Endophilin recruitment drives membrane curvature generation through coincidence detection of GPCR loop interactions and negative lipid charge.

Endophilin plays key roles during endocytosis of cellular receptors, including generating membrane curvature to drive internalization. Electrostatic interactions between endophilin's BIN/Amphiphysin/Rvs domain and anionic membrane lipids have been considered the major driving force in curvature generation. However, the SH3 domain of endophilin also interacts with the proline-rich third intracellular loop (TIL) of various G-protein-coupled receptors (GPCRs), and it is unclear whether this interaction has a direct role in generating membrane curvature during endocytosis. To examine this, we designed model membranes with a membrane density of 1400 receptors per µm2 represented by a covalently conjugated TIL region from the ß1-adrenergic receptor. We observed that TIL recruits endophilin to membranes composed of 95 mol% of zwitterionic lipids via the SH3 domain. More importantly, endophilin recruited via TIL tubulates vesicles and gets sorted onto highly curved membrane tubules. These observations indicate that the cellular membrane bending and curvature sensing activities of endophilin can be facilitated through detection of the TIL of activated GPCRs in addition to binding to anionic lipids. Furthermore, we show that TIL electrostatically interacts with membranes composed of anionic lipids. Therefore, anionic lipids can modulate TIL/SH3 domain binding. Overall, our findings imply that an interplay between TIL, charged membrane lipids, BAR domain, and SH3 domain could exist in the biological system and that these components may act in coordination to regulate the internalization of cellular receptors.

Fluorescence imaging of lamellipodin-mediated biomolecular condensates on solid supported lipid bilayer membranes.

Biomolecular condensates play a major role in numerous cellular processes, including several that occur on the surface of lipid bilayer membranes. There is increasing evidence that cellular membrane trafficking phenomena, including the internalization of the plasma membrane through endocytosis, are mediated by multivalent protein-protein interactions that can lead to phase separation. We have recently found that proteins involved in the clathrin-independent endocytic pathway named Fast Endophilin Mediated Endocytosis can undergo liquid-liquid phase separation (LLPS) in solution and on lipid bilayer membranes. Here, the protein solution concentrations required for phase separation to be observed are significantly smaller compared to those required for phase separation in solution. LLPS is challenging to systematically characterize in cellular systems in general, and on biological membranes in particular. Model membrane approaches are more suitable for this purpose as they allow for precise control over the nature and amount of the components present in a mixture. Here we describe a method that enables the imaging of LLPS domain formation on solid supported lipid bilayers. These allow for facile imaging, provide long-term stability, and avoid clustering of vesicles and vesicle-attached features (such as buds and tethers) in the presence of multi-valent membrane interacting proteins.

Universal protein misfolding intermediates can bypass the proteostasis network and remain soluble and less functional.

Some misfolded protein conformations can bypass proteostasis machinery and remain soluble in vivo. This is an unexpected observation, as cellular quality control mechanisms should remove misfolded proteins. Three questions, then, are: how do long-lived, soluble, misfolded proteins bypass proteostasis? How widespread are such misfolded states? And how long do they persist? We address these questions using coarse-grain molecular dynamics simulations of the synthesis, termination, and post-translational dynamics of a representative set of cytosolic E. coli proteins. We predict that half of proteins exhibit misfolded subpopulations that bypass molecular chaperones, avoid aggregation, and will not be rapidly degraded, with some misfolded states persisting for months or longer. The surface properties of these misfolded states are native-like, suggesting they will remain soluble, while self-entanglements make them long-lived kinetic traps. In terms of function, we predict that one-third of proteins can misfold into soluble less-functional states. For the heavily entangled protein glycerol-3-phosphate dehydrogenase, limited-proteolysis mass spectrometry experiments interrogating misfolded conformations of the protein are consistent with the structural changes predicted by our simulations. These results therefore provide an explanation for how proteins can misfold into soluble conformations with reduced functionality that can bypass proteostasis, and indicate, unexpectedly, this may be a wide-spread phenomenon.