technology (MudPIT) . Application of MudPIT to Saccharomyces cere-visiae was successful in the identification of 1484 proteins, of which 131 contained predicted transmembrane domains, and make the MudPIT approach superior to 2D gels for throughput. The authors used integral column/emitters packed with a biphasic bed (i.e., the columns were first packed with a RP, and then with a strong cation exchange (SCX), material). After initial sample loading of a shotgun (whole organism) proteome digest, peptides were step-eluted from the SCX portion of the column onto the RP bed. RP elution then followed, using a traditional water/acetonitrile gradient. This approach minimizes sample handling, a necessary requirement for the detection of low-abundance proteins. A number of variations of the multidimensional LC have been implemented . Lubman and co-workers employed isoelectric focusing as the first dimension with RP-HPLC as the second . Gygi and co-workers  used SCX as the first separation dimension with fraction collection. The second dimension was RP-HPLC with the vented column approach as described earlier. When the SCX is moved off-line from the RP-HPLC, the use of large-diameter SCX columns for milligram-scale sample injections is enabled and the dynamic range is increased.
Smith and co-workers have combined FTMS with ultrahigh-pressure LC for the use of 80-cm-long nanoscale columns for high throughput analysis (~105 peptide masses measured in one run)  with reliance on an accurate mass tag formalism. The relative protein expression ratios can be measured with the isotope-coded affinity-tag (ICAT) reagent and method . The ICAT strategy allows for relative comparison of proteomes from different cellular states and has been reviewed extensively elsewhere [67-69]. These contemporary approaches to MS-based proteomics will be especially powerful (and accessible to more nonexperts) for drug discovery efforts and in initiating studies that focus on a smaller scale through interplay with genetic data  and more sophisticated analysis of polymorphism . Further, the new nongel methods will enable more powerful biomarker diagnosis of (disease) phenotypes and facilitate a fundamental understanding of biological systems [63,69]. Much of this future work will be driven by the measurement of two-to-five peptides per protein and provide direct mass information on 5-50% of the primary sequence.
Although mass spectrometry has made significant strides in sensitivity and proteome coverage, a recent report by Weissman and co-workers indicates that significant challenges remain . With the use of a complete fusion, tandem-affinity-purification (TAP)-tag library for the genome of Saccharomyces cere-visiae, protein expression levels were quantitatively determined with Western blots by chemiluminescent detection. These results were compared to MS-based (MudPIT) results [54,73], and the MS data were strongly biased toward the detection of abundant proteins. For the 75% of the proteome that is represented by proteins present at fewer than 5000 copies per cell, only 8% were observed by MS. This landmark study will serve as an important benchmark for future MS method development.
Was this article helpful?