Broad Institute researchers have devised a method for serial enrichment and analysis of protein post-translational modifications.
The technique, detailed in a paper published this month in Nature Methods, enables detection and quantitation of multiple types of protein PTMs in the same sample via a single mass spec experiment.
Within the proteomics field "people haven't been sure whether it would have a negative effect [on PTM identification and quantification] if you did multiple serial enrichment steps on a single sample, Philipp Mertins, a Broad research scientist and first author on the study, told ProteoMonitor.
The group's work, Mertins added, is the first effort he knows of to use serial enrichment to analyze multiple types of PTMs in this way.
Previous researchers have, of course, looked at multiple classes of PTMs in samples. But, Mertins said, they have typically performed the analyses on replicate samples in parallel, and have typically used different workflows optimized for the specific PTMs they sought to measure.
"The novelty of our study is that we used one generic workflow that is useful for [analyzing] a number of different post-translational modifications," he said.
Mertins is a researcher within the Broad's Proteomics Platform, which is run by Steven Carr, senior author on the Nature Methods paper. One of the Platform's missions is to collaborate on proteomics work with the larger Broad community and outside researchers. It was these collaborations, in part, Mertins said, that led the group to develop the serial PTM approach.
"We work with a lot of collaboration partners who are interested in different types of PTMs and sometimes all of those different PTMs in one kind of biological sample," he said. "So we decided to develop a technique to allow us to serially enrich all these PTMs from the same sample."
In the study, Mertins and his colleagues developed the method on SILAC-labeled Jurkat cells treated with the proteasome inhibitor bortezomib. They subjected their samples to fractionation via high pH reverse phase chromatography followed by concatenation of these fractions prior to PTM enrichment and, ultimately, mass spec analysis on a Thermo Fisher Scientific Q Exactive instrument.
The researchers looked at three types of modifications – phosphorylation, ubiquitination, and acetylation – using an enrichment system which first pulled out phosphopeptides using IMAC then passed that sample flow-through on for antibody-based enrichment of ubiquitinated peptides and acetylated peptides.
Applying this workflow to three replicates, the Broad team quantified per replicate an average of 7,897 proteins, 20,800 phosphorylation-sites, 15,408 ubiquitination sites, and 3,190 acetylation sites. They achieved an overlap across the three replicates of 94 percent for the proteins, 66 percent for the phosphorylation sites, 44 percent for the ubiquitination sites, and 55 percent for the acetylation sites.
Based on these results, "we think the technique will work well in general for studying different PTMs and their co-regulation in a single model system," Mertins said.
As part of their analysis, the researchers also used the technique to determine what portion of the identified peptides exhibited multiple different types of modifications, finding – somewhat surprisingly, Mertins said – that such occurrences were quite rare.
Doing MS/MS searches for all three PTMs at the same time, they found that only 0.3 percent of all the modified tryptic peptides in their datasets contained multiple different types of modifications. This, the authors noted, suggests "that few peptides with multiple distinct modifications are being lost by serial enrichment, further supporting the efficacy" of the approach.
Beyond the work published in the Nature Methods paper, the researchers have thus far used the technique to investigate cross-talk between phosphorylation and ubiquitination and how it is influenced by certain E3 ubiquitin ligases, Mertins noted. He added that they are also considering the technique for work within the National Cancer Institute's Clinical Proteomic Tumor Analysis Consortium, of which Carr's lab is a member.
"We're running within the CPTAC program a large effort in which we are analyzing 100 breast cancer tumors for global proteome and phosphoproteome changes, and we might consider moving into the acetylome and possibly the ubiquitinome using similar workflows as those presented" in the recent study, he said.
Mertins added that the team expects their workflow should be suitable to modifications beyond the three analyzed in the study, though he noted that considerations like PTM abundance and method of enrichment would be key in determining how well the method might work with a given modification.
This is particularly the case with clinical samples like tumors, he said, where more limited labeling techniques and smaller sample sizes could present difficulties. For instance, Mertins said, because metabolic labeling methods like SILAC are not compatible with analysis of human tumors, the Broad researchers typically use isobaric labeling for quantitative analysis of such samples. However, particularly in the case of ubiquitination, this labeling method can interfere with the antibodies used for enrichment, requiring modifications to the protocol.
The small size of typical clinical samples could also prove a challenge, particularly in the case of low-abundance modifications, Mertins said. In the study, the researchers performed their analysis using three different amounts of sample input: 1 mg of protein lysate for what they called "low" proteome analysis, 2.5 mg for "medium" analysis, and 7.5 mg for "high" coverage."
"Phosphoproteomic or global proteome analysis is no problem" with clinical sample sizes, Mertins said. "But for [ubiquitination]... and lysine acetylation, we will have to see how much we can get out of [such] samples."
Adding new modifications to the method will "certainly be doable, but will required focused optimization of each enrichment method," he said.