TAZ stimulates exercise-induced muscle satellite cell activation via Pard3-p38 MAPK-TAZ signalling axis

Abstract

Background: Exercise stimulates the activation of muscle satellite cells, which facilitate the maintenance of stem cells and their myogenic conversion during muscle regeneration. However, the underlying mechanism is not yet fully understood. This study shows that the transcriptional co-activator with PDZ-binding motif (TAZ) stimulates muscle regeneration via satellite cell activation.

Methods: Taz^f/f mice were crossed with the paired box gene 7 (Pax7)^creERT2 mice to generate muscle satellite cell-specific TAZ knockout (sKO) mice. Mice were trained in an endurance exercise programme for 4 weeks. Regenerated muscles were harvested and analysed by haematoxylin and eosin staining. Muscle tissues were also analysed by immunofluorescence staining, immunoblot analysis and quantitative reverse transcription PCR (qRT-PCR). For the in vitro study, muscle satellite cells from wild-type and sKO mice were isolated and analysed. Mitochondrial DNA was quantified by qRT-PCR using primers that amplify the cyclooxygenase-2 region of mitochondrial DNA. Quiescent and activated satellite cells were stained with MitoTracker Red CMXRos to analyse mitochondria. To study the p38 mitogen-activated protein kinase (MAPK)-TAZ signalling axis, p38 MAPK was activated by introducing the MAPK kinase 6 plasmid into satellite cells and also inhibited by treatment with the p38 MAPK inhibitor, SB203580.

Results: TAZ interacts with Pax7 to induce Myf5 expression and stimulates mammalian target of rapamycin signalling for satellite cell activation. In sKO mice, TAZ depletion reduces muscle satellite cell number by 38% (0.29 ± 0.073 vs. 0.18 ± 0.034, P = 0.0082) and muscle regeneration. After muscle injury, TAZ levels (2.59-fold, P < 0.0001) increase in committed cells compared to self-renewing cells during asymmetric satellite cell division. Mechanistically, the polarity protein Pard3 induces TAZ (2.01-fold, P = 0.008) through p38 MAPK, demonstrating that the p38 MAPK-TAZ axis is important for muscle regeneration. Physiologically, endurance exercise training induces muscle satellite cell activation and increases muscle fibre diameter (1.33-fold, 43.21 ± 23.59 vs. 57.68 ± 23.26 μm, P = 0.0004) with increased TAZ levels (1.76-fold, P = 0.017). However, sKO mice had a 39% reduction in muscle satellite cell number (0.20 ± 0.03 vs. 0.12 ± 0.02, P = 0.0013) and 24% reduction in muscle fibre diameter compared to wild-type mice (61.07 ± 23.33 vs. 46.60 ± 24.29 μm, P = 0.0006).

Conclusions: Our results demonstrate a novel mechanism of TAZ-induced satellite cell activation after muscle injury and exercise, suggesting that activation of TAZ in satellite cells may ameliorate the muscle ageing phenotype and may be an important target protein for the drug development in sarcopenia.

Keywords: exercise; muscle regeneration; muscle satellite cell; sarcopenia.

Figure 1

TAZ depletion in satellite cells impairs skeletal muscle regeneration. (A) Preparation of satellite cell‐specific TAZ knockout mice (sKO). (B) Immunofluorescence staining of Pax7 (red) and TAZ (green) in tibialis anterior (TA) muscles of wt and sKO mice at 0 and 5 days after cardiotoxin (CTX) damage. The nuclei are stained with DAPI (blue). Scale bar, 50 μm. (C) Left: TAZ, Pax7, Myf5, embryonic myosin heavy chain (eMyHC) and α‐tubulin levels in the TA muscles of wt and sKO mice at 0 and 5 days after CTX damage were analysed by immunoblotting. Right: the transcript levels of Taz, Pax7, Myf5 and Myomaker were analysed by quantitative reverse transcription PCR in TA muscles of wt and sKO mice at 0 and 5 days after CTX damage (n = 5). (D) The number of Pax7‐positive cells relative to the number of fibres in (B) was quantified. (E) Immunofluorescence staining of eMyHC in TA muscles of wt and sKO mice at 5 days after CTX damage. Scale bar, 50 μm. (F) Left: H&E and β‐dystroglycan (green) immunofluorescence staining of TA muscles of wt and sKO mice at 10 days after CTX damage. Scale bar, 50 μm. Right: average myofibre size of TA muscles of wt and sKO mice was quantified at 10 days after CTX damage (n = 3). Data are presented as the mean ± SEM. *P < 0.05, ^** P < 0.01 and ^*** P < 0.001 (Student's t test).

Figure 2

TAZ stimulates Myf5 transcription, but YAP does not. (A) Experimental timeline showing time points of tamoxifen injection, cardiotoxin (CTX) injection and satellite cell isolation. (B) Left: TAZ, Pax7, Myf5 and α‐tubulin levels were analysed by immunoblotting in wt and sKO satellite cells under quiescent (Qui‐) and activated (Act‐) culture conditions. Quiescent and activated satellite cells were prepared as described previously. Right: the transcript levels of Taz, Pax7 and Myf5 were analysed by qRT‐PCR in wt and sKO satellite cells under quiescent and activated culture conditions (n = 3). (C) Left: YAP, Pax7, Myf5 and α‐tubulin levels were analysed by immunoblotting in wt and YAP knockdown (KD) satellite cells under quiescent and activated culture conditions. Right: Yap, Pax7, Myf5, Ctgf and Cyr61 transcription was analysed by qRT‐PCR in wt and YAP knockdown satellite cells under quiescent and activated culture conditions (n = 3). (D) Left: TAZ, Pax7, Myf5 and β‐actin levels were analysed by immunoblotting in wt, sKO and TAZ‐rescued sKO satellite cells. Right: Taz, Pax7 and Myf5 transcript levels were analysed by qRT‐PCR in wt, sKO and TAZ‐rescued sKO satellite cells (n = 3). Data are presented as the mean ± SD. *P < 0.05, ^** P < 0.01 and ^*** P < 0.001 (Student's t test).

Figure 3

TAZ induces Myf5 expression by interacting with Pax7. (A) HEK293T cells were transfected with His‐tagged TAZ or His‐tagged YAP and FLAG‐tagged Pax7 expression plasmids. After 24 h, cell lysates were immunoprecipitated using FLAG antibodies. The immunoprecipitates (IP:FLAG) and whole‐cell lysates (WCE) were analysed by immunoblotting. (B) Co‐immunoprecipitation analysis between endogenous Pax7 and TAZ or YAP in C2C12 cells expressing FLAG‐tagged TAZ or YAP. Cell lysates were immunoprecipitated using FLAG antibodies. The immunoprecipitates and whole‐cell lysates were analysed by immunoblotting. (C) Myc‐tagged TAZ plasmid and FLAG‐tagged Pax7 deletion constructs were transfected into HEK293T cells. After 24 h, cell lysates were immunoprecipitated using FLAG antibodies. The immunoprecipitates and whole‐cell lysates were analysed by immunoblotting. (D) FLAG‐tagged Pax7 plasmid and His‐tagged TAZ deletion construct were transfected into HEK293T cells. After 24 h, cell lysates were immunoprecipitated using FLAG antibodies. The immunoprecipitates and whole‐cell lysates were analysed by immunoblotting. (E) Chromatin immunoprecipitation was performed using control and FLAG‐tagged TAZ‐expressing C2C12 cells using anti‐FLAG antibodies. Immunoprecipitated chromatin fragments were analysed by PCR with a primer set spanning the −57.5kb Myf5 enhancer containing the Pax7‐binding site. (F) Schematic image of Myf5 luciferase reporter constructs containing the Pax7‐binding consensus (Myf5‐luc) or mutant (Myf5 mut‐luc) site. HEK293T cells were transfected with Myf5 luciferase vector, TAZ and Pax7 expression plasmids. After 24 h, luciferase activity was analysed and a Renilla luciferase plasmid was used to normalize transfection. Data are presented as the mean ± SD. *P < 0.05, ^** P < 0.01 and ^*** P < 0.001 (Student's t test).

Figure 4

TAZ activates satellite cells through mTOR signalling. (A) TAZ, Pax7, Myf5, Rheb, Rhebl1, phospho‐p70 S6K, p70 S6K, phospho‐4E‐BP, 4E‐BP and α‐tubulin levels were analysed by immunoblotting in wt and sKO satellite cells under quiescent (Qui‐) and activated (Act‐) culture conditions. (B) Taz, Rheb and Rhebl1 transcript levels were analysed by qRT‐PCR in wt and sKO satellite cells under quiescent and activated culture conditions (n = 3). (C) Mitochondrial DNA copy number was analysed by qRT‐PCR using genomic DNA of wt and sKO satellite cells under quiescent and activated culture conditions (n = 3). (D) Wt and sKO satellite cells under quiescent and activated culture conditions were immunostained with MitoTracker to analyse the mitochondrial potential (left). The nuclei are stained with DAPI (blue). Scale bar, 10 μm. The fluorescence intensity of each cell was quantified using ImageJ software (right). (E) Representative bright‐field microscopy images of wt and sKO satellite cells under quiescent and activated culture conditions (left). The average satellite cell size was quantified using ImageJ software (right, n = 25). Data are presented as the mean ± SD. *P < 0.05, ^** P < 0.01 and ^*** P < 0.001 (Student's t test).

Figure 5

Pard3 increases TAZ expression in committed cells during asymmetric division of satellite cells. (A) Single extensor digitorum longus (EDL) myofibres were isolated from wt mice, and 48 h later, asymmetrically dividing satellite cells on the myofibre were immunostained for TAZ, Pard3 and Myf5. The nuclei are stained with DAPI (blue). Scale bar, 20 μm. The fluorescence intensity of TAZ and Myf5 expression in asymmetrically dividing satellite cells was quantified using ImageJ software. (B) Satellite cells were transfected with Pard3‐expressing plasmid. After 24 h, Pard3, TAZ, Myf5, Pax7, Rheb, Rhebl1, phospho‐p70 S6K, p70 S6K, phospho‐4E‐BP, 4E‐BP and α‐tubulin levels were analysed by immunoblotting. (C) Pard3‐transfected satellite cells were treated with p38 MAPK inhibitor (SB203580, 10 μM) for 24 h, and the levels of Pard3, phospho‐p38, p38, TAZ, Myf5 and α‐tubulin were analysed by immunoblotting. (D) Quiescent (Qui‐) and activated (Act‐) satellite cells were treated with SB203580 (10 μM) for 24 h, and the levels of phospho‐p38, p38, TAZ, Myf5, Rhebl1, phospho‐4E‐BP, 4E‐BP and α‐tubulin were analysed by immunoblotting. (E) Quiescent and activated satellite cells were transfected with MKK6 expression plasmid to activate p38 MAPK signalling. After 24 h, the levels of phospho‐p38, p38, TAZ, Myf5, Rhebl1, phospho‐4E‐BP, 4E‐BP and α‐tubulin were analysed by immunoblotting. Data are presented as the mean ± SD. ^*** P < 0.001 (Student's t test).

Figure 6

Exercise increases the number of satellite cells and the diameter of myofibres. (A) Wt mice were subjected to endurance exercise training. Gastrocnemius (GA) muscles were harvested from these mice, and tissue immunostaining was performed. H&E and dystrophin (green) immunofluorescence staining of GA muscles from control and exercise‐trained wt mice. Scale bar, 50 μm. Muscle fibre diameters were quantified using ImageJ software. (B) TAZ, Pax7, Myf5, Cyclin D1, Rhebl1, phospho‐p70 S6K, p70 S6K, phospho‐4E‐BP, 4E‐BP and Vinculin levels in GA muscles of control and exercise‐trained wt mice were analysed by immunoblotting. (C) Taz, Pax7, Myf5 and Rhebl1 transcript levels were analysed by qRT‐PCR in GA muscles of control and exercise‐trained wt mice (n = 6). (D) Immunofluorescence staining of Pax7 (red) in GA muscles of control and exercise‐trained wt mice. Nuclei are stained with DAPI (blue). Scale bar, 50 μm. The number of Pax7‐positive cells was quantified relative to the number of fibres. (E) Immunofluorescence staining of Pax7 (red) and TAZ (green) in GA muscles of control and exercise‐trained wt mice. Scale bar, 50 μm. The ratio of the number of TAZ‐positive cells to the number of Pax7‐positive cells was quantified. (F) Immunofluorescence staining of Pax7 (red) and Myf5 (cyan) in GA muscles of control and exercise‐trained wt mice. Scale bar, 50 μm. The ratio of the number of Myf5‐positive cells to the number of Pax7‐positive cells was quantified. Data are presented as the mean ± SEM. *P < 0.05, ^** P < 0.01 and ^*** P < 0.001 (Student's t test).

Figure 7

TAZ depletion in vivo reduces the number of satellite cells and myofibre diameter after exercise. (A) Wt and sKO mice were subjected to endurance exercise training. The GA muscles were harvested from these mice, and tissue immunostaining was performed. H&E and dystrophin (green) immunofluorescence staining of GA muscles from exercise‐trained wt and sKO mice. Scale bar, 50 μm. Muscle fibre diameters were quantified using ImageJ software. (B) TAZ, Pax7, Myf5, Cyclin D1, Rhebl1, phospho‐p70 S6K, p70 S6K, phospho‐4E‐BP, 4E‐BP and Vinculin levels in GA muscles of exercise‐trained wt and sKO mice were analysed by immunoblotting. (C) Taz, Pax7, Myf5 and Rhebl1 transcript levels were analysed by qRT‐PCR in GA muscles of exercise‐trained wt and sKO mice (n = 5). (D) Immunofluorescence staining of Pax7 (red) in GA muscles of exercise‐trained wt and sKO mice. Nuclei are stained with DAPI (blue). Scale bar, 50 μm. The number of Pax7‐positive cells was quantified relative to the number of fibres. (E) Immunofluorescence staining of Pax7 (red) and Myf5 (cyan) in GA muscles of exercise‐trained wt and sKO mice. Scale bar, 50 μm. The ratio of the number of Myf5‐positive cells to the number of Pax7‐positive cells was quantified. Data are presented as the mean ± SEM. *P < 0.05, ^** P < 0.01 and ^*** P < 0.001 (Student's t test).

Figure 8

Proposed model. Quiescent satellite cells were activated by exercise or muscle injury. TAZ expression was upregulated in activated satellite cells, and the increased TAZ translocated to the nucleus and interacted with Pax7 and TEAD transcription factors. Myf5 transcription, which promotes the commitment of satellite cells to myogenic differentiation, was induced by Pax7 and TAZ interaction, and Rheb/Rhebl1 transcription, which stimulates cell cycle progression by promoting mTOR signalling activation, was induced by TEAD and TAZ interaction. As a result, TAZ stimulates satellite cell expansion and promotes skeletal muscle repair.