2.1 Cell culture

Human iPSC lines HuAiPSC-A1¹⁹ and HuAiPSC-A1-RHD^−/− were cultured in Essential 8 (Gibco, Grand Island, NY) on Matrigel-coated wells (Corning, New York). Cells were maintained daily and passaged every 4–6 days to maintain undifferentiated growth, as previously described.¹⁹ When colonies reached 70%–80% confluency, ReLeSR (StemCell Technologies, Vancouver, BC, Canada) was used to detach and dissociate large clones, and cells were passaged at a 1:10–1:20 ratio. Single cells were obtained using Accutase (StemCell Technologies) before plasmid transfection and fluorescence-activated cell sorting (FACS). A rock inhibitor (Y-27632 2HCl, Selleck, Houston, TX) was used to improve the cell survival rate during replating. All cultures were maintained at 37°C in a 5% CO₂ incubator (Thermo Scientific, Waltham, MA).

2.2 Guide RNA design

Guide RNAs (gRNA) were designed using online portals (crispr.mit.edu and CCTop the CRISPR/Cas9 target online predictor). Exons that encode key regions of the protein in question were selected. The selected gRNAs were as follows:

Exon 2: 5′-GCTGTGTCTCCGGAAACTCG-3′.

Exon 5-1: 5′-ACGGCATTCTTCCTTTCGAT-3′.

Exon 5-2: 5′-GCTGACTGCTACAGCATAGT-3′.

Exon 7-1: 5′-GGGCTACAACTTCAGCTTGC-3′.

Exon 7-2: 5′-GCTGAAGTTGTAGCCCATGA-3′.

Exon 7-3: 5′-GATACCGTCGGAGCCGGCAA-3′.

gRNAs were cloned into the LentiCRISPR v2 vector (ADDGENE, Massachusetts). K562 cells were used to identify successful gRNA combinations for the knockout of the selected genes.

2.3 Generation of the RHD knockout HuAiPSC-A1 cell line

To obtain the RHD knockout cell line, homologous recombination was used to repair the DNA sequence after gene editing. RHD-specific gRNA was cloned into the Cas9 expression vector (SH100, GeneCopoeia, Maryland) by replacing the original AAVS1 targeting gRNA to generate the pRHD-Exon 2-sgRNA-CRISPR-Cas9 vector. RHD-exon 2 left and right homologous arms, containing the sequence of the premature stop code, were cloned into the GFP/puromycin expression vector (SH200, GeneCopoeia) to produce the pRHD-HL-stop-HR-donor-puro-GFP vector (Figure S1). Specifically, homologous fragments were amplified from the genome of the hESC-H1 cell line and cloned into the pClone007 vector (TSINGKE, Beijing, China). Then, the vector SH200 was digested with HpaI/XhoI (NEB, New Jersey), the AAVS1 right homologous arm was removed, and a HpaI/SalI-digested RHD-exon 2 right homologous arm fragment from pClone007-RHD-Right was ligated into the donor vector. Subsequently, the RHD-exon 2 left homologous arm was digested with KpnI/XbaI (NEB) from pClone007-RHD-left and inserted into the KpnI/XbaI-digested donor vector.

Confluent HuAiPSC-A1 cells (40%–50%) were transfected with 500 ng of CRIPSR/Cas9 plasmids (250 ng pRHD-exon 2-sgRNA-CRISPR-Cas9 and 250 ng pRHD-HL-stop-HR-donor-puro-GFP) using Lipofectamine Stem Transfection Reagent (Invitrogen, Carlsbad, CA, USA). Untransfected HuAiPSC-A1 served as the negative control. After incubation for 2 days at 37°C, the cells were treated with 0.5 μg/mL puromycin (InvivoGen, California). After 2 weeks of drug selection, the puromycin-resistant clones were expanded. GFP-positive cells were sorted using FACS.

2.4 Sequencing analysis

To confirm RHD KO colonies, genomic DNA was extracted using the TIANamp Genomic DNA Kit (TIANGEN, Beijing, China) and polymerase chain reaction (PCR) was performed using the Q5® High-Fidelity 2xMaster Mix (NEB). The primers used are listed in Table S1.

2.5 Mycoplasma test

Mycoplasma was detected using a Mycoplasma Detection Set (Macgene, Beijing, China), according to the manufacturer's instructions.

2.6 Karyotype and short tandem repeat analyses

Cells were treated with colcemid (Selleck), harvested and treated with hypotonic solution for karyotype analysis, as previously described.¹⁹ At least 20 metaphases for each sample were analysed with regard to chromosome number and structural rearrangements by the Guangzhou Kingmed Center for Clinical Laboratory. Short tandem repeat (STR) analysis was performed with the detection of 21 loci by Microread Genetics Co. Ltd. (Beijing, China).

2.7 Alkaline phosphatase staining

Alkaline phosphatase (AP) activity was measured using an AP staining solution kit (Beyotime, Shanghai, China), according to the manufacturer's instructions.

2.8 Embryoid bodies (EBs) differentiation

Confluent (iPSCs) were detached from Matrigel using Accutase (StemCell Technologies) for 3–5 min at 37°C. To form EBs, cells were washed thrice, resuspended in Essential 8 (Gibco) supplemented with ROCK inhibitor (Y27632, 10 μmol/L), and plated at a density of 2 × 10⁵ cells/well in ultra-low attachment 6-well plates (Corning). After 24 h, the EBs were harvested and resuspended in Essential 6 medium (Gibco). The medium was changed every alternate day. After 7 days, the cells were transferred into Matrigel-coated (Corning) wells for adherent growth. The medium was replaced with fresh medium every alternate day. The cells were allowed to differentiate for 7 days for further immunostaining analysis.

2.9 Teratoma formation

The mice were manipulated and housed according to protocols approved by the Institutional Animal Care and Use Committee (IACUC) in compliance with the Beijing Medical Experimental Animal Care Commission. (Reference number: IACUC of AMMS-13-2016-016) 5 × 10⁶ cells were resuspended in 80 μL of precooled (4°C) Matrigel (Corning) and maintained on ice until intramuscular injection into NOD/SCID mice. Teratomas are typically formed within 8–10 weeks of injection. Teratomas were dissected, fixed in 4% paraformaldehyde, embedded with paraffin, sectioned, and then stained with haematoxylin and eosin (H&E) and subjected to immunostaining analysis.

2.10 Immunofluorescence

For immunofluorescence staining, cells were fixed with 4% paraformaldehyde, permeabilized with 0.2% Triton X-100, blocked in 3% bovine serum albumin (BSA), and further incubated with primary antibodies in Dulbecco's phosphate-buffered saline (DPBS) with 3% BSA overnight at 4°C. Cells were then incubated with the corresponding secondary antibodies for 45 min in the dark at room temperature. The cells were counterstained with 4′,6-diamidino-2-phenylindole (1:500 dilution) for 15 min, maintained in the dark at room temperature, and visualized using fluorescence microscopy. The antibodies used are listed in Table S2.

2.11 Induction of hiPSC differentiation into erythrocytes

The in vitro generation of erythrocytes comprises four phases: (I) mesoderm induction, (II) hemogenic endothelium (HE) commitment, (III) haematopoietic cell emergence and erythroid differentiation and (IV) erythrocyte maturation. Cells were differentiated into mesoderm and subsequently subjected to HE using a previously published protocol with modifications.^{20, 21}

Briefly, in Phase I (Days −1 to 2), 70%–80% confluent hiPSCs were treated with Accutase (StemCell Technologies) for 3–5 min in a 37°C incubator. Small, scraped clumps were collected, centrifuged, resuspended in Essential 8 (Gibco) supplemented with 10 μmol/L Y27632 (Selleck) and plated at a density of 2 × 10⁵ cells/well in ultra-low attachment 6-well plates (Corning). After 24 h (Day 0), cell aggregates were harvested and resuspended in differentiation medium (DM) 1 consisting of advanced Dulbecco's medium/F12 supplemented with 1% GlutaMAX, 1% penicillin/streptomycin, 50 μg/mL l-ascorbic acid 2-phosphate (AA2P), 25 ng/mL BMP4, 25 ng/mL bFGF and 25 ng/mL activin A. The medium was changed daily during this phase.

In phase II (Days 2–6), the EBs were collected and resuspended in DM2, consisting of advanced Dulbecco's medium/F12 supplemented with 1% GlutaMAX, 1% penicillin/streptomycin, 50 μg/mL AA2P, 25 ng/mL bFGF, 50 ng/mL VEGF₁₆₅ and 2 μmol/L SB431542 (TGFβ inhibitor). The medium was changed daily during this phase.

In phase III (Days 6–15), the EBs were collected and resuspended in DM3, consisting of BEL medium^{20, 22} (Table S3) supplemented with 50 ng/mL SCF, 20 ng/mL TPO, 20 ng/mL IL-3, 20 ng/mL Flt3L, 20 ng/mL VEGF₁₆₅, 5 U/mL EPO, 100 μg/mL transferrin and 10 μmol/L SB431542. The medium was changed every 2 days during this phase.

In phase IV (Days 15–18), for further maturation, cells were collected and passed through a 40 μm cell strainer, washed with DPBS, and resuspended in DM4 containing BEL (without BSA) supplemented with 2.5% AB serum, 5 U/mL EPO and 3 U/mL Heparin Solution. Cells were seeded at a density of 5 × 10⁶ cells/well in 6-well plates to allow non-erythroid cells to attach.²³ The medium was half-changed daily during this phase. The cultures were maintained at 37°C in 5% CO₂ and a designated concentration of oxygen in a humidified atmosphere according to the schedule.

On Day 18, non-adherent cells were collected by passing through a 40 μm cell strainer for erythrocyte analysis. Cells were maintained at 37°C in 5% CO₂ and 20% oxygen in a humidified atmosphere.

Cell number and morphology were assessed using a cell viability analyser (Beckman Counter, California) and Wright-Giemsa (BASO, Zhuhai, Guangdong, China) staining, respectively. At the end of the cultivation period, we assessed the number of total cells, erythroid cells, CD34⁺ cells, T cells, B cells, natural killer cells, mononuclear cells and neutrophils using the method described below.

The main reagents used for cell culture and differentiation are listed in Table S2.

2.12 Quantitative reverse-transcription PCR analysis

mRNA was extracted from cells using TRIzol reagent (Invitrogen) and reverse transcribed using ReverTra Ace quantitative reverse-transcription (qRT-)PCR Master Mix (TOYOBO, Osaka, Japan), according to the manufacturer's instructions. qRT-PCR was performed using the THUNDERBIRD SYBR qPCR Mix (TOYOBO) on a Bio-Rad CFX Connect™ (Bio-Rad, Hercules, CA). The primers used for detection are listed in Table S1.

2.13 Flow cytometry and FACS

The cultured cells were harvested on Days −1, 2, 6, 9, 15 and 18, and their immunophenotypes were determined as follows. On Days 2, 6 and 9, the cells were dissociated into single cells using Accutase (StemCell Technologies) with collagenase D (Sigma, St. Louis, MO) and DNase I (Sigma) and then resuspended in MACSima™ Running Buffer (Miltenyi Biotec, Bergish Gladbach, Germany). On days 15 and 18, the cells were directly passed through a 40 μm cell strainer, washed, and resuspended in Hanks' Balanced Salt Solution (HBSS). The cells were incubated with isotypic antibodies or the indicated antibodies for 30 min at 4°C. The cells were then washed and suspended in HBSS for analysis. A total of 10 nM Syto62 (BD Biosciences, San Jose, CA) was added to the corresponding samples, which were incubated for 10–30 min before analysis. The following antibodies were used: PE-Brachyury (R&D Systems, Aimolivel, California, CA), APC-KDR (R&D Systems), APC-SSEA4 (BD Biosciences), APC-CD34 (BD Biosciences), PE-CD45 (BD Biosciences), BV421-CD43 (BD Biosciences), APC-CD71 (BD Biosciences), BV421-CD235a (BD Biosciences), FVS510 (BD Biosciences), APC-CD31 (eBioscience, San Diego, CA), PECY7-CD71 (eBioscience), APC-Syto62 (Invitrogen) and PE-TRA-1-60 (MACS).

For intracellular staining of brachyury, the procedure was performed according to the manufacturer's instructions (R&D Systems). For Rh D analysis, the cells were blocked with 10% donkey serum for 1 h at room temperature. The cells were then incubated with Rh D antibody (Biorbyt; dilution1:100) for 1 h at 20°C. Next, the cells were washed with DPBS and stained with donkey anti-rabbit AF647 (Invitrogen) for 45 min at room temperature. Cells were washed and stained with BV421-CD235a, PECY7-CD71 and FVS510 for 30 min at 4°C. Live cells identified by FVS510 exclusion were analysed for surface marker expression using a BD FACSAria II cell analyser (BD Biosciences). The percentages of positive cells were determined and compared with isotype controls, and the analyses were performed using FlowJo software (Version 10; TreeStar, Ashland, OR). The antibodies used are listed in Table S2.

For FACS of immunolabelled cells, a BD FACSAria II Cell Sorter was used to isolate the GFP⁺/SSEA4⁺/TRA-1-60⁺ population.

2.14 Agglutination assay

Agglutination assay was performed as previously described.⁷ Briefly, differentiated HuAiPSC-A1 and HuAiPSC-A1-RHD^−/− cells as well as control RBCs were plated in 96-well plates at 1 × 10⁶ cells per well in 10 μL of DPBS. Next, 10 μL of anti-D blood grouping reagent was added to each well and mixed, incubated the mixture for 15 min at 37°C. The incubated cells were observed using a digital camera (Nikon Corporation, Tokyo, Japan).

2.15 Functional analysis of haemoglobin

After washing with DPBS, 5 × 10⁷ cells per group were analysed. Oxygen equilibrium curves were determined using an oxygenation-dissociation analyser (BLOODOX-2018 Analyzer, Softron Biotechnology, Beijing, China), as previously described.²⁴ Human adult peripheral blood cells were used as controls.

2.16 Statistical analysis

All data are expressed as the mean ± SD. Statistical significance was determined using a two-tailed Student's t-test. A p-value <0.05 was considered statistically significant.

3.1 Design and validation of gRNAs targeting RHD

To validate the activity of the designed gRNAs, we transfected plasmids encoding both Cas9 and gRNA into K562 cells, selected puromycin, and then performed qRT-PCR and flow cytometry (FCM) assays. qRT-PCR showed that the gRNA of exon 2-1, exon 5-1 and exon 7-3 significantly reduced RHD gene expression without off-target effects on the RHCE gene (Figure 1A). These three groups of gRNA-transfected K562 cells were subjected to single-cell cloning, followed by RT-qPCR and FCM assays. Finally, we determined that the gRNA targeting exon 2 was more efficient than the others (Figure 1B,C). gRNA of exon 2-1 was selected for further experiments.

3.2 Generation of hiPSC clones containing RHD mutations

A homology-directed repair (HDR)-based CRISPR/Cas9 system was used for RHD gene knockout in hiPSCs instead of the classical non-homologous end-joining (NHEJ)-based CRISPR/Cas9 system. The homologous arms were designed to be identical to the side sequence around the exon 2-1 gRNA target site of RHD gene loci, and a strong stop code together with the EF1α promoter driving GFP and puromycin was inserted to disrupt RHD expression (Figure 2A). After puromycin selection, followed by GFP-positive sorting, potential candidates for mutant cell clones were confirmed using electrophoresis and sequencing (Figure 2B,C; Figure S2). The unique electrophoresis band and sequencing results suggested the homogeneity of the selected hiPSC clone. Consequently, we obtained a mutated clone with a premature stop code (TAGaTAAcTGA) in the RHD coding sequence (Figure S2).

3.3 Initial characterization of RHD knockout iPSC colonies

RHD knockout hiPSC colonies (HuAiPSC-A1-RHD^−/−) showed a normal karyotype (46, XX) and no mycoplasma contamination (Figure 3A and S3). Short tandem repeat (STR) analysis confirmed that the hiPSC line shared a 100% matched genetic profile with wild-type (WT) HuAiPSC-A1 (Figure S4). HuAiPSC-A1-RHD^−/− cells exhibited a typical hiPSC-like morphology and AP activity (Figure 3B,C). The pluripotency of HuAiPSC-A1-RHD^−/− cells was confirmed using FCM analysis of pluripotency markers (SSEA-4 and TRA-1–60), immunostaining for pluripotency markers (SOX2, OCT4 and NANOG), and qRT-PCR for pluripotency genes (SOX2, OCT4 and NANOG; Figure 3D–F). This cell line had the potential to differentiate into all three germ layers in vitro in the EB differentiation assays (Figure 3G) and in vivo in the teratoma formation assays (Figure 3H,I).

3.4 Optimized differentiation scheme towards erythroid lineage

Studies have shown enhanced in vitro expansion of HSPCs and early development of erythroid clones when cultured under hypoxic conditions (1%–5% O₂).^{16, 17, 25} However, hypoxia inhibits the terminal expansion and maturation of erythroid precursors.¹⁸ To obtain more mature erythrocytes with high CD235a expression, we set up an 18-day stepwise induction protocol and analysed the influence of alterations in oxygen concentration during the process of induction (Figure 4A).

Three oxygen conditions, namely normoxia, hypoxia and hypoxia–normoxia culture conditions were established by setting the oxygen concentration at 20%, 5% and 5% for the first three phases and 20% for induction phase IV, respectively. By induction Day 18, the high levels of CD235a expression confirmed that these three conditions produced virtually pure populations of erythroid cells (Figure 4B). Significantly, we found that 5% oxygen followed by 20% oxygen resulted in more enucleated erythrocytes (CD235a⁺Syto16⁻) and higher expression of Rh D in CD235a⁺ cells than in cells under 20% oxygen or 5% oxygen (Figure 4B). The group treated with 5% oxygen followed by 20% oxygen had a higher HE marker (KDR/CD31) on Day 6 HSPC marker (CD34/CD43) on Day 9, and erythroid progenitor cell marker (CD71) on Day 15 than the other groups (Figure 4B). Moreover, the group treated with 5% oxygen followed by 20% oxygen had a higher expansion (measured by erythroid cells generated from one initially plated hiPSC) and cell vitality than the others (Figure 4C). Collectively, 5% oxygen followed by 20% oxygen resulted in higher haematopoietic and erythroid expansion efficiency as well as more mature erythrocytes compared with 5% and 20% oxygen.

3.5 Impact of RHD knockout on HuAiPSC-A1 differentiation and function

We evaluated whether CRISPR/Cas9-induced RHD knockout affects erythrocyte generation from iPSCs or oxygen-carrying functions. FCM analysis showed that at each stage of differentiation induction, the expression levels of the primitive streak (PS)/early mesoderm (Brachyury and KDR), HE (CD31), early haematopoietic (CD34 and CD43), early erythroid progenitor (CD71), and mature erythroid cell markers (CD235a) did not change between RHD knockout (HuAiPSC-A1-RHD^−/−) and WT hiPSC clones (HuAiPSC-A1; Figure 5A). At the end of differentiation, for both cell types, more than 93% of the cells expressed erythrocyte marker CD235a, whereas the majority of the cells did not express megakaryocytic antigens, myelomonocytic antigens, and other haematopoietic progenitor antigens (Figure S5), corroborating that RHD knockout did not affect haematopoietic differentiation. Furthermore, qRT-PCR analysis also indicated that the expression levels of pluripotency genes (OCT4 and SOX2), mesoderm genes (BRA and MIXL), haematopoietic-related genes (GATA1 and RUNX1) and erythroid-related genes (EKLF and EPOR) did not change between RHD knockout and WT hiPSC clones (Figure 5B), suggesting that CRISPR/Cas9-induced RHD knockout does not affect the erythroid differentiation process. Moreover, the average cell size during differentiation and the cell morphology at the end of differentiation were identical between RHD knockout and WT iPSCs (Figure 5C,D), suggesting that RHD knockout does not affect differentiation. For functional comparison between erythrocytes generated from RHD knockout and WT hiPSCs, we measured their oxygen-carrying capabilities. The oxygen binding and dissociation curves obtained were similar for both cell types (Figure 5E), indicating that CRISPR/Cas9-induced RHD knockout did not affect RBC function. Collectively, these results indicate that erythrocyte differentiation and the oxygen-carrying function of these derived erythrocytes are not affected by the CRISPR/Cas9-induced RHD mutation in HuAiPSC-A1.

3.6 Rh D antigen expression in HuAiPSC-A1-RHD^−/− erythrocytes

The primary goal of this study was to generate Rh-D-negative RBCs from genetically modified hiPSCs. Therefore, we evaluated Rh D antigen expression in erythrocytes generated from HuAiPSC-A1 and HuAiPSC-A1-RHD^−/− clones using FCM. A previous study has shown that FCM is a sensitive way to detect Rh D antigen expression with a nonspecific background signal of approximately 1%.⁷ In our study, among the CD235a⁺ cells, the Rh D antigen expression rate was 52.17 ± 5.914% and 0.907 ± 0.668% in HuAiPSC-A1-Ery and HuAiPSC-A1-RHD^−/−-Ery cells, respectively (Figure 6A). The Rh D antigen of RHD knockout clones was identical to the nonspecific background level, suggesting the absence of Rh D antigen expression in erythrocytes generated from HuAiPSC-A1-RHD^−/− clones. RHD gene knockout was also confirmed at the mRNA level, and compared to HuAiPSC-A1-Ery, the expression of the RHD gene was significantly reduced in HuAiPSC-A1-RHD^−/−-Ery (Figure 6B).

RHCE genes are highly homologous with RHD, and previous studies have commonly reported off-target effects on RHCE when RHD was targeted.^{7, 26} Next, we investigated whether the RHD knockout clones contained off-target mutations arising from gene editing. No significant difference was detected in the expression of RHCE genes between HuAiPSC-A1-Ery and HuAiPSC-A1-RHD^−/−-cells through qRT-PCR analysis (Figure 6B).

The blood type of the ‘universal’ RBCs derived from RHD gene knockout hiPSCs was finally verified using an agglutination test. HuAiPSC-A1 and HuAiPSC-A1-RHD^−/− cells were induced for erythroid differentiation for 18 days using the optimized protocol, and blood group serological tests were performed. We used Rh D-positive and Rh D-negative blood cells as positive and negative controls, which showed agglutination and no agglutination, respectively. HuAiPSC-A1-RHD^−/−-Ery cells did not agglutinate with Rh D antiserum, demonstrating that they were Rh D-negative cells (Figure 6C). The induced RBCs were tested as type O blood samples (Figure 6D).