Metabolomics is a relatively new area of systems biology based on detecting and quantifying metabolites and the activity of metabolic pathways. This is a daunting enterprise, because mammalian cells contain thousands of individual enzymatic reactions and metabolites, many of which are highly responsive to environmental cues and cell-intrinsic processes. Furthermore, hundreds of Mendelian genetic disorders are associated with profound defects in single enzymes, compounding metabolic complexity across populations. A highly simplified view of core metabolism that emphasizes essential metabolic functions is shown in Fig. 1.
Metabolism can be analyzed in cells, model organisms or human subjects. Any of a number of analytical techniques can be applied, depending on the complexity of the question. A main distinction is between “metabolomics” and “metabolic flux analysis” (Figure 2). The goal of metabolomics is to provide a detailed and quantitative catalog of metabolites present within a system at a given time. This is similar in principle to counting the cars in a parking lot at two different times of day and determining the differences. Although this approach produces a tremendous amount of detail, it provides only a static view of metabolite abundance with no information about the rates of metabolite turnover and flux through the pathways that produce and consume various molecules. Metabolic flux analysis, by contrast, while lacking some of the detail of metabolomics, adds the crucial component of measuring the activity of metabolic pathways, either in absolute terms or in relation to one another. These measurements are made by introducing a nutrient labeled with a stable isotope (e.g. 15N, 13C) into a biological system, and
measuring the distribution of the isotope into various metabolites using mass spectrometry and/or nuclear magnetic resonance (NMR). Mathematical models can then be used to determine quantitative flux rates throughout the network.
The choice of analytical approach depends in part on the hypothesis to be tested. Options include:
- Target metabolite detection. These simple approaches aim to quantify a small handful (sometimes one) metabolite with a high degree of accuracy. An example is the measurement of lactate in tissue culture medium from glycolytic cancer cells.
- Targeted Metabolomics. In targeted metabolomics, a set of metabolites (ranging from a small handful to several hundred) are analyzed simultaneously from a biological sample. For example, one could choose a set of one hundred metabolites from informative positions within a metabolic network, and then develop a mass spectrometry-based method to analyze this set in a highly quantitative and sensitive fashion. Targeted metabolomics can be used to determine the effects of a perturbation (e.g. activation of an oncogene) on the steady-state abundances of metabolites from known pathways.
- Unbiased metabolomics. In unbiased metabolomics, the goal is to detect all metabolites, known and unknown, within a biological sample. This approach is less constrained by a priori knowledge than targeted metabolomics, because it does not require the user to pre-select which metabolites to monitor.
- Metabolic flux analysis. As illustrated in Fig. 2, metabolic flux analysis seeks to analyze the activity of metabolic pathways rather than the abundance of metabolites per se. Two types of analysis related to metabolic flux include:
- Mass isotopomer distribution: This approach measures the population of isotope-enriched molecules within a population. For example, one might label cells with 13C-glucose and ask what fraction of a fatty acid contains 0, 1, 2, 3, 4, etc. 13C atoms after culture of a defined duration. 13C enrichment in the fatty acid is related to the overall rate of fatty acid turnover, the specific pathways used to produce fatty acid from precursors, and competing sources of lipogenic carbon.
- Quantitative flux analysis: These approaches build complex representations of metabolism based on quantitative inputs from growth rates, metabolite abundance, isotopic enrichment, and other parameters. The approach demands extensive mathematical modeling based on reconstructions of metabolic networks derived from empirically-validated measurements and values extrapolated from other sources. Data generated from wet experiments are fit to metabolic flux models to produce a much more comprehensive picture of metabolism than what can be provided from mass isotopomer distribution alone.