Checkpoint kinases CHK1 and CHK2 are activated in response to DNA damage that results in cell cycle arrest, allowing sufficient time for DNA repair. Agents that lead to abrogation of such checkpoints have potential to increase the efficacy of such compounds as chemo- and radiotherapies. Thiophenecarboxamide ureas (TCUs) were identified as inhibitors of CHK1 by high throughput screening. A structure-based approach is described using crystal structures of JNK1 and CHK1 in complex with 1 and 2 and of the CHK1-3b complex. The ribose binding pocket of CHK1 was targeted to generate inhibitors with excellent cellular potency and selectivity over CDK1and IKKβ, key features lacking from the initial compounds. Optimization of 3b resulted in the identification of a regioisomeric 3-TCU lead 12a. Optimization of 12a led to the discovery of the clinical candidate 4 (AZD7762), which strongly potentiates the efficacy of a variety of DNA-damaging agents in preclinical models.

A novel series of 5-aminopyrimidinyl quinazolines has been developed from anilino-quinazoline 1, which was identified in a high throughput screen for Aurora A. Introduction of the pyrimidine ring and optimisation of the substituents both on this ring and at the C7 position of the quinazoline led to the discovery of compounds that are highly specific Aurora kinase inhibitors. Co-crystallisation of one of these inhibitors with a fragment of Aurora A shows the importance of the benzamido group in achieving selectivity.

Modification of imidazo[1,2-a]pyridine CDK inhibitors lead to identification of less lipophilic imidazo[1,2-b]pyridazine series of CDK inhibitors. Although several equivalent compounds from these two series have similar structure and show similar CDK activity, the SAR of the two series differs significantly. Protein inhibitor structure determination has confirmed differences in binding mode and given some understanding of these differences in SAR. Potent and selective imidazo[1,2-b]pyridazine inhibitors of CDK2 have been identified, which show >1 muM plasma levels following a 2 mg/kg oral dose to mice. (C) 2004 Elsevier Ltd. All rights reserved.

Matrix metalloproteinases (MMPs) and their inhibitors are important in connective tissue re-modelling in diseases of the cardiovascular system, such as atherosclerosis. Various members of the MMP family have been shown to be expressed in atherosclerotic lesions, but MMP9 is consistently seen in inflammatory atherosclerotic lesions. MMP9 over-expression is implicated in the vascular re-modelling events preceding plaque rupture (the most common cause of acute myocardial infarction). Reduced MMP9 activity, either by genetic manipulation or through pharmacological intervention, has an impact on ventricular re-modelling following infarction. MMP9 activity may therefore represent a key mechanism in the pathogenesis of heart failure. We have determined the crystal structure, at 2.3 A resolution, of the catalytic domain of human MMP9 bound to a peptidic reverse hydroxamate inhibitor as well as the complex of the same inhibitor bound to an active-site mutant (E402Q) at 2.1 A resolution. MMP9 adopts the typical MMP fold. The catalytic centre is composed of the active-site zinc ion, co-ordinated by three histidine residues (401, 405 and 411) and the essential glutamic acid residue (402). The main differences between the catalytic domains of various MMPs occur in the S1' subsite or selectivity pocket. The S1' specificity site in MMP9 is perhaps best described as a tunnel leading toward solvent, as in MMP2 and MMP13, as opposed to the smaller pocket found in fibroblast collagenase and matrilysin. The present structure enables us to aid the design of potent and specific inhibitors for this important cardiovascular disease target.