Figure 2

Cyclic AMP agonists can protect APL and AML blasts against anthracycline-induced death. (a and b) Mv4-11 (a) and HL60 (b) AML cells were treated for 4 or 7 h, respectively, with DNR in the absence or presence of PGE₂/IBMX (2 μM/0.15 mM) or N⁶-Bz-cAMP (0.4 mM). The % dead cells was scored by microscopy (chromatin condensation), and is given as mean±S.E.M., n=3. Significance by Student t-test: **P<0.01. (c) Extracts isolated from HL60 cells treated with DNR (5 μM) for 3 or 5 h alone or in combination with PGE₂/IBMX (2 μM/0.15 mM) were analyzed by immunoblotting for Hsp90 cochaperone p23. Actin was used as loading control. (d–m) Blasts were isolated from peripheral blood of three patients with diagnosed APL (APL1-APL3) and seven AML patients (AML1-AML7) and incubated with DNR alone or together with PGE₂/IBMX (2 μM/0.3 mM) for 3 h (APL1, APL2 (8 μM DNR)) or 24 h (APL2 (0.2 and 0.8 μM DNR), APL3; AML 1–7). Samples were assessed for drug-induced cell death by FACS analysis of PI uptake (APL1, APL 2) or by Annexin V labeling and PI uptake (APL2, APL3, AML1-AML-7). When present, PGE₂/IBMX was added 0.5 h before DNR. Further experimental details are given in the Materials and Methods section. Similar results were obtained when PGE₂/IBMX was replaced by the cAMP analog N⁶-Bz-cAMP (data not shown). A Wilcoxon paired two-tailed comparison test of the ratio of dead cells after incubation with DNR alone and DNR+either PGE₂/IBMX or N⁶-Bz-cAMP (0.4 mM) for all the experiments performed on the APL1-3 blasts, yielded P<0.004 for a protective effect of the cAMP agonists against P<0.007 for the AML1-7 blasts