Abstract
BACKGROUND:
The transcription factor WT1 is a “pan-leukemic marker”. WT1 is overexpressed in a significant number of AML, CML, ALL, and MDS cells. Moreover, the persistence of WT1 expression after chemotherapy indicates residual leukemic stem cells. Hydroxyurea can decrease the white blood cell count in leukemic patients. Hydroxyurea inhibits ribonucleotide reductase and consequently depletes the cellular dNTP pool. This process induces replicative stress and activates the intra-S phase checkpoint. The prolonged arrest of DNA replication leads to double strand DNA breaks and apoptosis. The aim of our study was to investigate whether WT1 has an impact on the survival of AML and CML cells treated with hydroxyurea. Furthermore, we analyzed whether replicative stress affects WT1 expression. We also addressed whether clinically relevant drugs modulating WT1 influence apoptosis, autophagy, and DNA damage signaling in leukemic cells.
METHODS:
We used human AML and CML cells (MV4-11, RS4-11, NB4, HEL, K562) and primary AML cells with various genetic lesions (FLT3-ITD, MLL-AF4, PML-RARα, JAK2V617F, BCR-ABL; p53 mutant or null) or a complex karyotype. Cells were exposed to 0,5 mM hydroxyurea and analyzed by FACS, Western Blot and mRNA expression analyses. RNAi with morpholinos targeting the WT1 mRNA as well as the tyrosine kinase inhibitor imatinib (0,1-1 µM) and epigenetic modifiers of the histone deacetylase inhibitor family (HDACi: 1,5-5 µM entinostat and 10-100 nM panobinostat) were used to decrease WT1 levels.
RESULTS:
The stability of WT1 is causally linked to the survival of leukemic cells undergoing replicative stress. Hydroxyurea-treated leukemic cells retaining WT1 (K562 and RS4-11 cells) accumulate in S phase and survive. However, cells that lose WT1 (NB4, MV4-11, HEL, and primary AML cells) succumb to apoptosis (57-70% apoptosis after 24 h; p<0,001).
The specific elimination of WT1 in K562 cells with morpholino antisense nucleotides (10 nM) significantly sensitizes the cells to hydroxyurea-induced apoptosis (26-30% increased apoptosis rates after 24 h; p<0,005).
The hydroxyurea-induced loss of WT1 in sensitive cells is caspase-dependent but does not require the p53 tumor suppressor. The pan-caspase inhibitor ZVAD-FMK (20-50 µM) and the ectopic overexpression of mitochondrial anti-apoptotic proteins (BCL-XL/BCL2) prevents the hydroxyurea-induced depletion of WT1 and apoptosis in NB4 cells (from 62-69% to 9-14%; p<0,001).
Moreover, our data reveal that hydroxyurea promotes the occurrence of previously unrecognized WT1 cleavage products. These are different from the known HTRA2-induced WT1 fragments, as they contain the WT1 zinc fingers required for DNA binding. These novel cleavage products also contain the domain for transcriptional activation, but lack the domain for transcriptional repression. We demonstrate that caspases cleave the full-length WT1 protein into such fragments, which accumulate in the nucleus of hydroxyurea-treated NB4 cells.
Resistance of BCR-ABL-positive K562 CML cells to pro-apoptotic effects of hydroxyurea can be overcome with the tyrosine kinase inhibitor imatinib. This drug inhibits the WT1 promoter and sensitizes the cells to hydroxyurea (22-48% increased apoptosis rate after 24 h; p<0,005).
The HDACi entinostat and panobinostat suppress the WT1 gene and accelerate the proteasomal degradation of WT1 by an increased expression of the E2 ubiquitin conjugase UBCH8. The depletion of WT1 by HDACi renders K562 cells sensitive towards hydroxyurea-induced killing and causes a synergistic activation of cell death (apoptosis increased by 43-55% after 24 h; p<0,001).
In K562 cells hydroxyurea/entinostat combinations lead to a caspase-dependent cleavage of the DNA repair enzyme PARP-1 and to the accumulation of double-strand DNA breaks.
We further show that entinostat significantly shifts the balance from a hydroxyurea-induced autophagy program to apoptosis (from 60% to 16% autophagy; p<0,001) in K562 cells. We reveal that entinostat depletes proteins critical for autophagy and autophagosome-lysosome fusion (Beclin-1 and HDAC6) in these cells.
CONCLUSIONS:
Our data illustrate that WT1 is a novel biological target of hydroxyurea, independent of a specific genetic lesion and irrespective of a functional p53 pathway. The inhibition of WT1 might be therapeutically exploitable for targeting approaches against leukemic cells.
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.