Tools for Biomarker Discovery

Biomarker discovery experimental approach

Biomarkers are biological molecules—primarily proteins—that can be used to track a disease process, measure a drug response, or diagnose a disease. Biomarkers are of great interest to pharmaceutical and biotechnology companies who wish to accurately measure response to new treatments. Many drug developers, as well as contract research organizations that are associated with them, expend significant energy and resources to discover, identify, and measure novel biomarkers. “The idea behind biomarker discovery is that a researcher can take a biological sample and get more information out of it,” says Ian Pike, Ph.D., chief operating officer, Proteome Sciences, Surrey, U.K.

“With biomarker discovery, you have to set up an experimental approach up front,” says Jost Klawitter, Ph.D., chief technology officer, Eurofins Medinet, Washington, D.C., who adds that one could search for biomarkers in different biological systems, from cell culture to human; for example, plasma might be collected from a patient or an animal. Once collected, plasma typically has to undergo some sample processing steps that ultimately lead to the depletion of high abundance proteins such as albumin and globulins, which can mask the presence of less abundant, but more relevant, protein biomarkers in the sample. The most important criterion of any biomarker discovery approach is that there should be no assumptions about the biomarker to be discovered.

Tools for finding and identifying biomarkers

Biomarker discovery is accomplished by too many tools to cover in this article. However, this article will focus on bioanalytical technologies commonly known as -omics: a number of high throughput, unbiased approaches that, because of these features, are well suited for biomarker discovery. “Typically, ‘omics’ technologies are used as screening technologies for our biomarker discovery,” says Klawitter. “Principally, it is genomics, proteomics, and metabolomics.” In their proteomics approach, Klawitter and his colleagues focus on two-dimensional protein gel electrophoresis and two-dimensional liquid chromatography-based approaches.

Malcolm Ward, Ph.D., head of research, Proteome Sciences, Surrey, U.K., describes several different proteomics biomarker discovery workflows, as follows. Each workflow consists of essentially three components: protein separation, analysis by mass spectrometry (MS), and then bioinformatics in order to perform statistical and multivariate analysis on the data sets. For protein separation, the gold standard has historically been two-dimensional protein gel electrophoresis, which involves separation of proteins by isoelectric point (first dimension) and molecular weight (second dimension) by subjecting them to a pH gradient and an electric field, respectively, both of which are performed in a polyacrylamide gel. Following electrophoresis, the gels are typically treated with silver to visualize the proteins, which appear as spots of varying size, color, and intensity, and at different positions on the gel. “The relative intensity of a spot can be used to determine if there is a change in the concentration of a specific protein due to a treatment or disease process,” says Ward, who adds that the next step in the process is to identify the protein components within each spot using mass spectrometry. Typically this can be achieved by matching mass spectra to entries within protein databases—a bioinformatic approach.

“The second workflow involves the use of isobaric tags, which also involves the use of a mass spectrometer as the analytical tool,” says Ward. Here, the tags are designed to quantify the relative protein quantities between biological samples. The tags are first attached to the proteins and peptides at their free amino groups. Following separation by liquid chromatography, each individual peptide is then analyzed by tandem mass spectrometry. “In Tandem MS/MS experiments, the tags give rise to what we call reporter ions, which can be used for the quantification,” says Ward. 

Furthermore, “peptidomics is a subset of the proteomics world and is a common approach to biomarker discovery, for example in diabetes research as well as within the protease inhibitor space,” says Hans Dieter, chief technical officer, PS Research & Development GmbH, Germany. Peptidomics as proteomics need comprehensive screening tools, such as high-end mass spectrometry. After the discovery and validation of peptide biomarkers, they can be transformed into mass spectrometric assays or immunoassays useful for the diagnostic or pharmaceutical industries.

Biomarker validation, statistics

Potential biomarkers need to be validated using a number of methods. “You can perform statistical analysis to look for statistically significant differences in protein concentration between two different study groups,” says Klawitter, who adds that these statistical measures can then be used to determine which proteins are differentially expressed after treatment or disease, and the protein that is shown to be statistically significant is a potential biomarker. Further validation can be accomplished by using various targeted assays involving selective scanning mass spectrometry approaches or immunological techniques such as enzyme-linked immunosorbent assay (ELISA) and LC-MS/MS.

In summary, biomarker discovery is an experimental approach of great interest to the pharmaceutical industry. The primary tools of biomarker discovery are -omics technologies, which offer an unbiased method that often culminates in the generation of large amounts of data that must be analyzed using bioinformatic approaches. Statistical and biochemical methods are necessary to validate novel biomarkers.