Efficient RNA interference in patients' acute lymphoblastic leukemia cells amplified as xenografts in mice

RNAi-Mediated Screen of Primary AML Cells Nominates MDM4 as a Therapeutic Target in NK-AML with DNMT3A Mutations

DNA-methyltransferase 3A (DNMT3A) mutations belong to the most frequent genetic aberrations found in adult acute myeloid leukemia (AML). Recent evidence suggests that these mutations arise early in leukemogenesis, marking leukemic progenitors and stem cells, and persist through consolidation chemotherapy, providing a pool for AML relapse. Currently, there are no therapeutic approaches directed specifically against this cell population. To unravel therapeutically actionable targets in mutant DNMT3A-driven AML cells, we have performed a focused RNAi screen in a panel of 30 primary AML samples, all carrying a DNMT3A R882 mutation. As one of the strongest hits, we identified MDM4 as a gene essential for proliferation of primary DNMT3A^WT/R882X AML cells. We analyzed a publicly available RNA-Seq dataset of primary normal karyotype (NK) AML samples and found a trend towards MDM4 transcript overexpression particularly in DNMT3A-mutant samples. Moreover, we found that the MDM2/4 inhibitor ALRN-6924 impairs growth of DNMT3A^WT/R882X primary cells in vitro by inducing cell cycle arrest through upregulation of p53 target genes. Our results suggest that MDM4 inhibition is a potential target in NK-AML patients bearing DNMT3A R882X mutations.

Cell cycle-arrested tumor cells exhibit increased sensitivity towards TRAIL-induced apoptosis

Resting tumor cells represent a huge challenge during anticancer therapy due to their increased treatment resistance. TNF-related apoptosis-inducing ligand (TRAIL) is a putative future anticancer drug, currently in phases I and II clinical studies. We recently showed that TRAIL is able to target leukemia stem cell surrogates. Here, we tested the ability of TRAIL to target cell cycle-arrested tumor cells. Cell cycle arrest was induced in tumor cell lines and xenografted tumor cells in G0, G1 or G2 using cytotoxic drugs, phase-specific inhibitors or RNA interference against cyclinB and E. Biochemical or molecular arrest at any point of the cell cycle increased TRAIL-induced apoptosis. Accordingly, when cell cycle arrest was disabled by addition of caffeine, the antitumor activity of TRAIL was reduced. Most important for clinical translation, tumor cells from three children with B precursor or T cell acute lymphoblastic leukemia showed increased TRAIL-induced apoptosis upon knockdown of either cyclinB or cyclinE, arresting the cell cycle in G2 or G1, respectively. Taken together and in contrast to most conventional cytotoxic drugs, TRAIL exerts enhanced antitumor activity against cell cycle-arrested tumor cells. Therefore, TRAIL might represent an interesting drug to treat static-tumor disease, for example, during minimal residual disease.

Impact of the p53 status of tumor cells on extrinsic and intrinsic apoptosis signaling

Background

The p53 protein is the best studied target in human cancer. For decades, p53 has been believed to act mainly as a tumor suppressor and by transcriptional regulation. Only recently, the complex and diverse function of p53 has attracted more attention. Using several molecular approaches, we studied the impact of different p53 variants on extrinsic and intrinsic apoptosis signaling.

Results

We reproduced the previously published results within intrinsic apoptosis induction: while wild-type p53 promoted cell death, different p53 mutations reduced apoptosis sensitivity. The prediction of the impact of the p53 status on the extrinsic cell death induction was much more complex. The presence of p53 in tumor cell lines and primary xenograft tumor cells resulted in either augmented, unchanged or reduced cell death. The substitution of wild-type p53 by mutant p53 did not affect the extrinsic apoptosis inducing capacity.

Conclusions

In summary, we have identified a non-expected impact of p53 on extrinsic cell death induction. We suggest that the impact of the p53 status of tumor cells on extrinsic apoptosis signaling should be studied in detail especially in the context of therapeutic approaches that aim to restore p53 function to facilitate cell death via the extrinsic apoptosis pathway.

NOXA as critical mediator for drug combinations in polychemotherapy

During polychemotherapy, cytotoxic drugs are given in combinations to enhance their anti-tumor effectiveness. For most drug combinations, underlying signaling mechanisms responsible for positive drug-drug interactions remain elusive. Here, we prove a decisive role for the Bcl-2 family member NOXA to mediate cell death by certain drug combinations, even if drugs were combined which acted independently from NOXA, when given alone. In proof-of-principle studies, betulinic acid, doxorubicin and vincristine induced cell death in a p53- and NOXA-independent pathway involving mitochondrial pore formation, release of cytochrome c and caspase activation. In contrast, when betulinic acid was combined with either doxorubicine or vincristine, cell death signaling changed considerably; the drug combinations clearly depended on both p53 and NOXA. Similarly and of high clinical relevance, in patient-derived childhood acute leukemia samples the drug combinations, but not the single drugs depended on p53 and NOXA, as shown by RNA interference studies in patient-derived cells. Our data emphasize that NOXA represents an important target molecule for combinations of drugs that alone do not target NOXA. NOXA might have a special role in regulating apoptosis sensitivity in the complex interplay of polychemotherapy. Deciphering the differences in signaling of single drugs and drug combinations might enable designing highly effective novel polychemotherapy regimens.