Recent improvements in technology are enabling the simultaneous detection and quantitation of large numbers of metabolites from biological samples. Here, we examine the challenges and innovations in metabolomics technologies and find out how these tools are impacting on clinical biomarker discovery.
Liquid chromatography-mass spectrometry (LC-MS), which exploits the unique chemical properties of each molecule to separate them prior to analysis on a mass spectrometer, is emerging as the leading technique for untargeted metabolomics studies.
“Probably around three-quarters of analyses are now done using LC-MS, mainly due to its ability to measure hundreds of molecules in a single sample,” says Professor Daniel Raftery of the Mitochondria and Metabolism Center at the University of Washington.
The technique uses chromatography to separate the molecules within a complex mixture, such as a blood sample. For metabolomics studies, this is usually achieved in one of two ways – using either a hydrophobic or hydrophilic column.
“The molecules are separated and come out in order of their lack of preference for the surface,” explains Raftery.
Once leaving the column, the molecules are then detected in a mass spectrometer.
“The combination of the molecule’s retention time in the column and its mass are the two linked parameters that we use to characterize it,” says Dr Aalim Weljie, Research Assistant Professor of Pharmacology at the University of Pennsylvania.
A big advantage of LC-MS is its sensitivity - it can detect metabolites that are very low in concentration, providing a good coverage of the metabolome. But the downside is that the majority of these are unknown molecules – described by some researchers as the ‘dark metabolome’.
“If you’re doing an untargeted analysis, approximately anywhere between 10 and 20 per cent of what comes out is identifiable through databases and other searches – but that leaves up to 90 per cent that we have no idea what they are,” explains Weljie.
Some high-resolution instruments are now helping researchers to determine the chemical formula of a molecule, but its exact structure can still remain elusive. But newer approaches, such as tandem mass spectrometry, can help provide clues.
“We pick the parent molecule and break it into smaller fragments and look at the patterns,” says Weljie.
Help is also offered through recent innovation in cheminformatics, which uses computational approaches to couple structural information with experimental data.
Features or compounds? A data reduction strategy for untargeted metabolomics to generate meaningful data [Application Note]
HRAM LC-MS is a powerful tool for untargeted metabolomics analysis and detects all ionic species present in analyzed samples generating an exhaustive results list of molecular features. Taking all of these experimental artifacts into consideration during data processing enables a proactive approach to eliminating redundancies and unrelated ions.
Download this application note to learn how to differentiate a compound from a feature to reduce redundancies and accelerate data analysis using Thermo Scientific™ Compound Discoverer™ software.More Information
In the past, stability has also been a big challenge for LC-MS, limiting how many samples could be analyzed before the instrument changed its signal intensity.
“After you’d run 50 samples and if you then re-ran the first sample again, it would look different to how it did the first time,” explains Raftery.
This has meant a lack of comparability between data generated by different machines or laboratories. As Raftery describes: “In the old days, each mass spectrometer was its own creature!”
But things are now improving, due to a combination of more stable instruments and the development of better quality control methods.
“People are now talking about running samples and looking at presumably the same molecules and reporting the same results - and that’s started in the last several years,” says Raftery.
Advanced separation technologies
Ultraperformance liquid chromatography (UPLC), which allows the instrumentation to withstand much higher pressures, can improve speed, resolution and sensitivity. Faster run times are particularly advantageous for industrial laboratories, but many academic laboratories are also reaping the benefits.
“You can run larger studies and the mass spectrometer will be more stable because you’re just running faster,” explains Raftery.
More recently, micro- and nanoflow liquid chromatography are also helping to further improve sensitivity.
“UPLC and traditional chromatography both work in milliliters per minute flow rates, but these newer techniques work in micro- or nanoliters,” explains Weljie.
Nanoflow is extremely sensitive and so can be used to analyze molecules that have very low concentrations, but it can be challenging to perform due to difficulties with its connection to the mass spectrometry instrumentation.
“Microflow sits in between the two – so, you’ve got some of the advantages of the slower flow rate without exposing yourself to plumbing issues. It’s a major advance but it’s still not widespread yet,” says Weljie.
Exploring the secrets of blood lipids
“Lipid molecules are increasingly important and they seem to be quite sensitive to a number of disease states, especially neurological diseases and some cancers - for example, we did a study identifying blood lipid changes associated with ovarian cancer,” says Raftery.
Others are looking at the potential of using lipids as prognostic biomarkers.
“You place your sample in the NMR, which is a high field magnet that aligns the nuclear spins – for metabolomics, usually of the hydrogen atoms – and then uses radiofrequency pulses to detect them,” explains Raftery.
NMR is extremely quantitative, but it lacks sensitivity.
“You can recover information about the concentration of metabolites in a complex sample like blood,” explains Weljie. “But the number of metabolites that you can see per sample is relatively limited.”
But new technologies are finding ways to boost the signal. For example, many laboratories are starting to use cryogenically cooled NMR probes.
“Depending on the sample, you can get an increase of between three and five times in sensitivity,” says Weljie.
Into the clinic
However, the big hope for the future is around creating panels of biomarkers that can help spot metabolic changes within an individual – for example, to help to flag the onset of a disease far ahead of any symptoms or to select the right treatment. But despite a recent explosion in investigating potential biomarkers that are needed to move us towards precision medicine, more research will be needed to translate these into robust clinical diagnostic tools.
“That’s not because the technologies can’t do it, it’s just that the validation is challenging for biomarkers,” explains Raftery.
Large-scale studies are key for validation, with scientists following people taking part in population studies over many years. NMR can offer an advantage for analyzing these big sample sets because of its stability.
“The mass spectrometry it’s just not up to the level yet of running tens of thousands of samples, which you can do with NMR,” says Raftery.
Researchers are also trying to combine large datasets across all ‘omics information, which could provide far greater power and confidence behind their results.
“If you’re seeing a change in the metabolite and a change in the enzyme, then there’s probably a biologically verifiable change,” says Raftery.
Making medicine more precise
“There are lots of exciting things happening in metabolism, but there’s also still a lot of work to do to understand the changes in relation to different biological stresses,” says Raftery.
But recent advancements in the technologies available for exploring the metabolome are generating renewed enthusiasm for the field among researchers.
“The tools are just starting to mature so now we have to get into that space and ask those questions in a real way and I think that’s what’s exciting about this moment in time,” says Weljie.