Reengineering Ponatinib to Minimize Cardiovascular Toxicity

Figure 1.. Development and characterization of the ponatinib analogues. A, Chemical structures and molecular docking model of ponatinib and nilotinib to T315I mutant BCR-ABL. B, Chemical structures of the hybrid compounds 33a and 36a and molecular docking models showing entry into the binding pocket of T315I mutant BCR-ABL relative to that of ponatinib. C,Kd values of the compounds determined for WT and T315I mutant ABL. D, Schematic of the in vitro bioassays used to characterize the activities of the compounds.

Figure 2.. CML in vitro antiproliferative activity. Proliferation of CML (K562) cells carrying either WT (A) or CRISPR/Cas9 engineered T315I mutant (B) BCR-ABL oncogenes. Data points represent three biological replicates, with four technical replicates per dose in each biological replicate. GI50 values are indicated in Table 1.

Figure 3.. Effects of 33a and 36a on endothelial cell vasculogenesis. A, HMVEC-Cs were treated with compound for 6 hours and the extent of vascular network formation was quantified. B, Vascular network loop formation as a function of compound dose. Data points represent averages of three batches from different human donors, three replicates per batch. EC50 values are indicated in Table 1. C, Representative images of vascular networks at the indicated doses. Scale bar, 200 μm.

Figure 4.. Effects of 36a and 33a on human cardiomyocyte contractility. A, hiPSC-CMs were treated with compounds for 4 days and quantitative effects on contractility were measured by optical recording (100 frames/second) and analysis of motion. B and C, Peak contraction amplitude in response to compound treatment normalized to baseline in cardiomyocytes derived from two healthy donor hiPSC lines. Data points represent averages of three differentiation batches, four replicates per batch. EC50 values are indicated in Table 1. D, Heatmap representations of the peak contraction amplitude in the field of view at the indicated doses. E, Troponin-T levels in the media of treated cells analyzed by ELISA. Data points represent biological replicates (n = 5), each of which contained three technical replicates. **, P < 0.01.

Figure 5.. Kinase inhibition profile of 36a and 33a relative to ponatinib. A, Dendrogram representation of the kinase inhibition profiles of 33a, 36a, and ponatinib. B, Analysis of kinases nonselectively and selectively inhibited by 33a and 36a (relative to ponatinib). Note that all three compounds comparably inhibited mutant ABL kinases. However, a subset of kinases was selectively inhibited by ponatinib. See Supplementary Table S1. C, Physiologic effect on human cardiomyocyte contractility (peak contraction amplitude) by individual siRNA knockdown of the kinases that were selectively inhibited by ponatinib, shown relative to the effect of siRNA silencing of ABL. Silencing of most of these kinases degraded contractility, suggesting their potential involvement as mediators of ponatinib cardiotoxicity. D and E, Effect of siRNA knockdown of the significantly affected kinases in C on K562-WT and K562-T315I cell survival. Other than for ABL, knockdowns did not affect cell survival, suggesting that these kinases are not individually essential for tumor growth. *, P < 0.05; **, P < 0.01; ****, P < 0.0001. Ctrl, control.

Figure 6.. In vivo assessment of 33a and 36a. A and B, The PK study design (A) and table of resulting parameters (B). C and D, Determination of the maximum tolerated dose (MTD; C) and Kaplan–Meier survival curve (D). E, Xenograft study design. T315I mutant BCR-ABL K562 cells were implanted in Matrigel plugs and allowed to establish solid tumors for 4 days prior to compound (30 mg/kg) administration by daily oral gavage (n = 9 of each group). F–H, Tumor size over time (F), explanted tumors (G), and tumor weights (H) at termination (3 weeks). *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001. I, Body weight of the mice during the study. J, Serum Troponin-I levels at termination.