GLEEVEC-RESISTANT CML STEM AND PROGENITOR CELLS ARE SENSITIVE TO HIF-INHIBITING DRUGS

Introduction. We previously demonstrated that incubation in severe hypoxia of blast-crisis Chronic Myeloid Leukaemia (CML) cells results in the complete suppression of BCR/Ablprotein, but not mRNA, so that hypoxia-selected Leukaemia Stem Cells (LSC) of CML, while remaining genotypically leukaemic, are phenotypically independent of BCR/Abl signaling and thereby refractory to the treatment with Imatinib- mesylate (IM). This is in keeping with the notion that IM is extremely effective in ensuring remission, but not cure. Cure indeed relies on the suppression of LSC, which most likely reside in hypoxic stem cell niches in vivo and therein sustain Minimal Residual Disease (MRD). The main target of this study was to address the effects of the inhibition of Hypoxia-Inducible Factor 1 (HIF1 ) on the maintenance of hypoxia-selected LSC of CML. Methods. The BCR/Abl-independent maintenance of LSC in hypoxia has been assessed by the Culture-Repopulating Ability (CRA) assay. The maintenance of stem cell potential by CML cells selected in hypoxic primary cultures (LC1) is assessed by measuring their capacity to repopulate non-selective normoxic secondary liquid cultures (LC2). The CRA assay also allows to estimate the LC1 content with Leukaemia Progenitor Cells (LPC), by transferring to LC2 cells from LC1 at earlier times (days 2-3) than those necessary for LSC selection (day 7). The CRA of LC1 cells is estimated by determining the kinetics of LC2 repopulation. Two human stabilized CML cell lines were used, K562 and KCL-22. Three different drugs were used to inhibit HIF1: (a) a non-specific HIF inhibitor; (b) an inhibitor of coactivator binding to HIF complex; (c) an inhibitor of HIF1 dimerization with HIF1 . Cells were incubated in hypoxic LC1 (0,1 % O2) in the presence of one of the three drugs and transferred at different times into nonselective LC2. Results. Preliminary results obtained in our laboratory with K562 cells where HIF1 or HIF2 were suppressed by shRNA showed that BCR/Ablprotein-expressing LPC (rescued from day-2 LC1) were insensitive to HIF1 or HIF2 suppression, while BCR/Ablprotein-negative LPC (from day-3 LC1) were insensitive to the suppression of HIF2 but not HIF1 . The maintenance of hypoxia-selected LSC requires instead both HIF1 and HIF2. This indicates that the effects of shHIF are cell subsetspecific. The treatment of hypoxic LC1 at time 0 or day 1 with chemical inhibitors (a) or (b) did not reduce the capacity of LPC or LSC to repopulate LC2. Inhibitor (a), however, but not inhibitor (b), resulted effective when administered for three days starting from day 6. Finally, inhibitor (c), given to LC1 at any time, completely suppressed LC2 repopulation by LPC as well as LSC. Conclusions. The detection of cell subsetspecific effects of HIF1 inhibition led to propose some HIF1 inhibitors, together with IM, as a novel CML therapy targeting MRD via suppression of hypoxia-selected LSC. Data are being confirmed in vivo.