, and type IIa fibers, 6 and 36%, respectively.Citrate synthase activity, used as a marker of mitochondrial protein content (Table 2), apparently reflected the same pattern of fast-twitch glycolytic/ fast-twitch oxidative and slow-twitch oxidative classification of muscles by percentage of type fibers although we used a different animal breed.However, the metabolic characteristics and fiber composition of both animals and humans and how they correlate with the activity of a specific muscle group have been poorly explored in the literature.Furthermore, the mitochondrial content of the working muscle cells is thought to act as a key regulatory mechanism in substrate selection (32).
An interesting observation made here was that the white gastrocnemius muscle, which has fewer mitochondria, showed high catalytic activities of complexes I and IV in comparison with the other muscles, probably to compensate for its low oxidative potential.In agreement with our data, Howlett and Willis (32) observed that the isocitrate dehydrogenase (NAD-IDH) activity is higher in mitochondria from type IIb muscle, suggesting that type IIb mitochondria may rely on this enzyme as a regulatory site.This enzyme catalyzes the first dehydrogenase reaction in the TCA cycle and produces NADH + H + , which is the substrate of NADH dehydrogenase.This may indicate a higher activity of NADH dehydrogenase, demonstrated in the present study.
Conflicting results regarding the effect and the role of exhaustive physical exercise on mitochondria and their production of ROS have been reported in the literature (13,16,22,33,34).It seems that ROS play a primordial role in altering mitochondrial function in a variety of patho-physiological conditions (35), and in physical exercise situations mitochondrial alterations seem to depend on exercise schedule, previous training and   • • • • -formation (40).Thus, the decreased complex IV activity shown in our experiment supports the probability of ROS attack in mitochondrial membrane complexes.

Time
15) were submitted to this protocol until they refused to run, and were then divided into three groups, resting for 0, 3 and 6 h after the end of exhaustive physical exercise.

Figure 1 .
Figure 1.A, Colorimetric enzymatic staining on BN-PAGE gels of different rat muscle samples.Coomassie blue staining of all complexes (Blue), complex I (NADH dehydrogenase) and complex IV (cytochrome c oxidase, COX).SO = soleus, RG = red gastrocnemius, WG = white gastrocnemius, ST = semitendinosus.B, Colorimetric enzymatic staining on BN-PAGE gels of soleus muscles from exhaustively exercised rats.Coomassie blue staining of all complexes (Blue), complex I (NADH) and complex IV (COX).Control group (CO) and experimental samples were collected immediately (0), 3 and 6 h after exhaustive exercise.The Figure is representative of several experiments performed.Complex I V III

Table 1 .
Protocol of exhaustive physical exercise.

Table 2 .
Mitochondrial citrate synthase activity and protein content of different rat skeletal muscles.

Table 3 .
Complex I (NADH dehydrogenase) and IV (cytochrome c oxidase) activities in rat muscles determined by colorimetric enzymatic staining of BN-PAGE gel assays.

Table 4 .
(39)vity ratios of complexes I and IV in soleus muscle after exhaustive physical exercise.The values were obtained by densitometric analysis of the colorimetric enzymatic stained areas of complexes I and IV and the Coomassie blue-stained area of complex V of the control group (internal standard) run on the same gel.Quantitative data were determined by the ratio between the activity and the internal standard (N = 5).Results are reported as means ± SEM. *P < 0.05 compared to control group (one-way ANOVA).musclefibertype.Zhang et al.(36)showed mitochondrial enzyme inactivation as the first step of oxidative damage, which was further related to exercise-induced oxidative stress.Leeuwenburgh et al.(37)and Radák et al.(38)found increased levels of oxidized amino acids in muscle mitochondria but not in cytosolic proteins, suggesting that exercise is a physiological source of oxidative stress.It has also been suggested that oxidative injury might cause irreversible cell damage through the loss of homeostasis and the loss of mitochondrial function(39).Given the high oxygen consumption during exercise, mitochondria would be requested for energy supplying and might simultaneously produce ROS.These considerations agree with our results, which show decreasing activities of mitochondrial complexes I and IV no matter the time of sampling, and this may indicate that mitochondria are the main target of exercise-induced ROS production and attack.In addition, COX appears to have an indirect antioxidant action because of its affinity for O 2 , reducing O 2