Myofibrillar protein accumulation but reduced protein synthesis in PDCD4-depleted myotubes

Antibodies to PDCD4 (#9535), p62 (#5114), beclin-1 (#3738), microtubule-associated proteins 1A/1B light chain 3B (LC3B) (#3868), phosphorylated (p) FoxO3a^ser253 (#9466), p-AKT^Ser473 (#4060), p-S6^Ser235/236 (#4858), and its kinase S6K1^thr389 (#9234) were purchased from Cell Signaling Technology (Danvers, MA, USA). Antibodies to tropomyosin (CH1), troponin (JLT12) and myosin heavy chain-1 (MHC-1,MF-20) were purchased from Developmental Hybridoma (Iowa City, Iowa, USA). Antibody against the ubiquitin protein ligase muscle RING-finger protein-1 (MuRF1, #55456-1-AP) was obtained from Protein Tech (San Diego, CA, USA). Anti-ubiquitin antibody (#sc-8017) was purchased from Santa Cruz Technology (Dallas, TX, USA) while antibodies against γ-tubulin (#T6557), and puromycin (#MABE343) were purchased from Sigma Aldrich (St. Louis, MO, USA). We purchased ¹⁴C-valine (#ARC-0678) from American Radiolabeled Chemicals Inc. (St Louis, MO, USA). Alpha Modification of Eagle’s Medium 1X (AMEM, #310-010-CL) was obtained from Wisent Inc (Saint-Jean-Baptiste, QC, Canada). Fetal bovine serum (FBS, #12483-020), horse serum (HS: #26050-088), and antibiotic and antimycotic reagents (#15240-062) were obtained from Thermo Fisher Scientific (Mississauga, ON, Canada). Dulbecco’s Modified Eagle Medium (DMEM, high glucose, #D5796-500), protease inhibitor (#P8340) and phosphatase inhibitor (#P5726) cocktails were obtained from Millipore Sigma (St. Louis, MO, USA).

L6 myoblasts were cultured in 6 or 12-well plates (depending on the question being asked) in growth medium (GM: AMEM supplemented with 10% FBS and 1% antibiotic-antimycotic reagents) at 37°C and 5% CO₂. To induce myotube differentiation, confluent cells were shifted into the differentiation medium (DM: AMEM supplemented with 1% antibiotic-antimycotic reagents and 2% horse serum). Medium was replaced every other day until myotubes were formed. For C2C12 experiments, myoblasts were grown at 37°C and 5% CO₂ in DMEM supplemented with 10% FBS and 1% antibiotic-antimycotic reagents. Once confluent, cells were shifted into DMEM supplemented with 1% antibiotic-antimycotic reagents and 2% horse serum.

RNA interference (RNAi) was done as previously described [28]. L6 myotubes were incubated in either control (scrambled) or siRNA (designed against genes of interest) oligonucleotide transfection media, using lipofectamine RNAiMAX (Thermo Fisher, #2179825) following manufacturer’s instructions. Control transfection medium consisted of scramble siRNA oligonucleotides (Sigma Aldrich, #SIC001), Opti-MEM (Gibco, #31985−070) and 1mL of AMEM supplemented with 10% FBS (antibiotic-antimycotic free). We used the following PDCD4 siRNA oligonucleotides, all from Millipore Sigma, Oakville ON Canada: PDCD4 #1 sense (GUCUUCUACUAU UACCAUA [dT] [dT]), PDCD4 #1 antisense (UAUGG UAAUAGUAGAAGAC [dT] [dT]), PDCD4 #2 sense (CUACUAUUACCAUAGACCA [dT] [dT]), and PDCD4 #2 antisense (UGGUCUAUGGUAAUAGUAG [dT] [dT]. Twenty-four hours following transfection, 1mL of AMEM supplemented with 10% FBS and 1% antibiotic-antimycotic was added to each well. Forty-eight hours following transfection, cells were washed 2X in phosphate-buffered-saline (PBS) and harvested with 100 µL of lysis buffer (final concentration: 1mM ethylenediaminetetraacetic acid (EDTA), 2% sodium dodecyl sulfate (SDS), 25 mM Tris-HCl pH 7.5, 10 μL/mL protease inhibitor cocktail (Sigma Aldrich, #P8340-5ML),10 μL/mL phosphatase inhibitor cocktail (Sigma Aldrich, #P5726-5ML), 1 mM dithiothreitol (DTT)). Sample lysates were stored at −20°C until analysis. Western blotting was used to confirm successful knockdown of PDCD4 in myotubes. We routinely observed > 90% knockdown efficiency (Fig 1A). This transfection protocol was carried out similarly for C2C12 myotubes using oligonucleotides designed against the mouse PDCD4 (GAUAUGCUGCUUAAAGAGU[dT][dT], ACUCUUUAAGCAGCAUAUC[dT][dT]). In additional experiments, we used different sets of siRNA oligonucleotides to confirm the data from L6 (GGAUCUUGUGUAACACCUA[dT][dT]; UAGGUGUUACACAAGAUCC[dT][dT]) and C2C12 (GCAUUUGAGAAGACUCUAA[dT[dT]; UUAGAGUCUUCUCAAAUGC[dT][dT]) myotubes. Transfection was done as described above.

Scramble and PDCD4-depleted L6 myotubes cultured on cover slips (Fisher Scientific, #092815−9) within 6-well plates were fixed (4% paraformaldehyde in PBS, 10 min, room temp), permeabilized (0.03% Triton X-100 in PBS, 5 min, room temp) and blocked (10% horse serum in PBS) for 1h at 37°C. Myotubes were then incubated in primary antibody (2.5 μg/mL of anti MHC-1 antibody) in 1% bovine serum albumin (BSA)) overnight at 4°C. The following day, myotubes were washed 3 X 5 minutes in PBS at room temperature and incubated in secondary antibody (Texas Red anti-mouse IgG secondary antibody (1:100 with 1% BSA)) for 2h at room temperature. Next, 4′,6-diamidino-2-phenylindole (DAPI) staining (for nuclei) and mounting the cover slip on a microscope slide were completed. Slides were then imaged using the EVOS FL Auto microscope (Life Technologies) along with the EVOS FL Auto program for maintaining acquisition settings (light and brightness) in all experimental treatments as described previously [29,30]. Briefly, all images were transformed into an 8-bit gray scale image and the mean gray value of each sample was measured within a 0–255 range using ImageJ. Images were quantified using ImageJ.

Control and PDCD4-depleted L6 myotubes in 6-well plates were starved (serum-free, amino acid-free medium, US Biological, #R8999-04A) for 24h. They were then treated for 30 minutes with 1µM of puromycin diluted in differentiation medium as described before [31]. Myotubes were then harvested in a lysis buffer and proteins in the lysates were immunoblotted against an anti-puromycin antibody. Signals were normalized to Ponceau S (Millipore Sigma #P3504; 0.5% Ponceau S, 1% acetic acid) total protein quantification. For this, following protein transfer to PVDF membranes, the membranes were rinsed in TBST before being incubated in Ponceau S for 5 min at room temperature with gentle shaking. Membranes were then briefly destained by gentle shaking in water. Images of entire lanes were then acquired and quantified on BIO-RAD ChemiDoc XRS+ imager.

L6 myoblasts were seeded and differentiated in 6-well plates as described above. On day 4 of differentiation, myotubes were transfected with scrambled or PDCD4 siRNA oligonucleotides as described above. Immediately after adding the transfection mix, myotubes in each well were labeled with 5 μL of [¹⁴C]-valine (0.5 μCi/well) for 24 h. After, 1 mL of DM was added to the existing medium in each well, and the cells were incubated for an additional 24 h to allow further incorporation of the radiolabeled valine into proteins. After the 48-h labeling period, an aliquot of incubation medium was taken from each well and designated as medium radioactivity at the end of the labelling period (or pulse period) (S1 Fig). A batch of wells was harvested (designated as time 0 below) while the remaining myotubes were rinsed 2X in PBS before being incubated in DM that was supplemented with 2mM non-radioactive valine (‘chase’ solution) for 10 or 24h. At 0, 10 and 24 h of chase, samples were harvested by washing myotubes 2X in PBS and then incubating in 1 mL of cold 10% trichloroacetic acid (TCA, 15 min, on ice). Lysates were then collected into microcentrifuge tubes and kept on ice for 1 h. Samples were centrifuged at 9600xg for 2 minutes at 4°C, and the supernatant was discarded. Pellet was washed three times with cold 10% TCA. Following centrifugation (9600xg for 2 min at 4°C), the pellet was resuspended in 300 μL of a solution of 0.1 M NaOH with 0.1% sodium deoxycholate. Following, the samples were neutralized with 1M HCl. An aliquot of this was counted in a scintillation counter. We calculated proteolysis by two approaches: release of ¹⁴C valine from previously labelled cellular proteins into the incubation medium, and by the disappearance of ¹⁴C-valine radioactivity from protein pellets at different times during the chase. The former was calculated by counting medium radioactivity at different times and expressing the data as a % of radioactivity in proteins at the end of the labelling period. The latter was derived by measuring residual radioactivity in proteins at different times and expressing the data as a % of radioactivity in proteins at the end of the labelling period.

L6 myotubes were incubated in starvation medium (serum-free, amino acid-free medium, US Biological, #R8999-04A) as mentioned above for 24 hours. In the last 3h, starvation continued with or without the addition of the proteasome inhibitor MG132 (5µM). Myotubes were then harvested and immunoblotted for ubiquitin. Ponceau S staining was used to check equal protein loading.

C2C12 myoblasts were seeded and differentiated in 35 mm glass-bottom dishes (MatTek Corporation, Part No. P35G-1.5-14-C, Ashland, MA, USA) using DMEM-based DM. On day 4 of differentiation, myotubes were transfected with either PDCD4 siRNA oligonucleotides or scrambled control siRNA oligonucleotides using Lipofectamine RNAiMAX in Opti-MEM as described above. Two days later, dishes were removed from the incubator, and excess media was removed from the glass-bottomed petri dish, leaving 2mL of media remaining in the dish. Myotubes were then loaded with [3.5 μM] of indo-1 AM (Thermo Fisher Cat # I1223) and placed in a water-saturated incubator (37 °C) for 30 min. Once finished, the dishes were moved to a dark room and placed into a custom-made 3D printed chamber and mounted on an inverted microscope (Nikon Eclipse Ts2R-FL, Tokyo, Japan) and continuously super-fused via an analog pump (Reglo, Ismatec, Glattbrugg, Switzerland) with room temperature (~25°C) experimental Tyrode solution (mM): 121 NaCl, 5.0 KCl, 1.8 CaCl2, 0.5 MgCl₂, 0.4 NaH₂PO4, 24 NaHCO₃, 0.1 EDTA, 5.5 glucose, and ~0.2% fetal calf serum, bubbled with 95% O₂/5% CO₂ giving the bath a pH of ~7.4. After one hour of washing the dishes, background fluorescence reading was acquired by focusing on an empty section of the dish. Myotubes were electrically stimulated in a custom-made chamber (holding two parallel platinum electrodes positioned on each side of the light path, 1 cm apart) connected to an electrical stimulator (Model 701C, Aurora Scientific, Aurora, ON, Canada) controlled via Aurora Scientific 600A software (Aurora Scientific, Aurora, ON, Canada). Felix32 software (Horiba, London, ON, Canada) was used to acquire cytoplasmic free [Ca²+] ([Ca²+]_i) readings. Images were acquired using an inverted microscope equipped with a 40x objective. Indo-1 AM dye was excited at 346 ± 5 nm and the two emission wavelengths were recorded by two photomultipliers at 405 ± 5 nm and 495 ± 5 nm (Horiba Ratiomaster, London, ON, Canada) [32]. Myotubes were stimulated at 200mA for 300ms at 100 Hz. Tetanic [Ca²+]_i was measured as the mean Indo-1 AM fluorescence ratio during tetanic stimulation trains. The size/area of the myotube(s) in the field of view was determined by exporting a screenshot of the viewed area to ImageJ, where the area was calculated. [Ca²+]_i was calculated per area of the myotube.

We used an ATP Determination Kit (Thermo Fisher Canada, Cat. # A22066) according to the manufacturer’s instructions. This bioluminescence-based assay uses firefly luciferase, which emits light in the presence of its substrate D-luciferin and ATP. L6 and C2C12 myotubes were transfected with either PDCD4 siRNA oligonucleotides or scrambled siRNA oligonucleotides as described before, and incubated for 48 hours prior to ATP analysis. Two days after transfection, cells were lysed and supernatants collected. A standard reaction solution was prepared by combining D-luciferin, recombinant firefly luciferase, reaction buffer, DTT, and deionized water, as per manufacturer’s instructions. A series of ATP standards were prepared to develop a standard curve. Samples and standards were added to the reaction solution in opaque white 96-well plates, with a final volume of 100 µL per well. Luminescence was measured immediately using a luminometer. Background signal was subtracted from all wells, and ATP concentrations were calculated by interpolating from the standard curve. Final values were normalized to total protein content (µmol ATP/g protein), as determined by the Pierce BCA Protein Assay Kit (Thermo Fisher, Cat #23225).

This was done as described previously [33]. L6 myotubes were transfected with control or PDCD4 siRNA oligonucleotides as described above. Two days later, myotubes were washed and harvested in 10% TCA. Cell lysates were then centrifuged at 2.3g for 15 minutes and the supernatant (containing free amino acids) was collected. Samples were then diluted in a ratio of 1 (sample): 2 (potassium borate buffer): 1 (0.1N hydrochloric acid): 8 (HPLC grade water). Diluted samples were pre-column derivatized with a ratio of 1 (sample): 1 (o-phthalaldehyde, 29.28mM). Samples were then injected into a YMC-Triart C18 column (C18, 1.9μm, 75 x 3.0 mm; YMC America, Allentown, PA, USA) fitted onto an ultra HPLC system (Nexera X2, Shimadzu, Kyoto, Japan) connected to a fluorescence detector (Shimadzu, Kyoto, Japan; excitation: 340 nm; emission: 455 nm). A gradient solution derived from 20mM potassium phosphate buffer (a solution of 6.75mM K₂HPO4 and 13.26mM KH₂PO4 in HPLC grade water, pH 6.5) (Mobile phase A) and a solution made from HPLC grade water (15%), acetonitrile (45%) and methanol (40%) (Mobile phase B) at a flow rate of 0.8mL/min was used to elute the amino acids. The gradient, made up of 5%−100% mobile phase B, was run over 21 minutes. Amino acid standard (Millipore Sigma #AAS18) curves were used to calculate amino acid concentrations and all samples were normalized to total protein as previously described.

Myotubes were treated with either scrambled or PDCD4 siRNA as described above. Twenty-four h after siRNA treatment, myotubes were treated with 5μM of an AKT inhibitor (AKT Inhibitor VIII, Millipore-Sigma Cat #124018). Myotubes were then harvested 24h following treatment with the AKT inhibitor.

Following harvesting, protein concentration was determined using the pierce BCA protein assay method (Thermo Fisher, Cat #23225). Approximately 25 µg of protein (per sample) and molecular weight markers (Bio-Rad #1610373) were separated on 10 or 15% SDS-PAGE gels for 1-2h at 60−120 volts at room temperature. Proteins were then transferred onto polyvinylidene difluoride (PVDF) membranes (0.2µM, Bio-Rad #1620177) in 1X Tris-glycine buffer that contained 20% methanol at 4^oC for 18−22 h. Primary and secondary antibody incubations, imaging and quantification have been described [30,34]. Primary antibodies were diluted 1: 1000 in 5% bovine serum albumin (BSA) in 1X Tris-Buffered Saline + Tween 20 (TBST), except antibodies against troponin and tropomyosin (1: 400), MHC-1 (1: 500), and AKT^ser473 (1: 2000). Membranes were incubated in primary antibodies overnight at 4^oC. They were then washed (3 X 5 min in TBST) followed by incubation in horseradish peroxidase conjugated secondary antibodies (HRP-goat anti mouse (Cell Signaling Technology (CST) #7076) or HRP-goat anti rabbit (CST #7074), depending on the primary antibodies. Secondary antibodies were diluted 1: 10,000 in 2.5% in BSA in 1X TBST buffer. After, membranes were washed 3 × 5 min in TBST. To visualize protein bands, HRP chemiluminescent substrate (Clarity Western ECL Bio-Rad, #1705060S) was applied to each membrane. Signal detection and imaging were performed using the Bio-Rad ChemiDoc MP system, and all images were quantified using Image Lab Software v7.

All immunoblot analyses were quantified and adjusted to their corresponding γ-tubulin values, aside from ubiquitin, which was corrected to their respective ponceau S stain. Unpaired t-tests with a Welch correction were used to measure treatment differences between scrambled and PDCD4-depleted myotubes. For starvation and proteasome inhibition experiments, a one-way ANOVA with a Tukey’s post-hoc test was used. Significance was determined when p-value < 0.05. Results were expressed as mean ± standard error of the mean (SEM) of at least 3 independent experiments (biological replicates, with 3 technical replicates per experiment).

To investigate the effect of PDCD4 in myotubes, we used siRNA to deplete this protein first in L6 myotubes (Fig 1A). PDCD4-depleted myotubes had larger diameters (Fig 1B, C) and greater abundance of myofibrillar proteins MHC-1, troponin, and tropomyosin (Fig 1D–G). Similar results were obtained in C2C12 myotubes depleted of PDCD4 (Fig 2A–E). To confirm the observed effects on myofibrillar protein abundance, we studied additional rat and mouse siRNA oligonucleotides. Transfection of L6 myotubes with the new rat siRNA significantly increased MHC-1 abundance and tended to increase troponin level (S2 Fig). Also, transfection of C2C12 myotubes with a new mouse siRNA significantly increased the abundance of tropomyosin and marginally increased MHC-1 and troponin levels (S3 Fig).

We investigated changes that might explain the increased abundance of myofibrillar proteins upon PDCD4 depletion in L6 myotubes, starting with mTORC1 signaling, a regulator of protein synthesis [6,35]. Phosphorylation of the upstream mTORC1 activator, AKT^ser473, was increased in PDCD4-depleted myotubes (Fig 3A, B). Phosphorylation of a downstream target of mTORC1, S6K1^thr389, was also increased (Fig 3A, C). However, p-S6^ser235/236, a S6K1 substrate, was not different between conditions (Fig 3A, D). There was a trend for reduced 4E-BP1^thr37/46 phosphorylation (Fig 3A, E). The levels of unphosphorylated (total) AKT, S6K1 and S6 were not affected by treatments (Fig 3F, G). Contrary to what one might expect from the increased abundance of myofibrillar proteins, but consistent with our previous findings [28], PDCD4 depletion resulted in a decrease in protein synthesis, as measured with SUnSET analysis (Fig 3H).

Increased anabolic signaling without a corresponding increase in protein synthesis led us to hypothesize that L6 myotubes depleted of PDCD4 would exhibit decreased protein catabolism. There was a trend toward increased FoxO3a^ser253 phosphorylation (Fig 4A, B). Consistent with this, PDCD4-depleted myotubes had reduced expression of the muscle ubiquitin protein ligase, muscle ring finger protein-1 (MuRF1, Fig 4A, C). There were no detectable changes in basal levels of autophagy-related proteins beclin-1, microtubule-associated proteins 1A/1B light chain 3B (LC3BII/I), and p62 (Fig 4A, D–F). To further probe treatment effects on proteolysis, we pre-labelled L6 myotube proteins with ¹⁴C-valine and then ‘chased’ for different lengths of time. The amount of radioactivity released into the incubation medium relative to total protein-associated radioactivity at the end of the incubation with ¹⁴C valine (‘pulse’) (positively proportional to the rate of proteolysis, Fig 4G) or the amount of ¹⁴C-radioactivity retained in myotube proteins relative to total protein-associated radioactivity at the end of the ‘pulse’ (inversely proportional to the rate of proteolysis) should each reflect myotube global proteolysis. PDCD4 depletion did not significantly alter ¹⁴C-valine release from, or retention in myofibrillar proteins. To probe this further, we measured the abundance of ubiquitinated proteins, the substrates of the 26S proteasome. PDCD4 depletion had no effect on the abundance of ubiquitinated protein (Fig 4I, J). However, in cells treated with the proteasome inhibitor MG132, starvation-induced increase in the abundance of ubiquitinated proteins was attenuated in PDCD4-depleted cells (Fig 4G, H). Taken together, these findings point to altered protein turnover in PDCD4-depleted cells with suppression of protein synthesis and minimal effect on proteolysis, although it is possible that proteolysis may be affected in altered physiological states (for example, starvation).

We wondered if the increase in the accumulation of myofibrillar proteins in PDCD4-depleted myotubes conferred improvements to tetanic [Ca²⁺]_i, which is a major regulator of force generation [36]. However, there was no significant difference in raw or normalized tetanic [Ca²⁺]ᵢ between PDCD4-depleted and control myotubes (Fig 5A-C).

Although anabolic signaling is elevated in PDCD4-depleted myotubes, protein synthesis was suppressed. Combined with the fact that increased abundance of myofibrillar proteins in these myotubes did not affect tetanic [Ca²⁺]_i, we wondered if these observations were related to the energy status of the myotubes. A reduction in cellular ATP availability may limit the capacity for protein synthesis despite the activation of anabolic signaling. ATP concentration was lower (−25 to −62.5%) in PDCD4-depleted C2C12 (Fig 5D) and L6 (Fig 5E) myotubes when compared to the myotubes transfected with scrambled siRNA oligonucleotides.

Reduction in protein synthesis may also be related to altered amino acid availability. As can be seen in Fig 6A and B, 48 h after transfection, compared to control myotubes, PDCD4-depleted L6 myotubes had reduced intracellular concentrations of many amino acids, especially the branched-chain amino acids (valine, leucine, isoleucine), proline, and tyrosine.

Because of the marked increase in AKT phosphorylation in PDCD4-depleted cells and the report that AKT is activated in conditions when protein synthesis is inhibited [37], we wondered if AKT mediated the effect of PDCD4 depletion on myotube anabolism. Therefore, we treated L6 myotubes with PDCD4 siRNA and an AKT inhibitor either alone or in tandem. Inhibiting the activation of AKT (Fig 7A, B) had no effect on PDCD4 abundance following siRNA treatment (Fig 7A, C). However, treatment with the AKT inhibitor attenuated the increases in only MHC-1 myofibrillar protein content following PDCD4 depletion (Fig 7A, D–F).