HIV-1 GAG SPECIFICITY FOR PIP2-CONTAINING MEMBRANES MIGHT BE DRIVEN BY MACROMOLECULAR ELECTRIC PROPERTIES RATHER THAN MOLECULAR AFFINITIES

The HIV-1 assembly occurs at the plasma membrane, where the GAG polyprotein plays a crucial role. The GAG-membrane association is directed by the matrix domain (MA), which is myristoylated and has a highly basic region that interacts with the anionic lipids. Several evidence suggests that the presence of phosphatidylinositol-(4,5)-bisphosphate (PIP 2 ) highly influence this binding. In addition, MA also interacts with nucleic acids, which is proposed to be important for the specificity of GAG for PIP 2 -containing membranes. It is proposed that RNA could have a chaperone function when interacting with the MA domain, preventing GAG from associating with unspecific lipid interfaces. Here, we study the interaction of MA with monolayer and bilayer membrane systems, focused on the specificity for PIP 2 and on the possible effects of a GAG N-terminal peptide to impair the binding for either RNA or membrane. We found that RNA decreases the kinetics of the protein association with lipid monolayers but without any effect on the selectivity for PIP 2 . Interestingly, for bilayer systems, this selectivity increases in presence of both the peptide and RNA, even for highly negative charged compositions, where MA by itself doesn’t discriminate between membranes with or without PIP 2 . Therefore, we propose that the specificity of MA for PIP 2 -membranes might be related to the electrostatic properties of both membrane and protein local environments, rather than a simple difference in molecular affinities. This scenario gives a new understanding of the regulation mechanism with a macromolecular view instead of considering molecular interactions within a ligand-receptor model. Importance HIV-1 virions are formed at the PM of infected cells through a direct interaction of the viral GAG protein with lipids. This is a finely regulated process governed by the GAG N-terminal matrix domain, MA. Here, we obtained compelling evidence on how this process depends on the local dielectric environments of both, the membrane and MA. Using bio-membrane mimicking systems, we found how MA myristoylation is involved in the interfacial absorption and anchoring of the protein, where the interaction with RNA negatively regulates this process but in a lesser extent when traces of the PIP 2 lipid are present. Additionally, an N-terminal GAG-derived peptide competes with MA for the nucleic acid binding and impair the protein-membrane interaction when PIP 2 is absent. All these data allowed us to propose a model for MA association with lipid interfaces and how it depends on oligonucleotide binding, lipid composition and competing peptide presence.