Poor sample preparation, inexact results, a relative inability to show dynamic processes, and a number of other factors have all contributed to the paucity of mass spec-based proteomics work by cell biologists, according to researchers at McGill University and Göteborg University.
As proteomics lurches out of its infancy as a research field, many say they are seeing more biologists, especially cell biologists, entering the fray. But to date, mass spec proteomics is still dominated by analytical chemists with biologists distinctly in the minority.
In a paper in the current edition of Nature Methods, available here, researchers from McGill in Toronto and Göteborg in Sweden examine why biologists continue to have an aversion to proteomics in general, and mass spec-based proteomics research specifically.
The paper is one of several in the current issue of the journal exploring the use of mass spec for proteomics research.
In the paper the researchers say there is no one overarching reason why cell biologists continue to have little presence in mass spec-based proteomics. Rather, a number of factors have prevented cell biologist from embracing proteomics. First, they say mass spectrometry lacks the “charm” of technology more often found in cell biology, such as live-cell video microscopy, that puts the instrument at an immediate disadvantage.
“With imaging we can observe dynamic processes — something that proteomics has a hard time rivaling,” they write.
But aside from a missing coolness quotient, the quality of current proteomics research may be a barrier to recruiting more cell biologists into the field. In particular, the authors write, shoddy sample preparation has resulted in a skeptical view of proteomics among cell biologists.
While issues with sample preparation are “well-founded,” rather than trying to address the problem head-on, “the proteomics community has instead attempted to circumvent the need for rigorous sample preparation by developing advanced bioinformatics tools,” the authors write.
“But this is only a very limited substitute for proper sample preparation and when combined with problems of data presentation that effectively prohibit critical assessment, such efforts have had a limited impact in the cell biology community,” they continue.
Improving sample-preparation strategies could alleviate some of the limitations of the mass spec instrument, they claim. For example, using a high-salt wash to remove peripheral proteins and then using detergent extraction and detergent-phase separation can expand the proteome of the endoplasmic reticulum by nearly 500 proteins and increase the order of magnitude between the most abundant and least abundant protein by three.
By incorporating such techniques — which have existed in the biochemistry and cell biology fields for decades — the proteomics community can improve sample preparation for proteomics studies, the authors say. Until the proteomics community addresses problems associated with sample prep, not only will the field suffer from a lack of credibility among cell biologists but from the larger scientific community, they say.
They are not alone. In June, Joshua LaBaer, director of Harvard Medical School’s Institute of Proteomics, told ProteoMonitor that the field has been dominated by a “gear head” mentality that has stressed too much attention on development of technology and not enough on expertise of what is going on biologically and how to evaluate the results [See PM 06/28/07].
“They have to spend hours looking for something and cleaning it up, and they’re not going to do that.”
“That still remains to me one of the biggest challenges we face. We get this scad of data, and then how do you turn that into useful biological interpretation?” LaBaer said at the time. “They need to get in touch with the biologists who can understand what to be looking for.”
He and others agree that the number of biologists doing proteomics work is increasing, and some of the most prominent figures in proteomics have backgrounds in biology including LaBaer, a physician by training; John Yates at the Scripps Institute; Matthias Mann at the Max Planck Institute of Biochemistry; and John Bergeron at McGill, one of the authors of the Nature Methods paper.
In August, Bergeron told ProteoMonitor that one of the major obstacles for cell biologists has been poorly designed protein databases that make extracting any useful information a draconian task.
“They have to spend hours looking for something and cleaning it up, and they’re not going to do that,” Bergeron said.
In Nature Methods, Bergeron and his co-authors reiterate that problem saying that “all presently available databases are fraught with redundancies, inconsistencies in nomenclature, fused genes, inappropriately translated introns, and so on. In many cases, nondescriptive … or multiple names are erroneously assigned.”
Finally, they say that proteomics is “not an exact technique,” which has been a “source of confusion for cell biologists.” Proteomics should not be viewed as a stand-alone and absolute research method, they say, noting that the development of a spatial and quantitative map of the proteome will need input from cell biologists, mass spectrometrists, and bioinformaticians.
“No strategy presently exists to bring data together nor to ensure that they have a meaning tomorrow,” the authors write. “Who should do this? We think that the cell biologists, having some experience in bridging different cultures — like yeast genetics with biochemistry and developmental biology — should have an active role in addressing this challenge”