MetaProClust-MS1: an MS1 Profiling Approach for Large-Scale Microbiome Screening.

Metaproteomics is used to explore the functional dynamics of microbial communities. However, acquiring metaproteomic data by tandem mass spectrometry (MS/MS) is time-consuming and resource-intensive, and there is a demand for computational methods that can be used to reduce these resource requirements. We present MetaProClust-MS1, a computational framework for microbiome feature screening developed to prioritize samples for follow-up MS/MS. In this proof-of-concept study, we tested and compared MetaProClust-MS1 results on gut microbiome data, from fecal samples, acquired using short 15-min MS1-only chromatographic gradients and MS1 spectra from longer 60-min gradients to MS/MS-acquired data. We found that MetaProClust-MS1 identified robust gut microbiome responses caused by xenobiotics with significantly correlated cluster topologies of comparable data sets. We also used MetaProClust-MS1 to reanalyze data from both a clinical MS/MS diagnostic study of pediatric patients with inflammatory bowel disease and an experiment evaluating the therapeutic effects of a small molecule on the brain tissue of Alzheimer's disease mouse models. MetaProClust-MS1 clusters could distinguish between inflammatory bowel disease diagnoses (ulcerative colitis and Crohn's disease) using samples from mucosal luminal interface samples and identified hippocampal proteome shifts of Alzheimer's disease mouse models after small-molecule treatment. Therefore, we demonstrate that MetaProClust-MS1 can screen both microbiomes and single-species proteomes using only MS1 profiles, and our results suggest that this approach may be generalizable to any proteomics experiment. MetaProClust-MS1 may be especially useful for large-scale metaproteomic screening for the prioritization of samples for further metaproteomic characterization, using MS/MS, for instance, in addition to being a promising novel approach for clinical diagnostic screening. IMPORTANCE Growing evidence suggests that human gut microbiome composition and function are highly associated with health and disease. As such, high-throughput metaproteomic studies are becoming more common in gut microbiome research. However, using a conventional long liquid chromatography (LC)-MS/MS gradient metaproteomics approach as an initial screen in large-scale microbiome experiments can be slow and expensive. To combat this challenge, we introduce MetaProClust-MS1, a computational framework for microbiome screening using MS1-only profiles. In this proof-of-concept study, we show that MetaProClust-MS1 identifies clusters of gut microbiome treatments using MS1-only profiles similar to those identified using MS/MS. Our approach allows researchers to prioritize samples and treatments of interest for further metaproteomic analyses and may be generally applicable to any proteomic analysis. In particular, this approach may be especially useful for large-scale metaproteomic screening or in clinical settings where rapid diagnostic evidence is required.

STRATEGIES AND CHALLENGES IN METHOD DEVELOPMENT AND VALIDATION FOR THE ABSOLUTE QUANTIFICATION OF ENDOGENOUS BIOMARKER METABOLITES USING LIQUID CHROMATOGRAPHY-TANDEM MASS SPECTROMETRY.

Metabolomics is a dynamically evolving field, with a major application in identifying biomarkers for drug development and personalized medicine. Numerous metabolomic studies have identified endogenous metabolites that, in principle, are eligible for translation to clinical practice. However, few metabolomic-derived biomarker candidates have been qualified by regulatory bodies for clinical applications. Such interruption in the biomarker qualification process can be largely attributed to various reasons including inappropriate study design and inadequate data to support the clinical utility of the biomarkers. In addition, the lack of robust assays for the routine quantification of candidate biomarkers has been suggested as a potential bottleneck in the biomarker qualification process. In fact, the nature of the endogenous metabolites precludes the application of the current validation guidelines for bioanalytical methods. As a result, there have been individual efforts in modifying existing guidelines and/or developing alternative approaches to facilitate method validation. In this review, three main challenges for method development and validation for endogenous metabolites are discussed, namely matrix effects evaluation, alternative analyte-free matrices, and the choice of internal standards (ISs). Some studies have modified the equations described by the European Medicines Agency for the evaluation of matrix effects. However, alternative strategies were also described; for instance, calibration curves can be generated in solvents and in biological samples and the slopes can be compared through ratios, relative standard deviation, or a modified Stufour suggested approaches while quantifying mainly endogenous metabolitesdent t-test. ISs, on the contrary, are diverse; in which seven different possible types, used in metabolomics-based studies, were identified in the literature. Each type has its advantages and limitations; however, isotope-labeled ISs and ISs created through isotope derivatization show superior performance. Finally, alternative matrices have been described and tested during method development and validation for the quantification of endogenous entities. These alternatives are discussed in detail, highlighting their advantages and shortcomings. The goal of this review is to compare, apprise, and debate current knowledge and practices in order to aid researchers and clinical scientists in developing robust assays needed during the qualification process of candidate metabolite biomarkers. © 2019 John Wiley & Sons Ltd. Mass Spec Rev.

Quantitative determination of potential urine biomarkers of respiratory illnesses using new targeted metabolomic approach.

The diagnosis of asthma and chronic obstructive pulmonary disease (COPD) can be challenging due to the overlap in their clinical presentations in some patients. There is a need for a more objective clinical test that can be routinely used in primary care settings. Through an untargeted 1H NMR urine metabolomic approach, we identified a set of endogenous metabolites as potential biomarkers for the differentiation of asthma and COPD. A subset of these potential biomarkers contains 7 highly polar metabolites of diverse physicochemical properties. To the best of our knowledge, there is no liquid chromatography-tandem mass spectrometry (LC-MS/MS) method that evaluated more than two of the target metabolites in a single analytical run. The target metabolites belong to the families of monosaccharides, organic acids, amino acids, quaternary ammonium compounds and nucleic acids, rendering hydrophilic interaction liquid chromatography (HILIC) an ideal technology for their quantification. Since a clinical decision is to be made from patients data, a fully validated analytical method is required for biomarker validation. Method validation for endogenous metabolites is a daunting task since current guidelines were designed for exogenous compounds. As such, innovative approaches were adopted to meet the validation requirements. Herein, we describe a sensitive HILIC-MS/MS method for the quantification of the 7 endogenous urinary metabolites. Detection was achieved in the multiple reaction monitoring (MRM) mode with polarity switching, using quadrupole-linear ion trap instrument (QTRAP 6500) as well as single ion monitoring in the negative-ion mode. The method was fully validated according to the regulatory guidelines. Linearity was established between 6 and 21000 ng/mL and quality control samples demonstrated acceptable intra- and inter-day accuracy (85.7%-112%), intra- and inter-day precision (CV% <11.5%) as well as stability under various storage and sample processing conditions. To illustrate the method's applicability, the validated method was applied to the analysis of a small set of urine samples collected from asthma and COPD patients. Preliminary modelling of separation was generated using partial least square discriminant analysis (R2 0.752 and Q2 0.57). The adequate separation between patient samples confirms the diagnostic potential of these target metabolites as a proof-of-concept for the differentiation between asthma and COPD. However, more patient urine samples are needed in order to increase the statistical power of the analytical model.

Comparison of accuracy and precision between multipoint calibration, single point calibration, and relative quantification for targeted metabolomic analysis.

Targeted metabolomics requires accurate and precise quantification of candidate biomarkers, often through tandem mass spectrometric (MS/MS) analysis. Differential isotope labeling (DIL) improves mass spectrometric (MS) analysis in metabolomics by derivatizing metabolites with two isotopic forms of the same reagent. Despite its advantages, DIL-liquid chromatographic (LC)-MS/MS can result in substantial increase in workload when fully validated quantitative methods are required. To decrease the workload, we hypothesized that single point calibration or relative quantification could be used as alternative methods. Either approach will result in significant saving in resources and time. To test our hypothesis, six urinary metabolites were selected as model compounds. Urine samples were analyzed using a fully validated multipoint dansyl chloride-DIL-LC-MS/MS method. Samples were reprocessed using single point calibration and relative quantification modes. Our results demonstrated that the performance of single point calibration or relative quantification was inferior, for some metabolites, to multipoint calibration. The lower limit of quantification failed in the quantification of ethanolamine in most of participant samples using single point calibration. In addition, its precision was not acceptable in one participant during serine and ethanolamine quantification. On the other hand, relative quantification resulted in the least accurate data. In fact, none of the data generated from relative quantification for serine was comparable to that obtained from multipoint calibration. Finally, while single point calibration showed an overall acceptable performance for the majority of the model compounds, we cannot extrapolate the findings to other metabolites within the same analytical run. Analysts are advised to assess accuracy and precision for each metabolite in which single point calibration is the intended quantification mean.

Comparative analysis of creatinine and osmolality as urine normalization strategies in targeted metabolomics for the differential diagnosis of asthma and COPD.

INTRODUCTION: Urine is an ideal matrix for metabolomics investigation due to its non-invasive nature of collection and its rich metabolite content. Despite the advancements in mass spectrometry and 1H-NMR platforms in urine metabolomics, the statistical analysis of the generated data is challenged with the need to adjust for the hydration status of the person. Normalization to creatinine or osmolality values are the most adopted strategies, however, each technique has its challenges that can hinder its wider application. We have been developing targeted urine metabolomic methods to differentiate two important respiratory diseases, namely asthma and chronic obstructive pulmonary disease (COPD). OBJECTIVE: To assess whether the statistical model of separation of diseases using targeted metabolomic data would be improved by normalization to osmolality instead of creatinine. METHODS: The concentration of 32 metabolites was previously measured by two liquid chromatography-tandem mass spectrometry methods in 51 human urine samples with either asthma (n = 25) or COPD (n = 26). The data was normalized to creatinine or osmolality. Statistical analysis of the normalized values in each disease was performed using partial least square discriminant analysis (PLS-DA). Models of separation of diseases were compared. RESULTS: We found that normalization to creatinine or osmolality did not significantly change the PLS-DA models of separation (R2Q2 = 0.919, 0.705 vs R2Q2 = 0.929, 0.671, respectively). The metabolites of importance in the models remained similar for both normalization methods. CONCLUSION: Our findings suggest that targeted urine metabolomic data can be normalized for hydration using creatinine or osmolality with no significant impact on the diagnostic accuracy of the model.

Development of a validated LC- MS/MS method for the quantification of 19 endogenous asthma/COPD potential urinary biomarkers.

Obstructive airways inflammatory diseases sometimes show overlapping symptoms that hinder their early and correct diagnosis. Current clinical tests are tedious and are of inadequate specificity in special population such as the elderly and children. Therefore, we are developing tandem mass spectrometric (MS/MS) methods for targeted analysis of urine biomarkers. Recently, proton-nuclear magnetic resonance (1H-NMR) analysis proposed 50 urinary metabolites as potential diagnostic biomarkers among asthma and chronic obstructive pulmonary disease (COPD) patients. Metabolites are divided into 3 groups based on chemical nature. For group 1 (amines and phenols, 19 urinary metabolites), we developed and validated a high performance liquid chromatographic (HPLC)-MS/MS method using differential isotope labeling (DIL) with dansyl chloride. Method development included the optimization of the derivatization reaction, the MS/MS conditions, and the chromatographic separation. Linearity varied from 2 to 4800 ng/mL and the use of 13C2-labeled derivatives allowed for the correction of matrix effects as well as the unambiguous confirmation of the identity of each metabolite in the presence of interfering isomers in urine. Despite the challenges associated with method validation, the method was fully validated as per the food and drug administration (FDA) and the European medicines agency (EMA) recommendations. Validation criteria included linearity, precision, accuracy, dilution integrity, selectivity, carryover, and stability. Challenges in selectivity experiments included the isotopic contributions of the analyte towards its internal standard (IS), that was addressed via the optimization of the IS concentration. In addition, incurred sample analysis was performed to ensure that results from patient samples are accurate and reliable. The method was robust and reproducible and is currently being applied in a cohort of asthma and COPD patient urine samples for biomarker discovery purposes.

Mass spectrometric based approaches in urine metabolomics and biomarker discovery.

Urine metabolomics has recently emerged as a prominent field for the discovery of non-invasive biomarkers that can detect subtle metabolic discrepancies in response to a specific disease or therapeutic intervention. Urine, compared to other biofluids, is characterized by its ease of collection, richness in metabolites and its ability to reflect imbalances of all biochemical pathways within the body. Following urine collection for metabolomic analysis, samples must be immediately frozen to quench any biogenic and/or non-biogenic chemical reactions. According to the aim of the experiment; sample preparation can vary from simple procedures such as filtration to more specific extraction protocols such as liquid-liquid extraction. Due to the lack of comprehensive studies on urine metabolome stability, higher storage temperatures (i.e. 4°C) and repetitive freeze-thaw cycles should be avoided. To date, among all analytical techniques, mass spectrometry (MS) provides the best sensitivity, selectivity and identification capabilities to analyze the majority of the metabolite composition in the urine. Combined with the qualitative and quantitative capabilities of MS, and due to the continuous improvements in its related technologies (i.e. ultra high-performance liquid chromatography [UPLC] and hydrophilic interaction liquid chromatography [HILIC]), liquid chromatography (LC)-MS is unequivocally the most utilized and the most informative analytical tool employed in urine metabolomics. Furthermore, differential isotope tagging techniques has provided a solution to ion suppression from urine matrix thus allowing for quantitative analysis. In addition to LC-MS, other MS-based technologies have been utilized in urine metabolomics. These include direct injection (infusion)-MS, capillary electrophoresis-MS and gas chromatography-MS. In this article, the current progresses of different MS-based techniques in exploring the urine metabolome as well as the recent findings in providing potentially diagnostic urinary biomarkers are discussed. © 2015 Wiley Periodicals, Inc. Mass Spec Rev 36:115-134, 2017.

Analysis of a series of chlorogenic acid isomers using differential ion mobility and tandem mass spectrometry.

Chlorogenic acids are among the most abundant phenolics found in the human diet. Of these, the mono-caffeoylquinic acids are the predominant phenolics found in fruits, such as apples and pears, and products derived from them. In this research, a comprehensive study of the electrospray ionization (ESI) tandem mass spectrometric (MS/MS) dissociation behavior of the three most common mono-caffeoylquinic acids, namely 5-O-caffeoylquinic acid (5-CQA), 3-O-caffeoylquinic acid (3-CQA) and 4-O-caffeoylquinic acid (4-CQA), were determined using both positive and negative ionization. All proposed structures of the observed product ions were confirmed with second-generation MS(3) experiments. Similarities and differences between the dissociation pathways in the positive and negative ion modes are discussed, confirming the proposed structures and the established MS/MS fingerprints. MS/MS dissociation was primarily driven via the cleavage of the ester bond linking the quinic acid moiety to the caffeic acid moiety within tested molecules. Despite being structural isomers with the same m/z values and dissociation behaviors, the MS/MS data in the negative ion mode was able to differentiate the three isomers based on ion intensity for the major product ions, observed at m/z 191, 179 and 173. This differentiation was consistent among various MS instruments. In addition, ESI coupled with high-field asymmetric waveform ion mobility spectrometry-mass spectrometry (ESI-FAIMS-MS) was employed for the separation of these compounds for the first time. By combining MS/MS data and differential ion mobility, a method for the separation and identification of mono-caffeoylquinic in apple/pear juice samples was developed with a run time of less than 1 min. It is envisaged that this methodology could be used to identify pure juices based on their chlorogenic acid profile (i.e., metabolomics), and could also be used to detect juice-to-juice adulteration (e.g., apple juice addition to pear juice).

Validated spectrofluorimetric determination of two pharmaceutical antihypertensive mixtures containing amlodipine besylate together with either candesartan cilexetil or telmisartan.

Amlodipine besylate (AML) is available in fixed-dose combination tablets with either candesartan cilexetil (CAN) or telmisartan (TEL). This work describes a simple, selective and sensitive spectrofluorimetric method for analysis of AML/CAN and AML/TEL binary mixtures without prior separation. The method involves measurement of the native fluorescence of AML at excitation and emission wavelengths of 367 and 454 nm, respectively, in water without interference from either of the two drugs. By contrast, the intrinsic fluorescence of CAN was measured at excitation and emission wavelengths of 265 and 392 nm, respectively, in ethanol, while TEL was measured at 366 nm in 0.05 M sodium hydroxide solution using 294 nm as the excitation wavelength. The proposed spectrofluorimetric procedure was validated with respect to linearity, ranges, precision, accuracy, selectivity, robustness, detection and quantification limits. Regression analysis showed a good correlation between fluorescence intensity and concentration over the ranges 0.1-1.4, 0.025-0.25 and 0.0025-0.05 µg/mL for AML, CAN and TEL, respectively. Limits of detection were 0.034, 0.0063 and 0.0007 µg/mL for AML, CAN and TEL, respectively. The proposed method was successfully applied for the analysis of several synthetic binary mixtures of different ratios and laboratory-prepared tablets with good recoveries, and no interference from common pharmaceutical additives was observed.

New simple spectrophotometric method for determination of the binary mixtures (atorvastatin calcium and ezetimibe; candesartan cilexetil and hydrochlorothiazide) in tablets.

A new simple spectrophotometric method was developed for the determination of binary mixtures without prior separation. The method is based on the generation of ratio spectra of compound X by using a standard spectrum of compound Y as a divisor. The peak to trough amplitudes between two selected wavelengths in the ratio spectra are proportional to concentration of X without interference from Y. The method was demonstrated by determination of two drug combinations. The first consists of the two antihyperlipidemics: atorvastatin calcium (ATV) and ezetimibe (EZE), and the second comprises the antihypertensives: candesartan cilexetil (CAN) and hydrochlorothiazide (HCT). For mixture 1, ATV was determined using 10 µg/mL EZE as the divisor to generate the ratio spectra, and the peak to trough amplitudes between 231 and 276 nm were plotted against ATV concentration. Similarly, by using 10 µg/mL ATV as divisor, the peak to trough amplitudes between 231 and 276 nm were found proportional to EZE concentration. Calibration curves were linear in the range 2.5-40 µg/mL for both drugs. For mixture 2, divisor concentration was 7.5 µg/mL for both drugs. CAN was determined using its peak to trough amplitudes at 251 and 277 nm, while HCT was estimated using the amplitudes between 251 and 276 nm. The measured amplitudes were linearly correlated to concentration in the ranges 2.5-50 and 1-30 µg/mL for CAN and HCT, respectively. The proposed spectrophotometric method was validated and successfully applied for the assay of both drug combinations in several laboratory-prepared mixtures and commercial tablets.