Clustering of SARS-CoV-2 membrane proteins in lipid bilayer membranes

The accumulation of viral structural proteins along the endoplasmic reticulum-Golgi intermediate compartment (ERGIC) membrane drives SARS-CoV-2 self-assembly and budding through interactions among proteins, RNA, and the host membrane. The membrane (M) protein, the most abundant structural component, is thought to interact with other proteins and form clusters that induce membrane curvature and initiate virion formation. However, the relative roles of direct and membrane-mediated interactions between M proteins in this clustering process remain unclear. Here, we combine all-atom molecular dynamics (MD) simulations, continuum modeling, and experiments to demonstrate that M-M interactions alone are sufficient to drive clustering in ERGIC-like lipid bilayers, even in the absence of other proteins or RNA. From MD simulations, we quantify the membrane thinning induced by M proteins and the resulting membrane-mediated interaction energy. Integrating these results into a continuum model that describes the evolution of M protein density on a planar membrane, we identify a critical effective interaction energy required for cluster formation at a given protein density. Comparison with atomic force microscopy (AFM) measurements of M protein clusters enables quantitative estimation of the direct and membrane-mediated interaction energies, revealing that direct M-M interactions dominate through an effective oligomerization energy. Together, these findings establish that M protein interactions are sufficient to drive clustering and provide a quantitative framework for understanding the interplay of direct and membrane-mediated forces in coronavirus assembly and budding.