Lipid-Dependent Modulation of Desensitization in Pentameric Ligand-Gated Ion Channels

Abstract

Pentameric ligand-gated ion channels (pLGICs) are central to fast synaptic communication in the nervous system. Upon agonist binding, these receptors open an ion-conducting pore across the postsynaptic membrane, allowing selective ions to flow down their electrochemical gradient into the cell, and prolonged exposure to agonist drives the channels into a non-conducting desensitized state. While agonist binding initiates channel opening, the agonist-induced response can be influenced by various exogenous and endogenous compounds, including the lipids in the cell membrane. Although the lipid sensitivity of the prototypic pLGIC, the nicotinic acetylcholine receptor (nAChR) from Torpedo, is well characterized, the lipid sensitivities of other pLGICs, including the prokaryotic ELIC, have only recently gained considerable interest. Structural studies have identified a putative inner-leaflet lipid pocket in ELIC, shaped by a Trp-Arg-Pro (W-R-P) motif. Compelling evidence suggests that binding of phosphatidylethanolamine (PE) or phosphatidylglycerol (PG) to this site leads to slower rates of ELIC desensitization, likely by influencing the dynamics or position of the outermost transmembrane helix, M4. This motif is conserved in anion-selective pLGICs and in GLIC, raising the question of whether it plays a conserved functional role. Using site-directed mutagenesis and two-electrode voltage clamp (TEVC) electrophysiology, I probed the role of the W-R-P motif in ELIC and tested whether this lipid site plays a conserved functional role across pLGICs. I show that the W-R-P pocket plays different roles in modulating desensitization across pLGICs. Mutations to W-R-P pocket residues and to regions of M4 affect desensitization rates in both ELIC and the α1 glycine receptor (GlyR) but have no measurable effect in GLIC. ELIC also tolerates a broader range of M4 mutations than GlyR or GLIC. The phenotypes observed correlates with the density of aromatic interactions at the M4-M1/M3 interface, which is sparse in ELIC but extensive in α1 GlyR and GLIC. Focusing on ELIC, I then show that the strength of interactions at the M4-M1/M3 interface modulates the rate of desensitization. Eliminating the Trp224-Arg301 cation-π interaction accelerates desensitization, while reversing or substituting this interaction with a potential salt bridge restores slow desensitization. Strengthening M4-M1/M3 interactions through three aromatic substitutions partially restores slow-desensitizing phenotypes. My data identify three distinct interactions and/or features of the M4-M1/M3 interface that contribute to slow desensitization in ELIC including the Trp224-Arg301 cation-π interaction, an interaction between W220 on M1 and S271 on M3, and the Pro305 kink on M4. Collectively, my data shows that the W-R-P pocket plays different roles in shaping desensitization across pLGICs, and that the strength of interactions at the M4-M1/M3 interface modulate the rate of ELIC desensitization. These findings support a model in which lipid binding at the W-R-P pocket modulates ELIC desensitization through its effects on M4 structure or dynamics, while the broader architecture of the M4-M1/M3 interface determines whether this mechanism applies in other pLGICs.