Search

Acute myeloid leukemia (AML) is a myeloid progenitor-derived neoplasm of poor prognosis, particularly among the elderly, in whom age and comorbidities preclude the use of intensive therapies. Novel therapeutic approaches for AML are therefore critically needed. Adenosine monophosphate (AMP) activated protein kinase (AMPK) is a pleiotropic serine/threonine kinase promoting catabolism that represses anabolism and enhances autophagy in response to stress1. AMPK heterotrimers comprise catalytic α- and regulatory β- and γ-subunits, the latter harboring binding sites for AMP. Targets of AMPK include a host of metabolic pathway enzymes mediating carbohydrate, lipid and protein synthesis and metabolism. Accumulating evidence implicates AMPK in cancer biology, primarily as a tumor suppressor, although minimal AMPK activity may also be required for cancer cell growth under stress conditions2,3. Pharmacological activation of AMPK thus represents an attractive new strategy for targeting AML. We previously used the selective small molecule AMPK activator GSK621 to show that AMPK activation induces cytotoxicity in AML but not in normal hematopoietic cells, contingent on concomitant activation of the mammalian target of rapamycin complex 1 (mTORC1)4. However, the precise mechanisms of AMPK-induced AML cytotoxicity have remained unclear. We integrated gene expression profiling and bioinformatics proteomic analysis to identify the serine/threonine kinase PERK - one of the key effectors of the endoplasmic reticulum stress response - as a potential novel target of AMPK. We showed that PERK was directly phosphorylated by AMPK on at least two conserved residues (serine 439 and threonine 680) and that AMPK activators elicited a PERK/eIF2A signaling cascade independent of the endoplasmic reticulum stress response in AML cells. CRISPR/Cas9 depletion and complementation assays illuminated a critical role for PERK in apoptotic cell death induced by pharmacological AMPK activation. Indeed, GSK621 induced mitochondrial membrane depolarization and apoptosis in AML cells, an effect that was mitigated when cells were depleted of PERK or expressed PERK with a loss of function AMPK phosphorylation site mutation. We identified the mitochondrial enzyme aldehyde dehydrogenase 2 (ALDH2) as a downstream target of the AMPK/PERK pathway, as its expression was enhanced in PERK knockdown AML cells. Moreover, selective pharmacologic activation of ALDH2 by the small molecule ALDA-1 recapitulated the protective effects of PERK depletion in the face of pharmacological AMPK activation. Corroborating the impact of the AMPK/PERK axis on mitochondrial apoptotic function, BH3 profiling showed marked Bcl-2 dependency in AML cells treated with GSK621. This dependency was abrogated in PERK-depleted cells, suggesting a role for PERK in mitochondrial priming to cell death. In vitro drug combination studies further demonstrated synergy between the clinical grade Bcl-2 inhibitor venetoclax (ABT-199) and each of four AMPK activators (GSK621, MK-8722, PF-06409577 and compound 991) in multiple AML cell lines. Finally, the addition of GSK621 to venetoclax enhanced anti-leukemic activity in primary AML patient samples ex vivo and in humanized mouse models in vivo. These findings together clarify the mechanisms of cytotoxicity induced by AMPK activation and suggest that combining pharmacologic AMPK activators with venetoclax may hold therapeutic promise in AML. References 1. Lin S-C, Hardie DG. AMPK: Sensing Glucose as well as Cellular Energy Status. Cell Metabolism. 2018;27(2):299-313. 2. Hardie DG. Molecular Pathways: Is AMPK a Friend or a Foe in Cancer? Clinical Cancer Research. 2015;21(17):3836-3840. 3. Jeon S-M, Hay N. The double-edged sword of AMPK signaling in cancer and its therapeutic implications. Arch. Pharm. Res. 2015;38(3):346-357. 4. Sujobert P, Poulain L, Paubelle E, et al. Co-activation of AMPK and mTORC1 Induces Cytotoxicity in Acute Myeloid Leukemia. Cell Rep. 2015;11(9):1446-1457. Figure Disclosures Tamburini: Novartis pharmaceutical: Research Funding; Incyte: Research Funding.

While acute myeloid leukemia (AML) is still associated with a low cure rate, recent advances in understanding its molecular complexity have significantly improved therapy for subgroups of patients, including those harboring FLT3, IDH1 or IDH2 mutations1. However, more than half of AML cases still lack a druggable oncogenic target. Human cancers frequently harbor mutations in RAS oncogene family members, including NRAS, KRAS and HRAS, which drive oncogenesis by augmenting cellular proliferation and survival. These are small protein GTPases, regulated by a switch between active GTP-linked and inactive GDP-bound states governed by a complex network of guanine exchange factors (GEFs, favoring RAS-GTP) and GTPase activating factors (GAPs, favoring RAS-GDP). RAS activation - due to either extrinsic recruitment by transmembrane tyrosine kinase receptors or intrinsic mutations - propagates through the downstream RAF/MEK/ERK and PI3K/AKT signaling pathways. Besides RAS-activating mutations conferring independence from physiological regulators, human cancers harbor mutations in other RAS network genes such as NF1 (encoding neurofibromin, a RAS GAP), BRAF or PTPN11 (encoding the SHP2 tyrosine phosphatase involved in RAS activation). Somatic alterations of RAS pathway genes, notably NRAS, KRAS, PTPN11 (missense mutations) and NF1 (mutations and deletions), are reported in up to 20% of AML cases2. Generally arising as late driver events, RAS pathway mutations participate in leukemogenesis through mitogen activated protein kinase (MAPK) activation. The anti-tumor activity of MEK inhibitors in Nras-mutated AML in mice and in some NRAS or KRAS-mutated AML patients suggests that deregulated RAS pathway signaling may represent a bona fide therapeutic target. However, strategies to inhibit RAS - indirectly in most cases - have been hampered by signaling feedback, redundancy and tumor heterogeneity 3. We identified 127 cases of AML with unmet therapeutic need within a cohort from which we excluded those with European leukemia network (ELN) favorable prognosis or FLT3-ITD mutations. Targeted next-generation sequencing revealed RAS pathway alterations in 50 patients (39.3%) and NF1 mutations and deletions, mostly large cytogenetically detected deletions, in 17 (14.8%). NRAS, KRAS, PTPN11, CBL and BRAF variants were detected in 13 (10.4%), 10 (7.9%), 9 (7.2%), 5 (3.9%) and 2 (1.6%) cases, respectively. Mutations in RAF1, RASA1, SOS1 and MAP2K2 were observed in a single case each. RAS pathway alterations appeared in the putative main leukemic clone as well as in subclones inferred from variant allele frequencies. Concurrent RAS pathway mutations were observed in nine cases. Among 79 patients homogeneously treated with intensive induction chemotherapy, RAS pathway alterations correlated with higher clinical proliferation markers (elevated white blood cell count, blast cell percentage and LDH levels) and reduced survival probability, particularly within the ELN intermediate-risk subgroup. We established robust models of RAS/MAPK activation through genetic NF1 disruption or expression of NRASG12D or PTPN11D61Y in growth factor (GF)-dependent cell lines. We assessed oncogenic addiction to the RAS pathway in these cells through GF-independence, increased RAS activity, faster propagation in immunocompromised mice and an exquisite sensitivity to pharmacological MEK inhibition in vitro and in vivo. High-content pharmacological screens with FDA-approved molecules identified pyrvinium pamoate, an anti-helminthic agent, as preferentially active in RAS-activated cells. This compound significantly impaired cell viability and colony formation in primary AML samples with RAS pathway alterations. Moreover, the combination of trametinib and pyrvinium pamoate demonstrated synergy in cell line models and even primary samples. While pyrvinium pamoate strongly inhibited mitochondrial respiration and induced metabolic reprograming towards increased glycolysis, trametinib impaired glycolysis and mitochondrial respiratory capacity, suggesting a mechanistic basis for the synergy observed. These data highlight the translational opportunity in developing pyrvinium pamoate for RAS pathway mutated AML. References 1. Raj RV et al. Leuk. Res. 2018;74:113-120. 2. Simanshu DK et al. Cell. 2017;170(1):17-33. 3.Ryan MB et al. Nat Rev Clin Oncol. 2018;15(11):709-720. Figure Disclosures Hermine: AB Science: Membership on an entity's Board of Directors or advisory committees. Tamburini:Novartis pharmaceutical: Research Funding; Incyte: Research Funding.