The purpose of this investigation was to examine the temporal changes in uncoupling protein (UCP)-3 expression, as well as related adaptive changes in mitochondrial density and fast-to-slow fiber type transitions during chronically enhanced contractile activity. We examined the effects of 1–42 days of chronic low-frequency electrical stimulation (CLFS), applied to rat tibialis anterior (TA) for 10 h/day, on the expression of UCP-3 and concomitant changes in myosin heavy chain (MHC) protein expression and increases in oxidative capacity. UCP-3 protein content increased from 1 to 12 days, reaching 1.5-fold over control (P < 0.0005); it remained elevated for up to 42 days. In contrast, UCP-3 mRNA decreased in response to CLFS, reaching a level that was threefold lower than control (P < 0.0007). The activities of the mitochondrial reference enzymes citrate synthase (EC 126.96.36.199) and 3-hydroxyacyl-CoA-dehydrogenase (EC 188.8.131.52), which are known to increase in proportion to mitochondrial density, progressively increased up to an average of 2.3-fold (P < 0.00001). These changes were accompanied by fast-to-slow fiber type transitions, characterized by a shift in the pattern of MHC expression (P <0.0002): MHCI and MHCIIa expression increased by 1.7- and 4-fold, whereas MHCIIb displayed a 2.4-fold reduction. We conclude that absolute increases in UCP-3 protein content in the early adaptive phase were associated with the genesis of mitochondria containing a normal complement of UCP-3. However, during exposure to long-term CLFS, mitochondria were generated with a lower complement of UCP-3 and coincided with the emergence of a growing population of oxidative type IIA fibers.
- mitochondrial density
- fiber type transitions
uncoupling protein (UCP)-3 is a transmembrane carrier protein located on the inner mitochondrial membrane and is highly expressed in skeletal muscle (5). Because of its high sequence homology with UCP-1, it has been proposed to regulate thermogenesis and contribute to the basal tissue metabolic rate by uncoupling the mitochondrial electrochemical H+ gradient from ADP phosphorylation. In this regard, it has been estimated that UCP-3 activity may contribute as much as 50% of the resting metabolic rate (RMR) of skeletal muscle (43) and ∼20% of the overall RMR of the rat (6). Indeed, UCP-3 expression at the mRNA level was shown to positively correlate with the resting metabolic rate and body composition in humans (27, 49) and was upregulated in humans during a period of increased postexercise oxygen consumption (44).
It has also been proposed that UCP-3 is directly involved in the mitochondrial export of long-chain fatty acid anions as part of a futile cycle encompassing the two mitochondrial carnitine-palmitoyl transferases, as well as fatty acid synthase (17, 41). The role of this putative cycle is thought to be to maintain sufficient concentrations of free mitochondrial coenzyme A (CoASH) and NAD+ under conditions where fatty acid oxidation predominates (e.g., Ref. 17). More evidence for a metabolic role may be derived from recent observations of transgenic mice expressing a high copy number of the human UCP-3 gene (hUCP-3) (8). hUCP-3 mice exhibit lower fat mass, hyperphagia, and increased fat combustion; they also display increased mitochondrial uncoupling, higher intramuscular temperatures, elevated O2 uptake, and lower mitochondrial membrane potential. The latter has been proposed as an important basis for antioxidant defense (26). Interestingly, UCP-3-deficient mice are highly coupled (13, 54), display greater efficiency of ATP production, increases in the ATP-to-ADP ratio, impaired whole body fat combustion, and enhanced generation of reactive oxygen species (ROS) (3, 54).
Recently, it has been argued (12) that “mild uncoupling” associated with physiological levels of UCP-3 is more indicative of a role in antioxidant defense of the mitochondrial matrix. Even modest increases in UCP-3 content significantly lower the concentration of ROS by mitochondria (26). Finally, greater ROS production (7, 54), lower proton conductance (13, 54), and greater oxidative damage (7) were observed in mitochondria of UCP-3 knockout mice that also displayed unaltered body weights, respiration rates, or thermoregulatory properties (13, 54). The large increases in ROS production associated with exercise and electrical stimulation (2, 22, 25) might, therefore, indicate that greater UCP-3 expression is an important adaptive response leading to improved antioxidant defense.
In a preliminary presentation of data reported in the current study, we showed that UCP-3 protein content increased in proportion to mitochondrial biogenesis in response to 10 days of chronic low-frequency stimulation (CLFS) (39). Our findings were confirmed by Jones et al. (24) who reported that UCP-3 protein content increased in parallel with mitochondrial biogenesis induced by 1, 3, and 10 days of swimming exercise. However, UCP-3 mRNA content declined and was thus not strictly correlated to the changes observed at the protein level. Similarly, long-term single end-point exercise training studies reported UCP-3 mRNA expression remained unchanged (9) or even decreased (4, 45).
The purpose of this investigation was to extend the findings of previous exercise studies by investigating the comprehensive time course for contraction-induced changes in UCP-3 protein and mRNA expression, as related to changes in mitochondrial density and associated fast-to-slow fiber type transitions. CLFS was chosen over exercise training because it is the most suitable regimen for studying time-dependent changes of muscle in response to enhanced contractile activity (34, 35). We hypothesized that UCP-3 expression would increase in parallel with mitochondrial biogenesis and correspond to the emergence of highly oxidative type IIA fibers. CLFS was applied for 1 to 42 days; UCP-3 expression was examined at the mRNA level by RT-PCR and at the protein level by immunoblot analysis. The activities of reference enzymes of the citric acid cycle [citrate synthase (CS)], fatty acid oxidation (3-hydoxyacyl-CoA dehydrogenase), and glycolysis [glyceraldehydephosphate dehydrogenase (GAPDH)] were determined to establish adaptive changes in mitochondrial density and anaerobic metabolic capacity, respectively (15, 18, 19, 33, 40). Coordinate examination of the myosin heavy chain (MHC) content also made it possible to relate changes in UCP-3 expression with fiber type transitions.
Sixty-two adult male Wistar rats (Thomae, Biberach, Germany; Charles River Laboratories, Canada) were used for this study (380–425 g). Animals were cared for according to the previously established methods at the Animal Research Centers of the University of Konstanz and the University of Alberta and were in accordance with the guidelines of the Canadian Council for Animal Care. All animals were maintained under controlled environmental conditions (22°C and 12-h alternating light and dark cycles) and received standard chow and water ad libitum. Stimulation experiments were approved by the German government (Regierungspräsidium, Freiburg) and by the University of Alberta.
CLFS of the left tibialis anterior (TA) muscles was performed as previously described (50). Briefly, bipolar electrodes were implanted under general anesthesia lateral to the peroneal nerve of the left hindlimb. The electrodes were externalized at the animal's back and connected to a small, portable stimulator. To allow complete wound healing and recovery, CLFS (10 Hz, impulse width 380 μs, 10 h/day) was started 7 days later. In the case of the sham-operated group (n = 10), the unstimulated muscles were collected 8 days after electrode implantation. CLFS was applied for 1 day (n = 11), 5 ± 0.6 days (n = 10), 12 ± 0.6 days (n = 12), 24 ± 1.3 days (n = 11), or 42 days (n = 8). The 1-day-stimulated muscles were collected within 3–4 h after cessation of CLFS. Muscles from the remaining conditions were collected 14–18 h after completing CLFS; animals were killed and the TA muscles from both the stimulated (left) and contralateral control (right) legs were isolated, immediately frozen in melting isopentane (−159°C), and stored in liquid nitrogen. The time frame for muscle collection for the 1 day CLFS condition corresponded to a postactivity period in which the acute effects of exercise are known to persist (24). In contrast, muscles collected from the remaining time points corresponded to a period in which the effects of acute activity on UCP-3 gene expression are no longer detectable (24, 53) and thus represent long-term adaptive responses distinct from the effects of acute exercise. Unstimulated contralateral TA muscles served as internal controls for each animal. Normal control soleus muscles were also included (n = 10) to compare differences in UCP-3 expression between muscles possessing drastically different fiber type compositions.
Western Blot Analyses
UCP-3 protein content was quantified according to Ref. 38. Briefly, frozen muscle samples were powdered, homogenized in 10 volumes of a buffer containing 10% (vol/vol) glycerol, 1.5 ml 3% (wt/vol) SDS, 0.5% (vol/vol) β-mercaptoethanol, 0.25% (wt/vol) of protease inhibitor cocktail (Sigma-Aldrich, Oakville, ON, Canada) in 25 mM Tris-Cl (pH 6.8) and heated for 15 min at 96°C. After centrifugation (15 min, 13,000 g, 4°C), supernatant fractions were collected, and protein contents were determined (28). Samples were diluted with the homogenization buffer containing 0.1% (wt/vol) bromophenol blue and equal amounts of total protein loaded. Electrophoresis was performed on a 12% (wt/vol) polyacrylamide mini-gel (Protean II, Bio-Rad Laboratories, Mississauga, ON, Canada) for 105 min at 150 V. Separated proteins were transferred to a PVDF membrane (40 min, 200 mA), using a semi-dry technique (Hoefer, Multitemp III, Amersham Pharmacia Biotech, Montreal, Canada). Membranes were stained with Ponceau-S (Sigma-Aldrich) to confirm equal loading. They were subsequently destained, blocked in a buffer containing 2.5% (wt/vol) skim milk powder and 0.5% (wt/vol) protease-free and fat-free bovine serum albumin followed by incubation with the primary antibody. Parallel gels were also run in an identical fashion and stained with Amido-black (Sigma-Aldrich) to further confirm equal loading conditions, but were not used for Western blotting. Epitope-purified primary polyclonal rabbit anti-UCP-3 (IgG; Affinity Bioreagents, Golden, CO) (8) was applied overnight at 4°C in blocking solution (1:1,000). The primary antibody preparation is known to react strongly with human, mouse, and rat UCP-3. Membranes were washed and incubated with the secondary antibody, anti-rabbit IgG (Vector Laboratories, 1:2,000 in blocking solution) for 45 min and washed. Immunoreactivity was visualized with the ECL Plus chemiluminescence substrate (Amersham Pharmacia Biotech) and corresponded to a molecular weight of 30 kDa, as determined by comparison against standard molecular weight markers (Precision Plus Protein Standards, Bio-Rad Laboratories). All samples were evaluated in triplicate. Immunoblots were evaluated by integrating densitometry using GeneSnap and GeneTools (Chemigenius Gel Documention System, Syngene, UK). Equal loading was also confirmed by reprobing membranes with monoclonal anti-α-actinin (1:500) (clone EA-53; Sigma). Immunoreactivity was visualized as described above, after incubation with anti-mouse IgG (Vector Laboratories, 1:2,000 in blocking solution). As shown in Fig. 2A, α-actinin protein expression was not affected by CLFS.
UCP-3 and α-Actin mRNA Determination
UCP-3 mRNA expression was quantified according to Ref. 38. Briefly, muscles were powdered under liquid nitrogen, total RNA was extracted, and the concentration and purity were determined at A260 and A280. Reverse transcription was performed for 1 h at 37°C using 1 μl of total RNA, diluted to 1 μg/μl, and oligo (dT15) primers (Invitrogen, Life Technologies), and the Moloney murine leukemia virus DNA polymerase (Invitrogen Life Technologies). PCR was performed using the UCP-3 primers, 5′-AAG AGT GCA GAG CGT GCA GTA-3′ (forward) and 5′-ACA GAA ACC AGC TCC AAA G-3′ (reverse), which yielded a 537-bp amplicon corresponding to a unique nonhomologous transintronic region near the 5′-end of the UCP-3 mRNA. For α-actin, PCR was performed using the primers, 5′-GCGGTGCTGTCTCTCTATGC-3′ (forward) and 5′-CGGTGAGGATTTTCATCAGG-3′ (reverse) (Sigma Genosys), which yielded a single 127-bp amplicon corresponding to the translated region nt417 to nt589. All PCR products were sequenced to verify homology with the corresponding gene sequence (Molecular Biology Services Unit, University of Alberta). Reactions were carried out in an ICycler (BioRad Laboratories) using Taq DNA polymerase (Sigma-Aldrich). Linearity was confirmed for both the UCP-3 and α-actin assays between 15 and 35 cycles and between 0.2 and 2 μg of total RNA. Therefore, 28 cycles were used for each assay; conditions were similar to those previously reported (10, 32). The resulting UCP-3 and the corresponding α-actin amplicons for each sample were separated on 2% agarose gels in neighboring lanes for quantitative analyses. For ease of comparison, UCP-3 and α-actin amplicons generated from representative samples are shown here after resolution within the same lane (Fig. 3A). Amplicon size was confirmed by comparing to a 100-bp DNA Ladder (Invitrogen Life Technologies). All samples were analyzed in triplicate. Gels were stained with ethidium bromide and evaluated using GeneSnap and GeneTools (Syngene Chemigenius). Densitometric analyses of ethidium bromide-stained amplicons were carried out at levels below pixel saturation, ensuring accurate representation and quantification. Data are presented as the ratio of UCP-3 mRNA to α-actin mRNA content.
Enzyme Activity Measures
Frozen muscle was pulverized under liquid N2 and homogenized (1:10, wt/vol) in an ice-cold 100 mM KH2PO4-Na2HPO4 buffer (pH 7.2) containing 5 mM EDTA. The homogenate was sonicated five times for 30 s under intense cooling in an ice-salt mixture, stirred for 30 min on ice, and centrifuged for 15 min at 21,000 g at 4°C. The pellet was reextracted with the same volume of buffer, and the two supernatants were combined. The maximal activities of GAPDH (EC 184.108.40.206), CS (EC 220.127.116.11), and 3-hydroxyacyl-CoA dehydrogenase (HADH, EC 18.104.22.168) were determined at 30°C according to Bass et al. (1). In the case of HADH, however, the concentration of acetoacetyl-CoA in the assay mixture was reduced to 0.06 mM acetoacetyl-CoA. Samples were evaluated in duplicate.
Electrophoretic Analysis of MHC Isoforms
MHC isoforms were electrophoretically analyzed according to Hämäläinen and Pette (16) with slight modification. Briefly, muscle samples were homogenized in 7 vol of a buffer containing 100 mM Na4P2O7 (pH 8.5), 5 mM EGTA, 5 mM MgCl2, 0.3 M KCl, 10 mM DTT, and 5 mg/ml of a protease inhibitor cocktail (Complete, Roche Diagnostic). After being stirred for 30 min on ice, homogenates were centrifuged at 12,000 g for 5 min at 4°C, diluted 1:1 with glycerol, and stored at −20°C. Before gel loading, muscle extracts were diluted in Laemmli-lysis buffer and boiled for 5 min. MHC extracts were electrophoresed under denaturing conditions for 48 h (275 V and 10°C) on 7% (wt/vol) PAA separating gels containing glycerol. MHC isoforms were detected by silver staining (30) and evaluated by densitometry (Syngene Chemigenius).
Data are summarized as means ± SE and were analyzed by a two-way ANOVA with repeated measures between control and stimulated legs. Control soleus muscles were compared with TA muscles via one-way ANOVA. When a significant F ratio was observed, differences were located using the least significant difference post hoc test for planned comparisons. Differences in the patterns of MHC isoform expression were analyzed by a one-tailed Student's t-test based on establishment of a priori hypotheses that predicted the magnitude and direction of changes. Differences were considered significant at P < 0.05.
Activities of cytosolic and mitochondrial reference enzymes.
In agreement with previous observations (50), the cytosolic GAPDH activity progressively decreased throughout the stimulation period, reaching ∼50% of the control after 42 days, but was still approximately twofold greater than in soleus (Fig. 1A). One and five days of CLFS were associated with ∼30% reductions in basal GAPDH activity, although CLFS was without effect on GAPDH at those time points. In contrast, the mitochondrial CS and HADH activities progressively increased throughout the 42-day stimulation period (Fig. 1, B and C), being elevated by 2.47-fold and 2.15-fold (P < 0.00001), respectively, and exceeding maximal activities observed in the soleus.
UCP-3 protein levels.
Representative Western blot results for control and stimulated muscles are shown in Fig. 2A. Densitometric evaluation of the Western blots revealed 1.5 (P < 0.000001)-, 1.4 (P < 0.00005)-, and 1.3-fold (P < 0.0005) elevations in UCP-3 protein content over control in 12-, 24-, and 42-day-stimulated muscles, respectively (Fig. 2, A and B). UCP-3 protein content was not altered after 1 or 5 days of CLFS or by sham operations. The UCP-3 content of control soleus was ∼15% lower (P < 0.02) than observed in sham-operated or contralateral control TA muscles (Fig. 2, A and B).
Changes in UCP-3 content in relation to mitochondrial reference enzymes.
A comparison of the increases in UCP-3 protein content relative to the progressive and much larger increases in CS activity is shown in Fig. 2C. The increase in UCP-3 protein was proportional to increases in the activities of the CS only during the early phase of CLFS (i.e., up to 12 days of CLFS), whereas its relative mitochondrial content was clearly reduced with longer stimulation periods (i.e., from 24 to 42 days). UCP-3 protein expression relative to CS activity was reduced to 56 ± 2.8% (P < 0.00003) and 50 ± 3.7% (P < 0.00001) of control levels, after 24 and 42 days of CLFS, respectively (Fig. 2C). The same results were observed when changes in UCP-3 were expressed relative to the change in HADH activity. These changes were accompanied by continuous reductions in UCP-3 mRNA content that were lower than control TA and soleus after 42 days (see below).
Effects of CLFS on UCP-3 mRNA levels.
Sham-operated TA muscles did not differ from their respective contralateral controls (P > 0.69) or from contralateral control muscles of the 1, 12, 24, or 42 day groups (P > 0.57; Fig. 3, A and B). UCP-3 mRNA levels were 20% lower in stimulated TA muscles after 1 day compared with contralateral control muscles (P = 0.10). When this time point was analyzed individually using a Student's t-test, the 20% decrease was significant (P < 0.001). Thereafter, CLFS induced a progressive decrease in UCP-3 mRNA expression, reaching a level after 42 days that was 2.47-fold lower than the contralateral control TA muscles (P < 0.0007) and 1.88-fold lower than control soleus (P < 0.05). A sharp elevation in UCP-3 mRNA expression was noted in the contralateral control muscles of the 5 day group (P < 0.0005) (Fig. 3, A and B). Expression was 25% lower in normal soleus compared with unstimulated TA muscles from the sham-operated group and from the contralateral control muscles of the 1, 12, 24 day groups (P < 0.05).
Effect of CLFS on MHC isoforms.
In agreement with previous studies (21) of fast-twitch rat muscles, CLFS applied for up to 42 days induced transitions toward slower MHC isoforms (Fig. 4). A comparison of the 24- and 42-day-stimulated muscles did not reveal significant differences in the degree of fast-to-slow transformation, indicating that by 24 days most of the transformation had already occurred. Thus the data from these two time points were averaged. MHCI and MHCIIa were elevated 1.7 (P < 0.0005)- and 4.0-fold (P < 0.00001), respectively, at the expense of the fastest isoform, MHCIIb, which was reduced by 2.4-fold (P < 0.0002). The content of MHCIId/x remained unaltered (P > 0.77) and represented ∼30% of the total MHC present.
In the present study, CLFS was chosen to investigate effects of enhanced contractile activity on the expression of UCP-3 at the mRNA and protein levels. CLFS activates all motor units of the target muscle and thus reveals the full physiological adaptive potential of skeletal muscle fibers (34, 35). In this regard, CLFS is comparable to classical exercise training, resulting in qualitatively similar, but quantitatively greater, adaptive responses. CLFS also has the advantage over exercise training of being more reproducible, by ensuring each animal completes the same amount of contractile work and allowing simultaneous investigation of the unstimulated contralateral muscle as an internal control.
UCP-3 mRNA, UCP-3 Protein, and Mitochondrial Enzyme Activities
The short-term adaptive responses to CLFS (i.e., 1–12 days) are in agreement with a previous investigation by Jones et al. (24) who reported that UCP-3 protein expression increased in proportion to the increase in mitochondrial density in rat muscles during 10 days of swimming exercise. Unlike that study, however, we did not observe an acute increase in UCP-3 mRNA or protein after 1 day of CLFS, although our samples were collected within a time frame when increases in plasma FFA and catecholamines, resulting from a single bout of exercise, are known to induce greater UCP-3 (29, 47, 55). Thus the absence of changes in the stimulated leg after 1 day of CLFS suggests that the adaptive responses to CLFS did not encompass systemic regulation of UCP-3 protein expression by plasma FFA or catecholamines.
The transient elevation of UCP-3 mRNA in the control TA at day 5 was unexpected and probably resulted from increased weight bearing during the initial period of CLFS, where animals typically shift their weight to the contralateral control leg (36), which might elevate the rate of intramuscular triglyceride breakdown and thus account for the transient upregulation of UCP-3 mRNA (32, 42); nevertheless this did not result in increased translation of UCP-3 protein. When contrasted against the decline in UCP-3 mRNA in stimulated muscles, whereas absolute UCP-3 protein content was elevated within the first 12 days, it seems obvious that myogenic stimuli, associated with CLFS, represent important regulatory factors in posttranslational regulation of UCP-3 protein expression, perhaps in a manner similar to UCP-2 (31, 51).
An additional line of evidence supports regulation of UCP-3 expression by myogenic stimuli related to CLFS. Reichmann et al. (40) reported that 3-ketoacid-CoA transferase was elevated 30-fold by CLFS, although plasma concentrations of ketone bodies, which are normally the systemic stimulus for increased 3-ketoacid-CoA transferase expression, remained unaltered (D. Pette and H. Reichmann, unpublished). These observations point to the highly myogenic nature of the CLFS stimulus and argue against a major regulatory role by systemic factors. This is further emphasized when the size of the stimulated muscle group is considered in relation to the total muscle mass, especially compared with the muscle mass involved in exercise training. The CLFS model activates the TA, extensor digitorum longus, and the very small peroneus tertius muscles of the left hindlimb, which collectively have a mass of ∼1.2 g in a 400-g rat. Assuming that the total muscle mass corresponds to ∼40% of the total body weight (i.e., 160 g of muscle), the target muscles represent <1% of the total muscle mass, which seems to be insufficient to evoke systemic changes, such as increases in plasma free fatty acids, to affect changes in UCP-3 protein expression. Furthermore, the animals examined in the present investigation were free of stress and displayed normal patterns of feeding and weight gain.
Longitudinal end-point studies that investigated the effects of exercise training regimens on UCP-3 expression initially focused exclusively on changes in mRNA. Boss et al. (4) observed lower UCP-3 mRNA in the TA and soleus muscles of rats after 8 wk of treadmill running. In contrast, it was subsequently reported that 9 wk of voluntary wheel running did not alter UCP-3 mRNA levels in the same species (9) or in response to 6 wk of endurance training in humans (52). More recently it was reported that UCP-3 protein content was lower in the vastus muscles of endurance-trained cyclists than in untrained controls (46) and that immunofluorescent UCP-3 staining intensity and UCP-3 mRNA levels were lower in all fiber types after 6 wk of exercise training in humans (45). Collectively, these studies show unaltered or reduced levels of UCP-3 mRNA under conditions known to induce mitochondriogenesis.
According to our results, long-term CLFS (i.e., 24–42 days) resulted in absolute increases in UCP-3 protein content in whole muscle homogenates but obviously lower mitochondrial UCP-3 content (see results). These changes were accompanied by continuous reductions in UCP-3 mRNA. In this regard, the adaptive changes to UCP-3 protein and mRNA expression levels with long-term CLFS are qualitatively similar to recent exercise training studies in humans (e.g., Refs. 45, 46). However, the proportionally greater decline in UCP-3 mRNA compared with UCP-3 protein expression, relative to mitochondrial marker enzyme activities, points to a persistent element of posttranslational regulation. Thus differences in the findings of our investigation and previous exercise studies appear to be accounted for by the comparatively greater increase in mitochondrial content induced by the unique pattern of myogenic stimuli imposed by CLFS. Although changes in mitochondrial volume were not directly examined by previous exercise studies, the less extensive fast-to-slow fiber type transitions induced by exercise training protocols of similar duration (e.g., 45, 46) indicate that coordinate changes in mitochondrial density were also small.
Fiber Type Expression of UCP-3
CLFS not only induces increases in mitochondrial density but also leads to shifts in myofibrillar protein isoforms (34) in the slow direction with transitions from type IIB to type IIA fibers by 24 days. Thus coordinate changes in metabolic enzyme activities and MHC isoforms point to a growing population of highly “oxidative” fibers expressing MHCIIa after prolonged exposure to CLFS. These temporally related changes are relevant with regard to the limited data on the fiber type distribution of UCP-3 and the corresponding adaptive changes associated with muscle training. Russell et al. (45) reported that UCP-3 protein content in human vastus lateralis muscle was greatest within the IID/X fibers, intermediate in IIA fibers, and lowest in the type I fibers and that 42 days of exercise training resulted in lower content in all fiber types. In the present study, however, the greatest increase in UCP-3 protein (i.e., 12 and 24 days) coincided with the emergence of a population of highly oxidative IIA fibers.
Biochemical Implications of Changes in UCP-3
Although the present study did not examine the functional impact of absolute or relative changes in UCP-3 protein expression, it is interesting to note that whereas the reductions in UCP-3 protein relative to increases in mitochondrial density after long-term CLFS are consistent with a regulatory role in thermogenesis, with reduced uncoupling, and the potential for lower heat generation (8), the short-term adaptive changes are not. If thermogenic regulation was the primary function, the relative UCP-3 protein content per mitochondrial unit should have been most severely reduced during the early adaptive phase of CLFS (i.e., from 1 to 12 days). The fact that this did not occur and that mitochondria were generated with a normal complement of UCP-3 suggests an alternative role.
If, as indicated in recent studies, the primary role of UCP-3 relates to regulation of fatty acid metabolism with thermodynamic consequences (17, 48), then the early genesis of mitochondria with a normal complement of UCP-3 suggests this mitochondrial carrier retains an important regulatory function during the early phase of muscle training. It has been speculated that UCP-3 functions to defend mitochondria against toxic levels of fatty acids, and/or to sustain free concentrations of CoASH and NAD+. The latter function, however, seems unlikely. Assuming maximal flux through the Krebs cycle and β-oxidation (20), it is possible to calculate that the concentration of mitochondrial CoASH would remain above saturating levels for the flux-generating enzymes within those pathways. Furthermore, numerous studies have shown that in normoxia, mitochondria actually become more oxidized and free NAD+ concentration increases during contractile activity ranging in intensity from submaximal to maximal (e.g., 14, 23, 37). The decline in UCP-3 expression observed in human (45, 46) and rodent (4) studies further argues against this possibility. As the capacity to oxidize fatty acids increases with training, one would predict an increase per mitochondrial unit rather than a decrease.
The adaptive pattern of UCP-3 expression observed in our study supports a role for UCP-3 in antioxidant defense. The major source of ROS during muscle training, independent of muscle injury, is O2−· that forms between respiratory complexes I and III by the reaction of molecular oxygen with the transferring electron (11, 22). Although much of the O2−· is converted to H2O2 by mitochondrial superoxide dismutase (mSOD), these two chemical species may combine to form the highly reactive hydroxyl free radical (HO·) (12). In addition, nitric oxide may react with O2−· to form the strong oxidant peroxynitrite during CLFS (25). The deleterious consequences encompass peroxidation of DNA and lipids; they also include oxidation of SH groups and nitrosylation of tryptophan, tyrosine, and cytosine residues. The function of UCP-3 has been described as extruding hydrophilic fatty acid peroxides that form on the inner membrane to the outer leaflet to prevent further oxidative damage to mitochondrial DNA and proteins (12). Once in the outer leaflet, fatty acid peroxides are reduced by antioxidant enzymes, protonated, and may return to the inner membrane by a “flip-flop” mechanism. Once oriented toward the matrix, fatty acids are deprotonated, thus accounting for uncoupling properties associated with UCP-3 (12).
Exercise and CLFS are known to generate substantial amounts of ROS causing oxidative damage (22, 25). The expression of the antioxidant enzymes mSOD, catalase, and glutathione peroxidase has been shown to be elevated in response to exercise training and to display fiber type-specific patterns of expression, being highest in type I fibers and successively less in type IIA and type IIB fibers, respectively (for review see Ref. 22). Coincidentally, this corresponds to the inverse pattern of UCP-3 expression within the various fiber types (46) and in response to muscle training (45, 46). These observations suggest that the increase in mitochondrial content alone provides long-term protection against ROS. The generation of mitochondria with a lower complement of UCP-3 after 24 and 42 days of CLFS probably occurred secondary to such a reduction in ROS production. However, the production of mitochondria with a normal complement of UCP-3 from after 5- and 12-days of CLFS indicates that the concentration of ROS probably remained elevated during the early adaptive phase. Additionally, mitochondria that formed early in response to CLFS were probably concentrated within the transforming IID(X) and IIB fiber populations, which were presumably the greatest source of ROS production.
In summary, absolute UCP-3 protein content of rat TA muscle peaked after 12 days of CLFS and was sustained at the higher absolute level throughout the 42-day study period. However, the relative mitochondrial concentration of UCP-3 protein was lower after prolonged CLFS and corresponded to the observed decrease in UCP-3 mRNA expression levels. The major adaptive stimulus associated with overall increases in UCP-3 protein expression in response to CLFS seems to be myogenic in nature and to encompass posttranslational mechanisms. As judged from parallel changes in the pattern of MHC isoforms and comparative measures of UCP-3 protein and mRNA levels in soleus muscle, these changes coincided with the process of fiber type transition from fast type IIB and IID(X) to slower and highly oxidative type IIA fibers.
This study was funded by research grants from the Natural Sciences and Engineering Council of Canada, the Alberta Heritage Foundation for Medical Research (AHFMR), Alberta Agricultural Research Institute (AARI), and the Deutsche Forschungsgemeinschaft Du 260/2–2 and Fonds der Chemischen Industrie. M. J. Jendral was the recipient of an NSERC Scholarship. C. T. Putman is a Heritage Medical Scholar of AHFMR.
The authors thank Elmi Leisner and Karen Martinuk for completing animal surgeries.
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.
- Copyright © 2004 the American Physiological Society