J. L. Griffin and M. E. Bollard Pages 389 - 398 ( 10 )
The functional genomic techniques of transcriptomics and proteomics promise unparalleled global information during the drug development process. However, if these technologies are used in isolation the large multivariate data sets produced are often difficult to interpret, and have the potential of missing key metabolic events (e.g. as a result of experimental noise in the system). To better understand the significance of these megavariate data the temporal changes in phenotype must be described. High resolution 1H NMR spectroscopy used in conjunction with pattern recognition provides one such tool for defining the dynamic phenotype of a cell, organ or organism in terms of a metabolic phenotype. In this review the benefits of this metabonomics / nmetabolomics approach to problems in toxicology will be discussed. One of the major benefits of this approach is its high throughput nature and cost effectiveness on a per sample basis. Using such a method the consortium for metabonomic toxicology (COMET) are currently investigating ∼150 model liver and kidney toxins. This investigation will allow the generation of expert systems where liver and kidney toxicity can be predicted for model drug compounds, providing a new research tool in the field of drug metabolism. The review will also include how metabonomics may be used to investigate co-responses with transcripts and proteins involved in metabolism and stress responses, such as during drug induced fatty liver disease. By using data integration to combine metabolite analysis and gene expression profiling key perturbed metabolic pathways can be identified and used as a tool to investigate drug function.
metabolomics, multivariate pattern recognition, nmr spectroscopy, mass spectrometry, metabolic profiling, functional genomics
Department of Biochemistry, University of Cambridge, Tennis Court Road, Cambridge, UK.