Myostatin Deficiency Enhances Antioxidant Capacity of Bovine Muscle via the SMAD-AMPK-G6PD Pathway

During exercise, the body's organs and skeletal muscles produce reactive oxygen species (ROS). Excessive ROS can destroy cellular lipids, sugars, proteins, and nucleotides and lead to cancer. The production of nicotinamide adenine dinucleotide phosphate (NADPH) by the pentose phosphate pathway (PPP) is an auxiliary process of the cellular antioxidant system that supplements the reducing power of glutathione (GSH) to eliminate ROS in the cell. Myostatin (MSTN) is mainly expressed in skeletal muscle and participates in the regulation of skeletal muscle growth and development. Loss of MSTN leads to muscular hypertrophy, and MSTN deficiency upregulates glycolysis. However, the effect of MSTN on the PPP has not been reported. This study investigated the effect of MSTN on muscle antioxidant capacity from a metabolic perspective. We found that reducing MSTN modulates AMP-activated protein kinase (AMPK), a key molecule in cellular energy metabolism that directly regulates glucose metabolism through phosphorylation. Downregulation of MSTN promotes tyrosine modification of glucose-6-phosphate-dehydrogenase (G6PD) by AMPK and is regulated by the Smad signaling pathway. The Smad2/3 complex acts as a transcription factor to inhibit the AMPK expression. These results suggest that reduced MSTN expression inhibits the Smad signaling pathway, promotes AMPK expression, enhances the activity of G6PD enzyme, and enhances the antioxidant capacity of nonenzymatic GSH.


Introduction
MSTN is a member of the transforming growth factor-β superfamily that functions as a negative regulator of skeletal muscle mass. Inhibition of MSTN and knockdown of the MSTN gene cause muscle hypertrophy, which results in 'double-muscled' cattle, sheep, mice, rabbits, dogs, and humans [1]. MSTN is predominately expressed in skeletal muscles but is also expressed in other tissues [2,3]. As a signaling ligand, MSTN modulates muscle mass development through the MSTNactivin receptor IIB-(ActRIIB-) Smad2/3 signaling pathway [4,5]. Basic metabolic changes under the influence of MSTN indicate that MSTN plays an important role in the regulation of energy metabolism. In C2C12 cells, overexpression of MSTN results in decreased glucose transporter type 4 (Glut4) gene expression and glucose uptake, as well as decreased active phosphorylation of insulin receptor substrate 1 (IRS-1) and phosphoinositide 3-kinase (PI3K), decreased phosphorylation of AKT, and increased phosphorylation of glycogen synthase kinase 3β (GSK3β) [6]. Treatment with antimyostatin antibody increases insulin-stimulated glucose metabolism in elderly mice, possibly due to increased glucose uptake in insulin-stimulated skeletal muscle [7]. MSTN inhibits glycolysis and increases glycogen accumulation [8]. In MSTN −/− mice, enolase activity, a key component of glycolysis, increases along with serum lactate level in the resting state and after exercise, and the soleus muscle changes from the "slow oxidative" type to the "fast glycolytic" type [9]. In contrast, in the presence of 0.1-1.5 μg/ml recombinant MSTN, glycolysis in C2C12 cells is increased, and glycogen synthesis is inhibited [10]. MSTN seems to regulate glycolysis and glycogen synthesis through the AMPK pathway [8].
The pentose phosphate pathway (PPP) is a way of oxidative decomposition of glucose. The PPP reaction system starts with glucose-6-phosphate, which produces pentose, inorganic phosphoric acid, and NADPH in oxidative decomposition. NADPH can provide reducing capacity for GSH and enhance antioxidant capacity [11,12]. G6PD, the first rate-limiting PPP enzyme, controls NADPH production [13] and modulates the cytoplasmic redox state [14].
AMPK is a cellular energy status sensor and key regulator of mitochondrial function and oxidative stress in skeletal muscle. AMPKγ3 can affect phosphorylation of AMPK1/2 and prevent dephosphorylation of protein phosphatase Thr172 [15]. MSTN affects protein and fat metabolism by inhibiting the AMPK signaling pathway. Low concentrations of MSTN can inhibit the eukaryotic elongation factor-2 kinase (eEF2K)/eukaryotic elongation factor-2 (eEF2) pathway through AMPK signals and suppress protein synthesis, but low MSTN concentration does not affect the mTOR pathway [16]. Blocking MSTN with soluble activin receptor IIb (sActRIIB-Fc) enhances mTOR complex 1 (mTORC1) signaling, increases regulated in development and DNA damage responses 1 (REDD1) protein and eukaryotic translation initiation factor 2B subunit epsilon (eIF2Bε) protein, and improves phosphorylation of eukaryotic translation initiation factor 4E-binding protein 1 (4E-BP1) and AMPK [17]. Loss of MSTN suppresses fat accumulation with a 70% reduction in total body fat and approximately 25% and 40% reductions in gonadal and genital fat pads, respectively [18]. Molecular evidence suggests that MSTN mutations increase fatty acid oxidation and enhance brown fat formation by increasing lipolysis activity [19]. The transformation of white adipose tissue to brown adipose tissue is mediated by peroxisome proliferator-activated receptor γ coactivator-1α (PGC1α-irisin) [20] and the AMPK-PGC1α-fibronectin type III domain-containing protein 5 (Fndc5) pathway [21].
Although the contribution of MSTN to glycolysis has been extensively analyzed, similar analyses of the PPP and NADPH metabolism are lacking. In this study, we demonstrate that downregulation of MSTN enhances the antioxidant capacity of bovine muscle through PPP and Smad-AMPK-G6PD axis.

Materials and Methods
2.1. Cattle. We used CRISPR/Cas9 and somatic cell nuclear transfer to generate MSTN knockout cattle. The MSTN +/cattle were the offspring of MSTN -/bulls bred with wildtype (WT) cows [22,23]. Selected MSTN +/-(MT) and WT 24-month-old heifers were used in this study. All heifers were fed in the same pasture with the same feeding consistency. All experiments were performed in strict accordance with the guidelines of the Experimental Animal Management and Operation Standards of Inner Mongolia University (no. IMU-CATTLE-2020-034, 01 June 2020).

Cattle
Muscle Satellite Cell Culture. The longissimus dorsi muscles were dissected and cut into pieces. The cut tissue pieces were immersed in 75% alcohol for 10 s, followed by 3 immersions in PBS (FG701-01, TRANS, China) containing 2% Pen Strep (15140-122, Gibco, USA) for 20 s. The muscle pieces were then soaked in DMEM F12 (11330107, Gibco, USA), 20% FBS (10099141, Gibco, USA), 10% HS (S9050, Solarbio, China), and Pen Strep-for 10 s. Then, the tissue pieces were placed upside down in a 100 mm dish in an incubator (5% CO 2 , 38.5°C) for 2 h. Seven to eight milliliters of culture medium were added to the dishes, and the tissue was cultured until satellite cells appeared.
2.3. Quantitative Real-Time PCR. Total RNA was isolated from tissues using the RNAiso Plus kit (9109, Takara, Japan) according to the manufacturer's instructions. The RNA was reverse transcribed into cDNA using a cDNA reverse transcription kit (RR036A, Takara, Japan). All primer sequences are summarized in Table S1. Real-time PCR (qPCR) was performed using a real-time PCR detection system (7500, ABI, USA). The reaction mixture (20 μl) contained 1 μl cDNA, 0.5 μl of each primer, and 10 μl TB Green Supermix (RR820A, Takara, Japan). After the initial denaturation step at 95°C for 30 s, 40 cycles at 95°C for 15 s and 60°C for 30 s were performed. Gene expression was detected by 2 −ΔΔCT relative quantitative analysis. All experiments were repeated more than three times.
2.8. Coimmunoprecipitation. The coimmunoprecipitation (Co-IP) process followed the instructions of the Pierce classic IP kit (26146, Thermo, USA). To lyse the muscle tissues and cells, 1 ml ice-cold IP lysis buffer was added to the tissues and cells, which were incubated on ice for 5 min with periodic mixing. The supernatant was transferred to a new tube, and 1 ml of lysate was prepurified with 80 μl agarose resin. Ten micrograms of affinity-purified antibody were combined with 600 μl precleared cell lysate overnight in a 4°C microcentrifuge tube to form an immune complex. The immune complex was captured with protein A/G and agarose in the spin column and washed three times with 200 μl washing buffer. The immune complexes were eluted with sample buffer and incubated at room temperature for 5-10 min, and the eluents were collected by centrifugation (10000 g). Samples were left at -20°C prior to SDS-PAGE gel application.
2.9. Statistical Analyses. All data were expressed as the mean ± SD of at least three independent trials. In the graphs, bars represent means, while error bars represent one standard deviation. Welch's two-tailed t-test was used for statistical analysis when data from two groups with different standard deviations were compared. When comparing multiple groups, a repeatable one-way ANOVA was performed, followed by a post hoc analysis using Bonferroni's correction to adjust for multiple comparisons. * p < 0:05 and * * p < 0:01 were considered statistically significant. The histograms were drawn in Prism 7.0 (GraphPad, USA), and phototypesetting was done with Adobe Photoshop CS3.

The Reduction of MSTN Alters Muscle Transcriptome.
The soleus muscle had lesser mass in WT cattle compared to MT cattle (Figure 1(a)). Accordingly, the concentration of MSTN protein per volume of plasma in MT cattle was only 22:11 ± 1:09 ng/ml, which was significantly lower than in WT cattle 38:92 ± 3:48 ng/ml (p < 0:01, Figure S1A). The MSTN were also downregulated in MT cattle muscles (Figure 1(b)). Transcriptome sequencing analysis of seven MT cattle and three WT cattle showed significant differences in gene expression between the MT and WT cattle (Figure 1(c)). A total of 342 differentially expressed genes were identified, of which 312 were downregulated and 30 upregulated in MT muscle (Figure 1(c)). KEGG enrichment analysis was performed on metabolic pathways. Upregulated genes included genes involved in glycolysis and gluconeogenesis, biosynthesis of amino acids, carbon metabolism, and fructose and mannose metabolism (Figure 1(d)).

The Reduction of MSTN Promotes Glucose Catabolism.
The transcriptome data indicate that most of the upregulated genes are associated with glucose metabolism. To determine the effect of mutating MSTN on glucose metabolism, the levels and activity of products and enzymes involved in glucose catabolism were measured. MT cattle had significantly lower blood glucose than WT cattle (Figure 2(a)). However, muscle glucose levels were significantly higher in the MT group than in the WT group (Figure 2(b)). The expression of the glucose transporter gene GLUT1 is increased in MT cattle, while GLUT4 expression is similar to WT cattle (Figure 2(f)). The content of insulin, the regulator of glucose transporters, did not differ between MT and WT cattle ( Figure S2A). These results imply that more blood glucose enters muscle cells in the MT cattle vs. WT cattle.
During glycolysis, the activities of most rate-limiting enzymes are increased. Accordingly, hexokinase (HK) and pyruvate kinase (PK) were significantly higher in MT cattle than in WT cattle (Table 1). Phosphofructokinase (PFK), the most important rate-limiting enzyme in glycolysis, was significantly lower in MT cattle (144:10 ± 15:01 nmol/min/   (Table 1). PFK protein in the MT group was also lower than the WT group (Figure 2(f)). Fructose-1,6-diphosphate (FDP), a product of PFK acting on fructose 6-phosphate, was significantly lower in MT cattle (3:90 ± 0:33 mg/g) compared to WT cattle (6:07 ± 0:53 mg/g) (p < 0:01, Table 2). Triosephosphate isomerase (TPI), the conversion enzyme between dihydroxyacetone phosphate and glyceraldehyde-3-phosphate, was significantly lower in MT cattle muscle (Table 1), and the concentration of pyruvate (PA) in muscle and plasma was significantly higher (p < 0:01, Table 2, Figure S2B). The lactic acid (LA) content in MT cattle muscle was significantly upregulated ( Table 2), but we saw no significant difference in the content of LA in plasma between MT and WT cattle ( Figure S2C).
Muscular glucose catabolism involves glycolysis and the pentose phosphate pathway. The mRNA of key enzymes such as phosphoglucose dehydrogenase (G6PD), lactonase, and 6-phosphogluconate dehydrogenase (6PGDH) in the oxidation stage of the PPP were more highly expressed in  (Figure 2(c)). The enzymatic activity of G6PD was significantly higher in MT cattle compared to WT cattle (14:74 ± 0:79 vs. 8:79 ± 1:20 nmol/min/g), and the protein content was similarly increased (Figures 2(d) and 2(f)). Muscle NADPH concentration, the product of the PPP oxidation stage, was significantly higher in MT cattle (63:85 ± 6:3 vs. 37:52 ± 3:15 nmol/g), while NADP + was significantly lower (22:54 ± 2:03 vs. 35:03 ± 1:95 nmol/g) (p < 0:01, Figure 2(e)), suggesting that NADP + is more actively converted to NADPH in the PPP in MT vs. WT cattle. At the same time, the concentrations of key enzymes such as ribulose-5-phosphate epimerase (RPI), transketolase (TKL), and MT muscle protein in the nonoxidative phase of the PPP are significantly increased (Figure 2(f)). From the perspective of glycogen synthesis, the glycogen accumulation in the muscle of MT cattle is lower than in WT cattle ( Figure S2D). The mRNA expression of glycogen synthase kinase-3β (GSK3β), which negatively regulates the expression of the GS gene, was significantly increased compared to the control (p < 0:01, Figure S2E). Reduced MSTN expression inhibited the first stage of glycolysis, while upregulated PPP provided a substrate for the second stage of glycolysis, which together produced more pyruvate and LA.

The Reduction of MSTN Promotes Muscle Antioxidant
Capacity through G6PD in the Pentose Phosphate Pathway. The PPP provides reducing power for antioxidants. Based on the above results, the antioxidant capacity of muscles was studied. The content of oxygen free radicals (ROS) in the muscles of MT cattle was significantly lower than in WT cattle (124:2 ± 11:44 vs. 180:1 ± 9:18 U/g) (p < 0:01,  (Figures 3(b) and 3(c)). Cellular ROS is mainly produced by mitochondrial complex I and complex III. In MT cattle, the activity levels of complex I and complex III did not differ compared to WT cattle (Figures 3(d) and 3(e)). We tested the antioxidants of cattle muscle and found that the antioxidants related to enzymes and GSH were significantly increased in MT cattle. Glutathione peroxidase (GSH-PX), catalyzer of GSH into GSSG, reduces the toxic peroxides to nontoxic hydroxyl compounds. Mutation of MSTN enhanced GSH-PX activity (Figure 3(f)). Glutathione reductase (GR), the enzyme that catalyzes the transformation of glutathione disulfide (GSSG) to GSH, was higher in MT cattle than in WT cattle (Figure 3(g)). Superoxide dismutase (SOD) and catalase (CAT) are the most active members of the enzyme antioxidant system, and there was no significant difference in either between MT and WT cattle (Figures S3A, B). In the nonenzymatic antioxidant system, GSH is an important reducing agent in cells, and NADPH produced by the PPP promotes the production of GSH [11]. In MT cattle, the level of GSH was higher than in WT cattle (1:029 ± 0:06 vs. 0:695 ± 0:08 μmol/g) (p < 0:01, Figure 3(h)), as was the GSH/GSSG ratio ( Figure S3D). In contrast, GSSG, the oxidated state of GSH, was significantly lower in MT cattle (2:056 ± 0:24 vs. 3:79 ± 0:74 nmol/g) (p < 0:05, Figure 3(i)). The activity of superoxide radical scavenging (SRSA) was enhanced in MT cattle (0:22 ± 0:01% vs. 0:14 ± 0:02%) (p < 0:01, Figure S3C). ROS damage sodium-potassium (Na + -K + ) pumps and calcium-magnesium (Ca 2+ -Mg 2+ ) pumps [24, ]. The activities of Na + -K + pumps and Ca 2+ -Mg 2+ pumps in MT muscle were both higher than those in the WT cattle muscle (Figures S3E, F). These results suggest that reducing MSTN reduces muscle oxidative damage by increasing GSH.
3.4. The Reduction of MSTN Enhanced AMPK Interaction with G6PD. The loss of MSTN affects the TGF-β signaling pathway. In MT cattle, the protein level of p-AKT was significantly upregulated, while the levels of p-Smad2 and p-Smad3 were downregulated, and the level of p-p38 was unchanged (Figures 4(a)-4(c)). Smad2/3 is a transcription factor that regulates gene expression. ChIP-qPCR with anti-Smad2+Smad3 monoclonal antibody further demonstrated that Smad2/3 hardly binds to the G6PD promoter region (Figure 4(d)). These results indicate that MSTN does not regulate G6PD transcription in MT cattle via Smad2/3. Co-IP results showed that p-AKT interacts with G6PD, but knockdown of MSTN does not affect their interaction (Figure 4(e)). Loss or overexpression of MSTN affects AMPK [8,17,18], which is further proven by the finding that phosphorylation of AMPK in MT cattle was significantly increased (Figure 4(c)). Co-IP identification showed that knockdown of MSTN enhances the interaction between p-AMPKα1+α2 and G6PD (Figure 4(e)). This result suggests that the reduction of MSTN enhances the AMPK signaling pathway and increases the production of G6PD and that MSTN affects the expression of G6PD through AMPK.
3.6. MSTN Affects the Content of GSH through TGF-β-AMPK-G6PD. Protein sequence BLAST comparisons showed that G6PD is highly conserved, with over 90% similarity of sequences among bovines, mice, and humans ( Figure S4C). A total of 19 phosphorylation sites are located in the G6PD protein sequence, 18 of which are the same in cattle, mice, and humans. Only the 490 th amino acid is different, which is phosphorylated tyrosine in humans and mice and valine in cattle ( Figure S4C).
Co-IP results showed that MT cattle have significantly higher tyrosine phosphorylation levels G6PD than WT cattle, but we observed no difference in serine and threonine phosphorylation (Figures 6(a) and 6(b)). In WT muscle satellite cells, G6PD expression levels after treatment with AMPK inhibitor and Smad activator were significantly lower than in the untreated WT cattle group (Figures 6(c)). In WT

10
Oxidative Medicine and Cellular Longevity satellite cells, the expression level of G6PD in satellite cells treated with both Smad and AMPK inhibitors also showed a downward trend compared with untreated WT cattle muscle (Figures 6(c)). However, in WT satellite cells, when an AMPK activator and a Smad inhibitor were added, G6PD expression and tyrosine phosphorylation levels were significantly increased compared to the WT cattle untreated group (Figures 6(c) and 6(d)). The content of GSH in WT cattle muscle satellite cells was increased by adding an AMPK activa-tor and a Smad inhibitor (Figures 6(e)). These results suggest that reduction of MSTN modulates the pentose phosphate pathway through the Smad-AMPK-G6PD axis, thereby increasing the antioxidant capacity of the muscle (Figure 7).

Discussion
Muscle activity increases ROS while increasing the body's antioxidant content. Antioxidants neutralize free radicals by accepting unpaired electrons, thereby inhibiting the oxidation of other molecules [26]. Antioxidant systems that remove ROS can be divided into two categories: enzymatic antioxidant systems and nonenzymatic antioxidant systems (nonenzymatic: GSH, vitamin E, vitamin C) [27]. MSTN generates ROS through Smad3, NF-κB, and TNF-α signaling pathways [28,29]. This study found that the lack of MSTN enhances the antioxidant capacity of skeletal muscle by upregulating the content of GSH, an antioxidant that reacts with ROS to form GSSG, thereby reducing the release of ROS [30]. However, GR mainly uses NADPH produced by the pentose phosphate pathway to provide reducing power to convert GSSG into GSH and maintain the balance of intracellular GSH [31][32][33]. G6PD is a rate-limiting enzyme that controls the production of NADPH. The activity of G6PD is regulated by multiple signaling pathways during transcription, posttranslation, and intracellular localization [12]. At the transcription level, G6PD activity is regulated by transcription factors Yin-Yang 1 (YY1), Sp1, CREB, TAp73, and Nrf2 [34][35][36][37][38]. Insulin stimulation of G6PD mRNA expression involves the PI 3 K pathway [39,40], and arachidonic acid inhibits insulin by stimulating the AMPK and p38 MAPK pathways [41,42]. In this study, there was no significant change in serum insulin levels, Smad2+3 did not bind to the G6PD promoter, and knocking out MSTN did not affect the binding of AKT to G6PD. T cell leukemia/lymphoma protein 1A (Tcl1) increases G6PD pre-mRNA splice and protein expression in a heterogeneous nuclear ribonucleoprotein-(hnRNPK-) dependent manner. Meanwhile, phos-phatase and tensin homolog (PTEN) forms a complex with hnRNPK to inhibit G6PD pre-mRNA splice, and PTEN inactivates Tcl1 via GSK3β-mediated phosphorylation [25]. This experiment used qPCR analysis on GSK3β, and the results show that the expression of GSK3β is upregulated in MT cattle. However, whether GSK3β can regulate G6PD pre-mRNA splice and protein expression after MSTN deletion has not been determined.
AMPK is an important regulator of lipid and sugar metabolism, regulating enzyme activity through the phosphorylation of transcription factors, coactivators, and coinhibitors to achieve transcriptional regulation of metabolism, resulting in reduced anabolism and increased catabolism [40]. AMPK prevents oxidative stress by maintaining NADPH [41]. In AMPK and G6PD studies, changes in the PPP alter AMPK activity in cardiomyocytes [43]. However, 5-phosphoribulose (Ru-5-P), the product of the PPP, inhibits the LKB1-AMPK signaling pathway. Downregulation of G6PD does not affect AMPK activation [44]. In breast cancer cells, GL-V9 promotes AMPK expression and activity, leading to a decrease in G6PD [45]. This study found that mutation of MSTN enhances G6PD tyrosine phosphorylation by upregulating AMPK, leading to an increase in GSH content. The connection between MSTN and AMPK is achieved through the Smad2/3 transcription factor. Previous studies have found that MSTN reduces AMPK activity, Glut4 gene expression, and glucose uptake and that the addition of AMPK activator AICAR reverses these effects [6]. In this study, the Smad2/3 transcription factor binds to inhibit the transcription of the AMPKγ3  Figure 7: MSTN affects G6PD tyrosine phosphorylation through AMPK to improve muscle antioxidant capacity. MSTN regulates the AMPK pathway through the Smad2/3 pathway, and the upregulated AMPK activates the phosphorylation of G6PD tyrosine. The enhancement of G6PD upregulates the pentose phosphate pathway, thereby enhancing the content of nonenzymatic antioxidant GSH and improving the antioxidant capacity of muscles. In addition, the pentose phosphate pathway supplements the second stage of glycolysis to enhance its product yield.
subunit. The loss of MSTN weakens the formation of the Smad2/3 complex, thereby enhancing the expression of AMPK. After adding Smad3 inhibitor and AMPK activator to WT bovine muscle satellite cells, G6PD, tyrosine phosphate levels, and GSH content all showed an upward trend compared with untreated WT cells. These results suggest that reducing the expression of MSTN can regulate the tyrosine phosphorylation level of G6PD through the Smad-AMPK pathway to generate abundant NADPH to ensure high levels of GSH in the muscle tissue of MT cattle, thereby reducing ROS levels.

Conclusions
Inhibition of MSTN reduces the inhibitory effect of Smad on AMPK signaling, upregulates AMPK signaling, which affects G6PD activation, enhances PPP, produces more NADPH, increases the capacity of GR, and forms more reductive glutathione to remove ROS. Therefore, MSTN increases the antioxidant capacity of muscles.

Data Availability
The data generated and analyzed in this study are available from the corresponding author upon reasonable request.   Table S1: q-PCR primer sequences.