Blood includes cells of various structures and functions such as oxygen-transporting red blood cells, immunological white blood cells, and hemostasis-mediating platelets. All of these cells are derived from hematopoietic stem cells (HSCs) from various intermediate stages. HSCs are defined as cells possessing the ability to self-renew, in other words, the ability to reproduce themselves, as well as the ability to differentiate into all lineages of blood cells (Morrison et al., 1995, Weissman, 2000). The life of most blood cells is relatively short, and since they are turned-over continuously, HSCs are critical not only for developmental stages but also for maintenance of hematopoietic homeostasis. The dysfunction of HSCs/progenitor cells (PCs) has also been shown to lead to various blood disorders; therefore, HSCs/PCs can serve as therapeutic cellular targets for the treatment of hematological diseases.
One established therapy using HSCs is the HSC transplantation (HSCT). There are currently three different sources of HSCs for HSCT: umbilical cord blood (CB), bone marrow (BM), and mobilized peripheral blood (PB). Each source, however, provides only a limited number of HSCs/PCs. Therefore, advances in technologies such as ex vivo culture of HSCs and manipulation of embryonic stem cells (ESCs) and induced pluripotent stem cells (iPSCs) in order to derive HSCs will have direct benefits on both current and novel therapeutics.
In this article, we review methods to identify HSCs and attempts at ex vivo expansion of HSCs/PCs using cytokines, growth factors, niche cells that control the self-renewal, differentiation, and homing of HSCs, small-molecule compounds, and genetic modulation of transcription factors. We also focus on recent approaches to induce HSCs/PCs and mature blood cells from pluripotent stem cells.
DEVELOPMENT OF HSCs
During mammalian development, two waves of hematopoiesis occur in sequential stages: first, a transient wave of primitive hematopoiesis followed by definitive hematopoiesis. These stages are temporally and anatomically distinct, involving unique cellular and molecular regulators. Production of blood cells is started in primitive hematopoiesis that is first detected in the yolk sac as early as embryonic day 7.5 (E7.5) in mice (Palis et al., 1999) (Fig. 1). The first mature blood cells are found closely associated with endothelial cells in the yolk sac structures called blood islands. That hematopoietic and endothelial lineages are produced simultaneously at the same anatomical site led to the hypothesis that these cells are generated from a mesodermal precursor with both endothelial and hematopoietic potential, a cell termed the hemangioblast (Sabin, 1920, Murray, 1932). The extraembryonic yolk sac is considered as the first site of emergence of the hemangioblast. Hemangioblasts differentiate into a hemogenic endothelium intermediate, which gives rise to primitive hematopoietic cells (Lancrin et al., 2009). Primitive hematopoiesis primarily generates nucleated primitive erythrocytes transiently, followed by definitive hematopoiesis which generates all hematopoietic lineages, including enucleated definitive erythrocytes and HSCs with a long-term repopulating activity.
Within the embryo, definitive hematopoiesis takes place in various places (Fig. 1). HSCs emerge directly from a small population of endothelial cells in the conceptus, referred to as “hemogenic endothelium” (Bertrand et al., 2010; Boisset et al., 2010; Kissa et al., 2010). Hemogenic endothelium is located at all sites of HSC emergence, including the ventral aspect of the dorsal aorta in the aorta-gonad-mesonephros (AGM) region, the vitelline and umbilical arteries, the yolk sac, and the placenta. Recently, midgestation embryonic head was also characterized as a site for HSC development (Li et al., 2012). The process by which blood forms from hemogenic endothelium involves an endothelial-to-hematopoietic cell transition during which individual cells bud out and detach from the endothelial layer (Bertrand et al., 2010; Boisset et al., 2010; Kissa et al., 2010). Hemogenic endothelium is distinguished from all other endothelial cells by the presence of a transcription factor called Runx1 (Chen et al., 2009). Runx1 is expressed in hemogenic endothelial cells, in newly formed hematopoietic cell clusters, and in all functional HSCs (North et al., 2002, 1999). HSCs arising from the hemogenic endothelium migrate first to the fetal liver, and then to the spleen. Eventually, hematopoiesis shifts to the bone marrow (BM), where homeostatic blood formation is maintained throughout life (Wang et al., 2011).
DETECTION AND PURIFICATION OF ADULT HSCs
HSCs reside in a BM niche in a nondividing state from which they occasionally are aroused to undergo cell division on average of once every 1–2 months (Sudo et al., 2000; Foudi et al., 2009). The niche is a specialized anatomic compartment composed of supporting cells and extracellular matrices that are essential for maintaining HSCs/PCs in the BM. Various cell types have been characterized as contributing to the formation of HSC niches, including osteoblasts, endothelial cells, CXCL12 abundant reticular (CAR) cells, mesenchymal progenitor cells, nonmyelinating Schwann cells ensheathing autonomic nerves, macrophages, megakaryocytes, osteoclasts, and so on (Mercier et al., 2011; Wang and Wagers, 2011). The exact contribution of each of these cells to the niche has not been fully elucidated, but recent reports have suggested that these niches are not mutually exclusive in regulating HSC maintenance and mobilization. Although there is the unresolved question of how these complicated niche cells regulate HSCs within the BM, the niche is a source of a variety of signals essential to HSCs/PCs in the form of cytokines, chemokines, growth factors, hormones, cell adhesion proteins, and extracellular–matrix components.
The improvement of technologies using the combination of fluorescence activated cell sorters (FACS) and multistaining procedures with fluorophore-conjugated monoclonal antibodies has remarkably improved the efficiency of the identification and isolation of HSCs. Extensive analyses have been performed to identify the HSC fraction by evaluating the ability of the given fractions to reconstitute the entire hematopoietic system, particularly the three major blood lineages, myeloid cells, T lymphocytes, and B lymphocytes, following transplantation into myeloablated recipients. As a result, mouse HSCs appeared to be highly enriched in the CD34low/–c-Kit+Sca-1+Lineage-marker– (CD34–KSL) cell population, a very rare cell population (0.004% of the mouse BM) with single cell transplantations reconstituting hematopoiesis in more than one out of three recipients (Osawa et al., 1996). The CD150+CD48–CD41–Lineage-marker– cell population is also highly enriched for HSCs in mice (Kiel et al., 2005).
To evaluate human HSC activity, xenotransplantation systems have been developed using immunodeficient mice such as non-obese diabetic/severe combined immunodeficiency (NOD/SCID) mice, NOD/Shi-scid/IL-2Rγnull (NOG) mice, and NOD/SCID gamma (NSG) mice (Doulatov et al., 2012). The cells that exhibit long-term reconstitution and differentiation capacity in the immunodeficient mouse are regarded as the closest cells to human HSCs and referred to as SCID-repopulating cells (SRCs). Using these assays, the frequency of SRCs in human CB, BM, and mobilized PB was reported to be 1 in 9.3 × 105, 1 in 3 × 106, and 1 in 6 × 106, respectively (Wang et al., 1997). Among various cell surface markers, the most popular cell surface marker for human HSCs/PCs is CD34 that is expressed on CB and BM cells at a rate of <5% (Krause et al., 1996). Human CD34+ cells have been shown to have engraftment potential clinically in numerous autologous and allogeneic transplantation therapies (Vogel et al., 2000). The purification of human HSCs has progressed significantly by examining new antigens along with CD34. The frequency of SRCs in Lin–CD34+CD38–CD90+CD45RA– CB cells is as high as 1 in 10 cells (Majeti et al., 2007) and, notably, Notta et al. succeeded in purifying human HSCs at single-cell resolution in the Lin–CD34+CD38–CD90+ RholoCD45RA–CD49f+ fraction (Notta et al., 2011) (Fig. 2).
ATTEMPTS TO EXPAND HSCs/PCs
The frequency of SRCs in CB is higher compared to those in BM and mobilized PB, but the absolute number of CB cells infused to patients is much lower (Rocha et al., ), resulting in higher mortality due to infection, which is most often caused by delayed recovery of neutrophils after transplantation. Thus, efforts to develop techniques for ex vivo expansion of CB HSCs/PCs have been underway for the past two decades (Nishino et al., 2012). The ultimate goals of ex vivo expansion of CB HSCs/PCs are to increase the number of progenitor cells to establish rapid recovery of neutrophils and platelets, and to augment the number of HSCs in order to increase the chances of life-long hematopoietic reconstitution after transplantation. Various trials have been taken such as liquid culture with cytokines and coculture with stromal cells. However, the clinical trials utilizing these methods have not clearly demonstrated sufficient expansion of HSCs/PCs to overcome the delayed recovery of neutrophils and platelets. To overcome this barrier in HSC/PC expansion, strategies that target signal transduction pathways involved in the self-renewal of HSCs have been exploited by a variety of technologies, and a number of protein factors have also been thoroughly screened for the ability to enhance HSC self-renewal during ex vivo liquid culture (Table 1). These factors include Angiopoietin-like (ANGPTL)−5, insulin growth factor-binding protein (IGFBP)−2, fibroblast growth factor (FGF)−1 (Zhang et al., 2008), pleiotrophin (Himburg et al., 2010), and Notch ligands (Suzuki et al., 2006; Delaney et al., 2010).
Table 1. Expansion methods for HSCs from cord blood.
Small-molecule compounds (SMCs) are also attractive tools for HSC/PC expansion (Table 1). Tetraethylenepentamine (TEPA), a copper chelator, enhances the ex vivo expansion of HSCs/PCs and is under evaluation in a clinical trial (Peled et al., 2004; de Lima et al., 2008). The 6-bromoindirubin 3′-oxime (BIO), a chemical inhibitor of glycogen synthase kinase-3β (GSK-3β), has been shown to promote the engraftment of cultured CD34+ cells in NOD/SCID mouse (Ko et al., 2011). Valproic acid, a HDAC inhibitor, and Nicotinamide, an inhibitor of the NAD-dependent HDAC sirtuin 1 (SIRT1), have also been reported to promote the expansion of human HSCs ex vivo (De Felice et al., 2005; Peled et al., 2012). Among these chemical approaches, one of the most encouraging findings is the discovery of StemRegenin 1 (SR1) that promotes the self-renewal of human HSCs through its antagonizing effect on the aryl hydrocarbon receptor (AHR) (Boitano et al., 2010). CD34+ cells treated with SR1 in the presence of stem cell factor (SCF), thrombopoietin (TPO), Flt3 ligand, and IL-6 for 2 weeks contained >10-fold more short and long term SRCs as compared to uncultured cells or cells cultured with cytokines alone. The positive effect of SR1 on HSC/PC expansion was also confirmed in an automated culture system employing a controlled fed-batch media approach (Csaszar et al., 2012). SR1 has revealed a physiological role for AHR in controlling stemness and also has suggested that signaling molecules downstream of AHR may be promising targets for modulation by SMCs in ex vivo HSC/PC expansion.
We have also screened SMCs that are active on HSCs/PCs and identified NR-101, a novel small-molecule c-MPL agonist, as a stimulator of HSC/PC expansion more potent than TPO (Nishino et al., 2009). The signaling by TPO via its receptor, c-MPL or TPOR, plays a crucial role in the maintenance of HSCs. Small-molecule TPOR agonists have recently been shown to be beneficial in the treatment of thrombocytopenia. However, their effects on HSCs have not yet been explored. During a 7-day culture of CD34+ or CD34+CD38− HSCs/PCs, NR-101 efficiently promoted the expansion of HSCs/PCs, with a greater than twofold increase compared to culture with TPO. Correspondingly, SRCs were increased 2.9-fold during a 7-day culture with NR-101 compared to freshly isolated CD34+ cells, and 2.3-fold compared to that with TPO. Of note, NR-101 persistently activated STAT5 but not STAT3. Furthermore, NR-101 induced a long-term accumulation of hypoxia-inducible factor-1α (HIF1α) protein and enhanced activation of its downstream target genes. Our findings indicate that SMCs can promote HSC expansion through the modulation of signaling pathways in HSCs. We also identified Garcinol, a plant-derived histone acetyltransferase (HAT) inhibitor, as a novel stimulator of HSC/PC expansion (Nishino et al., 2011). During a 7-day culture of CD34+CD38− HSCs/PCs supplemented with Garcinol, numbers of CD34+CD38− HSCs/PCs increased more than 4.5-fold compared to controls. Culture with Isogarcinol, a derivative of Garcinol, resulted in a 7.4-fold increase. Furthermore, a 7-day culture of CD34+ HSCs/PCs with Garcinol expanded the number of SRCs 2.5-fold compared to control cells. We also demonstrated that the capacity of Garcinol and its derivatives to expand HSCs/PCs was closely correlated with their inhibitory effects on HATs. Our findings identified Garcinol as the first naturally synthesized product that acts on HSCs/PCs and suggests that the inhibition of HATs could be an alternative approach for manipulating HSCs/PCs. With these evolving technologies to expand HSCs/PCs, the applications of CB cells in transplantation and cell-based therapies is expected to continue to grow.
APPROACHES TO GENERATE HEMATOPOIETIC CELLS FROM HUMAN PLURIPOTENT STEM CELLS
ESCs are pluripotent stem cells derived from the inner cell mass of the blastocyst (Thomson et al., 1998), and iPSCs are ESC-like cells generated by reprogramming somatic cells, most often via forced expression of a combination of transcription factors, such as Oct3/4, Nanog, KLF4, c-Myc, LIN28, and SOX2 (Takahashi et al., 2007, Yu et al., 2007). Human pluripotent stem cells provide a good model for the analysis of the development of human hematopoietic cells and are expected to be a source of numerous hematopoietic cell types for regenerative medicine (Ye et al., 2012). In particular, it is thought that establishment of robust differentiation protocols for various types of blood cells using iPSCs could solve many of the problems of immunological rejection and ethical issues in ESC-based regenerative medicine, and could lead to a cure for various hematological diseases (Togarrati and Suknuntha, 2012). When using hematopoietic cells derived from pluripotent stem cells in regenerative medicine, it is important that the method is safe, efficient, cost-effective, and functional. Extensive trials have been undertaken to efficiently derive HSCs/PCs and mature blood cells from human pluripotent stem cells.
Human ESCs and iPSCs have been demonstrated to reproduce many aspects of embryonic hematopoiesis. Embryoid body (EB) formation (Doetschman et al., 1985) and coculture on stromal feeder cells, such as OP9 (Nakano et al., 1994), S17 (Tian et al., 2006), C3H10T1/2 (Hiroyama et al., 2006), and primary stromal cell lines derived from AGM, fetal liver and fetal BM (Ledran et al., 2008) are two major strategies to induce hematopoietic cells from pluripotent stem cells (Fig. 3). A recent study has provided evidence that hematopoietic differentiation of human ESCs/iPSCs progresses through sequential stages: hemogenic endothelium, then primitive hematopoiesis, and finally definitive hematopoiesis, resembling the development of physiological hematopoiesis (Kaufman, 2009).
Induction of Hematopoietic Stem Cells From Human ESC/iPSCs
It's already been a decade since Kyba et al. first succeeded in generating and expanding HSCs capable of long-term, multilineage reconstitution from mouse ESCs (Kyba et al., 2002). Using a tetracycline-inducible HoxB4 transgene system, they induced HoxB4 expression in EBs from days 4–6 of culture. After expansion of HoxB4-induced hematopoietic cells on OP9 stromal cells, they transplanted the cells into irradiated syngeneic mice. Recently, it was reported that overexpression of Lhx2 instead of HoxB4 also confers a long-term, multilineage reconstituting capacity to ESC/iPSC-derived hematopoietic cells (Kitajima et al., 2011).
Unlike in mice, however, induction of HSCs capable of long-term, multilineage engraftment from human ESCs has yet to be achieved robustly (Table 2). Although overexpression of HoxB4 promotes the proliferation of hematopoietic cells from human ESCs, it cannot induce functional HSCs (Wang et al., 2005). Overexpression of a number of genes besides HoxB4 has been tried, but no success has been achieved so far. Alternative approaches have also have been taken to induce HSCs from human ESCs (Wang et al., 2005; Narayan et al., 2006; Tian et al., 2006; Ledran et al., 2008). Wang and his colleagues succeeded in achieving engraftment of hematopoietic cells induced from hESCs by transplanting CD45negPFV (CD45negPECAM-1+FLK-1+VE-Cadherin+) cells differentiated in EBs directly into the BM of immunodeficient NOD/SCID mice (Wang et al., 2005). However, the level of human chimerism in the BM was remarkably low. So far, Lako's group has had the most success in the derivation of HSCs from hESCs (Ledran et al., 2008). They focused on the microenvironment of hematopoietic development during fetal life, and cocultured human ESC-derived hematopoietic cells on the monolayers of cells derived from mouse AGM or fetal liver, or on the stromal cell lines derived from these embryonic tissues. Among these trials, the human ESC-derived hematopoietic cells cocultured with the AGM-derived stromal cells established about 1.3–2.5% of chimerism in mouse BM. Nevertheless, the engraftment capacity of human ESC-derived hematopoietic cells is remarkably lower than that of CB HSCs. To be able to use ESC/iPSC-derived cells in regenerative medicine, more substantial breakthroughs are needed that are based on the precise understanding of the development of HSCs. We recently showed that SOX17 plays a key role in priming hemogenic potential in endothelial cells, thereby regulating hematopoietic development from hESCs/iPSCs. We further showed that overexpression of SOX17 in human ESC/iPSC-derived endothelial cells results in expansion of hemogenic endothelium-like cells (Nakajima-Takagi et al., 2013). These accumulating data may help in the development of new technologies.
Table 2. Attempts to generate repopulating cells from human ESCs.
Number of transplanted cells
S17, mouse bone marrow-derived stromal cell line; AM20.1 B4, stromal cell line generated from the aorta/mesenchyme part of the AGM region of 10 dpc murine embryos; UG26.1 B6, stromal cell line generated from urogenital ridge of the AGM region; EL08.1 D2, stromal cell line generated from the fetal liver of 11 dpc murine embryo; AGM, dorsal Aorta, Gonads, and Mesonephros; i.v., intravenous injection; IBMT, intra-bone marrow transplantation; i.p., intraperitoneal injection; N.E., no engraftment.
Induction of Mature Hematopoietic Cells From Human ESC/iPSCs Erythroid Cells
Red blood cells (RBCs) function to transport oxygen and carbon dioxide through the body, and are the most abundant cells in the blood. For regenerative and transfusion medicine, one of the important points of erythropoiesis from human pluripotent stem cells is globin gene regulation, which changes according to the developmental stages and mediates the function of RBCs. In vivo, embryonic-type ξ-and ε-globins are expressed in primitive hematopoiesis. In definitive hematopoiesis, these globins switch to fetal-type α- and γ-globins, respectively. Subsequently, γ-globin switches to adult-type β-globin around the time of birth (Peschle et al., 1985).
Recently, several groups described efficient differentiation of human ESCs into RBCs (Chang et al., 2011). In a methylcellulose-based colony-forming cell assay, Kaufman et al. demonstrated that ES-derived erythroid cells expressed α-and β-globin, but did not express fetal γ-globin. No embryonic (ε or ξ) globin gene expression was detected, either (Kaufman et al., 2001). But subsequent studies have shown expression of both embryonic and fetal globins by ESC-derived erythroid cells (Cerdan et al., 2004; Qui et al., 2005; Zambidis et al., 2005). For example, Oliver et al. succeeded in large-scale production of RBCs, but the resulting RBCs were similar to primitive erythroid cells present in the yolk sac of early human embryos and did not enucleate. They were fully hemoglobinized and express a mixture of embryonic and fetal globins but no β-globin (Oliver et al., 2006). Eventually, Lu et al. reported a method that induces RBCs with adult β-globin (Lu et al., 2008). RBCs were generated and expanded on a large scale (1 × 1010−1011 cells/six-well plate) from human ESCs via hemangioblasts by a four-step protocol. These cells were oxygen-carrying and showed increased expression of β-globin (from 0% to > 16%), although these cells mainly expressed fetal and embryonic globins. To date, technology in this field has had success in a large scale expansion of RBCs with features of primitive hematopoiesis. However, induction of adult type of RBCs with functional maturity is one issue to resolve in the future.
Megakaryocytes and Platelets
Platelets play an essential role in hemostasis and thrombosis. Platelets are derived from mature multinucleated megakaryocyte (MK) precursors through the cytoplasmic fragmentation and are produced at a rate of 1 × 1011 platelets each day in an adult human. Mature MKs were first differentiated from human ESCs by coculture with OP9 stroma cells in the presence of TPO (Gaur et al., 2006). After 15–17 days of coculture, 20–60% of the floating and loosely adherent cells were CD41a+CD42b+, characteristic of megakaryocyte lineage cells. However, no mature functional platelets were detected from these MKs. Subsequent studies reported a differentiation system using coculture with OP9 or C3H10T1/2 stroma cells (Takamaya et al., 2012, 2008). Unique sac-like structures (ES-sac) from the human ESCs/iPSCs appeared on day 14 of culture in the presence of VEGF. Although these ES-sacs included endothelial cells and hematopoietic progenitors, collection of only the hematopoietic progenitors followed by coculture for an additional 7–11days in the presence of TPO, SCF and heparin promoted differentiation into MKs. After culture, 50–60% of cells were CD41a+CD42a+CD42b+. Mature MKs from this system can release platelets with morphology and function similar to those isolated from fresh plasma. Lu et al. demonstrated a platelet generation system under serum-and feeder-free conditions using hemangioblasts/blast cells as intermediates (Lu et al., 2011). They demonstrated an efficient method for the differentiation of human ESCs into MKs, and that the platelets subsequently generated are functionally similar to normal blood platelets in vitro as well as in living animals. Because platelets contain no genetic material, they are ideal candidates for early clinical translation involving human pluripotent stem cells.
B Cells and T Cells
B cell lineage differentiation has been achieved by sequential exposure of human ESCs to OP9 and MS5 stromal cells (Vodyanik et al., 2005). CD34+ cells derived from coculture with OP9 stromal cells were further cultured on MS5 stromal cells in the presence of SCF, Flt3-L, IL-7, and IL-3, and differentiated into lymphoid (B and natural killer cells) as well as myeloid (macrophages and granulocytes) lineages. Using similar protocol, pre-B cells that exhibit multiple genomic D-JH rearrangements were induced from human iPSCs (Carpenter et al., 2011).
Galic et al. demonstrated that human ESCs could differentiate into the T lymphoid lineage, expressing CD4, CD8, CD3, CD7, and CD1a, for the first time by coculture with OP9 cells followed by engraftment into human thymic tissues in SCID-hu (thy/Liv) mice (Galic et al., 2006), a humanized mouse model constructed by insertion of small pieces of human fetal liver and thymus under the renal capsule of severe combined immunodeficient (SCID) mice (McCune et al., 1988, Namikawa et al., 1990). The same group showed that T cell progenitors could be derived from human ESCs cultured as EBs (Galic et al., 2009).
Differentiation systems using OP9 cells that ectopically express the Notch ligand Delta-like 1 (DL1) has been used very effectively to analyze the T-lineage potential. Timmermans et al. showed that human ESC also could generate T cells via a differentiation system using OP9-DL1 cells (Timmermans et al., 2009). When hematopoietic zones (HZs) formed from human ESCs on normal OP9 feeders, they were transferred to OP9-DL1 feeders and cultured in the presence of Flt-3L, IL7, and SCF. The cells expanded and differentiated into T cells, which displayed both the TCRαβ and TCRγδ phenotypes and proliferated and secreted cytokines in response to mitogens.
Although remarkable progress has been made in the purification of HSCs/PCs and techniques for the expansion of HSCs/PCs, further efforts are still required in order to assure the efficacy and safety of expanded HSCs for transplantation and cell-based therapies. Generation of HSCs and mature hematopoietic cells from human ESCs and iPSCs also requires more technical breakthroughs in order to apply these approaches to clinical therapies. To overcome these problems, it is necessary to keep trying to elucidate the pathways fundamentally supporting the development, maintenance, and differentiation of HSCs and to assess the suitability of these pathways for practical manipulation of HSCs. Improvement in the systems we use to evaluate human HSCs in immunodeficient mice will be also important in making pre-clinical assessments. It is especially desirable to improve the engraftment rate of human HSCs/PCs. With these evolving technologies to expand and produce HSCs, the manipulation of HSCs will secure its place as an instrumental tool in transplantation and cell-based therapies.
The authors thank George Wendt for critical reading of the manuscript.