UC Berkeley

Clathrin-mediated endocytosis (CME) depends on plasma membrane remodeling, which results in production of an endocytic vesicle approximately 100 nm across. Remodeling of a flat patch of membrane into a spherical vesicle occurs below the resolution limit of the light microscope, making elucidation of the molecular mechanisms underlying this shape change difficult. To study how the cell generates nanoscale curvature for CME, I developed a high-throughput nanofabrication process that mimics endocytic membrane curvature, and focused on how this curvature affects both normal and perturbed CME. I began by producing a mold with nanoscale ridges (nanoridges) through electron-beam lithography and reactive ion etching, and characterized the mold to ensure suitable structures were formed. I then used the mold to optimize a UV-nanoimprint lithography method of stamp-and-lift substrate manufacturing, allowing the production of one glass-like substrate every ~15 minutes. I characterized the substrates and verified that they induce a cellular response by reliably bending the ventral cell membrane into curved shapes and by reorganizing the endocytic machinery along the sites of high induced curvature.To determine how induced plasma membrane curvature affects endocytosis, I screened genome-edited cells expressing fluorescent CME proteins or fluorescent membrane markers and developed computational tools to quantify the enrichment of endocytic proteins on curved membranes, finding that proteins across stages of CME are strongly enriched by the highest curvature induced by our substrates (75 nm ridge diameter). I then characterized the response of endocytic sites to induced curvature, finding an increase in endocytic lifetimes with high curvature that decreases as the curvature decreases, as well as an increase in mean-square displacement of endocytic sites on nanoridges specifically at late stages of CME, while the fluorescence profiles of CME proteins on nanoridges were indistinguishable from control cells. Finally, I chose to examine how protein disruptions affected endocytosis on flat Ormocomp substrates, and whether induced curvature might rescue the resulting perturbations to CME. I found that induced curvature did not rescue CME from AP2 or FCHo1/2 knockdown, indicating that stable curvature cannot bypass nucleation or early stabilization of cargo-adaptor interactions. However, AP2 localizes to nanoridges even in the absence of FCHo1/2, thought to be key stabilizers of AP2, evidence that AP2 has some intrinsic affinity for high membrane curvature. I found that induced curvature rescues endocytic site localization after clathrin disruption, evidence that clathrin’s essential function in CME is curvature stabilization. I further characterized these clathrin-knockdown sites on regions of induced membrane curvature and found that the sites turn over with a nearly identical fluorescence intensity profiles to sites in control cells, albeit with a longer lifetime, and found that endocytic cargo uptake was partially rescued by induced curvature in clathrin knockdown cells. With these data, I put forward a model in which clathrin’s key role at an endocytic site is to stabilize the evolving curvature, but that with induced curvature, an endocytic site may proceed to scission without a coat. Indeed, I found evidence through electron microscopy for formation of exactly such vesicles in clathrin-knockdown cells grown on nanofabricated ridges. These data demonstrate the fundamental importance of membrane shape in control of cell biological processes, examples of which are predicted to include endocytosis of ECM proteins and viral particles.