IL-3 inhibits the binding of GM-CSF to AML blasts, but the two cytokines act synergistically in supporting blast proliferation.

Equilibrium binding of 125I-labeled recombinant granulocyte macrophage colony-stimulating factor (GM-CSF) to the blast cells of acute myeloblastic leukemia (AML) revealed the presence of two classes of binding components of high and low affinity, with dissociation constants (Kd) in the range of 5-10 pM and 1-10 nM, respectively. Specificity studies revealed that interleukin-3 (IL-3) could partially inhibit the binding of GM-CSF to AML blasts and to the cells of the leukemic lines M07-E, KG-1, and HL-60. The inhibition of GM-CSF binding by IL-3 was directly dependent on the presence of IL-3 receptors. Analysis of competition curves indicated that the Kd and the number of binding sites per cell of unlabeled and iodinated GM-CSF were identical. In contrast, the inhibition of GM-CSF binding by IL-3 was mediated by IL-3 occupancy of a high affinity receptor only, with the same number of sites as the high affinity GM-CSF receptor but a slightly higher Kd. Despite this competitive binding, IL-3 augmented AML blast proliferation in the presence of GM-CSF, indicating that the two growth factors have converging pathways in supporting blast proliferation. In striking contrast to AML blasts, GM-CSF binding to neutrophils was compatible with the presence of only one class of binding site of intermediate affinity (Kd approximately 100-160 pM). Furthermore, IL-3 does not compete for the binding of GM-CSF to neutrophils.

Comparison of antineoplastic activity of cytosine arabinoside and 5-aza-2'-deoxycytidine against human leukemic cells of different phenotype.

A comparison of the cellular and molecular pharmacology of the deoxycytidine analogues cytosine arabinoside (Ara-C) and 5-aza-2'-deoxycytidine (5-AZA-CdR) on human myeloid (HL-60), T-cell (Molt-3) and B-cell (RPMI-8392) leukemic cell lines was investigated. Ara-C was a more potent inhibitor of growth than 5-AZA-CdR. In a colony assay, 5-AZA-CdR was a more potent cytotoxic agent than Ara-C for both the myeloid and B-cell leukemic cells, but not for the T-cell leukemic cells. The total cellular uptake of 5-AZA-CdR was greater than Ara-C for the myeloid and B-cell leukemic cells, whereas for the T-cells the uptake of the arabinosyl analogue was greater. Ara-C produced a potent inhibition of DNA synthesis, whereas no inhibition was detected with 5-AZA-CdR during a short incubation. In contrast, 5-AZA-CdR produced a potent inhibition of DNA methylation whereas Ara-C produced a slight increase in the methylation of DNA. This study shows that there are significant differences in the antineoplastic activity of Ara-C and 5-AZA-CdR against human leukemic cell lines of different phenotype and that these differences are related to differences in the metabolism of these two deoxycytidine analogues and to their effects on DNA synthesis and methylation.

Cellular pharmacology of 1-beta-D-arabinofuranosylcytosine in human myeloid, B-lymphoid and T-lymphoid leukemic cells.

The in vitro inhibitory action and metabolism of 1-beta-D-arabinofuranosylcytosine (ara-C) on human myeloid (HL-60), B-lymphoid (RPMI-8392), and T-lymphoid (Molt-3) leukemic cells was compared. Ara-C produced greater inhibitory effects in Molt-3 cells than in either HL-60 or RPMI-8392 cells. At a 48 h exposure, ara-C was 7 and 10 times more cytotoxic to Molt-3 cells than to HL-60 and RPMI-8392 cells, respectively. The total ara-C uptake to nucleotides and the formation of 1-beta-D-arabinofuranosylcytosine 5'-triphosphate (ara-CTP) was about 5 times greater in Molt-3 cells than in either HL-60 or RPMI-8392 cells. The incorporation of ara-C into DNA was also higher in Molt-3 cells than in either HL-60 or RPMI-8392 cells. The mean intracellular half-life of ara-CTP was 31.7, 59.4, and 155 min for RPMI-8392, HL-60, and Molt-3 leukemic cells, respectively. The Km and Vmax values of ara-C for deoxycytidine kinase and the feedback inhibition of this enzyme by ara-CTP in the different leukemic cell lines could not explain the differences in metabolism of this analogue in these cells. These data indicate the increased sensitivity of T-lymphoid leukemic cells to ara-C than as compared with B-lymphoid and myeloid leukemic cells was due to an increased rate of formation and a longer half-life of ara-CTP in the T-cells.

Characterization of granulocyte-macrophage colony-stimulating factor receptor on the blast cells of acute myeloblastic leukemia.

Iodinated granulocyte-macrophage colony-stimulating factor (GM-CSF) was used to document the specific binding of GM-CSF to all acute myeloblastic leukemia (AML) samples examined in the present study. There was some heterogeneity in the number of GM-CSF binding sites per cell. To determine whether the low level of binding to some patient samples may be attributed to receptor occupancy by an endogenous source of GM-CSF, we devised an acid wash procedure that could remove surface-bound GM-CSF without affecting receptor properties. We thus document that GM-CSF specific binding to AML blasts before or after acid wash was the same, indicating that the observed heterogeneity in binding is not the result of receptor occupancy by an endogeneous source of GM-CSF. Saturation analyses are in favor of the presence of two classes of binding sites on AML blasts: a high-affinity receptor that binds GM-CSF with a dissociation constant (kd) of 3 to 73 pmol/L and a second class of low-affinity receptor that binds GM-CSF with a kd of 1 to 10 nmol/L. Binding studies with two established cell lines KG-1, and IRCM-8 also showed the presence of two classes of binding sites with high and low affinities. Analysis of GM-CSF titration curves in culture indicate that the median effective concentration required for stimulation of blast colony formation (EC50 = 5-36 pmol/L) were in the range of the kd of the high-affinity binding site, suggesting that this high-affinity binding site mediates the proliferative response.