Predicting Integral Membrane Protein Expression from Multiscale Simulation Model of Translocon-Mediated Membrane Integration

点击翻译按钮获取中文摘要

 DESCRIPTION (provided by applicant): Integral membrane proteins embed within the cellular membrane and perform crucial functions to modulate cellular transport, signaling, and anchoring. Although 20-30% of the proteins in the human proteome are membrane proteins, they represent less than 1% of the crystal structures in the Protein Data Bank despite of their biological importance. Increasing the number of solved membrane protein structures is extremely important to understand protein function, understand disease consequences due to protein misassembly, and increase potential drug targets. However, membrane proteins are insoluble in water and as a result their structural characterization is greatly inhibited by their inability to be reliably crystallized outside of the membrane environment. It is thus necessary to find novel methods to overexpress membrane proteins in order to facilitate structure determination. The hypothesis of this project is that protein expression can be increased by enhancing the efficiency with which membrane proteins are integrated into the bilayer during translation. In both prokaryotic and eukaryotic cells, proteins are inserted into the membrane during translation by the Sec translocon protein-conducting channel. In eukaryotes, the ribosome docks to the cytoplasmic opening of the channel and transfers a nascent polypeptide chain into the translocon during translation. Depending on the recognition of particular signals, the protein is directed to either integrate with the membrane or translocate through it. However, correctly directing nascent peptides to one of these outcomes depends on both thermodynamic and kinetic factors and as a result proteins are frequently sorted incorrectly or misassembled. If protein expression requires the correct integration of nascent membrane proteins, then designing protein mutants that enhance integration efficiency should increase expression yields by correcting errors in translocon-mediated insertion. We propose a hierarchical, multiscale simulation strategy to model the full pathway of co-translational membrane protein integration. The goals of this work are to: 1) predict experimentally-measurable protein expression yields using a coarse-grained model of membrane integration and 2) engineer mutants to optimize protein expression. To accomplish these goals, we will optimize a recently developed coarse-grained model that is able to reach the minute-long timescales necessary for quantifying integration efficiency. Results of this model will be compared to protein expression measurements in E. coli using protein homologs of different expression efficiency. Comparisons between simulation and experiments will be used to identify specific protein features that enhance expression. Iteratively refining the model and incorporating additional system properties, such as lipid composition and protein structure, will improve these quantitative predictions. These findings will have a broad impact on structural biology, drug targeting, and membrane protein characterization by significantly reducing the barrier to membrane protein crystallization.