Abstract
A novel dual SRC/ABL kinase inhibitor, BMS-354825, is showing promise for the treatment of imatinib-resistant chronic myeloid leukemia not only in vitro (Shah NP, et al., Science 305:399), but also in a phase I clinical trial (ASH abstract: Sawyers CL, et al.) Resistance to imatinib is increasingly found in patients due to point mutations in the BCR-ABL kinase domain that do not impair kinase activity but prevent drug binding. BMS-354825 is more potent than imatinib and retains activity against 14 of 15 imatinib-resistant BCR-ABL mutants in vitro. The compound’s ability to inhibit imatinib-resistant forms of BCR-ABL is presumed to be due to its relaxed binding requirements, whereas imatinib requires the adoption of a closed conformation of the kinase to bind. We addressed the hypothesis that the relaxed binding requirements of BMS-354825 would limit the range of BCR-ABL mutations that confer drug resistance. To address this question, we employed a saturation mutagenesis experiment as described by others (Azam M, et al., Cell 112:831) and found that the spectrum of BMS-354825-resistant mutants was reduced compared to that of imatinib. In a series of such screens, mutations at only four amino acids have been isolated, two of which account for the vast majority of resistant clones. In contrast, Azam et al. isolated over 20 mutations in a screen for imatinib resistance, a finding which has been generally reproduced in our lab. All four BMS-354825-resistant mutations map to known BMS-354825 contact residues as shown by co-crystallographic studies (ASH abstract: Tokarski JS et al., Bristol-Myers Squibb). Mutations at L248, T315, and F317 show BMS-354825 resistance and have been previously reported to confer imatinib resistance. Mutation at V299 represents a novel mode of resistance. Interestingly, some point mutations conferring BMS-354825 resistance were at positions known to be mutated in cases of imatinib resistance, but the mutated residues differed. Furthermore, the identity of the mutated residue was crucial in conferring sensitivity or resistance to an individual drug as shown by comparison of cellular IC50’s (see table). For example, F317L was shown previously to confer imatinib resistance. F317V, on the other hand, demonstrates relative BMS-354825-resistance but is still exquisitely sensitive to imatinib. In a screen for mutants simultaneously resistant to both drugs, we consistently recover 30–50 fold fewer mutant clones compared to single drug treatment. All such clones isolated to date encode for T315I. Kinase domain point mutation is becoming an increasingly encountered clinical problem in diseases treated with small molecule inhibitors. Our findings suggest that combination therapy with imatinib and BMS-354825 may be of clinical utility in CML, particularly by delaying the development of resistance.
IC50 for growth (nM)
Baf3 Clone . | imatinib . | BMS-354825 . |
---|---|---|
p210 wt | < 1,000 | < 5 |
T315I | > 10,000 | > 500 |
T315A | 1,000 | 100 |
F317L | 2,000 | 10 |
F317V | < 1,000 | 60 |
V299L | 1,000 | 20 |
L248R | > 10,000 | 20 |
Baf3 Clone . | imatinib . | BMS-354825 . |
---|---|---|
p210 wt | < 1,000 | < 5 |
T315I | > 10,000 | > 500 |
T315A | 1,000 | 100 |
F317L | 2,000 | 10 |
F317V | < 1,000 | 60 |
V299L | 1,000 | 20 |
L248R | > 10,000 | 20 |
Author notes
Corresponding author