Observed peptide pI and retention time shifts as a result of post-translational modifications in multidimensional...

Modified peptides constitute a sub-population among the tryptic peptides analyzed in LC-MS based shotgun proteomics experiments. For larger proteomes including the human proteome, the tryptic peptide pool is very large, which necessitates some form of sample fractionation. By carefully choosing the sample fractionation and separation methods applied as shown here for the combination of narrow-range immobilized pH gradient isoelectric focusing (IPG-IEF) and nanoUPLC-MS, significantly increased information content can be achieved. Relatively low standard deviations were obtained for such multidimensional separations in terms of peptide pI (<0.05 pI units) and retention time (<0.3 min for a 350 min gradient) for a selection of highly complex proteomics samples. Using narrow-range IPG-IEF, experimental and predicted pI were in relative good agreement. However, based on our data, retention time prediction algorithms need further improvements in accuracy to match state-of-the-art reversed-phase chromatography performance. General trends of peptide pI shifts induced by common modifications including deamidations and N-terminal modifications are described. Deamidations of glutamine and asparagines shift peptide pI by approximately 1.5 pI units, making the peptides more acidic. Additionally, a novel pI shift (+~0.4 pI units) was found associated with dethiomethyl Met modifications. Further, the effects of these modifications as well as methionine oxidation were investigated in terms of experimentally observed retention time shifts in the chromatographic separation step. Clearly, post-translational modification-induced influences on peptide pI and retention time can be accurately and reproducibly measured using narrow-range IPG-IEF and high-performance nanoLC-MS. Even at modest mass accuracy (±50 ppm), the inclusion of peptide pI (±0.2 pI units) and/or retention time (±20 min) criteria are highly informative for human proteome analyses. The applications of using this information to identify post-translationally modified peptides and improve data analysis workflows are discussed.