NEW YORK (GenomeWeb) – Researchers at the Institute for Systems Biology have developed a method for improved analysis of protein post-translational modifications using Swath-style data-independent acquisition mass spec.
Detailed in a paper published last month in Molecular & Cellular Proteomics, the approach allows for unbiased detection of protein PTMs with greater sensitivity than is possible using conventional shotgun mass spec, Robert Moritz, an ISB researcher and senior author on the paper, told GenomeWeb.
ISB is currently working to implement an automated, open-access version of the method. Moritz said that it is also working with Sciex on developing it as either a standalone software product or a tool that could be integrated in the company's OneOmics environment, its cloud computing collaboration with Illumina.
The use of Swath and related DIA methods has grown significantly in recent years, with researchers drawn by the technique's ability to provide reproducible quantitative data on thousands of proteins. However, Moritz noted, the method has not, to date, been particularly useful for analysis of protein PTMs.
This, he said, is due to the data analysis approach used in Swath experiments, wherein spectral libraries generated by an initial DDA run are used to query Swath datasets for the presence of particular peptides. If a modified peptide is not present in this spectral library, then it won't be detected in the Swath data, and the large number of potential modifications makes covering all the different possibilities a significant challenge.
"The problem is that you have to actually have every possible combination of modified components of a peptide in your spectral library," Moritz said. "So you are either going to have to [compile] that from natural sources, or you are going to have to do that [using synthetic peptides]. And that becomes a very difficult approach."
Moritz and colleagues including ISB computational biologist Andy Keller, first author on the MCP paper, tackled this problem by developing a tool, which they named SwathProphetPTM, that identified precursor ions in Swath data likely to harbor modifications and then re-extracts these ions using a precursor window adjusted to account for the mass added by the modification.
Essentially, the tool looks for likely modified peptides by identifying peak groups with missing fragment ion peaks, the notion being that these fragment ion peaks might be missing because they contain a modification and so did not appear in the expected m/z window. The software takes these candidate peak groups and then reanalyzes them using m/z windows appropriate for modified versions of the missing fragments ions.
This allows for unbiased, Swath-style detection of modified peptides within that new m/z window. "And then once we are able to detect a peptide with a specific mass difference, we can interpret from that mass difference what the modification could be," Moritz said.
In the MCP paper, the researchers tested the approach in several different samples, including a set of 1,055 heavy labeled synthetic peptides spiked into backgrounds of neat solvent or trypsinized human urine and a human metastatic breast cancer cell line lysate enriched for phosphopeptides via immobilized metal-affinity chromatography using Fe(III) and titanium oxide.
In the first experiment, they identified 209 modified synthetic peptides and 65 modified human urine peptides across three orders of magnitude. In the second experiment, they identified 210 phosphorylated peptides.
While the initial shotgun experiment used to generate the spectral library for the experiment with the phosphoenriched cell line identified significantly more phosphorylated peptides, 1,367, the SwathProphetPTM was, importantly, able to identify phosphorylated peptides not detected in the original shotgun experiment, and so not present in the spectral library used for the traditional portion of the Swath analysis.
This, the authors noted, demonstrates that the software "can identify previously undetected biological post-translational modifications such as phosphorylation events without anticipating them, on the basis of the unmodified peptide assays alone."
This is significant in that, while spectral libraries continue to grow, they will likely always remain incomplete given the large number of possible modifications and genetic polymorphisms, and these modifications not present in spectral libraries would go undetected using conventional Swath analysis methods.
Additionally, shotgun methods used to generate spectral libraries tend to favor higher abundance modifications, and so the SwathProphetPTM offers the possibility of more sensitive detection of low-abundance PTMs. The fact that the method looks for modifications at the MS2 as opposed to MS1 level could also allow for improved sensitivity.
Because the method is looking for PTMs in an unbiased manner across a wide m/z range, it also gives researchers the opportunity to identify less common modifications including non-biological modifications due to, for instance, laboratory contamination, Moritz said.
"One always assumes that the reagents you use are pure, but when you are doing your biochemical extractions you could be modifying the peptides and not knowing you are actually doing that using reagents you've had on the shelf for quite some time," he said. "And that can affect your quantitation, because if you are basing it purely on the detection of unmodified peptides then you could be under-sampling what you actually have."
Moritz said he expected the approach would at ISB be a standard method used as an additional analysis of all Swath data. It is currently available as a downloadable module within the Trans-Proteomic Pipeline, the mass spec informatics suite developed by ISB researchers, he said, adding that ISB was working with Sciex on a version of the method, as well.
The software "moves us in a global and comprehensive way toward, not only straightforward protein analyses, but also taking on the subtleties of post-translational modifications and the processing of proteins, the many different things that can happen to modify the primary structures," Leroy Hood, president and co-founder of ISB and co-author on the MCP paper, told GenomeWeb.
More generally, Hood noted the significant amount of ongoing investment in further fleshing out Swath-style mass spec on the computational side.
"It's one thing to be able to look at a window of fragments, but it is another to be able to interpret them in the context of what are the peptides and what are the proteins they represent," he said. "All of that requires strikingly challenging computational tools."