Transient states and barriers from molecular simulations and milestoning theory: kinetics in ligand-protein recognition and compound design

Abstract

Abstract:It is important but unclear how protein-ligand system motions affect free energy profiles, create energy barriers, and lead to slow residence time. We computed residence time (RT) and potential of mean force (PMF) of the dissociations of five structure-kinetic relationship (SKRs) series inhibitors of CDK8/CycC using metadynamics and milestoning theory. By using a novel way to represent the reaction coordinate based on principal component analysis (PCA), we found one-to-one mapping of hydrogen bond (H-bond) breakage and protein domain motions with the energy barriers in the PMFs. This work provides a novel and well-defined approach to study the dissociation kinetics of large and flexible systems.peer-reviewed) is the author/funder. All rights reserved. No reuse allowed without permission.The copyright holder for this preprint (which was not . http://dx.doi.org/10.1101/169607 doi: bioRxiv preprint first posted online Jul. 28, 2017; 2 Binding kinetics has become an important topic in molecular recognition because of the growing awareness of the correlation between kinetics and the drug efficacy.1-4 koff and residence time (RT) are particularly important as determinants of the efficacy of a drug candidate. 5 It is desirable to optimize these properties for drug discovery but, binding kinetics and its determinants in the ligand-receptor structures are not yet fully understood.It is also important to predict these quantities computationally when experimental measurements are not available. Therefore, computational methods, which are able to provide an atomistic description of temporal and spatial dissociation pathway details of ligand-receptor, can be employed to unravel the secret behind the RT and kinetics. The copyright holder for this preprint (which was not . http://dx.doi.org/10.1101/169607 doi: bioRxiv preprint first posted online Jul. 28, 2017; 3 this concern in mind, it is nature to solve this issue by using principal component analysis (PCA), a mathematical method that extracts the major motions from a collection of data. Therefore, it is desirable to take the advantage of PCA to suggest a more straightforward but robust way to define reaction coordinate for calculation using milestoning theory.In this work, we used metadynamics to provide dissociation pathways and computed PMF and RT for five ligands from the SKR compounds series 12 using milestoning theory with a novel way for defining the milestones. The details of methods are included in SI. Using Table 1. The relative ranking distinguishes ligand1 and 2 that have much slower RTs than the rest ligands.Among the ligands, ligand 1 has a RT on the millisecond level which reproduces the experimental value the best. The RTs of other ligands are on the micro-or submicro-level.Interestingly, the computed RT of ligand 2, whose experimental RT is slightly slower, is hundred times faster than ligand 1. We noticed that ligand 2 forms half less H-bonds with CDK8 and requires minor protein motions for it to dissociate compared to ligand 1.Therefore...

Machine Learning Analysis of τRAMD Trajectories to Decipher Molecular Determinants of Drug-Target Residence Times

“…To overcome this limitation, a range of enhanced sampling techniques has been explored recently (Bruce et al, 2018). Some of them are aimed at the reduction of the configurational space to be sampled for the computation of binding kinetic rates, e.g., metadynamics (Tiwary et al, 2015, 2017), weighted ensemble methods (Dickson and Lotz, 2016; Dixon et al, 2018), or milestoning (Tang and Chang, 2017) [a detailed review can be found elsewhere (Mollica et al, 2016; Dickson et al, 2017)]. Although these methods are designed for the prediction of the absolute values of binding and unbinding rates within a reasonable computation time, they are still very computationally demanding and require high user expertise, which impedes the implementation of these methods in drug design pipelines.…”

Markovian Weighted Ensemble Milestoning (M-WEM): Long-time Kinetics from Short Trajectories

“…Analyzing the milestone-to-milestone transition statistics via a statis-tical framework, 42 the kinetics and free energy profile are estimated. This method has also been implemented in the software tools miles, 47 ScMile 48 and SEEKR, 49 and has been used to study a variety of complex biological problems including protein allosteric transitions, 50 membrane permeation by small molecules, [51][52][53][54] protein small molecule interaction, 49,[55][56][57] simple ligand-receptor binding, 58 peptide transport through protein channels, 59,60 DNA protein interaction, 61 protein conformational dynamics 62 etc. Apart from the necessity of having a predefined reaction coordinate, the milestones need to be placed far apart to preserve the assumption of Markovianity.…”