Medical Device Link.

Originally Published IVD Technology March 2005

Commentary

Roger L. Lundblad, PhD, is an adjunct professor of pathology at the University of North Carolina at Chapel Hill, and Patricia M. Wagner is an assistant professor of English at Victor Valley College (Victorville, CA). The authors can be reached at lundbladr@bellsouth.net and wagnerp@vvc.edu, respectively.

Biochemistry has a new wunderkind: proteomics, which appears to have supplanted protein chemistry as a discipline. Commercial activity in proteomics has made great strides in developing analytical technologies.¹ Since a clear need exists for new diagnostic targets for such pathologies as ovarian cancer, bladder cancer, pancreatic cancer, and Alzheimer’s disease, it is hoped that proteomics will be able to identify such new targets. Even though current commercial activities in proteomics are focused on developing analytical technologies, there have been increasing efforts to develop clinical applications as well. For example, as of November 2, 2004, a PubMed search for clinical proteomics yielded 36 citations in 2001 and 159 in 2004. Such efforts to develop clinical applications are driven by a necessity to validate proteomics as a discipline and the perception of proteomic technologies as having the potential to provide new diagnostic tests.^2-5

Proteomic Technologies

The general approach in proteomics involves separating proteins using two-dimensional gel electrophoresis (2-D electrophoresis), or multidimensional liquid chromatography, and then identifying the separated proteins with mass spectrometry.⁵ Another approach in which mass spectrometry is used directly in a diagnostic test involves protein profiling using surface-enhanced laser desorption/ionization (SELDI) technology.⁶ In this approach, a population of proteins is isolated by binding to an affinity matrix and then analyzed by mass spectrometry, resulting in a protein profile.

While proteomics might consist of good or even great science, it remains to be seen whether it is uniquely useful for developing diagnostics. Nonetheless, the most promising results from clinical proteomics are related to diagnostic applications. This article will focus on the potential of proteomics to develop diagnostics.

In an interview published in IVD Technology, an IVD company executive emphasized the necessity of a new assay having a quantifiable effect on patient care.⁷ This is a critical consideration frequently missed by industry scientists who are seduced by the science at the expense of product development. Another way of looking at this is understanding that a new diagnostic should positively impact the course of therapy, and that it is not enough simply to obtain better data.

Identifying Biomarkers

A new diagnostic test could be based on identifying a biomarker by using proteomic technologies.⁸ A biomarker can be defined as a biological substance that changes in response to an underlying pathology. Some of the currently used protein biomarkers include analytes such as C-reactive protein, cholesterol, troponin, D-dimer, and prostate-specific antigen. Nucleic acid biomarkers include ERBB2 for synovial sarcoma and long DNA for colon cancer.⁹

While proteomic technologies might be uniquely useful in discovering protein biomarkers, such technologies as 2-D electrophoresis and mass spectrometry will unlikely prove to be directly applicable as general clinical diagnostic tools. Such technologies will more likely be used to identify a biomarker, followed by the development of an immunoassay, possibly within a microarray system.^{10, 11} Proteomics can therefore be pictured as technologies used for identifying new analytes for the diagnostics industry.

Developing Biomarkers into Diagnostics

Success in using proteomics to develop new diagnostics will require close collaborations between a strong basic science program and a strong clinical program. One group of researchers provides an example of such a successful collaboration.¹² Using the classical approach of 2-D electrophoresis coupled with mass spectrometry, these investigators identified a protein (calprotectin) that could differentiate between rheumatoid arthritis and osteoarthritis. An immunoassay was subsequently developed for clinical use. These investigators reduced the complexity of their 2-D electrophoretic analysis by using a narrow isoelectric focusing gradient gel from pH 4.0 to pH 7.0. The useful clinical application demonstrates the effective translation of complex proteomic data to the development of an enzyme-linked immunosorbent assay (ELISA), which was made possible by the close relationship between protein chemists and rheumatology experts. The clinical assay provides data critical for monitoring the use of etanercept, which is an expensive therapeutic.

Multidisciplinary Approaches

Combining data from other disciplines, such as histochemistry and cytochemistry, with proteomics is more valuable than using proteomic data alone for identifying biomarkers. The traditional approaches of histochemistry and cytochemistry are now represented by tissue microarrays and the use of specific probes.

Another group of investigators offers an example of the useful and essential interplay between these existing technologies and proteomics. These researchers reported the development of a technical triad for identifying biomarkers for cancer detection.¹³ This triad included immunohistochemistry, tissue microdissection, and protein chip technology, or SELDI mass spectrometry. Laser dissections provided samples of normal pharyngeal epithelium and tumor squamous epithelium that were extracted with a lysis buffer and analyzed by gel 2-D electrophoresis. After protease digestion/mass spectrometry, differentially expressed spots were established using a public database. Once the two proteins, calgranulin A and calgranulin B, were identified, monoclonal antibodies were then used to demonstrate the presence of these proteins in the tumor tissue.

This operational triad consisted of tissue microdissection (laser capture microdissection) to obtain samples for 2-D electrophoresis and mass spectrometry. Once the proteins are identified, monoclonal antibodies can be developed for immunohistochemical localization of the antigens in the tumors. This is another example of using proteomic technologies to identify biomarkers that could then be assayed with conventional technologies.

Profiling

The preceding examples are single proteins that might serve as biomarkers. The current data suggest that most pathologies result from changes in several proteins. Some changes might be generative, while others are more likely consequential. A pattern of multiple changes (up-regulation and down-regulation) can be obtained, most often with plasma or serum, and the subsequent analysis is referred to as profiling.¹⁴

While profiling might eventually prove useful, considerable skepticism emerged about this approach when a complex algorithm was used to correlate changes in multiple serum proteins as a diagnostic for ovarian cancer.¹⁵ While this example may represent an advance in diagnostics, more work is required, especially since there is an advantage over current diagnostic tests. Specifically, identifying individual components within the profile will permit the development of immunoassays, which would provide more reasonable assay information and be applicable within current laboratory information systems.

Validation Issues

Applying proteomic technologies will result in developing successful diagnostic products. However, using prefractionation prior to either high-pressure liquid chromatography or electrophoretic separation is critical for an effective use of proteomic technologies in biomarker identification and validation.¹⁶ A key is obtaining maximum reproducibility in the analytical process. Another major issue is defining what is normal for an analyte. Since normal levels of analytes in fluids such as blood or blood plasma are generally based on 10–40 subjects, depending on the variations in normal levels, a sufficient number of normal subjects should be included in a clinical proteomic study.

In addition, most of the early diagnostic approaches use analyte specific reagents, which brings another set of issues into the development of commercial diagnostics. Such issues concern the development of reagents, such as monoclonal antibodies or apatmers, that would be used in the assays.¹⁷ As noted above, current immunoassay technologies will most likely represent the vehicle for the assays of new biomarkers.

Conclusion

Proteomic technologies will offer effective tools in the development of diagnostic assays for new biomarkers. Such assays will be useful in both the diagnosis and prognosis of disease states that are using concepts being developed in personalized medicine and theranostics. However, despite such potential, the promise of proteomics should not be made at the expense of existing technologies.¹⁸ Existing technologies should not lose their value at the expense of new approaches.

References

1. DN Chakravarti, “From the Decline and Fall of Protein Chemistry to Proteomics,” Computational Proteomics Supplement 32 (2002): S2–S3.

2. LB Tabatabai, “Using Proteomics in Diagnostics,” IVD Technology 8, no. 7 (2002): 37–48.

3. IL Goldknopf, HR Park, and HM Kuerer, “Merging Diagnostics with Therapeutics Through Proteomics,” IVD Technology 9, no. 1 (2003): 39–43.

4. R Banks and P Selby, “Clinical Proteomics—Insights into Pathologies and Benefits to Patients,” Lancet 362 (2003): 415–416.

5. MA Baldwin, “Protein Identification by Mass Spectrometry,” Molecular & Cellular Proteomics 3 (2004): 1–9.
6. N Tang, P Tornatore, and SR Weinberger, “Current Developments in SELDI Affinity Technology,” Mass Spectrometry Reviews 23 (2004): 34–55.

7. HJ Loyda, “Putting New Assays to the Test,” IVD Technology 10, no. 2 (2004): 38–42.

8. SE Ilyin, SM Belkowski, and CR Plala-Salamón, “Biomarker Discovery and Validation: Technologies and Integrative Approaches,” Trends in Biotechnology 22 (2004): 411–416.

9. PD Wagner, M Verma, and S Srivastava, “Challenges for Biomarkers in Cancer Detection,” Annals of the New York Academy of Science 1022 (2004): 9–16.

10. R Park, “Clinical Applications Lead Growth in Immunoassay Testing,” IVD Technology 7, no. 9 (2001): 17.

11. W Kusnezow et al., “Antibody Microarrays: An Evaluation of Production Parameters,” Proteomics 3 (2003): 254–264.

12. S Drynda et al., “Proteome Analysis Reveals Disease-Associated Marker Proteins to Differentiate RA Patients from Other Inflammatory Joint Diseases with the Potential to Monitor Anti-TNFalpha Therapy,” Pathology Research and Practice 200 (2004): 165–171.

13. C Melle et al., “A Technical Triade for Proteomic Identification and Characterization of Cancer Biomarkers,” Cancer Research 64 (2004): 4099–4104.

14. JW Gillespie et al., “Molecular Profiling of Cancer,” Toxicologic Pathology 32, Supplement 1 (2004): 7–71.

15. KA Baggerly, JA Morris, and KR Coombes, “Reproducibility of SELFI-TOF Protein Patterns in Serum: Comparing Datasets from Different Experiments,” Bioinformatics 20 (2004): 777–785.

16. RL Moritz et al., “A Proteome Strategy for Fractionating Proteins and Peptides Using Continuous Free-Flow Electrophoresis Coupled Off-Line to Reversed-Phase High-Performance Liquid Chromatography,” Analytical Chemistry 76 (2004): 4811–4824.

17. S Gutman and DW Feigal Jr, “The Status of ASR Regulations,” IVD Technology 10, no. 4 (2004): 18–21.

18. R Longtin, “Out-of-Fashion: Non-Proteomic Research Vying for Attention,” Journal of the National Cancer Institute 95 (2003): 1034–1035.

Copyright ©2005 IVD Technology