National Institute for Nanotechnology, 11421 Saskatchewan Dr., Edmonton, AB, T6G 2M9, Canada;
Department of Mechanical Engineering, University of Alberta, Edmonton, AB, T6G 2G8, Canada
Title: Statistical-mechanical, 3D molecular theory of solvation for multiscale modeling in nanochemistry and biophysics
A genuine challenge of multiscale modeling is theoretical coupling of methods at different scales,
so that “observables” at lower-level scales are analytically linked to “force fields” of more coarsegrained
models at higher-level scales. Statistical mechanics itself is in a sense an example of such
theoretical coupling between microscopic variables and thermodynamic, macroscopic properties.
The statistical-mechanical, three-dimensional molecular theory of solvation (a.k.a. 3D-RISM-KH)
predicts from the first principles the solvation structure and thermodynamics of nanochemical and
biomolecular systems in solution.1 Rather than operating with trajectories of individual molecules,
it yields 3D correlation functions of species in the statistical ensemble by solving the integral
equations of liquid state theory, and then analytically gives the solvation thermodynamics in terms
of the correlation functions. The method properly accounts for chemical functionalities by
representing both electrostatic and non-polar features of the solvation structure, such as hydrogen
bonding, solvophobicity, salt bridges, structural solvent, associative and electrochemical effects.
We coupled the 3D-RISM-KH theory with ab initio methods (Kohn-Sham DFT and CASSCF) in a
self-consistent field description (including analytical gradients) of electronic structure, optimized
geometry, chemical reactions and nanocatalysis in solution.1,2 We implemented the KS-DFT/3DRISM-KH multiscale method in the ADF computational chemistry package and extensively
validated the method against experimental data for solvation thermochemistry, conformational
equilibria, and activation barriers for various nanosystems in different solvents3 and ionic liquids.4
We also coupled the 3D-RISM-KH theory contracting solvent degrees of freedom with MD
simulation of biomolecules, and implemented the MD/3D-RISM-KH method in the Amber
molecular dynamics package.5 This included a number of accelerating schemes with several cutoffs
for the interaction potentials and correlation functions, an iterative guess for the 3D-RISM
solutions, and extrapolating solvent-induced forces and applying them in large multi-time steps (up
to 20 fs) to enable simulation of large biomolecules. The MD/3D-RISM-KH method allows one to
study large biomolecular systems involving rare solvent events occurring on extremely long
timescales (e.g. solvent molecules, ions, and ligands in pockets and inner spaces), as the solvent
statistics is accounted for statistically-mechanically. It also replaces MM/PB(GB)SA postprocessing
using empirical non-polar terms with statistical-mechanical, MM/3D-RISM-KH
evaluation of the solvation thermodynamics.5,6 It yields at once 3D maps of binding affinity without
phenomenological approximations, and thus has significant potential for computer-aided drug
design.7 This talk will exhibit the KS-DFT/3D-RISM-KH and MD/3D-RISM-KH multiscale
methods and illustrate their application to various nanochemical and biomolecular systems.
1 A. Kovalenko, Three-dimensional RISM theory for molecular liquids and solid-liquid interfaces,
in: Molecular Theory of Solvation, F. Hirata (ed.). Series: Understanding Chemical Reactivity,
vol.24, (Kluwer, Dordrecht, 2003) pp.169-275, (and references therein).
2 S. Gusarov, T. Ziegler, A. Kovalenko, J. Phys. Chem. A 110, 6083 (2006).
3 D. Casanova, S. Gusarov, A. Kovalenko, T. Ziegler, J. Chem. Theory Comput. 3, 458 (2007).
4 M.Malvaldi, S.Bruzzone, C.Chiappe, S.Gusarov, A.Kovalenko, J.Phys.Chem.B 113, 3536 (2009).
5 T. Luchko, S. Gusarov, D. R. Roe, C. Simmerling, D. A. Case, J. Tuszynski, A. Kovalenko, J.
Chem. Theory Comput., 6, 607 (2010).
6 N. Blinov, L. Dorosh, D. Wishart, A. Kovalenko, Biophys. J. 98, 282 (2010).
7 T. Imai, K. Oda, A. Kovalenko, F. Hirata, A. Kidera, J. Am., Chem. Soc. 131, 12430 (2009).
(a) Gloeobacter violaceus pentameric ligand-gated ion channel (GLIC) in an open state [Bocquet et
al, Nature 457, 111 (2009)]. Top view at the ring of hydrophilic residues of GLIC M2 helices.
3D distributions of solvent in GLIC, obtained by the 3D-RISM-KH molecular theory of solvation:
(b) water O (in red) and H (in blue); (c) Na+ (in turquoise); (d) Cl- (in orange). Water forms a cap at
the ring of hydrophilic residues, Na+ fill in the channel, and Cl- are depleted in the channel.