NEW YORK (GenomeWeb) – A team led by researchers at the Leibniz Institute on Aging and proteomics firm Biognosys have compared the performance of data-independent acquisition (DIA) and tandem mass tag (TMT) labeling mass spectrometry workflows for quantitative proteomics.
Detailed in a paper published this month in the Journal of Proteome Research, the evaluation found that the two approaches are largely comparable, with both delivering coverage of more than 5,000 proteins with a small number (less than 2 percent) of missing values.
While this would indicate both approaches are well suited to quantitative proteomics experiments, a shift within the field towards experiments comprising larger numbers of samples and conditions will likely favor DIA in the future, suggested Alessandro Ori, a group leader at the Leibniz Institute and author on the study.
DIA and TMT labeling are two of the most commonly used methods in quantitative proteomics. In the first approach, the mass spectrometer selects broad m/z windows and fragments all the precursors in that window, which allows the machine to collect MS/MS spectra on all ions in a sample.
DIA typically provides less depth of coverage than traditional shotgun mass spec workflows, but because it collects spectra on all ions in a sample, it measures highly consistent sets of proteins across samples, making it well suited to quantitative experiments assessing protein levels in multiple samples under multiple conditions.
TMT labeling is a variety of isobaric labeling. TMT reagents were developed by Proteome Sciences and are sold by Thermo Fisher Scientific. A competing isobaric labeling regeant, iTRAQ, is sold by Sciex.
Isobaric labeling uses stable isotope tags attached to peptides of interest to enable relative or absolute quantitation of proteins via tandem mass spectrometry. Digested peptides are labeled with tags that fragment during MS2 to produce signals corresponding to the amount of peptide present in a sample. The approach is commonly used to multiplex samples, allowing researchers to run up to 10 samples in a single mass spec experiment, which improves throughput and reduces variation.
TMT technology was introduced in 2003, while the commonly used SWATH DIA approach was introduced in 2012 (though the concept underlying DIA was described in 2004 and Waters launched a version of the technology called MSE in 2006). Both approaches have been improved upon and refined since their introductions, and each has its adherents, but Ori noted that there is little data comparing the two methods head-to-head.
"There are a lot of people who are doing TMT and a lot of people who are doing or considering doing DIA, but there are not actually many labs that do both," he said, adding that while his lab has traditionally used TMT labeling for its quantitative proteomics work, it has recently begun setting up DIA workflows with help from Biognosys.
Ori said that he and his colleagues were interested in comparing the effectiveness of the two approaches and in assessing how transferrable they were across different laboratories.
In the study, the researchers compared the two approaches on a sample set consisting of 10 biological replicates of mouse cerebellum tissue with the UPSP protein standard (which contains 48 human proteins) spiked into the samples at five different concentrations.
They found that the TMT method identified 15 to 20 percent more proteins with somewhat better quantitative precision, while the DIA approach offered better quantitative accuracy. Both methods transferred well across labs.
Despite the relatively similar performance of the two approaches, several factors could make DIA a more attractive technology, particularly for labs looking to run large numbers of samples.
For instance, the study used a Thermo Fisher Orbitrap Fusion Lumos, which is the flagship of the Orbitrap line and one of the highest performance mass specs currently used in proteomics research. This instrument was necessary to effectively run the MS3 TMT workflow used in the study, and while a less expensive (and more widely accessible instrument) could run TMT experiments that stop at the MS2 level of fragmentation, this MS2 approach runs into issues of precursor interference that lower the method's quantitative accuracy.
The DIA method used in the study could be run equally successfully on a Fusion Lumos or a less expensive instrument like a Thermo Fisher Q Exactive, Ori said.
"When it comes to accuracy of quantification, DIA already had better accuracy even [compared to] the MS3 TMT, so if you were using MS2, that would become even more pronounced," he said, though he added that many groups have success with MS2 TMT, particularly in experiments where they are looking for differential protein expression between samples but don't need highly accurate quantitative data.
The authors noted that DIA also has an advantage in terms of the sample volume required for analysis. They used samples consisting of 2 µg of protein digest for the DIA analyses and 20 µg samples for the TMT work.
DIA experiments are also easier to scale as the method's high reproducibility across samples allows researchers to measure proteins consistently across large sample sets. TMT experiments, on the other hand, typically require the use of common reference samples to compare data from different experiments.
DIA "is simpler, no doubt," Ori said. He noted that while his lab still typically uses TMT for experiments that can be done in a single 10-plex due to the method's greater depth of coverage, he, like the proteomics field more generally, has been moving towards analyzing larger sample cohorts.
"The kind of studies people are moving to, the number of samples you are comparing is going to number in the hundreds in some cases," he said. "And that's where I think DIA has an advantage."