Thrust onto the frontline of biomarker discovery and personalized medicine, metabolomics is quickly yielding results powered by technological advancements. Studying metabolites involves separating complex mixtures of biomolecules, so the techniques commonly used are chromatographic separation — typically liquid chromatography (LC) — followed by mass spectrometry (MS). The diverse chemistries and abundances of metabolites present a challenge, but advances in instrumentation and software improvements are helping to overcome these. This article will discuss how such tools are helping metabolomics to become instrumental in biomarker discovery and personalized medicine.
Advances in essential tools
Tools for analyzing metabolites have become more sensitive, and today they can pull more data out of samples than ever before. “When Metabolon first started about 20 years ago, we were able to identify about 100 compounds,” says Annie Evans, director of R&D at Metabolon. “Today, due to the improved chromatographic and MS technology, we can identify over 2,000 compounds.”
This year Thermo Fisher Scientific launched two new mass spectrometers, the Thermo Scientific™ Orbitrap Exploris™ 240 and Thermo Scientific™ Orbitrap Exploris™ 120 instruments, which are well suited for analyzing metabolic signatures from complex samples. The Orbitrap Exploris 240 mass spectrometer includes the fully-automated AcquireX intelligent data acquisition workflow, which boosts the ability to annotate unknown metabolites.
Data analysis plays a key role in deriving meaning from raw metabolomics data. “The transfer of raw LC-MS data to actionable insights of metabolism requires confident assignments of putative compounds and identification of known metabolites,” says Amanda Souza, metabolomics program manager at Thermo Fisher. For example, LC-MS data requires preprocessing to account for several factors incurred by the technique itself, including drift in LC retention time, temporal drift, and signals from contaminants, adducts, and naturally occurring isotopes.
Further processing to make sense of the data requires advanced software, such as the Thermo Scientific™ Compound Discoverer™ software, which can process thousands of raw files for statistical analysis, data mining, visualization, biochemical pathway mapping, and interactive chemical networking. “The Compound Discoverer software provides numerous annotation tools like elemental composition prediction, database searching against the ChemSpider™ chemical structure database, MS2 spectral matching against the mzCloud™ spectral library, in silico fragmentation of proposed chemical structures, and an annotation source indicator to increased confidence of identifications,” says Souza.
Thermo Fisher Scientific recently collaborated with Timothy Garrett, associate professor of pathology at the University of Florida, who used the Orbitrap Exploris 240 MS to analyze plasma samples from people with either Laron (LS) or Guevara-Rosenbloom Syndrome (GRS), in an effort to identify biomarkers specific to each syndrome. Both syndromes are characterized by short stature, but people with LS are more likely to develop diabetes and cancer. “Employing a semi-targeted analytical approach, known metabolic differentiators were detected in addition to the discovery of previously unidentified biomarkers,” says Souza. Thermo Fisher is also collaborating with Olaris, using metabolic profiles of patients to uncover biomarkers for responders and non-responders to therapeutic drug treatment.
Better accuracy and precision
Indeed, one of the challenges in precision medicine is ascertaining how each person is different, and why some people respond to therapeutics while others don’t. “Metabolomics is well-suited to profiling the metabolic signatures of responders versus non-responders,” says Evans. “The idea is that certain people’s metabolic stress function is such that they cannot mount a strong enough immune system to actually respond to a specific drug, or to create the antibodies needed to treat the cancer.”
For example, Art Frankel’s lab at the University of South Alabama’s Mitchell Cancer Institute used Metabolon’s ultrahigh performance liquid chromatography-tandem mass spectroscopy (UPLC-MS/MS) platform for unbiased metabolomic profiling. They identified microbiota and metabolites from metastatic melanoma patients that were associated with the effectiveness of immune checkpoint inhibitor therapy. Metabolomic profiling revealed that patients who responded to the therapy had gut microbiomes enriched for specific bacterial species, and high levels of the metabolite anacardic acid. “The hope is that by identifying these metabolites and understanding the specific bacterial strain that they’re coming from, you can treat it with a probiotic, or a targeted antibiotic, to help influence that particular person’s responsiveness,” says Evans.
Evans believes that metabolomics is a powerful tool across the entire scope of precision medicine, moving from discovery into the clinic. To demonstrate that their platform has the accuracy and precision required for use in clinics, Metabolon validated their untargeted metabolomics platform by the Federal Drug Administration’s Clinical Laboratory Improvement Amendments (CLIA) certification and the College of American Pathologists (CAP) accreditation, which audit methods to certify that they meet standards for diagnostic testing of human samples. “So much of doing precision medicine is going to involve quality control and quality assurance,” says Evans. Metabolon is a member of the Metabolomics Quality Assurance & Quality Control Consortium (mQACC), a collaboration of academic, industry, and government institutions that addresses quality assurance and quality control in untargeted metabolomics. “We want everybody generating high-quality data, because we believe in the power of metabolomics,” she says.
Widespread interest
The growing interest in metabolomics is obvious to Drew Jones, director of the Metabolomics Core Resource Laboratory and assistant professor of biochemistry at NYU Langone Health, who notes that the Core receives requests from an increasing number of disciplines. The lab primarily performs analytical metabolomics and lipidomics using liquid chromatography and high-resolution mass spectrometry. They’ve programmed a custom Python-based pipeline for data analysis that includes library searches against mass spectral libraries, feature ID, and novel formula and structure determination. “We can do putative identification of metabolites based on their tandem spectra, and perform relative quantification of those metabolites across study samples,” says Jones, who expects to publish the methodology next year, and to release a platform on www.metabolomics.org. “We're going to make our whole workflow available to all academic researchers, so that they can upload their own raw data and use our analysis pipeline, sharing those results with journals and the public for interactive re-analysis.”
In addition to the Core lab, Jones’s own research lab uses metabolomics to study people’s responses to flu vaccines. “We’re discovering new metabolite biomarkers for immune responses to vaccination and influenza,” he says. His lab participates in the NIH’s Collaborative Influenza Vaccine Innovation Centers (CIVICs) project, which aims to improve the annual flu vaccine and ultimately create a universal flu vaccine.
Meanwhile, Jones is fielding more requests for collaborations in research on cancer, metabolic diseases, neurological diseases, as well as under-studied diseases. “For instance, we analyzed the metabolic profiles of parasitic worms to try to validate collaborators’ models of metabolic interplay between a parasitic worm and its symbiotic bacteria,” says Jones. The metabolic interaction itself is studied as a model system of the human microbiome in worms, and metabolomics provides an incisive tool. “Metabolomics also plays a very interesting role in the gut microbiome, because metabolism represents the interface of our bodies with the environment, and our ability to regulate our response to it.”
Continued advances in instrumentation will continue to power metabolomics in the lab and into the clinic. “Today’s instruments are incredibly robust and easy to run, with huge improvements in sensitivity and stability,” says Evans. “The hardware for metabolomics has absolutely improved over the last 15 years, and I don't see that slowing down.”