BCR-ABL Kinase Dynamics and Drug Resistance.

The aberrant signaling behavior caused by the expression of BCR-ABL is necessary and sufficient to cause chronic myeloid leukemia (CML), an observation which paved the way for the development of imatinib (GleevecTM), a small molecule inhibitor of the BCR-ABL kinase. Enthusiasm for the remarkable efficacy of imatinib has been tempered by the development of clinical resistance. The most common mechanisms for resistance are the development of kinase domain mutations and/or overexpression of the BCR-ABL gene, with mutations in the kinase accounting for ~90 % of all cases. The resistance-conferring lesions are found in regions of the kinase that are critical to its autoregulation, such as P-loop, C-helix, gatekeeper area, activation loop and the SH2-C-lobe interface. Mechanistically, these mutations effect either a steric blockade or a change in the dynamic equilibrium that favors the active kinase conformation that precludes imatinib binding. We have analyzed two dual Src-Abl kinase inhibitors, AP23464 and PD166326, against 58 BCR-ABL kinase variants conferring imatinib resistance. PD166326 binds to the Abl kinase domain in the open although enzymatically inactive conformation, while AP23464 targets the active conformation. Both of these compounds have effectively suppressed the cell growth of imatinib resistance variants, except for a recurrent mutation in the gatekeeper residue (T315I). The P-loop variants are more sensitive to AP23464 than PD166326. Interestingly, the imatinib resistant variants from the C-helix, hinge region, activation loop and SH2-C-lobe region, are hypersensitive to both compounds, as compared to native BCR-ABL. The BCR-ABL variants in the C-helix, gatekeeper area, and the activation loop are more sensitive to AP23464 than PD166326, while variants from the hinge region and the SH2-C-lobe interface are hypersensitive to PD166326. Altogether, these results define a differential requirement for a specific ABL conformation for drug binding of AP23464 and PD166326. In order to better understand their structure activity relationships and the patterns of resistance, we carried out an in-vitro mutagenesis-screen using different concentration of the drug either alone or in combination with imatinib. AP23464 mediates 2–3 time less resistance than PD166326. A higher concentration of all three compounds suppresses all resistance mutations, save for the notable exceptions, T315I and F317L/VandC. Resistance conferring mutations selected at 10–20 fold higher IC50 values are different. AP23464 efficiently suppresses the mutations from the P-loop (except E255K) and two mutations from the activation loop, while PD166326 remains refractory to the mutations in the C-helix and SH2-C-lobe interface. In combination with imatinib, AP23464 and PD166326 suppressed the emergence of most resistance mutations, with the notable exception of T315I. These in-vitro studies demonstrate that the combination of two or three different conformation specific inhibitors is needed to suppress the emergence of resistance. We are characterizing variants of AP23464 that we predict will show activity against the most challenging imatinib resistance mutant T315I.