Inhibitory properties of the P1 Tyr variant of antithrombin

Abstract

Antithrombin (AT) and protein Z-dependent protease inhibitor (ZPI) are among two physiological serpin inhibitors in plasma that are involved in the regulation of the clotting cascade. Unlike AT, which can inhibit the proteolytic activity of all coagulation proteases, ZPI has narrower protease specificity, inhibiting only factors Xa (fXa) and XIa. Unlike an Arg at the P1 site of the AT reactive center loop (RCL), this residue is a Tyr in ZPI. To investigate the contribution of P1 Tyr in restricting the specificity of ZPI, we engineered an AT mutant in which the P1 Arg of the RCL was replaced with the P1 Tyr of ZPI (AT-R393Y). The reactivity of AT-R393Y with fXa and thrombin was decreased 155- and 970-fold, respectively. However, the serpin mutant inhibited chymotrypsin with an efficiency higher by >4 orders of magnitude. By contrast, chymotrypsin did not exhibit any reactivity with ZPI. The substitution of Asp-189 of fXa with the corresponding residue of chymotrypsin (Ser) did not improve the reactivity of the protease mutant with AT-R393Y; however, the fXa mutant reacted normally with ZPI. These results suggest that the contribution of P1 Tyr to restricting the protease specificity of ZPI is RCL context-dependent and that in addition to P1 Tyr, other structural features within and/or outside the ZPI RCL are involved in determining the protease specificity of the serpin. The results further suggest that thrombin is less tolerant than fXa in accommodating the nonoptimal P1 Tyr of the AT mutant in its active-site pocket.

Figure 1

SDS-PAGE analysis of the stable protease-AT complexes. AT derivatives (2.5 μM each) was incubated with equimolar concentrations of either fXa (A) or thrombin (B) for 30 min in the absence and presence of 2-fold molar excess of heparin in 20 μL reactions at 37 °C. Five μL of non-reducing sample buffer was added to each sample, boiled for 5 min and loaded on 10% SDS-gel. (A) Lane 1, fXa; lane 2, AT-WT; lane 3, AT-R393Y; lane 4, AT-WT + fXa; lane 5, AT-WT + fXa + heparin; lane 6, AT-R393Y + fXa, lane 7, AT-R393Y + fXa + heparin; lane 8, molecular mass standards (kDa). (B) the same as above except that thrombin replaced fXa in the reactions.

Figure 2

Determination of stoichiometries of inhibition of AT-WT and AT-R393Y with proteases in the presence of heparin. A fixed concentration of either fXa (25 nM, panel A) or thrombin (25 nM, panel B) was incubated with increasing concentrations of AT derivatives (○, AT-WT; ●, AT-R393Y) in the presence of 0.5 μM heparin at room temperature for 1–8 h. After completion of the reactions, the residual amidolytic activities were measured and plotted versus the [AT]/[protease] ratios. The SI values were determined from the x-intercept of the linear regression fit of the kinetic data. SI values of 1.3 ± 0.16 and 1.3 ± 0.07 were obtained for the fXa reactions with AT-WT and AT-R393Y, respectively. The corresponding values for thrombin were 1.4 ± 0.08 for AT-WT and 1.4 ± 0.07 for AT-R393Y.

Figure 3

Binding of heparin to recombinant AT derivatives. The spectral changes were monitored by addition of 1–2 μL of a concentrated stock solution of unfractionated heparin to 50 nM AT in TBS (pH 7.5) containing 0.1% PEG-8000 and dissociation constants were calculated from the changes of the intrinsic protein fluorescence as described under “Materials and Methods”. The symbols are: (○) AT-wild type, and (●) AT-R393Y.

Figure 4

Dependence of k_obs values for inactivation of coagulation protease derivatives by AT derivatives on the heparin-serpin complex concentrations. A, the k_obs values for the heparin-catalyzed inactivation of wild-type fXa by AT-R393Y (○), the heparin-catalyzed inactivation of fXa-D189S by the wild-type AT (●) and the heparin-catalyzed inactivation of fXa-D189S by AT-R393Y (□) were determined at increasing concentrations of the heparin-serpin complex (shown on x-axis) in TBS/Ca²⁺ at room temperature by an amidolytic activity assay. B, the same as above except that the inactivation of thrombin and thrombin-D189S was monitored by the same serpins. The solid lines in both panels are least squares computer fits of data by a hyperbolic equation as described under “Materials and Methods”. The kinetic parameters are presented in Table 2.