Membranes and membrane proteins in healthy neurons and in brain diseases: recent advances from multiscale simulation and machine learning

Neuronal membranes are highly complex materials. They are composed of highly diverse phospholipids, cholesterol, and glycolipids. They contain ion channels, receptors, and transporters, often present not only as monomeric but also as multimeric structures. Neuronal membranes are crucial for brain function, from generating and propagating action potentials to provide the location for neurotransmitter release and receptor binding in neurotransmission. Membrane derangement is central to the development of brain diseases. From one hand, oxidative stress-based membrane damage and interactions with aggregates (interfering with neurotransmitter release) contribute to neurodegenerative diseases such as Alzheimer’s. Hence, ligands that countereact these aberrant processes are emerging as interesting therapeutic agents interfering with these diseases. From the other, abnormal membrane potentials may alter neuronal excitability and affecting channel function in epilepsy. Here, targeting dysfunctional ion channels with drugs could help alleviating membrane derangement associated with epilepsy[ro04]. Besides their role per se, membranes impact on neuroreceptors (dys)function: their transmembrane potential govern ion channels gating, receptor activation and they affect neurotransmitter/drug binding, and they cause proteins’ allosteric changes. Because of their relevance for basic brain science and for neurological /neurodegenerative diseases, the computational biophysics community is paying lots of attention to the structure and the dynamics of neuronal membranes, especially on their interactions with membrane proteins.