Evidence for a relationship between mouse hemopoietic stem cells and cells forming colonies in culture.

The spleen colony technique' has proved to be a useful tool for the enumeration of hemopoietic stem cells in the mouse and for the study of their properties. However, since the method depends on colony formation in the spleens of heavily irradiated' or genetically anemic2 mice, it has certain inherent limitations. For example, events occurring early in the growth of colonies cannot be observed, since they are obscured by the splenic environment; further, the spleencolony method has not been applied successfully to the study of human cells3, 4 and, therefore, the information obtained by use of this method cannot be applied directly to clinical problems. These limitations might be avoided by the use of cell-culture methods. Recently, Pluznik and Sachs' and Bradley and Metcalf6 have reported that cells derived from mouse hemopoietic tissue will form colonies in culture if they are suspended in dilute agar over a firm agar base containing a source of suitable "conditioning factor."7-9 If a relationship were established between the cells that give rise to these colonies in culture and the hemopoietic stem cells responsible for colony formation in the spleen, the two techniques might complement each other. Results obtained by the use of the culture technique might be related through the spleen-colony method to normal and abnormal hemopoiesis invivo, and the information gained in the mouse by using the spleen-colony technique might be extended to other species, including man, by using the culture method. The existence of a relationship between hemopoietic stem cells and the cells that give rise to colonies in culture is indicated by the properties of the latter: they are capable of extensive proliferation,1? 11 self-renewal,5 and granulopoietic differentiation.6 10 Further, they are found in the same organs as hemopoietic colony-forming cells and in approximately the same relative numbers.5, 6 In order to establish a relationship more directly, we have employed two approaches. First, we have used a chromosome marker techniquel2 to determine whether or not the cells that give rise to colonies in culture can belong to the same clone as cells that form colonies in vivo. Second, we have used both techniques in assaying cell suspensions derived from individual spleen colonies for their colonyforming activity. Previous work had shown that the distribution of spleen colony-forming cells among spleen colonies is very heterogeneous, with a 1000fold difference between colonies containing little and those containing much colony-forming potential.", 14 If individual colonies yielded similar results in both assay systems, a close relationship between the cells responsible for each activity might be assumed. The results of these experiments are presented in this report. They provide 1209