The Role of Cholesterol in M2 Clustering and Viral\nBudding Explained

The influenza A M2\nhomotetrameric channel consists of four transmembrane\n(TM) and four amphipathic helices (AHs). This viral proton channel\nis suggested to form clusters in the catenoid budding neck areas in\nraft-like domains of the plasma membrane, resulting in cell membrane\nscission and viral release. The channel clustering environment is\nrich in cholesterol. Previous experiments have shown that cholesterol\nsignificantly contributes to lipid bilayer undulations in viral buds.\nHowever, a clear explanation of membrane curvature from the distribution\nof cholesterol around the M2TM-AH clusters is lacking. Using coarse-grained\nmolecular dynamics simulations of M2TM-AH in bilayers, we observed\nthat M2 channels form specific, C2-symmetric, clusters with conical\nshapes driven by the attraction of their AHs. We showed that cholesterol\nstabilized the formation of M2 channel clusters by filling and bridging\nthe conical gap between M2 channels at specific sites in the N-termini\nof adjacent channels or via the C-terminal region of TM and AHs, with\nthe latter sites displaying a longer interaction time and higher stability.\nThe potential of mean force calculations showed that when cholesterols\noccupy the identified interfacial binding sites between two M2 channels,\nthe dimer is stabilized by 11 kJ/mol. This translates to the cholesterol-bound\ndimer being populated by almost 2 orders of magnitude compared to\na dimer lacking cholesterol. We demonstrated that the cholesterol-bridged\nM2 channels can exert a lateral force on the surrounding membrane\nto induce the necessary negative Gaussian curvature profile, which\npermits spontaneous scission of the catenoid membrane neck and leads\nto viral buds and scission.