letters nature structural biology • volume 8 number 1 • january 2001 37 Mechanism-based design of a protein kinase inhibitor Keykavous Parang 1,2 , Jeff rey H. Till 2,3 , Ararat J. Ablooglu 2,4 , Ronald A. Kohanski 4 , St evan R. Hubbard 3 and Philip A. Cole 1 1 The Johns Hopkins University School of Medicine, Department of Pharmacology and Molecular Sciences, Baltimore, Maryland 21205, USA. 2 These authors contributed equally. 3 Skirball Institute of Biomolecular Medicine and Department of Pharmacology, New York University School of Medicine, New York, New York 10016, USA. 4 Mount Sinai School of Medicine, Department of Biochemistry and Molecular Biology, New York, New York 10029, USA. Protein kinase inhibitors have applications as anticancer therapeutic agents and biological tools in cell signaling. Based on a phosphoryl transfer mechanism involving a dis- sociative transition state, a potent and selective bisubstrate inhibitor for the insulin receptor tyrosine kinase was syn- thesized by linking ATPγS to a peptide substrate analog via a two-carbon spacer. The compound was a high affinity com- petitive inhibitor against both nucleotide and peptide sub- strates and showed a slow off-rate. A crystal structure of this inhibitor bound to the tyrosine kinase domain of the insulin receptor confirmed the key design features inspired by a dissociative transition state, and revealed that the link- er takes part in the octahedral coordination of an active site Mg 2+ . These studies suggest a general strategy for the devel- opment of selective protein kinase inhibitors. Protein kinases play a critical role in cell signaling pathways by catalyzing the transfer of the γ-phosphoryl group from ATP to the hydroxyl groups of protein side chains 1 . The protein kinase superfamily is one of the largest as ∼2% of eukaryotic genes encode them. Because of their importance in contribut- ing to a variety of pathophysiologic states, including cancer, inflammatory conditions, autoimmune disorders, and cardiac diseases, intense efforts have been made to develop specific protein kinase inhibitors as biological tools and as therapeutic agents 2 . Considerable success has been achieved in developing potent and selective nucleotide analog-based inhibitors that interact with individual protein kinases at their nucleotide binding sites, and several compounds are in early phases of human clinical trials 2 . In general, the protein substrate binding site has not been exploited for inhibitor design 3 . Moreover, mechanism-based approaches to generating protein kinase inhibitors have been unsuccessful 4–8 . This stands in contrast to the situation with many other important enzyme classes, such as the proteases, where consideration of enzyme mechanism and structure has led to potent inhibitors, some of which are clinically useful drugs 9 . Protein kinases follow ternary complex kinetic mechanisms in which direct transfer of the phosphoryl group from ATP to protein substrate occurs 10 . For such mechanisms, designing covalently linked bisubstrate analogs can be a powerful approach for making potent enzyme inhibitors 9 . However, pre- vious attempts to employ this strategy with protein kinases have met with mixed results 4,8,11 . A sophisticated effort reported by Gibson and colleagues 8 linked ATP directly to the Ser oxygen of a protein kinase A (PKA) peptide substrate (kemptide) to generate an inhibitor (Fig. 1a, compound 1). Compound 1 was a weak inhibitor with an IC 50 of 226 μM (compared to the K m (ATP) of 10 μM and the K m (kemptide) of 15 μM) that was competitive versus ATP but noncompetitive versus peptide sub- strate. While not providing all the desired features of a bisub- strate analog, these results suggest that improvements in geometry and electronic character around the atoms equivalent to the entering nucleophile and reacting phosphate would ben- efit inhibitor design. Compound 1 has a Ser Oγ–P bond distance (1.7 Å) more compatible with a fully associative reaction mechanism for phosphoryl transfer by a protein serine kinase (Fig. 1b). In such a transition state, a bond is largely formed between the attack- ing oxygen and the reactive γ-phosphorus atom, while the bond to the departing ADP would not yet have been significantly broken (Fig. 1b). In contrast, studies of two protein tyrosine kinases in which the nucleophilicity of the attacking hydroxyl from the Tyr residue was varied have provided strong evidence for the occurrence of a dissociative or metaphosphate-like tran- sition state 6,12 . A dissociative transition state for phosphoryl transfer catalyzed by a protein kinase, analogous to an S N 1 reac- tion in organic chemistry, is one in which the importance of the nucleophilicity of the attacking hydroxyl is diminished and departure of the leaving group (ADP) is well advanced (Fig. 1b). Dissociative transition states are well established for nonenzymatic phosphate monoester phosphoryl transfer reac- tions but have been considered more controversial for the cor- responding enzyme catalyzed reactions 13,14 . A prediction for such a fully dissociative transition state is that the ‘reaction coordinate distance’ between the entering nucleophilic oxygen and the attacked phosphorus should be ≥4.9 Å. This is based on the assumptions that the γ-phosphoryl group moves toward the entering oxygen and that this nucleophile and the ADP are fixed, probably as they appear in the ground state ternary com- plex that precedes the transition state 14 . For a transition state with greater associative character, the reaction coordinate dis- tance could be ≤3.3 Å, indicating compression 14 . High resolu- tion X-ray and NMR structures of various protein kinase complexes with nucleotide and peptide substrate analogs have yielded conflicting values for such distances 14–16 , although these studies have, by necessity, employed unreactive substrate analogs that could have affected the results. In reconsidering the design of a bisubstrate analog based on the above parameters of (largely) dissociative reaction mecha- nisms for protein kinases, we hypothesized that a peptide–ATP bisubstrate analog (Fig. 1a, compound 2) in which the distance between the tyrosine nucleophilic atom and the γ-phosphorus was set to ∼5 Å (5.66 Å in an extended form; Fig. 1a) by a short linker might show more potent inhibition toward a protein tyrosine kinase than that of compound 1 toward PKA. A second design feature was based on the fact that proton removal from the hydroxyl of the Tyr residue occurs late in the dissociative reaction mechanism 6,12 , and that the phenolic hydroxyl serves as a hydrogen bond donor to the conserved catalytic Asp residue (Asp 1,132 of the insulin receptor). To exploit this interaction, the Tyr oxygen was replaced with a nitrogen atom that could serve as a hydrogen bond donor while simultaneous- ly being incorporated into a tether. Our choice of target enzyme for the inhibitor was the insulin receptor protein tyrosine kinase (IRK) because: (i) efficient peptide substrates (including IRS727) for this enzyme have been well characterized kinetical- ly 6 ; (ii) solution studies have provided direct evidence for a dis- © 2001 Nature Publishing Group http://structbio.nature.com © 2001 Nature Publishing Group http://structbio.nature.com