Computationally driven discovery of SARS-CoV-2 Mpro inhibitors: from design to experimental validation.

We report a fast-track computationally driven discovery of new SARS-CoV-2 main protease (Mpro) inhibitors whose potency ranges from mM for the initial non-covalent ligands to sub-µM for the final covalent compound (IC50 = 830 ± 50 nM). The project extensively relied on high-resolution all-atom molecular dynamics simulations and absolute binding free energy calculations performed using the polarizable AMOEBA force field. The study is complemented by extensive adaptive sampling simulations that are used to rationalize the different ligand binding poses through the explicit reconstruction of the ligand-protein conformation space. Machine learning predictions are also performed to predict selected compound properties. While simulations extensively use high performance computing to strongly reduce the time-to-solution, they were systematically coupled to nuclear magnetic resonance experiments to drive synthesis and for in vitro characterization of compounds. Such a study highlights the power of in silico strategies that rely on structure-based approaches for drug design and allows the protein conformational multiplicity problem to be addressed. The proposed fluorinated tetrahydroquinolines open routes for further optimization of Mpro inhibitors towards low nM affinities.

Response to comment on 'Valid molecular dynamics simulations of human hemoglobin require a surprisingly large box size'.

We recently reported that molecular dynamics simulations for hemoglobin require a surprisingly large box size to stabilize the T(0) state relative to R(0), as observed in experiments (El Hage et al., 2018). Gapsys and de Groot have commented on this work but do not provide convincing evidence that the conclusions of El Hage et al., 2018 are incorrect. Here we respond to these concerns, argue that our original conclusions remain valid, and raise our own concerns about some of the results reported in the comment by Gapsys and de Groot that require clarification.

Valid molecular dynamics simulations of human hemoglobin require a surprisingly large box size.

Recent molecular dynamics (MD) simulations of human hemoglobin (Hb) give results in disagreement with experiment. Although it is known that the unliganded (T[Formula: see text]) and liganded (R[Formula: see text]) tetramers are stable in solution, the published MD simulations of T[Formula: see text] undergo a rapid quaternary transition to an R-like structure. We show that T[Formula: see text] is stable only when the periodic solvent box contains ten times more water molecules than the standard size for such simulations. The results suggest that such a large box is required for the hydrophobic effect, which stabilizes the T[Formula: see text] tetramer, to be manifested. Even in the largest box, T[Formula: see text] is not stable unless His146 is protonated, providing an atomistic validation of the Perutz model. The possibility that extra large boxes are required to obtain meaningful results will have to be considered in evaluating existing and future simulations of a wide range of systems.

A Toolkit to Fit Nonbonded Parameters from and for Condensed Phase Simulations.

The quality of atomistic simulations depends decisively on the accuracy of the underlying energy function (force field). Of particular importance for condensed-phase properties are nonbonded interactions, including the electrostatic and Lennard-Jones terms. Permanent atomic multipoles (MTPs) are an extension to common point-charge (PC) representations in atomistic simulations. MTPs are commonly determined from and fitted to an ab initio Electrostatic Potential (ESP), and Lennard-Jones (LJ) parameters are obtained from comparison of experimental and computed observables using molecular dynamics (MD) simulations. For this a set of thermodynamic observables such as density, heat of vaporization, and hydration free energy is chosen, to which the parametrization is fitted. The current work introduces a comprehensive computing environment (Fitting Wizard (FW)) for optimizing nonbonded interactions for atomistic force fields of different qualities. The FW supports fitting of standard PC-based force fields and more physically motivated multipolar (MTP) force fields. A broader study including 20 molecules ranging from N-methyl-acetamide and benzene to halogenated benzenes, phenols, anilines, and pyridines yields a root mean squared deviation for hydration free energies of 0.36 kcal/mol over a range of 8 kcal/mol. It is furthermore shown that PC-based force fields are not necessarily inferior compared to MTP parametrizations depending on the molecule considered.

Vibrational relaxation and energy migration of N-methylacetamide in water: the role of nonbonded interactions.

Nonequilibrium molecular dynamics (MD) simulations together with physics-based force fields are used to follow energy flow between vibrationally excited N-methylacetamide (NMA) and water. The simulations are carried out with a previously validated force field for NMA, based on a multipolar representation of the electrostatics, and with a new fluctuating point charge model. For the water solvent, a flexible and a rigid model was employed to distinguish between the role of inter- and intramolecular degrees of freedom. On a 10 ps time scale about 90% of the available energy goes into the solvent. The remaining energy resides within internal NMA-degrees of freedom from where energy flow takes place on longer time scales. The total amount of energy transferred to the solvent on the 10 ps time scale does not depend on whether the water molecules are rigid or flexible during the simulations. Vibrational energy relaxation time scales include two regimes: one on the several 100 fs time scale and a longer one, ranging from 6 to 10 ps. This longer time scale agrees with previous simulations but overestimates the experimentally determined relaxation time by a factor of 2, which can be explained by the classical treatment of the vibrations. Including a previously determined quantum correction factor brings the long time scale into quite favorable agreement with experiment. Coupling to the bending vibration of the water molecules in H-bonding contact with the excited C═O chromophore is substantial. The equilibrium and nonequilibrium distribution of the bending angles of the water molecules in contact with the local oscillator are non-Gaussian, and one approaches the other on the subpicosecond time scale. Analysis of the water velocity distribution suggests that the C═O vibrational energy relaxes into the solvent water shells in an impulsive fashion on a picosecond time scale.

Spatial Averaging: Sampling Enhancement for Exploring Configurational Space of Atomic Clusters and Biomolecules.

Spatial averaging Monte Carlo (SA-MC) is an efficient algorithm dedicated to the study of rare-event problems. At the heart of this method is the realization that from the equilibrium density a related, modified probability density can be constructed through a suitable transformation. This new density is more highly connected than the original density, which increases the probability for transitions between neighboring states, which in turn speeds up the sampling. The first successful investigations included the diffusion of small molecules in condensed phase environments and characterization of the metastable states of the migration of the CO ligand in myoglobin. In the present work, a general and robust implementation including rotational and torsional moves in the CHARMM molecular modeling software is introduced. Also, a procedure to estimate unbiased properties is proposed in order to compute thermodynamic observables. These procedures are suitable to study a range of topical systems including Lennard-Jones clusters of different sizes and the blocked alanine dipeptide (Ala)2 in implicit and explicit solvent. In all cases, SA-MC is found to outperform standard Metropolis simulations in sampling configurational space at little extra computational expense. The results for (Ala)2 in explicit solvent are in good agreement with previous umbrella sampling simulations.