Novel Allosteric Inhibitors of Thrombin

Thrombin is a critical enzyme involved in blood coagulation and haemostasis. For this reason the study of its interactions with substrates, inhibitors and modulator is essential. It is also a unique enzyme in the serine protease family because unlike enzymes like trypsin and chymotrypsin its activity is modulated by various endogenous and exogenous ligands. This is due to the presence of “exosites” on the thrombin surface. Exosite II, unlike exosite-I, has not been characterized for its allosteric effect. In order to understand the structural basis of interaction and inhibition of inhibitor 4AS, which possibly interacts with exosite-II, native bovine thrombin crystals and human thrombin crystals grown in presence of 4-AS were prepared. X-ray diffraction data was collected on 4AS soaks of native bovine thrombin as well as human thrombin crystals. The data were phased by molecular replacement using appropriate search models. The structures were refined to R factors of 0.24 and 0.27 for native bovine thrombin-4AS soaks and human thrombin-4AS co-crystals respectively. Examination of a 2Fo-Fc electron density map revealed no density for 4-AS. Low affinity of the inhibitor may be the reason for its absence in the solved structures. In the process of solving these structures, unliganded native bovine thrombin in a new crystal form, previously unreported in literature, was solved. The structure shows an overall topology similar to that found in previously published thrombin molecules. Examination of the crystal packing shows that the exosite-II is solvent exposed. This crystal form can be used in the future to study interaction of exosite-II ligands. To characterize the interaction of sucrose octasulfate with thrombin, which may interact with thrombin exosite-II, fluorimetric equilibrium binding titrations were performed using the active site fluorescent probe para-amino benzamidine. At physiological salt concentrations, the KD was found to be ~22 μM, which is lower than heparin fragment of corresponding length. The higher affinity was attributed to the high charge density of the ligand. Measurement of KD at different salt concentrations showed a significant amount of contribution to the binding energy from ionic interactions. Based on the salt dependence experiments, the number of charged interactions per sucrose octasulfate molecule interacting with thrombin was found to be 3.5. Competitive experiments of sucrose octasulfate with FDs (a sulfated dehydro-polymer being investigated in the lab for its anticoagulant properties) for inhibition of thrombin activity showed competitive effects that did not appear to follow Dixon-Webb competitive phenomenon. It was found that sucrose octasulfate itself is a weak inhibitor of thrombin. To investigate the mode of interaction, co-crystals of sucrose-ocasulfate complexed with thrombin were prepared. High resolution data (2.2 Å) was collected. The structure solved using this data showed weak density for two sucrose octasulfate molecules. Sucrose octasulfate was modeled into this density and refined. The refined structure shows that two sucrose octasulfate molecules bind to two thrombin monomers of the asymmetric unit at exosite-II. One of the sucrose octasulfate molecules interacts with both monomers, and could be present as an artifact of crystal packing. The second molecule interacts with exosite-II of only one of the thrombin monomers. The key residues involved in the interaction are Lys236, His91, Arg93 and Arg101. The thrombin-sucrose octasulfate structure does not show any major deviation from unliganded structure. It is possible that the conformational change may have been masked due to crystal packing. Characterization of this novel interaction mode of sucrose octasulfate interaction with thrombin adds one more candidate to the list of compounds that interact with exosite-II in a manner very similar to heparin, but unlike heparin can inhibit thrombin activity