LONG BEACH, Calif. Mass spec-based MRM assays may serve as a good intermediate step to push proteomic biomarkers beyond the discovery phase and into the validation phase, according to Leigh Anderson, who gave the opening keynote lecture here this week at the Association of Biomolecular Resource Facilities conference.
"We need to do really interesting things in discovery, but if we don't do the validation, [the discovery] might as well not have been done," said Anderson, who is the founder and CEO of the Washington, DC-based Plasma Proteome Institute.
A multiple reaction monitoring, or MRM, assay is a triple quadrupole mass spec-based technique that enables researchers to simultaneously and quantitatively measure a number of proteins of interest. In a paper published in the December 2005 Molecular & Cellular Proteomics, Anderson and a collaborator at Applied Biosystems showed that about 50 proteins could be monitored at the same time using an MRM assay.
In his ABRF presentation, Anderson noted that while the number of potential biomarkers being discovered is increasing steadily, there are only a handful of proteins that are interesting to measure clinically, and none of those have been discovered through proteomics.
"We need to do really interesting things in discovery, but if we don't do the validation, [the discovery] might as well not have been done."
One reason for this is because "we don't have a pipeline for moving from biomarker discovery to a standard clinical test," said Anderson. "Nobody funds biomarker validation. The National Institutes of Health funds biomarker discovery, but not validation."
In moving beyond the discovery phase and into validation, one can take an immunoassay-based approach or a mass spectrometry-based approach, said Anderson.
Immunoassay-based validation approaches, such as antibody arrays and ELISA tests, are relatively easy to perform, but they are expensive and may take a long time to develop, especially if antibodies for the candidate biomarkers are not immediately available, said Anderson.
As an alternative to immunoassays, Anderson proposed using MRM assays, especially for whittling down the number of candidate biomarkers from many tens to a panel of about 10 biomarkers, which would be reasonable to deal with during validation.
"In discovery, we're generating a lot of biomarkers typically tens to hundreds of biomarkers for a single disease," said Christie Hunter, an application specialist at Applied Biosystems who collaborated with Anderson on a project to bring 177 candidate biomarkers for cardiovascular disease towards validation. "But we can only take a small portion of biomarker candidates to validation."
To facilitate that transition, Hunter and Anderson proposed an intermediate step, called a "candidate verification" step, which would use MRM assays.
An MRM experiment is accomplished by specifying the parent mass of a peptide for MS/MS fragmentation, and then specifically monitoring for a single fragment ion. MRM delivers a unique signal that can be monitored and quantified in the midst of a very complicated matrix. The assays' mass spectra plots are simple, usually containing only a single peak for each MRM. This characteristic makes the assay especially suitable for sensitive and specific quantitation.
"The advantage is that you can take a longer list of candidate biomarkers and find out in a relatively short period of time which of those biomarkers is most promising," said Hunter.
David Speicher, the director of the Wistar Institute's Proteomics Laboratory, agreed with Anderson and Hunter that MRM assays are a good strategy for intermediate validation.
"It's not a completely proven technology, but it seems that it allows you to do assays more quickly," he said. "We are exploring that [technique] ourselves."
In bringing candidate biomarkers from the proteomics discovery phase to the clinical assay phase, it is important to increase the reproducibility of quantitative biomarker tests, Anderson said.
Hunter and Anderson showed that 70 percent of their data from MRM assays had a coefficient of variation below 10 percent.
"The MRM's reproducibility is reasonably high," said Anderson.
"We don't have a pipeline for moving from biomarker discovery to a standard clinical test," said Anderson. "Nobody funds biomarker validation. The National Institutes of Health funds biomarker discovery, but not validation."
While any triple quadrupole can be used to do an MRM assay, Hunter said ABI's hybrid Q-TRAP is particularly good for doing the assays because it switches from triple quadrupole mode to linear ion trap mode once the MRM triggers a scan of the MS spectrum.
"A regular triple quad can also do MS/MS, but it's done with a quadrupole, which is slow and not very sensitive," said Tina Settineri, director of ABI's Q-TRAP proteomics product line. "With the ABI 4000 Q-TRAP, the third quadrupole can also function as an ion trap, which scans much faster and is much more sensitive."
Last November, ABI introduced its MIDAS workflow especially for streamlining MRM assays. The system, whose name stands for MRM-Initiated Detection and Sequencing, consists of a Tempo nano MDLC combined with a Q-TRAP and special software to help build MRM peptides. The high-end MIDAS system, which includes a 4000 Q-TRAP, costs $535,000, while the entry-level MIDAS system, which includes a 3200 Q-TRAP costs $355,000.
While ABI appears to be the only mass spec company so far that has developed special software for designing MRM peptides, other mass spec companies, including Thermo Electron, Agilent, and Waters said they have customers that do MRM assays using their instruments, and that specialized software is not a necessity.
"We don't have an MRM builder, but we don't see that as a real impediment to using our solution over an ABI solution [for MRM assays]," said Lester Taylor, the director of global product marketing for life sciences mass spectrometry at Thermo.
Ken Miller, the product marketing manager for proteomics at Thermo, said he would recommend Thermo's linear ion traps, such as its LTQ Orbitrap, for initial quantitative work, then when the customer's quantitative needs outweighs discovery needs, he would recommend transitioning to a triple quadrupole, such as the TSQ Quantum.
When asked how an MRM assay-based approach for protein or peptide biomarker validation compares with a protein chip-based approach, Graham Scott, the research and development manager for protein expression and proteomics at Sigma-Aldrich, said that the two methods are complementary, and that each approach has its advantages and disadvantages.
"With protein chips, the technology has a sensitivity limitation based on fluorescence," said Scott. "You would need brighter fluors or better reporter molecules in order to get better sensitivity in an array."
However chips are easy to use, and can potentially multiplex more proteins than Anderson's MRM-based method for validation, called SISCAPA, said Scott.
SISCAPA (Stable Isotope Standards and Capture by Anti-Peptide Antibodies) is Anderson's mass spec-based method for "next level" validation after candidate biomarkers have been whittled down, or "verified".
"SISCAPA could potentially alleviate the bottlenecks for building antibodies, and eliminate a lot of the need for exhaustively validating antibodies," said Scott. "But it's pretty much just an idea now. Leigh is the only one who has done any proof-of-principle of it."
Tien-Shun Lee ([email protected])