In recent years, significant progress has been made in the calculation of ion current through realistic biological ion channel proteins. The time scales involved in these processes make extraction of currents from brute force Molecular Dynamics difficult. However, it is possible to �feed up� molecular detail obtained from relatively short-time MD simulations into coarse-grained kinetic models, which then can be used to predict ion fluxes through channels with considerable accuracy and efficiency, scanning over applied transmembrane voltages, bathing solution concentrations and in silico mutations of key pore-lining amino acids. We will discuss two examples: 1) The Glutamate Chloride Receptor (GluCl), an invertebrate glutamate-gated chloride channel which has recently been crystallized in an open-pore state. This channel is homologous to the human Cys-loop receptor family of pentameric ligand-gated ion channels (pLGICs). Using relatively short-time (several nanosecond) all-atom MD simulations, we determined the Potential of Mean Force (PMF) for motion of a single ion through the channel, and the appropriate spatially dependent diffusivity profile. These inputs were fed into Brownian Dynamics simulations of ion flow through the channel, from which the ionic conductance of the channel was determined [J. Phys. Chem. B 116  , 13637 (2012)]. Our computations provide insight into the physico-chemical determinants of anion selectivity in GluCl. 2) Fe+2influx into ferritin, a ubiquitous ion storage protein. The ferritin protein is in essence a spherical shell with holes (pores) in it. Fe+2 passes through the protein shell via the appropriate pores. Once inside, it is converted to inert Fe2O3, which is stored as a mineral core. Starting from an extant X-ray structure of the ferritin protein, we performed extensive MD simulations to verify qualitatively the entry of Fe+2 through 3-fold symmetric pores in the ferritin protein shell [Proteins: Structure, Function, and Bioinformatics 81  , 1042 (2013)]. Subsequently, we utilized Poisson-Nernst-Planck (PNP) theory to compute the flux of Fe+2 ions into the interior of the protein, from which we estimated the transit time of an Fe+2 ion through the ferritin shell. This time is on the order of milliseconds, far beyond the reach of brute force all-atom MD simulation. Our 3D PNP model utilized the full geometric detail of the crystal structure of the protein and continuum electrostatic level estimation of the systematic driving forces (PMFs); in addition, short all atom MD simulations were used to obtain the appropriate diffusivity profile for the permeant Fe+2 species. Our results for the ion transit time under physiological conditions are in good agreement with experiment.

Calculation of ion permeation through biological ion channels via course-grained models employing atomic level input

247th National Meeting (2014)

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.