Analysis of iPS phenotype and hematopoietic potential. iPS cells were induced from adult human dermal fibroblasts as described, and ES-like phenotype was confirmed using markers specific to pluripotent stem cells. (A) Staining for TRA-1-81 (green) and SSEA4 (red) (original magnification ×10). (B) iPS colonies are stained for TRA-1-81 (red) and Nanog (green); original magnification, ×20. In all images, nuclear staining is with 4,6-diamidino-2-phenylindole (blue). After a 3-stage differentiation protocol, cell populations were further differentiated toward the hematopoietic lineage using complete methocult H4434; and after 14 days, we see the appearance of CFUs representative of the myeloid lineages. (C) CFU hematopoietic colonies formed from 3 separate iPS cell lines: c15, c18, and c19 (mean ± SEM). These cells were harvested and subsequently stained for CD13 (phycoerythrin) and CD235 (APC) before analysis by FACS (D) where we see distinct populations representing erythroid (CD13⁻CD235⁺) and monocytic lineages (CD13⁺CD235⁻) similar to that observed in MNC (data not shown). iPS-derived CD34⁺ cells isolated by MACS after differentiation were expanded in EGM-2 media and assessed for their endothelial function. (E) We demonstrate the ability of these endothelial cells to form tubules when cocultured with human dermal fibroblasts in EGM-2 media for 14 days, which are subsequently stained for CD31 (original magnification, ×10). All images were captured with a Nikon TE300 inverted fluorescent microscope using Hamamatsu Photonics CCD camera and processed with SimplePCI software (Hamamatsu Version 6.6) (Digital Pixel). Demonstration of B-cell lymphopoiesis from human iPS cells. To address whether human iPS cells are able to contribute to the lymphoid compartment, we cultured iPS cells with OP9 stroma for 10 days and cells were stained for CD34 and CD43 and assessed this by FACS. (F) We show (left) isotype and (middle) MNC controls for CD34⁺ and CD43⁺ staining with (right) representing populations observed from iPS cells differentiated on OP9. Next, we isolated these CD34⁺ cells by MACS and cultured on mouse MS-5 stroma for a further 21 days. We then isolated CD45⁺ cells, again by MACS separation, and stained these for CD45 and CD19. (G) We show (left) isotype control, (middle) staining for UCB-derived CD34⁺ cells cultured on MS-5 stroma for 21 days, and (right) staining for iPS-derived CD34⁺ cells also cultured under conditions to promote B-cell lymphopoiesis. (H) We show that a subpopulation of the CD45⁺ cells from human iPS sources stains double-positive for (left) CD45⁺ and CD10⁺ but negative for (middle) CD5 or (right) IgM. This experiment was repeated with 3 cell lines (c15, c18, and c19), thus representing 3 biologic replicates where CD19⁺CD45⁺ cells represent 4.67% ± 2.07% (mean ± SEM). B-cell formation was also performed in triplicate with the c18 line, which performed best previously, and here CD19⁺CD45⁺ cells represented 7.38% ± 0.86% (mean ± SEM; n = 3). With mRNA isolated from CD45⁺ cells from 3 separate iPS cell lines (c15, c18, and c19), we show (I) that these differentiated cells (D) are positive for B-cell specific transcripts, such as Pax5, IL7αR, λ-like, and VpreB receptor compared with their undifferentiated (U) counterparts. CD45⁺ cells from iPS cell line c18.8, which yielded the highest number of B cells by FACS, also had the highest levels B-cell specific transcripts when assessed by reverse-transcribed PCR. (J) We analyzed VDJ rearrangement by PCR amplification of genomic DNA, using primers specific for D-J_H recombination. In UCB-derived MNCs (control), we observe clonal rearrangements representative of D_N, D_XP, D_LR, and D_A to J_H recombinations; also shown is a positive clonal control for D_NJ rearrangement (DJ clonal control). Whereas NHDF and iPS cell genomic DNA are shown to be negative for these rearrangements, DNA from CD45⁺ cells from OP9/MS5-differentiated iPS cells (c18.8 CD45⁺) are positive for D_N, D_XP, and D_A rearrangements, although not D_LR. Amplicons were sequenced and confirmed to map to the IgH locus of the human genome. Full V_HJ_H rearrangement was not observed in any of the iPS-derived CD45⁺ populations assessed.