Multiscale Modeling of Membrane Remodeling by the Protein Epsin

The protein epsin plays a key role in both sensing and inducing curvature during clathrin-mediated endocytosis (CME), affecting a multitude of downstream cellular signals. Both the constitutive activation of cell surface receptors regulated by CME, including epidermal growth factor receptor (EGFR), and defects in the endocytic machinery itself may be oncogenic. Therefore, treatment of the resulting cancers requires molecular resolution of endocytosis. To this end, we employ multiscale modeling techniques to study the process by which the epsin N-terminal homology domain (ENTH) remodels the cell membrane. Coarse-grained molecular dynamics (CGMD) simulations under the MARTINI force field [1] capture the mechanism by which a single ENTH domain induces curvature in a large patch of the lipid bilayer. Calculation of the three-dimensional stress tensor [2] reveals both the distribution of stresses in the bilayer, and characterizes the curvature field induced by ENTH. Simulations with interacting pairs of ENTH in various configurations characterize the additivity of their curvature fields, and inform continuum mechanics models for vesicle budding [3]. To parameterize the coarse-grained model for epsin-membrane interactions, we must turn to more detailed all-atom simulations of both the protein-membrane dynamics and the action of a key lipid binding partner, phosphatidylinositol (4,5)- bisphosphate (PIP2), which associates with epsin's embedded helix. All-atom simulations of mixed lipid bilayers containing PIP2 under the CHARMM36 force field [4] provide structural and dynamic properties of PIP2 for use in CGMD simulations. Likewise, all-atom simulations of ENTH parameterize a heterogeneous elastic network model which reproduces the dynamics of ENTH embedded in the membrane. By imposing a variety of restraints on our model system, we probe the roles of membrane surface tension, interactions of multiple epsins, and association with PIP2. These methods apply equally well to other membrane-remodeling proteins, including the proteins N-BAR, I-BAR, and exo70, which are thought to generate either positive or negative curvature with distinct mechanisms. In each case, the flow of information between CGMD and AAMD levels explains how molecular interactions are translated to larger length scales.

[1] Marrink, S. J.; Risselada, H. J.; Yefimov, S.; Tieleman, D. P. & de Vries, A. H. The MARTINI Force Field: Coarse Grained Model for Biomolecular Simulations, Journal of Physical Chemistry B, 2007, 111, 7812-7824.

[2] Ollila, O. H. S.; Risselada, H. J.; Louhivuori, M.; Lindahl, E.; Vattulainen, I. & Marrink, S. J. 3D Pressure Field in Lipid Membranes and Membrane-Protein Complexes Physical Review Letters, American Physical Society, 2009, 102, 078101.

[3] Liu, J.; Tourdot, R.; Ramanan, V.; Agrawal, N. J. & Radhakrishanan, R. Mesoscale simulations of curvature-inducing protein partitioning on lipid bilayer membranes in the presence of mean curvature fields, Molecular Physics, 2012, In Press.

[4] Bjelkmar, P.; Larsson, P.; Cuendet, M. A.; Hess, B. & Lindahl, E. Implementation of the CHARMM Force Field in GROMACS: Analysis of Protein Stability Effects from Correction Maps, Virtual Interaction Sites, and Water Models, Journal of Chemical Theory and Computation, 2010, 6, 459-466.

Figure. Snapshot of CGMD simulation of ENTH domain embedded in a mixed DOPC(blue)-DOPS(purple) bilayer with PIP2 (green) clusters (water not pictured).