β2-adrenergic receptor agonist counteracts skeletal muscle atrophy and oxidative stress in uremic mice

In patients with chronic kidney disease, skeletal muscle dysfunction is associated with mortality. Uremic sarcopenia is caused by ageing, malnutrition, and chronic inflammation, but the molecular mechanism and potential therapeutics have not been fully elucidated yet. We hypothesize that accumulated uremic toxins might exert a direct deteriorative effect on skeletal muscle and explore the pharmacological treatment in experimental animal and culture cell models. The mice intraperitoneally injected with indoxyl sulfate (IS) after unilateral nephrectomy displayed an elevation of IS concentration in skeletal muscle and a reduction of instantaneous muscle strength, along with the predominant loss of fast-twitch myofibers and intramuscular reactive oxygen species (ROS) generation. The addition of IS in the culture media decreased the size of fully differentiated mouse C2C12 myotubes as well. ROS accumulation and mitochondrial dysfunction were also noted. Next, the effect of the β2-adrenergic receptor (β2-AR) agonist, clenbuterol, was evaluated as a potential treatment for uremic sarcopenia. In mice injected with IS, clenbuterol treatment increased the muscle mass and restored the tissue ROS level but failed to improve muscle weakness. In C2C12 myotubes stimulated with IS, although β2-AR activation also attenuated myotube size reduction and ROS accumulation as did other anti-oxidant reagents, it failed to augment the mitochondrial membrane potential. In conclusion, IS provokes muscular strength loss (uremic dynapenia), ROS generation, and mitochondrial impairment. Although the β2-AR agonist can increase the muscular mass with ROS reduction, development of therapeutic interventions for restoring skeletal muscle function is still awaited.

Sarcopenia is classically defined as a decline in the skeletal muscle mass, strength, and endurance. Nowadays, particularly low muscle strength is underscored as a critical characteristic and dysfunction to be treated 1 . Although it has been well studied in the context of ageing, a growing amount of epidemiological evidence has revealed a high incidence of sarcopenia in elderly patients with chronic kidney disease (CKD) as kidney function deteriorates [2][3][4][5][6][7] . The importance of uremic sarcopenia lies in its impact on mortality and morbidity, including fractures that affect quality of life for CKD patients 8 , cardiovascular events 9 , and overall survival 10 .
In addition to multiple contributors pinpointed by previous studies [11][12][13] , inappropriate accumulation of uremic toxin may yield uremic sarcopenia. Uremic toxins are usually filtered and excreted by healthy kidneys but are accumulated in the body when kidney function has deteriorated. Among the uremic toxins, the protein-bound uremic toxins, such as indoxyl sulfate (IS) and p-cresyl sulfate, are hardly ever removed by hemodialysis because of their high affinity to serum albumin 14,15 and are considered directly harmful to visceral systems such as the kidney [16][17][18] , cardiovascular system 19,20 , and immune system 21 . IS as a representative uremic toxin is generated in the liver via the metabolism of indole absorbed in the gut 22 . Accumulation of IS also exerts negative effects on myoblast proliferation and myotube size in vitro and skeletal muscle mass in vivo [23][24][25][26][27] . In addition, IS induced metabolic alterations resulting in mitochondrial dysfunction and ATP shortage in muscle cells 28 . However, skeletal muscle functioning after direct administration of IS has not been well documented in mice. Also, previous Protein synthesis is suppressed in atrophic skeletal muscles of mice injected with IS. Muscle quantity is determined by the homeostatic balance of protein synthesis and proteolysis 44 . To elucidate the mechanism of muscle atrophy in mice injected with IS, we first evaluated atrophy related gene expression associated with classic ubiquitin-proteasome pathways, such as myostatin and atrogenes (atrogin-1, and MuRF-1), which are involved in the pathomechanism of skeletal muscle atrophy in sarcopenia 45 . In IS treated mouse GC muscle, these atrophic gene expressions were not increased (Fig. 3A). Therefore, we conducted in vivo experiments using puromycin assays for analyzing protein synthesis, the surface sensing of translation technique 46,47 . Western blotting analysis revealed that the expression of peptides labeled with puromycin was noted only in mice injected with puromycin, as anticipated, and was decreased in mice treated with IS ( Fig. 3B), indicating suppression of muscle protein synthesis in these mice. Therefore, as a mechanism for reduction of skeletal muscle mass in the current model, IS treatment may restrict protein synthesis rather than promote proteolysis via atrogene-related ubiquitination.
Clenbuterol increases mouse muscle mass but does not improve weakness induced by IS. Next, to explore potential treatments for uremic sarcopenia, we focused on the ergogenic β2-AR agonists allegedly used for livestock in several countries 48,49 . Moreover, since β2-AR agonists induce slow-to-fast MHC isoform transition 50 , a β2-AR agonist could theoretically be a promising way to ameliorate skeletal muscle mass reduction and dysfunction in our model where predominantly fast-twitch myofibers have deteriorated. Indeed, clenbuterol treatment increased not only body weight (Fig. 4A) but GC muscle mass (Fig. 4B) in mice treated with IS. Furthermore, clenbuterol treatment led to an increase in cross-sectional areas (Fig. 4C) and MHC protein expression (Fig. 4D) of fast-twitch muscle fiber predominantly. However, intriguingly, clenbuterol treatment could not restore IS-induced muscle weakness (Fig. 4E).
IS, as a representative uremic toxin, induces reactive oxygen species (ROS) accumulation in tissues and organs [51][52][53] . Previous reports illustrated that ROS plays a pivotal role in skeletal muscle atrophy associated with ageing or cardiovascular diseases [54][55][56] . Therefore, we postulated that ROS is accumulated in mouse GC muscle treated with IS which might be attenuated by clenbuterol treatment. Immunofluorescent staining with 4-hydroxynonenal (4-HNE) revealed that lipid peroxidation was enhanced in mice treated with IS that was blunted by clenbuterol treatment (Fig. 4F). www.nature.com/scientificreports/ Clenbuterol does not restore mitochondrial function in cultured myotubes stimulated with IS. After confirming that C2C12 myotubes express AhR, an endogenous IS receptor (Fig. 5A), Since IS in the bloodstream forms a complex that binds to albumin, the absence of proper albumin quantities could result in a high concentration of free IS and induce exaggerated biologic responses with an overestimated toxic effect, as detailed elsewhere 15 . Therefore, we co-administered IS and albumin to the culture media (Fig. 5B). This treatment decreased the length and diameter of fully differentiated myotubes (Fig. 5C). Next, after confirmation of β2-AR expression in C2C12 myotube (Fig. 6A), we evaluated whether clenbuterol treatment may modulate the effect of IS on myotubes. Thus, clenbuterol and anti-oxidant L-ascorbic acid ameliorated IS-induced myotube size reduction (Fig. 6B). Stimulation with IS also induced cellular ROS accumulation www.nature.com/scientificreports/ in myotubes, which was confirmed by its partial cancellation after L-ascorbic acid treatment (Fig. 6C). Another anti-oxidant, N-Acetyl-L-cysteine, also attenuated the myotube size reduction as well as ROS accumulation that are both induced by IS ( Supplementary Fig. 5), suggesting that ROS scavenging contributes to the increase in cell size. Intriguingly, treatment of C2C12 myotubes with clenbuterol partially eliminated the intracellular ROS accumulation that had been induced by IS stimulation (Fig. 6C). Then, we focused on mitochondrial dysfunction which reportedly centers IS toxicity in the skeletal muscle and causes uremic sarcopenia 7,23,28 . Thus, MitoTracker Red staining of the C2C12 myotube reveled that IS treatment decreased the mitochondrial membrane potential that is critical for the energy production needed for muscle contraction (Fig. 6D). Then, clenbuterol treatment failed to restore the reduction of mitochondrial membrane potential in myotubes (Fig. 6D).

Discussion
Our data demonstrates that IS administration to mice following unilateral nephrectomy is sufficient to deteriorate four limbs grip strength while exercise endurance and capacity were spared. We also found that skeletal muscle size reduction mainly occurs in fast-twitch myofibers that play an important role in instantaneous muscle   www.nature.com/scientificreports/ strength and its quantity is associated with the risk of falls and hip fractures resulting in long hospitalizations and increased mortality [57][58][59][60] . In ageing sarcopenia, physiological and morphological changes in skeletal muscle are characterized by an overall decline in the size and number of skeletal muscle fibers, mainly the fast-twitch muscle fibers 59 . That applies also to secondary sarcopenia due to malnutrition and debilitating diseases such as cancer, diabetes, sepsis, and infection [29][30][31][32][33][34]36 . Therefore, preventing fast-twitch muscle fiber atrophy may be one of the treatment targets for uremic sarcopenia.
In the current study, we injected mice with IS for seven days after unilateral nephrectomy. Enoki et al. also clarified skeletal muscle atrophy with elevated atrogene expression in unilaterally nephrectomized mice injected with lower doses of IS for a longer period (100 mg/kg body weight for 12-weeks) 23 . Moreover, the muscle wet weight decreased in both GC muscle and soleus muscle rich in slow-twitch muscle fibers. Although our animal model failed to show the same results regarding GC muscle, the fast-twitch MHC protein expression and instantaneous muscle strength were significantly decreased. These changes may represent the characteristics of "dynapenia, " which was proposed in 2008 to define the age-related loss of muscle strength rather than muscle mass reduction 61 . Dynapenia is typical for age-related muscle weakness characterized by reduction in fast-twitch (type 2) muscle fibers and increased intramuscular adipose tissues 62,63 . Furthermore, a past study showed that dynapenia is more prevalent than sarcopenia in pre-dialysis CKD patients, and 50% of the dynapenic patients were in early stage of CKD 64 . Therefore, dynapenic changes with the decrease in predominantly fast-twitch muscle via impaired protein synthesis seen in our protocol might be the very early changes of IS-induced sarcopenia, followed by the whole fiber type reduction via proteolysis in the chronic phase. www.nature.com/scientificreports/ The gene expressions of myostatin and atrogenes such as atrogin-1 and MuRF-1 did not change in the current model although these genes were originally found to be up-regulated by fasting and inactivity in early phases of muscle wasting 65,66 . Increased expression of these genes was also reported after subtotal nephrectomy or longterm treatment of uremic toxin in experimental animals, accompanied by significant body weight loss 23,24,67 . However, the current animal model with IS injection did not develop significant loss of body weight and foodintake, thus explaining the discrepancy. Also, to be noted, it remains controversial whether atrogin-1 and MuRF-1 signaling is operative in muscle wasting during ageing. Atrogin-1 and/or MuRF1 mRNA levels are reportedly unchanged in aged muscle of human 68 or even decreased in that of rats 69,70 . Instead, our in vivo experiments show that IS suppressed skeletal muscle protein synthesis, rather than proteolysis, due to atrogene-related ubiquitination. Past reports demonstrated that C2C12 myotubes treated with IS show decreased insulin-stimulated protein expression of phospho-S6K1 28 . Also, in the mouse CKD model with subtotal nephrectomy, a deficiency in the nucleolar protein 66, selectively, in the skeletal muscle counteracts degradation of protein synthesis independently of myostatin signaling 71 . These findings are in line with our current concept that impairment of skeletal muscle protein synthesis contributes to uremic sarcopenia.
To explore therapeutic intervention for the IS toxic effects described above, a β2-AR agonist might have been suitable because it reinforces mainly fast MHC isoforms and up-regulates the Akt-mTOR protein synthesis pathway 50 . In our study, the β2-AR agonist increased the mouse GC muscle mass rich in fast-twitch myofibers and protein expression of MHC. Surprisingly, clenbuterol treatment also attenuated the ROS accumulation caused by IS just as anti-oxidant L-ascorbic acid treatment did, which has not been reported before. However, expression of anti-oxidant enzymes such as catalase and glutathione peroxidase 1 was not increased by clenbuterol treatment (data not shown). Therefore, the mechanism for clenbuterol role in diminishing IS-induced ROS accumulation still needs to be elucidated.
In a recent clinical guideline, low muscle strength was highlighted as a key characteristic of sarcopenia and a restoration of skeletal muscle function is recognized as one of the therapeutic goals 1 . In this regard, clenbuterol treatment alone is still suboptimal as we could not find significant improvement of skeletal muscle strength in uremic mice, which has also been reported elsewhere in humans 72,73 , possibly for the following reasons. First, as shown in our C2C12 experiments, clenbuterol was not sufficient for potentiation of the mitochondrial membrane potential that was impaired by IS. The result is in line with the current concept that mitochondrial dysfunction plays a pivotal role in uremic sarcopenia 74 . Second, IS treatment may directly affect muscle contraction independently of muscle mass. Skeletal muscle contraction requires calcium-dependent cellular responses, and the ER has a pivotal role in maintaining calcium concentration 75 . Moreover, IS induces ER stress on human proximal tubular cells and mouse C2C12 myocytes 17,76 . Therefore, IS-induced mitochondrial dysfunction and ER stress may disturb energy metabolism and calcium-dependent responses on myotube cells, respectively, and result in skeletal muscle weakness. Third, we observed the capacity of clenbuterol to cause amplification of fast-twitch fibers, which is consistent with previous reports 77, 78 . This uneven alteration of fiber type may be disadvantageous for skeletal muscle performance. Interestingly, when clenbuterol administration is accompanied by low endurance training 79 or electrical stimulation 72 , the slow-to fast-twitch fiber transformations are suppressed and then muscle performance are potentiated.
From a clinical viewpoint, since it is a decongestant and bronchodilator and is approved for treatment of bronchial asthma in several countries, drug repurposing for uremic sarcopenia is deemed not unrealistic. However, clenbuterol might promote arrhythmia and cardiomegaly particularly after long-term administration 73 , so, optimally, may not be used for patients with CKD who have a high risk of cardiovascular diseases 80 .
In conclusion, a representative albumin-bound form of uremic toxin, IS, directly impairs skeletal muscle strength and leads to dynapenic phenotype in mice. Clenbuterol treatment may be effective in increasing skeletal muscle mass via reduction of ROS accumulation. For uremic myopathy, work toward development of a new drug to intervene against impaired muscle strength needs to be continued.

Materials and methods
Reagents. Clenbuterol (C5423), L-ascorbic acid (A92902), N-Acetyl-L-cysteine (A7250), anti-actin (A2066), anti-skeletal slow myosin (M8421), anti-skeletal fast myosin (M4276), anti-puromycin (clone 12D10, MABE343, Millipore), and anti-laminin polyclonal antibody (AB19012, Millipore) were purchased from Sigma-Aldrich (St Louis, MO, USA). IS was purchased from Nacalai Tesque (#19,208-04, Kyoto, Japan). The other reagents were described in each experiment. Figure 6. IS induced cell size reduction and increased ROS accumulation that were suppressed by treatment with L-ascorbic acid and clenbuterol. (A) Expression of β2-AR was confirmed with electropheresis for PCR products of mouse C2C12 myotube cDNA. PCR products without reverse transcriptase (RT) was loaded as a negative control. (B) Representative micrographs of C2C12 myotubes stimulated with vehicle (VEH), IS, IS plus clenbuterol (Clen), and IS plus L-ascorbic acid (L-AsA) for 24 h were shown (left). The scale bar is 60 μm. Myotube morphology was quantitatively evaluated in terms of the length and the diameter (right). N = 100 and 200 myotubes per group, respectively. (C) Cellular ROS accumulation was evaluated with CM-H2DCFDA staining intensity in C2C12 myotubes treated with vehicle (VEH), IS, IS plus Clen, and IS plus L-AsA for 24 h (left). The scale bar is 100 μm. Signal intensity was quantitated (right). N = 3 per group. (D) Mitochonrial membrane potential was evaluated with MitoTracker Red (left). Staining intensity in C2C12 myotubes treated with VEH, Clen, and L-ascorbic acid (L-AsA) for 1 h. followed by incubation with or without IS for another 1 h was quantitated (right). The scale bar is 100 μm. N = 4-6 per group. Error bars indicate SEM. * p < 0.05, *** p < 0.001, #; p < 0.05 compared to VEH without IS treatment. www.nature.com/scientificreports/ HPLC analysis. IS levels in the plasma and gastrocnemius were measured by an HPLC method as described previously 24 . In brief, plasma or a gastrocnemius homogenate (homogenized in distilled water) was mixed with acetonitrile (1:9, v/v for the sample) and centrifuged at 12,000 g for 10 min. The supernatant was collected and was loaded to HPLC with 10 μL for the plasma sample or the tissue homogenate. The HPLC system consisted of Jasco PU-4180 pump and Jasco FP-4020 fluorescence spectrophotometer (Tokyo, Japan). A CAPCELL PAK C18 MG II column (SHISEIDO, Tokyo, Japan) was used for the stationary phase. The mobile phase consisted of 0.2 M acetate buffer (pH4.5)-acetonitrile (3:1, v/v) for IS. The flow rate was 1.0 mL/min. IS was detected by means of a fluorescence monitor with excitation/emission wavelengths set to 280 and 375 nm, respectively.
Animals. C57BL/6J female mice (7-8 weeks of age, 19.36 ± 0.15 g body weight) were purchased from CLEA Japan (Tokyo, Japan), and housed on a 12 h day/night cycle. Mice were anesthetized using an isoflurane-oxygen mixture (3% induction and 1.5-2.5% maintenance) and the left kidney was removed. At 1 week after surgery, mice were administrated with IS (600 mg/kg body weight, ip; intraperitoneal injection, everyday) for 7 days. IS was dissolved in phosphate-buffered saline (PBS). The control mice were administrated with PBS for 7 days at the same volume. In another protocol, clenbuterol (2 mg/kg body weight) was also intraperitoneally injected 30 min before the IS injection. The mouse body weight was measured every day. Treadmill exhaustion tests and four limbs grip strength tests were performed 12 h before sacrifice. At the end of the study, mice were anesthetized with the isoflurane-oxygen mixture (5% induction), and then mouse body weight was measured, and blood, kidney, heart, and skeletal muscles (soleus and gastrocnemius) were harvested. The concentrations of blood urea nitrogen (BUN) in mice blood serum were measured with commercial test kits (Wako Pure Chemical Industries, Ltd., Japan). Hemoglobin level was measured with an automatic blood cell counter. The animal experiments were carried out in accordance with the ARRIVE guidelines.
Treadmill exhaustion tests. Treadmill exhaustion tests were executed utilizing TMS-8B (Melquest Ltd, Japan). Mice were kept running individually on separated lanes with a shock grid (stimulatory shock of 0.3 mA) at the rear end. Before the test, mice were acclimated to treadmill running with the speed increased gradually Four limbs grip strength test. Peak force of the four limbs grip was measured using a grid attached to an isometric force transducer (IMADA, Toyohashi, Japan). Mice were allowed to grasp the metal grid with all four limbs. The maximum force exerted was recorded. This test was repeated 5 times and an average score was calculated. To achieve validation, all grip strengths (N, Newtons) were normalized to body weight (g, gram).
Muscle fiber-type analysis. GC muscles were harvested from C57BL/6J mice, embedded in the Tissue-

Target gene Forward Reverse
MuRF www.nature.com/scientificreports/ Analysis for skeletal muscle protein synthesis. Skeletal muscle protein synthesis was evaluated in vivo by the surface sensing of translation technique as previously described 46,47 . Under anesthesia, 0.04 μmol puromycin/g body weight (Sigma-Aldrich, St Louis, MO) diluted in PBS was intraperitoneally injected. Then, the GC muscle was removed 15 min. after puromycin administration. After homogenization, the supernatant was collected and used for western blotting.
Quantitative reverse-transcriptase PCR analysis. Total RNA was extracted using RNAiso Plus (Takara Bio Inc., Shiga, Japan) according to the manufacture's protocol. The cDNA was synthesized using Prime Script RT master mix (TaKaRa Bio Inc.). To confirm expression of AhR and β2-AR, the primers shown in Table 1 were used. Quantitative real-time RT-PCR analysis of myostatin, atrogin-1, MuRF-1, and GAPDH was performed in a CFX96 System (Bio-Rad Laboratories. Inc., USA) with THUNDERBIRD SYBR QPS-201 (Toyobo Co. Ltd., Japan). The primers used are shown in Myotube length and diameter measurement. C2C12 myoblast cells were seeded in a 6-well or 12-well plate and cultured in a differentiation medium as described above. When cells reached full differentiation after 5 days, the cultured medium was replaced with DMEM containing 4% bovine serum albumin (either with or without IS, clenbuterol (1 μM), L-ascorbic acid (100 μM) and N-Acetyl-L-cysteine (1 mM)) and cultured for 1 or 2 more days. The medium with IS and reagents were changed every day. Cell morphology was observed with microscopy (Olympus, Inc.). Myotube length and diameters were measured in at least 100 to 200 myotubes using ImageJ software (National Institute of Health (NIH), Bethesda, MD, USA).

ROS accumulation assay.
For cultured cells, CM-H2DCFDA (C6827, Thermo Fisher Scientific) fluorescent dye was utilized, mainly to detect hydrogen peroxide. C2C12 myoblast cells were seeded in a 6-well or 12-well plate and cultured as described above. After becoming fully differentiated, myotubes were co-incubated with 4% albumin and IS (1 mM) in the presence or absence of clenbuterol (1 μM), L-ascorbic acid (100 μM) and N-Acetyl-L-cysteine (1 mM) for 24 h. After treatment, myotubes were washed with serum-free DMEM and incubated with CM-H2DCFDA (1 μM) conjugated serum-free DMEM for 30 min at 37 ℃ under 5% CO2. After removing the medium, myotubes were washed with PBS twice. Then, image acquisition was performed with an inverted fluorescence microscope, BZ-X710 (Keyence, USA). The intensity of CM-H2DCFDA staining was quantitated by ImageJ software (NIH, Bethesda, USA). For animal tissue specimens, immunofluorescent staining with anti-4-HNE antibody (20 μg/mL; JaICA, Shizuoka, Japan) was utilized to detect lipid peroxidation derived from ROS generation. The overall methods are the same as in the muscle fiber-type analysis described above, but we changed the thickness of the GC muscle cryosections (10 μm-thick slices) and used Alexa Fluor 488 anti-mouse IgG antibody (1:2,000) as a secondary antibody. Image acquisition was performed with an inverted fluorescence microscope, BZ-X710 (Keyence, USA). The intensity of 4-HNE staining was quantitated by ImageJ software (NIH, Bethesda, USA).
MitoTracker Red staining. For cultured cells, MitoTracker Red (M7521, Thermo Fisher Scientific) fluorescent dye to detect viable mitochondrial membrane potential was utilized. C2C12 myoblast cells were seeded in a 12-well plate and changed to a differentiation medium as described above. After 5 days, having reached full differentiation, the cultured medium was changed to DMEM containing 4% bovine serum albumin. Differentiated myotubes were incubated with or without clenbuterol and L-ascorbic acid for 1 h, and then incubated with or without IS for another hour. Myotubes were washed with serum-free DMEM and incubated with the MitoTracker Red dye conjugated serum-free DMEM for 30 min at 37 ℃ under 5% CO2. After removing the www.nature.com/scientificreports/ medium, the myotubes were fixed with 4% paraformaldehyde for 5 min, and then washed with PBS twice. Images were acquired with an inverted fluorescence microscope, BZ-X710 (Keyence, USA) and quantitated by ImageJ software (NIH, Bethesda, USA).

Statistical analysis.
All data were expressed with the means and error bars representing SEM. Statistical analysis was performed using a two-tailed Student t-test. A Cochran-Armitage trend test was performed to examine the change in the frequency distribution of the cross-sectional area of muscle fiber. A p-value < 0.05 was considered statistically significant.
Ethics statement. All animal experiments were approved by the ethics committee of the University of Tokyo Graduate School of Medicine and performed following the guidelines for the care and use of laboratory animals.