In a systematic effort to design inhibitors of the
epidermal growth factor receptor (EGFR) family
protein tyrosine kinases (PTK) as anti-
cancer agents, we have constructed a three-dimensional homology model of the EGFR
kinase domain and used molecular modeling methods for the structure-based design of analogs of the active metabolite of
leflunomide (LFM) with potent and specific inhibitory activity against EGFR. These docking studies identified alpha-cyano-beta-hydroxy-beta-methyl-N-[4-(trifluoromethoxy)
phenyl]-p ropenamide (LFM-A12) as our lead compound, which was predicted to bind to the EGFR catalytic site in a planar conformation.
LFM-A12 inhibited the proliferation (IC50 = 26.3 microM) and in vitro invasiveness (IC50 = 28.4 microM) of EGFR positive human
breast cancer cells in a concentration-dependent fashion. Similarly, the model of the EGFR binding pocket was used in combination with docking procedures to predict the favorable placement of chemical groups with defined sizes at multiple modification sites on another class of EGFR inhibitors, the
4-anilinoquinazoline. This approach has led to the successful design of a dibromo
quinazoline derivative,
WHI-P97, which had an estimated Ki value of 0.09 microM from modeling studies and a measured IC50 value of 2.5 microM in EGFR
kinase inhibition assays.
WHI-P97 effectively inhibited the in vitro invasiveness of EGFR-positive human
cancer cells in a concentration-dependent manner. However, unlike
LFM-A12, the
quinazoline compounds are not specific for EGFR.