Synergistic effects of TGFβ2, WNT9a, and FGFR4 signals attenuate satellite cell differentiation during skeletal muscle development

Summary Satellite cells play a key role in the aging, generation, and damage repair of skeletal muscle. The molecular mechanism of satellite cells in these processes remains largely unknown. This study systematically investigated for the first time the characteristics of mouse satellite cells at ten different ages. Results indicated that the number and differentiation capacity of satellite cells decreased with age during skeletal muscle development. Transcriptome analysis revealed that 2,907 genes were differentially expressed at six time points at postnatal stage. WGCNA and GO analysis indicated that 1,739 of the 2,907 DEGs were mainly involved in skeletal muscle development processes. Moreover, the results of WGCNA and protein interaction analysis demonstrated that Tgfβ2, Wnt9a, and Fgfr4 were the key genes responsible for the differentiation of satellite cells. Functional analysis showed that TGFβ2 and WNT9a inhibited, whereas FGFR4 promoted the differentiation of satellite cells. Furthermore, each two of them had a regulatory relationship at the protein level. In vivo study also confirmed that TGFβ2 could regulate the regeneration of skeletal muscle, as well as the expression of WNT9a and FGFR4. Therefore, we concluded that the synergistic effects of TGFβ2, WNT9a, and FGFR4 were responsible for attenuating of the differentiation of aging satellite cells during skeletal muscle development. This study provided new insights into the molecular mechanism of satellite cell development. The target genes and signaling pathways investigated in this study would be useful for improving the muscle growth of livestock or treating muscle diseases in clinical settings.

The development of muscle stem cells can be divided into two important phases: embryonic and postnatal stages. At the embryonic stage, PAX7 is first expressed in the central dermomyotome and then colocalized with PAX3 in the myotome (Relaix, Rocancourt, Mansouri & Buckingham, 2005). The PAX3 + /PAX7 + cells then become embryonic muscle progenitor cells, which enter into the myogenic process and are differentiated into myoblasts with the expression of MYF5 and MYOD (Bober et al., 1991;Rudnicki et al., 1993;Sassoon et al., 1989). At the postnatal stage, satellite cells originate from PAX3 + /PAX7 + cells. In general, satellite cells gradually enter a quiescence state after birth (Chakkalakal, Jones, Basson & Brack, 2012;Charge & Rudnicki, 2004;Sato, Yamamoto & Sehara-Fujisawa, 2014). Once damage occurs, satellite cells became activated and triggered the regeneration and reconstruction of skeletal muscles (Charge & Rudnicki, 2004;Collins et al., 2005;Meeson et al., 2004). A previous study indicated that muscle regeneration was attenuated due to the depletion of satellite cells in adult muscle (Fry et al., 2015). The molecular mechanism of this phenomenon remains largely unknown. Understanding the molecular characteristics of satellite cells during development would aid the study of muscle regeneration after damage, particularly for aging muscle.
Previous studies indicated that several signaling pathways participated in the myogenesis of satellite cells. The FGF signaling pathway could induce the myogenic differentiation of muscle progenitor cells (Marics, Padilla, Guillemot, Scaal & Marcelle, 2002). Pax3 could activate the FGF signaling by upregulating the expression of Fgfr4 and Sprouty1 (Lagha et al., 2008). The WNT and TGFb signaling pathways could induce the fibrogenesis of satellite cells in dystrophic mice (Biressi, Miyabara, Gopinath, Carlig & Rando, 2014). The TNF, AKT, and MAPK signaling pathways participate in the proliferation and differentiation of satellite cells (Motohashi et al., 2013;Troy et al., 2012). However, the synergistic effects of different signaling pathways remain largely unknown.
This study mainly focused on the molecular mechanism of satellite cells at the postnatal stage. The results revealed that the number and differentiation capacity of satellite cells decreased during development. The results also indicated that the synergistic effects of TGFb2, WNT9a, and FGFR4 signals were responsible for attenuating the differentiation of satellite cells during postnatal development. This study provided new insights into the molecular mechanism of satellite cell development during the postnatal stage.
The genes and signaling pathways identified in this study would be useful targets for improving the muscle growth or clinical therapeutics of muscle diseases.

| Dynamic expression patterns of marker genes of satellite cells during postnatal development
To investigate the development of satellite cells in postnatal skeletal muscle, we examined the expression patterns of the marker genes.
The gastrocnemius muscle tissues at 10 different time points (Day 1, Day 8, Week 2, Week 4, Week 6, Week 8, Week 10, Week 12, Week 24, and Week 52) were obtained, followed by the detection of the expression of the marker genes through immunofluorescence analysis. The immunofluorescence results indicated that PAX7 + cells accounted for 19.7% on Day 1, and this value markedly decreased during development, accounting for <0.5% after Week 10 (Figure 1a, b, and Supporting Information Figure S1). MYF5 + cells only slightly decreased before Week 8 but sharply decreased at Week 10, and it remained at low levels (<20%) in the subsequent weeks ( Figure 1c and Supporting Information Figure S1). Myogenin + cells gradually declined from Day 1 to Week 2 but significantly increased at Week 4 and Week 6 ( Figure 1d and Supporting Information Figure S1).
MYOD-positive cells maintained low levels throughout the 10 different postnatal time points (Figure 1e and Supporting Information Figure S1).

| Differentiation capacity of satellite cells attenuated during development
To further elucidate the differentiation of satellite cells, an in vitro study was performed. First, skeletal muscle satellite cells were isolated from the hindlimb muscle of mice at six different time points (Week 2, Week 4, Week 6, Week 8, Week 10, and Week 12).
Immunofluorescence staining showed that more than 90% of the isolated cells were PAX7 and MYF5 double positive (Figure 2a,b).
Then, the isolated satellite cells were induced with differentiation for 24 or 48 hr. The differentiation capacity was evaluated through immunofluorescence staining and quantitative polymerase chain reaction (qPCR) methods. The result of differentiation for 24 hr indicated that the myosin expression decreased with development, especially after Week 6 ( Figure 2c). In the same way, the result of differentiation for 48 hr showed that the myosin expression levels were comparable at all stages, although the myotube size appeared more slender at Week 10 and Week 12 than those at the early stages (Supporting Information Figure S2A). qPCR results presented that the Mck expression level significantly decreased from Week 6 to Week 12 as compared to Week 2 (Figure 2d, Supporting Information Figure S2B). These results indicated that the differentiation capacity of satellite cells decreased with skeletal muscle development.

| Transcriptome of satellite cells during skeletal muscle development
To further understand the molecular mechanism of satellite cells during skeletal muscle development, the transcriptome profiles were detected using RNA-seq. First, the satellite cells were isolated from the hindlimb muscle at six time points (Week 2, Week 4, Week 6, Week 8, Week 10, and Week 12). After sequencing, the differentially expressed genes (DEGs) between any two time points were further analyzed (p < 0.01; FDR <0.05). A total of 2,907 DEGs were identified. Furthermore, WGCNA revealed that these DEGs were enriched in six main expression modules. The largest module  Figure S4). Therefore, the Tgfb, Wnt, and Fgf signaling pathways were identified as the key factors responsible for the differentiation of satellite cells.

| Tgfb2 and Tgfb3 genes inhibited the differentiation of satellite cells
The roles of Tgfb2 and Tgfb3 in the differentiation of satellite cells were further investigated. Satellite cells were first isolated from the hindlimb muscle of 4-week-old mice. Then, the expression levels of Tgfb2, Tgfb3, Akt2, and Mknk2 were detected during proliferation and differentiation using qPCR. The results showed that Tgfb2, Tgfb3, Akt2, and Mknk2 were significantly upregulated during differentiation (Supporting Information Figure S5A). Immunofluorescence staining results revealed that after differentiation for 24 hr, myosin increased in si-Tgfb2-and si-Tgfb3-transfected satellite cells as com- of pirfenidone treatment. The western blot results showed that the TGFb2 in the gastrocnemius muscle tissue was markedly lower than that in the control mice. In addition, FGFR4, WNT9a, PAX7, and MYOD were upregulated in the gastrocnemius muscle of pirfenidone-treated mice (Figure 7b). The pirfenidone-treated and control groups were compared in terms of the number of Pax7 + cells.
The immunofluorescence results indicated that Pax7 + cells were significantly reduced in the pirfenidone-treated mice (p < 0.01). A muscle-damaged mice model was created through the injection of cardiotoxin (CTX). On Day 6 of CTX injection, the number of Pax7 + cells in the damaged gastrocnemius muscle was significantly higher than that in uninjured muscle. Moreover, the number of Pax7 + cells in the pirfenidone-treated mice was significantly higher than that in the control mice (Figure 7c,d). The hematoxylin and eosin (H&E) staining results showed that the skeletal muscle regeneration in the pirfenidone-treated mice was preceded than that in the control mice Other genes are shown in the circle on the left. Weight value between genes is represented by width and transparency of edges. (b) qPCR was performed to validate the expression of genes. Tubulin was used as the internal control, and the relative fold change was compared to the expression in Week 2 satellite cells. Triplicate samples were analyzed for each treatment, and the results are presented as the mean AE SEM *p < 0.05, **p < 0.01 we first analyzed the expression patterns of the marker genes of satellite cells during skeletal muscle development. Then, we examined the differentiation capacity of satellites cells at different ages.
Our results indicated that the differentiation capacity of satellite cells was attenuated with development. Moreover, the molecular mechanism of the attenuation of differentiation was assessed by performing transcriptome and functional analyses. We concluded that the synergistic effects of TGFb2, WNT9a, and FGFR4 signals were responsible for attenuating the differentiation of skeletal muscle satellite cells.
In the study, we found that the number of PAX7 + cells was rapidly (e, f) Western blot results of TGFb2, TGFb3, WNT9a, FGFR4, AKT2, p-AKT2, and MKNK2 in proliferative satellite cells when TGFb2 and TGFb3 was inhibited using siRNA (e) or pirfenidone (f). Tubulin was used as the internal control for qPCR and western blot. Triplicate samples were analyzed for each treatment, and the results are presented as the mean AE SEM *p < 0.05; **p < 0.01 upstream genes, namely FGFR4, TGFb2, TGFb3, and WNT9a, were further investigated. FGFR4 could promote the differentiation of satellite cells, whereas TGFb2, TGFb3, and WNT9a displayed an opposite function. In vitro analysis showed that TGFb2 and WNT9a could increase PAX7 and inhibit MYOD at the protein level, whereas FGFR4 could downregulate PAX7 and upregulate MYOD at the protein level in satellite cells. In vivo analysis indicated that the inhibition of TGFb2 could downregulate the expression of PAX7 and promote the regeneration of skeletal muscles. Previous studies indicated that the activation of quiescent satellite cells was accompanied by reduced PAX7 and increased MYOD (Kostallari et al., 2015;Sato et al., 2014). The overexpression of TGFb2 and TGFb3 can decrease the myogenic differentiation of myoblasts (de Mello, Streit, Sabin & Gabillard, 2015;Schabort, van der Merwe & Niesler, 2011). Moreover, the knockout of FGFR4 attenuated the skeletal muscle regeneration (Zhao et al., 2006). Therefore, TGFb2, WNT9a, and FGFR4
The suspension was plated on a normal dish and then transferred to a dish coated with Matrigel (BD, USA, 356234) after 2.5 hr. The satellite cells were cultured at 37°C in a cell incubator with 5% CO 2 until they converged to 60%. Then, the second differential attachment experiment was performed. The differentiation medium was composed of Dulbecco's modified Eagle's medium (DMEM) and 3% (v/v) horse serum (Gibco). Pirfenidone (Selleck, USA;20 lg/ml;Burghardt et al., 2007) was used to stimulate the satellite cells continuously for 24 hr.

| Cell transfection
For RNAi assay, the isolated satellite cells were transfected with siRNA using Lipofectamine 2000 (Invitrogen, USA) in accordance with the manufacturer's recommendations after the cells converged to approximately 60%. The siRNA and the scrambled negative control were provided by RIBOBIO (RIBOBIO, P.R.C).
F I G U R E 8 Schema graph of TGFb2, WNT9a, and FGFR4 signals in the differentiation of satellite cells during skeletal muscle development Then, the gastrocnemius muscle was acquired for protein extraction and immunofluorescence. For satellite cell isolation, the total muscle of the hind leg was used (n = 8). Regeneration assay was also performed. In the same way, 4-week-old mice were divided into two groups (n = 9 for each group). Pirfenidone (250 mg kg À1 day À1 ) or vehicle (control) was orally administered daily. After 1 day of the first pirfenidone treatment, CTX was injected into gastrocnemius with 100ll of 10lM (Qiu et al., 2016). Then, the gastrocnemius was isolated on Day 2, Day 6, and Day 12 after CTX injection to analyze the tissue morphology. The cell nucleus was washed thrice with PBS and stained with 4 0 , 6diamidino-2-phenylindole (DAPI) (Wei et al., 2013). Images were captured using a Nikon Eclipse TE2000-S system (Nikon, Japan).

| RNA-seq
The total RNA was extracted from isolated satellite cells at six different time points using RNeasy Mini Kit (Qiagen, Germany, 74106) in accordance with the manufacturer's instructions. Qualified total RNA was further purified using the RNAClean XP Kit (Beckman Coulter, Inc., Kraemer Boulevard Brea, CA, USA, A63987) and the RNase-Free DNase Set (Qiagen, 79254). RNA and the library preparation integrity were verified with an Agilent Bioanalyzer 2100 (Agilent Technologies, Santa Clara, CA, USA). We accomplished the cluster and first dimension sequencing primer hybridization on cBot of Illumina sequencing machine in accordance with the cBot User Guide.
Sequencing was performed by Shanghai Biotechnology Corporation (P.R.C). Edger, which is an R package, was used to screen the DEGs.

| qPCR
Reverse transcription was performed to initiate cDNA synthesis using the Prime ScriptTM RT Reagent Kit with gDNA Eraser (TAKARA BIO INC, Otsu, Shiga, Japan). THUNDERBIRD SYBR qPCR Mix (TOYOBO, Japan) was used for qPCR, and the results were monitored using a CFX384 Real-Time PCR Detection System (Bio-Rad, USA). All primer sequences are listed in the supplementary data (Supporting Information Table S2).

| Western blot
The Mammalian Protein Extraction Reagent (Pierce, USA) was used to obtain the protein lysate. SDS-PAGE was used to separate the proteins, and a Mini Trans-Blotting Cell (Bio-Rad) was used to transfer protein onto polyvinylidene fluoride membranes (Millipore, USA).
An Image Quant LAS4000 mini (GE Healthcare Bio-Sciences, USA) was used to detect the signal produced by the Immobilon Western Chemiluminescent HRP Substrate (Millipore).

| Statistical analysis
All results are expressed as mean AE SEM. Unpaired Student's t tests were used to determine the statistical significance, and p < 0.05 indicated a significant difference. University Scientific and Technological Self-Innovation Foundation.

CONFLI CT OF INTEREST
The authors declare that they have no competing or financial interests.

AUTHOR CONTRI BUTIONS
Weiya Zhang, Lu Zhang, Yinlong Liao, Sheng Wang, and Binxu Yin conducted the experiments and prepared the materials involved in