Predicting the Interaction between Volatile Anesthetics and Cytoskeleton Proteins by Molecular Modelling

Anesthesia, the reversible pharmacological suspension of conscious brain activity, despite being the cornerstone of modern surgery, is to this date a biological puzzle. The peculiar selectivity of general anesthetics towards consciousness, sparing most of the other brain functions essential for homeostasis, introduces the broad problem of finding the underlying biological mechanisms of conscious perception on one hand, and on the other univocally determining the neural correlates of consciousness to be able to quantitatively measure it. While scientific efforts still have not managed to frame both aspects in a convincing theoretical framework, over the years various theories of general anesthetic action have been proposed. Initially, at the turn of the 20th century, potency of general anesthetics has been correlated to their lipid solubility. Subsequent theories somewhat coherently focused on the effects of anesthetics on the lipid environment of cell membranes but failed to discover a mechanism of action able to discriminate between anesthetics and non-anesthetics presenting similar lipid solubility and physicochemical properties. Research then focused on interactions with membrane proteins such as ion channels and receptors, but with inconclusive results. In the wake of studies providing evidence of the anesthetic-binding proteome in neurons, recent efforts have been focused on the traditionally overlooked cytoskeleton, specifically on microtubules. Their specific density and organization inside neurons is thought to be essential for memory formation and consciousness, in processes involving coherent quantum dipole oscillations. Following this theory, the present work is aimed at spatiotemporally characterizing the interaction between volatile anesthetics and the tubulin dimer – i.e. the constitutive element of microtubules – through the use of computational molecular modelling techniques. Different tubulin isotypes of human tubulin have been modeled and simulated in the presence of three different anesthetics ¬ desflurane, halothane and methoxyflurane – in the surrounding solvent. Compounds have also been docked to the dimers for comparison. The subsequent analysis highlighted distinct patterns of interaction of each anesthetic with each isotype, with a significant fraction of binding sites located on the luminal surface of the microtubule. The location of interaction sites is also compatible with alterations of quantum dipole oscillations, due to the proximity of π-electron clouds of aromatic residues, while the interaction time is in the order of tens of nanoseconds at the most. Further studies might explain how these distinctive patterns might be correlated to anesthetic potency and possible side effects, and if the anesthetics are indeed able to reach the lumen of microtubules.