Characterization of the Interaction between Thrombin and a Pentapeptide That Interferes with Anion Binding Exosite I of the Enzyme.

We have recently characterized a pentapeptide, DYDYQ, from coagulation factor V (Beck et. al. 2004J. Biol. Chem.279, 3084) that inhibits both factor V activation and prothrombinase function. The pentapeptide does not interfere with the active site of thrombin but rather interferes with substrate attachment. Our aim was to characterize at the molecular level the binding site of DYDYQ on thrombin. First we used computational methods (blind and focused docking) to propose a hypothetical binding site. Blind docking of the pentapeptide (structure obtained from a 20 ns molecular dynamics simulation) on thrombin was performed using a docking grid with large spacing. This approach provided us a favorable site (−4.8 kcal/mol) that was further investigated using a smaller spaced docking grid. Hydrogen bonding was analyzed between thrombin and DYDYQ. The final free energy of binding was −9.69 kcal/mol. Amino acids Y76R77I79I82 from thrombin anion binding exosite I (ABE-I), were identified to participate in the interaction of the enzyme with DYDYQ. We next investigated the Thrombin-DYDYQ interaction following cross-linking with 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide(EDC) and mass spectrometry. In these experiments purified thrombin was inhibited in the active site with a chloromethyl ketone, and treated with the DYDYQ peptide in the presence of EDC. Two bands were observed, one corresponding to thrombin cross linked to the peptide (CT) and another band corresponding to free thrombin (T). The proteins were either (i) stained with Coomassie blue for further digestion with trypsin or (ii) transferred to nitrocellulose membranes following by staining with Coomassie blue for treatment with cyanogen bromide (CNBr). Stained bands were isolated from the gel and subjected to trypsin digestion and liquid chromatography/mass spectrometry (LC-MS). Following trypsin digestion thrombin presence in both, T and CT samples was confirmed and the peptides identified in both samples covered approximately 63% of the entire thrombin sequence. The only difference observed between the sets of peptides obtained from T and CT following digestion with trypsin, was the peptide N78IEKISM*LEK87 (M* = oxidized Methionine), which was present in the T sample but was absent in the CT sample. These results suggest that the binding site of DYDYQ to thrombin is localized in the area of the above mentioned peptide protecting it from hydrolysis by trypsin. We next analyzed T and CT by LC-MS following CNBr digestion. Three important bands (peptide products from CNBr digestion) were detected in the sample containing the T, having an approximate molecular weight of ~4,500, ~7,500, and ~9,000. CNBr digestion products of CT lacked the median band (peptide mass: ~7,500). This band corresponded to the sequence L33…Y76ERN78IEKISM84 - as confirmed by ESI-ion trap mass spectrometry amino acid sequence of its first eleven amino acid residues. The difference between the two in gel CNBr digest profiles of T and CT, confirms the conclusion drawn from the MS analysis of the triptic digests which in turn was predicted by our computational analyses. Overall our data demonstrate that 1) amino acid residues Y76R77I79I82 from thrombin provide an interactive site for DYDYQ, and 2) our results from computational methods that identify protein-peptide interaction are valid and can be confirmed by mass spectrometry.