PerspectivesAre you interested in submitting a Perspective Article? Be sure to read The Science Advisory Board's Editorial Guides for Perspective Articles. Click here. Clinical Proteomics: Applying Protein Research for Disease Prevention by Jim Brady, Ph.D. Successful treatment of cancer hinges on early detection. For example, the 5-year survival rate for localized breast cancer is 97%, compared to a rate of 23% for women with distant metastases1. Unfortunately, many oncology patients are diagnosed too late for effective treatment because current detection methods, such as mammography, MRI, and biochemical or immunological testing, often lack sufficient accuracy or sensitivity to identify malignant cancers in their early stages. The recent completion of the human genome sequence and the emergence of sophisticated new technologies for identifying DNA polymorphisms offer the hope that genetic testing will be able to identify many individuals who are likely to develop certain types of cancer before those cancers occur. However, cancers can arise from a variety of mutations, many of which will fall beyond the scope of common genetic tests. Furthermore, individuals who appear to be at high risk for developing cancer based on genetic testing must face difficult decisions about whether to undergo extreme prophylactic measures such as preventative surgery. Rather than focusing on genetic alterations that may lead to cancer, many researchers believe that changes in protein expression patterns are the most accurate way to identify cancers in their early stages and to determine the most effective courses of treatment. Until recently, protein-based diagnostic assays have been limited in scope, focusing on a small repertoire of tumor-associated antigens, such as PSA and CA125 tests for prostate and ovarian cancer, respectively. However, encouraging research developments reported within the past several years suggest that we may soon see tremendous improvements in protein-based diagnostics. Advances in laboratory instrumentation and data management technologies have led to the emergence of a new field called clinical proteomics, which applies high-throughput protein analysis techniques to identify protein expression patterns that are indicative of disease states. In contrast to existing diagnostic assays, which examine protein biomarkers one at a time, clinical proteomics is based on creating protein profiles that simultaneously detect hundreds or even thousands of proteins in a single assay. The effectiveness of clinical proteomics hinges on two technological components: rapid, multiplex protein detection assays and data analysis systems that use artificial intelligence to assimilate vast amounts of protein expression data from healthy and diseased individuals into clinically relevant data sets. Until recently, two-dimensional polyacrylamide gel electrophoresis (2D PAGE) was the dominant methodology for protein profiling. While 2D PAGE provides the capability to analyze hundreds of different proteins in a single experiment, the technique has inherent limitations that restrict its usefulness in a clinical setting. Drawbacks to 2D PAGE include limited throughput capabilities, requirements for large sample volumes, gel-to-gel variability, and the inability to measure low abundance proteins. Although advanced techniques such as differential in-gel electrophoresis (DIGE), which uses differentially labeled protein populations to compare different samples on the same gel, are extending the applicability of 2D PAGE, the future of clinical proteomics appears to lie in two other technologies: mass spectrometry and protein chips or protein arrays. Mass spectrometry analyzes proteins based on the mass-to-charge (m/z) ratio of ionized peptides and proteins. A typical mass spectrometer consists of an ionization source, a mass analyzer, and a detector for counting the number of analytes at each m/z ratio. In addition, mass spectrometers are often coupled with separation devices such as liquid chromatography instrumentation. Aside from providing rapid data about the various proteins that are present in a biological sample, mass spectrometry also offers information about posttranslational modifications that may be associated with a particular disease state. Recent innovations in mass spectrometry include isotope-coded affinity tagging (ICAT) to facilitate the analysis of complex protein mixtures and imaging mass spectrometry, which provides spatial analysis of protein patterns in tissue sections. Although current mass spectrometers are limited in their abilities to create reliable protein profiles from unprocessed biological samples, it seems likely that instruments with higher mass accuracy, increased dynamic range, and better resolution will eventually appear, greatly extending the usefulness of mass spectrometry in a clinical setting. One of the most promising applications for mass spectrometry, from a clinical proteomics perspective, involves detecting proteins that are captured on protein chips, as exemplified by Ciphergen’s ProteinChip system. Using a variety of hydrophobic, ionic, or metal affinity chromatographic matrices and different reaction conditions, ProteinChips can effectively segregate proteins in biological samples based on the proteins’ physical properties. Protein identification is accomplished through surface enhanced laser desorption ionization (SELDI) coupled with a time-of-flight (TOF) mass analyzer. The potential usefulness of this system is evidenced by reported findings from the Clinical Proteomics Program that was established as a joint venture between the National Cancer Institute and the Food and Drug Administration. Although the program is only two years old, researchers already have achieved impressive success in the prognosis and diagnosis of ovarian cancer using protein chips and SELDI-TOF detection systems. Another type of protein chip that is commonly used for protein profiling is the protein array. Similar in concept to the DNA microarray, most protein arrays consist of a matrix of capture agents that are attached to the surface of glass slide. Capture agents can be peptides, nucleic acid aptamers, antibodies, or other types of proteins. Researchers often use the same instrumentation for producing and analyzing protein arrays as they do for DNA microarrays. However, there are significant differences between nucleic acids and proteins that add an extra degree of challenge to protein array studies. A major problem with protein arrays relates to the method by which protein capture agents are attached to glass slides or other matrices. Unlike DNA, which is a two-dimensional molecule, proteins frequently lose functionality or antigenicity if there are even minor perturbations in three-dimensional structures. Since proteins easily become denatured upon contact with the substrate, protein array producers have struggled to develop surface chemistries that will maintain the structure of attached capture molecules while reducing background resulting from non-specific binding of analyte proteins. Another issue with protein arrays involves the availability of capture agents. One of the most common types of protein array is the antibody array, which consists of small amounts of many different antibodies or antibody fragments spotted in a microarray format. While it is easy to obtain large amounts of individual antibodies, there are few suppliers that can provide small quantities of the thousands of different antibodies that are needed to produce a single antibody array. Furthermore, there are many proteins for which antibodies are not currently available. Most experts believe that traditional methods of antibody production (i.e., animals and hybridoma cells) are not suitable for meeting the needs of antibody arrayers. Instead, bacterially produced antibodies, such as those generated using phage display libraries, seem to be the most promising approach for creating the diversity of antibodies that are required for protein profiling. Aside from the chip- and slide-based arrays described above, there are several other types of protein arrays that are being used for protein profiling. For example, bead arrays consist of capture agents that are attached to labeled microspheres, rather than to a planar surface. In this type of system, captured analyte molecules are detected by flow cytometry. Other protein array formats include microfluidic lab-on-a-chip assays and systems that utilize innovative detection methodologies, such as surface plasmon resonance. Cancer detection is only one of many potential clinical applications for protein profiling. Autoimmune disorders, infectious diseases, cardiovascular disease, and Alzheimer’s are just a few of the other areas that are likely to benefit from advances in proteomic analysis. In addition, researchers are using protein profiling for drug discovery as well as for monitoring the efficacy and side effects of therapeutic drug treatments. However, before protein profiling is adopted as a routine clinical methodology, reproducibility standards for proteomic patterns must be established based on empirical data from many different patient samples. Toward this end, powerful bioinformatics tools are currently being used to assimilate output from mass spectrometry and protein array studies into retrospective or prospective data sets. Those data sets, in turn, serve as input for artificial intelligence systems, which “learn” to recognize proteomic patterns associated with particular disease states or physiological parameters. As levels of reference data increase and as the accuracy and reproducibility of protein detection technologies improve, the field of clinical proteomics may radically alter our current approaches to disease diagnosis and treatment. 1American Cancer Society. Cancer Facts & Figures 2003 ### << Previous Next >> [ View All Perspectives ] |
|
