One-step genetic correction of hemoglobin E/beta-thalassemia patient-derived iPSCs by the CRISPR/Cas9 system

Background Thalassemia is the most common genetic disease worldwide; those with severe disease require lifelong blood transfusion and iron chelation therapy. The definitive cure for thalassemia is allogeneic hematopoietic stem cell transplantation, which is limited due to lack of HLA-matched donors and the risk of post-transplant complications. Induced pluripotent stem cell (iPSC) technology offers prospects for autologous cell-based therapy which could avoid the immunological problems. We now report genetic correction of the beta hemoglobin (HBB) gene in iPSCs derived from a patient with a double heterozygote for hemoglobin E and β-thalassemia (HbE/β-thalassemia), the most common thalassemia syndrome in Thailand and Southeast Asia. Methods We used the CRISPR/Cas9 system to target the hemoglobin E mutation from one allele of the HBB gene by homology-directed repair with a single-stranded DNA oligonucleotide template. DNA sequences of the corrected iPSCs were validated by Sanger sequencing. The corrected clones were differentiated into hematopoietic progenitor and erythroid cells to confirm their multilineage differentiation potential and hemoglobin expression. Results The hemoglobin E mutation of HbE/β-thalassemia iPSCs was seamlessly corrected by the CRISPR/Cas9 system. The corrected clones were differentiated into hematopoietic progenitor cells under feeder-free and OP9 coculture systems. These progenitor cells were further expanded in erythroid liquid culture system and developed into erythroid cells that expressed mature HBB gene and HBB protein. Conclusions Our study provides a strategy to correct hemoglobin E mutation in one step and these corrected iPSCs can be differentiated into hematopoietic stem cells to be used for autologous transplantation in patients with HbE/β-thalassemia in the future. Electronic supplementary material The online version of this article (10.1186/s13287-018-0779-3) contains supplementary material, which is available to authorized users.


Background
HbE/β-thalassemia, a double heterozygosity of hemoglobin E (HbE) and β-thalassemia, is the most common thalassemic syndrome found in adults in Southeast Asia. The clinical manifestations are heterogeneous; at one end the mutation may be very mild, whereas at the other end it is very severe similar to homozygous β-thalassemia or thalassemia major. In HbE/β-thalassemia, one allele (β 0 ) produces no β-globin chain and the other allele (β E ) produces a HbE globin chain resulting from nucleotide substitution at codon 26 (GAG → AAG, glutamic acid to lysine) [1]. Previous reports have shown the success of genetic correction of a β-globin gene in β-thalassemia-specific induced pluripotent stem cells (iPSCs) using a lentiviral vector [2,3] or homologous recombination [4]. However, the lentiviral gene therapy results in random integration of the functional gene into the genome, leading to undesired mutations. It is therefore necessary to screen for the clones that integrate the transgene to genomic safe harbor sites. On the other hand, the classic homologous recombination is very inefficient, especially in human pluripotent stem cells [5]. The more efficient genome editing technologies using customengineered nucleases, zinc-finger nucleases (ZFNs) and transcription activator-like effector nucleases (TALENs), have been used to correct β-thalassemia iPSCs [6,7].
The RNA-guided clustered regularly interspaced short palindromic repeat (CRISPR)/Cas9 system has been used recently to correct HBB mutation in iPSCs derived from β-thalassemia [8][9][10][11] and sickle cell disease patients [12]. However, these studies relied on a donor plasmid containing a wild-type HBB gene and an antibiotic selection cassette for enrichment, thereby requiring subsequent excision and clonal selection steps. To overcome these limitations, a single-stranded DNA oligonucleotide (ssODN) donor template can be used to provide seamless correction [13,14]. In this study, we used the CRISPR/Cas9 system and the ssODN donor template to efficiently correct the HbE mutation in iPSCs derived from a patient with HbE/β-thalassemia, resulting in the corrected iPSCs, which is a β-thalassemia heterozygote. The corrected iPSCs are capable of differentiating into hematopoietic stem cells, which can be used for autologous transplantation to the patient in the future. In addition, our study further demonstrates that these cells can differentiate in vitro to reticulocytes, which can be developed for therapeutic use.

Sample collection and generation of induced pluripotent stem cells
The study was approved by the Siriraj Institutional Review Board (no. Si248/2011), in accordance with the Helsinki Declaration of 1975. All patients were provided with an explanation and with a participant information sheet and signed the informed consent. Skin biopsies were collected from HbE/β-thalassemia patients for further mutation analysis and isolation of fibroblasts. Briefly, the skin specimens were washed with sterile phosphate buffered saline (PBS) containing 25 U/ml penicillin, 25 μg/ml streptomycin, chopped into small pieces of 1 mm 3 and transferred into a T-25 tissue culture flask containing DMEM supplemented with 10% fetal bovine serum (FBS) (Lonza, Switzerland), 2 mM GlutaMAX™ and 25 U/ml penicillin, 25 μg/ml streptomycin. Fibroblasts were subcultured once every 5 days or whenever they reached 80% confluency by incubation with 0.25% Trypsin for 2 min. Generation and characterization of Eβ-iPSCs from a HbE/β-thalassemic patient's HDFs were performed as described previously [15]. iPSCs were maintained in mTeSR™1 medium (StemCell Technologies, Canada) on Matrigel™-coated (BD Bioscience, USA) plates and subcultured using 1 mg/ml Dispase (StemCell Technologies) according to the manufacturer's instructions.

Gene expression analysis
Total RNA was obtained using TRIzol® reagent (Invitrogen). cDNA was prepared using 2 μg of RNA and reverse-transcribed using the SuperScript III First-Strand Synthesis System and Oligo (dT) primers (Invitrogen). PCR analysis of pluripotent genes was performed on a T100™ Thermal Cycler (Bio-Rad, USA) using Platinum Taq DNA polymerase (Invitrogen). Primer sequences are presented in Additional file 1: Table S1. For quantitative RT-PCR (qRT-PCR) analysis, primers and probes were designed using the Universal Probe Library Assay Design Center (Roche Diagnostics). The qRT-PCR analysis was carried out on the CFX96™ Real-Time PCR detection system (Bio-Rad). Data were normalized with a housekeeping gene, GAPDH, and expressions were plotted against the undifferentiated normal iPSCs. Primer sequences are presented in Additional file 1: Table S2.

Multiplex PCR analysis for hemoglobin E
Genomic DNA was isolated using the Gentra® Puregene® Cell Kit (Qiagen) according to the manufacturer's instructions. HbE mutation was detected by multiplex PCR. The PCR reaction consisted of 1.5 U of DNA polymerase (Platinum Taq; Invitrogen), 1× PCR buffer, 1.5 mM MgCl 2 , 0.1 mM dNTPs, 0.4 μM HbE-Fc primer, 0.4 μM HbE-Rc primer, 0.5 μM HbE-Rn primer, 0.5 μM HbE-Fm primer and 2 μl of DNA sample in a total volume of 30 μl. The PCR was performed after an initial denaturation at 95°C for 15 min followed by 30 cycles of denaturation (94°C for 45 s), annealing (68°C for 45 s) and extension (72°C for 1 min), and a final extension step (72°C for 7 min). Primer sequences are presented in Additional file 1: Table S3.
Generation of CRISPR/Cas plasmid, single-guide RNA and single-stranded DNA oligonucleotide template CRISPR/Cas plasmids pSpCas9(BB)-2A-GFP (PX458) and pSpCas9(BB)-2A-Puro (PX459) were obtained from Addgene (Cambridge, USA). Seven gRNAs targeting the HbE mutation were designed and their potential offtarget sites were identified using the Crispr Design Tool (http://crispr.mit.edu/). Top and bottom strands of each gRNA were annealed, phosphorylated and cloned into the BbsI site of the PX459 plasmid according to a protocol published previously [16]. The ssODN donor template for HbE correction was designed to have homology arms of 90 nucleotides on either side of the point mutation (a total of 181 nucleotides). The gRNA sequences and the ssODN template are presented in Additional file 1: Tables S4 and S5.

Gene targeting of the HbE mutation in Eβ-iPSCs
Transfection of Eβ-iPSCs was performed using Amaxa 4D-nucleofector as described previously [17]. Briefly, 5 × 10 5 Eβ-iPSCs were resuspended in nucleofection mixture containing 20 μl of P3 Primary Cell Solution (Lonza) and 2 μg of DNA. The mixture was transferred to Nucleocuvette strips and nucleofection was performed using the CB-150 program according to the manufacturer's instructions. The PX459 plasmid was used for gRNA construction and the PX458 was used as a positive control. The transfected iPSCs were plated onto Matrigel™-coated 24-well plates and cultured in mTESR1 medium supplemented with 10 μM Y-27632. At 24 hours post nucleofection, the PX458 transfected cells were subjected to flow cytometry analysis and the percentage of GFP + cells was analyzed by FACS-Calibur™ (BD Biosciences). To determine the gRNA efficiency, the genomic DNA of PX459-gRNA transfected cells was extracted after being cultured for 5 days. The PCR product of 306 bp around the target site was amplified and digested with T7 endonuclease I (T7EI) enzyme (New England Biolabs) according to the manufacturer's instructions.
To target the HbE mutation in Eβ-iPSCs, 200 pmol of the ssODN template was cotransfected with 2 μg of the PX459-gRNA with the highest specificity to the HbE site, as determined by T7EI assay. At 3 days post nucleofection, we performed a clonal isolation by limiting dilution. Briefly, transfected iPSCs were pretreated with SMC4 (Corning) prior to dissociation with accutase (Merck) and plated into 96-well plates at a density of 20 cells/96-well plate in mTESR1 medium supplemented with SMC4 for 8 days. Single colonies were picked and screened for the HbE mutation. The DNA sequence of HbE negative clones was further confirmed by direct sequencing. For off-target analysis, five potential off-target sites were amplified by PCR. In brief, genomic DNA was extracted using QuickExtract DNA Extraction Solution (Epicentric) according to the manufacturer's instructions. The PCR reaction consisted of 0.2 U of Q5 High-Fidelity DNA polymerase (New England Biolabs), 1× Q5 PCR buffer with MgCl 2 , 0.2 mM dNTPs, 0.5 μM of each primer and 2 μl of DNA sample in a total volume of 50 μl. The PCR was performed with an initial denaturation at 98°C for 30 s followed by 35 cycles of denaturation (98°C for 10 s), annealing (68°C for 30 s) and extension (72°C for 30 s), followed by a final extension step (72°C for 2 min). Primer sequences are presented in Additional file 1: Table S6. DNA sequences were validated by Sanger sequencing.
Alternatively, iPSCs were also differentiated on OP9 mouse stromal cells. Briefly, small clumps of iPSCs were seeded onto overgrown OP9 cells in differentiation medium containing α-MEM (Invitrogen), 10% defined FBS (Hyclone), 100 mM MTG and 25 U/ml penicillin, 25 μg/ml streptomycin for 5 days. A half medium change was performed on day 3. iPSCs were harvested on day 6 using type IV collagenase (Invitrogen) for 20 min, followed by 0.25% trypsin for 15 min. CD34 + cells were isolated from differentiated cells using EasySep (StemCell Technologies) according to the manufacturer's instructions. To promote erythroid maturation, the purified CD34 + cells were cultured in a three-stage culture system according to Griffiths et al. [19]. Briefly, cells were seeded in stage 1 medium containing basic medium (IMDM (Biochrom), 3% AB serum, 2% defined FBS, 10 μg/ml insulin, 3 U/ml heparin, 3 U/ml EPO, 200 μg/ml transferrin and 100 U/ml penicillin/streptomycin) supplemented with 10 ng/ml SCF and 1 ng/ml IL-3 for 8 days, stage 2 medium containing basic medium supplemented with 10 ng/ml SCF for 3 days and stage 3 medium containing basic medium supplemented with extra transferrin to a final concentration of 500 μg/ml for 13 days. Morphological analysis was performed by Wright's staining at the indicated time points. The differentiated cells (1.5 × 10 4 -3 × 10 4 ) were spun onto glass slides at 1000 rpm for 5 min using a Cytospin™ 4 Cytocentrifuge (Thermo Scientific) and stained with Wright-Giemsa (Merck). Images were taken with a light microscope (Carl Zeiss, Axio Star Plus).

Western blotting analysis
Differentiated iPSCs at day 24 of erythroid liquid culture were harvested, washed with PBS and lysed in lysis buffer (Cell Signaling) for 1 hour on ice. The protein concentration was measured by the Pierce™ BCA Protein Assay Kit (Thermo Scientific). Samples were loaded onto 18% SDS-PAGE and transferred to a PVDF membrane. The membrane was blocked with 10% skimmed milk in Tris-buffered saline (TBS) with 0.1% Tween-20 (TBS-T) for 1 hour at room temperature and probed with the primary antibodies, anti-hemoglobin alpha and antihemoglobin beta (Santa Cruz) overnight at 4°C. Subsequently, the membrane was incubated with the HRP-conjugated secondary antibody for 1 hour at room temperature. After washing with TBS-T, the membrane was incubated with ECL substrate (Thermo Scientific) and the results were visualized by an ImageQuant LAS 4010 Biomolecular imager (GE Healthcare).

Generation and characterization of HbE/β-thalassemia patient-specific iPSCs
We obtained human dermal fibroblasts (HDFs) from a patient with HbE/β-thalassemia who had a 4-bp deletion (-TCTT) at codon 41/42 in one allele, and a point mutation (G → A) at codon 26 resulting in abnormal hemoglobin E (HbE) production in the other, and generated iPSCs from these HbE/β-thalassemia patient's HDFs. Several iPSC lines, designated Eβ-iPSCs, were obtained. Of these, two iPSC lines (Eβ-iPSC1 and Eβ-iPSC2) were expanded and characterized. To confirm the HbE mutation, multiplex PCR analysis was performed with the genomic DNA of the Eβ-iPSC1 and the Eβ-iPSC2 cells. Both Eβ-iPSCs had a product size of 529 bp, which represents the HbE point mutation, similar to that of fibroblasts derived from the patient's skin (Eβ-HDFs) but not in the wild-type HDFs or iPSCs (Fig. 1a). Further characterization showed that both Eβ-iPSCs expressed pluripotent markers and were able to differentiate both in vitro and in vivo into cells/tissues of three embryonic germ layers. The Eβ-iPSCs exhibited a normal karyotype of 46, XY at passage 19 (Additional file 2: Figure S1). We then selected the Eβ-iPSC2 cells for further analyses.

Correction of HbE mutation in Eβ-iPSC2 cells using the CRISPR/Cas9 system
To correct the mutation in the Eβ-iPSC2 cells, guide RNAs (gRNAs) were designed to target the HbE mutation at codon 26 where nucleotide substitution occurs (G → A) (Fig. 1b). We first designed seven 20nucleotide (nt) or truncated (18-nt or 19-nt) gRNAs targeting regions close to the point mutation ( Fig. 1c and Additional file 1: Table S4). Each gRNA was constructed into a PX459 plasmid, which contains the Cas9 nuclease and puromycin selection cassettes. We used PX458, which has the same backbone as PX459 except for GFP expression, instead of the puromycin selection cassette, to determine the transfection efficiency. At 24 hours post nucleofection, the PX458 transfected cells were harvested for flow cytometry analysis. We obtained approximately 30% transfection efficiency in the Eβ-iPSC2 cells (Additional file 3: Figure S2). We then used the same conditions for transfecting PX459-gRNA into the Eβ-iPSC2 cells. We harvested the genomic DNA of the PX459-gRNA transfected pool at 5 days post nucleofection and examined for gRNA-Cas9 cleavage efficiency and specificity using T7EI assay. Of these seven gRNAs, gRNA1, which targeted DSB at two nucleotides upstream of the HbE point mutation, gave the highest efficiency and was then selected for further experiments (Fig. 2a).
For HbE correction, we cotransfected PX459-gRNA1 and the ssODN template. At 3 days post nucleofection, the transfected pool was harvested and seeded at a density of 20 cells/well of a 96-well plate. Eight to 10 days after plating, one to five colonies were observed in each well. Single colonies were individually picked for multiplex PCR analysis for HbE mutation. From a total of 312 individual clones screened, we estimated that 93 clones were transfected (according to 30% transfection efficiency with the GFP plasmid control). We obtained 23 HbE-negative clones from PCR screening (7.4% DSB efficiency) (Fig. 2b). Of these, 14 clones showed indels (4.5% nonhomologous end joining (NHEJ)) and nine clones showed successful seamless correction of HbE mutation (2.9% homology-directed repair (HDR)) as confirmed by direct sequencing (Fig. 2c, d). We selected five clones (C22, C46, C134, C137 and C258) for karyotyping analysis. These corrected clones exhibited a normal karyotype (Fig. 2e and Additional file 4: Figure S3) and three clones (C22, C46 and C137) were randomly expanded for off-target analysis.

Off-target analysis
We selected five potential off-target sites of gRNA1 including the HBD gene and other genes that have similar homology with the gRNA1-PAM sequence using NCBI BLAST. PCR amplification of these regions was performed and DNA sequences of the PCR products were validated by direct sequencing as compared to that of the Eβ-iPSC2 cells. We detected no off-target cleavage in the corrected C46 cells, whereas a point mutation in the HBD gene was observed in both of the corrected C22 and C137 cells (Fig. 2f ).

Generation of hematopoietic cells from the corrected iPSCs
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 threedimensional structures, observed from day 8 onward. The Red arrowheads show mutation sites; black arrowhead shows cleavage site of Cas9 by gRNA1; purple and red arrows indicate primer pairs for T7E1 assay and multiplex PCR for HbE detection, respectively. c Sequence of seven gRNAs designed to target the HbE mutation. Red "A" indicates point mutation in the Eβ-iPSC2 cells and gRNAs at HBB locus. HBB beta hemoglobin 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 After genetic correction, both alleles contained the right nucleotide "G". e Representative karyotype of the corrected C46 cells, which exhibited a normal karyotype (46, XY). f Potential off-target sites for gRNA1 as identified by BLAST search. Mismatch nucleotides are indicated in red. Eβ-iPSC2 iPSC lines derived from a patient with HbE/β-thalassemia, HbE hemoglobin E, HBD delta hemoglobin hemoglobin (HBB) transcripts, when examined by qRT-PCR (Additional file 5: Figure S4b).
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 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/β-thalassemia, CFU-E colony-forming unit erythroid, EPO erythropoietin, BFU-E burst-forming unit erythroid, GM granulocyte, macrophage, GEMM granulocyte, erythrocyte, macrophage, megakaryocyte 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).

Discussion
HbE is the hallmark of Southeast Asia. Double heterozygosity of HbE and β-thalassemia, namely HbE/β-thalassemia, can cause a severe clinical syndrome similar to homozygous β-thalassemia [22]. These severe patients manifest with anemia requiring red cell transfusion during the first year of life. The only curative therapy at present is hematopoietic stem cell transplantation. However, unavailability of HLA-matched donors and risk of transplantrelated morbidity, mortality as well as immunologic complications, especially graft failure and graft versus host disease (GVHD), have limited the use of allogeneic transplantation [23]. Recently, in vitro genetic correction of hematopoietic stem cells (HSCs) has been successfully reported; however, culturing and maintaining HSCs remain difficult [24].
Advances in iPSC technology offer promise for autologous cell-based therapy as it provides a stem cell source, which can be continuously expanded in vitro and is amendable to genetic manipulation before differentiation into functional HSCs. We performed an efficient onestep seamless genetic correction of iPSCs from a HbE/βthalassemic patient using the CRISPR/Cas9 system and the ssODN template. The genetic correction of the HbE mutation in one allele is easier than correcting the β- 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/β-thalassemia, Hb hemoglobin globin gene mutation in the other allele in which the mutations are heterogeneous [25]. Our approach does not require an antibiotic selection cassette that may interfere with expression of the corrected gene [6,9,12].
We designed gRNAs to recognize only HbE mutation and used the ssODN template possessing the correct nucleotide at the center of the oligo, which provides the highest targeting efficiency especially when the mutation site is less than 10 bp away from the cutting sites [26]. The ssODN template is easy to design and synthesize. It is footprint-free after genetic correction and has a high targeting efficiency. Our results show that transfection of PX459-gRNA1 with ssODN template resulted in a gene disruption frequency (DSB) of 7.4% and homologous gene targeting frequency (HDR) of 2.9%. Our transfection efficiency varied among iPSC lines, ranging from 30 to 60%. The number of screened clones could be further reduced if the transfection efficiency increased. To improve HDR efficiency, small molecule inhibitors may be used to suppress the nonhomologous end joining (NHEJ) pathway, thus facilitating the downstream screening process [27,28].
Off-target mutagenesis is a major safety concern since genetic modifications are permanent and will have devastating consequences if the mutations are to be at important sites [29]. We identified five potential offtarget sites of gRNA1 including the HBD gene, which has similar homology to HBB gene, and confirmed them by Sanger sequencing. Of the three corrected clones screened, one had no indel mutation whereas two harbored a point mutation. Interestingly, the presence of the point mutation in the delta hemoglobin (HBD) gene did not affect the hematopoietic differentiation potential when using the feeder-free system. A previous report on the genetic correction of a beta hemoglobin gene to treat sickle cell disease in human hematopoietic stem/progenitor cells using ZFNs showed off-target cleavage in the highly homologous, but functionally dispensable, HBD gene [30]. Measures to improve targeting specificity of Cas9 include the use of truncated gRNA with shorter regions of target complementarity [31], or modifying the Cas9 components so that two gRNA/Cas9 complexes are required to cleave DNA. The latter can be achieved via nickases that induce single-strand breaks [32,33] or single-guide RNA (sgRNA)-guided catalytically inactive Cas9 (dCas9) fused to the FokI nuclease [34][35][36]. Recently, an alternative CRISPR system to Cas9 nuclease called Cpf1 nuclease was used for genome editing with simpler crRNA synthesis and more effective delivery and targeting results [37]. Nevertheless, whole genome sequencing is still required before using these corrected clones in clinical applications.
Future therapeutic use of ex vivo genetically corrected HSCs will require efficient conversion of these stem cells to differentiated hematopoietic precursors and safety testing. We used two in vitro differentiation schemes to assess the potential of our cells for conversion to hematopoietic cells making normal hemoglobin products. We first induced hematopoietic differentiation using a chemically defined, serum and feeder cell-free protocol based on aryl hydrocarbon receptor (AhR) activation [18]. This protocol has been shown to facilitate an expansion of both erythroid and megakaryocyte progenitor cells. We compared the number of hematopoietic progenitor and erythroid markers of the wild-type iPSCs (HDF-iPSCs), the Eβ-iPSC2 cells and the five corrected clones. The number of adherent and nonadherent cells expressing those markers was comparable in the wild-type iPSCs and the five corrected clones; however, we noticed varying numbers of multilineage CFUs among these cell types due to clonal variability. In contrast, the Eβ-iPSC2 cells could differentiate to some extent, as demonstrated by the presence of CD34 + and CD71 + cells in adherent cells, although these cells could not give rise to CFUs. Hematopoietic differentiation based on the OP9 coculture system seemed to be more supportive for the Eβ-iPSC2 cells. Morphological changes of the Eβ-iPSC2 cells on the OP9 feeder layer were similar to those of the corrected C46 cells. Both cells produced significant numbers of erythroblasts with occasional enucleated mature erythrocytes using the three-stage erythroid liquid culture system. The erythroblasts from both the Eβ-iPSC2 cells and the corrected C46 cells expressed KLF1 and BCL11A transcripts at lower levels compared to those of the peripheral blood-derived erythroblasts. Similarly, analysis of beta and alpha hemoglobin protein expressions showed that both the Eβ-iPSC2 cells and the corrected C46 cells expressed the proteins at similar levels; however, the beta hemoglobin levels were much lower than those of the peripheral blood-derived erythroblasts. A previous study demonstrated that the transcription factors KLF1 and BCL11A are required for the induction of beta hemoglobin levels [38]; therefore, low levels of KLF1 and BCL11A transcripts in the corrected C46 cell-derived erythroid cells could result in lower beta hemoglobin expression. In addition, the number of erythroid cells derived from iPSCs is still limited; further optimization of the differentiation protocol to increase the number of erythroid cells should be performed to further facilitate an in vitro functional assay. In this study, the anti-human beta hemoglobin antibody could not distinguish the one amino acid difference from abnormal hemoglobin E in the Eβ-iPSC2 cells. Therefore, validation with other techniques such as HPLC should be performed to confirm the result at protein levels. In addition, further development of long-term repopulating hematopoietic stem/progenitor cells is required for successful engraftment in immunodeficient mouse models, where the clinical efficacy of corrected human hematopoietic stem cells can be more properly assessed.

Conclusion
Our study provides a successful strategy to correct HbE mutation in one step and could be employed as a universal approach in the future correction of the HBB gene in iPSCs derived from other HbE/β 0 -thalassemia or β +thalassemia patients. The results indicate that genetic correction of HbE mutation in one allele is sufficient to restore HBB protein expression upon hematopoietic differentiation by the OP9 coculture system followed by an erythroid liquid culture. Similarly, previous studies demonstrated that genetic correction of mutation in one allele-to a heterozygous state-in iPSCs derived from homozygous β-thalassemia or sickle cell disease was achieved using CRISPR/Cas9 or TALENs [6,8,9,12]. Occasionally, homozygous corrections of homozygous β-thalassemia iPSCs were observed [10,13]. The knowledge and protocols obtained from this study will facilitate and be applicable to genetic correction of patientspecific iPSCs with other genetic disorders.