All of the small molecules involved in an organism’s metabolism make up its metabolome. When it comes to all of the ’omes—genomes, proteomes, lipidomes, and so on—the metabolome might be the most talked about in recent years. Although the term, metabolome, was coined almost 20 years ago, the research is rolling today more than ever.
This work often involves liquid chromatography (LC) and mass spectrometry (MS). “It’s common to detect thousands of signals or features in a typical LC-MS-based metabolomics experiment,” says biochemist Gary Patti at Washington University in St. Louis. “My thinking has really evolved in recent years about what all of these signals mean.” He used to think that lots of metabolomic signals just meant lots of metabolites and unknowns. Now, he says, “I see that the number of metabolites we are measuring in these experiments is much smaller than I expected.” He adds, “I think it is good news for the field, because it means that the number of unknowns isn’t nearly as overwhelming.” Patti’s work on detecting changes in an organism’s metabolome earned him the 2017 Agilent Early Career Professor Award.
Even with fewer metabolites than Patti expected, scientists need ways to sort through the signals. This started with software applications, like CAMERA, that combined signals that represent the same metabolite. “Recently, more advanced approaches have been developed for other signal redundancies and strategies, like credentialing, have emerged to find artifacts and contaminants,” Patti explains.
Exploring both cases of unknowns requires a growing set of resources.
To find unknown metabolites in a sample, a variety of tools can be combined. “The use of databases and fragmentation libraries help to facilitate the identification of known-unknowns, or those endogenous metabolites, which we know exist in nature,” says Amanda Souza, metabolomics program manager in chromatography and mass spectrometry at Thermo Fisher Scientific. “At the same time, there are unknown-unknown compounds that are truly novel and yet to be discovered.” Exploring both cases of unknowns requires a growing set of resources.
Sophisticated spectrometry
To put metabolomics to more use in healthcare, scientists seek more speed. “We believe, with the growing interest in precision medicine, biomedical researchers are becoming increasingly dependent on fast, robust, and accurate mass spectrometry-based technologies for comprehensive data collection on an industrialized scale,” says Tom Knapman, global marketing manager for academia/omics at SCIEX. “This work demands very-high throughput and accuracy for the rapid and reliable identification and quantification of every compound in thousands of samples. So, SCIEX is heavily focused on developing technologies for conducting targeted screening of known compounds or carrying out discovery research to globally study the metabolome.”
For example, SCIEX provides its data-independent SWATH Acquisition technology, which Knapman says, “simultaneously allows comprehensive identification and quantification of virtually every detectable compound in a sample from a single analysis.”
In a joint study performed by California-based Genentech—a subsidiary of Roche—and SCIEX, SWATH Acquisition was used to perform qualitative characterization and quantitative measurement of metabolic flux in cell-culture assays. “This shows that the SWATH Acquisition approach allows MS-MS of every single precursor and is not biased to an abundant compound,” Knapman explains.
Biocompare’s Mass Spec Search Tool
Find, compare and review mass
spectometers from different suppliers Search
Other methods also come in handy when looking for unknowns. “Particularly helpful is fragmentation or dissociation of ions,” Souza says. “Precursor ions are fragmented giving rise to product ions—MS/MS—of a certain pattern or profile, and these product ions can in turn be associated to chemical structures or portions of the structure.” Even the fragments ions can be further fragmented (MSn). “To enhance this capability, there are different dissociation techniques available that generate different fragmentation patterns, such as ultra-violet photodissociation (UVPD),” Souza explains. “Further, Orbitrap technology provides ultra-high resolution, and more recently up to one million resolving power, to aid in the determination of isotopic signatures to predict elemental composition providing a complimentary angle toward identifying unknowns.”
Advancing applications
Beyond advancing the technology behind metabolomics, scientists look for more ways to apply it. At Metabolon, chief scientific officer Mike Milburn sees clear trends emerging in the past couple years with metabolomics applied to: population studies to identify biomarkers of disease phenotype and disease mechanisms in cohorts with hundreds to thousands of individuals; precision medicine research and diagnostic use; and areas that have been predominantly about genomics technologies, such as microbiome research.
Metabolites play a fundamental role in a person’s phenotype. Image courtesy of Metabolon.
“Metabolomics is increasingly being used because there is a need to measure phenotype, rather than infer it or calculate a lifetime risk of developing a disease, which may or may not actually develop,” Milburn explains. “With metabolomics—particularly with a technology platform such as Metabolon’s that comprehensively measures metabolism—you have the capability to both discover and then monitor molecular phenotypic change longitudinally.” Longitudinal studies make it possible to track the development of diseases and possibly identify early indicators of an illness.
Metabolon’s platform employs fast LC and high-resolution MS. This provides “the sensitivity and power to identify about 1,500 metabolites in plasma,” Milburn says. “We combine that with an extensive software infrastructure that sifts through the massive amounts of data these instruments produce.” The company also provides standard sample-handling protocols, quality controls, and bioinformatics tools to ensure the reliability and reproducibility of the data.
Milburn points out that data interpretation makes all of the difference. He says that they have “built a knowledge infrastructure and tools that support a scientist’s ability to obtain actionable information from the data, rather than just looking at a large spreadsheet with a list of assorted metabolites.”
Moving ahead
To learn even more from the metabolome, scientists must separate signals from background. “I worry that a lot of time and money are being wasted trying to identify ‘unknown metabolites’ that are actually noise, contaminants, or artifacts,” Patti says. “It would be most useful if we could develop technologies to rapidly remove noise, contaminant, and artefactual signals from datasets, which I think might represent as many as 90% of the total signals in some LC/MS experiments.”
The next big steps in biological and medical research could come from combining metabolomics data with other kinds of information. “Integrated ’omics data is the holy grail for many researchers to give the whole individualized picture in areas such as precision medicine and disease research,” says Knapman. “With the industrialization of metabolomics data allowing very large datasets to be collected, there is a real need to integrate data generated using metabolomics approaches with both genomic and proteomic data.”
Approaches like Patti’s and those noted by the vendors interviewed here help scientists find meaningful information in metabolomics data, and these tools will change how that data can be used.
Image: Shutterstock Images