Identification of in vitro JMJD lysine demethylase candidate substrates via systematic determination of substrate preference

A major regulatory influence over gene expression is the dynamic post translational methylation of histone proteins, with major implications from both lysine methylation and demethylation. The KDM5/JARID1 sub-family of Fe(II)/2-oxoglutarate dependent lysine-specific demethylases is, in part, responsible for the removal of tri/dimethyl modifications from lysine 4 of histone H3 (i.e., H3K4me3/2), a mark associated with active gene expression. Although the relevance of KDM5 activity to disease progression has been primarily established through its ability to regulate gene expression via histone methylation, there is evidence that these enzymes may also target non-histone proteins. To aid in the identification of new non-histone substrates, we examined KDM5A in vitro activity towards a library of 180 permutated peptide substrates derived from the H3K4me3 sequence. From this data, a recognition motif was identified and used to predict candidate KDM5A substrates from the methyllysine proteome. High-ranking candidate substrates were then validated for in vitro KDM5A activity using representative trimethylated peptides. Our approach correctly identified activity towards 90% of high-ranked substrates. Here, we have demonstrated the usefulness of our method in identifying candidate substrates that is applicable to any Fe(II)- and 2-oxoglutarate dependent demethylase.