The cancer drug, Imatinib, is a selective Abl kinase inhibitor that does not
inhibit the closely related kinase c-Src. This one drug and its ability to
selectively inhibit Abl over c-Src has been a guiding principle in virtually all
kinase drug discovery efforts in the last 15 years. A prominent hypothesis
explaining the selectivity of Imatinib is that Abl has an intrinsic ability to
adopt an inactive conformation (termed DFG-out), whereas c-Src appears to pay a
high intrinsic energetic penalty for adopting this conformation, effectively
excluding Imatinib from its ATP pocket. This explanation of the difference in
binding affinity of Imatinib for Abl versus c-Src makes the striking prediction
that it would not be possible to design an inhibitor that binds to the DFG-out
conformation of c-Src with high affinity. We report the discovery of a series of
such inhibitors. We use structure-activity relationships and X-ray
crystallography to confirm our findings. These studies suggest that small
molecules are capable of inducing the generally unfavorable DFG-out conformation
in c-Src. Structural comparison between c-Src in complex with these inhibitors
allows us to speculate on the differential selectivity of Imatinib for c-Src and
Abl.
