An integrative approach for modelling large protein complexes using mass spectrometry-based strategies and computational analyses.

Strategic Research Priority: Industrial biotechnology and bioenergy
Abstract
This is to integrate a novel combination of mass spectrometry (MS)-based experiments with protein assembly modelling and bioinformatics tools for studying the structural dynamics of transient protein complexes. Structural data generated from MS - native MS, ion mobility (IM)-MS and hydrogen deuterium exchange (HDX)-MS - will be encoded into modelling restraints and subsequently analysed by protein-protein interaction networks for building 3D protein assembly models. We will validate our method on a benchmark of heteromeric complexes with known crystal structures and use it to characterize the assembly dynamics of the recently discovered antiviral defense system, clustered regularly interspaced palindromic repeats (CRISPR) complex.


Project
Cells contain macromolecular complexes of physically interacting proteins. Characterizing the structural dynamics of these complexes remains a challenge using single methods. Hybrid methods, which integrate multiple types of biophysical data[1], have emerged as a powerful tool for studying the architectures of otherwise intractable complexes. The development and assessment of a hybrid method for characterizing dynamic complexes forms the basis of our multi-disciplinary research hypothesis.

The method builds upon a ground-breaking MS-based strategy for structural modelling of protein complexes[2]. The innovation is: a) the performance and incorporation of HDX-MS experiments into our hybrid method. These experiments report time-dependent changes of HDX rates within different regions of proteins, yielding information about their solvent accessibility. HDX-MS measurements will be corroborated with calculated solvent density maps from dynamical behavior of water on protein surface predicted by MD simulations[3]; b) the treatment of dynamic rearrangements of the individual subunits as they assemble into functional complexes. We will intersect an MS-guided Monte Carlo sampling with solvent MD simulations enabling us to monitor the conformational changes of assemblies over time scales comparable to HDX-MS. Finally, protein-protein interaction network analyses[4] will allow us to annotate and infer confidence to the function of biological complexes observed in our experiments.

Prior to applying our method in unknown biological complexes, we will assess its ability in reproducing near-native models in a benchmark of diverse complexes. Next, we will target the CRISPR, an RNA-based immune system responsible for eliminating genetic parasites, for which purified samples for MS experiments are available from our collaborator (Malcolm White, St. Andrews). Previously, we modelled the CRISPR-CSM complex using biochemical methods, electron microscopy, MS and modelling[5]. Here we will target the CRISPR-CMR from Sulfolobus solfataricus composed of seven proteins. The detailed structural and functional characterisation of CMR will reveal its role in targeting viral nucleic acids.