A Predictive Compound Database Approach to the Tentative Identification and Semiquantitation of Volatile-Phenol Glycosides in Smoke‑Affected Grapes from Wildfires

Importance of the Topic

Volatile-phenol glycosides are odorless compounds that, upon hydrolysis during winemaking, release smoky aromas leading to significant wine quality defects known as smoke taint affecting wine composition and consumer acceptance.
The increasing occurrence of wildfires introduces volatile phenols into grape berries, highlighting the importance of comprehensive screening and monitoring strategies for these precursors.

Objectives and Study Overview

This study introduces an untargeted LC-MS workflow using a predictive compound database to tentatively identify and semiquantitate volatile-phenol glycosides in smoke-affected grapes and throughout winemaking. The approach aims to survey over 600 potential glycosides, compare ESI and APCI ion sources, and report the first evidence of trisaccharide glycosides in grapes.

Methodology and Instrumentation

Grape and winemaking samples were extracted by solid-phase extraction and spiked with salicin as an internal standard. An Agilent 1290 Infinity II LC coupled to a 6545 Accurate-Mass Q-TOF was used for analysis. A user-generated personal compound database guided untargeted screening against predicted structures. Electrospray ionization with an all ions fragmentation strategy was selected for final analyses. MSMS data were acquired and processed by Molecular Structure Correlator for structural elucidation. Semiquantitation was performed using MassHunter Quantitative Analysis with automatic peak integration and normalization to salicin.

Instrumental Setup

Liquid chromatography used a Poroshell 120 Phenyl Hexyl column (2.1 x 150 mm, 2.7 micron) at 40 C with 0.1 percent acetic acid in water and acetonitrile, a gradient from 5 to 95 percent organic over 18 minutes, and a flow rate of 0.42 mL per minute. The Q-TOF operated in negative mode with a mass range of 100 to 1000 m/z, MS acquisition at 3 spectra per second, MSMS fragmentation at collision energies of 0 and 20 eV, and continuous reference mass correction. ESI source parameters included drying gas at 150 C, sheath gas at 350 C, nebulizer at 35 psig, fragmentor at 120 V, and capillary at 3500 V. APCI was evaluated but showed reduced sensitivity for high-mass glycosides.

Main Results and Discussion

A total of 31 volatile-phenol glycosides were tentatively identified, including the first report of trisaccharide forms in grapes. Comparison of ESI and APCI demonstrated that ESI provided equal or higher sensitivity across mono-, di- and trisaccharides. Semiquantitative data revealed that about 75 percent of glycosides persisted into the finished wine, with most hydrolysis occurring during primary fermentation. The all ions workflow improved peak confirmation and minimized false positives for low-abundance compounds.

Benefits and Practical Applications

The untargeted database approach enables rapid screening of hundreds of glycoside precursors without the need for standards. Automation across MassHunter applications accelerates data acquisition and analysis, aiding research and quality control laboratories in assessing smoke taint risk and evaluating mitigation strategies.

Future Trends and Applications

Future developments may include expansion of spectral libraries with experimental standards, integration of machine learning for data interpretation, real-time monitoring of taint compounds, and adaptation of the workflow to other matrices such as juices and spirits.

Conclusion

The described LC-MS workflow with a predictive compound database effectively identifies and semiquantifies volatile-phenol glycosides in smoke-affected grapes and wine. Electrospray ionization with an all ions fragmentation strategy provided broad sensitivity, enabling the first evidence of trisaccharide glycosides. This method offers a robust tool for smoke taint research and quality assurance in enology.

References

1. Caffrey A; Lerno L; Rumbaugh A; Girardello R; Oberholster A; Ebeler SE. Changes in smoke-taint volatile phenol glycosides in wildfire smoke-exposed cabernet sauvignon grapes throughout winemaking. Amer J Enol Vitic 2019;70(4).
2. Parker M; Powers J; Soleas GJ; Jeffery DW. Contribution of volatile phenols and their glycoconjugates to smoke-related sensory properties of red wine. J Agric Food Chem 2012;60:2629-2637.
3. Kennison KR; Gump BJ; Pambianchi ER; Markides AL. Effect of timing and duration of grapevine exposure to smoke on composition and sensory properties of wine. Aust J Grape Wine Res 2009;15:228-237.
4. Hayasaka Y; Capone D; Jeffery DW. Investigation into formation of guaiacol conjugates in berries and leaves of cabernet sauvignon using HPLC-MS and MSMS. J Agric Food Chem 2010;58:2076-2084.
5. Noestheden M; Ristic R; Kolling K; Hayasaka Y; Jeffery DW. Detailed characterization of glycosylated sensory-active volatile phenols in smoke-exposed grapes and wine. Food Chem 2018;259:147-156.
6. Caffrey A; Hjelmeland A; Zweigenbaum J; Ebeler SE. Identification and semiquantitation of monoterpene glycosides in ripening muscat of alexandria grapes. Agilent Technologies Application Note 5991-9448EN, 2018.
7. Hayasaka Y; Maggs B; Jeffery DW. Assessing impact of smoke exposure in grapes: development and validation of an HPLC-MSMS method for quantitative analysis of smoke-derived phenolic glycosides in grapes and wine. J Agric Food Chem 2013;61(1):25-33.