miR-23a is decreased during muscle atrophy by a mechanism that includes calcineurin signaling and exosome-mediated export

Abstract

Skeletal muscle atrophy is prevalent in chronic diseases, and microRNAs (miRs) may play a key role in the wasting process. miR-23a was previously shown to inhibit the expression of atrogin-1 and muscle RING-finger protein-1 (MuRF1) in muscle. It also was reported to be regulated by cytoplasmic nuclear factor of activated T cells 3 (NFATc3) in cardiomyocytes. The objective of this study was to determine if miR-23a is regulated during muscle atrophy and to evaluate the relationship between calcineurin (Cn)/NFAT signaling and miR-23a expression in skeletal muscle cells during atrophy. miR-23a was decreased in the gastrocnemius of rats with acute streptozotocin-induced diabetes, a condition known to increase atrogin-1 and MuRF1 expression and cause atrophy. Treatment of C2C12 myotubes with dexamethasone (Dex) for 48 h also reduced miR-23a as well as RCAN1.4 mRNA, which is transcriptionally regulated by NFAT. NFATc3 nuclear localization and the amount of miR-23a decreased rapidly within 1 h of Dex administration, suggesting a link between Cn signaling and miR-23a. The level of miR-23a was lower in primary myotubes from mice lacking the α- or β-isoform of the CnA catalytic subunit than wild-type mice. Dex did not further suppress miR-23a in myotubes from Cn-deficient mice. Overexpression of CnAβ in C2C12 myotubes prevented Dex-induced suppression of miR-23a. Finally, miR-23a was present in exosomes isolated from the media of C2C12 myotubes, and Dex increased its exosomal abundance. Dex did not alter the number of exosomes released into the media. We conclude that atrophy-inducing conditions downregulate miR-23a in muscle by mechanisms involving attenuated Cn/NFAT signaling and selective packaging into exosomes.

skeletal muscle atrophy is a frequent consequence of catabolic conditions (e.g., mechanical ventilation, diabetes, chronic kidney disease, and glucocorticoid-induced insulin resistance) that decreases the quality of life for patients and increases their risk of mortality (13, 15, 25, 34). Muscle atrophy typically results from a coordinated program of changes in gene and protein expression that leads to increased activity of intracellular proteolytic systems, including the ubiquitin-proteasome pathway and autophagy (19, 20, 37). For example, expression of two muscle-specific E3 ligases, muscle atrophy F box (MAFbx)/atrogin-1 and muscle-specific RING finger-1 (MuRF1), is upregulated in multiple models of muscle atrophy (29, 31, 37). Knockout of atrogin-1 or MuRF1 in skeletal muscle attenuated muscle atrophy induced by various conditions, indicating a central role for the ubiquitin-proteasome system in muscle wasting (3, 10).

Recent studies demonstrated that microRNAs (miRs) can have a key regulatory role in protein expression in pathological states, including skeletal muscle atrophy (18, 43, 44). microRNAs are a class of short, noncoding RNAs that inhibit the translation and/or promote the degradation of specific mRNAs (2). An overview of microRNA biogenesis and transcription can be found in several recent reviews (21, 46, 47). Little is known about how specific microRNAs are regulated, especially under different physiological conditions.

miR-23a was recently shown to decrease atrogin-1 and MuRF1 expression in skeletal muscle by inhibiting the translation of their mRNAs (43). In the same study, overexpression of miR-23a protected muscles from atrophy in vitro and in vivo; however, the level of endogenous miR-23a was not evaluated. In other studies, the NFATc3 transcription factor, which is activated by the serine/threonine phosphatase calcineurin (Cn), induced miR-23a expression, implicating Cn as an upstream regulator (24). Consistent with these findings, Allen and Loh (11) reported that the miR-23a-27a-24-2 microRNA cluster from which miR-23a originates is responsive to Cn signaling in C2C12 myotubes. Overexpression of nuclear factor of activated T cells (NFATc3) or constitutively active Cn increased miR-23a reporter activity in C2C12 myotubes (43); however, for unexplained reasons, overexpression of constitutively active Cn in C2C12 myotubes did not increase miR-23a expression (43). Our group previously found that Cn/NFAT signaling was reduced in skeletal muscle during acute streptozotocin (STZ)-induced diabetes, a condition associated with increased atrogin-1 and MuRF1 expression and muscle wasting (5, 8, 10). In light of these previous findings, the purpose of this study was to investigate if miR-23a expression is decreased during atrophy and to evaluate the mechanisms responsible for downregulation of miR-23a expression in muscle cells.¹

METHODS

Rat model of acute STZ-induced diabetes.

All animal studies were approved by the Emory University Institutional Animal Care and Use Committee. Male Sprague-Dawley rats weighing 250–300 g were anesthetized using isoflurane and given a single tail-vein injection of STZ (125 mg/kg body wt; Sigma-Aldrich, St. Louis, MO) prepared fresh in 0.1 M sodium citrate buffer (pH 4.0); control rats were injected with vehicle alone. All animals were fed a standard diet ad libitum for 3 days. At the time of euthanization, animals were anesthetized, and the gastrocnemius muscles were removed, immediately frozen in liquid nitrogen, and stored at −80°C. Arterial blood was collected for glucose measurements. Average blood glucose was 79.7 ± 3.4 and 406.5 ± 25.8 (SE) mg/dl for control and diabetic rats, respectively.

Cultured myotube model.

Mouse C2C12 or rat L6 myoblasts (American Type Culture Collection, Manassas, VA) were cultured in growth medium [DMEM + 10% fetal bovine serum (Atlanta Biologicals, Lawrenceville, GA)] with 1% penicillin-streptomycin (Invitrogen, Carlsbad, CA); all experiments were conducted using cells from passages 3–7. Myoblasts were grown to ∼95% confluence in a six-well plate and then induced to differentiate into myotubes by replacement of growth medium with differentiation medium (DMEM supplemented with 2% horse serum + 1% penicillin-streptomycin) for 3 days. For experiments involving glucocorticoids, water-soluble dexamethasone (Dex; Sigma-Aldrich) was added (final concentration 0.1–1 μM) to myotubes after 3 days in differentiation medium. When myotubes were treated with Dex for 48 h, the same concentration of Dex was reintroduced after 24 h. For other experiments, 10 nM PMA and 300 nM thapsigargin (prepared in DMSO; Sigma-Aldrich) were added to myotubes to activate Cn; 2.5 μM cyclosporine A (CsA; Sigma-Aldrich) was used to inhibit Cn. DMSO was used as a vehicle control for control cell treatments.

CnAβ adenoviral infection of myotubes.

C2C12 myotubes were infected with adenoviruses as described previously (49). Briefly, differentiated cells were infected with an adenovirus that encodes green fluorescent protein (AdGFP) as a control or a FLAG-tagged CnAβ (AdcaCnAβ; Seven Hills Bioreagents, Cincinnati, OH) using a multiplicity of infection ≤50. After 48 h, infection was confirmed by Western blot analysis using anti-FLAG antibody, as well as visualization of GFP in control cells.

Primary muscle cell culture.

Primary myoblasts were isolated from the hindlimb muscles of late-juvenile, mixed-background wild-type (WT) mice, mice lacking the α-isoform of the CnA catalytic subunit (CnAα^−/−), and mice lacking the β-isoform of the CnA catalytic subunit (CnAβ^−/−), as described elsewhere (17). CnAα^−/− mice were created by Dr. J. Seidman (Howard Hughes Medical Institute, Harvard Medical School, Boston, MA) (48), while CnAβ^−/− mice were created by Dr. J. Molkentin (Cincinnati Children's Hospital, Cincinnati, OH) (4, 45). Hindlimb muscles from WT, CnAα^−/−, and CnAβ^−/− mice were provided for primary myoblast isolation by Dr. Jennifer Gooch. For experiments, myoblasts were cultured in growth medium (Ham's F-10 medium, 20% fetal bovine serum, 5 ng/ml basic fibroblast growth factor, and 1% penicillin-streptomycin) on collagen-coated plates until ∼60–70% confluence. To induce differentiation to myotubes, growth medium was replaced with differentiation medium (DMEM with 5.5 mM glucose supplemented with 2% horse serum and 1% penicillin-streptomycin) for 3 days.

Cultured myotube medium, exosome quantification, and exosome microRNA analysis.

C2C12 myoblasts were cultured as described above. After 3 days in differentiation medium, cells were washed briefly with serum-free medium and then incubated in serum-free medium (control) with or without 1 μM Dex for 6 h. After 6 h the medium was collected. Quantification of exosomes was performed using a previously described protocol (9, 41). Briefly, medium was centrifuged at 800 g for 10 min to pellet intact cells and debris. The supernatant was centrifuged at 16,000 g for 60 min to pellet microvesicles, membrane particles, ectosomes, and apoptotic vesicles, which are larger than exosomes. The pellet was discarded, and the supernatant was centrifuged at 110,000 g for 1 h to pellet exosomes. The pellet was resuspended in sterile PBS, and the number of exosomes was quantified by flow cytometry using 10-μm Flow-Count fluorescent beads (16).

To quantify exosomal microRNA, exosome-associated RNA, including microRNA, was isolated using a urine exosome RNA isolation kit (Norgen Biotek, Thorolod, ON, Canada) according to the manufacturer's instructions for isolating RNA from extracellular fluids. Subsequent quantitative PCR (qPCR) was performed as described below.

RNA isolation and quantitation.

Total RNA (including microRNAs) was extracted from cells using the Total RNA isolation kit (Norgen Biotek) according to the manufacturer's instructions. Total RNA (enriched in small RNAs, including microRNAs) was isolated from rat gastrocnemius muscle using the Animal Tissue RNA purification kit (Norgen Biotek) according to the manufacturer's instructions. Total RNA was reverse-transcribed using the NCode VILO miRNA cDNA synthesis kit (Invitrogen). For miR-23a and miR-1 analysis, cDNA was used as a template for the qPCR using a miR-23a- or miR-1-specific LNA PCR forward primer (Exiqon, Woburn, MA) and a universal reverse PCR primer provided in the NCode VILO miRNA cDNA synthesis kit. U6 small nuclear RNA was used as a normalization control for miR-23a and miR-1. For mRNA analysis, cDNA was used as the qPCR template with target-specific primer sets for RCAN1.4 [formally called MCIP1.4 (7)], as previously published (45). 18S rRNA was used to normalize RCAN1.4 mRNA measurements. All qPCR experiments were performed using the iCycler and iQ SYBR Green (Bio-Rad, Hercules, CA) with the following cycling parameters: 95°C for 2 min and 40 cycles at 95°C for 15 s and 60°C for 1 min. The data were analyzed for fold change (ΔΔC_t) using the iCycler software. Melting curve analyses were performed to analyze and verify the specificity of the reaction.

Western blot analysis.

Cytosolic and nuclear extracts were prepared as previously described (38). Briefly, differentiated L6 myotubes were washed with PBS, centrifuged at 1,500 g for 5 min, resuspended in a solution containing 0.01 M HEPES, pH 7.6, 1.5 mM MgCl₂, 2 mM KCl, and protease inhibitor cocktail (Roche, Indianapolis, IN), and kept on ice for 15 min. NP-40 was added to a final concentration of 0.5%, and the sample was vortexed immediately and centrifuged (14,000 rpm) for 1 min at 4°C. The supernatant (cytosolic fraction) was transferred to a new tube and stored at −80°C until further analysis. The pellet was resuspended in a solution containing 0.02 mM HEPES (pH 7.6–7.8), 1.5 mM MgCl₂, 2 mM KCl, 0.4 M NaCl, 0.2 mM EDTA, 0.5 mM DTT, 0.5 mM PMSF, 10 μg/ml aprotinin, 5 μg/ml leupeptin, 5 μg/ml pepstatin A, 1 mM sodium orthovanadate, and 25% glycerol, and samples were incubated on a shaking platform for 15–60 min at 4°C. Samples were centrifuged at 21,000 g for 20 min at 4°C, and the supernatant (nuclear fraction) was transferred to a new tube and stored at −80°C until further analysis. Protein concentrations were measured using a DC protein assay kit (Bio-Rad). Cytosolic and nuclear extract proteins were separated by polyacrylamide gel electrophoresis, transferred to a nitrocellulose membrane, and stained with Ponceau S to ensure equal loading. Thereafter, membranes were blocked and incubated with primary antibodies directed against the protein of interest. Primary antibodies included NFATc3 (Abcam, Cambridge, MA), pan Cn (catalog no. AB1695, Chemicon), histone H1 (Santa Cruz Biotechnology, Santa Cruz, CA), and GAPDH (Santa Cruz Biotechnology). Membranes were processed for enhanced chemiluminescence with the Immun-Star WesternC chemiluminescence kit (Bio-Rad) and exposed to film. Images were analyzed using National Institutes of Health ImageJ software.

Statistical analysis.

Group sample size was determined by a power analysis of preliminary data from our laboratory. A t-test was used to compare two groups. For more than two groups, comparisons were made by one-way ANOVA. When appropriate, a post hoc Tukey's honestly significant difference test was performed. Significance was established at P ≤ 0.05. At least three to five samples per treatment, acquired from at least three independent experiments, were quantified and analyzed for each outcome.

RESULTS

miR-23a and Cn/NFAT signaling are suppressed during muscle atrophy.

To determine if the level of miR-23a in skeletal muscle is changed during atrophy, we measured miR-23a in the gastrocnemius of rats with acute STZ-induced diabetes, a model of muscle atrophy that has been extensively characterized (26, 30, 32). miR-23a was significantly reduced by diabetes relative to control rats (Fig. 1A). Since glucocorticoids are required for diabetes-induced muscle atrophy (26), we next tested whether Dex treatment of C2C12 myotubes mimics the effects of diabetes on miR-23a. Dex induced a dramatic decrease in the level of miR-23a within 1 h of addition, and the reduction was sustained for ≥48 h (Fig. 1B). In contrast, a muscle-specific microRNA, miR-1, was increased after 6 h of Dex treatment (Fig. 1C). The latter result is similar to one recently reported by Shen et al. (40), who found that miR-1 was increased in C2C12 myotubes treated with a 500-fold-higher dose of Dex for 48 h.

Fig. 1.

MicroRNA-23a (miR-23a) and calcineurin (Cn) signaling are decreased by atrophy-inducing conditions. A: level of miR-23a in rat gastrocnemius 3 days after streptozotocin (STZ) administration to induce diabetes and muscle atrophy (n = 5). *P < 0.05 vs. control (Con). B: time course of changes in miR-23a after addition of 100 nM dexamethasone (Dex) to C2C12 myotubes. *P < 0.05 vs. Con. C: 1 μM Dex for 6 h increases intracellular miR-1 in C2C12 myotubes (n = 3). *P < 0.05 vs. Con. D: effect of 100 nM Dex for 48 h on RCAN1.4 mRNA in C2C12 myotubes (n = 3). *P < 0.05 vs. Con. E: effect of 100 nM Dex for 48 h on amount of nuclear Cn protein in C2C12 myotubes (n = 3). *P < 0.05 vs. Con. F: time course of changes in amount of cytoplasmic nuclear factor of activated T cells 3 (NFATc3) protein after addition of 100 nM Dex to C2C12 myotubes (n = 4). *P < 0.05 vs. 0 min. #P < 0.05 vs. 5 min. Values are means ± SE.

A recent study examining cardiomyocytes proposed a link between miR-23a and NFAT signaling (24). Previously, we demonstrated that Cn/NFAT signaling is suppressed in muscle during acute STZ-induced diabetes (34). Consistent with these observations, treatment of L6 myotubes with Dex (48 h) induced a decrease in the levels of RCAN1.4 mRNA (Fig. 1D), a transcriptional target of NFAT, and nuclear Cn catalytic A subunit protein (Fig. 1E). Furthermore, nuclear localization of NFATc3 was diminished within 1 h of Dex addition, indicating that Dex rapidly suppresses Cn signaling (Fig. 1F). These data demonstrate that miR-23a is reduced in muscle cells during atrophy and establish a correlational link between miR-23a expression and Cn/NFAT signaling.

Cn/NFAT regulates miR-23a expression.

To explore a causal link between Cn and miR-23a expression in muscle cells, we next utilized a combination of PMA and the Ca²⁺-ATPase inhibitor thapsigargin to increase cytosolic Ca²⁺ and activate Cn/NFAT signaling (9). Treatment of C2C12 myotubes with PMA-thapsigargin for 4 h increased RCAN1.4 mRNA (Fig. 2A). The increase was blocked by cotreatment with the Cn inhibitor CsA, thus verifying a requirement of Cn for the response (Fig. 2A). Similarly, PMA-thapsigargin treatment increased the level of miR-23a, while CsA attenuated the increase, indicating a direct link between Cn signaling and miR-23a (Fig. 2B). Next, we attempted to confirm this link by overexpressing a constitutively active form of Cn. Notably, Wada et al. (43) reported that overexpression of constitutively active CnAα did not increase miR-23a expression. Since we and others (14, 33) reported that NFAT is regulated by CnAβ, the other CnA catalytic subunit isoform in skeletal muscle, we overexpressed a constitutively active form of CnAβ by adenoviral infection (CnAβ AV) in C2C12 cells to test how it affected miR-23a. Similar to the findings of Wada et al., CnAβ AV did not increase the level of miR-23a; however, it did prevent the Dex-induced suppression of miR-23a (Fig. 2C). Although the reasons why activated CnAα and CnAβ do not increase miR-23a expression remain perplexing, the majority of our data still implicates Cn/NFAT signaling in the regulation of miR-23a.

Fig. 2.

Effect of Cn/NFAT signaling on miR-23a. A and B: levels of RCAN1.4 mRNA and miR-23a in C2C12 myotubes treated with 10 nM PMA-300 nM thapsigargin (PMA/Thap) in the presence or absence of cyclosporine A (CsA) for 4 h (n = 3). *P < 0.05 vs. Con. #P < 0.05 vs. PMA/Thap + CsA. C: miR-23a in C2C12 myotubes infected with adenoviruses to express green fluorescent protein (GFP; Con) or FLAG-tagged CnAβ [adenoviral CnAβ (CnAβ AV)] in the presence or absence of 100 nM Dex for the indicated times. *P < 0.05 vs. Con. Values are means ± SE. D: GFP fluorescence in C2C12 myotubes infected with adenovirus to express GFP. E: immunoblot analysis of FLAG-tagged protein in cells infected with adenovirus to express constitutively active FLAG-CnAβ and Ponceau S staining (n = 3).

Since ectopic expression of activated CnA did not stimulate miR-23a expression, we used a different approach: we tested whether a reduction of Cn activity via genetic knockout affects miR-23a expression. The level of miR-23a was lower in primary myotubes from CnAα^−/− and CnAβ^−/− than WT mice (Fig. 3). Moreover, the level of miR-23a in CnAα^−/− and CnAβ^−/− myotubes was not decreased further by Dex treatment. These data directly link the level of miR-23a and its regulation by Dex to the presence of Cn in muscle cells.

Fig. 3.

miR-23a is suppressed in primary myotubes from CnAα^−/− and CnAβ^−/− mice compared with wild-type (WT) mice; 100 nM Dex for 48 h did not further suppress the level of miR-23a. Values are means ± SE (n = 5). *P < 0.05 vs. WT.

miR-23a and exosomes.

miRNAs are reported to be very stable in tissues and biological fluids, even under conditions such as extremes in temperature and pH or RNase activity (27, 39). One mechanism that is thought to protect microRNAs from degradation is incorporation into exosomes (42). Since miR-23a was rapidly depleted from myotubes treated with Dex (Fig. 1B), we posited that miR-23a might be incorporated into exosomes that are released from the myotubes. This would represent a potential second mechanism by which the level of miR-23a is regulated in muscle cells. To test this hypothesis, we compared the level of miR-23a in exosomes in serum-free media of C2C12 myotubes treated with or without Dex for 6 h. The amount of exosomal miR-23a was increased by Dex treatment (Fig. 4A). Importantly, the amount of exosomal U6 small nuclear RNA was not different between treatment groups (data not shown). To verify that the exosomes originated from myotubes, we also measured miR-1 and found that it also was increased in exosomes in media from cells treated with Dex (Fig. 4B). Finally, we quantified the number of exosomes in the media from control and Dex-treated cells and found no difference between treatments (Fig. 4C). These findings suggest that Dex treatment increased the loading of miR-23a into exosomes.

Fig. 4.

Dex increases amount of miR-23a in C2C12 myotube exosomes. A and B: amounts of miR-23a and miR-1 in exosome preparations isolated from media of C2C12 cells treated with serum-free medium (Con) or serum-free medium + 1 μM Dex for 6 h (n = 3). C: Dex does not increase number of exosomes released into the medium (n = 3). Values are means ± SE; *P < 0.05 vs. Con.

DISCUSSION

Muscle atrophy is a debilitating consequence of a variety of diseases and pathological conditions that accelerate the rate of protein degradation. Recent studies demonstrate that microRNAs can play an important role in the regulation of specific target mRNAs and proteins in many human diseases, including muscle disorders (6, 35, 36). In regard to muscle atrophy, a glucocorticoid-induced reduction of miR-23a could be important, because miR-23a targets atrogin-1 and MuRF1, two atrophy-related, muscle-specific E3 ubiquitin ligases (43). The response is likely to attenuate the capacity of miR-23a to inhibit atrogin-1 and MuRF1 translation, thereby providing a mechanism to rapidly upregulate the level of these proteins independently of the well-described FoxO-mediated transcriptional mechanism.

Our data indicate that glucocorticoids regulate the level of miR-23a in myotubes by a multifaceted mechanism. The expression of miR-23a is regulated by Cn/NFAT signaling. Although Cn-NFAT signaling alone may not be sufficient to increase miR-23a expression, it is necessary to maintain miR-23a expression in skeletal muscle. Others have documented that NFAT increases miR-23a expression in cardiomyocytes and that Cn and NFATc3 increase miR-23a reporter gene activity (43). Consistent with our previous finding that Cn/NFAT signaling is suppressed in muscle during STZ-induced diabetes (34), the level of miR-23a was decreased in diabetic rat muscle. In addition, pharmacological augmentation of Cn activity in C2C12 myotubes induced an increase in miR-23a that was blocked by CsA. In a third experiment, we found that deficiency of CnAα or CnAβ in primary myotubes results in a lower steady-state level of miR-23a. Importantly, the level of miR-23a in CnAα^−/− and CnAβ^−/− primary muscle cells was not further suppressed by Dex, although overexpression of CnAβ did prevent the Dex-induced decrease in miR-23a. These studies establish a direct link between Cn signaling and miR-23a.

Curiously, overexpression of CnAβ in myotubes did not increase miR-23a but was sufficient to prevent the Dex-induced reduction of miR-23a. A similar finding was noted by Wada et al. (43), who overexpressed a constitutively active form of CnAα. Others also have reported disparate responses to manipulations of Cn signaling in skeletal muscle, depending on the experimental strategy and outcomes measured (reviewed in Ref. 12). For example, transgenic overexpression of activated Cn does not cause skeletal muscle hypertrophy, whereas stimulation of Cn signaling by IGF-I or β-adrenergic agonists induces muscle growth and overexpression of NFATc1 produces muscle hypertrophy (12). Thus a narrow range of Cn activity may be important for the regulation of miR-23a and muscle mass.

A second aspect of the regulatory mechanism that controls the amount of miR-23a during muscle atrophy involves exosomes. Exosomes are vesicles that are released from cells via secretion by endocytic compartments or by pinching off from the plasma membrane into the extracellular environment (41). In 2007, Valadi et al. (42) first proposed that exosome-mediated transfer of microRNAs represents a novel exchange system between cells. Recently, exosomes were shown to shuttle microRNAs between dendritic cells, and the exchanged microRNAs remained functional because they repressed target mRNAs in acceptor cells (28). In 2010, it was reported that C2C12 myotubes release exosomes, but there was no attempt to measure or identify the encapsulated microRNAs (11). In the present study, we demonstrate for the first time that glucocorticoid administration rapidly induces the release of exosomes highly enriched with microRNAs (i.e., miR-23a and miR-1) from C2C12 myotubes. Additionally, glucocorticoid administration does not alter the number of exosomes released into the medium, rather it appears to alter the loading of microRNAs into exosomes. This conclusion is based on several observations. First, the intracellular level of miR-23a is decreased over a time frame when exosomal miR-23a is increased. In contrast, the intracellular and exosomal levels of miR-1 increased in parallel. The Dex-induced increase in intracellular miR-1 is consistent with a recent report by Shen et. al (40). Thus the incorporation of microRNAs into exosomes appears to be a very selective and tightly regulated process. While future studies are needed to fully investigate this process, these findings identify a mechanism by which muscle cells can rapidly reduce the level of specific intracellular microRNAs. It may also represent a novel way for skeletal muscle to signal other tissues during muscle atrophy and other pathophysiological conditions.

In summary, our results demonstrate that the downregulation of miR-23a in muscle cells during glucocorticoid-induced atrophy results from a combination of decreased expression via reduced Cn/NFAT signaling and an increase in release of miR-23a via exosomes (Fig. 5). Since miR-23a acts to inhibit the translation of atrogin-1 and MuRF1, a rapid exosome-mediated decrease in miR-23a should facilitate an increase in expression of these key atrophy-inducing proteins while the FoxO transcription factors are being activated to increase the transcription of these and other atrophy-related genes. A consequence of increased atrogin-1 expression should be a reduction of Cn protein, because atrogin-1 facilitates the ubiquitination and degradation of Cn in cardiomyocytes (22, 23). Consistent with this prediction, Cn protein was reduced in Dex-treated C2C12 myotubes (Fig. 1D) and diabetic rat muscle (34). The resulting decrease in Cn/NFAT signaling should then contribute to the suppression of miR-23a. Together, these actions establish a potential feed-forward system that can sustain an attenuated level of miR-23a for as long as atrophy-inducing conditions persist (Fig. 5). The responses represent new mechanisms by which muscle cells regulate their size and respond to environmental signals. Furthermore, our results identify a new and novel role for glucocorticoids in the atrophy process.

Fig. 5.

Overview of regulation of miR-23a during muscle atrophy. Solid lines represent previously established signaling pathways or interactions identified by the current experiments. Dotted lines represent potential unverified responses. Black lines signify activation; red lines signify inhibition. MuRF, muscle-specific RING finger protein.

GRANTS

This work was supported by National Institutes of Health Grants T32 DK-007656 (to M. B. Hudson) and R01 DK-007656 and HL-109559 (to C. D. Searles), American Heart Association Grant 7660020 and VA Merit Award X01 BX001456 (to S. R. Price), and VA Merit Award I01 BX000704 (to C. D. Searles).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

M.B.H., M.E.W.-H., B.Z., J.A.R., M.A.B., J.L.G., C.D.S., and S.R.P. are responsible for conception and design of the research; M.B.H., M.E.W.-H., B.Z., J.A.R., M.A.B., J.L.G., and C.D.S. performed the experiments; M.B.H., B.Z., J.A.R., C.D.S., and S.R.P. analyzed the data; M.B.H., M.E.W.-H., C.D.S., and S.R.P. interpreted the results of the experiments; M.B.H. prepared the figures; M.B.H., M.A.B., and S.R.P. drafted the manuscript; M.B.H., M.E.W.-H., B.Z., J.A.R., M.A.B., J.L.G., C.D.S., and S.R.P. edited and revised the manuscript; M.B.H., M.E.W.-H., B.Z., J.A.R., M.A.B., J.L.G., C.D.S., and S.R.P. approved the final version of the manuscript.