Most of the biological processes that maintain cellular life depend on habitually interacting proteins, which are forming complexes that correlate with the type of biological processes they are involved in. In order to understand these processes, X-ray crystallography, cryogenic electron microscopy/tomography and mass spectrometry are commonly used to structurally characterize macromolecular protein assemblies. As briefly reviewed in Chapter 1, each of these methods has benefits and drawbacks associated, however most of the methods predominantly require a highly purified sample. The downside being that macromolecular protein assemblies are, up to more recently, rarely studied in their naïve cellular environment.
In this thesis, the versatility of cross-linking mass-spectrometry (XL-MS), to characterize protein assemblies in vivo is demonstrated.
We show that especially in mitochondria, for which cryo-ET identification is limited to the biggest or most "uniquely-shaped" complexes (Bauerlein and Baumeister, 2021) such as the ATP synthase XL-MS enables the sensitive identification of novel mitochondrial protein complexes.
Although XL-MS enables the characterization of protein complexes in their naïve cellular environment, established XL-MS workflows present drawbacks that significantly hamper an adequate structural characterization. The presence of coinciding respiratory chain protein assemblies in mitochondria (Wittig et al., 2006) for instance, prevent the confident identification of a cross-links origin and thereby an accurate structural characterization of a specific assembly. To tackle existing challenges of classical in-solution XL-MS, we therefore set out to develop a novel, easy-to use workflow (Chapter 2) termed in-gel cross-linking mass spectrometry (IGX-MS). IGX-MS circumvents time and sample consuming steps, and further provides a sensible solutions for differentiating cross-links obtained from co-occurring protein oligomers or complexes, leading up to an improved structural characterization.
Further, by combining XL-MS with complementary structural techniques, e.g. complexome profiling (CP-MS) or cryo-ET enabled us to identify and structurally characterize novel protein complexes in mitochondria (Chapter 3-5). By performing XL-MS and CP-MS for the structural analysis of macromolecular protein assemblies in bovine heart mitochondria (Chapter 3), we uncover that a substantial amount of dimeric apoptosis factor 1 (AIFM1) is associated with at least 10% of monomeric cytochrome c oxidase (COX).
In a collaborative effort with the Zeev-Ben-Mordehai group (Utrecht University), we highlight that ambiguity in cryo-ET experiments can be overcome by additionally performing XL-MS experiments (Chapter 4). Using sub-tomographic averaging, we identified that in mammalian sperm, mitochondria are wrapped around the flagellar cytoskeleton, which is mediated through conserved arrays of glycerol kinase-like proteins.
Finally yet importantly, in Chapter 5 we describe how XL-MS can be utilized for the characterization of protein complexes using computer-aided structural modeling. Like described in Chapter 3, XL-MS and CP-MS data is used for the detailed characterization of mitochondrial protein assemblies, providing essential information regarding complex members and assembly state prior to structural modeling. Combining data from both methods enabled us to identify MRPS36 as a novel component of the 2-oxoglutarate dehydrogenase complex (OGDHC), and subsequently was utilized to refine the AI-driven structural model. The final model provides new insights into the topology of this multi-component enzyme as well as details on its enzymatic function.