State-dependent protein-lipid interactions of a pentameric ligand-gated ion channel in a neuronal membrane

Abstract

Pentameric ligand-gated ion channels (pLGICs) are receptor proteins that are sensitive to their membrane environment, but the mechanism for how lipids modulate function under physiological conditions in a state dependent manner is not known. The glycine receptor is a pLGIC whose structure has been resolved in different functional states. Using a realistic model of a neuronal membrane coupled with coarse-grained molecular dynamics simulations, we demonstrate that the lipid-protein interactions are dependent on the receptor state, suggesting that lipids may regulate the receptor’s conformational dynamics. Comparison with existing structural data confirms known lipid binding sites, but we also predict further protein-lipid interactions including a site at the communication interface between the extracellular and transmembrane domain. Moreover, in the active state, cholesterol can bind to the binding site of the positive allosteric modulator ivermectin. These protein-lipid interaction sites could in future be exploited for the rational design of lipid-like allosteric drugs.Author SummaryIon channels are proteins that control the flow of ions into the cell. The family of ion channels known as the pentameric ligand gated ion channels (pLGICS) open in response to the binding of a neurotransmitter, moving the channel from a resting state to an open state. The glycine receptor is a pLGIC whose structure has been resolved in different functional states. It is also known that the response of pLGICs can also be modified by different types of lipid found within the membrane itself but exactly how is unclear. Here, we used a realistic model of a neuronal membrane and performed molecular dynamics simulations to show various lipid-protein interactions that are dependent on the channel state. Our work also reveals previously unconsidered protein-lipid interactions at a key junction of the channel known to be critical for the transmission of the opening process. We also demonstrate that cholesterol interacts with the protein at a site already known to bind to another compound that modulates the channel, called ivermectin. The work should be useful for future drug design.

A physiological role for GABAAreceptor desensitization – induction of long-term potentiation at inhibitory synapses

“…Consistent with this, recent structural work on the glycine receptor, and on the prokaryotic channel ELIC, has shown that the straightening of the M4 helix in a desensitized conformation disrupts this lipid binding site 69,70 . Furthermore, simulations of the analogous phosphatidylserine binding site on the glycine receptor have also shown that lipid binding is favoured in the inactive conformation relative to the active one 71 . Such a mechanism could effectively act as a signal of GABAAR activity at inhibitory synapses, potentially providing a more reliable indicator of synaptic activity than chloride fluxes or inhibition of excitatory signalling and thus enabling a greater range of inhibitory plasticity.…”

Polyunsaturated fatty acids inhibit a pentameric ligand-gated ion channel through one of two binding sites

“…The cost of these modifications, however, is a lack of atomic resolution and limited sampling of protein conformational changes due to structural restraints required for protein models (32). Nevertheless, CGMD has been useful for sampling protein-lipid interactions by permitting more lipid diffusion over the simulation time scale (33,34). We first quantified lipid sorting using the boundary lipid metric, B (Equation 2), where B > 1 indicates enrichment compared to bulk membrane and B < 1 indicates relative depletion (34).…”

Towards design of drugs and delivery systems with the Martini coarse-grained model

“…Recently, several CG MD investigations of proteinhttps://doi.org/10.1017/qrd.2022.16 Published online by Cambridge University Press ligand interactions in the transmembrane region of pharmacologically relevant targets have been reported. In general, the common computational strategy involves: (i) binding-site identification and structural characterization of the protein-ligand complex using, among other methods, unbiased CG MD simulations and ligand-density maps (Ferraro et al, 2016;Dämgen & Biggin, 2021); (ii) ranking of binding modes by binding affinity calculations based on equilibrium MD (Souza et al, 2020), potential of mean force (PMF), alchemical transformations, metadynamics (Corey et al, 2019), or binding saturation curves (Ansell et al, 2021); and (iii) structural refinement of the protein ligand complex via backmapping to atomistic models (Wassenaar et al, 2014).…”

“…A similar analysis of the chemokine receptor 3, a GPCR responsible for trafficking white blood cells, allowed for the identification of six cholesterol-binding sites, suggesting that recognition of cholesterol at these sites may modulate the affinity for agonists/antagonists allosterically via a rigidification of the protein structure (van Aalst, Koneri, & Wylie, 2021). Using CG MD simulations and lipid-density maps, Damgen and Biggin explored the affinity of cholesterol and different lipid types for the glycine receptor channel in its active and resting states and found that lipids may act as allosteric modulators because their strength of binding strongly depend(s) on the physiological state of the receptor (Dämgen & Biggin, 2021). In a similar study, protein-lipid interactions on the homologous nicotinic acetylcholine receptor were investigated using a complex quasi-neuronal membrane composed of 36 species of lipids, including cholesterol, in a binding competition assay (Sharp & Brannigan, 2021).…”