Advancing Crosslinking Mass Spectrometry as a tool for deciphering the spatial organization of protein assemblies: from molecular machines to biopolymers

The focus of this thesis is Crosslinking Mass Spectrometry (XL-MS) — a promising approach which received a lot of traction in recent years. This method allows capturing protein sites located in close proximity without the disturbing of protein native conformation. Here I describe our recent developments in the methodology together with XL-MS application to the complex biological systems. Chapter 1 contains a general overview of the Biomolecular MS and Proteomics over the years and covers its establishing as a robust and efficient method in protein characterization. With the focus on bottom-up proteomics, the chapter introduces instrumentation and main LC-MS/MS workflows which are routinely applied these days. Optimization of the protocol for crosslinking of proteins, proteins complexes, and system-wide crosslinking experiments is available in Chapter 2 and Appendix to this thesis. The approach is exemplified on two model systems: BSA protein as a low-complexity sample and whole PC9 cell lysate as a complex sample. In the case of BSA, the application of a set of proteases for protein digestion is described as a step towards more identified crosslinks. For the whole cell lysate experiment, the chapter reports the detection of thousands of unique crosslinked peptide pairs within a single experiment. Additionally, the detailed description of the crosslinking data analysis steps which are done with XlinkX node incorporated into Proteome Discoverer software suite together with the When applied to the complex biological systems, crosslinking outputs become not trivial in terms of data analysis and visualization. Chapter 3 describes the development and application of Cross-ID, the platform for analysis and visualization of complex XL-MS outputs. It describes the analysis of the proteome-wide crosslinking output generated in Chapter 2 for PC9 cells. Different visualization approaches, validation, and integration into DisVis pipeline for restraints-based interaction interface prediction are highlighted. After we have established and optimized the necessary steps in the crosslinking pipeline, we have started to investigate biological samples of interest. In Chapter 4 we describe the molecular architecture of the type I-F CRISPR adaptation complex comprised of Cas2-32 and Cas14. Integrative structural biology approach allows to correctly place the complex subunits and decipher interaction interface of the protein complex. First, complex mass and its stoichiometry are defined with SDS-PAGE, SEC and Native MS. Then by XL-MS with DSSO reagent, protein sites located close to each other are pinpointed. On the next step, XL-MS with formaldehyde reagent highlights the DNA-protein interaction sites and the final structure of the Cas2-34:Cas14-DNA complex is solved. Chapter 5 focuses on the XL-MS guided modeling and docking of the subunits of fibrin biological biopolymers. The high number of identified distance restraints within this extremely complex sample provides the basis for the modeling of the structures of the heparin-binding domain of β-fibrinogen, interactive and RGD-containing domains of α-fibrinogen as well as the α220-249 region. This chapter for the first time describes the model of the laterally aggregated Fibrin clot generated with the experimental data.