In any organism, the complexity of its biochemical cascade is dynamically influenced by environmental parameters, genetic disposition, etiologies, and/or drug administration, to name but a few factors. The ability to monitor the multi-parametric changes that an organism will experience is of great diagnostic and prognostic value to better understand the metabolic status of the organism and its relationship with induced stimuli [1]. In recent years, the field of metabonomics has evolved to effectively derive information from a combination of data-rich analytical techniques (NMR and more recently, mass spectrometry) and statistical multivariate analysis [2] for studying in vivo metabolic profiles. Metabonomics as a profiling technology was pioneered by Nicholson and colleagues. It can be defined as an approach to investigate complex metabolic consequences of patho-physiological or genetic modification in a multivariate space [3]. Metabonomics offers the advantage of being applicable to samples collected in non-invasive (urine), or minimally invasive ways (serum, tissues). The data provides qualitative and/or quantitative assessments of small endogenous molecules (∼50 to 1000 amu) and follows qualitative changes of unique biological macromolecules, i.e. lipoproteins [4,5]. The development of these information-rich techniques has prompted the scientific community to evolve from the original concept of a biomarker representing a single molecule, to a complex biomarker represented by a panel of molecules emanating from multi-parametric analysis. It has been demonstrated that the accuracy and predictability of a panel of biomarkers is of greater value than that of a single entity [6]. The holistic approach that systems biology (i.e. genomics, transcriptomics, proteomics, and metabonomics) brings to the study of biology assumes that each tier of the system depends on the other, and alterations in one tier may affect another. Compared to the other "omics", metabonomics focuses on the assessment of small endogenous metabolites. Since these cellular components of the metabolome represent the end products of gene expression and define the phenotype of a cell, tissue, or organism, metabonomics is well positioned to provide the most functional information amongst the "omics" technologies [7]. Biomarkers that change either in pattern or concentration, can relate to both site and mechanism of toxicity [8]. However, because of technical challenges to measure all metabolites present in biomatrices, metabonomics as a metabolic profiling tool has been delayed as a developing technology when compared to genomics, or proteomics. Rapid growth in the use of metabonomics is underway, and positive impacts to the study of biological systems are now being demonstrated [7]. A strategy aggressively pursued by the pharmaceutical industry is the discovery of specific, selective and robust biomarkers, as illustrated by the amount of work done with "omics" technologies. These efforts, driven by drug discovery needs, are intended to enhance the quality of lead prioritization decisions, decrease attrition by yielding better candidate selection, provide toxicity screening for better selection of backups, and provide a tool for continuous safety assessments that can be translated from early development to the clinical arena [9]. In cases where a particular toxicity has been observed in pre-clinical studies without a clear understanding of the relevance of these findings to humans, it is critical to possess the analytical tools that can be equally applied to pre-clinical and clinical situations. As our knowledge of data processing expands, allowing pattern recognition to evolve towards mathematical modeling, predictive assessments and metabolite identification/quantitation, scientists are starting to seek biomarkers that not only enable the elucidation of complex disease pathways, but also provide predictive safety assessments. This critical bridge is often the key to successful risk management strategies for the continuation of compounds in development. Metabonomics, although not a panacea for safety assessment or biomarker discovery, has been shown to provide an unbiased ability to differentiate genotypes based on metabolite levels that may, or may not produce visible phenotypes [10,11]. Numerous examples show the application of metabonomics to finding biomarkers of disease and efficacy, however these are out-of-scope for this chapter. It is our intent to focus on the latest applications of metabonomics to toxicity and safety assessments, while taking a cursory look at the analytical instrumentation and chemometric methods that are needed for producing and evaluating the vast amounts of generated data. © 2005 Elsevier Inc. All rights reserved.
