Essential Thrombin Residues for Inhibition by Plasminogen Activator Inhibitor-1 in the Absence and Presence of Heparin and Vitronectin.

The serine protease inhibitor (serpin), plasminogen activator inhibitor-1 (PAI-1) rapidly inactivates tissue plasminogen activator (tPA) and urokinase plasminogen activator to prevent plasminogen activation to plasmin. Although PAI-1 is a major regulator of fibrinolysis, PAI-1 also has a role in regulating coagulation due to its ability to inhibit thrombin. Previous studies have shown that replacing the 39-loop of thrombin with the 39-loop of tPA increases the rate of thrombin inhibition by PAI-1 suggesting that the 39-loop of thrombin is responsible for the relatively slow rate of inhibition by PAI-1 compared to tPA. Nevertheless, the role of other thrombin residues in the thrombin-PAI interaction [in the absence and presence of heparin and vitronectin (VN)] has not been fully investigated. We used 55 recombinant thrombin mutants in which solvent accessible residues are replaced with alanine to determine their effect on thrombin-PAI-1, thrombin-PAI-1-heparin and thrombin-PAI-1-VN interactions. Results from this study identified thrombin residues that either increased or decreased thrombin inhibition by PAI-1 relative to wild-type thrombin. First, we confirmed that Glu²⁵ (E25A, located in the 39-loop) had an enhanced rate of inhibition by PAI-1 in the presence and absence of heparin and vitronectin. Also, thrombin residues, Asn²¹⁶/Asn²¹⁷ (N216A/N217A, located in the 203–206 loop) and Lys¹⁴⁵/Thr¹⁴⁷/Trp¹⁴⁸ (K145A/T147A/W148A, located in the autolysis loop), showed increased rates of thrombin inhibition by PAI-1. These results suggest that these three thrombin regions contribute to the slow rate of thrombin inhibition by PAI-1. Second, we identified two anion-binding exosite-1 thrombin mutants, R68A and to a lesser extent Y71A, that showed decreased rates of inhibition by PAI-1 compared to wild-type thrombin. Consistent with this finding, there is a decrease in inhibition of γ-thrombin (α-thrombin proteolyzed in exosite-1) by PAI-1. These results suggest that Arg⁶⁸ and Tyr⁷¹ of thrombin exosite-1 are potential PAI-1 interacting residues since there is a decrease in inhibition in the absence and presence of heparin and VN. Third, we identified four anion-binding exosite-2 thrombin mutants (R89A/R93A/E94A, R98, R178A/R180A/D183A, R245A/K248A/Q251) that are resistant to PAI-1-heparin accelerated inhibition compared to wild-type thrombin, which implies that the thrombin residues important for antithrombin-heparin inhibition are also involved in the PAI-1-heparin inhibition reaction. By contrast, these exosite-2 thrombin mutants are not as resistant to VN-accelerated PAI-1 inhibition, which indicates that exosite-2 is more important for heparin interaction than for VN interaction. Lastly, active site thrombin mutants (W50A, D51A, E202A) and the sodium binding site thrombin mutants (E233A, R233A) were very resistant to PAI-1 inhibition in the absence and presence of heparin and VN. Considering that thrombin, PAI-1 and VN are localized in atherosclerotic arterial vessel wall, our results illustrate the importance of various thrombin domains for PAI-1 inhibition with and without heparin and VN.