This story has been updated, and upon the subject's request, several scientific points discussed in the conversation have been clarified.
KIRKLAND, Wash., Dec 19 - The Isotope Coded Affinity Tagging system, developed at the University of Washington and recently licensed by Applied Biosystems, is considered one of the most promising technological developments to hit proteomics yet.
By allowing researchers to precisely measure not only the identity of proteins, but minute changes in their quantities, the developers of the technology, who spent two years at an lab at the University of Washington funded by the National Science Foundation and Merck, say that the reagents will help scientists to decipher the mysteries lurking in the human proteome.
Recently, GenomeWeb spoke to David Goodlett of the Institute for Systems Biology, one of the developers of ICAT, about this new tool and other magic proteomics potions that the scientists at ISB have up their sleeves.
GW : What’s new about ICAT reagent system and how is it different from other approaches?
Goodlett : Before 1999, nobody had come up with [protein quantitation] on a global scale. They were doing it one peptide at a time, or by 2d -gels, [where] quantitation was roughly based on spot intensity and the identity of the protein had to be determined in a separate experiment.
Last year, there were several reports. One was Brian Chait’s lab at Rockefeller University in New York, where they took two pools of cells and [grew] them. One would run in normal media. The second pool [was] put in ammonium sulfate salts enriched for Nitrogen 14. The two pools of proteins would be mixed and all of the peptides would differ by some mass unit.
But it’s possible that the in vivo approach itself, where we are changing the [protein] growth also changes something about the protein expression that we want to measure. Furthermore, this approach isn’t amendable to looking at solid tumors or protein in human plasma.
With the ICAT approach, you harvest cells or proteins from whatever source you want. It could be Kennewick Man; it could be anything. You’re going to extract the proteins, compare them to some standard that you want to compare them to and then label them after the fact.
GW : How do ICAT reagents work?
Goodlett : You have two pools of proteins. Pool A is treated with an ICAT reagent that is chemically normal. Pool B is treated with a similar reagent that is chemically identical except eight protons are replaced with eight deuteriums. That reagent is 8 Daltons heavier than the normal reagent.
After labeling with these reagents, the proteins are harvested, pooled, and treated with trypsin, an enzyme that cuts them at the amino acids lysine and arginine. Then [you apply] chromatography to separate out these mixtures of peptides.
In this assay, we use mass spectrometry because mass spectrometers are universal detectors. You only need mass to charge as a readout [to detect the quantity of a particular protein]. Say you treated the A [pool] with the D-8 reagent. In the mass spec [readout] you know that any signal that’s heavier by 8 Daltons came from that pool. Then you can get a feel for [whether] that particular protein is up- or down-regulated in that cancer cell related to the normal cell line.
GW : How exactly do the reagents label the proteins?
Goodlett : Most proteins have at least one cysteine. [The ICAT] reagent analyzes all of the proteins that have cysteine [by tagging the cysteine]. The technology of identifying proteins by mass spectrometry is sophisticated in that you only need one peptide from any given protein to identify that parent protein. By isolating out peptides [with cysteine tags], we avoid redundant identification.
One reason this is important is [that] there is something like six orders of magnitude in terms of protein expression in a cell. There are lots of low copy numbers of proteins and high copy numbers of proteins. Mass spec works to select ions in a top-down approach. Without isolating cysteine-containing peptides, you redundantly identify the highest copy number of proteins again and again. This molecule with the affinity label lets us significantly reduce the complexity and get at some of these lower abundance proteins.
GW : How does the ICAT system differ from 2-d gels?
Goodlett: The first stage of the 2-d gel, isoelectric focusing, is pretty limited in the amount of protein you can load on that separation. You won’t see low copy proteins like transcription factors that are important in a regulatory sense, essentially because they are masked by things like actin. Reducing the sample complexity [through ICAT] gets us into that realm where we can identify and start to sequence these lower signals.
GW : What technology besides ICAT still needs to be developed to actually enable scientists to decode the proteome and figure out how to use it to cure or prevent disease?
Goodlett : [ICAT] is something that can be used in the first wave of proteomics to really target the functional proteins we want to go after. You still have a whole lot of basic biology ahead of you to figure out what that protein is doing.
You have to have functional assays. There would be other generations of these ICAT reagents that target a particular class of enzyme or a particular post-translational modification like phosphate, so we might not always in the future be just be pulling out just cysteine-containing peptides. We could pull out other structural classes like phosphorylated proteins, which are involved in signal transduction with inside cells.
GW : Are you working on these reagents now?
Goodlett : It’s very early on in the game, but, yes, those reagents are being worked on. The paper, which involves labeling phosphate groups, was sent to Nature Biotechnology and is currently under review.
GW : Where else are you going in terms of proteomics?
Goodlett : The next phase is a high-throughput proteomics group that is supposed to be an incubator [for] application of these ideas in industry. We scale it up to some point and Celera and others scale it up to some absurd level. [[With Celera,] It’s really just having enough money to hire enough people to buy enough toys to do all the separation science and the sample preparation that comes before ever getting it to the mass spectrometer and analyzing the data.
GW : Is there a way to make it more efficient?
Goodlett : There certainly is. We think we have a way that would make it applicable in just the basic biology lab. But I really can’t talk about that right now. There are some patents that will be filed really soon.
GW : So that could really revolutionize proteomics.
Goodlett : Potentially. There are going to be lots of different approaches. The 2-d gels aren’t going to go away. In fact, Celera has licensed a technology called a molecular scanner from GeneProt. The idea is to label proteins in their two different pools, run them out on a 2-d gel then analyze them on the gel. As the proteins are blotted out of the gel onto a membrane, the membrane is impregnated with trypsin. You end up on the membrane with peptides that can be immediately analyzed with a mass spectrometer.
GW : Proteomics seems to be the latest buzzword, and suddenly genomics is almost passé. What do you make of that?
Goodlett : In the last century, publications in the Journal of Biological Chemistry were essentially characterizing one protein at a time. Now with the human DNA sequence we can characterize whole systems of proteins and their relative genetic changes between states immediately in one experiment. So there’s something tangible that’s not hype that’s there. Is it being oversold? Is there going to be a new flashword tomorrow? Probably.