| |
Adobe Reader 6.0 or higher is required to view our PDF documents.
|
| |
Global Proteomic Methods
Arguably, as a hypothesis-free technique, global proteomic profiling has proven useful for biomarker discovery, where the research objective is to find proteins associated with a disease state1 or toxicological reaction2, and there is no a priori biomarker or mechanistic knowledge. However, global proteomic methods have not yet proven cost-effective enough for biomarker validation or robust enough for routine clinical use.
Advantages and Limitations
The advantage of global proteomic methods is that no hypothesis is required, other than there should be a measurable difference in one or more protein species between diseased and healthy samples. Global proteomic methods attempt to separate and quantify all the proteins (and isoforms) from a clinical sample. Such methods require a great deal of time and expense and have yet to yield a commercial success. Global proteomic methods have failed to identify many clinically-relevant isoforms because:
- They often lack sufficient dynamic range to see low abundance biomarkers in clinical samples.
- The isoform information can be lost during the sample simplification steps prior to mass spectrometric analysis.
- They are too costly to apply to the number of samples necessary to validate a biomarker or for use in clinical settings.
- Because of the complexity of the techniques involved and the large number of sample preparation and protein separation (or sample simplification) steps, they lack the robustness for clinical use.
The Dynamic Range Issue
The most critical of these issues is that of dynamic range, in which the higher abundance proteins in the sample inhibit detection of the lower abundance proteins. Proteins exhibit a very wide range in concentration, with a dynamic range of 105 in bacteria3 to 107–108 in human cells4 to at least 1012 in human plasma.5 Since there is no technique to amplify low-abundance proteins comparable to the polymerase chain reaction for nucleic acids, both gel and mass spectrometry (MS) methods often fail to detect them.6 The limitations imposed by this dynamic range limitation in MS methods is shown as the solid green areas in the Figure.
This has led to the development of strategies to deplete the high abundance proteins. However, after depletion of human serum albumin (HSA), immunoglobulin G (IgG), and transferrin (Tf) additional gains in protein detection are hard fought (hashed green area in the figure) because there are hundreds of proteins left with similar concentrations.7 Depletion strategies also suffer from the problem of co-depletion of lower abundance biomarkers of interest,8 potentially eliminating them from further analysis. There are also currently no depletion technologies for tissue samples.
The Sample Simplification Challenge
Another issue with global proteomics is the protein complexity in clinical samples. Mass spectrometric methods can resolve a maximum of 2,000 individual peptides at a time, but there are an estimated 10-35,000 proteins in any given human cell type. A proteome consisting of 10,000 proteins will generate approximately 350,000 peptides. Therefore, peptides have to be separated prior to mass spectral analysis. For many proteomes, however, the number of peptides easily exceeds the resolution capacity of front-end separation techniques.9
Another way to reduce the peptide complexity before MS analysis is affinity enrichment of specific amino acid residues (e.g., metal chelation of His residues)10 or residue-specific chemical affinity tags (e.g., isotope-coded affinity tags).11 These methods greatly reduce the number of peptides to be examined and allow more complete resolution of such peptides during front-end chromatography steps, but isoform information is lost during the affinity step. The relative abundance of Cys residues in proteins (1.7–1.8%) is approximately the same as that of His (0.65–2.2%)12, 13, which translates to seven Cys or His residues in the average 450 amino acid-containing protein and a sequence coverage of less than 20% for the average protein.
____________________
1Petricoin, E.F. et al., “Use of proteomic patterns in serum to identify ovarian cancer,” Lancet 359:572–577 (2002).
2Steiner, S. and Anderson, N.L., “Expression profiling in toxicology – potentials and limitations,” Toxicol. Lett. 112-113:467–471 (2000).
3Ingraham, J.L. et al., Growth of the Bacterial Cell (Sinauer Associates, 1983).
4Anderson, N.L. and Anderson, N.G., “Proteome and proteomics: new technologies, new concepts, and new words,” Electrophoresis, 19:1853–1861 (1998).
5Corthais, G.L. et al., “The dynamic range of protein expression: a challenge for proteomic research,” Electrophoresis 21:1104–1115 (2000).
6Kenyon, G.L. et al., “Defining the mandate of proteomics in the post-genomics era,” Mol. Cell. Proteomics 1:763–780 (2002).
7Anderson, presentation given at the ASMS Asilomar conference (Oct., 2005), http://www.plasmaproteome.org/plasmaframes.htm.
8Righetti, P.G., “Proteomic Approaches for Studying Chemotherapeutic Resistance in Cancer,” Expert Review of Proteomics, 2:215-228 (2005).
9Ren D, Penner NA, Slentz BE, Regnier FE., “Histidine-rich peptide selection and quantification in targeted proteomics,” J. Prot. Res. 3:37–45 (2004).
10Ren D, Penner NA, Slentz BE, Regnier FE., “Histidine-rich peptide selection and quantification in targeted proteomics,” J. Prot. Res. 3:37–45 (2004).
11Gygi SP, et al., “Quantitative analysis of complex protein mixtures using isotope-coded affinity tags,” Nature Biotechnol, 17:994–999 (1999).
12Ingraham JL, Maaløe O, Neidhardt FC.Growth of the Bacterial Cell. (Sinauer Associates, MA, USA, 1983).
13Creighton TE. Proteins: Structures and Molecular Properties. (WH Freeman & Company, NY, USA, 1993).
|
|
|
