Most primitive hematopoietic cells appear to be normally quiescent in vivo, whereas their leukemic counterparts in patients with chronic myeloid leukemia (CML) are maintained in a state of rapid turnover. This difference is also seen in the long-term culture system, where control of primitive hematopoietic progenitor proliferation is mediated by interactions of these cells with marrow-derived mesenchymal cells of the fibroblast lineage. We now show that exogenous addition of macrophage inflammatory protein 1 alpha (MIP-1 alpha) to normal long-term cultures can reversibly and specifically block the activation of "primitive" (high proliferative potential), but not "mature" (lower proliferative potential), progenitors in the adherent layer of these cultures. Moreover, addition of MIP-1 beta after primitive-progenitor activation can prevent the subsequent return of these cells to a quiescent state a few days later as shown previously in similar experiments using antibodies to transforming growth factor beta. This suggests that the level of MIP-1 alpha (or a related MIP-1 alpha agonist) produced in LTCs, like the level of transforming growth factor beta, may be necessary, but is not on its own sufficient, to mediate the inhibitory activity of the regulatory cells in the adherent layer. Addition of MIP-1 alpha to similar long-term cultures containing normal marrow adherent layers but supporting exclusively neoplastic (CML) hematopoiesis did not block the cycling of primitive neoplastic progenitors. A defect in the responsiveness of CML cells to MIP-1 alpha (or a similarly acting chemokine) would explain their deregulated proliferative behavior in this model and, by extrapolation to the in vivo setting, suggests a molecular mechanism whereby the leukemic clone may become amplified at the stem-cell level. In addition, these findings suggest approaches to the therapy of CML, using inhibitors such as MIP-1 alpha for the protection of primitive normal cells.