Human ESCs (H9 and HSF1) and iPSCs [iPS(IMR90) and iHUF4] were differentiated as adherent cultures with 20% fetal bovine serum medium supplemented with BMP-4, -7, and -8b. (A) Morphology of undifferentiated and differentiated cells; after 14 days of differentiation, cultures appeared confluent, with cell morphology distinct from undifferentiated cells for all cell lines. (B) Western blot analysis of germ cell-specific proteins VASA and DAZL in undifferentiated cells and cells differentiated for 7 and 14 days. Increased expression of both VASA and DAZL was observed for all cell lines with differentiation, with the level of expression being similar at day 7 and day 14 time points between all the cell lines. Notably, a low level of VASA and DAZL protein expression was detected for undifferentiated iHUF4 cells and VASA expression in one sample of undifferentiated iPS(IMR90) cells. GAPDH is shown as a loading control and 293T cells are used as a negative control. Two independent samples are shown for each time point for all cell lines. (C) The expression and cytoplasmic localization of the VASA protein was also detected by immunostaining for all cell lines after differentiation. In addition, occasional cells in undifferentiated cultures were stained positive for all cell lines. Representative images are shown for iHUF4 and iPS(IMR90) cells; green VASA, blue DAPI, merged image and lower magnification image showing the surrounding cells. Scale bars: 200 µm (A); 50 µm (C).

Gene expression analysis for undifferentiated cells and cells differentiated with BMPs for 7 and 14 days. The combined expression of endogenous and exogenous (derived from viral vectors) pluripotency markers OCT3/4 and NANOG was significantly different between all undifferentiated cell lines; iPSCs had higher expression relative to hESCs. At day 7 of differentiation, iPS(IMR90) and HSF1 had higher expression of OCT3/4 and iPS(IMR90) had higher expression of NANOG compared with other cell lines; however, at day 14, the levels decreased to the similar levels between all cell lines. As a further indication of differentiation, the expression of KRT7 (trophectodermal marker), GATA6 (endodermal marker), ACTC (mesodermal marker) and NCAM1 (ectodermal marker) increased similarly for all cell lines upon differentiation, although at day 7 the expression of GATA6 was significantly higher in HSF1 cells relative to iPS(IMR90) cells and at day 14 the expression of KRT7 and NCAM1 was higher in iHUF4 cells relative to H9 or HSF1 cells, respectively. In undifferentiated cells, iPS(IMR90) had significantly higher expression of KRT7 relative to H9 and iHUF4, and GATA6 relative to all other cell lines. The expression of ACTC was significantly higher in both undifferentiated iPSCs relative to hESCs, but the expression of NCAM1 was higher in H9 cells relative to HSF1 and iPS(IMR90) and in iHUF4 cells relative to HSF1 cells. The expression of early germ cell markers IFITM1, PELOTA and PRDM1A indicate the potential of all cell lines to differentiate toward the germ cell lineage, with similar expression levels between hESCs and iPSCs. However, the expression of IFITM1 at day 7 and PELOTA in undifferentiated cells was significantly higher in iPSCs relative to hESCs. In addition, the expression of IFITM1 in undifferentiated iPS(IMR90) cells was significantly higher relative to other cell lines, and in HSF1 relative to H9 and iHUF4 cells. The expression on PRDM1A was significantly higher at day 7 in HSF1 cells relative to H9 and iPS(IMR90) cells. Values are normalized to expression levels of two housekeeping genes, GAPDH and RPLPO, and results are shown as fold difference relative to undifferentiated H9 cells using the log scale. Statistical testing was performed comparing the values of different cell lines within a time group; undiff, day 7, and 14, for each gene separately. Error bars indicate SD between three technical replicates. *P < 0.05, one-way ANOVA.

Human iPSCs and ESCs were transduced with a lentiviral vector with VASA promoter driving eGFP expression. Cells were differentiated for 7 days with BMPs, and the GFP-positive cells were analyzed and sorted by flow cytometry. (A) FACS analysis of cells after 7 days of differentiation. Gating for cell sorting was setup using eGFP and PE parameters; eGFP-positive and PE-negative cells were collected for further analysis, excluding the double-positive autofluorescence cells that are located on the diagonal axis. (B) The percentage of GFP-positive cells for transduced undifferentiated cells and after 7 days of differentiation. The percentage of GFP-positive cells was ≤1% for all undifferentiated cell lines; however, after 7 days of differentiation, hESCs had ≥2% positive cells and iPSCs had ∼5% GFP-positive cells. (C) Immunostaining for the VASA protein in sorted GFP-positive and GFP-negative cells. All GFP-positive cells were positive for VASA staining, whereas no VASA-positive cells were found from the GFP-negative cell population. Representative images are shown for differentiated viPS(IMR90) and viHUF4 cells. (D) Gene expression analysis for GFP-positive and GFP-negative differentiated viPS(IMR90) and viHUF4 cells. The expression of germ cell markers DMC1, GCNF, IFITM1, PRDM1A, STELLAR and PELOTA was significantly higher in GFP-positive cell population compared with GFP-negative population, except for viPS(IMR90) for DMC1 and PRDM1A. In addition, the expression of OCT3/4 was significantly higher in GFP-positive population for viHUF4 cells. The expression of NANOG, GATA6, ACTC, NCAM1 and KRT7 was only detected in GFP-negative populations. Values are normalized to the expression levels of two housekeeping genes, GAPDH and RPLPO, and results are shown as fold difference relative to GFP-negative H9 cells. Error bars indicate SD between three technical replicates. *P < 0.05, unpaired t-test. Scale bar 50 µm (C).

Meiotic progression of iPSCs and hESCs. Meiotic spreads were prepared from undifferentiated cells, cells differentiated with supplementation of BMPs up to 14 days and cells with exogenous overexpression of human DAZ gene family members and then differentiated with BMPs up to 14 days. The meiotic spreads were analyzed by immunostaining against SCP3 (red), CENP-A (green) and DAPI (blue). (A) Meiotic spreads were classified as punctate or elongated SCP3 staining patterns, corresponding to the early leptotene stage (punctate) and the later zygotene, pachytene and diplotene stages (elongated) of meiotic prophase I. Representative images are shown for both groups. (B) Meiotic spreads from dH9, dHSF1, diPS(IMR90) and diHUF4 cell lines were classified as punctate or elongated SCP3 staining patterns and quantified by counting 800 cells from differentiated samples and 200 cells from undifferentiated samples. With overexpression, punctate and elongated SCP3 staining was detected in all cell lines; diPS(IMR90) cell line had both punctate and elongated staining percentages similar to dHSF1, diHUF4 cell line had a similar percentage of elongated staining percentage to dH9 cells, but a high percentage of punctate staining compared with all cell lines. Differentiated cells without overexpression had no elongated staining in any of the cell lines. However, rare cells with elongated staining were detected for both iPSC lines in undifferentiated cell cultures (W). When only morphologically good colonies were manually picked from undifferentiated cultures (G), no elongated staining was seen and the percentage of punctate staining decreased compared with whole culture (W). (C) Representative images are shown for undifferentiated and overexpressed cells. Some cells with punctate staining were observed for the undifferentiated H9 and HSF1 cells, for iPS(IMR90) and iHUF4 cell lines elongated SCP3 staining was also detected. Elongated staining was observed for all cell lines after overexpression. Scale bar 10 µm (A and C).

DNA content analysis for diPS(IMR90), diHUF4, dH9 and dHSF1 cells transduced with the DAZ gene family overexpression factors and differentiated with BMPs for 14 days. (A) Human semen was used as a positive control to identify the location of haploid cells (1N) by flow cytometer. A distinct peak of 1N population was not detected for any of the cell lines, but around 1.5% of cells were collected using the gating based on the human semen sample. Cells from the non-transduced HSF1 sample were also collected. (B) Fluorescent in situ hybridization for chromosome 16 in the sorted 1N cell population. Cells with one chromosome 16 are shown for each cell line. HSF1 control cells with no transduction did not have cells haploid for chromosome 16. (C) Cells with one chromosome 16 were identified and counted from 1N sorted cell populations; 13–21% of transduced cells had one chromosome 16. (D) A similar ACROSIN staining pattern seen in human spermatids was observed in some of the 1N sorted cells for all cell lines. Some positive cells were also observed in the non-transduced HSF1 cells. Undifferentiated H9 cells are shown as a negative control with low background staining. (E) Cells with ACROSIN-positive staining were identified and counted for 1N sorted cell samples. The percentage of positive cells varied between 35 and 72%. Scale bar 5 µm (B), 5 µm (and 40 µm for H9 undiff) (D).