Figure 3
Figure 3. Protease inhibition mechanism of ZPI mediated by PZ. (A) PZ/ZPI complex overlaid with antithrombin/FXa (S195A) Michaelis complex (PDB 2GD4) using the serpin fold. For clarity, the reactive center loop of ZPI (green) was removed. The reactive center loop of antithrombin is shown in black. The 3 acidic residues (E313, E231, and D233) near the FXa docking site are shown in sticks (blue), and the autolysis loop (143-154) of FXa is in red. (B) The electrostatic surface on the top of of ZPI shows a negatively charged surface on the left (circled), formed by D233, E231, and E313, and a positive patch on the right (in rectangle), formed by residues K253, K260, H265, K308, and R310 of ZPI. The autolysis loop of FXa or FXIa most likely binds to the negative patch (left), whereas the 36 loop will interact with the positive charges. The reactive center loop is shown in lines. Although it has been indicated previously that substitution of glutamates of the 36 loop with glutamines has little effect on ZPI/FXa interaction, we speculate these mutations will only weaken, but not abolish these exosite interactions. Because the reactive loop of ZPI is 4 residues shorter than that of antithrombin, it is expected that FXa will dock onto ZPI in a slightly different orientation to that in antithrombin-FXa complex, which would allow potential interactions between the 36 loop of FXa and the positive surface on the top of ZPI. (C) Model of the ternary complex of PZ/ZPI/FXa on the phospholipid membrane surface. The relative positions of Gla, EGF1, EGF2, and protease domain of FXa were modeled using the coordinates of full-length FIXa (PDB 1PFX).41 Full-length PZ was modeled similarly, but the Gla domain (brown) was rotated approximately 180° along the flexible linker between EGF1 and Gla domains, which allows its interactions (indicated by arrows) with the Gla domain of FXa on the membrane surface, as suggested.11,15 ZPI is colored in green with reactive loop in red and A-β-sheet in light blue. The N-terminal domains (Gla, EGF1, and EGF2) are in brown in PZ and in pink in FXa. Ca2+ is shown as blue dots. The head (phosphate) groups of phospholipids are shown as red dots, and the hydrophobic tails as black dashes. The disulfide bonds are shown in yellow sticks. In the ternary complex model, the EGF2 domain of PZ is in close proximity to the protease domain of FXa, but it is unclear whether they can form further stabilizing interactions within the complex.

Protease inhibition mechanism of ZPI mediated by PZ. (A) PZ/ZPI complex overlaid with antithrombin/FXa (S195A) Michaelis complex (PDB 2GD4) using the serpin fold. For clarity, the reactive center loop of ZPI (green) was removed. The reactive center loop of antithrombin is shown in black. The 3 acidic residues (E313, E231, and D233) near the FXa docking site are shown in sticks (blue), and the autolysis loop (143-154) of FXa is in red. (B) The electrostatic surface on the top of of ZPI shows a negatively charged surface on the left (circled), formed by D233, E231, and E313, and a positive patch on the right (in rectangle), formed by residues K253, K260, H265, K308, and R310 of ZPI. The autolysis loop of FXa or FXIa most likely binds to the negative patch (left), whereas the 36 loop will interact with the positive charges. The reactive center loop is shown in lines. Although it has been indicated previously that substitution of glutamates of the 36 loop with glutamines has little effect on ZPI/FXa interaction, we speculate these mutations will only weaken, but not abolish these exosite interactions. Because the reactive loop of ZPI is 4 residues shorter than that of antithrombin, it is expected that FXa will dock onto ZPI in a slightly different orientation to that in antithrombin-FXa complex, which would allow potential interactions between the 36 loop of FXa and the positive surface on the top of ZPI. (C) Model of the ternary complex of PZ/ZPI/FXa on the phospholipid membrane surface. The relative positions of Gla, EGF1, EGF2, and protease domain of FXa were modeled using the coordinates of full-length FIXa (PDB 1PFX).41  Full-length PZ was modeled similarly, but the Gla domain (brown) was rotated approximately 180° along the flexible linker between EGF1 and Gla domains, which allows its interactions (indicated by arrows) with the Gla domain of FXa on the membrane surface, as suggested.11,15  ZPI is colored in green with reactive loop in red and A-β-sheet in light blue. The N-terminal domains (Gla, EGF1, and EGF2) are in brown in PZ and in pink in FXa. Ca2+ is shown as blue dots. The head (phosphate) groups of phospholipids are shown as red dots, and the hydrophobic tails as black dashes. The disulfide bonds are shown in yellow sticks. In the ternary complex model, the EGF2 domain of PZ is in close proximity to the protease domain of FXa, but it is unclear whether they can form further stabilizing interactions within the complex.

Close Modal

or Create an Account

Close Modal
Close Modal