NEW YORK (GenomeWeb) – Harvard University researchers have developed a new form of isobaric tags that could enable considerably higher multiplexes than existing reagents.
In a paper published last month in Analytical Chemistry, the researchers used the tags for six-plex analysis of yeast whole cell lysate, finding that their performance approached that of conventional commercial isobaric tagging reagents.
The expectation is that this performance will remain the same as they move into higher multiplexes, Harvard researchers Craig Braun, Steve Gygi, and Wilhelm Haas told GenomeWeb.
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.
These reagents are sold as tandem mass tags, or TMT, by Thermo Fisher Scientific through a licensing agreement with Proteome Sciences, and as isobaric tags for relative and absolute quantitation, or iTRAQ, by AB Sciex.
Currently, 10 is the maximum multiplex enabled by commercial isobaric tagging reagents. The most obvious way to increase this multiplexing further, Haas noted, would be to increase the size of the tags to increase the number of different fragmentation patterns generated in MS2. However, he said, this approach is problematic in that peptide IDs decrease as tag size increases.
Haas cited in particular a 2010 Analytical Chemistry paper by the lab of Karl Mechtler at Vienna's Institute of Molecular Biotechnology.
That study "really opened our eyes to this problem," Haas said. The study compared a number of peptide IDs after labeling with iTRAQ four-plex (molecular weight of 144 Da), TMT six-plex (229 Da), and iTRAQ eight-plex (304 Da) reagents. The authors found that they detected 35 percent less peptides with the six-plex and 70 percent less with the eight-plex than with the four-plex.
"These are substantial numbers and clearly show the limits of adding channels through increasing the tag mass," Haas added.
To get around this problem, Haas and his colleagues adopted a different approach — altering the tag structure so that they release upon fragmentation not just one, but two reporter ions, a primary and secondary ion. Different tags can then be built to release different combinations of primary and secondary ions, allowing for a significant increase in multiplexing.
While the Analytical Chemistry paper looked only at six-plex reagents, the authors noted that the tags, which they have named Combinatorial isobaric Mass Tags (CMTs), as presented in the study could support multiplexing of up to 16 samples. And, with structural alterations to place the reporter ions more tightly together, the tags could allow for multiplexing of up to 28 samples at a time.
Haas said the researchers don't expect increased multiplexing to have "any effect on the number of IDs as the chemical structure of the reagents is exactly the same." He compared it to the move from TMT six-plex to TMT 10-plex reagents, which relies on use of the differences in 12C/13C and 14N/15N transitions, as opposed to increased tag size.
In that case, he said, "the number of IDs did not change, but we almost doubled the plexing capacity."
Haas said that the researchers have filed a patent on the reagents and are now "sorting out the best options to commercialize the reagents."
There remain some wrinkles to work out, however, most notably in terms of the reagents' reproducibility. While the six-plex CMT reagents tested in the Analytical Chemistry paper proved roughly equivalent to conventional six-plex TMT reagents in terms of peptides identified, the CMT reagents had roughly two-fold higher coefficients of variation than the TMTs.
This is particularly significant in that one of the main rationales for using isobaric tags is that they allow researchers to generate better quantitative data across multiple samples by running them all in a single multiplexed experiment.
As the authors noted in the paper, this reproducibility issue "limits the ability of CMT to detect significant protein expression differences, particularly when the fold-change is small."
Haas said that after looking into the question, he and his colleagues believed the source of the problem was likely "minor impurities of the reagents."
"A further purification step of the reagents already shows an increased reproducibility of the quantitative measurements," he said, adding that the researchers expect that they would soon be able to solve the problem and that it did not appear to scale with increased multiplexing.
The CMT reagents are one of several approaches being pursued in hopes of increasing multiplexing in proteomics experiments. Another notable ongoing effort is the development by University of Wisconsin-Madison research Joshua Coon of Neucode labeling, a method that uses the different nuclear binding energy of different carbon, hydrogen, nitrogen, and oxygen to incorporate distinct combinations that can be used in labeling of proteomic samples.
In a study published last year in Molecular & Cellular Proteomics, Coon and his colleagues used the reagents to multiplex 18 yeast samples in a single mass spec experiment. Coon noted at the time, however, that an 18-plex would not be typical given that the approach suffers from a lack of proteome coverage at such levels of multiplexing. For instance, in the MCP study, they quantified just 603 proteins in the 18-plex experiment.
More typical, Coon said, would be a nine- or 12-plex experiment, where the method allows for quantitation of between 1,000 and 1,500 proteins.
A high level of multiplexing is desirable in proteomics experiments for a variety of reasons. In experiments like biomarker discovery efforts aimed at detecting differences in protein expression across samples, large numbers of samples are needed for determining statistical significance, and running them in a single experiment allows researchers to better standardize sample prep and compensate for the stochastic sampling inherent in shotgun mass spec experiments.
The rise of other mass spec methods, data-independent acquisition methods like Swath, in particular, could offer a different route to reproducible quantitation of large numbers of samples, though. For instance, in a recent interview with GenomeWeb, University of Manchester researcher Tony Whetton noted that his group had moved away from isobaric tagging for biomarker discovery.
While isobaric tagging has traditionally been his group's main tool for looking for protein expression differences across samples, they have recently moved to DIA methods like Swath, Whetton said, noting that he finds this approach much faster and simpler and, in general, better suited to studying through the hundreds of samples he and his colleagues want to look at in their clinical work.
Haas said, however, that he expects multiplexing strategies to remain popular. He cited as one advantage the increased accuracy provided by stable isotopes such as those used in the CMTs.
Additionally, he said, "multiplexing gives us the means to increase the throughput in proteomics independently to new developments in the mass spec instrument technology."
"Mass spec time is the most expensive item when analyzing proteomics samples, and every added multiplex channel will bring down [those] costs," he said. "If we want to achieve similar things [in terms of sample throughput] as have been done in genomics and move proteomics to the clinic, we will have to reduce the costs. I think that increasing the multiplexing capacity is what will bring us there."
On this front, Coon's lab earlier this year presented an approach combining its Neucode labeling with Swath-style mass spec, allowing for implementation of sample multiplexing in concert with data-independent analysis.