BioWorld International

NORWICH, UK - Recent efforts to optimize metabolomics technology are beginning to bear fruit in the discovery of novel biomarkers for disease. In general, the concentrations of metabolites are amplified relative to the proteins that prompted their formation, making them potentially more sensitive, both in diagnostics and as drug targets, than protein biomarkers.

"Metabolites are downstream from the protein, so you must have an amplification in concentration," Douglas Kell, director of the Manchester Centre for Integrative Systems Biology, said at the BA Science Festival last week. "Furthermore, you don't need to know the gene sequence, there are fewer metabolites than proteins, and they are generic, so once you can measure a metabolite, it is the same in every organism - which is not the case for proteins."

It is estimated that there are about 3,000 metabolites in the human metabolome. The issue is that there is a wide concentration range, and it is very difficult to detect metabolites that occur at low concentrations.

Kell's group has developed an automated "robot scientist" that couples gas chromatography separation to mass spectrometry detection, and has optimized and refined the equipment to the point where it can discriminate 1,800 true metabolites.

Using the robot, blood samples from healthy and affected individuals have been analyzed to systematically uncover subsets of metabolites that are markers of particular diseases. For example, examining samples from women affected by pre-eclampsia in pregnancy and those who were not led to the discovery of three metabolites that distinguish sufferers.

"None of the [three metabolites] was visible [with our equipment] when we started; it is only as we improved the process that they were detected," Kell said.

Currently, pregnant women are sifted out as being at risk of pre-eclampsia by measuring blood pressure. "But this is more than a surrogate for blood pressure: You can look in the affected cohort only and follow the development of the disease," Kell said. More recently his group has found further metabolites linked to pre-eclampsia, making the biomarkers potentially even more sensitive.

Kell has discovered metabolites for Huntington's disease and for a number of cardiovascular diseases, also. "Many of these offer the possibility of novel interventions and of prognostic detection of diseases in their earliest stages, before they become life threatening," Kell said.

Last week it was announced that the robot scientist will be used in the first systematic attempt to find metabolites in Alzheimer's disease. The three-year study aims to find metabolites for diagnosis and for predicting and measuring the effectiveness of drugs designed to slow the advance of the disease. The study will compare 1,000 normal samples with 1,000 from Alzheimer's disease patients. The results will be made available in a public database.