Yashaswini Janardhanan, Marco J Herold, Rohit N. Haldar, Jonathan B. Baell, Steven W. Lane, Megan Bywater, Sebastien Jacquelin, Leanne Cooper, Siok-Keen Tey, Claudia Bruedigam, Victoria Ling, Jasmin Straube, Emily Cooper, Lars Bullinger, and William Godfrey
Chemotherapy-resistant acute myeloid leukemia (AML) manifesting as primary refractory or relapsed disease carries a dismal prognosis and is often driven by clonal evolution. We have performed a genome-wide CRISPR knockout screen investigating resistance to conventional AML therapy, cytarabine and doxorubicin (AraC/Dox), in 2 independent human AML cell lines. Chemoresistant populations were enriched with gRNAs disrupting cell cycle arrest genes, including cyclin dependent kinase inhibitor 2A (CDKN2A), checkpoint kinase 2 (CHEK2) and TP53. Here, we validate the direct contribution of these genes to chemoresistance and demonstrate that rationally designed therapeutic regimens targeting cell cycle enhances chemotherapy response in AML. Using CRISPR-mediated gene editing, we individually inactivated CDKN2A and CHEK2 in Cas9-expressing OCI-AML3 and MV4-11 cells. CDKN2A- and CHEK2-deficient cells demonstrated proliferative advantage in the presence of AraC/Dox in co-culture competition assays, confirming direct contribution of these gene knockouts to chemoresistance. Nil to modest reductions in apoptosis were seen in CDKN2A- and CHEK2-deficient cells treated with AraC/Dox for 72 hours compared to unedited controls. However, failure of accumulation of cells in the non-cycling G 0/G 1 proportion was seen in these edited cells after chemotherapy, corresponding with maintained DNA synthesis, as measured by BrdU incorporation, and a failure to downregulate phospho-Rb protein expression, indicating ongoing active cell cycling in spite of chemotherapy. These results confirm failure of cell cycle arrest as the major mechanism of resistance with inactivation of CDKN2A or CHEK2. To assess relevance of these genes in chemotherapy response in human AML, we analysed the effect of CDKN2A expression on prognosis. Reduced expression of CDKN2A conferred inferior overall survival in 3 independent clinical cohorts. Additionally, downregulation of CDKN2A and an increase in downstream cell cycling effector, cyclin-dependent kinase 6 (CDK6) was seen at relapse in paired diagnosis-relapse AML samples (Li et al. 2016, Nature Medicine). Further, CHEK2 mutations in clonal hematopoiesis are enriched in patients with solid organ cancers following chemo- or radiotherapy, functionally demonstrating chemoresistance over non-mutated cells. These data suggest CDKN2A and CHEK2 loss-of-function is relevant in promoting chemoresistance in human hematopoiesis. We therefore investigated whether therapeutically targeting cell cycle control pathways that converge on the G 1/S restriction point could synergise with cytotoxic chemotherapy, potentially circumventing chemoresistance. The addition of MDM2 inhibitor, nutlin-3a with AraC/Dox achieved striking synergism in TP53-competent cell lines in reducing viability and promoting apoptosis. Inhibition of CDK4/6 with palbociclib alone induced cell cycle arrest without apoptosis, however combination therapy with AraC/Dox significantly potentiated apoptosis. To functionally validate the role of CDKN2A itself, we examined the inhibitor of histone acetyltransferase KAT6A, WM-1119 (Baell et al. 2018, Nature), which transcriptionally upregulates CDKN2A. WM-1119 had anti-leukemic activity across multiple cell lines, corresponding with down regulation of cell cycling and upregulation of senescence signatures. The anti-leukemic activity was prevented by CDKN2A inactivation, indicating that CDKN2A upregulation is essential for WM-1119's anti-leukemic effect. WM-1119 enhanced activity of AraC/Dox in multiple cell contexts, mediated by cooperative induction of CDKN2A. Its efficacy and synergy with AraC/Dox was further demonstrated in MLL-AF9 primary murine AML ex vivo, demonstrating broader applicability of the approach. Our findings reveal defects in cell cycle arrest pathways as clinically relevant contributors to chemoresistance in AML. Combining agents to target cell cycle components improved in vitro anti-leukemic activity of chemotherapy and warrants further exploration of their translational potential. Disclosures Bullinger: Amgen: Honoraria; Bristol-Myers Squibb: Consultancy, Honoraria; Celgene: Consultancy, Honoraria; Menarini: Consultancy; Hexal: Consultancy; Jazz Pharmaceuticals: Consultancy, Honoraria, Research Funding; Bayer: Research Funding; Pfizer: Consultancy, Honoraria; Astellas: Honoraria; Gilead: Consultancy; Daiichi Sankyo: Consultancy, Honoraria; Seattle Genetics: Honoraria; Janssen: Consultancy, Honoraria; Novartis: Consultancy, Honoraria; Sanofi: Honoraria; Abbvie: Consultancy, Honoraria. Lane: BMS: Consultancy, Research Funding; Astellas: Membership on an entity's Board of Directors or advisory committees; Abbvie: Honoraria; Novartis: Consultancy; Geron: Consultancy.