By Tony Fong
The Institute for Systems Biology and Agilent Technologies this week said they will be working collaboratively to create a human multiple-reaction monitoring atlas, an effort that could spur on a strategy that only in recent years has begun to gain a toehold in proteomics.
At the end of the two-year project, the atlas would contain at least four peptides for each of the estimated 20,000 to 25,000 protein-coding genes in the human proteome. It also would include verified rapid and accurate MRM-based mass spectrometry assays for the identification and quantification of any protein in the human proteome in a multitude of samples.
According to a statement issued by ISB and Agilent, the effort is expected "to fuel important research gains in biomarker discovery and validation, the search for protein-based diagnostic tests, personalized medicine, and human health monitoring."
Funding for the project comes from the National Human Genome Research Institute of the National Institutes of Health which is providing $2.7 million in direct funding under the American Recovery and Reinvestment Act of 2009.
The European Research Council is contributing €2.7 million ($4.1 million) in direct funding to the project, as well.
According to Robert Moritz, director of proteomics at the ISB, the MRM atlas builds on the ISB's Peptide Atlas and work published during the summer describing a single-reaction monitoring atlas for yeast.
"The biggest problem with the mass spectrometry done today … is that you ask the mass spectrometer to try and identify anything that it sees per unit time," Moritz told ProteoMonitor. "And that's a really random process."
The idea behind an MRM workflow is that instead of random identification of proteins, specific proteins would be targeted for further analysis as candidate biomarkers. The MRM atlas, Moritz said, would provide researchers a tool to determine which proteins to target.
Ken Miller, director of LC-MS marketing for Agilent, said that protein quantitation now involves three steps: Identifying from a list of proteins the peptides to base assays on; determining an optimum transition to investigate with a mass spec; and setting up and optimizing a method to separate the peptides and perform an assay.
The MRM atlas would simplify that workflow, he said.
"We're going to have a standard methodology, so [researchers] won't have to worry about column selection, they won't have to worry about gradient conditions or anything else," Miller said. Additionally, the list of peptides that should be selected for each protein in the human proteome will be provided, as will all of the transition and collision energies for all the peptides.
"What we're really doing is dramatically reducing the amount of time, energy, and cost of getting involved in setting these assays up in the first place," he said.
The Peptide Atlas contains roughly 12,000 known human proteins. For each of these, "we will select the best four proteotypic peptides" for synthesis for the MRM atlas, Moritz said. For the remaining human proteins where no physical data exists, peptide prediction algorithms developed by ISB researchers and collaborators on the project will be used to choose at least four peptides for each protein "to complete the entire human proteome as a single dataset with verified MRM methods developed," he added.
Work on the atlas will be done using Agilent's 6460 triple-quadrupole LC-MS/MS platform, the 6530 quadrupole time-of-flight LC-MS, and the HPLC Chip-MS system. The HPLC Chip-MS system will be used to separate peptide mixtures on the nanoscale. The 6460 mass spec will be used to perform MRM-based quantitative assays of all the peptides, and the 6530 will be used to generate MS and MS/MS spectra for the peptides.
[ pagebreak ]
Moritz said that the fact that the atlas will contain information culled from a consistent platform is an important element of the project. Peptide Atlas contains information "from all different styles of machines from all manufacturers," but does not provide details about the machines used or their manufacturers.
But the MRM atlas will contain peptides that are measured on a "consistent platform so that we have this complete database on a fairly level footing," Moritz said.
The peptides for each protein will be validated by creating fragmentation profiles that can "pick up the most suitable transitions to use as the MRM analysis," he said. That information will then be programmed into the 6460 instrument to build the assays.
"That's where the MRM atlas becomes crucial: that we have that data collected from the measurement of the peptides … and then you can go back and say, 'From each of these peptides … I can choose the best or multiples of those transitions that would be able to uniquely identify those peptides in a complex mixture,'" Moritz said.
MRM: Workflow of Tomorrow?
The MRM atlas comes amid a trend in proteomics away from the warehousing of thousands upon thousands of proteins with little information about them aside from their identities, to a more quantitative approach, as well the verification of candidate biomarkers — a shift that could help move proteomics research along to the clinical setting.
Much of the attention to such a quantitative shift has focused on MRM-based workflows. In proteomics meetings and conferences, workshops have concentrated on using MRM-based strategies, while vendors also have designed instruments specifically for MRM work, including Agilent's 6460 platform being used in the current research.
Still, though MRM is well-established in pharmaceutical small molecule-based research and areas such as food safety and pesticide analysis, it is only just starting to make headway in proteomics.
According to Miller, the focus in proteomics has primarily been on identification by shotgun methods and only recently moved beyond that. But the field suffers from a general lack of robust and effective targeted protein quantitation methods.
With an ELISA-based technique, for example, antibody cross reactivity is a problem if a researcher wants to look at more than one protein at a time. And with mass spec-based methods, such as those utilizing iTRAQ technology, costs associated with labeling can be prohibitive.
MRM, however, allows for multiplexing, eliminates cross-reaction issues, and does not require any labeling technology.
"Once you have the triple-quad, the analysis is basically free," Miller said
The method may be poised for greater adoption. Three years ago, the NIH created its Clinical Proteomics Technologies for Cancer initiative, and, said Moritz, its director Henry Rodriguez has pushed heavily for targeted analytical methods such as MRM.
In addition, work by Ruedi Aebersold in recent years has shown to the proteomics community that such targeted proteomics can not only be done but can be applied to biological systems.
Aebersold, a co-founder of ISB and now a professor of systems biology at the Swiss Federal Institute of Technology is one of the co-PIs on the MRM atlas project along with Lee Hood, the president of ISB.
"It's really this whole idea of systems biology in which you can measure hundreds of proteins at any one time, rather than just trying to measure one or two proteins per unit time to get a complete picture of what's occurring within the cell, rather than small subsets," Moritz said.
[ pagebreak ]
This summer a milestone was reached when a team of researchers put together by CPTC published a paper that demonstrated for the first time interlaboratory reproducibility for a method coupling MRM with stable isotope dilution mass spectrometry, putting to rest one hurdle to greater acceptance of MRM-based methods [See PM 07/09/09].
Agilent also hopes that the MRM atlas will create demand for MRM-based research, and stands to benefit from its participation in the project in two ways. The first is feedback that it can use to assess the suitability of its instruments for an MRM workflow, and to make improvements to them based on the feedback, Miller said.
More importantly, the company plans to use the information resulting from the creation of the atlas to develop new software to enable a "widescale adoption of quantitative MRM-based methods," Miller said.
For example, a researcher may have a list of potential biomarkers or a particular pathway he or she wants to investigate. "Given this specific MRM atlas and access to an Agilent instrument, they should be able, then, to set these assays up to do very sensitive targeted protein quantitation for all or any of the [20,000] human proteins that we're going to look at as part of this project," he said.
While a researcher would be able to use a platform from any vendor to set up an MRM assay based on data from the MRM atlas, Agilent hopes to position itself "as a standard platform for performing these assays," Miller added.
Moritz added that one factor in the timing of the creation of the MRM atlas is the drop in costs to synthesize peptides to a point where it can now be done without having to mortgage the house. "At this point, it's feasible to consider orders of 100,000 peptides," he said, adding that the amount of time necessary to make them, about two or three months, has become similarly manageable.
After creating the basic atlas of four peptides per protein, the next step would be to "build further peptides in that would not only distinguish that protein but may distinguish proteins from SNPs," and modified proteins from unmodified ones, he added.
All information from the project will be open source.
The project, he stressed, is not meant to measure every protein in biological samples but to build assays that will allow researchers to do that themselves.
"This infrastructure will allow any researcher to start to apply that," Moritz said. "And so you can imagine researchers from around the world with all their different interests, taking that information and being able to readily create those assays and start doing those measurements, and then contributing that information back to the MRM atlas."