Metabolomics: Clearing the logjam for biomarkers
The word “metabolomics” (or “metanomics”) first appeared in journal articles in 2000. Although many analytical chemists and biochemists will claim they have been practicing the science of identifying small molecules in biological matrices long before 2000, most agree that as a robust discovery tool, the technology is new and rapidly evolving. Since then, there has been an explosion of papers describing the technology as well as a significant number of papers using the technology for the discovery of biomarkers.
The challenge for metabolomics is to develop a technology that can extract, identify, and quantitate the entire spectrum of the small molecules (MW<1500Da) in any biological sample. While this is clearly a significant goal, the exact number of small molecules in biological samples is a hotly debated subject and, in some sense, is similar to answering the question of how many angels can dance on the head of a pin. Estimates range from approximately 1000 compounds to slightly over 10,000 compounds for the human metabolome. These estimates do not take into account xenobiotics, or externally introduced agents, which are highly prevalent in the human body.
Most importantly, this number is much smaller than the other “omics” technologies such as genomics or proteomics and may represent a significant advantage for metabolomics in biomarker discovery. Smaller numbers of total observed measurements for any individual leads to the application of more robust statistical testing methodologies and results in fewer false discoveries; two factors that have plagued other biomarker discovery technologies.
Based on the analysis of small molecules, metabolomics analysis and interpretation is based on biochemistry, which is a highly developed field of science, established long before molecular biology and proteomics. While the newer technologies have received the vast majority of attention in recent years, it is interesting to note that many of the Nobel Prizes in Medicine before 1980 were awarded in biochemistry.
Today, we routinely depend upon metabolite-based disease diagnosis. For instance, high glucose in urine was one of the earliest tests for diabetes, and cholesterol is used to measure the risk of heart disease. Clearly, the field of metabolomics is positioned to take advantage of this repository of biochemical pathway knowledge.
The development of metabolomics as a tool to leverage this body of knowledge, however, has not been straightforward. Most organizations have focused on more targeted metabolomic analysis, specializing in the measurement of 20-100 different metabolites which, most commonly, are within a common class of compounds. For instance, a number of companies have developed methods for detecting lipid compounds and, although these represent a smaller set of the total biologically relevant metabolites, this data has proven very useful for biomarker discovery efforts.
Other groups have focused on methods to truly investigate all of the small molecules in samples. This approach has been plagued by several difficulties because the physical properties of small molecules can vary greatly, with significantly variable solubilities and molecular weights ranging from 20 to 1000 Da. It is, therefore, difficult to develop any single chromatography method to separate all of the compounds and even more difficult to analyze individual compounds without chromatographic separation. Further complications arise if studies are expected to be completed with a reasonable turn-around.
These issues are currently being addressed through advanced multi-system approaches where the best separation and detection systems are being developed to run in tandem. This approach allows for a comprehensive solution achieved by combining values offered by various best-in-breed technologies.
As this new technology develops and its use in biomarker detection studies increase, it is rapidly becoming clear that metabolomics will likely represent a highly impactful technology in various healthcare-related fields such as the diagnosis of disease, identifying drug targets, evaluating the effects of drugs, and selecting patients who are likely to respond to drug therapy (i.e., personalized medicine).
Many biology effects of drugs and disease result from the overall health of an individual, as well as his or her environment and lifestyle and diet. While genomics can play an important part of predisposing an individual to drug side effects and disease, the biochemistry of an individual is likely a more important measurement of one’s current state and condition.
Despite all the successes genomics and proteomics have had in helping to understand the fundamental nature of disease and guide therapeutic development, these fields will likely never become the universal source of biomarkers as originally envisioned. Will metabolomics be more powerful for biomarker discovery? If history is any measure of success than the answer could quite possibly be yes.