XCMS: processing mass spectrometry data for metabolite profiling using nonlinear peak alignment, matching, and identification

Metabolite profiling in biomarker discovery, enzyme substrate assignment, drug activity/specificity determination, and basic metabolic research requires new data preprocessing approaches to correlate specific metabolites to their biological origin. Here we introduce an LC/MS-based data analysis approach, XCMS, which incorporates novel nonlinear retention time alignment, matched filtration, peak detection, and peak matching. Without using internal standards, the method dynamically identifies hundreds of endogenous metabolites for use as standards, calculating 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/.     

Simultaneous screen for microsomal stability and metabolite profile by direct injection turbulent-laminar flow LC-LC and automated tandem mass spectrometry

A LC-LC/MS/MS method has been developed that significantly increases the throughput in metabolism screening of drug candidates during lead optimization in discovery. This was accomplished by the reduction of sample preparation time through an on-line extraction of a drug and its metabolites from microsomal proteins using turbulent flow chromatography. Following its injection onto a column at turbulent flow, the drug and its metabolites are backwashed onto a reverse-phase column via on-line column switching and resolved chromatographically at a laminar flow of 2 mL/min. This tandem turbulent-laminar flow chromatographic system in a total cycle time of 8 min can achieve adequate separation of isomeric metabolites of venlafaxine, haloperidol, or adatanserin. Further improvement in throughput can be achieved by multiplexing both microsomal stability assessment and metabolite profiling into a single analysis. This is made possible by the ability of the ion-trap mass spectrometer to perform simultaneously multiple-reaction monitoring for microsomal stability and data-dependent multiple-stage mass spectrometric analysis for metabolite profiling within a single LC analysis. Such a LC-LC/MS/MS approach can dramatically shorten the time for providing metabolism feedback to the drug discovery process.     

XCMS2: processing tandem mass spectrometry data for metabolite identification and structural characterization

Mass spectrometry based metabolomics represents a new area for bioinformatics technology development. While the computational tools currently available such as XCMS statistically assess and rank LC-MS features, they do not provide information about their structural identity. XCMS(2) is an open source software package which has been developed to automatically search tandem mass spectrometry (MS/MS) data against high quality experimental MS/MS data from known metabolites contained in a reference library (METLIN). Scoring of hits is based on a "shared peak count" method that identifies masses of fragment ions shared between the analytical and reference MS/MS spectra. Another functional component of XCMS(2) is the capability of providing structural information for unknown metabolites, which are not in the METLIN database. This "similarity search" algorithm has been developed to detect possible structural motifs in the unknown metabolite which may produce characteristic fragment ions and neutral losses to related reference compounds contained in METLIN, even if the precursor masses are not the same.     

Nonlinear data alignment for UPLC-MS and HPLC-MS based metabolomics: quantitative analysis of endogenous and exogenous metabolites in human serum

A nonlinear alignment strategy was examined for the quantitative analysis of serum metabolites. Two small-molecule mixtures with a difference in relative concentration of 20-100% for 10 of the compounds were added to human serum. The metabolomics protocol using UPLC and XCMS for LC-MS data alignment could readily identify 8 of 10 spiked differences among more than 2700 features detected. Normalization of data against a single factor obtained through averaging the XCMS integrated response areas of spiked standards increased the number of identified differences. The original data structure was well preserved using XCMS, but reintegration of identified differences in the original data reduced the number of false positives. Using UPLC for separation resulted in 20% more detected components compared to HPLC. The length of the chromatographic separation also proved to be a crucial parameter for a number of detected features. Moreover, UPLC displayed better retention time reproducibility and signal-to-noise ratios for spiked compounds over HPLC, making this technology more suitable for nontargeted metabolomics applications.     

Systematic identification of conserved metabolites in GC/MS data for metabolomics and biomarker discovery

Analysis of metabolomic profiling data from gas chromatography-mass spectrometry (GC/MS) measurements usually relies upon reference libraries of metabolite mass spectra to structurally identify and track metabolites. In general, techniques to enumerate and track unidentified metabolites are nonsystematic and require manual curation. We present a method and software implementation, freely available at http://spectconnect.mit.edu, that can systematically detect components that are conserved across samples without the need for a reference library or manual curation. We validate this approach by correctly identifying the components in a known mixture and the discriminating components in a spiked mixture. Finally, we demonstrate an application of this approach with a brief analysis of the Escherichia coli metabolome. By systematically cataloguing conserved metabolite peaks prior to data analysis methods, our approach broadens the scope of metabolomics and facilitates biomarker discovery.     

Systematic evaluation of solid-phase microextraction coatings for untargeted metabolomic profiling of biological fluids by liquid chromatography-mass spectrometry

In this study, we propose for the first time the use of solid-phase microextraction (SPME) in combination with liquid chromatography-mass spectrometry for untargeted metabolomic profiling of biological fluids. To achieve this goal, we first systematically evaluated 42 different SPME coatings for the extraction of 36 metabolites from different chemical classes and of widely varying polarities (log P range of -7.9 to 7.4) in order to identify SPME coatings which are the most suitable for metabolomic studies and to improve the extraction of polar metabolites over the existing commercial SPME devices. Three types of SPME coatings (mixed-mode coatings, polar-enhanced polystyrene-divinylbenzene, and phenylboronic acid) performed the best for simultaneous extraction of both hydrophilic and hydrophobic metabolites at physiological conditions, thus making them suitable for untargeted metabolomic profiling applications. A rapid and simple SPME method was then developed with single-use biocompatible mixed-mode coating for the metabolomic profiling of human plasma in combination with liquid chromatography-high-resolution mass spectrometry on a benchtop Orbitrap system. This optimized SPME method was evaluated versus ultrafiltration and solvent precipitation in terms of metabolite coverage and method precision. SPME detected 1592-3320 features versus 2082-3245 features detected by solvent precipitation methods and 2093-2686 detected for ultrafiltration using the same pooled human plasma sample. Method precision of SPME ranged between 11% and 18% (expressed as median relative standard deviation (RSD) of n = 7 replicates) versus 8-19% for solvent precipitation and 20-22% for ultrafiltration. The results demonstrate that the proposed SPME methodology reduces ionization suppression, provides free concentration information for hydrophobic analytes which are not detected by ultrafiltration methods, and can improve metabolite coverage over existing methodologies.     

Systematic Evaluation of Extraction Methods for Multiplatform-Based Metabotyping: Application to the Fasciola hepatica Metabolome

Combining data from multiple analytical platforms is essential for comprehensive study of the molecular phenotype (metabotype) of a given biological sample. The metabolite profiles generated are intrinsically dependent on the analytical platforms, each requiring optimization of instrumental parameters, separation conditions, and sample extraction to deliver maximal biological information. An in-depth evaluation of extraction protocols for characterizing the metabolome of the hepatobiliary fluke Fasciola hepatica , using ultra performance liquid chromatography and capillary electrophoresis coupled with mass spectroscopy is presented. The spectrometric methods were characterized by performance, and metrics of merit were established, including precision, mass accuracy, selectivity, sensitivity, and platform stability. Although a core group of molecules was common to all methods, each platform contributed a unique set, whereby 142 metabolites out of 14,724 features were identified. A mixture design revealed that the chloroform:methanol:water proportion of 15:59:26 was globally the best composition for metabolite extraction across UPLC-MS and CE-MS platforms accommodating different columns and ionization modes. Despite the general assumption of the necessity of platform-adapted protocols for achieving effective metabotype characterization, we show that an appropriately designed single extraction procedure is able to fit the requirements of all technologies. This may constitute a paradigm shift in developing efficient protocols for high-throughput metabolite profiling with more-general analytical applicability.     

In-vial dual extraction for direct LC-MS analysis of plasma for comprehensive and highly reproducible metabolic fingerprinting

Metabolic fingerprinting of biological tissues has become an important area of research, particularly in the biomarker discovery field. Methods have inherent analytical variation, and new approaches are necessary to ensure that the vast numbers of intact metabolites present in biofluids are detected. Here, we describe an in-vial dual extraction (IVDE) method and a direct injection method that shows the total number of features recovered to be over 4500 from a single 20 μL plasma aliquot. By applying a one-step extraction consisting of a lipophilic and hydrophilic layer within a single vial insert, we showed that analytical variation was decreased. This was achieved by reducing sample preparation stages including procedures of drying and transfers. The two phases in the vial, upper and lower, underwent HPLC-QTOF analysis on individually customized LC gradients in both positive and negative ionization modes. A 60 min lipid profiling HPLC-QTOF method for the lipophilic phase was specifically developed, enabling the separation and putative identification of fatty acids, glycerolipids, glycerophospholipids, sphingolipids, and sterols. The aqueous phase of the extract underwent direct injection onto a 45 min gradient, enabling the detection of both polarities. The IVDE method was compared to two traditional extraction methods. The first method was a two-step ether evaporation and IPA resuspension, and the second method was a methanol precipitation typically used in fingerprinting studies. The IVDE provided a 378% increase in reproducible features when compared to evaporation and a 269% increase when compared to the precipitate and inject method. As a proof of concept, the method was applied to an animal model of diabetes. A 2-fold increase in discriminant metabolites was found when comparing diabetic and control rats with IVDE. These discriminant metabolites accounted for around 600 entities, out of which 388 were identified in available databases.     

Optimizing the use of quality control samples for signal drift correction in large-scale urine metabolic profiling studies

The evident importance of metabolic profiling for biomarker discovery and hypothesis generation has led to interest in incorporating this technique into large-scale studies, e.g., clinical and molecular phenotyping studies. Nevertheless, these lengthy studies mandate the use of analytical methods with proven reproducibility. An integrated experimental plan for LC-MS profiling of urine, involving sample sequence design and postacquisition correction routines, has been developed. This plan is based on the optimization of the frequency of analyzing identical quality control (QC) specimen injections and using the QC intensities of each metabolite feature to construct a correction trace for all the samples. The QC-based methods were tested against other current correction practices, such as total intensity normalization. The evaluation was based on the reproducibility obtained from technical replicates of 46 samples and showed the feature-based signal correction (FBSC) methods to be superior to other methods, resulting in ~1000 and 600 metabolite features with coefficient of variation (CV) < 15% within and between two blocks, respectively. Additionally, the required frequency of QC sample injection was investigated and the best signal correction results were achieved with at least one QC injection every 2 h of urine sample injections (n = 10). Higher rates of QC injections (1 QC/h) resulted in slightly better correction but at the expense of longer total analysis time.     

Assessment of compatibility between extraction methods for NMR- and LC/MS-based metabolomics

Because of the wide range of chemically and structurally diverse metabolites, efforts to survey the complete metabolome rely on the implementation of multiplatform approaches based on nuclear magnetic resonance (NMR) and mass spectrometry (MS). Sample preparation disparities between NMR and MS, however, may limit the analysis of the same samples by both platforms. Specifically, deuterated solvents used in NMR strategies can complicate LC/MS analysis as a result of potential mass shifts, whereas acidic solutions typically used in LC/MS methods to enhance ionization of metabolites can severely affect reproducibility of NMR measurements. These intrinsically different sample preparation requirements result in the application of different procedures for metabolite extraction, which involve additional sample and unwanted variability. To address this issue, we investigated 12 extraction protocols in liver tissue involving different aqueous/organic solvents and temperatures that may satisfy the requirements for both NMR and LC/MS simultaneously. We found that deuterium exchange did not affect LC/MS results, enabling the measurement of metabolites by NMR and, subsequently, the direct analysis of the same samples by using LC/MS with no need for solvent exchange. Moreover, our results show that the choice of solvents rather than the temperature determined the extraction efficiencies of metabolites, a combination of methanol/chloroform/water and methanol/water being the extraction methods that best complement NMR and LC/MS analysis for metabolomic studies.     

Multiple ionization mass spectrometry strategy used to reveal the complexity of metabolomics

A multiple ionization mass spectrometry strategy is presented based on the analysis of human serum extracts. Chromatographic separation was interfaced inline with the atmospheric pressure ionization techniques electrospray ionization (ESI) and atmospheric pressure chemical ionization (APCI) in both positive (+) and negative (-) ionization modes. Furthermore, surface-based matrix-assisted laser desorption/ionization (MALDI) and desorption ionization on silicon (DIOS) mass spectrometry were also integrated with the separation through fraction collection and offline mass spectrometry. Processing of raw data using the XCMS software resulted in time-aligned ion features, which are defined as a unique m/z at a unique retention time. The ion feature lists obtained through LC-MS with ESI and APCI interfaces in both +/- ionization modes were compared, and unique ion tables were generated. Nonredundant, unique ion features, were defined as mass numbers for which no mass numbers corresponding to [M + H](+), [M - H](-), or [M + Na](+) were observed in the other ionization methods at the same retention time. Analysis of the extracted serum using ESI for both (+) and (-) ions resulted in >90% additional unique ions being detected in the (-) ESI mode. Complementing the ESI analysis with APCI resulted in an additional approximately 20% increase in unique ions. Finally, ESI/APCI ionization was combined with fraction collection and offline-MALDI and DIOS mass spectrometry. The parts of the total ion current chromatograms in the LC-MS acquired data corresponding to collected fractions were summed, and m/z lists were compiled and compared to the m/z lists obtained from the DIOS/MALDI spectra. It was observed that, for each fraction, DIOS accounted for approximately 50% of the unique ions detected. These results suggest that true global metabolomics will require multiple ionization technologies to address the inherent metabolite diversity and therefore the complexity in and of metabolomics studies.     

Methodology for profiling the steroid metabolome in animal tissues using ultraperformance liquid chromatography-electrospray-time-of-flight mass spectrometry

The advent of mass spectrometry-based metabolite profiling techniques should allow investigations into the behavior and regulation of many of the low abundant signaling molecules present in the animal metabolome such as hormones and so aid investigations into the mechanisms of endocrine system function and disorders. Examples of their potential applications include endocrine-related diseases such as some cancers, as well as endocrine disruption in wildlife and humans caused by some environmental contaminants. In the present study, a method was developed to profile a variety of vertebrate steroids and their conjugates in fish tissues using solid phase extraction (SPE) prior to ultraperformance liquid chromatography linked to electrospray-time-of-flight mass spectrometry (UPLC-TOF MS). Analysis of tissue extracts spiked with steroids revealed that most analytes could be detected at subnanogram per gram concentrations in liver or testes, but ovarian extracts caused ion suppression of estrogens. A comparison of the steroid metabolome of the ovaries and testes using partial least-squares discrimination analysis (PLS-DA) revealed that the androgens 11-ketotestosterone, 11-hydroxyandrostenedione, androstenedione, and two unidentified cortisol-type glucocorticoids were discriminatory steroid markers in testes extracts. In contrast, cortisol was the predominant glucocorticoid marker in ovary extracts, and cortisone was abundant in extracts of both testes and ovaries. A number of other unidentified nonsteroidal metabolites were also differentially expressed between the testes and ovarian tissues. These studies reveal that metabolite profiling using UPLC-TOF MS is a promising approach to investigate steroid homeostasis in animal tissues.     

Feature selection and classification of high-resolution NMR spectra in the complex wavelet transform domain

Successful identification of the important metabolite features in high-resolution nuclear magnetic resonance (NMR) spectra is a crucial task for the discovery of biomarkers that have the potential for early diagnosis of disease and subsequent monitoring of its progression. Although a number of traditional features extraction/selection methods are available, most of them have been conducted in the original frequency domain and disregarded the fact that an NMR spectrum comprises a number of local bumps and peaks with different scales. In the present study a complex wavelet transform that can handle multiscale information efficiently and has an energy shift-insensitive property is proposed as a method to improve feature extraction and classification in NMR spectra. Furthermore, a multiple testing procedure based on a false discovery rate (FDR) was used to identify important metabolite features in the complex wavelet domain. Experimental results with real NMR spectra showed that classification models constructed with the complex wavelet coefficients selected by the FDR-based procedure yield lower rates of misclassification than models constructed with original features and conventional wavelet coefficients.     

XCMS Online: a web-based platform to process untargeted metabolomic data

Recently, interest in untargeted metabolomics has become prevalent in the general scientific community among an increasing number of investigators. The majority of these investigators, however, do not have the bioinformatic expertise that has been required to process metabolomic data by using command-line driven software programs. Here we introduce a novel platform to process untargeted metabolomic data that uses an intuitive graphical interface and does not require installation or technical expertise. This platform, called XCMS Online, is a web-based version of the widely used XCMS software that allows users to easily upload and process liquid chromatography/mass spectrometry data with only a few mouse clicks. XCMS Online provides a solution for the complete untargeted metabolomic workflow including feature detection, retention time correction, alignment, annotation, statistical analysis, and data visualization. Results can be browsed online in an interactive, customizable table showing statistics, chromatograms, and putative METLIN identities for each metabolite. Additionally, all results and images can be downloaded as zip files for offline analysis and publication. XCMS Online is available at https://xcmsonline.scripps.edu.     

Reducing time and increasing sensitivity in sample preparation for adherent mammalian cell metabolomics

A simple, fast, and reproducible sample preparation procedure was developed for relative quantification of metabolites in adherent mammalian cells using the clonal β-cell line INS-1 as a model sample. The method was developed by evaluating the effect of different sample preparation procedures on high performance liquid chromatography- mass spectrometry quantification of 27 metabolites involved in glycolysis and the tricarboxylic acid cycle on a directed basis as well as for all detectable chromatographic features on an undirected basis. We demonstrate that a rapid water rinse step prior to quenching of metabolism reduces components that suppress electrospray ionization thereby increasing signal for 26 of 27 targeted metabolites and increasing total number of detected features from 237 to 452 with no detectable change of metabolite content. A novel quenching technique is employed which involves addition of liquid nitrogen directly to the culture dish and allows for samples to be stored at -80 °C for at least 7 d before extraction. Separation of quenching and extraction steps provides the benefit of increased experimental convenience and sample stability while maintaining metabolite content similar to techniques that employ simultaneous quenching and extraction with cold organic solvent. The extraction solvent 9:1 methanol: chloroform was found to provide superior performance over acetonitrile, ethanol, and methanol with respect to metabolite recovery and extract stability. Maximal recovery was achieved using a single rapid (～1 min) extraction step. The utility of this rapid preparation method (～5 min) was demonstrated through precise metabolite measurements (11% average relative standard deviation without internal standards) associated with step changes in glucose concentration that evoke insulin secretion in the clonal β-cell line INS-1.     

Expanding coverage of the metabolome for global metabolite profiling

Mass spectrometry-based metabolomics is the comprehensive study of naturally occurring small molecules collectively known as the metabolome. Given the vast structural diversity and chemical properties of endogenous metabolites, biological extraction and chromatography methods bias the number, property, and concentration of metabolites detected by mass spectrometry and creates a challenge for global untargeted studies. In this work, we used Escherichia coli bacterial cells to explore the influence of solvent polarity, temperature, and pH in extracting polar and nonpolar metabolites simultaneously. In addition, we explored chromatographic conditions involving different stationary and mobile phases that optimize the separation and ionization of endogenous metabolite extracts as well as a mixture of synthetic standards. Our results reveal that hot polar solvents are the most efficient in extracting both hydrophilic and hydrophobic metabolites simultaneously. In addition, ammonium fluoride in the mobile phase substantially improved ionization efficiency in negative electrospray ionization mode by an average increase in signal intensity of 5.7 and over a 2-fold increase in the total number of features detected. The improvement in sensitivity with ammonium fluoride resulted in 3.5 times as many metabolite hits in databases compared to ammonium acetate or formic acid enriched mobile phases and allowed for the identification of unique metabolites involved in fundamental cellular pathways.     

metaXCMS: second-order analysis of untargeted metabolomics data

Mass spectrometry-based untargeted metabolomics often results in the observation of hundreds to thousands of features that are differentially regulated between sample classes. A major challenge in interpreting the data is distinguishing metabolites that are causally associated with the phenotype of interest from those that are unrelated but altered in downstream pathways as an effect. To facilitate this distinction, here we describe new software called metaXCMS for performing second-order ("meta") analysis of untargeted metabolomics data from multiple sample groups representing different models of the same phenotype. While the original version of XCMS was designed for the direct comparison of two sample groups, metaXCMS enables meta-analysis of an unlimited number of sample classes to facilitate prioritization of the data and increase the probability of identifying metabolites causally related to the phenotype of interest. metaXCMS is used to import XCMS results that are subsequently filtered, realigned, and ultimately compared to identify shared metabolites that are up- or down-regulated across all sample groups. We demonstrate the software's utility by identifying histamine as a metabolite that is commonly altered in three different models of pain. metaXCMS is freely available at http://metlin.scripps.edu/metaxcms/.     

Cross-platform method for identifying candidate network biomarkers for prostate cancer

Discovering biomarkers using mass spectrometry (MS) and microarray expression profiles is a promising strategy in molecular diagnosis. Here, the authors proposed a new pipeline for biomarker discovery that integrates disease information for proteins and genes, expression profiles in both genomic and proteomic levels, and protein?protein interactions (PPIs) to discover high confidence network biomarkers. Using this pipeline, a total of 474 molecules (genes and proteins) related to prostate cancer were identified and a prostate-cancer-related network (PCRN) was derived from the integrative information. Thus, a set of candidate network biomarkers were identified from multiple expression profiles composed by eight microarray datasets and one proteomics dataset. The network biomarkers with PPIs can accurately distinguish the prostate patients from the normal ones, which potentially provide more reliable hits of biomarker candidates than conventional biomarker discovery methods.     

Discovery of enzymatic activity using stable isotope metabolite labeling and liquid chromatography-mass spectrometry

Stable isotope labeling of an intracellular chemical precursor or metabolite allows direct detection of downstream metabolites of that precursor, arising from novel enzymatic activity of interest, using metabolite profiling liquid chromatography-mass spectrometry. This approach allows the discrimination of downstream metabolites produced from novel enzymatic activity from the unlabeled form of the metabolite arising from native metabolic processes within the cell. Even for the case in which a given product of novel enzymatic activity is a transient, the novel enzymatic activity that produced it can be demonstrated to exist indirectly by identification of product-specific downstream metabolites. Therefore, direct or indirect discovery of novel enzymatic machinery engineered within a cell can be accomplished without a requirement for direct product purification or identification. The application of this approach to confirm the presence of a novel metabolic activity, alanine 2,3-aminomutase, obtained by mutagenesis and selection are discussed. The advantages of metabolite profiling approaches to metabolic engineering in terms of accelerating enzyme discovery and development of intellectual property will also be highlighted.