International Journal of

Progress in Hematology

HEMATOLOGY

Development of Hematopoietic Cells From Embryonic Stem Cells Akira Suzuki, Toru Nakano* Department of Molecular Cell Biology, Research Institute for Microbial Diseases, Osaka University, Osaka, Japan Received August 25, 2000; accepted September 7, 2000

Abstract Embryonic stem cells are pluripotent stem cells that can differentiate into all somatic cell lineages and germ lineage cells in vivo. In vitro differentiation capacity of the cells is rather limited compared with the in vivo pluripotency. However, differentiation into hematopoietic lineages is easily obtained, and it is a powerful tool to investigate hematopoietic development and differentiation. In this article, we describe a differentiation induction method that we established, the OP9 system, a unique method using the macrophage colony-stimulating factor–deficient stromal cell line OP9. The utility of the OP9 system includes hematopoietic development, differentiation, B-cell formation, osteoclast formation, and so on. The usefulness and limits of embryonic stem cell–derived hematopoietic cells in cell therapy are also discussed. Int J Hematol. 2001;73:1-5. ©2001 The Japanese Society of Hematology Key words: Embryonic stem cell; Cell differentiation; Hematopoiesis; Cell therapy

1. Introduction

2. Mouse ES Cells

The use of mouse embryonic stem (mES) cell lines brings about revolutionary progress in analyzing gene function at the cellular and animal levels. These cell lines, derived from the inner cell mass of preimplantation blastocysts, are particularly useful because the cells can remain immature and can give rise to not only all somatic cells but also germ cells in vivo, and because genetic manipulation, such as gene targeting by homologous recombination, can be carried out easily in the cells. Late in 1998, establishment of human ES (hES) cell lines was reported and many people, including scientists, clinicians, and even politicians, became interested in the possibility of using these cells in cell therapy, which has recently been a focus of regeneration medicine. In this article, recent progress regarding hematopoietic development from mES cells will be described, and benefits of and difficulties in the use of hES cells in hematopoietic cell transplantation will be discussed.

Stem cells are defined as immature cells possessing selfrenewal and differentiation capacities [1]. Many kinds of stem cells from various organs have been characterized, and they differ substantially in their self-renewing and differentiation abilities. mES Cells are the cells to which any strict criteria of stem cells can be applied. These cell lines are established from the inner cell mass of the blastocysts of 3.5 day-old embryos [2]. The cells are generally maintained on mouse embryonic fibroblasts and leukemia inhibitory factor (LIF), because signaling of LIF–glycoprotein 130 (gp130)– signal transducer and activator of transcription 3 (STAT3) is essential for keeping the cells immature. Once mES cells are reintroduced into the blastocysts, the cells participate in the normal developmental process and differentiate into all kinds of cells, including germline cells. Therefore, ES cells possess totipotency in cell differentiation.

3. In Vitro Differentiation of Mouse ES Cells Compared with the completeness of the in vivo differentiation capacity of mES cells, the in vitro differentiation capacity is rather limited. However, differentiation induction into various kinds of cells—including hematopoietic, vascular endothelial, cardiac muscle, neuronal, pigment, and pancreatic  cells—has been reported. There are a number of methods for differentiation induction from ES cells to hematopoietic cells. The conventional and most frequently used

*Correspondence and reprint requests: Toru Nakano, MD, PhD, Department of Molecular Cell Biology, Research Institute for Microbial Diseases, Osaka University, Yamadaoka 3-1, Suita Osaka 565-0871, Japan; 81-6-6879-8361; fax: 81-6-6879-8362 (e-mail: [email protected]).

1

2

Suzuki and Nakano / International Journal of Hematology 73 (2001) 1-5

Figure 1. OP9 hematopoietic differentiation induction system from embryonic stem (ES) cells gives rise to mature blood cells by subsequent differentiation via mesodermal cells and hematopoietic progenitors. Various applications of the method are described around the circle, but transplantable hematopoietic cells have not yet been developed by the system.

method is to simply remove the ES cells from the stimulation necessary for the immaturity of the cells (LIF and/or embryonic fibroblasts) and culture them in liquid or methylcellulose –containing media in bacteria-grade Petri dishes [3,4]. Under this culture condition, ES cells produce aggregates and generate colonies of differentiated cells known as embryoid bodies (EBs). An EB consists of 3 germ layers and resembles the egg cylinder of developing mouse embryos. Structures resembling blood islands in the yolk sac appear, and hematopoietic cells emerge. Both embryonic primitive (yolk sactype) erythrocytes and definitive hematopoietic progenitors emerge. The details of the EB differentiation induction method are omitted in this review because many excellent reviews on it have already been published [3,4]. The other method (OP9 system), which we developed, is a co-culture of ES cells on the stromal cell line OP9 [5-7]. The OP9 stromal cell line was established from newborn calvaria of op/op mice. The op/op mouse shows osteopetrosis that is caused by the deficiency of osteoclasts. The primary defect of the mutant mouse is a mutation in the coding region of the macrophage colony-stimulating factor (M-CSF) gene; the differentiation of osteoclasts is perturbed by the deficiency of MCSF [8]. Two independent research groups have shown that M-CSF has some deleterious effects on the early development of the hematopoietic system [5,9]. Differentiation induction by the OP9 system is as simple as the EB formation method.After LIF is removed, ES cells are co-cultured on OP9 cells. Then, ES cells begin to differentiate into mesodermal colonies, and the colonies are trypsinized on day 5 of the culture. Multipotential hematopoietic colonies begin to appear after reseeding onto OP9 cells. Only a small population (1 of 500-1000 cells) of the cells at day 5 of culture have hematopoietic capacity. These hematopoietic cells continue to proliferate and differentiate. Eventually, erythrocytes, macrophages,

granulocytes, megakaryocytes, and B-lineage cells develop around day 14. Below, we describe several experiments using the OP9 system and discuss its usefulness (Figure 1).

4. Development of Hematopoietic Stem Cells From Mouse ES Cells The biggest and most intriguing question is whether hematopoietic stem cells appear during differentiation induction. In steady-state hematopoiesis, such as that in bone marrow, a hierarchy of hematopoietic cells exists. At the top of the pyramid are hematopoietic stem cells with self-renewing capacity. There are hematopoietic progenitors below the hematopoietic stem cells, and they differentiate into mature blood cells. It seems reasonable that the order of appearance of hematopoietic cells during mouse development is the same as that of this steady-state hierarchy. It also seems reasonable to believe that hematopoietic stem cells would appear during the differentiation process from immature pluripotent ES cells to mature blood cells. However, no hematopoietic stem cells emerge during the process. How can this be explained? There are 2 possibilities. One is that differentiation induction from ES cells is an artificial process that cannot be seen in the normal developmental process. The other possibility is that the order of appearance of hematopoietic cells is the opposite of the hierarchy of steady-state hematopoiesis. Recent studies support the latter notion [10]. We and many other investigators believe that only the early phase of hematopoietic development can be recapitulated by differentiation induction from ES cells, and it is rather difficult to obtain self-renewing hematopoietic stem cells from ES cells. There are, however, 2 reports in which hematopoietic stem cells are seemingly developed from ES cells. One report shows that ES cell–derived lymphoid and myeloid cells were

Hematopoietic Development From ES Cells

detectable in the peripheral blood of the transplanted mice, although in low numbers [11]. The other reports the uses of stromal cells, conditioned medium of fetal liver stromal cells, and a cocktail of cytokines. In this case, the development of transplantable hematopoietic stem cells was very efficient, and their repopulating ability was very high [12]. To our knowledge, however, no other investigators have reproduced this result.

5. Common Hematopoietic Precursor for Primitive and Definitive Erythropoiesis? During mouse embryogenesis, the production of “primitive” erythrocytes (EryP) precedes the production of “definitive” erythrocytes (EryD) in parallel with the transition of the hematopoietic site from the yolk sac to the fetal liver [13]. It has long been controversial whether EryP and EryD originate from common progenitors. The OP9 system gives rise to EryP and EryD sequentially, with a time course similar to that seen in murine ontogeny [14]. From the results of the different growth factor requirements and limiting dilution analysis of precursor frequencies, we concluded that EryP and EryD developed from different precursors by way of distinct differentiation pathways. In contrast, a study using the EB formation method drew contradictory conclusions. Kennedy et al [15] showed that EryP and other hematopoietic lineages arose from a common multipotential precursor that developed within EB generated from differentiated ES cells. In response to vascular endothelial growth factor (VEGF) and c-Kit ligand, these precursors gave rise to colonies containing immature cells (blasts) expressing marker genes characteristic of hematopoietic precursors. The researchers demonstrated by kinetic study that the blast colony-forming cells represented a transient population, preceding the establishment of the primitive erythroid and other lineage-restricted precursors. How is this discrepant conclusion drawn? Kennedy et al [15] insisted that the reason is that our results were obtained at a relatively later stage of the differentiation induction; but we disagree. As we showed in our article [14], we carefully examined the time course of the appearance of the progenitor cells. We propose that the blast colony-forming cells of transient nature in the study of Kennedy et al are not common hematopoietic progenitors but common endothelial-hematopoietic progenitors, so-called hemangioblasts. In fact, Kennedy et al themselves demonstrated that blast colony-forming cells had the capacity to differentiate into both endothelial cells and blood cells [16]. There are common progenitors of EryP, EryD, and endothelial cells (hemangioblasts), but common hematopoietic progenitors that can differentiate into EryP and EryD but not endothelial cells do not seem to exist.

6. B-Lineage Development From Mouse ES Cells B-cell development from ES cells by the EB formation method is rather difficult, and special conditions, such as low oxygen, are necessary [17]. The OP9 system, however, could give rise to B-lineage cells reproducibly and efficiently [18]. Addition of Flt-3 ligand during the differentiation induction

3

facilitates the development of B-lineage cells. B-lineage cells generated in vitro from ES cells are functionally analogous to normal fetal liver–derived or bone marrow–derived B-lineage cells at 3 important developmental stages: first, they respond to Flt-3 ligand during an early lymphopoietic progenitor stage; second, they become targets for Abelson murine leukemia virus (A-MLV) infection at a pre–B-cell stage; third, they secrete immunoglobulin upon stimulation with lipopolysaccharide at a mature mitogen-responsive stage. Moreover, the ES cell–derived A-MLV–transformed pre-B cells are phenotypically and functionally indistinguishable from standard A-MLV–transformed pre-B cells derived from infection of mouse fetal liver or bone marrow. Notably, ES cell–derived A-MLV–transformed pre-B cells possess functional V(D)J recombinase activity. In particular, the generation of A-MLV transformants from ES cells will provide an advantageous system to investigate genetic modifications that will help to elucidate molecular mechanisms in V(D)J recombination and A-MLV–mediated transformation. Therefore, the OP9 system will facilitate the study of B-cell development and differentiation using targeted ES cell lines.

7. Development of Osteoclasts From ES Cells Osteoclasts belong to the family of hematopoietic lineage. A stepwise culture system with OP9 cells enables us to induce the differentiation from ES cells to osteoclasts [19]. Three phases are necessary for the differentiation induction: first, induction of hematopoiesis, along with the generation of osteoclast precursors; second, expansion of these precursors; third, terminal differentiation into mature osteoclasts in the presence of 1,25-dihydroxyvitamin D3. Although the transition of ES cells to the hematopoietic lineage was not blocked by an antibody to c-fms, later phases were dependent on signaling through this transmembrane receptor. Blockade of signaling through another tyrosine kinase–type receptor, c-Kit, did not affect any stages of osteoclastogenesis, although generation of other hematopoietic lineages was drastically reduced by the blocking antibody of c-Kit. This result is intriguing because osteoclasts in adult bone marrow are derived from c-Kit–dependent hematopoietic cells. cKit–independent osteoclastogenesis is ostensibly similar to the c-Kit–independent erythropoiesis of EryP. It may be possible that osteoclastogenesis would be categorized into primitive and definitive osteoclastogenesis.

8. Gene Expression Analysis Using the OP9 System The continuous generation of mature blood cells from hematopoietic progenitor cells requires a highly complex series of molecular events. To examine lineage-specific gene expression during the differentiation process, we developed a novel method combining LacZ reporter gene analysis with the OP9 system [20]. As was done in a model system using this method, we chose the erythrocytic and megakaryocytic differentiation pathways. Although erythrocytic and megakaryocytic cells possess distinct functional and morphologic features, these 2 lineages originate from bipotential erythromegakaryocytic progenitors and share common lineagerestricted transcription factors. A portion of the 5 flanking

4

Suzuki and Nakano / International Journal of Hematology 73 (2001) 1-5

region of the human glycoprotein IIb (IIb) integrin gene extending from base –598 to base 33 was examined in detail. Our data clearly showed that an approximately 200-base enhancer region extending from –598 to –400 was sufficient for megakaryocyte-specific gene expression. This experimental system has advantages over those using erythromegakaryocytic cell lines because the OP9 system recapitulates normal hematopoietic cell development and differentiation. Furthermore, this system is more efficient than transgenic analysis and can easily examine gene expression with null mutations of specific genes. We are continuing to improve this method, for example, through the introduction of a much longer reporter bacterial artificial chromosome (BAC) construct and use of EGFP instead of LacZ for easier analysis.

9. Differentiation Induction From Genetically Manipulated ES Cells: 2 Examples

tion of GATA-1.05 mutant ES cells along both primitive and definitive lineages was arrested in this ES cell culture system. Although the maturation-arrested primitive cells did not express detectable amounts of -globin messenger RNA (mRNA), the blast-like cells accumulated in the differentiated stage showed -globin mRNA expression at approximately 70% of that of the wild type. Erythroid-specific TER119 antigen was expressed and porphyrin was accumulated in the definitive cells. However, the levels of both were reduced to approximately 10%, indicating that maturation of definitive erythroid cells is arrested by the lack of GATA-1 with different timing from that of the primitive erythroid cells. The hematopoietic progenitor fraction of GATA-1.05 cells contains more colony-forming activity, termed CFUOP9. These results suggest that the GATA-1.05 mutation resulted in proliferation of proerythroblasts in the definitive lineage, which was not speculated from in vivo analysis.

10. OP9 System Versus EB Method In vitro differentiation induction from mutated ES cell lines is a fascinating method of analyzing gene function in hematopoietic development and differentiation. This method is especially powerful when the mutation causes a severe abnormality in erythropoiesis, because such a mutatation should be lethal to the embryo. We analyzed the function of 2 such genes, bcl-x and GATA-1. The bcl-x gene is a member of the bcl-2 gene family, which regulates apoptotic cell death in various cell lineages. There was circumstantial evidence suggesting that bcl-x might play a role in the apoptosis of cells of erythroid lineage, although there was no direct evidence. We used Bcl-X–null mES cells and showed that Bcl-X is indispensable for the production of both EryP and EryD at the end of their maturation [21]. In vivo, bcl-x–/– ES cells did not contribute to circulating EryD in adult chimeric mice that were produced by blastocyst microinjection of bcl-x–/– ES cells. bcl-x–/– EryP and EryD were produced by the OP9 system and further analysis was carried out. The emergence of immature EryP and EryD from bcl-x–/– ES cells was similar to that of bcl-x+/+ ES cells. However, prominent cell death of bcl-x–/– EryP and EryD occurred when the cells matured. The data show that the antiapoptotic function of bcl-x acts at the very end of erythroid maturation. These results suggest that erythropoietin acts to prevent apoptotic cell death of erythroid cells by 2 mechanisms—via Janus kinase-signal transducer and activation of transcription (JAKSTAT) pathway at the immature stage and via the accumulation of Bcl-X at the later maturation stage. Another good example is the analysis of the erythroidspecific zinc finger transcription factor GATA-1. Although the importance of GATA-1 in both primitive and definitive hematopoietic lineages was demonstrated in vivo, the precise roles played by GATA-1 during definitive hematopoiesis have not been clarified because of the difficulty posed by embryonic death of the GATA-1–null mice. Differentiation of GATA-1 promoter–disrupted (GATA-1.05) ES cells was analyzed using this system [22]. Because the GATA-1.05 mice die by 12.5 embryonic days because of the lack of primitive hematopoiesis, the in vitro analysis is an important approach to elucidate the roles of GATA-1 in differentiated hematopoiesis. Consistent with the in vivo observation, differentia-

Generally speaking, the OP9 system and the EB method provide similar results on the differentiation induction of double-knockout ES cells. However, flk-1–knockout ES cells show a clear discrepancy between the 2 methods [23]. Genetic studies in mice demonstrated an intrinsic requirement for the VEGF receptor Flk-1 in the early development of both the hematopoietic and endothelial cell lineages. ES cells homozygous for a targeted null mutation in flk-1 (flk1–/–) were examined for their hematopoietic potential in vitro during EB formation or when cultured on the stromal cell line OP9. Surprisingly, in EB cultures, flk-1–/– ES cells were able to differentiate into all myeloid-erythroid lineages, albeit at half the frequency of heterozygous lines. In contrast, although flk-1–/– ES cells formed mesodermal-like colonies on OP9 monolayers, they failed to generate hematopoietic clusters even in the presence of exogenous cytokines. However, flk-1–/– OP9 cultures did contain myeloid precursors, albeit at greatly reduced percentages. This defect was rescued by first allowing flk-1–/– ES cells to differentiate into EBs and then passaging these cells onto OP9 stroma. Thus, the in vivo phenotype of flk-1–/– is similar to the in vitro phenotype of the OP9 system rather than EB formation.

11. Establishment of Human ES Cells and Its Possible Application in Hematology The 2 types of mouse pluripotent stem cells that can differentiate into germ cells and all somatic cells are ES cells and embryonic germ (EG) cells, which are derived from primordial germ cells of developing embryos. Both ES and EG cells are established from not only mouse but also human sources [24-26]. It is impossible to determine if hES/EG cells have the same potential as mES/EG cells, because in vivo experiments are ethically prohibited. However, some characteristics of surface markers and their differentiation capacity in vitro and in immunocompromised mice strongly suggest that hES/EG cells have characteristics similar to those of mES/EG cells. It is reasonable to entertain the idea of using hES/EG cells for transplantation after the differentiation induction into some particular cell lineages, such as

Hematopoietic Development From ES Cells

neuronal cells or hematopoietic cells. However, we do not necessarily have an optimistic view on cell therapy after hematopoietic differentiation, because efficient differentiation induction into self-renewing hematopoietic stem cells has not been achieved, as discussed above. We feel that proper genetic manipulation would be necessary to obtain transplantable hematopoietic cells from ES cells. It is crucial to use self-renewing hematopoietic stem cells for cell therapy of ES/EG–derived hematopoietic cells, because the life span of mature blood cells is relatively short. In contrast, cell therapy using an ES/EG–derived neural lineage has some advantage, because progenitor or even postmitotic cells can be used for the transplantation. Actually, successful transplantation of ES-derived glial cells has already been reported in a rodent model [27].

12. Conclusion Hematopoietic development from ES cells is a novel and excellent experimental system to solve many important problems of hematopoietic development and differentiation, as discussed. ES/EG cell differentiation induction has progressed from the pure basic science to the clinical realm, with implications for future regeneration medicine. Although the clinical applications of the in vitro differentiation induction method are not fully defined, it is forseeable that the method will eventually become an important clinical technique.

References 1. Morrison SJ, Shah NM, Anderson DJ. Regulatory mechanisms in stem cell biology. Cell. 1997;88:287-298. 2. Robertson EJ. Derivation and maintenance of embryonic stem cell cultures. Methods Mol Biol. 1997;75:173-184. 3. Weiss MJ, Orkin SH. In vitro differentiation of murine embryonic stem cells: new approaches to old problems. J Clin Invest. 1996;97: 591-595. 4. Wiles MV. Embryonic stem cell differentiation in vitro. Methods Enzymol. 1993;225:900-918. 5. Nakano T, Kodama H, Honjo T. Generation of lymphohematopoietic cells from embryonic stem cells in culture. Science. 1994;265: 1098-1101. 6. Nakano T. Lymphohematopoietic development from embryonic stem cells in vitro. Semin Immunol. 1995;7:197-203. 7. Nakano T. In vitro development of hematopoietic system from mouse embryonic stem cells: a new approach for embryonic hematopoiesis. Int J Hematol. 1996;65:1-8. 8. Yoshida H, Hayashi S, Kunisada T, et al. The murine mutation osteopetrosis is in the coding region of the macrophage colony stimulating factor gene. Nature. 1990;345:442-444.

5

9. Mukouyama Y, Hara T, Xu M, et al. In vitro expansion of murine multipotential hematopoietic progenitors from the embryonic aorta-gonad-mesonephros region. Immunity. 1998;8:105-114. 10. Muller AM, Medvinsky A, Strouboulis J, et al. Development of hematopoietic stem cell activity in the mouse embryo. Immunity. 1994;1:291-301. 11. Hole N, Graham GJ, Menzel U, et al. A limited temporal window for the derivation of multilineage repopulating hematopoietic progenitors during embryonal stem cell differentiation in vitro. Blood. 1996;88:1266-1276. 12. Palacios R, Golunski E, Samaridis J. In vitro generation of hematopoietic stem cells from an embryonic stem cell line. Proc Natl Acad Sci U S A. 1995;92:7530-7534. 13. Zon LI. Developmental biology of hematopoiesis. Blood. 1995;86: 2876-2891. 14. Nakano T, Kodama H, Honjo T. In vitro development of primitive and definitive erythrocytes from different precursors. Science. 1996;272:722-724. 15. Kennedy M, Firpo M, Choi K, et al. A common precursor for primitive erythropoiesis and definitive haematopoiesis. Nature. 1997;386: 488-493. 16. Choi K, Kennedy M, Kazarov A, et al. A common precursor for hematopoietic and endothelial cells. Development. 1998;125:725-732. 17. Potocnik AJ, Nielsen PJ, Eichmann K. In vitro generation of lymphoid precursors from embryonic stem cells. Embo J. 1994;13: 5274-5283. 18. Cho SK, Webber TD, Carlyle JR, et al. Functional characterization of B lymphocytes generated in vitro from embryonic stem cells. Proc Natl Acad Sci U S A. 1999;96:9797-9802. 19. Yamane T, Kunisada T, Yamazaki H, et al. Development of osteoclasts from embryonic stem cells through a pathway that is c-fms but not c-kit dependent. Blood. 1997;90:3516-3523. 20. Era T, Takagi T, Takahashi T, et al. Characterization of hematopoietic lineage-specific gene expression by ES cell in vitro differentiation induction system. Blood. 2000;95:870-878. 21. Motoyama N, Kimura T, Takahashi T, et al. bcl-x prevents apoptotic cell death of both primitive and definitive erythrocytes at the end of maturation. J Exp Med. 1999;189:1691-1698. 22. Suwabe N, Takahashi S, Nakano T, et al. GATA-1 regulates growth and differentiation of definitive erythroid lineage cells during in vitro ES cell differentiation. Blood. 1998;92:4108-4118. 23. Hidaka M, Stanford WL, Bernstein A. Conditional requirement for the Flk-1 receptor in the in vitro generation of early hematopoietic cells. Proc Natl Acad Sci U S A. 1999;96:7370-7375. 24. Reubinoff BE, Pera MF, Fong CY, et al. Embryonic stem cell lines from human blastocysts: somatic differentiation in vitro. Nat Biotechnol. 2000;18:399-404. 25. Shamblott MJ, Axelman J, Wang S, et al. Derivation of pluripotent stem cells from cultured human primordial germ cells. Proc Natl Acad Sci U S A. 1998;95:13726-13731. 26. Thomson JA, Itskovitz-Eldor J, Shapiro SS, et al. Embryonic stem cell lines derived from human blastocysts. Science. 1998;282:1145-1147. 27. Brustle O, Choudhary K, Karram K, et al. Chimeric brains generated by intraventricular transplantation of fetal human brain cells into embryonic rats. Nat Biotechnol. 1998;16:1040-1044.