Proteomics systems for integrative structural biology

点击翻译按钮获取中文摘要

We present a program of research designed to provide new tools for characterizing and modeling the structure of complex protein systems. Our strategy for generating structure-rich biophysical data involves condensed-phase protein chemistries that code protein properties into quantities measurable by mass spectrometry, at all organizational scales (structural proteomics). There are four interrelated themes in the proposed program. First, we will investigate new protein bioconjugation strategies that involve fast-acting, light-triggered chemical reactions, and develop next-generation reagents for both covalent labeling mass spectrometry (CL-MS) and crosslinking mass spectrometry (XL-MS). We propose that separating the equilibrium protein solvation properties of these reagents from the photolytic process will provide unprecedented opportunities to increase the structural value of modeling restraints derived from both CL-MS (topographical mapping) and XL-MS (distance measurements). Second, we present concepts for implementing protein bioconjugation methods in complex mixtures and cells, to avoid reconstituting protein complexes. Reconstitution is a laborious and failure-prone process, thus overcoming this current requirement is critical step towards extending structural methods into a cellular context. The work in this theme will involve the production fast-isolation methods for removing complexes from in vitro translation media and cell lysates, as well as the development of in situ, proximity-based crosslinking methods. Third, the richness of structural data emerging from these activities will tax existing MS-based informatics, requiring new scoring algorithms, file management and reporting structures. We will augment our Mass Spec Studio software architecture with new plug-ins to accommodate the extraction of restraint data. All plug-ins will be made publicly available, and we will open-source the architecture for community participation. Fourth, new strategies will be developed, together with collaborators, that better utilize MS-generated structural data for accurately modeling large, complex protein systems. Current MS-based methods are rudimentary. We propose a strategy whereby structural proteomics data are used to transform free protein structures into their bound form, to enable a more accurate docking experiment further constrained by data. Routines will be built within our Studio framework to interface with key structure-modeling compute systems, in order to generate the first ever comprehensive data-to-structure resource in structural proteomics. Our research program will take an important step towards extending structural biology into the complex milieu in which proteins function. Progress will improve our understanding of biological mechanisms, and stimulate new strategies for the development of therapeutics.