INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS AND PROCEDURES RESULTS DISCUSSION REFERENCES

PROOXIDANT AND ANTIOXIDANT balance is critical for the survival of aerobic organisms. In the skeletal muscle of mammals, oxidative stress caused by generation of reactive oxygen species (ROS) has been shown to initiate a series of disorders related to modification of cellular constituents and destruction of muscle contractile function (18, 35). As a defensive strategy, muscle cells are capable of inducing antioxidant enzymes such as superoxide dismutase (SOD), glutathione peroxidase (GPX), and catalase (CAT), to remove harmful ROS (19, 20). Among the various nonenzymatic antioxidants, only glutathione (GSH) has been shown to increase in selective muscle fibers as a result of exercise training (29). No such adaptation has been reported for -tocopherol (vitamin E) or ascorbic acid (vitamin C) (25).

Superoxide radical (O₂·), in most occasions, is the first ROS produced during the one-electron reduction of molecular oxygen and is converted to H₂O₂ by SOD (6). In both healthy and diseased states, adequate protection by SOD is critical for the functionality of skeletal muscle (33). In mammalian tissues, SOD exists in two main forms, dependent on the metal ion bound to its active site. CuZn-containing SOD is a highly stable enzyme found primarily in the cytosol. CuZn SOD is a dimer (mol mass 32,000 Da) and is sensitive to cyanide and H₂O₂ inhibition (12). Mn SOD is a tetramer with a much larger molecular mass (88,000 Da) that is present in the mitochondrial matrix and is insensitive to cyanide and H₂O₂ (33). The two SODs have distinct characteristics in terms of protein turnover rate. Recombinant human SOD studies reveal that CuZn SOD has a half-life (t_1/2) of only 6-10 min, whereas Mn SOD has a much longer t_1/2 of 5-6 h (15). Both the relative abundance of mRNA and the catalytic activity of SOD have been shown to correlate with tissue metabolic rate, as well as oxidative capacity of various muscle fibers (20, 27, 32).

Endurance training has been shown to increase SOD activity in skeletal muscle (16, 23, 28, 34). However, most studies to date have been limited to measurements of enzyme activities, with few examinations of SOD gene expression (13, 31). Furthermore, the reported adaptive responses of CuZn and Mn SOD are inconsistent, possibly because of differences in training intensity, muscle fiber type, and the different enzyme assay methods employed. To gain more definitive information about the regulatory mechanisms involved, we need to not only measure enzyme activity but also investigate mRNA level and enzyme protein content of SOD. Oh-Ishi et al. (31) observed that CuZn SOD activity in rat soleus muscle (type I) was significantly increased with training, but the enzyme protein content and mRNA levels were not altered. In contrast, Mn SOD showed both an increased activity and protein content, yet steady-state mRNA levels were unaffected. Because antioxidant enzyme response to acute and chronic exercise is highly muscle fiber specific, it is not clear whether these findings derived from type I skeletal muscle can be generalized to all muscle fiber types. Because of the distinctive phenotypic and genotypic characteristics in skeletal muscle, investigation of fiber-specific responses of CuZn and Mn SOD would be of both physiological and clinical significance.

To gain further insight into the regulatory mechanisms involved in antioxidant enzyme adaptation, we examined the enzyme activity, enzyme protein content, and mRNA abundance for both CuZn and Mn SOD in rats in response to 10 wk of endurance training. These measurements were carried out in four types of hindlimb muscles: soleus (type I), deep vastus lateralis (DVL, type IIa), superficial vastus lateralis (SVL, type IIb), and plantaris (mixed type). We hypothesize that 1) training induction occurs only in type I and IIa but not in type IIb muscle; 2) endurance training induces Mn SOD but not CuZn SOD as a result of the former enzyme's mitochondrial location; 3) increased muscle SOD activity with training results from elevated SOD protein content; and 4) training induction of SOD is caused by an increase in the steady-state concentration of their respective mRNAs (i.e., via a pretranslational mechanism).

	METHODS AND PROCEDURES

TOP ABSTRACT INTRODUCTION METHODS AND PROCEDURES RESULTS DISCUSSION REFERENCES

Animals. Sixteen female Sprague-Dawley rats (age, 3 mo; body wt, 200-210 g) were housed individually in the animal facilities at the University of Wisconsin-Madison with a reverse 12:12-h light-dark cycle (0700-1900 dark; 1900-0700 light). Rats were fed a Purina chow diet and tap water ad libitum. Animal use protocol was approved by the University of Wisconsin-Madison Research Animals and Resource Center Review Committee. After a 1-wk acclimation period, rats were randomly divided into trained (T, n = 8) and untrained (U, n = 8) groups. Female rats were used because male rats are known to decrease their food intake and lose body weight during training, thereby confounding the results as a result of decreased antioxidant consumption. Female rats, on the other hand, maintain food intake and body weight during training, and demonstrated prominent muscle antioxidant enzyme adaptations (13, 29).

Exercise. Animals were introduced to treadmill exercise for 2 wk before initiation of the training protocol. During week 1, the animals were acclimated to treadmill running at 15 m/min, 0% grade, for 10 min/day, 5 day/wk. By the end of week 1, rats were able to run at 16.5 m/min and 0% grade for 30 min. Running speed, grade, and duration were progressively increased to 27 m/min, 12% grade, for 2 h/day by the end of week 5. This intensity was maintained for the rest of the 10-wk training period.

All rats were killed in the resting state. Trained rats were killed ~48 h after the end of their last training session to minimize acute exercise effects from the last training bout. Untrained rats were killed at the same time as their trained counterparts.

Tissue preparation and enzyme activity. Rats were killed by decapitation. After exsanguination, DVL, SVL, plantaris, and soleus muscles were immediately excised and divided into several portions. One part of the muscle was submerged in ice-cold 0.1 M Tris · HCl buffer (pH 7.4), minced, and homogenized at 0-4°C. The resulting muscle homogenates were subjected to a centrifugation at 500 g for 20 s. Cell debris and connective tissues in the pellets were discarded, and the supernatants were stored at 80°C for enzyme assays. Another portion of the muscle was frozen at 80°C for RNA extraction.

Total SOD activities in the various muscles were measured according to Sun and Zigman (36). Mn SOD was determined by inhibition of CuZn SOD with potassium cyanide as described by Higuchi et al. (16). Difference between total SOD and Mn SOD was defined as CuZn SOD activity. GPX activity was measured by the method of Flohe and Gunzler (11). Catalase activity was measured by means of the procedure described by Aebi (1) with some modification (21). Activity of the mitochondrial enzyme marker citrate synthase (CS) was measured as previously described (21). Protein content was measured by the Bradford method, with BSA as a standard (4).

Western analysis. SDS-PAGE in combination with an enhanced chemiluminescence (ECL) detection system was used to determine individual isozyme protein contents of Mn SOD and CuZn SOD (26). After electrophoresis, gels were equilibrated in transfer buffer (2.5 mM Tris base, 19.2 mM glycine) and transferred overnight at 20 V. Blots were blocked in 5% nonfat dry milk in PBS-T (80 mM Na₂HPO₄, 20 mM NaH₂PO₄, 100 mM NaCl, 0.1% Tween 20). Relative amounts of enzyme proteins were determined by means of monoclonal antibodies to rat Mn SOD and CuZn SOD (gifts of Naoyuki Taniguchi, Osaka University Medical School, Osaka, Japan). Incubation with secondary antibodies and detection of signal was performed in accordance with manufacturer's directions by means of an Amersham ECL detection system (Amersham, Arlington Heights, IL). Autoradiographic signals were assessed with a BioRad scanning densitometer (BioRad, Hercules, CA).

RNA analysis. Total RNA was isolated by means of the method of Chomczynski and Sacchi (7). The cDNA probes for CuZn SOD, Mn SOD, and -actin were labeled with random primer extension with [³²P]dCTP (10). Northern blot analysis was performed for each SOD to verify that the transcript size corresponded to the enzymes under investigation, as previously described (13). For quantification of relative mRNA abundance, slot blots were performed on the samples. Preliminary slot blots were performed with a range of RNA levels to determine linear autoradiographic response. For hybridization, slot blots were prehybridized for at least 3 h in a solution consisting of 50% formamide, 5× Denhardt's solution, 5× saline-sodium phosphate-EDTA, and 0.1% SDS at 42°C. Radiolabeled probes were added at a level of 10⁶ counts · min¹ · ml¹ prehybridization solution containing 10% dextran sulfate and allowed to hybridize overnight. Stringency washes consisted of two 20-min washes with 1× saline sodium citrate (SSC), 0.5% SDS, and two 20-min washes with 0.5× SSC, 0.5% SDS. All washes were performed at 42°C. Filters were wrapped in plastic and exposed to film at 80°C. Quantification of autoradiographic signals was performed as described earlier for Western analysis. Relative intensities of antioxidant probe signals were normalized to the -actin signals of the same samples.

Statistics. A student's t-test (Systat, version 5.03; Evanston, IL) was used to determine significant differences between the mean values of the U and T animals. P < 0.05 was considered significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS AND PROCEDURES RESULTS DISCUSSION REFERENCES

Body weight was not significantly different between the T and U rats (data not shown). Protein contents of the 500-g supernatant in the various types of skeletal muscle were not significantly different between treatment groups (Table 1). CS activity in soleus, DVL, and plantaris muscles was significantly increased by 65, 68, and 46% (P < 0.01) with training, respectively, indicating an increase in mitochondrial oxidative capacity. However, CS activity in SVL was unchanged. In addition to SOD, activities of GPX and CAT were evaluated in the various muscles. GPX activity was increased by 155% (P < 0.01) with training in DVL muscle (Table 1). In addition, CAT activity was 89% (P < 0.05) higher in T vs. U rats in DVL. Activities of these two enzymes were unchanged in the other muscles.

Training increased Mn SOD activity by 81% (P < 0.05) in DVL muscle (Table 2). Soleus muscle in T rats tended to have a higher Mn SOD activity (36%, P < 0.11) than that in U rats. Mn SOD activity was not different between T and U in SVL or plantaris. CuZn SOD activity showed no significant change with training in most muscle fibers. However, in DVL there was a trend toward a higher CuZn SOD activity in T vs. U rats (P < 0.12). Thus total SOD activity in DVL showed a tendency of 52% increase (P < 0.075) with training. No training effect on total SOD activity was observed in other muscle types.

Figure 1 illustrates a typical Northern blot for Mn SOD and CuZn SOD for the various muscle tissues from an untrained rat. Signals from gastrocnemius (type II) and heart muscles were included for comparison. The multiple bands in the Northern blot for Mn SOD represent various gene transcript products with size ranging from 1.0 to 4.0 kb, as described previously (17). For CuZn SOD, a single band of 0.7 kb was detected. Figures 2 and 3 illustrate the relative abundance of mRNA for Mn SOD and CuZn SOD in the various muscles examined. None of the muscles showed a significant change in Mn SOD mRNA abundance with training (Fig. 2). CuZn SOD mRNA levels in all muscle types appeared lower in T vs. U rats, but only plantaris reached a significant 50% decrease (P < 0.05; Fig. 3).

View larger version (73K): [in this window] [in a new window]   Fig. 1.   Northern blots of mRNA for Mn superoxide dismutase (SOD), CuZn SOD, and -actin in various tissues examined. DVL, deep vastus lateralis; SVL, superficial vastus lateralis; Gastro, gastrocnemius. Note that various Mn SOD gene transcript products are shown with size ranging from 1.0 to 4.0 kb, whereas for CuZn SOD a single band represents 0.7-kb transcript. For details, see text and Ref. 17.

View larger version (13K): [in this window] [in a new window]   Fig. 2.   Relative abundance of Mn SOD mRNA in various tissues examined. Plant, plantaris; Sol, soleus. Values are means ± SE for 8 (trained) and 8 (untrained) rats. Relative abundance of mRNA was normalized with -actin signal of same sample as a reference.

View larger version (13K): [in this window] [in a new window]   Fig. 3.   Relative abundance of CuZn SOD mRNA in various tissues examined. Values are means ± SE for 8 (trained) and 8 (untrained) rats. Relative abundance of mRNA was normalized with -actin signal of same sample as a reference. * P < 0.05, trained vs. untrained.

Muscles from T rats appeared to have higher Mn SOD protein levels than those from U rats; however, prominent differences were observed only in plantaris (42%) and SVL (54%; Fig. 4). For CuZn SOD (Fig. 5), only SVL showed a prominent 160% increase in enzyme protein content comparing T vs. U rats.

View larger version (21K): [in this window] [in a new window]   Fig. 4.   Immunoreactive Mn SOD protein content in various tissues examined. Each bar represents value of pooled muscle samples from 8 rats, derived from autoradiographic signals as shown at top.

View larger version (23K): [in this window] [in a new window]   Fig. 5.   Immunoreactive CuZn SOD protein content in various tissues examined. Each bar represents value of pooled muscle samples from 8 rats, derived from autoradiographic signals as shown at top.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS AND PROCEDURES RESULTS DISCUSSION REFERENCES

Despite extensive research over the years, the effect of endurance exercise on SOD is still not entirely clear as a result of inconsistent results derived from various studies (for a detailed review, see Ref. 33). The discrepancies may be related to differences in animal age, experimental diet, exercise mode, intensity and duration, as well as variable assay methods used to measure SOD activity (16, 23, 28, 29, 34). However, the most important factor may be related to the complexity of gene regulation for this enzyme. Furthermore, skeletal muscles are highly heterogeneous. Each muscle fiber type has distinct metabolic characteristics and oxidative potential, as well as antioxidant defense capacity. To gain more insights into the mechanisms involved in SOD training adaptation, it is critical to assess the fiber-specific alterations of SOD mRNA abundance and immunoreactive protein content in conjunction with enzyme activity.

Data in the current study partly supported our first hypothesis that endurance training increases SOD activity only in the oxidative but not glycolytic muscle fibers. Our results were in general agreement with those of Powers et al. (34) that training had no effect on total SOD activity in the white gastrocnemius (IIb) muscle, and only high-intensity training (30 m/min, 12-18% grade) could increase SOD activity in the red gastrocnemius (IIa) muscle. The data were also in agreement with our previous study, showing that training adaptation of total SOD activity occurred in DVL but not SVL muscle in rats (13). In the present study soleus muscle showed no adaptation in CuZn SOD, and the increase in Mn SOD activity with training did not reach statistical significance (P < 0.11). These findings did not support our original hypothesis regarding type I muscle fiber and contradicted some previous studies (31, 34). However, the data were consistent with our previous work, reporting no training adaptation of either CuZn or Mn SOD in soleus muscle when a similar training protocol was used (25 m/min, 10%) (29). Compared with Powers et al. (34), the exercise intensity in both of our studies was relatively low and might have limited the recruitment of and stimulation to soleus muscle. Training induction of antioxidant enzymes is generally believed to be a cellular adaptation to oxidative stress caused by free radical generation during heavy exercise (19, 20). DVL muscle exhibited the highest oxidative potential (marked by CS activity) among various fibers and has been shown to increase ROS production during an acute bout of exercise (3). However, it possesses only modest SOD, GPX, and CAT activities (Tables 1 and 2). These unfavorable antioxidant/oxidant profiles may have subjected the DVL to a greater oxidative stress. In comparison, soleus muscle had a similar oxidative capacity but much greater reserves of antioxidant enzyme levels. Furthermore, GSH content in soleus is two to three times higher than that in DVL (22, 29). Thus the differential training responses in favor of DVL muscle should not be surprising.

Of the two isoforms of SOD, only Mn SOD demonstrated a clear training adaptation in DVL, whereas CuZn SOD activity showed only a marginal increase with training. These findings confirmed an early study of Higuchi et al. (16), who found that training adaptation occurred only in Mn SOD but not in CuZn SOD. Because Mn SOD is located in the mitochondria, the question arises as to whether the increased Mn SOD activity was caused by the training-induced mitochondrial proliferation in skeletal muscle. Data in the current study did not support this notion. First, the increase in Mn SOD activity found in DVL (81%) was greater than that in CS activity (65%), a mitochondrial enzyme marker. Furthermore, CS activity was increased 68 and 46% with training in soleus and plantaris, respectively, whereas no adaptation of Mn SOD was observed in these two muscles. These data suggest that training adaptation of Mn SOD cannot be predicted solely by enhanced oxidative metabolism and is more likely induced by a specific mechanism. Although located in the mitochondria, Mn SOD is encoded with a nuclear gene, and the precursor enzyme synthesized in the cytosol is transported into the mitochondria by an energy-dependent process (39). Various forms of oxidative stress unrelated to exercise or increased metabolic rate, such as irradiation, endotoxin, and ROS, have been shown to induce Mn SOD (37-39). Indeed, Mn SOD protein content was elevated prominently in SVL and plantaris of trained rats, wherein no training adaptation of CS or change in steady-state concentration of Mn SOD mRNA took place. Similarly, Oh-Ishi et al. (32) reported a profound increase in immunoreactive Mn SOD content without alteration of mRNA abundance in rat soleus muscle after endurance training. There are two possible explanations for the discrepancy between enzyme protein and mRNA levels. First, eukaryotic cells are capable of increasing protein synthesis without prior activation of mRNA transcription, accomplished by increasing mRNA stability and/or efficiency for translation (9). However, we are not aware of such a mechanism operational in skeletal muscle. A second possibility is that each training bout in fact enhanced the transcription of Mn SOD gene, but the mRNA returned to the steady-state levels 48 h after exercise when the animals were killed. The subsequent elevation of Mn SOD protein synthesis might have persisted, resulting in a higher enzyme protein level. This explanation may be more plausible because mRNAs usually have much shorter half-lives than enzyme proteins and can be degraded rapidly after transient upregulation (15). In a previous study wherein trained rats were killed 24 h after their last exercise bout (13), we indeed found a significant increase in CuZn SOD mRNA level in the DVL, heart, and liver of trained rats. However, SOD protein content was not increased, presumably as a result of insufficient time for protein synthesis.

Cellular mechanisms pertaining to training induction of Mn SOD are still elusive. A potential regulatory site may be related to nuclear protein binding to putative DNA sequences on Mn SOD promoter (30). Mammalian Mn SOD gene contains nuclear factor B (NF-B) and activator protein-1 (AP-1) binding sites that can be activated by cytokines, toxins, and ROS (17, 24, 37, 39). Interleukin-1 and tumor necrosis factor (TNF)- have been shown to upregulate Mn SOD transcription via NF-B binding (8), whereas hyperbaric O₂ and radiation-induced O₂· and H₂O₂ increase Mn SOD mRNA transcription (2, 38). During and/or after a 2-h exercise session, ROS production has been shown to increase in rat DVL muscle (3). Furthermore, cytokine levels are expected to rise substantially in the blood of exercised rats (5). These factors could potentially enhance AP-1 and/or NF-B binding in muscle cells, as has been shown in a previous study in our laboratory (14), thereby affecting Mn SOD transcription. In contrast, CuZn SOD gene does not contain these putative binding sites. The different promoter sequences may account, at least in part, for the differential training responses between the two SOD isozymes.

The prominent increases in both Mn SOD (54%) and CuZn SOD (160%) protein content in the type IIb muscle SVL with training was unexpected, especially because SOD mRNA and activity were unaltered. Mitochondrial ROS production in SVL is expected to be low because of its low rate of oxidative metabolism. However, other ROS sources during strenuous muscular contraction have not been fully characterized. Some factors might have directly stimulated SOD protein synthesis without transcriptional activation. Much of the signaling pathway of SOD gene expression is still largely unknown, leaving little room for speculation.

In summary, individual skeletal muscle fibers are affected differently by endurance training, and the mechanisms involved in SOD training adaptation are complex and may be specific for each muscle fiber type. Nevertheless, the following conclusions can be drawn from the current data: 1) type IIa muscle demonstrates a greater training adaptation than type I or type IIb muscle as a result of a higher level of oxidative metabolism and ROS generation and a modest antioxidant defense capacity; 2) endurance training induces primarily Mn SOD with both enzyme protein and activity being elevated; and 3) training adaptation of Mn SOD is not associated with increased steady-state mRNA abundance in muscle. Whether there is a transient activation of Mn SOD gene expression between training sessions remains to be examined with time-course studies.

Perspectives