A non-ATP-competitive inhibitor of BCR-ABL overrides imatinib resistance

Abstract

Imatinib, which is an inhibitor of the BCR-ABL tyrosine kinase, has been a remarkable success for the treatment of Philadelphia chromosome-positive (Ph+) chronic myelogenous leukemias (CMLs). However, a significant proportion of patients chronically treated with imatinib develop resistance because of the acquisition of mutations in the kinase domain of BCR-ABL. Mutations occur at residues directly implicated in imatinib binding or, more commonly, at residues important for the ability of the kinase to adopt the specific closed (inactive) conformation to which imatinib binds. In our quest to develop new BCR-ABL inhibitors, we chose to target regions outside the ATP-binding site of this enzyme because these compounds offer the potential to be unaffected by mutations that make CML cells resistant to imatinib. Here we describe the activity of one compound, ON012380, that can specifically inhibit BCR-ABL and induce cell death of Ph+ CML cells at a concentration of <10 nM. Kinetic studies demonstrate that this compound is not ATP-competitive but is substrate-competitive and works synergistically with imatinib in wild-type BCR-ABL inhibition. More importantly, ON012380 was found to induce apoptosis of all of the known imatinib-resistant mutants at concentrations of <10 nM concentration in vitro and cause regression of leukemias induced by i.v. injection of 32Dcl3 cells expressing the imatinib-resistant BCR-ABL isoform T315I. Daily i.v. dosing for up to 3 weeks with a >100 mg/kg concentration of this agent is well tolerated in rodents, without any hematotoxicity.

Fig. 1.

BCR-ABL inhibitory activity of ON012380. (A) Structure of ON012380. (B and C) Ten nanograms of recombinant BCR-ABL protein was mixed with different concentrations of the indicated inhibitor, and kinase assays were performed by using Crk as a substrate to measure autophosphorylation and substrate (Crk) phosphorylation. (D) BCR-ABL kinase assays were performed as described in Materials and Methods by using c-Crk as a substrate. The reactions mixtures were spotted onto strips of P81 phosphocellulose paper, washed, and counted. In experiments for which a mixture of imatinib and ON012380 was used, the reaction mixtures contained a constant amount of imatinib (10 nM) and various amounts of ON012380. The values from individual samples were analyzed and plotted as a function of drug concentration. Data points represent an average of three independent experiments performed in duplicate.

Fig. 2.

Steady-state kinetic analysis of BCR-ABL kinase inhibition by ON012380. (A) BCR-ABL kinase inhibition assays were performed as described for Fig. 1 in a reaction mixture containing [γ-³²P]ATP and various concentrations (conc.) of ATP. The values from individual samples were analyzed and plotted as a function of inhibitor concentration. The IC₅₀ of ON012380 for kinase activity was calculated. (B) The curves represent calculated best fits to the Michaelis–Menton equation with a constant amount of substrate and various amounts of ATP and ON012380. (C) BCR-ABL kinase inhibition assays with different concentrations of ON012380 and various concentrations of substrate (Crk) were performed, and the values from individual samples were analyzed and plotted as a function of inhibitor concentration. (D) Michaelis–Menton curves for BCR-ABL with a curve fit derived by using nonlinear regression analysis is shown for data obtained by using a constant amount of ATP and various amounts of substrate and ON012380. (E) Inhibition assays with recombinant BCR-ABL protein and different concentrations of imatinib were performed in the presence of various concentrations of ATP as described for A. (F) The curves represent calculated best fits to the Michaelis–Menton equation with a constant amount of substrate and various amounts of ATP and imatinib. (G) Inhibition assays with recombinant BCR-ABL protein and different concentrations of imatinib were performed in the presence of various concentrations of substrate (Crk) as described for E, and the values from individual samples were analyzed and plotted as a function of drug concentration. (H) The curves represent calculated best fits to the Michaelis–Menton equation with a constant amount of ATP and various amounts of substrate and imatinib.

Fig. 3.

In vitro tumor-cell-killing activity of ON012380. (A and B) Effect of ON012380 and imatinib on the viability of Ph⁺ human CML K562 (A) and murine 32D/BCR-ABL (B) cells. The two cell lines were incubated with increasing concentrations of the indicated compounds, and the total number of viable cells was determined 72 h after treatment by trypan blue exclusion. (C and D) K562 cells were treated with various concentrations of imatinib or ON012380, and the BCR-ABL immunoprecipitates were subjected to in vitro kinase assays. BCR-ABL autophosphorylation (C) and substrate (Crk) phosphorylation (D) are shown. (E) Total cell lysates derived from K562 cells treated with DMSO (control), imatinib, or ON012380 were examined by Western blot analysis for the expression of phosphorylated (Upper) and nonphosphorylated (Lower) forms of STAT-5.

Fig. 4.

BCR-ABL^T315I inhibitory activity of ON012380. (A) BCR-ABL kinase assays were performed with T315I recombinant protein as described for Fig. 1 by using c-Crk as a substrate. The values from individual samples were analyzed and plotted as a function of inhibitor concentration. (B and C) Imatinib-resistant 32D/BCR-ABL^T315I cells were treated with various concentrations of imatinib or ON012380, and the BCR-ABL immunoprecipitates were subjected to in vitro kinase assays. BCR-ABL autophosphorylation (B) and substrate phosphorylation (Crk) (C) are shown. (D) Total cell lysates derived from 32D/BCR-ABL^T315I cells treated with DMSO (control), imatinib, or ON012380 were examined by Western blot analysis for the expression of phosphorylated STAT-5.

Fig. 5.

In vitro tumor-cell-killing activity of cells expressing an inatinib-resistant mutant of BCR-ABL by ON012380. The four representative imatinib-resistant cell lines T315I (A), E255K (B), Y253H (C), and G250E (D) were incubated with increasing concentrations of the indicated compounds, and the total number of viable cells was determined 72 h after treatment by trypan blue exclusion.

Fig. 6.

Effect of ON012380 on the in vivo growth of T315I cells. (A) Nude mice (10 mice per group) were injected i.v. through the tail vein with 1 × 10⁶ 32D/BCR-ABL^T315I cells. Treatment with daily i.p. injections of 100 mg/kg ON012380, 100 mg/kg imatinib, or an equal volume of saline was initiated 24 h later. Blood smears from each mouse were performed on days 7 and 14, and the number of T315I-expressing cells per 10 fields was determined. The data were plotted as the average number of 32D/BCR-ABL^T315I cells per 10 fields ± SEM (n = 10). (B) The total body weight of individual mice in the three groups was determined daily, and the average body weights were plotted as the percent of starting body weight. (C) CD-1 mice were injected i.v. (tail vein injection) with saline or ON012380 (200 mg/kg) dissolved in saline. Bone marrow cells were extracted from the mice after 24 h, and 2 × 10⁵ cells were plated on methycellulose containing appropriate cytokines for lineage-specific colony formation. Colonies were counted after 5–14 days of incubation. BFU-E, erythroid burst-forming unit; CFU-G, granulocyte colony-forming unit; CFU-M, macrophage colony-forming unit; CFU-GM, granulocyte/macrophage colony-forming unit; pre-B, pre-B lymphocyte.