What has proteomics accomplished?

The Barbados Principles, which emanated from the Barbados Conference in January 2007 and were reported in HUPO News (1), assigned to Ralph Bradshaw and John Bergeron the task of formulating a statement on “What has proteomics accomplished?” as a way of providing a progress report for the discipline and a guideline for future emphases. The amazing expansion in proteomics in the last five years is underscored by the number of international meetings, symposia, and workshops and on the growth in the scientific literature focused on this field. This is due in part to a major influx of physicists, chemists, and bioinformaticians who have helped catalyze the development and application of the principal supporting technologies utilized in proteomics research. At the same time, the biological community has made enormous strides in utilizing these tools to execute global analyses of protein expression with concomitant insight into the mechanisms of proteins of unknown function emanating from these analyses. Biologists are becoming comfortable with MS, which is the current backbone of the field, as well as other useful methods for protein separation and identification. Because a significant percentage of the genes in any genome are of unknown function, this remains a major challenge for proteomics, and it will, in turn, have a key role to play in filling these gaps in our knowledge. Budding yeast have been a surrogate model for the eukaryotic kingdom, and it is not surprising that the first comprehensive proteomics analyses have come from studies with yeast. The expression products of genes readily detected by a Western blot approach, as well as tagged versions of proteins made from corresponding open reading frames, have given a global analysis of protein expression (2). MS initially lagged behind the Western blot approach but is now able to match and even exceed it utilizing new quantitation technologies, as reported by Matthias Mann at a recent meeting in San Francisco (3) and in an article (4). That this is accomplished in a single MS experiment, as opposed to the 6,000 separate experiments for the Western blot data, is an amazing testimonial to the tremendous progress in current mass specbased technology and data analysis. Just as Magellan was the first to circumnavigate the globe with all the uncertainty of the resulting maps to document the voyage, navigation of the global protein-protein interaction networks in yeast has been one of the unassailable accomplishments of proteomics. As reported in these pages previously, from the HUPO congress highlights in Long Beach (5), there is only a small proportion (about 25%) of overlap between the two major efforts to accomplish this monumental journey through the yeast interactome (6, 7). However, considerable new insight into the biology emanating from these networks has been realized. This has especially been due to the superimposition of these interaction networks on global approaches to identify pathways by synthetic lethal screens. Here, the characterization of genes whose knock-out may be subtle or nonessential when coupled with the knock-out of another nonessential gene enables further insight into protein function (8). Striking biological discoveries are stemming from these approaches (e.g. Refs. 9–11). As the value and credibility (12) of this resource grows, these “gold” mines of data bases will undoubtedly lead to further discoveries. This is a particularly notable achievement of proteomics because the magnitude of protein-protein interactions in cells was simply grossly underestimated, if appreciated at all. As opposed to these large scale efforts, the generation of individual protein complexes by specific and careful sample preparation has also led to biological insight wherein MS is again the preferred technology to characterize these individual protein complexes. Two such efforts may be indicated as a sample of such discoveries. Using ICAT technology, the lab of Aebersold reported a new subunit of the RNA polymerase II complex (13). The detection of this subunit was at the limit of the technology of the time. Its importance has grown since the demonstration also by Aebersold and collaborators that this new subunit (TFB5) was mutated in a subset of patients suffering from trichothiodystrophy (14). A further noteworthy advance has been in the elucidation of proteins in association with mutant CFTR protein that leads to cystic fibrosis. Using a label-free quantitative approach, the Yates group with collaborators could elucidate a comprehensive characterization of all the molecular chaperones retaining CFTR in the endoplasmic reticulum (15). Since this leads to degradation of the CFTR protein, the patients suffer and die because of this over stringent quality control machinery in the endoplasmic reticulum. By using RNA interference technology the Yates collaborators, i.e. Balch and Kelly, could selectively remove each of the chaperones uncovered by MS methodology. The removal of one (Aha1) led to the astonishing observation that the mutant CFTR protein would now leave the endoplasmic reticulum and be successfully transported to the cell surface acting as a functional chloride channel and effectively “curing” the disease at the level of cells in culture. Of course, the deleterious effects of chaperone removal in the patient would outweigh the benefit of successfully transporting the mutant CFTR protein to the cell surface. Regardless, From ‡McGill University, Montreal, H3A 282, Canada and the §University of California, San Francisco, California 94143-0446 HUPO Views