How To Personalize Medicine

The measurement of individual organic molecules to determine inborn errors of metabolism, diagnose disease and monitor therapy and recovery is not new; indeed, it has a long and illustrious history. Can continuing advances in metabolomics answer growing calls for understanding ‘health’ more deeply and drive personalized medicine development? I believe so.

Molecules that are routinely evaluated in the clinic include testosterone and other hormones; vitamin D3; peptides; classes of compounds related to metabolic function, such as amino acids, organic acids, sugars and acylcarnitines; and many other organic molecules, such as lipids. Yet, there are hundreds, and more likely thousands, of molecules that are routinely ignored, either because a clear association has never been made to genetic variation, disease or well-being, or because they have been too difficult to measure in an affordable multiplexed fashion.

Metabolomics – the nonbiased identification and quantification of all metabolites in a biological system – is on the verge of transforming measurement in the clinic. Today, metabolomics of human biology is a rapidly growing field precisely because it is crucial for the development of personalized medicine. Two particularly striking recent papers (1, 2) epitomize the importance of whole metabolome analysis.

Some background

Nuclear magnetic resonance (NMR) was applied to the problem of multiplexed metabolite measurement from the mid-1980s, and particularly by Jeremy Nicholson of Imperial College London. Initial successes included insight into the diagnosis and treatment of diabetes mellitus. Though quantitative, the sensitivity and lack of large scale multiplex analysis has generally found this technology lacking, and the prohibitive system size and associated cost per analysis indicates that NMR will not enter routine clinical chemistry labs any time soon.

On the other hand, recent rapid advances in another technology have enabled striking increases in the number of metabolites that can be routinely measured in small biological samples. That technology is quantitative ultra-fast, highly sensitive, high resolution/high mass accuracy liquid chromatography with tandem mass spectrometry. Hundreds of metabolites are now accessible, and a few thousand may soon be routinely measurable using LC-HRAM-MS/MS.

Interestingly, these LC-MS/MS systems were predominantly developed for proteomic analysis, mainly for quantitative sequence analysis of peptides resulting from protein mixture digestion. Entire human cellular proteomes of >10,000 proteins, including post-translational modifications such as phosphorylation, can be measured using today’s instrumentation and software. 

The protein analysis problem, which once seemed intractable, is now seen as relatively routine because all analytes are peptides, the sequence of which can be determined from gene coding. The metabolome is arguably a much more difficult analytical problem; organic metabolite structures are extremely diverse, are not predictable from genetic information, and every sample contains a multitude of non-human metabolites, for example, from bacteria and other organic material, or from food, drugs, and a multitude of other organic contaminants. So, how does LC-HRAM-MS/MS stack up to this challenge?

Recent advances

The first paper (1) describes genome-wide association studies (GWAS) that have identified a number of genetic loci associated with blood metabolite concentrations. These findings provide new functional insights for many disease-related associations, including those for cardiovascular and kidney disorders, type 2 diabetes, cancer, gout, venous thromboembolism and Crohn’s disease. The study advances our knowledge of the genetic basis of metabolic individuality in humans and generates many new hypotheses for biomedical and pharmaceutical research. This study only employed LC ion-trap MS/MS analysis, without the application of HRAM. Upcoming publications, where the application of HRAM has been added to the MS/MS methodology, will expand these associations dramatically.

The second paper (2) presents an integrative personal omics profile (iPOP) analysis that combines extremely high coverage of genomic, transcriptomic, proteomic, metabolomic, and autoantibody profiles from a single individual over a 14 month period. The iPOP analysis revealed various medical risks, including type 2 diabetes. It also uncovered extensive, dynamic changes in diverse molecular components and biological pathways across healthy and diseased conditions. This study, which does employ LC-HRAM MS/MS, albeit not the latest technology, demonstrates that longitudinal iPOP can be used to interpret healthy and diseased states by connecting genomic information with additional dynamic omics activity, such as metabolomics.

To continue with this dramatic progress, the desires of metabolomics researchers are straightforward: they want to identify every metabolite in their samples, they want to do it quickly, and they want to do it quantitatively. To the instrument community, this translates into demand for more resolution, more reproducibility and the ability to identify unknowns, even at very low concentrations. Researchers can expect rapid advances in separations and mass spectrometry technology to continue, addressing major challenges like the tremendous structural diversity of these small molecules, the need for robust analysis across large numbers of samples and the need to identify unknowns. Moving this new advanced metabolomics capability into the clinic will be inevitable.