Neural stimulation suppresses mTORC1-mediated protein synthesis in skeletal muscle

Abstract

Skeletal muscle fibers are classified as glycolytic or oxidative, with differing susceptibilities to muscle wasting. However, the intracellular signaling pathways regulating fiber-specific muscle trophism remain unclear because of a lack of experimental models measuring protein synthesis. We developed a mouse model overexpressing a mutated transfer RNA synthetase in muscle fibers, enabling specific protein labeling using an artificial methionine substitute, which can be revealed through click chemistry. This model revealed that denervation increases protein labeling in oxidative muscle fibers through mammalian target of rapamycin complex 1 (mTORC1) activation, while deleting the mTORC1 scaffold protein Raptor reduces labeling in glycolytic fibers. On the other hand, increased muscle activity acutely decreases protein synthesis, accompanied by reduced mTORC1 signaling, glycogen depletion, and adenosine 5'-monophosphate kinase activation. Our findings identify nerve activity as an inhibitory signal for mTORC1-dependent protein synthesis in skeletal muscle, enhancing the understanding of fiber-specific responses to exercise and pathological conditions.

Fig. 1.. Muscle-specific MetRS mice allow for efficient labeling of newly synthesized proteins.

(A) Scheme showing the generation of muscle-specific MetRS mice and workflow for identification and visualization of newly synthesized proteins. (A¹) Schematic representation of the difference between mutated MetRS enzyme and WT. (B) Time-dependent increase in protein labeling after ANL administration (n = 4, means ± SEM; Kruskal-Wallis with Dunn’s post hoc test, **P < 0.01). A.U., arbitrary units. (C) Muscle cryosections stained for GFP (left panels, GFP in green and WGA in red) and FUNCAT (right panels, FUNCAT in red and WGA in green) in 7-day labeled MetRS animals versus WT. (D) Labeling intensity of ANL-containing proteins (BONCAT, left blot) and puromycin incorporation (right blot) in 1-month-old animals compared to 3-month-old ones followed by their quantifications (n = 4, means ± SEM, unpaired t test with statistical significance **P < 0.01 and ***P < 0.001). (E) Force production measurement (MetRS* versus WT after electrical stimulation through the sciatic nerve). (F) Hematoxylin and eosin muscle histology of WT animals (left) versus ANL-labeled Cre-MetRS animals (right). Scale bars, 100 μm. (G) Representative images (on the left) of differences in basal protein labeling (top image, FUNCAT in green and WGA in gray) in different fiber types (bottom image, MHC type 1 in blue, MHC type 2A in green, MHC type 2B in red, MHC type 2X in black, and WGA in gray). Scale bars, 100 μm. Fiber labeling quantification (graph on the left) (averages of 66 type 1 fibers, 190 type 2A fibers, 280 type 2B fibers, and 50 type 2X fibers were manually measured randomly across the slice for each biological replicate) and labeling versus CSA analysis in MetRS animals labeled with ANL for 1 week. An average of 8000 fibers was measured automatically for each biological replicate (means ± SEM, n = 5, unpaired t test with statistical significance **P < 0.01).

Fig. 2.. Denervation leads to an increase in protein labeling in type 1/2A fibers while decreasing in type 2B fibers.

(A) Denervation (Den) leads to 15 to 20% loss in muscle mass (left) and fiber CSA (right). (B) Overall protein labeling (BONCAT left, ponceau right) by Western blotting of gastrocnemius muscle lysates of innervated or denervated limbs [means ± SEM, n = 5 (for muscle weight and CSA) and n = 4 (for Western blot quantification)]. For CSA measurements, each biological replicate represents the mean CSA of an average of 7000 fibers automatically measured (paired t test with statistical significance *P < 0.05 and ***P < 0.001) (C) FUNCAT images show an increase in labeling of specific fibers in denervated muscles. The innervated limb shown in the left panel is compared to the denervated one of the same mouse in the right panel (FUNCAT in green and WGA in gray). (D) Average fluorescence intensity of all fibers. An average of 7000 fibers was measured automatically for each biological replicate (means ± SEM, n = 3, paired t test with no statistical difference). (E) Quantification of labeling intensity based on fiber CSA in innervated and denervated limbs. (F) Increased labeling of fibers during denervation correlates with mitochondrial content. FUNCAT F.I. means fluorescence intensity. n = 4, simple linear regression with ****P < 0.0001. (G) Representative images (on the left) of labeling intensity (FUNCAT in green and WGA in gray) (MHC type 1 in blue, MHC type 2A in green, MHC type 2B in red, MHC type 2X in black, and WGA in gray, right panels). Scale bars, 100 μm. Quantification of FUNCAT intensity in different fiber types (on the right) (means ± SEM, n = 3). About 50 type 1 fibers, 50 type 2A fibers, 150 type 2B fibers, and 50 type 2X fibers were manually measured across the slice. Wilcoxon matched-pairs t test with statistical significance ****P < 0.0001.

Fig. 3.. Denervation is accompanied by an increase in Akt-mTORC1 signaling and its downstream proteome.

(A) BONCAT workflow used to identify ANL-labeled proteins in innervated and denervated muscles. LC-MS, liquid chromatography–mass spectrometry. (B) Volcano plot showing a significant amount of labeled proteins increase (orange) or decrease (green) comparing the innervated and denervated muscles from the same animals (n = 4 per group). Labeled proteins are at least twofold regulated and show P < 0.05. (C) Enrichment analysis of up-regulated labeled proteins 1 week after denervation shows a highly significant presence of processes previously shown to be regulated by activated mTORC1 signaling in skeletal muscle (24), like RNA metabolism and translation factors. (D) Representative image showing P-S6–positive fibers (left image, in red) that correspond to those fibers with higher ANL incorporation after denervation (right image, FUNCAT in green) in denervated limb. Scale bars, 1 μm. (E) Western blotting analysis of Akt-mTORC1 signaling in denervated contralateral muscles compared to control innervated contralateral muscles. For quantification of Western blots, n = 4 animals were used. Data are shown as the means ± SEM by paired t test with statistical significance *P < 0.05 and **P < 0.01.

Fig. 4.. Increased labeling after denervation in type 1/2A fibers requires increased mTORC1 signaling.

(A) Western blotting analysis for labeled proteins after 1 week of denervation in rapamycin- and vehicle-treated animals. Quantification in fig. S2B. (B) Representative images of FUNCAT (images on the left) in innervated or denervated muscles after rapamycin or vehicle treatment (FUNCAT in green and WGA in red). Quantification (graph on the right) of mean fluorescence intensity of single fibers (n = 4). An average of 5900 fibers was automatically measured for each biological replicate. Statistical analyses were performed using two-way ANOVA with Tukey’s post hoc test (*P < 0.05 and **P < 0.01). (C) Rapamycin treatment prevents the increase in protein labeling in type 1/2A fibers after denervation while decreasing even further the labeling in type 2B fibers. Representative images of serial IHC stained for MHC (MHC type 1 in blue, MHC type 2A in green, MHC type 2B in red, MHC type 2X in black, and WGA in gray, left panels) and FUNCAT (FUNCAT in green and WGA in gray, right panels). Around 50 type 1 fibers, 160 type 2A fibers, 160 type 2B fibers, and 50 type 2X fibers were manually measured across the slice. Quantification of FUNCAT in different fiber types (n = 4). Scale bars, 100 μm. Data are shown as the means ± SEM by two-way ANOVA with Tukey’s post hoc test (*P < 0.05, ***P < 0.001, and ****P < 0.0001).

Fig. 5.. Acute, genetic loss of mTORC1 signaling in skeletal muscle reduces labeling predominantly in type 2B fibers.

(A) Schematic representation of the generation of Raptor ko MetRS mice. (B) Western blot showing protein labeling in Raptor ko MetRS mice, treated with vehicle or with rapamycin compared to labeled MetRS animals (n = 5). Statistical analysis: ordinary one-way ANOVA with Tukey post hoc test (*P < 0.05). (C) FUNCAT representative images showing a general labeling decrease when comparing vehicle-treated Raptor ko MetRS animals (left panel) to rapamycin-treated animals (right panel) (FUNCAT in green and WGA in magenta; scale bars, 1000 μm). (D) Mean fluorescence intensity quantification of all fibers (graph on the left) and CSA versus FUNCAT analysis (graph on the right) show labeling reduction in rapamycin-treated animals accompanied by a mild atrophy (CSA analysis, graph in the middle). For this quantification, an average number of 6047 fibers were measured for each biological replicate. Statistical analyses were performed on n = 4 WT animals and n = 5 Cre-Raptor ko MetRS animals using two-tailed Student’s t test (****P < 0.0001). (E) Representative images of vehicle- or rapamycin-treated Raptor ko MetRS* animals: serial sections stained for FUNCAT (top panels, FUNCAT in green and WGA in magenta) and MYH isoforms (bottom panels, MHC type 1 in blue, MHC type 2A in green, MHC type 2B in red, and MHC type 2X in black). Scale bars, 100 μm. (F) Fiber type–dependent fluorescence quantification showing a general decrease in labeling in all fiber types. Statistical analyses were performed on n = 4 WT animals and n = 5 Cre-Raptor ko MetRS animals using two-tailed Student’s t test (****P < 0.0001). (G) The relative loss in labeling compared to control animals is more pronounced in 2B fibers (graph on the right). Statistical analyses were performed on n = 4 WT animals and n = 5 Cre-Raptor ko MetRS animals using one-way ANOVA with Tukey’s post hoc test and statistical significance **P < 0.01.

Fig. 6.. Electrical stimulation impairs mTORC1 signaling and protein synthesis.

(A) Schematic representation of the electrical stimulation protocol used (upper part). Puromycin incorporation evaluated in three different time windows: the first overlapping with 20 min of stimulation (Stim), the second evaluated 20 min after the end of stimulation (Stim + 20 min), and the third evaluated 3 hours after the end of stimulation (Stim + 3 hrs). Representative image of puromycin incorporation immediately after stimulation (blot in the upper left image), 20 min after stimulation (Stim + 20 min, upper middle blot), and 3 hours after stimulation (Stim + 3 hrs, upper right blot). Representative images of blots for the Akt-mTORC1 pathway and elongation factors (lower part). Blot quantifications are in fig. S3A. (B) Representative images of serial stainings for glycogen content (PAS in upper panels) and SDH (lower panels) in control, stimulated muscles and muscles 3 hours after stimulation (Stim + 3 hrs). (C) Quantification of glycogen content versus fiber size (upper graphs) shows a clear reduction in glycogen content in bigger fibers after stimulation (upper graph on the left) compared to Stim +3 hrs (upper graph on the right). After stimulation, the average glycogen content is lower in stimulated fibers (lower graph on the left), while 3 hours after stimulation, the average glycogen content in stimulated fibers is higher than in the control muscle (lower graph on the right). Data are shown as the means ± SEM (n = 3, Wilcoxon matched-pairs t test, ****P < 0.0001). (D) Western blotting analysis for P-AMPK and its target P-ACC immediately after (Stim) and 3 hours after a bout of stimulation (Stim +3 hrs) (n = 4). Data are shown as the means ± SEM by two-way ANOVA with Tukey’s post hoc test (**P < 0.01 and ***P < 0.001). (E) Scheme depicting how contractile activity and mTORC1-dependent protein synthesis are two processes that are not occurring simultaneously, possibly due to competition for the same nutrient pool.