Previous studies have demonstrated that combinations of all-trans retinoic acid (ATRA) with either granulocyte-colony stimulating factor (G-CSF) or lithium chloride (LiCl) produced synergistic terminal differentiation of WEHI-3B myelomonocytic leukemia (D(+)) cells. It was found that steady-state retinoic acid receptor alpha (RARalpha) protein levels were markedly reduced in these cells after exposure to ATRA. Because the presence of receptors for a hormone ligand is required for its action, differentiation therapy with ATRA may be self-limiting. The combination of G-CSF with ATRA significantly attenuated the loss of RARalpha protein, and synergistic terminal differentiation occurred. LiCl was more effective than G-CSF in preserving RARalpha pools and synergized with ATRA more strongly than G-CSF. These findings suggested that the prevention of RARalpha protein loss by G-CSF or LiCl in ATRA-treated cells functioned to extend the differentiation response to the retinoid and was responsible, at least in part, for the observed synergism. D(+) cells transfected with an expression plasmid containing RARalpha cDNA had a 6- to 8-fold increase in steady-state RARalpha mRNA compared with vector-transfected cells and showed a 2- to 3-fold increase in RARalpha protein. ATRA caused a reduction, but not a complete loss, of RARalpha protein in these transfectants, which were considerably more responsive than parental D(+) cells to ATRA as a single agent, supporting the concept that the protection of RARalpha pools results in a heightened differentiation response to ATRA.