The proposed studies will provide a quantitative and predictive understanding of coagulation protease dynamics in relation to catalytic function and allosteric control. Cascades of serine proteases, the largest of all of the peptidase families, control coagulation. Control of function and regulation of the coagulation and anticoagulation proteases very likely involves dynamic allostery, however, no studies have measured the dynamics of coagulation proteases. Thrombin provides the switch between coagulation and anti-coagulation, and is allosterically regulated by thrombomodulin (TM) binding. In dynamic allostery, the protein exists in an ensemble of states that rapidly interconvert, and sub-populations of states are selected when an allosteric effector binds. NMR dynamics experiments are the only way to observe interconverting sub-states in allosteric proteins. We therefore propose NMR backbone and side chain dynamics experiments combined with enhanced sampling molecular dynamics (aMD) simulations to fully describe dynamic motions in apo-thrombin, PPACK-thrombin, two W215 mutants of thrombin and the thrombin-TM456 complex. The goal of the project is to obtain quantitative dynamic data that will be used to calibrate aMD simulations. Once calibrated, the simulations could predict dynamic allostery in other proteases involved in coagulation that are not amenable to NMR. The three complementary aims of the project are:
Aim 1. Determine the dynamic motions in thrombin that are important for catalytic function. Accelerated MD (aMD) simulations and NMR experiments will be performed to probe motions on time scales from ns to ms in the apo, and PPACK-bound forms of human ?-thrombin. Our hypothesis is that by comparing results from apo and PPACK-thrombin we will be able to discover those motions which change upon substrate binding, and are therefore likely to be important for catalytic activity.
Aim 2. Determine the dynamic motions in thrombin that are important for allosteric control. Extended aMD simulations will help interpret the experimental results and conversely the NMR results will provide quantitative information about rates of correlated backbone motions with which to validate the simulations. Two W215 mutant forms of thrombin and also the thrombin-TM456 complex will be studied. Our published 10 ns aMD simulations on thrombin-TM456 show that TM causes the disparate motions in thrombin to coalesce into correlated motions. Our hypothesis is that differences in dynamics between different allosteric forms of thrombin will reveal the mechanism of dynamic allostery in thrombin and other serine proteases.
Aim 3. Develop side chain dynamics experiments to probe the dynamic allostery between effector-binding sites and the active site of thrombin. The side chain dynamics results will be iteratively interpreted in light of extended aMD simulations. Hypothesis: That side chain dynamics, which have been shown to play a critical role in conformational entropy, also play a critical role in thrombin allostery.
This project will answer the question: How is the activity of coagulation serine proteases (in particular, thrombin) controlled? Serine proteases are difficult to study because they 'digest themselves' when analyzed in their native form. We have developed a catalytically dead version thrombin that is amenable to study by nuclear magnetic resonance, and have collected preliminary data showing that thrombin is incredibly dynamic. The proposed work is the first bona fide study of the dynamics of any serine protease. It is known that the activity of thrombin is controlled by binding of effector molecules, and our research will focus on how the binding of effectors alters the dynamics of thrombin. Our hypothesis is that by comparing results from apo-thrombin with results from various effector-bound forms, we will be able to discover effector-specific dynamic motions that will reveal the mechanism of its allosteric control. The detailed data we collect will also be used to calibrate computer simulations that recapitulate the slow dynamics we have observed. The simulation protocols could then be used on other coagulation serine proteases that are not amenable to experiments.
|Handley, Lindsey D; Fuglestad, Brian; Stearns, Kyle et al. (2017) NMR reveals a dynamic allosteric pathway in thrombin. Sci Rep 7:39575|