With the Human Proteome Project a deferred dream at the moment, a band of researchers hopes to launch a smaller-scale project to create a library of anti-peptide antibodies for measuring “defined” collections of proteins with high sensitivity and “absolute specificity.”
The project, which the researchers are calling the Human Proteome Detection and Quantitation initiative, or hPDQ, would serve as a “near-term tactical approach” for providing a more basic map of the human proteome than the HPP.
It would also aim to hasten the development of clinically useful protein biomarkers, the researchers wrote in a Jan. 7 article published online in Molecular & Cellular Proteomics describing hPDQ.
Unlike commercially available antibodies, which are directed against proteins, the hPDQ project seeks to develop “at least one antibody against peptides for each protein of the human proteome,” Christoph Borchers, a co-author of the MCP paper and director of the University of Victoria – Genome BC Proteomic Centre, told ProteoMonitor this week.
“The reason … [is] we all know that peptides are much [more] detectable using mass spectrometry than proteins,” he said, because after researchers detect the peptides, they can deduce the identity of the proteins.
Such anti-peptide antibodies would be created by two nearly identical methods independently developed by two authors of the MCP paper. One, called immuno-MALDi, or iMALDI, was developed by Borchers and is described here, while the other, called stable isotope standards and capture by anti-peptide antibodies, or SISCAPA, is based on electrospray mass spectrometry and was developed by Leigh Anderson, CEO of the Plasma Proteome Institute. It is described here.
The idea for both methods is to use anti-peptide antibodies to enrich the peptides. Similar to an ELISA, a first antibody enriches the peptide of interest, but instead of a secondary antibody for detection, iMALDI and/or SISCAPA uses mass spectrometry for detection, Borchers said, combining the enhanced sensitivity of an immunoassay with the specificity of mass spectrometry.
Ruedi Aebersold, a co-author of the paper and a professor at the Institute of Molecular Systems Biology, said that the two technologies capitalize on a trend away from shotgun proteomics to targeted discovery.
“There is no point in perpetually rediscovering the same peptides in every experiment, pretending we know nothing. The idea is to transit from a perpetual de novo discovery mode to targeting” specific peptides and proteins, he said.
Shotgun proteomics “is very powerful but it is still associated with very substantial informatics challenges mainly how do you know that the peptide fragment ion spectra are correctly assigned to sequences, how do you reassemble the proteins from the identified peptides, and so on,” Aebersold told ProteoMonitor. “These largely go away when you move toward a targeted approach with or without antibody peptide enrichment.”
In its pilot phase hPDQ would target 2,000 human proteins selected for biomarker potential. The initial project would take two years to complete at a cost of less than $50 million “through funding of existing academic and commercial resources in a distributed network,” according to the authors. Afterward, a project lasting five years and costing $250 million would target the remaining 18,500 proteins associated with protein-coding genes.
The hPDQ project, Borchers said, would benefit the research and clinical communities in a variety of ways, including biomarker discovery and validation that could lead to new diagnostic tools.
He and his co-authors wrote that current proteomics research platforms, which focus mainly on discovery, “do not yield an economical or accurate measurement of a defined set of proteins in every sample. There is thus a fundamental barrier to hypothesis testing in quantitative proteomics, where relationships between protein abundance and biology are sought.”
[ pagebreak ]
The shortcomings of current technology, they add, are particularly glaring in the field’s failure to establish clinically relevant biomarkers that can allow physicians to diagnose, treat, and monitor diseases.
An anti-peptide antibody library would allow researchers to “approach absolute quantitation,” Borchers said. “What we [want] is the concentration [of proteins], the copy number. … This is what the medical doctor wants to know. He doesn’t want to know [that] the PSA in your blood is three times more than your neighbor, but we don’t know what the neighbor’s condition is. What he wants to know is [whether] your PSA is above 4 ng/ml, so now you are at higher risk for developing prostate cancer. And this is what we can deliver with this kind of assay.”
Additionally, by designing hPDQ with a clear clinical bent, the chances of getting research funding increase. “It’s clear these days especially, [that] the NIH as a potential funding source will insist on a very close relation to disease,” Aebersold said. “However, it also should be pointed out that the general approach represented here is, of course, also valid equally for basic research, not only for human but for research in any species.”
Four resources would be needed to achieve the goals of hPDQ: a comprehensive database of proteotypic peptides for each of the human proteins, accompanied by data identifying the best peptides for mass spec measurement and “associated optimized mass spec instrument parameters;” at least two synthetic proteotypic peptides labeled with stable isotopes and available in accurately quantitated aliquots “for use as internal measurement standards for quantitation of each protein;” anti-peptide antibodies specific to the same two proteotypic peptides per target, and that should be able to bind the peptides with dissociation constants less than 1e-9; and robust and affordable instruments for quantitative analysis of small amounts of tryptic peptides and for sample preparation. Current triple-quadrupole platforms coupled with nanoflow LC systems, and MALDI systems suffice, the authors said.
Some important details about the project remain to be filled in, including where funding would come from. The authors say that a “substantial” share of financing for hPDQ would need to come from government and philanthropic sources, though there have not been any discussions yet with such sources.
Borchers added that he and his co-authors may explore the idea of partnering on the project with industry, particularly antibody and in vitro diagnostic firms. According to Borchers, what results from hPDQ could have commercial value because “we are making tools and these tools can be sold. … [T]hey can be very valuable for many other projects that are coming up.”
It also has not been determined yet who would coordinate hPDQ, though Borchers said he envisions a consortium comprising government agencies such as the National Institutes of Health or Genome Canada along with industry players taking the lead on the initiative.
Cheaper HPP Surrogate?
The hPDQ proposal comes as the HPP faces an unclear fate. The project was first proposed last spring by the Human Proteome Organization, which estimated it would span 10 years at a cost of $1 billion [See PM 05/09/08]. However, at HUPO’s annual conference in August, a variety of government and philanthropic funders said the project may be overly ambitious and its price too stupendous. [See PM 08/21/08].
As a result, focus has shifted to smaller and less-costly projects as alternatives to accomplishing some of HPP’s goals. For instance, HUPO’s new president, Young-Ki Paik, recently told ProteoMonitor that the South Korean government is expected soon to make a decision on a plan to map the protein-encoding genes of chromosome 13, the second-smallest in the genome [See PM 01/02/09].
[ pagebreak ]
Then there are existing projects, such as the Human Protein Atlas, which has for several years been using protein antibody-based methods to identify human proteins.
At the HUPO conference in August, Aebersold said, funders found HPP’s objectives “too diffuse and too broad, so they said it has to be tightened up, and I think [hPDQ] is a consequence of focusing down to one specific topic.”
hPDQ would serve also as a placeholder for HPP and would “enable individual biological researchers to measure defined collections of human proteins in biological samples with 1 ng/ml sensitivity and absolute specificity, at throughput and cost levels that permit study of meaningfully large populations,” or between 500 and 5,000 samples, a capability not currently achievable, the authors wrote.
Like HPP, hPDQ would not tackle things such as splice variants or post-translational modifications, but instead would focus on biological variation affecting protein expression.
“It is aimed at providing immediately useful capabilities of the human biology research community in a way that does not adversely impact funding for individual investigators and does not generate administrative constraints on their ability to set and change courses in the conduct of research,” the authors said.
De novo proteome-wide discovery, protein-protein interactions as well as protein interactions with other molecules, and the spatial arrangements of proteins in tissues and organs also would need to be addressed by a broader initiative such as HPP, they said.
But with HPP in purgatory, hPDQ could serve as an initiative to get it off the ground. “If you have all these antibodies we can do a much better job to tackle the Human Proteome Project,” Borchers said.