Identification and characterization of PDGFRα+ mesenchymal progenitors in human skeletal muscle

Abstract

Fatty and fibrous connective tissue formation is a hallmark of diseased skeletal muscle and deteriorates muscle function. We previously identified non-myogenic mesenchymal progenitors that contribute to adipogenesis and fibrogenesis in mouse skeletal muscle. In this study, we report the identification and characterization of a human counterpart to these progenitors. By using PDGFRα as a specific marker, mesenchymal progenitors can be identified in the interstitium and isolated from human skeletal muscle. PDGFRα(+) cells represent a cell population distinct from CD56(+) myogenic cells, and adipogenic and fibrogenic potentials were highly enriched in the PDGFRα(+) population. Activation of PDGFRα stimulates proliferation of PDGFRα(+) cells through PI3K-Akt and MEK2-MAPK signaling pathways, and aberrant accumulation of PDGFRα(+) cells was conspicuous in muscles of patients with both genetic and non-genetic muscle diseases. Our results revealed the pathological relevance of PDGFRα(+) mesenchymal progenitors to human muscle diseases and provide a basis for developing therapeutic strategy to treat muscle diseases.

Figure 1

Localization of PDGFRα⁺ cells in human skeletal muscle. Human muscle sections were stained with antibodies against M-cadherin, Pax7, and laminin (a); M-cadherin, CD56, and laminin (b); PDGFRα, Pax7, and laminin (c); PDGFRα, CD56, and laminin (d); and PDGFRα and laminin (e), followed by HE staining. Arrows indicate PDGFRα⁺ cells located in the interstitial spaces of muscle tissue, and arrowheads indicate satellite cells located beneath the basement membrane. Scale bars: 10 μm

Figure 2

Flow cytometric analysis of PDGFRα⁺ cells. (a) Human muscle-derived cells were analyzed for PDGFRα and CD56 expression. Representative data of 30 independent experiments are shown. (b) Positive gates were set by analyzing negative control samples stained with isotype control antibody or secondary reagent only. (c) Sorted PDGFRα⁺ cells were reanalyzed for PDGFRα and CD56 expression. Consistent results were obtained from two independent experiments. (d) Sorted CD56⁺ cells were reanalyzed for PDGFRα and CD56 expression. Consistent results were obtained from two independent experiments. Expressions of indicated cell surface antigens, PDGFRα⁺ cells (e), CD56⁺ cells (f), and bone marrow-derived mesenchymal stem cells (BMMSC, g) were analyzed. Positive gates were set by analyzing negative control samples stained with isotype control antibody (dotted line). The percentages of each cell population are shown in the panels

Figure 3

Myogenic markers are detected only in CD56⁺ population. (a) Reverse transcription-polymerase chain reaction (RT-PCR) analysis of indicated genes in the three populations indicated. RNA was extracted immediately after cell sorting, and RT-PCR was performed. (b) The three indicated populations were cultured for 3 days in growth conditions after cell sorting and then stained with antibodies against MyoD and Pax7. The percentages of positive cells are shown in the panels as means±s.d., n=20 fields from three independent preparations. Scale bar: 50 μm

Figure 4

Adipogenic differentiation of PDGFRα⁺ cells and myogenic differentiation of CD56⁺ cells after adipogenic induction. (a) The three populations indicated were sorted from human muscle-derived cells and cultured in growth condition for 3 days, and then cells were subjected to adipogenic condition. Cells were stained with antibodies against C/EBPα and peroxisome proliferator-activated receptor-γ (PPARγ) or Oil Red O. The percentages of positive cells are shown in the panels as means±s.d., n=15 fields from three independent preparations. Scale bar: 50 μm. (b) Adipogenic differentiation was evaluated by quantifying the intensity of Oil Red O staining. Values are represented as the ratio to CD56⁺ cells and shown as means±s.d. of eight independent preparations. *P<0.01, **P<0.05. (c) Reverse transcription-polymerase chain reaction (RT-PCR) analysis of indicated genes in undifferentiated PDGFRα⁺ cells and PDGFRα⁺ cell-derived adipocytes. (d) After adipogenic induction, the three populations indicated were stained with antibodies against MyoD and MyHC. Scale bar: 50 μm

Figure 5

Fibrogenic potential of PDGFRα⁺ cells and effects of TGF-β and PDGF signaling on PDGFRα⁺ cells. (a) Single PDGFRα⁺ cell-derived colonies were stained with antibodies against collagen I (Col1) and FABP4 or FABP4 and α-smooth muscle actin (α-SMA). FABP4 antibody labeled adipocytes with lipid droplet, whereas Col1 and α-SMA were detected only in fibrogenic cells. Scale bar: 50 μm. (b) CD56⁺ cells or PDGFRα⁺ cells were cultured with or without TGF-β1 (5 ng/ml) for 3 days, and the expressions of fibrosis-related genes were quantified by quantitative real-time polymerase chain reaction (qRT-PCR). Values are represented as the ratio to unstimulated CD56⁺ cells and shown as means±s.d. of three independent preparations. *P<0.01, **P<0.05. (c) Western blot analysis of collagen I, collagen III, and glyceraldehyde 3-phosphate dehydrogenase (GAPDH). Results from two independent preparations were shown. (d) Matrix metalloproteinase (MMP) activity in the cell culture supernatants. Values are represented as the substrate amounts that were cleaved by MMPs contained in the supernatants and shown as means±s.d. of three independent preparations. (e) Tissue inhibitors of metalloproteinase-1 (TIMP-1) concentration in the cell culture supernatants. Values are shown as means±s.d. of three independent preparations. (f) PDGFRα⁺ cells were cultured in SFM with or without PDGF-AA (10 ng/ml) for 2 days. Inhibitors were added 1 h before PDGF-AA stimulation. The extent of PDGFRα phosphorylation is represented as the ratio to unstimulated control cells (cont) and shown as means±s.d. of three independent preparations. *P<0.01, **P<0.05. (g) Proliferation of PDGFRα⁺ cells was quantified by measuring BrdU incorporation. Values are represented as the ratio to unstimulated control cells (cont) and shown as means±s.d. of three independent preparations. *P<0.01; versus control culture. (h and i) PDGFRα⁺ cells cultured in SFM were stimulated with PDGF-AA (10 ng/ml) for 15 min. Inhibitors were added 1 h before PDGF-AA stimulation. Phosphorylation of Akt and p44/42 MAPK was assessed by immunoblotting (h) and immunofluorescent staining (i). Scale bar: 50 μm in i

Figure 6

Pathological behavior of PDGFRα⁺ cells in human diseased muscle. DMD muscle sections exhibiting fatty and fibrous degeneration were subjected to immunofluorescent staining for PDGFRα (a and c), or PDGFRα, peroxisome proliferator-activated receptor-γ (PPARγ), and perilipin (b), and subsequently to HE staining. DMD muscle sections were subjected to immunofluorescent staining for CD31 and PDGFRα (d), or α-smooth muscle actin (α-SMA) and PDGFRα (e). Arrows indicate PDGFRα⁺ cells located around blood vessels. (f) DMD muscle sections were subjected to immunofluorescent staining for CD56, PDGFRα, and laminin. Arrowheads indicate CD56⁺ cells located inside of the basement membrane. Volkmann's contracture muscle samples were subjected to immunofluorescent staining for PDGFRα and perilipin, and subsequently to HE staining (g). Fluorescent image corresponds to boxed area in HE image. Scale bars: 20 μm in (a, c and g), 10 μm in (b, d, e and f)

Figure 7

Correlation between the extent of PDGFRα staining and severity of fibrosis. (a–c) Muscle sections from DMD (a and b) and control (c) patients were subjected to HE staining. Scale bar: 100 μm. (d–f) Muscle sections from DMD (d and e) and control (f) patients were subjected to immunofluorescent staining for PDGFRα and collagen I. Scale bar: 20 μm. (g) PDGFRα staining grades correlate positively with severity of fibrosis (Spearman's rank correlation coefficient=0.80, P=0.001)