Mechanisms of exercise-induced survival motor neuron expression in the skeletal muscle of spinal muscular atrophy-like mice

Abstract

Key points: Spinal muscular atrophy (SMA) is a health- and life-limiting neuromuscular disorder caused by a deficiency in survival motor neuron (SMN) protein. While historically considered a motor neuron disease, current understanding of SMA emphasizes its systemic nature, which requires addressing affected peripheral tissues such as skeletal muscle in particular. Chronic physical activity is beneficial for SMA patients, but the cellular and molecular mechanisms of exercise biology are largely undefined in SMA. After a single bout of exercise, canonical responses such as skeletal muscle AMP-activated protein kinase (AMPK), p38 mitogen-activated protein kinase (p38) and peroxisome proliferator-activated receptor γ coactivator 1α (PGC-1α) activation were preserved in SMA-like Smn^2B/- animals. Furthermore, molecules involved in SMN transcription were also altered following physical activity. Collectively, these changes were coincident with an increase in full-length SMN transcription and corrective SMN pre-mRNA splicing. This study advances understanding of the exercise biology of SMA and highlights the AMPK-p38-PGC-1α axis as a potential regulator of SMN expression in muscle.

Abstract: Chronic physical activity is safe and effective in spinal muscular atrophy (SMA) patients, but the underlying cellular events that drive physiological adaptations are undefined. We examined the effects of a single bout of exercise on molecular mechanisms associated with adaptive remodelling in the skeletal muscle of Smn^2B/- SMA-like mice. Skeletal muscles were collected from healthy Smn^2B/+ mice and Smn^2B/- littermates at pre- (postnatal day (P) 9), early- (P13) and late- (P21) symptomatic stages to characterize SMA disease progression. Muscles were also collected from Smn^2B/- animals exercised to fatigue on a motorized treadmill. Intracellular signalling and gene expression were examined using western blotting, confocal immunofluorescence microscopy, real-time quantitative PCR and endpoint PCR assays. Basal skeletal muscle AMP-activated protein kinase (AMPK) and p38 mitogen-activated protein kinase (p38) expression and activity were not affected by SMA-like conditions. Canonical exercise responses such as AMPK, p38 and peroxisome proliferator-activated receptor γ coactivator-1α (PGC-1α) activation were observed following a bout of exercise in Smn^2B/- animals. Furthermore, molecules involved in survival motor neuron (SMN) transcription, including protein kinase B (AKT) and extracellular signal-regulated kinases (ERK)/ETS-like gene 1 (ELK1), were altered following physical activity. Acute exercise was also able to mitigate aberrant proteolytic signalling in the skeletal muscle of Smn^2B/- mice. Collectively, these changes were coincident with an exercise-evoked increase in full-length SMN mRNA expression. This study advances our understanding of the exercise biology of SMA and highlights the AMPK-p38-PGC-1α axis as a potential regulator of SMN expression alongside AKT and ERK/ELK1 signalling.

Keywords: AMPK; Autophagy; PGC-1α; mRNA splicing.

Figure 1. Characterization of Smn^2B/− mice during disease progression

A, body mass of Smn^2B/+ and Smn^2B/− mice at postnatal day (P) 9, P13 and P21. B, Smn^2B/+ and Smn^2B/− mice in prone (left) and side‐lying (right) positions. C, mass of the tibialis anterior (TA; green), gastrocnemius (GAST; red), quadriceps (QUAD, blue), soleus (SOL, purple), and extensor digitorum longus (EDL, orange) muscles from Smn^2B/− mice displayed relative to Smn^2B/+ littermates. D, representative western blot of survival motor neuron (SMN) protein in QUAD muscles of Smn^2B/+ and Smn^2B/− mice. A Ponceau S stain is displayed below to indicate equal loading between samples. Protein ladder markers are expressed as kDa. ^* P < 0.05 vs. age‐matched Smn^2B/+; n = 10. [Color figure can be viewed at wileyonlinelibrary.com]

Figure 2. AMP‐activated protein kinase (AMPK), p38 mitogen‐activated protein kinase (p38) and peroxisome proliferator‐activated receptor γ coactivator‐1α (PGC‐1α) levels in skeletal muscle of SMA‐like mice

A, representative western blots of the phosphorylated form of AMPK (pAMPK), total AMPK, phosphorylated p38 (pp38), total p38 and PGC‐1α in QUAD muscles of Smn^2B/+ and Smn^2B/− animals at presymptomatic (P9), early symptomatic (P13) and late symptomatic time points (P21). A Ponceau S stain is shown below to indicate equal loading. Approximate molecular masses (kDa) are denoted to the right of blots. B–H, graphical summaries of pAMPK (B), AMPK (C), AMPK activation status (i.e. the phosphorylated form of the protein relative to its total amount within the same sample; D), pp38 (E), p38 (F), p38 activation status (G) and PGC‐1α (H). Values are displayed as a fold difference relative to P9 Smn^2B/+ animals. ^* P < 0.05 vs. age‐matched Smn^2B/+; n = 8.

Figure 3. Exercise performance and muscle damage in Smn^2B/− animals

A, overview of the experimental design utilized for the investigation of exercise‐induced signalling and gene expression in the skeletal muscle of SMA‐like animals. Tissues from sedentary (SED) Smn^2B/+ and Smn^2B/− mice, as well as from animals that were killed immediately after exercise (Smn^2B/− 0 h) or 3 h post‐exercise (Smn^2B/− 3 h) were harvested at P17. B, maximal treadmill run distance completed by Smn^2B/+ and Smn^2B/− mice. C, representative haematoxylin and eosin staining of SOL muscles from Smn^2B/+ SED, Smn^2B/− SED, Smn^2B/− 0 h, and Smn^2B/− 3 h animals. The scale bar represents 50 µm. ^* P < 0.05 vs. Smn^2B/+ SED, n = 7. [Color figure can be viewed at wileyonlinelibrary.com]

Figure 4. Exercise‐induced signalling in the skeletal muscle of SMA‐like animals

A, representative western blots of pAMPK, AMPK, pp38, p38 and PGC‐1α in QUAD muscles of Smn^2B/+ SED, Smn^2B/− SED, Smn^2B/− 0 h and Smn^2B/− 3 h mice. A Ponceau S stain is also displayed below to demonstrate equal loading across samples. Ladder markers are expressed as kDa. B–D, graphical summaries of pAMPK, AMPK and AMPK activation status (B), pp38, p38 and p38 activation status (C), and total myocellular PGC‐1α levels (D). E, PGC‐1α and nuclear respiratory factor‐1 (NRF‐1) mRNA expression in the TA muscles from all experimental groups. F, representative western blots of pAMPK and AMPK in TA muscles of Smn^2B/+ SED, Smn^2B/− SED and Smn^2B/− 3 h mice. Graphical summaries of pAMPK, AMPK and activation status are shown to the right. G, PGC‐1α and NRF‐1 mRNA expression in the QUAD muscles from Smn^2B/+ SED, Smn^2B/− SED, and Smn^2B/− 3 h animals. Values are expressed relative to Smn^2B/+ SED. ^* P < 0.05 vs. Smn^2B/+ SED; ^# P < 0.05 vs. Smn^2B/− SED; ^† P < 0.05 vs. Smn^2B/− 0 h; n = 5–7.

Figure 5. Subcellular localization of PGC‐1α in the muscle of exercised SMA‐like mice

A, immunofluorescence images of PGC‐1α in SOL muscles of Smn^2B/+ SED, Smn^2B/− SED, Smn^2B/− 0 h and Smn^2B/− 3 h mice. DAPI denotes myonuclei while laminin marks the sacrolemma. The PGC‐1α column displays representative, confocal immunofluorescence images from the four experimental cohorts. The merged images show the overlay of the four channels. White arrows denote PGC‐1α‐positive myonuclei. The scale bar represents 5 µm. B, left, graphical summary of PGC‐1α subcellular localization in cytosolic and nuclear compartments of SOL muscles from the four experimental groups. B, right, enlarged summary of the percentage nuclear accumulation of PGC‐1α from B, left. ^* P < 0.05 vs. Smn^2B/+ SED; ^# P < 0.05 vs. Smn^2B/− SED; n = 7. [Color figure can be viewed at wileyonlinelibrary.com]

Figure 6. Exercise‐induced expression and activity of SMN transcriptional regulators in the skeletal muscle of SMA‐like mice

A, representative western blots of the phosphorylated form of protein kinase B (pAKT), total AKT, phosphorylated form of ETS‐like gene 1 (pELK1) and ELK1, the phosphorylated form of extracellular signal‐regulated kinase (pERK), ERK, as well as the phosphorylated form of cAMP response element‐binding protein (pCREB) and total CREB the in QUAD muscles of Smn^2B/+ SED, Smn^2B/− SED, Smn^2B/− 0 h, and Smn^2B/− 3 h animals. A Ponceau S stain is shown below that demonstrates equal loading between samples. Protein ladder markers at right of blots are expressed as kDa. B–E, graphical summaries of protein expression and activation status of AKT (B), CREB (C), ELK1 (D) and ERK (E) from QUAD muscles. F, representative western blots of pERK and ERK in TA muscles of Smn^2B/+ SED, Smn^2B/− SED, and Smn^2B/− 3 h animals. Graphical summaries of pERK, ERK and activation status are shown to the right. Values are expressed relative to Smn^2B/+ SED. ^* P < 0.05 vs. Smn^2B/+ SED; ^# P < 0.05 vs. Smn^2B/− SED; ^† P < 0.05 vs. Smn^2B/− 0 h; n = 5–7.

Figure 7. Exercise‐induced SMN gene expression in skeletal muscle of Smn^2B/− mice

A, typical western blot of SMN protein content in QUAD muscles of Smn^2B/+ SED, Smn^2B/− SED, Smn^2B/− 0 h and Smn^2B/− 3 h mice. Ladder markers are expressed as kDa. A graphical summary of SMN protein expression is shown to the right. B, representative endpoint PCR gel of the full length SMN mRNA (SMN) and the alternatively spliced SMN mRNA lacking exon 7 (SMNΔ7) in TA muscles. Graphical summary is shown below. C, summary of TA muscle SMN mRNA expression, as determined using real‐time quantitative PCR analysis, in the four experimental cohorts. D, representative endpoint PCR gel of the full length SMN mRNA and SMNΔ7 in QUAD muscles of Smn^2B/+ SED, Smn^2B/− SED and Smn^2B/− 3 h animals. Graphical summary is shown below. E, summary of QUAD muscle SMN mRNA expression using real‐time quantitative PCR in Smn^2B/+ SED, Smn^2B/− SED and Smn^2B/− 3 h mice. Values are expressed relative to Smn^2B/+ SED. DNA ladder marker is at right of the representative gels in B and D. ^* P < 0.05 vs. Smn^2B/+ SED; ^# P < 0.05 vs. Smn^2B/− SED; n = 9.

Figure 8. Exercise‐induced proteolytic signalling in the skeletal muscle of SMA‐like animals

A, representative western blots of the phosphorylated form of Unc‐51‐like autophagy‐activating kinase 1 (pULK1), total ULK1, sequestosome‐1 (p62), as well as the cytosolic microtubule‐associated protein 1A/1B‐light chain 3 LC3 (i.e. LC3I) and membrane bound LC3 (i.e. LC3II) in QUAD muscles of Smn^2B/+ SED, Smn^2B/− SED, Smn^2B/− 0 h and Smn^2B/− 3 h animals. A Ponceau S stain, indicative of equal loading between samples, is also displayed below. Approximate molecular mass markers (kDa) are denoted at the right of the blots. B–D, graphical summaries of protein expression and activation status of ULK1 (B), as well as the levels of p62 (C) and the LC3II:LC3I ratio (D). E, summaries of ULK1, beclin‐1‐associated autophagy‐related key regulator (ATG14), BCL2/adenovirus E1B 19 kDa protein‐interacting protein 3 (BNIP3), p62, GABA_A receptor‐associated protein‐like 1 (Gabrapl1) mRNA expression in TA muscles from all experimental groups. F, summaries of muscle RING finger‐1 (MuRF1), and muscle atrophy F‐Box (MAFbx) mRNA expression in TA muscles from mice in the four experimental groups. G, gene expression summaries of ATG14 and p62 mRNA expression in QUAD muscles of Smn^2B/+ SED, Smn^2B/− SED and Smn^2B/− 3 h animals. H, gene expression summaries of MAFbx in QUAD muscles of Smn^2B/+ SED, Smn^2B/− SED and Smn^2B/− 3 h mice. All data are expressed relative to Smn^2B/+ SED. ^* P < 0.05 vs. Smn^2B/+ SED; ^# P < 0.05 vs. Smn^2B/− SED; n = 7.