A goal of our research is to induce the definitive hematopoietic stem cell (dHSC) in vitro from plurpotent cell lines. As ES cells and iPS as well are the sole cell line that can be maintained in vitro and is proven to give rise to dHSC, this is the most promising culture for producing a unlimited number of dHSC. At this moment, however, this is a difficult goal to attain We think that this difficulty is due to the lack of our knowledge about the actual differentiation process to dHPC in the embryo. Previous studies of embryonic HSC differentiation focused on either the molecules required for HSC differentiation or on the site where the first dHSC are generated. Studies in this line indeed facilitated our understanding on key players such as Runx1 that is involved in this process. On the other hand, only a little is known about the actual pathway for the dHSC specification nor about the phenotype of dHSC progenitors. Indeed, even the notion of hemangioblast is tottering. Thus, what to be clarified first should be the actual differentiation pathway of the dHSC, without which one can not specify the cells to be generated in the culture. We are addressing this issue by genetic approaches. One approach is to label the progenitors of dHSC at a given time point by the tamoxifen-inducible Cre-recombinase mediated activation of LacZ gene. The second approach also uses the tamoxifen-inducible Cre-recombinase, but this time Cre was used to reactivate runx1 gene in runx1-/- mouse. By these approaches, we specified that E7.5 Runx1+ cells in the YS are the progenitors of dHSC. While dHSC have been suggested to be generated from multiple sources in the embryo, our results suggest strongly that all of them are derived from this Runx1+ cells in the YS. Thus, the new scheme came out from our studies is completely distinct from what has been believed as consensus concerning the differentiation of dHSC. Once the actual pathway of dHSC is defined, it is possible to address whether or not this Runx1+ progenitors are induced from ES cells. Indeed, the population that is almost equivalent to this progenitors are able to be generated in ES cell culture. In addition, it should be noted that Runx1+ cells in E7.5 YS are unable to generate dHSC in vitro. Hence, it is likely that the failure of ES cell culture to generate the dHSC is not due to deficiency of ES cell culture in inducing the dHSC progenitors, but due to deficiency to support further differentiation of the Runx1+ progenitors to the stage that can reconstitute the irradiated recipient. Currently, we are trying to identify the molecular basis underlying this deficiency. In the symposium, we will present the latest results of this attempt.