The plasma membrane (PM) of eukaryotic cells consists of a crowded environment comprised of a high diversity of proteins in a complex lipid matrix. The lateral organization of membrane proteins in the PM is closely correlated with biological functions such as endocytosis, membrane budding and other processes which involve protein mediated shaping of the membrane into highly curved structures. Annexin A4 (ANXA4) is a prominent player in a number of biological functions including PM repair. Its binding to membranes is activated by Ca2+ influx and it is therefore rapidly recruited to the cell surface near rupture sites where Ca2+ influx takes place. However, the free edges near rupture sites can easily bend into complex curvatures and hence may accelerate recruitment of curvature sensing proteins to facilitate rapid membrane repair. To analyze the curvature sensing behavior of curvature inducing proteins in crowded membranes, we quantifify the affinity of ANXA4 monomers and trimers for high membrane curvatures by extracting membrane nanotubes from giant PM vesicles (GPMVs). ANXA4 is found to be a sensor of negative membrane curvatures. Multiscale simulations, in which we extract molecular information from atomistic scale simulations as input to our macroscopic scale simulations, furthermore predicted that ANXA4 trimers generate membrane curvature upon binding and have an affinity for highly curved membrane regions only within a well defined membrane curvature window. Our results indicate that curvature sensing and mobility of ANXA4 depend on the trimer structure of ANXA4 which could provide new biophysical insight into the role of ANXA4 in membrane repair and other biological processes. This journal is
|Status||Udgivet - 22. jan. 2021|
Bibliografisk noteFunding Information:
This work is financially supported by Danish Council for Independent Research, Natural Sciences (DFF-4181-00196), by a Novo Nordisk Foundation Interdisciplinary Synergy Program 2018 (NNF18OC0034936), by the Scientific Committee Danish Cancer Society (R90-A5847-14-S2) and by the Lundbeck Foundation (R218-2016-534). The aaMD simulations were carried out on the ICHEC Kay supercomputer under the DECI-15 PRACE award ANNEX, and on the Danish e-Infrastructure Cooperation (DeiC) National HPC Center, on ABACUS 2.0 at the University of Southern Denmark, SDU.
© 2021 The Royal Society of Chemistry.
Copyright 2021 Elsevier B.V., All rights reserved.