Figure 4.

Human GFRA2 Marks CPs from Human ESC Cultures

(A) qPCR analyses for human GFRA2, NKX2-5, TBX5, and MYL2. Note the peak of human GFRA2 expression is just before human embryoid bodies start to beat. Bar graph represents biological triplicates with technical duplicates as mean ± SEM.

(B) FACS isolation of human hGFRA2⁺/hPDGFRA⁺ (G⁺P⁺) cells. ICC analyses for TNNT2 demonstrate that isolated G⁺P⁺ population is highly cardiomyogenic. Scale bar, 100 μm.

(C and D) Quantitative analyses by Flo for TNNT2 show most of G⁺P⁺ cells (96.5% ± 1.1%) differentiated into cardiomyocytes. Bar graph represents mean ± SEM. n = 3.

(E) qPCR analyses for GFRA2, NKX2-5, ISL1, TBX5, and HCN4 in FACS-isolated populations. All cardiac marker gene expressions were significantly higher in the G⁺P⁺ population when compared to the G⁻P⁻ population (^∗p < 0.05, Student’s t test). Data are representative of biological triplicates with technical duplicates as mean ± SEM.

(F) Flow cytometry of human ESCs at differentiation day 4. hGFRA2⁺ cells (shown as green dots) were mostly included by hPDGFRA⁺/hKDR^low+/hKIT^neg population, which is reported as multipotent CPs (Yang et al., 2008). 15%–20% of hPDGFRA⁺/hKDR^low+/hKIT^neg cells were hGFRA2⁺.

(G) RT-PCR analyses of single-cell-derived colonies for each lineage marker gene. A single hGFRA2⁺/hKDR^low+/hPDGFRA⁺ cell at day 4 was clonally sorted and cultured on mouse embryonic fibroblasts (MEFs) for 2 weeks. RT-PCR of hTNNT2, hPECAM1, and hMYH11 represent the existence of cardiomyocytes, endothelial cells, and smooth muscle cells among the cells derived from a single cell, respectively. Note the existence of multiple lineages, which indicates the multipotency.

See also Figure S3 and Movie S2.