Derivation and expansion of PAX7-positive muscle progenitors from human and mouse embryonic stem cells

Abstract

Cell therapies treating pathological muscle atrophy or damage requires an adequate quantity of muscle progenitor cells (MPCs) not currently attainable from adult donors. Here, we generate cultures of approximately 90% skeletal myogenic cells by treating human embryonic stem cells (ESCs) with the GSK3 inhibitor CHIR99021 followed by FGF2 and N2 supplements. Gene expression analysis identified progressive expression of mesoderm, somite, dermomyotome, and myotome markers, following patterns of embryonic myogenesis. CHIR99021 enhanced transcript levels of the pan-mesoderm gene T and paraxial-mesoderm genes MSGN1 and TBX6; immunofluorescence confirmed that 91% ± 6% of cells expressed T immediately following treatment. By 7 weeks, 47% ± 3% of cells were MYH(+ve) myocytes/myotubes surrounded by a 43% ± 4% population of PAX7(+ve) MPCs, indicating 90% of cells had achieved myogenic identity without any cell sorting. Treatment of mouse ESCs with these factors resulted in similar enhancements of myogenesis. These studies establish a foundation for serum-free and chemically defined monolayer skeletal myogenesis of ESCs.

Figure 1

Ten Micromolar CHIR Applied for 2 Days Produces Optimal Levels of Paraxial and Premyogenic Mesoderm Gene Expression (A) hESCs were treated in monolayer with or without 2.5 or 10 μM CHIR for 2, 4, or 8 days. RNA was harvested at days 0, 2, and 8 to be analyzed by qPCR for markers of early and premyogenic mesoderm. Results are expressed as fold change over day 0 (n = 3 independent experiments, ^∗p ≤ 0.05 versus day 0, ^∗∗p ≤ 0.05 versus day 0 and control, ^∗∗∗p ≤ 0.05 2.5 versus 10 μM). (B) CHIR-differentiated cells were compared to BMP4/INHBA treatment by qPCR for markers of early mesoderm. (n = 3 independent experiments, ^∗p ≤ 0.05 versus day 0, ^∗∗p ≤ 0.05 versus day 0 and control, ^∗∗∗p ≤ 0.05 BMP4/INHBA versus CHIR). (C) Schematic overview of the CHIR directed differentiation of hESCs on a nonlinear scale timeline, along with highlighting key supplemental factors applied during the differentiation, and the predicted developmental stages of skeletal myogenesis. (D) Representative phase contrast live-cell images demonstrating the progressive changes in hESC morphology with and without CHIR treatment. 3D cell clusters (white arrowheads) and the presence of skeletal myocytes (black arrowheads) are indicated at specific time points (scale bar, 20 μm). Results are shown ±SEM.

Figure 2

CHIR Directed Differentiation of hESCs Induces up to a 90% Myogenic Population (A–G) hESCs were differentiated as described in Figure 1. Myogenic induction was quantified by staining with antibodies against (A) and (E) T at day 2, (B) and (F) PAX3 at day 8, (C) and (G) MYH and PAX7 at day 40, and (D) and (G) MYH and PAX7 at day 50. Target populations are expressed as a percentage of total number of cells, which was quantified by Hoechst staining (n = 3 independent experiments, ^∗p ≤ 0.05 versus control, ^∗∗p ≤ 0.05 day 40 versus day 50, scale bar, 20 μm). (H) Total number of cells was obtained by trypsinizing and counting one well of a 12-well culture dish at various time points of the differentiation protocol (n = 3 independent experiments). (I) Day 50 CHIR-treated cells were pulsed with EdU for 4 hr and fixed and stained additionally with PAX7. EdU and PAX7 single or double-positive cells were quantified and expressed as percentage of total number of cells (n = 3 independent experiments). Results are shown ±SEM.

Figure 3

qPCR Profiling of CHIR-Treated hESCs Highlights a Clear Progression through Specified Mesodermal Subtypes, Muscle Progenitor Stages, and Committed Myogenic Cultures hESCs were differentiated as described in Figure 1. RNA was harvested at days 0, 2, 8, 25, 40, and 50 to be analyzed by qPCR for markers several of (A) early and paraxial mesoderm, (B) premyogenic mesoderm, and (C) committed or differentiated skeletal muscle. Results are expressed as fold change over day 0 (n = 3 independent experiments, ^∗p ≤ 0.05 versus day 0, ^∗∗p ≤ 0.05 versus control). Results are shown ±SEM.

Figure 4

Low Levels of Neural Crest Transcripts Were Present in CHIR-Treated Cultures hESCs were differentiated, harvested, and analyzed as described above for markers of (A) pluripotency and neural ectoderm markers, (B) differentiated neural and traditional cardiac markers, (C) specific markers of mature cardiomyocytes, and (D) patterned dermomyotome genes (n = 3 independent experiments, ^∗p ≤ 0.05 versus day 0, ^∗∗p ≤ 0.05 versus control). In (C) RNA from serum-differentiated hESCs and human heart (HH) cardiomyocyte-positive controls were obtained (n = 1). All hESC fold changes (black bars) were plotted on the primary y axis, and HH fold change (gray bar) was plotted on the secondary y axis. Results are shown ±SEM.

Figure 5

CHIR Enhances Skeletal Myogenesis in a 15 Day Differentiation of mESCs mESCs were aggregated in differentiation medium for 2 days prior to treatment. CHIR, BMP4/INHBA, or vehicle control was applied from days 2 to 4. (A and B) Immunofluorescent staining was carried out at day 15 with antibodies against (A) PAX3 and PAX7 or (B) MYH (scale bar, 20 μm). (C) RNA was harvested at days 0 and 2 (prior to CHIR treatment) and days 4, 6, and 15 (following treatment). Samples were analyzed by qPCR for select markers of skeletal and cardiomyogenesis. Results are expressed relative to the maximum fold change over day 0 (n = 3 independent experiments, ^∗p ≤ 0.05 versus day 0, ^∗∗p ≤ 0.05 versus day 0 and control, ^∗∗∗p ≤ 0.05 BMP4/INHBA versus CHIR). Results are shown ±SEM.