Metabolites, the small non-polymeric molecules, which are the intermediate or final products of metabolic reactions, have emerged as powerful tools for biomedical research and precision medicine.
The term metabolome, introduced in 1998, represents the whole entity of metabolites within a cell, tissue or entire organism.1 Since then, metabolomics has gained impressive traction, especially due to its relevance in systems biology or the holistic approach of deciphering the complexity of biological processes. Metabolomics fits perfectly in that concept, with its strategic position downstream of genomics and transcriptomics in the omics cascade and also tightly connected and influenced by the surrounding environmental factors. Often, the metabolome provides insights where genomic profiling fails to explain a given phenotype, and for this, the field is gaining appreciation from a plethora of pathology-related areas, including cancer, diabetes and cardiac disease.2
From a technical point of view, metabolite profiling relies on the principles of analytical chemistry. The main methods are mass spectrometry (ms), applied in a targeted or untargeted manner, nuclear magnetic resonance (NMR) and stable isotope labeling for matabolomic flux analysis. Each of these methods has its pros and cons and the choice depends on the study's final goals. 3, 4
The diverse nature of the metabolome continually challenges researchers and creates a colorful array of opinions on which is the most comprehensive and precise approach.
Two ways to go – targeted versus untargeted metabolomics
Currently, there are two approaches in the field: targeted and untargeted metabolomics. Targeted metabolomics is the quantitative measurement of defined and annotated metabolites using validated internal controls. Untargeted metabolomics is the comprehensive measurement of up to 10 000 metabolites in biological samples, allowing also the identification of unknown entities. Both methods have their advantages and limitations, making them appropriate for different research contexts.
Professor Christopher Newgard, the founding director of the Duke Molecular Physiology Institute, has spent more than 40 years in the field, applying targeted metabolomics to clinical use. His work has delivered insights into the metabolic aspects in various pathology fields, including diabetes and cardiovascular diseases, and has brought the method to an impressive level of comprehensiveness. "We've built up the targeted modules. Today we measure more than 400 analytes by targeted methods,” says Newgard. He is a strong supporter of using targeted metabolomics, and advocates for measuring metabolites in a precise and quantitative way, allowing ready application wards answering a biological question or predict a disease outcome or predisposition.5 "What I care about is being able to measure things in a reliable way and do it in a way where you can understand what's going on in biological systems," says Newgard.
Professor Garry Patti, group leader at the Washington University in St. Louis, reflects on a strength of untargeted metabolomics relative to targeted metabolomics with the analogy of someone "looking for his lost keys under the street light, although he lost them somewhere else where it is dark and he can't see anything, so the person is only looking under the light".
"That's kind of the way targeted metabolomics is. You're looking for certain molecules, but you're only looking under the lights. And if it turns out that this is where you've lost your keys, then it works extremely well." Patti says. Nevertheless, the scientist highly appreciates the significance of targeted metabolomics: "Targeted metabolomics is extremely useful in the clinical arena and other areas of biology," but also acknowledges the importance of identifying novel metabolites and making new discoveries. "The advantage of untargeted metabolomics is the potential for discovering molecules or metabolites that are altered in a particular disease state or physiological condition that are not expected."
Patti's lab leverages on identifying new metabolites using untargeted metabolomics and he is no stranger to the obstacles this approach offers and how it differs from the targeted analysis. "When you do untargeted metabolomics, you don't know what you're looking for. The metabolic spaces are pretty ambiguous, and we can't optimize methods in the classical way that you would do in a targeted experiment," he reflects.
The current major challenges metabolomics is facing
Scientists may be divided in their opinions on the appropriate method of choice, but they are certainly on the same page when it comes to the challenges and limitations their field is facing.
"I think it's really the data processing and the biological interpretation of the results that remain the major challenge currently in metabolomics. And that's what I see people mostly getting hung up on," says Patti. Although more critical towards untargeted method, Newgard sees the same issue of data interpretation and understanding: "I think too often untargeted metabolomics is used and descriptive information is disseminated, and then it's not followed up on."
Both researchers agree that the technologies they dispose of today are quite mature and sophisticated, and there is one main challenging question– how to interpret and understand the generated data. "What's still pretty tricky is trying to understand what all of those data mean. How do you convert it into something that's biologically meaningful?" adds Patti.
The different approaches to confront them
In an effort to overcome this challenge, scientists are applying different approaches. "We need software, and we need metabolomic databases," definite is Patti. Indeed, his group contributes to the advancement of the field by introducing sophisticated software solutions with various and specific applications for the metabolomics field. The TOXcms, for example, is being developed to facilitate analysis of untargeted metabolomics studies aiming to evaluate a drug's dose-response and the biochemical mechanisms of off-target effects.6
X13CMS, on the other hand, is a software tool designed by Patti and colleagues to track isotopic labels and assess metabolic fluxes.7
Newgard's team is largely focused on integrating quantification of metabolites into flux analysis. His team has previously collaborated to use NMR for flux analysis and has now developed internal tools to do this with ms.8 "It's all based on giving a stable isotope-labeled substrate, and then tracing it into products," he explains.
Newgard has a clear vision for the future and the potential of his method: "Where I see us going is always to have that strong foundation of targeted and quantitatively reproducible and rigorous profiling, but then link it to what's happening in metabolic pathways, through the ability to do metabolic flux analysis.”
The future role of metabolomics
With to the introduction of the Warburg effect, describing the metabolic shift of tumor cells towards aerobic glycolysis, cancer research is the leader in understanding and applying metabolomics to basic and clinical research.9 Other diseases affecting organs and processes known to be highly dependent on metabolic events, such as cardiovascular diseases and diabetes, are not lagging far behind. Today, understanding the metabolic regulation of any pathological condition gives great promise for highly personalized treatments. "I don't think there's any chronic human disease where metabolomics can't be an important investigative tool," says Newgard. "Metabolism is connected to every biological process, and metabolomics could be useful to every disease state," adds Patti.
When asked about the potential of metabolomics for precision medicine, Newgard says: "I think that the metabolome is the chemical phenotype and it reads out on all of the upstream genetic and other variations. I think it's logically a very powerful tool for precision medicine if it's done properly. (...) The problem is assembling interventional cohorts that are sufficiently comprehensive."
Although somewhat divided into their approaches and tools, metabolomics researchers are unified by their vision for a metabolomics-based approach to personalized medicine in the future, and by the common challenges they encounter and recognize. Their collective efforts will undoubtedly pave the way for deciphering the code of metabolomics.
1. Oliver S. Systematic functional analysis of the yeast genome. Trends Biotechnol. 1998;16(9):373-378. doi:10.1016/S0167-7799(98)01214-1.
2. Newgard CB. Metabolomics and Metabolic Diseases: Where Do We Stand? Cell Metab. 2017;25(1):43-56. doi:10.1016/j.cmet.2016.09.018.
3. Zamboni N, Saghatelian A, Patti GJ. Defining the Metabolome: Size, Flux, and Regulation. Mol Cell. 2015;58(4):699-706. doi:10.1016/j.molcel.2015.04.021.
4. McGarrah RW, Crown SB, Zhang G-F, Shah SH, Newgard CB. Cardiovascular Metabolomics. Circ Res. 2018;122(9):1238-1258. doi:10.1161/CIRCRESAHA.117.311002.
5. Kraus WE, Muoio DM, Stevens R, et al. Metabolomic Quantitative Trait Loci (mQTL) Mapping Implicates the Ubiquitin Proteasome System in Cardiovascular Disease Pathogenesis. Lusis AJ, ed. PLOS Genet. 2015;11(11):e1005553. doi:10.1371/journal.pgen.1005553.
6. Yao C-H, Wang L, Stancliffe E, et al. Dose-Response Metabolomics To Understand Biochemical Mechanisms and Off-Target Drug Effects with the TOXcms Software. Anal Chem. 2020;92(2):1856-1864. doi:10.1021/acs.analchem.9b03811.
7. Huang X, Chen Y-J, Cho K, Nikolskiy I, Crawford PA, Patti GJ. X 13 CMS: Global Tracking of Isotopic Labels in Untargeted Metabolomics. Anal Chem. 2014;86(3):1632-1639. doi:10.1021/ac403384n.
8. Lu D, Mulder H, Zhao P, et al. 13C NMR isotopomer analysis reveals a connection between pyruvate cycling and glucose-stimulated insulin secretion (GSIS). Proc Natl Acad Sci. 2002;99(5):2708-2713. doi:10.1073/pnas.052005699.
9. Warburg O, Wind F, Negelein E. The Metabolism Of Tumors in the body. J Gen Physiol. 1927;8(6):519-530. doi:10.1085/jgp.8.6.519.