ACS Publications. Most Trusted. Most Cited. Most Read
Credentialing Features: A Platform to Benchmark and Optimize Untargeted Metabolomic Methods
More

OR SEARCH CITATIONS

My Activity
Publications

Figure 1Loading Img
Download Hi-Res ImageDownload to MS-PowerPointCite This:Anal. Chem. 2014, 86, 19, 9583-9589
  • Open Access
Article

Credentialing Features: A Platform to Benchmark and Optimize Untargeted Metabolomic Methods
Click to copy article linkArticle link copied!

View Author Information
† Department of Chemistry, Washington University in St. Louis, St. Louis, Missouri 63130, United States
‡ Departments of Genetics and Medicine, Washington University School of Medicine, St. Louis, Missouri 63110, United States
*E-mail: [email protected]
Open PDFSupporting Information (1)

Analytical Chemistry

Cite this: Anal. Chem. 2014, 86, 19, 9583–9589
Click to copy citationCitation copied!
https://doi.org/10.1021/ac503092d
Published August 26, 2014

Copyright © 2014 American Chemical Society. This publication is licensed under these Terms of Use.

Request reuse permissions

Abstract

Click to copy section linkSection link copied!
High Resolution Image
Download MS PowerPoint Slide

The aim of untargeted metabolomics is to profile as many metabolites as possible, yet a major challenge is comparing experimental method performance on the basis of metabolome coverage. To date, most published approaches have compared experimental methods by counting the total number of features detected. Due to artifactual interference, however, this number is highly variable and therefore is a poor metric for comparing metabolomic methods. Here we introduce an alternative approach to benchmarking metabolome coverage which relies on mixed Escherichia coli extracts from cells cultured in regular and 13C-enriched media. After mass spectrometry-based metabolomic analysis of these extracts, we “credential” features arising from E. coli metabolites on the basis of isotope spacing and intensity. This credentialing platform enables us to accurately compare the number of nonartifactual features yielded by different experimental approaches. We highlight the value of our platform by reoptimizing a published untargeted metabolomic method for XCMS data processing. Compared to the published parameters, the new XCMS parameters decrease the total number of features by 15% (a reduction in noise features) while increasing the number of true metabolites detected and grouped by 20%. Our credentialing platform relies on easily generated E. coli samples and a simple software algorithm that is freely available on our laboratory Web site (http://pattilab.wustl.edu/software/credential/). We have validated the credentialing platform with reversed-phase and hydrophilic interaction liquid chromatography as well as Agilent, Thermo Scientific, AB SCIEX, and LECO mass spectrometers. Thus, the credentialing platform can readily be applied by any laboratory to optimize their untargeted metabolomic pipeline for metabolite extraction, chromatographic separation, mass spectrometric detection, and bioinformatic processing.

Copyright © 2014 American Chemical Society

Subjects

what are subjects

The objective of untargeted metabolite profiling is to assay as many endogenous small molecules in a biological sample as possible. (1) Mass spectrometry-based metabolomics represents an established analytical platform that has been widely applied toward this goal and has already yielded many fundamental biological insights. (2-5) Nevertheless, experimental strategies to maximize the number of metabolites profiled are still being developed. (6-8) A major challenge in optimizing metabolomic methodologies has been the difficulty in comparing the number of metabolites profiled in each. Given that the size and identity of the complete metabolome is unknown, it is currently not possible to assess metabolome coverage directly. Consequently, the most common metric used to compare different experimental approaches has been the number of features detected in a sample. (6, 9-12)
We show here that a method detecting a maximal number of features does not necessarily provide the greatest metabolome coverage. We present a solution for the evaluation of untargeted metabolomic method performance that enables us to distinguish between two types of features: artifactual features and biologically derived features. Artifactual features are peaks in metabolomic data that arise from contaminants, chemical noise, and bioinformatic noise. In contrast, biologically derived features are peaks that arise from metabolites in the biological sample being analyzed. We refer to the process of distinguishing artifactual features from features of biological origin as “credentialing”. In the credentialing workflow (Figure 1), standard samples are prepared from Escherichia coli grown in either natural-abundance media or uniformly 13C (U-13C) enriched media. After performing metabolomic experiments utilizing the methods to be compared, our algorithm finds and credentials features based on expected isotope-intensity ratios. This number of credentialed features represents a more reliable metric of metabolome coverage than total feature count because credentialed features are known to be of biological origin and hence are representative of true metabolites. Upon optimizing our bioinformatic workflow by counting credentialed features, we reduce noise features by 15% and increase properly detected and grouped features by 20%. Further, we select several biological features for tandem mass spectrometry (MS/MS) analysis without any prior knowledge of their identity or physiological significance. It is important to emphasize that the credentialing platform described herein is not intended to identify differences between various biological phenotypes (discovery profiling). Rather, the credentialing platform is designed only to compare the performance of different untargeted metabolomic methods. We provide a step-by-step protocol for performing credentialing with E. coli. While other cell types could potentially be used, E. coli is a simple model system whose optimized results will be applicable to the vast majority of metabolomic optimizations.

Figure 1

Figure 1. Overview of the feature credentialing process. A sample is generated from two cultures of E. coli grown in parallel, one grown on natural-abundance glucose and a second grown on 13C-glucose as the sole carbon source. These two cultures are mixed in distinct ratios prior to harvesting, here 1:1 and 1:2. Extraction and LC/MS analysis is then performed on the standard samples. The resulting data are searched for pairs of coeluting peaks which satisfy the following requirements: (i) the intensities of the peaks must reflect the mixing ratio, (ii) the U-13C peak must predict a feasible number of carbons for the mass in question, and (iii) the exact masses of the peaks must predict an integer number of carbons. These requirements define a “credentialed space” in which the apex of a second peak must be found to qualify as an acceptable isotope. These candidate peaks are then aligned and grouped between the two samples. Each peak pair is compared across samples and a second, stricter intensity check is performed. This requires that the ratios of each sample (Ia12/Ia13 and Ib12/Ib13) are proportional to the mixed ratios of each sample. Peaks that pass these filters are considered credentialed.

High Resolution Image
Download MS PowerPoint Slide

Background

Click to copy section linkSection link copied!
Metabolomic studies are complex, multistep experiments with a large number of parameters to optimize. The choice of sample extraction, chromatography, and ionization method strongly influences which metabolites are detected. Establishing protocols which survey the broadest number of metabolites during untargeted profiling has received detailed attention in recent years. (6, 12-15) Previous studies have explored a multitude of experimental variations to improve global metabolome coverage that include the addition of ammonium fluoride and ion-pairing reagents to chromatographic mobile phases, separation strategies ranging from reversed-phase to hydrophilic interaction liquid chromatography (HILIC), different mass analyzers such as time-of-flight and the Orbitrap, and various informatic software solutions for subsequent data processing. (12, 15-18) The extensive list of mutually exclusive experimental possibilities is confounding, particularly to scientists just entering the field of untargeted metabolomics. Yet, to date, comparisons of different methods have been impractical because there is no robust metric for performance evaluation.
Most published comparisons of mass spectrometry-based, untargeted metabolomic methods are evaluated by counting the total number of features detected. A feature is defined as a peak in the metabolomic data set with a unique retention time and mass-to-charge ratio. The number of features detected depends on numerous factors including sample type, metabolite extraction protocols, analyte separation, mass analyzer, and bioinformatic processing. For liquid chromatography/mass spectrometry (LC/MS)-based metabolomics, it is common to detect thousands of features from a biological sample. Importantly, a single metabolite often leads to many features (19) due to: (i) isotopic peaks from naturally occurring 13C, (ii) adduct formation such as hydrogen, ammonium, and sodium adducts, (iii) neutral-loss fragments (loss of a hydroxyl group as water or a carboxylate as carbon dioxide), (iv) other fragmentation (breakage at labile bonds such as esters), (v) multiple-charge states, and (vi) chromatographic effects which result in a single metabolite eluting at more than one retention time.
Informatic solutions have been established to annotate isotopes, adducts, and neutral losses in untargeted metabolomic data sets. (17, 20, 21) Although these approaches are effective, they cannot distinguish signals as endogenous or artifactual. Thus, even after data reduction, a subset of the remaining features are likely the result of contaminants introduced during sample preparation, carryover from previous experiments, chemical noise, or bioinformatic error. These highly variable artifactual signals found in untargeted metabolomic data sets make it challenging to estimate the number of true biologically derived metabolites that are assayed by a particular untargeted LC/MS-based metabolomic experiment. There is therefore a great need to develop a robust metric to the evaluate performance of untargeted metabolomic methods.

Experimental Section

Click to copy section linkSection link copied!
Our filtering process relies on the generation of standard samples derived from a mixture of E. coli grown on 100% natural-abundance glucose and E. coli grown on 100% U-13C-glucose as the sole carbon source. Two standard samples are required for the filtering process; these are generated by mixing natural-abundance E. coli cultures and U-13C-glucose E. coli cultures at either 5 mL/5 mL or 3 mL/6 mL ratios, respectively. The mixed E. coli samples are then extracted, yielding a standard sample for analysis and optimization.

Materials

U-13C-d-Glucose was purchased from Cambridge Isotope Laboratories Inc. (Andover, MA). E. coli strain K12, MG1655 was purchased from ATCC (Manassas, VA). Lennox LB broth powder, 5× M9 salts, and all LC/MS-grade solvents were purchased from Sigma-Aldrich (St. Louis, MO). Cell culture was performed with ultrapure water provided by a Milli-Q system (Millipore).

Growth of E. coli Standards

Cultures were grown in a rotary shaker at 37 °C and 250 rpm. A preculture of E. coli was grown in LB broth for 16 h. Prior to inoculation, 3 mL of preculture was pelleted and resuspended to OD600 = 0.6 in M9 salts. M9 salts were prepared with the following concentrations in sterile Erlenmeyer flasks: 6.8 g/L Na2HPO4·7H2O; 3 g/L KH2PO4; 1 g/L NH4Cl; 0.5 g/L NaCl; 240 mg/L MgSO4; 11 mg/L CaCl2. Salts were divided into two 100 mL aliquots, and to each aliquot, 2 mL of 20% glucose was added with a fresh-filtered syringe. The filter was rinsed with 2 mL of ultrapure water to ensure complete transfer of glucose. One aliquot received U-13C-glucose and the second received natural-abundance glucose. The M9 media was then inoculated with 1 mL of the resuspended preculture per 100 mL of media. Cultures were grown to OD600 = 0.6, at which point they were harvested as described below.

Harvesting of E. coli Standards

Upon reaching OD600 = 0.6, flasks were removed from the shaker and placed on ice. Appropriate volumes of the 12C and 13C cultures were pipetted together into 15 mL centrifuge tubes, also on ice, generating samples with ratios of 1/1 of 1/2 12C/13C culture. These mixtures established two distinct ratios of 12C to 13C feature intensities that could then be used in our credentialing algorithm, described below. Cells were pelleted by centrifugation at 2000g for 10 min at 4 °C. The supernatant was removed via pipet, and the cell pellets were snap-frozen in liquid nitrogen. In addition to the mixed 12C and 13C cultures, natural-abundance (12C) cultures were used as controls. We refer to the mixed samples as “labeled” and the natural-abundance extracts alone as “unlabeled.”

Metabolite Extraction

The mixed E. coli pellets were extracted as previously described. (6) Briefly, cells were lysed by three freeze–thaw cycles in 2/2/1 methanol/acetonitrile/water along with sonication and vortexing. The soluble portion was then vacuum concentrated and reconstituted in 100 μL of 1/1 acetonitrile/water for LC/MS analysis.

LC/MS Analysis

The data shown herein were obtained from an Agilent 6540 UHD QTOF interfaced with an Agilent 1260 Capillary LC. The column used for separation was a Phenomenex Luna NH2 (150 mm × 1 mm, 3 μm). HILIC solvents were A, 95% water in acetonitrile with 10 mM ammonium acetate/10 mM ammonium hydroxide (pH 9.8), and B, 95% acetonitrile in water. HILIC was performed at 45 μL/min with the following linear gradient (minutes, %B): 0, 100%; 5, 100%; 45, 0%; 50, 0%; 51, 100%; 60, 100%. For all experiments, 5 μL of extract was injected. MS parameters were as follows: gas, 300 °C 9 L/min; nebulizer, 35 psi 1000 V; sheath gas, 350 °C 11 L/min; capillary, 3500 V; fragmentor, 175 V; scan rate, 1 scan/s.
To demonstrate the wide applicability of our credentialing approach to other metabolomic platforms, we also analyzed our samples and subsequently validated correct credentialing with multiple chromatographic and mass spectrometrometric technologies. In addition to the Agilent QTOF, we credentialed data from the Thermo QE, the AB SCIEX TripleTOF, and the LECO Pegasus GC-HRT. Chromatographic methods we credentialed include reversed-phase LC and HILIC. Effective parameters for credentialing each of these experimental platforms are listed in the Supporting Information Table S-3.

Data Analysis

Analysis was performed with a custom filtering script that utilizes the XCMS (16) and CAMERA (20) R (22) packages as well as the METLIN (23) database. The script is available on our laboratory Web site at http://pattilab.wustl.edu/software/credential/. The algorithm identifies features of biological origin through two rounds of data filtering, as depicted in Figure 1. Prior to filtering, features are detected from the MS raw data with the XCMS findPeaks.centWave algorithm. In the first round of filtering, coeluting peaks within a single sample are assessed for potential isotopologue pairs differing by [(n)1.003355/z] Da in mass, where n is a whole number, z is the ion’s charge, and the constant is the mass difference between 12C and 13C. Upper and lower bounds of n for each m/z in question were calculated from the distribution of mass per carbon number from the compounds in ECMDB (24) (E. coli Metabolome Database, Supporting Information Figure S-2). The ratios of the putative 12C and 13C peak intensities are then evaluated. Each measured ratio that is not within a set percentage of the mixture ratio of the 12C and 13C culture is disqualified. For credentialing, the default value of 400% is effective.
The two filtered samples with distinct mixture ratios of 12C and 13C are then taken together for a final round of filtering. Peaks from each sample are aligned and grouped. Surviving features found in both samples are evaluated such thatwhere xi is the 12C/13C mixing ratio of the ith sample, ri is intensity ratio (I12C/I13C) of the ith sample, and e (ratio_tol) sets the acceptable tolerance for the intensity ratio relative to the mixing ratio. This two-round intensity filter allows for features with varying 12C and 13C intensity ratios (due to the kinetic isotope effect or carbon fixation of atmospheric CO2) to pass the relaxed first round and stricter second round as long as their intensities vary systematically between samples. All passing features are termed credentialed. Credentialed features are output as a summary table that includes all U-12C peaks determined to be of biological origin.

Results and Discussion

Click to copy section linkSection link copied!
Each step of the untargeted metabolomic workflow can introduce artifactual signals that are not endogenous to the biological sample being analyzed. It is generally not possible to discriminate features of biological origin from artifactual features a priori, and thus, artifactual signals significantly complicate interpretation of untargeted metabolomic results. These artifactual signals can arise from sample contamination during metabolite extraction, carryover from previous experiments, background noise detected by the MS, or misannotation of data during bioinformatic processing. While efforts are made to minimize artifactual signals, it is not possible to completely eliminate them from the features list. We therefore attempted to filter out artifactual signals by using isotopic signatures of cellular metabolism that are easily identified by informatic analysis. We utilized the widely available and extensively characterized E. coli strain K12 to generate isotopically enriched biological extracts. Two cultures were prepared in parallel, one containing 12C (natural-abundance) glucose and the other containing 13C glucose as the sole carbon source in M9 minimal media. The cultures were mixed in defined ratios and processed through the metabolomic workflow together. By searching the resulting features list for pairs of unlabeled and fully labeled isotopologues and comparing their intensities to the values expected from the culture volume ratios, signals of biological origin can be distinguished from artifactual ones. The output of the approach is a list of credentialed features arising from the biological sample of interest. These features reflect the extent to which the methodology employed was able to capture the metabolome.
The power of stable isotope labeling in conjunction with mass spectrometry has long been leveraged to improve quantitative measurements. Mixing labeled and unlabeled samples has proven to be an effective approach to perform quantitation in proteomics, (25-27) and similar approaches have recently been extended to metabolomics. (28) Mashego et al. developed “mass isotopomer ratio analysis of U-13C labeled extracts” (MIRACLE) in which U-13C labeled metabolites obtained from yeast grown in defined culture medium are mixed with unlabeled sample extracts to improve quantitation. (29) More recently, an innovative variation of 12C–13C metabolite mixing was developed in which cells are grown in either 5% or 95% randomly enriched 13C glucose. This experimental strategy, termed isotopic ratio outlier analysis or IROA, leads to a diagnostic isotopic pattern for naturally occurring compounds that can be used for quantitation and metabolite identification during untargeted profiling. (30, 31) Here, we introduce another experimental approach which involves mixing 12C and 13C metabolic extracts. We then use the unique isotopic signals that result from the metabolic transformation of the label as a mechanism to identify features of biological origin.

Contrasting the Credentialing and IROA Platforms

It is worth distinguishing IROA from our credentialing approach. Fundamental to the distinction is that mixing a natural-abundance sample with a U-13C labeled sample in a single ratio does not provide a specific enough signature to effectively discriminate features of biological origin from artifactual features. IROA introduces additional specificity to the isotopic pattern by enriching one sample with 5% 13C and a second sample with 95% 13C, instead of using natural-abundance and U-13C samples. In contrast, credentialing introduces additional specificity to the isotopic pattern by mixing different ratios of natural-abundance and U-13C samples. In credentialing, one sample is made by mixing natural-abundance and U-13C cells at a ratio of 1/1 and a second sample is made by mixing natural-abundance and U-13C cells at a ratio of 1/2. There are experimental benefits of each approach that make the platforms better suited for each of their unique experimental applications. IROA has been used to identify and quantitate differences between biological phenotypes during untargeted profiling. Given that the relative ratio of any given peak between biological phenotypes is unknown during untargeted profiling, the credentialing strategy based on defined ratios is incompatible with this type of discovery analysis. The objective of credentialing, on the other hand, is to identify features of biological origin exclusively from standard E. coli samples. While IROA could be used for this purpose in principle, the credentialing platform is not constrained by the aim of discovery analysis and therefore offers several advantages. First, media needed to produce labeled E. coli samples for credentialing is easily synthesized in any laboratory, whereas IROA media can only be obtained commercially. Second, the credentialing platform is better suited to identify low-intensity features of biological origin. In IROA, the signal intensity of any given metabolite is shifted away from the U-12C peak and the U-13C peak as a function of carbon number. For a metabolite with 10 carbons, as an example, 50% of the signal intensity is lost from the U-12C peak or the U-13C peak. This decrease in signal intensity prevents low-abundance E. coli derived metabolites that are detected in unlabeled samples from being detected with IROA. Because the credentialing platform only uses natural-abundance and U-13C samples, it is not subject to this limitation. Indeed, detection of low-abundance metabolites is of particular importance when optimizing metabolomic methods as these compounds are the most challenging to measure, but can be of great biological importance. Finally, because credentialing only uses E. coli samples, the analysis of the resulting isotopic data can exploit the known relationship between mass and carbon number derived from ECMDB (Supporting Information Figure S-2).

Parameters for Credentialing

To accomplish the filtering of artifactual signals, we created a simple R package. The core function, credential(), has several adjustable parameters allowing various chromatographic and instrumental platforms to be credentialed. These parameters include (i) iso_ppm, the ppm tolerance when searching for 13C isotopes, (ii) iso_rt, the retention-time window in which a peak and its isotope must elute, (iii) mix_tol, the tolerance for the intensity ratio of the 12C and 13C peak, (iv) ratio_tol, the tolerance for the ratio of the intensity ratios between two samples, and (v) mpc_tol, the tolerance for compounds with unusually high or low mass compared to the number of carbons they contain. (Details concerning the calculation of mass per carbon based on the ECMDB can be found in Supporting Information Figure S-2.)
We have determined effective parameters for reversed-phase and hydrophilic interaction liquid chromatography as well as for the Agilent QTOF, Thermo QE, AB SCIEX TripleTOF 5600+, and the LECO Pegasus GC-HRT. These parameters have been experimentally validated and are listed in the Supporting Information Table S-3.
Evaluation of the filtering effectiveness was accomplished by comparing the number of credentialed features found in unlabeled and labeled extracts. In addition to the labeled extracts, natural-abundance (unlabeled) extracts were generated as controls. An unlabeled extract should have no credentialed features if it is not mixed with a labeled extract. Therefore, the number of passing features in an unlabeled extract represents the false positive rate. Initial experiments indicated that filtering based on a single mixed-extract sample was not sufficiently selective to remove the majority of artifactual peaks. We found that a two-sample, relative-intensity filter was most effective. As shown in Table 1, this filtering process is selective. The process credentialed only 0.6% of the negative-control features, whereas 9% of the 12C/13C mixture features were credentialed.
Table 1. Performance of Feature Credentialinga
sample typetotal featurescredentialed featurespercentage credentialed (%)
no injection1564130.8
extraction blank2736180.7
natural-abundance E. coli186431200.6
12C/13C standard sample2356721929.3
a

A summary of the results of the credentialing process after being applied to several different data sets. The rows labeled “no injection” and “extraction blanks” represent credentialed peaks due to carryover from previous credentialing runs. Natural-abundance E. coli is a negative control that estimates the false positive rate of the credentialing process.

To further validate the filtering process, we examined the natural isotopic peaks that were credentialed in our 12C/13C sample. Consider that in a 12C sample many peaks will contain a natural-abundance M + 1 peak which by definition satisfies the mass requirement to be an isotope. The filtering process credentials some of these natural isotopes along with the monoisotopic peak. These are easily detected and removed by established deisotoping methods, but these peaks allowed us to assess how often an M + 1 is credentialed when the M + 0 is not. If this occurs often, it would indicate that the algorithm is inappropriately disqualifying features. We detected 385 credentialed natural isotopes in our mixture sample. Out of the 385 credentialed, natural isotopes only one did not have a corresponding U-12C in the final credentialed features list. This indicates the filtering approach is performing reliably.

Application: Reoptimization of a Previously Published XCMS Method

With an established method to credential features as biological in origin and exclude various noise sources, we set out to optimize our XCMS-based informatic workflow. XCMS is a widely used informatic package suited for the analysis of untargeted LC/MS data sets. The general XCMS workflow involves peak finding, peak grouping across samples, and retention-time alignment. Settings for each step in this process affect the quality of features returned and therefore the overall performance of the untargeted metabolomic workflow. For example, we found that settings for peak picking that cause the annotation of spurious noise peaks as features lower the quality of peak grouping and retention-time alignment (data not shown). Further, using poor grouping parameters can lead to XCMS splitting a single peak into multiple groups, thereby resulting in erroneous statistics.
To generate data for XCMS optimization, a previously published method was replicated. (6) The same LC/MS system, extraction method, and chromatography protocols were utilized as published and described in the Experimental Section. When processing the data, however, we varied several parameters of the XCMS functions findPeaks.centWave(), group(), and retcor(). As the filtering depends on each of these functions, the final number of credentialed features is representative of the quality of XCMS data processing. Previous approaches to optimizing untargeted metabolomic parameters such as these have relied on counting the total number of features detected. Here, we applied our filtering approach to instead count the number of credentialed features and use this as a benchmark for parameter optimization. Our results show that the published method parameters based on total number of features are suboptimal (Table 2). The published parameters do return a greater number of total features, but the number of features of biological origin accurately detected and grouped is substantially lower with these settings. These data highlight that a larger feature number does not necessarily indicate better metabolome coverage and therefore an improved untargeted metabolomic method.
Table 2. Reoptimization of Published XCMS Parametersa
XCMS parameterpublished parameters with optimized peak finding with optimized retcor and group
ppm151212
peak width10, 12015, 14015, 140
mzwid0.0150.0150.015
bw5510
gapInit  0.6
total features320102726027260
credentialed features147517761817
a

Parameters used and the results of each step in the optimization process are shown. Published parameters were taken from a previously published method (ref 6). The column labeled “with optimized peak finding” shows results for the optimization of findPeaks.centWave().

Reoptimization of XCMS parameters resulted in a substantial improvement. Our XCMS parameters led to an increase of 20% in credentialed features (an increase of 342 features), while reducing the total number of features by 15% (a decrease of 4750 features). Parameters for findPeaks.centWave() were determined to be the most critical to the analysis, while further optimization of group and retcor qualified only an additional 41 peaks. It is notable that, prior to optimizing findPeaks.centWave(), optimization of group() parameters increased the number of credentialed features, partially overcoming the negative impact of artifactual signals.

Characterizing Features in Untargeted Metabolomic Data Sets

To translate metabolomic data into biochemical insight, the features generated in a typical untargeted experiment must first be structurally characterized. The standard workflow for structurally characterizing features requires matching MS/MS data of the features of interest to the MS/MS data of authentic standards. Identifying features is the most time-demanding step of the untargeted metabolomic workflow and is generally performed in a targeted manner. That is, MS/MS data are only acquired and interpreted for a handful of features determined to be interesting, usually on the basis of statistical thresholds. While this worfklow is often applied to identify tens of metabolites in a metabolomic study, attempting to identify each of the thousands of features detected in a typical sample with this approach is impractical. New technologies to reduce the time required to establish metabolite identifications are an active area of research, but high-throughput methods to structurally characterize metabolites are not widely available. Moreover, many of the MS/MS data are challenging to interpret. When the MS/MS pattern of a feature does not match any of the MS/MS patterns in metabolite databases, it is difficult to determine if the MS/MS data correspond to an unknown metabolite or merely MS/MS data from an artifactual feature.
The feature credentialing approach offers a mechanism to rapidly filter features that should not be pursued for identification, namely, those features that do not correspond to signals of biological origin. When we applied credentialing to E. coli extract, we reduced the number of features that represent candidates for MS/MS from 23 567 to 2192. The resulting subset of credentialed features can be targeted for MS/MS analysis with standard workflows. As an example, we performed targeted MS/MS on 250 compounds in a single experimental run. These data illustrate that MS/MS experiments could be performed on every feature of biological origin over a minimal and feasible number of analytical runs. Select data are presented in Figure 2A–C. The MS/MS data collected on these features were matched to the METLIN metabolite database and resulted in the identification of three metabolites: uracil, ADP (adenosine diphosphate), and UDP-GlcA (uridine diphosphate glucuronic acid). MS1 spectra and chromatograms for these compounds can be found in Supporting Information Figure S-4.

Figure 2

Figure 2. MS/MS spectra from six representative credentialed features. MS/MS spectra were collected at four collision energies (0, 10, 20, and 40 V) on six credentialed ions. Three of these ions (A) uracil, (B) ADP, and (C) UDP-GlcA were identified based on accurate mass, carbon number, and METLIN database hits. These identifications were confirmed by comparing the experimental MS/MS spectra to the METLIN MS/MS reference spectra as shown. The upper spectrum of each plot is the experimental data, and the lower spectrum is the METLIN reference data. Unmatched peaks are depicted in red. The second three ions (D) 578.0093, (E) 1169.3011, and (F) 848.7473 were classified as unknowns as they did not match any METLIN database entries as either a fragment or parent mass. The MS/MS spectrum of each ion is displayed as normalized intensity at the same four collision energies.

High Resolution Image
Download MS PowerPoint Slide
In addition to generating MS/MS data for metabolites included in databases, it is possible to reliably generate MS/MS data on biological peaks which currently cannot be annotated by metabolomic databases. Because credentialed features have passed our filtering rounds, we know that they are true metabolites of biological origin even if they do not return any database hits. Of the 1827 credentialed features, 392 were not found in METLIN or the METLIN fragment databases. Three such example features are seen in Figure 2D–F. Previously these features may have been discarded as artifacts, but the credentialing platform provides confidence in their authenticity such that they can be reported and referenced in future experiments.

Conclusion

Click to copy section linkSection link copied!
The feature credentialing strategy presented here is a powerful platform to discriminate biological features from the various noise sources prevalent in untargeted metabolomic data. The process is experimentally straightforward and can be easily implemented in any metabolomic laboratory. Feature credentialing reliably removes artifactual features such as those arising from chemical and informatic noise, thereby resulting in a valuable list of features of biological origin. These credentialed features address many of the drawbacks associated with feature counting in comparing method performance on the basis of metabolome coverage. As such, counting credentialed features can be used in the development and optimization of untargeted metabolomic approaches as demonstrated by the reoptimization of XCMS parameters. Credentialing features is also an effective data reduction strategy for untargeted metabolomic results such that a smaller number of peaks can be targeted for MS/MS analysis. In summary, the feature credentialing platform introduced here represents a step toward defining optimal untargeted metabolomic platforms and provides a standard metric to facilitate collaboration between different metabolomic laboratories.

Supporting Information

Click to copy section linkSection link copied!

Additional information as noted in text. This material is available free of charge via the Internet at http://pubs.acs.org.

Terms & Conditions

Most electronic Supporting Information files are available without a subscription to ACS Web Editions. Such files may be downloaded by article for research use (if there is a public use license linked to the relevant article, that license may permit other uses). Permission may be obtained from ACS for other uses through requests via the RightsLink permission system: http://pubs.acs.org/page/copyright/permissions.html.

Author Information

Click to copy section linkSection link copied!
  • Corresponding Author
    • Gary J. Patti - Department of Chemistry, Washington University in St. Louis, St. Louis, Missouri 63130, United States ;  Departments of Genetics and Medicine, Washington University School of Medicine, St. Louis, Missouri 63110, United States;  Email: [email protected]
  • Authors
    • Nathaniel Guy Mahieu - Department of Chemistry, Washington University in St. Louis, St. Louis, Missouri 63130, United States ;  Departments of Genetics and Medicine, Washington University School of Medicine, St. Louis, Missouri 63110, United States
    • Xiaojing Huang - Departments of Genetics and Medicine, Washington University School of Medicine, St. Louis, Missouri 63110, United States
    • Ying-Jr Chen - Department of Chemistry, Washington University in St. Louis, St. Louis, Missouri 63130, United States ;  Departments of Genetics and Medicine, Washington University School of Medicine, St. Louis, Missouri 63110, United States
  • Notes
    The authors declare no competing financial interest.

Acknowledgment

Click to copy section linkSection link copied!

This work was supported by the National Institutes of Health Grants R01 ES022181 (G.J.P.), L30 AG0 038036 (G.J.P.), and the Alfred P. Sloan Foundation.

References

Click to copy section linkSection link copied!

This article references 31 other publications.

  1. 1
    Patti, G. J.; Yanes, O.; Siuzdak, G. Nat. Rev. Mol. Cell Biol. 2012, 13, 263– 269
    Google Scholar
    1
    Innovation Metabolomics: the apogee of the omics trilogy
    Patti, Gary J.; Yanes, Oscar; Siuzdak, Gary
    Nature Reviews Molecular Cell Biology (2012), 13 (4), 263-269CODEN: NRMCBP; ISSN:1471-0072. (Nature Publishing Group)
    A review. Metabolites, the chem. entities that are transformed during metab., provide a functional readout of cellular biochem. With emerging technologies in mass spectrometry, thousands of metabolites can now be quant. measured from minimal amts. of biol. material, which has thereby enabled systems-level analyses. By performing global metabolite profiling, also known as untargeted metabolomics, new discoveries linking cellular pathways to biol. mechanism are being revealed and are shaping our understanding of cell biol., physiol. and medicine.
  2. 2
    Patti, G. J.; Yanes, O.; Shriver, L. P.; Courade, J.-P.; Tautenhahn, R.; Manchester, M.; Siuzdak, G. Nat. Chem. Biol. 2012, 8, 232– 234
    Google Scholar
    2
    Metabolomics implicates altered sphingolipids in chronic pain of neuropathic origin
    Patti, Gary J.; Yanes, Oscar; Shriver, Leah P.; Courade, Jean-Phillipe; Tautenhahn, Ralf; Manchester, Marianne; Siuzdak, Gary
    Nature Chemical Biology (2012), 8 (3), 232-234CODEN: NCBABT; ISSN:1552-4450. (Nature Publishing Group)
    Neuropathic pain is a debilitating condition for which the development of effective treatments has been limited by an incomplete understanding of its chem. basis. We show by using untargeted metabolomics that sphingomyelin-ceramide metab. is altered in the dorsal horn of rats with neuropathic pain and that the upregulated, endogenous metabolite N,N-dimethylsphingosine induces mech. hypersensitivity in vivo. These results demonstrate the utility of metabolomics to implicate unexplored biochem. pathways in disease.
  3. 3
    Khan, A. P.; Rajendiran, T. M.; Ateeq, B.; Asangani, I. A.; Athanikar, J. N.; Yocum, A. K.; Mehra, R.; Siddiqui, J.; Palapattu, G.; Wei, J. T.; Michailidis, G.; Sreekumar, A.; Chinnaiyan, A. M. Neoplasia (N. Y., NY, U. S.) 2013, 15, 491– 501
    Google Scholar
    3
    The role of sarcosine metabolism in prostate cancer progression
    Khan, Amjad P.; Rajendiran, Thekkelnaycke M.; Ateeq, Bushra; Asangani, Irfan A.; Athanikar, Jyoti N.; Yocum, Anastasia K.; Mehra, Rohit; Siddiqui, Javed; Palapattu, Ganesh; Wei, John T.; Michailidis, George; Sreekumar, Arun; Chinnaiyan, Arul M.
    Neoplasia (Ann Arbor, MI, United States) (2013), 15 (5), 491-501CODEN: NEOPFL; ISSN:1522-8002. (Neoplasia Press Inc.)
    Metabolomic profiling of prostate cancer (PCa) progression identified markedly elevated levels of sarcosine (N-Me glycine) in metastatic PCa and modest but significant elevation of the metabolite in PCa urine. Here, we examine the role of key enzymes assocd. with sarcosine metab. in PCa progression. Consistent with our earlier report, sarcosine levels were significantly elevated in PCa urine sediments compared to controls, with a modest area under the receiver operating characteristic curve of 0.71. In addn., the expression of sarcosine biosynthetic enzyme, glycine N-methyltransferase (GNMT), was elevated in PCa tissues, while sarcosine dehydrogenase (SARDH) and pipecolic acid oxidase (PIPOX), which metabolize sarcosine, were reduced in prostate tumors. Consistent with this, GNMT promoted the oncogenic potential of prostate cells by facilitating sarcosine prodn., while SARDH and PIPOX reduced the oncogenic potential of prostate cells by metabolizing sarcosine. Accordingly, addn. of sarcosine, but not glycine or alanine, induced invasion and intravasation in an in vivo PCa model. In contrast, GNMT knockdown or SARDH overexpression in PCa xenografts inhibited tumor growth. Taken together, these studies substantiate the role of sarcosine in PCa progression.
  4. 4
    Tang, W. H. W.; Wang, Z.; Levison, B. S.; Koeth, R. A.; Britt, E. B.; Fu, X.; Wu, Y.; Hazen, S. L. N. Engl. J. Med. 2013, 368, 1575– 1584
    Google Scholar
    4
    Intestinal microbial metabolism of phosphatidylcholine and cardiovascular risk
    Tang, W. H. Wilson; Wang, Zeneng; Levison, Bruce S.; Koeth, Robert A.; Britt, Earl B.; Fu, Xiaoming; Wu, Yuping; Hazen, Stanley L.
    New England Journal of Medicine (2013), 368 (17), 1575-1584CODEN: NEJMAG; ISSN:0028-4793. (Massachusetts Medical Society)
    BACKGROUND Recent studies in animals have shown a mechanistic link between intestinal microbial metab. of the choline moiety in dietary phosphatidylcholine (lecithin) and coronary artery disease through the prodn. of a proatherosclerotic metabolite, trimethylamine-N-oxide (TMAO). We investigated the relationship among intestinal microbiota-dependent metab. of dietary phosphatidylcholine, TMAO levels, and adverse cardiovascular events in humans. METHODS We quantified plasma and urinary levels of TMAO and plasma choline and betaine levels by means of liq. chromatog. and online tandem mass spectrometry after a phosphatidylcholine challenge (ingestion of two hard-boiled eggs and deuterium [d9]-labeled phosphatidylcholine) in healthy participants before and after the suppression of intestinal microbiota with oral broad-spectrum antibiotics. We further examd. the relationship between fasting plasma levels of TMAO and incident major adverse cardiovascular events (death, myocardial infarction, or stroke) during 3 years of follow-up in 4007 patients undergoing elective coronary angiog. RESULTS Time-dependent increases in levels of both TMAO and its d9 isotopologue, as well as other choline metabolites, were detected after the phosphatidylcholine challenge. Plasma levels of TMAO were markedly suppressed after the administration of antibiotics and then reappeared after withdrawal of antibiotics. Increased plasma levels of TMAO were assocd. with an increased risk of a major adverse cardiovascular event (hazard ratio for highest vs. lowest TMAO quartile, 2.54; 95% confidence interval, 1.96 to 3.28; P<0.001). An elevated TMAO level predicted an increased risk of major adverse cardiovascular events after adjustment for traditional risk factors (P<0.001), as well as in lower-risk subgroups. CONCLUSIONS The prodn. of TMAO from dietary phosphatidylcholine is dependent on metab. by the intestinal microbiota. Increased TMAO levels are assocd. with an increased risk of incident major adverse cardiovascular events.
  5. 5
    Jain, M.; Nilsson, R.; Sharma, S.; Madhusudhan, N.; Kitami, T.; Souza, A. L.; Kafri, R.; Kirschner, M. W.; Clish, C. B.; Mootha, V. K. Science 2012, 336, 1040– 1044
    Google Scholar
    5
    Metabolite profiling identifies a key role for glycine in rapid cancer cell proliferation
    Jain, Mohit; Nilsson, Roland; Sharma, Sonia; Madhusudhan, Nikhil; Kitami, Toshimori; Souza, Amanda L.; Kafri, Ran; Kirschner, Marc W.; Clish, Clary B.; Mootha, Vamsi K.
    Science (Washington, DC, United States) (2012), 336 (6084), 1040-1044CODEN: SCIEAS; ISSN:0036-8075. (American Association for the Advancement of Science)
    Metabolic reprogramming has been proposed to be a hallmark of cancer, yet a systematic characterization of the metabolic pathways active in transformed cells is currently lacking. Using mass spectrometry, we measured the consumption and release (CORE) profiles of 219 metabolites from media across the NCI-60 cancer cell lines, and integrated these data with a preexisting atlas of gene expression. This anal. identified glycine consumption and expression of the mitochondrial glycine biosynthetic pathway as strongly correlated with rates of proliferation across cancer cells. Antagonizing glycine uptake and its mitochondrial biosynthesis preferentially impaired rapidly proliferating cells. Moreover, higher expression of this pathway was assocd. with greater mortality in breast cancer patients. Increased reliance on glycine may represent a metabolic vulnerability for selectively targeting rapid cancer cell proliferation.
  6. 6
    Ivanisevic, J.; Zhu, Z.-J.; Plate, L.; Tautenhahn, R.; Chen, S.; O’Brien, P. J.; Johnson, C. H.; Marletta, M. A.; Patti, G. J.; Siuzdak, G. Anal. Chem. 2013, 85, 6876– 6884
    Google Scholar
    6
    Toward 'Omic Scale Metabolite Profiling: A Dual Separation-Mass Spectrometry Approach for Coverage of Lipid and Central Carbon Metabolism
    Ivanisevic, Julijana; Zhu, Zheng-Jiang; Plate, Lars; Tautenhahn, Ralf; Chen, Stephen; O'Brien, Peter J.; Johnson, Caroline H.; Marletta, Michael A.; Patti, Gary J.; Siuzdak, Gary
    Analytical Chemistry (Washington, DC, United States) (2013), 85 (14), 6876-6884CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
    Although the objective of any omic science is broad measurement of its constituents, such coverage has been challenging in metabolomics because the metabolome is comprised of a chem. diverse set of small mols. with variable phys. properties. While extensive studies have been performed to identify metabolite isolation and sepn. methods, these strategies introduce bias toward lipophilic or water-sol. metabolites depending on whether reversed-phase (RP) or hydrophilic interaction liq. chromatog. (HILIC) was used, resp. Here the authors extend the authors' consideration of metabolome isolation and sepn. procedures to integrate RPLC/MS and HILIC/MS profiling. An aminopropyl-based HILIC/MS method was optimized on the basis of mobile-phase additives and pH, followed by evaluation of reproducibility. When applied to the untargeted study of perturbed bacterial metabolomes, the HILIC method enabled the accurate assessment of key, dysregulated metabolites in central carbon pathways (e.g., amino acids, org. acids, phosphorylated sugars, energy currency metabolites), which could not be retained by RPLC. To demonstrate the value of the integrative approach, bacterial cells, human plasma, and cancer cells were analyzed by combined RPLC/HILIC sepn. coupled to ESI pos./neg. MS detection. The combined approach resulted in the observation of metabolites assocd. with lipid and central carbon metab. from a single biol. ext., using 80% org. solvent (ACN:MeOH:H2O 2:2:1). It enabled the detection of >30,000 features from each sample type, with the highest no. of uniquely detected features by RPLC in ESI pos. mode and by HILIC in ESI neg. mode. Therefore, when time and sample are limited, the max. amt. of biol. information related to lipid and central carbon metab. can be acquired by combining RPLC ESI pos. and HILIC ESI neg. mode anal.
  7. 7
    Bajad, S. U.; Lu, W.; Kimball, E. H.; Yuan, J.; Peterson, C.; Rabinowitz, J. D. J. Chromatogr., A 2006, 1125, 76– 88
    Google Scholar
    7
    Separation and quantitation of water soluble cellular metabolites by hydrophilic interaction chromatography-tandem mass spectrometry
    Bajad, Sunil U.; Lu, Wenyun; Kimball, Elizabeth H.; Yuan, Jie; Peterson, Celeste; Rabinowitz, Joshua D.
    Journal of Chromatography A (2006), 1125 (1), 76-88CODEN: JCRAEY; ISSN:0021-9673. (Elsevier B.V.)
    A key unmet need in metabolomics is the ability to efficiently quantify a large no. of known cellular metabolites. Here the authors present a liq. chromatog. (LC)-electrospray ionization tandem mass spectrometry (ESI-MS/MS) method for reliable measurement of 141 metabolites, including components of central carbon, amino acid, and nucleotide metab. The selected LC approach, hydrophilic interaction chromatog. with an amino column, effectively separates highly water sol. metabolites that fail to retain using std. reversed-phase chromatog. MS/MS detection is achieved by scanning through numerous selected reaction monitoring events on a triple quadrupole instrument. When applied to exts. of Escherichia coli grown in [12C]- vs. [13C]glucose, the method reveals appropriate 12C- and 13C-peaks for 79 different metabolites.
  8. 8
    Lu, W.; Bennett, B. D.; Rabinowitz, J. D. J. Chromatogr., B 2008, 871, 236– 242
    Google Scholar
    8
    Analytical strategies for LC-MS-based targeted metabolomics
    Lu, Wenyun; Bennett, Bryson D.; Rabinowitz, Joshua D.
    Journal of Chromatography B: Analytical Technologies in the Biomedical and Life Sciences (2008), 871 (2), 236-242CODEN: JCBAAI; ISSN:1570-0232. (Elsevier B.V.)
    A review. Recent advances in mass spectrometry are enabling improved anal. of endogenous metabolites. Here the authors discuss several issues relevant to developing liq. chromatog.-electrospray ionization-mass spectrometry methods for targeted metabolomics (i.e., quant. anal. of dozens to hundreds of specific metabolites). Sample prepn. and liq. chromatog. approaches are discussed, with an eye towards the challenge of dealing with a diversity of metabolite classes in parallel. Evidence is presented that heated electrospray ionization (ESI) generally gives improved signal compared to the more traditional unheated ESI. Applicability to targeted metabolomics of triple quadrupole mass spectrometry operating in multiple reaction monitoring (MRM) mode and high mass resoln. full scan mass spectrometry (e.g., time-of-flight, Orbitrap) are described. The authors suggest that both are viable solns., with MRM preferred when targeting a more limited no. of analytes, and full scan preferred for its potential ability to bridge targeted and untargeted metabolomics.
  9. 9
    Masson, P.; Alves, A. C.; Ebbels, T. M. D.; Nicholson, J. K.; Want, E. J. Anal. Chem. 2010, 82, 7779– 7786
    Google Scholar
    There is no corresponding record for this reference.
  10. 10
    Zelena, E.; Dunn, W. B.; Broadhurst, D.; Francis-McIntyre, S.; Carroll, K. M.; Begley, P.; O’Hagan, S.; Knowles, J. D.; Halsall, A.; HUSERMET Consortium; Wilson, I. D.; Kell, D. B. Anal. Chem. 2009, 81, 1357– 1364
    Google Scholar
    10
    Development of a Robust and Repeatable UPLC-MS Method for the Long-Term Metabolomic Study of Human Serum
    Zelena, Eva; Dunn, Warwick B.; Broadhurst, David; Francis-McIntyre, Sue; Carroll, Kathleen M.; Begley, Paul; O'Hagan, Steve; Knowles, Joshua D.; Halsall, Antony; Wilson, Ian D.; Kell, Douglas B.
    Analytical Chemistry (Washington, DC, United States) (2009), 81 (4), 1357-1364CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
    A method for performing untargeted metabolomic anal. of human serum has been developed based on protein pptn. followed by Ultra Performance Liq. Chromatog. and Time-of-Flight mass spectrometry (UPLC-TOF-MS). This method was specifically designed to fulfill the requirements of a long-term metabolomic study, spanning more than 3 years, and it was subsequently thoroughly evaluated for robustness and repeatability. The authors describe here the obsd. drift in instrumental performance over time and its improvement with adjustment of the length of anal. block. The optimal setup for the authors' purpose was further validated against a set of serum samples from 30 healthy individuals. The authors also assessed the reproducibility of chromatog. columns with the same chem. of stationary phase from the same manufacturer but from different prodn. batches. The results have allowed the authors to prep. std. operating procedures for "fit for purpose" long-term UPLC-MS metabolomic studies, such as are being employed in the HUSERMET (Human Serum Metabolome) project. This method allows the acquisition of data and subsequent comparison of data collected across many months or years.
  11. 11
    Geier, F. M.; Want, E. J.; Leroi, A. M.; Bundy, J. G. Anal. Chem. 2011, 83, 3730– 3736
    Google Scholar
    There is no corresponding record for this reference.
  12. 12
    Yanes, O.; Tautenhahn, R.; Patti, G. J.; Siuzdak, G. Anal. Chem. 2011, 83, 2152– 2161
    Google Scholar
    There is no corresponding record for this reference.
  13. 13
    Nordström, A.; Want, E.; Northen, T.; Lehtiö, J.; Siuzdak, G. Anal. Chem. 2008, 80, 421– 429
    Google Scholar
    There is no corresponding record for this reference.
  14. 14
    Buescher, J. M.; Moco, S.; Sauer, U.; Zamboni, N. Anal. Chem. 2010, 82, 4403– 4412
    Google Scholar
    There is no corresponding record for this reference.
  15. 15
    Lu, W.; Clasquin, M. F.; Melamud, E.; Amador-Noguez, D.; Caudy, A. A.; Rabinowitz, J. D. Anal. Chem. 2010, 82, 3212– 3221
    Google Scholar
    15
    Metabolomic Analysis via Reversed-Phase Ion-Pairing Liquid Chromatography Coupled to a Stand Alone Orbitrap Mass Spectrometer
    Lu, Wenyun; Clasquin, Michelle F.; Melamud, Eugene; Amador-Noguez, Daniel; Caudy, Amy A.; Rabinowitz, Joshua D.
    Analytical Chemistry (Washington, DC, United States) (2010), 82 (8), 3212-3221CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
    The authors present a liq. chromatog.-mass spectrometry (LC-MS) method that capitalizes on the mass-resolving power of the Orbitrap to enable sensitive and specific measurement of known and unanticipated metabolites in parallel, with a focus on water-sol. species involved in core metab. The reversed phase LC method, with a cycle time 25 min, involves a water-methanol gradient on a C18 column with tributylamine as the ion pairing agent. The MS portion involves full scans from 85 to 1000 m/z at 1 Hz and 100,000 resoln. in neg. ion mode on a stand alone Orbitrap ("Exactive"). The median limit of detection, across 80 metabolite stds., was 5 ng/mL with the linear range typically ≥100-fold. For both stds. and a cellular ext. from Saccharomyces cerevisiae (Baker's yeast), the median inter-run relative std. deviation in peak intensity was 8%. In yeast ext., the authors detected 137 known compds., whose 13C-labeling patterns could also be tracked to probe metabolic flux. In yeast engineered to lack a gene of unknown function (YKL215C), the authors obsd. accumulation of an ion of m/z 128.0351, which the authors subsequently confirmed to be oxoproline, resulting in annotation of YKL215C as an oxoprolinase. These examples demonstrate the suitability of the present method for quant. metabolomics, fluxomics, and discovery metabolite profiling.
  16. 16
    Smith, C. A.; Want, E. J.; O’Maille, G.; Abagyan, R.; Siuzdak, G. Anal. Chem. 2006, 78, 779– 787
    Google Scholar
    16
    XCMS: Processing Mass Spectrometry Data for Metabolite Profiling Using Nonlinear Peak Alignment, Matching, and Identification
    Smith, Colin A.; Want, Elizabeth J.; O'Maille, Grace; Abagyan, Ruben; Siuzdak, Gary
    Analytical Chemistry (2006), 78 (3), 779-787CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
    Metabolite profiling in biomarker discovery, enzyme substrate assignment, drug activity/specificity detn., and basic metabolic research requires new data preprocessing approaches to correlate specific metabolites to their biol. origin. Here we introduce an LC/MS-based data anal. approach, XCMS, which incorporates novel nonlinear retention time alignment, matched filtration, peak detection, and peak matching. Without using internal stds., the method dynamically identifies hundreds of endogenous metabolites for use as stds., calcg. a nonlinear retention time correction profile for each sample. Following retention time correction, the relative metabolite ion intensities are directly compared to identify changes in specific endogenous metabolites, such as potential biomarkers. The software is demonstrated using data sets from a previously reported enzyme knockout study and a large-scale study of plasma samples. XCMS is freely available under an open-source license at http://metlin.scripps.edu/download/.
  17. 17
    Chokkathukalam, A.; Jankevics, A.; Creek, D. J.; Achcar, F.; Barrett, M. P.; Breitling, R. Bioinformatics 2013, 29, 281– 283
    Google Scholar
    There is no corresponding record for this reference.
  18. 18
    Mishur, R. J.; Rea, S. L. Mass Spectrom. Rev. 2012, 31, 70– 95
    Google Scholar
    18
    Applications of mass spectrometry to metabolomics and metabonomics: Detection of biomarkers of aging and of age-related diseases
    Mishur, Robert J.; Rea, Shane L.
    Mass Spectrometry Reviews (2012), 31 (1), 70-95CODEN: MSRVD3; ISSN:0277-7037. (John Wiley & Sons, Inc.)
    A review. Every 5 years or so new technologies, or new combinations of old ones, seemingly burst onto the science scene and are then sought after until they reach the point of becoming commonplace. Advances in mass spectrometry instrumentation, coupled with the establishment of standardized chem. fragmentation libraries, increased computing power, novel data-anal. algorithms, new scientific applications, and com. prospects have made mass spectrometry-based metabolomics the latest sought-after technol. This methodol. affords the ability to dynamically catalog and quantify, in parallel, femtomole quantities of cellular metabolites. The study of aging, and the diseases that accompany it, has accelerated significantly in the last decade. Mutant genes that alter the rate of aging have been found that increase lifespan by up to 10-fold in some model organisms, and substantial progress has been made in understanding fundamental alterations that occur at both the mRNA and protein level in tissues of aging organisms. The application of metabolomics to aging research is still relatively new, but has already added significant insight into the aging process. In this review we summarize these findings. We have targeted our manuscript to two audiences: mass spectrometrists interested in applying their tech. knowledge to unanswered questions in the aging field, and gerontologists interested in expanding their knowledge of both mass spectrometry and the most recent advances in aging-related metabolomics. © 2011 Wiley Periodicals, Inc., Mass Spec Rev 31:70-95, 2012.
  19. 19
    Brown, M.; Dunn, W. B.; Dobson, P.; Patel, Y.; Winder, C. L.; Francis-McIntyre, S.; Begley, P.; Carroll, K.; Broadhurst, D.; Tseng, A.; Swainston, N.; Spasic, I.; Goodacre, R.; Kell, D. B. Analyst 2009, 134, 1322– 1332
    Google Scholar
    19
    Mass spectrometry tools and metabolite-specific databases for molecular identification in metabolomics
    Brown, M.; Dunn, W. B.; Dobson, P.; Patel, Y.; Winder, C. L.; Francis-McIntyre, S.; Begley, P.; Carroll, K.; Broadhurst, D.; Tseng, A.; Swainston, N.; Spasic, I.; Goodacre, R.; Kell, D. B.
    Analyst (Cambridge, United Kingdom) (2009), 134 (7), 1322-1332CODEN: ANALAO; ISSN:0003-2654. (Royal Society of Chemistry)
    The chem. identification of mass spectrometric signals in metabolomic applications is important to provide conversion of anal. data to biol. knowledge about metabolic pathways. The complexity of electrospray mass spectrometric data acquired from a range of samples (serum, urine, yeast intracellular exts., yeast metabolic footprints, placental tissue metabolic footprints) has been investigated and has defined the frequency of different ion types routinely detected. Although some ion types were expected (protonated and deprotonated peaks, isotope peaks, multiply charged peaks) others were not expected (sodium formate adduct ions). In parallel, the Manchester Metabolomics Database (MMD) has been constructed with data from genome scale metabolic reconstructions, HMDB, KEGG, Lipid Maps, BioCyc and DrugBank to provide knowledge on 42,687 endogenous and exogenous metabolite species. The combination of accurate mass data for a large collection of metabolites, theor. isotope abundance data and knowledge of the different ion types detected provided a greater no. of electrospray mass spectrometric signals which were putatively identified and with greater confidence in the samples studied. To provide definitive identification metabolite-specific mass spectral libraries for UPLC-MS and GC-MS have been constructed for 1,065 com. available authentic stds. The MMD data are available at http://dbkgroup.org/MMD.
  20. 20
    Kuhl, C.; Tautenhahn, R.; Böttcher, C.; Larson, T. R.; Neumann, S. Anal. Chem. 2012, 84, 283– 289
    Google Scholar
    20
    CAMERA: An Integrated Strategy for Compound Spectra Extraction and Annotation of Liquid Chromatography/Mass Spectrometry Data Sets
    Kuhl, Carsten; Tautenhahn, Ralf; Boettcher, Christoph; Larson, Tony R.; Neumann, Steffen
    Analytical Chemistry (Washington, DC, United States) (2012), 84 (1), 283-289CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
    Liq. chromatog. coupled to mass spectrometry is routinely used for metabolomics expts. In contrast to the fairly routine and automated data acquisition steps, subsequent compd. annotation and identification require extensive manual anal. and thus form a major bottleneck in data interpretation. Here the authors present CAMERA, a Bioconductor package integrating algorithms to ext. compd. spectra, annotate isotope and adduct peaks, and propose the accurate compd. mass even in highly complex data. To evaluate the algorithms, the authors compared the annotation of CAMERA against a manually defined annotation for a mixt. of known compds. spiked into a complex matrix at different concns. CAMERA successfully extd. accurate masses for 89.7% and 90.3% of the annotatable compds. in pos. and neg. ion modes, resp. Furthermore, the authors present a novel annotation approach that combines spectral information of data acquired in opposite ion modes to further improve the annotation rate. The authors demonstrate the utility of CAMERA in two different, easily adoptable plant metabolomics expts., where the application of CAMERA drastically reduced the amt. of manual anal.
  21. 21
    Alonso, A.; Julià, A.; Beltran, A.; Vinaixa, M.; Díaz, M.; Ibañez, L.; Correig, X.; Marsal, S. Bioinformatics 2011, 27, 1339– 1340
    Google Scholar
    21
    AStream: an R package for annotating LC/MS metabolomic data
    Alonso, Arnald; Julia, Antonio; Beltran, Antoni; Vinaixa, Maria; Diaz, Marta; Ibanez, Lourdes; Correig, Xavier; Marsal, Sara
    Bioinformatics (2011), 27 (9), 1339-1340CODEN: BOINFP; ISSN:1367-4803. (Oxford University Press)
    AStream, an R-statistical software package for the curation and identification of feature peaks extd. from liq. chromatog. mass spectrometry (LC/MS) metabolomics data, is described. AStream detects isotopic, fragment and adduct patterns by identifying feature pairs that fulfill expected relational patterns. Data redn. by AStream allows compds. to be identified reliably and subsequently linked to metabolite databases. AStream provides researchers with a fast, reliable tool for summarizing metabolomic data, notably reducing curation time and increasing consistency of results.
  22. 22
    R Core Team. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, 2014; http://www.R-project.org.
    Google Scholar
    There is no corresponding record for this reference.
  23. 23
    Smith, C. A.; Maille, G.; Want, E. J.; Qin, C.; Trauger, S. A.; Brandon, T. R.; Custodio, D. E.; Abagyan, R.; Siuzdak, G. Ther. Drug Monit. 2005, 27, 747– 751
    Google Scholar
    23
    METLIN. A metabolite mass spectral database
    Smith, Colin A.; O'Maille, Grace; Want, Elizabeth J.; Qin, Chuan; Trauger, Sunia A.; Brandon, Theodore R.; Custodio, Darlene E.; Abagyan, Ruben; Siuzdak, Gary
    Therapeutic Drug Monitoring (2005), 27 (6), 747-751CODEN: TDMODV; ISSN:0163-4356. (Lippincott Williams & Wilkins)
    Endogenous metabolites have gained increasing interest over the past 5 years largely for their implications in diagnostic and pharmaceutical biomarker discovery. METLIN (http://metlin.scripps.edu), a freely accessible web-based data repository, has been developed to assist in a broad array of metabolite research and to facilitate metabolite identification through mass anal. METLIN includes an annotated list of known metabolite structural information that is easily cross-correlated with its catalog of high-resoln. Fourier transform mass spectrometry (FTMS) spectra, tandem mass spectrometry (MS/MS) spectra, and LC/MS data.
  24. 24
    Guo, A. C.; Jewison, T.; Wilson, M.; Liu, Y.; Knox, C.; Djoumbou, Y.; Lo, P.; Mandal, R.; Krishnamurthy, R.; Wishart, D. S. Nucleic Acids Res. 2013, 41, D625– D630
    Google Scholar
    There is no corresponding record for this reference.
  25. 25
    Ong, S.-E.; Blagoev, B.; Kratchmarova, I.; Kristensen, D. B.; Steen, H.; Pandey, A.; Mann, M. Mol. Cell. Proteomics 2002, 1, 376– 386
    Google Scholar
    25
    Stable isotope labeling by amino acids in cell culture, SILAC, as a simple and accurate approach to expression proteomics
    Ong, Shao-En; Blagoev, Blagoy; Kratchmarova, Irina; Kristensen, Dan Bach; Steen, Hanno; Pandey, Akhilesh; Mann, Matthias
    Molecular and Cellular Proteomics (2002), 1 (5), 376-386CODEN: MCPOBS; ISSN:1535-9476. (American Society for Biochemistry and Molecular Biology, Inc.)
    Quant. proteomics has traditionally been performed by two-dimensional gel electrophoresis, but recently, mass spectrometric methods based on stable isotope quantitation have shown great promise for the simultaneous and automated identification and quantitation of complex protein mixts. Here we describe a method, termed SILAC, for stable isotope labeling by amino acids in cell culture, for the in vivo incorporation of specific amino acids into all mammalian proteins. Mammalian cell lines are grown in media lacking a std. essential amino acid but supplemented with a non-radioactive, isotopically labeled form of that amino acid, in this case deuterated leucine (Leu-d3). We find that growth of cells maintained in these media is no different from growth in normal media as evidenced by cell morphol., doubling time, and ability to differentiate. Complete incorporation of Leu-d3 occurred after five doublings in the cell lines and proteins studied. Protein populations from exptl. and control samples are mixed directly after harvesting, and mass spectrometric identification is straightforward as every leucine-contg. peptide incorporates either all normal leucine or all Leu-d3. We have applied this technique to the relative quantitation of changes in protein expression during the process of muscle cell differentiation. Proteins that were found to be up-regulated during this process include glyceraldehyde-3-phosphate dehydrogenase, fibronectin, and pyruvate kinase M2. SILAC is a simple, inexpensive, and accurate procedure that can be used as a quant. proteomic approach in any cell culture system.
  26. 26
    Wiese, S.; Reidegeld, K. A.; Meyer, H. E.; Warscheid, B. Proteomics 2007, 7, 340– 350
    Google Scholar
    26
    Protein labeling by iTRAQ: a new tool for quantitative mass spectrometry in proteome research
    Wiese, Sebastian; Reidegeld, Kai A.; Meyer, Helmut E.; Warscheid, Bettina
    Proteomics (2007), 7 (3), 340-350CODEN: PROTC7; ISSN:1615-9853. (Wiley-VCH Verlag GmbH & Co. KGaA)
    A novel, MS-based approach for the relative quantification of proteins, relying on the derivatization of primary amino groups in intact proteins using isobaric tag for relative and abs. quantitation (iTRAQ) is presented. Due to the isobaric mass design of the iTRAQ reagents, differentially labeled proteins do not differ in mass; accordingly, their corresponding proteolytic peptides appear as single peaks in MS scans. Because quant. information is provided by isotope-encoded reporter ions that can only be obsd. in MS/MS spectra, the authors analyzed the fragmentation behavior of ESI and MALDI ions of peptides generated from iTRAQ-labeled proteins using a TOF/TOF and/or a QTOF instrument. The authors obsd. efficient liberation of reporter ions for singly protonated peptides at low-energy collision conditions. In contrast, increased collision energies were required to liberate the iTRAQ label from lysine side chains of doubly charged peptides and, thus, to observe reporter ions suitable for relative quantification of proteins with high accuracy. The authors then developed a quant. strategy that comprises labeling of intact proteins by iTRAQ followed by gel electrophoresis and peptide MS/MS analyses. As proof of principle, mixts. of five different proteins in various concn. ratios were quantified, demonstrating the general applicability of the approach presented here to quant. MS-based proteomics.
  27. 27
    Hebert, A. S.; Merrill, A. E.; Bailey, D. J.; Still, A. J.; Westphall, M. S.; Strieter, E. R.; Pagliarini, D. J.; Coon, J. J. Nat. Methods 2013, 10, 332– 334
    Google Scholar
    27
    Neutron-encoded mass signatures for multiplexed proteome quantification
    Hebert, Alexander S.; Merrill, Anna E.; Bailey, Derek J.; Still, Amelia J.; Westphall, Michael S.; Strieter, Eric R.; Pagliarini, David J.; Coon, Joshua J.
    Nature Methods (2013), 10 (4), 332-334CODEN: NMAEA3; ISSN:1548-7091. (Nature Publishing Group)
    We describe a protein quantification method called neutron encoding that exploits the subtle mass differences caused by nuclear binding energy variation in stable isotopes. These mass differences are synthetically encoded into amino acids and incorporated into yeast and mouse proteins via metabolic labeling. Mass spectrometry anal. with high mass resoln. (>200,000) reveals the isotopologue-embedded peptide signals, permitting quantification. Neutron encoding will enable highly multiplexed proteome anal. with excellent dynamic range and accuracy.
  28. 28
    Birkemeyer, C.; Luedemann, A.; Wagner, C.; Erban, A.; Kopka, J. Trends Biotechnol. 2005, 23, 28– 33
    Google Scholar
    There is no corresponding record for this reference.
  29. 29
    Mashego, M. R.; Wu, L.; Van Dam, J. C.; Ras, C.; Vinke, J. L.; Van Winden, W. A.; Van Gulik, W. M.; Heijnen, J. J. Biotechnol. Bioeng. 2004, 85, 620– 628
    Google Scholar
    There is no corresponding record for this reference.
  30. 30
    De Jong, F. A.; Beecher, C. Bioanalysis 2012, 4, 2303– 2314
    Google Scholar
    30
    Addressing the current bottlenecks of metabolomics: Isotopic Ratio Outlier Analysis®, an isotopic-labeling technique for accurate biochemical profiling
    de Jong, Felice A.; Beecher, Chris
    Bioanalysis (2012), 4 (18), 2303-2314CODEN: BIOAB4; ISSN:1757-6180. (Future Science Ltd.)
    A review. Metabolomics or biochem. profiling is a fast emerging science; however, there are still many assocd. bottlenecks to overcome before measurements will be considered robust. Advances in MS resoln. and sensitivity, ultra pressure LC-MS, ESI, and isotopic approaches such as flux anal. and stable-isotope diln., have made it easier to quantitate biochems. The digitization of mass spectrometers has simplified informatic aspects. However, issues of anal. variability, ion suppression and metabolite identification still plague metabolomics investigators. These hurdles need to be overcome for accurate metabolite quantitation not only for in vitro systems, but for complex matrixes such as biofluids and tissues, before it is possible to routinely identify biomarkers that are assocd. with the early prediction and diagnosis of diseases. In this report, we describe a novel isotopic-labeling method that uses the creation of distinct biochem. signatures to eliminate current bottlenecks and enable accurate metabolic profiling.
  31. 31
    Stupp, G. S.; Clendinen, C. S.; Ajredini, R.; Szewc, M. A.; Garrett, T.; Menger, R. F.; Yost, R. A.; Beecher, C.; Edison, A. S. Anal. Chem. 2013, 85, 11858– 11865
    Google Scholar
    There is no corresponding record for this reference.

Cited By

Click to copy section linkSection link copied!
Citation Statements
Explore this article's citation statements on scite.ai
powered by  Scite

This article is cited by 85 publications.

  1. Boryana Petrova, Arzu Tugce Guler. Recent Developments in Single-Cell Metabolomics by Mass Spectrometry─A Perspective. Journal of Proteome Research 2025, 24 (4) , 1493-1518. https://doi.org/10.1021/acs.jproteome.4c00646
  2. Tashi C. E. Liebergesell, Ethan G. Murdock, Aaron W. Puri. Detection of Inverse Stable Isotopic Labeling in Untargeted Metabolomic Data. Analytical Chemistry 2024, 96 (41) , 16330-16337. https://doi.org/10.1021/acs.analchem.4c03528
  3. Brady G. Anderson, Pavlo Popov, Amanda R. Cicali, Adanna Nwamba, Charles R. Evans, Robert T. Kennedy. In-Depth Chemical Analysis of the Brain Extracellular Space Using In Vivo Microdialysis with Liquid Chromatography-Tandem Mass Spectrometry. Analytical Chemistry 2024, 96 (41) , 16387-16396. https://doi.org/10.1021/acs.analchem.4c03806
  4. Julius Agongo, Scott F. Grady, Kevin Cho, Gary J. Patti, Benjamin J. Bythell, Christopher K. Arnatt, James L. Edwards. Discovery and Identification of Three Homocysteine Metabolites by Chemical Derivatization and Mass Spectrometry Fragmentation. Analytical Chemistry 2024, 96 (29) , 11639-11643. https://doi.org/10.1021/acs.analchem.4c01706
  5. Brady G. Anderson, Alexander Raskind, Rylan Hissong, Michael K. Dougherty, Sarah K. McGill, Ajay S. Gulati, Casey M. Theriot, Robert T. Kennedy, Charles R. Evans. Offline Two-Dimensional Liquid Chromatography–Mass Spectrometry for Deep Annotation of the Fecal Metabolome Following Fecal Microbiota Transplantation. Journal of Proteome Research 2024, 23 (6) , 2000-2012. https://doi.org/10.1021/acs.jproteome.4c00022
  6. Yiguang Qian, Ziyu Chen, Jiahui Wang, Man Peng, Shenghua Zhang, Xiaoyu Yan, Xiaole Han, Xiaohui Ou, Jie Sun, Siyue Li, Ke Chen. H/D Exchange Coupled with 2H-labeled Stable Isotope-Assisted Metabolomics Discover Transformation Products of Contaminants of Emerging Concern. Analytical Chemistry 2023, 95 (33) , 12541-12549. https://doi.org/10.1021/acs.analchem.3c02833
  7. Ethan Stancliffe, Gary J. Patti. PeakDetective: A Semisupervised Deep Learning-Based Approach for Peak Curation in Untargeted Metabolomics. Analytical Chemistry 2023, 95 (25) , 9397-9403. https://doi.org/10.1021/acs.analchem.3c00764
  8. Ke Chen, Yuhui Xiang, Xiaoyu Yan, Zhenghui Li, Rui Qin, Jie Sun. Global Tracking of Transformation Products of Environmental Contaminants by 2H-Labeled Stable Isotope-Assisted Metabolomics. Analytical Chemistry 2022, 94 (20) , 7255-7263. https://doi.org/10.1021/acs.analchem.2c00500
  9. Ritchie Ly, Nicholas Ly, Kazunori Sasaki, Makoto Suzuki, Kenjiro Kami, Yoshiaki Ohashi, Philip Britz-McKibbin. Nontargeted Serum Lipid Profiling of Nonalcoholic Steatohepatitis by Multisegment Injection–Nonaqueous Capillary Electrophoresis–Mass Spectrometry: A Multiplexed Separation Platform for Resolving Ionic Lipids. Journal of Proteome Research 2022, 21 (3) , 768-777. https://doi.org/10.1021/acs.jproteome.1c00682
  10. James N. Dodds, Lingjue Wang, Gary J. Patti, Erin S. Baker. Combining Isotopologue Workflows and Simultaneous Multidimensional Separations to Detect, Identify, and Validate Metabolites in Untargeted Analyses. Analytical Chemistry 2022, 94 (5) , 2527-2535. https://doi.org/10.1021/acs.analchem.1c04430
  11. Kristina Tötsch, John C. Fjeldsted, Sarah M. Stow, Oliver J. Schmitz, Sven W. Meckelmann. Effect of Sampling Rate and Data Pretreatment for Targeted and Nontargeted Analysis by Means of Liquid Chromatography Coupled to Drift Time Ion Mobility Quadruple Time-of-Flight Mass Spectrometry. Journal of the American Society for Mass Spectrometry 2021, 32 (10) , 2592-2603. https://doi.org/10.1021/jasms.1c00217
  12. Melanie T. Odenkirk, David M. Reif, Erin S. Baker. Multiomic Big Data Analysis Challenges: Increasing Confidence in the Interpretation of Artificial Intelligence Assessments. Analytical Chemistry 2021, 93 (22) , 7763-7773. https://doi.org/10.1021/acs.analchem.0c04850
  13. James N. Dodds, Nancy Lee M. Alexander, Kaylie I. Kirkwood, MaKayla R. Foster, Zachary R. Hopkins, Detlef R. U. Knappe, Erin S. Baker. From Pesticides to Per- and Polyfluoroalkyl Substances: An Evaluation of Recent Targeted and Untargeted Mass Spectrometry Methods for Xenobiotics. Analytical Chemistry 2021, 93 (1) , 641-656. https://doi.org/10.1021/acs.analchem.0c04359
  14. Taylor M. Domenick, Emily L. Gill, Vinata Vedam-Mai, Richard A. Yost. Mass Spectrometry-Based Cellular Metabolomics: Current Approaches, Applications, and Future Directions. Analytical Chemistry 2021, 93 (1) , 546-566. https://doi.org/10.1021/acs.analchem.0c04363
  15. Evelyn Rampler, Yasin El Abiead, Harald Schoeny, Mate Rusz, Felina Hildebrand, Veronika Fitz, Gunda Koellensperger. Recurrent Topics in Mass Spectrometry-Based Metabolomics and Lipidomics—Standardization, Coverage, and Throughput. Analytical Chemistry 2021, 93 (1) , 519-545. https://doi.org/10.1021/acs.analchem.0c04698
  16. Wenyun Lu, Xi Xing, Lin Wang, Li Chen, Sisi Zhang, Melanie R. McReynolds, Joshua D. Rabinowitz. Improved Annotation of Untargeted Metabolomics Data through Buffer Modifications That Shift Adduct Mass and Intensity. Analytical Chemistry 2020, 92 (17) , 11573-11581. https://doi.org/10.1021/acs.analchem.0c00985
  17. Shanshan Guan, Michael R. Armbruster, Tianjiao Huang, James L. Edwards, Benjamin J. Bythell. Isomeric Differentiation and Acidic Metabolite Identification by Piperidine-Based Tagging, LC–MS/MS, and Understanding of the Dissociation Chemistries. Analytical Chemistry 2020, 92 (13) , 9305-9311. https://doi.org/10.1021/acs.analchem.0c01640
  18. Miriam Sindelar, Gary J. Patti. Chemical Discovery in the Era of Metabolomics. Journal of the American Chemical Society 2020, 142 (20) , 9097-9105. https://doi.org/10.1021/jacs.9b13198
  19. Erin S. Baker, Gary J. Patti. Perspectives on Data Analysis in Metabolomics: Points of Agreement and Disagreement from the 2018 ASMS Fall Workshop. Journal of the American Society for Mass Spectrometry 2019, 30 (10) , 2031-2036. https://doi.org/10.1007/s13361-019-02295-3
  20. Edward D. Kantz, Saumya Tiwari, Jeramie D. Watrous, Susan Cheng, Mohit Jain. Deep Neural Networks for Classification of LC-MS Spectral Peaks. Analytical Chemistry 2019, 91 (19) , 12407-12413. https://doi.org/10.1021/acs.analchem.9b02983
  21. Jonathan L. Spalding, Fuad J. Naser, Nathaniel G. Mahieu, Stephen L. Johnson, Gary J. Patti. Trace Phosphate Improves ZIC-pHILIC Peak Shape, Sensitivity, and Coverage for Untargeted Metabolomics. Journal of Proteome Research 2018, 17 (10) , 3537-3546. https://doi.org/10.1021/acs.jproteome.8b00487
  22. Nathaniel G. Mahieu and Gary J. Patti . Systems-Level Annotation of a Metabolomics Data Set Reduces 25 000 Features to Fewer than 1000 Unique Metabolites. Analytical Chemistry 2017, 89 (19) , 10397-10406. https://doi.org/10.1021/acs.analchem.7b02380
  23. Nathaniel G. Mahieu, Jonathan L. Spalding, Susan J. Gelman, and Gary J. Patti . Defining and Detecting Complex Peak Relationships in Mass Spectral Data: The Mz.unity Algorithm. Analytical Chemistry 2016, 88 (18) , 9037-9046. https://doi.org/10.1021/acs.analchem.6b01702
  24. Zhendong Li, Jaspaul Tatlay, and Liang Li . Nanoflow LC–MS for High-Performance Chemical Isotope Labeling Quantitative Metabolomics. Analytical Chemistry 2015, 87 (22) , 11468-11474. https://doi.org/10.1021/acs.analchem.5b03209
  25. H. Paul Benton, Julijana Ivanisevic, Nathaniel G. Mahieu, Michael E. Kurczy, Caroline H. Johnson, Lauren Franco, Duane Rinehart, Elizabeth Valentine, Harsha Gowda, Baljit K. Ubhi, Ralf Tautenhahn, Andrew Gieschen, Matthew W. Fields, Gary J. Patti, and Gary Siuzdak . Autonomous Metabolomics for Rapid Metabolite Identification in Global Profiling. Analytical Chemistry 2015, 87 (2) , 884-891. https://doi.org/10.1021/ac5025649
  26. Caroline H. Johnson, Julijana Ivanisevic, H. Paul Benton, and Gary Siuzdak . Bioinformatics: The Next Frontier of Metabolomics. Analytical Chemistry 2015, 87 (1) , 147-156. https://doi.org/10.1021/ac5040693
  27. Erick Helmeczi, Haley Pandya, Katie O’Hearn, Dayre McNally, Philip Britz-McKibbin. Treatment response variations to a single large bolus of enteral cholecalciferol in vitamin D deficient critically Ill children: Metabolomic insights for precision nutrition. The Journal of Steroid Biochemistry and Molecular Biology 2025, 250 , 106720. https://doi.org/10.1016/j.jsbmb.2025.106720
  28. Rodolfo A. Salido, Haoqi Nina Zhao, Daniel McDonald, Helena Mannochio-Russo, Simone Zuffa, Renee E. Oles, Allegra T. Aron, Yasin El Abiead, Sawyer Farmer, Antonio González, Cameron Martino, Ipsita Mohanty, Ceth W. Parker, Lucas Patel, Paulo Wender Portal Gomes, Robin Schmid, Tara Schwartz, Jennifer Zhu, Michael R. Barratt, Kathleen H. Rubins, Hiutung Chu, Fathi Karouia, Kasthuri Venkateswaran, Pieter C. Dorrestein, Rob Knight. The International Space Station has a unique and extreme microbial and chemical environment driven by use patterns. Cell 2025, 188 (7) , 2022-2041.e23. https://doi.org/10.1016/j.cell.2025.01.039
  29. Yonghui Dong, Liron Feldberg, Asaph Aharoni, Uwe Heinig. Metabolite Annotation through Stable Isotope Labeling. TrAC Trends in Analytical Chemistry 2024, 181 , 118037. https://doi.org/10.1016/j.trac.2024.118037
  30. Xiaojie Liu, Ting Peng, Miaochun Xu, Shitong Lin, Bai Hu, Tian Chu, Binghan Liu, Yashi Xu, Wencheng Ding, Li Li, Canhui Cao, Peng Wu. Spatial multi-omics: deciphering technological landscape of integration of multi-omics and its applications. Journal of Hematology & Oncology 2024, 17 (1) https://doi.org/10.1186/s13045-024-01596-9
  31. Selina Hemmer, Sascha K. Manier, Lea Wagmann, Markus R. Meyer. Comparison of reversed-phase, hydrophilic interaction, and porous graphitic carbon chromatography columns for an untargeted toxicometabolomics study in pooled human liver microsomes, rat urine, and rat plasma. Metabolomics 2024, 20 (3) https://doi.org/10.1007/s11306-024-02115-0
  32. Selina Hemmer, Sascha K. Manier, Lea Wagmann, Markus R. Meyer. Impact of four different extraction methods and three different reconstitution solvents on the untargeted metabolomics analysis of human and rat urine samples. Journal of Chromatography A 2024, 1725 , 464930. https://doi.org/10.1016/j.chroma.2024.464930
  33. Julio César España Amórtegui, Susanne Ekroth, Heidi Pekar, Jairo Arturo Guerrero Dallos. A green-footprint approach for parallel multiclass analysis of contaminants in roasted coffee via LC-HRMS. Analytical and Bioanalytical Chemistry 2024, 416 (7) , 1541-1560. https://doi.org/10.1007/s00216-024-05157-4
  34. Shuzhao Li, Amnah Siddiqa, Maheshwor Thapa, Yuanye Chi, Shujian Zheng. Trackable and scalable LC-MS metabolomics data processing using asari. Nature Communications 2023, 14 (1) https://doi.org/10.1038/s41467-023-39889-1
  35. Philip Schippers, Sari Rasheed, Yu Mi Park, Timo Risch, Lea Wagmann, Selina Hemmer, Sascha K. Manier, Rolf Müller, Jennifer Herrmann, Markus R. Meyer. Evaluation of extraction methods for untargeted metabolomic studies for future applications in zebrafish larvae infection models. Scientific Reports 2023, 13 (1) https://doi.org/10.1038/s41598-023-34593-y
  36. Lorenzo Favilli, Corey M. Griffith, Emma L. Schymanski, Carole L. Linster. High-throughput Saccharomyces cerevisiae cultivation method for credentialing-based untargeted metabolomics. Analytical and Bioanalytical Chemistry 2023, 415 (17) , 3415-3434. https://doi.org/10.1007/s00216-023-04724-5
  37. Sergey Girel, Davy Guillarme, Szabolcs Fekete, Serge Rudaz, Víctor González-Ruiz. Investigation of several chromatographic approaches for untargeted profiling of central carbon metabolism. Journal of Chromatography A 2023, 1697 , 463994. https://doi.org/10.1016/j.chroma.2023.463994
  38. Travis E. Hartman, Hannah Jane Lees. Introduction of Metabolomics: An Overview. 2023, 1-37. https://doi.org/10.1007/978-3-031-39094-4_1
  39. Yasin El Abiead, Christoph Bueschl, Lisa Panzenboeck, Mingxun Wang, Maria Doppler, Bernhard Seidl, Jürgen Zanghellini, Pieter C. Dorrestein, Gunda Koellensperger. Heterogeneous multimeric metabolite ion species observed in LC-MS based metabolomics data sets. Analytica Chimica Acta 2022, 1229 , 340352. https://doi.org/10.1016/j.aca.2022.340352
  40. Bingwen Zhou, Mengke Jia, Fan Zhang, Jin Qi, Boyang Yu. A new data processing strategy combined with a convolutional neural network for rapid and accurate prediction of geographical classifications of natural products. Chemometrics and Intelligent Laboratory Systems 2022, 227 , 104594. https://doi.org/10.1016/j.chemolab.2022.104594
  41. Alisa B. Nelson, Lisa S. Chow, Curtis C. Hughey, Peter A. Crawford, Patrycja Puchalska. Artifactual FA dimers mimic FAHFA signals in untargeted metabolomics pipelines. Journal of Lipid Research 2022, 63 (5) , 100201. https://doi.org/10.1016/j.jlr.2022.100201
  42. Jeremy P Koelmel, Hongyu Xie, Elliott J Price, Elizabeth Z Lin, Katherine E Manz, Paul Stelben, Matthew K Paige, Stefano Papazian, Joseph Okeme, Dean P Jones, Dinesh Barupal, John A Bowden, Pawel Rostkowski, Kurt D Pennell, Vladimir Nikiforov, Thanh Wang, Xin Hu, Yunjia Lai, Gary W Miller, Douglas I Walker, Jonathan W Martin, Krystal J Godri Pollitt. An actionable annotation scoring framework for gas chromatography-high-resolution mass spectrometry. Exposome 2022, 2 (1) https://doi.org/10.1093/exposome/osac007
  43. Morgan B. Harris, Mahbobeh Lesani, Zongyuan Liu, Laura-Isobel McCall. Molecular networking in infectious disease models. 2022, 341-375. https://doi.org/10.1016/bs.mie.2021.09.018
  44. Martin Giera, Oscar Yanes, Gary Siuzdak. Metabolite discovery: Biochemistry’s scientific driver. Cell Metabolism 2022, 34 (1) , 21-34. https://doi.org/10.1016/j.cmet.2021.11.005
  45. Corey M. Griffith, Adhish S. Walvekar, Carole L. Linster. Approaches for completing metabolic networks through metabolite damage and repair discovery. Current Opinion in Systems Biology 2021, 28 , 100379. https://doi.org/10.1016/j.coisb.2021.100379
  46. Li Chen, Wenyun Lu, Lin Wang, Xi Xing, Ziyang Chen, Xin Teng, Xianfeng Zeng, Antonio D. Muscarella, Yihui Shen, Alexis Cowan, Melanie R. McReynolds, Brandon J. Kennedy, Ashley M. Lato, Shawn R. Campagna, Mona Singh, Joshua D. Rabinowitz. Metabolite discovery through global annotation of untargeted metabolomics data. Nature Methods 2021, 18 (11) , 1377-1385. https://doi.org/10.1038/s41592-021-01303-3
  47. Roger Giné, Jordi Capellades, Josep M. Badia, Dennis Vughs, Michaela Schwaiger-Haber, Theodore Alexandrov, Maria Vinaixa, Andrea M. Brunner, Gary J. Patti, Oscar Yanes. HERMES: a molecular-formula-oriented method to target the metabolome. Nature Methods 2021, 18 (11) , 1370-1376. https://doi.org/10.1038/s41592-021-01307-z
  48. Robert A. Quinn, Kehau A. Hagiwara, Ken Liu, Maryam Goudarzi, Wimal Pathmasiri, Lloyd W. Sumner, Thomas O. Metz, . Bridging the Gap between Analytical and Microbial Sciences in Microbiome Research. mSystems 2021, 6 (5) https://doi.org/10.1128/msystems.00585-21
  49. Kelly H. Telu, Ramesh Marupaka, Nirina R. Andriamaharavo, Yamil Simón‐Manso, Yuxue Liang, Yuri A. Mirokhin, Tallat H. Bukhari, Renae J. Preston, Lila Kashi, Zvi Kelman, Stephen E. Stein. Creation and filtering of a recurrent spectral library of CHO cell metabolites and media components. Biotechnology and Bioengineering 2021, 118 (4) , 1491-1510. https://doi.org/10.1002/bit.27661
  50. Brett T. Doherty, Jeremy P. Koelmel, Elizabeth Z. Lin, Megan E. Romano, Krystal J. Godri Pollitt. Use of Exposomic Methods Incorporating Sensors in Environmental Epidemiology. Current Environmental Health Reports 2021, 8 (1) , 34-41. https://doi.org/10.1007/s40572-021-00306-8
  51. Kevin Cho, Michaela Schwaiger-Haber, Fuad J. Naser, Ethan Stancliffe, Miriam Sindelar, Gary J. Patti. Targeting unique biological signals on the fly to improve MS/MS coverage and identification efficiency in metabolomics. Analytica Chimica Acta 2021, 1149 , 338210. https://doi.org/10.1016/j.aca.2021.338210
  52. Sean W. Harshman, Kraig E. Strayer, Christina N. Davidson, Rhonda L. Pitsch, Latha Narayanan, Alexander M. Scott, Nicole M. Schaeublin, Taylor L. Wiens, Mandy S. Phelps, Maegan L. O'Connor, Nicholas S. Mackowski, Kristyn N. Barrett, Samantha M. Leyh, Jason J. Eckerle, Adam J. Strang, Jennifer A. Martin. Rate normalization for sweat metabolomics biomarker discovery. Talanta 2021, 223 , 121797. https://doi.org/10.1016/j.talanta.2020.121797
  53. Yunping Qiu, Irwin J. Kurland. High Accurate Mass Gas Chromatography–Mass Spectrometry for Performing Isotopic Ratio Outlier Analysis: Applications for Nonannotated Metabolite Detection. 2021, 77-88. https://doi.org/10.1007/978-1-0716-0864-7_7
  54. Asja Ćeranić, Christoph Bueschl, Maria Doppler, Alexandra Parich, Kangkang Xu, Marc Lemmens, Hermann Buerstmayr, Rainer Schuhmacher. Enhanced Metabolome Coverage and Evaluation of Matrix Effects by the Use of Experimental-Condition-Matched 13C-Labeled Biological Samples in Isotope-Assisted LC-HRMS Metabolomics. Metabolites 2020, 10 (11) , 434. https://doi.org/10.3390/metabo10110434
  55. Lisa A. Peterson, Silvia Balbo, Naomi Fujioka, Dorothy K. Hatsukami, Stephen S. Hecht, Sharon E. Murphy, Irina Stepanov, Natalia Y. Tretyakova, Robert J. Turesky, Peter W. Villalta. Applying Tobacco, Environmental, and Dietary-Related Biomarkers to Understand Cancer Etiology and Evaluate Prevention Strategies. Cancer Epidemiology, Biomarkers & Prevention 2020, 29 (10) , 1904-1919. https://doi.org/10.1158/1055-9965.EPI-19-1356
  56. Michael G. Leeming, Andrew P. Isaac, Luke Zappia, Richard A.J. O’Hair, William A. Donald, Bernard J. Pope. HiTIME: An efficient model-selection approach for the detection of unknown drug metabolites in LC-MS data. SoftwareX 2020, 12 , 100559. https://doi.org/10.1016/j.softx.2020.100559
  57. Tobias Depke, Janne Gesine Thöming, Adrian Kordes, Susanne Häussler, Mark Brönstrup. Untargeted LC-MS Metabolomics Differentiates Between Virulent and Avirulent Clinical Strains of Pseudomonas aeruginosa. Biomolecules 2020, 10 (7) , 1041. https://doi.org/10.3390/biom10071041
  58. Sascha K. Manier, Markus R. Meyer. Impact of the used solvent on the reconstitution efficiency of evaporated biosamples for untargeted metabolomics studies. Metabolomics 2020, 16 (3) https://doi.org/10.1007/s11306-019-1631-1
  59. Jennifer Cable, Lydia Finley, Benjamin P. Tu, Gary J. Patti, Trudy G. Oliver, Santosha Vardhana, Miyeko Mana, Russell Ericksen, Sanika Khare, Ralph DeBerardinis, Brent R. Stockwell, Aimee Edinger, Marcia Haigis, William Kaelin. Leveraging insights into cancer metabolism—a symposium report. Annals of the New York Academy of Sciences 2020, 1462 (1) , 5-13. https://doi.org/10.1111/nyas.14274
  60. Luis Galvez, Mate Rusz, Michaela Schwaiger-Haber, Yasin El Abiead, Gerrit Hermann, Ute Jungwirth, Walter Berger, Bernhard K Keppler, Michael A Jakupec, Gunda Koellensperger. Preclinical studies on metal based anticancer drugs as enabled by integrated metallomics and metabolomics. Metallomics 2019, 11 (10) , 1716-1728. https://doi.org/10.1039/c9mt00141g
  61. Jan Stanstrup, Corey Broeckling, Rick Helmus, Nils Hoffmann, Ewy Mathé, Thomas Naake, Luca Nicolotti, Kristian Peters, Johannes Rainer, Reza Salek, Tobias Schulze, Emma Schymanski, Michael Stravs, Etienne Thévenot, Hendrik Treutler, Ralf Weber, Egon Willighagen, Michael Witting, Steffen Neumann. The metaRbolomics Toolbox in Bioconductor and beyond. Metabolites 2019, 9 (10) , 200. https://doi.org/10.3390/metabo9100200
  62. Elizabeth M. Llufrio, Kevin Cho, Gary J. Patti. Systems-level analysis of isotopic labeling in untargeted metabolomic data by X13CMS. Nature Protocols 2019, 14 (7) , 1970-1990. https://doi.org/10.1038/s41596-019-0167-1
  63. Claire L. Boulangé, Ilse M. Rood, Joram M. Posma, John C. Lindon, Elaine Holmes, Jack F. M. Wetzels, Jeroen K. J. Deegens, Manuja R. Kaluarachchi. NMR and MS urinary metabolic phenotyping in kidney diseases is fit-for-purpose in the presence of a protease inhibitor. Molecular Omics 2019, 15 (1) , 39-49. https://doi.org/10.1039/C8MO00190A
  64. Lingjue Wang, Fuad J. Naser, Jonathan L. Spalding, Gary J. Patti. A Protocol to Compare Methods for Untargeted Metabolomics. 2019, 1-15. https://doi.org/10.1007/978-1-4939-8769-6_1
  65. Timothy M.D. Ebbels, Jake T.M. Pearce, Noureddin Sadawi, Jianliang Gao, Robert C. Glen. Big Data and Databases for Metabolic Phenotyping. 2019, 329-367. https://doi.org/10.1016/B978-0-12-812293-8.00011-6
  66. Liyan Chen, Yan Gao, Louis Zizhao Wang, Ning Cheung, Gavin S.W. Tan, Gemmy Chiu Ming Cheung, Roger W. Beuerman, Tien Yin Wong, Eric Chun Yong Chan, Lei Zhou. Recent advances in the applications of metabolomics in eye research. Analytica Chimica Acta 2018, 1037 , 28-40. https://doi.org/10.1016/j.aca.2018.01.060
  67. Nicholas J. W. Rattray, Nicole C. Deziel, Joshua D. Wallach, Sajid A. Khan, Vasilis Vasiliou, John P. A. Ioannidis, Caroline H. Johnson. Beyond genomics: understanding exposotypes through metabolomics. Human Genomics 2018, 12 (1) https://doi.org/10.1186/s40246-018-0134-x
  68. Zhucui Li, Yan Lu, Yufeng Guo, Haijie Cao, Qinhong Wang, Wenqing Shui. Comprehensive evaluation of untargeted metabolomics data processing software in feature detection, quantification and discriminating marker selection. Analytica Chimica Acta 2018, 1029 , 50-57. https://doi.org/10.1016/j.aca.2018.05.001
  69. Albert Elmsjö, Jakob Haglöf, Mikael K.R. Engskog, Ida Erngren, Marika Nestor, Torbjörn Arvidsson, Curt Pettersson. Method selectivity evaluation using the co-feature ratio in LC/MS metabolomics: Comparison of HILIC stationary phase performance for the analysis of plasma, urine and cell extracts. Journal of Chromatography A 2018, 1568 , 49-56. https://doi.org/10.1016/j.chroma.2018.05.007
  70. Elizabeth M. Llufrio, Lingjue Wang, Fuad J. Naser, Gary J. Patti. Sorting cells alters their redox state and cellular metabolome. Redox Biology 2018, 16 , 381-387. https://doi.org/10.1016/j.redox.2018.03.004
  71. Fuad J. Naser, Nathaniel G. Mahieu, Lingjue Wang, Jonathan L. Spalding, Stephen L. Johnson, Gary J. Patti. Two complementary reversed-phase separations for comprehensive coverage of the semipolar and nonpolar metabolome. Analytical and Bioanalytical Chemistry 2018, 410 (4) , 1287-1297. https://doi.org/10.1007/s00216-017-0768-x
  72. Dajana Vuckovic. Improving metabolome coverage and data quality: advancing metabolomics and lipidomics for biomarker discovery. Chemical Communications 2018, 54 (50) , 6728-6749. https://doi.org/10.1039/C8CC02592D
  73. Rachel Spicer, Reza M. Salek, Pablo Moreno, Daniel Cañueto, Christoph Steinbeck. Navigating freely-available software tools for metabolomics analysis. Metabolomics 2017, 13 (9) https://doi.org/10.1007/s11306-017-1242-7
  74. Leonardo Perez de Souza, Thomas Naake, Takayuki Tohge, Alisdair R Fernie. From chromatogram to analyte to metabolite. How to pick horses for courses from the massive web resources for mass spectral plant metabolomics. GigaScience 2017, 6 (7) https://doi.org/10.1093/gigascience/gix037
  75. Kristine K. Dennis, Elizabeth Marder, David M. Balshaw, Yuxia Cui, Michael A. Lynes, Gary J. Patti, Stephen M. Rappaport, Daniel T. Shaughnessy, Martine Vrijheid, Dana Boyd Barr. Biomonitoring in the Era of the Exposome. Environmental Health Perspectives 2017, 125 (4) , 502-510. https://doi.org/10.1289/EHP474
  76. Albert Elmsjö, Jakob Haglöf, Mikael K.R. Engskog, Marika Nestor, Torbjörn Arvidsson, Curt Pettersson. The co-feature ratio, a novel method for the measurement of chromatographic and signal selectivity in LC-MS-based metabolomics. Analytica Chimica Acta 2017, 956 , 40-47. https://doi.org/10.1016/j.aca.2016.12.022
  77. Susan J. Gelman, Gary J. Patti. Profiling cancer metabolism at the ‘omic’ level: a last resort or the next frontier?. Cancer & Metabolism 2016, 4 (1) https://doi.org/10.1186/s40170-016-0144-x
  78. Karin Ortmayr, Tim J. Causon, Stephan Hann, Gunda Koellensperger. Increasing selectivity and coverage in LC-MS based metabolome analysis. TrAC Trends in Analytical Chemistry 2016, 82 , 358-366. https://doi.org/10.1016/j.trac.2016.06.011
  79. Nikoline J. Nielsen, Giorgio Tomasi, Jan H. Christensen. Evaluation of chromatographic conditions in reversed phase liquid chromatography-mass spectrometry systems for fingerprinting of polar and amphiphilic plant metabolites. Analytical and Bioanalytical Chemistry 2016, 408 (21) , 5855-5865. https://doi.org/10.1007/s00216-016-9700-z
  80. P.A. Vorkas, J. Shalhoub, M.R. Lewis, K. Spagou, E.J. Want, J.K. Nicholson, A.H. Davies, E. Holmes. Metabolic Phenotypes of Carotid Atherosclerotic Plaques Relate to Stroke Risk: An Exploratory Study. European Journal of Vascular and Endovascular Surgery 2016, 52 (1) , 5-10. https://doi.org/10.1016/j.ejvs.2016.01.022
  81. Nathaniel G Mahieu, Jessica Lloyd Genenbacher, Gary J Patti. A roadmap for the XCMS family of software solutions in metabolomics. Current Opinion in Chemical Biology 2016, 30 , 87-93. https://doi.org/10.1016/j.cbpa.2015.11.009
  82. Nathaniel G. Mahieu, Jonathan L. Spalding, Gary J. Patti. Warpgroup: increased precision of metabolomic data processing by consensus integration bound analysis. Bioinformatics 2016, 32 (2) , 268-275. https://doi.org/10.1093/bioinformatics/btv564
  83. Biswapriya B. Misra, Justin J. J. van der Hooft. Updates in metabolomics tools and resources: 2014–2015. ELECTROPHORESIS 2016, 37 (1) , 86-110. https://doi.org/10.1002/elps.201500417
  84. Susan J. Gelman, Nathaniel G. Mahieu, Kevin Cho, Elizabeth M. Llufrio, Timothy A. Wencewicz, Gary J. Patti. Evidence that 2-hydroxyglutarate is not readily metabolized in colorectal carcinoma cells. Cancer & Metabolism 2015, 3 (1) https://doi.org/10.1186/s40170-015-0139-z
  85. Nicola Zamboni, Alan Saghatelian, Gary J. Patti. Defining the Metabolome: Size, Flux, and Regulation. Molecular Cell 2015, 58 (4) , 699-706. https://doi.org/10.1016/j.molcel.2015.04.021
Download PDF
Go to Analytical Chemistry
Get e-Alerts
Get e-Alerts

Analytical Chemistry

Cite this: Anal. Chem. 2014, 86, 19, 9583–9589
Click to copy citationCitation copied!
https://doi.org/10.1021/ac503092d
Published August 26, 2014

Copyright © 2014 American Chemical Society. This publication is licensed under these Terms of Use.

Request reuse permissions

Article Views

5534

Altmetric

-

Citations

85
Learn about these metrics

Article Views are the COUNTER-compliant sum of full text article downloads since November 2008 (both PDF and HTML) across all institutions and individuals. These metrics are regularly updated to reflect usage leading up to the last few days.

Citations are the number of other articles citing this article, calculated by Crossref and updated daily. Find more information about Crossref citation counts.

The Altmetric Attention Score is a quantitative measure of the attention that a research article has received online. Clicking on the donut icon will load a page at altmetric.com with additional details about the score and the social media presence for the given article. Find more information on the Altmetric Attention Score and how the score is calculated.

  • Abstract

    High Resolution Image
    Download MS PowerPoint Slide

    Figure 1

    Figure 1. Overview of the feature credentialing process. A sample is generated from two cultures of E. coli grown in parallel, one grown on natural-abundance glucose and a second grown on 13C-glucose as the sole carbon source. These two cultures are mixed in distinct ratios prior to harvesting, here 1:1 and 1:2. Extraction and LC/MS analysis is then performed on the standard samples. The resulting data are searched for pairs of coeluting peaks which satisfy the following requirements: (i) the intensities of the peaks must reflect the mixing ratio, (ii) the U-13C peak must predict a feasible number of carbons for the mass in question, and (iii) the exact masses of the peaks must predict an integer number of carbons. These requirements define a “credentialed space” in which the apex of a second peak must be found to qualify as an acceptable isotope. These candidate peaks are then aligned and grouped between the two samples. Each peak pair is compared across samples and a second, stricter intensity check is performed. This requires that the ratios of each sample (Ia12/Ia13 and Ib12/Ib13) are proportional to the mixed ratios of each sample. Peaks that pass these filters are considered credentialed.

    High Resolution Image
    Download MS PowerPoint Slide

    Figure 2

    Figure 2. MS/MS spectra from six representative credentialed features. MS/MS spectra were collected at four collision energies (0, 10, 20, and 40 V) on six credentialed ions. Three of these ions (A) uracil, (B) ADP, and (C) UDP-GlcA were identified based on accurate mass, carbon number, and METLIN database hits. These identifications were confirmed by comparing the experimental MS/MS spectra to the METLIN MS/MS reference spectra as shown. The upper spectrum of each plot is the experimental data, and the lower spectrum is the METLIN reference data. Unmatched peaks are depicted in red. The second three ions (D) 578.0093, (E) 1169.3011, and (F) 848.7473 were classified as unknowns as they did not match any METLIN database entries as either a fragment or parent mass. The MS/MS spectrum of each ion is displayed as normalized intensity at the same four collision energies.

    High Resolution Image
    Download MS PowerPoint Slide

  • References

    This article references 31 other publications.

    1. 1
      Patti, G. J.; Yanes, O.; Siuzdak, G. Nat. Rev. Mol. Cell Biol. 2012, 13, 263– 269
      1
      Innovation Metabolomics: the apogee of the omics trilogy
      Patti, Gary J.; Yanes, Oscar; Siuzdak, Gary
      Nature Reviews Molecular Cell Biology (2012), 13 (4), 263-269CODEN: NRMCBP; ISSN:1471-0072. (Nature Publishing Group)
      A review. Metabolites, the chem. entities that are transformed during metab., provide a functional readout of cellular biochem. With emerging technologies in mass spectrometry, thousands of metabolites can now be quant. measured from minimal amts. of biol. material, which has thereby enabled systems-level analyses. By performing global metabolite profiling, also known as untargeted metabolomics, new discoveries linking cellular pathways to biol. mechanism are being revealed and are shaping our understanding of cell biol., physiol. and medicine.
    2. 2
      Patti, G. J.; Yanes, O.; Shriver, L. P.; Courade, J.-P.; Tautenhahn, R.; Manchester, M.; Siuzdak, G. Nat. Chem. Biol. 2012, 8, 232– 234
      2
      Metabolomics implicates altered sphingolipids in chronic pain of neuropathic origin
      Patti, Gary J.; Yanes, Oscar; Shriver, Leah P.; Courade, Jean-Phillipe; Tautenhahn, Ralf; Manchester, Marianne; Siuzdak, Gary
      Nature Chemical Biology (2012), 8 (3), 232-234CODEN: NCBABT; ISSN:1552-4450. (Nature Publishing Group)
      Neuropathic pain is a debilitating condition for which the development of effective treatments has been limited by an incomplete understanding of its chem. basis. We show by using untargeted metabolomics that sphingomyelin-ceramide metab. is altered in the dorsal horn of rats with neuropathic pain and that the upregulated, endogenous metabolite N,N-dimethylsphingosine induces mech. hypersensitivity in vivo. These results demonstrate the utility of metabolomics to implicate unexplored biochem. pathways in disease.
    3. 3
      Khan, A. P.; Rajendiran, T. M.; Ateeq, B.; Asangani, I. A.; Athanikar, J. N.; Yocum, A. K.; Mehra, R.; Siddiqui, J.; Palapattu, G.; Wei, J. T.; Michailidis, G.; Sreekumar, A.; Chinnaiyan, A. M. Neoplasia (N. Y., NY, U. S.) 2013, 15, 491– 501
      3
      The role of sarcosine metabolism in prostate cancer progression
      Khan, Amjad P.; Rajendiran, Thekkelnaycke M.; Ateeq, Bushra; Asangani, Irfan A.; Athanikar, Jyoti N.; Yocum, Anastasia K.; Mehra, Rohit; Siddiqui, Javed; Palapattu, Ganesh; Wei, John T.; Michailidis, George; Sreekumar, Arun; Chinnaiyan, Arul M.
      Neoplasia (Ann Arbor, MI, United States) (2013), 15 (5), 491-501CODEN: NEOPFL; ISSN:1522-8002. (Neoplasia Press Inc.)
      Metabolomic profiling of prostate cancer (PCa) progression identified markedly elevated levels of sarcosine (N-Me glycine) in metastatic PCa and modest but significant elevation of the metabolite in PCa urine. Here, we examine the role of key enzymes assocd. with sarcosine metab. in PCa progression. Consistent with our earlier report, sarcosine levels were significantly elevated in PCa urine sediments compared to controls, with a modest area under the receiver operating characteristic curve of 0.71. In addn., the expression of sarcosine biosynthetic enzyme, glycine N-methyltransferase (GNMT), was elevated in PCa tissues, while sarcosine dehydrogenase (SARDH) and pipecolic acid oxidase (PIPOX), which metabolize sarcosine, were reduced in prostate tumors. Consistent with this, GNMT promoted the oncogenic potential of prostate cells by facilitating sarcosine prodn., while SARDH and PIPOX reduced the oncogenic potential of prostate cells by metabolizing sarcosine. Accordingly, addn. of sarcosine, but not glycine or alanine, induced invasion and intravasation in an in vivo PCa model. In contrast, GNMT knockdown or SARDH overexpression in PCa xenografts inhibited tumor growth. Taken together, these studies substantiate the role of sarcosine in PCa progression.
    4. 4
      Tang, W. H. W.; Wang, Z.; Levison, B. S.; Koeth, R. A.; Britt, E. B.; Fu, X.; Wu, Y.; Hazen, S. L. N. Engl. J. Med. 2013, 368, 1575– 1584
      4
      Intestinal microbial metabolism of phosphatidylcholine and cardiovascular risk
      Tang, W. H. Wilson; Wang, Zeneng; Levison, Bruce S.; Koeth, Robert A.; Britt, Earl B.; Fu, Xiaoming; Wu, Yuping; Hazen, Stanley L.
      New England Journal of Medicine (2013), 368 (17), 1575-1584CODEN: NEJMAG; ISSN:0028-4793. (Massachusetts Medical Society)
      BACKGROUND Recent studies in animals have shown a mechanistic link between intestinal microbial metab. of the choline moiety in dietary phosphatidylcholine (lecithin) and coronary artery disease through the prodn. of a proatherosclerotic metabolite, trimethylamine-N-oxide (TMAO). We investigated the relationship among intestinal microbiota-dependent metab. of dietary phosphatidylcholine, TMAO levels, and adverse cardiovascular events in humans. METHODS We quantified plasma and urinary levels of TMAO and plasma choline and betaine levels by means of liq. chromatog. and online tandem mass spectrometry after a phosphatidylcholine challenge (ingestion of two hard-boiled eggs and deuterium [d9]-labeled phosphatidylcholine) in healthy participants before and after the suppression of intestinal microbiota with oral broad-spectrum antibiotics. We further examd. the relationship between fasting plasma levels of TMAO and incident major adverse cardiovascular events (death, myocardial infarction, or stroke) during 3 years of follow-up in 4007 patients undergoing elective coronary angiog. RESULTS Time-dependent increases in levels of both TMAO and its d9 isotopologue, as well as other choline metabolites, were detected after the phosphatidylcholine challenge. Plasma levels of TMAO were markedly suppressed after the administration of antibiotics and then reappeared after withdrawal of antibiotics. Increased plasma levels of TMAO were assocd. with an increased risk of a major adverse cardiovascular event (hazard ratio for highest vs. lowest TMAO quartile, 2.54; 95% confidence interval, 1.96 to 3.28; P<0.001). An elevated TMAO level predicted an increased risk of major adverse cardiovascular events after adjustment for traditional risk factors (P<0.001), as well as in lower-risk subgroups. CONCLUSIONS The prodn. of TMAO from dietary phosphatidylcholine is dependent on metab. by the intestinal microbiota. Increased TMAO levels are assocd. with an increased risk of incident major adverse cardiovascular events.
    5. 5
      Jain, M.; Nilsson, R.; Sharma, S.; Madhusudhan, N.; Kitami, T.; Souza, A. L.; Kafri, R.; Kirschner, M. W.; Clish, C. B.; Mootha, V. K. Science 2012, 336, 1040– 1044
      5
      Metabolite profiling identifies a key role for glycine in rapid cancer cell proliferation
      Jain, Mohit; Nilsson, Roland; Sharma, Sonia; Madhusudhan, Nikhil; Kitami, Toshimori; Souza, Amanda L.; Kafri, Ran; Kirschner, Marc W.; Clish, Clary B.; Mootha, Vamsi K.
      Science (Washington, DC, United States) (2012), 336 (6084), 1040-1044CODEN: SCIEAS; ISSN:0036-8075. (American Association for the Advancement of Science)
      Metabolic reprogramming has been proposed to be a hallmark of cancer, yet a systematic characterization of the metabolic pathways active in transformed cells is currently lacking. Using mass spectrometry, we measured the consumption and release (CORE) profiles of 219 metabolites from media across the NCI-60 cancer cell lines, and integrated these data with a preexisting atlas of gene expression. This anal. identified glycine consumption and expression of the mitochondrial glycine biosynthetic pathway as strongly correlated with rates of proliferation across cancer cells. Antagonizing glycine uptake and its mitochondrial biosynthesis preferentially impaired rapidly proliferating cells. Moreover, higher expression of this pathway was assocd. with greater mortality in breast cancer patients. Increased reliance on glycine may represent a metabolic vulnerability for selectively targeting rapid cancer cell proliferation.
    6. 6
      Ivanisevic, J.; Zhu, Z.-J.; Plate, L.; Tautenhahn, R.; Chen, S.; O’Brien, P. J.; Johnson, C. H.; Marletta, M. A.; Patti, G. J.; Siuzdak, G. Anal. Chem. 2013, 85, 6876– 6884
      6
      Toward 'Omic Scale Metabolite Profiling: A Dual Separation-Mass Spectrometry Approach for Coverage of Lipid and Central Carbon Metabolism
      Ivanisevic, Julijana; Zhu, Zheng-Jiang; Plate, Lars; Tautenhahn, Ralf; Chen, Stephen; O'Brien, Peter J.; Johnson, Caroline H.; Marletta, Michael A.; Patti, Gary J.; Siuzdak, Gary
      Analytical Chemistry (Washington, DC, United States) (2013), 85 (14), 6876-6884CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
      Although the objective of any omic science is broad measurement of its constituents, such coverage has been challenging in metabolomics because the metabolome is comprised of a chem. diverse set of small mols. with variable phys. properties. While extensive studies have been performed to identify metabolite isolation and sepn. methods, these strategies introduce bias toward lipophilic or water-sol. metabolites depending on whether reversed-phase (RP) or hydrophilic interaction liq. chromatog. (HILIC) was used, resp. Here the authors extend the authors' consideration of metabolome isolation and sepn. procedures to integrate RPLC/MS and HILIC/MS profiling. An aminopropyl-based HILIC/MS method was optimized on the basis of mobile-phase additives and pH, followed by evaluation of reproducibility. When applied to the untargeted study of perturbed bacterial metabolomes, the HILIC method enabled the accurate assessment of key, dysregulated metabolites in central carbon pathways (e.g., amino acids, org. acids, phosphorylated sugars, energy currency metabolites), which could not be retained by RPLC. To demonstrate the value of the integrative approach, bacterial cells, human plasma, and cancer cells were analyzed by combined RPLC/HILIC sepn. coupled to ESI pos./neg. MS detection. The combined approach resulted in the observation of metabolites assocd. with lipid and central carbon metab. from a single biol. ext., using 80% org. solvent (ACN:MeOH:H2O 2:2:1). It enabled the detection of >30,000 features from each sample type, with the highest no. of uniquely detected features by RPLC in ESI pos. mode and by HILIC in ESI neg. mode. Therefore, when time and sample are limited, the max. amt. of biol. information related to lipid and central carbon metab. can be acquired by combining RPLC ESI pos. and HILIC ESI neg. mode anal.
    7. 7
      Bajad, S. U.; Lu, W.; Kimball, E. H.; Yuan, J.; Peterson, C.; Rabinowitz, J. D. J. Chromatogr., A 2006, 1125, 76– 88
      7
      Separation and quantitation of water soluble cellular metabolites by hydrophilic interaction chromatography-tandem mass spectrometry
      Bajad, Sunil U.; Lu, Wenyun; Kimball, Elizabeth H.; Yuan, Jie; Peterson, Celeste; Rabinowitz, Joshua D.
      Journal of Chromatography A (2006), 1125 (1), 76-88CODEN: JCRAEY; ISSN:0021-9673. (Elsevier B.V.)
      A key unmet need in metabolomics is the ability to efficiently quantify a large no. of known cellular metabolites. Here the authors present a liq. chromatog. (LC)-electrospray ionization tandem mass spectrometry (ESI-MS/MS) method for reliable measurement of 141 metabolites, including components of central carbon, amino acid, and nucleotide metab. The selected LC approach, hydrophilic interaction chromatog. with an amino column, effectively separates highly water sol. metabolites that fail to retain using std. reversed-phase chromatog. MS/MS detection is achieved by scanning through numerous selected reaction monitoring events on a triple quadrupole instrument. When applied to exts. of Escherichia coli grown in [12C]- vs. [13C]glucose, the method reveals appropriate 12C- and 13C-peaks for 79 different metabolites.
    8. 8
      Lu, W.; Bennett, B. D.; Rabinowitz, J. D. J. Chromatogr., B 2008, 871, 236– 242
      8
      Analytical strategies for LC-MS-based targeted metabolomics
      Lu, Wenyun; Bennett, Bryson D.; Rabinowitz, Joshua D.
      Journal of Chromatography B: Analytical Technologies in the Biomedical and Life Sciences (2008), 871 (2), 236-242CODEN: JCBAAI; ISSN:1570-0232. (Elsevier B.V.)
      A review. Recent advances in mass spectrometry are enabling improved anal. of endogenous metabolites. Here the authors discuss several issues relevant to developing liq. chromatog.-electrospray ionization-mass spectrometry methods for targeted metabolomics (i.e., quant. anal. of dozens to hundreds of specific metabolites). Sample prepn. and liq. chromatog. approaches are discussed, with an eye towards the challenge of dealing with a diversity of metabolite classes in parallel. Evidence is presented that heated electrospray ionization (ESI) generally gives improved signal compared to the more traditional unheated ESI. Applicability to targeted metabolomics of triple quadrupole mass spectrometry operating in multiple reaction monitoring (MRM) mode and high mass resoln. full scan mass spectrometry (e.g., time-of-flight, Orbitrap) are described. The authors suggest that both are viable solns., with MRM preferred when targeting a more limited no. of analytes, and full scan preferred for its potential ability to bridge targeted and untargeted metabolomics.
    9. 9
      Masson, P.; Alves, A. C.; Ebbels, T. M. D.; Nicholson, J. K.; Want, E. J. Anal. Chem. 2010, 82, 7779– 7786
      There is no corresponding record for this reference.
    10. 10
      Zelena, E.; Dunn, W. B.; Broadhurst, D.; Francis-McIntyre, S.; Carroll, K. M.; Begley, P.; O’Hagan, S.; Knowles, J. D.; Halsall, A.; HUSERMET Consortium; Wilson, I. D.; Kell, D. B. Anal. Chem. 2009, 81, 1357– 1364
      10
      Development of a Robust and Repeatable UPLC-MS Method for the Long-Term Metabolomic Study of Human Serum
      Zelena, Eva; Dunn, Warwick B.; Broadhurst, David; Francis-McIntyre, Sue; Carroll, Kathleen M.; Begley, Paul; O'Hagan, Steve; Knowles, Joshua D.; Halsall, Antony; Wilson, Ian D.; Kell, Douglas B.
      Analytical Chemistry (Washington, DC, United States) (2009), 81 (4), 1357-1364CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
      A method for performing untargeted metabolomic anal. of human serum has been developed based on protein pptn. followed by Ultra Performance Liq. Chromatog. and Time-of-Flight mass spectrometry (UPLC-TOF-MS). This method was specifically designed to fulfill the requirements of a long-term metabolomic study, spanning more than 3 years, and it was subsequently thoroughly evaluated for robustness and repeatability. The authors describe here the obsd. drift in instrumental performance over time and its improvement with adjustment of the length of anal. block. The optimal setup for the authors' purpose was further validated against a set of serum samples from 30 healthy individuals. The authors also assessed the reproducibility of chromatog. columns with the same chem. of stationary phase from the same manufacturer but from different prodn. batches. The results have allowed the authors to prep. std. operating procedures for "fit for purpose" long-term UPLC-MS metabolomic studies, such as are being employed in the HUSERMET (Human Serum Metabolome) project. This method allows the acquisition of data and subsequent comparison of data collected across many months or years.
    11. 11
      Geier, F. M.; Want, E. J.; Leroi, A. M.; Bundy, J. G. Anal. Chem. 2011, 83, 3730– 3736
      There is no corresponding record for this reference.
    12. 12
      Yanes, O.; Tautenhahn, R.; Patti, G. J.; Siuzdak, G. Anal. Chem. 2011, 83, 2152– 2161
      There is no corresponding record for this reference.
    13. 13
      Nordström, A.; Want, E.; Northen, T.; Lehtiö, J.; Siuzdak, G. Anal. Chem. 2008, 80, 421– 429
      There is no corresponding record for this reference.
    14. 14
      Buescher, J. M.; Moco, S.; Sauer, U.; Zamboni, N. Anal. Chem. 2010, 82, 4403– 4412
      There is no corresponding record for this reference.
    15. 15
      Lu, W.; Clasquin, M. F.; Melamud, E.; Amador-Noguez, D.; Caudy, A. A.; Rabinowitz, J. D. Anal. Chem. 2010, 82, 3212– 3221
      15
      Metabolomic Analysis via Reversed-Phase Ion-Pairing Liquid Chromatography Coupled to a Stand Alone Orbitrap Mass Spectrometer
      Lu, Wenyun; Clasquin, Michelle F.; Melamud, Eugene; Amador-Noguez, Daniel; Caudy, Amy A.; Rabinowitz, Joshua D.
      Analytical Chemistry (Washington, DC, United States) (2010), 82 (8), 3212-3221CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
      The authors present a liq. chromatog.-mass spectrometry (LC-MS) method that capitalizes on the mass-resolving power of the Orbitrap to enable sensitive and specific measurement of known and unanticipated metabolites in parallel, with a focus on water-sol. species involved in core metab. The reversed phase LC method, with a cycle time 25 min, involves a water-methanol gradient on a C18 column with tributylamine as the ion pairing agent. The MS portion involves full scans from 85 to 1000 m/z at 1 Hz and 100,000 resoln. in neg. ion mode on a stand alone Orbitrap ("Exactive"). The median limit of detection, across 80 metabolite stds., was 5 ng/mL with the linear range typically ≥100-fold. For both stds. and a cellular ext. from Saccharomyces cerevisiae (Baker's yeast), the median inter-run relative std. deviation in peak intensity was 8%. In yeast ext., the authors detected 137 known compds., whose 13C-labeling patterns could also be tracked to probe metabolic flux. In yeast engineered to lack a gene of unknown function (YKL215C), the authors obsd. accumulation of an ion of m/z 128.0351, which the authors subsequently confirmed to be oxoproline, resulting in annotation of YKL215C as an oxoprolinase. These examples demonstrate the suitability of the present method for quant. metabolomics, fluxomics, and discovery metabolite profiling.
    16. 16
      Smith, C. A.; Want, E. J.; O’Maille, G.; Abagyan, R.; Siuzdak, G. Anal. Chem. 2006, 78, 779– 787
      16
      XCMS: Processing Mass Spectrometry Data for Metabolite Profiling Using Nonlinear Peak Alignment, Matching, and Identification
      Smith, Colin A.; Want, Elizabeth J.; O'Maille, Grace; Abagyan, Ruben; Siuzdak, Gary
      Analytical Chemistry (2006), 78 (3), 779-787CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
      Metabolite profiling in biomarker discovery, enzyme substrate assignment, drug activity/specificity detn., and basic metabolic research requires new data preprocessing approaches to correlate specific metabolites to their biol. origin. Here we introduce an LC/MS-based data anal. approach, XCMS, which incorporates novel nonlinear retention time alignment, matched filtration, peak detection, and peak matching. Without using internal stds., the method dynamically identifies hundreds of endogenous metabolites for use as stds., calcg. a nonlinear retention time correction profile for each sample. Following retention time correction, the relative metabolite ion intensities are directly compared to identify changes in specific endogenous metabolites, such as potential biomarkers. The software is demonstrated using data sets from a previously reported enzyme knockout study and a large-scale study of plasma samples. XCMS is freely available under an open-source license at http://metlin.scripps.edu/download/.
    17. 17
      Chokkathukalam, A.; Jankevics, A.; Creek, D. J.; Achcar, F.; Barrett, M. P.; Breitling, R. Bioinformatics 2013, 29, 281– 283
      There is no corresponding record for this reference.
    18. 18
      Mishur, R. J.; Rea, S. L. Mass Spectrom. Rev. 2012, 31, 70– 95
      18
      Applications of mass spectrometry to metabolomics and metabonomics: Detection of biomarkers of aging and of age-related diseases
      Mishur, Robert J.; Rea, Shane L.
      Mass Spectrometry Reviews (2012), 31 (1), 70-95CODEN: MSRVD3; ISSN:0277-7037. (John Wiley & Sons, Inc.)
      A review. Every 5 years or so new technologies, or new combinations of old ones, seemingly burst onto the science scene and are then sought after until they reach the point of becoming commonplace. Advances in mass spectrometry instrumentation, coupled with the establishment of standardized chem. fragmentation libraries, increased computing power, novel data-anal. algorithms, new scientific applications, and com. prospects have made mass spectrometry-based metabolomics the latest sought-after technol. This methodol. affords the ability to dynamically catalog and quantify, in parallel, femtomole quantities of cellular metabolites. The study of aging, and the diseases that accompany it, has accelerated significantly in the last decade. Mutant genes that alter the rate of aging have been found that increase lifespan by up to 10-fold in some model organisms, and substantial progress has been made in understanding fundamental alterations that occur at both the mRNA and protein level in tissues of aging organisms. The application of metabolomics to aging research is still relatively new, but has already added significant insight into the aging process. In this review we summarize these findings. We have targeted our manuscript to two audiences: mass spectrometrists interested in applying their tech. knowledge to unanswered questions in the aging field, and gerontologists interested in expanding their knowledge of both mass spectrometry and the most recent advances in aging-related metabolomics. © 2011 Wiley Periodicals, Inc., Mass Spec Rev 31:70-95, 2012.
    19. 19
      Brown, M.; Dunn, W. B.; Dobson, P.; Patel, Y.; Winder, C. L.; Francis-McIntyre, S.; Begley, P.; Carroll, K.; Broadhurst, D.; Tseng, A.; Swainston, N.; Spasic, I.; Goodacre, R.; Kell, D. B. Analyst 2009, 134, 1322– 1332
      19
      Mass spectrometry tools and metabolite-specific databases for molecular identification in metabolomics
      Brown, M.; Dunn, W. B.; Dobson, P.; Patel, Y.; Winder, C. L.; Francis-McIntyre, S.; Begley, P.; Carroll, K.; Broadhurst, D.; Tseng, A.; Swainston, N.; Spasic, I.; Goodacre, R.; Kell, D. B.
      Analyst (Cambridge, United Kingdom) (2009), 134 (7), 1322-1332CODEN: ANALAO; ISSN:0003-2654. (Royal Society of Chemistry)
      The chem. identification of mass spectrometric signals in metabolomic applications is important to provide conversion of anal. data to biol. knowledge about metabolic pathways. The complexity of electrospray mass spectrometric data acquired from a range of samples (serum, urine, yeast intracellular exts., yeast metabolic footprints, placental tissue metabolic footprints) has been investigated and has defined the frequency of different ion types routinely detected. Although some ion types were expected (protonated and deprotonated peaks, isotope peaks, multiply charged peaks) others were not expected (sodium formate adduct ions). In parallel, the Manchester Metabolomics Database (MMD) has been constructed with data from genome scale metabolic reconstructions, HMDB, KEGG, Lipid Maps, BioCyc and DrugBank to provide knowledge on 42,687 endogenous and exogenous metabolite species. The combination of accurate mass data for a large collection of metabolites, theor. isotope abundance data and knowledge of the different ion types detected provided a greater no. of electrospray mass spectrometric signals which were putatively identified and with greater confidence in the samples studied. To provide definitive identification metabolite-specific mass spectral libraries for UPLC-MS and GC-MS have been constructed for 1,065 com. available authentic stds. The MMD data are available at http://dbkgroup.org/MMD.
    20. 20
      Kuhl, C.; Tautenhahn, R.; Böttcher, C.; Larson, T. R.; Neumann, S. Anal. Chem. 2012, 84, 283– 289
      20
      CAMERA: An Integrated Strategy for Compound Spectra Extraction and Annotation of Liquid Chromatography/Mass Spectrometry Data Sets
      Kuhl, Carsten; Tautenhahn, Ralf; Boettcher, Christoph; Larson, Tony R.; Neumann, Steffen
      Analytical Chemistry (Washington, DC, United States) (2012), 84 (1), 283-289CODEN: ANCHAM; ISSN:0003-2700. (American Chemical Society)
      Liq. chromatog. coupled to mass spectrometry is routinely used for metabolomics expts. In contrast to the fairly routine and automated data acquisition steps, subsequent compd. annotation and identification require extensive manual anal. and thus form a major bottleneck in data interpretation. Here the authors present CAMERA, a Bioconductor package integrating algorithms to ext. compd. spectra, annotate isotope and adduct peaks, and propose the accurate compd. mass even in highly complex data. To evaluate the algorithms, the authors compared the annotation of CAMERA against a manually defined annotation for a mixt. of known compds. spiked into a complex matrix at different concns. CAMERA successfully extd. accurate masses for 89.7% and 90.3% of the annotatable compds. in pos. and neg. ion modes, resp. Furthermore, the authors present a novel annotation approach that combines spectral information of data acquired in opposite ion modes to further improve the annotation rate. The authors demonstrate the utility of CAMERA in two different, easily adoptable plant metabolomics expts., where the application of CAMERA drastically reduced the amt. of manual anal.
    21. 21
      Alonso, A.; Julià, A.; Beltran, A.; Vinaixa, M.; Díaz, M.; Ibañez, L.; Correig, X.; Marsal, S. Bioinformatics 2011, 27, 1339– 1340
      21
      AStream: an R package for annotating LC/MS metabolomic data
      Alonso, Arnald; Julia, Antonio; Beltran, Antoni; Vinaixa, Maria; Diaz, Marta; Ibanez, Lourdes; Correig, Xavier; Marsal, Sara
      Bioinformatics (2011), 27 (9), 1339-1340CODEN: BOINFP; ISSN:1367-4803. (Oxford University Press)
      AStream, an R-statistical software package for the curation and identification of feature peaks extd. from liq. chromatog. mass spectrometry (LC/MS) metabolomics data, is described. AStream detects isotopic, fragment and adduct patterns by identifying feature pairs that fulfill expected relational patterns. Data redn. by AStream allows compds. to be identified reliably and subsequently linked to metabolite databases. AStream provides researchers with a fast, reliable tool for summarizing metabolomic data, notably reducing curation time and increasing consistency of results.
    22. 22
      R Core Team. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, 2014; http://www.R-project.org.
      There is no corresponding record for this reference.
    23. 23
      Smith, C. A.; Maille, G.; Want, E. J.; Qin, C.; Trauger, S. A.; Brandon, T. R.; Custodio, D. E.; Abagyan, R.; Siuzdak, G. Ther. Drug Monit. 2005, 27, 747– 751
      23
      METLIN. A metabolite mass spectral database
      Smith, Colin A.; O'Maille, Grace; Want, Elizabeth J.; Qin, Chuan; Trauger, Sunia A.; Brandon, Theodore R.; Custodio, Darlene E.; Abagyan, Ruben; Siuzdak, Gary
      Therapeutic Drug Monitoring (2005), 27 (6), 747-751CODEN: TDMODV; ISSN:0163-4356. (Lippincott Williams & Wilkins)
      Endogenous metabolites have gained increasing interest over the past 5 years largely for their implications in diagnostic and pharmaceutical biomarker discovery. METLIN (http://metlin.scripps.edu), a freely accessible web-based data repository, has been developed to assist in a broad array of metabolite research and to facilitate metabolite identification through mass anal. METLIN includes an annotated list of known metabolite structural information that is easily cross-correlated with its catalog of high-resoln. Fourier transform mass spectrometry (FTMS) spectra, tandem mass spectrometry (MS/MS) spectra, and LC/MS data.
    24. 24
      Guo, A. C.; Jewison, T.; Wilson, M.; Liu, Y.; Knox, C.; Djoumbou, Y.; Lo, P.; Mandal, R.; Krishnamurthy, R.; Wishart, D. S. Nucleic Acids Res. 2013, 41, D625– D630
      There is no corresponding record for this reference.
    25. 25
      Ong, S.-E.; Blagoev, B.; Kratchmarova, I.; Kristensen, D. B.; Steen, H.; Pandey, A.; Mann, M. Mol. Cell. Proteomics 2002, 1, 376– 386
      25
      Stable isotope labeling by amino acids in cell culture, SILAC, as a simple and accurate approach to expression proteomics
      Ong, Shao-En; Blagoev, Blagoy; Kratchmarova, Irina; Kristensen, Dan Bach; Steen, Hanno; Pandey, Akhilesh; Mann, Matthias
      Molecular and Cellular Proteomics (2002), 1 (5), 376-386CODEN: MCPOBS; ISSN:1535-9476. (American Society for Biochemistry and Molecular Biology, Inc.)
      Quant. proteomics has traditionally been performed by two-dimensional gel electrophoresis, but recently, mass spectrometric methods based on stable isotope quantitation have shown great promise for the simultaneous and automated identification and quantitation of complex protein mixts. Here we describe a method, termed SILAC, for stable isotope labeling by amino acids in cell culture, for the in vivo incorporation of specific amino acids into all mammalian proteins. Mammalian cell lines are grown in media lacking a std. essential amino acid but supplemented with a non-radioactive, isotopically labeled form of that amino acid, in this case deuterated leucine (Leu-d3). We find that growth of cells maintained in these media is no different from growth in normal media as evidenced by cell morphol., doubling time, and ability to differentiate. Complete incorporation of Leu-d3 occurred after five doublings in the cell lines and proteins studied. Protein populations from exptl. and control samples are mixed directly after harvesting, and mass spectrometric identification is straightforward as every leucine-contg. peptide incorporates either all normal leucine or all Leu-d3. We have applied this technique to the relative quantitation of changes in protein expression during the process of muscle cell differentiation. Proteins that were found to be up-regulated during this process include glyceraldehyde-3-phosphate dehydrogenase, fibronectin, and pyruvate kinase M2. SILAC is a simple, inexpensive, and accurate procedure that can be used as a quant. proteomic approach in any cell culture system.
    26. 26
      Wiese, S.; Reidegeld, K. A.; Meyer, H. E.; Warscheid, B. Proteomics 2007, 7, 340– 350
      26
      Protein labeling by iTRAQ: a new tool for quantitative mass spectrometry in proteome research
      Wiese, Sebastian; Reidegeld, Kai A.; Meyer, Helmut E.; Warscheid, Bettina
      Proteomics (2007), 7 (3), 340-350CODEN: PROTC7; ISSN:1615-9853. (Wiley-VCH Verlag GmbH & Co. KGaA)
      A novel, MS-based approach for the relative quantification of proteins, relying on the derivatization of primary amino groups in intact proteins using isobaric tag for relative and abs. quantitation (iTRAQ) is presented. Due to the isobaric mass design of the iTRAQ reagents, differentially labeled proteins do not differ in mass; accordingly, their corresponding proteolytic peptides appear as single peaks in MS scans. Because quant. information is provided by isotope-encoded reporter ions that can only be obsd. in MS/MS spectra, the authors analyzed the fragmentation behavior of ESI and MALDI ions of peptides generated from iTRAQ-labeled proteins using a TOF/TOF and/or a QTOF instrument. The authors obsd. efficient liberation of reporter ions for singly protonated peptides at low-energy collision conditions. In contrast, increased collision energies were required to liberate the iTRAQ label from lysine side chains of doubly charged peptides and, thus, to observe reporter ions suitable for relative quantification of proteins with high accuracy. The authors then developed a quant. strategy that comprises labeling of intact proteins by iTRAQ followed by gel electrophoresis and peptide MS/MS analyses. As proof of principle, mixts. of five different proteins in various concn. ratios were quantified, demonstrating the general applicability of the approach presented here to quant. MS-based proteomics.
    27. 27
      Hebert, A. S.; Merrill, A. E.; Bailey, D. J.; Still, A. J.; Westphall, M. S.; Strieter, E. R.; Pagliarini, D. J.; Coon, J. J. Nat. Methods 2013, 10, 332– 334
      27
      Neutron-encoded mass signatures for multiplexed proteome quantification
      Hebert, Alexander S.; Merrill, Anna E.; Bailey, Derek J.; Still, Amelia J.; Westphall, Michael S.; Strieter, Eric R.; Pagliarini, David J.; Coon, Joshua J.
      Nature Methods (2013), 10 (4), 332-334CODEN: NMAEA3; ISSN:1548-7091. (Nature Publishing Group)
      We describe a protein quantification method called neutron encoding that exploits the subtle mass differences caused by nuclear binding energy variation in stable isotopes. These mass differences are synthetically encoded into amino acids and incorporated into yeast and mouse proteins via metabolic labeling. Mass spectrometry anal. with high mass resoln. (>200,000) reveals the isotopologue-embedded peptide signals, permitting quantification. Neutron encoding will enable highly multiplexed proteome anal. with excellent dynamic range and accuracy.
    28. 28
      Birkemeyer, C.; Luedemann, A.; Wagner, C.; Erban, A.; Kopka, J. Trends Biotechnol. 2005, 23, 28– 33
      There is no corresponding record for this reference.
    29. 29
      Mashego, M. R.; Wu, L.; Van Dam, J. C.; Ras, C.; Vinke, J. L.; Van Winden, W. A.; Van Gulik, W. M.; Heijnen, J. J. Biotechnol. Bioeng. 2004, 85, 620– 628
      There is no corresponding record for this reference.
    30. 30
      De Jong, F. A.; Beecher, C. Bioanalysis 2012, 4, 2303– 2314
      30
      Addressing the current bottlenecks of metabolomics: Isotopic Ratio Outlier Analysis®, an isotopic-labeling technique for accurate biochemical profiling
      de Jong, Felice A.; Beecher, Chris
      Bioanalysis (2012), 4 (18), 2303-2314CODEN: BIOAB4; ISSN:1757-6180. (Future Science Ltd.)
      A review. Metabolomics or biochem. profiling is a fast emerging science; however, there are still many assocd. bottlenecks to overcome before measurements will be considered robust. Advances in MS resoln. and sensitivity, ultra pressure LC-MS, ESI, and isotopic approaches such as flux anal. and stable-isotope diln., have made it easier to quantitate biochems. The digitization of mass spectrometers has simplified informatic aspects. However, issues of anal. variability, ion suppression and metabolite identification still plague metabolomics investigators. These hurdles need to be overcome for accurate metabolite quantitation not only for in vitro systems, but for complex matrixes such as biofluids and tissues, before it is possible to routinely identify biomarkers that are assocd. with the early prediction and diagnosis of diseases. In this report, we describe a novel isotopic-labeling method that uses the creation of distinct biochem. signatures to eliminate current bottlenecks and enable accurate metabolic profiling.
    31. 31
      Stupp, G. S.; Clendinen, C. S.; Ajredini, R.; Szewc, M. A.; Garrett, T.; Menger, R. F.; Yost, R. A.; Beecher, C.; Edison, A. S. Anal. Chem. 2013, 85, 11858– 11865
      There is no corresponding record for this reference.

  • Supporting Information

    Supporting Information

    Additional information as noted in text. This material is available free of charge via the Internet at http://pubs.acs.org.


    Terms & Conditions

    Most electronic Supporting Information files are available without a subscription to ACS Web Editions. Such files may be downloaded by article for research use (if there is a public use license linked to the relevant article, that license may permit other uses). Permission may be obtained from ACS for other uses through requests via the RightsLink permission system: http://pubs.acs.org/page/copyright/permissions.html.