ON012380, a putative BCR-ABL kinase inhibitor with a unique mechanism of action in imatinib-resistant cells

Imatinib is very effective in controlling BCR-ABL-expressing leukemias, but resistance remains a concern and has led to the development of additional agents with unique activity against BCR-ABL and imatinib-resistant disease. A compound with a novel mechanism of action, ON012380, was recently described that inhibited both wild-type and T315I mutant BCR-ABL.¹ The distinct mechanism of action was of potential clinical significance since the evidence of activity against both unmutated and T315I-mutant BCR-ABL kinase activity and efficacy against T315I-BCR-ABL cells in animal studies were described. However, although BCR-ABL-directed kinase inhibition was described in detail with isolated enzyme, there was only limited information regarding the effects of this compound on BCR-ABL activity in intact cells and in chronic myeloid leukemia (CML) patient specimens.

To further characterize this compound, we synthesized (Supplementary Text File 1, Supplementary Figure 1) and characterized (Supplementary Figure 2) ON012380 and compared its activity with that of imatinib and dasatinib. We first examined its specificity and efficacy on interleukin (IL)-3-dependent and BCR-ABL-transformed (unmutated or T315I-mutant) BaF3 cells. As shown in Figure 1a, both imatinib and dasatinib reduced the viability of BaF3 cells transformed by wild-type BCR-ABL but did not affect the viability of IL-3-dependent BaF3 cells or those transformed with T315I-BCR-ABL. In contrast, ON012380 not only reduced the viability of BaF3 cells transformed with either unmutated or T315I-BCR-ABL, but also reduced the viability of IL-3-dependent BaF3 cells. These results suggested that the antitumor activity of ON012380 was not solely dependent on BCR-ABL kinase inhibition.

Figure 1.

Effect of tyrosine kinase inhibitors on survival and signaling in IL-3-dependent and BCR-ABL-transformed BaF3 cells. (a) BaF3 cells maintained in IL-3 or transformed by unmutated (w/t, wild type) or T315I-mutant BCR-ABL were incubated with the indicated concentration of inhibitor before analysis of cell growth and viability. Each data point represents the average of four replicates with no more than 5% variance in any point. The results are representative of three additional independent experiments. (b) BaF3 (left) or BaF3 cells transformed by BCR-ABL (right) were incubated with 5 μM imatinib or ON012389 (ON) at the indicated concentration for 2 h before lysates were prepared and analyzed for Stat5 and CrkL tyrosine phosphorylation and protein levels by immunoblotting. Lysates prepared after 24 h of incubation were analyzed for PARP cleavage (bottom) as a measure of apoptosis.

The effects of ON012380 on BCR-ABL signaling and activation of caspase cascades were also compared with imatinib in IL-3-dependent and BCR-ABL-transformed BaF3 cells. Lysates were collected from cells treated with these compounds for 2 or 24 h and probed for early (2 h) changes in Stat5 and CrkL tyrosine phosphorylation and activation of caspase cascades (24 h). In BCR-ABL-transformed cells, imatinib suppressed both CrkL and Stat5 phosphorylation, whereas ON012380 had only minor effects on these substrates after 2 (Figure 1b) or 24 h (data not shown). Both compounds induced poly (ADP-ribose) polymerase (PARP) cleavage (24 h). Although neither Stat5 nor CrkL were highly tyrosine phosphorylated or modulated by inhibitors in IL-3-dependent BaF3 cells, PARP cleavage was activated in ON012380-treated cells. These results suggest that ON012380-induced apoptosis was not dependent on BCR-ABL transformation or associated with inhibition of BCR-ABL signaling.

Imatinib-resistant clonal variants derived from imatinib-sensitive cell lines were also treated with ON012380 and other inhibitors to assess target-specific effects and antitumor activity. BV-173R cells have 3–4 BCR-ABL fusion signals by fluorescent in situ hybridization analysis (Supplementary Text File 2), and although unmutated BCR-ABL was detected (Supplementary Figure 3 and legend), these cells predominantly express the T315I-mutant form of BCR-ABL² and retain surface markers present on parental BV-173 cells (Supplementary Table 1). K562R cells are BCR-ABL kinase mutation negative but overexpress Lyn kinase.^3,4 As shown in Figure 2a (top), imatinib dose-dependently reduced the growth and survival of K562 and BV-173 cells (IC50 ~0.2 μM) but had minimal effects on BV-173R or K562R cell survival (IC50 >5 μM). Dasatinib dose-dependently reduced the viability of K562, K562R and BV-173 cells but had 100-fold less activity against BV-173R cells (dasatinib IC50 ~0.02 μM in BV-173 vs ~2 μM in BV-173R). Dasatinib concentrations necessary to reduce growth and survival of T315I-expressing BV-173R cells are not likely to be clinically achieved by current dasatinib regimens.⁵ In contrast, ON012380 dose-responsiveness was equal in K562, BV-73 and BV-173R cells (IC50~300 nM), whereas K562R cells displayed only limited sensitivity (IC50>5 μM). These results support earlier reports of ON012380 activity in CML cells expressing mutant or imatinib-insensitive BCR-ABL kinase.¹

Figure 2.

Effect of tyrosine kinase inhibitors on survival and signaling in imatinib-sensitive and -resistant CML cells. (a) Four cell lines representing isogenetic variants of imatinib-sensitive and -resistant cells were incubated with the indicated concentration of inhibitor for 72 h before analysis of cell growth and viability. Each data point represents the average of four replicates with no more than 5% variance in any point. The results are representative of three additional independent experiments. (b) BV-173 (left) and BV-173R cells (right) expressing unmutated and T315I-mutant BCR-ABL, respectively, were incubated with imatinib or dasatinib at the indicated concentration for 2 h before lysates were prepared and analyzed for CrkL electrophoretic mobility, Hck/Lyn activation and phosphotyrosine protein levels by immunoblotting. Actin was immunoblotted as a protein-loading control. (c) BV-173 (left) and BV-173R cells (right) were incubated with imatinib or ON012380 at the indicated concentration for 2 h before lysates were prepared and analyzed for phosphotyrosine protein levels and Stat5 activation or protein levels by immunoblotting. Lysates prepared after 24 h of incubation were analyzed for PARP cleavage (bottom) as a measure of apoptosis. (d) K562 and K562R cells were treated with the indicated concentration of inhibitor for 2 h before cell lysates were prepared and analyzed for phosphotyrosine levels by immunoblotting. Lysates prepared after 24 h of incubation were analyzed for PARP cleavage (bottom) as a measure of apoptosis.

To define mediators of biological responsiveness, inhibitor-treated cells were screened for changes in total and target-specific tyrosine phosphorylation. In BV-173 cells (Figure 2b,) imatinib and dasatinib reduced total tyrosine phosphoprotein levels, increased the electrophoretic mobility of CrkL, indicative of its dephosphorylation and reduced activation of kinases in the Src family (Hck, Lyn). In BV-173R cells, imatinib had no effect on total tyrosine phosphorylation levels, CrkL mobility or Src-family kinases. Dasatinib incubation reduced Src family kinase activation but had no effect on CrkL mobility or BCR-ABL tyrosine phosphorylation in BV-173R cells. In contrast, ON012380 had no effect on BCR-ABL, Stat5 or cellular tyrosine phosphorylation levels. However, although caspase activation was associated with kinase inhibition by imatinib in these cells, ON012380 activated caspase cascades in both cell lines in the absence of tyrosine kinase inhibition (Figure 2c).

In K562 cells, both imatinib and dasatinib reduced tyrosine phosphorylation and activated cleavage of PARP (Figure 2d, left). In K562R cells, imatinib had no effect on tyrosine phosphorylation or caspase activation, whereas dasatinib effectively reduced tyrosine phosphorylation and Lyn activation, resulting in caspase activation with PARP cleavage (Figure 2d, right). In agreement with the results in other cell lines, caspase activation was detected in ON012380-treated K562 and K562R cells in the absence of changes in tyrosine phosphorylation. Together, these results suggest that ON012380-induced apoptosis and growth inhibition are not associated with inhibition of BCR-ABL or Lyn kinase.

The T315I BCR-ABL mutation was detected in three CML patients who developed progressive disease on imatinib, and the effect of ex vivo incubation with kinase inhibitors on tyrosine phosphorylation, BCR-ABL signaling (p-CrkL) and apoptosis were examined. Variable levels of basal tyrosine phosphorylation were detected and only dasatinib was active in suppressing phosphorylation of one or more proteins in all cell samples (Figure 3). BCR-ABL and tyrosine phosphorylation were examined 4 h after compound addition and apoptosis was determined by annexin staining after 48 h of incubation. In patient A, dasatinib reduced BCR-ABL phosphorylation and signaling through an unknown mechanism, possibly involving inhibition of Src family kinases^3,6 and other recently described kinase targets.⁷ In patient B, dasatinib had minimal effects on total phosphotyrosyl proteins but did suppress tyrosine phosphorylation of an unknown protein of ~80 kDa (arrow). In patient C, dasatinib reduced phosphorylation of multiple proteins, including those that migrate in the range common for Src-family kinases (~53–62 kDa). Importantly, ON012380 reduced cell survival in all specimens without altering tyrosine or CrkL phosphorylation. Overall, these results suggest that the antileukemic effects of ON012380 were not associated with specific suppression of BCR-ABL or Src-family kinase activity in patients with T315I-mutant BCR-ABL.

Figure 3.

Effect of tyrosine kinase inhibitors on tyrosine phosphorylation and survival of cells derived from imatinib-resistant CML patients expressing T315I-mutant BCR-ABL. Three CML patients (a–c) who progressed on imatinib therapy were found to express only T315I-mutant BCR-ABL. Mononuclear cell preparations from each patient were incubated with imatinib (5 μM), dasatinib (0.5 μM) or ON012380 (5 μM) for 4 h before equal protein lysates were subjected to immunoblotting for phosphotyrosine, pY-CrkL and CrkL. Untreated K562R cell lysate was used as a positive control for specific phosphoproteins. Cells incubated for 48 h with inhibitors were stained for markers of apoptosis and analyzed by flow cytometry. Percent survival is based on relative changes in apoptosis in treated vs vehicle-only (control) treated cells. The values represent the average±s.d. of three replicates.

ON012380 did not reduce phosphoproteins or BCR-ABL substrates previously noted to be responsive to ON012380 incubation (Stat5), supporting a novel mechanism of apoptotic action of ON012380 in CML cells that does not involve kinase inhibition. ON012380 activity in clinical specimens of imatinib-resistant CML was similar to the results obtained with cell lines. Reduced cell survival was consistently measured in ON012380-treated cells and was not associated with changes in tyrosine phosphoproteins expressed at high- or low-basal levels (Figure 3). Imatinib was nearly inactive in reducing either phosphorylation or survival in any specimen, whereas dasatinib did not induce a predictable or uniform response in any of these specimens, although all expressed T315I-mutant BCR-ABL based on sequencing. Distinctions in activity may be related to the multiplicity of targets for dasatinib and the potential cross talk between kinases expressed in CML.^3,7 These observations may also explain the unexpected activity of dasatinib in some clinical studies of imatinib resistance associated with the presence of T315I-BCR-ABL.⁸ A detailed phosphoproteomic analysis in additional clinical cases may be necessary to clarify the activity of dasatinib in T315I imatinib-resistant patients.

In this report, we attempted to confirm the BCR-ABL-inhibitory and anti-CML activity of ON012380. Our results are only in partial agreement with previous observations of ON012380 activity. We chemically synthesized this compound and subsequently compared its activity with ON012380 provided by Dr E Premkumar Reddy (Temple University, Philadelphia, PA, USA) with identical results. The reasons for the disparities are unknown but may involve several factors, including the assays and cell models used to define the specificity and activity of ON012380.¹ We noted apoptotic activity in T315I-positive CML cells and clinical specimens, an activity that did not seem to be related to BCR-ABL kinase inhibition. Although the precise mechanism of ON012380 action remains unresolved, ON012380 did induce apoptosis in CML cells and was able to reduce the number of leukemic cells in animal models without toxicity.¹ These activities alone may support further development and testing of ON012380 in human disease.

Supplementary Material

Supplement Figure Legends

Supplement figure 1. Synthesis of ON012380

Supplement figure 2. Crystal structure of compound 6 (supplement fig. 1). The compound is racemic at C20A position.