Sca-1 is negatively regulated by TGF-β1 in myogenic cells

Abstract

Sca-1 (stem cell antigen-1) is a member of the Ly-6 family of proteins and regulates cell proliferation, differentiation, and self-renewal in multiple tissues. In skeletal muscle, Sca-1 inhibits both proliferation and differentiation of myogenic cells. Sca-1 expression is dynamically regulated during muscle regeneration, and mice lacking Sca-1 display increased fibrosis following muscle injury. Here, we show that Sca-1 expression is negatively regulated by TGF-β1 and that this inhibition is dependent on Smad3. We demonstrate that levels of TGF-β1 in skeletal muscle rapidly increase on injury and that the majority of this TGFβ1 is produced by infiltrating macrophages. Sca-1 is expressed in multiple cell types, and we demonstrate that TGF-β1 represses Sca-1 expression in T cells and other immune cell populations derived from the spleen, indicating that regulation by TGF-β1 is a general feature of Sca-1 expression in multiple cell types. Elucidation of the mechanisms by which Sca-1 expression is regulated may aid in the understanding of muscle homeostasis, potentially identifying novel therapeutic targets for muscle diseases.—Long, K., Montano, M., Pavlath, G. K. Sca-1 is negatively regulated by TGF-β1 in myogenic cells.

Efficient regeneration of skeletal muscle requires a balance between myofiber regeneration and growth of connective tissue (1). Loss of this balance results in pathological fibrosis, leading to impaired myofiber regeneration and prevention of complete muscle regeneration (2). Fibrosis following myofiber degeneration is a prominent feature of muscular dystrophies and is characterized by excessive production and decreased degradation of extracellular matrix (ECM) components (3, 4).

We recently identified Sca-1 (stem cell antigen 1) as a regulator of ECM remodeling during muscle regeneration (5). Sca-1 is a member of the Ly-6 family of small (12–15 kDa) GPI-linked proteins and is expressed in a wide variety of tissues (6). Originally identified by its up-regulation in activated lymphocytes, Sca-1 has since been implicated in progenitor cell self-renewal, cell proliferation, and cell differentiation (5, 7–12). Sca-1 is not expressed in myogenic cells present in uninjured tissue; however, we have recently shown that Sca-1 expression is up-regulated in a subset of myogenic cells 2 d following injury and that Sca-1 is no longer detectable in these cells 24 h later (5, 10, 13, 14). Sca-1^−/− mice display increased fibrosis following muscle injury, suggesting that this dynamic expression pattern is required for efficient muscle regeneration (5). Currently, the only well-characterized factors known to up-regulate Sca-1 expression are IFNα/β and IFNγ; the mechanisms by which Sca-1 expression is decreased are unknown (15). Therefore, a greater understanding of Sca-1 regulatory factors in muscle may further illuminate mechanisms of muscle homeostasis.

In this study, we identify transforming growth factor β (TGF-β) as a negative regulator of Sca-1 expression in myogenic cells. The TGF-β family of cytokines regulates a wide variety of cellular processes, including cell proliferation, migration, differentiation, adhesion, and apoptosis (16). The precise effects of TGF-β are highly dependent on cell context. In skeletal muscle, TGF-β is a potent inhibitor of myogenesis, preventing premature differentiation of myogenic precursors during growth and regeneration (17). Excessive TGF-β present in diseased muscle also contributes to fibrosis by stimulating the production of ECM components (18). As TGF-β has such profound effects on skeletal muscle, our results suggest that regulation of Sca-1 by TGF-β may play an important role in muscle growth and maintenance.

MATERIALS AND METHODS

Animals

C57BL/6 mice were purchased from Charles River Laboratories (Wilmington, MA, USA). Myf5-nLacZ mice were obtained from Shahragim Tajbakhsh (Pasteur Institute, Paris, France; ref. 19). Adult mice 8–12-wk old were used. All animals were handled in accordance with the institutional guidelines of Emory University and Boston Medical Center.

Antibodies

α-Sca-1 PE (1 μg/10⁶ cells), α-CD11b PE (0.2 μg/10⁶ cells), α-CD3 FITC (1 μg/10⁶ cells), α-CD44 PE (1 μg/10⁶ cells), α-Syn4 PE (1 μg/10⁶ cells), and appropriate isotype controls were purchased from BD Biosciences (San Jose, CA, USA). αTGF-β1,-β2,-β3 neutralizing antibody (clone 1D11) was purchased from R&D Systems (Minneapolis, MN, USA) and used at the indicated concentrations. αSmad3 (1:1000) antibody was purchased from Cell Signaling Technology (Danvers, MA, USA) and detected using an HRP-conjugated IgG (1:5000; Jackson ImmunoResearch, West Grove, PA, USA). α-BrdU (5-bromo-2-deoxyuridine; 1:500) was purchased from Accurate Chemical and Scientific Corporation (Westbury, NY, USA) and visualized using a FITC-conjugated donkey α-rat IgG (1:250; Jackson ImmunoResearch). α-MyoD (NCL-MyoD1; 1:20; NovoCastra, Newcastle on Tyne, UK) was detected using a FITC-conjugated goat α-mouse IgG (1:250; Jackson ImmunoResearch).

Primary myoblast culture

Primary myoblasts were isolated from mouse hind limb muscles, as described previously (20, 21). Cells were suspended in growth medium (GM; Ham's F-10, 20% FBS, 5 ng/ml bFGF, 100 U/ml penicillin G, and 100 μg/ml streptomycin) and grown on collagen-coated dishes in a humidified 5% CO₂ incubator at 37°C. All cultures were >95% myogenic cells, as assessed by MyoD immunostaining. Where indicated, primary myoblasts were treated for 24 h with TGF-β1 (Sigma, St. Louis, MO, USA), IL-4 (R&D Systems), IL-6 (Sigma), or TNF-α (Sigma) at the indicated concentrations.

Determination of TGF-β1 expression during muscle regeneration

Regeneration was induced in the gastrocnemius muscles of C57/B6 mice by injection of 1.2% BaCl₂ in PBS (20, 22). At the indicated times, uninjured and regenerating muscles were collected in cold RIPA-2 buffer (25 mM Tris, pH 7.6; 150 mM NaCl; 1% Nonidet P-40; 1% sodium deoxycholate; and 0.1% SDS) with Complete mini protease inhibitor tablets (Roche, San Francisco, CA, USA) added. Muscles were homogenized and clarified by centrifugation. Protein concentration was determined by BCA analysis (Pierce, Rockford, IL, USA). TGF-β1 levels were determined using a TGF-β1 Quantikine ELISA (R&D Systems) and normalized to protein content. Three samples for each time point were analyzed, and each sample was read in triplicate.

Isolation of CD11b⁺ cells from regenerating muscle

Regeneration was induced as described above. At 24 h after injury, the regenerating muscle was collected. The contralateral muscle was collected as an uninjured control and placed in TRIzol reagent (Invitrogen, Carlsbad, CA, USA) for isolation of RNA. Half of the regenerating muscle was also immediately placed in TRIzol reagent. The remainder of the muscle was digested as described previously (20, 21), and CD11b⁺ cells were isolated using CD11b microbeads and magnetic cell separation columns (Miltenyi Biotec, Auburn, CA. USA). Sorted cells were placed in TRIzol prior to RNA isolation. Purity of the sorts was assessed via flow cytometry using a PE-conjugated Cd11b antibody, and all sorts were 90–95% CD11b⁺.

Isolation of splenocytes

Spleens were removed and diced into 3 to 4 pieces. Splenocytes were isolated by gently pressing each spleen against a 70-μm filter into cold FACS buffer (PBS, 0.5% BSA, and 2 mM EDTA) using the rubber end of a 1-ml syringe plunger. This process was repeated until the splenic capsule became white. The collected cells were passed over a second 70-μm filter and pelleted by centrifugation, and red blood cells were lysed in buffer (0.2% Tris, pH 7.6, and 0.747% NH₄Cl) for 3 min. Following lysis, cells were again centrifuged and resuspended in RPMI supplemented with 10% FBS, 100 U/ml penicillin G, and 100 μg/ml streptomycin. Cells were placed in a humidified incubator with 5% CO₂. Where indicated, 5 ng/ml TGF-β1 was added for 24 h, and Sca-1 and CD3 expression was analyzed via flow cytometry.

RNA purification and real-time PCR

RNA was isolated from myoblasts using TRIzol reagent followed by RNA cleanup using the RNeasy Mini Kit with on-column DNase digestion (Qiagen, Valencia, CA, USA). Sca-1 gene expression was quantified using the iCycler iQ5 real-time detection system (Bio-Rad, Hercules, CA, USA). cDNA was generated by reverse transcription using 1 μg RNA. PCR reactions were performed with a 10-min denaturation step at 95°C followed by 40 cycles of 95°C for 15 s and 60°C for 60 s. SYBR Green fluorescence was measured after each extension cycle. Sca-1 expression was normalized to expression of a housekeeping gene, hypoxanthine guanine phosphoribosyl transferase I (HPRT), and fold change relative to control was determined. TGF-β1 gene expression was quantified using an ABI Prism 7000 real-time PCR system (Applied Biosystems, Foster City, CA, USA) and the Express One-Step SYBR Green qRT-PCR kit (Invitrogen), and expression was normalized to the housekeeping gene GAPDH (glyceraldehyde 3-phosphate dehydrogenase). RNA samples were included to ensure that no DNA contamination was present. Primers for Sca-1, TGF-β1, GAPDH, and HPRT were purchased from SA Biosciences (Frederick, MD, USA). All samples were analyzed in triplicate, and 3 independent replicates were performed for each condition.

Single myofiber isolation and culture

Single myofibers were isolated from gastrocnemius muscles as described previously (10). Briefly, the gastrocnemius was dissected and digested in DMEM containing 25 mM HEPES and 0.1% collagenase type I (Worthington, Lakewood, NJ, USA) with gentle agitation for 90 min. Single myofibers were extracted individually into clean plates, transferred to 15-ml conical tubes, and rinsed 3 times with medium to remove contaminating cells and debris. Myofibers were transferred to 100-mm dishes prior to plating. For MyoD immunostaining, myofibers were transferred to 24-well dishes coated with 10% growth factor reduced Matrigel (BD Biosciences). αTGF-β antibody or control IgG (0.5 μg/ml) was included in the medium at all steps of isolation and culture. At 24 h after plating, myofibers were fixed with 3.75% formaldehyde and immunostained for MyoD as described previously (10). For flow cytometry, myofibers were isolated from Myf5-nLacZ mice and plated 15–20/well in Matrigel-coated 6-well dishes. bFGF (12 ng/ml) was added to the medium to inhibit differentiation of myoblasts. Myofibers were cultured for 6 d, with 2 μg/ml α-TGF-β antibody added for the final 24 h. Myogenicity of myofiber cultures was determined as described previously (5); only cultures > 95% β-galactosidase⁺ cells were used. After plating, myofibers were spun at 1100 g to facilitate adherence to the Matrigel. Myofibers were incubated in a humidified incubator at 37°C, 5% CO₂.

Flow cytometry

To analyze Sca-1 expression by flow cytometry, cells were immunostained with a PE-conjugated antibody and analyzed on a FACSCalibur (Becton-Dickinson, Franklin Lakes, NJ, USA). For analysis of myofiber explants, cultures were trypsinized and passed over a 100-μm filter to remove myofibers. Muscle-derived macrophages and splenocytes were immunostained with the appropriate antibodies and analyzed on an LSRII (Becton-Dickinson). Ten thousand cells were analyzed for each sample (5000 for myofiber explant cultures), and propidium iodide was used to gate out dead cells. All analysis was performed using FlowJo v. 9.0.1 (TreeStar, Ashland, OR, USA).

Analysis of cell proliferation

Primary myoblasts were incubated for 12 h with 25 μM BrdU and 2 μg/ml α-TGF-β or control antibody. Cells were fixed with 3.75% paraformaldehyde and immunostained with an antibody to BrdU as described previously (23).

siRNA analysis

Stealth siRNA duplexes (Invitrogen) were used for all siRNA experiments. The sequence for Smad3 siRNA is 5′-AGGAUGAAGUGUGUGUAAAUCCUUA-3′. The medium negative control duplex (Invitrogen) was used as a control. Each siRNA duplex (80 μM) was transfected into primary myoblasts using Lipofectamine 2000 (Invitrogen). After 24 h, 5 ng/ml TGF-β1 or vehicle was added to cultures. After a further 24 h, cells were collected, and Sca-1 expression was analyzed. Knockdown was verified by Western blot analysis. Three independent replicates were performed. For rescue of RNA knockdown, primary myoblasts were infected with control (pBABE) or Smad3 retrovirus (LPCX Smad3) (24). Following selection with 0.75 μg/ml puromycin, cells were transfected with the appropriate siRNA duplex and analyzed as described above.

Western blot analysis

Cells were lysed in RIPA-2 buffer (50 mM Tris, pH 7.6; 150 mM NaCl; 1 mM EDTA; 1 mM EGTA; 1% Triton X-100; 1% sodium deoxycholate; and 0.1% SDS) with protease inhibitors added (Complete Mini protease inhibitor tablets; Roche). Lysate (20 μg) was loaded onto an 8% polyacrylamide gel. Immunoblots were performed using the indicated antibodies.

Statistics and image assembly

Significance between two groups was determined using Student's t test. Analyses of multiple groups were performed using a 1-way ANOVA with Bonferroni's post hoc test. All analyses were carried out using GraphPad Prism 4.0 for Macintosh (GraphPad, San Diego, CA, USA). A confidence level of P < 0.05 was accepted for statistical significance. Images were assembled using Adobe Photoshop and Illustrator CS4 (Adobe Systems, San Jose, CA, USA) and were not modified other than uniform adjustments to size, color levels, brightness, and contrast.

RESULTS

TGF-β1 decreases Sca-1 expression

During skeletal muscle regeneration, Sca-1 expression is transiently up-regulated in myoblasts 2 d postinjury. This Sca-1⁺ population is no longer evident 3 d following injury (5). The rapidity with which Sca-1 expression declines suggests the presence of factors in injured muscle capable of down-regulating Sca-1 expression. We employed real-time PCR to examine the ability of the cytokines TGF-β, IL-4, IL-6, and TNF-α, which are present in injured or diseased muscle, for their ability to down-regulate Sca-1 gene expression (25–27). As shown in Fig. 1, treatment with IL-4 or IL-6 had no effect on Sca-1 levels. However, treatment with either TGF-β1 or TNF-α resulted in significant decreases (58 and 77%, respectively) in Sca-1 mRNA levels. As TGF-β is critical to the development of muscle fibrosis, and our previous data suggest a role for Sca-1 in fibrosis, we chose to focus on TGF-β as a regulator of Sca-1 expression (3, 5).

Figure 1.

TGF-β1 decreases Sca-1 mRNA. Primary myoblasts were treated for 24 h with vehicle, 5 ng/ml TGF-β1, 5 ng/ml IL-4, 40 ng/ml IL-6, or 20 ng/ml TNF-α. Following treatment, RNA was isolated, and Sca-1 mRNA levels were analyzed by quantitative real-time PCR. Sca-1 mRNA levels were normalized to a housekeeping gene (HPRT). Fold change in expression relative to vehicle is shown; n = 3. *P < 0.05 vs. vehicle control.

Sca-1 expression is regulated by endogenous TGF-β signaling

We next investigated whether the effects of TGF-β1 on Sca-1 mRNA are reflected in changes in Sca-1 protein expression. Primary myoblasts were treated with 0, 1, or 5 ng/ml TGF-β1 for 24 h, and Sca-1 expression was determined by flow cytometry. Treatment with 1 ng/ml resulted in a 22% decrease in the number of Sca-1⁺ cells, while 5 ng/ml resulted in a 51% decrease (Fig. 2A, B). To ensure that the addition of TGF-β1 did not cause a general decrease in protein expression, we also analyzed expression of two other cell surface proteins expressed in myoblasts, CD44 and Syndecan-4, in response to TGF-β1 (Fig. 2C). Expression of neither protein was affected by TGF-β1 treatment, suggesting that the effects on Sca-1 are specific. To determine whether endogenous TGF-β signaling regulates Sca-1 expression, cells were treated with a pan-TGF-β-neutralizing antibody or an IgG control for 0.5, 1, 2, or 3 d. Relative to control, treatment with the TGF-β-neutralizing antibody for 12, 24, and 48 h resulted in 21, 25, and 21% increases, respectively, in the number of Sca-1⁺ cells, indicating that TGF-β produced by myoblasts is capable of regulating Sca-1 expression (Fig. 2D). The effect of the neutralizing antibody diminished with time, an effect previously reported with c-kit expression in hematopoietic cells (28). The increase in the number of Sca-1⁺ cells following 12 h of antibody treatment was not due to changes in cell proliferation, as measured by the percentage of BrdU⁺ cells (Fig. 2E).

Figure 2.

Sca-1 protein levels are regulated by endogenous TGF-β signaling. A) Representative flow cytometry plots of primary myoblasts incubated with vehicle or 5 ng/ml TGF-β1 for 24 h. Sca-1 was detected using a PE-conjugated Sca-1 antibody. B) Quantitation of the effects of TGF-β1 on the number of Sca-1⁺ cells. Myoblasts were treated with the indicated concentrations of TGF-β1 for 24 h, and the number of Sca-1⁺ cells was analyzed as in A. C) TGF-β1 treatment does not affect expression of other cell surface proteins. Following treatment with 5 ng/ml TGF-β1 for 24-h, myoblasts were immunostained with FITC-conjugated antibodies to CD44 and Syndecan-4 and analyzed by flow cytometry. D) Endogenous TGF-β signaling regulates Sca-1 expression. Myoblasts were treated with 2 μg/ml of a pan-TGF-β neutralizing antibody or an IgG control for the indicated times, and the numbers of Sca-1⁺ cells were analyzed by flow cytometry. E) Treatment with the TGF-β neutralizing antibody for 12 h does not affect cell proliferation. Myoblasts were incubated with IgG or the anti-TGF-β antibody for 12 h with 25 μM BrdU. The number of Sca-1⁺ cells was analyzed by flow cytometry; cells receiving BrdU were fixed and immunostained with an antibody to BrdU, and then the percentage of BrdU⁺ cells was determined. For all experiments, n = 3. *P < 0.05 vs. control.

TGF-β signaling alters Sca-1 expression in satellite cells

Cultured myoblasts have undergone multiple rounds of expansion in vitro during which certain populations of cells may not survive. To determine whether endogenous TGF-β signaling plays a role in regulating Sca-1 expression in the earliest myogenic precursors, single myofibers were isolated from Myf5:LacZ mice and cultured for 6 d. In these mice, β-galactosidase is expressed from the Myf5 locus, thereby allowing verification of the myogenic purity of the explant cultures via X-gal staining. In all experiments, the myogenic purity exceeded 95%. During the final 24 h of culture, control IgG or α-TGF-β antibody was added. In these early progenitors, as with cultured myoblasts, the addition of TGF-β-neutralizing antibody resulted in a 2-fold increase in Sca-1 fluorescence (Fig. 3A, B).

Figure 3.

TGF-β signaling regulates Sca-1 expression in satellite cells. A) Single myofibers were isolated from gastrocnemius muscles and cultured for 6 d, with 2 μg/ml IgG or anti-TGF-β antibody added for the final 24 h. Mononuclear cells were isolated and immunostained with a PE-conjugated Sca-1 antibody. Sca-1 expression was analyzed by flow cytometry. Representative histogram is shown. B) Quantitation of the fold increase in Sca-1 fluorescence shown in A; n = 3. *P < 0.05. C) Anti-TGF-β treatment does not result in increased satellite cell activation. Single myofibers were isolated from gastrocnemius muscles and cultured for 24 h. IgG or anti-TGF-β antibody (2 μg/ml) was included in all steps of myofiber isolation and culture. Myofibers were immunostained with an antibody to MyoD to identify activated satellite cells, and the number of MyoD⁺ cells per myofiber was determined. Representative image of a MyoD⁺ cell is shown. Scale bar = 15 μm. D) Quantitation of average number of MyoD⁺ cells per myofiber in explants treated with IgG or anti-TGF-β antibody. E) Frequency distribution of MyoD⁺ cells per myofiber. IgG, n = 2; myofiber, n = 71; anti-TGF-β, n = 2; myofiber, n = 68.

To ensure that the increase in Sca-1 expression following α-TGF-β treatment was not due to increased satellite cell activation, single myofibers were isolated and cultured for 24 h in the presence of control IgG or α-TGF-β. Myofibers were fixed and immunostained with an antibody to MyoD to identify activated satellite cells (Fig. 3C). No difference was observed in the average number or frequency of MyoD⁺ cells in α-TGF-β-treated myofibers relative to control (Fig. 3D, E), indicating that TGF-β signaling does not alter satellite cell activation.

TGF-β1-mediated repression of Sca-1 expression is Smad3 dependent

Activation of the TGF-β receptor results in the phosphorylation and activation of two downstream effector proteins, Smad2 and Smad3 (29, 30). Once phosphorylated, Smad2 and Smad3 form a heterodimer with the coreceptor Smad, Smad4, and this complex then enters the nucleus, where it is recruited to target DNA through interactions with other DNA binding proteins, as well as Smad-DNA interactions (31). Although the majority of TGF-β signaling results in transcriptional up-regulation, in several instances, TGF-β-mediated repression occurs. On TGF-β stimulation of myoblasts, Smad3 binds MyoD and prevents its interactions with E-box proteins. MyoD is therefore unable to bind the DNA of target genes, and myogenesis is inhibited (32). Smad3 inhibits transcription in other cell types by binding critical transcriptional regulators, such as CBFA1 and C/EBP (adipocytes), androgen receptor (prostatic epithelial cells), and MEF2 (myoblasts) (24, 33–35).

Although TGF-β signaling is mediated by Smad2 and Smad3 (29), all instances of TGF-β-mediated gene repression identified are dependent solely on Smad3 (24, 33, 34, 36). To determine whether TGF-β-dependent repression of Sca-1 is also mediated by Smad3, Smad3 expression was knocked down in primary myoblasts using siRNA, and the effects of TGF-β on Sca-1 expression in these cells were analyzed. Efficient knockdown of Smad3 was achieved, as determined by Western blot analysis (Fig. 4A). Knockdown of Smad3 did not significantly alter Smad2 expression (Fig. 4A). At 24 h after transfection with control or Smad3 siRNA duplexes, 5 ng/ml TGF-β1 was added, and Sca-1 expression was analyzed by flow cytometry after a further 24 h. As shown in Fig. 4B, knockdown of Smad3 altered the effect of TGF-β on Sca-1 expression, with TGF-β treatment resulting in only a 9.3% decrease in Sca-1⁺ cells, compared to a 21% decrease in cells transfected with control siRNA (Fig. 4B).

Figure 4.

TGF-β1 negatively regulates Sca-1 expression through a Smad3-dependent mechanism. A) Western blot analysis demonstrating knockdown of Smad3 in primary myoblasts using siRNA. Tubulin was used as a loading control. Expression of Smad2 is not significantly affected by Smad3 siRNA. B) Knockdown of Smad3 reduces the effects of TGF- β1 on Sca-1 expression. Myoblasts were transfected with the indicated siRNA duplexes and treated with vehicle or 5 ng/ml TGF-β1 for 24 h, and Sca-1 expression was analyzed by flow cytometry. Percentage change in number of Sca-1⁺ cells relative to vehicle is shown. *P < 0.05 vs. control. C) Overexpression of human Smad3 rescues the effect of Smad3 siRNA. Myoblasts were infected with vector or human Smad3, selected with puromycin, and transfected with control (open bars) or Smad3 (solid bars) siRNA duplexes. Cells were treated with vehicle or 5 ng/ml TGF-β1 for 24 h, and Sca-1 expression was determined by flow cytometry. Percentage change in number of Sca-1⁺ cells relative to vehicle is shown; n = 3. *P < 0.01 vs. control. D) Representative Western blot showing knockdown of endogenous, but not overexpressed (human), Smad3 by siRNA. Two different exposures are shown; tubulin was used to ensure equal loading. E, F) Knockdown (E) or overexpression (F) of Smad3 has no effect on Sca-1 expression in the absence of TGF-β1. Myoblasts were transfected with control or Smad3 siRNA, or infected with vector or human Smad3, and Sca-1 expression analyzed by flow cytometry after 48 h; n = 3.

To ensure that the observed effects of Smad3 knockdown were specific to reduced Smad3 expression and not a nonspecific effect of the siRNA, we performed a rescue of Smad3. Primary myoblasts were retrovirally infected with either control or human Smad3-expressing vectors (24). Following puromycin selection, cells were transfected with control or Smad3 siRNA, and TGF-β1 was added 24 h later. Smad3 overexpression and knockdown were verified by Western blot analysis (Fig. 4D). Sca-1 expression was analyzed by flow cytometry after a further 24 h. As shown in Fig. 4C, in cells infected with control vector alone, knockdown of Smad3 reduced the TGF-β1-dependent decrease in Sca-1⁺ cells from 43.7 to 7.7%. However, in cells infected with human Smad3, transfection with the Smad3 siRNA did not significantly alter the response of myoblasts to TGF-β1 relative to control siRNA. These data indicate that knockdown of Smad3 specifically alters the effect of TGF-β1 on myoblasts. Interestingly, neither knockdown nor overexpression of Smad3 affected Sca-1 expression in the absence of exogenously added TGF-β1 (Fig. 4E, F), suggesting that Smad3 is a critical regulator of Sca-1 expression only in the presence of elevated TGF-β.

TGF-β1 expression in regenerating muscle is predominantly derived from infiltrating macrophages and correlates with Sca-1 expression

We have previously shown that Sca-1 expression is transiently increased in myogenic cells on injury; Sca-1 expression is not observed in myogenic cells derived from uninjured muscle but is transiently increased in a subset of these cells 2 d postinjury; by 3 d postinjury Sca-1, expression in this population is no longer detected (5). While TGF-β is known to be up-regulated in response to injury (37), we wished to determine whether TGF-β1 levels during muscle regeneration correlated with changes in Sca-1 expression. Gastrocnemius muscles were collected 0, 1, 3, and 5 d postinjury, and the amount of TGF-β1 was determined by ELISA. As shown in Fig. 5A, TGF-β1 levels rapidly increase on injury; within 24 h postinjury, TGF-β1 levels are 4.8-fold higher than in uninjured tissue, and this increase rises to 6.8-fold by d 5 postinjury (Fig. 5A). TGF-β1 is therefore present at elevated levels in the muscle during the period when Sca-1 is expressed and then down-regulated.

Figure 5.

Macrophages contribute significantly to the increase in TGF-β1 observed during muscle regeneration. A) Muscle homogenates were generated from regenerating muscle collected 0, 1, 3, and 5 d postinjury. TGF-β1 protein levels were measured by ELISA. B) To determine the contribution of macrophages (mφ) to TGF-β1 expression, macrophages were isolated from muscle 1 d after injury using CD11b magnetic beads. RNA was generated from these cells, and TGF-β1 expression was determined by qRT-PCR. For comparison, expression of TGF-β1 was also measured from uninjured and d 1 regenerating muscle (SkM). C) Representative flow cytometry plots demonstrating the purity of the CD11b⁺ population. Purity of 90–95% CD11b⁺ was achieved.

Inflammation is a critical component of muscle regeneration (38). Following injury, muscle is infiltrated by inflammatory cells, primarily macrophages, which are critical in the production of various growth factors needed to promote regeneration and also to participate in the clearance of cellular debris (39, 40). We therefore investigated the contribution of macrophages to the increased TGF-β1 expression observed during regeneration, hypothesizing that macrophage-derived TGF-β1 may be an important contributor to down-regulation of Sca-1. CD11b⁺ cells (macrophages) were isolated from regenerating muscle 1 d postinjury using magnetic cell separation, and TGF-β1 expression was determined by real-time PCR analysis. TGF-β1 expression was also analyzed from whole uninjured and regenerating muscle as a comparison. While CD11b is expressed on both macrophages and natural killer (NK) cells, the number of NK cells present in injured muscle is negligible (data not shown), and we therefore consider CD11b to be a marker for macrophages within the muscle.

As shown in Fig. 5B, whole regenerating muscle expressed 1.5-fold more TGF-β1 than uninjured muscle. However, the relative amount of TGF-β1 produced by macrophages from regenerating muscle is 23-fold greater than TGF-β1 expression in regenerating whole muscle, indicating that a significant amount of TGF-β1 present in regenerating muscle is derived from infiltrating macrophages (Fig. 5B). Purity of the Cd11b⁺ cells postsorting was 90–95%, as assessed by flow cytometry (Fig. 5C).

TGF-β1 represses Sca-1 expression in multiple cell populations

Having demonstrated that TGF-β1 represses Sca-1 expression in myogenic cells, we next ascertained whether this repression occurs only in muscle or whether it represents a more general regulatory mechanism. Because Sca-1 is expressed in a variety of immune populations, and is particularly important for T-cell development and function (41–43), we chose to investigate the effects of TGF-β1 on Sca-1 expression in T cells. Cells were isolated from mouse spleens and the ability of TGF-β1 to down-regulate Sca-1 expression in T cells was determined. Splenic cells were incubated with vehicle or 5 ng/ml TGF-β1 for 24 h, and Sca-1 expression in CD3⁺ cells (T cells) was determined. While TGF-β1 treatment results in a decrease in the number of Sca-1⁺ T cells, this difference is not significant (Fig. 6B). However, TGF-β1 does significantly decrease the mean Sca-1 fluorescence intensity of T cells (47% decrease relative to vehicle), indicating that TGF-β1 also regulates Sca-1 expression in T cells (Fig. 6C). Interestingly, TGF-β1 also significantly reduces the Sca-1 fluorescence intensity of CD3⁻ cells (51% decrease relative to vehicle) (Fig. 6D). This population of cells comprises non-T-cell immune cells, such as B cells and macrophages, demonstrating that TGF-β1 is capable of regulating Sca-1 in multiple cell populations. T and B cells do not infiltrate damaged muscle in significant numbers (data not shown), and Sca-1 expression in these cell populations is therefore unlikely to affect regeneration. However, macrophages, which also express Sca-1, play a significant role in muscle regeneration (39, 40). Interestingly, we were unable to identify any differences in the numbers, kinetics, or phenotype of infiltrating macrophages in Sca-1^−/− mice compared to WT (data not shown), suggesting that Sca-1 is not involved in these aspects of macrophage function during regeneration.

Figure 6.

TGF-β1 also regulates Sca-1 expression in splenocytes. A) Cells were isolated from spleens of normal mice and incubated with vehicle (V) or 5 ng/ml TGF-β1 for 24 h. Representative flow cytometry plots of total splenocytes are shown. Sca-1 expression was detected using a PE-conjugated Sca-1 antibody. T cells were detected using a FITC-conjugated CD3 antibody. B) Quantitation of the effects of TGF-β1 on the number of Sca-1⁺ T cells. C) Quantitation of the effects of TGF-β1 on mean Sca-1 fluorescence in T cells. D) Mean Sca-1 fluorescence of CD3⁻ (non-T) cells treated with vehicle or TGF-β1. n = 3 mice. *P < 0.003.

DISCUSSION

In this study, we identify TGF-β as a novel negative regulator of Sca-1 expression in myoblasts and demonstrate that the effects of TGF-β are dependent on Smad3. The observed TGF-β-dependent decrease in Sca-1 expression was not due to changes in cell proliferation, as BrdU incorporation was not altered in TGF-β-treated cells relative to vehicle-treated cells. In addition, the negative effects of TGF-β were present in the earliest myogenic precursors observed in myofiber explant cultures, suggesting that this effect was not a result of prolonged time in culture.

Expression of Sca-1 in multiple tissues is highly dynamic, and proper regulation of expression is required for tissue development and homeostasis. For example, dynamic Sca-1 expression is required for T-cell development; Sca-1 is expressed in hematopoietic stem cells, down-regulated in prothymocytes, and reexpressed by mature T cells. Constitutive expression of Sca-1 results in impaired generation of thymocytes (42). In addition, we have previously shown that Sca-1 expression in myogenic cells is transiently up-regulated following muscle injury; mice lacking Sca-1 display increased fibrosis during regeneration, due in part to decreased activity of matrix metalloproteinases (MMPs) and reduced turnover of the extracellular matrix (ECM) (5). We hypothesize that expression of Sca-1 during muscle regeneration increases MMP activity, thereby enhancing ECM remodeling and promoting efficient muscle regeneration. Here, we expand on our previous observations by demonstrating that TGF-β levels rapidly increase following muscle injury, largely generated by infiltrating macrophages, and that the timing of TGF-β up-regulation correlates with Sca-1 expression and subsequent down-regulation. These data suggest that the transient nature of Sca-1 expression during regeneration is largely due to the repressive effects of TGF-β, and that down-regulation of Sca-1 results in reduced MMP activity and decreased ECM turnover. TGF-β plays a critical role in the development of fibrosis, both by directly increasing transcription of collagen and other ECM genes, and by repressing transcription of MMPs (1, 38, 44), and our results suggest that modulation of Sca-1 expression could be another mechanism by which TGF-β influences fibrosis. Although Sca-1 expression in myogenic cells is not apparent until 2 d postinjury (5), TGF-β1 expression is significantly elevated within 24 h. How then is Sca-1 expression up-regulated in the presence of a repressor? The effects of TGF-β vary widely depending on the cellular context. For example, TGF-β can be proliferative or antiproliferative, proapoptotic or antiapoptotic, and prodifferentiative or antidifferentiative depending on its concentration, the local environment, and the differentiation state of the cell (45, 46). Local concentrations of TGF-β may be insufficient to repress Sca-1 expression in rapidly proliferating cells until 2 or 3 d postinjury. Alternatively, myogenic cells may not express the appropriate repertoire of proteins required for TGF-β to exert its repressive effect.

Our results also indicate that the repression of Sca-1 in response to elevated levels of TGF-β1 is dependent on Smad3. These data are in concurrence with other instances of TGF-β-mediated repression, in which Smad3 functions to repress gene transcription by binding to and inhibiting critical transcriptional regulators (24, 33–35). Although the exact mechanism by which Smad3 represses Sca-1 expression is unknown, we hypothesize that, as in other cell types, Smad3 binds to and inhibits a transcription factor critical in the regulation of Sca-1 expression. In many instances of Smad3-mediated repression, Smad3 interacts with target transcription factors directly on the DNA, although Smad3 prevents myogenesis by binding to MyoD in the cytosol (24, 32, 33). In many tissues, Sca-1 expression is highly up-regulated by IFNα/β and IFNγ, and while multiple interferon response elements have been identified in the Sca-1 promoter, the mechanisms of Sca-1 regulation by interferons have not been fully elucidated. The transcription factors STAT1, Oct1, Oct2, and HMGI(Y) are required for Sca-1 up-regulation by IFNγ, but whether these proteins are necessary for Sca-1 regulation on additional stimuli is unknown (15). To date, the transcription factors that regulate Sca-1 expression in muscle has not been identified.

Interestingly, the effects of TGF-β on Sca-1 expression are not limited to myogenic cells, suggesting the TGF-β regulation of Sca-1 is a general phenomenon. We show here that TGF-β also repressed Sca-1 expression in splenic T cells. Given the observation that TGF-β is a critical regulator of T-cell development (47), these data suggest that the effects of TGF-β on T cells may be due to regulation of Sca-1 expression, itself an important regulator of T-cell function and development (42). Proper regulation of Sca-1 expression is critical for development of T cells. Sca-1 is expressed in immature CD4⁻ CD8⁻ CD3⁻ thymocytes and is subsequently down-regulated in developing CD4⁺CD8⁺ cells. Sca-1 is then up-regulated in mature CD4⁺ and CD8⁺ cells (41). Thymocytes from mice in which Sca-1 is constitutively overexpressed are permanently arrested at the immature stage, indicating that correct timing of Sca-1 expression is essential for T-cell development (42). The mechanism by which Sca-1 is regulated during T-cell development is unknown; however, given the ability of TGF-β to regulate Sca-1 expression in splenic T cells and the importance of TGF-β in all stages of hematopoiesis (45), we hypothesize that regulation of Sca-1 by TGF-β plays an important role in T-cell development and function. Because T cells are not recruited to injured muscle in significant numbers, it is unlikely that TGF-β1 repression of Sca-1 in these cells plays a role in muscle regeneration. Rather, these data indicate that regulation of Sca-1 by TGF-β1 is a general phenomenon and is likely to play a role in other Sca-1 dependent functions, such as T-cell development.

The ability of TGF-β to regulate Sca-1 expression in other cell types also has implications in muscle regeneration. Multiple cell types present in skeletal muscle express Sca-1, including endothelial cells, macrophages, and the recently identified fibro-adipogenic precursors (FAPs) and Sca-1 expression in these cells is likely down-regulated by TGF-β as well (6, 48). For example, FAPs promote differentiation of myogenic cells and generate fibroblasts required to repair muscle structure (48), and decreased expression of Sca-1 may affect the function of these cells, resulting in decreased myogenesis or increased fibrosis.

The effects of TGF-β on Sca-1 expression in myoblasts are particularly intriguing given the critical role of TGF-β in skeletal muscle growth and homeostasis. TGF-β inhibits myogenic differentiation, and elevated levels of TGF-β present in multiple myopathies result in increased fibrosis and inefficient regeneration (3, 17). In addition, a large-scale genomic analysis has implicated TGF-β signaling as a critical component of muscle atrophy (49). On the basis of data presented here, we hypothesize that down-regulation of Sca-1 by TGF-β in myogenic cells is critical for the effects of the cytokine observed in skeletal muscle, by affecting MMP activity and ECM turnover through its effects on Sca-1 expression. Because the excess TGF-β present in disease states (such as muscular dystrophies) has such deleterious effects, TGF-β is an attractive therapeutic target (3). However, studies of mice lacking TGF-β1 suggest that long-term antagonism of TGF-β may have detrimental side effects. These mice exhibit a severe wasting syndrome and a generalized inflammatory response leading to tissue necrosis and organ failure (50). If down-regulation of Sca-1 by TGF-β is an integral part of this response than therapies that interfere only with the ability of TGF-β to affect Sca-1 expression may improve muscle regeneration without severe negative side-effects.