Chronic myeloid leukemia (CML) has long served as a prototype malignancy for basic as well as clinical studies aimed at developing curative cancer treatment protocols. Well established features of chronic phase CML are its origin in a pluripotent stem cell, a now well defined molecular genetic basis involving the creation of a BCR-ABL fusion gene and evidence of resultant abnormalities in the mechanisms that normally control primitive hemopoietic cell proliferation. We have recently shown how the long-term marrow culture system can be adapted to quantitate and characterize a very primitive cell type in normal blood and marrow samples, as well as their normal and leukemic counterparts in patients with CML. This system has also been used to dissect mechanisms of normal progenitor regulation and to identify specific anomalies affecting leukemic (CML) progenitors. Our studies show that cells detected by their ability to initiate long-term cultures (LTC) of leukemic cells (i.e., CML LTC-initiating cells or LTC-IC) are differently distributed between marrow and blood by comparison to LTC-IC in normal individuals and, although functionally similar in terms of the number and differentiation types of clonogenic cells they produce, CML LTC-IC exhibit defective self-maintenance. Phenotypically these primitive leukemic cells are heterogeneous; the majority display features of activated/proliferating cells but a significant proportion do not. We have also documented heterogeneity in primitive CML cell responses to two factors that specifically and reversibly arrest the cycling of primitive normal hemopoietic cells; i.e., TGF-beta and MIP-1 alpha, to which CML cells are normally responsive and abnormally unresponsive, respectively.(ABSTRACT TRUNCATED AT 250 WORDS)