To evaluate whether the genetic correction in Eβ-iPSCs could restore the

HBB

expression, hematopoietic differentiation of the wild-type iPSCs (HDF-iPSCs), the Eβ-iPSC2 cells and the corrected clones (C22, C46, C134, C137 and C258) was performed in a feeder-free condition (Fig.

3a

) at the following stages of hematopoietic development: mesoderm progenitor, hematovascular specification, endothelial hematopoietic transition and hematopoietic progenitor cells. At days 5–6 of culture, the differentiated cells appeared to be a monolayer of endothelial-like cells, which later formed three-dimensional structures, observed from day 8 onward. The nonadherent cells started to emerge from both monolayer and three-dimensional structures on days 8–12 (Fig.

3b

). During hematopoietic differentiation, the Eβ-iPSC2 cells showed impaired hematopoietic differentiation as indicated by a lower number of cells expressing hematopoietic progenitor and erythroid markers, CD43 and CD71, when compared to the HDF-iPSCs in the adherent cell population. In contrast to the HDF-iPSCs, which could give rise to the nonadherent cell population that highly expressed CD43, CD71 and CD235, the Eβ-iPSC2 cells produced a very low number of nonadherent cells, which were mainly nonviable. After genetic correction, all of the five corrected clones were able to differentiate into hematopoietic progenitor cells which expressed CD34 and CD43, and erythroid markers CD71 and CD235a at comparable levels to those of the HDF-iPSCs in both adherent and nonadherent cell population (Fig.

3c, d

). We also examined the gene expression profile of these corrected cells during hematopoietic specification by quantitative real-time PCR. All corrected clones expressed

SOX17

and

RUNX1

, which play an important role in blood formation from hemogenic endothelium, and

GATA1

and

KLF1

, which are erythroid-specific markers (Additional file

5

: Figure S4a). We harvested the differentiated floating cells from the HDF-iPSCs and the corrected clones on day 12 and seeded them onto methylcellulose plates. After 2 weeks, the HDF-iPSCs and three of the corrected clones (C22, C46 and C137) gave rise to all types of colonies, mainly CFU-E and CFU-GM, confirming the functional characteristic of hematopoietic progenitor cells. However, two of the corrected clones (C134 and C258) could only give rise to CFU-GM and a small number of CFU-E (Fig.

3e, f

). In contrast to the corrected cells, the Eβ-iPSC2 cells could not produce any CFU colonies. The BFU-E obtained from the corrected clones expressed high levels of fetal gamma hemoglobin (

HBG

) and low levels of adult beta hemoglobin (

HBB

) transcripts, when examined by qRT-PCR (Additional file

5

: Figure S4b).

Fig. 3

Hematopoietic differentiation of iPSCs using the feeder-free system. a Schematic of the feeder-free hematopoietic differentiation protocol used in this study. b Morphological changes of the corrected C46 cells during hematopoietic differentiation. c, d Numbers of CD34, CD43, CD235a and CD71-expressing cells in adherent and nonadherent systems at day 12 of differentiation. e Representative images of CFU from HDF-iPSCs. Differentiated cells at day 12 were harvested and seeded in MethoCult media. f Numbers of CFU colonies counted on day 14 of culture in MethoCult media. Data obtained from two independent experiments. Scale bars = 200 μm. IMDM Iscove modified Dulbecco medium, BMP-4 bone morphogenetic protein 4, VEGF vascular endothelial growth factor, KOSR knockout serum replacement, bFGF basic fibroblast growth factor, BSA bovine serum albumin, SCF stem cell factor, TPO thrombopoietin, IL interleukin, FICZ 6-formylindolo[3,2-b]carbazole, HDF human dermal fibroblasts, iPSC induced pluripotent stem cell, Eβ-iPSC2 iPSC lines derived from a patient with HbE/βthalassemia, CFU-E colony-forming unit erythroid, EPO erythropoietin, BFU-E burst-forming unit erythroid, GM granulocyte, macrophage, GEMM granulocyte, erythrocyte, macrophage, megakaryocyte

Since the Eβ-iPSC2 cells seemed to be refractory to the hematopoietic differentiation protocol under the feeder-free condition, we turned to the OP9 coculture system for hematopoietic differentiation followed by an erythroid liquid culture (Fig.

4a

) [

20

]. The supportive OP9 stromal cells have been shown to efficiently induce hematopoietic differentiation [

21

]. We selected the corrected C46 cells, which differentiated well under the feeder-free condition and contained no off-target mutation, for comparison with the Eβ-iPSC2 cells. Small clumps of iPSCs were seeded onto overgrown OP9 cells and cultured for 6 days. In contrast to the feeder-free hematopoietic differentiation system, both the Eβ-iPSC2 cells and the corrected C46 cells were able to differentiate into sac-like structures (Fig.

4b

). We isolated CD34

+

cells from the differentiated cells on day 6 of the OP9 coculture system and further expanded erythroid cells using the three-stage culture system [

19

]. Upon erythroid culture, both the Eβ-iPSC2 cells and the corrected C46 cells gradually changed their morphology from that representing proerythroblasts/basophilic erythroblasts on day 13 of culture to that representing polychromatic/orthochromatic erythroblasts on day 23 and finally became orthochromatic erythroblasts/reticulocytes on day 29 of culture. Analysis of gene expression during the erythroid liquid culture demonstrated that both differentiated Eβ-iPSC2 and the corrected C46 cells at day 19 of differentiation (when the morphological stages are equivalent to those of day 13 erythroid cells derived from peripheral blood progenitors) expressed lower levels of erythroid-associated transcription factors

KLF1

and

BCL11A

as compared to the cultured erythroblasts from peripheral blood CD34

+

cells (Fig.

4c

). We harvested the differentiated cells at day 30 and analyzed hemoglobin protein expression. Both differentiated Eβ-iPSC2 and the corrected C46 cells expressed similar levels of beta hemoglobin and alpha hemoglobin proteins, indicating successful hematopoietic differentiation under the OP9 coculture system (Fig.

4d

).

Fig. 4

Hematopoietic differentiation of iPSCs using the OP9 coculture system and erythroid liquid culture. a Schematic of hematopoietic differentiation protocol used in this study. b Morphological changes of the Eβ-iPSC2 cells and the corrected C46 cells during hematopoietic differentiation on day 6 of OP9 coculture (scale bar = 500 μm), and Wright’s staining on days 13, 23 and 29 of differentiation. c Quantitative RT-PCR analysis of erythroid-associated transcription factors at day 19 of differentiation (equivalent to day 13 of erythroid liquid culture) of the Eβ-iPSC2 cells and the corrected C46 cells as compared to peripheral blood CD34+ cell-derived erythroblasts at day 13 (PB). Data presented as mean ± SD of triplicate samples from a representative experiment. d Western blot analysis of alpha and beta hemoglobin expression of the Eβ-iPSC2 cells and the corrected C46 cells at day 30 of differentiation as compared to peripheral blood CD34+ cell-derived erythroblasts at day 24 of erythroid liquid culture. MEM minimal essential medium, IMDM Iscove modified Dulbecco medium, FBS fetal bovine serum, SCF stem cell factor, IL interleukin, EPO erythropoietin, Eβ-iPSC2 iPSC lines derived from a patient with HbE/βthalassemia, Hb hemoglobin