Cross-linked Peptide Data Obtained by Ion Mobility used to Sort Correct/Incorrect in silico Protein Conformation Predictions

摘要

Definition of the atomic interfaces in protein-protein complexes often depends on use of physical techniques like chemical cross-linking and mass spectrometry. Typical workflows include: chemically cross-linking two or more proteins, digesting proteins to peptides, proteolysis in H2O16 and H2O18 followed by MS analysis to detect and identify cross-linked peptides as we recently published (Gao et al. JBC & Anal Chem 2006). As with detection and sequencing of phosphopeptides, problematic to this approach is that cross-linked peptides are normally at low stoichiometry relative to all other peptides in the mixture and are thus very difficult to find. We investigated the ability of a new commercial ion mobility instrument to detect and identify cross-linked peptides. We also investigated the utility of using MS identified cystine-containing peptides to sort correct from incorrect in silico protein structure predictions made using tools like Rosetta from Professor David Baker at the University of Washington.