NEW YORK (GenomeWeb) – A team led by researchers at the Van Andel Research Institute has developed a functional proteomics platform for identifying protein substrates of lysine methyltransferase.
In a study published this week in Science Advances, the researchers used the platform to identify peptide sequences likely to be modified by methylation and found that traditional mass spectrometry workflows are poorly suited for the identification of these sequences, which suggests that protein methylation may be more widespread than previous studies would indicate, said Scott Rothbart, an assistant professor at Van Andel and senior author on the study.
Lysine methylation, in which a lysine methyltransferase (KMT) adds methyl groups to a protein lysine residue, is known to be an important process in transcriptional regulation and methylation of histone proteins in particular has become an area of focus in proteomics research.
However, while much work on protein methylation has centered around histones, many of the roughly 60 known KMTs do not act on histones. Additionally, as Rothbart and his coauthors noted, researchers have identified around 6,000 lysine methylation sites on roughly 3,000 human proteins but have linked few of these sites to a specific KMT.
The lack of information around the non-histone substrates of KMTs has become a more significant issue as interest in targeting KMTs for drug development has grown, Rothbart said.
"We and others have appreciated for a time now that these enzymes that install and remove and interact with modifications on histones also modify other proteins," he said. "And this has really come into the limelight in recent years with the push in drug discovery efforts toward targeting this class of proteins."
Such efforts have raised questions about the targets of KMTs and whether, for instance, the therapeutic effects of agents targeting these enzymes can be explained entirely by their influence on histones and chromatin function or if their action is more multifaceted, Rothbart said.
"And when we try to really get at that question, we realize that we don't even know what the repertoire of targets for these enzymes is," he said.
He noted that proteomic studies have managed to generate some limited information on KMT signaling networks and identified some of their non-histone substrates, but many of these substrates remain unknown.
To identify KMT substrates, Rothbart and his colleagues developed a synthetic peptide library consisting of roughly 47 million unique nine-amino acid peptides, each of which contained a lysine residue at the fifth position. They then screened this library against the KMTs G9a, SET7/9, and SMYD2, measuring the activity of each against the different peptides in the library to identify potential substrate motifs.
The analysis, the authors wrote, "identified and extended known sequence motifs" for each.
They then looked more in depth at SMYD2, which, they noted is the focus of several drug development efforts due to the fact that it is known to methylate the tumor suppressor protein p53, which inhibits that protein's function.
"The knowledge of all SMYD2 targets will be important for understanding the clinical outcomes of SMYD2 inhibition," they wrote.
Through an analysis of SMYD2's substrate selectivity, the researchers determined an optimal substrate for this KMT, which, notably, does not exist in the human proteome due to the fact that it contains a tryptophan residue. The finding, they noted, suggests that the peptide library could prove a useful tool for identifying ideal scaffolds for small molecule KMT inhibitors.
The platform could also allow researchers to analyze the effect of missense mutations on lysine methylation. In the Science Advances study, Rothbart and his colleagues identified a series of mutations present in human cancer cells that their platform predicted would alter SMYD2-substrate interactions.
They also found evidence indicating that previous mass spec-based studies of lysine methylation likely underrepresent the abundance of this PTM. They noted that a recent mass spec study identified 35 proteins as likely SMYD2 substrates but that none of these 35 substrates were among the top scoring 50 SMYD2 substrates based on their peptide library platform. They further noted that while the 50 top substrates identified via their study were all enriched for lysine and arginine, the 35 substrates identified in the mass spec study had lower levels of these amino acids.
Protein sequences rich in lysine and arginine will produce small, hydrophilic peptides upon trypsin digestion that are poorly suited to LC-MS/MS analysis, Rothbart noted. This suggests that mass spec experiments will be biased against such sequences, even though, according to the Science Advances study, they likely comprise the bulk of lysine methylation sites.
Rothbart said that when he and his colleagues looked at existing mass spec lysine methylation data sets, they found that they contained relatively few methylation sites with lysine rich motifs.
"We think, at least from our in vitro analysis, that these [lysine- and arginine-rich] motifs are the most robust targets for these enzymes, and they are escaping detection in standard proteomic pipelines," he said. "So we think we think we are really just cracking the surface, and that the full extent of lysine methylation is likely going to rival that of the phosphorylation world."
Rothbart noted that while the potential difficulty of identifying lysine methylation using mass spec has not been widely discussed with regard to non-histone proteins, it is a well-known issue in histone research.
"This is really common in the histone field," he said. "Histone proteins are highly basic. And there are tricks that you have to do in your sample preparation in order to get coverage of even many of the most well-studied modification sites on histone proteins."
He suggested that researchers might now borrow some of these approaches to improve mass spec coverage of lysine methylation on non-histone proteins.
Rothbart said that moving forward his team now hoped to expand their platform to other classes of enzymes. Thus far they have developed an assay to look at substrates of lysine demethylase, he said.
They also plan to do more extensive screening of the KMT family as well as look at existing mass spec data on lysine methylation to determine the KMTs most likely to modify previously identified sites.