Transforming growth factor-beta1 induces the differentiation of myogenic cells into fibrotic cells in injured skeletal muscle: a key event in muscle fibrogenesis

Abstract

Transforming growth factor-beta1 (TGF-beta1) is thought to play a crucial role in fibrotic diseases. This study demonstrates for the first time that TGF-beta1 stimulation can induce myoblasts (C2C12 cells) to express TGF-beta1 in an autocrine manner, down-regulate the expression of myogenic proteins, and initiate the production of fibrosis-related proteins in vitro. Direct injection of human recombinant TGF-beta1 into skeletal muscle in vivo stimulated myogenic cells, including myofibers, to express TGF-beta1 and induced scar tissue formation within the injected area. We also observed the local expression of this growth factor by myogenic cells, including regenerating myofibers, in injured skeletal muscle. Finally, we demonstrated that TGF-beta1 gene-transfected myoblasts (CT cells) can differentiate into myofibroblastic cells after intramuscular transplantation, but that decorin, an anti-fibrosis agent, prevents this differentiation process by blocking TGF-beta1. In summary, these findings indicate that TGF-beta1 is a major stimulator that plays a significant role in both the initiation of fibrotic cascades in skeletal muscle and the induction of myogenic cells to differentiate into myofibroblastic cells in injured muscle.

Figure 1

Autocrine expression of TGF-β1 in myogenic cells in vitro. Unstimulated C2C12 cells do not express TGF-β1. We detected TGF-β1 transcripts (371 bp) in the extract of the TGF-β1-treated C2C12 cells stimulated with high concentrations of hrTGF-β1 (1.0 ng/ml and 5.0 ng/ml) at 1 hour (a, top) and with all tested concentrations of hrTGF-β1 at 2 hours (a, middle). c and d: We also detected the expression of TGF-β1 protein (25KD) in C2C12 cells after 8 hours of stimulation with a high concentration of hrTGF-β1. b to d: We found more activated TGF-β1 in stimulated C2C12 myoblasts when we treated the samples with acid. b: Enzyme-linked immunosorbent assay results indicated that, in comparison to C2C12 cells, CT clone cells secrete a large amount of TGF-β1 in a time-dependent manner. a to c: The expression of TGF-β1 by CT cells was because of TGF-β1 gene transfection and was used as the positive control.

Figure 2

TGF-β1 stimulates fibrotic protein production but down-regulates myogenic protein expression in myogenic cells in vitro. a: C2C12 cells expressed low levels of fibronectin and vimentin but did not express α-SMA. However, TGF-β1 induced the expression of fibrotic proteins (ie, α-SMA, fibronectin, and vimentin) in C2C12 myoblasts after 8 hours of incubation (a and b). c: C2C12 cells expressed several myogenic proteins, including myogenin, MyoD, and desmin. However, the expression of these myogenic-related proteins in C2C12 cells was down-regulated after 8 hours of hrTGF-β1 incubation (c and d).

Figure 3

Autocrine expression of TGF-β1 by the myogenic cells in muscle injected with TGF-β1. We labeled collagen type IV to outline the basal lamina of muscle fibers (a to f, green). Normal skeletal muscle does not express TGF-β1 (e); however, we detected TGF-β1 in the skeletal muscle in which hrTGF-β1 was injected (a to d and f, red). We also detected the autocrine expression of TGF-β1 in myofibers 3 hours after injection (a, asterisks). Numerous myofibers were positive for TGF-β1 12 hours after the injection of hrTGF-β1 (b, asterisks). TGF-β1 continued to be expressed in myofibers for up to 5 days after injection (d and f, asterisks). At 5 days after injection, several muscle fibers remained positive for TGF-β1 (d and f, asterisks), and many mononucleated cells that were positive for TGF-β1 had appeared (d and f, arrows). We also detected increased numbers of TGF-β1-positive myofibers from 3 hours to 12 hours after injection (g). The number of TGF-β1-positive myofibers remained high for 48 hours, had decreased by 3 to 5 days later, and was undetectable thereafter (g).

Figure 4

TGF-β1-positive myofibers are replaced by mononucleated cells in a time-dependent manner. Many of the TGF-β1-positive myofibers had been replaced by TGF-β1-expressing mononucleated cells (a, arrow) 5 days after injection. These mononucleated cells were continuously positive for TGF-β1 on days 5 through 7 after injection (a and b, arrows); however, few TGF-β1-positive myofibers were visible during this time period (a and b). The mononucleated cells eventually replaced the myofibers and differentiated into fibrotic cells that were visible at 7, 14, and 21 days after injection (b to d, arrowheads). By using trichrome staining, we confirmed that a large amount of collagen was deposited in the TGF-β1-positive location (e to h, arrows and arrowheads).

Figure 5

Detection of CD11b-positive cells within TGF-β1-injected and injured skeletal muscles. A few CD11b-positive cells (ie, macrophages) remained in the TGF-β1-injected muscle 7 days after injection (a and c, red). These CD11b-positive cells also expressed fibrotic markers such as α-SMA (b and c, green). By counting these CD11b-positive cells, we found that the number of macrophages in the TGF-β1-injected muscle decreased at time points after 7 days of injection (d). A time-dependent infiltration of CD11b-positive cells that co-expressed α-SMA also was detected in the injured skeletal muscle (cardiotoxin, e to g; laceration, i to k). Many more CD11b-positive cells were detected 5 days after injury by cardiotoxin (h) or laceration (l) than at other time points after injury. The asterisks identify blood vessels in the skeletal muscle (e to g).

Figure 6

Co-expression of TGF-β1 and neonatal MyHC in myofibers after injury. We observed TGF-β1-positive myofibers in the injured skeletal muscle at 3 days after cardiotoxin (a to c) or laceration (d to f) injury. The TGF-β1-expressing myofibers also were positive for neonatal MyHC (a, c, d, and f), suggesting that TGF-β1 was expressed in regenerating myofibers.

Figure 7

The number of TGF-β1-expressing myofibers decreases gradually in a time-dependent manner after muscle laceration injury. We could still detect some TGF-β1-expressing myofibers at 3 days after laceration (a, asterisks). By 5 days after laceration, these TGF-β1-positive myofibers had been replaced by mononucleated cells that were positive for TGF-β1 (b, arrowheads). The connective tissue grew gradually throughout time and was positive for TGF-β1 (b to e, arrows). The myofibers, including regenerated myofibers in the injured skeletal muscle, gradually became smaller after laceration (a to f, green), as evidenced by the reduced diameter of these myofibers at subsequent time points (g). A large amount of scar tissue had formed by 3 weeks after injury (f, asterisks). A negative control experiment performed without the TGF-β1 primary antibody demonstrated a lack of autofluorescence for TGF-β1 immunostaining (h).

Figure 8

TGF-β1 expression promotes myoblast differentiation into fibrotic cells in vivo. Histology and LacZ staining revealed that the transplanted C2C12 cells that were genetically engineered to express the β-galactosidase reporter gene had survived and regenerated numerous myofibers in the injected skeletal muscle of SCID mice at 1, 2, and 3 weeks after transplantation (a, c, e, and g). Although some of the transplanted CT clone cells (C2C12 cells expressing TGF-β1) also had regenerated a few myofibers by 1 week after injection (b), the number of these LacZ-positive myofibers in the injected area had significantly decreased by 2 weeks after injection (d). We detected a large amount of scar tissue in the CT cell-transplanted area 3 weeks after transplantation (f and h). By counting the LacZ-positive myofibers in the C2C12 and CT cell-transplanted areas within the injected muscle at different time points after transplantation, we determined that the number of LacZ-positive myofibers in the C2C12 groups remained stable (i, blue bar); however, the number of LacZ-positive myofibers in the CT cell-injected areas decreased significantly with time after injection (i, red bar). The scar tissue found in the CT cell-injected muscle was positive for β-galactosidase (j and l, green) and strongly expressed α-SMA (m and o, green) and vimentin (n and o, red), but was negative for desmin (k and l, red), suggesting that the CT cells had differentiated into fibrotic cells by 3 weeks after transplantation. We also performed a negative control experiment (p) in which the first antibody (rabbit anti-desmin) was omitted from the immunohistochemistry; no immunostaining was observed.

Figure 9

Decorin reduces the expression of fibrosis-related proteins by CT cells and decreases TGF-β1-induced fibrosis in skeletal muscle. The high level of fibrosis-related protein expression in CT cells decreased after treatment with decorin in vitro (a and b). LacZ and eosin staining revealed that the CT cells had led to the development of scar tissue in the injected skeletal muscle by 3 weeks after injection (c and d); however, the CT cells persisted and regenerated myofibers after decorin treatment (CT+Decorin; 1 week later) in vivo (e and f). Results based on the number of LacZ-positive myofibers revealed significantly more LacZ-expressing myofibers in the group treated with decorin 1 week after CT cell transplantation (CT+Decorin; 1 week later) than in the group that was injected with a combination of CT cells and decorin (CT+Decorin; same time) or in the nontreated (control) group (g).