Crizotinib acts as ABL1 inhibitor combining ATP-binding with allosteric inhibition and is active against native BCR-ABL1 and its resistance and compound mutants BCR-ABL1T315I and BCR-ABL1T315I-E255K

Abstract

Resistance remains the major clinical challenge for the therapy of Philadelphia chromosome-positive (Ph+) leukemia. With the exception of ponatinib, all approved tyrosine kinase inhibitors (TKIs) are unable to inhibit the common "gatekeeper" mutation T315I. Here we investigated the therapeutic potential of crizotinib, a TKI approved for targeting ALK and ROS1 in non-small cell lung cancer patients, which inhibited also the ABL1 kinase in cell-free systems, for the treatment of advanced and therapy-resistant Ph+ leukemia. By inhibiting the BCR-ABL1 kinase, crizotinib efficiently suppressed growth of Ph+ cells without affecting growth of Ph- cells. It was also active in Ph+ patient-derived long-term cultures (PD-LTCs) independently of the responsiveness/resistance to other TKIs. The efficacy of crizotinib was confirmed in vivo in syngeneic mouse models of BCR-ABL1- or BCR-ABL1^T315I-driven chronic myeloid leukemia-like disease and in BCR-ABL1-driven acute lymphoblastic leukemia (ALL). Although crizotinib binds to the ATP-binding site, it also allosterically affected the myristol binding pocket, the binding site of GNF2 and asciminib (former ABL001). Therefore, crizotinib has a seemingly unique double mechanism of action, on the ATP-binding site and on the myristoylation binding pocket. These findings strongly suggest the clinical evaluation of crizotinib for the treatment of advanced and therapy-resistant Ph+ leukemia.

Keywords: Allosteric inhibition; BCR-ABL1; Compound mutations; Crizotinib; Philadelphia chromosome–positive leukemia; TKI resistance.

Fig. 1

Inhibition of the ABL1 kinase activity by crizotinib blocks the factor-independent growth of Ba/F3 cells mediated by BCR-ABL1 and BCR-ABL1^T315I. a Western blot analysis of lysates of Ba/F3 cells expressing BCR-ABL1 and BCR-ABL1^T315I using antibodies directed against the following targets: c-Abl, Abl-Y245 (anti-phospho-ABL1), Crkl, phosphorylated Crkl, Stat5, phosphorylated Stat5 (anti-phospho-STAT5), and β-tubulin (anti-β-tubulin). Molecular mass reference (kDa) values are presented, and c-Abl and β-tubulin were used as loading controls. To avoid bias of stress-induced signaling by factor withdrawal, we performed these experiments in the presence of IL-3. b The effect of crizotinib on the factor-independent growth of Ba/F3 cells expressing BCR-ABL1 or BCR-ABL1^T315I was assessed in cells selected by IL-3 withdrawal. These cells were seeded in semi-solid medium and exposed to the indicated concentrations of GNF-2 (allosteric inhibitor of ABL1), imatinib, nilotinib, crizotinib, and dasatinib. At day 14, colonies were stained. c The effect of crizotinib on cell proliferation and viability of factor-independent Ba/F3 cells upon the expression of BCR-ABL1, BCR-ABL1^T315I, BCR-ABL1^Y253F, and BCR-ABL1^F317L was assessed with an XTT assay. The lack of cytotoxic effects was confirmed in empty vector-transduced Ba/F3 cells in the presence of IL-3 at a concentration of less than 1 μM. The mean of three experiments ± SD is given. d Crizotinib inhibits human Ph+ patient-derived cell lines and primary Ph+ ALL PD-LTCs. Proliferation/cytotoxicity assays using XTT were performed on human Ph+ cell lines derived from Ph+ ALL or CML patients. SupB15 (Ph+ ALL) expressing p185^BCR-ABL1 or BV-173 cells expressing p210^BCR-ABL1 (lymphocytic CML-BC cells) were exposed to increasing concentrations of crizotinib. The means of three experiments ± SD each performed in triplicates are given. e Proliferation of PD-LTCs - PH (sens - TKI sensitive), BV (res - TKI-resistant), and KÖ (expressing BCR-ABL1^T315I) - XTT assays upon exposure to increasing concentrations of crizotinib were performed. The Ph− PD-LTCs HP (Ph−) was used as a control. The means ± SD of three experiments each performed in triplicates are given. f Interaction of His-ABL with biotin-myristoyl-peptide (100%) and the displacement of the interaction by GNF2 and crizotinib

Fig. 2

The efficacy of crizotinib in vivo in models of Ph+ leukemia. a For the induction of CML-like disease, sub-lethally irradiated C57BL/6N mice were transplanted intravenously with 1 × 10⁵ Sca1⁺-positive BM cells expressing BCR-ABL1 or BCR-ABL1^T315I. Eight mice/group were treated orally either with crizotinib (100 mg/kg) or ponatinib (25 mg/kg) once daily for 20 days (treatment). For the design of in vivo experimentation, see the Supplementary Information. b Crizotinib prolongs the survival of mice with BCR-ABL1-derived ALL. 5 × 10⁴ spleen cells from ALL mice (frozen stock in liquid N₂) were transplanted into sub-lethally (4.5 Gy) irradiated recipients. The mice were treated with crizotinib (100 mg/kg) or ponatinib (25 mg/kg) by gavage for 20 days. c Response of the compound mutation BCR-ABL1^T315I-E255K to crizotinib and ponatinib. The effect of crizotinib on the factor-independent growth of Ba/F3 expressing BCR-ABL1 or BCR-ABL1^T315I-E255K was performed on cells selected by IL-3 withdrawal. These cells were exposed to the indicated concentrations of PF-114, ponatinib, and asciminib. Cell proliferation and viability were assessed by XTT assays. The means ± SD of three experiments are given