Efficient generation of β-globin-expressing erythroid cells using stromal cell-derived induced pluripotent stem cells from patients with sickle cell disease

Abstract

Human embryonic stem (ES) cells and induced pluripotent stem (iPS) cells represent an ideal source for in vitro modeling of erythropoiesis and a potential alternative source for red blood cell transfusions. However, iPS cell-derived erythroid cells predominantly produce ε- and γ-globin without β-globin production. We recently demonstrated that ES cell-derived sacs (ES sacs), known to express hemangioblast markers, allow for efficient erythroid cell generation with β-globin production. In this study, we generated several iPS cell lines derived from bone marrow stromal cells (MSCs) and peripheral blood erythroid progenitors (EPs) from sickle cell disease patients, and evaluated hematopoietic stem/progenitor cell (HSPC) generation after iPS sac induction as well as subsequent erythroid differentiation. MSC-derived iPS sacs yielded greater amounts of immature hematopoietic progenitors (VEGFR2+GPA-), definitive HSPCs (CD34+CD45+), and megakaryoerythroid progenitors (GPA+CD41a+), as compared to EP-derived iPS sacs. Erythroid differentiation from MSC-derived iPS sacs resulted in greater amounts of erythroid cells (GPA+) and higher β-globin (and βS-globin) expression, comparable to ES sac-derived cells. These data demonstrate that human MSC-derived iPS sacs allow for more efficient erythroid cell generation with higher β-globin production, likely due to heightened emergence of immature progenitors. Our findings should be important for iPS cell-derived erythroid cell generation.

Introduction

Red blood cell (RBC) transfusion is central to the management of a number of severe congenital and acquired anemias, including hematological malignancies, aplastic anemia, thalassemias, and hemoglobinopathies. RBC transfusion-related complications were significantly reduced by infectious screening and blood group matching; however, every transfusion has associated risks, such as alloimmunization, transmitting infectious disease, and immediate transfusion reactions, among others [1]. Efforts to find potential alternative sources for RBC transfusion have recently been pursued by the development of in vitro erythroid differentiation techniques from human CD34+ cells, peripheral blood mononuclear cells (PBMCs), embryonic stem (ES) cells, and induced pluripotent stem (iPS) cells [2, 3]. Reprogramming methods with genome editing techniques may allow the creation of identical and, if necessary, genetically corrected RBCs for transfusion, especially for diseases such as sickle cell disease (SCD) [4-9]. Autologous iPS cell-derived RBCs can circumvent the significant problem of alloimmunization in bone marrow (BM) failure or hemoglobinopathy patients.

In mammalian development, primitive hematopoiesis begins in the yolk sac (YS), which directly generates primitive RBCs expressing ε-globin. Subsequently, definitive hematopoiesis commences in the aorta-gonad-mesonephros (AGM) region, fetal liver, and BM, where definitive RBCs expressing γ-globin or β-globin are produced [10-15]. In the AGM region, hemangioblasts produce both endothelial cells and hematopoietic cells through hemogenic endothelia. The hemogenic endothelia give rise to hematopoietic stem/progenitor cells (HSPCs) [16-20]. Therefore, hemangioblast formation during in vitro differentiation of ES/iPS cells might be crucial for the derivation of definitive erythroid cells [21-23].

In traditional embryoid body (EB)-based in vitro differentiation methods, iPS cell-derived erythroid cells predominantly produce ε-globin and γ-globin without β-globin expression, even though small amounts of β-globin production is observed in ES cell-derived erythroid cells [24-31]. We recently demonstrated that ES cell-derived sacs (ES sacs), known to express hemangioblast markers, allow for efficient erythroid cell generation with β-globin production [23, 32]. The ES sac-derived definitive erythroid cells with β-globin expression were mainly derived from CD34+ HSPCs in ES sacs [32].

We speculated that the iPS cells are more efficiently differentiated to target cells when the iPS cells are generated from a similar source of cells due to epigenetic memory [33]. In addition, difference among iPS cell clones may affect the differentiation abilities. Our initial hypothesis for this study was that erythroid progenitor (EP)-derived iPS cells are more efficiently differentiated to erythroid cells. On the other hand, erythroid specific epigenetic memory might induce direct erythroid differentiation during iPS sac generation and ε-globin expressing primitive erythroid cell generation. It raised the second hypothesis that BM stromal cell (MSC)-derived iPS cells more efficiently generate iPS sacs and emerge immature HSPCs, leading to the generation of definitive erythroid cells expressing higher levels of β-globin.

In the current study, we investigated erythroid cell generation using iPS cell lines derived from different starting cell types, including cells from SCD patients, through iPS sac induction.

Methods

Transgene-free iPS cell generation with lentiviral transduction

We generated several iPS cell lines which were derived from either (1) EPs (6 clones) which were differentiated from PBMCs or (2) MSCs (5 clones) from SCD patients (Supplementary table). All human subject materials were collected under a protocol approved by the Institutional Review Board of the National Heart, Lung, and Blood Institute (07-H-0113). All patients gave written informed consent. PBMCs were separated from blood of an SCD patient with the homozygous sickle mutation (HbSS), and 2×10e6 PBMCs were differentiated to CD36+CD71+ EPs (>90% of positivity) for 9-12 days [34], and these EPs were transduced with an Oct4, Klf4, Sox2, and c-Myc encoding lentiviral vector (hSTEMCCA-loxP) at multiplicity of infection (MOI) 5 [35, 36]. BM cells from another SCD patient with HbSS were cultured in α-minimum essential medium (α-MEM; Life Technologies, Grand Island, NY, USA) containing 20% fetal bovine serum (FBS; Thermo Fisher Scientific, Waltham, MA, USA) and 2 mM L-glutamine (Life Technologies) [37], and the MSCs were confirmed as a spindle-shaped morphology of attached cells (which were separated from blood cells with a round shape of suspension cells). The 1.0×10e5 MSCs were transduced with the same reprogramming lentiviral vector at MOI 15 or 25, and 3-5 days after transduction, the transduced cells were cultured on irradiated mouse embryonic fibroblast feeder cells (CF1-MEF, GlobalStem, Gaithersburg, MD, USA) in Dulbecco’s Modified Eagle Medium / Nutrient Mixture F-12 (Life Technologies) containing 20% KnockOut Serum Replacement (Life Technologies), 10 ng/ml basic fibroblast growth factor (PeproTech, Rocky Hill, NJ, USA), 0.1mM Non Essential Amino Acids (Life Technologies), 1mM L-glutamine, and 0.1 mM 2-mercaptoethanol (Life Technologies). 3-6 weeks later, we picked iPS cell-like colonies, and the reprogramming cassette was later excised by Cre recombinase [35]. The iPS cells were evaluated by immunostaining (Nanog, Oct4, SSEA4, Tra1-60, and Tra1-81), alkaline phosphatase stain, karyotyping, and teratoma assay, as previously described [38].

As pluripotent control cells, we used the human H1 ES cell line (WiCell, Madison, WI, USA) and established human iPS cell lines (CTRL-2 c1 and CTRL-1 c3) [38]. The CTRL-2 c1 cells were derived from BJ1 neonatal foreskin fibroblasts (American Type Culture Collection (ATCC), Manassas, VA, USA), and CTRL-1 c3 cells were derived from adult dermal fibroblasts from a healthy donor.

Human ES/iPS cell-derived erythroid cell differentiation through ES/iPS sacs

Human ES/iPS cell-derived erythroid cells were generated through ES/iPS sacs using a four-step culture method, as we previously described (Figure 2A) [23, 32]. Briefly, in step 1 (ES sac culture phase), small clumps of ES/iPS cells (1.0×10e5 cells per 100 mm dish) were cultured on irradiated C3H10T1/2 feeder cells (1.0×10e6 cells per 100 mm dish) (ATCC) for 15 days in Iscove’s Modified Dulbecco’s Medium (IMDM; Sigma Aldrich, Saint Louis, MO, USA) supplemented with 0.2 mg/ml insulin (Lilly, Indianapolis, IN, USA), 0.11 mg/ml transferrin (Sigma Aldrich), 0.1 µg/ml sodium selenite (Sigma Aldrich), 0.45 mM α-mono-thioglycerol (Sigma Aldrich), 50 μg/ml ascorbic acid (Sigma Aldrich), 20 ng/ml human vascular endothelial growth factor (VEGF; Perpro Tech), 2 mM L-glutamine, and 15% FBS. The ES sac culture media were replaced on days 3, 6, 9, 11, and 13. After 15 days of culture, we observed ES/iPS sacs structures. In step 2 (transfer phase), ES/iPS sacs were gently crushed using a 1000 μl pipette, passed through a 40 µm cell strainer, and cultured on irradiated OP9 feeder cells (ATCC) for 2 days in the same media used in the culture phase supplemented with 50 ng/ml stem cell factor (SCF; R&D systems, Minneapolis, MN, USA), 50 ng/ml fms-related tyrosine kinase 3 ligand (R&D systems), 50 ng/ml thrombopoietin (R&D systems), 5 μg/ml interleukin-3 (IL3; R&D systems), 10 ng/ml bone morphogenetic protein 4 (R&D systems) and 5 U/ml erythropoietin (EPO; AMGEN, Thousand Oaks, CA, USA). In step 3 (erythroid differentiation phase), only suspension cells were collected and transferred onto fresh OP9 feeder cells for 5 days in IMDM media containing 10 ng/ml SCF, 1.0 ng/ml IL3, 2.0 U/ml EPO, 1.0 μM dexamethasone (VETone, Boise, ID, USA), 1.0 μM estradiol (Pfizer, NY, NY, USA), and 20% FBS [39, 40]. In step 4 (erythroid maturation phase), the erythroid differentiation media were replaced with IMDM media containing 2.0 U/ml EPO, 10 ng/ml insulin, 0.56 mg/ml transferrin, 2% bovine serum albumin (BSA; Roche, Indianapolis, IN, USA), 2 mM L-glutamine, and 20% FBS [39, 40]. The cells were cultured for 8 days, and the erythroid maturation media were replaced on days 25 and 28. After the ES/iPS sac-derived erythroid differentiation culture (day 30), we evaluated the morphology and enucleation by Wright-Giemsa staining, as previously described [32].

Figure 2. More efficient generation of β-globin-expressing erythroid cells using MSC-derived iPS cells.

(A) The MSC- and EP-derived iPS cells and controls (fibroblast (FB)-derived iPS cells and ES cells) were used to generate ES/iPS cell-derived sacs (ES/iPS sacs) supplemented with vascular endothelial growth factor (VEGF) for 15 days [23]. After 2 day culture of ES/iPS sac-derived spherical cells on OP9 feeder cells, the suspension cells were differentiated into erythroid cells for 13 days [39]. (B) At the end of ES/iPS sac maturation (15 days), greater amounts of CD34+CD45+ hematopoietic stem/progenitor cells (HSPCs) emerged in both MSC- and EP-derived iPS sacs, compared to FB-derived iPS sacs. (C) After an additional 2 weeks of erythroid differentiation, we observed greater amounts of glycophorin A (GPA)+ erythroid cells from both MSC- and EP-derived iPS sacs, compared to FB-derived iPS sacs. Interestingly, MSC-derived iPS sacs resulted in greater amounts of GPA+ erythroid cells, compared to EP-derived iPS sacs. (D) Higher β-globin RNA expression was observed in erythroid cells from MSC-derived iPS sacs, compared to EP- and FB-derived iPS sacs, which was comparable to ES sac-derived erythroid cells. (E) We detected βS-globin protein (β-globin with the sickle mutation) production in erythroid cells from MSC-derived iPS sacs by a reversed phase high-performance liquid chromatography, which amounts are similar to β-globin protein in ES sac-derived erythroid cells. However, either βS- or β-globin protein was not detectable in erythroid cells from EP- and FB-derived iPS sacs. (F) Sickle hemoglobin (HbS) in MSC-derived erythroid cells was detected by hemoglobin electrophoresis. SCF: stem cell factor, FL: fms-related tyrosine kinase 3 ligand, TPO: thrombopoietin, IL3: interleukin 3, EPO: erythropoietin, BMP4: bone morphogenetic protein 4, eHb: embryonic hemoglobin, HbF: fetal hemoglobin, HbA: adult hemoglobin. **p<0.01, *p<0.05 evaluated by Tukey’s HSD test.

Colony forming unit (CFU) assay

Following the transfer phase (day 17), we cultured the suspension cells (1.0x10e5 per 35-mm dish) in semi-solid media (MethoCult H4434 Classic; STEMCELL Technologies, Vancouver, BC, Canada), as previously described [41]. After 2 week culture, the CFUs were counted by microscope.

Flow cytometry

We performed cell surface analysis using a FACSCalibur flow cytometer (Becton Dickinson, East Rutherford, NJ, USA). The following monoclonal antibodies were used for identifying and characterizing cell subsets: CD31 (clone WM59), CD34 (clone 581 or 563), CD41a (clone HIP8), CD43 (clone 1G10), CD45 (clone HI30), CD71 (clone M-A712), CD73 (clone AD2), vascular endothelial cadherin (VE-cadherin (CD144); clone 55-7H1), glycophorin A (GPA (CD235a); clone GA-R2), vascular endothelial growth factor receptor-2 (VEGFR2 (CD309); clone 89106) (all from Becton Dickinson) as well as delta like ligand 4 (DLL4; clone MHD4-46) (Miltenyi, San Diego, CA, USA). Apoptotic cells were evaluated by Annexin V: FITC Apoptosis Detection Kit I (Becton Dickinson). The surface marker-positive (or negative) cell numbers were calculated by multiplying the percentage of surface markers by total cell amounts at the time of evaluation. All cell number data were obtained from the same initial amounts of ES/iPS cells (1.0×10e5).

Reverse transcription quantitative polymerase chain reaction (RT-qPCR)

After a 13 day erythroid differentiation culture (day 30 of ES/iPS sac generation protocol), erythroid cells were collected and evaluated to determine RNA expression levels of ε-globin, γ-globin, β-globin, and α-globin as we previously described [42]. Quantitative PCR assay was performed using gene-specific primers and probes in the Mx3000P (Agilent Technologies, Santa Clara, CA, USA). The following primer and probe sequences were used: ε-globin forward primer, 5’- TGG CAA GGA GTT CAC CCC T -3’; ε-globin reverse primer, 5’- AAT GGC GAC AGC AGA CAC C -3’; ε-globin probe, 5-ROX- TGC AGG CTG CCT GGC AGA AGC -IBRQ-3’; γ-globin forward primer, 5’- GGC AAC CTG TCC TCT GCC TC -3’; γ-globin reverse primer, 5’- GAA ATG GAT TGC CAA AAC GG -3’; γ-globin probe, 5’-Cy5- CAA GCT CCT GGG AAA TGT GCT GGT G -IBRQ-3’; β-globin forward primer, 5’- CTC ATG GCA AGA AAG TGC TCG -3’; β-globin reverse primer, 5’- AAT TCT TTG CCA AAG TGA TGG G -3’; β-globin probe, 5’-FAM- CGT GGA TCC TGA GAA CTT CAG GCT CCT -IBRQ-3’, α-globin forward primer, 5’- TCC CCA CCA CCA AGA CCT AC -3’, α-globin reverse primer, 5’- CCT TAA CCT GGG CAG AGC C -3’, α-globin probe, 5’-HEX- TCC CGC ACT TCG ACC TGA GCC A -IBRQ-3’ [43-46]. We used a control plasmid containing one copy of ε-, γ-, β-, and α-globin cDNA, and calculated relative amounts of ε-, γ-, and β-globin RNA which were standardized by α-globin signals.

Reverse phase high pressure liquid chromatography (RP-HPLC)

For globin protein analysis, we collected erythroid cells following 13 day erythroid differentiation (day 30) from ES/iPS sacs. After washing 3 times with phosphate buffered saline (Corning, One Riverfront Plaza, NY, USA), the cell pellet was resuspended in 100μl HPLC grade water (Sigma-Aldrich) and vortexed to lyse the cells followed by a centrifugation at 16,000 g for 20 minutes at 4°C. The 90 μl supernatant was mixed to 10μl of 100mM Tris (2-carboxyethyl) phosphine (Thermo Fisher Scientific) and incubated for 5 minutes at room temperature. After incubation, we added 85μl solution containing 0.1% trifluoroacetic acid (TFA) (Thermo Fisher Scientific) and 32% acetonitrile (Honeywell Burdick & Jackson, Morris Plains, NJ, USA) and well mixed it by a vortex. After a centrifugation at 16,000 g for 5 minutes at 4°C, the supernatant was transferred to a sample vial (SUN-Sri, Rockwood, TN, USA).

10 μl of the samples were injected and analyzed in 0.8 ml/minute flow for 45 minutes using the Agilent 1100 HPLC (Agilent Technologies) equipped with a reverse phase column, Aeris 3.6μm Widepore C4 200 (250x4.6mm, Phenomenex, Torrance, CA, USA) with two solvents: solvent A, 0.12% TFA in water and solvent B, 0.08% TFA in acetonitrile. The gradient for the separation of globin protein was started with 35% of solvent B and changed % of solvent B as follows: 3 minutes at up to 41.2%, 3 minutes at up to 41.6%, 5 minutes at up to 42%, 4 minutes at up to 42.4%, 6 minutes at up to 42.8%, 6 minutes at up to 44.4%, 6 minutes at up to 47%, 7 minutes at up to 75% and re-equilibrated for 10 minutes at 35%. The globin types were detected at 215 nm and confirmed by Agilent HPLC-6224 mass spectrometer equipped with an ESI interface and a time-of-flight (TOF) mass detector (Agilent Technologies) as described [47, 48].

Hemoglobin electrophoresis

After a 13 day erythroid differentiation culture (day 30), erythroid cells were collected and evaluated to determine the hemoglobin types by hemoglobin electrophoresis (HELENA LABORATORIES, Beaumount, TX, USA) according to the manufacturer’s instructions.

Statistical analysis

Statistical analyses were performed using the JMP 11 software (SAS Institute Inc., Cary, NC, USA). The sample sizes (n) were defined by biological replicates. All experiments were performed in triplicate. The averages in various conditions were evaluated by Tukey’s honest significant difference (HSD) test (one-way analysis of variance among all groups) using all raw data. The variabilities among iPS clones were evaluated by standard deviations of average data in each clone, which were shown as error bars in all figures. The variances among cell sources and clones (or MOIs) were evaluated by analysis of variance with random effects. A p value of <0.01 or 0.05 was deemed significant.

Results

Generation of iPS cells derived from BM stromal cells (MSCs) and peripheral blood erythroid progenitors (EPs) in SCD patients

To test our two hypotheses; (1) EP-derived iPS cells are more efficiently differentiated to erythroid cells, and (2) MSC-derived iPS cells more efficiently emerge immature HSPCs which results in greater erythroid cell generation, we generated several iPS cell lines which were derived from both MSCs (5 clones) and peripheral blood EPs (6 clones) in SCD patients (Figures 1A and B). We transduced MSCs and EPs with a reprogramming lentiviral vector, and obtained iPS cell-like colonies followed by excision of the reprogramming cassette [35]. Pluripotent stem cell markers were detected by immunostaining, which were comparable to H1 ES cells (Figure 1C). In teratoma assays, all clones of iPS cells were differentiated to various tissues including three germ layers (Figure 1D). We also observed a normal karyotype in all of iPS cell clones.

Figure 1. Generation of induced pluripotent stem (iPS) cells derived from bone marrow stromal cells (MSCs) and peripheral blood erythroid progenitors (EPs) in sickle cell disease (SCD) patient.

(A and B) We generated several clones of iPS cells which were derived from (A) MSCs (5 clones) and (B) peripheral blood EPs (6 clones) which were differentiated from peripheral blood mononuclear cells (PBMCs) in SCD patients. We transduced MSCs and EPs with a reprogramming lentiviral vector, and obtained iPS cell-like colonies followed by excision of the reprogramming cassette. (C) Pluripotent stem cell markers were detected by immunostaining (Nanog, Oct4, SSEA4, Tra1-60, and Tra1-81) and alkaline phosphatase (ALP) staining. (D) All clones of iPS cells were differentiated to various tissues including three germ layers in teratoma assays. We observed a normal karyotype in all clones of iPS cells. ES cells: embryonic stem cells, MOI: multiplicity of infection.

More efficient generation of β-globin-expressing erythroid cells using stromal cell-derived iPS cells

The MSC- and EP-derived iPS cells, previously established control iPS cells (fibroblast (FB)-derived iPS cell clones), and H1 ES cells were used to generate ES/iPS sacs supplemented with VEGF for 15 days [23]. After 2 day culture of ES/iPS sac-derived spherical cells on OP9 feeder cells, the suspension cells were differentiated into erythroid cells for 13 days (Figure 2A) [39]. After a 13 day erythroid differentiation culture (day 30), we observed eosinophilic erythroid cells with a high density of chromatin (0.5-1.4% of enucleated cells), which were differentiated from iPS sacs as well as ES sacs (Supplementary figure 1).

At the end of ES/iPS sac maturation (15 days), 3.5-4.8 fold greater amounts of CD34+CD45+ HSPCs emerged in both MSC- and EP-derived iPS sacs, compared to FB-derived iPS sacs (p<0.01) (Figure 2B). After an additional 2 weeks of erythroid differentiation, we observed 4.5-8.7 fold greater amounts of GPA+ erythroid cells from both MSC- and EP-derived iPS sacs, compared to FB-derived iPS sacs (p<0.01) (Figure 2C). Interestingly, MSC-derived iPS sacs resulted in 1.4-2.0 fold greater amounts of GPA+ erythroid cells (p<0.01), compared to EP-derived iPS sacs (Figure 2C). The ES/iPS sac-derived erythroid cell generation was more strongly affected by cell sources than variations among iPS cell clones (80.8% vs. 19.2%) as well as MOIs used in reprograming vector transduction (64.5% vs. 35.5%) (Supplementary figure 2). We observed 2.6-4.9 fold higher β-globin (21.5±4.3%, p<0.01), similar γ-globin (1.0-1.2 fold), and 6.8-18.6 fold lower ε-globin RNA expression in erythroid cells from MSC-derived iPS sacs, compared to EP- and FB-derived iPS sacs, which were comparable to ES sac-derived erythroid cells (Figure 2D). The βS-globin protein (β-globin with the sickle mutation) production was detected by an RP-HPLC in erythroid cells from MSC-derived iPS sacs, and the amounts were similar to β-globin protein (1.9 fold) in ES sac-derived erythroid cells (Figure 2E). However, either βS- or β-globin protein was not detectable in erythroid cells from EP- and FB-derived iPS sacs. Importantly, sickle hemoglobin in MSC-derived erythroid cells was easily detectable by hemoglobin electrophoresis (Figure 2F), demonstrating the robustness of this approach as an in vitro model of sickle erythropoiesis. These data demonstrated that MSC-derived iPS sacs allow for more efficient erythroid cell generation with higher β-globin production, compared to EP- and FB-derived iPS sacs.

Increased immature hematopoietic progenitor cells in stromal cell-derived iPS sacs

We evaluated total cell counts during erythroid differentiation from ES/iPS sacs, and found that MSC-derived iPS cells expanded more robustly at the late phase of erythroid differentiation, compared to EP- and FB-derived cells (Figure 3A). At 5 days after erythroid differentiation (day 22), both MSC- and EP-derived cells contained 11.9-12.2 fold greater amounts of CD45+ hematopoietic cells, compared to FB-derived cells (Figure 3B). These data suggest that MSC-derived iPS sacs require a longer culture duration to be differentiated to erythroid cells which might be mediated by partially differentiated CD45+ cells [49].

Figure 3. More efficient generation of immature HSPCs in MSC-derived iPS sacs.

(A) The MSC-derived cells expanded more robustly during the late phase of erythroid differentiation, as compared to EP- and FB-derived cells, suggesting that MSC-derived iPS sacs might contain greater amounts of immature progenitor cells. (B) At 5 days after erythroid differentiation (day 22), both MSC- and EP-derived cells contained relatively greater amounts of CD45+ hematopoietic cells, compared to FB-derived cells. (C and D) We evaluated immature HSPC markers (including hemogenic endothelia) at day 15, and observed greater amounts of VEGFR2+GPA- cells (C) and slightly greater amounts of CD31+CD34+ cells in MSC-derived iPS sacs (D), as compared to EP- and FB-derived iPS sacs. (E) Relatively greater amounts of GPA+CD41a+ megakaryoerythroid progenitors were observed in MSC-derived iPS sacs, compared to EP- and FB-derived iPS sacs (day 15); however, there was no significant difference. (F) When we evaluated more specific hemogenic endothelium markers at day15, greater amounts of VE-cad+CD43-CD73-DDL4-GPA- cells were observed in EP-derived iPS sacs, as compared to MSC- and FB-derived iPS sacs. **p<0.01, *p<0.05 evaluated by Dunnett’s test (compared to fibroblast-derived iPS cells) in Figure 3A, and evaluated by Tukey’s HSD test in Figures 3B-F.

Since we observed an overall increased erythroid output from MSC-derived iPS sacs, we hypothesized that MSC-derived iPS sacs might contain greater amounts of immature HSPCs (including hemogenic endothelia) and/or immature EPs (including megakaryoerythroid progenitors), and we sought to examine the quantitative and qualitative differences of iPS sac differentiation in MSC-derived iPS cells. We evaluated immature HSPC markers (including those found on hemogenic endothelia) at the end of ES/iPS sac maturation [50], since the cell population (day 15) should contain both suspension cells and adhesion cells (maybe including hemogenic endothelia). We observed 7.7 fold greater amounts of VEGFR2+GPA- cells (p<0.01) (Figure 3C) and 1.3-1.4 fold greater amounts of CD31+CD34+ cells in MSC-derived iPS sacs (Figure 3D), compared to EP- and FB-derived iPS sacs (not detectable VEGFR2+GPA- cells in EP-derived iPS sacs). 1.8-2.4 fold greater amounts of GPA+CD41a+ megakaryoerythroid progenitors were also observed in MSC-derived iPS sacs, compared to EP- and FB-derived iPS sacs (day 15); however, there was no significant difference (Figure 3E). Interestingly, when we evaluated more specific hemogenic endothelium markers (more immature) at day15, 3.4-18.2 fold greater amounts of VE-cad+CD43-CD73-DDL4-GPA- cells were observed in EP-derived iPS sacs, as compared to MSC- and FB-derived iPS sacs and ES sacs (p<0.01) (Figure 3F) [29, 51]. These data suggest that EP-derived iPS sacs contains greater amounts of hemogenic endothelia; however, these cells could not efficiently generate HSPCs.

Next, we evaluated ES/iPS sac-derived suspension cells 2 days after harvest of spherical cells from ES/iPS sacs (day 17), to test whether suspension cells might contain more hematopoietic cells and less endothelial cells. We observed 10.9-13.9 fold greater amounts of CD34+CD45+ HSPCs in both MSC- and EP- derived iPS sacs, compared to FB-derived iPS sacs (p<0.01) (Figure 4A). 3.2-16.4 fold greater amounts of GPA+CD41a+ megakaryoerythroid progenitors were also observed in MSC-derived iPS sacs, compared to EP- and FB-derived iPS sacs (p<0.05) (Figure 4B). Additionally, we observed 6.2-22.5 fold greater amounts of GPA-CD41a+ megakaryocyte progenitors in MSC-derived iPS cells, as compared to EP- and FB-derived iPS sacs and ES sacs (p<0.01) (Supplementary figure 3), suggesting more efficient generation of megakaryoerythroid progenitors. In CFU assays, 1.8-40.4 fold greater amounts of erythroid, myeloid, and mixed colonies were observed in MSC-derived iPS sacs, compared to EP- and FB-derived iPS sacs (p<0.01) (Figure 4C). In addition, 1.7-8.4 fold greater amounts of apoptotic cells were observed in EP-derived cells, compared to MSC- and FB-derived cells (Figure 4D), suggesting that EP-derived iPS sacs contains greater amounts of cells which could not be differentiated to erythroid cells. These data suggest that MSC-derived iPS sacs more efficiently produce immature HSPCs and immature EPs, which results in more efficient generation of erythroid cells with β-globin production.

Figure 4. More efficient generation of megakaryoerythroid progenitors in MSC-derived iPS sacs.

(A) We evaluated ES/iPS sac-derived suspension cells 2 days after harvest from ES/iPS sacs (day 17). Greater amounts of CD34+CD45+ HSPCs were observed in both MSC- and EP- derived iPS sacs, compared to FB-derived iPS sacs. (B) Greater amounts of GPA+CD41a+ megakaryoerythroid progenitors were observed in MSC-derived iPS sacs, compared to EP- and FB-derived iPS sacs. (C) In colony forming unit assays, greater amounts of erythroid, myeloid, and mixed colonies were observed in MSC-derived iPS sacs, compared to EP- and FB-derived iPS sacs. (D) Relatively greater amounts of apoptotic cells were observed in EP-derived cells. **p<0.01, *p<0.05 evaluated by Tukey’s HSD test.

Discussion

We generated MSC- and EP-derived iPS cells from SCD patients (Figure 1) based upon the notion that EP-derived iPS cells would generate erythroid cells more efficiently, and surprisingly demonstrated that MSC-derived iPS sacs allow for more efficient erythroid cell generation with higher β-globin production, as compared to not only FB-derived iPS cells but also EP-derived iPS cells (Figure 2C). Even when compared to H1 ES cells, efficient production of erythroid cells (54.6%) can be obtained from MSC-derived iPS sacs, which express equivalent amounts of β-globin (21.5%) and ε-globin (0.63%) (Figure 2D). The MSC-derived iPS cells more efficiently produce cells with an immature HSPC phenotype in iPS sacs (Figures 3C and D), which could result in the generation of β-globin expressing definitive erythroid cells. Furthermore, the erythroid cell generation is strongly affected by the cell source for iPS cells generation (Supplementary figure 2). Our findings could be meaningful for development of alternative RBC transfusion methods using in vitro erythroid differentiation. Importantly, the erythroid differentiation methods that we have developed should prove useful for the evaluation of potential novel therapeutic applications in RBC disease, such as an iPS cell-based RBC disease model, i.e., genome editing at the β-globin locus to treat iPS cells derived from cells in SCD patients.

Before starting this study, we expected that EP-derived iPS cells should be the best resource to generate erythroid cells in vitro. However, MSC-derived iPS cells generated greater amounts of definitive erythroid cells with higher β-globin expression, as compared to EP-derived iPS cells (Figures 2D and E). We previously demonstrated that CD34+ cells with an HSPC phenotype contained within ES sacs can be differentiated to definitive erythroid cells with higher β-globin expression, while GPA+ cells with an erythroid progenitor phenotype within ES sacs predominantly produces ε-globin after further maturation [32]. These data suggest that greater emergence of immature HSPCs may be more crucial to generate definitive erythroid cells than direct differentiation from iPS cells to erythroid cells (not mediated by HSPCs). In this study, MSC-derived iPS cells more expanded during a late phase of erythroid differentiation (after day 22), as compared to EP- and FB-derived iPS cells, and this pattern was similar to the cell growth curve of ES cells (Figure 3A). In addition, we detected higher amounts of β-globin and lower amounts of ε-globin in erythroid cells differentiated from MSC-derived iPS cells, as compared to EP- and FB-derived iPS cells, which was comparable to ES cell-derived erythroid cells (Figures 2D and E). These data suggest that MSC-derived iPS sacs could more efficiently induce definitive hematopoiesis (or erythropoiesis), and efficient definitive erythropoiesis might result from greater emergence of immature HSPCs on a niche-like place within iPS sacs.

We then investigated the hypothesis that MSC-derived iPS sacs could more efficiently produce immature hematopoietic progenitors including hemogenic endothelia and megakaryoerythroid progenitors. We observed greater amounts of VEGFR2+GPA- immature HSPCs (day 15) (Figures 3C and D) and GPA+CD41a+ megakaryoerythroid progenitors (day 17) (Figures 3E and 4B) in MSC-derived iPS cells, as compared to EP-derived iPS cells, which may explain the more efficient definitive erythroid generation with higher β-globin expression we observed. In addition, greater amounts of erythroid, myeloid and mixed colonies were observed in MSC-derived iPS cells, as compared to EP- and FB-derived iPS cells (Figure 4C). These data also suggest that MSC-derived iPS sacs contain greater amounts of hematopoietic (and erythroid) progenitor cells. Interestingly, when we evaluated more specifically hemogenic endothelium population (day 15) using detailed cell surface analysis (VE-cad+CD43-CD73-DDL4-GPA-) [29, 51], greater amounts of hemogenic endothelia were observed in EP-derived iPS sacs, as compared to MSC- and FB-derived iPS sacs and ES sacs (Figure 3F), while EP-derived iPS sacs produced fewer amounts of CD34+CD45+ HSPCs (Figure 2B) and GPA+ erythroid cells (Figure 2C), as compared to MSC-derived iPS sacs. These data suggest that greater amounts of hemogenic endothelia were generated in EP-derived iPS sacs, while the hemogenic endothelia were less efficiently differentiated to HSPCs as compared to MSC-derived iPS sacs.

We observed greater amounts of VEGFR2+GPA- immature HSPCs in MSC-derived iPS sacs as compared to ES cells, while fewer amounts of erythroid cells were obtained from MSC-derived iPS sacs as compared to ES cells. The iPS cell-derived erythroid cell production was reported to decrease by apoptosis during erythroid differentiation, when using a traditional EB-based differentiation method [52]. In our iPS sac-based method, apoptotic cell amounts relatively increased in EP- and MSC-derived cells among the suspension cells at day 17, as compared to ES cells (Figure 4D). The apoptosis may explain the less efficient erythroid cell generation from EP- and MSC-derived iPS cells than ES cells. In addition, we observed relatively less apoptosis in MSC-derived iPS cells as compared to EP-derived iPS cells, which may explain the greater amounts of erythroid cell generation from MSC-derived iPS cells.

The process of ES/iPS sac generation could be suitable for not only erythroid cell generation but also platelet generation [23]. In the current study, we observed greater amounts of GPA-CD41a+ megakaryocyte progenitors in MSC-derived iPS cells at day 17, as compared to EP- and FB-derived iPS cells and ES cells (Supplementary figure 2). These data also suggest that MSC-derived iPS cells can efficiently generate megakaryoerythroid progenitors. MSC-derived iPS sacs might therefore be suitable for in vitro platelet generation.

We established an RP-HPLC method to separately analyze β-globin protein and βS-globin protein in ES/iPS-sac derived erythroid cells. We confirmed βS-globin protein production in erythroid cells from MSC-derived iPS sacs (Figure 2E). The βS-globin protein amounts are similar to β-globin amounts in ES sac-derived erythroid cells, while we could not detect the βS- or β-globin protein in erythroid cells in EP- and FB-derived iPS sacs. Now, we are performing gene correction of the sickle mutation in an MSC-derived iPS cell line, and the RP-HPLC should allow us to evaluate both β-globin and βS-globin production in iPS sac-derived erythroid cells with gene correction.

SCD is caused by the point mutation in the β-globin intron 1. β-globin expression is confined to erythroid cells, and specific conditions (such as high-concentration βS-globin and low oxygen) are required for sickling in vitro [53]. Previous data demonstrated a gene correction in iPS cells with the sickle mutation, noted similar erythroid differentiation in gene-corrected and uncorrected iPS cells as compared to normal iPS cells [54]. These data suggest that the sickle mutation does not strongly affect iPS cell biology.

When we generated erythroid cells from MSC-derived iPS sacs, we obtained 88 (70-107) erythroid cells per single iPS cell, which are around half amounts of ES sac-derived erythroid cells (160 per cell) and comparable to in vitro erythroid differentiation from mobilized CD34+ cells [30]. In this study, we observed more efficient definitive erythroid differentiation from MSC-derived iPS sacs (Figure 2C), while greater amounts of hemogenic endothelia were obtained in EP-derived iPS sacs (Figure 3F). If hemogenic endothelium generation in EP-derived iPS sacs could be followed by definitive erythroid differentiation in MSC-derived iPS sacs, total amounts of definitive erythroid cells might be improved. Regarding the erythroid differentiation culture, around 10-fold more efficient erythroid cell production from human CD34+ cells was reported in a previous study [55], as compared to our erythroid cell production in the same protocol (maybe due to the lot of FBS). Our iPS sac-derived erythroid differentiation protocol is ideal as an in vitro erythroid cell production assay at present; however, further optimization to increase expansion would be required for clinical application.

In summary, we demonstrated that human MSC-derived iPS sacs allow for more efficient erythroid cell generation with higher β-globin production, and we provide evidence that this occurs through a heightened emergence of immature hematopoietic progenitors (likely including hemogenic endothelia) in MSC-derived iPS sacs. The reliable production of β-globin is important for modeling erythropoiesis in vitro as efforts to produce erythroid cells from iPS cells from patients with SCD has been hampered by the lack of production of the β-globin equivalent in SCD leading to sickle hemoglobin production. This system will now allow comprehensive testing of genetic strategies aimed at correction of the SCD mutation, an important advance for eventual clinical translation of these strategies. Our findings should also be important for in in vitro iPS cell-derived erythroid cell generation with high β-globin expression.

Supplementary Material

Supp Fig S1. Supplementary figure 1. The morphology of ES/iPS sac-derived erythroid cells.

(A) After 30 day culture, we observed eosinophilic erythroid cells with a high density of chromatin which were differentiated from iPS sacs as well as ES sacs, evaluated by Wright-Giemsa staining. (B) There was no significant difference of enucleated cell percentages among all groups.

Supp Fig S2. Supplementary figure 2. ES/iPS sac-derived erythroid cell generation more strongly affected by cell sources.

We examined the amounts of GPA+ erythroid cells among each clone of ES/iPS cells at day 30. The ES/iPS sac-derived erythroid cell generation was more strongly affected by cell sources (80.8%) than variations among iPS cell clones (19.2%). In addition, we evaluated the amounts of GPA+ erythroid cells from MSC-derived iPS sacs between MOIs 15 and 25 (which were used in reprograming vector transduction to generate iPS cells). However, the ES/iPS sac-derived erythroid cell generation was still more strongly affected by cell sources (64.5%) than variations among MOIs (35.5%). The standard error of the mean was shown as error bars.

Supp Fig S3. Supplementary figure 3. More efficient emergence of megakaryocyte progenitors in MSC-derived iPS cells.

We observed greater amounts of GPA-CD41a+ megakaryocyte progenitors in MSC-derived iPS cells before erythroid differentiation (day 17), as compared to EP- and FB-derived iPS cells and ES cells. **p<0.01, *p<0.05 evaluated by Tukey’s HSD test.