Integrating analytical resolutions in non-targeted wine metabolomics

High-Resolution Mass Spectrometry-Based Metabolomics for Increased Grape Juice Metabolite Coverage

The composition of the juice from grape berries is at the basis of the definition of technological ripeness before harvest, historically evaluated from global sugar and acid contents. If many studies have contributed to the identification of other primary and secondary metabolites in whole berries, deepening knowledge about the chemical composition of the sole flesh of grape berries (i.e., without considering skins and seeds) at harvest is of primary interest when studying the enological potential of widespread grape varieties producing high-added-value wines. Here, we used non-targeted DI-FT-ICR-MS and RP-UHPLC-Q-ToF-MS analyses to explore the extent of metabolite coverage of up to 290 grape juices from four Vitis vinifera grape varieties, namely Chardonnay, Pinot noir, Meunier, and Aligoté, sampled at harvest from 91 vineyards in Europe and Argentina, over three successive vintages. SPE pretreatment of samples led to the identification of more than 4500 detected C,H,O,N,S-containing elemental compositions, likely associated with tens of thousands of distinct metabolites. We further revealed that a major part of this chemical diversity appears to be common to the different juices, as exemplified by Pinot noir and Chardonnay samples. However, it was possible to build significant models for the discrimination of Chardonnay from Pinot noir grape juices, and of Chardonnay from Aligoté grape juices, regardless of the geographical origin or the vintage. Therefore, this metabolomic approach opens access to a remarkable holistic molecular description of the instantaneous composition of such a biological matrix, which is the result of complex interplays among environmental, biochemical, and vine growing practices.

Direct introduction MALDI FTICR MS based on dried droplet deposition applied to non-targeted metabolomics on Pisum Sativum root exudates

Non-targeted metabolomic approaches based on direct introduction (DI) through a soft ionization source are nowadays used for large-scale analysis and wide cover-up of metabolites in complex matrices. When coupled with ultra-high-resolution Fourier-Transform ion cyclotron resonance (FTICR MS), DI is generally performed through electrospray (ESI), which, despite the great analytical throughput, can suffer of matrix effects due to residual salts or charge competitors. In alternative, matrix assisted laser desorption ionization (MALDI) coupled with FTICR MS offers relatively high salt tolerance but it is mainly used for imaging of small molecule within biological tissues. In this study, we report a systematic evaluation on the performance of direct introduction ESI and MALDI coupled with FTICR MS applied to the analysis of root exudates (RE), a complex mixture of metabolites released from plant root tips and containing a relatively high salt concentration. Classic dried droplet deposition followed by screening of best matrices and ratio allowed the selection of high ranked conditions for non-targeted metabolomics on RE. Optimization of MALDI parameters led to improved reproducibility and precision. A RE desalted sample was used for comparison on ionization efficiency of the two sources and ion enhancement at high salinity was highlighted in MALDI by spiking desalted solution with inorganic salts. Application of a true lyophilized RE sample exhibited the complementarity of the two sources and the ability of MALDI in the detection of undisclosed metabolites suffering of matrix effects in ESI mode.

Untargeted metabolomic analysis by ultra-high-resolution mass spectrometry for the profiling of new Italian wine varieties

The chemical composition of wine samples comprises numerous bioactive compounds responsible for unique flavor and health-promoting properties. Thus, it's important to have a complete overview of the metabolic profile of new wine products in order to obtain peculiar information in terms of their phytochemical composition, quality, and traceability. To achieve this aim, in this work, a mass spectrometry-based phytochemical screening was performed on seven new wine products from Villa D'Agri in the Basilicata region (Italy), i.e., Aglianico Bianco, Plavina, Guisana, Giosana, Malvasia ad acino piccolo, Colata Murro and Santa Sofia. Ultra-high-resolution mass spectrometry data were processed into absorption mode FT-ICR mass spectra, in order to remove artifacts and achieve a higher resolution and lower levels of noise. Accurate mass-to-charge ratio (m/z) values were converted into putative elemental formulas. Therefore, 2D van Krevelen diagrams were used as a tool to obtain molecular formula maps useful to perform a rapid and more comprehensive analysis of the wine sample metabolome. The presence of important metabolite classes, i.e., fatty acid derivatives, amino acids and peptides, carbohydrates and phenolic derivatives, was assessed. Moreover, the comparison of obtained metabolomic maps revealed some differences among profiles, suggesting their employment as metabolic fingerprints. This study shed some light on the metabolic composition of seven new Italian wine varieties, improving their value in terms of related bioactive compound content. Moreover, different metabolomic fingerprints were obtained for each of them, suggesting the use of molecular maps as innovative tool to ascertain their unique metabolic profile.

Microbial Interactions in Kombucha through the Lens of Metabolomics

Kombucha is a fermented beverage obtained through the activity of a complex microbial community of yeasts and bacteria. Exo-metabolomes of kombucha microorganisms were analyzed using FT-ICR-MS to investigate their interactions. A simplified set of microorganisms including two yeasts (Brettanomyces bruxellensis and Hanseniaspora valbyensis) and one acetic acid bacterium (Acetobacter indonesiensis) was used to investigate yeast-yeast and yeast-acetic acid bacterium interactions. A yeast-yeast interaction was characterized by the release and consumption of fatty acids and peptides, possibly in relationship to commensalism. A yeast-acetic acid bacterium interaction was different depending on yeast species. With B. bruxellensis, fatty acids and peptides were mainly produced along with consumption of sucrose, fatty acids and polysaccharides. In opposition, the presence of H. valbyensis induced mainly the decrease of polyphenols, peptides, fatty acids, phenolic acids and putative isopropyl malate and phenylpyruvate and few formulae have been produced. With all three microorganisms, the formulae involved with the yeast-yeast interactions were consumed or not produced in the presence of A. indonesiensis. The impact of the yeasts' presence on A. indonesiensis was consistent regardless of the yeast species with a commensal consumption of compounds associated to the acetic acid bacterium by yeasts. In detail, hydroxystearate from yeasts and dehydroquinate from A. indonesiensis were potentially consumed in all cases of yeast(s)-acetic acid bacterium pairing, highlighting mutualistic behavior.

Non-Targeted Metabolomic Analysis of the Kombucha Production Process

Kombucha is a traditional fermented beverage obtained from the transformation of sugared black tea by a community of yeasts and bacteria. Kombucha production recently became industrialized, but its quality standards remain poorly defined. Metabolomic analyses were applied using FT-ICR-MS to characterize the impacts of production phases and the type of tea on the non-volatile chemical composition of kombucha. Independently from tea type, the first phase of acidification in open vessel was characterized by the release of gluconate and gallate from acetic acid bacteria metabolism and probably from polymeric polyphenols, respectively. The second phase of carbonation in closed vessel induced a consumption or transformation of oleic acid that could be consecutive of oxygen limitation. The first phase had the most impact on molecular diversity, but tea type mainly influenced the global composition in polyphenol profile. Black tea polyphenols were more impacted by microbial activity compared to green tea polyphenols.

Different Wines from Different Yeasts? "Saccharomyces cerevisiae Intraspecies Differentiation by Metabolomic Signature and Sensory Patterns in Wine"

Alcoholic fermentation is known to be a key stage in the winemaking process that directly impacts the composition and quality of the final product. Twelve wines were obtained from fermentations of Chardonnay must made with twelve different commercial wine yeast strains of Saccharomyces cerevisiae. In our study, FT-ICR-MS, GC-MS, and sensory analysis were combined with multivariate analysis. Ultra-high-resolution mass spectrometry (uHRMS) was able to highlight hundreds of metabolites specific to each strain from the same species, although they are characterized by the same technological performances. Furthermore, the significant involvement of nitrogen metabolism in this differentiation was considered. The modulation of primary metabolism was also noted at the volatilome and sensory levels. Sensory analysis allowed us to classify wines into three groups based on descriptors associated with white wine. Thirty-five of the volatile compounds analyzed, including esters, medium-chain fatty acids, superior alcohols, and terpenes discriminate and give details about differences between wines. Therefore, phenotypic differences within the same species revealed metabolic differences that resulted in the diversity of the volatile fraction that participates in the palette of the sensory pattern. This original combination of metabolomics with the volatilome and sensory approaches provides an integrative vision of the characteristics of a given strain. Metabolomics shine the new light on intraspecific discrimination in the Saccharomyces cerevisiae species.

On the Trail of the German Purity Law: Distinguishing the Metabolic Signatures of Wheat, Corn and Rice in Beer

Here, we report a non-targeted analytical approach to investigate the influence of different starch sources on the metabolic signature in the final beer product. An extensive sample set of commercial beers brewed with barley, wheat, corn and/or rice were analyzed by both direct infusion Fourier transform ion cyclotron mass spectrometry (DI-FTICR MS, 400 samples) and UPLC-ToF-MS (100 samples). By its unrivaled mass resolution and accuracy, DI-FTICR-MS was able to uncover the compositional space of both polar and non-polar metabolites that can be traced back to the use of different starch sources. Reversed phase UPLC-ToF-MS was used to access information about molecular structures (MS²-fragmentation spectra) and isomeric separation, with a focus on less polar compounds. Both analytical approaches were able to achieve a clear statistical differentiation (OPLS-DA) of beer samples and reveal metabolic profiles according to the starch source. A mass difference network analysis, applied to the exact marker masses resolved by FTICR, showed a network of potential secondary metabolites specific to wheat, corn and rice. By MS²-similarity networks, database and literature search, we were able to identify metabolites and compound classes significant for the use of the different starch sources. Those were also found in the corresponding brewing raw materials, confirming the potential of our approach for quality control and monitoring. Our results also include the identification of the aspartic acid-conjugate of N-β-D-glucopyranosyl-indole-3-acetic acid as a potential marker for the use of rice in the brewing industry regarding quality control and food inspection purposes.

Terroir and vintage discrimination of Malbec wines based on phenolic composition across multiple sites in Mendoza, Argentina

This study evaluated the phenolic profiles of Malbec wines made from grapes of 23 parcels distributed in 12 geographical indications (GIs) from Mendoza, Argentina. Wines were elaborated under standardized winemaking conditions over three consecutive vintages (2016–2018). Data discriminated wines from different GIs and parcels, based on an integrative data analysis by chemometric tools. Vintage effect and specific phenolic compounds were associated with some GIs or parcels. As well, regional climate conditions allowed partial discrimination of the GIs (and also some parcels). A random forest analysis correctly identified 11 out of 23 individual parcels across the different vintages. The most notorious compounds associated with such classification were p-coumaric acid, delphinidin-3-O-glucoside, caffeic acid, quercetin and peonidin-3-O-glucoside. The presented research allows to individualize, through phenolic profiles, parcels with unique characteristics over years. This is the first report characterizing Malbec wines coming from several GIs (and individual parcels) in different vintages. These results are strongly related to terroir features of wines, contributing to a better communication to consumers and to position Argentinean wines.

An overview of chemical contaminants and other undesirable chemicals in alcoholic beverages and strategies for analysis

The presence of chemical contaminant in alcoholic beverages is a widespread and notable problem with potential implications for human health. With the complexity and wide variation in the raw materials, production processes, and contact materials involved, there are a multitude of opportunities for a diverse host of undesirable compounds to make their way into the final product-some of which may currently remain unidentified and undetected. This review provides an overview of the notable contaminants (including pesticides, environmental contaminants, mycotoxins, process-induced contaminants, residues of food contact material [FCM], and illegal additives) that have been detected in alcoholic products thus far based on prior reviews and findings in the literature, and will additionally consider the potential sources for contamination, and finally discuss and identify gaps in current analytical strategies. The findings of this review highlight a need for further investigation into unwanted substances in alcoholic beverages, particularly concerning chemical migrants from FCMs, as well as a need for comprehensive nontargeted analytical techniques capable of determining unanticipated contaminants.

Yeast-Yeast Interactions: Mechanisms, Methodologies and Impact on Composition

During the winemaking process, alcoholic fermentation is carried out by a consortium of yeasts in which interactions occurs. The consequences of these interactions on the wine matrix have been widely described for several years with the aim of controlling the winemaking process as well as possible. In this review, we highlight the wide diversity of methodologies used to study these interactions, and their underlying mechanisms and consequences on the final wine composition and characteristics. The wide variety of matrix parameters, yeast couples, and culture conditions have led to contradictions between the results of the different studies considered. More recent aspects of modifications in the composition of the matrix are addressed through different approaches that have not been synthesized recently. Non-volatile and volatile metabolomics, as well as sensory analysis approaches are developed in this paper. The description of the matrix composition modification does not appear sufficient to explain interaction mechanisms, making it vital to take an integrated approach to draw definite conclusions on them.

Exploring yeast interactions through metabolic profiling

As a complex microbial ecosystem, wine is a particularly interesting model for studying interactions between microorganisms as fermentation can be done by microbial consortia, a unique strain or mixed culture. The effect of a specific yeast strain on its environments is unique and characterized by its metabolites and their concentration. With its great resolution and excellent mass accuracy, ultrahigh resolution mass spectrometry (uHRMS) is the perfect tool to analyze the yeast metabolome at the end of alcoholic fermentation. This work reports the change in wine chemical composition from pure and mixed culture fermentation with Lachancea thermotolerans, Starmerella bacillaris, Metschnikowia pulcherrima and S. cerevisiae. We could clearly differentiate wines according to the yeast strain used in single cultures and markers, which reflect important differences between the yeast species, were extracted and annotated. Moreover, uHRMS revealed underlining intra species metabolomics differences, showing differences at the strain level between the two Starmerella bacillaris. Non volatile metabolomics analysis of single and sequential fermentations confirmed that mixed fermentations lead to a different composition. Distinct metabolites appeared in wines from sequential fermentation compared to single fermentation. This suggests that interactions between yeasts are not neutral.

Wine aging: a bottleneck story

The sporadic oxidation of white wines remains an open question, making wine shelf life a subjective debate. Through a multidisciplinary synoptic approach performed as a remarkable case study on aged bottles of white wine, this work unraveled a yet unexplored route for uncontrolled oxidation. By combining sensory evaluation, chemical and metabolomics analyses of the wine, and investigating oxygen transfer through the bottleneck/stopper, this work elucidates the importance of the glass/cork interface. It shows unambiguously that the transfer of oxygen at the interface between the cork stopper and the glass bottleneck must be considered a potentially significant contributor to oxidation state during the bottle aging, leading to a notable modification of a wine’s chemical signature.

Influence of cell-cell contact between L. thermotolerans and S. cerevisiae on yeast interactions and the exo-metabolome

Sequential fermentation of grape must inoculated with L. thermotolerans and then S. cerevisiae 24 h later (typical wine-making practice) was conducted with or without cell-cell contact between the two yeast species. We monitored cell viability of the two species throughout fermentation by flow cytometry. The cell viability of S. cerevisiae decreased under both conditions, but the decrease was greater if there was cell-cell contact. An investigation of the nature of the interactions showed competition between the two species for nitrogen compounds, oxygen, and must sterols. Volatile-compound analysis showed differences between sequential and pure fermentation and that cell-cell contact modifies yeast metabolism, as the volatile-compound profile was significantly different from that of sequential fermentation without cell-cell contact. We further confirmed that cell-cell contact modifies yeast metabolism by analyzing the exo-metabolome of all fermentations by FT-ICR-MS analysis. These analyses show specific metabolite production and quantitative metabolite changes associated with each fermentation condition. This study shows that cell-cell contact not only affects cell viability, as already reported, but markedly affects yeast metabolism.

Juice Index: an integrated Sauvignon blanc grape and wine metabolomics database shows mainly seasonal differences

INTRODUCTION

Although Sauvignon Blanc (SB) grapes are cultivated widely throughout New Zealand, wines from the Marlborough region are most famous for their typical varietal combination of tropical and vegetal aromas. These wines differ in composition from season to season as well as among locations within the region, which makes the continual production of good quality wines challenging. Here, we developed a unique database of New Zealand SB grape juices and wines to develop tools to help winemakers to make blending decisions and assist in the development of new wine styles.

METHODS

About 400 juices were collected from different regions in New Zealand over three harvest seasons (2011-2013), which were then fermented under controlled conditions using a commercial yeast strain Saccharomyces cerevisiae EC1118. Comprehensive metabolite profiling of these juices and wines by gas chromatography-mass spectrometry (GC-MS) was combined with their detailed oenological parameters and associated meteorological data.

RESULTS

These combined metabolomics data clearly demonstrate that seasonal variation is more prominent than regional difference in both SB grape juices and wines, despite almost universal use of vineyard irrigation to mitigate seasonal rainfall and evapotranspiration differences, Additionally, we identified a group of juice metabolites that play central roles behind these variations, which may represent chemical signatures for juice and wine quality assessment.

CONCLUSION

This database is the first of its kind in the world to be available for the wider scientific community and offers potential as a predictive tool for wine quality and innovation when combined with mathematical modelling.

Wine and grape marc spirits metabolomics

INTRODUCTION

Mass spectrometry (MS)-based and nuclear magnetic resonance (NMR) spectroscopic analyses play a key role in the field of metabolomics due to their important advantages. The use of metabolomics in wine and grape marc spirits allows a more holistic perspective in monitoring and gaining information on the making processes and thus it can assist on the improvement of their quality.

OBJECTIVES

This review surveys the latest metabolomics approaches for wine and grape marc spirits with a focus on the description of MS-based and NMR spectroscopic analytical techniques.

METHODS

We reviewed the literature to identify metabolomic studies of wine and grape marc spirits that were published until the end of 2017, with the key term combinations of 'metabolomics', 'wine' and 'grape marc spirits'. Through the reference lists from these studies, additional articles were identified.

RESULTS

The results of this review showed that the application of different metabolomics approaches has significantly increased the knowledge of wine metabolome and grape marc spirits; however there is not yet a single analytical platform that can completely separate, detect and identify all metabolites in one analysis.

CONCLUSIONS

The authentication and quality control of wines and grape marc spirits has to be taken with caution, since the product's chemical composition could be affected by many factors. Despite intrinsic limitations, NMR spectroscopy and MS based strategies remain the key analytical methods in metabolomics studies. Authenticity, traceability and health issues related to their consumption are the major research initiatives in wine and grape marc spirits metabolomics analysis.

Grape and Wine Metabolomics to Develop New Insights Using Untargeted and Targeted Approaches

Chemical analysis of grape juice and wine has been performed for over 50 years in a targeted manner to determine a limited number of compounds using Gas Chromatography, Mass-Spectrometry (GC-MS) and High Pressure Liquid Chromatography (HPLC). Therefore, it only allowed the determination of metabolites that are present in high concentration, including major sugars, amino acids and some important carboxylic acids. Thus, the roles of many significant but less concentrated metabolites during wine making process are still not known. This is where metabolomics shows its enormous potential, mainly because of its capability in analyzing over 1000 metabolites in a single run due to the recent advancements of high resolution and sensitive analytical instruments. Metabolomics has predominantly been adopted by many wine scientists as a hypothesis-generating tool in an unbiased and non-targeted way to address various issues, including characterization of geographical origin (terroir) and wine yeast metabolic traits, determination of biomarkers for aroma compounds, and the monitoring of growth developments of grape vines and grapes. The aim of this review is to explore the published literature that made use of both targeted and untargeted metabolomics to study grapes and wines and also the fermentation process. In addition, insights are also provided into many other possible avenues where metabolomics shows tremendous potential as a question-driven approach in grape and wine research.

Characterization of Cultivar Differences of Blueberry Wines Using GC-QTOF-MS and Metabolic Profiling Methods

A non-targeted volatile metabolomic approach based on the gas chromatography-quadrupole time of fight-mass spectrometry (GC-QTOF-MS) coupled with two different sample extraction techniques (solid phase extraction and solid phase microextraction) was developed. Combined mass spectra of blueberry wine samples, which originated from two different cultivars, were subjected to orthogonal partial least squares-discriminant analysis (OPLS-DA). Principal component analysis (PCA) reveals an excellent separation and OPLS-DA highlight metabolic features responsible for the separation. Metabolic features responsible for the observed separation were tentatively assigned to phenylethyl alcohol, cinnamyl alcohol, benzenepropanol, 3-hydroxy-benzenethanol, methyl eugenol, methyl isoeugenol, (E)-asarone, (Z)-asarone, and terpenes. Several of the selected markers enabled a distinction in secondary metabolism to be drawn between two blueberry cultivars. It highlights the metabolomic approaches to find out the influence of blueberry cultivar on a volatile composition in a complex blueberry wine matrix. The distinction in secondary metabolism indicated a possible O-methyltransferases activity difference among the two cultivars.

Usage of FT-ICR-MS Metabolomics for Characterizing the Chemical Signatures of Barrel-Aged Whisky

Whisky can be described as a complex matrix integrating the chemical history from the fermented cereals, the wooden barrels, the specific distillery processes, aging, and environmental factors. In this study, using Fourier transform ion cyclotron resonance mass spectrometry (FT-ICR-MS) and liquid chromatography coupled with tandem mass spectrometry (LC-MS/MS), we analyzed 150 whisky samples from 49 different distilleries, 7 countries, and ranging from 1 day new make spirit to 43 years of maturation with different types of barrel. Chemometrics revealed the unexpected impact of the wood history on the distillate's composition during barrel aging, regardless of the whisky origin. Flavonols, oligolignols, and fatty acids are examples of important chemical signatures for Bourbon casks, whereas a high number of polyphenol glycosides, including for instance quercetin-glucuronide or myricetin-glucoside as potential candidates, and carbohydrates would discriminate Sherry casks. However, the comparison of barrel aged rums and whiskies revealed specific signatures, highlighting the importance of the initial composition of the distillate and the distillery processes.

The impact of SO2 on wine flavanols and indoles in relation to wine style and age

Wine has one of the broadest chemical profiles, and the common oenological practice of adding the antioxidant and antimicrobial sulfur dioxide has a major impact on its metabolomic fingerprint. In this study, we investigated novel discovered oenological reactions primarily occurring between wine metabolites and sulfur dioxide. The sulfonated derivatives of epicatechin, procyanidin B2, indole acetic acid, indole lactic acid and tryptophol were synthesized and for the first time quantified in wine. Analysis of 32 metabolites in 195 commercial wines (1986-2016 vintages) suggested that sulfonation of tryptophan metabolites characterised white wines, in contrast to red wines, where sulfonation of flavanols was preferred. The chemical profile of the oldest wines was strongly characterised by sulfonated flavanols and indoles, indicating that could be fundamental metabolites in explaining quality in both red and white aged wines. These findings offer new prospects for more precise use of sulfur dioxide in winemaking.

A GC-MS based metabolic profiling of fermented sausage supplemented with pineapple

A GC-MS based metabolomic study was performed to understand metabolic changes during sausage fermentation and to investigate how the incorporation of pineapple affects the metabolic profiles of fermented sausages. Principal component analysis models showed clear metabolic differences among the fermented sausages according to the fermentation periods and the pineapple addition. Increased amounts of amino acids and organic acids except for citric acid, along with decreased levels of sugars were observed after fermentation. Higher levels of sugars and citric acid in the pineapple supplemented sausages dramatically decreased during the early stage of fermentation. The contents of lactic acid, phosphoric acid, succinic acid, ribonic acid, valine, leucine, isoleucine, glycine, threonine, glutamic acid, glucose, and sucrose were significantly increased in the 2% pineapple addition sausages. GC-MS and PCA analytical methods provide a new approach to understand of the metabolic changes in fermented sausages during fermentation.

Applications of Fourier Transform Ion Cyclotron Resonance (FT-ICR) and Orbitrap Based High Resolution Mass Spectrometry in Metabolomics and Lipidomics

Metabolomics, along with other "omics" approaches, is rapidly becoming one of the major approaches aimed at understanding the organization and dynamics of metabolic networks. Mass spectrometry is often a technique of choice for metabolomics studies due to its high sensitivity, reproducibility and wide dynamic range. High resolution mass spectrometry (HRMS) is a widely practiced technique in analytical and bioanalytical sciences. It offers exceptionally high resolution and the highest degree of structural confirmation. Many metabolomics studies have been conducted using HRMS over the past decade. In this review, we will explore the latest developments in Fourier transform mass spectrometry (FTMS) and Orbitrap based metabolomics technology, its advantages and drawbacks for using in metabolomics and lipidomics studies, and development of novel approaches for processing HRMS data.

Wine metabolomics reveals new sulfonated products in bottled white wines, promoted by small amounts of oxygen

The impact of minute amounts of oxygen in the headspace on the post-bottling development of wine is generally considered to be very important, since oxygen can either damage or improve the quality of wine. This project aimed to gain new experimental evidence about the chemistry of the interaction between wine and oxygen. The experimental design included 216 bottles of 12 different white wines produced from 6 different cultivars (Inzolia, Muller Thurgau, Chardonnay, Grillo, Traminer and Pinot gris). Half of them were bottled using the standard industrial process with inert headspace and the other half without inert gas and with extra headspace. After 60 days of storage at room temperature, the wines were analysed using an untargeted LC-MS method. The use of a detailed holistic analysis workflow, with several levels of quality control and marker selection, gave 35 metabolites putatively induced by the different amounts of oxygen. These metabolite markers included ascorbic acid, tartaric acid and various sulfonated compounds observed in wine for the first time (e.g. S-sulfonated cysteine, glutathione and pantetheine; and sulfonated indole-3-lactic acid hexoside and tryptophol). The consumption of SO2 mediated by these sulfonation reactions was promoted by the presence of higher levels of oxygen on bottling.