Session: Parallel session 2 - Interactomics

Cross-Linking Mass Spectrometry for Investigating Protein Structure and Protein–Protein Interactions with Success

The past decade has seen cross-linking mass spectrometry (XL-MS) rapidly evolve from a technique that is applied on mainly purified proteins or their complexes, to large scale proteome-wide studies on complex samples such as lysates, to in vivo cross-linking on bacteria cells. The application of XL-MS on a large-scale non-purified samples has allowed for the unbiased determination of protein interaction partners as there are no prior purification steps that can eliminate many dynamic or transient interactions.

Despite its many advantages, XL-MS does have its challenges. Perhaps the biggest challenge that is faced in this technique is the abundance of cross-linked peptides. Because a careful amount of cross-linker is used as to not over-cross-link a system, the majority of peptides obtained after the digestion step does not contain the cross-linker and are not useful for analysis. Consequently, it is has become common practice to introduce an enrichment step of the cross-linked peptides before MS analysis. To this aim, our lab introduced a few years ago a novel bio-orthogonal enrichable cross-linker called NNP91. Building on this new reagent, we developed an in vivo XL-MS strategy to study bacterial protein-protein interactions with a focus on virulence systems. We used Neisseria meningitidis, a gram-negative bacterium which strictly infects humans, as our model organism. Specifically, we were interested in gaining interaction and structure information on its type IV piliation (T4P) machinery which is a large and highly dynamic complex spanning both the inner and outer membrane. Although the data obtained were in general of excellent quality2, the presence of highly abundant cytosolic proteins led to only a scarce description of the piliation machinery. Because this issue is frequently encountered by researchers turning towards XL-MS to probe their complexes of interest, we got interested in defining, for the first time, a clear cut-off rule in protein abundance as to what is considered “abundant enough” for good XL-MS results. We show that a simple iBAQ analysis performed from trypsin digest data can provide a good understanding if proteins of interest are abundant enough to achieve successful cross-linking data3. Comparing our findings to large-scale data on diverse system from several other groups, we show that proteins of interest should be at least in the top 20 % abundance range to expect more than one cross-link found per protein. We foresee that this guideline is a good starting point for researchers who would like to use XL-MS to study their protein of interest and help ensure a successful cross-linking experiment from the beginning.