Impact of acid pH on neurologically relevant systems explored by molecular simulation

Extracellular acidification may impact the function of neurobiologically relevant biomolecules, such as transmembrane channels and transporters. This is the case of mouse type 2 hyperpolarization-activated cyclic nucleotide-gated (mHCN2) channel, investigated by Prof. M. Spehr here at RWTH-Aachen University. HCN channels are key mediators in regulating synaptic weight, dendritic integration and filtering, gain control as well as neuronal and network oscillations. Providing insight on the protonation state and on the H-bond pattern upon acidification is very difficult experimentally. From the computational point of view, the method of choice is arguably Constant pH Molecular Dynamics (CpHMD) Simulations, which modify the protonation state of ionized residues during the dynamics according to their chemical environment. Here we apply the method to the neuropeptide Y (NPY) in solution, a small system and yet very important for regulating key processes in the brain such as memory formation and feeding. It participates in cell signaling paths that delay aging. In the peripheral nervous system, it is involved in vasoconstriction. It is also one of the extracellular neurotransmitters that modulate HCN channels. Its interaction with the cell membrane, where it binds to specialized receptors with key physiological roles, likely depends on the pH. However, the protonation states of the carboxylic and histidine residues of the peptide, and the interplay between protonation states and peptide conformational dynamics, have not been explored. Here we perform CpHMD simulations and graph-based analyses to investigate dynamics and H-bond pattern of the NPY in the range from 3.0 to 7.0. The structural ensembles of the porcine and human peptides, solved by NMR acid pH in aqueous solution, indicated an α-helical core with unstructured termini. We find that an α-helical core, as observed in the NMR experiments, is presented at all pH values, though its length can vary by 2-3 residues depending on the pH. The pKa of Asp16, part of the α-helix, and of Asp11 may shift by more than on pH unit. Based on these findings, we suggest that performing constant pH simulations may be required to describe accurately the interactions of the peptide with its cellular partners at the pH values of interest. Being successful in this first application, we have started a computational study mHCN2 based on the abundant structural information on human and rabbit HCN channels. Our work sets the stage for CpHMD, which might be able to provide the molecular basis of the pH-driven shift in voltage-dependence of Ih activation that causes the opening of HCN channels at rest, observed experimentally by Prof. Spehr and co-workers.