TLE4 regulates muscle stem cell quiescence and skeletal muscle differentiation

The mammalian adult skeletal muscle exhibits remarkable regenerative capability mediated by the skeletal muscle stem (satellite) cells, resident between the myofiber plasma membrane and basal lamina. Satellite cells in their niche are normally quiescent, but are activated upon muscle injury, leading to their proliferation, migration to the injury site and differentiation to repair the damage. Satellite cells are necessary for skeletal muscle regeneration, and in their absence, regeneration is severely compromised (Lepper et al., 2011; Murphy et al., 2011; Sambasivan et al., 2011). Pax7 is a key paired-domain transcription factor expressed by satellite cells (Seale et al., 2000). Germline Pax7-null mice exhibit reduced satellite cell numbers, compromised proliferation, increased apoptosis and aberrant muscle regeneration (Oustanina et al., 2004; Seale et al., 2000). Conditional Pax7 deletion showed that it is crucial at the perinatal stage, and required for satellite cell maintenance, expansion and adult muscle regeneration (Günther et al., 2013; Lepper et al., 2009; von Maltzahn et al., 2013).

Pax7 is essential for satellite cell function and regulates the expression of target genes that facilitate self-renewal, activation and differentiation (Soleimani et al., 2012). Notably, Pax7 regulates the expression of myogenic regulatory factors (MRFs), which commit muscle stem cells to a myogenic fate and promote their differentiation (Kuang et al., 2007; McKinnell et al., 2008; Zammit et al., 2006). The expression of Myf5, a MRF, is directly regulated by Pax7 in satellite cells via a −111 kb Myf5 enhancer (Soleimani et al., 2012; Zammit et al., 2004a). Whereas satellite cells are normally quiescent in the G0 cell cycle phase, but they are ready to get activated and proliferate if they receive the necessary cues (Dumont et al., 2015). Although all satellite cells express Pax7, they are heterogenous with respect to Myf5 expression (Beauchamp et al., 2000; Gayraud-Morel et al., 2012; Kuang et al., 2007). Satellite cells divide asymmetrically to give rise to Pax7⁺Myf5⁻ and Pax7⁺Myf5⁺ cells; Pax7⁺Myf5⁺ cells are more likely to differentiate compared to Pax7⁺Myf5⁻ cells which are thought to self-renew the satellite cell pool (Kuang et al., 2007).

Here, we identify the transducin-like enhancer of split-4 (TLE4) as a critical regulator of satellite cell quiescence. TLE family corepressors regulate fundamental cellular processes, such as growth, development and homeostasis (Agarwal et al., 2015). A recent study has reported that another TLE family member, TLE3, regulates myogenic differentiation by disrupting the interaction between MyoD and E proteins (Kokabu et al., 2017). Repression by TLE family proteins has been proposed to be mediated in different ways, such as by interacting with and recruiting histone deacetylases, direct histone binding and interacting with the transcriptional machinery (reviewed in Agarwal et al., 2015). We find that TLE4 is expressed in quiescent satellite cells, which is transiently downregulated during activation. We show for the first time that TLE4 interacts with Pax7 and occupies the −111 kb Myf5 enhancer element where Pax7 binds in satellite cells. Our results indicate that TLE4 is a corepressor that inhibits Pax7-mediated Myf5 activation in quiescent satellite cells, and whose transient downregulation during satellite cell activation leads to Myf5 upregulation and consequent differentiation. Accordingly, loss of TLE4 function leads to faster differentiation following muscle injury. Thus, our results demonstrate that TLE4 is a novel regulator of satellite cell function during skeletal muscle regeneration.

Since TLE family proteins play essential roles in cell fate specification and differentiation, we explored whether TLEs are expressed in the skeletal muscle. We found that TLE4 is expressed in the adult mouse skeletal muscle; TLE4 is co-expressed in the satellite cells labeled with Pax7 (Fig. 1A–A‴). TLE4 is also expressed in myonuclei and interstitial cells, which we have not explored further in this study (Fig. 1B–B‴). Most Pax7⁺ cells were TLE4⁺ under homeostasis, indicating that quiescent satellite cells express TLE4 (Fig. 1A–A‴). Using Myf5^nLacZ/+ mice (Tajbakhsh et al., 1996), where nLacZ labels quiescent satellite cells (Beauchamp et al., 2000), we found that LacZ⁺ nuclei are TLE4⁺, confirming that quiescent satellite cells express TLE4 (Fig. S1A–A″). We validated our findings using C2C12 mouse myogenic cells, where TLE4 was expressed by mononuclear cells during differentiation (Fig. S1B–E″). TLE4 expression in C2C12 cells was relatively high at early stages up to day 5 of differentiation, declining thereafter (Fig. S1F–H). Since TLE4 is expressed by satellite cells in vivo and its expression is relatively high at early differentiation stages in vitro, we hypothesized that it might be important in satellite cell function during muscle regeneration.

TLE4 levels increased significantly during regeneration following muscle injury (Fig. 1C; Fig. S1I). The increase in TLE4 expression was most prominent at 3 and 5 days post-injury (dpi), which is similar to the expression of Pax7, and declined to that of uninjured levels by 28 dpi (Fig. 1C–E; Fig. S1I). TLEs interact with homeodomain proteins in diverse contexts (Agarwal et al., 2015). Since Pax7⁺ satellite cells express TLE4, we tested whether TLE4 interacts with Pax7. Interestingly, Myc–Pax7 immunoprecipitated FLAG–TLE4, indicating that TLE4 and Pax7 interact (Fig. 1F). This interaction was validated endogenously using Pax7 and TLE4 antibodies in C2C12 cells (Fig. 1G).

Since TLE4 and Pax7 proteins interact, we analyzed their expression in satellite cells during regeneration. We found that >96% Pax7⁺ cells were TLE4⁺ and <4% Pax7⁺ cells were TLE4⁻ in uninjured tibialis anterior (TA), indicating that most satellite cells express TLE4 under homeostasis (Fig. 1H–H‴,K; Fig. S1J). In injured muscles at 3 dpi, ∼76% Pax7⁺ cells were TLE4⁺ (∼24% Pax7⁺ cells were TLE4⁻) (Fig. 1K; Fig. S1J). At 5 dpi, ∼84% Pax7⁺ cells were TLE4⁺ and ∼16% were TLE4⁻ (Fig. 1I–I‴, where arrowhead points to Pax7⁺TLE4⁻ nucleus, K; Fig. S1J). By 7 dpi, ∼94% Pax7⁺ cells were TLE4⁺ and ∼6% were TLE4⁻, whereas by 14 dpi Pax7⁺TLE4⁺ and Pax7⁺TLE4⁻ cells had returned to uninjured proportions (Fig. 1J–J‴,K; Fig. S1J). These results indicate that most satellite cells express TLE4 under homeostasis, while a significant fraction downregulate TLE4 up to day 5 post-injury, beyond which TLE4 expression is restored in most satellite cells.

We observed that ∼41% of TLE4⁻ satellite cells are proliferating [positive for phosphorylated histone H3 (PHH3⁺)] at 5 dpi in Myf5^nLacZ/+ mice (Fig. S1K,L). Since a significant proportion of proliferating satellite cells do not express TLE4, this suggested that activated satellite cells are the ones that downregulate TLE4. To confirm our observations, we performed mouse muscle fiber cultures. Muscle fibers along with resident satellite cells can be isolated and cultured to study satellite cell commitment to differentiation versus self-renewal. We found that a subset of satellite cells were TLE4⁻, with 14% at 0 h, 25% at 12 h, 19% at 24 h and 0% at 48 h (Fig. 2A–G). During muscle fiber isolation, satellite cells get activated. Thus, a transient downregulation of TLE4 occurs early during activation of satellite cells. These observations strengthen the notion that satellite cells that downregulate TLE4 might be the ones that are activated and committed to differentiation.

MyoD (encoded by Myod1), an MRF, is expressed by satellite cells upon activation; ∼89% of satellite cells on isolated fibers are MyoD⁺ by 24 h (Zammit et al., 2006). However, MyoD is expressed in activated satellite cells as early as 3 h post-injury (Cooper et al., 1999). Since 14% satellite cells downregulated TLE4 at 0 h of culture (Fig. 2G), we tested how early MyoD is detectable. At 0 h, ∼9% satellite cells are MyoD⁺, 94% at 12 h and 99% by 24 h (Fig. 2H–K). It is intriguing that 9% satellite cells get activated rapidly and are MyoD⁺ at 0 h. The majority of satellite cells are activated within 12 h and almost all are activated by 24 h of fiber culture.

To directly test whether TLE4 downregulation correlates with satellite cell activation, we labeled satellite cells with MyoD and TLE4. At 0 h, ∼67% of MyoD⁺ cells are TLE4⁻, indicating that activated satellite cells indeed downregulate TLE4 (Fig. 2L–M″,P). By 12 h, only ∼2% MyoD⁺ cells are TLE4⁻ and by 24 h there are no MyoD⁺TLE4⁻ satellite cells, suggesting that this TLE4 downregulation is transient (Fig. 2N–P). About one-third of MyoD⁺ satellite cells are TLE4⁺ at 0 h, possibly because they are past the brief window of TLE4 downregulation. These results support our hypothesis that transient downregulation of TLE4 takes place during, and could be essential for, satellite cell activation.

By 72 h, more than half the satellite cell pool and progeny continue to express Pax7, with ∼23% expressing Pax7 without MyoD (Zammit et al., 2004b). Previous studies suggest that Pax7⁺MyoD⁻ satellite cells are located in clusters, and arise from Pax7⁺MyoD⁺ cells which downregulate MyoD to regain quiescence (Zammit et al., 2004b). We find that almost all Pax7⁺ satellite cell clusters are TLE4⁺ at 72 h, indicating that satellite cells that repopulate the niche and become quiescent express TLE4 (Fig. S1M–M″).

To characterize the precise role of TLE4 in muscle differentiation, we performed Tle4 knockdown on C21C2 cells and studied the effect on key regulators of muscle differentiation. A sequence-specific siRNA effectively silenced TLE4 at days 5 and 7 of differentiation (Fig. 3A,B,H; Fig. S2A–B″). Myf5 and MyoD are the two early regulators of myogenic differentiation; we observed a striking ∼6–7-fold upregulation of Myf5 levels upon Tle4 knockdown (Fig. 3A,C). MyoD was upregulated to a lesser extent (Fig. 3A,D), while myogenin (MyoG) levels initially declined and then increased (Fig. 3A,E). Levels of the terminal differentiation markers myosin heavy chain-embryonic (MyHC-emb, encoded by Myh3) and MyHC-slow (encoded by Myh7) exhibited a prominent reduction (Fig. 3A,F,G).

Since TLE4 is a corepressor, and Myf5, MyoD and MyoG protein levels increased upon Tle4 knockdown, we tested whether TLE4 transcriptionally represses these MRFs to regulate differentiation. A striking, ∼5-fold (day 5) and ∼3-fold (day 7) upregulation of Myf5 transcript levels was seen upon Tle4 knockdown, suggesting that TLE4 represses Myf5 activation (Fig. 3I). Interestingly, this repressive effect was Myf5-specific, since MyoD levels were unchanged upon Tle4 knockdown (Fig. 3J). Levels of the late myogenic differentiation regulator MyoG was significantly reduced, possibly due to impaired differentiation (Fig. 3K). Confirming this, MyHC-emb and MyHC-slow levels were also significantly reduced (Fig. 3L,M). These results were validated using a second Tle4 siRNA (Fig. S2C,D). Elevated MyoD protein levels (Fig. 3A,D), without any significant change in MyoD transcript levels (Fig. 3J) suggest that MyoD protein stability is altered upon Tle4 knockdown. Although we have not studied this further, one possibility is that the impaired differentiation resulting in reduced levels of terminal myogenic differentiation markers upon Tle4 knockdown leads to a stabilization of MyoD protein to promote differentiation.

Since the protein and transcript levels of the terminal differentiation markers MyHC-emb and MyHC-slow were significantly decreased upon Tle4 knockdown, we tested the effect of TLE4 on differentiation capability. A striking decrease in differentiated myofibers and a 4–6-fold reduction in fusion index was observed upon Tle4 knockdown in C2C12 cells (Fig. 3N–R). To validate this, primary myoblasts isolated from wild-type neonatal mice were transfected with control or Tle4 siRNA and allowed to differentiate. Confirming our C2C12 cell experiment results, we detected a decreased number of differentiated myofibers and a significantly reduced fusion index upon Tle4 knockdown in primary myoblasts (Fig. 3S–U). These results demonstrate that TLE4 is essential for proper myogenic differentiation.

While caspase-3 levels were unaffected upon Tle4 knockdown, indicating that the observed effects were not mediated by altered cell death, PHH3 levels showed an increase at day 3 of differentiation suggesting that increased proliferation might play a role in the effect of TLE4 on differentiation (Fig. S2E–H).

Our results indicate that loss of TLE4 function leads to upregulation of Myf5, which could cause the observed differentiation defects. Therefore, we wished to test whether downregulating Myf5 would rescue the effect of loss of TLE4 function. To achieve this, we transfected C2C12 cells with either Tle4 siRNA, Myf5 siRNA or a combination of both Tle4 and Myf5 siRNAs and allowed the cells to differentiate. Interestingly, we found that the protein levels of the terminal differentiation marker MyHC-emb, which is significantly decreased upon Tle4 knockdown, is rescued at least partially by Tle4+Myf5 siRNA treatment at day 5 of differentiation (Fig. 4A,B). The C2C12 cell fusion index upon Tle4+Myf5 siRNA treatment at day 5 of differentiation was significantly increased compared to transfection of Tle4 siRNA alone (Fig. 4C; Fig. S2J–M); however, although this rescue of fusion index was quite noticeable compared to Tle4 siRNA-treated cells, it was not complete, compared to the fusion index of the control siRNA treated cells (Fig. 4C; Fig. S2J–M). Overall, these results indicate that downregulation of Myf5 at least partially rescues the fusion index defects caused by loss of TLE4 function (Fig. 4C; Fig. S2J–M). We also find that the protein levels of MyHC-emb are significantly elevated upon Tle4 knockdown at day 3 of differentiation, indicating that at early timepoints, loss of TLE4 function leads to increased differentiation (Fig. S2I,I′). Overexpression of Myf5 by transfecting a Myf5–Myc construct into C2C12 cells resulted in increased MyHC-emb levels at days 3 and 5 of differentiation and increased fusion index at day 3 of differentiation, indicating that upregulation of Myf5 can lead to effects on myogenic differentiation (Fig. S3A–D).

To test the effect of upregulation of TLE4, we transfected a FLAG-TLE4 construct that resulted in a 2-fold overexpression of Tle4 and significant downregulation of Myf5 transcript levels (Fig. 4D,E; Fig. S3E). This also led to elevated MyoD, MyHC-emb and MyHC-slow levels, and an increased fusion index at days 3 and 5 of differentiation (Fig. 4F–N). Thus, while loss of TLE4 function caused Myf5 upregulation, TLE4 overexpression led to its repression, indicating that TLE4 is critical for Myf5 regulation and myogenic differentiation.

A Myf5 enhancer element 111 kb upstream of the Myf5 transcription start is bound by Pax7 to activate Myf5 transcription in satellite cells (Soleimani et al., 2012; Zammit et al., 2004a) (Fig. 5A). Since TLE4 interacts with Pax7, and Tle4 knockdown causes Myf5 upregulation, we hypothesized that TLE4 represses Myf5 by occupying the −111 kb Myf5-enhancer. To test this, we carried out chromatin immunoprecipitation (ChIP) and found that TLE4 occupies the −111 kb Myf5 enhancer, compared to a control region (Fig. 5B). In luciferase assays transfecting the −111 kb Myf5-luciferase along with Pax7 into C2C12 cells, we observed a significant increase in luciferase activity due to Pax7-mediated activation of Myf5 (Soleimani et al., 2012). Upon co-transfecting Tle4 along with Pax7 and Myf5-luciferase, a significant decrease in luciferase activity was observed, validating that TLE4 represses Pax7-mediated activation of Myf5 via the −111 kb Myf5 enhancer (Fig. 5C). Thus, we have identified a new mechanism by which Myf5 expression is regulated during muscle differentiation. When the −111 kb Myf5-luciferase and Pax7 were electroporated with or without Tle4 into the TA muscle, which was then injured, the presence of Tle4 led to a significant decrease in Myf5-luciferase activity confirming that TLE4 represses Pax7-mediated activation of Myf5 transcription in satellite cells (Fig. 5D). Interestingly, TLE4 also occupied a −57.5 kb Myf5 enhancer bound by Pax7 that activates Myf5 expression during development and in proliferating myogenic cells (Soleimani et al., 2012) (Fig. S3F–H). Thus, TLE4 might function as a repressor of Pax7-mediated Myf5 activation in multiple contexts during development and regeneration.

We next tested TLE4 function during regeneration, where satellite cell activation and differentiation are critical. We electroporated Tle4 or control siRNA into adult mice TA (Yoshida et al., 2014) and found that TLE4 levels were significantly reduced by >50% at 3, 5 and 7 dpi (Fig. 5E,F,L,M,S,T). Tle4 knockdown led to a significant upregulation of Myf5 levels at all time points, confirming its essential role in repressing Myf5 (Fig. 5E,G,L,N,S,U). This effect was specific to Myf5 and not MyoD, except at 3 dpi (Fig. 5E,I,L,P,S,W).

Myf5 upregulation has been reported to lead to activation of satellite cells (Kawabe et al., 2012; Kuang et al., 2007; Soleimani et al., 2012). Since TLE4 depletion led to elevated Myf5 levels, we next tested the effect on Pax7 and observed a significant increase in Pax7 levels (Fig. 5E,H,L,O,S,V). To distinguish whether the elevated Pax7 levels resulted from increased Pax7 expression or an increase in Pax7⁺ cell numbers, we quantified the number of satellite cells following Tle4 knockdown. Pax7⁺ cell numbers more than doubled compared to controls, indicating that increased satellite cell numbers contributed at least partially to elevated Pax7 levels upon TLE4 depletion (Fig. S5A–I). Satellite cell activation causes reentry into cell cycle and increased proliferation, which explains the increase in satellite cell numbers following Tle4 knockdown (Dumont et al., 2015). Interestingly, Myf5 directly regulates cyclin D1, a critical cell cycle regulator of satellite cell proliferation (Panda et al., 2016). This is validated by increased proliferation (PHH3 levels) following Tle4 knockdown (Fig. 5E,K,L,R,S,X). This fits well with the in vitro data, where PHH3 levels increased at 3 days of differentiation and thereafter remained unchanged upon Tle4 knockdown (Fig. S2F,H).

MyHC-emb, a differentiation marker for regenerating fibers, levels increased significantly at 3 dpi and decreased at 5 and 7 dpi upon Tle4 knockdown (Fig. 5E,J,L,Q,S,Y). Similarly, MyHC-emb⁺ regenerating muscle fiber numbers increased at 3 dpi, and decreased by ∼50% at 5 and 7 dpi upon Tle4 knockdown (Fig. 6A–F,I; Fig. S4A–F‴). Thus, loss of TLE4 function leads to Myf5 upregulation, and accelerated differentiation at early stages of regeneration (increased MyoD, MyHC-emb levels and MyHC-emb⁺ fiber numbers at 3 dpi) compared to later stages (decrease in MyHC-emb levels and MyHC-emb⁺ fiber numbers at 5 and 7 dpi). This is comparable to the in vitro data with C2C12 cells, where we observed increased MyHC-emb levels at day 3 of differentiation (Fig. S2I,I′) and reduced levels at days 5 and 7 (Fig. 3A,F; Fig. 4A,B) of differentiation, upon Tle4 knockdown. The area occupied by connective tissue (Fig. 6G,H,J) and the frequency of myofibers with larger area (1000–1600 µm²) (Fig. 6K,K′) were significantly increased upon Tle4 knockdown, indicating that regeneration is affected. Since Tle4 knockdown by electroporation is not specific to satellite cells, it is possible that some of the effects we observe might be due to Tle4 depletion in other cell types in the regenerating muscle. Overall, these in vivo experiments confirm that TLE4 function is essential for proper skeletal muscle regeneration.

In this study, we have identified a novel mechanism by which the corepressor TLE4 regulates muscle stem cell quiescence and differentiation by means of its interaction with the homeodomain transcription factor Pax7. Several reports indicate that TLEs interact with homeodomain family proteins (Agarwal et al., 2015). For example, Pax3 interacts with TLE4 to repress target genes and regulate melanocyte stem cell differentiation (Lang et al., 2005). TLE4 interacts with Pax5 in B cells, enabling Pax5 switching between activator and repressor roles (Eberhard et al., 2000; Linderson et al., 2004). TLE4 interacts with Pax2 to regulate its transactivation via the methyltransferase PRMT5 and polycomb proteins (Cai et al., 2003; Patel et al., 2012). This is the first study to identify the interaction between TLE4 and Pax7, which we report to be critical for satellite cell quiescence and differentiation.

Our study finds interesting parallels in TLE4 function during C2C12 in vitro differentiation and in vivo muscle regeneration following injury. Levels of the terminal differentiation marker MyHC-emb increase significantly at early stages of differentiation (day 3 of differentiation in vitro and 3 dpi in vivo), and decline thereafter (days 5 and 7 of differentiation in vitro; 5 and 7 dpi in vivo), upon loss of TLE4 function. This suggests that the early effect of loss of TLE4 is acceleration of differentiation, which means that TLE4 expression is required to prevent differentiation. The reduced differentiation seen at later stages could be due to the premature differentiation at early timepoints, which can lead to exhaustion of the stem cell pool that is available for differentiation. On the other hand, it is also possible that TLE4 plays a separate role in promoting differentiation at later stages.

Maintenance of quiescence by satellite cells is critical to prevent their premature differentiation and for maintaining regenerative potential. Interestingly, the Tle4 transcript was enriched, along with transcripts of genes involved in maintaining quiescence, such as Bcl2, Cdkn1b and PDK4, in quiescent satellite cells in a recent study comparing quiescent and activated satellite cell transcriptomes (van Velthoven et al., 2017). Our findings not only validate that TLE4 is expressed in quiescent satellite cells, but also identify the mechanism by which TLE4 regulates stem cell quiescence and differentiation. We propose a model where TLE4 interacts with Pax7 and occupies the −111 kb Myf5 enhancer in quiescent satellite cells, to prevent transcriptional activation of Myf5. Since activation of Myf5 is crucial for satellite cells to initiate differentiation, these findings indicate that TLE4 is critical to maintain satellite cell quiescence under homeostasis. In activated satellite cells, a transient downregulation of TLE4 permits Pax7-mediated Myf5 transcriptional activation, allowing satellite cells to differentiate and regenerate damaged muscle fibers (Fig. 6L). Intriguingly, the −111 kb Myf5 enhancer occupied by TLE4 is reported to be hypermethylated in embryonic stem cells, and this enhancer underwent partial demethylation in proliferating satellite cells (Carrio et al., 2015). Reports indicate that TLE4 suppresses Pax-mediated gene activation by inhibiting H3K4 methylation and promoting H3K27 methylation (Patel et al., 2012). In uninjured muscle, a small fraction (3–4%) of satellite cells do not express TLE4. Lineage tracing experiments demonstrated that satellite cells contribute to muscle fibers during homeostasis (Keefe et al., 2015); the small proportion of TLE4⁻ satellite cells might be the activated cells embarking on differentiation during homeostasis. The mechanism regulating the transient TLE4 downregulation to permit Myf5 transcriptional activation and satellite cell differentiation is currently unknown and would be of great interest in future studies. This should have enormous implications on our understanding of muscle regeneration following injury or in diseases such as Duchenne muscular dystrophy.

C57Bl/6 wild-type and Myf5^nLacZ (Tajbakhsh et al., 1996) mice were used in this study. All animal procedures in this study were reviewed and approved by the Institutional Animal Ethics Committee (RCB/IAEC/2016/008 and RCB/IAEC/2016/021). Muscle injury was induced by injecting 25 μl of 1.2% BaCl₂ in normal saline into the right TA muscle of mice (Murphy et al., 2011). The left TA served as the uninjured control. The TAs were collected at 3, 5, 7, 14 and 28 dpi, flash frozen and stored at −80°C (Murphy et al., 2011). Myofibers were isolated from the extensor digitorum longus (EDL) muscle of 6–8-week-old male mice. Isolated myofibers were either fixed and processed immediately (0 h) or cultured in suspension for 12, 24, 48 and 72 h (Pasut et al., 2013). Myofibers were fixed in 4% paraformaldehyde and processed for immunofluorescence. Myofibers from at least five mice were isolated and a minimum of 15 fibers per mouse were analyzed per time point.

For primary myoblast isolation, hind limb muscles were dissected from wild-type mice at postnatal day 0 (P0) and digested in 2 mg/ml type II collagenase (Gibco, 17101-015) in Dulbecco's modified Eagle's medium (DMEM; Gibco, 11995065) supplemented with 1% penicillin-streptomycin (Gibco, 15140122) for 90 min. The cell suspension was passed through 70 µm and 40 µm strainers and the filtered lysate centrifuged at 1400 g for 10 min at room temperature. The pellet was resuspended in culture medium containing DMEM supplemented with 10% (v/v) fetal bovine serum (Sigma-Aldrich, F2442) and 1% penicillin-streptomycin and pre-plating was done on cell culture-treated plastic 100 mm dishes. After 3 h of pre-plating, the nonadherent cells were transferred to 10% Matrigel-coated 60 mm Petri dishes with 10 ng/ml FGF2 (Sigma-Aldrich, SRP4038). After 24 h, the cells were trypsinized and transferred to wells of a 24-well dish containing gelatin-coated coverslips and transfected with control siRNA or Tle4 siRNA, similar to how it is performed for C2C12 cells (see next section). Culture medium was changed to differentiation medium (see next section for compositions) at 24 h post-transfection and the cells were cultured for 4 days with 50% daily replacement of differentiation medium. The coverslips with differentiated primary myoblasts were processed for immunofluorescence labeling to quantify the fusion index as described for C2C12 cells below.

C2C12 mouse myogenic cells were cultured and maintained as recommended by the American Type Culture Collection (ATCC). For differentiation, ∼30,000 cells per well were plated in complete medium (DMEM containing 10% fetal bovine serum and 2% penicillin-streptomycin) in a 24-well dish. The day of plating was considered day 0. After 48 h (day 2), the complete medium was replaced with differentiation medium (DMEM containing 5% horse serum and 2% penicillin-streptomycin). 50% of the medium was replaced with fresh differentiation medium at day 3 onwards.

Gene knockdown in C2C12 myoblasts was carried out by reverse transfection of siRNA, using Tle4, Myf5, Tle4+Myf5 or control siRNA. The transfection mix comprised 100 µl Opti-MEM (Gibco; Cat# 31985070), 50 nM of Tle4 (Ambion; Cat# 4390816, ID: s232475 and Eurogentec SR-NP001-001 Custom siRNA), 50 nM of Myf5 (Ambion; Cat# 4390771, ID:s70242) or control siRNA (Ambion; Cat# 4390816, ID: 4390847) and 2 µl Lipofectamine RNAiMAX (Invitrogen; Cat# 13778150) (Agarwal et al., 2020). NIH-3T3 cells were cultured and maintained as recommended by the ATCC and transfected using Lipofectamine 2000 (Invitrogen; Cat# 11668019) as per the manufacturer's protocol.

Total RNA was extracted from cells using RLT lysis buffer (Qiagen) according to the manufacturer's instructions, followed by cDNA synthesis using reverse transcriptase and oligo-dT primer. The cDNA samples were used as template for quantitative PCR (qPCR) on an ABI 7500 Fast Real-Time System (Applied Biosystems) using primers specific to each gene (Table S1). Expression levels of the genes examined were normalized to Gapdh expression for each sample. Taqman probes were used to quantify the expression of Tle4 (Applied Biosystems; Cat# Mm 00925591_m1), Myf5 (Applied Biosystems; Cat# Mm 00435125_m1), and Gapdh was used as normalizing control (Applied Biosystems; Cat# 99999915_g1).

Whole TA samples were embedded in optimal cutting temperature (OCT) cryomatrix, flash frozen in 2-methyl butane cooled in liquid nitrogen and 10 µm transverse sections cut on a cryomicrotome (Thermo Fisher Scientific; Microm HM 550). TA sections were fixed in 4% paraformaldehyde (PFA), blocked in 5% goat serum (Himedia; Cat# RM10701-500 ml), incubated with primary antibodies, and treated with biotin-conjugated secondary antibodies (list of antibodies used provided in Table S2). This was followed by treatment with fluorophore coupled to streptavidin; where required, the signal was amplified using a TSA kit (Perkin Elmer, NEL741B001KT). The sections were mounted using Fluoromount-G (Southern Biotech; Cat# 0100-01).

For histological analysis, slides were incubated with 0.1% Sirius Red (Sigma; Cat# 365548) solution dissolved in aqueous saturated picric acid for 1 h, washed in water, dehydrated and mounted with DPX (Sigma; Cat# 06522). Sirius Red is a histological dye which marks the connective tissue red and muscle fibers yellow (Murphy et al., 2011).

For immunofluorescence of cultured cells, C2C12 cells were cultured on gelatin-coated coverslips in a 24-well dish. Cells were fixed by incubating with 4% PFA for 20 min at room temperature, blocked in 5% goat serum, treated with primary antibodies overnight at 4°C and secondary antibodies for 2 h at room temperature and mounted using Fluoromount-G. For fusion index analysis, the coverslips with cells were harvested on days 5 and 7 of myogenic differentiation and labeled using a mixture of myosin heavy chain antibodies (Mathew et al., 2011). TLE4 antibody specificity was verified by labeling C2C12 cells treated with control or Tle4 siRNA with the TLE4 antibody (Fig. S2A–B″).

For immunofluorescence of isolated fiber preparations, myofibers were fixed by adding 4% PFA, washed with PBS, blocked in 5% goat serum for an hour and incubated with primary antibodies overnight at 4°C. Next day, fibers were washed with PBS, incubated with secondary antibodies at room temperature and mounted using Fluoromount-G.

Cell pellets and muscle tissue (TA) were lysed in RIPA buffer (Sigma; Cat# R0278-500 ml) containing 1% protease inhibitor (Sigma, P8340-5 ml). Samples were then subjected to SDS-PAGE, followed by transfer to PVDF membrane (Millipore; Cat# iPVH00010) and subsequent immunoblotting. All antibodies used are listed in Table S2.

For quantification of total protein levels in uninjured and injured TA muscles, protein lysates were separated by SDS-PAGE, imaged on a LICOR Odyssey infrared imaging system in the red channel and total band intensity in each lane quantified (Eaton et al., 2013). This was used to normalize band intensity of western blots in uninjured to injured TA comparisons.

Cell lysates in a total volume of 500 µl were immunoprecipitated with 2 µg of each respective antibody and 50 µl of protein G Sepharose 4 fast flow beads (GE Healthcare; Cat# 17-0618-01). The samples were incubated on a nutator at 4°C overnight. After the beads were washed four times with lysis buffer, the pellets were analyzed by SDS-PAGE.

The plasmids pCMV3tag-3b-Tle4-FLAG and pCMV-3tag-4b-Myc-Pax7 were generated by cloning mouse Tle4 and Pax7 into the pCMV3tag-3b-FLAG and pCMV-3tag-4b-Myc empty plasmids (Agilent) respectively. pGL4.27-luciferase and pGL4.27-Myf5 (−57.5 kb) enhancer-luciferase plasmids were kindly provided by John Epstein (Perelman School of Medicine, University of Pennsylvania, USA; Bajard et al., 2006; Manderfield et al., 2014). pGL4.10-Myf5 (−111 kb) enhancer-luciferase plasmid was a generous gift from Prof. Michael Rudnicki, Ottawa Health Research Institute, Canada (Soleimani et al., 2012). For the luciferase assay, C2C12 cells were transfected with the pGL4.27-Myf5 (−57.5 kb) luciferase or pGL4.10-Myf5 (−111 kb) enhancer-luciferase reporter vectors alone, with pCMV-3tag-4b-Myc-Pax7, with pCMV3tag-3b-Tle4-FLAG, or with both pCMV-3tag-4b-Myc-Pax7 and pCMV3tag-3b-Tle4-FLAG using Amaxa Cell line nucleofector Kit V (Lonza; Cat# VCA 1003) on a Nucleofector 2b (Lonza; Cat# AAB-1001). The total plasmid quantity transfected was kept constant (6 μg per 2×10⁶ cells) by adding empty pCMV3tag-3b-FLAG plasmid. phRL-CMV (Promega) was used to normalize transfection efficiency. After 48 h, cells were washed with PBS, lysed and luciferase assays performed using the Dual-Glo luciferase assay system (Promega; Cat# E2920).

For the overexpression experiments, C2C12 cells were transfected with pCMV3tag-3b-FLAG (control) or pCMV3tag-3b-Tle4-FLAG (Tle4 overexpression) and pCMV-3tag-4b-Myc (control) or pCMV-3tag-4b-Myc-Myf5 (Myf5 overexpression) using Lipofectamine 2000, at a concentration of 6 μg per 2×10⁶ cells.

Undifferentiated C2C12 cells were fixed in 1% formaldehyde solution and quenched using 0.125 M glycine. Chromatin from fixed cells were fragmented to a size range of 300–500 bp by sonication (Notani et al., 2010). Solubilized fragmented chromatin was immunoprecipitated using the TLE4 antibody. Antibody–chromatin complexes were pulled down using protein G beads (Dynabeads, Invitrogen; Cat# 10003D), washed and eluted in elution buffer. Reverse crosslink was performed subsequently at 65°C overnight. Fragmented DNA was then isolated using phenol-chloroform and precipitated with chilled ethanol. DNA fragments acquired were subjected to qPCR using Myf5-enhancer-specific primers and control primers from a −315 bp region upstream of Myf5 on chromosome 10 (McKinnell et al., 2008). Bead-only negative controls were used and a minimum of three independent experiments were carried out.

For in vivo electroporation of siRNA into the TA, 6–8-week-old mice were anesthetized using an appropriate dose of xylazine/ketamine, and 35 μl of siRNA (20 µM) was injected into the TA with a 22-gauge syringe. Tle4 siRNA was electroporated into the right TA and control siRNA into the left TA. Transcutaneous standard square wave pulses were delivered using stainless steel electrodes according to manufacturer's instructions (ECM830 Square Wave Electroporation System, BTX). Ten pulses of 170 V/cm were administered to the muscle with each pulse of 20 microsecond duration, and interval of 1 s each (Yoshida et al., 2014). The electroporated TAs were injured by injecting BaCl₂ as described, after 24 h of electroporation. The TAs were harvested at 3, 5 and 7 days post injury (dpi).

Reporter and plasmid DNA constructs in a volume of 30 μl (40 μg) were electroporated into the TA under similar settings and conditions. TA muscles were electroporated with the pGL4.27-Myf5 (−111 kb) enhancer-luciferase reporter vector alone, with pCMV-3tag-4b-Myc-Pax7, with pCMV3tag-3b-Tle4-FLAG, or with both pCMV-3tag-4b-Myc-Pax7 and pCMV3tag-3b-Tle4-FLAG, injured using BaCl₂ 48 h after electroporation, and harvested 48 h-post-injury for luciferase assay.

Fluorescence microscopy was performed using a Leica TCS SP8 confocal microscope, or an Olympus BX63F epifluorescence microscope with an ORCA-Flash4 monochrome camera (Hamamatsu). Sirius Red-stained images were acquired on a Nikon Eclipse Ti microscope using a DS-FI2 camera. For C2C12 cell fusion index quantification, Tle4 and control siRNA-treated cells were cultured, stained for phalloidin, myosin heavy chains (mixture of My32 and MyHC-slow antibodies) and DAPI, and images captured and maximally projected. The number of MyHC-positive fibers and DAPI⁺ nuclei within fibers were quantified using ImageJ (Schindelin et al., 2012; Schneider et al., 2012). The total number of DAPI⁺ nuclei were counted using spot and annotation functions in Imaris software (http://www.bitplane.com/). Muscle cross-sectional area was measured using CellSens Dimension 1.16. Satellite cell counts on fiber preps and cross-sections for Pax7⁺TLE4⁺, Pax7⁺TLE4⁻, Pax7⁺MyoD⁺, Pax7⁺MyoD⁻, TLE4⁺MyoD⁺ and TLE4⁻MyoD⁺ were carried out manually in Adobe Photoshop following image capture. Control and Tle4 siRNA electroporated TA muscle fiber count and fiber area were measured using the Semi-automatic muscle analysis using segmentation of histology software (SMASH) (Smith and Barton, 2014). Laminin labeled fiber numbers were counted using fiber typing function, and muscle fiber area was measured using the fiber properties function of SMASH software, respectively, in TA muscles of control and Tle4 siRNA-electroporated mice.

Each experiment includes data from a minimum of three independent replicates, with the data represented as mean±s.e.m. Data from all the experiments were analyzed using a parametric, two-tailed unpaired t-test using the Graphpad Prism software. The P-value is indicated on the graph, with P-values ≤0.05 considered significant and marked with asterisks.

We acknowledge valuable suggestions and help from Dr Suchitra Gopinath, THSTI, for muscle fiber preparation and culture experiments. We thank Dr Dimple Notani, NCBS, for inputs on the ChIP assay. The pGL4.27-luciferase and pGL4.27-Myf5 (−57.5 kb) luciferase plasmids were generous gifts from Prof. John Epstein, University of Pennsylvania, and the pGL4.10-Myf5 (−111 kb) enhancer-luciferase plasmid was a generous gift from Prof. Michael Rudnicki, Ottawa Health Research Institute. We thank the RCB microscopy facility for imaging help and the small animal facility (SAF) at the NCR Biotech Science Cluster for help with the animal work. We thank Pankaj Kumar for help and suggestions with the in vivo electroporation experiments. We also acknowledge past and present members of the SJM lab for valuable suggestions and inputs. Authors acknowledge the support of DBT e-Library Consortium (DeLCON) for providing access to e-resources.

The peer review history is available online at https://journals.biologists.com/jcs/article-lookup/doi/10.1242/jcs.256008.