1 International Journal of HEMATOLOGY Progress in Hematology 1. Introduction The derivation of human embryonic stem (ES) cells [1] has ignited a gold rush of research activity into the therapeu- tic potential of ES-derived cells for all manner of degenera- tive conditions. At the root of this optimism lies an attractive logic: When transplanted back into a blastocyst or morula embryo, ES cells can differentiate into any cell type of the animal proper; therefore once we know the details of their differentiation, we should be able to generate any of these cell types in vitro, including those damaged or missing in degenerative diseases. The devil is, of course, in the details. How well do we really know the cell type that we are inter- ested in, or what governs its origin and differentiation, and how do we know that the cell we have generated in vitro cor- responds precisely to the cell type of interest? From the perspective of transplantation-based therapy, the hematopoietic stem cell (HSC) is likely to play a pivotal role. It would serve as a therapeutic in its own right for patients with blood disorders, but it may also evolve into a critical enabling tool for transplantation of tissues derived from nonisogenic ES cells, which require that the recipient’s immune system be tolerized to the nonisogenic graft. Although it is possible to generate isogenic ES cells through Genesis of Hematopoietic Stem Cells In Vitro and In Vivo: New Insights into Developmental Maturation Michael Kyba Center for Developmental Biology, University of Texas Southwestern Medical Center, Dallas,Texas, USA Received December 2, 2004; accepted December 20, 2004 Abstract Hematopoietic stem cells first arise in the mammalian embryo in a primitive state, not capable of reconstituting hemato- poiesis in irradiated adult recipients.As development proceeds, these cells eventually mature to acquire definitive, adult char- acteristics, including adult reconstitution ability. Mouse embryonic stem cells induced to undergo hematopoiesis in vitro readily generate primitive hematopoietic stem cells but rarely generate the definitive type. Recent work has stimulated a new appreci- ation of the events involved in the developmental maturation of hematopoietic stem cells. Application of this knowledge to in vitro differentiation systems will be critical to the successful development of hematopoietic therapies from embryonic stem cells. Int J Hematol. 2005;81:xxx-xxx. doi: 10.1532/IJH97.04192 ©2005 The Japanese Society of Hematology Key words: Embryonic stem cells; Hematopoietic stem cells; CD41; Placenta; HoxB4 Correspondence and reprint requests: Michael Kyba, Center for Developmental Biology, University of Texas Southwestern Medical Center, Dallas, TX 75390-9133, USA; 1-214-648-1169; fax: 1-214-648-1960 (e-mail: michael.kyba@ utsouthwestern.edu). reprogramming somatic cells by nuclear transfer [2], it is unlikely that this process will become routine in the clinic, at least with current methods, because of the time and effort involved to generate such a cell line and because the human oocytes required for the process are not available in large quantities. It is, however, within the reach of current technol- ogy to generate a bank of 3000 or more genetically distinct human ES cell lines, enough to provide 95% of the popula- tion with 4 of 6 or more HLA antigen–matched donor cells [3], a level of match that is currently considered acceptable for unrelated cord blood transplantation [4,5]. Because HSC transplantation can induce tolerance to grafts of other tissues isogenic to HSCs [6-9], mixed chimerism with HSCs from an ES cell line, selected from such a bank for the best HLA anti- gen match, would then facilitate tolerance to therapeutic cells or tissues of other types derived from the same best matched ES cell line. The ability of ES cells to differentiate into blood in vitro was documented by Doetschman and colleagues in 1985 [10]. Those authors used the embryoid body (EB) method of differentiation, in which ES cells are detached from the 2-dimensional surface on which they are growing in self- renewal mode and aggregated in 3-dimensional clumps. Soon after aggregation, a superficial epithelial layer resembling primitive endoderm differentiates, and this process is fol- lowed closely by apoptosis of cells in the core of the EB. These processes directly recapitulate early events in the dif- ferentiation of the inner cell mass of the murine blastocyst, namely, formation of an egg cylinder followed by cavitation to generate the proamniotic cavity [11,12]. After cavitation, the internal cells of the EB that escape apoptosis go on to