SMAD2 ubiquitination through PY motif regulates skeletal muscle mass and fibrotic degeneration

Abstract

Transforming growth factorβ (TGFβ) signaling regulates diverse aspects of vertebrate skeletal muscle tissue including differentiation, homeostasis, regeneration and pathogenic degeneration. Ubiquitination of SMAD2, an intracellular transducer of TGFβ signaling, is a well-studied negative feedback regulation of the signaling pathway in the field of cell biology, but it’s relevance in skeletal muscle tissue has been elusive. In this study, to elucidate the in vivo role of SMAD2 ubiquitination, we generated Smad2dPY mutant mice in which a 15 bp sequence encoding the PY motif of SMAD2 protein is deleted from Smad2 gene. By removing this motif, the SMAD2 protein escapes from protein-protein interaction with NEDD4 family E3 ligases and thus is devoid of ubiquitination-dependent negative regulation. Smad2dPY mice showed no obvious abnormality in development, growth or fertility, indicating that SMAD2 ubiquitination through PY motif is dispensable for these processes. The skeletal muscle of Smad2dPY mice demonstrated reduced weight and myofiber size reduction at 12 months old. SMAD2 protein level was increased in the skeletal muscle of Smad2dPY mice while SMAD2 ubiquitination was reduced. Primary myoblasts of Smad2dPY mice displayed higher TGFβ responsiveness and suppressed terminal differentiation, which may explain the reduced muscle mass. The TGFβ responsiveness of the interstitial fibroblast population was also increased. Fibrotic tissue remodeling triggered by cardiotoxin injection was exacerbated in Smad2dPY mice. Altogether, our study identified SMAD2 ubiquitination through PY motif as an important regulatory mechanism operating in skeletal muscle tissue to maintain the TGFβ signaling pathway at the desired level in homeostasis and tissue remodeling.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-37582-z.

Introduction

Vertebrate skeletal muscle is a dynamic tissue that changes its volume in response to various stimuli such as exercise or disuse, and the molecular and cellular mechanisms underlying the plasticity of skeletal muscle mass has been well studied¹. Among these molecules, transforming growth factorβ (TGFβ) signaling is known as a potent negative regulator of skeletal muscle mass. Mutant mice with mutations in its positive components show increased muscle volume², whereas mice lacking negative regulators of TGFβ signaling demonstrate muscle atrophy³. Muscle TGFβ expression increases with age or in chronic injury^4–6 and the signaling pathway component SMAD2 accumulates upon immobilization, subsequently inducing muscle atrophy or fibrosis⁷. At the cellular level, the inhibitory activity of TGFβ signaling has been reported in both muscle stem cell (MuSC) and myofibers; myofiber-specific Acvr1b and Tgfbr1 knockout mice show hypertrophy of type IIB muscle fibers⁸, and TGFβ treatment negatively regulates myoblasts cell fusion⁹. Interstitial PDGFRα + mesenchymal progenitor cells (or fibro-adipogenic progenitors, FAPs) are also reported to be regulated by TGFβ signaling for their proliferation, differentiation and ECM deposition^10–13. Therefore, maintaining TGFβ signaling at the desired level in skeletal muscle tissue is crucial for its homeostasis as well as preventing malfunction. However, how the TGFβ signaling pathway is regulated in skeletal muscle tissue is not fully understood.

TGFβ ligand stimulus is transduced through a type I and type II receptor complex which in turn phosphorylates the c-terminus of SMAD2 and SMAD3, the intracellular transducers of the signaling pathway¹⁴. Phosphorylated SMAD2/3 then form complex with SMAD4, translocate into the nucleus and regulate target gene expression^14,15. Phosphorylated SMAD2/3 are then either mono-ubiquitinated and dissociated from DNA, or poly-ubiquitinated and degraded to terminate the signaling activity^16,17. Among various E3 ubiquitin ligases that are reported to ubiquitinate SMAD2/3 (R-SMADS)^18,19, the C2-HECT-WW domain type, also known as NEDD4 family E3 ligases, ubiquitinate SMAD2/3 through the interaction of their WW domain with the P-P-X-Y motif (PY motif) of SMAD2/3 ^17,20–22. Deletion of PY motif inhibits the interaction between NEDD4 family E3 ligases and SMAD2/3 and thus SMAD2/3 ubiquitination^17,23–25.

The importance of SMAD2/3 ubiquitination as a regulatory mechanism for TGFβ signaling has been repeatedly shown in cultured cell^18,26 and in early development of lower vertebrates. In Xenopus embryo, modifying Smurf2 activity affected mesodermal and neural gene expression through TGFβ signaling²⁷, while loss of USP15, a deubiquitinating enzyme of SMAD2/3 ubiquitination, reduced TGFβ signaling-dependent mesoderm induction, showing the indispensable role of SMAD2/3 ubiquitination in Xenopus early development¹⁶. However, in vivo relevance SMAD2/3 ubiquitination in mammalian tissue remains elusive. In this study, we approached this question by generating mice lacking the PY motif which would prevent SMAD2 ubiquitination by NEDD4 family E3 ligases and thus promote TGFβ signaling in cells in which ubiquitination of SMAD2 through PY motif has a significant role. We found that Smad2dPY mice showed an age-dependent reduction in skeletal muscle mass. TGFβ treatment of myoblasts increased the amount of pSMAD2 in Smad2dPY mice compared to WT. These results indicated that ubiquitination of SMAD2 via the PY motif is important for skeletal muscle regulation.

Results

Generation of Smad2dPY mice

To gain insight into the in vivo role of SMAD2 ubiquitination, we decided to generate a mouse line with a small genetic deletion to the locus that encodes PY motif (P-P-P-G-Y) of SMAD2 (Fig. 1A). We took advantage of the double nicking approach²⁸ and designed a pair of gRNAs that partially overlap with the locus that encodes PY motif (Fig. 1A). The in vitro synthesized gRNA and hCas9D10A mRNA, together with an ssOligo DNA lacking the 15 base pair (bp) sequence corresponding to the PY motif, were microinjected into fertilized eggs and transferred to pseudo-pregnant females. Delivered pups were genotyped, and a single pup with the intended 15 bp deletion was obtained (Fig. 1B). The 15 bp deletion allele (hereafter Smad2^dPY allele) was transmitted to subsequent generations, thus establishing the Smad2^dPY allele. Mice with homozygous Smad2^dPY allele (Smad2^dPY/dPY, hereafter Smad2dPY mice) were born at a Mendelian ratio, were fertile, and showed no apparent overall abnormalities at 9 weeks of age (Fig. 1C). The body weight of Smad2dPY mice was not significantly different from that of wildtype (WT) at this age (Fig. 1D). These results indicate that the PY motif of SMAD2 is not essential for development, postnatal growth and fertility in mice.

Fig. 1.

Generation of Smad2dPY mice. (A) A schematic drawing of the deletion of the sequence encodes the PY motif from mouse Smad2 locus. Top panel: Smad2 locus in mouse Chr.18. Black and white boxes indicate coding and non-coding exons, respectively. Bottom panel: design of gRNAs, ssOligo DNA and exon6 of mouse Smad2 gene. Blue and red boxes indicate gRNAs in which protospacer adjacent motif (PAM) is colored in blue. The dashed line in ssOligo DNA indicates the lack of sequence that corresponds with the sequence encoding PY motif. (B) Sequence comparison of WT and dPY allele. Bold letters indicate the position of the PY motif. The dashed line indicates the deleted bases. Underlines indicate the position of gRNAs. (C) Gross appearance of WT and Smad2dPY female mice at 8 weeks old. (D) Body weight of WT and Smad2dPY male mice at 9 weeks old. Paired two- tailed Student’s t-tests was used for the statistical analyses (*p < 0.05). n.s., not significant.

Smad2dPY mice show reduced muscle growth

As we have been studying TGFβ signaling in skeletal muscle tissue^29,30, next we investigated the adult skeletal muscle of Smad2dPY mice. First, we dissected hindlimb muscles (Tibialis Anterior: TA, Extensor Digitorum Longus: EDL, Soleus: SOL, Quadriceps: QU) of 9-weeks-old male mice and their weights were compared. SOL muscle was significantly smaller in Smad2dPY mice compared to WT mice (Fig. 2C), while no significant differences were found between Smad2dPY and WT mice in other muscles (Fig. 2A, B, D). Histological observation of TA muscle of Smad2dPY mice was indistinguishable from that of WT mice (Fig. 2E), and the cross-sectional area (CSA) of myofibers also showed no significant difference (Fig. 2F). Next, we compared the hindlimb muscle weights of 12-month-old male mice. The TA, EDL, and QU muscles, but not SOL muscle were significantly smaller in Smad2dPY mice compared to WT mice (Fig. 2G–J). CSA of myofibers of Smad2dPY TA muscle was significantly smaller than that of WT (Fig. 2K, L). These results indicated that Smad2dPY mice exhibited reduced muscle volume compared to WT mice, which becomes more apparent with age. Quantitative RT-PCR analysis revealed that the expression level of most of TGFβ signaling components, E3 ligase and deubiquitinating enzyme of SMAD2, or atrogenes were not significantly changed between WT and Smad2dPY mice (Fig. 3A–E). We found Smad3 mRNA level was higher, and Mstn and Tgfbr2 mRNA levels were lower in Smad2dPY mice, which may have an indirect effect on TGFβ outcome. Currently we do not have a clear mechanism behind this mRNA level change. While the Smad2 mRNA was at the same level, SMAD2 protein level was higher in Smad2dPY mice (Fig. 3F). We also noted that in Smad2dPY mice, band of SMAD2 protein was slightly lower than that of WT, reflecting the PY motif deletion. In line with this result, ubiquitination of SMAD2 protein was higher in WT than Smad2dPY mice (Fig. 3G). Altogether, Smad2dPY mice demonstrated reduced muscle volume in which SMAD2 ubiquitination level is lower and SMAD2 protein level is higher.

Fig. 2.

Smad2dPY mice show reduced muscle weights. (A–D) The hindlimb muscle weights of (A) TA, (B) EDL, (C) SOL, (D) QU from 9-weeks-old WT or Smad2dPY male mice (WT: n = 6, Smad2dPY: n = 6). (E) HE staining and Laminin immunofluorescent staining of TA muscle cross-section in 9-weeks-old WT or Smad2dPYmale mice. Nuclei were stained with DAPI. Scale bar: 50 μm. (F) CSA frequency distribution of TA muscles in 9-weeks-old WT (light blue) or Smad2dPY (orange) mice. (G–J) The hindlimb muscle weights of (G) TA, (H) EDL, (I) SOL, (J) QU from 12-month-old WT or Smad2dPY male mice (EDL of WT: n = 9, EDL of Smad2dPY: n = 8, TA, SOL and QU of WT: n = 10, TA, SOL and QU of Smad2dPY: n = 9). (K) HE staining and Laminin immunofluorescent staining of TA muscle cross-section in 12-month-old WT or Smad2dPY male mice. Nuclei were stained with DAPI. Scale bar: 50 μm. (L) CSA frequency distribution of TA muscles in 12-month-old WT (light blue) or Smad2dPY (orange) mice. For the muscle weight, data are shown as box plots and two-tailed Student’s t-test was used for the statistical analyses. For the CSA analysis, Mann–Whitney nonparametric test was used for statistical analyses. *p < 0.05, **p < 0.01, n.s., not significant.

Fig. 3.

SMAD2 protein, but not mRNA was increased in Smad2dPY muscle. (A–E) Quantitative RT-PCR of (A) Smads, (B) TGFβ receptors, (C) TGFβ ligands, (D) SMAD2 ubiquitination/deubiquitination enzymes, and (E) Atrogenes in TA muscle of wildtype (WT) and Smad2dPY mice. The names of each gene are indicated at the bottom of the panels. Data are shown as the mean ± SD. Two-tailed Student’s t-test was used for the statistical analyses. *p < 0.05, n.s., not significant. (F) Western blotting analysis of TA muscle of wildtype (WT) and Smad2dPY (dPY) mice. Asterisks (*) indicate SMAD3 protein bands. GAPDH serves as an internal control. Numbers indicate molecular weight (kDa). (G) Immunoprecipitation and western blotting of TA muscle of wildtype (WT) and Smad2dPY (dPY) mice. IgG used for precipitation are shown at the top. Black bracket indicates Ubiquitinated SMAD2 signals. White arrowheads indicate SMAD2 protein band and asterisks indicate IgG bands. Numbers indicate molecular weight (kDa).

Myoblasts response to TGFβ is promoted in Smad2dPY myoblasts

Next, to confirm whether the increased SMAD2 protein promotes the signaling pathway in myogenic cells of Smad2dPY mice, we isolated primary myoblasts from skeletal muscle tissue. Myoblasts were obtained from 8 to10-weeks-old GA muscles by a modified pre-plating method (Fig. 4A, see details in Materials and Methods). Collagenase-digested GA muscles were directly seeded onto Geltrex-coated dishes and fibroblasts were excluded by pre-plating with collagen-coated dishes at each passage to enrich myoblasts (Fig. 4A). To evaluate the degree of myoblast enrichment, the percentage of myogenic cells (i.e. Pax7 and/or MyoD+ cells) were compared before and after four-rounds of pre-plating (Fig. 4B). As a result, the Pax7/MyoD+ fraction increased from 61.47 ± 9.32% to 93.88 ± 4.81%, while few PDGFRβ + or NG2 + mesenchymal progenitor cells, fibroblasts, or mesangioblast were observed after the pre-plating (Fig. 4B, C). From these results, we considered the cells after four pre-platings/passages as the myoblast population. Myoblasts isolated from Smad2dPY mice showed a similar level of Ki67 staining to WT, indicating that the cell proliferation is not affected (Fig. 4D). To investigate the differentiation potential of myoblasts isolated from Smad2dPY mice, myoblasts were induced to differentiate for 3 days, and the fusion index was examined. As a result, myoblasts from Smad2dPY mice showed a lower fusion index than those of WT mice (Fig. 4E, F). Consistently, the ratio of mono-nucleated myocytes was higher in Smad2dPY samples, while myotubes with ≥ 7 nuclei were fewer (Fig. 4G). These results implied that myotube differentiation ability is suppressed in the myoblasts of Smad2dPY mice. Next, we examined the TGFβ responsiveness of these myoblasts. In response to the ligand treatment, the expression of TGFβ target genes, such as Smad7 and PAI1, were higher in the myoblasts from Smad2dPY mice than in those from WT mice (Fig. 4K–M). These results indicated that the TGFβ responsiveness of the myoblasts lacking SMAD2 PY motif were enhanced. Western blotting analysis showed that in myoblasts from Smad2dPY mice, the amount of phosphorylated SMAD2 (pSMAD2) upon TGFβ ligand treatment was higher than in WT mice, while the total amount of SMAD2 was also slightly increased (Fig. 4H, arrowhead). Moreover, a pulse-chase experiment showed that pSMAD2 protein was more stable in Smad2dPY myoblast than WT myoblast, indicating that PY motif dependent ubiquitination negatively regulates pSMAD2 in myoblasts (Fig. 4I, J). These results are in line with the previous studies showing phosphorylated SMADs are targeted by ubiquitination through PY motif¹⁷.

Fig. 4.

TGFβ response was promoted in Smad2dPY myoblasts. (A) Schematic diagram of modified-plating method (see details in Materials and Methods). (B) Representative image of Pax7/MyoD and PDGFRβ/NG2 immunofluorescent staining in cells before and after pre-plating. Nuclei were stained with DAPI. Scale bar: 50 μm (C) Quantification of the cells stained Pax7/MyoD. (D) Quantification of the cells stained Ki67. (E) Immunofluorescent staining of myosin heavy chain (MHC) of myotubes. Scale bar: 50 μm. (F) Fusion index of myotubes shown in (E). (G) Frequency of the number of nuclei per myotube. (H) Western blotting analysis of myoblasts treated with 5 µM SB505124, 0.1 ng/ml or 1 ng/ml TGFβ for 6 h. The orange arrowhead indicates phosphorylated SMAD2 (pSMAD2) bands. GAPDH serves as loading control. (I) Western blotting analysis of myoblasts with TGFβ pulse chase. SB: SB505124 treatment. The orange arrowhead indicates pSMAD2 bands. (J) Time course of relative pSMAD2/SMAD2 level. (K–M) Quantitative RT-PCR of (K) Smad7, (L) PAI1, (M) α-Sma in myoblast treated with 5 µM SB505124, 0.1 ng/ml TGFβ, 1 ng/ml TGFβ for 6 h. All the experiments were performed at least twice with similar results and a representative result is shown. Data are shown as the mean ± SD. Two-tailed Student’s t-tests was used for the statistical analyses. *p < 0.05, **p < 0.01, ***p < 0.001, n.s., not significant.

TGFβ response is promoted in fibroblasts of Smad2dPY mice

As the interstitial fibroblasts of skeletal muscle are also regulated by TGFβ signaling, next we also examined the TGFβ responsiveness of the muscle fibroblasts of Smad2dPY mice. To this end, we purified fibroblasts from hindlimb muscles by the conventional pre-plating method (Fig. 5A, see details in Materials and Methods). Hindlimb muscles of 8-weeks-old male WT or Smad2dPY mice were dissected, digested with collagenase, seeded onto a non-coated dish and cultured for 2 h. The supernatant was collected and seeded onto a collagen-coated dish for 1 day and the adherent cells on this plate were analyzed (Fig. 5A). These cells contained mesenchymal progenitor cells, fibroblasts, or mesangioblasts that were positive for PDGFRα, PDGFRβ and NG2, while very few Pax7 and MyoD+ cells were detected (Fig. 5B). Thus we considered these cells as the fibroblast population of the skeletal muscle. Upon TGFβ treatment, the fibroblasts from Smad2dPY mice demonstrated higher expression levels of Smad7, PAI1 and α-Sma than those from WT mice (Fig. 5C–E). These results showed that in Smad2dPY mice muscle tissue, TGFβ responsiveness is also promoted in the fibroblasts population. These results are consistent with the promoted TGFβ responsiveness in the primary fibroblast cells of Smurf2 knockout mice and in USP15 overexpressed cells^16,25.

Fig. 5.

TGFβ response was promoted in Smad2dPY fibroblasts. (A) Schematic diagram of pre-plating method (see details in Materials and Methods). (B) Representative image of Pax7/MyoD, PDGFRα/NG2 and PDGFRβ immunofluorescent staining of fibroblasts obtained by pre-plating. Scale bar: 50 μm. (C–E) Quantitative RT-PCR of (C) Smad7, (D) PAI1, (E) α-Sma in fibroblasts treated with 5 µM SB505124, 0.1 ng/ml TGFβ, 1 ng/ml TGFβ for 6 h. All the experiments were performed at least twice with similar results and a representative result is shown. Two-tailed Student’s t-tests was used for the statistical analyses. *p < 0.05, **p < 0.01, ***p < 0.001, n.s., not significant.

Fibrotic degeneration is promoted in Smad2dPY mice

Finally, we examined whether the inhibition of SMAD2 ubiquitination also affects the dynamic tissue remodeling process of skeletal muscle in which TGFβ/SMAD2 signal plays a role. To this end, we applied a cardiotoxin (CTX) induced muscle injury model. Eight to ten-weeks-old male mice of WT and Smad2dPY were injected with CTX into TA muscles. Myofibers with central nuclei were observed in TA muscles at 3 weeks post CTX-injection indicating that the regeneration process occurred in both genotypes (Fig. 6A). CSA of myofibers of Smad2dPY after CTX injection was significantly smaller than that of WT, indicating impaired regenerative process in Smad2dPY mice (Fig. 6B). Interestingly, the COL1+ area in Smad2dPY mice were significantly broader than that of WT only after the CTX treatment (Fig. 6C, D). Consistently, the number of α-SMA+ myofibroblast was higher than that of WT mice at 3 weeks post CTX-injection (Fig. 6E, F). These results indicate that in Smad2dPY mice, fibrotic remodeling of skeletal muscle tissue upon injury is exacerbated, due to the increased ECM depositing interstitial cells.

Fig. 6.

Fibrotic Tissue Remodeling is promoted in Smad2dPY mice. (A) HE staining of TA muscle cross-section at 21-days post CTX injection. Scale bar: 50 μm. (B) CSA frequency distribution of TA muscles in WT (light blue) or Smad2dPY (orange) mice. The upper panel shows the result of the control (CTRL) side and the lower panel shows the result of the CTX-injected side. (C) COL1 immunofluorescent staining of TA muscle cross-section at 21-days post CTX injection. Scale bar: 50 μm. (D) Quantification of COL1+ area of samples shown in (A). (E) αSmooth muscle actin (α-SMA) and CD31 immunofluorescent staining of TA muscle cross-section at 21-days post CTX injection. White and black arrows indicate CD31+/α-SMA+ and CD31-/α-SMA+ cells, respectively. Scale bar: 50 μm. (F) Quantification of CD31-/α-SMA+ cells of samples shown in (E). For the CSA analysis, Mann–Whitney nonparametric test was used for statistical analyses. ***p < 0.001. For the COL1+ area and α-SMA quantification, a one-way ANOVA followed by Tukey’s post hoc test was used for the statistical analyses. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, n.s., not significant.

Discussion

SMAD2 ubiquitination has been shown to be an important regulatory mechanism in TGFβ signal in cell biology, but its in vivo relevance in mammals remained elusive. While several mutant mice of TGFβ regulators showed defects in their early development^31–34, Smad2dPY mice were born and grew without clear abnormality. This indicates that PY motif-dependent SMAD2 ubiquitination is not an essential regulation for mouse development, at least under normal laboratory conditions. This is consistent with reports of knockout mice lacking the E3 ligase and deubiquitinating enzyme of SMAD2, namely Smurf2 and USP15, did not show abnormality in their early development^25,35,36. Further investigation is required to conclude the role of R-SMAD ubiquitination during development including the involvement of SMAD3 or other E3 ligases.

We found that removing PY motif resulted in reduced SMAD2 ubiquitination and increased SMAD2 protein level in skeletal muscle, indicating SMAD2 protein is regulated by PY motif dependent ubiquitination in this tissue. In Smad2dPY mice, skeletal muscle mass was reduced at the tissue level and TGFβ responsiveness was increased at cellular level. These results are consistent with the well-known role of TGFβ signaling as a negative regulator of skeletal muscle mass^2,3,8,37. Although the TGFβ responsiveness of myoblasts and fibroblasts from 8 to10-weeks-old mice are already higher than WT, the reduced muscle mass phenotype of Smad2dPY mice became more clear in 12-month-old mice than in 8-weeks-old mice. This might reflect the increasing TGFβ level in aged skeletal muscle tissue; GDF11/Myostatin expression or SMAD3 and SMAD4 protein levels have been reported to increase with age^5,6. CTX induced injury model revealed that also muscle tissue of Smad2dPY mice responded differently to the acute tissue injury probably reflecting higher TGFβ responsiveness. Considering that post-translational modification often functions as a feedback/buffering mechanism for extracellular signaling, our result implies that the ubiquitination of SMAD2 may curb the age-related increase of tissue TGFβ signaling and subsequently muscle mass loss. Further studies are necessary to examine this hypothesis in future studies. Evaluation of SMAD2 ubiquitination in in vivo skeletal muscle tissue would be also an important next issue, although currently it is technically difficult to detect the ubiquitination of specific endogenous protein. Novel methods such as biFC or Nanobit reporter could be applied for this purpose in the future studies. In addition, although most of the studies on SMAD2 PY motif focused on its interaction with the NEDD4 family of E3 ligases, other proteins such as SETD2 was reported to possibly interact with SMAD2 PY motif³⁸. The effect of depleting this interaction should be evaluated in future studies.

At the cellular level, Smad2dPY myoblasts demonstrated increased TGFβ responsiveness and a reduced fusion index. Ectopic TGFβ treatment or loss of TGFβ receptor has been reported to inhibit or promote the myoblast fusion^9,39 and loss of Smad4 in myoblasts has been shown to promote its terminal differentiation⁴⁰, which are consistent with our observation in Smad2dPY myoblasts. Recently, a positive role of SMAD2 on terminal differentiation of myoblast was reported. In this report, SMAD2 protein level positively correlated with Myogenin level and myoblast terminal differentiation, independent of ligand stimulus⁴¹. In Smad2dPY mice, both phosphorylated SMAD2 and total SMAD2 were increased, possibly resulting in positive and negative effects on myogenic process. The phenotypes we found in Smad2dPY mice were inhibition of muscle differentiation, implying that the effect of the promotion of ligand dependent phosphorylated Smad2 signaling was stronger than that of ligand independent Smad2 activity. In CTX induced injury and repair process, reduced myogenic activity resulted in impaired myofiber regeneration. In addition to myoblasts, fibroblasts of Smad2dPY mice also showed increased TGFβ responsiveness. Among the fibroblast population, PDGFRα+ mesenchymal progenitor cells (or FAPs) play a regulatory role in skeletal muscle. Because TGFβ signaling has been reported to regulate proliferation, myofibroblast differentiation, and ECM deposition by FAPs^4,10,12, increased TGFβ responsiveness in FAPs could be the cause of fibrotic phenotype in Smad2dPY mice. The increased number of α-SMA+ myofibroblasts in Smad2dPY muscle is in line with this hypothesis. In cardiac pressure overload model, conditional knockout of Smad7 in myofibroblast resulted in accentuated fibrosis in heart tissue associated with macrophage expansion⁴². A similar process might be involved in skeletal muscle of Smad2dPY mice. As the dPY mutation could affect all cell types in the tissue, it will be also important to examine whether the TGFβ responsiveness of myofibers as well as macrophages of Smad2dPY is increased and thus contribute for the phenotypes of Smad2dPY mice.

In summary, this study identified the in vivo role of ubiquitination of SMAD2 as a novel regulatory mechanism of skeletal muscle, both under homeostasis and tissue remodeling. Further studies on when and how this regulation operates will provide a more precise understanding of the regulation of the pleiotropic roles of TGFβ signaling in skeletal muscle tissue.

Materials and methods

Animals

The Smad2^dPY allele was generated using the double nicking method of the CRISPR-Cas9 system, as previously described^28,43. A pair of guide RNAs was designed in the vicinity of target sites with the desired spacing. Superovulated BDF1 females and BDF1 males were mated to obtain zygotes and hCas9D10A mRNA and guide RNAs targeting Smad2 exon 6 were microinjected. An allele with a 15 bp deletion that corresponds to the PY motif was selected from the obtained F0 mice as a founder of a Smad2^dPY line. The Smad2dPY mice used in this study were backcrossed for at least five generations to C57BL/6 purchased from Sankyo-lab service (Tokyo Japan). For the injury model, 100 µl of 10 µM cardiotoxin (Latoxan, Portes-lès-Valence, France) was injected into the TA muscle. The mice were housed under controlled environmental conditions with free access to water and food. Light was provided between 06:00 and 20:00. All animal procedures were performed in accordance with the Guidelines for Appropriate Implementation of Animal Experiments of the Science Council of Japan and were approved by the Animal Care and Use Committee of the National Research Institute for Child Health and Development (Permission Number: A2004-003-C09) and Meiji University (Permission Numbers: 17-0007, MUIACUC-20-113, MUIACUC2022-04). The authors confirm that animal experiment procedures complied with the ARRIVE 2.0 guidelines. Sequences of gRNA used are listed in Table S1.

Isolation of myoblasts and fibroblasts

For the conventional pre-plating method for fibroblast isolation, hindlimb muscles (TA, EDL, SOL, GA, QU, Plantaris, and Hamstrings) of 8-weeks-old WT or Smad2dPY mice were dissected and placed in a phosphate buffered saline (PBS) containing penicillin/streptomycin (P/S, Nacalai Tesque, Kyoto, Japan). All subsequent manipulations were performed in a clean bench. Any visible tendons, fat, and nerves were removed with tweezers and the muscles were minced with scissors. Minced tissues were transferred to 20 ml beakers contained 4 ml of 0.2% Collagenase Type II (Worthington, Lakewood, NJ, USA) /HBSS and digested in the incubator at 37 °C with a stirrer. After 30 min of digestion, the muscle tissue was passed through a 18G needles several times and digested for another 30 min (total 60 min). The digested muscles were passed through a 18G needle again and the homogenate was filtered using a 100 μm and 40 μm cell strainer. The flow-through was centrifuged at 760g for 5 min and the cells were resuspended in growth medium (GM, Dulbecco’s Modified Eagle Medium (DMEM, SIGMA-ALDRICH, MO, USA), 20% fetal bovine serum (FBS), 5 ng/ml bFGF (PEPROTECH, NJ, USA), 1000 U/ml LIF (Merk Millipore, MA, USA), 1% penicillin/streptomycin/glutamine (PSG, Nacalai Tesque)) and seeded onto uncoated plates. After 2 h, the supernatant was collected and centrifuged at 760g for 5 min. The collected cells were resuspended in GM and seeded onto type I collagen (Nitta Gelatin, Tokyo, Japan)-coated plates. Adhering cells on this plate were seeded into multi-well plates and used for assays as fibroblasts. For modified-plating method, GA muscle of 8 to 10-weeks-old WT or Smad2dPY mice were dissected and placed in PBS containing P/S. All subsequent manipulations were performed in a clean bench. Any visible tendons, fat, and nerves were removed with tweezers and the muscles were minced with scissors. GA were transferred to 5 ml tubes containing 4 ml 0.2% Collagenase Type II/DMEM and minced with scissors and micro shears. Minced tissues were passed through 18G needle several times and digested at 37 °C for 60 min. Digested tissues were centrifuged at 300g for 5 min and seeded onto Geltrex (Thermo Fisher Scientific, MA, USA)-coated dish directly. Cells were passaged every 2 days with 2 h of pre-plating steps on collagen coated dish. Cells after 4 times of pre-plating/passage were seeded onto multi-well plates and used for assays as myoblasts.

Cell culture and Immunofluorescence

Myoblasts were cultured with DMEM supplemented with 20% FBS, 10% horse serum (HS), PSG, 0.5% Chicken Embryonic Extract (CEE) and 10 ng/ml bFGF under 10% CO₂ condition. Fibroblasts were cultured with DMEM supplemented with 20% FBS, 5 ng/ml bFGF, 10³ U/ml LIF, L-Glutamine, and P/S under 5% CO₂ condition. For TGFβ treatment, fibroblasts were pre-treated with low serum medium (DMEM supplemented with 0.2% FBS and PSG) overnight and treated with 0.1 or 1 ng/ml TGFβ1 (PEPROTECH) or 5 µM SB505124 (SIGMA-ALDRICH) for indicated period. Immunofluorescence of cultured cell were performed as previously described¹⁶ primary antibodies used are α-MyoD1 (Novus, CO, USA, NBP1-54153, 1:100), α-Pax7 (DSHB, IOWA, USA, 1:100), α-NG2 (Millipore, AB5320, 1:100), α-PDGFRα (BioLegend, CA, USA, 135902, 1:100), α-PDGFRβ (Thermo Fisher Scientific, 14-1402-82, 1:100), Ki67 (BD Biosciences, CA, USA, 550609, 1:200) and α-MHC (Sigma-Aldrich, My-32, 1:500). Secondary antibodies used are: Donkey anti-mouse IgG-Alexa fluor 488, anti-rabbit IgG-Alexa fluor 488, and anti-rabbit IgG-Alexa fluor 488 (Jackson Immunoresearch, PA, USA). All secondary antibodies were diluted 1:500. The ratio of myogenic cell population was calculated by dividing the number of MyoD/Pax7 + nuclei by DAPI+ nuclei. The average of 7 randomly selected fields was calculated for each well. Fusion Index was calculated by (1) manually counting the number of two or more nuclei in MHC+ myotubes, and the total number of nuclei in the field of view (2) calculated (two or more nuclei in MHC+ myotubes)/(total number of nuclei). The average of 10 randomly selected fields was calculated for each well, and the average of three wells are shown as the fusion index of the sample.

Quantitative reverse transcription polymerase chain reaction (qRT-PCR)

qRT-PCR analysis was performed as described previously. Briefly, total RNA was isolated from tissues or cells using ISOGEN-II (NIPONGENE, Tokyo, Japan) and was reverse-transcribed with oligo(dT) primers and SuperScript II Reverse Transcriptase (Thermo Fisher Scientific). Quantitative PCR (qPCR) were performed using Power SYBR® Green Master Mix (Thermo Fisher Scientific) and Step-One system (Thermo Fisher Scientific). Primers used are listed in Table S1.

Western blotting

Western blotting was performed as described previously⁴⁴. Briefly, the tissues and cells were lysed with RIPA buffer (50 mM Tris-HCl, pH8.0, 150 mM Sodium Chloride, 0.5% Sodium Deoxycholate, 0.1% SDS, 1% NP-40) and lysis buffer (50 mM HEPES pH 7.8, 200 mM NaCl, 5 mM EDTA, 1% NP40, 5% Glycerol, 1 mM DTT, 20 mM N-ethylmaleimide), respectively, supplemented with protease inhibitor cocktail (cOmplete; SIGMA-Aldrich). Protein concentration was quantified using DC protein assay kit (#500-0111, Bio-Rad, CA, USA) and 10 μg protein/sample were loaded for SDS-PAGE. After SDS-PAGE, the samples were transferred to PVDF membranes (Immobilon-P; Merck Millipore) for immunoblotting. The primary antibodies used were α-Smad2/3 (BD, NJ, USA, #610843, 1:500), α-phosyphorylated Smad2 (Cell Signaling, MA, USA, #3108, 1:500), α-β-actin (MBL, Nagoya, Japan, M177-3, 1:2000), α-GAPDH (Merk Millipore, 1:1000), and α-Ubiquitin (Proteintech #80992-1-RR). The secondary antibodies used were α-rabbit IgG-Peroxidase (SIGMA-Aldrich, A0545, 1:2000) and α-mouse IgG-Peroxidase (SIGMA–Aldrich, A2304, 1:2000). For immunoprecipitation, α-Smad2 antibody (Proteintech #12570-1-AP) was used. Signal detection was performed with ECL Western Blotting Detection Reagents (GE Healthcare, IL, USA) or SuperSignal West Femto Maximum Sensitivity Substrate (Thermo Fisher Scientific). The proteins bands were quantified by densitometry using the ImageJ software.

Histology

The TA muscles were surgically removed, mounted on tragacanth-gum, and snap-frozen in liquid nitrogen-cooled isopentane. The TA muscles were sectioned at 10 μm and subjected for hematoxylin and eosin (HE) staining⁴⁵ and immunofluorescence. Anti-laminin antibody (Sigma-Aldrich, L9393, 1:500), anti-type I collagen (Southern Bio, 1310-01, 1:200), and anti-α-Smooth muscle actin (SIGMA, #A2547, 1:200) were used as primary antibodies. Anti-rabbit IgG-Alexa fluor 488 or anti- goat IgG-Alexa488 (Jackson Immunoresearch) were used as secondary antibody. Slides were coverslipped with Fluoroshield Mounting Medium with DAPI (ImmunoBioScience Corp., WA, USA). The cross-sectional area (CSA) of myofibers were measured from fluorescence images of anti-laminin-stained sections using ImageJ/Fiji software program and Trainable Weka Segmentation plugin⁴⁶. Briefly, immunofluorescence images were classified into laminin+ (membrane) and negative (cytosol of myofiber) domains, and the area of laminin negative regions were extracted by Analyze Particle tool. COL1 + area was measured from fluorescence images of α-COL1-stained sections using ImageJ/Fiji software. The average of three randomly chosen area was used as a value of each sample.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

K.S., Y.Y., and M.I. wrote the main manuscript text and K.S., Y.Y., S.Y., Y.K., A.Kubo, A.Kondow, and K.H. prepared Figs. 1, 2, 3, 4, 5 and 6. All authors reviewed the manuscript.

Funding

This research was supported by the Ministry of Education, Culture, Sports, Science and Technology (MEXT) KAKENHI (22H02636, 23K23899, and 24K22257) to M.I., the Naito Foundation to M.I., and the Cooperative Research Project Program of Joint Usage/Research Center at the Institute of Development, Aging and Cancer, Tohoku University to A.Kubo. and M.I.

Data availability

All data are contained within the manuscript.

Competing interests

The authors declare no competing interests.