Finding how a trace molecule's concentration varies when it's mixed in a jumble of others is not straightforward but this is just what a metabolomics scientist does every day. Now an international team is combining several analytical techniques to make the process easier and applying its method to study diabetes.

Metabolomics is the study of metabolites - the small molecules formed by specific processes inside a single cell. As variations in levels of some of these may be linked to metabolic disorders such as diabetes, identifying these so-called metabolic markers has the potential to provide diagnostic information and to lead to better therapies. 

With this in mind, Geoffrey Gipson at Drexel University, Philadelphia, US, and colleagues in the US and UK, combined spectroscopic methods such as nuclear magnetic resonance (NMR) and liquid chromatography-mass spectrometry (LC-MS) to analyse the metabolites in urine and tissue samples from mice deficient in the hormone protein leptin. Leptin has a key role in regulating metabolism and the leptin-deficient mice display several genetic traits associated with type 2 diabetes. By comparing the metabolite levels with those in control mice, the researchers confirmed that many pathways, including fatty acid metabolism, are altered in 'diabetic' mice. Gipson says that these pathways could potentially be targeted for diabetes treatments.

Leptin-deficient mice (left) show genetic traits associated with diabetes and obesity © ONRL

While Gipson admits that there are limitations in using animals to model human conditions, he explains that metabolism is a fundamental biological process which has been highly conserved over evolutionary time. 'As such, there are many cross-species molecular similarities which allow researchers to gain critical insights into human disease,' he adds. Given this, the biggest challenge, says Gipson, is in applying the new knowledge gained through metabolic profiling research into therapies for patients living with diabetes.

Jules Griffin, an expert in metabolomics at the University of Cambridge, UK, commends the team's integrated approach. 'While NMR has been used widely in mammalian disease models,' says Griffin, 'the use of LC-MS based approaches has lagged behind, in part because of the greater challenges in technical reproducibility. By picking a well-characterised model of type 2 diabetes the team has been able to validate its approach and then extend its analysis to new metabolite changes not previously described.'

Kathleen Too