The Asp-Phe-Gly (DFG) motif plays an important role in the regulation of kinase activity. Structure-based drug design was performed to design compounds able to interact with the DFG motif; epidermal growth factor receptor (EGFR) was selected as an example. Structural insights obtained from the EGFR/2a complex suggested that an extension from the meta-position on the phenyl group (ring-5) would improve interactions with the DFG motif. Indeed, introduction of an N,N-dimethylamino tail resulted in 4b, which showed almost 50-fold improvement in inhibition compared to 2a. Structural studies confirmed this N,N-dimethylamino tail moved toward the DFG motif to form a salt bridge with the side chain of Asp831. That the interactions with the DFG motif greatly contribute to the potency of 4b is strongly evidenced by synthesizing and testing compounds 2a, 3g, and 4f: when the charge interactions are absent, the inhibitory activity decreased significantly.

HTS hit 7 was modified through hybrid design strategy to introduce a chiral side chain followed by introduction of Michael acceptor group to obtain potent EGFR kinase inhibitors 11 and 19. Both 11 and 19 showed over 3 orders of magnitude enhanced HCC827 antiproliferative activity compared to HTS hit 7 and also inhibited gefitinib-resistant double mutant (DM, T790M/L858R) EGFR kinase at nanomolar concentration. Moreover, treatment with 19 shrinked tumor in nude mice xenograft model.

The inhibition of FMS-like tyrosine kinase 3 (FLT3) activity using small-molecule inhibitors has emerged as a target-based alternative to traditional chemotherapy for the treatment of acute myeloid leukemia (AML). In this study, we report the use of structure-based virtual screening (SBVS), a computer-aided drug design technique for the identification of new chemotypes for FLT3 inhibition. For this purpose, homology modeling (HM) of the DFG-in FLT3 structure was carried using two template structures, including PDB ID: 1RJB (DFG-out FLT3 kinase domain) and PDB ID: 3LCD (DFG-in CSF-1 kinase domain). The modeled structure was able to correctly identify known DFG-in (SU11248, CEP-701, and PKC-412) and DFG-out (sorafenib, ABT-869 and AC220) FLT3 inhibitors, in docking studies. The modeled structure was then used to carry out SBVS of an HTS library of 125,000 compounds. The top scoring 97 compounds were tested for FLT3 kinase inhibition, and two hits (BPR056, IC50 = 2.3 and BPR080, IC50 = 10.7 μM) were identified. Molecular dynamics simulation and density functional theory calculation suggest that BPR056 (MW: 325.32; cLogP: 2.48) interacted with FLT3 in a stable manner and could be chemically optimized to realize a drug-like lead in the future.

Aurora kinases have emerged as attractive targets for the design of anticancer drugs. Through structure-based virtual screening, novel pyrazole hit 8a was identified as Aurora kinase A inhibitor (IC(50) = 15.1 microM). X-ray cocrystal structure of 8a in complex with Aurora A protein revealed the C-4 position ethyl carboxylate side chain as a possible modification site for improving the potency. On the basis of this insight, bioisosteric replacement of the ester with amide linkage and changing the ethyl substituent to hydrophobic 3-acetamidophenyl ring led to the identification of 12w with a approximately 450-fold improved Aurora kinase A inhibition potency (IC(50) = 33 nM), compared to 8a. Compound 12w showed selective inhibition of Aurora A kinase over Aurora B/C, which might be due to the presence of a unique H-bond interaction between the 3-acetamido group and the Aurora A nonconserved Thr217 residue, which in Aurora B/C is Glu and found to sterically clash with the 3-acetamido group in modeling studies.

No abstract