Poster Presentation HUPO 2019 - 18th Human Proteome Organization World Congress

Targeted validation and quantification of cross-links from endogenous protein complexes   (#516)

Marco Faini 1 , Emanuela Milani 1 , Charlotte Nicod 1 , Alexander Leitner 1 , Michael Ewing 1 , Audrey Van Drogen 1 , Ludovic Gillet 1 , Bernd Wollscheid 1 , Rudolf Aebersold 1
  1. ETH Zurich, Zurich, ZURICH, Switzerland

Background

Protein complexes represent the functional modules of the cell. Studying their assembly and structure in a close to native environment is fundamental to understanding cellular processes and their dynamics. Affinity purification mass spectrometry (AP-MS) provides bait interactors but requires multiple experiments to attribute high interaction confidence. Conversely, cross-linking mass spectrometry (XL-MS) can pinpoint protein neighbourhood and structural restraints between single amino-acid residues but requires large sample amounts and high purity. We combined the speed and throughput of AP-MS with the topological and structural information of XL-MS in a quantitative method. 

Methodologies

We developed qAXL-MS, combining affinity purification from tagged human cell lines with on-beads cross-linking. After identification by spectral matching, we validate and quantify cross-linked peptides by targeted proteomics (PRM). Similarly to AP-MS, the protocol is fast, sensitive and amenable to high throughput. We implemented a data analysis pipeline in an open access R package.

Findings

First, we validated qAXL-MS on the TriC chaperonin purified from HEK293 cells. We could identify almost 3 times more cross-links than previous methods, including cross-links specific for each of the open and closed chaperonin conformation. Next, we characterized Cul4A complexes that facilitate ubiquitination. Cul4A alternates between a resting state, interacting with CAND1, and a primed state that requires neddylation. Comparing cross-links abundance between the resting and de-neddylated conditions we could confirm previous physical interactions and structural models of the two neddylation states. The accurate multi-level quantitation enabled the discrimination between compositionally and conformationally derived links. We could thus describe a set of cross-links diagnostic for several conformational changes in DDB1, CAND1 and the COP9 signalosome.

Conclusions

We developed a quantitative method to define, for a given protein complex, interaction topologies, physical interactions, distance restraints and structural markers from different cellular states from as little as ten million human cells.