Ultra-performance liquid chromatography coupled to mass spectrometry (UPLC/MS) has been used increasingly for measuring changes of low molecular weight metabolites in biofluids/tissues in response to biological challenges such as drug toxicity and disease processes. Typically samples show high variability in concentration, and the derived metabolic profiles have a heteroscedastic noise structure characterized by increasing variance as a function of increased signal intensity. These sources of experimental and instrumental noise substantially complicate information recovery when statistical tools are used. We apply and compare several preprocessing procedures and introduce a statistical error model to account for these bioanalytical complexities. In particular, the use of total intensity, median fold change, locally weighted scatter plot smoothing, and quantile normalizations to reduce extraneous variance induced by sample dilution were compared. We demonstrate that the UPLC/MS peak intensities of urine samples should respond linearly to variable sample dilution across the intensity range. While all four studied normalization methods performed reasonably well in reducing dilution-induced variation of urine samples in the absence of biological variation, the median fold change normalization is least compromised by the biologically relevant changes in mixture components and is thus preferable. Additionally, the application of a subsequent log-based transformation was successful in stabilizing the variance with respect to peak intensity, confirming the predominant influence of multiplicative noise in peak intensities from UPLC/MS-derived metabolic profile data sets. We demonstrate that variance-stabilizing transformation and normalization are critical preprocessing steps that can benefit greatly metabolic information recovery from such data sets when widely applied chemometric methods are used.

Small molecules are central to biology, mediating critical phenomena such as metabolism, signal transduction, mating attraction, and chemical defense. The traditional categories that define small molecules, such as metabolite, secondary metabolite, pheromone, hormone, and so forth, often overlap, and a single compound can appear under more than one functional heading. Therefore, we favor a unifying term, biogenic small molecules (BSMs), to describe any small molecule from a biological source.In a similar vein, two major fields of chemical research,natural products chemistry and metabolomics, have as their goal the identification of BSMs, either as a purified active compound (natural products chemistry) or as a biomarker of a particular biological state (metabolomics). Natural products chemistry has a long tradition of sophisticated techniques that allow identification of complex BSMs, but it often fails when dealing with complex mixtures. Metabolomics thrives with mixtures and uses the power of statistical analysis to isolate the proverbial “needle from a haystack”, but it is often limited in the identification of active BSMs. We argue that the two fields of natural products chemistry and metabolomics have largely overlapping objectives: the identification of structures and functions of BSMs, which in nature almost inevitably occur as complex mixtures.Nuclear magnetic resonance (NMR) spectroscopy is a central analytical technique common to most areas of BSM research. In this Account, we highlight several different NMR approaches to mixture analysis that illustrate the commonalities between traditional natural products chemistry and metabolomics. The primary focus here is two-dimensional (2D) NMR; because of space limitations, we do not discuss several other important techniques, including hyphenated methods that combine NMR with mass spectrometry and chromatography.We first describe the simplest approach of analyzing 2D NMR spectra of unfractionated mixtures to identify BSMs that are unstable to chemical isolation. We then show how the statistical method of covariance can be used to enhance the resolution of 2D NMR spectra and facilitate the semi-automated identification of individual components in a complex mixture. Comparative studies can be used with two or more samples, such as active vs inactive, diseased vs healthy, treated vs untreated, wild type vs mutant, and so on. We present two overall approaches to comparative studies: a simple but powerful method for comparing two 2D NMR spectra and a full statistical approach using multiple samples. The major bottleneck in all of these techniques is the rapid and reliable identification of unknown BSMs; the solution will require all the traditional approaches of both natural products chemistry and metabolomics as well as improved analytical methods, databases, and statistical tools.

Background: In renal Fanconi's syndrome, dysfunction in proximal tubular cells leads to renal losses of water, electrolytes, and low-molecular-weight nutrients. For most types of isolated Fanconi's syndrome, the genetic cause and underlying defect remain unknown. Methods: We clinically and genetically characterized members of a five-generation black family with isolated autosomal dominant Fanconi's syndrome. We performed genomewide linkage analysis, gene sequencing, biochemical and cell-biologic investigations of renal proximal tubular cells, studies in knockout mice, and functional evaluations of mitochondria. Urine was studied with the use of proton nuclear magnetic resonance (1H-NMR) spectroscopy. Results: We linked the phenotype of this family's Fanconi's syndrome to a single locus on chromosome 3q27, where a heterozygous missense mutation in EHHADH segregated with the disease. The p.E3K mutation created a new mitochondrial targeting motif in the N-terminal portion of EHHADH, an enzyme that is involved in peroxisomal oxidation of fatty acids and is expressed in the proximal tubule. Immunocytofluorescence studies showed mistargeting of the mutant EHHADH to mitochondria. Studies of proximal tubular cells revealed impaired mitochondrial oxidative phosphorylation and defects in the transport of fluids and a glucose analogue across the epithelium. 1H-NMR spectroscopy showed elevated levels of mitochondrial metabolites in urine from affected family members. Ehhadh knockout mice showed no abnormalities in renal tubular cells, a finding that indicates a dominant negative nature of the mutation rather than haploinsufficiency. Conclusions: Mistargeting of peroxisomal EHHADH disrupts mitochondrial metabolism and leads to renal Fanconi's syndrome; this indicates a central role of mitochondria in proximal tubular function. The dominant negative effect of the mistargeted protein adds to the spectrum of monogenic mechanisms of Fanconi's syndrome. (Funded by the European Commission Seventh Framework Programme and others.)