Membrane Protein Structure Determination Through Assembly of Trans-Membrane Helices Guided by Limited Experimental Data

Abstract

We present BCL::Fold, a new computational algorithm that rapidly predicts the topology of helical membrane proteins through assembly of secondary structure elements. A symmetry folding mode allows for the prediction of homomultimers. Final models are evaluated using a knowledge-based energy potential. In a benchmark test of 40 proteins with up to 1200 residues, the method samples the correct topology in 39 cases. We will demonstrate the combination of this algorithm with limited NMR, EPR, and cryo-EM data for membrane protein structure determination. Although great progress in determining the structure of integral, helical, multi-span membrane proteins was made in recent years, only 10% of the expected 1000 membrane protein folds are represented by at least one example in the protein databank. For the 900 membrane proteins topologies that remain to be determined, billions of topologies are possible with an increasing number of trans-membrane spans. This large structural space is contrasted by experimental datasets that - if available at all - are limited in resolution of density map or number of structural restraints. It is the objective of BCL::Fold to enable and accelerate structure determination for these proteins.