L-leucyl-L-leucine methyl ester (LLME) is a lysosomotropic agent that in microMolar concentrations has been found to be selectively toxic to human and murine precursor and effector cytotoxic cells, irrespective of their surface membrane phenotype. We describe a new method of synthesis of LLME and evaluate the effects of this preparation on human lymphoid and hematopoietic progenitor cells. The new method of synthesis did not change the previously characterized activities of LLME. Consistent with previous observations, NK effectors, LAK precursors and effectors, and allospecific CTL (aCTL) effectors were completely ablated by treatment with 50-250 microM LLME, while the activities of helper T cells and B cells were preserved after treatments of up to 1000 microM LLME. The effects of LLME treatment on human marrow-derived erythroid, myeloid, and monocyte progenitors have not been previously described. We found that the growth of each of these committed precursors was reduced or eliminated following treatment with 100-250 microM LLME. Admixture of LLME-treated marrow with marrow depleted of T cells and other mature cellular elements resulted in increased growth of myeloid and erythroid colonies suggesting that cells that could provide colony-enhancing activities were preserved. In contrast to previous studies in humans, we found a minority of individuals to have aCTL precursors that were partially resistant to LLME. PBL from 10 of 15 individuals tested showed nearly complete ablation of aCTL precursors following treatment with 375 microM LLME. The remaining 5 individuals demonstrated significant aCTL precursor activity after identical treatment. The resistance to LLME was restricted to aCTL precursors, and neither increasing the dose of LLME nor prolonging the time of treatment completely overcame the resistance. The pattern of susceptibility (sensitive versus resistant) was found to be independent of the degree or type of HLA disparity between responder and stimulator. LLME-treated cultures with and without CTL activity contained a predominance of CD4+ T cells. However, in the subjects tested LLME-resistant aCTL was shown to be CD8+. In vitro priming of aCTL precursors from sensitive individuals did not consistently result in the development of resistance to LLME. These data indicate that further studies are needed to evaluate the effects of LLME on human stem cells and to determine the potential role of resistant aCTL precursors in GvHD prior to application of this technique as a form of selective T cell depletion in humans.