BORTEZOMIB SPECIFICALLY TARGETS A SUBSET OF HYPOXIA-SELECTED CML STEM CELLS

Introduction. We previously demonstrated that severe hypoxia selects LSC from CML or other leukaemias and that hypoxia-adaptation is a common feature of HSC and LSC. We also found that BCR/Abl protein is suppressed (although mRNA is not) in hypoxia-selected CML cells, so that these cells are independent of BCR/Abl signalling for their maintenance. Hypoxia-resistant LSC of CML thereby resulted refractory to Imatinib- mesylate (IM) due to the lack of its molecular target. This study was targeted to determine the effects of the proteasome inhibitor Bortezomib (BZ) on hypoxia-selected, IM-refractory LSC of CML. Methods. Cells of the stabilized CML line K562 were used. BCR/Abl protein expression was determined by SDS-PAGE/western blotting. The maintenance of stem cell potential in hypoxic (0.3% O2) primary cultures (LC1) was assessed by the Culture-Repopulating Ability (CRA) assay on the basis of the capacity of LC1 cells to repopulate secondary liquid cultures (LC2) incubated in normoxia. Results. The addition of BZ at time-zero (t0) of incubation in hypoxia (as well as in normoxia) markedly reduced the number of viable cells, whereas that at day 1 of hypoxia was completely ineffective. In hypoxic LC1 treated with BZ from t0, BCR/Abl suppression was faster than in untreated cultures, where BCR/Abl was still well expressed at day 2 of incubation. In cells treated with BZ at day 1, BCR/Abl suppression was delayed to day 3, indicating that it was driven by the proteasome, and also that BZ was devoid of aspecific toxicity. In hypoxia, before its suppression, BCR/Abl was cleared from cytosol to accumulate into the nucleus in its enzymatically-inactive form. Treatment with BZ at day 1 did not interfere with nuclear translocation of BCR/Abl, extending its survival therein. The CRA assay was carried out with cells recovered from hypoxic LC1 at days 2, 3 or 7. Cells recovered at day 2 started to repopulate LC2 immediately, due to the maintenance of BCR/Abl, to peak at day 10. LC2 repopulation by cells from day-3 LC1 was delayed, starting from day 7 and peaking at day 21. BZ did not alter either of these kinetics. The kinetics of LC2 repopulation by cells from day-7 hypoxic LC1 was identical to that of day-3 LC1 cells, but in this case BZ (added at either t0 or day 1) significantly reduced LC2 repopulation with respect to untreated controls. Conclusions. Incubation of K562 cells in hypoxia for 1 day protects completely cells bulk from the toxic effects of BZ. In hypoxia, BCR/Abl translocates into the nucleus and is suppressed, at least in part, via proteasome. BZ delays, but not prevents, disappearance of BCR/Abl from nucleus. The CRA assay defined 3 progenitor subsets: A, progenitors selected in hypoxic LC1 at day 2, BCR/Abl+ and insensitive to BZ; B, LSC selected at day 3, BCR/Abl- and insensitive to BZ; C, LSC selected at day 7, BCR/Abl- and insensitive to BZ. Thus, BZ seems selectively active on the CML progenitor subset which is typically refractory to IM.