Knapp J.S., Cabrera W.L. Metabolomics: Metabolites, Metabonomics, and Analytical Technologies

Подождите немного. Документ загружается.

From Metabolic Profiling to Metabolomics 137

Another hybrid instrument used in the metabolic profile is the linear ion trap-triple

quadrupole mass spectrometer (Q-Trap) [157,158,168,212]. The Q-Trap mass spectrometer is

a modified triple quadrupole where the Q3 region can be operated either as a conventional

quadrupole mass filter or as a linear ion trap with axial ion ejection. Thus, the instrument

encompasses the functionality of an ion trap mass spectrometer, with its associated high

sensitivity for product ion scanning, and that of a triple-quadrupole mass spectrometer. The

system also has MS

capabilities which are useful to determine the origin of the fragments

[213]. The most powerful MS ion analyzers, in terms of resolution, are the FT-ICR

and the

Orbitrap; these instruments can also perform MS/MS. Nevertheless their application, coupled

with LC in metabolomics is rare [214-217], and an increasing number of applications,

especially for LC-Orbitrap can be possible in a near future.

Figure 4. Workflow of a metabolomics GC-MS or LC-MS(/MS) platform.

Ion mobility spectrometry (IMS) was introduced in the ’70s as a low resolution ion-

separation and detection device [218]. IMS is based on the fact that ions with different shapes

travel at different speeds when they are pulled by a weak electric field through a drift cell

filled with a buffer gas. Coupling of IMS to MS results in a 2D, orthogonal separation

technique. IMS was interfaced with ToF-MS in the late ’90s [219], permitting the

Chiara Cavaliere, Eleonora Corradini, Patrizia Foglia et al. 138

simultaneous acquisition of ion mobility spectra and mass spectra in a single run. The

combined LC-IM-MS approach using Q-ToF with electrospray ionization, has been recently

tested for urine metabolome study, and demonstrates the potential for high coverage, high

throughput analysis [220].

Whatever the technological platform (hardware) was chosen, a metabolomic experiment

include, in addition to the experimental design, a series of operational steps as reported, for

example, in Figure4.

Direct Infusion MS and MS/MS

Direct MS (DMS) analyses of complex metabolite mixtures in conjunction with

chemometric data analysis offer a viable solution when high-throughput screening is

mandatory. The high-throughput capacity of DMS fingerprinting of complex mixtures is

similar to that of NMR fingerprinting.

Direct infusion MS (DIMS), or flow injection of complex metabolic extracts without

chromatographic separation via ESI provides a sensitive, high-throughput method for

metabolic fingerprinting. The greater sensitivity compared to NMR makes it a very useful

approach for large-scale screening. Obviously, DI-ESI-MS analysis is much more susceptible

to ionization suppression than LC-ESI-MS, thus it is not usually advocated as a quantitative

method. The utilization of nano-ESI reduces ionization suppression effects due to the

increased ionization efficiency of nano-DIMS, and chip-based nanospray emitters provide a

fully automated platform for high-throughput DIMS metabolite measurements [221]. DIMS is

unable to differentiate among isomeric compounds, however DIMS

using ion traps produces

fragment ions that often enable the differentiation of isomeric structures. Recently a rapid

metabolic profiling method using both untargeted and targeted DIMS

with a relatively low

resolution linear ion trap mass spectrometer has been shown to yield sufficient precision and

accuracy for application in genetical metabolomics provided that suitable software tools were

developed [222].

FT-ICR-MS is a powerful tool for DIMS because of its very high mass resolution (10

)

and mass accuracy (<1 ppm), and it has been successfully applied for metabolite-

fingerprinting studies [223,224]. However, large ion populations influence negatively the

mass-measurement accuracy and limit the dynamic range of FT-ICR when a wide m/z range

is tapped. Narrow m/z ranges are therefore commonly acquired separately to increase the

dynamic range and mass accuracy for metabolic profiling. An optimized strategy, namely

high sensitivity selected ion monitoring (SIM)-stitching approach, is usually followed. Each

wide-scan mass spectrum is recorded as a series of overlapping selected ion monitoring (SIM)

windows that are stitched together using novel algorithms [225]. This, reducing space-charge

effects, increases the dynamic range and maintains high mass accuracy.

Ion suppression during ESI, ion–ion interactions in the detector cell, and thermally-

induced white noise remain major challenges for this approach [226]. Orbitraps may be a

good alternative to the more expensive and higher maintenance FT-ICR mass spectrometers,

with similar resolving powers and mass accuracies of 2–5 ppm.

From Metabolic Profiling to Metabolomics 139

Desorption and Imaging MS

MALDI-MS is a popular analytical technique for protein and peptide analysis. Although

the high throughput nature of MALDI-MS makes it an ideal tool for large-scale metabolomic

studies, its application in the field has been rather limited. Due to the elevated chemical

background signals generated by the matrix in the low molecular-weight region that obscures

the detection of metabolites in the range, MALDI-MS has been confined to the targeted

analysis of high-molecular weight metabolites [227,228]. Recent advancements in laser

desorption techniques include matrices that have minimal background signals in the low-

molecular-weight region (‘‘ionless matrices’’), yet still assisting an efficient

ionization/desorption of the analytes [229,230], and DI-MS from porous silicon chips [231].

These offer exciting new opportunities for the utilization of desorption-ionization in

metabolite screening and fingerprinting; however, the technique is still subject to ion

suppression and yields poor quantitative detection of metabolites [232].

Figure 5a. DESI (desorption electrospray) scheme.

Desorption ESI (DESI) is a ionization technique that combines features of ESI and

desorption ionization to permit analysis directly from a surface with virtually no sample

preparation. An electrospray emitter is used to generate a spray of charged micro droplets that

is directed towards an ambient sample surface. Molecules on the surface are subsequently

desorbed, ionized, desolvated and directed to the MS inlet [233]. Virtually no sample

preparation is required for DESI, thus allowing the direct analysis of tissues or biological

fluids that can be deposited on an inert surface and then analyzed. However, DESI

experimental conditions typically require optimization for each sample type, so time must be

invested initially in optimizing the experimental parameters such as the chemical and physical

nature of the surface and the nature of the spraying solvent [234]. Also the geometry of the

system sample surface- sprayer tip- MS inlet, directly affects the ionization efficiency and

sensitivity. DESI seems to have a higher tolerance to sample-matrix effects than ESI,

however the quantitative precision of DESI, as of other surface ionization techniques, is less

Chiara Cavaliere, Eleonora Corradini, Patrizia Foglia et al. 140

than that of ESI. The application of DESI in metabolomics is relatively new, but its ambient

DI properties as well as its high throughput capacity make it an attractive tool for

metabolomics. Figure 5a shows a DESI scheme, and Figure 5b the images produced from an

analysis of lipids in a rat brain tissue sample.

Figure 5b. Images produced from an analysis of lipids in a rat brain tissue sample. The first (up left) is

an optical image, and the others are ion images created from desorption electrospray ionization

analysis.

Extractive ESI (EESI) is another new ESI technique that uses two separate sprayers. One

sprayer nebulizes the sample solution that intersects with a second electrospray containing

charged micro-droplets of the ionizing reagent solvent, usually an acidic aqueous methanol.

Analyte molecules are ionized following collision with the reagent micro-droplets and then

mass analyzed. EESI is related to DESI, but was developed for the direct analysis of trace

compounds in the solution phase, especially when complex mixtures are of interest [235].

EESI utilizes two spray sources, eliminating the use of a surface on which the analyte is first

collected. One spray source nebulizes the sample while the other provides charged solvent

droplets. The two spray sources are set at an angle to each other and to the mass spectrometer

inlet so as to introduce the analyte of interest directly to the source [236]. The advantage of

EESI is its ability to analyze complex biological samples, such as urine and serum, directly,

with minimum or no sample preparation for an extended period of time. The direct infusion of

such complex biological samples to conventional ESI ion sources causes an irrecoverable loss

in signal intensity due to the formation of salt adducts, sample carry-over, or cumulative

build-up of non-volatile components in the ion source. The long-term spray stability of

untreated biological samples in EESI is very promising for high throughput metabolomics, as

EESI significantly reduces data-collection interruptions due to frequent ion source cleaning to

From Metabolic Profiling to Metabolomics 141

remove non-volatile accumulations associated with the ESI-DIMS of crude biological

samples [143]. EESI may represents a valid, more sensitive alternative to NMR in

metabolomics; recently, it has been demonstrate that NMR data and EESI mass spectral data

can be cross-validated [237].

Imaging MS (IMS) is an other emerging technology that permits the direct analysis and

determination of the distribution of molecules in tissue sections. Tissues are analyzed intact

and thus spatial localization of molecules within a tissue is preserved [238]. To investigate the

spatial distribution of specific biomolecules, an increasing number of work groups are aiming

to combine the sensitivity and specificity of mass spectrometry with imaging capabilities.

This has included both MALDI and ESI, and has provided fresh impetus to secondary ion

mass spectrometry (SIMS) imaging. The potential of IMS has long been recognized. The first

elemental imaging experiments using SIMS were made in the ’60s [239,240]. However,

widespread bimolecular IMS had to wait for the advent of MALDI and SIMS methodologies

capable of generating ions from biomolecules with sufficient sensitivity. The high mass

capabilities of MALDI enabled it to be readily applied for imaging protein distributions

within tissue sections [241], whereas the principle SIMS application in semi-conductor

research directed its development toward higher spatial resolution. The success of imaging

MALDI has begun to steer SIMS developments toward high mass molecules while retaining

the high spatial resolution capabilities. However, as a rule, MALDI imaging and SIMS

imaging provide spatial information about different classes of biological compounds. MALDI

provides proteomics information, and SIMS that of lipids and other surface active, relatively

low MW species. Hence, MALDI can record the spatial distribution of high mass molecules

using the chemically specific molecular ions; however, typical spatial resolutions are

approximately 25 µm or more (10 µm sources are now available). SIMS is able to provide

high spatial resolution images, sub-micron is routine and 50 nm is commercially available

[242];

however, the molecular ion mass range is much lower than that of MALDI, most

imaging experiments use ions of m/z <500.

MALDI can analyze hundreds of proteins directly from tissue sections, and combining

this with spatial coordinates, allows the spatial distribution of these proteins to be determined

in parallel, without a label and within practical time-scales [243]. The sample preparation,

spatial resolution and sensitivity of the ionization step, are all important parameters that affect

the type of information obtained. Recently, significant progress has been made in each of

these steps for both SIMS and MALDI imaging of biological samples. Mass resolution is an

important feature because it defines the degree of chemical specificity and, through

improving the precision of mass measurement, helps improve mass accuracy. The ultra-high

mass resolution and mass accuracy of the Fourier transform ion cyclotron resonance mass

spectrometer, resolution >100,000 and sub part-per-million mass accuracy, is beginning to be

developed for imaging mass spectrometry experiments [244,245].

The advantage of a SIMS analysis is that tissue sections did not need any further sample

preparation steps: the sample plate can be mounted and the analysis performed. Impressive

biological images of atomic ions and low mass fragments can be obtained directly from

tissue. The sensitivity improvements provided by polyatomic primary ions, added as very thin

films, show sufficient intensity for imaging of intact biomolecules [246-248].

DESI, being an atmospheric pressure (AP) surface ionization technique, can be used as a

chemical IMS. Many lipid species are easily ionized by DESI, making them attractive target

molecules from which to create molecular images of thin tissue sections. DESI-MS has been

Chiara Cavaliere, Eleonora Corradini, Patrizia Foglia et al. 142

used to construct chemical images of tissue sections of mouse pancreas, rat brain, and

metastatic human liver adenocarcinoma, as well as whole tissue analysis of adipose tissue

surrounding a chicken heart giving strong signals from the major lipid components of

biological membranes [249-253]. Theoretically, DESI may be coupled with all the MS and

MS/MS analyzer and, although the IMS application of DESI has been limited to

phospholipids, the technique seems to have the potential for a wider metabolite profiling.

AP-IMS techniques are particularly attractive because, in principle, working at AP also

enables the study of live specimens. Laser ablation ESI (LAESI) [254,255] is a new AP-IMS

technique. LAESI IMS is realized as a combination of lateral imaging and depth profiling by

tissue ablation with single laser pulses of ca. 2.94 μm wavelength. The O-H vibrations of the

native water molecules in the tissue samples readily absorbed the laser pulse energy leading

to ablation, i.e., the ejection of a microscopic volume of the sample in the form of neutral

particulates and/or molecules. This plume was then intercepted by an electrospray, and the

ablated material was efficiently post-ionized. Tandem MS was performed on numerous ions

to help with metabolite structure assign. A recent, novel development of LAESI was the

combination of lateral imaging and molecular depth profiling capabilities, which enabled a

truly 3D metabolite distribution imaging. Although 3D ambient imaging with LAESI has

been proven feasible in plant tissues, further improvements are needed in spatial resolution

[255].

Data Handling

A great drawback of all “omics” sciences where a large number of qualitative and

quantitative data can be generated per single run is the problem of data analysis. Hence, there

is a demand for computational tools to handle and interpret the large amounts of data. For

metabolomics to be successful, raw analytical data must be converted into metabolites

(named chemicals) and their concentrations, or variation of concentration, data that can be

usefully interpreted for biological research.

The data-processing techniques for the deconvolution of chromatographic peaks, library

search, and GC retention index developed originally by Kovats [256] have been in use since

the ’70s to help identification of peaks recorded by GC-MS. [257] Metabolite profiling by

chromatography-MS and statistical analysis still relies on efficient data-processing

procedures, and minimum reporting requirements have recently been suggested [258,259]. In

all published metabolomic studies based on GC-MS, the starting point of the data analysis has

been the deconvolution of chromatograms, followed by peak matching, and then

identification of the differences between samples. Multivariate methods have been developed

and used to clarify chromatographic and spectral profiles from overlapping chromatographic

peaks obtained using various types of hyphenated chromatography systems. The multivariate

curve resolution methods can be divided into iterative, non-iterative, and hybrid approaches,

and all present both advantages and disadvantages [134]. Software programs have been

developed capable of spectral filtering (noise elimination), peak detection finding the peaks

corresponding to the same compounds, m/z alignment (aiming at matching the corresponding

peaks across multiple sample runs) and normalization (adjusting the intensities within each

sample run by reducing the systematic error) [260-264]. Software packages have augmented

their capability during the years, and many can be downloaded from the Internet.

From Metabolic Profiling to Metabolomics 143

After data alignment and correction of retention times, a data matrix can be generated

from the peak lists as output for subsequent multivariate statistical analysis, including

supervised classification for fingerprinting and principal component analysis for data

visualization [153]. Visualization of complex mass spectrometric data sets is becoming

increasingly important in metabolomics. In a recent paper a versatile tool suitable for many

frequently occurring tasks handling LC-MS data is described [265]. Depending on the task at

hand, different views may be used: single spectra visualized as a plot of intensity against

mass-to-charge ratio (1D view); LC-MS maps displayed in a 2D view from a bird’s-eye

perspective with color-coded intensities; selected regions of LC-MS maps displayed in a 3D

view.

The chemical identification of metabolites is fundamental for the extraction of biological

context from the data. It is not easy to identify a metabolite in a metabolome, especially low

level metabolites which are at or slightly above the noise level or are masked by other

metabolites. An estimated total number of possible metabolites ranges from 200,000 to

1,000,000 [4,266]. A number of strategies are being tried out to assist in the chemical

identification of the unknowns, including the development of metabolite-specific mass

spectral libraries and databases [267-270]. The basis on which metabolites are identified

varies among the metabolomics community. In an effort to standardize the reporting and

interpretation of metadata, the Metabolomics Standard Initiative (MSI) [271] has formed five

working groups that follow the general workflow model in metabolomics: biological context;

chemical analysis; data processing; ontology; and data exchange.

Two types of identification are possible: putative or preliminary identification and

definite identification. A global metabolomic study involves the search for all metabolites in a

biological specimen. If the search is for known metabolites, the identification involves

comparing the experimental data with that of pure standards. Experimentally determined

accurate mass or electron-impact mass spectrum are typically applied for putative

identification. In LC-MS and DIMS the accurate mass is used to define molecular formulae

from which suitable metabolites can be derived by searching electronic resources. However,

isomers have the same accurate mass and therefore require a separate, orthogonal property for

definite identification of all potential isomers. Most metabolite identifications reported are

typically non-novel as they have been previously characterized, identified, and reported at a

rigorous level in literature. Thus, non-novel metabolites not being identified for the first time

are typically identified through the co-characterization with authentic chemical standards. The

Chemical Analysis Working Group has recently proposed a guideline for the identification of

non-novel metabolites [258], in which a minimum of two independent and orthogonal data

relative to an authentic standard compound, typically retention time or index and

fragmentation mass spectrum analyzed under identical experimental conditions, are

considered necessary for metabolite identification.

However, if the metabolites are not known, metabolite identification is much more

difficult. This effort would involve multiple chromatographic separations using different

column chemistries and mobile phases. It also would require high resolution MS and MS/MS

for accurate mass analysis, statistical analysis software for data mining, metabolite databases,

and access to an index of hundreds of pure compounds for confirmation. Databases are still at

an early level of development, and techniques, such as chemical-formula determination via

accurate mass, often serve only to narrow the potential search field rather than provide an

unambiguous formula. Biomarker identification therefore remains a potentially time-

Chiara Cavaliere, Eleonora Corradini, Patrizia Foglia et al. 144

consuming process and a limitation for this technique. Some freely and commercially

available web-resources for MS-based metabolomics are mainly reported in references

[261,272] and also in the literature quoted in this section.

Future Perspectives

LC-MS is certainly going to become a key technology for the provision of global

metabolomics. Major advances in available technologies and a totally new approach to

analysis are essential before a true metabolomics, not achievable with the current state of

development in analytical science, can be performed. This will include developments in all

steps of metabolite analysis, from sampling, sample storage and preparation, to data

acquisition, storage and processing, coupled with a greater understanding and application of

bioinformatics.

Future developments in exhaustive mining of the metabolome will no doubt evolve

through improvements in chromatographic technologies and sensitive MS instruments with a

high capability for mass accuracy. Continued improvements in miniaturized MS, multi-

dimensional chromatography, and multiplexed MS approaches will offer new opportunities in

metabolomics. An increase in the number of chemically identifiable metabolites is a

fundamental key to the interpretation and understanding of the biological context of

metabolomics experiments. High throughput techniques, such as DESI and EESI, are very

attractive and will certainly undergo major improvements in sensitivity, dynamic range, and

high resolution MS/MS coupling. Finally, chemical imaging techniques based on MS are very

exciting and, although not a novelty, still in their infancy. In vivo analysis of intact samples is

an attractive proposition, and in this context, NMR-based technologies have a distinct

advantage over instruments based on MS. But every gap may be overcome.

References

[1] van der Werf, MJ; Jellema, RH; Hankemeier, T. Towards replacing closed with open

target selection strategies. J. Ind. Microbiol. Biotechnol., 2005 32, 234-252.

[2] van der Werf, MJ. Microbial metabolomics: replacing trial-and-error by the unbiased

selection and ranking of targets. Trends Biotechnol., 2005 23, 11-16.

[3] Sumner, LW; Mendes, P; Dixon, RA. Plant metabolomics: large-scale phytochemistry

in the functional genomics era. Phytochemistry, 2003 62, 817-36.

[4] Fiehn, O; Kopka, J; Dormann, P; Altmann, T; Trethewey, RN; Willmitzer, L.

Metabolite profiling for plant functional genomics. Nat. Biotechnol., 2000 18, 1157-

1161.

[5] Raamsdonk, LM; Teusink, B; Broadhurst, D; Zhang, N; Hayes, A; Walsh, MC; Berden,

JA; Brindle, KM; Kell, DB; Rowland, JJ; Westerhoff, HV; van Dam, K; Oliver, SG. A

functional genomics strategy that uses metabolome data to reveal the phenotype of

silent mutations. Nat. Biotechnol., 2001 19, 45-50.

[6] Clish, CB; Davidov, E; Oresic, M; Plasterer, TN; Lavine, G; Londo, T; Meys, M; Snell,

P; Stochaj, W; Adourian, A; Zhang, X; Morel, N; Neumann, E; Verheij, E; Vogels, JT;

From Metabolic Profiling to Metabolomics 145

Havekes, LM; Afeyan, N; Regnier,F; van der Greef, J; Naylor, S. Integrative biological

analysis of the APOE3-leiden transgenic mouse. Omics, 2004 8, 3-13.

[7] Goodacre, R; Vaidyanathan, S; Dunn, WB; Harrigan, GG; Kell, DB. Metabolomics by

numbers: acquiring and understanding global metabolite data. Trends Biotechnol., 2004

22, 245-252.

[8] Brindle, JT; Antti, H; Holmes, E; Tranter, G; Nicholson, JK; Bethell, HW; Clarke, S;

Schofield, PM; McKilligin, E; Mosedale, DE; Grainger, DJ. Rapid and noninvasive

diagnosis of the presence and severity of coronary heart disease using 1H-NMR-based

metabonomics. Nat. Med., 2002 8, 1439-1444.

[9] Sabatine, MS; Liu, E; Morrow, DA; Heller, E; McCarroll, R; Wiegand, R; Berriz, GF;

Roth, FP; Gerszten, RE. Metabolomic identification of novel biomarkers of myocardial

ischemia. Circulation, 2005 112, 3868-3875.

[10] Kawashima, H; Oguchi, M; Ioi, H; Amaha, M; Yamanaka, G; Kashiwagi, Y;

Takekuma, K; Yamazaki, Y; Hoshika, A; Watanabe, Y. Primary biomarkers in cerebral

spinal fluid obtained from patients with influenza-associated encephalopathy analyzed

by metabolomics. Int J. Neurosci., 2006 116, 927-936.

[11] [11] Ippolito, JE; Xu, J; Jain, S; Moulder, K; Mennerick, S; Crowley, JR; Townsend,

RR; Gordon, JI. An integrated functional genomics and metabolomics approach for

defining poor prognosis in human neuroendocrine cancers. Proc. Natl. Acad. Sci. 2005

102, 9901-9906.

[12] Ryan, D; Robards, K. Metabolomics: The greatest omics of them all? Anal. Chem.,

2006 78, 7954-7958.

[13] Oliver, SG; Winson, MK; Kell, DB; Baganz, F. Systematic functional analysis of the

yeast genome. Trends Biotechnol., 1998 16, 373-378.

[14] Goodacre, R. Making sense of the metabolome using evolutionary computation: seeing

the wood with the trees. J. Exp. Bot., 2005 56, 245-254.

[15] Fiehn, O. Combining genomics, metabolome analysis, and biochemical modelling to

understand metabolic networks. Comp. Funct. Genom., 2001 2, 155-168.

[16] Nicholson, JK; Lindon, JC; Holmes, E. “Metabomics": understanding the metabolic

responses of living systems to pathophysiological stimuli via multivariate statistical

analysis of biological NMR spectroscopic data. Xenobiotica, 1999 29, 1181-1189.

[17] Fiehn, O. Metabolomics - the link between genotypes and phenotypes. Plant Mol. Biol.,

2002 48, 155-71.

[18] Lind

on, JC; Holmes, E; Nicholson, JK. So what's the deal with metabonomics? Anal.

Chem., 2003 75, 384A-391A.

[19] MacKenzie, DA; Defernez, M; Dunn, WB; Brown, M; Fuller, LJ; de Herrera, S;

Guenther, A; James, SA; Eagles, J; Philo, M; Goodacre, R; Roberts, IN. Relatedness of

medically important strains of Saccharomyces cerevisiae as revealed by phylogenetics

and metabolomics. Yeast, 2008 25, 501-512.

[20] Smedsgaard, J; Nielsen, J. Metabolite profiling of fungi and yeast: from phenotype to

metabolome by MS and informatics. J. Exp. Bot., 2005 56, 273-286.

[21] Allwood, JW; Ellis, DI; Heald, JK; Goodacre, R; Mur, LAJ. Metabolomic approaches

reveal that phosphatidic and phosphatidyl glycerol phospholipids are major

discriminatory non-polar metabolites in responses by Brachypodium distachyon to

challenge by Magnaporthe grisea. Plant J., 2006 46, 351-368.

Chiara Cavaliere, Eleonora Corradini, Patrizia Foglia et al. 146

[22] Hall, RD. Plant metabolomics: from holistic hope, to hype, to hot topic. New Phytol.,

2006 169, 453-468.

[23] Kell, DB. Systems biology, metabolic modelling and metabolomics in drug discovery

and development. Drug Discov. Today., 2006 11, 1085-1092.

[24] Dunn, WB; Broadhurst, DI; Deepak, SM; Buch, MH; McDowell, G; Spasic, I; Ellis, D;

Brooks, N; Kell, DB; Neyses, L. Serum metabolomics reveals many novel metabolic

markers of heart failure, including pseudouridine and 2-oxoglutarate. Metabolomics,

2007 3, 413-426.

[25] Kenny, LC; Broadhurst, D; Brown, M; Dunn, WB; Redman, CWG; Kell DB; Baker,

PN. Detection and identification of novel metabolomic biomarkers in preeclampsia.

Reprod. Sci., 2008 15, 591-597.

[26] Lindon, JC; Holmes, E; Nicholson, JK. Metabonomics in pharmaceutical R & D. FEBS

J., 2007 274, 1140-1151.

[27] Tanaka, Y; Higashi, T; Rakwal, R; Wakida, S-Ii; Iwahashi, H. Quantitative analysis of

sulfur-related metabolites during cadmium stress response in yeast by capillary

electrophoresis-mass spectrometry. J. Pharm. Biomed. Anal., 2007 44, 608-613.

[28] Viant, MR. Metabolomics of aquatic organisms: the new 'omics' on the block. Mar.

Ecol.-Prog. Ser., 2007 332, 301-306.

[29] Herrgard, MJ; Swainston, N; Dobson, P; Dunn, WB; Arga, KY; Arvas, M; Bluthgen,

N; Borger, S; Costenoble, R; Heinemann, M; Hucka, M; Le Novere, N; Li, P;

Liebermeister, W; Mo, ML; Oliveira, AP; Petranovic, D; Pettifer, S; Simeonidis, E;

Smallbone, K; Spasic, I; Weichart, D; Brent, R; Broomhead, DS; Westerhoff, HV;

Kirdar, B; Penttila, M; Klipp, E; Palsson, BO; Sauer, U; Oliver, SG; Mendes, P;

Nielsen, J; Kell, DB. A consensus yeast metabolic network reconstruction obtained

from a community approach to systems biology. Nat. Biotechnol., 2008 26, 1155-1160.

[30] Wishart, DS; Knox, C; Guo, AC; Eisner, R; Young, N; Gautam, B; Hau, DD;

Psychogios, N; Dong, E; Bouatra, S; Mandal, R; Sinelnikov, I; Xia, J; Jia, L; Cruz, JA;

Lim, E; Sobsey, CA; Shrivastava, S; Huang, P; Liu, P; Fang, L; Peng, J; Fradette, R;

Cheng, D; Tzur, D; Clements, M; Lewis, A; De Souza, A; Zuniga, A; Dawe, M; Xiong,

Y; Clive, D; Greiner, R; Nazyrova, A; Shaykhutdinov, R; Li, L; Vogel, HJ; Forsythe, I.

HMDB: a knowledge-base for the human metabolome. Nucleic Acids Research, 2009

37, D603-D610.

[31] Brown, M; Dunn, WB; Ellis, DI; Goodacre, R; Handl, J; Knowles, JD; O’Hagan, S;

Spasic, I; Kell, DB. A metabolome pipeline: from concept to data to knowledge.

Metabolomics, 2005 1, 39-51.

[32] Dunn, WB. Current trends and future requirements for the mass spectrometric

investigation of microbial, mammalian and plant metabolomes. Phys. Biol., 2008 5, 1-

24.

[33] Lu, W; Bennett BD; Rabinowitz, JD. Analytical strategies for LC-MS-based targeted

metabolomics. J. Chromatogr. B, 2008 871, 236-242.

[34] Griffin, JL. Metabonomics: NMR spectroscopy and pattern recognition analysis of

body fluids and tissues for characterisation of xenobiotic toxicity and disease diagnosis.

Curr. Opin. Chem. Biol., 2003 7, 648-654.

[35] Want, EJ; N

ordstrom, A; Morita, H; Siuzdak, G. From Exogenous to Endogenous: The

Inevitable Imprint of Mass Spectrometry in Metabolomics. J. Proteome Res., 2007 6,

459-468.