The human serum metabolome

Continuing improvements in analytical technology along with an increased interest in performing comprehensive, quantitative metabolic profiling, is leading to increased interest pressures within the metabolomics community to develop centralized metabolite reference resources for certain clinically important biofluids, such as cerebrospinal fluid, urine and blood. As part of an ongoing effort to systematically characterize the human metabolome through the Human Metabolome Project, we have undertaken the task of characterizing the human serum metabolome. In doing so, we have combined targeted and non-targeted NMR, GC-MS and LC-MS methods with computer-aided literature mining to identify and quantify a comprehensive, if not absolutely complete, set of metabolites commonly detected and quantified (with today's technology) in the human serum metabolome. Our use of multiple metabolomics platforms and technologies allowed us to substantially enhance the level of metabolome coverage while critically assessing the relative strengths and weaknesses of these platforms or technologies. Tables containing the complete set of 4229 confirmed and highly probable human serum compounds, their concentrations, related literature references and links to their known disease associations are freely available at http://www.serummetabolome.ca.     

Genetic Influences on Metabolite Levels: A Comparison across Metabolomic Platforms

Metabolomic profiling is a powerful approach to characterize human metabolism and help understand common disease risk. Although multiple high-throughput technologies have been developed to assay the human metabolome, no technique is capable of capturing the entire human metabolism. Large-scale metabolomics data are being generated in multiple cohorts, but the datasets are typically profiled using different metabolomics platforms. Here, we compared analyses across two of the most frequently used metabolomic platforms, Biocrates and Metabolon, with the aim of assessing how complimentary metabolite profiles are across platforms. We profiled serum samples from 1,001 twins using both targeted (Biocrates, n = 160 metabolites) and non-targeted (Metabolon, n = 488 metabolites) mass spectrometry platforms. We compared metabolite distributions and performed genome-wide association analyses to identify shared genetic influences on metabolites across platforms. Comparison of 43 metabolites named for the same compound on both platforms indicated strong positive correlations, with few exceptions. Genome-wide association scans with high-throughput metabolic profiles were performed for each dataset and identified genetic variants at 7 loci associated with 16 unique metabolites on both platforms. The 16 metabolites showed consistent genetic associations and appear to be robustly measured across platforms. These included both metabolites named for the same compound across platforms as well as unique metabolites, of which 2 (nonanoylcarnitine (C9) [Biocrates]/Unknown metabolite X-13431 [Metabolon] and PC aa C28:1 [Biocrates]/1-stearoylglycerol [Metabolon]) are likely to represent the same or related biochemical entities. The results demonstrate the complementary nature of both platforms, and can be informative for future studies of comparative and integrative metabolomics analyses in samples profiled on different platforms.     

Determination of metabolic fluxes in a non-steady-state system

Estimation of fluxes through metabolic networks from redistribution patterns of (13)C has become a well developed technique in recent years. However, the approach is currently limited to systems at metabolic steady-state; dynamic changes in metabolic fluxes cannot be assessed. This is a major impediment to understanding the behaviour of metabolic networks, because steady-state is not always experimentally achievable and a great deal of information about the control hierarchy of the network can be derived from the analysis of flux dynamics. To address this issue, we have developed a method for estimating non-steady-state fluxes based on the mass-balance of mass isotopomers. This approach allows multiple mass-balance equations to be written for the change in labelling of a given metabolite pool and thereby permits over-determination of fluxes. We demonstrate how linear regression methods can be used to estimate non-steady-state fluxes from these mass balance equations. The approach can be used to calculate fluxes from both mass isotopomer and positional isotopomer labelling information and thus has general applicability to data generated from common spectrometry- or NMR-based analytical platforms. The approach is applied to a GC-MS time-series dataset of (13)C-labelling of metabolites in a heterotrophic Arabidopsis cell suspension culture. Threonine biosynthesis is used to demonstrate that non-steady-state fluxes can be successfully estimated from such data while organic acid metabolism is used to highlight some common issues that can complicate flux estimation. These include multiple pools of the same metabolite that label at different rates and carbon skeleton rearrangements.     

Quantitative analysis of metabolic fluxes in Escherichia coli, using two-dimensional NMR spectroscopy and complete isotopomer models.

Complete isotopomer models that simulate distribution of label in 13C tracer experiments are applied to the quantification of metabolic fluxes in the primary carbon metabolism of E. coli under aerobic and anaerobic conditions. The concept of isotopomer mapping matrices (IMMs) is used to simplify the formulation of isotopomer mass balances by expressing all isotopomer mass balances of a metabolite pool in a single matrix equation. A numerically stable method to calculate the steady-state isotopomer distribution in metabolic networks in introduced. Net values of intracellular fluxes and the degree of reversibility of enzymatic steps are estimated by minimization of the deviations between experimental and simulated measurements. The metabolic model applied includes the Embden-Meyerhof-Parnas and the pentose phosphate pathway, the tricarboxylic acid cycle, anaplerotic reaction sequences and pathways involved in amino acid synthesis. The study clearly demonstrates the value of complete isotopomer models for maximizing the information obtainable from 13C tracer experiments. The approach applied here offers a completely general and comprehensive analysis of carbon tracer experiments where any set of experimental data on the labeling state and extracellular fluxes can be used for the quantification of metabolic fluxes in complex metabolic networks.     

Modeling isotopomer distributions in biochemical networks using isotopomer mapping matrices

Within the last decades NMR spectroscopy has undergone tremendous development and has become a powerful analytical tool for the investigation of intracellular flux distributions in biochemical networks using (13)C-labeled substrates. Not only are the experiments much easier to conduct than experiments employing radioactive tracer elements, but NMR spectroscopy also provides additional information on the labeling pattern of the metabolites. Whereas the maximum amount of information obtainable with (14)C-labeled substrates is the fractional enrichment in the individual carbon atom positions, NMR spectroscopy can also provide information on the degree of labeling at neighboring carbon atom positions by analyzing multiplet patterns in NMR spectra or using 2-dimensional NMR spectra. It is possible to quantify the mole fractions of molecules that show a specific labeling pattern, i.e., information of the isotopomer distribution in metabolite pools can be obtained. The isotopomer distribution is the maximum amount of information that in theory can be obtained from (13)C-tracer studies. The wealth of information contained in NMR spectra frequently leads to overdetermined algebraic systems. Consequently, fluxes must be estimated by nonlinear least squares analysis, in which experimental labeling data is compared with simulated steady state isotopomer distributions. Hence, mathematical models are required to compute the steady state isotopomer distribution as a function of a given set of steady state fluxes. Because 2(n) possible labeling patterns exist in a molecule of n carbon atoms, and each pattern corresponds to a separate state in the isotopomer model, these models are inherently complex. Model complexity, so far, has restricted usage of isotopomer information to relatively small metabolic networks. A general methodology for the formulation of isotopomer models is described. The model complexity of isotopomer models is reduced to that of classical metabolic models by expressing the 2(n) isotopomer mass balances of a metabolite pool in a single matrix equation. Using this approach an isotopomer model has been implemented that describes label distribution in primary carbon metabolism, i.e., in a metabolic network including the Embden-Meyerhof-Parnas and pentose phosphate pathway, the tricarboxylic acid cycle, and selected anaplerotic reaction sequences. The model calculates the steady state label distribution in all metabolite pools as a function of the steady state fluxes and is applied to demonstrate the effect of selected anaplerotic fluxes on the labeling pattern of the pathway intermediates. (c) 1997 John Wiley & Sons, Inc. Biotechnol Bioeng 55:831-840, 1997.     

Numerical bias estimation for mass spectrometric mass isotopomer analysis

Mass spectrometric (MS) isotopomer analysis has become a standard tool for investigating biological systems using stable isotopes. In particular, metabolic flux analysis uses mass isotopomers of metabolic products typically formed from ¹³C-labeled substrates to quantitate intracellular pathway fluxes. In the current work, we describe a model-driven method of numerical bias estimation regarding MS isotopomer analysis. Correct bias estimation is crucial for measuring statistical qualities of measurements and obtaining reliable fluxes. The model we developed for bias estimation corrects a priori unknown systematic errors unique for each individual mass isotopomer peak. For validation, we carried out both computational simulations and experimental measurements. From stochastic simulations, it was observed that carbon mass isotopomer distributions and measurement noise can be determined much more precisely only if signals are corrected for possible systematic errors. By removing the estimated background signals, the residuals resulting from experimental measurement and model expectation became consistent with normality, experimental variability was reduced, and data consistency was improved. The method is useful for obtaining systematic error-free data from ¹³C tracer experiments and can also be extended to other stable isotopes. As a result, the reliability of metabolic fluxes that are typically computed from mass isotopomer measurements is increased.     

Microbial metabolomics: toward a platform with full metabolome coverage

Achieving metabolome data with satisfactory coverage is a formidable challenge in metabolomics because metabolites are a chemically highly diverse group of compounds. Here we present a strategy for the development of an advanced analytical platform that allows the comprehensive analysis of microbial metabolomes. Our approach started with in silico metabolome information from three microorganisms-Escherichia coli, Bacillus subtilis, and Saccharomyces cerevisiae-and resulted in a list of 905 different metabolites. Subsequently, these metabolites were classified based on their physicochemical properties, followed by the development of complementary gas chromatography-mass spectrometry and liquid chromatography-mass spectrometry methods, each of which analyzes different metabolite classes. This metabolomics platform, consisting of six different analytical methods, was applied for the analysis of the metabolites for which commercial standards could be purchased (399 compounds). Of these 399 metabolites, 380 could be analyzed with the platform. To demonstrate the potential of this metabolomics platform, we report on its application to the analysis of the metabolome composition of mid-logarithmic E. coli cells grown on a mineral salts medium using glucose as the carbon source. Of the 431 peaks detected, 235 (=176 unique metabolites) could be identified. These include 61 metabolites that were not previously identified or annotated in existing E. coli databases.     

Metabolomics – an overview. From basic principles to potential biomarkers (part 2)

The metabolome, which represents the complete set of molecules (metabolites) of a biological sample (cell, tissue, organ, organism), is the final downstream product of the metabolic cell process that involves the genome and exogenous sources. The metabolome is characterized by a large number of small molecules with a huge diversity of chemical structures and abundances. Exploring the metabolome requires complementary analytical platforms to reach its extensive coverage. The metabolome is continually evolving, reflecting the continuous flux of metabolic and signaling pathways. Metabolomic research aims to study the biochemical processes by detecting and quantifying metabolites to obtain a metabolic picture able to give a functional readout of the physiological state. Recent advances in mass spectrometry (one of the mostly used technologies for metabolomics studies) have given the opportunity to determine the spatial distribution of metabolites in tissues. In a two-part article, we describe the usual metabolomics technologies, workflows and strategies leading to the implementation of new clinical biomarkers. In this second part, we first develop the steps of a metabolomic analysis from sample collection to biomarker validation. Then with two examples, autism spectrum disorders and Alzheimer's disease, we illustrate the contributions of metabolomics to clinical practice. Finally, we discuss the complementarity of in vivo (positron emission tomography) and in vitro (metabolomics) molecular explorations for biomarker research.     

A novel stable isotope labelling assisted workflow for improved untargeted LC–HRMS based metabolomics research

Many untargeted LC–ESI–HRMS based metabolomics studies are still hampered by the large proportion of non-biological sample derived signals included in the generated raw data. Here, a novel, powerful stable isotope labelling (SIL)-based metabolomics workflow is presented, which facilitates global metabolome extraction, improved metabolite annotation and metabolome wide internal standardisation (IS). The general concept is exemplified with two different cultivation variants, (1) co-cultivation of the plant pathogenic fungi Fusarium graminearum on non-labelled and highly ¹³C enriched culture medium and (2) experimental cultivation under native conditions and use of globally U-¹³C labelled biological reference samples as exemplified with maize and wheat. Subsequent to LC–HRMS analysis of mixtures of labelled and non-labelled samples, two-dimensional data filtering of SIL specific isotopic patterns is performed to better extract truly biological derived signals together with the corresponding number of carbon atoms of each metabolite ion. Finally, feature pairs are convoluted to feature groups each representing a single metabolite. Moreover, the correction of unequal matrix effects in different sample types and the improvement of relative metabolite quantification with metabolome wide IS are demonstrated for the F. graminearum experiment. Data processing employing the presented workflow revealed about 300 SIL derived feature pairs corresponding to 87–135 metabolites in F. graminearum samples and around 800 feature pairs corresponding to roughly 350 metabolites in wheat samples. SIL assisted IS, by the use of globally U-¹³C labelled biological samples, reduced the median CV value from 7.1 to 3.6 % for technical replicates and from 15.1 to 10.8 % for biological replicates in the respective F. graminearum samples.     

Metabolomic profiling and genomic analysis of wheat aneuploid lines to identify genes controlling biochemical pathways in mature grain

Metabolomics is becoming an increasingly important tool in plant genomics to decipher the function of genes controlling biochemical pathways responsible for trait variation. Although theoretical models can integrate genes and metabolites for trait variation, biological networks require validation using appropriate experimental genetic systems. In this study, we applied an untargeted metabolite analysis to mature grain of wheat homoeologous group 3 ditelosomic lines, selected compounds that showed significant variation between wheat lines Chinese Spring and at least one ditelosomic line, tracked the genes encoding enzymes of their biochemical pathway using the wheat genome survey sequence and determined the genetic components underlying metabolite variation. A total of 412 analytes were resolved in the wheat grain metabolome, and principal component analysis indicated significant differences in metabolite profiles between Chinese Spring and each ditelosomic lines. The grain metabolome identified 55 compounds positively matched against a mass spectral library where the majority showed significant differences between Chinese Spring and at least one ditelosomic line. Trehalose and branched‐chain amino acids were selected for detailed investigation, and it was expected that if genes encoding enzymes directly related to their biochemical pathways were located on homoeologous group 3 chromosomes, then corresponding ditelosomic lines would have a significant reduction in metabolites compared with Chinese Spring. Although a proportion showed a reduction, some lines showed significant increases in metabolites, indicating that genes directly and indirectly involved in biosynthetic pathways likely regulate the metabolome. Therefore, this study demonstrated that wheat aneuploid lines are suitable experimental genetic system to validate metabolomics–genomics networks.     

The Bovine Ruminal Fluid Metabolome

The rumen is a unique organ that serves as the primary site for microbial fermentation of ingested plant material for domestic livestock such as cattle, sheep and goats. The chemical composition of ruminal fluid is thought to closely reflect the healthy/unhealthy interaction between rumen microflora and diet. Just as diet and feed quality is important for livestock production, rumen health is also critical to the growth and production of high quality milk and meat. Therefore a detailed understanding of the chemical composition of ruminal fluid and the influence of diet on its composition could help improve the efficiency and effectiveness of farming and veterinary practices. Consequently we have undertaken an effort to comprehensively characterize the bovine ruminal fluid metabolome. In doing so, we combined NMR spectroscopy, inductively coupled plasma mass-spectroscopy (ICP-MS), gas chromatography-mass spectrometry (GC-MS), direct flow injection (DFI) mass spectrometry and lipidomics with computer-aided literature mining to identify and quantify essentially all of the metabolites in bovine ruminal fluid that can be routinely detected (with today’s technology). The use of multiple metabolomics platforms and technologies allowed us to substantially enhance the level of metabolome coverage while critically assessing the relative strengths and weaknesses of these techniques. Tables containing the set of 246 ruminal fluid metabolites or metabolite species, their concentrations, related literature reference and links to their known diet associations for the bovine rumen metabolome are freely available at http://www.rumendb.ca.     

Strategy for the fine characterization of glycosyltransferase specificity using isotopomer assembly

Glycosylation, which represents the most complex posttranslational modification (PTM) event during protein maturation, has a vital role in biological processes. Glycan biosynthesis is orchestrated by numerous glycosyltransferases, each displaying different selectivities for multiple reaction sites. The precise specificities of these enzymes have been difficult to study because of the lack of available substrates of defined structure and problems associated with the analyses. Moreover, the analysis of glycans is extremely difficult owing to the structural complexity of the glycan chain. Here we describe a new strategy for the fine characterization of enzyme specificity using substrate isotopomer assemblies. Because isotopomer assemblies contain a sugar residue that is position-specifically labeled with a stable isotope, we can use tandem mass spectrometry (MS/MS) to assign the structure of positional isomers generated by glycosylation. We demonstrated the analysis of substrate specificities of five beta4-galactosyltransferases (beta4GalT-I, -II, -III, -IV and -V) using our strategy.     

Application of metabolome data in functional genomics: a conceptual strategy

A gene with yet unknown physiological function can be studied by changing its expression level followed by analysis of the resulting phenotype. This type of functional genomics study can be complicated by the occurrence of 'silent mutations', the phenotypes of which are not easily observable in terms of metabolic fluxes (e.g., the growth rate). Nevertheless, genetic alteration may give rise to significant yet complicated changes in the metabolome. We propose here a conceptual functional genomics strategy based on microbial metabolome data, which identifies changes in in vivo enzyme activities in the mutants. These predicted changes are used to formulate hypotheses to infer unknown gene functions. The required metabolome data can be obtained solely from high-throughput mass spectrometry analysis, which provides the following in vivo information: (1) the metabolite concentrations in the reference and the mutant strain; (2) the metabolic fluxes in both strains and (3) the enzyme kinetic parameters of the reference strain. We demonstrate in silico that changes in enzyme activities can be accurately predicted by this approach, even in 'silent mutants'.     

Measurement of intracellular metabolites of primary metabolism and adenine nucleotides in chemostat cultivated Penicillium chrysogenum

An experimental platform has been developed for rapid sampling and quenching of chemostat cultivated Penicillium chrysogenum broth for metabolome analysis in highly dynamic experiments, aimed at the elucidation of the in vivo kinetic properties of metabolism. The sampling and quenching protocol available from Saccharomyces cerevisiae had to be modified for Penicillium chrysogenum mainly because of its filamentous character. Intracellular metabolites of glycolysis, TCA cycle, and adenine nucleotides were measured with isotope dilution mass spectrometry (IDMS) using a U-(13)C-labeled metabolite mix produced from yeast cells as internal standard. By addition of the U-(13)C internal standard mix prior to the metabolite extraction procedure, partial degradation of metabolites as well as non-linearity and drift of the LC-MS/MS could be successfully compensated for. It was found that there is a serious matrix effect on metabolite extraction between different organisms, which is however completely corrected for by the IDMS approach. Intracellular metabolites could be analyzed with standard deviations of around 5%. A comparison of the metabolite levels between Saccharomyces cerevisiae and Penicillium chrysogenum showed both significant similarities and large differences, which seem to be related to the presence of the penicillin pathway.     

Cellular metabolomics of Escherchia coli

Escherichia coli is among the simplest and best-understood free-living organisms. It has served as a valuable model for numerous biological processes, including cellular metabolism. Just as E. coli stood at the front of the genomic revolution, it is playing a leading role in the development of cellular metabolomics: the study of the complete metabolic contents of cells, including their dynamic concentration changes and fluxes. This review briefly describes the essentials of cellular metabolomics and its fundamental differentiation from biomarker metabolomics and lipidomics. Key technologies for metabolite quantitation from E. coli are described, with a focus on those involving mass spectrometry. In particular emphasis is given to the cell handling and sample preparation steps required for collecting data of high biological reliability, such as fast metabolome quenching. Future challenges, both in terms of data collection and application of the data to obtain a comprehensive understanding of metabolic dynamics, are discussed.     

Bidirectional reaction steps in metabolic networks: III. Explicit solution and analysis of isotopomer labeling systems

The last few years have brought tremendous progress in experimental methods for metabolic flux determination by carbon-labeling experiments. A significant enlargement of the available measurement data set has been achieved, especially when isotopomer fractions within intracellular metabolite pools are quantitated. This information can be used to improve the statistical quality of flux estimates. Furthermore, several assumptions on bidirectional intracellular reaction steps that were hitherto indispensable may now become obsolete. To make full use of the complete measurement information a general mathematical model for isotopomer systems is established in this contribution. Then, by introducing the important new concept of cumomers and cumomer fractions, it is shown that the arising nonlinear isotopomer balance equations can be solved analytically in all cases. In particular, the solution of the metabolite flux balances and the positional carbon-labeling balances presented in part I of this series turn out to be just the first two steps of the general solution procedure for isotopomer balances. A detailed analysis of the isotopomer network structure then opens up new insights into the intrinsic structure of isotopomer systems. In particular, it turns out that isotopomer systems are not as complex as they appear at first glance. This enables some far-reaching conclusions to be drawn on the information potential of isotopomer experiments with respect to flux identification. Finally, some illustrative examples are examined to show that an information increase is not guaranteed when isotopomer measurements are used in addition to positional enrichment data.