A Combination of Retinoic Acid and Proteasome Inhibitors for the Treatment of Leukemias Is Potentially Dangerous

To the Editor:

Retinoic acid (RA) is used in the treatment of leukemias; however, the phenomenon of RA resistance is often the cause of relapse in some patients. In a recent report, Fanelli et al1 showed, using an acute promyelocytic leukemia–derived cell line NB4, that the treatment of RA-resistant cells with proteasome inhibitors restores RA sensitivity.

Proteasome inhibitors are already being tested in phase I of clinical trials as anticancer agents2 and are proposed to be useful in the treatment of chemoresistant and radioresistant leukemias.3 One might speculate that combination therapy of RA and proteasome inhibitors can be advantageous and may potentially overcome RA-resistant cases. A selective inhibitor of the proteasome [PSI;N-benzyloxycarbonyl-Ile-Glu-(O-t-butyl)-Ala-leucinal] induces apoptosis of murine lymphocytic leukemia L1210 cells,4 while RA treatment inhibits their proliferation,5 abolishes clonogenicity, and causes G1 arrest.6 L1210 is sensitive to RA, regardless of the fact that it does not express PML/RARα or RA binding proteins.5,7 

We pretreated L1210 cells for 72 hours with either a control solvent (dimethyl sulfoxide [DMSO]) or 5 μmol/L and 50 μmol/L RA. Cells were then further treated for 24 hours with a combination of those agents and 500 nmol/L PSI or DMSO alone as control. Treated cells were fixed in 70% ethanol, stained with acridine orange, and analyzed using a FACSCalibur flow cytometer (Becton Dickinson, San Diego, CA).8 

We have found that treatment with 5 μmol/L RA, which by itself is not cytotoxic, has a cytotoxic effect when combined with 500 nmol/L PSI (Table 1). However, a combination of 50 μmol/L RA, which is by itself cytotoxic, with 500 nmol/L PSI surprisingly diminishes the cytotoxic effects on L1210 cells, allowing a survival of 21.4% of the cells compared with 12.2% for PSI alone (χ²P < .0001). Moreover, while RA or PSI alone caused a G0/G1 block, this effect was partially abolished by a combination of both drugs (Table 2). PSI or 5 μmol/L RA treatment did not increase granulosity of the cells, as judged by the flow cytometric SSC-H parameter, while 50 μmol/L RA increased mean granulosity by 62%. The combination of 50 μmol/L RA and PSI increased mean granulosity by 50%.

Although 5 μmol/L RA does not induce differentiation—it causes only a moderate G1 phase block or retardation—50 μmol/L RA treatment most probably induces differentiation of a fraction of L1210 cells, as judged by increased granulosity, and these cells became resistant to proteasome inhibition. Although proteasome inhibitors cause apoptosis of proliferating cells,4 they prevent apoptosis of some differentiated cells, such as neurons9 or thymocytes.10 Our results suggest that any attempt to treat leukemia with a combination of RA and proteasome inhibitors should be made with great caution. It potentially allows the survival of cells, which after the withdrawal of the drugs can cause a fast relapse of the leukemic disease, regardless of the presence of the PML/RMRα fusion protein.