The Daft Approach for in Silico Prediction of Transmembrane Peptide Complexes

Abstract

Transmembrane helix-helix interactions play a vital role in signaling, and the details of the complexes involved are a valuable complement to biochemical studies. Unfortunately, it is difficult to obtain structural models experimentally, and computational approaches such as protein-protein docking are problematic due to the membrane environment. Here we present an innovative method that combines coarse-grained molecular dynamics simulations, without biasing potentials, with a docking approach to quickly sample the interaction energy landscape. Starting structures are set up using schemes specific for the number of components, promoting encounters, and processed using a fully automated MARTINI-based workflow, facilitating high-throughput simulations.The method has been applied to ErbB1/ErbB2 homo- and heterodimers, validating the results against available experimental data. In addition, the method was used to investigate temperature controlled signaling by the DesK derived minimal sensor, and to provide a view on the interplay of TM dimers and trimers involved in TNF receptor activation. The results give insight in the mechanisms underlying signaling by simple transmembrane helix systems. This insight offers opportunities for the design of custom membrane-based sensor systems and for assessment of signaling by more complicated helix based systems.View Large Image | View Hi-Res Image | Download PowerPoint Slide