There’s More to Life Than DNA
Genomics often steals the headlines, but could metabolomics be the true deliverer of personalized medicine?
Kévin Contrepois, Michael Snyder |
The promise of personalized medicine is to accurately predict, diagnose and treat diseases for each individual by taking into account their biological uniqueness. To date, personalized genomes have been the main focus of such efforts, allowing the identification of disease-causing and drug-interacting variants and thus informing potential therapeutic decisions. This approach has been widely successful in the diagnosis and treatment of cancer and rare Mendelian and idiopathic diseases.
Although successful, the personal genomics approach is limited. Why? Because it does not capture biological information from the patient’s environment or lifestyle. Most common diseases, including metabolic and mental disorders such as diabetes and autism, are influenced by a variety of factors and cannot be predicted from DNA sequence only. For these reasons, there is significant value in including additional personalized information derived from transcriptomics (gene expression), proteomics (protein expression), and metabolomics (metabolite profile), which together help monitor biochemical changes reflective of genetic, environmental and lifestyle influences. Despite the fact that these technologies provide independently more accurate measurements of a current physiological state than genome sequencing, their combination is essential in unraveling the molecular mechanisms contributing to clinical outcomes, which could reveal more efficient patient-tailored therapeutic targets.
As a proof-of-principle, we have published a study integrating detailed molecular information from RNAs, proteins and metabolites with genomic sequences of a single individual monitored over a 14-month period (1). This multi-omics approach uncovered dynamic molecular and pathway changes across healthy and pre-diseased states, and resulted in the actionable diagnosis of medical risks. The same approach is currently being applied to a larger cohort of patients at risk for type 2 diabetes (2) and is expected to reveal integrative multi-omic signatures associated with the development of insulin resistance and glucose misregulation.
Metabolomics has a number of advantages over other established ‘omics’. First, it measures the closest molecules to biochemical phenotypes in a fast, inexpensive and unbiased way. Along with the realization that most common health problems (diabetes and cancer, for example) are associated with unique metabolic abnormalities/signatures, the field of metabolomics is attracting increased interest and holds great promises for understanding and diagnosing disease. In addition, metabolites are not only regulated by the host genome but also by RNA and protein activities, which means metabolomics contains integrated multi-level information. Importantly, when performed on body fluids (blood or urine), metabolomics captures information not only from the host, but also from their microbiome (the commensal microbes present in or on the host). This is critical, because the gut microbiome is now recognized as a major modulator of human health and a crucial player in regulating health/disease states. For instance, the gut microbiome is essential in digestion and synthesizes hundreds of essential metabolites, such as vitamins.
Despite its great potential, we feel metabolomics has been understudied compared with other ‘omics’. Indeed, we do not even know how large the human metabolome really is... Part of the challenge is that metabolites have highly diverse chemical properties, making it difficult to assess the complete metabolome of an individual with a single technology or experiment. In this context, we recently developed an optimized liquid-chromatography coupled to mass spectrometry (LC-MS) analytical platform to robustly profile urine and plasma metabolites (3). Such state-of-the-art LC-MS experiments can now routinely generate robust data sets that contain more than 10,000 metabolic features derived either from the host or the gut microbiota, among which most are still uncharacterized. Hence, a major remaining challenge in the field rests in the accurate identification of non-annotated peaks.
Thanks to its unique ability to monitor metabolites that not only contain dynamic information from the host, but also from their microbiome, metabolomics will undoubtedly become a key player for enabling personalized medicine. It will most likely enable the discovery of useful biomarkers for the onset and pathogenicity of human disease, but will also help us better understand human health and treat disease. Importantly, small molecule measurements are relatively tractable to adapt for clinical assays. For example, rapid home-test measurements of glucose and sodium levels are now possible, and several continuous monitoring assays now exist for the former.
In addition, information from many of these devices can be monitored directly on smartphones, enabling rapid feedback on metabolic state. In our view, metabolomics will lead to novel, useful biochemical tests that will profile critical components in individuals in real-time. Moreover, the integration of metabolite profiles with genomic, transcriptomic and proteomic data will produce a much clearer picture of health and disease states and significantly improve personal health management.
- R Chen, et al., “Personal omics profiling reveals dynamic molecular and medical phenotypes,” Cell 148, 1293-1307 (2012).
- “The integrative human microbiome project: Dynamic analysis of microbiome-host omics profiles during periods of human health and disease,” Cell Host & Microbe 16, 276-289 (2014).
- K Contrepois, L Jiang and M Snyder, “Optimized analytical procedures for the untargeted metabolomic profiling of human urine and plasma by combining hydrophilic interaction and reverse-phase liquid chromatography - mass spectrometry,” Molecular & Cellular Proteomics 14: 10.1074/mcp.M114.046508, 1–12, (2015).
Kévin Contrepois is a postdoctoral research scientist in the Snyder laboratory at Stanford University, California, USA, where he leads the metabolite and lipid profiling research. By integrating multi-omics data sets, he is interested in the discovery of biomarkers and in understanding the molecular mechanisms contributing to disease onset, with a special emphasis on host-gut microbiome interactions. He recently developed an optimized analytical platform involving the combination of HILIC- and RPLC-MS for the untargeted metabolic profiling of human urine and plasma. He received his Ph.D. from the University of Paris-Sud (France) where under the supervision of Carl Mann and François Fenaille, his doctoral research focused on identifying chromatin modifications associated with cellular senescence and understanding their biological implications. He developed an UHPLC-MS methodology on intact proteins to characterize and quantify histone variants and post-translational modifications in a high-throughput manner.
Michael Snyder is a Stanford Ascherman professor, the chair of the department of genetics and director of the Stanford Center for Genomics and Personalized Medicine, Stanford University, California, USA. He received his Ph.D. training at the California Institute of Technology and carried out his postdoctoral training at Stanford University. He is recognized internationally for major contributions that have revolutionized the field of genetics. He launched the first large-scale characterization of eukaryotic genes using a transposon tagging strategy. He is a co-inventor of Chip-chip and Chip-seq as well as the first chromosome tiling arrays. Snyder is also the inventor of RNA-Seq and paired-end sequencing for eukaryotic cells using high-throughput sequencing methods. Recently, he performed the first detailed longitudinal integrative personal omics profile (iPOP) of a person and used this information to assess disease risk and monitor disease states for personalized medicine.
