Exploring the Free Energy Landscape of Ion Channel-Modulator Binding with Molecular Dynamics Simulations

Ion channels play a critical role in the proper function of the body and are a major target for pharmaceutical compounds. A better understanding of their function, particularly in response to modulating compounds, will aid in both future medical advancements and scientific understanding. General anaesthesia is a method of medically inducing unconsciousness, allowing invasive surgeries to be performed on patient without causing them pain. While anaesthesia has been utilised medically for over a century, the molecular mechanisms causing the effect have not been fully resolved. Experimental evidence currently points to the modulation of Cys-loop ligand-gated ion channels as the cause. Similarly, many of the voltage-gated ion channels present in the peripheral nervous system possess highly specific roles in signal transmission. Naturally occurring venom compounds found in cone snails demonstrate that these channels can be targeted with high selectivity. A more complete understanding of the molecular mechanisms these venom peptides utilise could allow for highly targeted therapies, including the inhibition of pain signals. This thesis makes use of molecular dynamics (MD) and a series of atomically solved structures of ligand- and voltage-gated ion channels. Extensive simulations were performed of the pentameric ligand-gated channel GLIC with the general anaesthetic molecule propofol. These simulations made use of the metadynamics accelerated-sampling technique, although techniques such as unbiased sampling, umbrella sampling and free energy perturbation each contributed to this investigation. The simulations performed in this study were able to capture every possible propofol binding position in the GLIC transmembrane domain, generating a comprehensive free energy surface of interaction in both the open and closed states. This completely described the position of every site and pathway of the drug in the channel. The interaction energy between propofol and each of the residues inside of every site was also analysed. This identified the key contributors to the formation of the propofol binding sites. Through extensive simulations and analysis, it was demonstrated that propofol does not modulate target channels by way of free energy differences. Instead, it was shown that propofol binding to the inhibitory sites in open state GLIC altered the M2-M1(-) helix separation distances. This variable is descriptive of the radius of the GLIC pore and defines whether the channel is open or closed. Binding in the closed state had no impact, except to prevent the return of the channel structure to the open state. Similarly, a series of metadynamics simulations of the NaVAb voltage-gated channel bound to the KIIIA μ-conotoxin peptide were performed. These were intended to investigate the binding free energy of each of the possible KIIIA-NaVAb orientations. Binding between the two macromolecules was extremely strong, which presented a challenge to sampling. However, with the application of MD methods to separate the two molecules before reorienting and rebinding, the problem was partially overcome. This allowed for the estimation of the binding free energy of KIIIA and to identify the optimum binding orientation for the toxin. In this position, residue pairs between KIIIA and NaVAb that were experimentally identified as stabilising the interaction were seen to be in close proximity to each other. The use of MD simulations throughout this thesis were able to determine previously unknown information on the behaviour of anaesthetic modulation and conotoxin-channel binding. The ability to sample drug binding at all points in a large protein such as GLIC or between strongly bound complexes such as KIIIA-NaVAb, presents an ongoing challenge to research. This study addressed these issues with the extensive use of metadynamics to accelerate sampling, along with a broad number of other MD and post-analysis techniques to support the results. The new information learned regarding ion channel drug binding mechanisms could potentially lead to material improvements in the design of medically relevant anaesthetics and analgesics.