Here, we present human pluripotent stem cell (hPSC) culture protocols, used to differentiate hPSCs into CD34+ hematopoietic progenitors. This method uses stage-specific manipulation of canonical WNT signaling to specify cells exclusively to either the definitive or primitive hematopoietic program.
One of the major goals for regenerative medicine is the generation and maintenance of hematopoietic stem cells (HSCs) derived from human pluripotent stem cells (hPSCs). Until recently, efforts to differentiate hPSCs into HSCs have predominantly generated hematopoietic progenitors that lack HSC potential, and instead resemble yolk sac hematopoiesis. These resulting hematopoietic progenitors may have limited utility for in vitro disease modeling of various adult hematopoietic disorders, particularly those of the lymphoid lineages. However, we have recently described methods to generate erythro-myelo-lymphoid multilineage definitive hematopoietic progenitors from hPSCs using a stage-specific directed differentiation protocol, which we outline here. Through enzymatic dissociation of hPSCs on basement membrane matrix-coated plasticware, embryoid bodies (EBs) are formed. EBs are differentiated to mesoderm by recombinant BMP4, which is subsequently specified to the definitive hematopoietic program by the GSK3β inhibitor, CHIR99021. Alternatively, primitive hematopoiesis is specified by the PORCN inhibitor, IWP2. Hematopoiesis is further driven through the addition of recombinant VEGF and supportive hematopoietic cytokines. The resulting hematopoietic progenitors generated using this method have the potential to be used for disease and developmental modeling, in vitro.
Human pluripotent stem cells (hPSCs) are defined as encompassing both human embryonic stem cells (hESCs) and human induced pluripotent stem cells (hiPSCs), and have the unique capability of not only undergoing self-renewal under appropriate growth conditions, but also, the capacity to differentiate into all cell types derived from the three germ layers: endoderm, mesoderm, and ectoderm1. Due to these unique abilities, hPSCs hold great promise for regenerative medicine, disease modeling, and cell-based therapies2. While multiple cell types have been successfully differentiated from hPSCs, one significant challenge is the in vitro specification of exclusively adult-like hPSC-derived hematopoietic stem cells (HSCs) and definitive hematopoietic progenitors.
One likely barrier to the development of human HSCs from hPSCs is the presence of multiple hematopoietic programs within the human embryo3. The first program which emerges, termed "primitive hematopoiesis," originates within the extraembryonic yolk sac tissue and is best characterized by its transient production of erythroblast progenitors (EryP-CFC), macrophages, and megakaryocytes. Notably, this program does not give rise to HSCs, nor does it give rise to T and B lymphoid progenitors. However, the yolk sac does transiently give rise to restricted definitive hematopoietic progenitors, such as the erythro-myeloid progenitor (EMP4,5,6,7,8) and the erythroid-deficient lymphoid-primed multipotent progenitor (LMPP9). However, neither EMPs nor LMPPs are fully multipotent, or capable of HSC-like engraftment in adult recipients. In contrast, later in development, the classically defined "definitive" hematopoietic program is specified in the aorta-gonad-mesonephros region of the embryo proper, giving rise to all adult hematopoietic lineages, including the HSC. The specification of these intra-embryonic definitive hematopoietic cells occurs in a Notch-dependent fashion, via an endothelial-to-hematopoietic transition from hemogenic endothelium (HE)3,10,11,12,13,14. Aside from reconstitution capacity, the multilineage potential and Notch-dependence of these cells can be used to distinguish these definitive hematopoietic progenitors from the EMP and the LMPP (reviewed in references3,13).
Understanding the mechanism(s) governing primitive and definitive hematopoietic specification from hPSCs is likely critical to the reproducible production of definitive hematopoietic progenitors across a variety of hPSC lines. Until recently, hPSC differentiation protocols that could separate multipotent primitive and definitive hematopoietic progenitors did not exist15,16,17,18,19,20,21,22,23,24,25. Many approaches using fetal bovine serum (FBS) and/or stromal co-culture first outlined the hematopoietic potential of hPSC differentiation, with mixtures of primitive and definitive hematopoietic potential15,16,17,19,22,23,25. Further, many serum-free hematopoietic protocols have described the signal requirements for the specification of mesoderm from hPSCs that harbors hematopoietic potential18,20,21,24. However, as these methods still gave rise to heterogeneous mixtures of both programs, their use in clinical applications and understanding developmental mechanisms may be limited.
We have recently built on these studies, having outlined the stage-specific signal requirements for ACTIVIN/NODAL and WNT signaling in primitive and definitive hematopoietic specification from hPSC-derived mesoderm18,26. The latter was particularly unique, as its use of stage-specific WNT signal manipulation allows for the specification of either exclusively primitive or exclusively definitive hematopoietic progenitors26. During mesoderm specification, the inhibition of canonical WNT signaling with the PORCN inhibitor IWP2 results in the specification of CD43+ EryP-CFC and myeloid progenitors, with no detectable lymphoid potential. In sharp contrast, stimulation of canonical WNT signaling with the GSK3β inhibitor, CHIR99021, during the same stage of differentiation resulted in the complete absence of detectable CD43+ EryP-CFC, while simultaneously leading to the specification of CD34+CD43− HE. This population possessed myeloid, HBG-expressing erythroid, and T-lymphoid potential. Subsequent analyses identified this HE as lacking the expression of CD7327,28 and CD18428, and its hematopoietic potential was NOTCH-dependent28. Further, single-cell clonal analyses demonstrated that these definitive hematopoietic lineages could be derived from multipotent single cells28. Taken together, these studies indicate that stage-specific WNT signaling manipulation can specify either pure primitive hematopoietic progenitors, or multipotent NOTCH-dependent definitive hematopoietic progenitors.
Here, we outline our differentiation strategy that yields exclusively primitive or definitive hematopoietic progenitors, via manipulation of canonical WNT signaling during mesodermal patterning, and their downstream hematopoietic lineage assays. This protocol is of great value to investigators who are interested in the production of either primitive or definitive hematopoietic progenitors from hPSCs for regenerative medicine applications.
1. Reagents
2. Mesoderm Differentiation of hPSCs
3. Specification of CD34+ Hematopoietic Progenitors
4. Enzymatic Dissociation of EBs and Hematopoietic Progenitor Isolation
5. Endothelial-to-hematopoietic Transition
6. CFC Assay
7. T Cell Assay to Establish Definitive Hematopoietic Potential
A schematic depicting the induction of primitive and definitive hematopoietic progenitors from hPSCs is illustrated in Figure 1. Mesoderm patterning by canonical WNT signaling occurs during days 2 – 3 of differentiation, followed by hematopoietic progenitor specification.
Representative flow cytometric analysis and colony forming methylcellulose assays of hPSC-derived hematopoietic differentiation cultures are shown in Figure 2. IWP2-treated differentiation cultures will yield a distinct CD34+CD43− and CD34low/−CD43+ population, while CHIR-treated differentiation cultures will have <1% CD43+ cells on Day 8 of differentiation (Figure 2A,B)26. Primitive hematopoietic progenitors derived under IWP2 conditions isolated on Day 9 will give rise to predominantly primitive erythroid (EryP-CFC) and myeloid progenitors in the methylcellulose assays, while CHIR-treated cultures will be largely devoid of CFC potential at this stage (Figure 2C,F). Instead, the CD34+CD43– population derived from the CHIR99021 treatment is a heterogeneous population, containing endothelial progenitors, as well as CD34+CD43–CD73–CD184– HE (Figure 2B), which when isolated by FACS, will initially form a monolayer of endothelial-like cells (Figure 2D, left panel), that then undergoes an endothelial-to-hematopoietic transition, resulting in round, refractile non-adherent hematopoietic cells (Figure 2D, right panel). These hematopoietic cells can be assessed by flow cytometry for the expression of CD34 and CD45 (Figure 2E). Hematopoietic progenitors isolated from the EHT assays can be used in methylcellulose assays and give rise to large burst-like erythroid colony-forming units (BFU-E), small erythroid colony-forming units (CFU-E), and myeloid progenitors (CFU-M; Figure 2F).
Figure 3 depicts representative flow cytometric analyses of the T-lymphocyte assays potential from CHIR99021-derived CD34+CD43−CD73−CD184− hematopoietic progenitors (Figure 3A) and IWP2-derived (Figure 3B) CD34+CD43− hematopoietic progenitors, following 21 days of OP9-DL4 co-culture. The presence of a CD45+CD56−CD4+CD8+ population in CHIR-derived progenitors is indicative of definitive hematopoietic potential within the input CD34+ population. CD34+ and CD43+ progenitors isolated from IWP2-derived differentiations do not give rise to T-lymphocytes in this assay, under these differentiation conditions18,26.
Figure 1: Schematic diagram of protocol for the specification of definitive and primitive hematopoietic progenitors. Depiction of the major experimental steps and timelines of the differentiation from EBs to definitive and primitive hematopoietic progenitors. EBs are generated on Day 0 from hPSCs on MAT-coated plasticware. EBs are grown in SFD media containing BMP4 in a multi-gas incubator. On Day 1 of differentiation, cultures are fed with media supplemented with BMP4 and bFGF. On Day 2, a media change occurs with WNT manipulation through the addition of CHIR99021 (CHIR) for definitive hematopoietic progenitors and IWP2 for primitive hematopoietic progenitors. On Day 3, media is changed to SP34, supplemented with VEGF and bFGF. At Day 6, cultures are fed with media supplemented with hematopoietic cytokines. On Day 8 of the primitive hematopoietic specification, the media is changed and cells are moved to a 5% CO2 incubator for 24 h, followed by the hematopoietic progenitor cell (HPC) assay. On Day 8 of definitive hematopoietic specification, CD34+CD43–CD73–CD184– cells are isolated by FACS and assayed for definitive hemogenic endothelial potential, or T-lymphoid potential (not depicted). Please click here to view a larger version of this figure.
Figure 2: Analysis of differentiation cultures and generation of primitive hematopoietic progenitors. (A) Representative flow cytometric analyses of primitive hematopoietic progenitors from IWP2-treated EBs on Day 8. (B) Representative flow cytometric analyses from CHIR99021-treated EBs on Day 8. Hemogenic endothelium (HE) can be identified and isolated as a CD34+CD43–CD73–CD184– population. (C) Colony-forming potential of day 9 hematopoietic progenitors from IWP2- and CHIR99021- (CHIR) treated differentiation cultures. n = 3, error bars represent SD of the mean, asterisks indicate statistical significance, using student's t-test *** p <0.0001. (D) Representative photographs of isolated HE cells after 4 days (left) or 7+ days (right) of growth on MAT-coated plasticware. Original magnification, 100X. Scale bars = 100 µm. (E) Representative flow cytometry of CD34 and CD45 expression from definitive hematopoietic progenitors after 9 days of culture. (F) Representative photograph showing morphology of EryP-CFC (left), CFU-M (middle), BFU-E and CFU-E (right). Original magnification, 40X. Scale bars = 100 µm. Arrows indicate named colonies. Please click here to view a larger version of this figure.
Figure 3: Analysis of definitive hematopoietic potential. Representative flow cytometric analyses of T lymphocyte potential of CHIR-derived CD34+CD43−CD73−CD184− hematopoietic progenitors (A) or IWP2-derived (B) CD34+CD43− hematopoietic progenitors, 28 days after co-culture with OP9-DL4 cells. The T-lymphocyte potential from a sample is considered positive if there are more than 100 CD45+ events detected. Please click here to view a larger version of this figure.
SFD Media | Day 0 | Day 1 | Day 2 Definitive | Day 2 Primitive |
BMP4 | 10 ng/mL | 10 ng/mL | 10 ng/mL | 10 ng/mL |
bFGF | – | 10 ng/mL | 5 ng/mL | 5 ng/mL |
Activin A | – | – | – | 1 ng/mL |
CHIR99021 | – | – | 3 µM | – |
IWP2 | – | – | – | 3 µM |
Table 1: SFD-based media for Days 0–2.
SP34 Media | Day 3 | Day 6 | Day 8 | HE |
VEGF | 15 ng/mL | 15 ng/mL | – | 5 ng/mL |
bFGF | 5 ng/mL | 5 ng/mL | – | 5 ng/mL |
SCF | – | 200 ng/mL | 100 ng/mL | 100 ng/mL |
EPO | – | 4 IU | 2 IU | 2 IU |
IL-6 | – | 20 ng/mL | 10 ng/mL | 10 ng/mL |
IL-11 | – | 10 ng/mL | 5 ng/mL | 5 ng/mL |
IGF-1 | – | 50 ng/mL | 25 ng/mL | 25 ng/mL |
TPO | – | – | 30 ng/mL | 30 ng/mL |
Flt-3L | – | – | 10 ng/mL | 10 ng/mL |
IL-3 | – | – | 30 ng/mL | 30 ng/mL |
BMP4 | – | – | – | 10 ng/mL |
SHH | – | – | – | 20 ng/mL |
Angiotensin II | – | – | – | 10 µg/mL |
Losartan potassium | – | – | – | 100 µM |
Table 2: SP34-based media for Days 3–8 and HE.
This protocol describes a rapid, serum-free, stroma-free method for the differentiation of either primitive or definitive hematopoietic progenitors. Mesodermal specification of either primitive or definitive hematopoietic progenitors can be reliably achieved using our protocol, which uniquely exploits small molecule inhibitors of canonical WNT signaling. Stage-specific WNT activation by the GSK3β inhibitor CHIR9902133 gives rise to definitive hematopoietic mesoderm, whereas WNT inhibition by the PORCN inhibitor IWP234 specifies primitive hematopoietic mesoderm26. Of note, this method does not robustly give rise to an engraftable HSC-like population (not shown). However, the definitive hematopoietic progenitors generated with this approach are amenable to transgene-induced engraftment potential35, highlighting their potential in future translational applications.
A crucial determinant to successful hPSC differentiation is the initial size of the EBs that are formed from hPSCs. Ideally, the EBs should be around 6 – 10 cells in size. The differentiation cultures can aberrantly specify a mixture of both programs if the EBs are too large, possibly due to improper signal activation within the EB. Differentiation cultures should be visually inspected for optimal EB size within the first 24 h of differentiation. Alternatively, if the hPSCs are completely dissociated to single cells, the hPSCs instead undergo anoikis cell death36, resulting in poor hematopoietic differentiation.
A successful differentiation will give rise to exclusively primitive hematopoietic progenitors when IWP2 is used during mesodermal specification on Days 2 and 3 of this differentiation protocol. On Day 8 of differentiation, the IWP2-derived culture should be comprised of distinct CD34+CD43−, CD34mid/lowCD43+, and CD34−CD43+ populations (Figure 2A). The CD34mid/lowCD43+ population gives rise to primitive erythroid and myeloid progenitors; while the CD34−CD43+ population primarily gives rise to erythroid progenitors with limited myeloid potential18. The primitive hematopoietic potential can be confirmed by the methylcellulose assay for the presence of EryP-CFC (Section 6), which will predominately express the embryonic globin HBE in comparison to fetal HBG26,28,37. Similarly, the presence of CD43+ hematopoietic progenitors at this stage can be reliably used to indicate the presence of primitive hematopoietic progenitors18,23,26. It should be noted that CD43 expression is not exclusive to progenitors of the primitive hematopoietic program18,23,26, and cannot be used as the sole metric to assess primitive hematopoietic potential, as the immunophenotype and NOTCH-dependence of the human EMP remains uncharacterized (reviewed in3).
When CHIR99021 is used during mesodermal patterning during days 2 and 3 of differentiation, a population of exclusively definitive hematopoietic progenitors will be specified. On Day 8 of the differentiation protocol, CHIR99021-derived cultures should be comprised of a distinctive CD34+CD43− population, with little to no CD43 expression26. The CD34+CD43− population is a heterogeneous population of endothelial cells with hematopoietic, venous, and arterial potential that can be demarcated based on CD73 and CD184 expression, with HE cells lacking the expression of both27,28. When definitive HE is isolated after undergoing a NOTCH-dependent endothelial-to-hematopoietic transition (Section 5)28 it will yield CD34+CD45+ hematopoietic progenitors that will give rise to BFU-E and myeloid cells in methylcellulose, but not EryP-CFC26,28 (Figure 2). In addition, BFU-E colonies can be isolated and assessed for globin analysis, which will predominately express the fetal globin HBG in comparison to embryonic HBE (not shown)26,28,37. In contrast, the lymphoid potential can be directly assessed from isolated HE, as the NOTCH-dependent endothelial-to-hematopoietic transition occurs on the OP9-DL4 stroma18,26,28. The definitive HE specified with this method contains erythro-myelo-lymphoid progenitors at a clonal level28, which functionally distinguishes it from our current understanding of EMP or LMPP progenitors3. As such, the presence of T-lymphoid and HBG-expressing erythroid potential, coupled with an absence of EryP-CFC potential, can be used to reliably indicate the definitive hematopoietic specification from hPSCs. Of note, solely assessing for progenitors that can give rise to HBG-expressing erythroblasts may not accurately determine the definitive potential, as, similar to that described above, the human EMP, and its signal requirements, has not been characterized to-date (reviewed in reference3).
This simple differentiation strategy is very powerful as it allows for the generation of exclusively primitive or exclusively definitive hematopoietic progenitors, enabling disease modeling comparative studies of the two programs from patient-derived iPSCs or gene-modified hPSCs. Similarly, human developmental processes can be interrogated, such as discovering what additional signal pathway(s) are required to confer self-renewal capacity, and hence HSC-like potential, from the definitive hematopoietic progenitors.
In summary, WNT manipulation during mesodermal patterning in hPSCs specifies primitive CD43+ or definitive CD34+CD45+ hematopoietic progenitors, which can be isolated and used to study hPSC-derived hematopoiesis.
The authors have nothing to disclose.
This work has been supported by the Department of Internal Medicine, Division of Hematology, Washington University School of Medicine. CD was supported by T32HL007088 from the National Heart, Lung, and Blood Institute. CMS was supported by an American Society of Hematology Scholar Award.
Iscove's Modified Dulbecco's Medium (IMD) | Corning | 10-016 | |
Fetal Bovine Serum (FBS), ES cell rated | Gemini Bioproducts | 100-500 | |
Fetal Bovine Serum (FBS) | Hyclone | SH30396.03 | |
L-glutamine, 200 mM solution | Life Technologies | 25030-081 | |
Penicillin-streptomycin | Life Technologies | 15070-063 | |
0.25% Trypsin-EDTA | Life Technologies | 25200056 | |
0.05% Trypsin-EDTA | Life Technologies | 25300054 | |
Gelatin, porcine skin, Type A | Sigma-Aldrich | G1890 | |
Alpha-MEM | Life Technologies | 12000-022 | |
DMEM-F12 | Corning | 10-092-CV | |
Knock-out serum replacement | Life Technologies | 10828028 | "KOSR" |
Non-essential amino acids (NEAA) | Life Technologies | 11140050 | |
b-mercaptoethanol, 55 mM solution | Life Technologies | 21985023 | |
Hydrochloric acid | Sigma-Aldrich | H1758 | |
Fraction V, Bovine Serum Albumin | Fisher Scientific | BP1605 | |
Ham's F12 | Corning | 10-080 | |
N2 supplement | Life Technologies | 17502048 | |
B27 supplement, no vitamin A | Life Technologies | 12587010 | |
Stempro-34 (SP34) | Life Technologies | 10639011 | "SP34" |
Growth factor reduced Matrigel | Corning | 354230 | "MAT" |
L-absorbic acid | Sigma-Aldrich | A4403 | |
Human serum transferrin | Sigma-Aldrich | 10652202001 | |
Monothioglycerol (MTG) | Sigma-Aldrich | M6145 | |
Collagenase B | Roche | 11088831001 | |
Collagenase II | Life Technologies | 17101015 | |
DNaseI | Calbiochem | 260913 | |
Phosphate Buffered Saline (PBS) | Life Technologies | 14190144 | |
bFGF | R&D Systems | 233-FB | |
BMP4 | R&D Systems | 314-BP | |
Activin A | R&D Systems | 338-AC | |
VEGF | R&D Systems | 293-VE | |
SCF | R&D Systems | 255-SC | |
IGF-1 | R&D Systems | 291-G1 | |
IL-3 | R&D Systems | 203-IL | |
IL-6 | R&D Systems | 206-IL | |
IL-7 | R&D Systems | 207-IL | |
IL-11 | R&D Systems | 218-1L | |
TPO | R&D Systems | 288-TP | |
EPO | Peprotech | 100-64 | |
Flt-3 ligand (FLT3-L) | R&D Systems | 308-FK | |
CHIR99021 | Tocris | 4423 | |
IWP2 | Tocris | 3533 | |
Angiotensin II | Sigma-Aldrich | A9525 | |
Losartan Potassium | Tocris | 3798 | |
CD4 PerCP Cy5.5 Clone RPA-T4 | BD Biosciences | 560650 | Dilution 1:100; T cell assay |
CD8 PE Clone RPA-T8 | BD Biosciences | 561950 | Dilution 1:10; T cell assay |
CD34 APC Clone 8G12 | BD Biosciences | 340441 | Dilution 1:100; EHT assay |
CD34 PE-Cy7 Clone 8G12 | BD Biosciences | 348801 | Dilution 1:100; Hemogenic endothelium |
CD43 FITC Clone 1G10 | BD Biosciences | 555475 | Dilution 1:10; Hemogenic endothelium |
CD45 APC-Cy7 Clone 2D1 | BD Biosciences | 557833 | Dilution 1:50; T cell assay |
CD45 eFluor450 Clone 2D1 | BD Biosciences | 642284 | Dilution 1:50; EHT assay |
CD56 APC Clone B159 | BD Biosciences | 555518 | Dilution 1:20; T cell assay |
CD73 PE Clone AD2 | BD Biosciences | 550257 | Dilution 1:50; Hemogenic endothelium |
CD184 APC Clone 12G5 | BD Biosciences | 555976 | Dilution 1:50; Hemogenic endothelium |
4',6-diamidino-2-phenylindole (DAPI) | BD Biosciences | 564907 | Dilution 1:10,000; T cell assay |
OP9 DL4 cells | Holmes, R. and J.C. Zuniga-Pflucker. Cold Spring Harb Protoc, 2009. 2009(2): p. pdb prot5156 | ||
MethoCult H4034 | Stemcell Technologies | 4034 | "MeC" |
Milli-Q water purification system | EMD Millipore | ||
5% CO2 incubator | Set at 37 C | ||
Multigas incubator | Set at 37 C, 5% CO2, 5% O2 | ||
6 well tissue culture plate | Corning | 353046 | |
24 well tissue culture plate | Corning | 353226 | |
6 well low-adherence tissue culture plate | Corning | 3471 | |
24 well low-adherence tissue culture plate | Corning | 3473 | |
35 mm tissue culture dishes | Corning | 353001 | |
Blunt-end needle, 16 gauge | Corning | 305198 | |
3 cc syringes | Corning | 309657 | |
5 mL polypropylene test tube | Corning | 352063 | |
5 mL polystyrene test tube | Corning | 352058 | |
15 mL polypropylene conical | Corning | 430791 | |
50 mL polypropylene conical | Corning | 430921 | |
2 mL serological pipette | Corning | 357507 | |
5 mL serological pipette | Corning | 4487 | |
10 mL serological pipette | Corning | 4488 | |
25 mL serological pipette | Corning | 4489 | |
Cell scrapers | Corning | 353085 | |
2.0 mL cryovials | Corning | 430488 | |
5 mL test tube with 40 µM cell strainer | Corning | 352235 | |
40 µM cell strainer | Corning | 352340 | |
Cell culture centrifuge | |||
Biosafety hood | |||
FACS AriaII or equivalent | |||
LSRii or equivalent | |||
FlowJo software | TreeStar | ||
Water bath | Set at 37 C | ||
0.22 µM filtration system | Corning | ||
Autoclave | |||
4 C refrigerator | |||
-20 C Freezer | |||
-80 C Freezer |