Currently investigators are pursuing three different approaches to develop biomarkers with increased sensitivity and specificity. The first is to improve on a currently used biomarker. For instance, specificity and sensitivity of PSA may be improved by measurement of its complex with alpha(1)-antichymotrypsin; patients with benign prostate conditions have more free PSA than bound, while patients with cancer have more bound PSA than free.29 This difference is thought to result from differences in the type of PSA released into the circulation by benign and malignant prostatic cells. Researchers are also trying to improve the specificity and sensitivity of PSA by incorporating age- and race-specific cut points and by adjusting serum PSA concentration by prostatic volume (PSA density). The second approach is to discover and validate new biomarkers that have improved sensitivity and specificity. Many investigators are actively pursuing new biomarkers using a variety of new and old technologies. The third approach is to use a panel of biomarkers, either by combining several individually identified biomarkers or by using mass spectrometry to identify a pattern of protein peaks in sera that can be used to predict the presence of cancer or other diseases. High-throughput proteomic methodologies have the potential to revolutionize protein biomarker discovery and to allow for multiple markers to be assayed simultaneously.
In the past, researchers have mostly used a one-at-time approach to biomarker discovery. They have looked for differences in the levels of individual proteins in tissues or blood from patients with disease and from healthy individuals. The choice of proteins to examine was frequently based on biological knowledge of the cancer and its interaction with surrounding tissues. This approach is laborious and time consuming, and most of the biomarkers discovered thus far do not have sufficient sensitivity and specificity to be useful for early cancer detection. A mainstay of protein biomarker discovery has been two-dimensional gel electrophoresis (2DE). The traditional 2DE method is to separately run extracts from control and diseased tissues or cells and to compare the relative intensities of the various protein spots on the stained gels. Proteins whose intensities are significantly increased or decreased in diseased tissues are identified using mass spectrometry. For example, 2DE was recently used to identify proteins that are specifically overexpressed in colon cancer.30 The limitations of the 2DE approach are well known: the gels are difficult to run reproducibly, a significant fraction of the proteins either do not enter the gels or are not resolved, low-abundance proteins are not detected, and relatively large amounts of sample are needed. A number of modifications have been made to overcome these limitations, including fractionation of samples prior to 2DE, the use of immobilized pH gradients, and labeling proteins from control and disease cells with different fluorescent dyes and then separating them on the same gel (differential in-gel electrophoresis; DIGE). An additional difficulty is contamination from neighboring stromal cells that can confound the detection of tumor-specific markers. Laser capture microdissection (LCD) can be used to improve the specificity of 2DE, as it allows for the isolation of pure cell populations; however, it further reduces the amount of sample available for analysis. Even with these modifications, 2DE is a relatively low throughput methodology that only samples a subset of the proteome, and its applicability for screening and diagnosis is very limited.
A number of newer methods for large-scale protein analysis are being used or are under development. Several of these rely on mass spectrometry and database interrogation. Mass spectrometers work by imparting an electrical charge to the analytes (e.g., proteins or peptides) and then sending the charged particles though a mass analyzer. A time of flight (TOF) mass spectrometer measures the time it takes a charged particle (protein or peptide) to reach the detector; the higher the mass the longer the flight time. A mixture of proteins or peptides analyzed by TOF generates a spectrum of protein peaks. TOF mass spectrometers are used to analyze peptide peaks generated by protease digestion of proteins resolved on 2DE. A major advance in this methodology is matrix-assisted laser desorption ionization (a form of soft ionization), which allows for the ionization of larger biomolecules such as proteins and peptides. TOF mass spectrometers are also used to identify peptides eluted from HPLC columns.
With tandem mass spectrometers (MS/MS), a mixture of charged peptides is separated in the first MS according to their mass-to-charge ratios, generating a list of peaks. In the second MS, the spectrometer is adjusted so that a single mass-to-charge species is directed to a collision cell to generate fragment ions, which are then separated by their mass-to-charge ratios. These patterns are compared to databases to identify the peptide and its parent protein. Liquid chromatography combined with MS or MS/MS (LC-MS and LC-MS/MS) is currently being used as an alternative to 2DE to analyze complex protein mixtures. In this approach, a mixture of proteins is digested with a protease, and the resulting peptides are then fractionated by liquid chromatography (typically reverse-phase HPLC) and analyzed by MS/MS and database interrogation. A major limitation to this approach is the vast number of peptides generated when the initial samples contain a large number of proteins. Even the most advanced LC-MS/MS systems cannot resolve and analyze these complex peptide mixtures, and currently it is necessary to either prefractionate the proteins prior to proteolysis or to enrich for certain types of peptides (e.g., phosphorylated, glycoslylated, or cysteine containing) prior to liquid chromatography.
Although the use of mass spectrometry has accelerated the pace of protein identification, it is not inherently quantitative and the amounts of peptides ionized vary. Thus, the signal obtained in the mass spectrometer cannot be used to measure the amount of protein in the sample. Several comparative mass spectrometry methods have been developed to determine the relative amounts of a particular peptide or protein in two different samples. These approaches rely on labeling proteins in one sample with a reagent containing one stable isotope and labeling the proteins in the other sample with the same reagent containing a different stable isotope. The samples are then mixed, processed, and analyzed together by mass spectrometry. The mass of a peptide from one sample will be different by a fixed amount from the same peptide from the other sample. One such method (isotope-coded affinity tags; ICAT) modifies cysteine residues with an affinity reagent that contains either eight hydrogen or eight deuterium atoms.31 Other methods include digestion in 16O and 18O water and culturing cells in 12C- and 13C-labeled amino acids.
Although the techniques described thus far are useful for determining proteins that are differently expressed in control and disease, they are expensive, relatively low throughput, and not suitable for routine clinical use. Surface-enhanced laser description ionization time-of-flight (SELDI-TOF) and protein chips are two pro-teomic approaches that have the potential to be high throughput and adaptable to clinical use. In the SELDI-TOF mass spectrometry approach, protein fractions or body fluids are spotted onto chromatographic surfaces (ion exchange, reverse phase, or metal affinity) that selectively bind a subset of the proteins (Ciphergen® ProteinChip Arrays). After washing to remove unbound proteins, the bound proteins are ionized and analyzed by TOF mass spectrometry. This method has been used to identify disease-related biomarkers, including the alpha chain of haptoglobin (Hp-alpha) for ovarian cancer32 and alpha defensin for bladder cancer. Other investigators are using SELDI-TOF to acquire proteomic patterns from whole sera, urine, or other body fluids. The complex patterns of proteins obtained by the TOF mass spectrometer are analyzed using pattern recognition algorithms to identify a set of protein peaks that can be used to distinguish disease from control. With this approach, protein identification and characterization are not necessary for development of clinical assays, and a SELDI protein profile may be sufficient for screening. For example, this method has been reported to identify patients with Stage I ovarian cancer with 100% sensitivity and 95% specificity.27 Similar, albeit less dramatic, results have been reported for other types of cancer.2833-36 At this time, it is uncertain whether SELDI protein profiling will prove to be as valuable a diagnostic tool as the initial reports have suggested. A major technical issue is the reproducibility of the protein profiles. Variability between SELDI-TOF instruments, in the extent of peptide ionization, in the chips used to immobilize the proteins, and in sample processing, can contribute to the lack of reproducibility. There is concern that the protein peaks identified by SELDI and used for discriminating between cancer and control are not derived from the tumor per se but rather from the body's response to the cancer (epiphenomena) and that they may not be specific for cancer; inflammatory conditions and benign pathologies may elicit the same bodily responses.37,38 Most known tumor marker proteins in the blood are on the order of ng/ml (PSA above 4 ng/ml and alpha fetoprotein above 20 ng/ml are considered indicators of, respectively, prostate and hepatocellular cancers). The SELDI-TOF peptide peaks typically used to distinguish cancer from control are relatively large peaks representing proteins present in the serum on the order of ^g to mg/ml; these protein peaks may result from cancer-induced proteolysis or posttranslational modification of proteins normally present in sera. Although identification of these discriminating proteins may not be necessary for this "black-box" approach to yield a clinically useful diagnostic test, identifying these proteins may help elucidate the underlying pathology and lead to improved diagnostic tests. Potential advantages of the SELDI for clinical assays are that it is high throughput, it is relatively inexpensive, and it uses minimally invasive specimens (blood, urine, sputum).
Interest in protein chips in part reflects the success of DNA microarrays. While these two methodologies have similarities, a number of technical and biological differences exist that make the practical application of protein chips or arrays challenging. Proteins, unlike DNA, must be captured in their native conformation and are easily denatured irreversibly. There is no method to amplify their concentrations, and their interactions with other proteins and ligands are less specific and of variable affinity. Current bottlenecks in creating protein arrays include the production (expression and purification) of the huge diversity of proteins that will form the array elements, methods to immobilize proteins in their native states on the surface, and lack of detection methods with sufficient sensitivity and accuracy. To date, the most widely used application of protein chips are antibody microarrays that have the potential for high-throughput profiling of a fixed number of proteins. A number of purified, well-characterized antibodies are spotted onto a surface and then cell extracts or sera are passed over the surface to allow for the antigen to bind to the specific, immobilized antibodies. The bound proteins are detected either by using secondary antibodies against each antigen or by using lysates that are tagged with fluorescent or radioactive labels. A variation that allows for direct comparison between two different samples is to label each extract with a different fluorescent dye, which is then mixed prior to exposure to the antibody array. A significant problem with antibody arrays is lack of specificity; the immobilized antibodies cross react with proteins other than the intended target. The allure of protein chips is their potential to rapidly analyze multiple protein markers simultaneously at a moderate cost.
As discussed earlier, most currently available cancer biomarkers lack sufficient sensitivity and specificity for use in early detection, especially to screen asymptomatic populations. One approach to improve sensitivity and specificity is to use a panel of biomarkers. It is easy to envision how combining biomarkers can increase sensitivity if they detect different pathological processes or different stages of cancer, and one factor to consider in developing such a panel is whether the markers are complementary. However, simply combining two biomarkers will more than likely decrease specificity and increase the number of false positives. Reducing their cutoff values (the concentration of a biomarker that is used as an indication of the presence of cancer) can be useful to reduce the number of false positives. A useful test for evaluating a single biomarker or panel of biomarkers is the receiver operating characteristic (ROC) curve. An ROC curve is a graphical display of false-positive rates and true-positive rates from multiple classification rules (different cutoff values for the various biomarkers). Each point on the graph corresponds to a different classification rule. In addition to analyzing individually measured markers, ROC curves can be used to analyze SELDI-TOF proteomic profiles.39
The measurement and analysis of biomarker panels will be greatly facilitated by high-throughput technologies such as protein arrays, microbeads with multiple antibodies bound to them, and mass spectrometry. It is in these areas that a number of companies are concentrating their efforts, as not only must a biomarker or panel of biomarkers have good specificity and sensitivity, there must be an efficient and cost-effective method to assay them.
Was this article helpful?
Learning About 10 Ways Fight Off Cancer Can Have Amazing Benefits For Your Life The Best Tips On How To Keep This Killer At Bay Discovering that you or a loved one has cancer can be utterly terrifying. All the same, once you comprehend the causes of cancer and learn how to reverse those causes, you or your loved one may have more than a fighting chance of beating out cancer.