Mouse Hematopoietic Stem Cells and the Interaction of c-kit Receptor and Steel Factor

Transcription

1 Concise Review International Journal of Cell Cloning 9: (1991) Mouse Hematopoietic Stem Cells and the Interaction of c-kit Receptor and Steel Factor kbichi Ikuta", Diane E. Ingoliaq JdFriedman", Shelly Heimfe db, Irving L. Weissman a ahoward Hughes Medical Institute and Departments of hthology and Developmental Biology, Stanford University School of Medicine, Stanford, C!aUbrnia, USA; bcellpro, Bothell, Washington, USA Key Words. Hematopoiesis * Hematopoietic stem cell Anemia * Progenitor * Growth factor c-kit * Steel factor * Abstract. Hematopoietic stem cells (HSCs) are distinguished from other hematopoietic progenitors in bone m m by their unique ability to undergo multilineage differentiation and self-renewal. Two mouse mutations, dominant spotting (W) and steel (a), have pleiotropic effects on hematopoiesis, gametogenesis, and melanoblast development. These two mutations have been shown to be intrinsic (W) and microenvironmental (Sl) defects. Recently, molecular studies revealed that the Wand Sl loci encode the c-kit receptor and steel factor (SLF), respectively. The c-kir receptor is expressed on HSCs and hematopietic progenitors, while SLF is produced by stromal cells. SLF acts on hematopoietic progenitors synergistically with other growth factors. Here we review the effect of these mutations on mouse hematopoiesis, and show that SLF acts on HSCs and other myeloerythroid progenitors, but that it, in our hands, does not play a critical role in HSC generation or selfrenewal. Rather, SLF is the most potent co-mitosn (with IL-1, IL-3, IL-6, G-CSF, GM-CSF, or M-CSF) found that acts on these cells, but the effect of such treatments is the rather specific and massive expansion of myelcerythropoiesis, not lymphopoiesis, and perhaps at the expense of HSC self-renewal. Introduction Hematopoietic stem cells (HSCs) play an essential role in hematopoiesis. It is generally accepted that HSCs have two major characteristics: the capacity for self-renewal and the capability of multilineage differentiation. The process by which this occurs has been intensively studied because of its clinical importance, Correspondence: Dr. Koichi Ikuta, Howard Hughes Medical Institute, Beckman Center for Molecular and Genetic Medicine, Stanford University School of Medicine, Stanford, CA 94305, USA. Received July 9, 1991; accepted for publication July 9, $2.00/0 0AlphaMed Press

2 Interaction of c-kir Receptor and Steel Factor 452 and because of its utility as a model system for the study of cell differentiation. Among many mouse genetic loci whose mutations cause hereditary anemias, two loci, Wand Sl, have been analyzed intensively because of their unique effects on hematopoiesis and HSCs. These two mutations have been known to be intrinsic ( W) and microenvironmental ($1) defects. Recently, molecular studies revealed that the Wand S1 loci encode c-kit receptor and SLF, respectively. These findings have led to studies on the effect of SLF on hematopoietic progenitors and should aid in the study of HSC self-renewal and differentiation. Here we focus on recent advances in HSC identification and function and the stages of hematolymphoid maturation that utilize c-kit receptor/slf interactions. Isolation of Hematopoietic Stem Cells HSCs are distinguished from other hematopoietic progenitors in the bone marrow by their unique ability to undergo multilineage (T, B, myeloerythroid, etc.) differentiation and self-renewal. Therefore, in the strictest definition of HSCs, HSC activity can be measured experimentally by long-term, multilineage reconstitution. Initially, colony-forming units in the spleen (CFU-s) and in vitro colony assays (CFU-c) were thought to be stem cell assays, although their outcomes were nonlymphoid [I]. Later, the precursor activity for CFU-s (pre-cfu-s) was proposed as a more likely stem cell activity than CFU-s itself [2]. HSCs have been enriched from adult bone marrow and fetal liver in many laboratories using many different techniques [3-91. These include density centrifugation and fluorescence-activated cell sorting, employing as parameters the forward and right angle light scatter as well as wheat germ agglutinin (WGA) binding and H-2K or (a granulocyte-monocyte marker) expression. WGA' H-2K+ or WGA' cells were shown to be highly enriched for HSCs [4]. Rhodamine 123 (Rhl23) dull subpopulation of these cells was further shown to contain more highly enriched precursors for CFU-s [4]. By the combination of Thy-1 antigen expression at low levels (Thy-ll"), the absence or low expression of lineage markers (Lin-), (TER-W [erythroid], B220 [B cell], Mac-1 [monocyte], Gr-1 [granulocyte], CD4, and CD8 [T cell]), and Sca-1 expression (Sca-I+), Thy-1'" Lin- Sca-1' cells were isolated from adult bone marrow and fetal liver, and shown to be highly enriched for multipotent HSCs [6, 91. This population was also shown to allow long-term reconstitution of lethally irradiated mice [6, lo]. Thy-1'" Lin- Sca-1' cells can be further subdivided into Rhl23 bright and dull subpopulations, and it has been found that the Rhl23 dull population contains a 20-fold higher level of pre-cfu-s [ll]. Wand S1 Mutations Two mouse mutations, dominant white spotting ( W) and steel (Sl), have pleio-

3 Ikuta/Ingolia/Friedman/Heimfeld/Weissman 453 tropic effects on hematopoiesis, gametogenesis, and melanoblast development (Table I) [12, 131. The Wmouse has a defect intrinsic to hematopoiesis, mast cells, primordial germ cells, and melanoblasts, while the S1 mouse has a defect in the stromal cells that support the proliferation and/or migration of these cell types [ The Wand Sl defects appear to be more profound for erythroid and mast cells than lymphoid cells. Hematological studies have further defined the nature of these mutations (Table II) [ The transplantation of normal bone marrow cells into W/W recipients results in full reconstitution of all hematopoietic lineages, including T and B lymphoid cells, while W/w bone marrow cells cannot save lethally irradiated wild type mice [ On the other hand, the anemia of SZ/Sld mice is not curable by the transplantation of normal bone marrow cells, while the transplantation of S1/Sld bone marrow cells results in full reconstitution of irradiated wild type mice [23,26]. Bone marrow cells from the W/W mice give rise to macroscopic spleen colonies at a reduced frequency (ll200 of normal) [22]. In W/W spleens, normal HSCs can give rise to normal numbers of spleen colonies [22]. Conversely, S1/Sld mice contain normal numbers of CFU-s in bone marrow [23, 241. Normal HSCs can only give rise to microscopic colonies in S1/Sld spleens. Both W and Sl mice are highly radiosensitive, presumably because of the reduced efficiency of hematopoietic cell development from HSCs and hematopoietic progenitors [12]. The hematopoietic defects in W and Sl mice have been duplicated in vitro [30]. When normal bone marrow cells were cultured on nonnal or W/W stromal cells, CFU-s and CFU-c were maintained for several months. However, neither WIW bone marrow cells nor Sl/Sld stromal cells could maintain the cultures. The c-kit Receptor and SLF The similarities in these two defects suggest that some common developmental mechanism may be shared between the Wand S1 mutations. Recent studies have shown that the W locus encodes a transmembrane tyrosine protein kinase receptor, c-kit [ It has particularly high homology with the platelet derived growth factor receptor (PDGFR) and the M-CSF receptor (c-fms). On the other hand, the S1 gene encodes a growth factor protein, SLF, which binds to the c-kit receptor [ It is heavily glycosylated, and probably exists as a dimer. This is interesting because PDGF and M-CSF also function as dimers. SLF seems to be present in both transmembrane and soluble forms. A recent study has suggested that the two forms are determined by alternative splicing [MI. The product of the Sld allele is a SLF in which the transmembrane form is missing. Because the Sld/Sld mice display moderate defects, the transmembrane form of SLF is likely to have a critical role in vivo. It has also been suggested that the transmembrane form of SLF not only stimulates proliferation of mast cells, but also medi-

4 ~ ~ Interaction of c-kit Receptor and Steel Factor 454 Table I. Effects of W and Sl mutations Locus Defect Gene Chromosome Disorders W intrinsic C-kit 5 macrocytic anemia mast cell decrease white coat color sterility homozygote (sub)lethality s1 microenvironmental steel factor 10 macrocytic anemia mast cell decrease white coat color sterility homozygote (sub)lethality ates cell-cell adhesion [MI. These molecular characterizations provide a basis for explaining the intrinsic (W) versus environmental (SI) defects. S1 mice cannot produce the growth factor but have the c-kit receptor. Thus, Sl stem cells can respond when transplanted to a wild type animal which produces the SLF. Normal HSCs introduced into Sl mice will not have the SLF present and therefore cannot correct the defect. Conversely, Wstem cells lack the c-kit receptor but W mice can produce the SLF. W cells placed into normal mice cannot respond to SLF, while normal cells introduced into W mice will grow appropriately. Transplantation of normal bone marrow cells into unirradiated W/W' mice leads to long-term reconstitution [El. These bone marrow transplantation studies imply that n o d HSCs and/or other hematolymphoid progenitors have some advantage which allows their repopulation and/or self-renewal over W/W' progenitors. Because these earlier studies were done with total bone marrow cells, it is still unclear whether the bone marrow reconstitution was due to single HSCs or a set of long-lived hematopoietic progenitors. In other words, it is not clear whether SLF can act at the level of HSCs or hematopoietic progenitors. Therefore, it is of interest to check the long-term reconstitution of unirradiated W/W' mice with limiting numbers of purified HSCs from wild type mice, preferably using a congenic system to allow discrimination between donor and host bone marrow cells [a]. If a single HSC can repopulate W W recipients in all cell lineages for long term, it would provide indirect evidence that SLF may play a role in self-renewal as well as in differentiation of HSCs. The Expression of c-kit on HSCs and Hematopoietic Progenitors Previous studies have suggested that the interaction of the c-kit receptor and SLF should occur at the level of HSCs mble III). Recently, it has been shown,

5 Ikuta/Ingolia/Friedman/Heimfeld/Weissman 455 Table II. Effect of the W and Sl mutations on hematopoietic stem cells W mutation -cure of anemia of W/W' mice by histocompatible +I+ stem cells -no radioprotection of lethally irradiated normal mice by W/W" stem cells -low number (1I200) of CFU-s in W/W' bone marrow (many microscopic colonies) -normal spleen-colony formation in W/W' spleen by +/+ stem cells -radiosensitivity Sl mutation -no cure of anemia of Sl/Sld mice by histocompatible +/+ stem cells -radioprotection of lethally irradiated normal mice by Sl/Sld stem cells -normal number of CFU-s in S1/Sld bone marrow -no spleen-colony formation in S1/Sld spleen by +I+ stem cells (ody microscopic colonies) -radiosensitivity using anti-c-kit monoclonal antibodies, that the c-kit receptor is expressed on hematopoietic progenitors [45]. In this study, it was shown that the c-kit receptor is expressed on 8% of total bone marrow cells; hematopoietic progenitors responsive to IL-3, GM-CSF, and M-CSF, as well as the CFU-s themselves express the c-kit receptor. In vivo injection of an anti-c-kit antibody with blocking activity resulted in the elimination of hematopoietic progenitors, suggesting that the c-kit receptor is required for the maintenance of these cells. In another study, it was shown that the c-kit receptor is expressed on HSCs with long-term reconstitution activity, as well as on cells with CFU-s activity (K. Ikuta and Z.L. Weissman, in preparation). In this study, Thy-1" Lin- Sca-1' cells from adult bone marrow, which are known to contain both the HSCs and are separated into c-kit' (70%) and c-kit- (30%) subpopulations. CFU-s assays and long-term reconstitution experiments show that only the Thy-1"' Lin- Sca-1' c-kit' subpopulations manifest these activities. The c-kit staining pattern of Thy-1" Lin- Sca-1' cells from day 14 fetal liver was similar to that from adult bone marrow. The c-kit transcript is detectable in the Thy-1" Lin- Sca-1' cells by PCR (0. Zngolia and Z.L. Weissmun, in preparation). These results suggest that the c-kit receptor seems to be expressed on bone marrow cells from the earliest stage, i.e., HSCs to the later stages of hematopoietic progenitors of at least the myeloerythroid lineages. The Effects of SLF on Hematopoietic Progenitors The in vivo effects of SLF have been studied [40]. When S1/SZd mice were treated with high doses of recombinant SLF, their macrocytic anemia was partially reversed. Concomitant with the decrease in the red cell volume, the number of erythrocytes increased. Increased representation of myeloid lineages in the peripheral blood was also observed. Very little effect was observed on lympho-

6 Interaction of c-kit Receptor and Steel Factor 456 Table III. Possible hernatopoietic cells expressing c-kit receptor Pluripotent hernatopoietic stern cells CFU-s Myeloid and B progenitors Erythroid progenitors including BFU-e and CFU-e Mast cells poiesis. Injection of SLF also resulted in the appearance of mast cells at the injection site. Both purified native and/or recombinant SLF show strong synergistic actions with other factors on whole bone marrow, such as IL-3, IL-1, IL-6, G-CSF, GM-CSF, and erythropoietin (Epo) in various CFU-c assays [39, 42, 431. SLF has a profound effect on the erythroid lineage in costimulation assays with Epo. In cultures of 5-FU treated bone marrow, SLF alone can stimulate colony formation, and it is synergistic with IL-6 and IL-1 [39]. Recombinant human SLF, in combination with IL-3, GM-CSF, or G-CSF, also enhance colony formation by CD34' human hematopoietic progenitors [46]. In another study [47], mouse hematopoietic progenitors were fractionated into WGA' and either Rhl23 bright or dull populations. These populations constituted 0.03% of total bone marrow cells. Day 14 CFU-s were enriched 130- and 240-fold in Rhl23 bright and dull populations, respectively (approximately 2- to 3-fold less than Thy-1'" Lin- Sca-1' cells comparing published data [6]). Recombinant SLF alone stimulated the proliferation of these cells, and acted synergistically with IL-3 and IL-la. Culture of the WGA' Rh123 bright population with SLF and IL-3 results in a net increase of day 14 CFU-s (12-fold after 11 days). These results indicate that SLF may act on day 14 CFU-s to cause self-renewal, or that it may stimulate the pre-cfu-s to differentiate into CFU-s. The response of Thy-1" Lin- Sca-1' cells to the SLF has also been tested. These cells do not proliferate in response to SLF alone, but combinations of SLF and other cytokines give a very high response frequency (S. Heimfeld, et al., in preparation). The predominant types of cells generated in vitro are granulocytes and macrophages. Interestingly, there is a progressive loss of HSC activity in these cultured cells, as measured by day l2 CFU-s and long-term reconstitution. By day 14, HSC activity was gone from such cultures in the face of continued vigorous myelopoiesis. These results indicate that SLF is a factor which may act on HSCs as well as on their clonogenic progeny. It seems to promote HSC proliferation and differentiation into the non-hsc pool, and is the most potent hematopoietic co-mitogen yet found. However, the studies from our laboratory provide no evidence that SLF promotes self-renewal of HSCs in vitro, especially for HSCs with long-term reconstitution activity. Is SLF required for the generation and proliferation of HSCs?

7 Ikuta/Ingolia/Friedman/Heimfeld/Weissman 457 This question may be addressed by following the development of HSCs in SL/SL mice, which have totally deleted the Sl gene. SUSL homozygote mice usually die at day 15 to day 16 of gestation. Surprisingly, during fetal life hematopoiesis is taking place in the SL/SL fetal liver. Although the total number of SUS1 fetal liver cells is about 20% of that of the normal fetus, it does increase during fetal development. The number of Thy-, Lin- Sca-1 cells, containing HSCs, in SUS1 fetal liver is 30 to 40% of normal, and also increases between days 13 and 14 of gestation (K. Ikuta and I.L. Weissmn, in preparation). Day 12 CFU-s activity shows a similar increase during this time. These results may indicate that SLF is not essential for the hitiation of hematopoiesis or the early self-renewal of HSCs. Future Directions It is possible that there may be some unknown factor(s) or mechanism which plays a crucial role on the initiation of hematopoiesis. SLF is produced by many different types of cells, such as fibroblasts, Sertoli cells, and hair follicle cells, as well as by the stromal cells in adult bone marrow and fetal liver. Therefore, the fact that hematopoiesis is restricted to the hematopoietic organs, such as bone marrow, ktal liver, and yolk sac, cannot be explained by the presence of SLF alone. The effect of SLF seems to be more profound on early erythroid cells than on HSCs, and fetal HSCs can be generated in the absence of SLF. It is possible that HSCs may depend on other factors for their initial generation andor self-renewal. It is still necessary to test whether a combination of factors, such as IL-6, IL-7, IL-11, M-CSF, G-CSF, and GM-CSF, known to be produced by bone marrowderived stromal cells, can lead to HSC self-renewal, or some as yet unknown selfrenewal factor(s) for HSCs. It will be interesting to know whether putative HSC self-renewal factor(s) also determines the tissue specificity of hematopoiesis. It will also be important to determine what microenvironmental factors exist that cause the appearance and generation of large numbers of HSCs in the mouse fetus. Embryonic stem (ES) cells derived from the inner cell mass of the mouse blastocyst could provide a useful in vitro system to study the early development of HSCs. ES cells can be induced to differentiate into cystic embryoid bodies that contain hematopoietic cells such as blood island cells [ During the course of differentiation, the c-kit receptor and SLF have been shown to be expressed at the RNA level [50]. It will be interesting to know the role of the c-kit receptor/ SLF interaction on the initiation of hematopoiesis and the development of the first HSCs from their precursors in this system. References 1 Till JE, McCulloch EA. A direct measurement of the radiation sensitivity of normal mouse bone marrow cells. Radiation Res 1961;14:

8 Interaction of c-kit Receptor and Steel Factor Hodgson GS, Bradley TR. Properties of haematopoietic stem cells surviving 5-fluorouracil treatment: evidence for a pre-cfu-s cell? Nature 1979;281: Spangrude GJ. Enrichment of murine haemopoietic stem cells: diverging roads. Immunol Today 1989;10: Visser JWM, Van Bekkum DW. Purification of pluripotent hemopoietic stem cells: past and present. Exp Hematol 1990;18: Visser JWM, Bol SJL, van den Engh G. Characterization and enrichment of murine hemopoietic stem cells by fluorescence activated cell sorting. Exp Hematol 1981;9: Spangrude GJ, Heimfeld S, Weissman IL. Purification and characterization of mouse hematopoietic stem cells. Science 1988;241: Nicola NA, Metcalf D, von Melchner H, Burgess AW. Isolation of murine fetal hemopoietic progenitor cells and selective fractionation of various erythroid precursors. Blood 1981 ;58: Jordan CT, McKearn JP, Lemischka IR. Cellular and developmental properties of fetal hematopoietic stem cells. Cell 1990;61: Ikuta K, Kina T, MacNeil I, et al. A developmental switch in thymic lymphocyte maturation potential occurs at the level of hematopoietic stem cells. Cell 1990;62: Smith LG, Weissman IL, Heimfeld S. Clonal analysis of hematopoietic stem-cell differentiation in vivo. Proc Natl Acad Sci USA 1991;88: I Spangrude GJ, Johnson GR. Resting and activated subsets of mouse multipotent hematopoietic stem cells. Proc Natl Acad Sci USA 1990;87: Russell ES. Hereditary anemias of the mouse: a review for geneticists. Adv Genet 1979 ;20: Sarvella PA, Russell LB. Steel, a new dominant gene in the house mouse. J Hered 1956;47: Mayer K, Green MC. An experimental analysis of the pigment defect caused by mutations at the Wand Sl loci in mice. Dev Biol 1968;18: Kitamura Y, Go S, Hatanaka K. Decrease of mast cells in W/W' mice and their increase by bone marrow transplantation. Blood 1978;52: Fujita J, Nakayama H, Onoue H, et al. Fibroblast-dependent growth of mouse mast cells in vitro: duplication of mast cell depletion in mutant mice of W/w' genotype. J Cell Physiol 1988;134: Fujita J, Onoue H, Ebi Y, Nakayama H, Kanakura Y. In vitro duplication and in vivo cure of mast-cell deficiency of SZ/Sld mutant mice by cloned 3T3 fibroblasts. Proc Natl Acad Sci USA 1989;86: Mekori T, Phillips RA. The immune response in mice of genotypes W/Wv and SZ/Sld. Proc Soc Exp Biol Med l969;132: Landreth KS, Kincade PW, Lee G, Harrison DE. B lymphocyte precursors in embryonic and adult W anemic mice. J Immunol l984;132: Ebbe S, Phalen E, Stohlman Jr F. Abnormalities of megakaryocytes in W/Wv mice. Blood 1973;42: Ebbe S, Phalen E, Stohlman Jr F. Abnormalities of megakaryocytes in Sl/Sld mice. Blood 1973;42: McCulloch EA, Siminovitch L, Till JE. Spleen-colony formation in anemic mice of genotype WW". Science 1964;144: McCulloch EA, Siminovitch L, Till JE, Russell ES, Bernstein SE. The cellular basis of the genetically determined hemopoietic defect in anemic mice of genotype Sl/Sld. Blood 1965;26: Alms MS, Bernstein SE, Russell ES, Carsten AL, Upton AC. Defect extrinsic to stem cells in spleens of steel anemic mice. Proc Soc Exp Biol Med 1971;138:

9 Ikuta/Ingolia/Fri~m/Heimfeld/Weissman Seller MJ. Donor haemoglobin in anaemic mice of the W-series transplanted with haematopoietic tissue from an unrelated donor. Nature 1966;2l2: Bernstein SE, Russell ES, Keighley G. Two hereditary mouse anemias (Sl/Sld and Wlw ) deficient in response to erythropoietin. Ann NY Acad &i 1968;149: Seller MJ. Transplantation of anaemic mice of the W-series with haemopoietic tissue bearing marker chromosomes. Nature 1968;220: Harrison DE, Astle CM. Population of lymphoid tissues in cured W-anemic mice by donor cells. Transplant 1976;22: Fleischman RA, Mintz B. Prevention of genetic anemias in mice by microinjection of normal hematopoietic stem cells into the fetal placenta. Proc Natl Acad Sci USA 1W9;76: Dexter TM, Moore MAS. In vitro duplication and cure of haemopoietic defects in genetically anaemic mice. Nature 1977;269: Qiu F, Ray P, Brown K, et al. Primary structure of c-kit: relationship with the CSF-11 PDGF receptor base family-oncogenic activation of v-kit involves deletion of extracellular domain and C terminus. EMBO J 1988; Yarden Y, Kuang W-J, Yang-Feng T, et al. Human proto-oncogene c-kit: a new cell surface receptor tyrosine kinase for an unidentified ligand. EMBO J l987,6: Chabot B, Stephenson DA, Chapman VM, Besmer P, Beinstein A. The promncogene c-kit encoding a transmembrane tyrosine kinase receptor maps to the mouse Wlocus. Nature 1988;335 : Geissler EN, Ryan MA, Housman DE. The dominant-white spotting (W) locus of the mouse encodes the c-kit proto-oncogene. Cell 1988;55: Wdliams DE, Eisenman J, Baird A, et al. Identification of a ligand for the c-kir protooncogene. Cell 1990;63: Copeland NG, Gilbert DJ, Cho BC, et al. Mast cell growth factor maps near the steel locus on mouse chromosome 10 and is deleted in a number of steel alleles. Cell 1990; 63 ~ Flanagan JG, Leder P. The kir ligand: a cell surface molecule altered in steel mutant fibroblasts. Cell 1990;63: Zsebo KM, Wypych J, McNiece M, et al. Identification, purification, and biological characterization of hematopoietic stem cell factor from Buffalo rat liver-conditioned medium. Cell 1990;63: Martin FH, Suggs SV, Langley KE, et al. Primary structure and functional expression of rat and human stem cell factor DNAs. Cell l990;63:203-2ll. Zsebo KM, Williams DA, Geissler KM, et al. Stem cell factor is encoded at the SI locus of the mouse and is the ligand for the c-kit tyrosine kinase receptor. Cell 1990; 63:2l Huang E, Nocka K, Beier DR, et al. The hematopoietic growth factor KL is encoded by the St locus and is the ligand of the c-kit receptor, the gene product of the Wlocus. Cell 1990;63: Anderson DM, Lyman SD, Baird A, et al. Molecular cloning of mast cell growth factor, a hematopoietin that is active in both membrane bound and soluble forms. Cell 1990;63 : Nocka K, Buck J, Levi E, Besmer P. Candidate ligand for the c-kit transmembrane base receptor: KL, a fibroblast derived growth factor stimulates mast cells and erythroid progenitors. EMBO J 1990;9: Flanagan JG, Chan DC, Leder P. Transmembrane form of the kit ligand &rowth factor is determined by alternative splicing and is missing in the Sld mutant. Cell 1991; 64:

10 Interaction of c-kit Receptor and Steel Factor Ogawa M, Matsuzaki Y, Nishikawa S, et al. Expression and function of c-kit in hemopoietic progenitor cells. J Exp Med 1991;174: Bernstein ID, Andrews RG, Zsebo KM. Recombinant human stem cell factor enhances the formation of colonies by CD34 and CD34+lin- cells, and the generation of colonyforming cell progeny from CD34+lin- cells cultured with interleukin-3, granulocyte colony-stimulating factor, or granulocyte-macrophage colony-stimulating factor. Blood 1991i de Vries P, Brasel KA, Eisenman JR, Alpert AR, Williams DE. The effect of recombinant mast cell growth factor on purified murine hematopoietic stem cells. J Exp Med 1991;173:L205-L Doetschman TC, Eistetter H, Katz M, Schmidt W, Kemler R. The in vitro development of blastocyst-derived embryonic stem cell lines: formation of visceral yolk sac, blood islands and myocardium. J Embryo1 Exp Morphol 1985;87: Wiles MV, Keller G. Multiple hematopoietic lineages develop from embryonic stem (ES) cells in culture. Development 1991;111: Schmitt RM, Bruyns E, Snodgrass HR. Hematopoietic development of embryonic stem cells in vitro: cytokine and receptor gene expression. Genes Dev l991;5: