Skeletal muscle is a heterogeneous tissue. To further elucidate this heterogeneity, we probed relationships between myosin heavy chain (MHC) isoform composition and abundance of GLUT4 and four other proteins that are established or putative GLUT4 regulators [Akt substrate of 160 kDa (AS160), Tre-2/Bub2/Cdc 16-domain member 1 (TBC1D1), Tethering protein containing an UBX-domain for GLUT4 (TUG), and RuvB-like protein two (RUVBL2)] in 12 skeletal muscles or muscle regions from Wistar rats [adductor longus, extensor digitorum longus, epitrochlearis, gastrocnemius (mixed, red, and white), plantaris, soleus, tibialis anterior (red and white), tensor fasciae latae, and white vastus lateralis]. Key results were 1) significant differences found among the muscles (range of muscle expression values) for GLUT4 (2.5-fold), TUG (1.7-fold), RUVBL2 (2.0-fold), and TBC1D1 (2.7-fold), but not AS160; 2) significant positive correlations for pairs of proteins: GLUT4 vs. TUG (R = 0.699), GLUT4 vs. RUVBL2 (R = 0.613), TUG vs. RUVBL2 (R = 0.564), AS160 vs. TBC1D1 (R = 0.293), and AS160 vs. TUG (R = 0.246); 3) significant positive correlations for %MHC-I: GLUT4 (R = 0.460), TUG (R = 0.538), and RUVBL2 (R = 0.511); 4) significant positive correlations for %MHC-IIa: GLUT4 (R = 0.293) and RUVBL2 (R = 0.204); 5) significant negative correlations for %MHC-IIb vs. GLUT4 (R = −0.642), TUG (R = −0.626), and RUVBL2 (R = −0.692); and 6) neither AS160 nor TBC1D1 significantly correlated with MHC isoforms. In 12 rat muscles, GLUT4 abundance tracked with TUG and RUVBL2 and correlated with MHC isoform expression, but was unrelated to AS160 or TBC1D1. Our working hypothesis is that some of the mechanisms that regulate GLUT4 abundance in rat skeletal muscle also influence TUG and RUVBL2 abundance.
- glucose transport
- fiber type
- physical activity
skeletal muscle is a heterogeneous tissue, and different muscles in the same individual can vary greatly with regard to metabolic characteristics, including the capacity for insulin-mediated glucose uptake. Henriksen et al. (19) found that the GLUT4 protein abundance of four different isolated rat skeletal muscles varied by ∼4-fold, with the following rank-order: flexor digitorum brevis (FDB) > soleus > extensor digitorum longus (EDL) > epitrochlearis. On the basis of previously published data for the fiber type compositions (based on myosin ATPase staining) of the FDB (6), epitrochlearis (28), soleus, and EDL (2), it was evident that the two muscles that were known to be largely composed of slow-oxidative (SO) or fast-oxidative-glycolytic (FOG) fibers (soleus and FDB, respectively) were characterized by greater GLUT4 abundance compared with the other two muscles that were predominantly composed of fast-glycolytic (FG) fibers (epitrochlearis and EDL). The most comprehensive study to measure both GLUT4 abundance and fiber type composition in multiple rat skeletal muscles was performed by Megeney et al. (27) who evaluated six different muscles or muscle regions and correlated percent fiber type composition of the muscles (determined by ATPase staining in the same six muscles from other rats) with GLUT4 abundance. They found a significant positive correlation with percent of oxidative fibers (SO + FOG) and a significant negative correlation with the percent of FG fibers. We have extended the knowledge from earlier studies by measuring both GLUT4 and fiber type (based on myosin heavy chain isoform, MHC-I, MHC-IIa, MHC-IIb, and MHC-IIx expression) in 12 different rat muscles or regions of muscles.
Although GLUT4 is the ultimate mediator of muscle glucose transport capacity, a number of other proteins participate in insulin's regulation of GLUT4 function. Akt substrate of 160 kDa (also known as TBC1D4) and TBC1D1 are two paralog Rab GTPase (GAP) proteins that were recently recognized as key signaling proteins that can control GLUT4 trafficking and glucose transport (8, 10–11, 33). Taylor et al. (40) reported that the abundance of each of these proteins varied greatly among three mouse skeletal muscles. AS160 protein abundance was ∼10-fold greater in the soleus compared with the tibialis anterior (TA) and the EDL. TBC1D1 abundance was much greater for the TA compared with the EDL (∼3-fold) and the soleus (∼10-fold). They also noted that TBC1D1 expression appeared to track with the MHC-IIx content of these muscles, but they did not quantitatively analyze the relationship of TBC1D1 or AS160 with each other or with MHC isoform expression. The relationship of AS160 and TBC1D1 abundance to each other or to fiber type of rat skeletal muscle has also not been reported.
Xie et al. (43) recently identified RuvB-like protein two (RUVBL2) as a protein that is physically associated with AS160 in 3T3-L1 adipocytes. Genetic depletion of RUVBL2 in 3T3-L1 cells resulted in a decrease in both phosphorylated AS160 and insulin-stimulated glucose uptake. In addition, the epididymal fat pads of obese diabetic KKAy mice compared with normal controls had greatly reduced RUVBL2 expression. The relationship between RUVBL2 abundance and muscle MHC isoform composition has not been reported.
Bogan et al. (4) identified the Tethering protein containing a UBX-domain for GLUT4 (TUG) as a putative GLUT4 tethering protein that functions as part of the system that retains GLUT4 intracellularly in 3T3-L1 cells in the absence of insulin. This model has been further supported by subsequent studies in 3T3-L1 adipocytes (44, 45). Under basal conditions (without insulin), TUG is associated with GLUT4 in either L6 myotubes or skeletal muscle of transgenic mice with muscle-specific overexpression of GLUT4myc (34). Insulin causes GLUT4 to rapidly disassociate from TUG in both L6 cells and in skeletal muscle from GLUT4myc transgenic mice. Although GLUT4 and TUG are known to be binding partners, their relative expression levels in skeletal muscles with diverse MHC composition are unknown.
The first major goal of this study was to extend knowledge about the relationship between MHC isoform composition and the abundance of five proteins (GLUT4, AS160, TBC1D1, RUVBL2 and TUG) that are either established or putative regulators of glucose transport in rat skeletal muscle. The second major goal was to determine if the abundance of any of these proteins were significantly associated with each other. Notable aspects of the experimental design were 1) the inclusion of a large number of skeletal muscles (12 muscles or regions of muscles) with diverse fiber type compositions; 2) the assessment of relative MHC isoform levels and abundance of GLUT4 and four other proteins (TUG, RUVBL2, TBC1D1, and AS160) for which very little was known with regard to relative expression by rat skeletal muscles; and 3) immunoblotting for each of the proteins and the determination of relative MHC isoform levels were all performed using the same muscles from the same rats. We hypothesized that 1) the abundance of each of the five proteins would vary among the 12 different muscles; 2) %MHC-I and %MHC-IIa would be positively correlated with GLUT4 and TUG; 3) %MHC-IIb would be negatively correlated with GLUT4 and TUG; 4) GLUT4 would be positively correlated with TUG; 5) %MHC-IIx would be positively correlated with TBC1D1 and negatively correlated with AS160 and RUVBL2; 6) AS160 would be positively correlated with RUVBL2; and 7) TBC1D1 would be negatively correlated with AS160 and RUVBL2.
The reagents and apparatus for SDS-PAGE and immunoblotting were purchased from Bio-Rad (Hercules, CA). Bicinchoninic acid protein assay reagent (no. 23227) and West Dura Extended Duration Substrate were from Pierce Biotechnology (Rockford, IL). Anti-AS160 (no. 07–741) was from Millipore (Billerica, MA). Anti-RUVBL2 (no. ab36569) was from Abcam (Cambridge, MA). Anti-TBC1D1 was provided by Dr. Makoto Kanzaki (Tohoku University). Anti-TUG was previously described (4). Anti-GLUT4 was provided by Dr. Samuel Cushman (NIH, Bethesda, MD).
Procedures for animal care were approved by the University of Michigan Committee on Use and Care of Animals. Male Wistar rats (150–220 g; Harlan, Indianapolis, IN) were provided with rodent chow (Lab Diet; PMI Nutritional International, Brentwood, MO) and water ad libitum. Rats were anesthetized with an intraperitoneal injection of pentobarbital sodium (50 mg/kg wt). While rats were under deep anesthesia, the following muscles or portions of muscles were dissected out and freeze-clamped: adductor longus (AL), extensor digitorum longus (EDL), epitrochlearis (EPI), mixed gastrocnemius (GASM), red gastrocnemius (GASR), white gastrocnemius (GASW), plantaris (PLAN), soleus (SOL), red tibialis anterior (TAR), white tibialis anterior (TAW), tensor fasciae latae (TFL), and white vastus lateralis (VLW). The deep red (GAS and TA) and superficial white (GAS, TA, and VL) regions of the relevant muscles were identified based on visual inspection and dissected out. The GASM was dissected from the muscle's medial head.
Frozen muscles were weighed, transferred to prechilled glass tissue grinding tubes (Kontes, Vineland, NJ), and homogenized in ice-cold lysis buffer (20 mM Tris·HCl, 150 mM NaCl, 1% Triton X-100, 1 mM Na3VO4, 1 mM EDTA, 1 mM EGTA, 2.5 mM NaPP, 1 mM β-glycerophosphate, 1 μg/ml leupeptin, 1 mM PMSF at 1 ml/muscle) using a glass pestle attached to a motorized homogenizer (Caframo, Wiarton, ON). Homogenates were then rotated at 4°C for 1 h and an aliquot of the homogenate (whole homogenate) was taken for myosin heavy chain (MHC) analysis prior to being centrifuged (1,000 g for 10 min at 4°C). Portions of the supernatant and the whole homogenate were used to determine protein concentration according to the manufacturer's protocol (Pierce Biotechnology no. 23227). The remaining supernatant was stored at −80°C until further analyzed. A portion of the whole homogenate was immediately used to determine MHC isoform composition as described below.
Relative abundance of MHC isoforms.
Laemmli sample buffer was added to 5 μg of the whole homogenate prior to heating for 10 min at 90°C. Samples where then run at 45 V for 24 h at 4°C on an 8% acrylamide-bis (50:1), 30% glycerol gel as described by Talmadge and Roy (39). Gels were then trimmed and stained with Coomassie Blue for 1 h while gently rotating, followed by destaining for 3–4 h in 20% methanol and 10% acetic acid solution. MHC bands were quantified using densitometry.
Homogenized muscle lysates were boiled for 5–10 min in sodium dodecyl sulfate loading buffer then separated via PAGE and electrophoretically transferred to nitrocellulose membranes. Samples were then rinsed with Tris-buffered saline plus Tween (TBST; 140 mM NaCl, 20 mM Tris base, pH 7.6, and 0.1% Tween), blocked with 5% nonfat dry milk in TBST for 2 h at room temperature, washed three times for 5 min at room temperature, and treated with the appropriate primary antibody (1:1,000 in TBST + 5% BSA) overnight at 4°C. Blots were then washed six times for 5 min with TBST, incubated with the secondary antibody, goat anti-rabbit IgG horse horseradish peroxidase conjugate (1:20,000 in TBST + 5% milk) for 1 h at room temperature. They were washed again six times for 5 min with TBST then four times for 5 min TBS and subjected to enhanced chemiluminescence (West Dura Extended Duration Substrate; #34075; Pierce). The chemiluminescence of protein bands on nitrocellulose membranes was quantified using electrically cooled CCD camera technology (Alpha Innotech, San Leandro, CA). The individual values for the samples were normalized to the mean value for all of the samples on the blot.
The statistical analyses were performed using Sigma Plot (San Rafael, CA) version 11.0. Data are expressed as means ± SE. P ≤ 0.05 was considered to be statistically significant. The data used for comparison of protein abundance among the 12 different muscles or muscle regions were analyzed by one-way ANOVA, and the Bonferroni post hoc test was used to identify the source of significant variance for data. If data failed the normality test, the Kruskal-Wallis one-way ANOVA on ranks and the Tukey post hoc test were used. The nonparametric Spearman rank order correlation (for individual data rather than on mean values for each of the 12 muscles) was used to assess the relationships between abundance of each pair of the proteins studied (e.g., between GLUT4 vs. TUG) and between abundance of each protein and percent of each MHC isoform (MHC-I, MHC-IIa, MHC-IIb, and MHC-IIx).
Relative abundance of myosin heavy chain isoforms.
Relative protein abundance.
In the 12 muscles or muscle regions studied the relative abundance of GLUT4 differed as follows: SOL was significantly greater (P < 0.05) than the EDL, EPI, GASM, GASW, TAW, TFL, and VLW; AL was significantly greater (P < 0.05) than the EPI, GASM, GASW, TAW, TFL, and VLW; the TAR and GASR were significantly greater (P < 0.05) than the EPI, GASW, TAW, and VLW; PLAN was significantly greater (P < 0.05) than the EPI, GASW, and TAW; EDL was significantly greater (P < 0.05) than the EPI and VLW (Fig. 2). In the 12 muscles studied, the GLUT4 abundance had a range of 2.5-fold (the VLW had the lowest and the SOL had the highest values).
TUG protein abundance in the AL and SOL was significantly greater (P < 0.05) than all of the other muscles studied (Fig. 3). The range of TUG abundance was 1.7-fold (the EPI had the lowest and the SOL had the highest values).
The relative abundance of RUVBL2 in the 12 muscles or muscle regions studied differed as follows: SOL was significantly greater (P < 0.05) than the EDL, EPI, GASW, TAW, and VLW; AL was significantly greater (P < 0.05) than the EDL, GASW, TAW, and VLW; both PLAN and TFL were significantly greater (P < 0.05) than TAW and VLW; GASR was significantly greater (P < 0.05) than the VLW (Fig. 4). The range of RUVBL2 abundance was 2.0-fold (the VLW had the lowest and the SOL had the highest values).
AS160 protein abundance did not differ significantly among any of the muscles or muscle regions studied (Fig. 5).
TBC1D1 protein abundance in both the GASW and EDL was significantly greater (P < 0.05) than both the TFL and EPI (Fig. 6). The GASR and GASM both had significantly greater (P < 0.05) TBC1D1 abundance compared with the TFL. The range for TBC1D1 abundance was 2.7-fold (the TFL had the lowest and GASR had the highest values).
As expected, the %MHC-I values were high for the AL (90%) and SOL (88%). These muscles compared with the other muscle studied were also characterized as having higher values for GLUT4, TUG, or RUVBL2. In an earlier study, Megeney et al. (27) evaluated six rat skeletal muscles (including the SOL, but not the AL) for correlations between GLUT4 and fiber type. For some comparisons, they identified the SOL as an outlier and excluded the SOL data from the correlation analysis. In the current study, correlations were performed both with and without exclusion of the AL and SOL data to make it possible to assess the data both with and without the influence of these muscles that have exceptional MHC profiles.
The following pairs of proteins were significantly and positively correlated with each other either with all of the data or with the AL and SOL data excluded: GLUT4 vs. TUG (Fig. 7A), GLUT4 vs. RUVBL2 (Fig. 7B), TUG vs. RUVBL2 (Fig. 7C), AS160 vs. TBC1D1 (Fig. 7D,) and AS160 vs. TUG (Fig. 7E).
There were significant positive correlations for %MHC-I and the abundance of the following proteins with either all of the data or excluding the AL and SOL data: TUG (Fig. 8B) and RUVBL2 (Fig. 8C). %MHC-I vs. GLUT4 was significantly positively correlated only when all of the data were in the analysis (Fig. 8A).
There were significant positive relationships between %MHC-IIa and the abundance of the following proteins either with all of the data or excluding the AL and SOL data: GLUT4 (Fig. 9A) and RUVBL2 (Fig. 9C). TUG was significantly positively correlated with %MHC-IIa only when the AL and SOL data were excluded (Fig. 9B).
There were significant negative correlations for %MHC-IIb and the abundance of the following proteins either using either all of the data or excluding the AL and SOL data: GLUT4 (Fig. 10A), TUG (Fig. 10B), and RUVBL2 (Fig. 10C).
There was a significant negative correlation between %MHC-IIx and the abundance of TUG using all of the data; however, when the AL and SOL data were excluded, the correlation between %MHC-IIx and TUG was significant, but positive (Fig. 11B). In addition, when the AL and SOL data were excluded there were significant positive correlations for %MHC-IIx with either GLUT4 (Fig. 11A) or RUVBL2 (Fig. 11C).
Neither AS160 nor TBC1D1 was significantly correlated with the relative abundance of any of the MHC isoforms based on analysis using either all of the data or excluding the AL and SOL data (Figs. 12⇓⇓–15).
This study provides novel information about five proteins (GLUT4, TUG, RUVBL2, AS160, and TBC1D1) that are either established or putative regulators of glucose transport. The results supported some, but not all, of the hypotheses about the relationships of these proteins with each other and with relative MHC isoform expression. The hypotheses that related to GLUT4 and TUG were largely supported: the abundance of both GLUT4 and TUG varied significantly among the muscles that were studied; the %MHC-I abundance was positively correlated to either GLUT4 or TUG; the %MHC-IIb abundance was negatively correlated with either GLUT4 or TUG; and GLUT4 and TUG were positively correlated with each other. As was hypothesized, several of the muscles were significantly different with regard to TBC1D1 abundance. However, none of the other hypotheses related to TBC1D1 or AS160 were supported by the results: AS160 and TBC1D1 were positively rather than negatively correlated with each other, neither TBC1D1 nor AS160 was significantly correlated with the relative abundance of any MHC isoform, and no significant differences were detected for AS160 abundance among the muscles studied. The results also revealed several unanticipated insights about RUVBL2, including significant positive correlations for RUVBL2 with GLUT4, TUG, %MHC-I, and %MHC-IIa, and a significant negative correlation for RUVBL2 with %MHC-IIb. However, RUVBL2 was not significantly correlated with either AS160 or TBC1D1.
As expected, the 12 different muscles or muscle regions that were studied were diverse with regard to the relative abundance of MHC isoforms. The results for relative MHC isoform abundance were in good agreement with previously published values for the AL, SOL (20), GASM (39), GASR, GASW, EDL (35), EPI (1), PLAN (29), and VLW (26). The current study was apparently the first to report MHC isoform values for the TFL, TAR, or TAW from rats.
Previous studies that have assessed the relationship between fiber type composition of skeletal muscle and GLUT4 abundance have relied on histological assessment of myosin ATPase for fiber typing. Measurement of MHC isoform by SDS-PAGE provides important advantages for quantitative analysis of skeletal muscle compared with fiber type composition based on myosin ATPase activity (30). The rank order for relative values of GLUT4 by the different muscles or muscle regions were similar to the results from several previous publications that evaluated GLUT4 abundance for three or more of the rat skeletal muscles assessed in the current study, including the SOL, EPI, and EDL (19); GASR, GASW, PLAN, EDL, and SOL (5); GASR, GASW, PLAN, and SOL (41); TAR, TAW, and EDL (21); and SOL, PLAN, EDL, GASR, and GASW (27). Megeney et al. (27) assessed GLUT4 abundance and fiber type composition (using myosin ATPase staining) in rat skeletal muscles. They found a significant positive correlation for %SO + FOG fibers vs. GLUT4 and a significant negative correlation for FG fibers vs. GLUT4. In the current study, there were significant positive correlations for GLUT4 with either %MHC-I or %MHC-IIa vs. GLUT4 and a significant negative correlation for %MHC-IIb vs. GLUT4. The ∼2.5-fold range from the lowest to highest values in the current study compares to ranges from ∼2- to ∼4-fold reported in earlier studies that evaluated several rat skeletal muscles with a range of fiber type compositions (5, 19, 21, 27, 41).
A key outcome of this study was that the relative expression of three proteins (GLUT4, TUG, and RUVBL2) clustered together based on comparisons of the muscles that were studied. These results raise the possibility that the expression of the three proteins may be modulated, in part, by shared mechanisms. Skeletal muscle MHC isoform expression is influenced by conditions that alter neuromuscular activity, including spinal cord injury, hindlimb unloading, chronic bed rest, spaceflight, compensatory overloading, chronic electrical stimulation, and exercise training (38). Although the MHC isoform expressed by skeletal muscle is unlikely to directly determine GLUT4 expression levels, the abundance of GLUT4 in skeletal muscle has also been reported to be responsive to many of these same interventions. For example, chronic electrical stimulation of skeletal muscle (21, 23) or endurance exercise training (5, 32, 36) can result in an increase in GLUT4 abundance, and denervation can reduce muscle GLUT4 abundance (27). To date, the effects of increased or decreased neuromuscular activity on muscle TUG or RUVBL2 abundance have not been reported, but it seems possible that their expression might also be altered in a manner that is similar to activity-related shifts in GLUT4 abundance.
The significant correlation between TUG and GLUT4 abundance in skeletal muscles was a novel, but not unexpected, finding. Because TUG and GLUT4 can physically associate with each other, and because TUG-GLUT4 binding appears to play a role in the subcellular localization of GLUT4, altering the abundance of one of the binding partners may potentially influence the abundance of the other. Consistent with a role for TUG in controlling GLUT4 protein levels, genetic depletion of TUG by siRNA in 3T3-L1 cells resulted in a marked decrease in GLUT4 protein abundance that appeared to be attributable to greater GLUT4 protein degradation (45). However, the possibility that altered GLUT4 abundance might influence TUG abundance has not been assessed.
RUVBL2 was evaluated because it was recently reported to be physically associated with AS160 in 3T3-L1 adipocytes and because genetic depletion of RUVBL2 in 3T3-L1 cells induced decrements in both phosphorylated AS160 and insulin-stimulated glucose uptake without altering the abundance of Akt or AS160 protein (43). However, RUVBL2 has received much more scrutiny for its roles in the regulation of DNA structure and function. It is a member of the AAA+ (ATPase associated with diverse cellular activities) family of DNA helicases and has been implicated in the response to DNA double-strand breaks and the regulation of gene expression (16). Although RUVBL2 has been shown to have important functions in the nucleus, it is localized in both the cytosol and the nucleus (43). The current study, apparently the first to assess RUVBL2 protein abundance in skeletal muscle, revealed significant positive correlations for RUVBL2 with %MHC-I or MHC-IIa and a significant negative correlation for RUVBL2 with %MHC-IIb. The causes and functional consequences of these differences in RUVBL2 expression levels are not currently known.
Taylor et al. (40) found a 10-fold range in the relative abundance of TBC1D1 in skeletal muscles from mice, with the rank order of TA > EDL > SOL. TBC1D1 abundance also varied among the rat skeletal muscles evaluated in the current study, with 2.7-fold greater values for GASR (highest) vs. TFL (lowest) muscle. TBC1D1 abundance for the rat EDL was very similar to the GASR. The TBC1D1 values for the SOL, TAR, and TAW were intermediate, but not significantly different than the EDL (Fig. 6). Taylor et al. (40) also measured MHC isoform abundance and noted that TBC1D1 appeared to track with the type MHC-IIx content of the TA, EDL, and SOL muscles, although quantitative analysis for this relationship was not reported. However, there was no evidence that TBC1D1 abundance in rat skeletal muscle was related with relative expression of MHC-IIx or with any of the other MHC isoforms. An et al. (1a) reported that increasing TBC1D1 expression of mouse TA muscle by ∼7-fold did not alter glucose uptake (basal, insulin stimulated, or contraction stimulated) or abundance of AS160 or GLUT4. The causes and functional consequences of the different levels of TBC1D1 protein abundance in rat skeletal muscles remain to be determined.
AS160 and TBC1D1 have substantial structural similarities and appear to have overlapping functional properties, including the regulation of GLUT4 trafficking in skeletal muscle (8). There were no significant differences for AS160 abundance among the rat skeletal muscles in the current study. Consistent with these results, Gupte et al. (17) found no difference between the EPI and SOL of rats for AS160 abundance. However, the current results are in contrast to substantial muscle-specific differences in the mouse, in which AS160 protein abundance was ∼10-fold greater for the SOL compared with TA or EDL muscles (40). Increasing AS160 abundance in mouse TA muscle by ∼8-fold (i.e., to levels nearly as great as the endogenously high values found in mouse SOL) did not alter basal or insulin-stimulated glucose uptake or expression of GLUT4 protein (24). However, contraction-stimulated glucose uptake was reduced ∼24% with AS160 overexpression compared with empty vector controls (24). Endurance-trained rats fed a high-fat diet compared with sedentary controls eating the same diet had a small (15%) but significant decrease in muscle AS160 abundance concomitant with a moderately large (50%) increase in muscle GLUT4 protein content (25). Endurance training by lean or obese humans also resulted in elevated muscle GLUT4 protein levels (∼20–50%), but in contrast to the results for rats, there were also small (∼20–30%) training-induced increases in AS160 abundance (14, 42). Either humans with Type 2 diabetes compared with nondiabetic controls (22) or insulin-resistant obese Zucker rats compared with lean controls (3) have reduced insulin-mediated phosphorylation of AS160 without differences in AS160 protein abundance in skeletal muscle. Transfection of skeletal muscle of either lean or obese rats with PGC1α induced a significant increase in GLUT4 abundance without concurrent changes in AS160 content (3). There are various mechanisms for regulating AS160 function without large changes in AS160 protein abundance. For example, AS160 is acutely regulated by insulin-induced phosphorylation of key sites and subcellular localization of AS160, and there is also evidence that AS160 function is modulated by its association with other proteins, e.g., 14–3-3 (31) and possibly RUVBL2 (43). Clarification of how muscles differ with regard to these regulatory processes will be needed to fully appreciate AS160's role in determining muscle-related differences for glucose transport with insulin and/or exercise.
The current study evaluated GLUT4 and related proteins in more muscles than previous publications. However, we opted not to include the flexor digitorum brevis (FDB). In our experience (7, 9), other muscles offer advantages compared with the FDB for studying GLUT4, insulin signaling proteins, and glucose uptake. Advantages of the epitrochlearis or soleus vs. FDB include a larger data base of publications for comparing the results, the presence of less connective tissue in the muscle, and relatively greater fold increases in insulin-mediated glucose uptake.
It is important to interpret the correlations between protein abundance and relative levels of MHC isoforms with care. Obviously, correlations do not prove causality. Furthermore, it is inappropriate to assume that expression of a given protein is uniform in all of the fibers expressing a certain MHC isoform. For example, in pooled single fibers expressing MHC-I collected from sedentary humans, GLUT4 abundance was greater for the fibers taken from the vastus lateralis compared with fibers collected from either the soleus or triceps brachii (12). In addition, hybrid muscle fibers that express more than one MHC isoform can be relatively common in rat skeletal muscles (37). The correlation for a given protein's abundance with the percent of MHC isoforms with relatively lower expression (MHC-IIa and MHC-IIx) can be highly influenced by the levels of the more abundant isoforms (MHC-I and MHC-IIb) in the group of muscles being studied. For example, the correlation between MHC-IIx and TUG was significantly negative when all 12 muscles were included in the analysis, but the correlation was significantly positive when the analysis was performed after excluding the muscles with extremely high levels of MHC-I (SOL and AL, which have 0 or 1% MHC-IIx). Nonetheless, careful examination of the correlations between the abundance of a given protein and the relative levels of the MHC isoforms can provide useful insights.
In conclusion, an important result of this study was the identification of a cluster of three proteins (GLUT4, TUG, and RUVBL2) that tracked together in the skeletal muscles that were evaluated. GLUT4, TUG, and RUVBL2 were each also positively correlated with %MHC-I and inversely correlated with the %MHC-IIb levels. The paralog proteins AS160 and TBC1D1 are 47% identical, share several important functional domains, and are implicated as regulators of glucose transport (40). AS160 and TBC1D1 were positively correlated with each other, but neither of these proteins was significantly correlated with the relative levels of any of the MHC isoforms. Earlier research has clearly documented that altered neuromuscular activation of skeletal muscle can markedly alter GLUT4 protein levels (5, 13, 21, 23, 36). Training effects on TUG and RUVBL2 abundance have apparently not been assessed to date. Given that SOL compared with EDL muscles from sedentary male Wistar rats have much greater levels of motor unit activation (18), it is notable that the SOL compared with the EDL had significantly greater values for GLUT4, TUG, and RUVBL2. Our working hypothesis is that TUG and RUVBL2 protein content in rat skeletal muscle are regulated by mechanisms that, at least in part, are similar to those that control GLUT4 protein abundance and that each is influenced by the level of neuromuscular activation.
This research was supported by grants from the National Institutes of Health (DK-071771 and AG-10026 to G. D. Cartee and DK-075772 to J. S. Bogan).
No conflicts of interest, financial or otherwise, are declared by the authors.
The authors thank Dr. Samuel Cushman for providing the GLUT4 antibody.
- Copyright © 2011 the American Physiological Society