Assigning Peptide Structure from Ion-Mobility Mass Spectrometry Collision Cross Section Data

The ion mobility (IM) technique coupled with traditional mass spectrometry (IM-MS) has introduced a practical tool for the characterization of peptide analyte ions by exploiting the difference in their collision cross section (CCS) values. CCS holds molecular level information that can be used to computationally assign the conformation(s) of a molecular system. However, a reliable and accurate method for peptide structure prediction remains a challenge because peptides exhibit dynamic gas-phase intramolecular interactions, and hence, the methods used will need to account for this. In this work, we systematically assessed the performance of a computational workflow involving both classical and density functional theory (DFT) steps to elucidate the peptide structure. Not unexpectedly, extensive enumeration of available peptide conformations was critical to obtain high-quality results. Due to the size of the systems studied and the large numbers of conformers that needed to be optimized, we initially chose the D3-B3LYP/6-31G(d) level of theory and obtained good agreement between experimental and computed CCS values. However, in several cases, suboptimal accuracies were observed, but we found that increasing the basis set used to 6-31G(d,p) was able to improve our agreement with experiment. Altogether, we demonstrated that accurate peptide structure assignment is achievable with adequate sampling of the conformational space and using the appropriate quantum mechanical level of theory to account for intramolecular interactions in the gas phase.