Sox6 Differentially Regulates Inherited Myogenic Abilities and Muscle Fiber Types of Satellite Cells Derived from Fast- and Slow-Type Muscles

Adult skeletal muscle is primarily divided into fast and slow-type muscles, which have distinct capacities for regeneration, metabolism and contractibility. Satellite cells plays an important role in adult skeletal muscle. However, the underlying mechanisms of satellite cell myogenesis are poorly understood. We previously found that Sox6 was highly expressed in adult fast-type muscle. Therefore, we aimed to validate the satellite cell myogenesis from different muscle fiber types and investigate the regulation of Sox6 on satellite cell myogenesis. First, we isolated satellite cells from fast- and slow-type muscles individually. We found that satellite cells derived from different muscle fiber types generated myotubes similar to their origin types. Further, we observed that cells derived from fast muscles had a higher efficiency to proliferate but lower potential to self-renew compared to the cells derived from slow muscles. Then we demonstrated that Sox6 facilitated the development of satellite cells-derived myotubes toward their inherent muscle fiber types. We revealed that higher expression of Nfix during the differentiation of fast-type muscle-derived myogenic cells inhibited the transcription of slow-type isoforms (MyH7B, Tnnc1) by binding to Sox6. On the other hand, Sox6 activated Mef2C to promote the slow fiber formation in slow-type muscle-derived myogenic cells with Nfix low expression, showing a different effect of Sox6 on the regulation of satellite cell development. Our findings demonstrated that satellite cells, the myogenic progenitor cells, tend to develop towards the fiber type similar to where they originated. The expression of Sox6 and Nfix partially explain the developmental differences of myogenic cells derived from fast- and slow-type muscles.


Introduction
Most mammalian skeletal muscle fibers are classified into four myosin heavy chain (MyHC) isoforms (slow type I, oxidative type IIA, fast type IIB and intermediate type IID/X) and are divided into two muscle fiber types (the slow type, which is red, oxidative and resistant to fatigue, and the fast type, which is white, glycolytic and fatigable [1]), based on their proportion of the isoforms, advantage of contraction speed or endurance and dominant metabolic way [1,2]. Adult skeletal muscle has a remarkable capability for regeneration following muscle damage mainly due to the function of satellite cells, one of the myogenic progenitors localized between the muscle fiber membrane and the basal 2 of 20 lamina [3]. Muscle satellite cells, are myogenic stem cells, which play an essential role in muscle regeneration and hypertrophy by differentiating into myofibers and contributing to myonuclear accretion [3,4]. Number maintenance and functional exertion of satellite cells are essential for muscle regeneration ability and body homeostasis [5].
Many studies have reported the contribution of satellite cells to skeletal muscle maturation, regeneration, health, disease, aging and exercise adaptation in various species [6]. Moreover, satellite cells were also found to be required for hypertrophic growth in young adults [7]. Previous research reported that Pax7(+) cells required for fetal myogenesis [8] and adult myogenesis (muscle growth, fiber maturation and regeneration) are mediated by adult myogenic progenitors [9,10]. Further, abnormal small fibers and markedly decreased muscle mass with few myo-nuclei were reported in skeletal muscle of mice surviving the deletion of Pax7 [11]. These studies highlighted the crucial role of satellite cells in adult myogenesis. A recent discovery showed that satellite cells possess a high population heterogeneity in homeostasis, aging and disease [6]. Moreover, satellite cells derived from fast-or slow-type muscles are functionally heterogeneous cells with different cell proliferation and differentiation potentials [12,13]. It has been reported that the fate of satellite cells during muscle regeneration is governed by a complex network of intrinsic and extrinsic regulators [6,14]. However, the underlying molecular mechanisms of these intrinsic regulations are not well understood.
In poultry, the pectoral muscle (PM) is generally considered to be a fast-type muscle and the leg muscle (LM) to be a slow-type muscle [15,16]. Immunofluorescence staining and western blot with specific myosin heavy chain isoforms antibodies indicate that the chicken PM is mainly composed of fast-type fibers and LM is mainly composed of slow-type fibers [17,18]. Our previous work found that SRY-box 6 (Sox6) transcription factor was highly expressed in the adult chicken fast-type muscle PM and had a regulatory role in muscle-derived myoblasts [19]. Sox6 plays a critical role in the regulation of skeletal muscle fiber types [20]. However, little is known about the function of Sox6 in satellite cell development and whether Sox6 contributes to the postnatal muscle fiber type regulation by regulating satellite cell development.
In this study, we isolated satellite cells from fast-type PM and slow-type LM separately. We observed that myoblasts from PM-derived satellite cells (PM-MBs) and myoblasts from LM-derived satellite cells (LM-MBs) tended to develop towards their origins. Concerning the role of Sox6 in satellite cells development, Sox6 regulated the myogenesis of the satellite cells, directly promoting the expression of fast-type fiber isoforms (MyH1A, Tnnc2, Tnni2 and Tnnt3). In particular, during myogenic differentiation, the differential expressions of Nfix between myoblasts derived from fast-type and slow-type muscles led to the different roles of Sox6 in slow-type muscles-derived myoblasts. These observations partially explained the myogenic functional differences between the satellite cells derived from fast-type and slow-type muscles, indicating the essential roles of Sox6 in the regulation of myogenic cells.

Different Myogenic Potentials in Satellite Cells of Distinct Muscle Fiber Types
We isolated satellite cells from PM and LM to validate the myogenic differences between satellite cells derived from slow-and fast-type muscles ( Figure 1A). Edu staining demonstrated that there was a higher number of EdU(+) cells in PM-MBs compared to LM-MBs ( Figure 1B), indicating that the PM-MBs proliferated faster than the LM-MBs. Quiescent satellite cells continuously express Pax7 and down-regulate Pax7 when activated [21]. Here, the expression levels of Pax7 gradually declined in differentiation medium (DM), indicating that the satellite cells exited the quiescent state and differentiated into myotubes ( Figure 1C). According to the qPCR results, with cell differentiation, the expression of MyH1A (myosin, heavy chain 1A) was significantly higher than MyH7B (myosin heavy chain 7B) in myotubes formed from PM-MBs (PM-MTs) ( Figure 1C). In contrast, the myotubes formed from LM-MB (LM-MTs) highly expressed MyH7B ( Figure 1C). Double immunofluorescence staining showed that slow MyHC (MyH7B)(+) fibers were in a higher proportion than fast MyHC (MyH1A)(+) fibers in LM-MTs ( Figure 1D). In addition, the ratio of the fast MyHC(+) fibers to slow MyHC(+) fibers in PM-MTs was higher than that in LM-MTs ( Figure 1D). Furthermore, qPCR results indicated that PM-MTs highly expressed fast-type isoforms Tnni2 (troponin I2,) and Tnnt3 (troponin T3), while the expression of slow-type isoforms sMyHC1 (slow myosin heavy chain 1), Tnni1 (troponin I1) and Tnnt1 (troponin T1) were low ( Figure 1E). As the differentiation proceeded, the expression of slow-type isoform Tnnc1 (troponin C1) increased rapidly and exceeded the Tnni2 and Tnnt3 in LM-MTs ( Figure 1E). Therefore, the satellite cells intrinsically tended to differentiate into specific fiber types of the muscle from which they derived. medium (DM), indicating that the satellite cells exited the quiescent state and differentiated into myotubes ( Figure 1C). According to the qPCR results, with cell differentiation, the expression of MyH1A (myosin, heavy chain 1A) was significantly higher than MyH7B (myosin heavy chain 7B) in myotubes formed from PM-MBs (PM-MTs) ( Figure 1C). In contrast, the myotubes formed from LM-MB (LM-MTs) highly expressed MyH7B ( Figure  1C). Double immunofluorescence staining showed that slow MyHC (MyH7B)(+) fibers were in a higher proportion than fast MyHC (MyH1A)(+) fibers in LM-MTs ( Figure 1D). In addition, the ratio of the fast MyHC(+) fibers to slow MyHC(+) fibers in PM-MTs was higher than that in LM-MTs ( Figure 1D). Furthermore, qPCR results indicated that PM-MTs highly expressed fast-type isoforms Tnni2 (troponin I2,) and Tnnt3 (troponin T3), while the expression of slow-type isoforms sMyHC1 (slow myosin heavy chain 1), Tnni1 (troponin I1) and Tnnt1 (troponin T1) were low ( Figure 1E). As the differentiation proceeded, the expression of slow-type isoform Tnnc1 (troponin C1) increased rapidly and exceeded the Tnni2 and Tnnt3 in LM-MTs ( Figure 1E). Therefore, the satellite cells intrinsically tended to differentiate into specific fiber types of the muscle from which they derived.  When quiescent satellite cells were activated, the cells down-regulate Pax7, then the satellite cell-derived myogenic cells (myoblasts) enter the cell cycle and undergo limited proliferation, most of them differentiating into myotubes [22]. The rest exit the cell cycle, highly re-express Pax7, remain in the G0-phase and no longer proliferate and differentiate, entering back to the quiescent state to maintain the satellite cell pool and completing the cell self-renewal [22]. We cultured PM-MBs and LM-MBs in the DM for 5 days to investigate the retention number and cell cycle status of satellite cells in the later stage of cell differentiation. We detected Pax7(+)EdU(+) cells (proliferating satellite cells), Pax7(−)EdU(+) cells (cells that had been activated to proliferate and differentiate) and Pax7(+)EdU(−) cells (satellite cells entered back to the quiescent state or maintained in G0-phase) in PM-MTs and LM-MTs. We found that the number of Pax7(+)EdU(−) cells was higher in LM-MTs than in PM-MTs ( Figure 1F), implying that more satellite cells stay or return to the quiescent stage in LM-MTs and the satellite cells derived from slow-type LM possessed a higher self-renewal potential.

Sox6 Promotes the Myogenesis and Development of Satellite Cells Derived from Fast Type-Enriched Pectoral Muscles towards Intrinsic Tendency
Our previous study found that Sox6 was highly expressed in adult chicken PM; here, we found that the expression of Sox6 in PM-MTs gradually increased during the differentiation. However, the Sox6 expression in LM-MTs peaked at DM day 3 and then slightly declined ( Figure 1C). Then we overexpressed or suppressed Sox6 in PM-MBs to assess its effect (Figures 2A and S1A). Proliferation assay and cell cycle analysis showed that Sox6 overexpression significantly promoted the PM-MB proliferation ( Figure  levels of Sox6 in Sox6-overexpressing PM-MBs were quantified by qPCR (mean ± SEM; ** p < 0.01; n = 3; two-tailed Student's t-test). (B) Proliferation of Sox6-overexpressing PM-MBs were assessed by Edu (mean ± SEM; ** p < 0.01; n = 3; two-tailed Student's t-test). (C) Cell cycle analysis of Sox6-overexpressing PM-MBs (mean ± SEM; * p < 0.05; ** p < 0.01; n = 3; two-tailed Student's t-test). (D) After inducing differentiation for 3 days, the relative expression levels of MyH1A and MyH7B in Sox6-overexpressing PM-MTs were quantified by qPCR (mean ± SEM; * p < 0.05; n = 3; two-tailed Student's t-test). (E) After inducing differentiation for 3 days, the relative expression levels of fast-and slow-type fiber isoforms in Sox6-overexpressing PM-MTs were quantified by qPCR (mean ± SEM; * p < 0.05; ** p < 0.01; n = 3; two-tailed Student's t-test). (F) Western blot on lysates from Sox6-overexpressing PM-MTs and control groups myotubes after inducing differentiation for 3 days. β-actin was used to normalize. (G) After inducing differentiation for 3 days, myotubes derived from Sox6-overexpressing PM-MTs were stained against S58 (slow-type MyHC, green) and F59 (fast-type MyHC, red) and nuclei were counterstained with DAPI. The slow MyHC(+) nuclei or fast MyHC(+) nuclei were counted. The proportion of slow MyHC(+) cells − fast MyHC(+) cells among total nuclei and the proportion of cells in slow MyHC(+) myotubes or fast MyHC(+) myotubes among total nuclei are presented as mean ± SEM (* p < 0.05; n = 3; two-tailed Student's t-test). (H) After inducing differentiation for 3 days, the relative expression levels of Myf5, Pax7 and MyoD in Sox6-overexpressing PM-MBs were quantified by qPCR (mean ± SEM; * p < 0.05; ** p < 0.01; n = 3; two-tailed Student's t-test). (I) Sox6-overexpressing PM-MBs were cultured in DM for 5 days and EdU was added to the culture medium 24 h prior to harvest. In the end, the cells were stained against Pax7 (green) and EdU (red) and the numbers were counted (mean ± SEM; ** p < 0.01; n = 3; two-tailed Student's t-test). White arrows represented part of the typical Pax7(+)EdU(−) cells and red arrows represented part of the typical Pax7(+)EdU(+) cells.
After inducing the differentiation, Sox6 overexpression and inhibition was performed in PM-MTs. qPCR results revealed that Sox6 overexpression significantly increased the expression of MyH1A and other fast-type isoforms (Tnnc2, Tnni2, Tnnt3), while attenuated MyH7B and other slow-type isoforms' (sMyHC1, Tnnc1) expression in PM-MTs ( Figure 2D,E). In contrast, inhibition of Sox6 in PM-MTs showed reduced expression of MyH1A, Tnnc2, Tnni2 and Tnnt3, whereas the expression of MyH7B, sMyHC1, Tnnc1 and MYBPC1 were increased ( Figure S1D,E). Similar to the qPCR results, western blotting demonstrated increased MyH1A and decreased MyH7B protein levels in Sox6-overexpressing PM-MTs compared to control ( Figure 2F). Inhibiting Sox6 in PM-MTs decreased MyH1A and increased MyH7B protein levels ( Figure S1F). To obtain a further validation, we used double immunofluorescence staining on the myotubes. Results showed that Sox6 overexpression significantly reduced the disparity in the proportion between slow (MyH7B) and fast (MyH1A) MyHC(+) myotubes in PM-MTs ( Figure 2G), supporting the above finding that Sox6 increased the fast-type fibers and decreased the slow-type fibers in PM-MTs. In contrast, suppression of Sox6 in PM-MTs significantly increased the disparity in the proportion between slow and fast MyHC(+) myotubes ( Figure S1G).
In addition, the myogenic differentiation (all types of MyHC(+) per total nuclei) was significantly improved in Sox6-overexpressing PM-MTs, while inhibition of Sox6 reduced the myogenic differentiation ( Figures 2G and S1G). Besides, qPCR results also demonstrated significantly increased expression of MyoD (myogenic differentiation 1), Myf5 (myogenic factor 5) and down-regulation of Pax7, which regulate the myogenic cell differentiation [23], in Sox6-overexpressing PM-MTs ( Figure 2H). The opposite result was revealed in Sox6-inhibited PM-MTs ( Figure S1H). Further, we investigated the retention number and cell cycle status of satellite cells after 5 days of cell differentiation. We found that, the Sox6-overexpressing PM-MTs contained a significantly lower number of Pax7(+)EdU(−) cells and Sox6-inhibited PM-MTs contained an increased number of Pax7(+)EdU(−) cells ( Figures 2I and S1I). The reduction in the number of Pax7(+)EdU(−) cells in Sox6-overexpressing PM-MTs reflected a diminished cell self-renewal potential.
(F) Western blot on lysates from Sox6-overexpressing LM-MTs and control groups myotubes after inducing differentiation for 3 days. β-actin was used to normalize. (G) After inducing differentiation for 3 days, myotubes derived from Sox6-overexpressing LM-MTs were stained against S58 (slow-type MyHC, green) and F59 (fast-type MyHC, red) and nuclei were counterstained with DAPI. The slow MyHC(+) nuclei or fast MyHC(+) nuclei were counted. The proportion of slow MyHC(+) cells − fast MyHC(+) cells among total nuclei and the proportion of cells in slow MyHC(+) myotubes or fast MyHC(+) myotubes among total nuclei are presented as mean ± SEM (** p < 0.01; n = 3; two-tailed Student's t-test). (H) After inducing differentiation for 3 days, the relative expression levels of Myf5, Pax7 and MyoD in Sox6-overexpressing LM-MBs were quantified by qPCR (mean ± SEM; ** p < 0.01; n = 3; two-tailed Student's t-test). (I) Sox6-overexpressing LM-MBs were cultured in DM for 5 days and EdU was added to the culture medium 24 h prior to harvest. In the end, the cells were stained against Pax7 (green) and EdU (red) and the numbers were counted (mean ± SEM; n = 3; two-tailed Student's t-test). White arrows represented part of the typical Pax7(+)EdU(−) cells and red arrows represented part of the typical Pax7(+)EdU(+) cells.

Differential Expression of Nfix between LM-MBs and PM-MBs at the Differentiation Stage Leads to Different Roles of Sox6 in the Regulation of Slow-Type Fibers
In order to identify the possible mechanism by which Sox6 differentially regulates the transcription of slow-type isoforms, we focused on the transcription factor Nfix. Nfix, a known regulator for the slow twitching fibers [24], had been found to influence the role of Sox6 [25]. First, we examined Nfix expression in differentiating PM-MTs and LM-MTs. qPCR results showed that the expression of Nfix significantly upregulated with the differentiation in PM-MTs ( Figure 4A), while the differences of Nfix expression in LM-MTs was not significant with myogenic differentiation. Further validation on protein levels demonstrated that the Nfix expression upregulated with the PM-MTs differentiation ( Figure 4B). While Nfix also expressed in LM-MTs and upregulated with myogenic differentiation, the expression level was relatively low and the increase was limited ( Figure 4B). Further, confocal microscopy results showed that, the PM-MTs in differentiating contained more Nfix(+) cells than LM-MTs ( Figure 4C).
To further elucidate the mechanisms between Nfix and Sox6 in the regulation of slowtype fibers, we performed dual-luciferase assays on DF-1 cells, non-skeletal muscle-derived cells, transfected with reporter vectors containing different lengths of 5 -upstream region of slow-type isoform genes ( Figure 4D). There was no luciferase activity difference of MyH7B and Tnnc1 in Sox6-overexpressing DF-1 compared to the control ( Figure 4E). However, with the co-overexpression of Nfix and Sox6, the luciferase activities of MyH7B and Tnnc1 were significantly down-regulated ( Figure 4E). To further validate whether the suppressive effect of the slow-type isoforms comes from Nfix, we investigated the luciferase activities of MyH7B and Tnnc1 under the overexpression of Nfix in DF-1. However, no significant change was detected ( Figure 4F). Our results indicated that neither Sox6 nor Nfix alone could directly affect the transcriptions of MyH7B and Tnnc1, but Sox6 could suppress the transcriptions of slow-type isoforms by acting on their proximal promoters in a Nfixdependent way ( Figure 4D,E). Further, we performed a co-immunoprecipitation (Co-IP) assay for Sox6 in PM-MTs and LM-MTs after overexpressing Sox6. Co-IP results revealed that Sox6 could bind to Nfix in PM-MTs and LM-MTs ( Figure 4G). Three-dimensional stereogram through confocal microscopy also showed a co-localization pattern between the proteins of Sox6 and Nfix, while myo-nuclei were fusing at this location ( Figures 4H and S3). (C) After inducing differentiation for 5 days, myotubes derived from PM and LM were performing double immunofluorescence staining against Nfix (green) and MyHC (red) and nuclei were counterstained with DAPI (mean ± SEM; ** p < 0.01; n = 3; two-tailed Student's t-test). (D) Dual-Luciferase report assay on DF-1 transfected with reporter vectors containing different lengths of 5′-upstream region of MyH7B and Tnnc1 (mean ± SEM; *** p < 0.001; n = 6; one-way analysis of variance with Tukey's multiple comparison test). (E) Dual-Luciferase report assay on DF-1 transfected with reporter vectors containing different lengths of 5′-upstream region of MyH7B and Tnnc1 after overexpressing Sox6 or Sox6 and Nfix (mean ± SEM; ** p < 0.01; n = 6; one-way analysis of variance with Tukey's multiple comparison test). (F) Dual-Luciferase report assay on DF-1 transfected with reporter vectors containing different lengths of 5′-upstream region of MyH7B and Tnnc1 after overexpressing Nfix (mean ± SEM; n = 6; two-tailed Student's t-test). (G) Lysates of PM-MTs and LM-MTs overexpressing Sox6 were immunoprecipitated with Sox6 antibody and then western blotted with Nfix antibody. IgG, negative control; IP, immunoprecipitated; IB, immunoblotting. (H) Myotubes derived from PM-MTs were performing double immunofluorescence staining against Sox6 (red) and Nfix (green). The three-dimensional view of Sox6 and Nfix localization in PM-MTs was presented through confocal microscopy.

Sox6 Indirectlyupregulates the Slow-Type Isoforms through the Activation of Mef2C
Since Sox6 alone did not directly bind to the promoters of MyH7B and Tnnc1 ( Figure 4E), implying the promoted effect of Sox6 on slow-type isoforms in LM-MTs might be exerted through other factors. We focused on a transcription regulator Mef2C, which has been found to up-regulate the slow-type fibers [26,27]. According to previous research, the regulation of Mef2C through Sox6 could be influenced by Nfix [25]. Therefore, we investigated the relationship between Sox6, Nfix and Mef2C in satellite cells. After inducing the myoblasts differentiation, we overexpressed and suppressed Sox6 in PM-MTs and LM-MTs separately. We found that Mef2C expression was up-regulated in Sox6-overexpressing PM-MTs, but the effect was not significant ( Figure 5A). Inhibition of Sox6 in PM-MTs did not influence the expression of Mef2C ( Figure 5A). However, the expression of Mef2C was significantly increased in Sox6-overexpressing LM-MTs and suppression of Sox6 in LM-MTs significantly reduced the expression of Mef2C ( Figure 5B).

Sox6 Indirectlyupregulates the Slow-Type Isoforms through the Activation of Mef2C
Since Sox6 alone did not directly bind to the promoters of MyH7B and Tnnc1 ( Figure  4E), implying the promoted effect of Sox6 on slow-type isoforms in LM-MTs might be exerted through other factors. We focused on a transcription regulator Mef2C, which has been found to up-regulate the slow-type fibers [26,27]. According to previous research, the regulation of Mef2C through Sox6 could be influenced by Nfix [25]. Therefore, we investigated the relationship between Sox6, Nfix and Mef2C in satellite cells. After inducing the myoblasts differentiation, we overexpressed and suppressed Sox6 in PM-MTs and LM-MTs separately. We found that Mef2C expression was up-regulated in Sox6-overexpressing PM-MTs, but the effect was not significant ( Figure 5A). Inhibition of Sox6 in PM-MTs did not influence the expression of Mef2C ( Figure 5A). However, the expression of Mef2C was significantly increased in Sox6-overexpressing LM-MTs and suppression of Sox6 in LM-MTs significantly reduced the expression of Mef2C ( Figure 5B). Further investigation using dual-luciferase assays on DF-1 cells transfected with reporter vectors containing different lengths of 5 -upstream region of Mef2C ( Figure 5C). Dual-luciferase assays showed that significantly increased luciferase activity in all Mef2C reporter vectors in Sox6-overexpressed cells ( Figure 5D). However, Sox6 overexpression did not change the luciferase activities of all reporter vectors under the co-overexpression with Nfix ( Figure 5D). These results indicated that Sox6 directly promoted the transcription of Mef2C by acting on the proximal promoter region ( Figure 5C,D), but Nfix could interrupt the promotion. Further, we observed significant increases of luciferase activities in all slowtype isoform gene reporter vectors in DF-1 overexpressing Mef2C (Figure 5E), indicating that Mef2C could directly act on the proximal promoters of MyH7B and Tnnc1 to promote the transcriptions. Given the differential Nfix expression between PM-MTs and LM-MTs, these results indicated that high expression of Nfix in PM-MTs bind to Sox6 to hinder its promotion of Mef2C, while the expression of Nfix in LM-MTs is insufficient to exert the same effect.

Sox6 Decreases Slow-Type Isoforms through Binding with Nfix in PM-MTs or Increases Slow-Type Isoforms by Activating Mef2C in LM-MTs
To get on a further validation, we simultaneously overexpressed Sox6 and suppressed Nfix in PM-MTs ( Figure 6A). Expectedly, qPCR result showed a significant increase of Mef2C ( Figure 6A), which was consistent with the effect of Sox6 overexpression on LM-MTs ( Figure 5B). In addition, we overexpressed Mef2C in another group of PM-MTs to validate whether its effect on slow-type isoforms was the same as the overexpression of Sox6 in Nfix-inhibited PM-MTs ( Figure S4A). Using qPCR, we found significantly increased expression of MyH1A, MyH7B, sMyHC1, Tnnc1, Tnnc2, Tnni2, Tnnt3 and MyBPC1 in Sox6-overexpressing PM-MTs with Nfix inhibition (Figure 6B,C). Western blotting results showed increased protein levels of MyH1A and MyH7B in Sox6-overexpressing PM-MTs with Nfix inhibition ( Figure 6D). On the other hand, Mef2C overexpression in PM-MTs also significantly elevated the expression of MyH7B, sMyHC1, Tnnc1 and MyBPC1 ( Figure S4B,C). Western blotting results showed increased protein levels of MyH7B in Mef2C-overexpressing PM-MTs ( Figure S4D). Double immunofluorescence staining showed no significant increase of the disparity in the proportion between slow and fast MyHC(+) myotubes in Sox6-overexpressing PM-MTs with Nfix inhibition (Figure 6E), which was consistent with the simultaneous increases in fast and slow type fibers ( Figure 6B-D). Mef2C overexpression significantly increased the disparity between slow and fast MyHC(+) myotubes and promoted the myogenic differentiation in PM-MTs ( Figure S4E). These results indicated that Sox6 upregulated the slow-type fibers through the promotion of Mef2C, but it was interrupted by Nfix in PM-MTs, while the fast-type fibers were improved by Sox6.
Next, we compared the effect of Nfix and Sox6 co-overexpression with the effect of Mef2C inhibition on LM-MTs ( Figures 7A and S5A). After inducing the differentiation, Nfix, Sox6 co-overexpression and Mef2C inhibition was performed in LM-MTs separately. qPCR results showed that Sox6 and Nfix co-overexpression in LM-MTs significantly increased the expression of MyH1A, Tnnc2, Tnni2 and Tnnt3, while decreasing the expression of MyH7B, sMyHC1 and Tnnc1 ( Figure 7B,C). Western blotting results demonstrated increased MyH1A and decreased MyH7B protein levels in LM-MTs with Sox6 and Nfix co-overexpression ( Figure 7D). Double immunofluorescence staining showed significant decrease of the disparity in the proportion between slow and fast MyHC(+) myotubes in LM-MTs with Sox6 and Nfix co-overexpression ( Figure 7E). These results were reminiscent of the results observed in Sox6-overexpressing PM-MTs ( Figure 2D-G). Inhibition of Mef2C in LM-MTs significantly decreased the expression of MyH7B, sMyHC1, Tnnc1 and MYBPC1 (Figure S5B,C). Western blotting results showed decreased MyH7B protein levels in LM-MTs with inhibition of Mef2C ( Figure S5D). And inhibition of Mef2C in LM-MTs significantly decrease of the disparity in the proportion between slow and fast MyHC(+) myotubes ( Figure S5E). These results consolidated that the Sox6 promoting effect on slow-type isoforms in LM-MTs was due to its low expression of Nfix, which was insufficient to hinder the promotion of Sox6 on Mef2C. Next, we compared the effect of Nfix and Sox6 co-overexpression with the effect of Mef2C inhibition on LM-MTs ( Figures 7A and S5A). After inducing the differentiation, Figure 6. Overexpression of Sox6 in Nfix-inhibited PM-MTs promotes the expression of Mef2C and slow-type fibers. (A) After inducing differentiation for 3 days, the relative expression levels of Sox6, Nfix and Mef2C in Sox6-overexpressing PM-MTs with Nfix inhibition were quantified by qPCR (mean ± SEM; ** p < 0.01; n = 3; two-tailed Student's t-test). (B) After inducing differentiation for 3 days, the relative expression levels of MyH1A and MyH7B in Sox6-overexpressing PM-MTs with Nfix inhibition were quantified by qPCR (mean ± SEM; * p < 0.05; ** p < 0.01; n = 3; two-tailed Student's t-test). (C) After inducing differentiation for 3 days, the relative expression levels of fastand slow-type fiber isoforms in Sox6-overexpressing PM-MTs with Nfix inhibition were quantified by qPCR (mean ± SEM; * p < 0.05; n = 3; two-tailed Student's t-test). (D) Western blot on lysates from Sox6-overexpressing PM-MTs with Nfix inhibition and control group myotubes after inducing differentiation for 3 days. β-actin was used to normalize. (E) After inducing differentiation for 3 days, myotubes derived from Sox6-overexpressing PM-MTs with Nfix inhibition were performing immunofluorescence double staining against S58 (slow-type MyHC, green) and F59 (fast-type MyHC, red) and nuclei were counterstained with DAPI. The slow MyHC(+) nuclei or fast MyHC(+) nuclei were counted. The proportion of slow MyHC(+) cells − fast MyHC(+) cells among total nuclei and the proportion of cells in slow MyHC(+) myotubes or fast MyHC(+) myotubes among total nuclei are presented as mean ± SEM (** p < 0.01; n = 3; two-tailed Student's t-test). MYBPC1 ( Figure S5B,C). Western blotting results showed decreased MyH7B protein levels in LM-MTs with inhibition of Mef2C ( Figure S5D). And inhibition of Mef2C in LM-MTs significantly decrease of the disparity in the proportion between slow and fast MyHC(+) myotubes ( Figure S5E). These results consolidated that the Sox6 promoting effect on slowtype isoforms in LM-MTs was due to its low expression of Nfix, which was insufficient to hinder the promotion of Sox6 on Mef2C.  In addition, we noticed that Sox6 promoted the expression of fast-type isoforms in PM-MTs and LM-MTs with or without the presence of Nfix. Therefore, we performed dual-luciferase assays on DF-1 transfected with reporter vectors containing different lengths of 5 -upstream region of fast-type isoform genes ( Figure 8A). Dual-luciferase assays showed increased luciferase expression in the overexpression of Sox6 with the vector carrying the −2317 bp 5 upstream sequence of the MyH1A ( Figure 8B). However, the increased luciferase expression by Sox6 was lost without the −2317 to −1823 bp 5 upstream sequence of MyH1A on the vector (Figure 8B), indicating that Sox6 directly acted on this site to enhance the MyH1A expression in a Nfix-independent way. In addition, Sox6 alone could directly act on the proximal promoters to promote the expression of Tnnc2, Tnni2 and Tnnt3 ( Figure 8A,B). These results indicated that Sox6 independently promoted the fast-type fiber formation by directly acting on regulatory elements or proximal promoters of the fast-type isoforms.
In addition, we noticed that Sox6 promoted the expression of fast-type isoforms in PM-MTs and LM-MTs with or without the presence of Nfix. Therefore, we performed dual-luciferase assays on DF-1 transfected with reporter vectors containing different lengths of 5′-upstream region of fast-type isoform genes ( Figure 8A). Dual-luciferase assays showed increased luciferase expression in the overexpression of Sox6 with the vector carrying the −2317 bp 5′ upstream sequence of the MyH1A ( Figure 8B). However, the increased luciferase expression by Sox6 was lost without the −2317 to −1823 bp 5′ upstream sequence of MyH1A on the vector (Figure 8B), indicating that Sox6 directly acted on this site to enhance the MyH1A expression in a Nfix-independent way. In addition, Sox6 alone could directly act on the proximal promoters to promote the expression of Tnnc2, Tnni2 and Tnnt3 (Figure 8A,B). These results indicated that Sox6 independently promoted the fast-type fiber formation by directly acting on regulatory elements or proximal promoters of the fast-type isoforms.

Discussion
It was generally considered that the fiber types of muscles are primarily influenced by innervation or age and external stimulus [28,29]. However, a previous study reported and Nfix and transfected with reporter vectors containing different lengths of 5 -upstream region of MyH1A, Tnnc2, Tnni2 and Tnnt3 (mean ± SEM; ** p < 0.01; n = 6; one-way analysis of variance with Tukey's multiple comparison test).

Discussion
It was generally considered that the fiber types of muscles are primarily influenced by innervation or age and external stimulus [28,29]. However, a previous study reported that myoblasts isolated from soleus (SOL) and tibialis anterior (TA) muscles of mice retained their contractile and metabolic properties in vitro [30]. Although, compared to model species, avian body structure and genetic mechanisms are not exactly the same. However, many studies have also proved that it is a good way to study the molecular genetic regulation mechanisms using avian as model [31,32]. In the present study, we isolated the satellite cells from the fast-type muscle and slow-type muscle and found that the PMderived cells have a faster proliferation rate. A previous review reported that multiple fiber types were generally intermingled within a single skeletal muscle, while different skeletal muscles possessed distinct proportions of fiber types instead of a single fiber type [33]. Here we found that the relative ratio of slow-type fibers was high in LM-MTs and the relative proportion of fast-type fibers in PM-MTs was highly elevated compared to LM-MTs. Thus, our study once again demonstrated that the development of muscle fiber types is regulated by the endogenous genetic mechanisms. In addition, the presence of a considerable proportion of slow-type fibers in PM-MTs might be similar to the fact that, most of the fast-type muscles in the newborn mice contain significant amounts of slow-type I fibers and the majority of them would disappear during the growth and development [34].
The self-renewal ability of satellite cells determines the capacity to restore satellite cell number [35], which is essential for the muscle regeneration and development. Rare surviving mice with a few number of satellite cells exhibited reduced body growth and marked muscle atrophy [36]. The present study implied that satellite cells derived from LM possessed a higher potential for self-renewal compared to PM. As previously reported, compared with fast-type muscle fibers, slow-type muscle fibers undergo a preferential satellite cell expansion and myonuclear addition during hypertrophy, indicating the potential differences in satellite cell requirement between different muscle fiber types [37].
Disruption of Sox6 exons in humans with delayed speech development and attention deficit hyperactivity disorder is associated with generalized dystonic and pectus carinatum [38]. Sox6 gene disruption in mouse resulted in growth retardation, myopathy [39]. It was shown that the Sox6 mutant influenced the MyHC isoform expression patterns in myotubes formed from myoblasts in the absence of innervation [40]. Here we showed that Sox6 could promote the PM-MB proliferation, increase the fast-type fiber formation and decrease the slow-type fiber formation in PM-MTs. A previous study demonstrated that Sox6-mutant fetal mice maintained slow fiber characteristics in all muscle fibers [40], indicating that Sox6 is required for normal fast-type fiber differentiation in developing fetal skeletal muscle. Here we showed that Sox6 could also contribute to the growth and property maintenance of fast-type muscles in adult skeletal muscle via the regulation of satellite cells. However, unlike in PM-MTs, Sox6 led to an up-regulation of slow-type fibers in LM-MTs. Therefore, we further investigated Nfix, a known regulator of the Sox6 function [25]. Recent studies reported that Nfix is related to skeletal muscle regeneration and satellite cell development [24,41]. Satellite cells derived from different muscles exhibited distinct gene expression profiles [12]. Sox6 gene structure dictates its functional versatility and critical requirement for co-factors to exert its transcriptional regulation [42]. Our present study showed that Nfix in PM-MTs and LM-MTs possessed distinct expression profiles during differentiation, bound to Sox6 to influence its effect on satellite cells-derived slow-type fibers. There was a higher expression of slow MyHC in myotubes derived from muscle satellite cells in Nfix-null mouse in DM [24], which supported our finding that Nfix-low expressed LM-MTs intrinsically generated slow muscles.
Furthermore, Nfix-null mice display delayed regeneration and knocking out Nfix in satellite cells causes muscle regeneration to be delayed [24,43]. Our results implied a higher self-renewal potential in Nfix-low expressed LM-MTs, indicating more satellite cells stay or return to the quiescent stage and less satellite cells participate in muscle repair. These might partially explain the delayed muscle regeneration by Nfix-null satellite cells. Moreover, NFATc4, which would be downregulated by Nfix [44], had been found to improve the pool of reserve cells [45]. Therefore, we argued that a complex regulatory network governs the inherent characteristics in satellite cells.
Mef2C had been documented to play a vital role in slow-twitch myofibers [26,46]. Previous studies reported that multiple factors regulated Mef2C to control muscle cells [47,48]. Here we showed that Sox6 activated the transcription of Mef2C by directly acting on its proximal promoter, but the presence of Nfix inhibited this activation. Moreover, we found that Mef2C was able to bind to the slow-type isoforms promoter to promote its transcriptions. These findings implied that the indirect promotion of slow fibers by Sox6 was due to the expression of Nfix in LM-MTs, which was insufficient to interfere with the activation of Mef2C by Sox6. Previous research demonstrated that Mef2C could rescue muscle atrophy [49]. Combined deletion of MEF2A, C and D in mouse hind limb muscles-derived satellite cells prevented the muscle regeneration [50], indicating a positive effect of Mef2C on cell self-renewal ability. Hence, the insufficient Sox6 influence in cell self-renewal potential of LM-MTs might be due to the upregulation of Mef2C by Sox6. Previous studies have found that the expression of Sox6 and Nfix is closely related to the regulation of fast and slow muscles in skeletal muscle development during embryonic to fetal stages [25,44]. Our results provided a new understanding that Sox6 could also contribute to the muscle fiber types regulation after birth by interacting with Nfix in regulating the myogenesis development of satellite cells.
We also performed functional similarity experiments to validate the roles of Nfix and Mef2C in the slow fiber regulation of Sox6. We found that Sox6 overexpression in Nfix-inhibited PM-MTs promoted slow-type fiber formation, which was consistent with the effects of Mef2C overexpression in PM-MTs and Sox6 overexpression in LM-MTs. In contrast, Sox6 overexpression in Nfix-overexpressing LM-MTs downregulated the slow-type fiber formation, which was consistent with the effect of Mef2C inhibition in LM-MTs and Sox6 overexpression in PM-MTs. These results confirmed that the different effects of Sox6 on slow fibers were implemented partially through Mef2C, which are affected by the expression of Nfix in PM-MTs and LM-MTs. A recent study showed that Sirt6 expression was elevated in chronically exercised humans and Sirt6 could downregulate Sox6 by increasing the transcription of CREB to increase the slow-type fibers and exercise endurance, which indicated that Sirt6 activation may offer an exercise mimetic therapy [51]. Here, whether our research findings about Sox6 can provide a certain reference for human exercise mimetic therapy remains to be further studied.

Ethics Standards
All animal experimental protocols were carried out in accordance with "The Instructive Notions with Respect to Caring for Laboratory Animals" issued by the Ministry of Science and Technology of the People's Republic of China and approved by the South China Agricultural University Institutional Animal Care and Use Committee (2022F150).

Isolation of Muscle Satellite Cells and Cell Culture
All chickens were obtained from a single hatch and raised indoors. The chickens were fed a diet containing 2837 kcal ME (metabolic energy)/kg and 200 g CP (crude protein)/kg. Satellite cells were isolated from pectoral and leg muscles of 3-day-old chickens as previously described [52]. Briefly, skeletal muscles were isolated, cut and minced into smooth pulp. The minced tissues were digested with 0.1% collagenase I (Solarbio, Beijing, China) for 30 min at 37 • C. After centrifugation and removing the supernatant, the tissues were digested with 0.25% trypsin (Solarbio, Beijing, China) for 1 h at 37 • C. The digestion was terminated with complete medium containing DMEM/F12 (Invitrogen, Carlsbad, CA, USA), 20% fetal bovine serum (FBS, Gibco, Bethesda, MD, USA), 100 units/mL penicillin and 100 µg/mL streptomycin (final concentration, Gibco, Bethesda, MD, USA). The suspension was transferred onto a cell strainer (70 µm) until it passed through. The cell suspension was centrifuged at 1500 rpm for 8 min then the supernatant was discarded. The cells were resuspended with complete medium and cultured at 37 • C with 100% humidity under 5% CO 2 . After 2 h of culture, the satellite cells were in the cell suspension. Then the cell suspension was transferred to a new cell culture plate and cultured at 37 • C with 100% humidity under 5% CO 2 . The cells were cultured in differentiation medium (DM) containing DMEM/F12, 5% horse serum (Gibco, Bethesda, MD, USA), 100 units/mL penicillin and 100 µg/mL streptomycin to induce myoblast differentiation.

Edu and Flow Cytometry Assays
Cell-Light EdU Apollo In Vitro Kit (RiboBio, Guangzhou, China) was used for Edu assay according to the manufacturer's protocols. Cell nuclei were stained by DAPI. Flow cytometry assays were performed using FxCycle™ PI/RNase Staining Solution (Invitrogen, USA) was used according to the manufacturer's protocols and analyzed by BD Accuri C6 flow cytometer (BD Biosciences, San Jose, CA, USA).

RNA Extraction, cDNA Synthesis and Quantitative Real-Time PCR
Total RNA was extracted from cultured cells and myotubes using RNAiso plus reagent (Takara, Otsu, Japan) according to the traditional protocols. Complementary DNA (cDNA) synthesis for messenger RNA (mRNA) was carried out using the PrimerScript RT Reagent Kit with gDNA Eraser (Perfect Real Time) (Takara). Quantitative real-time PCR assays were conducted on QuantStudio 5 Real-Time PCR System (Applied Biosystems Inc., Foster, Waltham, MA, USA) using the iTaq Universal SYBR Green Supermix Kit (Bio-Rad Laboratories Inc., Hercules, CA, USA) in triplicate, as described previously [53]. The relative expression levels were calculated using the 2 − Ct relative quantitative method, as described previously [54]. Chicken β-actin was used as an internal control. Primers for qRT-qPCR are shown in Table S1.
Promoter reporter plasmid: different lengths of Mef2C, MyH1A, Tnnc2, Tnni2, Tnnt3, MyH7B and Tnnc1 promoter sequences were amplified from chicken leg muscle DNA by PCR. The PCR products were isolated from agarose gel and ligated into the pGL3-basic luciferase reporter vector (Promega, Madison, WI, USA). Primers for promoter sequence cloning are shown in Table S2.
Small interfering RNA (siRNA) against Sox6, Nfix and Mef2C were designed and synthesized by Ribobio (Table S3, Guangzhou, China), nonspecific duplex siNC was used as control and provided by Ribobio. Cell transfection was carried out using Lipofectamine 3000 reagent (Invitrogen) according to the manufacturer's protocols. Lipofectamine 3000 reagents and nucleic acids were diluted in OPTI-MEM with Reduced Serum Medium (Gibco).

Dual-Luciferase Reporter Assay
The chicken embryonic fibroblast cell line (DF-1) was maintained in medium containing Dulbecco's modified Eagle's medium (Invitrogen) with high glucose, 10% fetal bovine serum (Gibco), 100 units/mL penicillin and 100 µg/mL streptomycin (final concentration, Gibco) and incubated at 37 • C with 100% humidity under 5% CO 2 . For essential promoter validation, the DF-1 cells were transfected with a series of the promoter reporter plasmids described above and the TK-Renilla reporter (Promega) was co-transfected as internal control. For interaction assays, the promoter reporter plasmids were co-transfected with the gene overexpression vectors in DF-1 and the TK-Renilla reporter (Promega) was co-transfected as internal control. At 48 h after transfection, the luciferase activities and Renilla luciferase of the cells were measured using the Dual-Glo ® Luciferase Assay System (Promega) and Synergy™ Neo 2 Multi-Mode Microplate Reader (Biotek, Winooski, VT, USA) with Gen5 software (Biotek, Winooski, VT, USA). The levels of firefly luciferase activity were normalized to Renilla luciferase activity.

Co-Immunoprecipitation Assay
Co-immunoprecipitation assays were performed using Pierce Co-Immunoprecipitation (Co-IP) Kit (Thermo Scientific, MA, USA) in accordance with the manufacturer's protocols. Briefly, cultured cells were washed by 1× Modified Dulbecco's PBS and incubated with ice-cold IP Lysis/Wash Buffer on ice for 5 min with periodic mixing. Lysates were centrifuged at 13,000× g for 10 min to pelletize the cell debris. Antibody immobilizations were prepared by adding appropriate amount of AminoLink Plus Coupling Resin, Coupling Buffer, Sodium Cyanoborohydride Solution, ultrapure water and 75 µg of anti-Sox6 or IgG antibody (bs-0310G, Bioss, Beijing, China) to a tube and incubating on a mixer at room temperature for 120 min. For immunoprecipitation, protein lysates and antibodies-binding resin were mixed and incubated with gentle mixing at 4 • C overnight. After centrifugation and removing the flow-through, the mixture was washed with cold IP Lysis/Wash Buffer and the antibody-binding protein complexes were eluted from the resin mixture using 50 µL Elution Buffer. Proteins were denatured by adding 12.5 µL 5× SDS-PAGE Sample Buffer and boiling at 100 • C for 5 min, then applied to SDS-PAGE analysis.

Statistical Analysis
The statistical analyses were performed using the two-tailed Student's t-test for comparing two groups, or the one-way analysis of variance with Tukey's multiple comparison test for comparing among multiple groups. All data were presented as mean values with error bars representing standard error of mean. Probability of less than 5% (p < 0.05) was considered to be statistically significant.

Conclusions
In conclusion, we demonstrated that the satellite cell populations from different muscle fiber types are heterogeneous and they possessed an intrinsically predetermined property to generate specific muscle fiber types similar to their original muscle. Furthermore, Sox6 could directly promote the fast fiber isoform transcriptions and the different effects of Sox6 on slow-type fibers were due to the distinct expression patterns of Nfix between PM-MTs and LM-MTs. Highly expressed Nfix in PM-MTs was bound with Sox6 to form a complex, inhibiting slow-type fiber formation. However, Nfix in LM-MTs was insufficient to exhibit the same role, resulting in the Mef2C activation by Sox6 and increased regulation of slow fiber isoforms by Mef2C (Figure 9). These findings could partially explain the myogenic differences of satellite cells from different muscles, helping in broadening the understanding of molecular mechanisms involved in the satellite cell myogenesis and chicken skeletal muscle development. erty to generate specific muscle fiber types similar to their original muscle. Furthermore, Sox6 could directly promote the fast fiber isoform transcriptions and the different effects of Sox6 on slow-type fibers were due to the distinct expression patterns of Nfix between PM-MTs and LM-MTs. Highly expressed Nfix in PM-MTs was bound with Sox6 to form a complex, inhibiting slow-type fiber formation. However, Nfix in LM-MTs was insufficient to exhibit the same role, resulting in the Mef2C activation by Sox6 and increased regulation of slow fiber isoforms by Mef2C (Figure 9). These findings could partially explain the myogenic differences of satellite cells from different muscles, helping in broadening the understanding of molecular mechanisms involved in the satellite cell myogenesis and chicken skeletal muscle development. Figure 9. The regulation mechanism of Sox6 in satellite cells from fast-type and slow-type muscles. In brief, Sox6 promotes the proliferation and differentiation of PM-MBs and LM-MBs, improves the fast-type fiber formation. The upregulation of Nfix in PM-MTs inhibits the slow-type fiber formation via the collaboration with Sox6. In LM-MTs, Sox6 activates Mef2C transcription to promote the slowtype fiber formation.

Supplementary Materials:
The following supporting information can be downloaded at: www.mdpi.com/xxx/s1.

Informed Consent Statement: Not applicable.
Data Availability Statement: All data generated or analyzed during this study are included in this published article and its additional file.

Conflicts of Interest:
The authors declare no conflict of interest.