Exploring the active site of human factor Xa protein by NMR screening of small molecule probes Lee Fielding,* a Dan Fletcher, a Samantha Rutherford, a Jasmit Kaur † b and Jordi Mestres‡ b a Department of Analytical Chemistry, Organon Laboratories Ltd, Newhouse, Lanarkshire, UK ML1 5SH. E-mail: l.fielding@organon.co.uk; Fax: +44 1698 736187; Tel: +44 1698 736182 b Department of Medicinal Chemistry, Organon Laboratories Ltd, Newhouse, Lanarkshire, UK ML1 5SH Received 26th August 2003, Accepted 13th October 2003 First published as an Advance Article on the web 31st October 2003 A collection of small molecules (MW < 350 Da) was screened for binding to human factor Xa using saturation transfer difference NMR spectroscopy to detect binding. The NMR screening experiments identified four hits. Binding isotherms constructed from NMR linewidth data showed that the binding affinities of the hits were all in the 30–210 μM range. Competition binding experiments showed that three of the ligands were displaced by a known μM inhibitor of factor Xa. The success of the method for identifying new ligands and the relevance of this information to the design of new factor Xa inhibitors are discussed. Introduction Some recent advances in the field of NMR spectroscopy that are of interest to the drug discovery enterprise are techniques that use NMR as a direct tool to detect small molecule– receptor interactions. 1 Technologies for using NMR as a biophysical screening tool are particularly relevant to lead dis- covery because they have the potential to identify interesting low affinity, low molecular weight compounds that may other- wise be overlooked by conventional bioassay based high throughput screening techniques. 2,3 NMR has the additional advantage of providing structural information that can then be exploited to guide the progression of those low affinity compounds to novel potent leads. A methodology that looks to be particularly promising is that of screening modest collections of small molecules for low affinity compounds. 4,5 In this strategy a small but diverse library of low molecular weight soluble compounds is assembled. This library is generally composed of compounds representing fragments commonly found in drug molecules, 6 often combined with structural motifs present in known active molecules for the target of interest or a closely related target of the same family. Then, screening of the library against a protein target by NMR allows for detecting compounds with weak binding (K D ∼ μM– mM). This makes the technology very attractive at the early stages of lead discovery to identify, on the one hand, hit pre- cursors that can be further optimised using structure-guided chemistry and, on the other hand, privileged fragments that can then be incorporated or enriched in the larger compound collections undergoing high-throughput screening. Cardiovascular disease (myocardial infarction, stroke, deep vein thrombosis and pulmonary embolism etc.) is a major cause of mortality in the western world. The immediate cause of these disease conditions is an occlusive blood clot and hence anti- thrombotic therapy is a crucial component for the treatment and management of these diseases. The blood clotting process is a natural response to stop the loss of blood in higher organ- isms following vascular injury. Initially, platelets are activated and aggregate at the site of injury, then in a sequence of events † Present address: OSI Pharmaceuticals, Watlington Road, Oxford, UK OX4 6LT. ‡ Present address: Universitat Pompeu Fabra, Passeig Marítim de la Barceloneta 37-49, 08003 Barcelona, Spain. termed the coagulation cascade, the blood thickens and fibrin formation leads to an insoluble clot. Therapy usually consists of a combination of anticoagulant and antiplatelet agents, and hence drug discovery activities have focused on finding orally active inhibitors of proteins in the coagulation cascade. 7 Of all the proteins involved in the coagulation cascade, factor Xa is a prime target in cardiovascular drug discovery. Factor Xa is a trypsin like serine protease that converts prothrombin to catalytically active thrombin – the protein at the centre of events in the coagulation process, and is therefore key to several of the processes necessary to form stable clots. As such, factor Xa is an attractive target for new anticoagulant agents. 8,9 Factor Xa (51 kDa) consists of a heavy chain (251 amino acids) and a light chain (139 amino acids), held together by a single disulfide bond. The heavy chain incorporates the catalytic triad composed of His-57, Asp-102 and Ser-195. The light chain contains two epidermal growth factor like domains and the chymotrypsin cleavage site (Tyr-44, Lys-45). The crystal structure of factor Xa is known, 10 and several crystal structures of factor Xa in complex with a variety of ligands have been deposited in the Protein Data Bank. 11 Four binding pockets have been identified within the active site, labelled S1 to S4, and the S1 and S4 binding pockets are the important ones that are exploited by factor Xa inhibitors. The S1 pocket is a narrow cleft defined by Asp-189, Ala-190 and Gln-192 and favours positively charged moieties such as amine, guanidine and benzamidine. A hydrogen bond formed between the aspartate residue at the bottom of the cleft (Asp-189) and benzamidine based inhibitors is a very important piece of the SAR, but is not an absolute prerequisite for high affinity. 12 The S4 pocket is a shallow groove formed by aromatic residues Tyr- 99, Phe-174 and Trp-215 and is sufficiently rich in π electrons that it is not only a hydrophobic pocket, but also forms a cation recognition site, thus favouring both hydrophobic and basic moieties in ligands. 13 The present study reports the results obtained from a struc- ture-based NMR screening approach to factor Xa lead gener- ation. The small molecule screening collection was specifically designed to include both diverse fragments from known drugs and targeted fragments from known factor Xa inhibitors. The study was also conceived to test entirely NMR based method- ologies from the initial screen to follow-up experiments desig- ned to measure K D and to validate the specificity of the hits. An overview of the strategy is shown in Scheme 1. DOI: 10.1039/ b310265c 4235 This journal is © The Royal Society of Chemistry 2003 Org. Biomol. Chem. , 2003, 1, 4235–4241