Background: Protein S (PS) is a vitamin K-dependent anticoagulant whose importance is distinguished by many hematological disorders in PS-deficient individuals. PS deficiency is associated with venous and arterial thrombosis, cerebral thrombophlebitis, myocardial infarction, ischemic stroke, and vascular calcification. The 75-kDa PS contains a γ-carboxyglutamic acid module, a thrombin sensitive region, four epidermal growth factor-like modules, and a sex hormone binding globulin-like region that consists of two laminin G domains (LG1 and LG2). PS has three different functions: (a) a cofactor for activated protein C; (b) a cofactor for tissue factor pathway inhibitor, and (c) a direct inhibitor of factor IXa (FIXa). Our laboratory discovered that PS binds FIXa with a dissociation constant (Kd) ~40 nM and inhibits FIXa's function. Currently, we are ascertaining the domain of PS responsible for binding FIXa.

Result and Discussion:

By using molecular modeling and biochemical assays, we demonstrated that the laminin G domains of PS are crucial for binding FIXa. However, we do not have any information regarding conformational or structural changes that occur when the LG domains and FIXa interact. To confirm our hypothesis, we used circular dichroism (CD), isothermal titration calorimetry (ITC) and biochemical assays to locate the region of the LG domains responsible for FIXa binding.

CD spectroscopy gives details of protein secondary structure and conformational change. The conformational change was greatest for binding of LG1+2 to FIXa (LG1+2>LG1>LG2); thus, FIXa apparently has a greater affinity for LG1+2 than for either of the separate LG domains.

ITC measures the stoichiometry and thermodynamics (binding affinity, enthalpy, entropy, and Gibbs free energy of protein-protein interaction. We observed that the binding affinity between LG1+2 and FIXa was higher than the affinity between either LG1 or LG2 and FIXa, with a trend of LG1+2>LG1>LG2. In addition, the differences in Gibbs free energy supported the same trend. The binding stoichiometry was 1:1 in all three cases, which implied that each LG construct binds one FIXa molecule. All three interactions were enthalpy driven, their binding was exothermic, and binding was favored by a negative enthalpy.

In addition to the foregoing biophysical study, we performed biochemical assays to assess the activities of the LG domains. Thrombin generation assays were performed with PS-deficient plasma supplemented with each LG domain construct and with intact PS. The reactions were initiated with a low tissue factor concentration (0.4 pm) to selectively generate thrombin in the intrinsic pathway. With PS-deficient plasma, the amount of thrombin produced decreased from 255 nM to 155 nM on addition of LG1+2. This result was comparable to the effect of adding intact PS (148 nM thrombin produced). Overall, the structural studies and the plasma studies indicated that LG1+2 is required to mimic the function of PS in binding to and inhibiting FIXa. Similarly, we previously observed that the LG1+2 domain inhibited generation of FXa by FIXa (by ~55%), which was similar to the inhibition of FIXa by intact PS. We conclude that the LG1+2 domain mimics the function of PS in inhibiting FIXa.

Future Direction:

To exactly identify the region in the PS LG domains responsible for binding to FIXa, we have created multiple single and double amino acid substitutions in the LG domains. Currently, we are identifying particular PS mutants unable to bind FIXa to precisely map the FIXa interaction region in PS.

Disclosures

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.