To examine the relationship between FoxO1 and Atrogin-1/MurF-1 expression, we generated mice with muscle-specific knockout of FoxO1 (MFKO). These mice grow normally and their muscles exhibit a normal distribution of myofiber sizes (Figure 1A and D), Using western blots and immunostaining, we confirmed that FoxO1 is absent in myofibers from MFKO mice (Figure 1A). Importantly, deletion of FoxO1 did not alter the expression of FoxO3a or of FoxO4 (Figure 1B) and when CKD was induced in MFKO mice, there was no significant loss of muscle mass in tibialis anterior (TA), extensor digitorum longus (EDL) or soleus (Sol) muscles. In contrast, CKD did decrease the weights of muscles from lox/lox, control mice (Figure 1 C). These changes were confirmed by an analysis of cross-sectional areas of myofibers in TA muscles (Figure 1, E).

To explore why CKD did not induce muscle atrophy in MFKO mice, we measured the protein synthesis and degradation rates EDL and Sol muscles. CKD did not significantly suppress protein synthesis in muscles from lox/lox or MFKO mice when compared to values in muscles of the respective control mice (Figure 2A). In contrast, CKD stimulated protein degradation in muscles of lox/lox mice but not in muscles of MFKO mice (Figure 2B). In TA and Sol muscles of lox/lox mice, CKD also induced the expression Atrogin-1/MAFbx and MuRF-1 (Figure 2C and Supplemental Figure 1B). As with protein degradation, the expression of these E3 ligases was suppressed in mice with muscle-specific FoxO1 knockout. Importantly, these responses occurred even though the muscle levels of p-Akt or the presence of FoxO3a and FoxO4 did not change (Figure 2D).

In another model of muscle catabolism, we treated MFKO mice with a pharmacologic dose of dexamethasone (Dex, 5mg/kg/day) [²³]. After 2-weeks, the distribution of myofiber sizes in TA muscles of lox/lox mice were shifted to the left. This response was largely blocked in muscle of MFKO mice (Supplemental Figure 2A). In muscles of lox/lox mice, Dex elicited the expected increase in Atrogin-1/MAFbx and MurF-1 mRNAs. These catabolic responses were largely suppressed in muscles of MFKO mice (Supplemental Figure 2B). Thus, FoxO1 is a dominant transcription factor acting to stimulate muscle atrophy in at least two catabolic conditions (CKD or high dose of glucocorticoid administration).

Because PTEN [²] and FoxO1 (Figure 1) influence muscle protein wasting, we hypothesized that enhancing miR-486 in muscle might improve muscle protein turnover. To examine this hypothesis, we first introduced a miR-486 “mimic” into primary cultures of mouse skeletal muscle cells and measured the rates of protein synthesis and degradation. In transfected primary culture of myotubes, the miR-486 mimic increased more than 12-fold compared to cells transfected with control miRNA (Figure 3A). Following treatment with Dex, the miR-486 mimic significantly dampened proteolysis (Figure 3B). Dex also suppressed protein synthesis in cultured myotubes and transfection of these cells with miR-486 mimic tended to counteract this response (Figure 3C). These results demonstrate that enhancing miR-486 can suppress Dex-stimulated muscle protein degradation.

Since the E3 ubiquitin ligases, Atrogin-1/MAFbx and MurF-1, are intimately involved in the UPS-mediated muscle proteolysis, we examined whether enhancing miR-486 in muscle cells would suppress expression of these E3 ubiquitin ligases. In primary cultures of muscle cells, Dex produced the expected increase in Atrogin-1/MAFbx and MurF-1 mRNAs in myotubes. The increase in these E3 ubiquitin ligases was suppressed by the miR-486 mimic (Figure 4A). Likewise, introduction of the miR-486 mimic blocked the suppression of Akt phosphorylation induced by Dex (Figure 4B). Because miR-486 can target the 3′-UTR of PTEN [²²], we measured the PTEN protein in myotubes. The miR-486 mimic significantly decreased the level of PTEN protein (Figure 4C). A decrease in PTEN induced by the miR-486 mimic would raise p-Akt by inhibiting PTEN expression and this response in turn would be a mechanism that suppresses Atrogin-1/MAFbx expression.

The downstream target of p-Akt, FoxO1 has been shown to influence MuRF-1 transcription by binding to MuRF-1 promoter or acting synergistically with the glucocorticoid receptor [²⁴]. We found that Dex not only decreased the levels of p-FoxO1 but also up-regulated FoxO1 in muscle cells; transfection with the miR-486 mimic reversed these responses (Figure 4D). The results confirm that FoxO1 and PTEN are targets of miR-486 in skeletal muscle cells. They also demonstrate that miR-486 can suppress both Atrogin-1/MAFbx and MuRF-1 expression by repressing FoxO1.

We examined if CKD influences miR-486 expression in muscle and found significant decrease in miR-486 in muscles of CKD mice (Supplemental Figure 3A). Because our in vitro experiments showed that miR-486 suppresses muscle protein degradation by enhancing PI3K/Akt signaling, we hypothesized that enhancing miR-486 might block muscle wasting in CKD mice. To test this possibility, we transfected either the miR-486 mimic or a control miRNA into TA muscles of normal and CKD mice using electroporation. First we validated this process by electroporating Dy547-labeled C. elegans microRNA (a control miRNA) into TA muscles and examined the fluorescence intensity in muscle fibers. As shown in Figure 5, few of the myofibers exhibited fluorescence without electroporation. With electroporation, however, the fluorescence of Dy547-labeled control miRNA was present in more than 70% of myofibers, the expression of this control miRNA lasted at least 3 weeks as assessed by real-time PCR (Figure 5F, bar graph). We also detected a >10-fold increase in miR-486 in TA muscles at two weeks after electroporation of the miR-486 mimic (Figure 6A). As with the control miRNA, the miR-486 mimic was detected in > 70% of myofibers by in situ hybridization (Figure 6B and Supplemental Figure 3B). Importantly, muscle mass (assessed as the ratio of TA muscle weight to tibia bone length) in CKD mice was significantly increased when muscles were electroporated with the miR-486 mimic (Figure 6C). This result was confirmed by finding a rightward shift in the distribution of the cross-sectional areas of myofibers when compared to results in muscles electroporated with the control miRNA (Figure 6D).

The expression of both Atrogin-1/MAFbx and MurF-1 also decreased in muscles electroporated with the miR-486 mimic (Figure 7A). To assess how it might br related to PI3K/Akt signaling in skeletal muscle, we examined FoxO1 and PTEN proteins in TA muscles electroporated with the miR-486 mimic. There was a decrease in both FoxO1 and PTEN proteins compared to results in muscles treated with control miRNA (Figure 7B, C). This was accompanied by phosphorylation of Akt at the serine 473 site plus the expected increase in phosphorylation of FoxO1 (Figure 7C, D). We also analyzed the expression of miR-1, miR-133, miR-29 and 206, because these muscle-enriched miRNAs are reported to affect skeletal muscle growth [²⁰;²⁵]. Expression of these miRNAs did not significantly change when the miR-486 mimic was electroporated into muscle compared to results obtained in TA muscles electroporated with control miRNA (Figure 7E). Thus, the improvement in muscle mass in CKD mice following treatment with miR-486 mimic was mainly due to an enhanced level of miR-486.

Mice were housed with a 12-h light-dark cycles and all animal procedures were approved by the Baylor College of Medicine Institutional Animal Care and Use Committee. MFKO mice were obtained by crossing loxp flanked FoxO1 mice (kindly provided by Dr SY Chen, Baylor College of Medicine) with MCK-Cre mice (Jackson lab, Bar Harbor, ME), both strains are on FVS background. Lox+/+ plus Cre+ mice were designated as MFKO; lox+/+ with Cre-mice severed as controls. CKD was induced in 3 months old MFKO or lox/lox control male mice by 2 stage nephrectomy under anesthesia [³⁵]. Briefly, the left kidney was removed and 1 week later ~70% of the right kidney was removed and homeostasis was achieved by tissue adhesive (3M, St. Paul, MN) to prevet bleeding. After one week of 24% dietary protein, mice were switched to a 40% protein diet (Harlan Teklab, Indianapolis, IN, USA) for at least 2-weeks. CKD was confirmed by serum creatinine (Fig S1A) and BUN (average 84±3.3 mg/dl in CKD groups). Serum creatinine level was determined using Cayman’s creatinine kit (Cayman chemical, Ann Arbor, MI). Sham-operated, control mice underwent surgery without damaging the kidneys and were fed the same diets. Electroporation was accomplished by injecting 0.5 nmol of microRNA in 20ul PBS into tibialis anterior (TA) muscle. Electroporation was performed at 80 volt for 10 pulses (100ms each pulse, 200 ms intervals) [²²]. The long-lasting miR-486 mimic and the control, C. elegans microRNA (cel-miR-67) labeled with fluorescent Dy547 were purchased form Dharmcon (Thermo Scientific, Chicago, IL). At 14-day after electroporation, mice were anesthetized and TA muscles as well as serum were collected for experiments or the samples were frozen at −80 °C.

Extensor digitorum longus (EDL) or soleus (Sol) muscles were maintained at resting length and incubated in Krebs-Henseleit bicarbonate buffer with 10 mM glucose as described [⁴²]. Protein synthesis in EDL muscles was measured as the rate of incorporation of L-[¹⁴C] phenylalanine while protein degradation was measured the release of tyrosine. These amino acids were used as neither is synthesized or degraded in muscle. For primary culture of myotubes, the rates of protein synthesis were measured in during a 16-h incubation with 3 uCi L-[(3,5)-³H] tyrosine (MP Biomedicals). Subsequently, myotubes were washed 3X with ice-cold PBS before adding 10% trichloroacetic acid (TCA) to precipitate proteins. After three additional PBS washings, the pellets were dissolved in 0.5 ml of 0.15 mol/l NaOH, and the incorporation of radiolabeled tyrosine and protein content (Bio-Rad DC protein assay kit; Bio-Rad, Hercules, CA) were measured. To measure protein degradation, myotubes were prelabeled with -[(3,5)-³H] tyrosine and incubated in DMEM media containing 2% horse serum. At different times, the TCA–soluble radioactivity released from proteins was measured to calculate the rate of protein degradation [⁴²].

Total RNA was extracted using TRIzol (Sigma) and precipitated in isopropanol for overnight. cDNAs were synthesized using MicroRNA cDNA kit ( Exiqon, Woburn, MA) . SYBR Green Real-time qPCR was performed with the CFX96 System (Bio-Rad Laboratories, Hercules, CA). All miRNA expressions were normalized to non-coding small nuclear RNA component of U6 (U6). The LNA™ miRNAs primers (miR-486, miR-1, miR-29a, miR-133, miR-206, cel-miR-67 and U6) were perchased from Exiqon. Northern blot analysis of Atrogin1/MaFbx and MuRF-1 was completed as described [²²].

To assess differences in cross-sectional areas of myofibers, 5-μm sections of TA muscles were stained with an anti-laminin antibody (Sigma-Aldrich) or anti-FoxO1 (Cell Signaling, Danvers, MA). The diameters of at least 300 myofibers per TA muscle were assessed using Image J software (National Institutes of Health, Frederick, MD). In situ hybridization was performed using a digoxigenin-labeled miR-486, LNA™ probe (Exiqon) incubated at 42°C overnight as described [⁴³]. After washing, the sections were incubated with anti-digoxigenin antibody for 16h and signals were visualized with nitro blue tetrazolium/5-bromo-4-chloro-3-indolyl phosphate (Roche Applied Science, Indianapolis, IN). A scrambled control probe LNA™ (Exiqon) was used as a negative control.

Satellites cells were isolated as described [²⁰;²⁸]. Myoblasts were maintained in F-10/DMEM with 10% FBS (HyClone, Logan, UT), penicillin (200 units/ml), and streptomycin (50 ug/ml) (Invitrogen, Carlsbad, CA). Myotubes were induced from myoblast by switching to DMEM plus 2% horse serum (Sigma) for 72 h. miR-486 mimic (0.01 nmol/well in 6 well dish) was transfected into myotubes with liporfectin 2000 (Invitrogen); Control miRNA (cel-miR-67 ) served as the control. After 36 h, rates of protein synthesis and degradation were measured as described [⁴²]. For western blotting, lysates of myotubes were prepared in RIPA buffer (20 mM Tris, pH 7.5, 5 mM EDTA, 150 mM NaCl, 1% NP40, 0.5% Na-deoxycholate, 0.025% SDS, 1 mMNa-orthovanadate, 10 mM NaF, 25 μM β-glycerophosphate) containing proteases inhibitor (Roche). Skeletal muscle lysates were prepared from ~50 mg muscle by homogenizing in 0.5 ml RIPA buffer. After centrifugation at 15,000xg for 15 min at 4°C, the supernatants were subjected to western blotting as described [²²].

Results are presented as mean ± SEM. Statistical analysis was performed by ANOVA followed by Tukey’s or Student-Newman-Keul’s tests. P < 0.05 was considered statistically significant. Experiments were repeated at least three times.