Abstract
Lysine methylation is one of the important post-translational modifications (PTMs) that regulate protein functions. Up to now, proteomic identification of this PTM remains a challenge due to the lack of effective enrichment methods in mass spectrometry experiments. To address this challenge, we present here a systematic approach to predicting peptides in which lysine residues may be methylated to mediate protein-protein interactions. We used the chromodomain of the polycomb protein in Drosophila melanogaster as a model system to illustrate the success of this approach. We started with molecular dynamics simulations and free energy analyses on the histone peptides complexed with the polycomb chromodomain to understand how the binding specificity is achieved. We next conducted virtual mutagenesis to quantify each domain and peptide residue's contribution to the domain-peptide recognition, based on which scoring scheme was developed to evaluate the possibility of any lysine-containing peptides to be methylated and recognized by the chromodomain. A peptide microarray experiment on a panel of conserved histone peptides showed a satisfactory prediction accuracy of the scoring scheme. Next, we implemented a bioinformatics pipeline that integrates multiple lines of evidence including conservation, subcellular localization, and mass spectrometry data to scan the fly proteome for a systematic identification of possible methyllysine-containing peptides. These putative chromodomain-binding peptides suggest unknown functions of the important regulator protein polycomb and provide a list of candidate methylation events for follow-up investigations.