The matrix protein M1 of the Influenza A virus is considered to mediate viral assembly and budding at the plasma membrane (PM) of infected cells. In order for a new viral particle to form, the PM lipid bilayer has to bend into a vesicle towards the extracellular side. Studies in cellular models have proposed that different viral proteins might be responsible for inducing membrane curvature in this context (including M1), but a clear consensus has not been reached. In this study, we use a combination of fluorescence microscopy, cryogenic transmission electron microscopy (cryo-TEM), cryo-electron tomography (cryo-ET) and scanning fluorescence correlation spectroscopy (sFCS) to investigate M1-induced membrane deformation in biophysical models of the PM. Our results indicate that M1 is indeed capable to cause membrane curvature in lipid bilayers containing negatively-charged lipids, in the absence of other viral components. Furthermore, we prove that simple protein binding is not sufficient to induce membrane restructuring. Rather, it appears that stable M1-M1 interactions and multimer formation are required in order to alter the bilayer threedimensional structure, through the formation of a protein scaffold. Finally, our results suggest that, in a physiological context, M1-induced membrane deformation might be modulated by the initial bilayer curvature and the lateral organization of membrane components (i.e. the presence of lipid domains). FIGURE 4: M1 layer interacting with the membrane is stable even after lipid removal. A-B: LSM confocal image of a typical GUV composed of DOPC:cholesterol:DOPS 50:20:30, after incubation with 5 ”M M1-Alexa488 (green channel, panel A). The lipid bilayer is visualized via the addition of 0.05 mol% Rhodamine-DOPE (red channel, panel B). C-D: LSM confocal image of a typical GUV after treatment for 5 min with detergent (e.g. Triton X-100 1.7 mM). Panel C represent the signal from M1-Alexa488. Panel D represents the signal from Rhodamine-DOPE. The excitation laser power used to acquire the image shown in panel D was ~10 times higher than the power used to acquire the image shown in panel B. GUVs contained 150 mM sucrose in their lumen and were suspended in a phosphate buffered protein solution (pH 7.4) with similar osmolarity (see Materials and Methods). Scale bars are 5 ”M. Images were acquired at 23°C.