Skeletal Muscle-specific G Protein-coupled Receptor Kinase 2 Ablation Alters Isolated Skeletal Muscle Mechanics and Enhances Clenbuterol-stimulated Hypertrophy*

GRK2, a G protein-coupled receptor kinase, plays a critical role in cardiac physiology. Adrenergic receptors are the primary target for GRK2 activity in the heart; phosphorylation by GRK2 leads to desensitization of these receptors. As such, levels of GRK2 activity in the heart directly correlate with cardiac contractile function. Furthermore, increased expression of GRK2 after cardiac insult exacerbates injury and speeds progression to heart failure. Despite the importance of this kinase in both the physiology and pathophysiology of the heart, relatively little is known about the role of GRK2 in skeletal muscle function and disease. In this study we generated a novel skeletal muscle-specific GRK2 knock-out (KO) mouse (MLC-Cre:GRK2fl/fl) to gain a better understanding of the role of GRK2 in skeletal muscle physiology. In isolated muscle mechanics testing, GRK2 ablation caused a significant decrease in the specific force of contraction of the fast-twitch extensor digitorum longus muscle yet had no effect on the slow-twitch soleus muscle. Despite these effects in isolated muscle, exercise capacity was not altered in MLC-Cre:GRK2fl/fl mice compared with wild-type controls. Skeletal muscle hypertrophy stimulated by clenbuterol, a β2-adrenergic receptor (β2AR) agonist, was significantly enhanced in MLC-Cre:GRK2fl/fl mice; mechanistically, this seems to be due to increased clenbuterol-stimulated pro-hypertrophic Akt signaling in the GRK2 KO skeletal muscle. In summary, our study provides the first insights into the role of GRK2 in skeletal muscle physiology and points to a role for GRK2 as a modulator of contractile properties in skeletal muscle as well as β2AR-induced hypertrophy.

diverse processes throughout the body by responding to a vast array of extracellular stimuli including hormones, neurotransmitters, and photons of light (1). Equally important to the stimulation and activation of GPCRs is the desensitization and "shutting-off" of the receptor. This task is primarily conducted by the GPCR kinase (GRK) family of proteins (2,3). GRK2 is ubiquitously expressed throughout the tissues of the body, perhaps most notably in the heart where its regulation of adrenergic receptors (ARs) is critical to physiological heart function. Cardiac overexpression of GRK2 in mice suppresses contractility, whereas cardiac overexpression of the GRK2 inhibitor ␤ARKct (a C-terminal peptide that competes with GRK2 binding to G ␤␥ ) enhances contractile function (4). Catecholamine overdrive during heart failure drives increased GRK2 expression in the cardiomyocyte, ultimately leading to excessive desensitization of ␤ARs, loss of receptor density, and a drop in inotropic reserve (5)(6)(7). Indeed, myocardial inhibition or deletion of GRK2 can prevent and even reverse heart failure in numerous animal models (4, 8 -13).
Although we have a firm understanding of the role of GRK2 in the physiology and pathophysiology of the heart, relatively little is known about the function of this kinase in skeletal muscle. Many of the prominent effects of GRK2 in the heart are mediated by regulation of the ␤ 2 AR (2,14). This receptor is also expressed in skeletal muscle and modulates various aspects of skeletal muscle physiology. ␤AR agonists have long been known to induce hypertrophy of skeletal muscle and have been studied as a potential therapeutic for muscle wasting diseases (15). In particular, clenbuterol administration has been shown to induce skeletal muscle hypertrophy via a ␤ 2 AR-dependent mechanism (16). In addition, as is the case in the heart, ␤ 2 AR agonists can modulate the contractile properties of skeletal muscle (17).
Given the aforementioned roles of the ␤ 2 AR in skeletal muscle and the fact that GRK2 is a regulator of the ␤ 2 AR, we sought to determine the genetic requirement for GRK2 in skeletal muscle physiology. Specifically, using skeletal muscle-specific GRK2 knock-out (KO) mice, we assessed exercise performance and contractile properties of isolated muscles. Furthermore we investigated whether GRK2 ablation in skeletal muscle would enhance the pro-hypertrophic effects of a ␤ 2 AR agonist.

Generation of a Skeletal Muscle-specific GRK2 KO Mouse-
To study the role of GRK2 in skeletal muscle function and contractility, we generated a skeletal muscle-specific GRK2 KO mouse by crossing mice with loxP sites flanking exons 3-6 of GRK2 (GRK2 fl/fl ) (18) with MLC-Cre mice. The MLC-Cre mice express Cre-recombinase under control of the myosin light chain 1f (MLC1f) promoter (19). The MLC1f locus is active only in type II fast-twitch fibers. In agreement with others, we found the greatest levels of target gene knockdown in the gastrocnemius and EDL muscles (ϳ83 and 82% knockdown of GRK2 respectively), both of which contain a high proportion of type II fibers (Fig. 1, A and B) (20,21). In contrast, the soleus muscle, which contains fewer type II fibers and a higher proportion of type I slow fibers showed ϳ65% GRK2 knockdown, reflecting lower MLC1f promoter activity in this muscle (Fig. 1C). To convince ourselves of the skeletal muscle specificity of the MLC-Cre:GRK2 fl/fl mice, we assessed GRK2 levels in heart tissue lysate by Western blot and found no change with respect to GRK2 fl/fl animals (Fig. 1D).
Skeletal Muscle GRK2 Knockdown Does Not Impair Exercise Performance-GRK2 activity levels in the heart correspond directly with contractile function (4). To assess whether GRK2 ablation in the skeletal muscle has a direct functional impact, we challenged GRK2 fl/fl and MLC-Cre:GRK2 fl/fl mice with involuntary treadmill running to exhaustion. After a weeklong acclimatization period, mice were run using a protocol of increasing speed over time until they were unable to continue despite receiving a mild electrical shock from a platform at the rear of the treadmill. Using this protocol, we found no difference in time taken to reach exhaustion between GRK2 fl/fl and MLC-Cre:GRK2 fl/fl mice (13.33 min and 12.52 min, respec-FIGURE 1. Skeletal muscle-specific knockdown of GRK2 in MLC-Cre:GRK2 fl/fl mice. GRK2 protein levels were measured via Western blot in lysates from gastrocnemius (A), soleus (B), EDL (C), and heart (D). GAPDH levels were assessed as a loading control. Bars depict the mean signal intensity of GRK2 normalized to GAPDH, and error bars represent S.D. **, p Ͻ 0.01; ***, p Ͻ 0.001 between GRK2 fl/fl and MLC-Cre:GRK2 fl/fl animals; n ϭ 4 and 5 per group, respectively. A.U., arbitrary units. tively) ( Fig. 2A). Likewise we found no difference in the maximum speed attained during the treadmill protocol with GRK2 fl/fl mice reaching 25.56 m/min and MLC-Cre:GRK2 fl/fl mice reaching an average of 24.0 m/min (Fig. 2B). The average total distance run was also calculated for both groups, and again we observed no difference between GRK2 fl/fl and MLC-Cre: GRK2 fl/fl mice, covering 267.84 m and 250.39 m, respectively (Fig. 2C). Finally, we also kept a tally of the number of times each mouse engaged the shock grid at the rear of the treadmill and once again found no difference between groups with GRK2 fl/fl mice engaging the grid an average of 2.75 times/min and MLC-Cre:GRK2 fl/fl mice averaging 2.89 times/min (Fig.  2D). In conclusion we find no effect of muscle GRK2 ablation on exercise capacity in four separate parameters of treadmill performance.
Skeletal Muscle GRK2 Knockdown Differentially Modulates ex Vivo Mechanics of Isolated Soleus and Extensor Digitorum Longus (EDL) Muscles-To obtain a more detailed characterization of skeletal muscle function in the absence of GRK2, we next studied the ex vivo mechanics of soleus and EDL muscles isolated from GRK2 fl/fl and MLC-Cre:GRK2 fl/fl mice. Whole muscle mechanical measurements using a force transducer provide a highly sensitive and reproducible assessment of muscle function. Muscle fiber cross-sectional area (CSA), twitch, and tetanus did not differ between GRK2 fl/fl and MLC-Cre: GRK2 fl/fl groups in either soleus or EDL muscle (Fig. 3, A-C). We did, however, find a statistically significant decrease in EDL muscle-specific force of contraction in MLC-Cre:GRK2 fl/fl compared with GRK2 fl/fl mice (17.5 Ϯ 0.54 N/cm 2 and 19.9 Ϯ 0.69 N/cm 2 , respectively) (Fig. 3D). By contrast, we found an opposite trend in slow-twitch soleus muscle in which there was a marginal, albeit non-significant, increase in specific force in the MLC-Cre:GRK2 fl/fl compared with GRK2 fl/fl mice (18.66 Ϯ 0.57 newtons/cm 2 and 17.49 Ϯ 0.58 newtons/cm 2 respectively) (Fig. 3D). In addition, EDL muscle twitch:tetanus ratio was significantly higher in MLC-Cre:GRK2 fl/fl compared with GRK2 fl/fl mice (0.3 Ϯ 0.004 and 0.26 Ϯ 0.01, respectively), whereas no difference was detected between groups in the soleus (Fig. 3E).
We next made force frequency measurements in the soleus of GRK2 fl/fl and MLC-Cre:GRK2 fl/fl mice to detect any physiological shift in the fiber type composition (Fig. 3F). We were unable to detect any difference between the force frequency profiles of soleus muscle from GRK2 fl/fl and MLC-Cre:GRK2 fl/fl mice. Similarly, we found no difference in the percent fatigability of soleus muscles from these mice during a 10-min fatigue protocol (Fig. 3G). Both the force frequency and fatigability results suggest the ablation of GRK2 in skeletal muscle has little effect on fiber type composition. Indeed, fiber type composition assessed via immunostaining was identical in GRK2 fl/fl and MLC-Cre:GRK2 fl/fl animals (Fig. 4, A and B). We found a close correlation between the percentage of fast-twitch fibers in the soleus (ϳ67%) and the degree of GRK2 knockdown (ϳ65%; Fig.  1B), confirming the fidelity of the MLC1f driven Cre used to generate the knock-out mice. Overall, these results suggest that GRK2 deletion may differentially alter specific force of contraction in the fast-twitch EDL versus slow-twitch soleus muscles without causing a fiber type switch.
Skeletal Muscle GRK2 Knockdown Does Not Alter Ca 2ϩ Transients in Isolated Myotubes-In the heart, ␤AR signaling modulates sarcolemmal calcium regulation, which in turn influences inotropy and chronotropy. We, therefore, next asked whether deletion of GRK2, a key regulator of ␤ARs, could alter Ca 2ϩ transients in skeletal muscle and thus explain the altered  OCTOBER 14, 2016 • VOLUME 291 • NUMBER 42 contractility observed in GRK2 ablated EDL muscle (Fig. 3D). We isolated single skeletal muscle myotubes via enzymatic digestion of the fast-twitch flexor digitorum brevis (FDB) muscle. Myotubes, loaded with the fluorescent Ca 2ϩ indicator Fluo-4, were electrically paced at 0.2 Hz, and cytosolic Ca 2ϩ transients were assessed at baseline and, after isoproterenol stimulation (100 nM, 5 min), via fluorescence microscopy.

GRK2 and Skeletal Muscle Physiology
Baseline Ca 2ϩ transients from GRK2 fl/fl and MLC-Cre: GRK2 fl/fl myotubes produced almost identical traces (Fig. 5, A and B). Isoproterenol stimulation caused a significant increase in Ca 2ϩ transient amplitude in both GRK2 fl/fl and MLC-Cre: GRK2 fl/fl myotubes compared with their respective non-stimulated control groups (Fig. 5, A-C). We found no statistical significance between the Ca 2ϩ transient amplitudes of the isoproterenol-treated groups; however, MLC-Cre:GRK2 fl/fl myotubes trended higher than GRK2 fl/fl . Further analysis found no significant difference between groups in the time to peak (Fig.  5D) or in the time rate of decay of Ca 2ϩ transients (Fig. 5E). From these results, we conclude that neither baseline nor isoproterenol-pretreated Ca 2ϩ transients are significantly influenced by GRK2 deletion in skeletal muscle myotubes.
Clenbuterol-stimulated Skeletal Muscle Hypertrophy Is Enhanced by Skeletal Muscle-specific GRK2 Knockdown-Previous studies have identified clenbuterol administration as an effective way to stimulate muscle hypertrophy and inhibit atrophy (15). Furthermore, this effect of clenbuterol is ␤ 2 AR-dependent (16). Given that GRK2-mediated phosphorylation is instrumental in ␤ 2 AR desensitization, we next asked whether ablation of GRK2 in the skeletal muscle would enhance clenbuterol-induced hypertrophy. GRK2 fl/fl and MLC-Cre:GRK2 fl/fl mice were administered clenbuterol or PBS continuously for 14 days via subcutaneous osmotic minipump. The effect of clenbuterol on body and muscle mass is shown in Table 1.
Chronic clenbuterol treatment (14 days) caused a significant increase in body weight in both GRK2 fl/fl and MLC-Cre: GRK2 fl/fl mice (2.17 Ϯ 0.16 g and 2.33 Ϯ 0.14 g, respectively) although -fold change in body weight, comparing PBS to clenbuterol treatment was significantly greater in MLC-Cre: GRK2 fl/fl compared with GRK2 fl/fl mice (2.83 Ϯ 0.17-fold and 2.11 Ϯ 0.15-fold, respectively). Clenbuterol treatment caused a significant increase in the soleus weight of both GRK2 fl/fl and MLC-Cre:GRK2 fl/fl mice (0.37 Ϯ 0.01 g and 0.41 Ϯ 0.01 g, respectively). However, both the absolute weight gain and -fold change in soleus weight comparing PBS to clenbuterol treat-ment were significantly greater in MLC-Cre:GRK2 fl/fl compared with GRK2 fl/fl mice (1.31 Ϯ 0.02-fold and 1.15 Ϯ 0.03fold, respectively). In addition, clenbuterol treatment caused a significant increase in TA and gastrocnemius weight relative to PBS treatment in MLC-Cre:GRK2 fl/fl mice, whereas the same muscles in GRK2 fl/fl mice trended toward a clenbuterol-stimulated increase in mass yet did not reach significance. As with the soleus, -fold change in gastrocnemius weight comparing PBS to clenbuterol treatment was significantly greater in MLC-Cre: GRK2 fl/fl compared with GRK2 fl/fl mice (1.16 Ϯ 0.03-fold and 1.03 Ϯ 0.03-fold, respectively).
To support the muscle weight data, we analyzed fiber crosssectional area of TA muscle from these mice (Fig. 6, A and B). Clenbuterol treatment significantly increased the cross-sectional area of both GRK2 fl/fl and MLC-Cre:GRK2 fl/fl muscle fibers compared with PBS-treated groups. Importantly, the clenbuterol stimulated increase in fiber cross-sectional area was greater in MLC-Cre:GRK2 fl/fl compared with GRK2 fl/fl mice (1436.29 Ϯ 88.70 m 2 versus 1166.57 Ϯ 64.27 m 2 , respectively). Overall, we found GRK2 ablation increases clenbuterol-induced hypertrophy of skeletal muscle.
Pro-hypertrophic Akt Signaling Is Elevated in Clenbuteroltreated MLC-Cre:GRK2 fl/fl Mice-␤ 2 -Adrenergic agonists are known to stimulate hypertrophy, in part via activation of the pro-hypertrophic Akt signaling pathway; we, therefore, hypothesized that increased Akt signaling contributes to the  OCTOBER 14, 2016 • VOLUME 291 • NUMBER 42  enhanced hypertrophy response in clenbuterol-treated MLC-Cre:GRK2 fl/fl mice. We examined Akt signaling in soleus muscle harvested from mice after 4 h of acute clenbuterol stimulation (1 mg/ml). p-Akt levels were significantly increased in both GRK2 fl/fl and MLC-Cre:GRK2 fl/fl soleus muscle after clenbuterol stimulus (Fig. 7, A  and B); however, -fold increase in p-Akt in clenbuterol-treated mice relative to PBS controls was significantly greater in MLC-Cre:GRK2 fl/fl mice (Fig. 7C). Levels of p-GSK3␤, a downstream target of Akt, were likewise significantly increased in clenbuterol-treated MLC-Cre:GRK2 fl/fl mice compared with PBS controls (Fig. 7, A and D). p-GSK3␤ levels trended higher in clenbuterol-treated GRK2 fl/fl mice yet fell short of statistical significance (Fig. 7, A and D). As with p-Akt, -fold increase in p-GSK3␤ levels in clenbuterol-treated mice relative to PBS controls was significantly greater in MLC-Cre:GRK2 fl/fl mice (Fig. 7E). In conclusion, clenbuterol-stimulated Akt signaling is enhanced in GRK2-ablated skeletal muscle, which may in part explain the pro-hypertrophic phenotype in these mice.

Discussion
Research over the last few decades has identified dichotomous roles for GRK2 in physiology and pathophysiology of the heart. GRK2 activity is instrumental in the physiological homeostasis of the heart by phosphorylating and thereby desensitizing activated GPCRs, predominantly ␤ARs (2). As a consequence, GRK2 activity levels in the heart correlate with contractile function (4,22). However, after cardiac insult, increased GRK2 expression directly contributes to injury and  OCTOBER 14, 2016 • VOLUME 291 • NUMBER 42 progression to heart failure via excessive ␤AR desensitization and increased myocyte apoptosis (5)(6)(7)23).

GRK2 and Skeletal Muscle Physiology
Given the well characterized roles of GRK2 in heart function and disease, in this study we sought insight into the role of GRK2 in skeletal muscle function. Although not identical in nature, skeletal muscle and cardiac muscle still share many features, including many of the same mechanisms and proteins required for excitation-contraction coupling. To examine the role of GRK2 in skeletal muscle we generated a skeletal musclespecific GRK2 KO mouse. To our knowledge this is the first study to directly investigate the role of GRK2 in skeletal muscle function.
We studied the contractile properties of isolated soleus and EDL muscle and found that skeletal muscle GRK2 KO caused a decrease in specific force of contraction in the fast-twitch EDL muscle. Furthermore the twitch:tetanus ratio was significantly elevated in GRK2-ablated EDL. In contrast, we found no significant changes in the slow-twitch soleus mechanics from these animals; in fact, contrary to the EDL, the specific force trended higher in the soleus of skeletal muscle GRK2 KO mice. The EDL mechanics results were somewhat unexpected, as deletion of GRK2 increases isoproterenol-stimulated contractility and calcium transients in isolated cardiomyocytes (24). In this study we found no difference in baseline Ca 2ϩ transients in GRK2 KO and wild-type (WT) myotubes. Isoproterenol pretreatment increased Ca 2ϩ transient amplitude in both GRK2 KO and WT myotubes, and although GRK2 KO amplitude values trended higher than those of WT cells, we ultimately found no statistical difference between isoproterenol-treated groups. This would indicate that GRK2 KO alters EDL contractility independently of Ca 2ϩ in the skeletal muscle. In GRK2-ablated cardiomyocytes, Raake et al. (24) found that elevated Ca 2ϩ transient amplitudes were caused, at least in part, by enhanced phosphorylation of the L-type Ca 2ϩ channel (LTCC) by protein kinase A (PKA), a downstream effector of ␤AR activation . Phosphorylation of the LTCC increases Ca 2ϩ influx through the channel, which in turn heightens calcium-induced calcium release through the ryanodine receptor (RyR). This GRK2 dependent effect is irrelevant in skeletal muscle, since RyR opening is regulated by direct physical interaction with the LTCC and not by calcium-induced calcium release. Furthermore, Raake et al. (24) found that phosphorylation of phospholamban, a key regulator of sarco/endoplasmic reticulum Ca 2ϩ -ATPase (SERCA)-mediated Ca 2ϩ re-uptake into the sarcoplasmic reticulum, was decreased in GRK2-ablated cardiomyocytes, resulting in faster Ca 2ϩ transient decay rates. Phospholamban is not expressed in type-II skeletal muscle fibers; hence, any effect of GRK2 deletion on phospholamban phosphorylation is not relevant to this cell type. Given these key distinctions in Ca 2ϩ handling machinery, it is perhaps not surprising that isoproterenol-mediated changes in the Ca 2ϩ transients of GRK2 ablated cardiomyocytes are largely absent in the skeletal muscle myotube.
In the heart, ␤AR stimulation generates an increase in inotropy and chronotropy (25). Research into the effects of ␤AR stimulation on skeletal muscle contraction has yet to reach a consensus; results in this tissue seem to vary based upon species, muscle type, and even dose of agonist (17). This may also help to explain why GRK2 deletion does not have the same effect on contractile properties in the EDL as it does in the heart. It is interesting to note, however, that similar to the heart the soleus trended toward increased contractility after GRK2 ablation. In contrast to the fast-twitch EDL, the soleus is largely composed of mitochondria-rich, type I muscle fibers, which are similar in nature to the slow oxidative fibers constituting the heart. It is, therefore, possible that the effects of GRK2 deletion on skeletal muscle contractile properties are dependent upon fiber type.
GRK2 can translocate to the mitochondria, and although the consequences of this localization are not completely understood, our laboratory has shown that GRK2 disrupts mitochondrial function in the cardiomyocyte by altering substrate utilization for ATP production (26). In addition, our laboratory and others have found GRK2 to be a negative regulator of glucose uptake (27,28). This may be of particular functional relevance to the glycolytic EDL muscle; it is possible that altering glucose homeostasis by GRK2 ablation in the skeletal muscle may result in a metabolic reprogramming, which in turn could affect energy production and muscle contractile properties as a consequence. Further studies examining this hypothesis are warranted.
We next conducted force frequency and fatigue tests on isolated soleus muscle to detect any physiological shift in the fiber type composition. Skeletal muscle GRK2 KO mice were indistinguishable from their WT counterparts in both tests, suggesting GRK2 expression in skeletal muscle has no effect on fibertype composition. Immunostaining and quantification of muscle fibers in soleus sections confirmed this finding. ␤ 2 AR agonists can promote slow-to-fast fiber-type switching in skeletal muscle (29,30). Our data suggest that reducing the desensitization of ␤ 2 ARs by GRK2 ablation alone is not sufficient to promote a fiber type shift in the absence of an exogenous ␤-agonist.
The lack of fiber type shift or difference in soleus fatigability between GRK2KO and WT mice helps to explain why we also see no difference in involuntary treadmill running to exhaustion in these animals. However, given the apparent difference in EDL-specific force generation between our mice, future studies may warrant additional experiments more suited to identifying differences in muscle strength, such as a grip strength test.
Chronic stimulation of ␤ARs with ␤-agonists has long been known to cause a "repartitioning effect," decreasing body fat while increasing skeletal muscle mass (15). This has led to extensive study of ␤-agonists as a potential therapeutic for muscle wasting diseases such as muscular dystrophy and cancer cachexia. The anabolic effect of clenbuterol on skeletal muscle is due to decreased protein degradation while simultaneously increasing protein synthesis (31,16). The latter action seems to be primarily mediated by activation of Akt, a well characterized mediator of muscle hypertrophy (32). In our study we found that the anabolic effects of chronic clenbuterol administration (14 days) were potentiated in skeletal muscle by the deletion of GRK2, most likely due to decreased desensitization of activated ␤ 2 AR. Furthermore, this appears to be mediated in part by enhanced Akt activation that can occur when GRK2 activity is lowered via knockdown in skeletal muscle cells, although other mechanisms including gene regulatory changes downstream of the ␤ 2 AR cannot be ruled out. Interestingly, in cardiomyocytes, GRK2 deletion prevents hypertrophy after injury; however, this can be explained by the lower levels of cardiomyocyte apoptosis and increased retention of contractile mass after injury in GRK2 KO hearts, reducing the necessity for compensatory hypertrophy (24). Overall, we find that inhibition of GRK2 may increase the effectiveness of ␤ 2 AR-mediated strategies to grow skeletal muscle mass, which could have translational significance.
In summary, our findings provide novel insight into the function of GRK2 in skeletal muscle. These results highlight a role for GRK2 as a modulator of contractile properties in skeletal muscle. Furthermore, our data show that GRK2 deletion in skeletal muscle potentiates the anabolic effect of clenbuterol administration, in part by enhancing Akt activity. These obser- OCTOBER 14, 2016 • VOLUME 291 • NUMBER 42 vations may prove useful in the understanding of skeletal muscle physiology and in developing more effective therapeutic strategies for the treatment of skeletal muscle wasting diseases.

Experimental Procedures
Mice-All animal studies were conducted with the approval of the Animal Care and Use Committee at Temple University. To obtain skeletal muscle-specific GRK2 KO mice (MLC-Cre: GRK2 fl/fl ), MLC-Cre mice expressing Cre-recombinase under control of the myosin light chain 1f (MLC1f) promoter (19) were crossed with GRK2 fl/fl mice (18). GRK2 fl/fl mice were used as WT controls for all experiments. All mice used were between 10 and 16 weeks of age.
Treadmill Exercise-Animals were acclimatized to treadmill running (Columbus Instruments) over 5 days starting with a 5-min session at a speed of 6 m/min on day 1, gradually increasing to a speed of 15 m/min by day 5. Mice were given two full days rest after acclimatization before undergoing the treadmill test to exhaustion. The test was conducted as follows: 3 min at 10 m/min, 3 min at 20 m/min, then increasing speed by 2 m/min every 3 min until a final speed of 28 m/min was reached. A 0% gradient was used for all treadmill running. Mice were considered to be exhausted when they engaged a platform at the rear of the treadmill for more than 5 s despite receiving a mild electric shock. Running time, maximum speed reached, distance covered, and number of times mice engaged the shock grid were recorded.
Muscle Functional Testing-Soleus and EDL muscles were subjected to isolated mechanical measurements using a previously described apparatus (Aurora Scientific, Ontario, Canada) (33) and bathed in Ringer's solution gas-equilibrated with 95% O 2 , 5% CO 2 . Optimum muscle length (Lo) was determined with iterative manual adjustments of length to achieve maximum twitch force with supramaximal stimulation.
Maximum isometric twitch was measured in the muscles followed after 20 s by maximum isometric tetanus during a 500-ms stimulation. The maximum twitch and tetanus were measured a total of 3 times with 5-min intervals between tests, and the individual maximum value was used. The soleus was subjected to additional functional measures. First, a force-frequency test was employed using stimulation frequencies of 10,20,30,50,70,90, and 100 Hz. Force generation at 90 and 100 Hz did not differ in any muscle studied, and the force was normalized to the isometric tetanus at 100 Hz to generate force-frequency curves. Second, soleus muscles were subjected to a fatigue test, as previously described (34). Briefly, muscles were stimulated once per second for 10 min (200-s pulse, 100 Hz, 330-ms duration) to determine resistance to fatigue. Upon completion of functional testing, muscles were blotted, weighed, and rapidly frozen for subsequent analysis. Specific force was determined based on the physiological cross-sectional area (PCSA) using the formula, where m is muscle mass, Lo is muscle length, Lf/Lo is the ratio of fiber length to optimal muscle length, and density of muscle () ϭ 1.06 g/cm 3 . Lf/Lo was 0.45 for EDL and 0.69 for soleus.
Micro-osmotic Pumps-Chronic infusion of clenbuterol (clenbuterol hydrochloride, Sigma) was done using Alzet 14-day micro-osmotic pumps (model 1002, DURECT Corp.). Pumps were filled following the manufacturer's specifications with sterile PBS or clenbuterol (3 mg/kg per day) and inserted as previously described (35). Briefly, mice were anesthetized with isoflurane (2.5% v/v), and pumps were implanted subcutaneously through a subscapular incision, which was then closed using 4.0 silk suture (Ethicon). The contents of the pumps were delivered at a rate of 0.25 l/h for 14 days.
Acute Clenbuterol Treatment-Acute clenbuterol stimulation of mice was achieved via a single, subcutaneous injection of clenbuterol (1 mg/kg) or sterile PBS (control). 4 h post-injection muscles were isolated and rapidly frozen for subsequent analysis.
Fiber Cross-sectional Area-TA muscles were paraffin-embedded, and 7-m sections were cut at the center of the muscle. Muscle sections were stained with H&E; images were taken using a Nikon DS-Ri1 and quantified in a blind manner using ImageJ. 100 fibers across 5 separate fields were measured per mouse.
Fiber-type Immunohistochemistry-After harvest, soleus muscles were immediately embedded in Tissue-Tek OCT compound (Sakura Finetek), and 10-m sections were cut using a cryotome. Sections were blocked in 5% BSA for 45 min before primary antibody incubation overnight at 4°C in a humid chamber. All sections were incubated with laminin primary antibody (Thermo, #RB-082-A0) in addition to one of the following fiber-type primary antibodies: MyHCI, MyHCIIa, or MyHCIIb (all MyHC antibodies were from the Developmental Studies Hybridoma Bank, #BF-F3, SC-71, and BA-F8). Secondary antibodies (Life Technologies A21434 and A21202; Immu-noResearch Laboratories 715-545-140) were applied for 1 h at room temperature. Sections were treated with mounting media containing DAPI and coverslipped. Sections were visualized using a Nikon-Ti fluorescence microscope.
Isolation and Dissociation of FDB Muscle-Single skeletal muscle myotubes were obtained via enzymatic digestion of FDB muscles. Following dissection, isolated FDB muscles were placed in 35-mm dishes containing dissociation media composed of DMEM, 2% fetal bovine serum, and 2 mg/ml collagenase II (Worthington Biochemical Corp.) and incubated at 37°C, 5% CO 2 for 1.5-2 h. Next, muscles were transferred to a new 35-mm dish containing incubation media composed of DMEM, 2% fetal bovine serum, and 1% penicillin/streptomycin. The muscles were gently triturated with a sterile, wide-bore P1000 pipette until single myotubes were dissociated. Myo-tubes were then seeded onto 35-mm collagen-coated, glassbottomed dishes (MatTek Corp.) at 50 -60% confluence and allowed to settle and adhere overnight in a 37°C, 5% CO 2 incubator. Experiments were conducted the following day.
Myotube Cytosolic Ca 2ϩ Transient Measurements-Isolated myotubes were loaded with 5 M Fluo-4 AM (Invitrogen) for 20 min at room temperature. Loaded myotubes were placed in a 37°C heated chamber on an inverted microscope stage. Myotubes were perfused with a normal physiological Tyrode's buffer (150 mM NaCl, 5.4 mM KCl, 1.2 mM MgCl 2 , 10 mM glucose, 2 mM sodium pyruvate, and 5 mM HEPES, pH 7.4) containing 2 mM Ca 2ϩ . Myotubes were paced at 0.2 Hz and continuously recorded for Ca 2ϩ transients using a Zeiss Observer Z1 fluorescent microscope at 490/20ex and 535/50em. Ca 2ϩ transients were measured at baseline and after pretreatment with isoproterenol (100 nM, 5 min). For intracellular Ca 2ϩ fluorescence measurements, the F 0 (or baseline) was measured as the average fluorescence of the myotube 100 ms before stimulation. The maximal Fluo-4 fluorescence (F) was measured for peak amplitude. Time to peak was calculated as the time from the beginning of the transient to peak amplitude. Tau was measured as the decay rate of the Ca 2ϩ transient traces.
Statistics-All the values in the text and figures are presented as the mean Ϯ S.E. Statistical significance was determined by Student's t test or ANOVA. p values of Ͻ0.05 were considered significant.