Trace mineral deficiencies tend to have profound effects on the integrity of formed blood elements. Anemia and neutropenia are commonly seen in copper (Cu) deficiency. We therefore developed a serum-free medium to examine the trace mineral requirements, in particular iron and Cu, for proliferation and retinoic acid (RA)-induced differentiation of HL-60 cells. This defined medium (DFM) was composed of Iscove's Modified Dulbecco's Medium (IMDM) supplemented with insulin and human apo-transferrin (each at 5 micrograms/ml) and 1.4 microM FeSO4. The iron concentration range for optimal cellular proliferation was narrow (2-3 microM). HL-60 cells could be maintained in DFM for 15 passages with a doubling time of 38-40 hr. The Cu content of IMDM was very low. Thus, by the fourth passage in DFM, the activity of cuproenzymes (cytochrome c oxidase, CCO; and copper-zinc superoxide dismutase, CuZnSOD) began to decline. Supplementation of DFM with CuSO4 (50 nM) restored enzyme activities. Treatment of cells with a Cu chelator (tetrathiomolybdate, 1 microM) rapidly reduced the activities of both CCO and CuZnSOD. Over the Cu concentration range examined (5-350 nM), Cu supplementation had little effect on HL-60 proliferation. Cell retained the ability to differentiate along the granulocytic pathway when treated with RA, but seemed to be less sensitive to the inducing agent except at the highest concentration tested (1 microM). This decreased sensitivity to RA did not seem to be related to the Cu status of the cells but rather to the absence of a component of serum. Indeed, cells grown in DFM regained their sensitivity to RA when allowed to differentiate in IMDM with 5% serum. These data indicate that the processes of growth and terminal differentiation in HL-60 cells are not greatly influenced by Cu. Thus, it seems likely that the insult resulting in neutropenia which is associated with Cu deficiency may occur earlier than the promyelocytic stage. However, the possibility that the mechanisms contributing to neutropenia may be unrelated to primary defects in the biochemistry of neutrophil maturation cannot be ruled out.