Nano to Meso and Back Again: Capturing and Exploiting Dynamic Heterogeneities in Biological Membranes via Large Scale Simulations

Membranes are of fundamental importance to biology: they form the 'surrounding walls' of all cells, and control the flow of materials and information into and out of cells. The proteins of membranes are embedded in a lipid bilayer, which is an oil-like film separating the cellular contents from its watery external environment. These membrane proteins are responsible for many key biological functions, including communication between cells, and consequently are targets for many drugs, including some antibiotics. However, membranes are more than a simple 'sum of their parts'. Rather, they are complex and dynamic assemblies of proteins and lipids, and their biological functions are emergent properties of the membrane system as a whole. The overall aim of this proposal is to develop methods to understand how the various components of membranes are dynamically organized so as to carry out their biological functions.

Computer simulations of the interacting molecules (proteins and lipids) which make up cell membranes may be used to study their dynamic behaviour and organization. Simulation studies to date have been limited to the very small (nano)scale, and have focused on e.g. the interactions of single membrane proteins with immediately neighbouring lipid molecules. However, larger scale simulations are needed to address (sub)cellular scales of the dynamic organization of cell membranes. These larger scale simulations will enable direct comparison of predictions emerging from simulations with experimental observations. We will develop larger scale simulations using 'meso' scale models which can be built using the outcomes of smaller scale models and simulations which have already been successfully employed to study isolated membrane proteins.

We will use advanced computer simulations, benefiting from UK investment in national supercomputing facilities, to:
1. Develop biologically realistic models of dynamic spatial heterogeneities in cell membranes; and
2. Link smaller scale (coarse-grained) and larger scale (meso) models with experimental data in an integrated approach allowing us to bridge between structural and cellular studies of membranes.

Thus, the overall outcome of these studies will be a powerful new computer simulation method for biological membranes, which will have applications relevant to the pharmaceutical and biotechnology sectors.

Membranes are heterogeneous environments, even at the nanoscale, and these heterogeneities undergo dynamic fluctuations over a range of timescales. Since membranes play a key role in processes such as drug permeation and as components of biosensors, we need to understand the impact of the dynamic nanoscale heterogeneities presented by cell membranes.

Coarse-grained (CG) molecular dynamics (MD) simulations are of central importance to understanding the dynamic organisation of membranes and their constituent proteins. Such simulations typically encompass time scales of tens of microseconds and length scales of tens of nanometers. However, experimental techniques, such as fluorescence microscopy, typically track single proteins on much larger time and lengthscales (seconds and micrometres).

In this project we propose the development of a 'meso' model, parameterized according to output from conventional CG-MD simulations (such as protein diffusion, protein clustering behaviour and bilayer undulations). Together, atomistic, CG and meso model simulations will be used in a serial multiscale approach to develop biologically realistic models of membranes bridging from the nano to the meso scale, akin to the zooming in and out of a microscope. With the increase in length and timescale afforded by the meso model simulations we will be able to directly link our simulations to single molecule tracking data, providing a powerful and novel approach to the interpretation of such experimental observations.
We aim to deliver outcomes in three main areas: (1) clustering and functional activity and cooperativity of receptors and ion channels; (2) drug permeation through spatially complex membranes; and (3) fluctuating membranes on surfaces for biosensor applications.

The development of this integrated, multiscale approach will have important implications both for our understanding of membrane biology and for pharmaceutical, bioctehnological and healthcare applications.

This project will have broad social and economic benefits. These will arise primarily from two healthcare related areas: (i) improvements in the rational design of drugs based on an improved physical mechanistic understanding of how they permeate through membranes; and (ii) advances in exploitation of membrane proteins in biosensors. It is estimated that ca. 40% of drug targets correspond to membrane proteins. Thus, the principal beneficiaries will be the pharmaceutical and healthcare sectors.

One of the main outcomes of this research will be improved computational approaches to predicting drug permeation across membranes. This will aid researchers in the pharmaceutical industry in understanding the pharmacokinetics of novel compounds.

A second major outcome will be an enhanced understanding of how complex membranes interact with solid support surfaces. This will aid design of novel biosensors based on membrane proteins. Given the importance of membrane proteins in e.g. nanopore based technologies, and the expertise of the UK in graphene technologies, helping to bring these two aspects together has considerable economic potential for the future.

The project will expose a postdoctoral researcher to a combination of cutting edge computational and theoretical methods and their applications, enabling her to contribute more effectively to the wider economy in the future. In this context, the project will help to build the UK capacity in successful applications of HPC for biology. This is an area which is currently growing as evidenced by e.g. the current large-scale collaboration in the USA between DoE computational researchers and NCI/NIH biologists in one of the high profile 'cancer moonshot' projects. Via our ongoing outreach activities we will use this project to highlight to the general public how large scale computing resources can be focussed on real world biological problems.