Loss of muscle mass and function occurs in various diseases. Myostatin blocking can attenuate muscle loss, but downstream signaling is not well known. Therefore, to elucidate associated signaling pathways, we used the soluble activin receptor IIb (sActRIIB-Fc) to block myostatin and activins in mice. Within 2 wk, the treatment rapidly increased muscle size as expected but decreased capillary density per area. sActRIIB-Fc increased muscle protein synthesis 1–2 days after the treatment correlating with enhanced mTORC1 signaling (phosphorylated rpS6 and S6K1, r = 0.8). Concurrently, increased REDD1 and eIF2Bε protein contents and phosphorylation of 4E-BP1 and AMPK was observed. In contrast, proangiogenic MAPK signaling and VEGF-A protein decreased. Hippo signaling has been characterized recently as a regulator of organ size and an important regulator of myogenesis in vitro. The phosphorylation of YAP (Yes-associated protein), a readout of activated Hippo signaling, increased after short- and longer-term myostatin and activin blocking and in exercised muscle. Moreover, dystrophic mdx mice had elevated phosphorylated and especially total YAP protein content. These results show that the blocking of myostatin and activins induce rapid skeletal muscle growth. This is associated with increased protein synthesis and mTORC1 signaling but decreased capillary density and proangiogenic signaling. It is also shown for the first time that Hippo signaling is activated in skeletal muscle after myostatin blocking and exercise and also in dystrophic muscle. This suggests that Hippo signaling may have a role in skeletal muscle in various circumstances.
- mammalian target of rapamycin complex 1
- mitogen-activated protein kinase
- muscle hypertrophy
- muscle dystrophy
adequate size and function of skeletal muscle is of paramount importance for health (4, 45, 68). Muscle size is regulated by myostatin and activins, which belong to the TGFβ superfamily of proteins (1, 31, 32). They are ligands of type IIb activin receptors (ActRIIB) that are expressed in skeletal muscle (23, 31). Strong myostatin and activin blocking can be accomplished by systemic injection of a soluble ligand-binding domain of ActRIIB fused to the Fc domain of IgG [soluble activin receptor IIb (sActRIIB-Fc)] (24, 32). sActRIIB-Fc treatment has increased muscle mass and function and has alleviated the disease phenotype in a number of mouse models for neuromuscular disorders (40, 45) and in cancer cachexia (68). However, a lack or inhibition of myostatin may decrease oxidative/aerobic capability of muscle (2, 37, 47, 48).
In healthy adult muscle, blocking of myostatin and other TGFβ family members increases protein synthesis (58, 61) and muscle size (18, 24, 27, 32, 45, 50, 61, 68). This occurs at least partially through the inhibition of the mechanistic target of rapamycin complex 1 (mTORC1) pathway (27, 50, 61). A recent microarray study suggested that decreased oxidative capacity occurs after blocking of myostatin and activins by sActRIIB-Fc (47), but no further signaling analysis was conducted. At the moment, the downstream signaling in the myostatin and activin inhibition or blocking situations is not well known (1). In addition, novel pathways regulating cell growth have been found in recent years. One of those is Hippo signaling (13, 22). In vitro studies suggest its importance at least in muscle cell proliferation and differentiation (26, 57). However, no in vivo studies of Hippo signaling have been conducted in skeletal muscle.
We present here a comprehensive signaling analysis in skeletal muscle subjected to in vivo blocking of myostatin and activins. Furthermore, we show that capillary density decreases during the fast muscle growth process and that Hippo signaling is altered in skeletal muscle in various circumstances.
MATERIALS AND METHODS
The treatment of the animals was in strict accordance with the European Convention for the Protection of Vertebrate Animals Used for Experimental and Other Scientific Purposes. The protocol was approved by the National Animal Experiment Board (permit no. ESLH-2009-08528/Ym-23).
In the main experiments, we used 6- to 7-wk-old male C57Bl/10SnJ mice. Additionally, in the longer 7-wk experiment we used mdx mice on a C57Bl/10SnJ background and healthy C57Bl/10SnJ mice as controls. All mice were obtained from the Jackson Laboratory (Bar Harbor, ME). Mice were housed in standard conditions (temperature 22°C, light from 8 AM to 8 PM) and had free access to tap water and food pellets (R36, 4% fat, 55.7% carbohydrate, 18.5% protein, 3 kcal/g; Labfor, Stockholm, Sweden).
In the 2-wk experiment, mice were randomly divided into three groups: 1) PBS, 2) sActRIIB-Fc intraperitoneally (ip) once/wk, and 3) sActRIIB-Fc twice/wk (n = 6 in each). Both the once/wk and twice/wk groups were further divided into 5 and 10 mg/kg subgroups. Because no difference in body and muscle mass was found between these doses, the results were pooled for the further analysis (see Fig. 1A for body mass as an example). To study signaling further, we conducted another experiment where mice were divided into three groups (n = 6 in each group): 1) euthanized 1 or 2 days after a single injection of PBS (ip), 2) euthanized 1 day, or 3) euthanized 2 days after sActRIIB-Fc (10 mg/kg ip).
Because we found changes in Hippo pathway that have not been investigated previously in skeletal muscle, we also analyzed muscle samples from another experiment. In this 7-wk experiment, we used dystrophic mdx mice that were randomly divided into four groups: 1) PBS control, 2) PBS exercise, 3) sActRIIB-Fc, and 4) sActRIIB-Fc exercise (n = 8 in each). A fifth group consisting of wild-type control mice (with PBS) served as healthy controls (n = 5). sActRIIB-Fc (5 mg/kg) or PBS was injected ip once/wk. Exercise was voluntary wheel running, since it may offer benefits in mdx mice (10, 30), instead of heavy forced exercise (20). Mice were housed in cages, where they had free access to custom-made running wheels (diameter 24 cm, width 8 cm) 24 h/day. Total running distance was recorded 24 h daily. One mouse that ran substantially less was taken out from the further analysis to get a more homogenous group of exercising mice. Sedentary animals were housed in similar cages without the running wheel.
The mice were euthanized by cervical dislocation, and blood and tissue samples were collected. Before euthanization, mice were weighed and grip strength was measured.
The recombinant fusion protein was produced in house. It has been reported previously to normalize (LPS- or activin A overexpression-induced) lung pathology in a mouse model (5) and to increase muscle size in mdx mice (24). The ectodomain (ecd) of human sActRIIB was amplified via PCR with the following primers: 5′-GGACTAGTAACATGACGGCGCCCTGG-3′ and 5′-CCAGATCTGCGGTGGGGGCTGTCGG-3′ from a plasmid containing the human ActRIIB sequence (in pCR-Blunt II-TOPO AM2-G17 ActRIIB, IMAGE clone no. 40005760; The IMAGE Consortium). A human IgG1 Fc domain with a COOH-terminal His6 tag was amplified by PCR (5′-GCAGATCTAATCGAAGGTCGTGGTGATCCCAAATCTTGTGAC-3′ and 5′-TCCCTGTCTCCGGGTAAACACCATCACCATCACCATTGAGCGGCCGCTT-3′) from the pIgPlus expression plasmid. The subcloning of these products was done into the pGEM-T easy (Promega) vectors, sequenced, and fused before cloning into the expression vector pEFIRES-p. For the final protein production, chinese hamster ovary (CHO) cells were transfected with the above-mentioned ActRIIBecd-FcHis6 expression vector via lipofection (Fugene 6; Roche) and selected with puromycin (Sigma-Aldrich, Lyon, France). During selection, cells were grown in DMEM supplemented with 2 mmol/l l-glutamine, 100 μg/ml streptomycin, 100 IU/ml penicillin, and 10% FCS. For large-scale expression, cells were adapted to CD OptiCHO medium (Gibco) supplemented with 2 mmol/l l-glutamine and grown in suspension in an orbital shaker. Cell culture supernatants were clarified by filtration through a 0.22-μm membrane (Steritop; Millipore). Next, NaCl and imidazole were added, and the solution was pumped through a Ni2+-loaded HiTrap Chelating column (GE Healthcare Life Sciences, Uppsala, Sweden) at 4°C. Protein was eluted by raising imidazole concentrations, dialyzed against PBS, and finally concentrated with Amicon Ultra concentrator (30 000 MWCO; Millipore). The purity of our sActRIIB-Fc preparation after IMAC purification was estimated to be >90% based on silver-stained SDS-PAGE.
Hindlimb muscles, soleus, gastrocnemius, musculus quadriceps femoris, extensor digitorum longus, and tibialis anterior, were removed, weighed, and frozen in liquid nitrogen. The muscle masses reported are average weights of the left and right leg.
Muscle protein synthesis: in vivo surface sensing of translation.
To measure muscle protein synthesis, we used a surface sensing of translation (SUnSET) method that was developed recently (51). In this SUnSET technique, the injected antibiotic puromycin (a structural analog of tyrosyl-tRNA), when used at low concentrations, is detected by a puromycin-specific antibody (51). The accumulation of puromycin-conjugated peptides into nascent peptide chains reflects the rate of protein synthesis in many different in vitro and in vivo conditions (19, 43, 51).
Measurement of puromycin incorporation is a very strict process (anesthetic treatment, timing of injections, and freezing etc.), and thus we selected it for only the acute 1- to 2-day experiment. Mice were anesthetized 1 or 2 days after sActRIIB-Fc or PBS injection with carefully standardized inhalation anesthesia using isoflurane (MSS vaporizer with Fluovac absorbing unit; Harvard Apparatus). Subsequently, mice were injected ip with 0.040 μmol/g puromycin (Calbiochem, Darmstadt, Germany) dissolved in 200 μl of PBS (19). At exactly 25 min after the injection of puromycin, mice were euthanized by cervical dislocation. Gastrocnemius muscle was isolated and snap-frozen in liquid nitrogen exactly 30 min after the puromycin injection.
Tissue processing for the protein analysis.
The gastrocnemius muscle was pulverized in liquid nitrogen and homogenized in ice-cold buffer with proper inhibitors and further treated as reported previously (25). One part of the homogenate was taken for the protein puromycin incorporation examination. For that purpose, the sample was centrifuged at 500 g for 5 min to remove cell debris. For the signaling analysis, the rest of the homogenate was centrifuged at 10,000 g for 10 min. Total protein content was determined using the bicinchonic acid protein assay (Pierce Biotechnology, Rockford, IL) with an automated KoneLab analyzer (Thermo Scientific, Vantaa, Finland).
Western immunoblot analyses.
Muscle homogenates were solubilized in Laemmli sample buffer and heated at 95°C to denaturate proteins. Samples containing 30 μg of protein were separated by SDS-PAGE and transferred to PVDF membrane. The membrane was blocked in TBS with 0.1% Tween 20 (TBST) containing 5% nonfat dry milk and incubated overnight at 4°C with primary antibodies. Membrane was then washed in TBST and incubated with secondary antibody (horseradish peroxidase-conjugated anti-rabbit or anti-mouse IgG; Jackson ImmunoResearch Europe) for 1 h, followed by washing in TBST. Proteins were visualized by ECL (SuperSignal west femto maximum sensitivity substrate; Pierce Biotechnology) and quantified (band intensity × volume) using a ChemiDoc XRS device in combination with Quantity One software (version 4.6.3; Bio-Rad Laboratories).
To detect puromycin incorporation, the mouse monoclonal anti-puromycin antibody (clone 12D10) (51) was incubated overnight at 4°C (19). An automatic band detection of the Quantity One software was used to analyze intensities of all of the visible bands of puromycin-incorporated proteins as well as ubiquitinated proteins (see below). The sum of all the bands was taken into analysis. The result was confirmed by using different sensitivities of band recognition or quantifying the intensity of the whole lane (data not shown). Myosin heavy chain (MyHC) protein synthesis was estimated by quantifying the strong band at ∼200 kDa thought to contain almost exclusively MyHC isoforms (Fig. 2B).
The uniformity of protein loading was confirmed by staining the membrane with Ponceau S and by reprobing the membrane with an antibody against GAPDH (Abcam, Cambridge, MA) or in some runs against α-tubulin Sigma-Aldrich (Lyon, France).
Antibodies recognized phosphorylated Akt at Ser473 and Thr308, rpS6 at Ser240/244, rpS6 at Ser235/236, p38 MAPK at Thr180/Tyr182, ERK1/2 MAPK at Thr202/Tyr204, S6K1 at Thr389, AMPKα at Thr172, 4E-BP1 at Thr37/46, Yes-associated protein (YAP) at Ser127, MKK4 at Ser257, FoxO1 at Thr24, Smad2 at Ser465/467, Smad3 at Ser423/425, acetyl-CoA carboxylase (ACC) at Ser79, and eEF2 at Thr56 (Cell Signaling Technology, Beverly, MA). Antibody against eIF2Bε was from Cell Signaling Technology, REDD1 antibody was from ProteinTech Group (Chicago, IL), VEGF-A (Santa Cruz Biotechnology) against COOH-terminal myostatin (detecting myostatin at size ∼30 kDa) was from Chemicon/Millipore (AB 3239), and androgen receptor was from Sigma-Aldrich. Moreover, total proteins of S6K1 (Santa Cruz Biotechnology), Smad2 (Invitrogen, Life Technologies, NY), rpS6, Akt, AMPK, YAP, ERK1/2, p38, FoxO1, ACC, and eEF2 were analyzed using specific antibodies (Cell Signaling Technology). This was conducted by reprobing the membrane after the membrane was carefully stripped with Restore Western Blot Stripping Buffer (Pierce Biotechnology). For ubiquitinated proteins, horseradish peroxidase-conjugated anti-ubiquitin antibody was used (Santa Cruz Biotechnology).
Gastrocnemius muscle cross-sections cut with cryomicrotome were stained for sarcolemma using antibody against caveolin 3 (ab2912; Abcam) and capillaries with isolectin-GS-IB4 containing fluorescent Alexa Fluor 488 label (Molecular Probes, Eugene, OR), as described earlier (28). Image analysis was performed with Image J software (National Institutes of Health, Bethesda, MD).
Differences between groups were evaluated by analysis of variance or unpaired t-test where appropriate. When the Shapiro-Wilk test revealed that Western blot data were not normally distributed, the results of those variables were log-transformed. ASW statistics version 18.0 was used for statistical analyses (SPSS, Chicago, IL). The level of significance was set at P ≤ 0.05. Data are expressed as means ± SE. Correlations were analyzed using Pearson's Product Moment Coefficient.
sActRIIB-Fc increases muscle mass but decreases capillary density per area.
Mice that were treated for 2 wk with sActRIIB-Fc were euthanized to understand what the effect of sActRIIB-Fc is during this short period. To examine whether the frequency of injections matters, mice were injected with sActRIIB-Fc either once or twice/wk for 2 wk. The twice/wk dosage increased body and muscle mass slightly more than the once/wk administration (Fig. 1). sActRIIB-Fc administration increased mean myofiber size in gastrocnemius and induced a shift toward larger cross-sectional myofiber area (Fig. 1C). sActRIIB-Fc also increased grip strength (not shown). Capillary density per area decreased (Fig. 1D), which can be explained simply by the increase in fiber size, as there was no change in capillaries per fiber (not shown). No effect on heart or fat mass existed, but liver mass increased (not shown), although not when adjusted to body weight.
sActRIIB-Fc increases muscle protein synthesis and mTORC1 signaling.
To elucidate potential molecular mechanisms and responsive signaling cascades underlying the observed changes in muscles with sActRIIB-Fc administration, mice were euthanized 1 or 2 days after sActRIIB-Fc or PBS injection. Global muscle protein synthesis was increased in gastrocnemius muscle ∼30% 2 days after the sActRIIB-Fc injection (Fig. 2). The estimated MyHC protein synthesis was slightly increased already 1 day postinjection but more robustly at 2 days postinjection (Fig. 2). These changes occurred together with the increased phosphorylation (p) of rpS6 at Ser240/244 and Ser235/236 and phosphorylation of S6K1 and 4E-BP1, indicating activated mTORC1 signaling (Fig. 3, A and B). p-rpS6 at Ser240/244 and p-S6K1 correlated positively with protein synthesis on day 1 and especially on day 2 (r = 0.7–0.9, P < 0.01). On day 2, p-4E-BP1 also showed a similar association (r = 0.9). After 2 wk of injection, a trend for increased mTORC1 activation was observed, but only in the mice injected twice/wk (Fig. 3, A and B). In that group, the last injection was 4 days before the muscles were collected.
The protein content of eukaryotic initiation factor 2Bε (eIF2Bε) also increased 2 days after the sActRIIB-Fc administration (Fig. 3C). The phosphorylation of Akt at Ser473 increased after 2 wk of the sActRIIB-Fc injection but was unchanged at the 1- and 2-day time points (Fig. 3C). No significant change was observed in the phosphorylation of Akt at Thr308 (Fig. 3C) or eukaryotic elongation factor 2 (eEF2) in relation to total eEF2 protein (not shown).
The phosphorylation of FoxO1 was unchanged acutely but was decreased 2 wk after the sActRIIB-Fc administration (Fig. 3D). In the group of two injections per week, this decrease was also significant when calculated relative to total FoxO1. However, these changes did not lead to changes in total ubiquitinated proteins (not shown).
Changes in MAPK signaling and VEGF-A protein.
The phosphorylation of MAPKs p38 (MAPK14) and Erk 1/2 (MAPK3/MAPK1) (Fig. 4A) as well as MKK4 (JNKK2; Fig. 4D) decreased and AMPK (Fig. 4B) increased 2 days after sActRIIB-Fc injection. No changes were seen after 2 wk. Interestingly, a decrease in proangiogenic VEGF-A was seen (Fig. 4D). At these time points, total and phosphorylated Smad2 were unchanged (Fig. 4B), although our sActRIIB-Fc strongly inhibits myostatin and activin A and B-induced Smad-dependent CAGA reporter activity (not shown). It is possible that changes in Smad3 phosphorylation would have been observed instead. We could, however, observe only very faint p-Smad3 bands in our samples with no major changes due to sActRIIB-Fc (not shown) similarly as has been reported after 14 wk of myostatin blocking in healthy mice (42).
Altered hippo signaling and REDD1 protein.
To find more novel proteins, we searched published microarray data for possible candidates showing a response to sActRIIB-Fc treatment (47) that might have a role in the regulation of cell size. The injection of sActRIIB-Fc increased the phosphorylated YAP at Ser127 acutely, and the increase was still observed after 2 wk (Fig. 4C). Also, total YAP was increased at 2 days (Fig. 4C), as was the case earlier in mRNA level (47). Because of these changes in the Hippo pathway, we also examined whether the altered Hippo signaling exists after longer myostatin- and activin-blocking treatment. Moreover, the possible effect of exercise or muscle dystrophy was investigated. Both myostatin/activin blocking and exercise for 7 wk increased YAP phosphorylation, but the combination did not have an additive effect (Fig. 5). Interestingly, dystrophic mdx mice had elevated phosphorylated and especially total YAP compared with similarly treated wild-type control mice (Fig. 5). Also, a small (1.2-fold, P = 0.05) increase was seen in YAP mRNA in microarray data in the dystrophic compared with wild-type control muscle, but as was the case with total YAP protein, no effects of exercise or sActRIIB-Fc were seen (unpublished data). Upstream to YAP, mRNA expression of FRMD6 (1.4-fold), RASSF2 (1.7-fold), and RASSF4 (2.7-fold) was increased significantly (P < 0.001) in dystrophic muscle. However, some other genes in the Hippo pathway, such as Taz, Mst1, and Lats2, were unchanged (unpublished data). We also reanalyzed our previous microarray data from skeletal muscle in streptozotocin-induced diabetes (25). It was found that the mRNA level of YAP is not affected significantly in diabetic muscle atrophy or in exercised muscle (microarray data are available under accession no. GSE1659 at http://www.ncbi.nlm.nih.gov/geo/).
Rahimov et al. (47) also reported a significant increase in androgen receptor mRNA after sActRIIB-Fc. We found that the protein content of androgen receptors tended to increase after 2 days; however, this did not reach statistical significance (not shown). We found earlier that REDD1 protein and mRNA contents were strongly increased in streptozotozin-induced diabetic atrophy (25). In the present study, the protein content of REDD1 was increased after 1 and 2 days of sActRIIB-Fc injection (Fig. 4C). Mdx mice had lower REDD1 compared with similarly treated wild-type control mice, but exercise and sActRIIB-Fc treatment reduced this difference by increasing REDD1 (Fig. 5). The myostatin “immunoreactive protein” at ∼30 kDa probed with COOH-terminal antibody was slightly but nonsignificantly decreased after 2 wk (not shown).
Muscle loss is a major risk factor for many diseases and for early death (4, 45, 68). In the present study, our main goal was to elucidate signaling cascades downstream of myostatin and activins by blocking them using sActRIIB-Fc. It was found that myostatin and activin blocking rapidly increased muscle size. Thus the next step was to investigate possible changes in muscle protein synthesis. As expected, muscle protein synthesis increased, especially 2 days after sActRIIB-Fc administration. Increased muscle protein synthesis supports the results observed after myostatin blocking (58, 64) or in inhibition/deletion situations (49, 54, 58, 64). A similar 2-day delay for the maximal effect occurred in most of the signaling changes. Conceivably, skeletal muscle requires some time to change its signaling response to a decreased concentration of free myostatin/activin concentration since at least part of the endogenous signaling seemed to be still active after 1 day (14).
Various signaling pathways known to regulate muscle size were investigated. An important pathway in muscle hypertrophy, mTORC1, was activated 2 days after sActRIIB-Fc injection, as assessed by increases in S6K1 and rpS6 phosphorylation. These markers correlated well with muscle protein synthesis. This supports the evidence that mTORC1 has a role in the muscle growth that is induced by myostatin/activin blocking (27, 50, 61). However, pathways other than rapamycin-sensitive mTORC1 also regulate muscle protein synthesis (58) and muscle growth (27, 50) when myostatin/activins are blocked or inhibited. sActRIIB-Fc also increased the phosphorylation of 4E-BP1, a protein less responsive to rapamycin (12), and eIF2Bε protein. eIF2Bε is required for translation initiation, and its overexpression has induced ∼20% hypertrophy of myofibers in mice (38). In contrast, the phosphorylation of Akt at Ser473 increased after only 2 wk, when muscle mass was already increased, and at Thr308 it remained unchanged. This supports previous results showing that sActRIIB-Fc-induced increase in muscle mass may be Akt independent (18), perhaps in contrast to follistatin-induced hypertrophy (27, 61). Further studies are requested to investigate upstream regulation of mTOR signaling and muscle protein synthesis after myostatin and activin blocking. Earlier studies (50, 60) and our present results suggest that in healthy adult muscles the role of altering protein degradation by blocking myostatin is minor. This may be in contrast with situations in which protein degradation (64) or myostatin levels (35) are increased. Actually, our data showed a decrease in the phosphorylation of FoxO1 after sActRIIB-Fc, which would in theory increase the transcription of ubiquitin ligases (9). Some studies have indeed found unexpected increases in ubiquitin ligase transcripts as a response to myostatin knockout/blockade (41, 60) and, on the other hand, decreased levels by increased myostatin (55), but not all studies agree with this (18). However, no changes in ubiquitinated proteins were seen in the present study. We also analyzed phosphorylated FoxO3a at Thr32, and no changes were observed (not shown).
Decreased capillary density per muscle area after blocking myostatin and activins was observed, supporting the evidence from myostatin-deficient mice (2, 37, 48). Previously, a decreased mitochondria count (2, 47) has also been reported in the skeletal muscle of myostatin-deficient mice. It is possible that the observed decrease in angiogenic transcription factor VEGF-A and proangiogenic MAPK signaling may have slowed down the increase in angiogenesis (46, 67) relative to the increase in global muscle protein synthesis and increase in muscle fiber size. This eventually led to a muscle phenotype in which capillaries can be considered as “diluted” in the muscle. The formation of new blood vessels is a fast process (3). Thus it can be speculated that without reduced VEGF-A and MAPK signaling, maintenance of the capillary density could have been seen.
The clear time course of decreased MAPK signaling after blocking of myostatin and activin was found. This explains possibly why the previous in vivo results in muscle with myostatin deficiency have been inconsistent (34, 59), even though increased myostatin has been shown repeatedly to increase MAPK signaling (62, 64). In addition to angiogenesis as mentioned above, decreased MAPK signaling may also affect muscle function through having both anabolic and catabolic roles in muscle (21, 33, 44, 52) and mediate antidifferentiation effects of myostatin (62). Further studies are needed to investigate the importance of MAPK signaling in mediating the effects of modulated TGFβ signaling. Also, the importance of the decreased capillary density on muscle function needs further study. Our unpublished data suggest that during the fast muscle and body mass growth period mice are voluntarily less active and run less. Therefore, additional treatments may be mandatory for more optimal muscle adaptation at least during the first few weeks when myostatin/activins are blocked.
As a novel finding, myostatin and activin blocking, exercise, and muscle dystrophy all increased YAP phosphorylation. Moreover, YAP protein content was robustly increased and also YAP mRNA slightly increased in dystrophic muscle and the protein content also to some extent 2 days after ActRIIB-Fc injection. This suggests transcriptional/translational regulation in skeletal muscle of this protein in addition to direct phosphorylation. The overexpression of this Hippo pathway protein strongly increases organ size (13, 22), but no in vivo studies in skeletal muscle have been reported. Phosphorylation at Ser127 localizes YAP into cytoplasm and thus decreases YAP activity as a transcription factor in nuclei (65). During muscle differentiation in C2C12 cells, YAP phosphorylation at this site is strongly increased, which was important for myoblast differentiation (57). A recent study from the same group showed that YAP expression increased when satellite cells were activated but decreased when differentiation started and that the constitutive expression of YAP promoted proliferation but prevented differentiation (26). In contrast, knocking down YAP strongly decreased the proliferation of myoblasts (26). The upstream mechanisms for the present findings are only speculations, but our unpublished microarray evidence suggests that transcription of many genes upstream to YAP are increased in dystrophic muscle. It can be speculated that slightly hypertrophic dystrophic muscle with large amounts of inflammatory and regenerated/degenerated cells (hematoxylin and eosin staining; not shown) activates the Hippo pathway. The Hippo pathway may sense increased cell density or other factors in vitro, as has been published previously (66). On the other hand, YAP may function as an inhibitor of TGFβ/Smad signaling (17, 56). This suggests a direct connection between myostatin and activin blocking with Hippo signaling. As mentioned above, signaling downstream to Akt (but not Akt itself) was increased after myostatin and activin blocking. Despite alternative evidence (6), a recent comprehensive study suggests that activation of Akt may not explain increased phosphorylation of YAP (66).
Another interesting protein, REDD1, was decreased in dystrophic mice that are slightly more muscular than wild-type control mice because of a pseudohypertrophy (29). This is in contrast to a strong increase in REDD1 that we reported recently in an animal model of type 1 diabetes (25). We also found that both sActRIIB-Fc injection and exercise slightly increased REDD1 content. The overexpression of REDD1 specifically in muscle has been shown to cause a 10% decrease in muscle fiber size (16). Thus, in theory, the small increase in REDD1 induced by sActRIIB-Fc would not be anabolic but may instead be catabolic in nature. Possible candidates for observed increased REDD1 may be DNA damage (15), energy stress (39), and hypoxia (8, 53). Another possibility in the present study may be just a nonspecific increase in REDD1 protein synthesis during myostatin/acivin blocking-induced muscle growth. In summary, our current study identifies the Hippo signaling and to some extent possibly also REDD1 as pathways involved in regulating skeletal muscle adaptation in response to myostatin and activin blocking, muscle dystrophy, and exercise.
The phosphorylation of AMPK increased 2 days after sActRIIB-Fc injection. In skeletal muscle cells in vitro, myostatin increases AMPK phosphorylation (11), but in contrast, myostatin-null mice also have an increased AMPK activity (63). Thus, the connection of myostatin/activin inhibition and AMPK in vivo may be indirect, possibly due to the changes in cellular energy state. Previously, myostatin blocking for 6 wk has enhanced glucose homeostasis (7). This response may have benefited in part from AMPK activation (36), but it was not investigated in that study. However, the phosphorylation of AS160, a protein downstream of AMPK (36), remained unchanged in the present study, as was the case with the phosphorylation of ACC (unpublished data). Therefore, the consequence of the increased phosphorylation of AMPK remains unclear.
The present results show that some of the effects of blocking myostatin/activins are anabolic in nature for muscle (e.g., mTORC1 signaling). On the other hand, some responses have both anabolic and catabolic roles depending on the circumstances (MAPK signaling), whereas some are thought to be mainly catabolic (AMPK and REDD1). In the case of blocking myostatin and activins, anabolic signaling is overriding catabolic stimulus to induce muscle growth. Future studies should investigate the importance of these pathways during myostatin/activin blocking on muscle phenotype and function.
Our data indicate that blocking of myostatin and activins rapidly increases muscle mass but decreases capillary density per muscle area. The results also show that many signaling pathways are altered with a specific time course after myostatin and activin blocking. Of those, we report for the first time that Hippo signaling is increased after myostatin and activin blocking and exercise and in dystrophic muscle. This suggests that Hippo signaling has an important and previously unappreciated regulatory function in skeletal muscle.
This work was supported by the Academy of Finland (Decision No. 137787 to J. J. Hulmi) and the Paulo Foundation (H. Kainulainen).
None of the authors have any conflicts of interest, financial or otherwise.
J.J.H., H.K., and O.R. did the conception and design of the research; J.J.H., B.M.O., M.S., and H.M. performed the experiments; J.J.H., B.M.O., W.M.H., and A.P. analyzed the data; J.J.H., B.M.O., W.M.H., H.K., and O.R. interpreted the results of the experiments; J.J.H. prepared the figures; J.J.H. drafted the manuscript; J.J.H., M.S., W.M.H., H.K., and O.R. edited and revised the manuscript; J.J.H., B.M.O., M.S., W.M.H., H.M., P.P., A.P., H.K., and O.R. approved the final version of the manuscript.
We thank Kaisa-Leena Tulla, Mervi Matero, Simon Walker, Risto Puurtinen, Aila Ollikainen, Paavo Rahkila, Sanna Lensu, Sira Torvinen, and Mia Horttanainen for the provided help. We also thank the students who assisted in the data collection.
- Copyright © 2013 the American Physiological Society