The structure of residues 7-16 of the A alpha-chain of human fibrinogen bound to bovine thrombin at 2.3-A resolution.

New insights into the regulation of the blood clotting cascade derived from the X-ray crystal structure of bovine meizothrombin des F1 in complex with PPACK

“…The thrombin domains of mzTBN-F1 molecules I and II have rms overlaps for the Cα atoms [38] of 0.42 Å or less relative to bovine and human thrombin [39,40]. Two chain segments with well defined electron density differ significantly from free thrombin, relative to the estimated 0.4 Å error in the structures [41].…”

“…The second region in the thrombin domain that differs significantly from free thrombin is centered at the Tyr-Pro-Pro-Trp (YPPW loop; residues 367 Tyr60A to 370 Trp60D) and the adjacent Lys97 segment containing residues 412 Trp96 and 413 Lys97. These two chain segments exhibit Cα differences of up to 2.5 Å and 1.8 Å, respectively, relative to thrombin (Figure 7) and form part of the S9 subsite of thrombin [39] (according to the convention of Schechter and Berger for naming protease substrate residues and the corresponding enzyme subsites that bind them [51]). The Lys97 segment in molecule I interacts with its own kringle domain (Table 1) and YPPW loop ( 412 Trp96 contacts 368 Pro60B).…”

“…One of the reasons for the greater substrate specificity of thrombin, as compared to trypsin, is the presence of the YPPW loop inserted near the active site where it sterically restricts access and forms one side of the unique S2 specificity site of thrombin [39,40]. The original impression that the YPPW loop is a rigid part of the active site [34] is contradicted by the drastic difference in Cα positions (2.5 Å for 370 Trp60D) between mzTBN-F1 and thrombin.…”

“…The model was constructed from the structure of mzTBN-F1 molecule I and the structure of calcium-loaded bovine F1 [28] The structure was solved by a combination of molecular replacement with X-PLOR [60][61][62], solvent leveling [63], and graphics with TOM/FRODO [64,65]. A rotation search with thrombin [39] as the model (7-4 Å 4σ data) followed by Patterson correlation refinement (eight steps/peak) of the resulting 331 rotation function peaks produced two unambiguous solutions with correlation values that were the highest (0.053) and fourth highest (0.041) in the list. Both orientations gave unambiguous solutions to the translation function in P4 1 2 1 2 that had correlation coefficients of 0.16 or higher, were at least 10.5σ above the mean, and were 3.8σ or more above the next highest peak.…”

“…Figure 6The γ-autolysis loop. (a) The mzTBN-F1 γautolysis loop is shown in stereo superimposed with the same loop in thrombin[39]. The loop has moved over 6 Å between the two structures.…”