Knowing the identity of a protein is not enough anymore. Because the concentration of a protein is vital to its function in the cell, a change in the level of a specific protein(s) can be indicative of a cellular process. Thus, it’s important that scientists be able to measure relative and absolute protein concentration. It used to be that a scientist would have to run two-dimensional (2D) gel electrophoresis, cut out the band, and measure the amount of protein in that band using mass spectrometry. The problem is that this method was neithervery sensitive nor very accurate.
“We actually used 2D gels when we started doing proteomics, says Hong Li, PhD, associate professor and director, Center for Advanced Proteomics Research, University of Medicine and Dentistry of New Jersey, Newark. “The volume of information coming from 2D gels is not what most PIs [principal investigators] would expect because most of the proteins that have changed are metabolic, heat shock, orhousekeeping proteins,” says Li.
However, improvements in proteomic methods based on the highly sensitive and accurate tandem mass spectrometry approach are making it possible to obtain relative and absolute quantities of proteins without using a gel. iTRAQ and iCAT are two of these major new advances in proteomics, and they represent a shift in the quantitative proteomics arena. But because these are new methods, there are stillsome kinks in them.
The iTRAQ kit contains four isobaric (identical mass) amine-reactive reagents that can label peptides in a protein digest and consequentlyallow for accurate identification and quantification of peptides via tandem massspectrometry. “The kit contains everythingyou need to do five sets of reactions, even thetrypsin to do the digest,” says Dominic Gostick,PhD, director of the proteomics TOF-MSproduct line for Applied Biosystems, FosterCity, Calif.
The general procedure for using iTRAQ is as follows. The proteins are cleaved into peptides, which are then differentially labeled with iTRAQ reagents. The labeled samples are then pooled, so that a comparison can be made between them. After combining the samples, they are typically subjected to MudPIT, which involves 2D liquid chromatography followed by tandem mass spectrometry. Each peptide that comes off the column is subject to identification via Applied Biosystems’ software packages, MASCOT and Protein Pilot, which base the identification on the pattern of specific peptide ion fragments in the MS spectrum.
After digestion, each protein might be represented by a number of peptides, says Darryl Pappin, PhD, scientific fellow, Applied Biosystems, the creator of iTRAQ. According to Pappin, each of these peptides will be labeled with all four iTRAQ reagents, each of which will be considered an independent measurement; all of the measurements are averaged. However, says Pappin, some proteins might be identified based on only a single peptide measurement. “At that end, it’s up to the investigator to decide where he wants to draw the line.”
A useful tool
“The experimental procedure [for iTRAQ] is very straightforward because the labeling chemistry works very well. And labeling efficiency is very important for how well the quantification works,” says Li. And of course, the name of the game in quantification is reproducibility. So if a quantitative method is not reproducible, it’s worthless.
“The reliability and reproducibility [of iTRAQ] is certainly good. We are starting to get coefficients of variation in the 9% region. So the accuracy and precision of that technique are a lot better than some of the other tools out there,” says Phillip Wright, PhD, professor of chemical and process engineering at the University of Sheffield, Sheffield, United Kingdom.
In addition to looking at coefficients of variation, it is possible to do a rough validation of iTRAQ results. “We have validated iTRAQ using the Enzyme-linked ImmunoSorbent Assay (ELISA) in two different studies,” says Jan Hirsch, MD, a physician and researcher in the Department of Anesthesia and Perioperative Care at the University of California at San Francisco. He goes on to say that this validation approach has its own pitfalls. ELISA might not detect all of the peptides detected by mass spectrometry, or mass spectrometry might assign additional peptides to the protein, he says.
iTRAQ has advantages over the previous technologies. “I do believe that iTRAQ is a very good method for quantitative proteomics, but I think the technology of isotope ratio mass spectrometry needs to be developed further,” says Hirsch. A general disadvantage of these techniques, such as iCAT and iTRAQ, is that they only provide relative quantification. So you can only compare relative abundances. Therefore, researchers can compare two different samples, but they only get a relative value for the concentration of an individual protein. A normalized control can help with the quantification, but will often provide information for only one, or a couple of proteins. “Isotope ratio mass spectrometry can however provide valuable information, for example, if samples (such as cell lysates) from differentially treated animals are compared.”
“Because we are interested in making quantitative comparisons, iTRAQ seems to be ideally suited [for us] because it allows four samples to be compared simultaneously and it not only allows for quantification, but also identification,” says Beerelli Seshi, MD, head of hematopathology, Harbor-UCLA Medical Center and Los Angeles Biomedical Research Institute, and professor of pathology at the David Geffen School of Medicine at University of California, Los Angeles.
iTRAQ is also more sensitive than previous methods for protein quantification. “We are viewing a lot of the protein changes that we were not able to see before,” says Li. This is one of the problems with 2D gel electrophoresis because of its resolution and low sensitivity. And high sensitivity is needed to be able to see changes in low-abundance proteins.
Of course, a major hallmark of the global proteomic approach is that it is unbiased. “We are not biased for the result we get,” says Seshi. He uses iTRAQ to label the proteomes of stromal cells from leukemic and normal bone marrow grown ex vivo. By comparing the differences in the levels of specific proteins from bone marrow singular stromal cells, he hopes to gain some insight into the influence of bone marrow microenvironment on the development of leukemia. Furthermore, because cancer is a multigenic disease, it is important to be able to study more than one protein at a time. Thus, a strong proteomic approach is needed.
Another major advantage of using iTRAQ is that it will allow for identification of any type of protein, including high molecular weight proteins, acidic proteins, and basic proteins, all of which are problematic when using alternative methods such as 2D gel electrophoresis. Furthermore, 2D gel electrophoresis is not particularly useful for resolving insoluble proteins such as membrane proteins, but the coupling of the iTRAQ-labeling system and mass spectrometry solves this problem.
Improvement needed
Proteomic studies produce a lot of data. “I think the data can become overwhelming,” says Seshi, “so we tried to tackle the data mining issues, streamlining the approaches to data analysis.” This is where the ability to do a powerful statistical analysis can come in handy.
“The data output doesn’t even include the basic information about the proteins. That is, molecular weight and isoelectric point. To get this information, I have to go around and around and it should not be that way.” Both of these parameters can be obtained via 2D gel electrophoresis very easily, but the information obtained from the coupling of iTRAQ labeling and tandem mass spectrometric analysis is so much richer.
Of course, this richness begets a high degree of data complexity, especially when looking at the total proteome, says Seshi. He explains that one of the disadvantages of proteomic technologies is that they are not geared for reaching the entire proteome because high-abundance proteins interfere with the detection and identification of low-abundance proteins.
Confounding the high volume of proteomics data is another problem associated with doing a shotgun-type workflow [such as iTRAQ]: the investigator has to identify the entire proteome every time they do an experiment, says Wright. He further explains that with 2D gels, one runs the gels, performs an image analysis, and then selects only a handful of protein spots for MS analysis. With 2D gels, he says, “you are not mapping the entire [proteome], but you are seeing differential changes.”
However, iTRAQ is not only useful for labeling whole proteomes but subcellular proteomes— those found within organelles and membranes—as well. “We are using iTRAQ in conjunction with phosphopeptide enrichment,” says Alexandra Jones, PhD, project scientist at the Sainsbury Laboratory, Norwich, UK. Jones is trying to characterize Arabidopsis phosphoproteomes and, for some fractions, such as plasma membranes, she labels them first with iTRAQ. “Our enrichment procedures are quite complex, with a lot of steps. But as long as you tag [the proteomes] early on, it does not matter how many steps you do subsequently.”
Despite some of its weaknesses, iTRAQ is a powerful tool for proteomics research. Continual improvements in its usability and technical specifications should make iTRAQ a musthave for anyone doing proteomics. PA.
iTRAQ workflow
iTRAQ Reagent Design
