Dissecting the Autophagy Tumor Suppressor Pathway Network in Acute Promyelocytic Leukemia Therapy

Acute promyelocytic leukemia (APL) is a rare disease that is successfully treated with targeted therapy in the clinic using all-transretinoic acid (ATRA) in combination with chemotherapy or arsenic trioxide (ATO). APL is invariably characterized by a translocation involving the retinoic acid receptor alpha (RARA) and genes encoding proteins with self-aggregation motifs. The translocation t(15;17) leads to the expression of the oncofusion protein PML-RARA that results in a differentiation arrest, an increase of self-renewal potential and an anti-apoptotic phenotype. Unfortunately, resistance to ATRA and ATO treatment are described in patients due to mutations in the RARA or PML moiety, respectively. Therefore, a better understanding of APL therapy induced molecular responses will allow to address treatment resistances. We and others described that current APL therapies induce autophagy. Autophagy is a proteolytic self-degradation process characterized by the formation of double-membraned vesicles, so called autophagosomes, which engulf cytoplasmic contents such as protein aggregates or defective organelles. In APL therapy autophagy is among others involved in PML-RARA degradation. We also found that key autophagy-related proteins such as WIPI1, DRAM1, ATG7 and ATG5 are essential for APL cells differentiation. A comprehensive study of APL therapy induced autophagy is needed to design meaningful combination therapies with autophagy modulating drugs.

We previously found significantly lower Death-Associated Protein Kinase 2 (DAPK2) expression levels in acute myeloid leukemia (AML) patients with particular low levels in APL (Humbert et al, J Leukoc Biol 2014). DAPK2 is a positive mediator of autophagy and cell death in different cellular systems. To test if DAPK2 is also involved in autophagy and cell death responses to APL therapy, we first determined DAPK2 levels upon ATRA or ATO treatment. Both treatments significantly induced DAPK2 expression in NB4 and HT93 APL cell lines model paralleled by autophagy and cell death induction. Next, we used an immunoprecipitation screening approach to identify DAPK2 binding partners during APL therapy. We found that the known tumor suppressor p73 as well as the autophagy-related (ATG) protein ATG5 and a short form of Beclin1 that is involved in apoptosis co-precipitate with DAPK2. Then, we knocked down DAPK2 in NB4 cells via shRNAs and determined autophagic and cell death responses upon ATRA and ATO treatment. Autophagic flux was determined by endogenous Light Chain 3 (LC3)B-II puncta formation and levels by immunofluorescence microscopy and by western blot in presence or absence of the lysosomal inhibitor, Bafilomycin A1. Silencing of DAPK2 led to decreased autophagic activity upon ATRA treatment. In addition, DAPK2 depletion in NB4 cells caused a decrease in ATG5-ATG12 complex formation in both APL therapies as well as an accumulation of full-length Beclin1. Surprisingly, inhibiting DAPK2 expression did not significantly impair autophagic activity upon ATO treatment but attenuated ATO-induced apoptosis (p<0.05). Since ATRA induces a Beclin-1 independent, non-canonical, autophagy, we propose that DAPK2 regulates autophagy via ATG5-ATG12 complex formation and apoptosis via Beclin1 cleavage regulation. Moreover, knocking down p73 resulted in autophagy and apoptosis inhibition during ATRA and ATO treatment, potentially due to its downstream targets DRAM1, ATG5 and CDKN1A. Furthermore, we showed that DAPK2 stabilizes p73 protein levels within the nucleus and that silencing DAPK2 led to a significant reduction of the p73 transcriptional target gene, CDKN1A (p<0.05). Additionally, DAPK2 transcription is directly regulated by p73 thereby creating a positive feedback loop during APL treatment. Together, we identified novel protein-protein interaction networks involved in ATRA and ATO induced autophagy and cell death responses.

In conclusion, we provide strong evidence that DAPK2 is at the nexus of apoptosis (DAPK2-Beclin1 short) and autophagy (DAPK2-ATG5) pathways in response to APL therapies. Thus, we identified an autophagy-related network involved in APL therapy responses that represents a promising new target in therapeutic strategies to treat APL.