Reliable identification of protein-protein interactions by crosslinking mass spectrometry

Abstract

Crosslinking mass spectrometry is widening its scope from structural analyzes of purified multi-protein complexes towards systems-wide analyzes of protein-protein interactions. Assessing the error in these large datasets is currently a challenge. Using a controlled large-scale analysis of Escherichia coli cell lysate, we demonstrate a reliable false-discovery rate estimation procedure for protein-protein interactions identified by crosslinking mass spectrometry.Crosslinking mass spectrometry has become a key technology for structural biology by providing distance restraints on purified multi-protein complexes 1 . Proteins can also be crosslinked in complex mixtures. Charting protein-protein interactions (PPIs) in cell lysates, organelles or even whole cells has, therefore, become the next frontier for this technique 2-13 . To avoid reporting large numbers of spurious PPIs, it is important that the false discovery rate (FDR) of the reported interactions is reliably estimated. Doing this correctly is a challenge for the field 12-15 and to date, a consensus has not yet emerged (Table S1).

Retention Time Prediction Using Neural Networks Increases Identifications in Crosslinking Mass Spectrometry

“…Such multi-dimensional chromatography workflows can yield in the order of 10,000 CSM at 1-5% false discovery rate (FDR). [14][15][16][17] The identification of crosslinked peptides from spectra is however still challenged by the uneven fragmentation of the two peptides and the large search space that increase the odds of random matches. This is especially the case for heteromeric crosslinks as the size of their search space exceeds that of self-links, i.e.…”

“…This is especially the case for heteromeric crosslinks as the size of their search space exceeds that of self-links, i.e. links falling within a protein or homomer 16 . Typically, database search tools use the precursor mass and fragmentation spectrum for the identification of peptides to compute a single final score for each CSM.…”

Cleavable Cross-Linkers and Mass Spectrometry for the Ultimate Task of Profiling Protein–Protein Interaction Networks in Vivo

“…In particular, for nonexperts, it is still hard to know which enrichment technique, analysis software, and software settings to choose. However, recent efforts regarding selective enrichment strategies (e.g., refs ( 44 and 45 )), the capture of lower abundant proteins, 49 the harmonization of standards, 126 clever FDR control, 86 , 88 as well as the improved sensitivity of mass spectrometers and increased computational power show that within the next 2–5 years, scientists will truly be able to dig deeper than ever before in the interactome of cells, organelles, or tissues. As mentioned in the Applications and Bottlenecks section, the combination with FAIMS 79 or ion mobility (caps-PASEF 46 ) as an additional separation dimension will further alleviate issues in the dynamic range and therefore boost sensitivity toward cross-link detection in future studies.…”

“…As demonstrated by the Rappsilber group 31 this leads to a potentially much higher error for the actual cross-linked residue pairs of interest. As demonstrated in a preprint, 88 combining CSMs with unique cross-linked sites increases the actual FDR up to 47% (also dependent on the size of the search database used). It is demonstrated that a separate calculation of inter-, intra-, and monolinked spectra as well as the merging of CSMs to their respective protein–protein interaction (PPI) sites prior to the FDR calculation improve the reliability, leading to correct FDR estimations.…”

Toward Increased Reliability, Transparency, and Accessibility in Cross-linking Mass Spectrometry

“…Second, the search spaces of self and heteromeric cross-links are of different sizes. Therefore, the error of self and heteromeric links must be considered separately (Lenz et al, 2020;Walzthoeni et al, 2012). Third, for samples of limited complexity, when using only a small sequence database, and dependent on the cross-linking chemistry and data analysis strategy, there may be an insufficient number of decoy hits for an accurate determination of FDRs.…”