Methods for stable multimeric molecular dynamics simulation and modeling of cardiac SK2 channel function.

Abstract

Small conductance Ca2+-activated K+ (SK) channels are unique since they are gated solely by changes in intracellular Ca2+ and are known to be upregulated in heart failure. SK channels are regulated via the ubiquitous Ca2+ sensing protein, calmodulin (CaM). Currently there are no structures for SK1-3 channels bound to CaM, thus, significant need for a method to generate stable SK channel models to assess conduction and membrane interactions exists. Recent cryo-EM structures with nearly complete tetrameric SK4-CaM complex (PDB: 6CNM, 6CNN, 6CNO) allows the generation of nearly full-length SK2-CaM homology models (SK2-CaM). Molecular dynamic simulations (MD) performed on SK2-CaM models generated with established methods, however, critical loops and subunits were unstable due to multimeric and size of complex. We have developed a method for each state of SK4-CaM to be used as an input with output being 1μs+ stable MD simulation ofSK2-CaM complex. This process includes many novel approaches using Rosetta, NAMD, and Amber along with specific regional restraints during specific steps; the procedure can differ for various states. MD designed to assess stability and conduction of the relevant SK4-CaM PDBs listed above and comparing results to the similar MD simulations of the final SK2-CaM models, we propose potential mechanisms for ion conduction in SK2 and SK4. In addition, basic residue clustering was detected near the lower leaflet headgroup of the membrane, implying anionic lipid binding locations. SK2-CaM was also stable in all 3 tested states with POPC membrane and varying concentrations of anionic lipids, allowing for the exploration of binding/interacting mechanisms. Using our expertise on the SK2-CaM complex as a template, we propose a new protocol that can be used to create stable MD simulations of homology models for SK1-3 using SK4 as an input.