Our understanding of biology has been greatly improved through recent developments in mass spectrometry, which is providing detailed information on protein and metabolite composition as well as protein-metabolite interactions. The high sensitivity and resolution of mass spectrometry achieved with liquid or gas chromatography allows for detection and quantification of hundreds to thousands of molecules in a single measurement. Where homogenization-based sample preparation and extraction methods result in a loss of spatial information, mass spectrometry imaging technologies provide the in situ distribution profiles of metabolites and proteins within tissues. Mass spectrometry–based analysis of metabolite abundance, protein-metabolite interactions, and spatial distribution of compounds facilitates the high-throughput screening of biochemical reactions, the reconstruction of metabolic networks, biomarker discovery, determination of tissue compositions, and functional annotation of both proteins and metabolites.

Mass spectrometry (MS) is a rapidly growing technology for the comprehensive profiling of small molecules (metabolites) and proteins (1,2). Information obtained from the metabolome is particularly useful given that small molecules represent the downstream outcome of cellular machinery (i.e., enzymes) and can provide a metabolic phenotype of a biological system. Another advantage of studying the metabolome (versus the proteome or genome) is that the analysis can be performed independently of a genome sequence or large expressed sequence tag databases and therefore can be applied to virtually any biological system.

In most cases, the first step in a metabolite profiling experiment is to extract metabolites from the biological matrix (Figure 1). Extracted metabolites are often separated using gas chromatography/MS (GC/MS) (3), liquid chromatography/MS (LC/MS) (4), or capillary electrophoresis/MS (CE/MS) (5). Critical to all MS-based approaches is the efficient desorption and ionization of metabolites, where the resulting gas phase ions can be separated by mass analyzers such as quadrupole, time-of-flight, and ion trap. Ions are typically detected using a micro-channel plate and photomultiplier tube and identified through comparison of exact mass, retention time, and fragmentation information with genuine standards and spectral databases. This review introduces general concepts and technical approaches to metabolite profiling, characterization of metabolite-protein interactions, and metabolite imaging.

Figure 1. Conventional untargeted metabolomics workflow. (Click to enlarge)

Sample preparation

Sample preparation is critical to metabolite analysis. While methods vary depending on the experimental goal, they are typically divided into the following steps: (i) quenching to halt metabolism, (ii) cell harvesting through medium removal (in the case of microorganism and animal cell cultures), (iii) cell lysis, and (iv) metabolite extraction (6). Commonly, cellular metabolism is quenched through the addition of cold methanol to a culture broth or flash-freezing of tissues in liquid nitrogen (7). This procedure can be skipped by utilizing a filtration method that enables fast separation of cells from the medium (8). For metabolite extraction, a wide range of solvent systems and temperatures are used (from boiling ethanol to cold mixtures of organic solvents); however, methods of choice include cold methanol and mixtures of methanol and chloroform (9-13).

Ion source technology

MS requires formation of gas phase ions that can be resolved through the manipulation of electromagnetic fields. This is commonly achieved using electron impact ionization (EI), electrospray ionization (ESI), or matrix-assisted laser desorption/ionization (MALDI); more detail can be found in Siuzdak's excellent text (14). Briefly, in EI, volatile metabolites are first separated using GC and then ionized by bombardment with an electron beam; this process results in formation of radical cations and extensive fragmentation. Many metabolites have insufficient vapor pressure, even at high temperature (i.e., sugars, amino acids), and must be chemically modified (derivatized typically via silylation or alkylation) to increase their volatility and stability. However, chemical modification is not well-suited for thermally labile molecules or for chemicals lacking derivatizable groups (e.g., amino, hydroxyl, or carboxyl groups). Lee and Fiehn's article offers detailed discussion and protocols for GC/MS metabolite analysis (9).

The development of ESI has been a break-through for MS analysis of intact biomolecules. The technique allows desorption and ionization of a wide range of molecules directly from the liquid phase. Therefore, it can be directly interfaced with either LC or CE. ESI is based on the formation and drying of charged liquid droplets. Fine droplets are formed through a charged nebulizer needle. As solvent evaporates, the charges are concentrated on the surface. When surface charge repulsion exceeds surface tension, droplets break up into smaller droplets. This consecutive concentration process continues until gas phase ions are ultimately formed.

MALDI is another widely used ionization approach for biomolecules. MALDI is performed using a large excess of compounds (matrix) that absorb light and sublime into the gas phase. These matrices efficiently assist desorption and ionization of a broad range of analytes for subsequent mass spectrometric analysis. However, the matrix itself ionizes and is detected, which greatly complicates analysis for ions with mass-to-charge ratio (m/z) <500 Da. It should be noted that the composition of the MALDI matrix can be modified to reduce these matrix background effects (15-17), and there have been recent innovations in matrix-free technologies. For example, nanostructure-initiator MS (NIMS) is a new technology for metabolite analysis. Here, metabolites are adsorbed onto a vacuum-compatible initiator liquid-coated nanostructured surface (18). This surface, when irradiated with a laser, vaporizes the initiator causing the desorption and ionization of analytes. It should be noted that the adsorption of nonpolar metabolites to a hydrophobic NIMS surfaces has been found to reduce signal suppression in complex biological samples (18).