TGF-β-driven downregulation of the transcription factor TCF7L2 affects Wnt/β-catenin signaling in PDGFRα⁺ fibroblasts

The remarkable long-term capacity of adult skeletal muscle to grow and regenerate is largely based on the presence of tissue-resident muscle stem cells (MuSCs, formerly called satellite cells) (Lepper et al., 2011; Murphy et al., 2011; Sambasivan et al., 2011). However, another tissue-resident population of quiescent mesenchymal stromal cells (MSCs, also known as mesenchymal stem cells) is also essential for effective regeneration and maintenance of the skeletal muscle and its connective tissue (CT). Following acute damage, PDGFRα⁺ cells provide regenerative cues, regulate extracellular matrix (ECM) remodeling, and MuSC behavior, thus facilitating muscle regeneration and maintenance (Contreras, 2019d; Heredia et al., 2013; Joe et al., 2010; Scott et al., 2019; Uezumi et al., 2010; Wosczyna et al., 2019). These stromal mesenchymal progenitors are also called fibro-adipogenic progenitors (FAPs) in mouse and human skeletal muscle (Agley et al., 2013; Dulauroy et al., 2012; Uezumi et al., 2014a; Vallecillo-García et al., 2017; Wosczyna and Rando, 2018). Muscle FAPs have the potential to differentiate in vivo and in vitro into myofibroblasts, adipocytes, chondrogenic cells and osteogenic cells (Agley et al., 2013; Contreras et al., 2019c; Oishi et al., 2013; Uezumi et al., 2014b, 2010, 2011; Wosczyna et al., 2012). PDGFRα⁺ mesenchymal progenitors are found in most mammalian tissues, where they regulate tissue homeostasis, repair and regeneration (Carr et al., 2019; Riquelme-Guzmán and Contreras, 2020; Lemos and Duffield, 2018; Lynch and Watt, 2018; Rognoni et al., 2018). Despite their required normal activity during muscle regeneration, we and others have suggested dysregulated behavior of these mesenchymal precursor cells in models of chronic muscle damage, muscular dystrophy (MD), neurodegenerative diseases and aging (Acuña et al., 2014; Contreras et al., 2016, 2019c; González et al., 2017; Ieronimakis et al., 2016; Kopinke et al., 2017; Madaro et al., 2018; Mahmoudi et al., 2019; Lemos et al., 2015; Lukjanenko et al., 2019; Uezumi et al., 2014a). Fibrosing disorders are a common outcome of the dysregulation of these fibroblastic cells and include non-malignant fibroproliferative diseases with high morbidity and mortality (Lemos and Duffield, 2018; Wynn and Ramalingam, 2012). However, the participation of mesenchymal progenitors in repair and regeneration remains poorly understood.

The Wnt signaling pathway has a key role in many aspects of developmental biology, tissue homeostasis, stem cell fate, organogenesis, cancer and tissue fibrogenesis. Secreted Wnt ligands, which are 19 secreted lipid-modified glycoproteins, bind to Frizzled receptors and the co-receptors LRP5 and LRP6 on the cell surface to initiate a signaling pathway that regulates the proteostasis of cytoplasmic β-catenin. In the absence of canonical Wnt ligands, β-catenin is a target for degradation mediated by the ubiquitin–proteasome system (UPS) (Aberle et al., 1997). Once the canonical cascade starts, the accumulation of β-catenin leads to its translocation to the nucleus, where it recognizes and binds the T-cell factor (TCF) or lymphoid enhancer factor (LEF) transcription factors (TFs), and recruits transcriptional partners and chromatin remodeling complexes, which in concert regulate the expression of TCF/LEF target genes. These TFs recognize TCF/LEF-binding sites in the regulatory elements of target genes to regulate the expression of thousands of genes (Cadigan and Waterman, 2012; Clevers, 2006; Schuijers et al., 2014). Therefore, the TCF/LEF TFs are the final effectors of the canonical Wnt/β-catenin signaling cascade in metazoans (Nusse and Clevers, 2017; Schuijers et al., 2014; Tang et al., 2008). Mammalian cells express four TCF/LEF protein-coding genes: TCF1 (also known as TCF7), LEF1, TCF7L1 (formerly named TCF3), and TCF7L2 (formerly named TCF4) (Cadigan and Waterman, 2012; Korinek et al., 1997; van de Wetering et al., 1991). They have critical roles regulating body plan establishment, cell fate specification, proliferation, survival and differentiation, which are predominant features in fast and constantly renewing tissues (Clevers, 2006; Korinek et al., 1998; Van De Wetering et al., 2002). Accumulating evidence indicates that pathologically activated canonical Wnt signaling plays a major role in the pathogenesis of fibrosis in multiple tissues (Chilosi et al., 2003; Cisternas et al., 2014; Colwell et al., 2006; Cosin-Roger et al., 2019; He et al., 2009, 2010; Henderson et al., 2010; Konigsho et al., 2008; Liu et al., 2009; Surendran et al., 2002; Trensz et al., 2010; Wei et al., 2011). For example, the conditional genetic loss of β-catenin in cardiac fibroblast–myofibroblast lineages (TCF21⁺ or periostin⁺) causes the reduction of interstitial fibrosis and attenuates heart hypertrophy induced by cardiac pressure overload (Xiang et al., 2017). Furthermore, upregulated Wnt activity during aging accounts, in part, for the declining regenerative potential of muscle with age and increased fibrosis (Brack et al., 2007). Studies of transgenic mice overexpressing canonical WNT10B demonstrated that activation of this pathway is sufficient to induce fibrosis in vivo (Akhmetshina et al., 2012). Mechanistically, it has been demonstrated that the Wnt/β-catenin pathway regulates the expression of several ECM genes in fibroblasts from different organs (Akhmetshina et al., 2012; Hamburg-Shields et al., 2015; Xiang et al., 2017). Altogether, these studies suggest that Wnt/β-catenin signaling in stromal cells is required for pathogenic ECM gene expression and collagen deposition during fibrogenesis.

Transforming growth factor beta (TGF-β) regulates the differentiation program of a variety of cell types (for review see David and Massagué, 2018; Derynck and Budi, 2019), including MSCs. Mechanistically, TGF-β signaling stimulates proliferation of PDGFRα⁺ cells, ECM production and myofibroblast differentiation (Contreras et al., 2019b; Lemos et al., 2015; Uezumi et al., 2011). Three TGF-β isoforms – TGF-β1, TGF-β2, and TGF-β3 – are expressed in mammals (Massagué, 1998). These isoforms and their mediated signaling pathways are exacerbated in Duchenne muscular dystrophy (DMD) patients (Bernasconi et al., 1999; Smith and Barton, 2018), skeletal muscle of mdx mice (Gosselin et al., 2004; Lemos et al., 2015) and during fibrogenesis in several organs (Kim et al., 2018). Thus, it is recognized that dysregulated TGF-β activity is a driver for reduced muscle regeneration, impaired tissue function and fibrosis (Juban et al., 2018; Mann et al., 2011; Pessina et al., 2015; Vidal et al., 2008). Interestingly, the inhibition of TGF-β improves the pathophysiology of MD (Accornero et al., 2014; Acuña et al., 2014; Ceco and McNally, 2013; Cohn et al., 2007; Danna et al., 2014).

Here, we show that TCF/LEF members are differentially expressed in tissue-resident PDGFRα⁺ cells and demonstrate that Tcf7l2 is a novel TGF-β target gene. The UPS participates in both the modulation of TCF7L2 proteostasis and TGF-β-mediated downregulation of TCF7L2. We also found that TGF-β requires the activity of histone deacetylases (HDACs) to repress Tcf7l2 gene expression and to alter Wnt signaling. Finally, these effects were not observed in myogenic cells, suggesting that this crosstalk is specific to fibroblasts and MSCs.

We first determined the relative expression of the Wnt TCF/LEF members in the MSC cell line C3H/10T1/2 and mouse embryonic fibroblasts (MEFs) (Fig. 1A). Tcf7l1 and Tcf7l2 were the two most highly expressed members of this family, whereas Lef1 and Tcf7 were almost not expressed in MSCs or MEFs (Fig. 1A). Then, we evaluated the expression of these transcription factors in ex vivo FACS-isolated skeletal muscle PDGFRα^H2BEGFP+ FAPs, which express the H2B–eGFP fusion gene from the Pdgfra locus (Fig. 1B; Fig. S1A). Our results using muscle PDGFRα⁺ cells further corroborated the observations described above (Fig. 1C). We further corroborated these results using RNA-seq data of mononuclear muscle fractions (Scott et al., 2019). We focused on a Lin⁻ (CD31⁻CD45⁻Ter119⁻) LY6A/Sca-1⁺ population and observed that Tcf7l2 and Tcf7l1 transcripts were enriched in the Lin⁻Sca-1⁺ fraction (Fig. 1D,E). Thus, these results suggest that among the Wnt TCF/LEF members Tcf7l1 and Tcf7l2, but not Lef1 or Tcf7, are highly expressed in PDGFRα⁺ cells. We next evaluated TCF7L2 protein expression in skeletal muscle tissue and MSCs. The protein products detected by western blotting in most tissues and cells are ∼78 and ∼58 kDa in size (Tang et al., 2008; Weise et al., 2010). We found that these TCF7L2 protein isoforms are present in limb muscles and MSCs, representing the extended (E) and medium or short (M/S) isoforms (Fig. 1F; Fig. S1B) (Jin, 2016). We also found that the TCF7L2 TF is predominantly expressed in the nucleus of muscle-resident interstitial cells of the undamaged diaphragm muscle, isolated EGFP⁺ FAPs from PDGFRα^H2BEGFP mice and mdx;PDGFRα^H2BEGFP dystrophic mice, and MEFs (Fig. 1G; Fig. S1B–E). The tissue-specific expression of the TCF/LEF members and their relative expression in MSCs was further corroborated by additional single-cell RNA-seq data (Fig. S1F,G) (The Tabula Muris Consortium et al., 2018). Next, we evaluated the expression of TCF7L2 in transiently activated PDGFRα⁺ FAPs and confirmed that the TCF7L2 TF is present in these cells 3 d after notexin injury (Fig. 1H,I). Stromal PDGFRα⁺ fibroblasts are also found in the heart, playing supportive roles in cardiac development and repair (Chong et al., 2011; Farbehi et al., 2019; Fu et al., 2018; Furtado et al., 2016; Soliman et al., 2020; Tallquist and Molkentin, 2017). Similar to what we found in MSC cell lines and muscle PDGFRα⁺ FAPs, Tcf7l1 and Tcf7l2 were the two highest expressed members of this family in cardiac PDGFRα⁺ fibroblasts, whereas Lef1 and Tcf7 have very low expression levels (Fig. 1J). We further corroborated these results using RNA-seq data from cardiac fibroblasts (Fu et al., 2018). We also found that Tcf7l2 and Tcf7l1 transcripts were enriched in the heart PDGFRα⁺-Tcf21⁺ fibroblast population (Fig. 1K). Similar to muscle, both E and M/S TCF7L2 protein isoforms were also found in cardiac tissue and PDGFRα⁺ fibroblasts (Fig. 1L,M). Altogether, these results establish that specific TCF/LEF transcription factors, namely TCF7L2 and TCF7L1, are highly expressed in PDGFRα⁺ MSCs and fibroblasts as well as in total skeletal muscle and cardiac tissue.

We, and other investigators, have previously demonstrated that sub-populations of stromal MSCs co-express both TCF7L2 and PDGFRα during skeletal muscle development and also in adulthood (Contreras et al., 2016, 2019a,b; Murphy et al., 2011; Vallecillo-García et al., 2017). As a consequence of the stromal expansion caused by injury, both TCF7L2 and PDGFRα protein levels are increased in dystrophic tissue, denervated muscle and after chronic barium chloride-induced damage (Contreras et al., 2016). Interestingly, the number of TCF7L2⁺ cells and TCF7L2 bulk protein levels correlate with the extension of damage, fibrosis and TGF-β levels in muscles of dystrophic mdx mice (Contreras et al., 2016, 2019a). Additionally, the expansion of TCF7L2⁺ MSCs in human regenerating muscle closely associates with regenerating myofibers in vivo (Mackey et al., 2017). Thus, to study whether the expression of TCF7L2 changes during regeneration and repair, we used two previously described muscle injury models (Uezumi et al., 2010; Kopinke et al., 2017). First, we aimed to study whole tissue TCF7L2 protein levels during skeletal muscle regeneration following glycerol-induced acute damage, a model that leads to transient expansion of PDGFRα⁺ cells (Contreras et al., 2019c; Kopinke et al., 2017). As expected, glycerol acute damage caused an increase in total TCF7L2 protein levels, which transiently peaked at day 3 following glycerol intramuscular injection (Fig. 2A,B). PDGFRα bulk protein levels increased with similar kinetics of expression as TCF7L2 after acute injury (Fig. 2A,B). We also found that β-catenin, the transcriptional partner of TCF7L2, increased after glycerol damage (Fig. 2A,B). Myosin heavy chain expression (specifically MyHC-2X, also known as MYH1), a marker of regenerating myofibers, peaked at day 7 after damage (Fig. 2A,B). As mentioned above, we previously described that the proportion of TCF7L2⁺ stromal cells and TCF7L2 bulk expression increases in the dystrophic diaphragm, and we observed similar differences here (Fig. 2C,D) (Contreras et al., 2016). Intriguingly, at the single cell level we found a reduction in the expression levels of TCF7L2 in the expanded population of TCF7L2⁺ cells in the dystrophic diaphragm, where the proportion of TCF7L2⁺ cells expressing low levels of TCF7L2 was larger compared to levels in undamaged muscle. In the latter case, TCF7L2 expression was generally high (Fig. 2C). To further explore our observation, we used confocal microscopy and z-stack reconstructions of transverse muscle sections to analyze the cellular amount of TCF7L2 in TCF7L2⁺ cells in the dystrophic diaphragm (Fig. 2C,E; Fig. S1H). Cells expressing high levels of TCF7L2 were distributed throughout the muscle interstitium, consistent with the findings of previous studies of fibroblast distribution (Fig. 2C; Fig. S1H) (Contreras et al., 2016, 2019c; Mathew et al., 2011; Merrell et al., 2015; Murphy et al., 2011). On the other hand, the number of cells expressing medium or low levels of TCF7L2 expanded, and their expansion corresponded to increased phosphorylation of SMAD3, an indicator of activated TGF-β signaling in interstitial cells and muscle fibers, as well as in collagen type 1-enriched areas (Fig. 2F–H; Fig. S1H). We further corroborated our results using RNA-seq data from undamaged PDGFRα⁺ cardiac fibroblasts and after myocardial infarction (MI) (Fu et al., 2018). Genes encoding proteins involved in proliferation, TGF-β signaling and modification of the ECM, as well as matrisome components were most highly induced at the initial stages (days 3 and 7) but then they were downregulated 2 and 4 weeks after MI (Fig. S2). However, we observed that the expression of Tcf7l2 and Tcf7l1 was highest in quiescent fibroblasts and became downregulated upon fibroblast activation, while expression of Lef1 followed the opposite trend (Fig. S2). Pdgfra and Tcf21 followed a similar kinetic pattern to that of Tcf7l2 and Tcf7l1. Altogether, these results suggest that the magnitude of expansion of cells expressing medium and low levels of TCF7L2 depends on the extent of tissue inflammation, TGF-β levels, damage and fibrosis, and hence they were abundant in the inflammatory dystrophic model and after MI.

Chronic upregulation of TGF-β is found in several models of organ damage, where it is known to regulate the severity of tissue fibrosis. Activated extracellular TGF-β not only promotes MSC survival and proliferation but also primes these progenitor cells to become myofibroblasts (Contreras et al., 2019b,c; Cho et al., 2018; Kim et al., 2018; Lemos et al., 2015). TCF7L2⁺ cells expand near CD68⁺ macrophages in dystrophic muscles (one of the major cell sources of TGF-β in DMD) (Contreras et al., 2016; Juban et al., 2018; Tidball and Villalta, 2010). Our previous results suggest that the cell-specific expression of TCF7L2 negatively correlates with damage and inflammation in dystrophic muscle. Hence, damage-associated signaling pathways might regulate TCF7L2 expression and function endogenously during tissue repair. Therefore, we aimed to investigate the role of TGF-β signaling on the expression of TCF7L2 in fibroblast-related cell types and tissue-resident PDGFRα⁺ cells. We first used the multipotent mesenchymal progenitor cell line C3H/10T1/2 as an in vitro model of MSCs, because these cells have been extensively used to study mesenchymal biology (Braun et al., 1989; Contreras et al., 2019c; Reznikoff et al., 1973; Riquelme-Guzmán et al., 2018; Singh et al., 2003). TGF-β1 treatment diminished TCF7L2 protein expression in C3H/10T1/2 MSCs in a concentration-dependent manner, with effects already noticeable at the pathophysiological concentration of 0.5 ng/ml (Fig. 3A,B; Fig. S3A,B). We observed that C3H/10T1/2 cells respond to TGF-β1, in a concentration-dependent manner, by increasing the expression of ECM-related proteins [fibronectin, β1 integrin, CCN2 (also known as CTGF)] and αSMA (ACTA2, a myofibroblast marker), as well as by reducing the expression of PDGFRα (Fig. 3A,B) (Contreras et al., 2019c). Moreover, TGF-β1-mediated reduction of TCF7L2 expression began 8 h after stimulation, reaching a maximum at 24 and 48 h. At this point the expression of TCF7L2 had decreased by ∼85% (Fig. 3C,D). Similar effects were also seen in the embryonic fibroblast NIH-3T3 cell line (Fig. 3E,F; Fig. S3A,B). Taken together, these results suggest that TGF-β1 downregulates TCF7L2 protein expression in a concentration- and time-dependent manner in MSCs and fibroblasts. Consistently with the abovementioned results, we found that TGF-β1 treatment reduced the expression of TCF7L2 by ∼50% in FACS-isolated muscle PDGFRα⁺ FAPs (Fig. 3G,H). As expected, TGF-β1 treatment increased the expression of the ECM protein fibronectin (Fig. 3G) (Uezumi et al., 2010; Contreras et al., 2019c). Taken together, our results indicate that, concurrently with the induction of fibroblast activation and differentiation, TGF-β1 inhibits the expression of TCF7L2 in skeletal muscle PDGFRα⁺ FAPs, and in two different mesenchymal cell lines, C3H/10T1/2 and NIH-3T3 fibroblasts.

Not only were TCF7L2 protein levels decreased in response to TGF-β1, but the relative levels of Tcf7l2 mRNA were also diminished 2 h, 8 h and 24 h after TGF-β1 stimulation (Fig. 3I). Tcf7l2 gene expression was also sensitive to the inhibition of mRNA synthesis with actinomycin D treatment, which suggests active Tcf7l2 gene transcription and TCF7L2 translation in MSCs (Fig. 3J; Fig. S3C–E). The expression of PDGFRα was also very sensitive to the inhibition of mRNA synthesis with actinomycin D (Fig. 3J; Fig. S3C). Next, we evaluated whether TGF-β2 and TGF-β3 cytokines also impair TCF7L2 expression in MSCs. Similar to TGF-β1, both TGF-β2 and TGF-β3 strongly reduced TCF7L2 protein levels in MSCs (Fig. S3F,G). Thus, the three TGF-β ligands decrease the expression of TCF7L2 in PDGFRα⁺ cells. Next, to investigate whether the function of the Wnt-effector TF TCF7L2 was also altered by TGF-β, we determined, by quantitative PCR, the gene expression levels of several validated TCF7L2 target genes in response to TGF-β1. We found that the expression of Sox9, Axin2, and Tcf7l1 was repressed, while Nfatc1, Lef1, and Tcf7 expression was increased in response to TGF-β1 (Fig. S3H). We did not find changes in Ctnnb1 (which encodes β-catenin) or Ccnd1 mRNA levels (Fig. S3H). Furthermore, RNA-seq analyses of TGF-β-treated idiopathic pulmonary fibrosis (IPF) lung (Jones et al., 2019) and cardiac fibroblasts (Schafer et al., 2017) show that TGF-β1 alters the expression of several validated target genes of the Wnt/β-Catenin/TCF7L2 pathway. Dkk1, Dkk3, Lrp5, Lef1, Tcf7l1, Tgfbr2, Ctbp2, Sox9, Cxcl12 and Nfatc4 expression is reduced, whereas expression of the Fn1, Col1a1, Ctgf, Tgfb1, Foxo1, Wnt5a, Wnt5b, Wnt11, Wnt9a, Myc, and Oat TCF7L2-target genes is increased after TGF-β1 treatment (Fig. 3K,L). Altogether, these data suggest that TGF-β signaling reduces the expression of TCF7L2, and therefore alters its TF function in PDGFRα⁺ cells such as FAPs, MSCs and fibroblasts. Because TGF-β-induced myofibroblast differentiation of MSCs correlates with reduced TCF7L2 expression and impaired function, we investigated the impact that in vitro differentiation towards other lineages, such as adipocytes and osteocytes, might have on Tcf7l2 expression. We used the expression of Adipoq and Runx2 as markers of effective adipogenic and osteogenic commitment in MEFs (Fig. S4). Adipogenic differentiation did not alter Tcf7l2 mRNA expression (Chen et al., 2018), whereas increased expression of Tcf7l2 was found after osteogenic differentiation (Fig. S4A). Interestingly, expression of Tcf7l1, Lef1 and Tcf7 also varied in response to adipogenic or osteogenic differentiation (Fig. S4A). These data suggest that different molecular mechanisms are involved in the regulation of Tcf7l2 gene expression during differentiation to diverse cell lineages.

Having detected a drop in the total levels of TCF7L2 TF, we next determined the extent of decrease of TCF7L2 protein levels in the nuclei of TGF-β-treated cells. A subcellular fractionation method (see Materials and Methods) allowed us to determine that TCF7L2 was relatively abundant in the nucleus, although TCF7L2 protein was also detected in the cytoplasm (Fig. S4B,C). TGF-β1 stimulation downregulated the expression of TCF7L2 in the nuclear and cytoplasmic fractions (Fig. 4A; Fig. S4B,C). Next, we studied further the association between TGF-β-mediated fibroblast activation and reduced TCF7L2 levels. Using confocal microscopy, we confirmed that TCF7L2 nuclear expression was decreased in response to TGF-β1 and that TCF7L2 repression correlated with a TGF-β-induced myofibroblast phenotype in FAPs and NIH-3T3 fibroblasts (Fig. 4B,C; Fig. S4D). In addition, TGF-β signaling did not alter the subcellular distribution of TCF7L2 protein (Fig. 4D,E). Overall, these data indicate that TGF-β, in addition to inducing the differentiation of MSCs and fibroblasts into myofibroblasts, also represses the expression of the Wnt effector TCF7L2. Then, to investigate the role of TGF-β receptors in the regulation of TCF7L2 by TGF-β, we used the TGFBR1-specific small-molecule inhibitor SB525334 (Callahan et al., 2002). The binding of TGF-β ligand to the type II receptor TGFBR2 leads to recruitment and phosphorylation of the type I receptor TGFBR1 (also known as ALK-5). Treatment with the TGFBR1 inhibitor completely abolished the downregulation of TCF7L2 expression by TGF-β1 in wild-type PDGFRα⁺ FAPs and C3H/10T1/2 MSCs, without affecting the total levels of the TCF7L2-binding partner β-catenin (Fig. 4F,G). These results suggest that TGF-β-mediated TCF7L2 downregulation requires the activation of TGF-β receptors and the associated downstream signaling cascade.

The TGF-β family acts via SMAD and non-SMAD signaling pathways to regulate several cellular responses (Derynck and Budi, 2019; Zhang, 2017). To investigate the role of SMAD and non-SMAD cascades in the regulation of TCF7L2 by TGF-β, we used specific small-molecule inhibitors of SMAD3, p38 MAPKs, JNKs, and ERK1/2 (also known as MAPK3 and MAPK1, respectively) (Fig. 5A). None of these inhibitors were able to abolish the effect of TGF-β1 on the expression of TCF7L2 (Fig. 5A,B). Although we found in a previous study that p38 MAPK participates in TGF-β-mediated downregulation of PDGFRα expression (Contreras et al., 2019c), we did not detect any significant effect of co-treatment with the p38 MAPK inhibitor SB203580 on TGF-β-induced TCF7L2 downregulation (Fig. 5A,B; Fig. S5A). Interestingly, we found that co-treatment with SIS3, a SMAD3 inhibitor, augmented the inhibitory effect of TGF-β on the expression of TCF7L2 (Fig. 5A,B; Fig. S5B). Overall, these results suggest that another, not yet identified, SMAD-independent pathway may participate in TGF-β-mediated downregulation of TCF7L2 expression or that rather than SMAD3 activation, SMAD2/SMAD4 cofactors could be participating in TGF-β-mediated TCF7L2 downregulation.

It has been suggested that TCF7L2 might be a non-histone target of HDACs. This was based on the observation that treatment with trichostatin A (TSA), a commonly used histone deacetylase inhibitor, reduces the protein expression of TCF7L2 by 50% in the human colon cancer cell type HCT116 (Götze et al., 2014). Since HDACs regulate transcription of many genes (Bolden et al., 2006; Greer et al., 2015; Seto and Yoshida, 2014), we first evaluated whether the TSA affects TCF7L2 protein levels in MSCs. Indeed, TSA treatment reduced TCF7L2 levels, although to a lesser extent than TGF-β (Fig. 5C,D). Based on recent findings that demonstrated that TGFβ-mediated fibroblast activation requires HDAC-mediated transcriptional repression (Jones et al., 2019), we examined whether inhibiting HDACs with TSA modified the TGF-β-mediated reduction of TCF7L2 expression. Therefore, cells were treated with TGF-β1 and/or TSA for 8 h and then TCF7L2 protein levels were evaluated by western blotting. Interestingly, whereas TGF-β1 treatment decreased TCF7L2 expression, TSA treatment partially blocked TGF-β-mediated repression of TCF7L2 expression (Fig. 5E,F). RNA-seq analysis from recent work on IPF lung fibroblasts helped us to corroborate our results (Fig. 5G) (Jones et al., 2019). The pan-HDAC inhibitor pracinostat attenuated the TGF-β-mediated repression of Tcf7l2 gene expression and induction of expression of genes encoding ECM-associated proteins (Fig. 5G). Furthermore, TGF-β also impairs TCF7L2-mediated target gene expression and Wnt signaling via HDACs (Fig. S6). Taken together, these data suggest that HDACs participate in the regulation of TGF-β-mediated Tcf7l2 repression.

To gain knowledge of the mechanisms involved in TGF-β-induced TCF7L2 repression, we performed a timecourse analysis of treatment with the protein synthesis inhibitor cycloheximide. Thus, we determined that the half-life of TCF7L2 is quite short, being ∼3.5 h in both proliferating C3H/10T1/2 MSCs and PDGFRα⁺ FAPs (T_1/2=3.5 h) (Fig. 6A–C). Then, to study the molecular mechanism governing TGF-β-mediated TCF7L2 downregulation, we evaluated the human TCF7L2 interactome using BioGRID datasets (Stark et al., 2006). Remarkably, several protein-quality-control-related proteins are interacting partners of TCF7L2, among them: RNF4, RNF43, RNF138, NLK, UHRF2, UBE2I, UBE2L6, USP4, UBR5 and XIAP (Fig. 6D). Because these TCF7L2 interacting proteins belong to or are related to the UPS, we performed in silico analyses and identified several top-ranked potential ubiquitination sites along the TCF7L2 protein sequence (Fig. 6E). Next, we used MG132, a potent proteasome inhibitor that reduces the degradation of ubiquitin-conjugated proteins (Nalepa et al., 2006), to evaluate the participation of the UPS in the downregulation of TCFL72 induced by TGF-β. Mechanistically, MG132 completely blocked TGF-β-mediated downregulation of TCF7L2 protein after 9 h of co-treatment in MSCs and FAPs (Fig. 6F,G; Fig. S7A,B). MG132 also increased TCF7L2 basal protein levels when administered alone, which suggests that an intrinsic ubiquitin-mediated protein degradation mechanism controls TCF7L2 steady-state levels or proteostasis (Fig. 6F,G; Fig. S7). Remarkably, separation on a 12% SDS–PAGE gel allowed us to identify an unknown higher molecular weight form (∼65–70 kDa) of TCF7L2, which was not present at steady-state but was sensitive to proteasome-mediated proteolysis inhibition by MG132 (Fig. S7B). Finally, because our BioGRID analysis suggested the interaction of TCF7L2 with deubiquitinating enzymes (e.g. USP4), we evaluated the amount of TCF7L2 protein in C3H/10T1/2 MSCs treated with the ubiquitin-specific protease 7 (USP7) inhibitor HBX 41108 (HBX) in proliferating conditions (Colland et al., 2009; de la Vega et al., 2020; Yuan et al., 2018). Treatment with only HBX diminished TCF7L2 protein levels in a time-dependent manner (Fig. 6H). We also found that USP7-mediated TCF7L2 degradation was sensitive to the inhibition of protein synthesis (Fig. 6H). Taken together, these results suggest the participation of the UPS and ubiquitin-specific proteases in the regulation of TCF7L2 steady-state proteostasis and TGF-β-mediated repression of TCF7L2 protein expression.

Myoblasts are known to respond to TGF-β signaling (Droguett et al., 2006; Massagué et al., 1986; Riquelme et al., 2001; Riquelme-Guzmán et al., 2018; Schabort et al., 2009). Therefore, we investigated whether TCF7L2 downregulation by TGF-β signaling was stromal cell-type-specific or whether it also occurred in myoblasts. C2C12 myoblasts express very low levels of TCF7L2 and β-catenin compared to C3H/10T1/2 MSCs (Fig. 7A,B; Fig. S8A,B). Similar to our findings using stromal cells, the TCF7L2 protein was expressed predominantly in the nuclei of myoblasts (Fig. S8A). However, TGF-β1 stimulation did not alter the expression of TCF7L2, both at the protein and mRNA levels, in C2C12 myoblasts at the different concentrations used for 24 or 48 h (Fig. 7C–E; Fig. S8C,D). However, myoblasts did respond to TGF-β by increasing the expression of the two ECM components fibronectin and the matricellular protein CCN2 (Fig. 7C,E; Fig. S8C), as previously reported (Riquelme-Guzmán et al., 2018). Finally, we did not find changes in TCF7L2 subcellular distribution in myoblasts after TGF-β treatment (Fig. 7F,G; Fig. S8E). Thus, although TGF-β induces myoblast activation it does not change the expression of the TCF7L2 TF, which is expressed at relatively low levels in these cells compared to expression levels in PDGFRα⁺ fibroblasts.

In this study, first we reported differential gene expression of the four Wnt TCF/LEF TFs in PDGFRα⁺ fibroblasts from skeletal muscle and cardiac tissue, MSC cell lines and MEFs. Second, we established that TCF7L2 is a novel TGF-β target gene. Therefore, in addition to the strong reduction in gene and protein expression of TCF7L2 by TGF-β signaling, the expression of several TCF7L2-target genes is also altered upon TGF-β stimulation (Fig. 8). Third, we showed, via analysis of the TCF7L2 BioGRID-based interaction network, in silico prediction of TC7L2 ubiquitination residues and proteasome inhibition, that TGF-β regulates TCF7L2 protein stability via the UPS (Fig. 8). We also found that two pan-HDAC inhibitors, TSA and pracinostat, counteracted TGF-β-mediated repression of TCF7L2 (Fig. 8). Finally, we observed that TGF-β-driven TCF7L2 downregulation is fibroblast-specific, as this effect did not occur in C2C12 myoblasts.

TCF/LEF TFs can act as activators or repressors of gene transcription, like other high-mobility group box-containing proteins, mostly depending on their binding cofactors, activators or repressors and in a tissue- and cell-specific fashion (Frietze et al., 2012; Jin, 2016; Lien and Fuchs, 2014; Tang et al., 2008; Weise et al., 2010). TCF7L2 emerges as an interesting TF to study because TCF7L2⁺ MSCs are increased in number in fibrotic muscles from mdx mice and DMD (Contreras et al., 2016; Pessina et al., 2015), and also in damaged and atrophied muscles of the symptomatic amyotrophic lateral sclerosis (ALS) transgenic mice hSOD^G93A (Gonzalez et al., 2017). TCF7L2⁺ cells peak at day 5 following acute injury and then return to their basal levels once the damage is resolved (Murphy et al., 2011). In addition, TCF7L2 is a known transcriptional regulator in several cancers (Clevers, 2006; Cosin-Roger et al., 2019; Jin, 2016; Ravindranath et al., 2008; Tang et al., 2008; van de Wetering et al., 2002) and, to date, is the gene most closely associated with risk of type 2 diabetes (Grant et al., 2006; Jin, 2016). Nevertheless, most knowledge of canonical Wnt signaling concerns β-catenin translocation to the nucleus or dysregulated activity of upstream β-catenin-stability regulators like GSK-3β or CSNK1, but the precise contribution of TCF/LEF members to fibrogenesis has not yet been addressed. Even though TCF7L2 has been described as a marker for muscle CT fibroblasts (Contreras et al., 2016; Kardon et al., 2003; Mathew et al., 2011; Merrell et al., 2015), increasing evidence suggests that these cells share many properties with muscle PDGFRα⁺ FAPs (Contreras et al., 2019b; Malecova et al., 2018; Wosczyna and Rando, 2018).

A recent report demonstrated that TCF7L2 protein but not mRNA levels are increased during adipocyte differentiation, thereby TCF7L2 plays a regulatory role in fibroblast–adipocyte fate decisions in vitro and in vivo (Chen et al., 2018). In agreement with Chen et al. (2018) we did not find that Tcf7l2 gene expression changes after adipogenic differentiation. However, the precise role of TCF7L2 in regulating the fate and lineage restriction of fibroblast still needs to be addressed. It is important to mention that our work may have some experimental limitations. For example, we have not explored the role of different TCF/LEF TFs on the fate of MSCs. Here, we also reported that Tcf7l1 (Tcf3) is a TCF/LEF gene with high expression in PDGFRα⁺ cells. Regulation of the expression of Tcf7l1 seems to resemble that of Tcf7l2 but its role in fibroblast biology still needs to be further addressed. Interestingly, Tcf7 and Lef1 were expressed at very low levels under resting conditions, but they were strongly upregulated by tissue damage and TGF-β treatment. This might be explained as a compensatory regulation of expression or as a TCF7L2-mediated response. Tcf7 and Lef1 are target genes of TCF7L2 (Frietze et al., 2012; Lien et al., 2014), which may explain why these two members are upregulated when TCF7L2 is downregulated by TGF-β. However, future studies are needed to address the questions raised above.

The UPS has emerged as an essential component in cell and molecular biology due to its role in regulation of cellular proteostasis in homeostasis and disease (Nalepa et al., 2006; Yuan et al., 2018). Here, we demonstrated the participation of both the UPS and the deubiquitinase USP7 in the regulation of TCF7L2 steady-state levels and stability in response to TGF-β. Although we did not address the ubiquitination status of TCF7L2, to our knowledge this is the first study suggesting potential cross-modulation between TGF-β and the Wnt pathway through deubiquitinating enzymes and the UPS. Post-translational modifications of TCF7L2 occur in the forms of phosphorylation (Ishitani et al., 2003), acetylation (Elfert et al., 2013) and SUMOylation (Yamamoto et al., 2003). Although recent reports demonstrated the ubiquitination of some TCF/LEF members (Han et al., 2017; Song et al., 2018; Yamada et al., 2006), the extent and importance of ubiquitin-mediated regulatory mechanisms for activity of these TFs are still elusive. TCF7L1 half-life was reported to be longer than 12 h, with the DNA-binding of TCF7L1 regulating its protein stability, and therefore, Wnt signaling (Shy et al., 2013). Because TGF-β signaling is a known chromatin-modifying factor (Massagué, 2012), an intriguing possibility is that TCF7L2 DNA-binding regulates its stability and that TGF-β-induced chromatin remodeling might negatively affect TCF7L2 stability. However, the complexity of different molecular events maintaining TCF7L2 protein or mRNA levels in PDGFRα⁺ cells remains unclear. A limitation of this study is that we did not investigate the role of TCF7L2 in TGF-β-mediated ECM remodeling and myofibroblast differentiation. Because TCF7L2 is known to regulate the expression of thousands of genes in a cell-type-specific manner (Frietze et al., 2012; Schuijers et al., 2014; Tang et al., 2008; Weise et al., 2010), we propose that TGF-β-induced TCF7L2 downregulation should have a massive and profound impact on the transcriptome of fibroblasts.

Recently, histone deacetylase inhibitors (HDACi) have emerged as potential compounds for use in pre-clinical and clinical studies to improve tissue regeneration and repair in DMD (Bettica et al., 2016; Minetti et al., 2006). Central to this idea is that PDGFRα⁺ cells are specifically targeted by HDACi in muscular dystrophies (Mozzetta et al., 2013; Saccone et al., 2014). HDACi reduce DMD progression by increasing muscle regeneration while inhibiting fibro-fatty differentiation of PDGFRα⁺ cells. Mechanistically, HDACi favor intercellular communication between PDGFRα⁺ FAPs and myogenic progenitors during regeneration and repair (Bettica et al., 2016; Mozzetta et al., 2013; Saccone et al., 2014). Here, we observed that two well-characterized, known pan-HDACi reduce TGF-β-mediated ECM gene expression and also block TGF-β-induced downregulation of Tcf7l2 expression. This corresponds with our results showing that HDACi modulate TGF-β-mediated changes in expression of Wnt/TCF7L2 downstream genes. Further studies should unravel the mechanism by which HDACi regulate the Wnt/TCF7L2-dependent gene network.

Canonical Wnt signaling activation through WNT1 upregulates collagen deposition in the ECM and promotes fibroblast differentiation into myofibroblasts (Akhmetshina et al., 2012). Similarly, TGF-β activates the canonical Wnt cascade by inducing nuclear accumulation of β-catenin and increasing the activity of TCF/LEF-responsive elements in reporter assays. Therefore, TGF-β mediates reduction in expression of the Wnt inhibitor DKK1 (Akhmetshina et al., 2012). This seems to be in agreement with our results showing that TGF-β reduces the expression of several TCF7L2 target genes, including Dkk1, Dkk3, Lrp5, Lef1, Tcf7l1, Tgfbr2, Ctbp2, Sox9, Cxcl12 and Nfatc4, in IPF lung and heart fibroblasts. Furthermore, pharmacological inhibition of Wnt/β-catenin signaling by the small molecule ICG-001, reduces proliferation and TGF-β1-induced myofibroblast differentiation of lung resident MSCs in vitro and triggers organ protection via attenuation of bleomycin-induced lung fibrosis in vivo (Cao et al., 2018). These studies, and others, suggest that the canonical Wnt cascade is required for TGF-β-mediated effects, and vice versa (Cosin-Roger et al., 2019; Działo et al., 2018; Girardi and Le Grand, 2018; Piersma et al., 2015). Here, we hypothesize that extracellular TGF-β release after injury primes PDGFRα⁺ progenitor cells into differentiating TCF7L2 non-expressing myofibroblasts, which are then refractory to the self-renewal Wnt signals. Central to this idea is the significance of restraining an exacerbated stromal response to control scar formation. Thus, Wnt–TCF7L2 and TGF-β signaling could interact in a cell-specific and complex arrangement during regeneration and repair.

Emerging studies suggest a functional crosstalk between the Wnt cascade and TGF-β signaling in modulating MSC activation and fate (Burgy and Königshoff, 2018; Cosin-Roger et al., 2019; Vallée et al., 2017; Hamburg-Shields et al., 2015). For example, the canonical WNT3A ligand upregulates TGF-β signaling via SMAD2 in a β-catenin-dependent mechanism, and therefore, promotes the differentiation of fibroblasts into myofibroblasts (Carthy et al., 2011). Moreover, Wnt/β-catenin/TCF7L2 pathway mediates the transcriptional regulation of the TGF-β-target gene, Tmepai (also known as Pmepa1; Nakano et al., 2010). TGF-β promotes the secretion of Wnt proteins, via transforming growth factor-beta-activated kinase 1 (also known as MAP3K7) activation, which in turn activates the canonical Wnt cascade (Blyszczuk et al., 2016). Therefore, this TGF-β-dependent Wnt secretion induces myofibroblast formation and myocardial fibrosis progression (Blyszczuk et al., 2016). Elevated canonical Wnt/β-catenin signaling is found in dystrophic mdx muscles and controls MuSC fate via crosstalk with TGF-β2 (Biressi et al., 2014). Also, TGF-β induces the production of the canonical ligand WNT3A, which in turn increases TGF-β secretion, establishing an amplification circuit between TGF-β and Wnt signaling pathways in cardiac fibroblasts (Seo et al., 2019). Also, it has been demonstrated that TGF-β3-induced epithelial–mesenchymal transformation proceeds via transcription complexes of SMAD2–SMAD4–LEF1 that directly inhibit E-cadherin (Cdh1) gene expression (Nawshad et al., 2007). Tian et al. (2013) reported that β-catenin is a co-factor of SMAD3 during TGF-β1-mediated epithelial–mesenchymal transformation. Therefore, these studies and ours suggest a novel interplay between TGF-β and the canonical Wnt pathway.

The study of the heterogeneity and plasticity of tissue-resident mesenchymal stromal populations emerges as an attractive field to understand the balance between regeneration and degenerative fibrosis (Riquelme-Guzmán and Contreras, 2020; Mahmoudi et al., 2019; Lemos and Duffield, 2018; Lynch and Watt, 2018). Different MSC populations and their lineages may have intrinsic properties that favor either permanent scar formation or regeneration via scar regression (Driskell et al., 2013; Malecova et al., 2018; Plikus et al., 2017; Rinkevich et al., 2015; Rognoni et al., 2018; Soliman et al., 2020; Furtado et al., 2016). Thus, investigating different sub-populations of fibroblasts, with particular niches and genetic programs, is important to understand how these cells and their progeny influence wound healing, tissue repair and regeneration. Dynamic downregulation of fibroblast markers following damage has also been identified in resident PDGFRα⁺ cardiac fibroblasts (Asli et al., 2019 preprint; Farbehi et al., 2019; Kanisicak et al., 2016; Tallquist and Molkentin, 2017; Soliman et al., 2020) and skeletal muscle (Contreras et al., 2019c; Malecova et al., 2018). We have recently reported that the in vivo and in vitro expression of PDGFRα is strongly downregulated by damage-associated TGF-β signaling in skeletal muscle and heart PDGFRα⁺ fibroblasts (Contreras et al., 2019c). Conversely, TGF-β mediates the differentiation of PDGFRα⁺ cells into myofibroblasts at the expense of their adipogenic differentiation (Contreras et al., 2019c; Uezumi et al., 2014a, 2011). In summary, the work of others and our results suggest that stromal stem cell and/or progenitor markers are often downregulated by TGF-β, probably after a complex array of niche signals during regeneration or disease, as cells change phenotypically towards an activated cell state. These mesenchymal progenitor and/or fibroblast downregulated markers include Sca-1, PDGFRα, TCF21 and HIC1 (Asli et al., 2019 preprint; Contreras, 2019d,c; Fu et al., 2018; Kanisicak et al., 2016; Scott et al., 2019; Soliman et al., 2020). With our work, we add TCF7L2 to this growing list.

Our understanding of the cellular and molecular determinants of fibrosis has advanced immensely (Ceco and McNally, 2013; Kim et al., 2018; Piersma et al., 2015; Smith and Barton, 2018). Nevertheless, the lack of successful antifibrotic therapy to date prompts us to continue the search for new potential candidates to fight against non-malignant proliferative disorders. Compelling evidence from this work raises the idea that TCF7L2 should be explored as a therapeutic target in fibrotic diseases, as the Wnt/β-catenin pathway emerges as a novel and attractive signaling cascade to target for the improvement of tissue function in myopathies, fibrosis-related disorders and aging.

Housing, husbandry and experimental protocols were conducted in strict accordance and with the formal approval of the Animal Ethics Committee of the Pontificia Universidad Católica de Chile (doctoral ID protocol: 160512005) and following institutional and national guidelines at the University of British Columbia, Canada. Mice were housed in standard cages under 12 h light–dark cycles and fed ad libitum with a standard chow diet. Five-month-old C57BL/10ScScJ male mice (hereafter referred to as wild type; stock #000476) and dystrophic C57BL/10ScSn-Dmd^mdx/J mice (stock #001801) male mice (both from Jackson Laboratories) were used in experiments for Figs 1E and 2, and Fig. S3H. Pdgfra^{tm11(EGFP)Sor} mice (herein referred to as PDGFRα^H2BEGFP mice) were purchased from Jackson Laboratories (stock #007669 B6.129S4-Pdgfra^{tm11(EGFP)Sor}/J; Hamilton et al., 2003). For FAP detection in mdx muscles, we crossed male C57Bl/10ScSn-Dmd^mdx mice with hemizygous female B6.129S4-Pdgfra^{tm11(EGFP)Sor}/J mice. We used the F1 male mdx;PDGFRα^H2BEGFP offspring (5- to 6-month-old), and the comparisons were performed among siblings. All surgeries were performed after the mice had been anesthetized with 2.5–3% of isoflurane gas in pure oxygen. The mice were euthanized with cervical dislocation at the ages indicated in each figure, and the tissues were immediately processed, either by direct freezing in liquid nitrogen for protein and RNA extraction or in 2-methylbutane cooled with liquid nitrogen for histological analysis as described below.

For acute glycerol injury, the tibialis anterior (TA) muscle from 2- to 3-month-old PDGFRα^H2BEGFP mice was injected with 50 μl 50% v/v glycerol. Tissue collection was performed as indicated in the corresponding figures after glycerol injections. Notexin muscle damage was induced by intramuscular injection of 0.15 μg notexin snake venom (Latoxan) into the TA muscle (Contreras et al., 2019c; Joe et al., 2010; Lemos et al., 2015). Non-injected muscles from the contralateral limb were used as controls. Muscles were isolated and collected for analysis at the indicated time points in the corresponding figures.

Tissue preparation for skeletal muscle and heart FAPs was performed as described previously (Contreras et al., 2019c; Soliman et al., 2020). One-step digestion of tissue for FAPs was performed mainly as described previously with some modifications (Judson et al., 2017; Lemos et al., 2015). All the steps were performed on ice unless otherwise specified. Briefly, skeletal muscle from both hindlimbs (limb FAPs) and diaphragm (diaphragm FAPs) was carefully dissected, washed with 1×PBS and cut into small pieces with scissors until homogeneous. Collagenase D (Roche Biochemicals) 1.5 U/ml and Dispase II (Roche Biochemicals) 2.4 U/ml, in 2.5 mM CaCl₂, was added to every two hindlimbs in a total volume of 3 ml per mouse, and the preparation was placed at 37°C for 45 min with rotation. Preparations were passed through a 70 μm, and then 40 μm cell strainer (Becton Dickinson), and washed with FACS buffer (PBS, 2% FBS, 2 mM EDTA pH 7.9). Resulting single cells were collected by centrifugation at 1000 g for 5–10 min. Cell preparations were incubated with primary antibodies for 20–30 min at 4°C in FACS buffer at ∼3×107 cells/ml. We used the following monoclonal primary antibodies: anti-CD31 (clone MEC13.3, Cat. no. 553372, 1:400, Becton Dickinson; clone 390, Cat. no. CL8930F-3, 1:500, Cedarlane Laboratories), anti-CD45 (clone 30-F11, Cat. no. 557659, 1:400, Becton Dickinson), anti-CD45.1 (clone A20, Cat. no. 553775, 1:400, Becton Dickinson), anti-CD45.2 (clone 104, Cat. no. 11-0454-85, 1:200, eBiosciences), anti-Sca-1 (1:2000–1:5000, clone D7, Cat. no. 25-5981-82, Invitrogen) and anti-α7 integrin (1:11,500; Clone R2F2, Cat. no. 67-0010-05, AbLab). For all antibodies, we performed a fluorescence minus one control by staining with appropriate isotype control antibodies (rat anti-IgG2a kappa, PE-Cyanine7, clone eBR2a, Cat. no. 25-4321-82, eBioscience, 1:400; mouse anti-IgG2a κ, FITC, clone G155-178, Cat. no. 553456, BD, 1:100; rat anti-IgG2b κ, APC, clone eB149/10H5, Cat. no. 17-4031-82, eBioscience, 1:100). To assess viability, cells were stained with propidium iodide (1 μg ml^–1) and Hoechst 33342 (2.5 μg ml^–1) and resuspended at ∼1×10⁶ cells ml^–1 immediately before sorting or analysis. The analysis was performed on an LSRII (Becton Dickinson) flow cytometer equipped with three lasers. Data were collected using FacsDIVA software (Becton Dickinson). Cell sorting was performed on a FACS Vantage SE (Becton Dickinson), BD Influx flow cytometer (Becton Dickinson), or FACS Aria (Becton Dickinson), all equipped with three lasers, using a 100 μm nozzle at 18 p.s.i. to minimize the effects of pressure on the cells. Sorting gates were strictly defined based on isotype control (fluorescence minus one) stains. All flow cytometry data were analyzed using FlowJo 10.5.3v.

The TGFBR1 inhibitor SB525334 (used at 5 μM; S8822, Sigma-Aldrich), p38 MAPK inhibitor SB203580 (used at 20 μM; 5633, Cell Signaling Technology), MEK1/2 inhibitor (used to subsequently inhibit ERK1/2 kinases) UO126 (used at 10 μM; 9903, Cell Signaling Technology), SMAD3 inhibitor SIS3 (used at 6 μM; 1009104-85-1, Merck Calbiochem), JNK activity inhibitor SB600125 (used at 20 μM; Cell Signaling Technology), trichostatin A (used at 10 μM; T8552, Sigma-Aldrich; kindly provided by Dr Martín Montecino, Instituto de Ciencias Biomédicas, Universidad Andrés Bello, Santiago, Chile) and the inhibitor of USP7 activity HBX 41108 (used at 10 μM; 4285; Tocris; kindly provided by Dr Hugo Olguín, Pontificia Universidad Católica de Chile, Santiago, Chile) were all diluted in DMSO. DMSO treatment alone was used as a control. Cycloheximide (C104450, Sigma-Aldrich) was diluted in ethanol and used at 30 μg/ml final concentration. All the inhibitors used were added at the same time and co-incubated with TGF-β1. Other reagents, unless indicated otherwise, were purchased from Sigma-Aldrich.

The murine mesenchymal stromal cell (MSC) cell line C3H/10T1/2, Clone 8, and the embryonic fibroblast cell line NIH-3T3 were obtained from the American Type Culture Collection (ATCC, VA) and grown at 37°C in 5% CO2 in growth medium (GM): high-glucose Dulbecco's modified Eagle's medium (DMEM; Invitrogen) with 10% fetal bovine serum (FBS; Hyclone) and supplemented with antibiotics (Gutiérrez et al., 2015). The murine C2C12 myoblast cell line (ATCC) was cultured at 37°C in 8% CO2 in growth medium (GM): DMEM high glucose (Invitrogen, CA) with 10% fetal bovine serum (FBS; Hyclone, UT) and supplemented with antibiotics. Cells were treated with recombinant hTGF-β1 (#580702, Biolegend, USA), recombinant hTGF-β2 (#583301, Biolegend, USA), recombinant hTGF-β3 (#501123524, eBioscience, CA) in DMEM supplemented with 2% (v/v) FBS and penicillin/streptomycin in a 5% CO2 atmosphere at the concentration and time indicated in the corresponding figure legend. Adipogenic or osteogenic differentiation of MEFs was induced for 28 d using a MesenCult Adipogenic Differentiation kit (Mouse) (STEMCELL Technologies, Canada) and MesenCult Osteogenic Stimulatory Kit (Mouse), respectively. Our cell cultures were periodically tested to ensure no mycoplasma contamination using PCR. Cell fractionation was performed mainly according to the principles of the rapid, efficient and practical (REAP) method for subcellular fractionation with a few modifications (Suzuki et al., 2010). Briefly, the initially PBS-scraped and pelleted cells were resuspended by pipetting and by vortexing for 5 s, using 900 µl of ‘cytoplasmic buffer’ [10 mM Tris-HCl pH 7.5, 3 mM MgCl₂, 100 mM NaCl₂, 1 mM EGTA, 0.25% (v/v) Nonidet P-40]. Next, 300 µl of the initial lysate was removed and kept as whole cell lysate (referred to here as Ho) (Fig. 4A). The remaining (∼600 μl) material was centrifuged twice for 10 s in 1.5 ml micro-centrifuge tubes and 300 μl of the supernatant was removed and kept as the cytosolic fraction. After the remaining supernatant was removed, the pellet (∼20 μl) was resuspended with 180 μl of 1×Laemmli sample buffer and designated as the nuclear fraction. 100 μl of 4×Laemmli sample buffer was added to both the whole cell lysate and the cytosolic fractions. Finally, samples were sonicated and 30 μl, 30 μl and 15 μl of whole cell lysate, cytoplasmic and nuclear fractions, respectively, were loaded and electrophoresed using SDS­–PAGE (see below).

PDGFRα⁺ FAPs were FACS-isolated from either wild-type or PDGFRα^H2BEGFP/+ mice and grown in high-glucose DMEM (Invitrogen), supplemented with 10% FBS, 1% sodium pyruvate and 2.5 ng/ml bFGF (Invitrogen) at a density of 15,000 cell/cm² in a 48-well plate or 24-well plate. Cells were isolated from undamaged muscles. For the TGF-β1 treatment experiment, after 72 h–96 h in culture and 70–80% confluence, FAPs were stimulated with 5 ng/ml TGF-β1 (Contreras et al., 2019c). Cells were then collected for further analyses.

Protein extracts from cells were obtained using RIPA 1× lysis buffer (9806, Cell Signaling, MA) plus protease and phosphatase inhibitors (#P8340, #P0044, Sigma-Aldrich). Whole-muscle and heart extracts were obtained by homogenization at 4°C in RIPA 1× lysis buffer plus protease and phosphatase inhibitors. The cells were sonicated for 10 s and centrifuged at 9000 g. Whole-tissue homogenates were clarified by centrifugation at 10,000 g for 5 min. Proteins were quantified with the Micro BCA assay kit, following the manufacturer's instructions (Pierce, IL). Protein extracts (30–60 μg) were subjected to SDS–PAGE in 9–10% (or 12% in Fig. S7B) polyacrylamide gels, transferred to PDVF membranes (Millipore, CA), and probed with primary antibodies: goat anti-PDGFRα (1:1000; AF1062, R&D Systems), rabbit anti-TCF4/TCF7L2 (C48H11, 1:1000; 2569, Cell Signaling), rabbit anti-c-Jun (60A8, 1:1000; 9165, Cell Signaling), rabbit anti-β-catenin (1:2000; 9562, Cell Signaling), rabbit anti-histone H3 (1:1000; 9715, Cell Signaling), mouse anti-alpha smooth muscle actin (αSMA; 1:2000; Cat. no. A5228, Sigma-Aldrich), goat anti-CCN2/CTGF (1:500; Cat. no. sc-14939, Santa Cruz Biotechnology), rabbit anti-integrin β1 (M-106, 1:1000; sc-8978, Santa Cruz Biotechnology), rabbit anti-fibronectin (1:2000; F3648, Sigma-Aldrich), mouse anti-myosin skeletal fast (1:1000; M4276, Sigma-Aldrich), rabbit anti-myogenin (1:500; sc-576, Santa Cruz Biotechnology), mouse anti-GAPDH (1:5000; MAB374, Millipore), mouse anti-α-tubulin (1:5000; T5168, Sigma-Aldrich), Then, primary antibodies were detected with a secondary antibody conjugated to horseradish peroxidase: mouse anti-goat IgG (1:5000; 31400), goat anti-rabbit IgG (1:5000; 31460) and goat anti-mouse IgG (1:5000; 31430), all from Pierce. All immunoreactions were visualized by enhanced chemiluminescence Super Signal West Dura (34075, Pierce) or Super Signal West Femto (34096, Pierce) using a ChemiDoc-It HR 410 imaging system (UVP). Western blot densitometry quantification was done using Fiji software (ImageJ version 2.0.0-rc/69/1.52n). Briefly, minimum brightness thresholds were increased to remove background signal. Remaining bands were bracketed, plot profiles generated, and area under histograms auto traced. Protein levels were normalized against the levels of the loading control. Ponceau S Red Staining Solution [0.1% (w/v) Ponceau S in 5% (v/v) acetic acid] was used to detect protein transfer.

Cell immunofluorescence was performed as previously described (Contreras et al., 2018). For tissue section immunofluorescence, flash-frozen muscles were sectioned at 7 μm, fixed for 15 min in 4% paraformaldehyde, and washed in PBS. Cells and tissue sections were blocked for 30–60 min in 1% BSA with 1% fish gelatin in PBS, incubated overnight at 4°C in primary antibody: rat anti-laminin-α2 (1:250, Cat. no. L0663, Sigma-Aldrich), rabbit anti-collagen type I (1:250, Cat. no. A34710, Abcam), rabbit anti phospho-SMAD3 antibody (1:100; Cat. no. 9520S, Cell Signaling), mouse anti-alpha smooth muscle actin (αSMA; 1:250, Cat. no. A5228, Sigma-Aldrich). Then, samples were washed in PBS, incubated for 1 h or at room temperature with a secondary antibody (Alexa-Fluor-568 donkey anti-rabbit IgG (H+L), Cat. no. A10042, Life Technologies; Alexa-Fluor-594 donkey anti-rabbit IgG (H+L), Cat. no. A21207, Life Technologies; Alexa-Fluor-488 goat anti-rabbit IgG (H+L), 1:500, Cat. no. A11008, Life Technologies; Alexa-Fluor-555 donkey anti-mouse IgG (H+L), Cat. no. A31570, Invitrogen; Alexa-Fluor-568 goat anti-rat IgG (H+L), Cat. no. A11077, Life Technologies; all used at 1:500) and washed in PBS. Hoechst 33342 stain (2 mg/ml) and wheat germ agglutinin (WGA) Alexa Fluor 594 conjugate (#W11262, Invitrogen) were added for 10 min in PBS before the slides were mounted, according to the supplier's instructions. Slides were then washed in PBS and mounted with fluorescent mounting medium (DAKO, USA). To stain F-actin, Alexa Fluor 568 Phalloidin was added to the cells according to supplier’s instructions for 10 min (#A12380, Thermo Fisher Scientific). Cells were imaged on a Nikon Eclipse C2 Si Confocal Spectral Microscope or Nikon Eclipse Ti Confocal Microscope using Nikon NIS-Elements AR software 4.00.00 (build 764) LO, 64 bit. Confocal images were acquired at the Unidad de Microscopía Avanzada (UMA), Pontificia Universidad Católica de Chile, using a Nikon Eclipse C2 Si confocal spectral microscope. Plan-Apochromat objectives were used (Nikon, VC 20× DIC N2 NA 0.75, 40× OIL DIC H NA 1.0 and VC 60× OIL DIC NA 1.4). Cytospin confocal microscopy was performed using a Nikon eclipse Ti Confocal Microscope with a C2 laser unit. Confocal microscopy images shown in Figs 1G, 2C and Fig. S1G were composed using maximum-intensity projection z-stack reconstructions (0.3 μm each stack) of 7-μm-thick transversal sections or cultured cells. Then, we automatically analyzed the intensity of fluorescence (amount) of TCF7L2 in TCF7L2⁺ cells using the analyze particles plugin, and manually counted cells using the cell counter plugin from Fiji software (ImageJ version 2.0.0-rc/69/1.52n, NIH, MD). Counts of 6–8 randomly chosen fields were averaged from four independent experiments.

Total RNA from cultured cells was isolated using TRIzol (Invitrogen, CA) according to the manufacturer's instructions. RNA integrity was corroborated as described previously (Contreras et al., 2018). A total of 2 μg RNA was reverse transcribed into cDNA using random primers and M-MLV reverse transcriptase (Invitrogen, CA). Reverse transcription qPCR (RT-qPCR) was performed in triplicate with the Eco Real-Time PCR System (Illumina, CA, USA), using primer sets (see Table S1) for detection of Tcf7, Lef1, Tcf7l1 (Tcf3), Tcf7l2 (Tcf4), Ccnd1 (encoding cyclin-D1), Sox9, Axin2, Nfatc1, Ctnnb1 (encoding β-catenin) and the housekeeping gene 18s (used as a reference gene). The ΔΔCt method was used for quantification, and mRNA levels were expressed relative to the mean level of the control condition in each case. We analyzed and validated each RT-qPCR expected gene product using a 2% agarose gel. Digital droplet PCR was performed as previously described (Soliman et al., 2020). Gene expression analysis was performed using Taqman Gene Expression assays (Applied Biosystems), on a 7900HT Real Time PCR system (Applied Biosystems). Sequence information for the primers contained in the Taqman assays is as follows. Taqman probes (Thermo Fisher Scientific): Tcf7l2 mouse (Mm00501505_m1), Tcf7l1 mouse (Mm01188711_m1), Lef1 mouse (Mm00550265_m1), Tcf7 mouse (Mm00493445_m1), Runx2 mouse (Hs01047973_m1), Adipoq (Hs00605917_m1) and the housekeeping gene Hprt mouse (Mm03024075_m1). Data were acquired and analyzed using SDS 2.0 and SDS RQ Manager software (Applied Biosystems).

The image in Fig. 6D was generated using BioGRID, based on the human TCF7L2 interactome (Stark et al., 2006). The data from figures and tables in the BioGRID webpage (https://thebiogrid.org/) can be searched and sorted. For post-translational modification detection and delineation of ubiquitination at a site-specific level we used the UbiSite webpage (http://csb.cse.yzu.edu.tw/UbiSite/) (Akimov et al., 2018), and we corroborated the findings using UbPred (http://www.ubpred.org) and BDM-PUB (http://bdmpub.biocuckoo.org/). The murine Tabula Muris open database was used to generate the figures shown in Fig. S1F,G (The Tabula Muris Consortium et al., 2018). Data were extracted and analyzed for total tissues, diaphragm, and limb muscles.

Mean±s.e.m. values, as well as the number of experiments performed, are indicated in each figure. All datasets were analyzed for normal distribution using the D’Agostino normality test. Statistical significance of the differences between the means was evaluated using the one-way analysis of variance (ANOVA) test followed by post-hoc Dunnett's multiple comparison test, and the significance level set at P<0.05. A two-tailed Student's t-test was performed when two conditions were compared. Differences were considered significant with P<0.05. Data were collected in Microsoft Excel, and statistical analyses were performed using Prism 8 software for MacOS (GraphPad).

We are grateful to the Unidad de Microscopía Avanzada (UMA) of Pontificia Universidad Católica de Chile for its support in image acquisition. We acknowledge the services provided by the UC CINBIOT Animal Facility funded by the PIA CONICYT ECM-07 Program for Associative Research, of the Chilean National Council for Science and Technology. We also acknowledge the animal unit staff and genotyping core facility at the Biomedical Research Centre (University of British Columbia), especially Mr Taka Murakami, Mrs Krista Ranta, and Mr Wei Yuan. We thank Mr Andy Johnson and Mr Justin Wong of UBC Flow Cytometry. We also acknowledge Martin Arostegui from Michael T. Underhill's Lab (Biomedical Research Centre, University of British Columbia) for helping with MEF culture. We acknowledge the laboratory of Dr Hugo Olguín (Pontificia Universidad Católica de Chile) for kindly providing the USP7 small-molecule inhibitor HBX 41108, and also the laboratory of Dr Martín Montecino (Instituto de Ciencias Biomédicas, Universidad Andrés Bello) for kindly providing trichostatin A. For administrative assistance, we thank Ms Vanessa Morales and Mrs Vittoria Canale. We finally acknowledge Mr Eduardo Ramirez and Ms Darling Vera for their technical support.

The peer review history is available online at https://jcs.biologists.org/lookup/doi/10.1242/jcs.242297.reviewer-comments.pdf