Expression of cell surface antigens in human fetal skeletal mononuclear cells. (A–C) Flow cytometry analysis of 18-wk human fetal skeletal muscle cells. Cells were costained with 3 μg/ml Hoechst 33342, anti–CD34-FITC, anti–CD133-PE, and PI. (A) Cells stained with Hoechst 33342. 7.49% of the cells are detected in the SP gate and 62.9% in the MP gate. (B) CD34-positive cells stained with Hoechst 33342. 3.66% of the cells are detected in the SP gate and 67.2% in the MP. (C) CD133-positive cells stained with Hoechst 33342. 0.56% of the cells are detected in the SP gate and 53.3% in the MP gate. (D–I) Expression of ABCG2 (D) 3T3 cells transfected with a mock vector. 99.6% of the cells are ABCG2-negative. (E) 3T3 cells transfected with human ABCG2. 60.7% of the cells are ABCG2-positive. (F–I) Three-color flow cytometry (Hoechst 33342, PI, and ABCG2-PE) staining profiles of 18-wk-old human fetal skeletal muscle. Cells were costained with 3 μg/ml Hoechst 33342, anti–ABCG2-FITC, and PI. (F) 2% of human fetal skeletal myoblasts express ABCG2. (G–H) Hoechst/PI profiles of all the cells stained in the presence (G) or absence (H) of reserpine. 2.25% of cells were detected as SP. Hoechst/PI profile of ABCG2-positive cells (I) showed only one ABCG2-positive cell detected in the SP gate.

Effects of BMP4 and Gremlin on the differentiation and proliferation of human SP and MP cells. (A) Brightfield photographs of cultured SP and MP cells. Under all conditions, muscle SP cells were nonadherent and morphologically round throughout the culture period, whereas MP cells differentiated into myotubes. Arrows point to mononuclear cells (SP and MP) and myotubes (MP, day 14). (B) Effects of BMP4 and Gremlin on human fetal muscle MP cell differentiation. Myogenic differentiation was visualized by immunostaining with anti–myogenin-FITC (green). (a) MP cells cultured under normal differentiation conditions without addition of factors. (b) MP cells cultured in the presence of 25 ng/ml BMP4. (c) MP cells cultured in the presence of 25 ng/ml BMP4 and 2 μg/ml Gremlin. (d) MP cells cultured in the presence of 2 μg/ml Gremlin. Nuclei are stained in blue with DAPI. (C) [3H]thymidine incorporation by SP cells (left) and MP cells (right) cultured as described above, performed at day 7 of culture. Addition of BMP4 to MP cells resulted in a significant increase of [3H]thymidine incorporation compared with nontreated controls, whereas it had no effect on proliferation of SP cells. Addition of Gremlin to BMP4-treated MP cells significantly decreased their proliferation rate compared with MP cells treated with BMP4 alone. Gremlin alone inhibited proliferation of MP and SP cells compared with untreated control cultures. Error bars indicate SD. *, P < 0.05. Bars, 0.2 mm.

Effect of SP cells on myogenic differentiation. (A) Purified human fetal muscle MP cells were cocultured with irradiated SP cells. Myogenic differentiation was visualized by immunostaining for myogenin (green) and MHC (green) at day 7 of culture. Control MP cultures had 22.9 ± 4.4% (mean ± SD) of myogenin-positive cells and 22.4 ± 6.9% (mean ± SD) of MHC-positive cells, whereas MP cells cocultured with irradiated SP cells demonstrated no myogenin or MHC expression. (B) Effect of purified muscle SP cells on proliferation of BMPR1a⁺ (left) and BMPR1a⁻ (right) cells. BMPR1a⁺ and BMPR1a⁻ cells were cultured with no treatment, cocultured with irradiated SP cells, cocultured with irradiated SP cells and BMP4 blocking antibody, or cocultured with irradiated SP cells and IgG2b isotype control antibody. Proliferation was assessed using a [3H]thymidine incorporation assay after 5 d of culture, and proliferation indexes were established by calculating the ratios of thymidine incorporation in treated versus untreated cells. Error bars indicate SD.

Immunohistochemistry of human embryonic skeletal muscle, 20-wk gestation. A–F represent consecutive sections. Insets in A, C, and E correspond precisely to the high magnifications (1,000×) portrayed in B, D, and F, respectively. (A and B) Dystrophin staining defines mature muscle fibers, represented in cross section. Inset denotes interstitial space containing dystrophin-negative nucleated cells. (C and D) BMP4-reactive cells are rounded in contours and have smaller cross-sectional diameters than mature BMP4-negative skeletal muscle fibers. BMP4-positive cells (arrowheads) focally rim interstitial spaces formed by mature muscle fibers. Black arrows (D) define the area of BMPR1a⁺ cells shown in F. (E and F) Occasional cells within interstitial spaces show variable membrane reactivity for BMPR1a (arrows), where they are found as clusters surrounded by BMP4-positive cells (D and F). (G and H) Sequential sections of 20-wk fetal skeletal muscle stained for BMP4 and Gremlin. (G) Staining for Gremlin (brown; 400×); (H) staining for BMP4 (brown; 400×). Nuclei are stained violet. Note Gremlin staining in mature myofibers (MF) and scattered rounded cells in the interstitium (IS; G, arrows). Staining for BMP4 is present only in rounded cells that are most prominent in the interstitium (H, arrows). Inset (1,000×) is double labeled (nuclei are unstained), showing mutually exclusive staining for BMP4 (blue) and Gremlin (brown) by interstitial cells. Bars: (A–H) 0.2 mm; (H, inset) 0.02 mm.

Expression of BMP4 and Gremlin in human fetal skeletal muscle SP and MP cells. (A) Quantitative real-time PCR of BMP4 and Gremlin mRNA expression in SP (green) and MP (blue) cells. mRNA expression was measured as Ct values. Ct values for each gene were normalized to the respective values for GAPDH. (B) Expression of BMP4 (top) and Gremlin (bottom) proteins in SP and MP cells. BMP4- and Gremlin-positive cells are stained in green (Alexa Fluor 488), and nuclei are stained in blue (DAPI). Three representative fields are shown for each population. (C) Quantitative immunofluorescence of BMP4 (top) and Gremlin (bottom) protein expression in SP and MP cells, measured as mean pixel intensity per unit area. For BMP4, the mean pixel intensity per unit area for SP cells (n = 96) is 506 ± 30, versus 113 ± 8 pixels/μm² for MP cells (n = 111; mean ± SEM). Gremlin expression is significantly higher in human MP cells (n = 113; 455 ± 22 pixels/μm² [mean ± SEM]) than in human SP cells (n = 87; 261 ± 17 pixels/μm² [mean ± SEM]). *, P < 0.0001

Effect of BMP4 stimulation on BMPR1a-expressing cells. (A) Expression of BMPR1a in human fetal skeletal myoblasts. Three-color flow cytometry (Hoechst 33342, PI, and BMPR1a-FITC) staining profiles of an 18-wk gestation sample. Cells were costained with 3 μg/ml Hoechst 33342, anti–BMPR1a- FITC, and PI. (1 and 2) Cells stained with Hoechst 33342 in the presence (1) or absence (2) of reserpine. 1.96% of the cells are detected in the SP gate. (3) 6.79% of cells are BMPR1a⁺. (4) Approximately 99% of BMPR1a⁺ cells are located outside of the SP gate. (B) [3H]thymidine incorporation by BMPR1a⁺ and BMPR1a⁻ cells cultured with and without BMP4 for 4 d. Addition of BMP4 to BMPR1a⁺ cultures resulted in a significant increase of [3H]thymidine incorporation compared with nontreated controls. BMP4 had no effect on the proliferation of BMPR1a⁻ cells. *, P < 0.05. Error bars indicate SD. (C) Expression of Myf5, MyoD, and Pax7 mRNA in freshly isolated BMPR1a⁺ and BMPR1a⁻ cells. mRNA expression levels were normalized to GAPDH for each condition. Bars represent on a logarithmic scale the level of expression of each myogenic factor in BMPR1a⁺ and BMPR1a⁻ relative to the expression of GAPDH, which is arbitrarily considered to be 1. The baseline expression of Myf5 is twofold higher in BMPR1a⁺ cells than in BMPR1a⁻ cells, whereas MyoD and Pax7 mRNA were expressed 1.5-fold higher in BMPR1a⁻ cells. (D) Expression of Pax7 and MyoD on cytospins of purified human fetal skeletal muscle BMPR1a⁺ and BMPR1a⁻ cells. By total cell count, 1.04% of 192 analyzed BMPR1a⁺ cells and 12.9% of 743 analyzed BMPR1a⁻ cells expressed Pax7. None of 205 analyzed BMPR1a⁺ cells and 7.4% of 594 analyzed BMPR1a⁻ cells expressed MyoD. BMPR1a⁺ cells are stained in green (FITC), Pax7- and MyoD-positive cells are stained in red (TRITC), and nuclei are stained in blue with DAPI. The arrows indicate Pax7- and MyoD-positive cells. The arrowheads in the insets in the top panels highlight the same Pax7- and MyoD-positive cells in the bottom panels, which are marked by arrows. Bars, 0.2 mm.

Expression of BMP4 in rare interstitial human fetal muscle cells. (A–D) Arrowheads point to single BMP4-positive cells (green, FITC) located between dystrophin positive myofibers (stained in red, Texas red). Nuclei are stained in blue (DAPI). (E and G) Clusters of interstitial BMP4-positive cells with corresponding hematoxylin/eosin staining (F and H) at lower magnification. Black squares in F and H are shown at higher magnification in E and G, respectively.

Schematic illustration of the proposed cellular hierarchy within skeletal muscle. BMP4^high quiescent SP cells (A, red) induce the proliferation of BMPR1a⁺ Myf5^high MP cells (B, purple) via BMP4 secretion, and BMPR1a⁻ Myf5^low Gremlin^high committed MP cells (C, blue) may inhibit the stimulatory effect of BMP4 by secreting Gremlin. Cells are displayed schematically according to their respective positions in the Hoechst/PI profile.