NEW YORK(GenomeWeb) – Researchers at the University of Pennsylvania and AB Sciex are working to apply Swath-style data independent acquisition mass spec to the analysis of histone post-translational modifications.
The method's ability to provide reproducible, proteome-scale MS/MS level quantitation could make it a useful addition to mass spec-based histone analysis, which, to date, has largely relied on a combination of traditional data-dependent acquisition mass spec and targeted multiple-reaction monitoring experiments, said Ben Garcia, a University of Pennsylvania researcher and leader of the effort.
In a study published last week in Molecular & Cellular Proteomics, Garcia and his colleagues used AB Sciex's Swath workflow on the company's TripleTOF 6600 instrument to analyze histone H3 peptides from human embryonic stem cells, finding that they could accurately quantify these peptides across four orders of magnitude and could discriminate isobaric and co-eluting peptides with high sensitivity and precision.
Histones are one of the key components of chromatin and are involved in gene expression, DNA repair, and cell duplication. Their function is significantly controlled by their post-translational modifications, which also play an important epigenetic role as they can be inherited through cell division.
Given the importance of histone PTMs, they are a key area of interest for proteomics researchers studying these proteins. However, given the large number of PTMs on histones, their analysis is more difficult than for other proteins.
In particular, isobaric modifications – peptides with modifications of the same mass but in different locations – are a common problem in histone analysis. High-quality MS/MS data is necessary to get good identifications and quantification of these isobarically modified peptides, Garcia said. However, because mass specs operating in DDA mode can't scan quickly enough to select every ion in a sample for MS/MS fragmentation, the peptides selected for MS/MS vary with each run, making it difficult to obtain reproducible quantitation. This is particularly a problem with low-abundance peptides, which are most likely to go unselected.
In DIA mass spec, on the other hand, the instrument selects broad m/z windows and fragments all precursors in that window, allowing the machine to collect MS/MS spectra on all ions in a sample.
Given this, the approach seemed potentially promising for histone analysis, Garcia said.
"In DDA, you often don't sample the lowest-level species for MS/MS, and that kind of puts a lot of strain [on your analysis] when you try to quantify the species, because even if you have the accurate mass and retention time, if you don't have the MS/MS of that species, there's always a little bit of doubt in your mind that perhaps you're quantifying a [different] analyte," he said.
"I felt like the DIA method – like everyone has talked about – is kind of between MRM and shotgun proteomics," he added. "So it would be a good sweet spot to be in."
Specifically, the approach would potentially allow Garcia and his colleagues to avoid the tedious method development required to confidently measure isobarically modified peptides either by shotgun mass spec or MRM.
For instance, noted Simone Sidoli, a post-doc in Garcia's lab and first author on the MCP paper, to overcome the issue of DDA's stochastic sampling, the researchers programmed the mass spec with inclusion lists telling it to target specific peptides known to have isobaric forms.
However, as Sidoli told GenomeWeb, it is difficult to consider in advance all the isobaric forms that might be present in a sample.
DIA "allows us to re-mine isobaric forms that we didn't consider previously," he said.
Additionally, Garcia said, "when you add all these targeted MS/MS values [to an inclusion list] it eats away at your scan time."
It also requires very consistent retention times, he noted, meaning that if the researchers switched LC columns or otherwise adjusted their chromatography and slightly altered these times, they could miss the isobaric species on which they intended to perform MS/MS.
"Sometimes it didn't quite work if some parameter was slightly off with the chromatography," he said. "So it was always a little bit of a pain, and with Swath, you completely eliminate that."
By the same token, Swath analysis requires much less method development than selecting transitions for MRM experiments targeting large numbers of isobarically modified peptides, Garcia added.
That said, DDA still typically provides better depth of analysis, he noted. "We did look for some species that are pretty rare, and in most cases we didn't get enough confident MS/MS information from the DIA experiment to say that they were truly there, [while] with DDA we have been able to see some of them."
DIA will likely prove most useful for "rigorous, accurate, reproducible quantification of things that more or less you are expecting or are familiar with across many data sets rather than discovery," Garcia said.
Further development is needed to optimize the approach for histone PTM analysis, particularly on the informatics side of things, he said, noting that a major point of his group's collaboration with AB Sciex on the MCP paper was optimizing its software for histone work.
"Histones are actually very difficult to search with commercial software because there are so many different types of modifications and combinations that false discovery rates go up quite a bit," he said. "So what applies to normal peptides doesn't really apply to histones."
In particular, DIA data analysis, which uses spectral libraries as opposed to conventional peptide databases for making identifications, presents a challenge, Garcia said.
Conventional MS/MS searching is a problem, "but something that could be remedied pretty quickly with existing commercial software – there just has to be new rules and scoring in place for histone peptides," he said. "The Swath data analysis was actually pretty difficult, though. [DIA] is pretty new, so there just aren't enough options yet to really analyze that data really comprehensively."
He noted that for the study Sidoli had to adapt the University of Washington's widely used Skyline targeted proteomics software for histone analysis.
One tool that could help is a new DIA analysis package developed by University of Michigan researcher Alexey Nesvizhskii that allows researchers to search DIA data using conventional DDA database searches. The software uses m/z and retention times to detect and match precursor and fragment ion levels in DIA MS1 and MS2 level data and then uses these groupings to generate pseudo-MS/MS spectra that can be searched traditionally.
Garcia said he had not yet looked into using the software for his lab's histone work, but that it might prove useful.
In addition to AB Sciex, Garcia and his team are also working with Thermo Fisher Scientific to develop and refine methods for DIA analysis of histones. In particular, they are focusing on middle-down DIA analysis – looking at peptides longer than conventional tryptic peptides but shorter than full length proteins. The hope is that such an approach could simplify and improve analysis of features like PTMs and enable more complete sequence coverage.
One limitation is that middle-down typically fragments peptides using electron transfer dissociation, which is not currently available on the AB Sciex TripleTOF instruments.
It is available on Thermo Fisher's Orbitrap Fusion, though, which, Garcia said, is the instrument they are focusing on for middle-down DIA development. "If it's going to happen, it's going to be on that type of instrument," he said.