The Pennsylvania State University the Graduate School College Of

The Pennsylvania State University

The Graduate School

College of Agricultural Sciences

STUDIES ON THE REACTION OF WINE FLAVONOIDS

WITH EXOGENOUS ACETALDEHYDE

A Dissertation in

Food Science

Marlena K. Sheridan

Submitted in Partial Fulfillment of the Requirements for the Degree of

Doctor of Philosophy

May 2016

The dissertation of Marlena K. Sheridan was reviewed and approved* by the following:

Ryan J. Elias Associate Professor of Food Science Dissertation Advisor Chair of Committee

Joshua D. Lambert Associate Professor of Food Science

John Coupland Professor of Food Science

Michela Centinari Assistant Professor of Horticulture

Robert F. Roberts Professor of Food Science Head of the Department of Food Science

*Signatures are on file in the Graduate School

ABSTRACT

Red wine quality is known to improve with oxidation, typically as a result of exposure to oxygen. These benefits are based on the reaction of acetaldehyde with flavonoids in red wine.

Acetaldehyde forms ethylidene bridges between flavonoids resulting in polymeric pigments and modified tannins, contributing to color stability and improved mouthfeel of red wine.

Winemakers often use oxygenation techniques in order to take advantage of the benefits of acetaldehyde. These methods are based on the reduction of oxygen in several metal-catalyzed steps resulting in acetaldehyde; undesirable, reactive intermediates are also formed as side products of this mechanism. As these reactive species can result in deleterious effects, oxygen exposure is a relatively risky technique in order to gain the benefits of acetaldehyde.

As a replacement for oxygenation techniques, I investigated the use of exogenous acetaldehyde additions to improve the color stability and mouthfeel of red wine. I first determined the viability of exogenous acetaldehyde treatment of a red wine by examining the effects of the treatment during alcoholic fermentation. Two levels of acetaldehyde (100 mg/L and 1000 mg/L) were added to a red wine over eight days of fermentation. After the completion of fermentation, wines were analyzed for their color stability and protein precipitation. High acetaldehyde treatment significantly increased the concentration of polymeric pigments and decreased the amount of protein precipitated by tannin. These results demonstrate the ability of exogenous acetaldehyde treatment to improve color stability and mouthfeel of a red wine.

In order to understand the role of wine components in the reaction of acetaldehyde with wine flavonoids, I assessed the effect of pH, dissolved oxygen, and sulfur dioxide (SO2) in model wine solutions. The rate of reaction of acetaldehyde with catechin was significantly increased with lower pH and was not affected by dissolved oxygen. Interestingly, the reaction of iii

acetaldehyde with flavonoids was slowed but not prevented by the addition of SO2 as determined by monitoring the rate of reaction and the formation of polymeric pigments. These results demonstrate that acetaldehyde is reactive, not inert as previously assumed, in its sulfonate form.

Based on the efficacy of acetaldehyde in a system with an equimolar concentration of bisulfite in the previous study, I then explored the reactivity of aldehydes from bisulfite adducts. I synthesized α-hydroxyalkylsulfonates from bisulfite and several aldehydes found in wine: formaldehyde, acetaldehyde, propionaldehyde, isobutyraldehyde, and benzaldehyde. The reactivity of aldehydes from their free and bound (sulfonate) forms with catechin was determined.

The results demonstrate a clear relationship between reactivity of an aldehyde from its sulfonate and the dissociation constant (Kd) of that sulfonate. The bridged catechin oligomers from these aldehydes were also characterized by MALDI-TOF MS, several for the first time. This work was the first evidence of the reactivity of aldehydes, including acetaldehyde, from their sulfonates.

In order to confirm the application of exogenous acetaldehyde in red wine production, I investigated the treatment of red wine with exogenous acetaldehyde, the acetaldehyde-bisulfite adduct, and oxygenation after alcoholic fermentation. A reasonably low concentration of exogenous acetaldehyde (500 µM, 22 mg/L) significantly improved all measures of color stability evaluated, including color density and polymeric pigments. The sulfonate did not have the same effect as acetaldehyde, but did increase several parameters of color stability. The comparison between exogenous acetaldehyde and oxygenation demonstrated the inefficiency of the formation of acetaldehyde from oxygen. While both treatments consumed monomeric anthocyanins, only exogenous acetaldehyde significantly increased the concentration of polymeric pigments.

Overall, I demonstrated that exogenous acetaldehyde treatment is a viable alternative to oxygenation for improving the color stability of red wines. At low concentrations (22 mg/L), winemakers could use acetaldehyde to significantly improve color stability and wine quality.

Acetaldehyde is also likely to contribute to beneficial reactions of flavonoids in wine when found as a sulfonate, as I showed its reactivity upon the addition of bisulfite and when added as a sulfonate. While further work is needed to optimize its application during red wine production, this work suggests that exogenous acetaldehyde treatment could be an effective method of improving color stability for winemakers.

TABLE OF CONTENTS

List of Figures ...... x

List of Tables ...... xiii

Acknowledgements...... xv

Chapter 1 Literature Review...... 1

1.1 Introduction...... 1 1.1.1 Description of Acetaldehyde...... 2 1.2 Wine Oxidation...... 3 1.2.1 Oxygen in Wine ...... 3 1.2.1.1 Micro-Oxygenation of Wine...... 4 1.2.2 Formation of Acetaldehyde...... 5 1.2.3 Oxidation Side Reactions...... 7 1.2.4 Sulfur Dioxide in Wine...... 9 1.2.4.1 Reaction of Bisulfite with Acetaldehyde ...... 10 1.3 Reactions of Acetaldehyde with Wine Flavonoids...... 12 1.3.1 Anthocyanins ...... 14 1.3.1.1 Description of Wine Anthocyanins...... 14 1.3.1.2 Role of Anthocyanins in Wine...... 15 1.3.1.3 Reactions of Anthocyanins with Acetaldehyde ...... 18 1.3.2 Tannins...... 21 1.3.2.1 Description of Wine Tannins ...... 21 1.3.2.2 Role of Tannins in Wine ...... 23 1.3.2.3 Reactions of Tannins with Acetaldehyde ...... 25 1.3.3 Impact of Reactions of Acetaldehyde with Flavonoids on Wine Quality...... 27 1.4 Significance and Hypotheses ...... 29 1.4.1 Significance...... 29 1.4.2 Hypotheses and Aims ...... 30

Chapter 2 Exogenous Acetaldehyde as a Tool for Modulating Wine Color and Astringency during Fermentation...... 31

2.1 Abstract ...... 31 2.2 Introduction...... 32 2.3 Materials and Methods...... 35 2.3.1 Materials ...... 35 2.3.2 Wine Production ...... 35 2.3.3 Acetaldehyde Measurement...... 37 2.3.4 Pigment Analysis ...... 37 2.3.5 Total Phenolics Measurement...... 37 2.3.6 Protein Precipitation Analysis...... 38 2.3.7 Salivary Protein Precipitation Analysis ...... 38 2.3.8 Statistical Analysis...... 38 2.4 Results and Discussion ...... 39 2.4.1 Acetaldehyde Concentrations in Must and Wine...... 39 2.4.2 Effect of Exogenous Acetaldehyde on Wine Color...... 40 2.4.3 Effect of Exogenous Acetaldehyde on Tannin-Protein Interactions...... 43 2.5 Conclusions...... 44 2.6 Acknowledgements...... 45

Chapter 3 Reaction of Acetaldehyde with Wine Flavonoids in the Presence of Sulfur Dioxide ..46

3.1 Abstract ...... 46 3.2 Introduction...... 47 3.3 Materials & Methods ...... 49 3.3.1 Materials ...... 49 3.3.2 Reaction Mixture Preparation ...... 49 3.3.3 Flavonoid Analysis ...... 50 3.3.4 Pigment Analysis ...... 51 3.3.5 Reaction Product Characterization...... 52 3.3.6 Statistical Analysis...... 52 3.4 Results and Discussion ...... 53 3.4.1 Reaction of Acetaldehyde with Flavanols ...... 53 3.4.2 Reaction of Acetaldehyde with M3G and Flavanols ...... 57 3.4.3 Characterization of Products with MALDI-TOF MS...... 63 3.5 Acknowledgements...... 72 vii

Chapter 4 Reactions of Free and SO2-Bound Aldehydes with (+)-Catechin...... 73

4.1 Abstract ...... 73 4.2 Introduction...... 74 4.3 Materials and Methods...... 77 4.3.1 Materials ...... 77 4.3.2 Bisulfite Adduct Synthesis...... 77 4.3.3 Reaction Mixture Preparation ...... 78 4.3.4 Catechin Analysis ...... 79 4.3.5 Reaction Product Characterization...... 79 4.3.6 Statistical Analysis...... 80 4.4 Results and Discussion ...... 80 4.4.1 Consumption of Catechin by Aldehydes ...... 80 4.4.2 Consumption of Catechin by Aldehyde-Bisulfite Adducts ...... 83 4.4.3 Characterization of Catechin-Aldehyde Products...... 86 4.5 Acknowledgements...... 95

Chapter 5 Improving Red Wine Color Stability with Exogenous Acetaldehyde ...... 96

5.1 Abtract...... 96 5.2 Introduction...... 97 5.3 Materials and Methods...... 100 5.3.1 Materials ...... 100 5.3.2 Bisulfite Adduct Synthesis...... 100 5.3.3 Model Reaction Mixture Preparation...... 101 5.3.4 Catechin Analysis ...... 101 5.3.5 Treatment of Red Wine with Free and Bound Acetaldehyde ...... 102 5.3.6 Treatment of Red Wine with Endogenous and Exogenous Acetaldehyde ...... 102 5.3.7 Pigment Characterization...... 103 5.3.8 Statistical Analysis...... 104 5.4 Results and Discussion ...... 104 5.4.1 Activity of Acetaldehyde-Bisulfite Adduct in Model Solutions...... 104 5.4.2 Effect of Acetaldehyde and Acetaldehyde-Bisulfite Adduct on Red Wine Color...... 106 5.4.3 Effect of Exogenous Acetaldehyde and Oxygenation on Red Wine Color ...... 110 5.5 Acknowledgements...... 114 viii

Chapter 6 Conclusions and Recommendations for Future Work ...... 115

6.1 Reaction of Flavonoids with Acetaldehyde ...... 115 6.2 Reactivity of Aldehydes from α-Hydroxyalkylsulfonates...... 116 6.3 Exogenous Acetaldehyde as Wine Treatment ...... 117 6.4 Concluding Remarks...... 119

References...... 120

List of Figures

Figure 1.1. Formation of acetaldehyde from oxygen (Adapted from Danilewicz).31 ...... 6

Figure 1.2. Formation of reactive intermediates during the reduction of H2O2 (Adapted from Danilewicz).10 ...... 7

Figure 1.3. Reaction of quinone with wine components (Adapted from Kreitman).51 ...... 8

Figure 1.4. Reactions of the hydroxyl radical (•OH) and 1-hydroxyethyl radical (1- HER)...... 9

Figure 1.5. Equilibrium and pKa values for sulfur dioxide...... 10

Figure 1.6. Reactions of bisulfite in wine ...... 10

Figure 1.7. Reaction of acetaldehyde with bisulfite to form 1-hydroxyethanesulfonate...... 11

Figure 1.8. Flavonoid Structures...... 12

Figure 1.9. Reaction of acetaldehyde with a representative flavonoid...... 13

Figure 1.10. Equilibrium of anthocyanin forms...... 17

Figure 1.11. Reaction of malvidin-3-glucoside with bisulfite...... 18

Figure 1.12. Formation of polymeric pigments from malvidin-3-glucoside...... 19

Figure 1.13. Flavan-3-ol monomers found in grapes...... 22

Figure 1.14. Flavan-3-ol monomer and oligomer (i.e., a condensed tannin)...... 23

Figure 2.1. Mechanism of astringency by tannin-protein complex formation, aggregation, and precipitation...... 33

Figure 2.2. Direct and indirect condensation products of flavanols and anthocyanins occurring in wine...... 33

Figure 2.3. Tannin content of BSA-tannin precipitates quantified by ferric chloride reaction in (+)-catechin equivalents (CE). Error bars represent one standard deviation of the mean, and results in with different letters (a, b) are significantly different (p < 0.05)...... 43

Figure 2.4. Tannin content of saliva-tannin precipitates quantified by ferric chloride reaction in (+)-catechin equivalents (CE). Error bars represent one standard

deviation of the mean, and results in with different letters (a, b) are significantly different (p < 0.05)...... 44

Figure 3.1. Reaction of acetaldehyde with bisulfite and with representative flavonoids to form an ethylidene-bridged adduct...... 48

Figure 3.2. Catechin concentrations in model wine after treatment with 20 mg/L acetaldehyde at pH 2.0, 2.5, 3.0, and 3.5 as determined by HPLC-DAD...... 54

Figure 3.3. Catechin concentrations in model wine after treatment with 20 mg/L acetaldehyde with and without oxygen present at pH 2.5...... 55

Figure 3.4. Catechin concentrations in model wine (pH 2.5) after treatment with acetaldehyde and SO2 additions under aerobic (A) and anaerobic (B) conditions. Low SO2 samples (A+Low SO2) contained 40 mg/L TSO2 and High SO2 samples (A+High SO2) contained 80 mg/L TSO2...... 56

Figure 3.5. Catechin concentrations in all samples containing catechin over 12 days: catechin only (C), catechin with acetaldehyde (C+A), catechin with acetaldehyde and 40 mg/L SO2 (C+A+SO2), M3G and catechin with acetaldehyde (M3G+C+A), and M3G and catechin with acetaldehyde and 40 mg/L SO2 (M3G+C+A+SO2)...... 58

Figure 3.6. Malvidin 3-glucoside (M3G) concentrations in all samples containing M3G over 12 days: M3G only (M3G), M3G with acetaldehyde (M3G+A), M3G with acetaldehyde and 40 mg/L SO2 (M3G+A+SO2), M3G and catechin with acetaldehyde (M3G+C+A), M3G and catechin with acetaldehyde and 40 mg/L SO2 (M3G+C+A+SO2), M3G and GSE with acetaldehyde (M3G+G+A), and M3G and GSE with acetaldehyde and 40 mg/L SO2 (M3G+G+A+SO2)...... 59

Figure 3.7. Select color parameters from the modified Somers assay for control and treatment groups (acetaldehyde and acetaldehyde+SO2) at 12 days: A) Total anthocyanins, B) Degree of ionization of anthocyanins, C) Color density. Values represent the average of three experimental replicates ± standard deviation. Columns in the same group with different letters indicate significant differences (p<0.05)...... 62

Figure 3.8. MALDI-TOF mass spectrum recorded in positive reflectron mode of catechin treated with acetaldehyde...... 65

Figure 3.9. Reaction of catechin with acetaldehyde to form ethylidene-bridged oligomers followed by cleavage to form vinyl catechin moieties...... 67

Figure 3.10. Reaction of catechin with M3G and acetaldehyde to form an ethylidene- bridged polymeric pigment and subsequent rearrangement to a pyranoanthocyanins. ... 69

Figure 4.1. Mechanism of reaction of aldehydes with catechin...... 75

Figure 4.2. Reactions of aldehydes with bisulfite and catechin...... 76

Figure 4.3. Consumption of catechin by aldehydes. Values represent the average of three experimental replicates ± standard deviation...... 81

Figure 4.4. Consumption of catechin by aldehyde-bisulfite adducts. Values represent the average of three experimental replicates ± standard deviation...... 84

Figure 4.5. Catechin consumed after 28 days by aldehydes and aldehyde-bisulfite adducts. Values represent the average of three experimental replicates ± standard deviation. Values with different letters indicate significant differences (p<0.05)...... 85

Figure 4.6. Correlation between –log Kd of aldehydes-bisulfite adducts and the percent of catechin consumed by the bisulfite adduct compared to the free aldehyde...... 86

Figure 4.7. MALDI-TOF mass spectrum recorded in positive reflectron mode of catechin treated with formaldehyde...... 87

Figure 4.8. MALDI-TOF mass spectrum recorded in positive reflectron mode of catechin treated with acetaldehyde...... 89

Figure 4.9. MALDI-TOF mass spectrum recorded in positive reflectron mode of catechin treated with propionaldehyde...... 91

Figure 4.10. MALDI-TOF mass spectrum recorded in positive reflectron mode of catechin treated with isobutyraldehyde...... 93

Figure 4.11. MALDI-TOF mass spectrum recorded in positive reflectron mode of catechin treated with benzaldehyde...... 94

Figure 5.1. Reaction of malvidin-3-glucoside with catechin and acetaldehyde...... 97

Figure 5.2. Formation of acetaldehyde from oxygen (Adapted from Danilewicz).31 ...... 98

Figure 5.3. Oxidation reactions leading to acetaldehyde formation (Adapted from Danilewicz).10...... 99

Figure 5.4. Catechin consumption at pH 2 – 4 by acetaldehyde (A) and 1- hydroxyethanesulfonate (B)...... 105

xii

List of Tables

Table 1.1. Concentrations of acetaldehyde in alcoholic beverages.6,7 ...... 2

Table 1.2. Sources for introduction of atmospheric oxygen in wine.1,11,13–15 ...... 3

Table 1.3. Anthocyanins found in V. vinifera grapes.71 ...... 15

Table 2.1. Color parameters from the HTP modified Somers assay of treatment groups post-primary fermentation.a...... 41

Table 3.1. Color parameters from the modified Somers assay for control and treatment a groups (Acetaldehyde or Acetaldehyde and SO2) at 12 days...... 61

Table 3.2. Predicted and observed m/z values as recorded in positive reflectron mode MALDI-TOF MS of catechin treated with acetaldehyde...... 66

Table 3.3. Predicted and observed m/z values recorded in positive reflectron mode MALDI-TOF MS of catechin and M3G treated with acetaldehyde...... 69

Table 3.4. Predicted and observed m/z values recorded in positive reflectron mode MALDI-TOF MS of GSE treated with acetaldehyde...... 70

Table 3.5. Predicted and observed m/z values recorded in positive reflectron mode MALDI-TOF MS of GSE and M3G treated with acetaldehyde...... 71

Table 4.1. Odor thresholds and characteristics of aldehydes ...... 74

61 Table 4.2. Dissociation constants (Kd) of aldehyde-bisulfite adducts...... 76

Table 4.3. Aldehyde-bisulfite adducts synthesized and their synthetic yields...... 78

Table 4.4. Predicted and observed m/z values as recorded in positive reflectron mode MALDI-TOF MS of catechin treated with formaldehyde...... 88

Table 4.5. Predicted and observed m/z values as recorded in positive reflectron mode MALDI-TOF MS of catechin treated with acetaldehyde...... 90

Table 4.6. Predicted and observed m/z values as recorded in positive reflectron mode MALDI-TOF MS of catechin treated with propionaldehyde...... 92

Table 4.7. Predicted and observed m/z values as recorded in positive reflectron mode MALDI-TOF MS of catechin treated with isobutyraldehyde...... 93

Table 4.8. Predicted and observed m/z values as recorded in positive reflectron mode MALDI-TOF MS of catechin treated with benzaldehyde...... 94

xiii

Table 5.1. Color parameters from the modified Somers assay for control, bisulfite treated, acetaldehyde treated, and sulfonate treated wines after 8 weeks...... 107

Table 5.2. Color parameters from the Harbertson-Adams assay for control, bisulfite treated, acetaldehyde treated, and sulfonate treated wines after 8 weeks.a ...... 109

Table 5.3. Color parameters from the modified Somers assay for control, acetaldehyde treated, and oxygenated (periodic aeration) wines after 8 weeks.a ...... 111

Table 5.4. Color parameters from the Harbertson-Adams assay for control, acetaldehyde treated, and oxygenated (periodic aeration) wines after 8 weeks.a ...... 112

xiv

Acknowledgements

I would like to express my sincere gratitude to my advisor, Dr. Ryan Elias, for giving me the opportunity to join his lab and explore what was a brand new field to me. I appreciated his guidance throughout this process and his trust in allowing me the freedom to pursue my own interests.

I would also like to thank my committee members, Dr. Josh Lambert, Dr. John Coupland, and Dr. Michela Centinari for their guidance. Their diverse knowledge and insights helped me to envision a larger context for my project and challenged me to expand my own thinking.

I would like to acknowledge Ms. Denise Gardner for her unending support and help during wine production. Her impeccable organization, work ethic, and optimism made the wine production for this project possible, and even enjoyable. I would also like to acknowledge Dr.

Tatiana Laremore for her assistance with MALDI-TOF analyses and for her constant supply of chocolates.

Finally, I would like to thank my lab mates and classmates for teaching me, supporting me, and commiserating with me during my time at Penn State. I have the utmost appreciation for my family and friends who have encouraged me from near and far. Thank you for believing in me and making these years in State College great ones.

Chapter 1

Literature Review

1.1 Introduction

In recent years, wine research has progressed as analytical techniques and knowledge of wine have improved. A large portion of this research has focused on underlying mechanisms for observed benefits of wine processing in order to better their outcomes. Oxygen management in particular has been a focus due to its significant impact on wine quality. Oxygenation is known to improve red wine characteristics associated with color and mouthfeel based on the oxidation and modification of flavonoids. However, oxygen also leads to deleterious effects associated with chemical and microbial instability. As a risky yet rewarding treatment, oxygenation has been investigated in an effort to enhance the ability of winemakers to take advantage of this process.

Developments in oxygenation include micro-oxygenation as a means to achieve precise and predictable oxidation. This technique allows winemakers greater control of oxygen concentrations but does not eliminate many of the risks associated with its presence. Accelerated aging techniques have also been attempted including heating, ultrasound, irradiation, high pressure, and electrolysis.1,2 These methods all include undesirable side effects. Acetaldehyde is in an intermediate oxidation product that results in the modification of red wine polyphenols to improve mouthfeel and color stability and has not been thoroughly investigated. Addition of exogenous acetaldehyde could be used as a tool for winemakers that would replace risky oxygen exposure as a less expensive alternative.

1.1.1 Description of Acetaldehyde

Acetaldehyde is a volatile compound that is ubiquitous in nature. It is present in the atmosphere as a product of plant respiration and combustion.3 Acetaldehyde is also formed in the body during the metabolism of alcohol.3 Acetaldehyde is commonly used as a flavor additive and is generally recognized as safe (GRAS).4 It has an odor threshold of 0.5 mg/L and this odor is typically fruity and pleasant at low concentrations but pungent at higher concentrations.3,5 A common source of acetaldehyde is alcoholic beverages as acetaldehyde is formed during fermentation and chemical oxidation of ethanol. The flavor threshold of acetaldehyde in wine is

100-125 mg/L.6 A summary of acetaldehyde levels in these beverages is found in Table 1.1.6,7

Table 1.1. Concentrations of acetaldehyde in alcoholic beverages.6,7

Total Beverage Acetaldehyde (mg/L) Red Wine 4 – 212 White Wine 11 – 493 Sweet Wine 188 – 248 Sherry 50 – 523 Port 22 – 800 Brandy 63 – 308 Cognac 0 – 211 Beer 5 – 63 Apple Wine/Cider 15 – 133 Whisky 10 – 110 Rum 0 – 68 Vodka 0 – 13 Tequila 0 – 670 Sake 15 – 60

1.2 Wine Oxidation

Acetaldehyde is formed as an ethanol oxidation product in wine typically after oxygen exposure. Direct oxidation of phenolics does not occur at wine pH (3-4) since the phenolate anion

8–10 form is not present (pKa 9-10). Instead, oxidation of wine results in products like acetaldehyde that can react with phenolics. The reduction of oxygen to form acetaldehyde is the goal of deliberate oxygen exposure, though this mechanism involves many other reactions and intermediates that should not be discounted.

1.2.1 Oxygen in Wine

Oxygen is easily incorporated into wine throughout production during the course of several processing steps and deliberate oxygenation techniques. Upon exposure to air, oxygen will be dissolved quickly; wine stirred with air will become saturated in ca. 30 seconds.11,12 The dissolved oxygen (DO) concentration in wine saturated with air is 8.4 mg/L at 20 °C. Wine production steps can introduce DO levels that vary from very low concentrations (e.g., pumping) to saturation (crushing or pressing) (Table 1.2).11,13,14

Table 1.2. Sources for introduction of atmospheric oxygen in wine.1,11,13–15

Increase in Dissolved Operation Oxygen (mg/L) Crushing, Pressing Saturation Racking 0.3 – 5 Pumping 0.1 – 0.2 Centrifugation 1.0 Filtration 0.3 – 0.7 Bottling 0.8 Cold Stabilization 1.3

Oxygen is also quickly consumed in red wine; after saturation, DO of a red wine will be reduced to below 1 mg/mL in about 6 days at 30 °C.8,16 The DO of wine in storage tanks or barrels is typically between 0.02 and 0.05 mg/L.11 Red wine has an oxygen consumption capacity of up to 800 mg/L.11 While techniques like barrel aging are used to incorporate a relatively controlled amount of oxygen, the exposure of red wine to oxygenation is still highly variable.

Instead, winemakers can use micro-oxygenation (MOX) under highly controlled conditions in order to introduce small amounts of DO over time.

1.2.1.1 Micro-Oxygenation of Wine

MOX involves the deliberate, gradual addition of DO before malolactic fermentation

(MLF) (10-30 mL/L/month) or after MLF (1-5 mL/L/month).17,18 MOX is used to improve wine color, aroma, and mouthfeel.18 MOX is based on the premise that the rate of oxygen dosing is less than the rate of oxygen consumption in the wine; therefore, in theory, no increase in DO should be measureable during an appropriately administered MOX application.19 However, it has been shown that even low doses of oxygen (1 mL/L/month) result in increased concentrations of DO in red wine.20 When applied before MLF, MOX is more effective due to an abundance of monomeric anthocyanins. This allows better color stabilization and also limits tannin lengthening as anthocyanins form terminal subunits.21 Additionally, MLF has been shown to decrease the concentration of residual acetaldehyde from MOX as malolactic bacteria can metabolize acetaldehyde.22 This may be detrimental to wines if there is insufficient time for acetaldehyde to react before it is depleted by MLF.23,24

Studies on the use of MOX with red wine have been reviewed recently.18,19,25,26

Generally, MOX is considered to be beneficial for wines that have high phenolic contents and

especially for wines that have a high proportion of free anthocyanins.21,27 Wines with relatively low phenolic contents are more prone to the risk of over-treatment, which leads to the loss of desirable aromas, undesirable changes in wine color, and growth of aerobic organisms like acetic acid bacteria.18,27 These deleterious effects are also a risk if the rate of oxygen addition is not chosen correctly. MOX is also expensive, especially for the initial investment, with prices around

$1,000 per tank and total costs up to a couple hundred thousand dollars for large wineries.28,29

Even when MOX is implemented correctly, it is difficult to get reproducible results and to monitor the process without the use of specialized, sensitive equipment.20 Though MOX is a useful tool for controlling oxidation, there are still many risks and costs of its use in winemaking.

Periodic aeration has recently been explored as an alternative to MOX.30 In this treatment, oxygen is added weekly by aerating a small volume of wine to DO saturation and then reintroducing it to the bulk wine. In this way, the same dose of total oxygen can be added without the use of MOX equipment but with the overall benefits of oxidation. Periodic aeration may be a useful and less expensive alternative to MOX for winemakers.

1.2.2 Formation of Acetaldehyde

Acetaldehyde in wine can result from either biological (enzymatic) or strictly chemical

(non-enzymatic) means. Under the latter scenario, acetaldehyde is formed by several metal- catalyzed steps that lead to the eventual oxidation of ethanol (Figure 1.1). Metal ions first take part in the reaction by forming an Fe(II)-O2 adduct, wherein an electron transfer occurs to form

31 an Fe(III)-superoxo complex, and subsequently hydrogen peroxide (H2O2). Hydrogen peroxide is further reduced by the Fenton reaction to form a hydroxyl radical (•OH). This radical is sufficiently reactive to interact with organic material in a concentration-based manner. The

hydroxyl radical therefore reacts with ethanol to yield acetaldehyde, as ethanol is the most abundant organic compound present in wine (~2 M).8–10,31,32 Acetaldehyde is also formed as a byproduct of yeast metabolism.6 Sugar is the primary substrate for the formation of acetaldehyde, which takes place mainly during the growth period of yeast. There are large differences in the amount of acetaldehyde produced between species and strains of Saccharomyces cerevisiae.6,33

Fe(II) Fe(II)

2+ O2 Fe(II)-O2 2Fe(III) + H2O2 Fe(II) Fe(II) Fe(III) Fe(III) H2O HO 0.5 H2O2 0.5 O2

CH CHO CH CHOH EtOH 3 3

Figure 1.1. Formation of acetaldehyde from oxygen (Adapted from Danilewicz).31

In studies on MOX or other sources of DO, there is no consensus on resulting acetaldehyde concentrations. Due to the complexity of the reactions that lead to acetaldehyde generation and its quick consumption, the concentration of acetaldehyde in a wine sample is not representative of the acetaldehyde formed up to that point. Previous studies have shown that the acetaldehyde concentration remains low during oxygenation until it reaches a point where it begins to increase significantly (1 mg/L/day).34 This lag phase may represent the consumption of free sulfur dioxide or available polyphenols, but this has not been investigated. It is, however, clear that large amounts of acetaldehyde are formed with the excessive application of MOX.35

The benefits of MOX have been seen even when acetaldehyde concentrations do not appear to 6

increase with treatment.36 Once oxygen is no longer being introduced, the acetaldehyde in a wine will be consumed.37

1.2.3 Oxidation Side Reactions

Oxygen is reduced to water in a series of steps that involve several reactive intermediates

(Figure 1.2). While acetaldehyde is formed as a product of these reactions, quinones, hydroxyl radicals, and hydroxyethyl radicals are also generated, which can be detrimental to wine quality.10,38,39

R O

Fe(II) H2O2 O H2O

R OH CH3CHOH CH3CHO Fe(III) HO EtOH OH

Figure 1.2. Formation of reactive intermediates during the reduction of H2O2 (Adapted from Danilewicz).10

Quinones are highly electrophilic reactive oxygen species formed during the oxidation of some phenolic compounds. These species react quickly with nucleophiles found in wine (e.g.,

- amino acids, bisulfite (HSO3 ), thiols (RSH), ascorbic acid, and flavanols) (Figure 1.3). Reaction of quinones with flavanols can lead to browning as they form polymers.40,41 In the presence of bisulfite or ascorbic acid, quinones will be reduced back to hydroquinones.42,43 The reaction of quinones with amino acids is followed by Strecker degradation to form aldehydes including methional and phenylacetaldehyde.44 The reaction of quinones with thiols significantly impacts

wine aroma; the consumption of thiols as catechol adducts removes these desirable aroma compounds.45–49 Reaction with bisulfite, glutathione (GSH), and ascorbic acid has been shown to occur at a very fast rate followed by other thiols and amino acids.50

Strecker Degradation Aldehydes HO3S OH OH

R OH R OH - Amino 3 Acids HSO

O RS OH RSH R OH R O Ascorbic Acid Flavanols OH R R OH

HO OH OH R OH

Figure 1.3. Reaction of quinone with wine components (Adapted from Kreitman).51

Hydroxyl radicals are formed by the breakdown of hydrogen peroxide (Figure 1.2).

These radicals are highly reactive and therefore react in a non-selective manner with wine components based on concentration.10,52 Ethanol is therefore the most likely target followed by glycerol and tartaric acid. The reaction products of ethanol, glycerol, and tartaric acid are acetaldehyde, glyceraldehyde, and glyoxylic acid, respectively (Figure 1.4).53,54 1-Hydroxyethyl 8

radical (1-HER), an intermediate in the oxidative pathway from ethanol to acetaldehyde, is also a reactive oxygen species.10,55 1-HER also reacts with the α,β-unsaturated side chains of cinnamic acids, including caffeic acid and ferulic acid, forming an allylic alcohol.56 Furthermore, recent studies have demonstrated that 1-HER can also contribute to the oxidative loss of thiols that are important to wine aroma.57 The reactions of these two radicals, as well as quinones, contribute to side reactions and possible deleterious effects of oxidation in wine.

O OH

OH OH OH HO OH CO O OH 2 Glycerol Cinnamic Acid OH Glyceraldehyde O Ethanol OH OH Acetaldehyde OH O 1-HER O HO O OH OH R-SH O OH Thiol R-S Glyoxylic Acid OH Tartaric Acid

Figure 1.4. Reactions of the hydroxyl radical (•OH) and 1-hydroxyethyl radical (1-HER).

1.2.4 Sulfur Dioxide in Wine

Sulfur dioxide (SO2) is added to wine to prevent chemical and microbial spoilage. SO2 is found in equilibrium of three forms: molecular SO2, bisulfite, and sulfite (Figure 1.5). The molecular form is responsible for microbial preservation of wine. The majority of SO2, however, is found as the bisulfite ion at wine pH; 96% will be the bisulfite form at pH 3.2.1 Bisulfite provides chemical stability, as it is able to quench oxidation reactions as well as to react with oxidation products. Bisulfite reacts readily with hydrogen peroxide and quinones (Figure 1.6).

Reaction with hydrogen peroxide prevents the oxidation of ethanol to acetaldehyde and reaction

42,58,59 with quinones regenerates their catechol. SO2 therefore protects other wine components from oxidation by quenching reactive oxygen species. Bisulfite also reacts with anthocyanins, bleaching their color. The role of this reaction will be discussed in greater detail in later sections.

pKa = 1.86 pKa = 7.2 - + 2- + SO2 + H2O HSO3 + H SO3 + 2H Molecular SO 2 Bisulﬁte Sulﬁte

Figure 1.5. Equilibrium and pKa values for sulfur dioxide.

Anthocyanin-SO H H SO 3 Anthocyanin HOOH 2 4

- HSO3 O O R H O O HO3S OH OH H3C S O OH O R OH R OH

Figure 1.6. Reactions of bisulfite in wine

1.2.4.1 Reaction of Bisulfite with Acetaldehyde

Bisulfite is known to react with compounds to prevent possible aromatic faults associated with oxidation by forming non-volatile adducts. Bisulfite readily reacts with carbonyl compounds, including acetaldehyde, to form α-hydroxyalkylsulfonates (Figure 1.7). Excessive acetaldehyde can lead to a characteristically oxidized aroma and is perceived as a fault in wine, though the sulfonate is reported to have no detectable aroma.60 In this way, bisulfite can both 10

prevent and mask wine oxidation. The portion of SO2 that has reacted with acetaldehyde, other carbonyls, or other wine components (e.g., anthocyanins) is referred to as bound SO2. All other forms constitute the free SO2 portion; together, bound and free SO2 comprise total SO2. During production, wine will have free SO2 present in order to protect the wine from chemical and microbial instability.

O S H O O HO O H3C S O H3C H OH O

Figure 1.7. Reaction of acetaldehyde with bisulfite to form 1-hydroxyethanesulfonate.

The reaction of acetaldehyde with bisulfite is fast and results in the strongly bound

-6 1,61 adduct, 1-hydroxyethanesulfonate (Kd = 2.4 x 10 M). Most research has therefore assumed that 1-hydroxyethanesulfonate is inert with respect to acetaldehyde reactivity. However, this acetaldehyde-bisulfite adduct has been shown to have antioxidant activity in beer; the formation of radicals from ethanol oxidation was slowed by the sulfonate indicating similar activity to SO2 itself.62

In most wines, there is excess SO2 to acetaldehyde so that there will be free SO2 present.

While further oxidation consumes SO2 and may free some acetaldehyde, reaction products from acetaldehyde are seen in wines over time with free SO2 present throughout. This suggests that acetaldehyde is able to react with wine components in the presence of free SO2. The reactivity of acetaldehyde from its bound form has not been explored.

1.3 Reactions of Acetaldehyde with Wine Flavonoids

Reactions of acetaldehyde, specifically those with phenolics, are complicated but are critical to understanding the complexities of wine chemistry. Acetaldehyde mediates many oxidation reactions as the direct oxidation of phenolics is unlikely at wine pH.8–10 Direct condensation reactions can occur via quinone intermediates (Figure 1.3). The reactions of acetaldehyde with specific polyphenols will be reviewed here.

The total polyphenol content of red grapes is in the range of 175-209 mg/100 g (fresh weight basis).63 Red wine contains a large number of polyphenols, between 1000 and 3000 mg/L gallic acid equivalents total polyphenol content.64 These polyphenols include flavonoids, which are important for red wine chemistry and quality.

Some of the most important reactions of acetaldehyde in wine are those with flavonoids.

Flavonoids are compounds with a characteristic three-ring structure: two aromatic rings (A and

B) connected by a pyran ring (C) (Figure 1.8). The saturation of the C ring determines the class of flavonoids. Anthocyanins contain a fully unsaturated C ring, also known as a pyrilium cation.

Flavanols contain a fully saturated C ring. Anthocyanins and flavanols are the major classes of flavonoids found in grapes and wine.65,66 These flavonoids and their reaction products will be discussed in greater detail in later sections.

3' 2' 4' B 8 1 1' O 5' O O 7 2 A C 6' 6 3 OR OH 5 4 Flavonoid Anthocyanin Flavanol

Figure 1.8. Flavonoid Structures 12

Flavonoids can react with acetaldehyde by the mechanism shown in Figure 1.9. The first step is the acid-catalyzed protonation of acetaldehyde to form an electrophilic carbocation. The nucleophilic A ring of a flavonoid, represented here as a substituted resorcinol, then adds to the protonated acetaldehyde to form a carbocation after the loss of water. The addition of another nucleophilic flavonoid forms an ethyl-bridged flavonoid dimer. This ethylidene bridge between flavonoids is characteristic of the reaction of acetaldehyde with flavonoids. However, acid- catalyzed cleavage of that bridge results in the formation of a vinyl moiety on one of the bridged flavonoids. By this mechanism, bridged anthocyanins and flavanols can form as well as those with vinyl moieties.

R2 OH

HO R1 HO R HO H 1 + O H+ O -H HO H+ R2 R2 H C H H C H 3 3 OH OH

HO OH HO R1 OH R HO R 4 1 HO R H+ 2 R3 + OH R2 R OH OH 3 -H+ OH R 4 R 3 R4

Figure 1.9. Reaction of acetaldehyde with a representative flavonoid.

1.3.1 Anthocyanins

1.3.1.1 Description of Wine Anthocyanins

Anthocyanins are a class of pigmented flavonoids that contribute colors of red, purple, or blue. Their environment (e.g., pH, complexation with phenolics) and intrinsic structural characteristics determine their observed color.67,68 The amount of anthocyanins found in grapes varies significantly from 30 to 750 mg/100 g depending on the variety and growing conditions.69

Anthocyanins have been studied for their health benefits including their ability to act as anticancer, antimicrobial, and antiviral agents.70

Anthocyanins are found in the skin of red grapes and are derived from six aglycones, or anthocyanidins – pelargonidin, cyanidin, delphinidin, peonidin, petunidin, and malvidin (Table

1.3).67,71,72 These grape anthocyanins are found as 3-glucosides in Vitis vinifera grapes and their

3,5-diglucosides are also found in American species (V. riparia, V. rupestris, etc.) and French-

American hybrids.68 The aglycone anthocyanidins differ in their substitution on the B-ring with variations in the number and position of hydroxyl and methoxyl groups present. The anthocyanins with more methoxyl groups appear to be more red, thus malvidin is the reddest pigment present.68

Malvidin-3-glucoside is also the most abundant anthocyanin in young, red, V. vinifera wines accounting for 50-90% of the anthocyanins present depending on the variety, growing conditions, and winemaking practices.71 Acylated monoglucoside anthocyanins are also present in wine, which are acylated by p-coumaric, caffeic and acetic acids. The reactions of acylated monoglucosides will not be reviewed here but have been discussed elsewhere.1,71,73,74

Table 1.3. Anthocyanins found in V. vinifera grapes.71

R1 OH

HO O R2

OR3 OH

Name R1 R2 R3 Pelargonidin H H H Pelargonidin-3-glucoside H H Glu Cyanidin OH H H Cyanidin-3-glucoside OH H Glu Delphinidin OH OH H Delphinidin-3-glucoside OH OH Glu

Peonidin OCH3 H H

Peonidin-3-glucoside OCH3 H Glu

Petunidin OCH3 OH H

Petunidin-3-glucoside OCH3 OH Glu

Malvidin OCH3 OCH3 H

Malvidin-3-glucoside OCH3 OCH3 Glu

1.3.1.2 Role of Anthocyanins in Wine

Anthocyanins are readily extracted from grape skins during red wine production since they are water-soluble. The breakdown of grape skin and cell membranes begins during crushing, while maceration, heat of fermentation, ethanol content, and other fermentation conditions encourage further extraction of anthocyanins.75,76 The total concentration of anthocyanins in young, red wines is about 500 mg/L, though this concentration can be as high as 2000 mg/L.1,71 15

However, as will be discussed in greater detail in this review, monomeric anthocyanins react quickly with other components of wine. Only a few days after crushing, the rate of reaction of anthocyanins exceeds the rate of extraction, and the concentration of solution phase monomeric anthocyanins begins to decrease.77 Approximately 50 to 70% of anthocyanins will be incorporated into polymeric pigments after one year.75,78–80 Anthocyanins are also lost to degradation and precipitation reactions. The degradation of anthocyanins generates colorless, low molecular weight compounds.81 The most important role of anthocyanins in wine is by contributing color, though they may also contribute to taste and mouthfeel under certain conditions.82

Anthocyanins exist in four forms in solution whose distribution is dependent on pH

(Figure 1.10). Only one of these forms is the desired red form, the flavylium cation, which is favored at low pH.68,83,84 At pH 3.4, only 15% of anthocyanins are in their colored, flavylium

1 state. This proportion decreases at higher pH. The hydration constant (pKh) of malvidin-3- glucoside is 2.6.85 Therefore, at wine pH, the colorless, hydrated form is favored.86

OCH3 OCH3 OH OH +H2O OH HO O HO O OCH3 OCH3 -H+ O-Glu O-Glu OH OH Flavylium cation Carbinol pseudobase (red) (colorless)

-H+

OCH3 OCH3 OH OH OH O O HO O OCH3 OCH3

O-Glu O-Glu OH OH Quinoidal base Chalcone (blue) (pale yellow)

Figure 1.10. Equilibrium of anthocyanin forms

Anthocyanins also react quickly with the bisulfite form of SO2 forming a colorless adduct

86 in a reaction known as SO2 bleaching (Figure 1.11). At wine pH, bisulfite is by far the most common form (96%) of SO2 so the concentration of total SO2 is closely related to the amount of anthocyanins that remain colored. Consequently, at a constant SO2 concentration, higher concentrations of anthocyanins will lead to greater color intensity.71 The anthocyanin-bisulfite

66,87 adduct is stable at wine pH as its dissociation constant (pKs) is 5. However, resistance to SO2 bleaching is seen in anthocyanin-derived pigments formed by oxidation reactions.

OCH3 OCH3 OH OH

HO O +HSO3 HO O OCH3 OCH3

O-Glu O-Glu OH OH SO3H Flavylium cation Bisulﬁte adduct (red) (colorless)

Figure 1.11. Reaction of malvidin-3-glucoside with bisulfite.

The color of anthocyanins in their environment is mainly affected by pH and SO2 concentration as described above. Their color can also be stabilized by copigmentation, an inter- or intramolecular association of anthocyanins with other moieties that increases their color density. Copigmentation has been extensively reviewed elsewhere.1,71,88–90 Perhaps the more important process for stabilization, however, is the conversion of anthocyanins to polymeric pigments. This can occur by several mechanisms associated with oxidation74, but this review will focus on the role of acetaldehyde in improving color stability of red wine pigments.

1.3.1.3 Reactions of Anthocyanins with Acetaldehyde

Experiments by Timberlake and Bridle first showed the important role that acetaldehyde plays in increasing the degree of ionization of anthocyanins and lowering the amount of monomeric anthocyanins in red wine.91 Anthocyanins readily take part in condensation reactions involving acetaldehyde. The hydrated form of the monomeric anthocyanin can act as a nucleophile from the A ring, similar to flavanols. As shown in Figure 1.9, an anthocyanin adds to protonated acetaldehyde to form a carbocation. The addition of another nucleophilic flavonoid 18

leads to the formation of an ethylidene bridge between the anthocyanin and flavonoid. This bridging by acetaldehyde was proposed by Timberlake and Bridle in 1976 and confirmed by

Fulcrand et al. in 1996.92,93 Anthocyanins can react by this acetaldehyde-mediated mechanism with other anthocyanins to form dimers or with other flavonoids, including flavanols. This mixed group of products formed is referred to as polymeric pigments (Figure 1.12). Ethyl-bridged polymeric pigments have been observed in model systems and in red wine.94–98

OH OGl

OCH HO O 3 OCH3 OH OH HO O OCH3 OCH3 OCH3 OGl OH OH HO O OCH3 OCH 3 OGl OH O OH HO O H3C H OCH3 OH OCH3 OGl OGl OH OH OH OCH HO O HO O 3 OCH OH 3 OH HO O OGl OCH3 O OH OH OH OH HO O HO O OH OH

OH OH OH OH

Figure 1.12. Formation of polymeric pigments from malvidin-3-glucoside.

These ethylidene bridges formed by reaction with acetaldehyde are unstable and susceptible to acid-catalyzed cleavage (Figure 1.9).86,99,100 Further reaction of the products from that cleavage form another class of polymeric pigments referred to as pyranoanthocyanins

(Figure 1.12).66,100 These pyranoanthocyanins include vitisins, derivatives of monomeric anthocyanins with an additional pyran ring formed between C4 and the hydroxyl group of

C5.15,66,101 Pyranoanthocyanin-flavanol oligomers were first proposed based on model studies by

Francia-Aricha et al. and were observed to be more resistant to discoloration by pH changes and 19

96 SO2. These products can be formed either from the reaction of an anthocyanin with a vinyl flavanol, or the reaction of a flavanol with a pyranoanthocyanin already formed by reaction with acetaldehyde. Anthocyanins are believed to act primarily as terminal subunits in reactions with flavanol oligomers as they react mainly from C8.102

Reactions between tannins, anthocyanins, and acetaldehyde in model solutions have shown that polymerization of anthocyanins happens readily with tannins present and slowly without them, with color shifts towards violet.91,93,96,103–110 Products of these condensation reactions with anthocyanins also undergo shifts in color from the starting monomeric anthocyanin. Ethyl-linked anthocyanin species are typically more purple (absorbance between

528 and 540 nm) and the rearranged, stable derivatives, pyranoanthocyanins, are orange

(absorbance between 480 and 510 nm).91,95,111,112 Other adducts range from red (absorbance between 515 and 526 nm) for tannin-anthocyanin adducts to blue (absorbance at 575 nm) for flavanyl-vinylpyranoanthocyanins (portisins).113–116 Some anthocyanin-tannin adducts, however, may be colorless.117

Polymeric pigments formed contribute to increased color stability of red wine. These

15,101 polymeric pigments are resistant to bleaching as they are weak binders of SO2. In fact, flavanol-ethyl-anthocyanin adducts are not bleached by SO2 to any degree, although their color is known to be affected by pH with a shift towards blue as pH increases.95 Pyranoanthocyanins are relatively more stable to SO2 and pH changes due to the addition of a pyran moiety on the C ring

84,101,112,118 which prevents the addition of water or SO2. Polymeric pigments may also have increased color intensity due to intramolecular copigmentation.86 Overall, reactions of acetaldehyde lead to a shift in color from red/purple to tawny/brick red, enhancement of color intensity, and resistance to bleaching by pH changes and SO2 as pyranoanthocyanins and flavanyl-pyranoanthocyanins are formed from monomeric anthocyanins.1,15,66,73,119

1.3.2 Tannins

1.3.2.1 Description of Wine Tannins

Condensed tannins, or proanthocyanidins, are oligomeric polyphenols composed of flavan-3-ol subunits (catechins). While tannins are common to many foods including tea, chocolate, and many fruits, they are especially important to grape and wine quality.120 They are also known to confer many health benefits upon consumption.120–122 Grapes contain between 1 and 4 g/L condensed tannin which are found in the skin and seeds of grapes.1 These tannins vary in size (degree of polymerization or DP) and subunit composition (e.g., stereochemistry, substitution).123 The most common catechin subunits found in grape tannins are (+)-catechin, (-)- epicatechin, (-)-epicatechin gallate (ECG), and (-)-epigallocatechin (EGC) (Figure 1.13). The composition of tannins depends on their location in the grape; seed tannins contain no EGC while skin tannins do, skin tannins have 5% ECG while seed tannins have 30%.86

OH OH OH OH

HO O HO O

OH OH OH OH

(+)-Catechin (-)-Epicatechin OH OH OH OH HO O

HO O OH O OH OH OH O OH OH OH

(-)-Epigallocatechin (-)-Epicatechin gallate

Figure 1.13. Flavan-3-ol monomers found in grapes.

Tannins are present in grapes in a wide range of DP, from the monomer (DP 1) to large oligomers up to DP 80 found in grape skin (Figure 1.14).98,124,125 Tannins in grapes are B-type proanthocyanidins, indicating that they have a single bond (C4-C6 or C4-C8) between subunits.

However, A-type proanthocyanidins have been observed in wine.126 These A-type compounds contain an additional ether bond (C5 or C7 and C2) between subunits and can be formed from B- type via a radical process.1

OH OH

HO O

4 OH OH OH OH OH n OH 8 HO O HO O

4 OH OH 3 OH OH OH OH 8 HO O

OH OH

Figure 1.14. Flavan-3-ol monomer and oligomer (i.e., a condensed tannin).

1.3.2.2 Role of Tannins in Wine

Tannins are extracted into red wine during the primary alcoholic fermentation as an increase in ethanol concentration enables dissolution of these large, relatively nonpolar molecules. Tannin extraction requires higher alcohol concentration and temperature than other polyphenols (e.g., anthocyanins) due to their lower aqueous solubility.77 Practices such as extended maceration can be used to increase tannin extraction from grapes during wine

77 production. Higher alcohol content, SO2, temperature, and skin contact time are known to increase tannin extraction.75 Tannin concentration in wine has been reported to be between 300 and 700 mg/L catechin equivalents, although quantification methods for tannins are too inconsistent to give reliable concentrations in wine.64,127,128 23

Tannins, and other polyphenols, are important in red wine for their ability to eliminate free radicals and to chelate metal ions, thus contributing to the chemical stability of wine.1,123

Tannins are also capable of interaction with many proteins. Their interaction with proteins is involved in many of the benefits of tannins for plants and humans.120–122 Tannin-protein interactions can also lead to haze formation in white wine.120 The ability of tannins to bind with proteins is arguably most important for their role in the astringency of red wine based on their interaction with salivary proteins.129

Tannins contribute to the taste and mouthfeel of wine. In particular, larger tannins are perceived as astringent. Astringency is the sensation of drying and/or puckering in the mouth, a sensation that is desirable to a certain extent. Astringent components in wine contribute to a velvety mouthfeel but, when present at exceedingly high concentrations, can lead to wines that are perceived as “out of balance” and having a grainy or powdery mouthfeel.130,131 Astringency has been an area of intense interest due to the complexities in its characterization and molecular mechanism.132 The sensation of astringency is likely due to a loss of lubrication upon the precipitation of protein-tannin complexes.133,134 The proteins that take part in this interaction are primarily proline-rich proteins (PRPs)77,135 that are common in saliva with over 20 previously identified. Proline, glycine, and glutamine make up 70 to 80% of PRPs.136 PRPs are generally unstructured and more likely to be randomly coiled.137

Astringency is dependent on the tannins in a wine as well as the matrix of the wine.138

Tannin size is an important determining factor for the perceived astringency. Small tannins are typically more bitter than astringent while large tannins are more astringent and only slightly bitter.82,139,140 The interaction between tannins and proteins has been shown to be a function of

DP.141–143,143,144 While astringency has been shown to be a function of tannin length, other characteristics of tannin structure can influence the interaction with proteins.82 Astringency also

increases with the degree of galloylation.77,145 Differences in the conformation of tannins and their colloidal state have been shown to affect protein interactions.138,146–148 Ethanol concentration, pH, acids, and polysaccharides have also been shown to affect perceived astringency due to their effect on tannin interactions.149–151

Astringency is difficult to characterize, even with trained panelists, due to its complexity.152 Assays based on protein and polysaccharide precipitation have been shown to have high correlation with sensory ratings.128,153,154 The Harbertson-Adams (HA) assay is one assay that has proven to be relatively simple to implement and has been found to have correlations as high as 0.9044 with sensory ratings of drying.128 The HA assay is based on the precipitation of

Bovine Serum Albumin (BSA) by tannins in wine samples.155,156 The HA Assay has also been used to corroborate data on the role of tannin size in determining protein precipitation.141 In vitro assays are useful methods for characterizing astringency for winemakers and researchers without involving the complications of sensory analysis.

While native tannins are integral to the mouthfeel of wine, the tannins present transform significantly with time and oxidation. These reactions are critical to the important role that tannins play in color stability. Tannin modifications also lead to decreasing the astringency and increasing the overall complexity of the mouthfeel of red wine. Acetaldehyde is essential to these beneficial changes for the improved quality of red wine.

1.3.2.3 Reactions of Tannins with Acetaldehyde

The benefits of acetaldehyde for red wine were once thought to be mainly based on the effect on color.91 However, it is now known that reactions of tannins with acetaldehyde are critical to astringency softening as well as color stability of red wine. The effect of acetaldehyde

on lowering astringency in red wine was first based on results seen in the ripening of persimmon fruits, where the reaction of flavanols with acetaldehyde led to their precipitation.157 The principal mechanism for astringency softening was assumed to be due to reaction between tannins and anthocyanins rather than between tannins and other flavanols. However, Fulcrand et al. first showed the polymerization of flavan-3-ols due to bridging by acetaldehyde by observing oligomers up to the hexamer from epicatechin.93 Saucier then showed the presence of an ethyl- bridged catechin dimer in red wine confirming the important role of acetaldehyde in red wine.158,159

Flavanols have two reactive sites on the A ring which are available for bridging, C6 and

C8.160 The reaction of flavanols with acetaldehyde is known to be faster at lower pH due to the important step of protonation of the aldehyde (Figure 1.3). Epicatechin reacts more quickly than catechin, though both flavanols will form oligomers readily in the presence of acetaldehyde.161 In mixtures, these monomers can form heterogeneous oligomers and can also undergo cleavage to vinyl catechins and subsequent rearrangement.160,161 Like native tannins, ethyl-bridged flavanols are perceived as astringent.82

Direct evidence of acetaldehyde-mediated modifications of tannins is difficult due to their large size and complexity. Drinkine et al. created an indirect method to quantify the number of ethylidene bridges in tannins using a modified phloroglucinolysis method.162 Using this method, Drinkine et al. were able to correlate the proportion of ethylidene linkages to wine age and pigmented polymers.163 This aids in illustrating the relationship between acetaldehyde reactions and the desirable benefits of oxidation and aging.

Tannin modifications from oxidation have been shown to decrease astringency.82,86,134,153

Since polymerization should lead to an increase in MW and astringency, other mechanisms may be important. Acid-catalyzed cleavage of tannins to generate small species may contribute to the 26

loss in astringency.86 Cleavage products also serve as intermediates in polymerization reactions with flavanols or anthocyanins.164 Reactions leading to the incorporation of anthocyanins into tannins may contribute to lower astringency though this has not been shown conclusively.94 This may be due to anthocyanins acting as terminal subunits, preventing further polymerization and contributing to the formation of lower MW species.21,102 Oxidation may also decrease astringency by forming large tannins that precipitate and therefore do not contribute to astringency.161,165

These reactions could even lead to changes in the overall shape and availability of hydroxyl groups upon modification of tannin structure which would alter their protein interaction.

Several factors make it difficult to pin down the exact mechanisms occurring in wine.

The estimation of tannin mDP becomes increasingly inaccurate in older wines as modifications occur.86 Oxidized wines also contain a highly variable mixture of new species formed with new compounds still being identified.98 The effect of oxidation, including reactions with acetaldehyde, on astringency is likely a combination of precipitation of large polymers formed, acid-catalyzed cleavage to form smaller species, and the incorporation of anthocyanins as terminal subunits.

However, the overall improvements in red wine based on oxidation reactions have been demonstrated in many wine studies.

1.3.3 Impact of Reactions of Acetaldehyde with Flavonoids on Wine Quality

The overall impact of acetaldehyde on red wine is difficult to characterize due to a lack of studies that directly examine this issue. Instead, there is a large body of work looking at the effect of MOX, an indirect method of acetaldehyde generation, on wine quality.18,35 In the studies that monitor acetaldehyde concentration from oxygenation, acetaldehyde appears to be readily consumed as concentrations do not increase significantly without excessive oxygen

exposure.34,36,37 There are several important factors that affect the success of MOX treatment.

Lower pH conditions, lower SO2 concentrations, and higher phenolic concentrations have all been shown to lead to greater improvements from MOX.21,166–168 The ratio of tannins to anthocyanins also affects the changes seen.21,169

Studies have shown that MOX improves color stability as seen by decreases in monomeric anthocyanins and increases in polymeric pigments and color density and intensity.21,36,119,153,166,167,170–173 Pyranoanthocyanins (vitisins) and ethyl-linked polymeric pigments have been observed as a result of MOX.73,119,166,174 Ethylidene-bridged flavanols have also been seen due to MOX treatment.166,169 The improvements in color stability seen after MOX persist through aging.36,171

Generally, MOX also improves the mouthfeel of red wine.153,175 However, studies report that tannin size may increase167,175 or decrease37 in conjunction with astringency softening. These conflicting reports are likely due to difficulties in characterizing tannins with greater oxidation.

DP of wine tannins is frequently measured using acid-catalyzed cleavage with subsequent analysis of known monomeric subunits (e.g., thiolysis, phloroglucinolysis).176,177 Decreases in mDP of tannins determined by phloroglucinolysis can indicate a larger proportion of small MW species or lower conversion of tannins to known phloroglucinol-flavanol subunits. This includes anthocyanins, which are not included in the number of subunits, and any subunits bound by linkages that are not acid-labile. While these methods are indirect, monitoring the amount of tannin converted to known subunits (i.e., percent yield or conversion) can be a helpful measure of tannin modification.37,178

Few studies have looked at the direct role of acetaldehyde in red wine systems or the implications of exogenous acetaldehyde. Aleixandre-Tudo et al. have examined the use of acetaldehyde addition (17.8 mg/L in 22 additions) as a means of simulating MOX post-bottling. 28

They observed improvements in nearly all parameters relating to color stability and tannin profile, including increased polymeric pigments and decreased astringency (gelatin index).179,180 There still remain many unanswered questions pertaining to the use of exogenous acetaldehyde in red wine production and the reactivity of acetaldehyde in a wine system.

1.4 Significance and Hypotheses

1.4.1 Significance

Wine is an economically significant product whose value – perhaps more so than other food products – is strongly linked to perceived quality. The global wine industry has grown in recent years with a total revenue of $281 billion in 2014 and an average growth of 3% per year between 2010 and 2014.181 Wine quality is an especially important driver for the industry as it shapes consumer choices.182 In order to achieve high standards for quality, winemakers use techniques to optimize characteristics of their wine – taste, aroma, color, and mouthfeel. Novel methods for improving wine attributes, especially those that are inexpensive and easy to implement, are of great interest for those in wine research and production.

Currently, winemakers frequently use oxygenation in order to improve wine color stability and mouthfeel. Oxygen exposure is typically achieved in the most controlled manner that is feasible (e.g., barrel aging, micro-oxygenation). These methods are meant to take advantage of the formation of acetaldehyde so that it will react with wine flavonoids. However, there are many risks associated with the intentional introduction of oxygen into wine (or any food product for that matter), including the loss of desirable aroma, non-enzymatic browning, and microbial instability. Methods of oxygenation are also typically expensive in terms of labor and

infrastructure, and require skill and knowledge to implement correctly. The studies described here address the use of exogenous acetaldehyde as a technique for winemakers to obtain the benefits of oxidation without the risks of oxygen exposure.

1.4.2 Hypotheses and Aims

I hypothesize that exogenous acetaldehyde will improve the color stability of red wine.

Additionally, exogenous acetaldehyde will contribute to advantageous oxidation reactions in the presence of sulfur dioxide. Furthermore, exogenous acetaldehyde treatment will prove to be as effective as, if not more effective than, oxygenation of a red wine in observed improvements in pigment characteristics.

In order to test these hypotheses, I will investigate the following aims:

1. Determine the ability of exogenous acetaldehyde treatment in a red wine to improve color stability and protein precipitation

2. Characterize the effect of wine components on reaction of acetaldehyde with flavonoids

3. Evaluate the reactivity of aldehydes from the sulfonate adducts of aldehydes with bisulfite

4. Assess exogenous acetaldehyde treatment compared to oxygenation and the acetaldehyde- bisulfite adduct for improving red wine color stability

Chapter 2

Exogenous Acetaldehyde as a Tool for Modulating Wine Color and Astringency during Fermentation

Published as:

Sheridan, M.K. & Elias, R.J. Exogenous acetaldehyde as a tool for modulating wine color and astringency during fermentation. Food Chem. 2015, 177, 17–22.

2.1 Abstract

Wine tannins undergo modifications during fermentation and storage that can decrease their perceived astringency and increase color stability. Acetaldehyde acts as a bridging compound to form modified tannins and polymeric pigments that are less likely to form tannin- protein complexes than unmodified tannins. Red wines are often treated with oxygen in order to yield acetaldehyde, however this approach can lead to unintended consequences due to the generation of reactive oxygen species. The present study employs exogenous acetaldehyde at relatively low and high treatment concentrations during fermentation to encourage tannin modification without promoting potentially deleterious oxidation reactions. The high acetaldehyde treatment significantly increased polymeric pigments in the wine without increasing concentrations of free and SO2-bound acetaldehyde. Protein-tannin precipitation was also significantly decreased with the addition of exogenous acetaldehyde. These results indicate a possible treatment of wines early in their production to increase color stability and lower astringency of finished wines.

2.2 Introduction

The tannin profile of wine (i.e., tannin composition, size, and shape) changes during the course of winemaking, and can be altered within the fruit, through fermentation, and during maturation. These changes, primarily consequences of non-enzymatic oxidation reactions, are known to affect the perception of astringency in the mouth. Astringency is the sensation of drying and/or puckering in the mouth, a sensation that is desirable to a certain extent. Astringent components in wine contribute to a velvety mouthfeel but, when present at exceedingly high concentrations, can lead to wines that are perceived as “out of balance” and having a grainy or powdery mouthfeel.183,184 The main polyphenols contributing to astringency in wine are tannins, namely monomeric and polymeric flavan-3-ols.185,186 Previous studies have shown a direct relationship between astringency and tannin size expressed as mean degree of polymerization

(mDP) and an inverse relationship with storage and oxidation.185,187–189

The effect of tannin modification on astringency perception is likely due to the mechanism of protein precipitation in the mouth.189 Salivary proteins are found as random, loose coils and are known to be proline-rich proteins (PRP). The exposed proline residues of the salivary protein interact with tannin hydroxyl groups via hydrogen bonds while hydrophobic interactions (Van der Waals and π-π stacking) contribute to the formation of protein-tannin complexes. Coiling of the PRP around tannins leads to a decrease in the space occupied by the protein-tannin complex, thus yielding a more compact and spherical shape overall. This interaction usually involves multiple binding sites for tannins on the protein. Next, cross-linking between the protein-tannin complexes causes aggregation and, finally, further aggregation leads to complexes that are large and insoluble (Figure 2.1).190,191

Tannin

Salivary Protein Protein-Tannin Aggregation Precipitation of Large Aggregates Complex

Figure 2.1. Mechanism of astringency by tannin-protein complex formation, aggregation, and precipitation.

The sensation of astringency from these tannin interactions is known to decrease with oxidative modification and with subsequent changes in the availability of tannin hydroxyl groups.

These modifications include degradation through cleavage reactions, internal or external condensation reactions with other flavanols, and condensation with anthocyanins.186,192–195

Condensation reactions can occur directly by formation of a covalent bond between flavanol or anthocyanin subunits, or indirectly via a molecular bridge (e.g., acetaldehyde). When monomeric anthocyanins are bound to other compounds directly or indirectly, polymeric pigments are formed

(Figure 2.2).186,196

OH OH OH OH OH HO O HO O OH OH H C CH OH OH 3 OH OH OH HO O HO O O-gl OH O-gl OH

flavanol-anthocyanin flavanol-ethyl-anthocyanin Figure 2.2. Direct and indirect condensation products of flavanols and anthocyanins occurring in wine.

Acetaldehyde is formed in wine both as an enzymatically-derived byproduct of yeast metabolism, and as a non-enzymatic oxidation product of ethanol.197 Acetaldehyde concentrations, therefore, rise as the fermentation progresses and as a wine is exposed to oxygen.

Most of this acetaldehyde reacts quickly with free bisulfite (when present), with the remainder either volatilizing or reacting with other compounds such as tannins.198–201 Acetaldehyde-bridged tannins, especially those with anthocyanins incorporated with them, have been shown to have lower perceived astringency than unmodified tannins.186,187 Polymeric pigments – the result of anthocyanin condensation reactions – also have the trait of being resistant to bleaching by sulfur

200,202 dioxide (SO2). This characteristic is used as a measure of tannin modification and wine age.

Acetaldehyde has been used indirectly to modify tannins through “controlled” oxidation during micro-oxygenation operations.203–205 While micro-oxygenation is thought to offer a more controlled introduction of oxygen compared to barrel aging, it can also yield deleterious effects if the rates of oxygen consuming reactions do not exceed oxygen addition rates. Such conditions can harm wine color, cause the loss of desirable aromas and formation of undesirable aromas, and even promote aerobic bacteria.206 A method of treatment where the desirable reactions of acetaldehyde are achieved without the risks of oxygen addition may be the key to controlled oxidation of red wines.

Though studies have shown the link between acetaldehyde condensation products and decreased astringency, no studies have used exogenous acetaldehyde as a treatment during wine production to directly affect astringency. Treatment with acetaldehyde during a red wine fermentation should mimic model studies in the formation of stable color and modified tannins.

This study aims to examine the effect of exogenous acetaldehyde during a wine fermentation on the formation of polymeric pigments and on astringency.

2.3 Materials and Methods

2.3.1 Materials

Folin & Ciocalteu’s phenol reagent, gallic acid, sodium dodecyl sulfate, triethanolamine, maleic acid, sodium chloride, (+)-catechin hydrate, and bovine serum albumin were purchased from Sigma Aldrich (St. Louis, MO). Sodium carbonate and ferric chloride hexahydrate were purchased from Mallinckrodt Pharmaceuticals (St. Louis, MO). Potassium metabisulfite, acetaldehyde, and tartaric acid were purchased from Alfa Aesar (Ward Hill, MA). Sodium hydroxide (10.00 N) and 200 proof ethanol were purchased from VWR (Radnor, PA). Glacial acetic acid was purchased from J.T. Baker (Phillipsburg, NJ). Hydrochloric acid was purchased from EMD Chemicals (Gibbstown, NJ). Water was purified through a Millipore Q-Plus

(Millipore Corp., Bedford, MA) purification train.

Microscale fermenters were used, which were constructed based on a previously reported design.207 Fermenters were constructed using 4-L jars (Packaging Options Direct, St. Louis, MO),

Teflon-lined caps (QEC, Beaver, WV), a fermentation airlock (Midwest Supplies, St. Paul, MN), and a perforated LDPE lid fit to the jars so as to act as a screen. The fermenters were not completely filled to prevent overflow.

2.3.2 Wine Production

Hand-harvested Cabernet Franc (V. vinifera) grapes were sourced from the Pennsylvania

State University’s experimental research vineyard in North East, PA. Grapes were processed on the same day they were received. The fruit (42 kg) was crushed and destemmed (Eno-15,

Enoitalia, San Martino Buon Albergo, Italy) before separation. The starting must chemistry for all

wines was characterized: pH 3.28; titratable acidity, 5.78 g/L as tartaric acid; soluble solids, 20.3

°Brix; yeast assimilable nitrogen, 144 mg N/L. The must was then separated into skin and juice fractions before being further divided for each group by weight; this was done to ensure that an equivalent ratio of skins, seeds, and juice was maintained between treatments. Equal weights of skins and juice were used in each of the three groups and placed in a fermenter to begin the fermentation. Fermentations were initiated before separation into replicates to ensure similar initial fermentation rates were achieved. All groups were treated with GoFerm yeast nutrient (30 g/hL) and inoculated with LCV GRE yeast (25 g/hL, Lallemand). The low and high acetaldehyde treatment groups received their first addition of acetaldehyde (25 mg/L and 250 mg/L respectively) from a concentrated stock solution of acetaldehyde (75 g/L). Juice samples were taken for analysis and frozen at -80 °C.

After two days of fermentation at 21 – 23 °C, 20 mL aliquots were taken from each fermenter and the treatment groups were separated into four 4 L microfermenters each (i.e., four replicates). Musts were separated into juice and skins before being divided by weight as previously described and acetaldehyde treatments were added.

Samples (20 mL) were taken from the fermenting must and acetaldehyde additions were continued every two days for the treatment groups (4x25 mg/L and 4x250 mg/L total over 8 days of fermentation). The fermentation was monitored by following the loss of soluble solids (°Brix) by hydrometry; residual sugar was analyzed at the end of fermentation using a reducing sugar assay (Clinitest, Bayer, Tarrytown, NY). Once dry (<1/2% RS), wines were individually pressed by hand in order to remove the pomace. All samples were stored at -80 °C until analysis.

2.3.3 Acetaldehyde Measurement

Acetaldehyde was measured enzymatically using a commercial test kit (Megazyme, UK) according to the kit’s instructions. Wine samples were analyzed in triplicate for free and SO2- bound acetaldehyde.

2.3.4 Pigment Analysis

The modified Somers assay was performed using a high throughput procedure.208 Wine samples were centrifuged at 2500 x g for 5 min prior to analysis. Solutions were incubated in 1.5- mL capacity microcentrifuge tubes before being transferred to 96-well plates (Greiner Bio One

UV-Star, Monroe, NC) for absorbance readings at 280, 420, and 520 nm using a Multiskan GO microplate reader (Thermo-Scientific, Waltham, MA). All Somers color parameters were calculated as previously reported. Wine samples were analyzed in duplicate.

2.3.5 Total Phenolics Measurement

Total phenolics were measured according to previously described methods and were expressed in g/L of gallic acid equivalents (GAE).209 Wines were analyzed in triplicate by reaction with Folin-Ciocalteau reagent in 1.5-mL microcentrifuge tubes. After the addition of sodium carbonate solution and subsequent incubation, samples were transferred to a 96-well microplate for measurement at 765 nm using a Multiskan GO microplate reader.

2.3.6 Protein Precipitation Analysis

The Harbertson-Adams assay of Bovine Serum Albumin (BSA) precipitation was used as a method to determine tannin-protein precipitation. Wine samples were analyzed using the microplate method previously published.210 Readings at 280, 420, 520 nm were taken in 96-well microplates using a Multiskan GO microplate reader.

2.3.7 Salivary Protein Precipitation Analysis

Resting saliva samples were collected from eight non-smoking volunteers (four males and four females) after being asked not to consume food or beverage for 1 h before sample collection. All procedures for saliva collection were approved by the Pennsylvania State

University Institutional Review Board (protocol number #45774). Samples were stored on ice before being combined and centrifuged at 10,000 x g for 10 min. The supernatant was removed and stored at -80 °C in 1.5-mL aliquots for use in subsequent binding assays. Binding assays of saliva with wine samples were completed in triplicate as described previously.211 The supernatant was removed and the pellet from the saliva-tannin precipitation was analyzed using the method of ferric chloride reaction as described in the Harbertson-Adams assay and expressed in (+)-catechin equivalents.210

2.3.8 Statistical Analysis

One-way ANOVA analysis was paired with Tukey’s test for determining significance between samples. Differences of p < 0.05 were considered significant. Statistical analysis was performed using GraphPad Prism 6.0 (GraphPad, La Jolla, CA).

2.4 Results and Discussion

All fermentations began with the same Cabernet Franc juice so as to limit any differences due to initial must chemistry. SO2 was not added to wines at any point of the study.

Acetaldehyde was added starting at the beginning of fermentation to minimize reaction between

SO2 and acetaldehyde. Although exogenous SO2 was not added, SO2 formed by yeast during fermentation would bind with acetaldehyde and hinder its reaction with tannins.198 All treatment groups fermented at the same rate as determined by soluble solids analysis (data not shown), indicating that exogenous acetaldehyde at these concentrations does not inhibit yeast fermentation.

2.4.1 Acetaldehyde Concentrations in Must and Wine

Acetaldehyde concentrations were chosen based on sensory thresholds and U.S. federal regulations. The total acetaldehyde added to the low acetaldehyde treatment (4x25 mg/L) is below sensory detection thresholds of 100 mg/L and the federal limit of 300 mg/L.212 The high acetaldehyde treatment (4x250 mg/L) involved adding more acetaldehyde than is allowed by law.

Acetaldehyde was measured in the completed wines (post-primary fermentation) using an enzymatic assay of both free and SO2-bound acetaldehyde to give a total acetaldehyde concentration. There were no significant differences in total acetaldehyde concentration between treatment groups. Acetaldehyde concentrations post-primary fermentation were (mean ± standard deviation) 29.8 ± 7.2 mg/L for the control group, 32.6 ± 5.7 mg/L for the low acetaldehyde group, and 30.8 ± 1.6 mg/L for the high acetaldehyde group. The remainder of the exogenous acetaldehyde added to the fermenting wines is assumed to have volatilized during the course of fermentation or reacted with tannins and anthocyanins, leaving a small amount present in the

wine. The amount of acetaldehyde incorporated into tannins and polymeric pigments was not directly measured; however, tannin modifications were observed, likely due to the formation of acetaldehyde bridges and are discussed below. These results also indicate that acetaldehyde addition over the legal limit of 300 mg/L does not significantly increase the amount of acetaldehyde in the finished wine.

2.4.2 Effect of Exogenous Acetaldehyde on Wine Color

The modified Somers assay was used as a measure of the degree of polymeric pigment formation due to direct and indirect (i.e., acetaldehyde-mediated) condensation reactions. There were expected differences in the chemical age and SO2-resistant pigment parameters due to the ability of acetaldehyde to bridge anthocyanins and tannins. The chemical age parameters are a ratio of polymeric pigments to monomeric anthocyanins, which indicates the extent of direct and indirect condensation reactions that anthocyanins have taken part in.200 Chemical age describes the ratio of absorbances at 520 nm: absorbance in the presence of SO2 to absorbance in the presence of acetaldehyde (chemical age 1) or absorbance in the presence of HCl (chemical age

2).208 As these reactions are typically associated with wine maturation, the parameters are referred to as chemical ages (unitless). SO2-resistant pigments are a similar parameter but represent a more simple and direct measure of polymeric pigments. The results of the modified Somers assay are shown in Table 2.1.

Table 2.1. Color parameters from the HTP modified Somers assay of treatment groups post- primary fermentation.a

Control Low Acetaldehyde High Acetaldehyde

Chemical age 1 0.277 ± 0.006 a 0.273 ± 0.007 a 0.295 ± 0.18 b Chemical age 2 0.038 ± 0.005 a 0.044 ± 0.005 a,b 0.049 ± 0.005 b Degree of ionization of 11.6 ± 2.1 13.3 ± 1.6 12.9 ± 1.8 anthocyanins (%) Total anthocyanins (mg/L) 516 ± 69 442 ± 68 450 ± 36 Color density (au) 6.4 ± 0.2 6.3 ± 0.3 6.6 ± 0.6 Color density, SO - 2 6.1 ± 0.1 a 6.1 ± 0.2 a,b 6.6 ± 0.6 b corrected (au) Hue 0.606 ± 0.008 0.604 ± 0.004 0.611 ± 0.011 a a b SO2-resistant pigments (au) 1.03 ± 0.01 1.03 ± 0.03 1.19 ± 0.08 Total phenolics (au) 144.9 ± 3.7 a 141.4 ± 3.5 a,b 140.5 ± 2.9 b

aAveraged results are shown ± one standard deviation of the mean, and results in the same row with different letters (a, b) are significantly different (p < 0.05).

Significant differences were observed both in terms of chemical ages and SO2-resistant pigments. For chemical age 1, the high acetaldehyde group was significantly higher at 0.295 than both the control and low acetaldehyde groups at 0.277 and 0.273, respectively. For chemical age

2, the high acetaldehyde group was significantly higher at 0.049 than the control at 0.038. These results indicate a shift from monomeric anthocyanins to SO2-resistant pigments with exogenous acetaldehyde treatment. Similarly, the high acetaldehyde group had a significantly higher amount of SO2-resistant pigments at 1.19 au than both the control and low acetaldehyde groups, both at

1.03 au. These data suggest that the high acetaldehyde treatment led to more extensive tannin modification, specifically the condensation of anthocyanins with themselves or tannins.

There were no significant differences in the calculated values for other pigment variables: the degree of ionization of anthocyanins, total anthocyanins, color density, and hue. There was a minor but statistically significant difference in the values for SO2-corrected color density.

However, the total phenolics in the high acetaldehyde group were significantly lower at 140.5 au than the control group at 144.9 au, although we note the magnitude of this difference is small.

Tannin modifications could lead to larger oligomers that would precipitate out of the wine and acetaldehyde may increase the degree to which larger oligomers are seen in the wine. Though this is a separate mechanism than changes in the structures of the tannins, precipitation of tannins could be another route to decreased astringency through acetaldehyde treatment. While the

Somers assay uses absorbance at 280 nm as a measure of total phenolics, the Folin-Ciocalteau assay for total phenols was used as a more accurate analysis of any possible loss of phenolics.

The results of the Folin-Ciocalteau assay show that there is no significant difference in the total phenols between the groups. Total phenol concentrations, expressed as gallic acid equivalents, post-primary fermentation were (mean ± standard deviation) 123.7 ± 5.9 mg/L for the control group, 121.6 ± 5.9 mg/L for the low acetaldehyde group, and 118.0 ± 7.2 mg/L for the high acetaldehyde group. Loss of tannin by precipitation is therefore not a major factor in any changes in tannin content or astringency. Changes in total phenols may be of concern with greater acetaldehyde treatment or over time due to the observed trend, but there were no significant differences in total phenols seen at the end of the primary fermentation.

SO2-corrected color density values observed were significantly different; however, given the magnitude of this difference (6.6 ± 0.6 au compared to 6.1 ± 0.1 au), its practical meaning could be argued. Overall, the Somers parameters indicate significant changes in the tannin profile of the high acetaldehyde treated wine, especially as these data point to an increase in polymeric pigments.

2.4.3 Effect of Exogenous Acetaldehyde on Tannin-Protein Interactions

Protein precipitation assays were also performed on the finished wine samples in order to assess changes in astringency as a function of acetaldehyde treatment. The Harbertson-Adams assay was used as a basis for measuring protein-tannin interactions with a standard method. This assay uses BSA as a model protein to precipitate tannins from a wine sample. The amount of tannin in the precipitated pellet as measured by ferric chloride reaction has been shown to have high correlation with sensorial astringency measurements – as high as R2=0.9044.213 The results of the Harbertson-Adams assay using the wine samples show a significant difference between the control and high acetaldehyde groups (Figure 2.3).

300

a 200 ab b

100

Tannin Concn (mg/L CE) (mg/L Concn Tannin 0

Low High Control

The high acetaldehyde group had significantly lower tannin-BSA precipitate compared to the control. This high acetaldehyde treatment led to decreased interactions between tannins and

BSA, which one could extrapolate to a potential decrease in perceived astringency, as well.

However, there may be differences between tannin interaction with BSA and an actual salivary protein. Salivary proteins are unique in their composition – 70% of salivary proteins are PRPs.189

Their high proline content is known to play a significant role in the mechanism of precipitation with tannins. BSA, on the other hand, is not a PRP and therefore may interact with tannins by different mechanisms compared to salivary proteins. To assess the interaction with salivary proteins, a modified assay was used that combined a sensorially relevant binding assay with tannin quantification by ferric chloride reaction.211

These results (Figure 2.4) corroborate our finding that the high acetaldehyde treatment group has significantly lower amounts of tannin-protein precipitation. The trend and values are similar to precipitation with BSA, though small differences may be enough to increase the correlation with sensorial astringency measurements.

300 a ab

200 b

100

Tannin Concn (mg/L CE) (mg/L Concn Tannin 0

Low High Control

Figure 2.4. Tannin content of saliva-tannin precipitates quantified by ferric chloride reaction in (+)-catechin equivalents (CE). Error bars represent one standard deviation of the mean, and results in with different letters (a, b) are significantly different (p < 0.05).

2.5 Conclusions

A high acetaldehyde treatment (4 additions; 250 mg/L per addition) during the fermentation of a red wine resulted in increased tannin modification compared to an acetaldehyde-free control. Exogenous acetaldehyde treatment to fermenting musts contributed to an increase in polymeric pigments, which is likely due to acetaldehyde-mediated condensation

reactions between tannins and anthocyanins. The low acetaldehyde treatment was seemingly too low to affect any significant changes from the control group though significant differences were seen in the high acetaldehyde treatment group. Acetaldehyde treatment did not result in a higher amount of free and SO2-bound acetaldehyde in the wine. Furthermore, high acetaldehyde treatment lowered tannin-protein precipitation in both the Harbertson-Adams assay and a modified saliva precipitation assay. These results suggest that the addition of exogenous acetaldehyde can be used to modulate wine astringency and color, and supports previous assumptions that tannin modifications and the formation of polymeric pigments via non- enzymatic oxidation lead to a decrease in astringency.

2.6 Acknowledgements

The authors thank Denise Gardner for her assistance with wine production as well as for her insightful discussions throughout the study.

Chapter 3

Reaction of Acetaldehyde with Wine Flavonoids in the Presence of Sulfur Dioxide

3.1 Abstract

Acetaldehyde is responsible for many of the beneficial changes seen in red wines with oxidation. Ethylidene bridges are formed between flavonoids upon their reaction with acetaldehyde. These bridged compounds contribute to improvements in color stability and SO2- resistant pigments. The reactions of acetaldehyde with flavonoids (catechin, tannins from grape seed extract, and malvidin-3-glucoside) were examined in a model wine system. Lower pH significantly increased the rate of reaction with acetaldehyde while dissolved oxygen did not affect the rate. In systems containing SO2, the rate of reaction of acetaldehyde with catechin was slowed but was not prevented until SO2 was in great excess. There were also significant improvements in color stability observed after treatment with acetaldehyde and equimolar SO2.

These results demonstrate that acetaldehyde is reactive, not inert as previously assumed, in its sulfonate form. Products of the reaction of flavonoids with acetaldehyde were characterized using

MALDI-TOF MS. Ethyl-bridged catechin nonamers were observed as well as anthocyanin and pyranoanthocyanin derivatives of catechin and tannin oligomers. The results of this work illustrate the significance of acetaldehyde reactions in forming stable pigments in wine and the reactivity of acetaldehyde from its sulfonate form.

3.2 Introduction

Flavonoids in red wine are directly related to many important indicators of wine quality.

Changes in these compounds during wine production are also critical to improving quality parameters including mouthfeel and color stability.86 Condensed tannins and anthocyanins take part in reactions which lead to decreases in astringency and the formation of stable, polymeric pigments.66,86 Many of the flavonoid modifications that occur in red wine are due to oxidation wherein condensation reactions bind molecules together. Acetaldehyde, itself a major product of wine oxidation, drives some of the beneficial reactions of flavonoids by forming ethyl-bridged condensation products.8 Oxygen is reduced by several metal-catalyzed steps leading to the formation of acetaldehyde.8,10,31 This process forms reactive intermediates including ortho- quinones and 1-hydroxyethyl radicals which can lead to deleterious effects on wine aroma and color.10,38,39 Yeast metabolism also produces acetaldehyde as a side product of the alcoholic fermentation.6

Recent studies in our group and elsewhere have shown that exogenous acetaldehyde can be used to improve red wine color stability and astringency (Chapter 2).24,179,180,214 The use of exogenous acetaldehyde provides benefits for the wine without the risks of reactive intermediates formed from oxygen exposure. Adducts formed by acetaldehyde are found to be bridged by an ethylidene moiety (Figure 3.1),92,93 which can occur via C6 or C8 on the A ring of the flavonoid molecule.74,160,215 Ethylidene bridging reactions can involve tannins and anthocyanins to form modified tannins as well as polymeric pigments.74,92 Bridged products from reaction with acetaldehyde have been found in red wine and in model wine systems,93,159,163,216 and evidence for ethylidene bridged products of native tannins has been observed indirectly by pholoroglucinolysis;162 however, these compounds are difficult to observe without fragmentation during ionization for MS. 47

HO R1 O R2 OH S HO O H O HO R2 O HO OH H3C S O H3C H OH OH O

R3 R4

Figure 3.1. Reaction of acetaldehyde with bisulfite and with representative flavonoids to form an ethylidene-bridged adduct.

Another important fate of acetaldehyde under wine conditions is its reaction with bisulfite ions (Figure 3.1). Sulfur dioxide (SO2) is added to wine to prevent faults due to chemical and microbial instability. Bisulfite insures the chemical stability of wine by quenching oxidation reactions; it reacts readily with hydrogen peroxide and quinones formed during the reduction of oxygen.42,58,59 One of the main fates of bisulfite in wine is the binding of aldehydes, especially acetaldehyde as it is the most common.32 This reaction is known to be incredibly fast (98% bound

-6 61,217 in 90 min at pH 3.3) and the resulting adduct is strongly bound (Kd 2.06 x 10 at pH 3.5).

Due to the large preference for the formation of the acetaldehyde-bisulfite adduct, 1- hydroxyethanesulfonate, previous studies have largely discounted the possibility that

42,119,167 acetaldehyde remains reactive in the presence of SO2. However, there has been evidence for the antioxidant activity of this sulfonate in beer; the formation of radicals from ethanol

62 oxidation was slowed by the sulfonate indicating activity similar to SO2 itself. The activity of acetaldehyde from the sulfonate has not been examined.

The present work aims to characterize the reactions of acetaldehyde with flavonoids in a model wine system using wine-relevant concentrations of acetaldehyde, model flavanols, and anthocyanins. For the first time, the activity of acetaldehyde in bridging flavonoids will be

assessed in the presence of equimolar bisulfite, when it has been assumed to be in an inert form as a sulfonate.

3.3 Materials & Methods

3.3.1 Materials

(+)-Catechin hydrate was purchased from Sigma Aldrich (St. Louis, MO). Malvidin-3- glucoside was purchased from Extraynthese (Genay, France). Grape seed extract was donated from San Joaquin Valley Concentrates (Fresno, CA). Potassium metabisulfite, acetaldehyde, and tartaric acid were purchased from Alfa Aesar (Ward Hill, MA). Sodium iodide and LC/MS grade formic acid were purchased from Fisher Scientific (Pittsburgh, PA). Acetone, sodium hydroxide

(10.00 N), 200 proof ethanol, 2,5-dihydroxybenzoic acid (DHB), and HPLC grade methanol were purchased from VWR International (Radnor, PA). Glacial acetic acid was purchased from J.T.

Baker (Phillipsburg, NJ). Hydrochloric acid and phosphoric acid were purchased from EMD

Chemicals (Gibbstown, NJ). Water was purified through a Millipore Q-Plus (Millipore Corp.,

Bedford, MA) purification train.

3.3.2 Reaction Mixture Preparation

Model wine was composed of 5% w/v acetic acid, 12% ethanol v/v, and pH adjusted using 10 N NaOH. Preliminary experiments showed that there was no effect of organic acid

(tartaric acid versus acetic acid) on the reaction of acetaldehyde with catechin. Acetic acid was therefore used for all experiments described here to simplify the preparation of samples for

MALDI-TOF MS as it is volatile and easily removed. 49

Model solutions were prepared with 20 mg/L acetaldehyde added from a stock solution,

1000 mg/L (+)-catechin or grape seed extract (GSE), and 500 mg/L M3G when needed. M3G

74,98 was used as a model anthocyanin as it is the most common in red wine. SO2 was added from a freshly prepared stock solution of sodium metabisulfite. Concentrations of SO2 in each sample were confirmed using an enzymatic assay for total SO2 (Megazyme, Chicago, IL). All experiments were completed in triplicate. Glassware was soaked in 5% nitric acid overnight and rinsed to minimize trace metals before use. Each reaction mixture was separated into 1 mL aliquots and stored at room temperature. Timepoints for each experiment were stored at -80 °C and thawed immediately prior to analysis.

All samples, except those specified as aerobic, were stored in an anaerobic chamber

(Anaerobe Systems, Morgan Hill, CA) with 95% argon/ 5% hydrogen gas and a palladium catalyst to remove residual oxygen. Anaerobic status of the chamber was confirmed using a

PreSens oxygen meter (PreSens Precision Sensing GmbH, Regensburg, Germany). All solutions were deoxygenated using argon gas before being introduced into the chamber and all materials were left in the anaerobic chamber overnight prior to starting experiments.

3.3.3 Flavonoid Analysis

Flavonoid concentrations were determined by HPLC using a Shimadzu system with

10ADvp pumps and a SIL-20AC HT temperature-controlled autosampler (4 °C). Separation was achieved on a reverse phase Supelcosil LC-18 column (4.6 mm x 150 mm, 5 µm; Supelco, Inc.,

Bellefonte, PA). Samples were filtered through 0.45 µm PTFE syringe filters prior to analysis.

The injection volume was 10 µL, column temperature was 30 °C, and flow rate was held at 1

mL/min for all samples. Concentrations were calculated based on external calibration curves for the analytes, (+)-catechin and M3G.

For catechin analysis, the mobile phase consisted of 1% phosphoric acid in water (A) and methanol (B). Catechin was eluted using an isocratic method of 20% B and detected at 280 nm.

For M3G analysis, the mobile phase consisted of 5% formic acid in water (A) and acetonitrile (B). M3G was eluted according to the following binary gradient: 0 min, 10% B; 14 min, 65% B; 14-16 min, 65% B, followed by reequilibration at 10% B prior to the next injection.

In samples containing GSE and M3G, the gradient was modified: 0 min, 10% B; 14 min, 65% B;

16 min, 90% B; 16-18 min 90% B, followed by reequilibration at 10% B prior to the next injection. M3G was detected at 520 nm.

3.3.4 Pigment Analysis

Samples containing M3G were analyzed using a modified Somers assay to characterize pigment changes. The modified Somers assay was performed using a high throughput procedure.218 Solutions were incubated in 1.5-mL capacity microcentrifuge tubes before being transferred to 96-well plates (Greiner Bio One UV-Star, Monroe, NC) for absorbance readings at

280, 420, and 520 nm using a Multiskan GO microplate reader (Thermo-Scientific, Waltham,

MA). All Somers color parameters were calculated as previously reported. Model wine samples were analyzed in duplicate.

3.3.5 Reaction Product Characterization

Samples for MALDI-TOF were prepared in acetic acid model wine with 200 mg/L acetaldehyde. After 4 weeks, aliquots of samples were dried under nitrogen and redissolved in methanol to 10 mg/mL catechin or GSE. Samples were spotted from a 2,5-dihydroxybenzoic acid

(DHB) matrix solution. DHB was recrystallized prior to sample preparation. Sample (1 µL, 10 mg/mL in methanol), sodium iodide (1 µL, 10 mg/mL in water), and DHB (10 µL, 200 mg/mL in acetone) were mixed and 0.5 µL was spotted on a polished stainless steel plate (Bruker corp,,

Santa Barbara, CA) for analysis. MALDI-TOF mass spectra were collected in positive reflectron mode using a Bruker ultrafleXtreme instrument (Bruker Corp., Santar Barbara, CA). Laser intensity was adjusted to optimize the resolution and signal for each spectrum. MALDI spectra were externally calibrated and acquired as an average of 1000 shots collected in random walk mode from m/z 500 to 4000.

3.3.6 Statistical Analysis

Two-way ANOVA analysis was paired with Tukey’s test for determining significance between samples. For all samples, the amount of catechin consumed at the final timepoint was used for the comparison. Differences of p < 0.05 were considered significant. Statistical analysis was performed using GraphPad Prism 6.0 (GraphPad, La Jolla, CA).

3.4 Results and Discussion

3.4.1 Reaction of Acetaldehyde with Flavanols

Catechin concentrations were monitored over time to examine the effect of several matrix parameters on its reaction with acetaldehyde. As has been shown by other researchers,161 there is a significant effect of pH on the rate of consumption of catechin (Figure 3.2). The greatest degree of monomeric catechin consumption after 14 days was observed at pH 2.0, whereas little to no catechin reacted at pH 3.5 during this time; however, it is probable that the catechin would have reacted at pH 3.5 by the same mechanism (albeit at a slower rate) given sufficient time. The significant increase in the rate of reaction at lower pH is due to the first step in the reaction of acetaldehyde with catechin, the protonation of acetaldehyde.93 Our results, along with others, show that lower pH leads to increased rate of reaction of acetaldehyde with flavonoids in real and model wine systems. In the interest of expediency, the following experiments described were performed at pH 2.5 so that an acceptable rate of catechin consumption could be achieved within a reasonable time frame.

1000

950

pH 2.0 pH 2.5 900 pH 3.0 pH 3.5 Catechin (mg/L) 850

800 0 4 8 12 Time (Days)

Figure 3.2. Catechin concentrations in model wine after treatment with 20 mg/L acetaldehyde at pH 2.0, 2.5, 3.0, and 3.5 as determined by HPLC-DAD.

In order to confirm that catechin consumption was due to its reaction with exogenous acetaldehyde and not by other mechanisms (e.g., oxidation), catechin reactions were monitored under aerobic and anaerobic (< 20 µg/L O2) conditions. As such, no significant difference between catechin consumption rates were observed under aerobic versus anaerobic conditions

(Figure 3.3), which confirms that the mechanism of reaction with acetaldehyde does not involve oxygen as a reactant. These results also confirm that the consumption of catechin observed is not due to the formation of endogenous acetaldehyde or direct oxidation of catechin, as these would require oxygen as a reactant.

1050

Aerobic 1000 Anaerobic

950 Catechin (mg/L) 900

0 3 6 9 12 Time (Days)

Figure 3.3. Catechin concentrations in model wine after treatment with 20 mg/L acetaldehyde with and without oxygen present at pH 2.5.

SO2 was then examined for its effect on the reaction of acetaldehyde with flavonoids due to its important role in the formation and fate of acetaldehyde. SO2 is known to prevent the metal- catalyzed oxidation of ethanol to form acetaldehyde.8,42 Bisulfite can react with ortho-quinones and hydrogen peroxide, reactive intermediates formed during the reduction of oxygen to acetaldehyde.46,64,65 Because bisulfite is so reactive with these species, acetaldehyde is not formed form oxygen when SO2 is present in wine. Bisulfite is also highly reactive with acetaldehyde, forming a sulfonate adduct (1-hydroxyethanesulfonate). This strongly bound adduct is typically assumed to be irreversible under wine conditions due to its low dissociation constant.61 Previous 55

studies have examined the effect of SO2 on acetaldehyde formation and products of endogenous acetaldehyde in wine.32,167,219,220 However, the reactivity of acetaldehyde when it is assumed to be bound as a sulfonate and inert, i.e. in the presence of SO2, has not been directly assessed.

Catechin consumption was measured under aerobic and anaerobic conditions under low

(40 mg/L) and high (80 mg/L) total SO22 (TSO2) concentrations. As was seen without any SO2 present, there was no significant difference between samples stored under aerobic or anaerobic conditions. However, there was a significant effect of SO2 concentration on catechin consumption. At the higher concentration of SO2, approximately double the concentration of bisulfite compared to acetaldehyde, the consumption of catechin was stopped. This is due to the fact that the acetaldehyde is preferentially bound to the excess of bisulfite and therefore unable to react with catechin (Figure 3.4).

A B 1050 1050

1000 1000

Catechin A+High SO 950 950 2 A+Low SO2 Catechin+A Catechin (mg/L) Catechin (mg/L) 900 900

0 3 6 9 12 0 3 6 9 12 Time (Days) Time (Days)

At the lower concentration of SO2 (i.e., a slight molar excess of bisulfite compared to acetaldehyde), the consumption of catechin was decreased but not completely prevented. The concentration of catechin under these conditions was significantly lower than the control and high

SO2 treatment samples; the concentration of catechin was also significantly higher after 14 days than in the sample containing acetaldehyde alone. To our knowledge, this is the first clear evidence for the reactivity of acetaldehyde when it assumed to be bound to bisulfite under wine conditions. These results suggest that the small amount of free acetaldehyde in equilibrium with its sulfonate is able to react with catechin.

3.4.2 Reaction of Acetaldehyde with M3G and Flavanols

After confirming these effects in model systems containing catechin only, malvidin-3- glucoside (M3G; 500 mg/L) was added to the system. Catechin concentrations were measured to determine if M3G had an effect on the consumption of catechin (Figure 3.5). There was a small but significant difference in the consumption of catechin between samples with and without

M3G. This difference could be due to competition between M3G and catechin for acetaldehyde.

This experiment was then repeated in the presence of SO2 (40 mg/L TSO2), under which conditions SO2 was observed to retard the consumption of catechin compared to SO2-free samples. However, there was significant consumption of catechin by acetaldehyde in the presence of SO2 compared to a control. There was no significant difference in catechin consumption between SO2 containing samples with or without M3G.

1025

1000

C 975 M3G+C+A+SO2 C+A+SO2 950 M3G+C+A

Catechin (mg/L) C+A

925

900 0 3 6 9 12 Time (Days)

The consumption of M3G was followed in order to determine the effect of SO2 on reactions of acetaldehyde with M3G as well as the effect of the type of flavanol present (i.e., monomeric catechin or tannin from grape seed extract) (Figure 3.6). No significant consumption of M3G with acetaldehyde present was observed compared to a control without acetaldehyde.

The reaction of M3G with acetaldehyde and itself is slower than with catechin, and may not be occurring at a rate that is sufficient for the effect to be seen in this time frame. The presence of

SO2 did not have a significant effect on M3G consumption in any samples. Since the effect of

SO2 was seen in the consumption of catechin, it is possible the consumption rate of M3G was sufficiently slow that no effect of SO2 could be observed.

550

500 M3G M3G+A 450 M3G+A+SO2 M3G+G+A 400 M3G+G+A+SO2 M3G+C+A M3G+C+A+SO2 350 Malvidin-3-Glucoside (mg/L)

300 0 3 6 9 12 Time (Days)

The presence of flavanols in the system did significantly affect M3G consumption.

Samples containing catechin had greater consumption of M3G than those containing GSE and those with M3G alone. These results indicate that the catechin monomer, in conjunction with acetaldehyde, is more reactive towards M3G than the oligomeric tannins found in the GSE. The

slow and non-significant reaction of M3G with acetaldehyde and itself illustrates that the rate of that reaction is significantly slower than that involving flavanols.

The reaction of M3G towards catechin compared to GSE illustrates the increased reactivity of monomers over oligomers. Monomeric catechin has two functional sites (C6 and C8) that are able to participate in ethylidene bridging, while oligomeric tannins should only have one functional site on each terminal subunit. Assuming most of the tannins in GSE are larger oligomers, the sites for ethylidene bridging will be lower in equivalent masses as compared to catechin.

Samples containing M3G were also characterized by a modified Somers assay in order to assess changes in color and color stability due to acetaldehyde treatment (Figure 3.7; Table 3.1).

Control samples that did not contain acetaldehyde were compared to samples after 12 days of acetaldehyde treatment or acetaldehyde and SO2 treatment, as described above. There were significant differences in several parameters based on experimental conditions and treatment time. In general, acetaldehyde treatment significantly improved color stability, and acetaldehyde treatment with SO2 present improved color stability, but to a lesser extent than acetaldehyde alone. The effects of both treatments (acetaldehyde or acetaldehyde and SO2) were greater when

M3G was treated with flavanols (catechin or GSE) than M3G treated alone.

Table 3.1. Color parameters from the modified Somers assay for control and treatment groups a (Acetaldehyde or Acetaldehyde and SO2) at 12 days.

Chemical Chemical Color-Density - SO SO -Resistant Total 2 Hue 2 Age 1 (au) Age 2 (au) Corrected (au) Pigments (au) Phenolics (au)

M3G Samples

Control 0.098 ± 0.004 a 0.0146 ± 0.0002 5.29 ± 0.03 a 0.3614 ± 0.0004 a 3.84 ± 0.03 a 11.81 ± 0.06 a

Acetaldehyde 0.113 ± 0.003 b 0.0164 ± 0.0002 4.63 ± 0.43 b 0.3638 ± 0.0006 b 3.61 ± 0.08 b 11.54 ± 0.18 ab (12 days) Acetaldehyde and SO 2 0.107 ± 0.021 ab 0.0170 ± 0.0035 5.19 ± 0.04 a 0.3689 ± 0.0009 c 3.73 ± 0.04 c 11.32 ± 0.04 b (12 days)

M3G + C Samples

Control 0.104 ± 0.001 a 0.0162 ± 0.0001 a 5.59 ± 0.08 a 0.3748 ± 0.0004 a 4.13 ± 0.03 a 20.89 ± 0.12 a

Acetaldehyde 0.124 ± 0.001 b 0.0264 ± 0.0002 b 6.32 ± 0.25 b 0.4075 ± 0.0002 b 4.69 ± 0.02 b 19.12 ± 0.10 b (12 days) Acetaldehyde and SO 2 0.127 ± 0.002 b 0.0258 ± 0.0003 b 6.16 ± 0.04 b 0.4144 ± 0.0010 c 4.38 ± 0.02 c 18.63 ± 0.22 c (12 days)

M3G + GSE Samples

Control 0.121 ± 0.001 a 0.0204 ± 0.0002 a 5.80 ± 0.06 a 0.4257 ± 0.0006 a 4.22 ± 0.02 a 23.47 ± 0.20 a

Acetaldehyde 0.178 ± 0.002 b 0.0350 ± 0.0003 b 5.91 ± 0.41 a 0.4413 ± 0.0007 b 4.46 ± 0.07 b 22.75 ± 0.31 b (12 days) Acetaldehyde and SO 2 0.165 ± 0.003 b 0.0328 ± 0.0003 b 6.46 ± 0.07 b 0.4489 ± 0.0006 c 4.58 ± 0.03 c 22.90 ± 0.13 b (12 days) a Values represent the average of three experimental replicates ± standard deviation. Values in the same experimental group column with different letters indicate significant differences (p < 0.05).

There was a significant decrease in the concentration of monomeric anthocyanins in all

treated samples. With respect to the degree of ionization of anthocyanins, there was a significant

increase in all acetaldehyde-treated samples. In samples containing M3G alone, the increase was

modest but significant in both the presence and absence of SO2, although there was no significant

difference in degree of anthocyanin ionization between the two acetaldehyde treatments. This is

likely due to the relatively small amount of M3G that participates in these reactions, which was

confirmed by measuring M3G concentrations. In samples containing M3G and catechin, there

was a significant and much larger increase in degree of ionization with acetaldehyde treatment.

Acetaldehyde treatment with SO2 resulted in a significant increase in anthocyanin ionization

compared to the control; however, this value was significantly lower than that for the

corresponding treatment with acetaldehyde alone. Samples containing M3G with GSE had a

similar trend with respect to change in degree of ionization of anthocyanins compared to M3G

with catechin samples. These trends were also observed with respect to increases in color density

associated with acetaldehyde treatment (Figure 3.7).

A B C 25 A 600 B b 8 C b b c a a a c b c b b c b b 20 b b 600 b 25 b 8 b b a c 6 a a c a a a b b c b c b b c 400 b b a a a b b 20 15 b a a c b b 6 a 400 a a a 4 15 10 4 200 10 anthocyanins (%) Color Density (au)

200 Degree of ionization of 2 5 Total Anthocyanins (mg/L) Anthocyanins Total 2

anthocyanins (%) 5 Color Density (au) Degree of ionization of

Total Anthocyanins (mg/L) Anthocyanins Total 0 0 0 0 0 0 M3G M3G+CM3G M3G+CM3G+GSEM3G+GSE M3GM3G M3G+CM3G+CM3G+GSEM3G+GSE M3GM3G M3G+CM3G+CM3G+GSEM3G+GSE

Control Acetaldehyde Acetaldehyde+SO2

Figure FigureX. Select 3.7 color. Select parameters color parameters from the from modified the modified Somers Somers assay assay at 0 forand control 12 day and time treatment points: A) Total anthocyanins,groups (acetaldehydeB) Degree of and ionization acetaldehyde+SO of anthocyanins,2) at 12 days: C) Color A) Total density. anthocyanins, Values represent B) Degree the of average of threeionization experimental of anthocyanins, replicates C)± standardColor density. deviation. Values Columnsrepresent the in theaverage same of threegroup experimental with different letters indicatereplicates significant ± differencesstandard deviation. (p<0.05). Columns in the same group with different letters indicate significant differences (p<0.05).

Taken together, these results suggest that acetaldehyde contributes to a decrease in

monomeric anthocyanins while increasing the degree of ionization of anthocyanins and color

density. This is consistent with previous results that were performed under similar

91,92,221 conditions. Furthermore, increases in SO2-resistant pigments were seen in acetaldehyde-

treated samples of M3G with catechin and GSE (Table 3.1). There was a slight but significant

decrease in total phenolics for most samples, which could indicate precipitation of higher molecular weight oxidation products formed by ethylidene bridging.

In the model studies, pH was shown to have a significant effect on the rate of reaction of catechin with acetaldehyde with higher rates of consumption at lower pH, which is consistent with previous reports.161 This is due to the important mechanistic step of protonation of the acetaldehyde. However, oxygen did not have a significant effect on the rate of reaction of acetaldehyde with catechin; which confirms that this mechanism does not involve the participation of oxygen and also that the loss of catechin observed is not due to other oxidative mechanisms.

Finally, acetaldehyde was found to react with catechin and M3G in the presence of an equimolar concentration of bisulfite. The rate of reaction was relatively slower compared to the

SO2-free system, as measured by following the loss of monomeric catechin; however, SO2 did not affect the concentration of free M3G. Most importantly, there were statistically significant improvements in several color stability parameters in the presence of SO2 for systems containing

M3G with catechin or GSE. While SO2 may slow the formation of ethylidene-bridged catechin oligomers, bisulfite had little or no effect on polymeric pigment formation when it was present as equimolar concentrations, as was evident by measuring M3G concentrations and color stability indices.

3.4.3 Characterization of Products with MALDI-TOF MS

Samples were characterized by MALDI-TOF MS in order to directly observe the high molecular weight phenolic compounds formed by the acetaldehyde bridging reactions described above. MALDI was selected here as it allows for a more gentle ionization of compounds, which

prevents fragmentation of larger molecules (e.g., tannins). Tannins and anthocyanin-derived pigments from grapes and wine have been previously observed by MALDI-TOF MS.222–224

Analysis of the samples described here required minimal sample preparation since MALDI-TOF can be used without separation of analytes. By using a completely volatile mixture (with the exception of the analytes), samples were dried and immediately mixed with the matrix components for analysis.

The experimental conditions were different from those used to characterize the consumption of catechin and M3G. Higher concentrations of acetaldehyde (200 mg/L) and longer times (1 month) were used to prepare these samples so that concentrations of the resulting products were sufficiently abundant to be seen by MALDI-TOF without any separation steps. All products discussed below were found in treated samples after 1 month and were not seen in control MALDI-TOF spectra of the mixtures at the beginning of the experiment.

In samples containing catechin alone with acetaldehyde, oligomers up to the nonamer were observed by MALDI-TOF MS (Figure 3.8; Table 3.2). Oligomers are formed by the addition of a catechin subunit along with an ethylidene bridge due to reaction with acetaldehyde

(i.e., the addition of 316 mass units). This sequence of increasing oligomer size is clearly seen from the dimer to the nonamer. These results are consistent with those previously seen using

MALDI-TOF MS in systems containing catechin and formaldehyde225, and by ESI and LSI-MS methods in systems contaning catechins and acetaldehyde.160,161 Another sequence of mass additions was also observed corresponding to additional acetaldehyde products. These are believed to represent vinyl additions to catechin subunits, corresponding to an increase in mass of

26 units. Up to four addional vinyl catechin moieties were observed in the larger oligomers.

Figure 3.8. MALDI-TOF mass spectrum recorded in positive reflectron mode of catechin treated with acetaldehyde.

Table 3.2. Predicted and observed m/z values as recorded in positive reflectron mode MALDI- TOF MS of catechin treated with acetaldehyde.

Vinyl DP Predicted Observed Additions 2 0 629 629.188 1 655 655.223 2 681 681.286

3 0 945 945.469 1 971 971.501 2 997 997.536

4 0 1261 1261.691 1 1287 1287.717 2 1313 1313.744 3 1339 1339.775

5 0 1577 1577.875 1 1603 1603.886 2 1629 1629.909 3 1655 1655.940

6 1 1919 1920.026 2 1945 1946.048 3 1971 1972.072 4 1997 1998.102

7 2 2261 2262.172 3 2287 2288.195 4 2313 2314.218

8 3 2603 2604.309 4 2629 2630.330

9 3 2919 2922.430 4 2945 2947.441

Oligomers of catechin can undergo depolymerization in wine and model wine systems.

Acid-catalyzed cleavage of ethylidene-bridged catechins have been shown to result in the formation of vinyl catechins.163 These vinyl moieties can be found on carbon-6 and carbon-8 of catechins, including those in oligomers (Figure 3.9). Vinyl-catechin, bivinyl-catechin, and vinyl oligomers have been observed by LSI-MS.160 These reactive vinyl species are likely to be involved in further modification reactions.74 Based on vinyl catechin products previously seen, we would only expect up to two vinyl moieties to be found in oligomers, those on terminal catechin subunits. The presence of sequential mass increases of 26 over the assumed two possible sites may indicate further functional sites for acetaldehyde reaction on catechin beyond the A ring or potentially another reaction product.

OH OH HO HO

O OH O OH HO OH HO OH HO HO OH OH OH OH O O O HO O HO OH HO OH H3C H n n-1 OH OH OH OH OH OH OH OH

HO O HO O

OH OH OH OH

Figure 3.9. Reaction of catechin with acetaldehyde to form ethylidene-bridged oligomers followed by cleavage to form vinyl catechin moieties.

The masses predicted based on ethyl-bridged catechin subunits and vinyl moieties corresponded closely with the observed spectrum (Table 3.2). The observation of multiple vinyl moieties is evidence that acetaldehyde may be responsible for greater changes in the flavonoid composition of wine than previously assumed. Based on the observed sequence of masses corresponding to vinyl catechin subunits, there may be another available functional site on

catechin under certain conditions. These results also illustrate that determining the fate of acetaldehyde under wine conditions is complicated based on the complexities of ethylidene- bridged oligomers and vinyl catechins. There are likely higher molecular weight oligomers formed that should be explored in future research.

Samples containing M3G alone with acetaldehyde showed evidence for the formation of

M3G dimers bridged by an ethylidene moiety. The dimer observed had m/z 1011.595 and corresponds to a previously described ethyl-linked M3G dimer with one neutral quinoidal base and one flavylium cation.95 Previous studies have argued for the existance of anthocyanin trimers in grape pomace,224 however, these compounds were not observed in the present study. Unlike catechin, which forms long oligomeric chains, M3G was not observed to polymerize beyond its dimer. This corroborates previous work demonstrating that anthocyanins may act as terminal subunits, thus preventing further increases in tannin size.

The combination of catechin and M3G treated with acetaldehyde yielded a relatively more complicated mixture of products than either flavonoid alone. In this mixture, catechin oligomers up to the tetramer were observed as well as catechin oligomers up to the trimer that also contained a M3G subunit. Products containing M3G also have the potential to rearrange to pyranoanthocyanins, resulting in a net decrease of 4 mass units (Table 3.3). Pyranoanthocyanins are cyclized forms of anthocyanin adducts formed by the reaction of vinyl catechins with an anthocyanin. The pyranoanthocyanin formed from an ethylidene bridged adduct can be considered a rearrangement of the adduct; acid-catalyzed cleavage of the ethylidene bridge produces a vinyl catechin and releases the anthocyanin, M3G in this case. Reaction of the vinyl catechin with M3G forms an intermediate that recyclizes into a pyranoanthocyanin-catechin adduct (Figure 3.10).74,226,227 Pyranoanthocyanin adducts have been observed in model wine and in red wine and are known to be more stable than their anthocyanin counterparts.73,96,228–230

Table 3.3. Predicted and observed m/z values recorded in positive reflectron mode MALDI-TOF MS of catechin and M3G treated with acetaldehyde.

M3G Vinyl Predicted DP of Catechin Observed Additions Additions Adduct/Rearranged 1 1 0 809 809.409 1 835/831 835.446 831.407

2 1 0 1125 1125.657 1 1151/1147 1151.691 2 1177/1173 1173.694

3 0 0 945 945.519 1 971/967 971.556

1 0 1441 1441.847 1 1467/1463 1467.881 1463.857 2 1493/1489 1489.889

4 0 1 1287 1287.774

OCH3

OH OH OCH3 OGl OH HO O OCH3 OCH HO O OH HO O 3 OCH OGl 3 OH OH OH OGl HO O OCH3 O O OH OH OH H HO O HO O OH H3C OH OH

OH OH OH OH

Figure 3.10. Reaction of catechin with M3G and acetaldehyde to form an ethylidene-bridged polymeric pigment and subsequent rearrangement to a pyranoanthocyanins.

It should be noted that the peak at m/z 809.409, corresponding to a catechin-ethyl-M3G adduct, was by far the most abundant in the spectrum for this mixture. Again, vinyl catechin moieties were observed in this reaction mixture of M3G and catechin. While we did not observe adducts larger than the tetramer, they are likely present in the reaction mixture. The detector may 69

have been saturated due to the presence of smaller m/z compounds, which would include M3G monomers as well as any doubly charged adducts formed that would contain multiple M3G subunits. It is conceivable that prior chromatographic separation of the components in advance of

MALDI-TOF MS analysis would allow for improved characterization of the mixture.

Samples containing tannins from GSE were treated with acetaldehyde to observe the effect on native tannins. In this mixture of products, tannins up to the pentamer were observed in their native and modified states. These predicted masses and the corresponding observed m/z values are summarized (Table 3.4). Procyanidins and their gallate equivalents were all found to be modified by vinyl and/or catechin additions. It is likely that larger modified tannins would form given longer treatment times, resulting in concentrations that could be more easily detected.

Table 3.4. Predicted and observed m/z values recorded in positive reflectron mode MALDI-TOF MS of GSE treated with acetaldehyde.

Gallate Vinyl Catechin DP Predicted Observed Groups Additions Additions 2 0 0 0 601 601.145 1 1 943 943.485

3 0 0 0 889 889.39 1 1 1231 1231.668 1 0 0 1041 1041.481 1 1 1383 1383.731

4 0 0 0 1177 1177.587 1 1 1519 1519.801 1 0 0 1329 1329.651 1 1 1671 1671.863

5 0 0 0 1465 1465.74 1 0 1491 1491.783 1 0 0 1617 1617.785 1 0 1643 1643.82

The combination of GSE and M3G treated with acetaldehyde resulted in complicated product mixtures, as was seen in catechin and M3G samples. These samples contained native tannins as well as those that have been modified to include M3G subunits. The MALDI-TOF spectrum showed evidence for native tannins up to the tetramer modified by an additional M3G subunit. These predicted and observed mass values are summarized (Table 3.5). Additionally, the rearranged pyranoanthocyanin version of adducts containing M3G were observed in these oligomers. As was seen in samples of GSE alone with acetaldehyde, there would likely be larger, modified tannins formed that could be observed by MALDI-TOF with longer reaction time.

Table 3.5. Predicted and observed m/z values recorded in positive reflectron mode MALDI-TOF MS of GSE and M3G treated with acetaldehyde.

Gallate M3G Vinyl Predicted DP of Tannin Observed Groups Additions Additions Adduct/Rearranged 1 0 0 0 313 1 0 809/805 809.392 805.356 1 1 835/831 835.438 831.393 1 0 0 465 1 0 961/957 961.491 957.463

2 0 0 0 601 601.145 1 0 1097/1093 1097.613 1093.564 1 1 1123/1119 1119.597 1 0 0 753 753.259 1 0 1249/1245 1249.675 1245.64

3 0 0 0 889 889.39 1 0 1385/1381 1381.731 1 1 1411/1407 1407.77 1 0 0 1041 1041.481 1 0 1537/1533 1533.797

4 0 0 0 1177 1177.587 1 0 1673/1669 1669.83

MALDI-TOF characterization confirmed the formation of many products resulting from the reaction of grape flavonoids with acetaldehyde. Acetaldehyde was shown to form polymeric pigments by adding M3G subunits to native tannins from GSE and to catechin oligomers formed by ethylidene bridges. These polymeric pigments existed in their ethyl-linked forms and as rearranged pyranoanthocyanins. M3G was also observed to form dimers bridged by ethylidene moieties in the presence of acetaldehyde. The ability of acetaldehyde to formed bridged catechin oligomers was confirmed by the observation of nonamers in samples containing only catechin and acetaldehyde. These results further demonstrate the advantages that MALDI confer for the analysis of complex wine flavonoids, especially given the minimal sample preparation required compared to other MS techniques.

In summary, these results illustrate the important role that acetaldehyde plays, even in the presence of equimolar bisulfite. Though past research has focused on free acetaldehyde, the bound form should not be considered a true end point. Instead, future work should examine the best ways to take advantage of the activity of acetaldehyde in wine that is protected by SO2.

Additionally, MALDI-TOF MS characterization of several reaction products confirms the complexity of adducts formed by reaction with acetaldehyde – including the polymeric pigments formed by the addition of anthocyanins to monomeric and oligomeric condensed tannins.

3.5 Acknowledgements

The authors thank the Proteomics and Mass Spectrometry Core Facility at Penn State

University for access to the MALDI-TOF instrument and Dr. Tatiana N. Laremore for her help with MALDI-TOF analysis.

Chapter 4

Reactions of Free and SO2-Bound Aldehydes with (+)-Catechin

4.1 Abstract

Aldehydes are formed during fermentation and chemical oxidation of wine and wine spirits. Their reaction with flavonoids leads to the formation of bridged oligomers and stable pigments. In this study, the reactivity of several aldehydes – formaldehyde, acetaldehyde, propionaldehyde, isobutyraldehyde, and benzaldehyde – with catechin was determined. Bisulfite adducts of these aldehydes, α-hydroxyalkylsulfonates, were then synthesized and studied in a model system with catechin. The reactivity of each aldehyde from its sulfonate was examined and found to correlate with the dissociation constant of the sulfonate. The bridged catechin oligomers formed by each aldehyde were also characterized by MALDI-TOF MS with oligomers of DP 2 to

9 observed. The results of this work demonstrate the significant reactivity of free and SO2-bound aldehydes with catechin and their potential impact in wine.

4.2 Introduction

Aldehydes in wines, which can be generated during fermentation, processing, and aging, are important with respect to wine quality (e.g., aroma, color, mouthfeel). Most short chain aliphatic aldehydes are formed by yeast during fermentation;231 however, aldehydes can also result in wine due to Strecker degradation of amino acids, and the oxidation of ethanol and certain fatty acids.232 Acetaldehyde is the most abundant aldehydic compound in wine, comprising about

90% of aldehydes in wine.231 However, there are many other aldehydes present in wine and wine spirits, such as formaldehyde, propionaldehyde, isobutyraldehyde, and benzaldehyde.44,233–235 The odor thresholds and descriptors of these species are included in Table 4.1.

Table 4.1. Odor thresholds and characteristics of aldehydes

Detection Threshold Aldehyde Descriptors in Water (µg/L) Formaldehyde 500 236 Hay, straw-like, pungent 237,238 Acetaldehyde 1.5 – 34 5,236,239 Green, sweet, pungent, ethereal 238,240 Propionaldehyde 1 – 10 236,239 Roasted coffee, pungent 240 Isobutyraldehyde 0.35 – 2.3 5,236,240 Malty 5,240 Benzaldehyde 200 – 350 234,239 Bitter almond 240

Aldehydes are known to react with flavonoids in red wine, which can lead to quality improvements in terms of color stability and mouthfeel. As it is the most abundant, acetaldehyde and its role in these reactions are well characterized. Acetaldehyde has been shown to form ethyl- bridged flavonoids and pyranoanthocyanins.66,93,95,96 Glyceraldehyde53,241, glyoxylic acid102,215,216,242, pyruvic acid243, furfural102,244, and 5-(hydroxymethyl)furfural102,244 have been

shown to react by similar mechanisms to react with flavonoids and improve wine characteristics.

Pissara et al. have looked at reactions of acetaldehyde, formaldehyde, propionaldehyde, 2- methylbutyraldehyde, isobutyraldehyde, isovaleraldehyde, and benzaldehyde in model solutions containing malvidin-3-glucoside and (+)-catechin.230,245–247 The mechanism of reaction of the aldehydes studied in the present work is shown in Figure 4.1.

OH OH HO HO O OH O OH HO OH OH HO OH OH HO R OH R OH O H H+ O -H+ HO O H+ HO O R H R H OH OH OH OH OH Formaldehyde: R = H HO

Acetaldehyde: R = CH3 O OH OH Propionaldehyde: R = CH CH HO 2 3 R OH -H+ OH Isobutyraldehyde: R = CH2(CH3)2 HO O

Benzaldehyde: R = C6H5 OH OH

Figure 4.1. Mechanism of reaction of aldehydes with catechin.

Bisulfite, the major form of sulfur dioxide (SO2) in red wine, reacts with aldehydes to form an adduct and therefore diminishes their reactivity. These aldehyde-bisulfite adducts, α- hydroxyalkylsulfonates, may be very strongly bound or more likely to dissociate.61 The binding of an aldehyde to bisulfite affects its reactivity and aroma activity, as only the free form can contribute (Figure 4.2). The dissociation constant (Kd) of the α-hydroxyalkylsulfonate, therefore, is important for predicting the reactivity of an aldehyde in the presence of bisulfite. The Kd values

for the aldehydes studied here are reported in Table 4.2, with the exception of isobutyraldehyde, whose Kd has not been found using similar methods to the other aldehydes in this table.

OH HO O S O OH H O HO O O Catechin OH HO R R S O OH R H OH OH K O d HO O

OH OH

Figure 4.2. Reactions of aldehydes with bisulfite and catechin.

61 Table 4.2. Dissociation constants (Kd) of aldehyde-bisulfite adducts.

K Aldehyde d (pH 3.5) Formaldehyde 1.09x10-7 Acetaldehyde 2.06x10-6 Propionaldehyde 7.70x10-6 Isobutyraldehyde - Benzaldehyde 2.83x10-3

Some aldehyde-bisulfite adducts can be recovered as solid sodium salts. These isolated salts allow the analysis of aldehydes added in their 100% bound form. Based on the low amount of these adducts expected to dissociate under wine conditions, it is expected that there would be very little activity of the free aldehyde. However, Andersen et al. showed that the acetaldehyde- bisulfite adduct (sodium 1-hydroxyethanesulfonate) had similar antioxidant activity to free bisulfite.62 The activity of acetaldehyde in the presence of equimolar bisulfite has been studied in our lab where we demonstrated that the reaction of acetaldehyde was slowed, but not prevented 76

(Chapter 3). The reactivity of other aldehydes from their bisulfite adducts or its relationship with their Kd has not been explored.

The present work examines the reactivity of five aldehydes in a model wine system: formaldehyde, acetaldehyde, propionaldehyde, isobutyraldehyde, and benzaldehyde. In this study, the reaction of these aldehydes from their free form and from their respective synthesized bisulfite adduct salts was assessed and products of these aldehydes with catechin were characterized.

4.3 Materials and Methods

4.3.1 Materials

(+)-Catechin hydrate and sodium metabisulfite were purchased from Sigma Aldrich (St.

Louis, MO). Acetaldehyde was purchased from Alfa Aesar (Ward Hill, MA). Propionaldehyde, sodium iodide, and LC/MS grade formic acid were purchased from Fisher Scientific (Pittsburgh,

PA). Isobutyraldehyde, benzaldehyde, formaldehyde (36.5% solution), acetone, sodium hydroxide (10.00 N), 200 proof ethanol, and 2,5-dihydroxybenzoic acid (DHB) were purchased from VWR International (Radnor, PA). Glacial acetic acid was purchased from J.T. Baker

(Phillipsburg, NJ). Hydrochloric acid was purchased from EMD Chemicals (Gibbstown, NJ).

Water was purified through a Millipore Q-Plus (Millipore Corp., Bedford, MA) purification train.

4.3.2 Bisulfite Adduct Synthesis

Aldehyde-bisulfite adducts were synthesized according to the protocol described by

Andersen et al.62 Each aldehyde (24 mmol) was dissolved in cold ethanol (25 mL) and stirred 77

over ice. A solution of 40% aqueous sodium metabisulfite (3.75 mL, 20 mmol bisulfite) was added dropwise to the stirred aldehyde solution. Ethanol (25 mL) was then added to the solution.

The reaction mixture was kept at 4 °C overnight. The precipitate was then collected by cold, vacuum filtration, washed several times with cold ethanol, and then dried under a gentle stream of nitrogen. All aldehyde-bisulfite adducts were isolated as white powders. The synthetic yields for each product are summarized in Table 4.3.

Table 4.3. Aldehyde-bisulfite adducts synthesized and their synthetic yields.

Synthetic Aldehyde Bisulfite Adduct Yield Formaldehyde Sodium 1-hydroxymethanesulfonate 47% Acetaldehyde Sodium 1-hydroxyethanesulfonate 67% Propionaldehyde Sodium 1-hydroxypropanesulfonate 54% Isobutyraldehyde Sodium 1-hydroxy-2-methylpropanesulfonate 93% Benzaldehyde Sodium 1-hydroxy(phenyl)methanesulfonate 70%

4.3.3 Reaction Mixture Preparation

Model wine was composed of 5% w/v acetic acid, 12% ethanol v/v, at pH 2. Acetic acid was used as the organic acid in this system due to its volatility and suitability for MALDI-TOF

MS analysis. The low pH was chosen in order to achieve higher rates of reaction so that differences between treatments could be observed. Model experiments were prepared with 500

µM treatment (aldehyde or sulfonate) and 1000 mg/L (+)-catechin. All experiments were completed in triplicate. Glassware was soaked in 5% nitric acid overnight and rinsed to remove trace metals before use. Each reaction mixture was separated into 1 mL aliquots and stored at room temperature in an anaerobic chamber (Anaerobe Systems, Morgan Hill, CA). The anaerobic 78

chamber was filled with 95% argon/ 5% hydrogen gas and had a palladium catalyst to remove residual oxygen. Anaerobic status of the chamber was confirmed using a PreSens oxygen meter

(PreSens, Regensburg, Germany). All solutions were deoxygenated using argon gas before being introduced into the chamber, and all materials were left in the anaerobic chamber overnight prior to starting experiments. Samples collected at each experimental time point were stored at -80 °C and thawed immediately prior to analysis.

4.3.4 Catechin Analysis

Catechin concentrations were determined by HPLC using a Shimadzu system with

10ADvp pumps and a SIL-20AC HT temperature-controlled autosampler set to 4 °C. Separation was achieved on a reverse phase Agilent Zorbax SB-Aq column (2.1 mm x 150 mm, 3.5 µm;

Agilent Tech., Santa Clara, CA). Samples were filtered through 0.45 µm PTFE syringe filters prior to analysis. The injection volume was 10 µL, column temperature was 30 °C, and flow rate was held at 0.2 mL/min for all samples. The mobile phase consisted of 0.1% formic acid in water

(A) and methanol (B). Catechin was eluted using an isocratic method of 25% B and detected at

280 nm. Concentrations were calculated based on external calibration curve of catechin.

4.3.5 Reaction Product Characterization

Samples for MALDI-TOF were prepared in acetic acid model wine as described above with 5 mM of each aldehyde. After 4 weeks, aliquots of samples were dried under nitrogen and re-dissolved in methanol to achieve a 10 mg/mL catechin concentration. Samples were spotted from a 2,5-dihydroxybenzoic acid (DHB) matrix solution. DHB was recrystallized prior to sample preparation. Sample (1 µL, 10 mg/mL in methanol), sodium iodide (1 µL, 10 mg/mL in 79

water), and DHB (10 µL, 200 mg/mL in acetone) were mixed and 0.5 µL was spotted on a polished stainless steel plate for analysis (Bruker Corp., Santa Barbara, CA). MALDI-TOF mass spectra were collected in positive reflectron mode using a Bruker ultrafleXtreme instrument

(Bruker Corp., Santa Barbara, CA). Laser intensity was adjusted to optimize the resolution and signal for each spectrum. MALDI spectra were externally calibrated and acquired as an average of 1000 shots collected in random walk mode from m/z 500 to 4000.

4.3.6 Statistical Analysis

4.4 Results and Discussion

4.4.1 Consumption of Catechin by Aldehydes

The results of the analysis of the reaction of the aldehydes studied with catechin are shown in Figure 4.3. These results clearly show significant differences with respect to aldehyde reactivity. There was no significant difference in the amount of catechin consumed after 28 days between acetaldehyde, propionaldehyde, and benzaldehyde. There was also no significant difference in the consumption by isobutyraldehyde and formaldehyde; however, there was a clear and significant difference between these groups of aldehydes.

1050

Formaldehyde 950 Acetaldehyde Propionaldehyde Isobutyraldehyde 850 Benzaldehyde Catechin (mg/L)

750 0 7 14 21 28 Time (Days)

Figure 4.3. Consumption of catechin by aldehydes. Values represent the average of three experimental replicates ± standard deviation.

These differences in reactivity of the aldehydes can be explained by the mechanism of their reaction with catechin (Figure 4.1). The first step of this reaction is the protonation of the aldehyde to form a carbocation, the stability of which varies based on the substituent, with alkyl groups stabilizing the carbocation by inductive effects.248 Acetaldehyde, propionaldehyde, and isobutyraldehyde have similar abilities to stabilize their corresponding carbocation. The phenyl group of benzaldehyde stabilizes the carbocation by mesomeric effects wherein the charge is shared among resonance structures.248 In the next step of the mechanism, catechin acts as a nucleophile from C6 or C8 to add to the electrophilic carbocation. This protonated adduct is dehydrated to form another carbocation which then reacts with another available catechin. These addition steps are influenced by other characteristics of the aldehydes, namely the size of the substituent. Steric hindrance from the larger groups (e.g., isobutyraldehyde, benzaldehyde) could slow the rate of their reaction with catechin.

The results shown here demonstrate the influence of these inductive, mesomeric, and steric effects on the reaction rate of aldehydes with catechin. Formaldehyde’s reactivity is slowed by the instability of its carbocation, whereas the steric hindrance of isobutyraldehyde slows its reaction compared to other alkyl groups, and benzaldehyde’s relative steric hindrance seems to be overcome by the role of its carbocation stability. There may be further effects of hyperconjugation and equilibrium states of the aldehydes, which could also affect these reactions.

Additionally, differences in the rate of acid-catalyzed cleavage of the bridged oligomers may influence the observed monomeric catechin concentration.

While the reaction rates of these aldehydes to form bridged catechin oligomers has not been previously reported to our knowledge, there has been research done on their reaction with malvidin-3-glucoside and catechin.230,245–247 Pissara et al. found that the rate of reaction differed where acetaldehyde was the fastest followed by propionaldehyde, formaldehyde, benzaldehyde, and isobutyraldehyde.245 The notable difference from our observations is the rate of the benzaldehyde reaction; we found the fast reaction of benzaldehyde was comparable to acetaldehyde and propionaldehyde while Pissara et al. found significantly slower reaction rates.

This may indicate that steric hindrance is a greater obstacle for the reaction of catechin with malvidin-3-glucoside than with itself. However, this work did not take into account all of the complexities of the reactions of malvidin-3-glucoside and catechin involving ethylidene bridge formation and cleavage, so there could be an additional influence of the rate of pyranoanthocyanin formation. There were also differences in the experimental conditions as they used higher temperatures (35 °C), an unbuffered solution, a large excess of aldehyde, and performed their experiments in the presence of oxygen. These differences could affect the reaction mechanism and the rates observed for each aldehyde. However, their results with ours

indicate there is more research needed to fully understand the reactions of these aldehydes with catechin, anthocyanins, and other compounds in wine.

4.4.2 Consumption of Catechin by Aldehyde-Bisulfite Adducts

As aldehydes react readily with bisulfite, it is important to understand the reactivity of aldehydes from their bound form. This effect was examined by using aldehyde-bisulfite adduct salts in reaction mixtures with catechin. The consumption of catechin by aldehyde-bisulfite adducts illustrates the importance of the reactivity of the aldehyde as well as its dissociation from bisulfite. The aldehyde-bisulfite adducts used in these experiments were chosen because they could be synthesized as solid sodium salts for comparison in these experiments.249–252 By adding solid aldehyde-bisulfite adducts, each aldehyde was added only in its bound form, which means that any reactivity with catechin observed was due to release of the free aldehyde from the adduct. These treatments were added to achieve equimolar concentrations (500 µM) to the aldehyde treatments so that they could be accurately compared.

The results for the analysis of catechin concentration in samples treated with aldehyde- bisulfite adducts are shown in Figure 4.4. There is a clear, significant difference in consumption by the benzaldehyde-bisulfite adduct compared to all others. As expected, the adduct with the largest dissociation constant resulted in the greatest magnitude of catechin consumption. The next highest degree of consumption was seen with the propionaldehyde-bisulfite adduct, followed by the acetaldehyde-bisulfite adduct. These adducts have lower dissociation constants than benzaldehyde-bisulfite so we expected this significant effect of the adduct compared to the free aldehyde. Isobutyraldehyde-bisulfite also had significantly lower catechin consumption compared to free isobutyraldehyde. Finally, formaldehyde-bisulfite resulted in no significant consumption of catechin after 28 days compared to the starting concentration. This suggests that formaldehyde- 83

bisulfite, with the lowest dissociation constant of the adducts studied, did not free enough formaldehyde to react with catechin.

1050

Form-Bisulfite 950 Acet-Bisulfite Prop-Bisulfite Ibut-Bisulfite 850 Benz-Bisulfite Catechin (mg/L)

750 0 7 14 21 28 Time (Days)

Figure 4.4. Consumption of catechin by aldehyde-bisulfite adducts. Values represent the average of three experimental replicates ± standard deviation.

In order to compare all aldehydes and their bisulfite adducts, the total amount of catechin consumed after 28 days with each treatment was calculated (Figure 4.5), which allows one to clearly see the effect of dissociation constant on the reactivity of the aldehyde. There were no significant differences in consumption between acetaldehyde, propionaldehyde, and benzaldehyde; however, their bisuflite adducts behaved differently. The benzaldehyde-bisulfite adduct resulted in the same degree of catechin consumption as free benzaldehyde, demonstrating that adding the bound form had no significant effect on the aldehyde’s activity. Acetaldehyde- bisulfite and propionaldehyde-bisulfite, on the other hand, resulted in significantly lower catechin consumption compared to their free aldehydes, with acetaldehyde-bisulfite having significantly lower consumption than propionaldehyde-bisulfite.

Isobutyraldehyde and formaldehyde had similar amounts of catechin consumed, although their bisulfite adducts differed. There was significantly less catechin lost with bound isobutyraldehyde. As discussed previously, formaldehyde-bisulfite had no significant consumption of catechin in 28 days, confirming the strong binding of formaldehyde by bisulfite due to its low dissociation constant.

250

e e 200 e e

150 d d 100 c b 50 ab a

Catechin Consumed (mg/L) 0 Form Acet Prop Ibut Benz

Aldehyde Bisulfite Adduct

The relationship between the catechin consumption seen in these experiments and the dissociation constants of the bisulfite adducts is shown in Figure 4.6, which compares the degree of catechin consumption (%) by the bisulfite adduct (vs. aldehyde) to the –logKd for the known

Kd values. Despite the fact that only four aldehydes can be compared here, there is a clear

2 correlation (r = 0.9791) between percent catechin consumed by the bisulfite adduct and –logKd.

We would expect other aldehydes and their adducts to follow this relationship, demonstrating the role of dissociation of the bisulfite adduct in the activity of the free aldehyde.

100

80 r2=0.9791 60

%Consumption 20 by Bisulfite Adduct Bisulfite by

0 2 4 6 8

- log Kd

Figure 4.6. Correlation between –log Kd of aldehydes-bisulfite adducts and the percent of catechin consumed by the bisulfite adduct compared to the free aldehyde.

4.4.3 Characterization of Catechin-Aldehyde Products

Products of the reaction of each aldehyde with catechin were examined using MALDI-

TOF MS. This technique was used as it allows a gentle ionization of the larger oligomers formed than is seen with other MS methods. As will be discussed here, oligomers up to the nonamer (m/z

3000) were observed from the reaction of aldehydes with monomeric catechin. To our knowledge, this is the first evidence for catechin oligomer formation in model wine conditions beyond the dimer for most of these aldehydes.

Formaldehyde has been previously shown to form catechin oligomers.225 The oligomers formed show sequential mass increases of 302, representing the addition of a methyl-bridged catechin (Figure 4.7, Table 4.4). While some aldehydes allow the formation of vinyl moieties on

catechin, the simple formaldehyde structure cannot. Formaldehyde can, however, form large oligomers with up to nine methyl-bridged catechin subunits. These oligomers were among the largest seen from any of the aldehydes in this study. This is especially interesting, as formaldehyde had among the lowest amount of catechin consumed indicating that the catechin that is consumed is going into larger oligomers. This could indicate that the methylidene linkage is less susceptible to acid-catalyzed cleavage than others, but this should be investigated further.

Figure 4.7. MALDI-TOF mass spectrum recorded in positive reflectron mode of catechin treated with formaldehyde.

Table 4.4. Predicted and observed m/z values as recorded in positive reflectron mode MALDI- TOF MS of catechin treated with formaldehyde.

DP Predicted Observed 2 615 615.163 3 917 917.438 4 1219 1219.660 5 1521 1521.835 6 1823 1823.980 7 2125 2126.112 8 2427 2428.243 9 2729 2730.365

The catechin oligomers formed by acetaldehyde have been observed previously by our lab and others (Chapter 3).93,159,160 Our results show the formation of oligomers up to the nonamer with increases of 316 upon the addition of an ethyl-bridged catechin (Figure 4.8, Table 4.5).

However, there is a second, significant sequence of products with mass increases of 26 representing a vinyl moiety on catechin resulting from acid-catalyzed cleavage of an ethylidene linkage to another subunit.86 Products were observed with up to 6 additional vinyl catechin moieties on the nonamer product. The formation of ethyl-bridged oligomers and their cleavage to vinyl catechins is important for many reactions in wine including the formation of pyranoanthocyanins.

Figure 4.8. MALDI-TOF mass spectrum recorded in positive reflectron mode of catechin treated with acetaldehyde.

Table 4.5. Predicted and observed m/z values as recorded in positive reflectron mode MALDI- TOF MS of catechin treated with acetaldehyde.

Vinyl DP Predicted Observed Additions 2 0 629 629.197 1 655 655.236 3 1 971 971.536 2 997 997.563 3 1023 1023.595 4 1 1287 1287.775 2 1313 1313.800 3 1339 1339.823 5 1 1603 1603.972 2 1629 1629.992 3 1655 1656.013 4 1681 1682.041 6 2 1945 1946.159 3 1971 1972.179 4 1997 1998.206 7 3 2287 2288.339 4 2313 2314.355 5 2339 2340.377 8 4 2629 2630.496 5 2655 2656.525 9 4 2945 2947.622 5 2971 2973.690 6 2997 3000.686

To our knowledge, the oligomers of catechin from propionaldehyde observed here are the first to be characterized (Figure 4.9, Table 4.6). Dimeric products have been seen previously.245

Propyl-linked oligomers were observed up to the pentamer with sequential increases of 330 with each subunit. As was seen with acetaldehyde, there were additional vinyl products (mass of 40)

seen due to cleavage reactions. Based on the spectrum acquired, there may be other oligomeric products in this mixture that differ from our predicted structures and mass values.

Figure 4.9. MALDI-TOF mass spectrum recorded in positive reflectron mode of catechin treated with propionaldehyde.

Table 4.6. Predicted and observed m/z values as recorded in positive reflectron mode MALDI- TOF MS of catechin treated with propionaldehyde.

Vinyl DP Predicted Observed Additions 2 0 643 643.271 1 683 683.293 2 723 723.341 3 0 973 973.563 1 1013 1013.607 2 1053 1053.650 4 1 1343 1343.858 2 1383 1383.897 3 1423 1423.943 5 2 1713 1714.104 3 1753 1754.139 4 1793 1794.176

The reaction mixture of isobutyraldehyde with catechin contained more varied products than with other aldehydes (Figure 4.10, Table 4.7). We were able to identify an isobutyl-linked dimer and a dimer with an additional vinyl moiety. The dimer has been observed by others.245 As can be seen in the mass spectrum, there appears to be a sequence of increasing oligomers, though the masses do not correspond with any predicted values. The mechanism of reaction with isobutyraldehyde could be more complicated due to carbocation rearrangement or another bridging configuration not seen with other aldehydes and their vinyl catechin products. Based on the variety of products observed, further work should be done to understand the mechanism of the reaction of isobutyraldehyde and catechin.

Figure 4.10. MALDI-TOF mass spectrum recorded in positive reflectron mode of catechin treated with isobutyraldehyde.

Table 4.7. Predicted and observed m/z values as recorded in positive reflectron mode MALDI- TOF MS of catechin treated with isobutyraldehyde.

Vinyl DP Predicted Observed Additions 2 0 658 657.283 1 712 711.339

As was seen with our results for the consumption of catechin, benzaldehyde proved to be more reactive than we expected. We observed bridged octamers with sequential mass increases of

378 with each additional catechin subunit (Figure 4.11, Table 4.8). This is the first example of any benzaldehyde-catechin oligomers beyond the dimer to our knowledge.245 Benzaldehyde did not produce any vinyl-linked compounds, as their formation would require the loss of aromaticity of the phenyl ring. The large benzaldehyde products may be observed because they are less susceptible to acid-catalyzed cleavage and should be investigated further. 93

Figure 4.11. MALDI-TOF mass spectrum recorded in positive reflectron mode of catechin treated with benzaldehyde.

Table 4.8. Predicted and observed m/z values as recorded in positive reflectron mode MALDI- TOF MS of catechin treated with benzaldehyde.

DP Predicted Observed 2 691 691.275 3 1069 1069.608 4 1447 1447.863 5 1825 1826.067 6 2203 2204.242 7 2581 2582.410 8 2959 2961.595

The MALDI-TOF MS characterization of the products of catechin reaction with various aldehydes confirmed their ability to form bridged catechin oligomers. Isobutyraldehyde and propionaldehyde appear to be able to form vinyl moieties as has been described for acetaldehyde.

Formaldehyde and acetaldehyde formed large oligomers (nonamers) as has been seen previously.

Benzaldehyde also formed large oligomers, up to the octamer, which has not been observed prior to this work.

The results of this work show that several aldehydes – formaldehyde, acetaldehyde, propionaldehyde, isobutyraldehyde, and benzaldehyde – react with catechin to form bridged oligomers. These aldehydes differ in their reactivity based on structural differences. The aldehydes, except for formaldehyde, are also reactive when added as aldehyde-bisulfite adduct salts. The amount of reactivity of the free aldehyde is based on the dissociation constant of its bisulfite adduct.

The activity of these aldehydes from their bisulfite adducts has significant importance in wine and wine spirits. Aldehydes play many roles including their aroma and chemical activity. In wine, bisulfite is always present as it is formed endogenously during fermentation and is added for protection from spoilage. This work demonstrates that bisulfite will impact aldehydes differently depending on the stability of the complex formed (Kd). Further research is needed to examine this impact in wine and to explore other aldehydes relevant to wine aroma and chemistry.

4.5 Acknowledgements

The authors thank the Proteomics and Mass Spectrometry Core Facility at Penn State

University for access to the MALDI-TOF instrument and Dr. Tatiana N. Laremore for her help with MALDI-TOF analysis. 95

Chapter 5

Improving Red Wine Color Stability with Exogenous Acetaldehyde

5.1 Abtract

Oxidation reactions involving acetaldehyde lead to the formation of stable pigments in red wine. Oxygen exposure is typically used as a method of forming endogenous acetaldehyde.

The reduction of oxygen to acetaldehyde, however, also forms reactive intermediates that can lead to deleterious effects on wine. Acetaldehyde is also assumed to be inert when bound to sulfur dioxide due to the low dissociation constant of this adduct. The present study evaluates the use of exogenous acetaldehyde in place of oxygenation for red wine post-fermentation and the activity of acetaldehyde from its bisuflite adduct, 1-hydroxyethanesulfonate. Exogenous acetaldehyde

(500 µM) significantly increased color density and the concentration of polymeric pigments in red wine. The sulfonate also showed significant activity in model and wine systems. Compared to oxygenation, exogenous acetaldehyde increased the formation of SO2-resistant pigments from monomeric anthocyanins. These results demonstrate the possible use of exogenous acetaldehyde treatment as a method of improving color stability of red wine.

5.2 Introduction

Oxidation improves the color of red wine by several mechanisms with the most significant contributions from the reactions of acetaldehyde.119 Acetaldehyde is known to improve the color stability of red wine by converting monomeric anthocyanins to stable, polymeric pigments. Timberlake and Bridle first showed the effect of acetaldehyde on color and its mechanism has since been elucidated.91,92 Reactions of acetaldehyde are mediated by the formation of ethylidene bridges between flavonoids (Figure 5.1).93 Ethyl-bridged polymeric pigments have been observed in model systems and in red wine.94–97,169 Ethyidene bridges are susceptible to acid-catalyzed cleavage, which contributes to the formation of

86,99–101 pyranoanthocyanins. Pyranoanthocyanin color is stable to both bleaching by SO2 and pH,

84,95,101,112,118 and ethyl-bridged pigments are stable to SO2 bleaching. Formation of these pigments from anthocyanins also contributes to shifts in color, as pyranoanthocyanins are typically orange

(480 to 510 nm).91,95,111,112

OH OCH3 OGl OH O OCH3 HO O OCH3 HO O OCH H 3 OH H3C OH OGl HO O OH OCH3 O OCH3 OH OH OH OGl HO O OH HO O HO O OH OH

OH OH OH OH OH OH

Figure 5.1. Reaction of malvidin-3-glucoside with catechin and acetaldehyde.

Acetaldehyde is found in red wine as a byproduct of yeast fermentation and as a result of oxidation.6,8,9,32 Oxygen reacts via several metal-catalyzed steps to form acetaldehyde from the oxidation of ethanol (Figure 5.2).10,31 Oxygenation techniques (e.g., micro-oxygenation, aging in oak cooperage) take advantage of this chemistry through the gradual addition of small amounts of

97 98 oxygen with the overall goal of acetaldehyde formation but without an increase in dissolved oxygen concentration.18 Micro-oxygenation has been shown to significantly improve the color stability of treated wines as monomeric anthocyanins are oxidized to polymeric pigments.18,21,26,27

Recently, periodic aeration has been investigated wherein the same dose of oxygen can be added, but through scheduled aeration of small amounts of the bulk wine and mixing. This technique was shown to confer similar benefits to red wine compared to micro-oxygenation.30

Fe(II) Fe(II)

2+ O2 Fe(II)-O2 2Fe(III) + H2O2 Fe(II) Fe(II) Fe(III) Fe(III) H2O HO 0.5 H2O2 0.5 O2

CH CHO CH CHOH EtOH 3 3

Figure 5.2. Formation of acetaldehyde from oxygen (Adapted from Danilewicz).31

The reduction of oxygen to acetaldehyde involves several reactive intermediates, which can promote reactions that are deleterious with respect to wine quality (Figure 5.3).10 Hydroxyl radicals and ortho-quinones react readily with other wine components.8,10,11 These reactive intermediates contribute to some of the possible deleterious effects of oxygen exposure including the loss of desirable aromas and unfavorable changes in wine color.18,27 Thiols, which are common aroma compounds, have been found to be especially susceptible to consumption by reaction with reactive intermediates..45–49,57 The use of oxygenation to form endogenous acetaldehyde is risky due to the non-selective reactivity of reactive oxygen species, namely the

98 99 hydroxyl radical. Other oxidation reactions of anthocyanins, e.g., direct condensation with flavanols, can also occur that may cause a loss of monomeric anthocyanins without the formation of pH and bisulfite stable pigments.66,253

R O

Fe(II) H2O2 O H2O

R OH CH3CHOH CH3CHO Fe(III) HO EtOH OH

Figure 5.3. Oxidation reactions leading to acetaldehyde formation (Adapted from Danilewicz).10

One of the key factors affecting the reactivity of acetaldehyde is the presence of sulfur dioxide (SO2). Bisulfite, the major form of SO2 at wine pH, binds quickly and strongly with

-6 61 acetaldehyde to yield 1-hydroxyethanesulfonate (Kd = 2.4 x 10 M). This sulfonate has been shown to have antioxidant activity comparable to bisulfite itself.62 Acetaldehyde has also been shown to react with flavonoids in the presence of equimolar bisulfite and from the addition of the acetaldehyde-bisulfite adduct in model systems (Chapters 3 & 4).

The present work examines the role of acetaldehyde in oxidation reactions of flavonoids.

Exogenous acetaldehyde addition is examined in a model system and in red wine post- fermentation. This treatment is compared to the addition of the acetaldehyde-bisulfite adduct and to the use of oxygenation.

99 100

5.3 Materials and Methods

5.3.1 Materials

(+)-Catechin hydrate, sodium dodecyl sulfate, triethanolamine, maleic acid, sodium chloride, bovine serum albumn, and sodium metabisulfite were purchased from Sigma Aldrich

(St. Louis, MO). Acetaldehyde, potassium metabisulfite, and tartaric acid were purchased from

Alfa Aesar (Ward Hill, MA). LC/MS grade formic acid was purchased from Fisher Scientific

(Pittsburgh, PA). Sodium hydroxide (10.00 N) and 200 proof ethanol were purchased from VWR

International (Radnor, PA). Glacial acetic acid was purchased from J.T. Baker (Phillipsburg, NJ).

Hydrochloric acid was purchased from EMD Chemicals (Gibbstown, NJ). Water was purified through a Millipore Q-Plus (Millipore Corp., Bedford, MA) purification train.

5.3.2 Bisulfite Adduct Synthesis

Sodium 1-hydroxyethanesulfonate was synthesized according to the protocol described by Andersen et al.62 Acetaldehyde (1.0 g, 24 mmol) was dissolved in cold ethanol (25 mL) and stirred over ice. A solution of 40% aqueous sodium metabisulfite (3.75 mL, 20 mmol bisulfite) was added dropwise to the stirred solution. Ethanol (25 mL) was then added to the solution. The reaction mixture was kept at 4 °C overnight. The precipitate was then collected by cold, vacuum filtration, washed several times with cold ethanol, and then dried under a gentle stream of nitrogen. Sodium 1-hydroxyethanesulfonate was isolated as a white powder (67% yield).

100 101

5.3.3 Model Reaction Mixture Preparation

Model wine was composed of 5% w/v tartaric acid, 12% ethanol v/v with pH adjusted using 10 N NaoH. Model experiments were prepared with 500 µM treatment (bisulfite, acetaldehyde, or sulfonate) added from a stock solution or solid (sulfonate only) and 1000 mg/L

(+)-catechin. All experiments were completed in triplicate. Glassware was soaked in 5% nitric acid overnight and rinsed to remove trace metals before use. Each reaction mixture was separated into 1 mL aliquots and stored at room temperature in an anaerobic chamber (Anaerobe Systems,

Morgan Hill, CA). The anaerobic chamber was filled with 95% argon/ 5% hydrogen gas and had a palladium catalyst to remove residual oxygen. Anaerobic status of the chamber was confirmed using a PreSens oxygen meter (PreSens, Regensburg, Germany). All solutions were deoxygenated using argon gas before being introduced into the chamber and all materials were left in the anaerobic chamber overnight prior to starting experiments. Timepoints for each experiment were stored at -80 °C and thawed immediately prior to analysis.

5.3.4 Catechin Analysis

Concentrations of catechin were determined by HPLC using a Shimadzu system with

10ADvp pumps and a SIL-20AC HT temperature-controlled autosampler (4 °C). Separation was achieved on a reverse phase Agilent Zorbax SB-Aq column (2.1 mm x 150 mm, 3.5 µm; Agilent

Tech., Santa Clara, CA). Samples were filtered through 0.45 µm PTFE syringe filters prior to analysis. The injection volume was 10 µL, column temperature was 30 °C, and flow rate was held at 0.2 mL/min for all samples. The mobile phase consisted of 0.1% formic acid in water (A) and methanol (B). Catechin was eluted using an isocratic method of 25% B and detected at 280 nm.

Concentrations were calculated based on external calibration curve of catechin.

101 102

5.3.5 Treatment of Red Wine with Free and Bound Acetaldehyde

Wine was produced at pilot scale from Cabernet Franc (Vitis vinifera) grapes that were hand-harvested from the New York State Agricultural Experiment Station (Cornell University) in

Geneva, NY without any sulfur dioxide or tannin additions. Immediately following the completion of the alcoholic fermentation (<0.25% residual sugar), the wine was sterile-filtered

(0.45 µm) and had the following composition prior to starting the experiments: pH 3.26, 8.3 g/L titratable acidity, 12.3% v/v ethanol. All materials used for these experiments were sterilized prior to use. Experiments were conducted in an anaerobic chamber as described above. Wine was deoxygenated using argon gas before being introduced into the chamber and all materials were left in the anaerobic chamber overnight prior to starting experiments.

Bisulfite, acetaldehyde, and sulfonate treatments were prepared at 500 µM concentrations. Bisulfite and acetaldehyde were added to wine from stock solutions. Sodium 1- hydroxyethanesulfonate was added as a solid and was synthesized as described above. All treatments were prepared in 500-mL volumetric flasks, poured into 12 ml GC vials, and capped with minimal headspace for storage. After 8 weeks, samples were frozen in aliquots at -80 °C and aliquots were thawed immediately prior to each analysis.

5.3.6 Treatment of Red Wine with Endogenous and Exogenous Acetaldehyde

The Cabernet Franc wine described above was used to compare acetaldehyde and oxygen treatments. Oxygen and acetaldehyde were added as equimolar treatments based on an ideal conversion of oxygen to acetaldehyde. Oxygenation was achieved using periodic aeration.30

Acetaldehyde was added according to a similar schedule. A total oxygen or acetaldehyde concentration of 500 µM was delivered to the wine as four separate additions. 102 103

All wines were stored in triplicate 500 mL ground glass flasks in an anaerobic chamber.

An aliquot of wine from each replicate was taken from the flask and anaerobic chamber in a sterilized 1 L media bottle with cap. The bottle was thoroughly shaken for 1 minute to achieve saturation with dissolved oxygen (DO). This wine was then reintroduced to the bulk sample in the anaerobic chamber. An equivalent amount of acetaldehyde (125 µM) was added from a concentrated stock solution according to the same schedule.

8 weeks after the beginning of the experiment, or 4 weeks after the final aeration and acetaldehyde addition, samples were frozen in aliquots at -80 °C and aliquots were thawed immediately prior to each analysis.

5.3.7 Pigment Characterization

Samples from each wine experiment taken at 8 weeks were characterized using the modified Somers assay. The modified Somers assay was performed using a high throughput procedure.218 Wine samples were centrifuged at 2500×g for 5 min prior to analysis. Solutions were incubated in 1.5-mL capacity microcentrifuge tubes before being transferred to 96-well plates (Greiner Bio One UV-Star, Monroe, NC) for absorbance readings at 280, 420, and 520 nm using a Multiskan GO microplate reader (Thermo-Scientific, Waltham, MA). All Somers color parameters were calculated as previously reported. Wine samples were analyzed in duplicate.

Wine samples were also characterized using the Harbertson-Adams assay to differentiate pigments by protein precipitation. Wine samples were analyzed using the microplate method previously published.156 Readings at 280, 420, and 520 nm were taken in 96-well microplates using a Multiskan GO microplate reader. Wine samples were analyzed in duplicate.

103 104

5.3.8 Statistical Analysis

5.4 Results and Discussion

5.4.1 Activity of Acetaldehyde-Bisulfite Adduct in Model Solutions

The addition of 1-hydroxyethanesulfonate was used to determine the reactivity acetaldehyde from this bound form. Previous work in our lab showed that acetaldehyde reacts in the presence of equimolar bisulfite and from its sulfonate when it is assumed to be bound and inert (Chapters 3 & 4). In the present study, a sodium salt of the acetaldehyde-bisulfite adduct was synthesized. The sodium 1-hydroxyethane sulfonate was first added to model wine solutions containing (+)-catechin as a representative flavanol in wine.

The reactivity of acetaldehyde from its free and bound forms was measured by the loss of monomeric catechin at several pH values between pH 2 and 4. Monomeric catechin is consumed as it forms large ethyl-bridged oligomers (Chapter 3). The results of these experiments are shown in Figure 5.4. As has been seen previously by our lab and others, there is a clear effect of pH on the rate of reaction of acetaldehyde with catechin (Chapter 3).161 This is due to the important first step of acid-catalyzed protonation of acetaldehyde to form a carbocation. We would expect to see an effect of pH on the reactivity of acetaldehyde from its bound form due to this mechanism as well. There is an effect of pH on the dissociation constant (Kd) of the sulfonate, though values are essentially constant in the pH range examined here.254

104 105

A B 1050 1025

1000 950

975

850

Catechin (mg/L) 950 Catechin (mg/L)

750 925 0 10 20 pH 4 30 0 10 20 30 Time (Days) Time (Days) pH 3.5 pH 3 pH 4 pH 2.5 pH 3.5 pH 2 pH 3 pH 2.5 pH 2 Figure 5.4. Catechin consumption at pH 2 – 4 by acetaldehyde (A) and 1-hydroxyethanesulfonate (B).

There was statistically significant consumption of catechin by the sulfonate at pH 2.0 and

2.5 (Figure 5.4B). At higher pH levels, there was no significant consumption of catechin compared to a control containing no acetaldehyde. At pH 2.0 and 2.5, the catechin consumption by the sulfonate was significantly lower than that by free acetaldehyde. These results demonstrate that acetaldehyde is indeed reactive when added as its bisulfite-bound form, 1- hydroxyethanesulfonate; however, as expected, the activity of acetaldehyde from its sulfonate form is significantly less than that of free acetaldehyde. The effect of pH on the reactivity illustrates the importance of time for these reactions to take place. As pH only increases the rate of the reaction and not the actual mechanism of reaction, the sulfonate is likely to be effective at pH values above 2.5, but over a longer time scale than was studied here.

These results demonstrate that acetaldehyde bound to bisulfite can participate in many desired aging reactions for red wine. While the bound form of acetaldehyde is not as reactive as 105 106 free acetaldehyde, it should not be considered inert as is commonly thought. Given the significant lengths of time (i.e., months to years) that red wines are often allowed to age between vinification and consumption, bound acetaldehyde could contribute significantly to the beneficial oxidation reactions for color stability.

5.4.2 Effect of Acetaldehyde and Acetaldehyde-Bisulfite Adduct on Red Wine Color

Based on our results in model wine solutions, we then compared the effects of free and bound acetaldehyde on the color stability of a red wine. The wine, prepared with Cabernet Franc grapes, was at a higher pH (3.26) than the model solutions where we saw significant catechin consumption. However, this pH level is directly relevant to other red wines and we would still expect to see the effect of reactions of acetaldehyde, albeit at a slower rate. The results shown here are from wine samples taken 8 weeks after the addition of each treatment. Each treatment was designed to provide equimolar (500 µM) levels of acetaldehyde and sulfonate compared to a control containing the equivalent molarity of bisulfite and a control with no bisulfite or acetaldehyde additions.

Wines were analyzed using a modified Somers assay for specific color parameters.218 The results of this assay are summarized in Table 5.1. Acetaldehyde treatment was observed to affect all parameters; sulfonate treatment produced significant differences from control for some. There was a significant effect of bisulfite on all parameters as well indicating that there were ongoing oxidation reactions in the control with no additions.

106 107

Table 5.1. Color parameters from the modified Somers assay for control, bisulfite treated, acetaldehyde treated, and sulfonate treated wines after 8 weeks.

Control Bisulfite Acetaldehyde Sulfonate

Chemical age 1 0.394 ± 0.002 b 0.331 ± 0.001 a 0.441 ± 0.006 d 0.405 ± 0.003 c

Chemical age 2 0.1242 ± 0.0016 b 0.0872 ± 0.0006 a 0.1679 ± 0.0036 d 0.1311 ± 0.0015 c Degree of ionization of 24.8 ± 0.7 b 16.2 ± 0.2 a 30.1 ± 0.4 c 24.8 ± 0.4 b anthocyanins (%) Total anthocyanins 173.5 ± 3.0 b 215.9 ± 1.2 c 142.6 ± 2.9 a 174.5 ± 1.7 b (mg/L) Color density (au) 5.80 ± 0.09 b 4.74 ± 0.03 a 6.19 ± 0.05 c 5.92 ± 0.04 b Color density, SO - 2 5.73 ± 0.03 b 5.40 ± 0.04 a 6.17 ± 0.08 d 5.92 ± 0.03 c corrected (au) Hue 0.6537 ± 0.0005 c 0.6669 ± 0.0010 d 0.6278 ± 0.0015 a 0.6328 ± 0.0013 b SO -resistant pigments 2 3.45 ± 0.02 b 3.33 ± 0.02 a 3.77 ± 0.05 d 3.62 ± 0.02 c (au) Total phenolics (au) 19.5 ± 0.4 b 20.7 ± 0.1 c 18.6 ± 0.3 a 20.0 ± 0.3 b a Averaged results are shown ± one standard deviation of the mean, and results in the same row with different letters (a, b) are significantly different (p < 0.05).

Acetaldehyde treatment resulted in statistically significant improvements (i.e., increases in chemical age values, color density values, and the degree of ionization of anthocyanins) for every parameter compared to both controls. There was a slight but significant decrease in total phenolics suggesting that some tannin precipitation had occurred, as would be expected when tannin is allowed to react with acetaldehyde. There was also a significant change in hue, which indicates the formation of a different profile of pigments. Acetaldehyde treatment resulted in a decrease in the concentration of monomeric anthocyanins. As polymeric pigments are formed, these unstable monomers are consumed and converted to more stable pigments. These stable pigments are seen by an increase in the concentration of SO2-resistant pigments.

107 108

Acetaldehyde contributes to the formation of SO2-resistant pigments by reacting with anthocyanins as well as flavanols. The main fraction of pigments contributing to this effect is the pyranoanthocyanins (Figure 5.1), which are formed as a result of the acid-catalyzed cleavage of ethylidene bridges from reaction with acetaldehyde. Pyranoanthocyanin formation contributes to

84,95,101,112,118 greater color stability to pH and SO2 bleaching seen with red wine oxidation.

Treatment of red wine with acetaldehyde resulted in a significant effect on all color parameters when acetaldehyde was added at a concentration that is relevant under wine conditions. The amount of acetaldehyde added is low compared to the range that is found in red wines (8 – 212 mg/L).6,7 The potential impact for low concentrations of endogenous acetaldehyde from fermentation and oxidation or exogenous acetaldehyde on color stability is significant and should continue to be explored.

The reactivity of acetaldehyde from its sulfonate form was markedly lower compared to its free form; however, there were significant improvements in some color parameters based on the Somers assay. There were significant differences in all parameters compared to the bisulfite treatment (i.e., wine with the same total SO2 concentration). There were no significant differences between the sulfonate treatment and the control without additions in terms of the degree of ionization of anthocyanins, color density, total phenolics, or total anthocyanins. There was a significant increase in both chemical ages indicating a shift from monomeric anthocyanins to polymeric pigments. There was also a significant change in hue and SO2-corrected color density.

The most important improvement is that in SO2-resistant pigments. While the increase in color stable pigments is mitigated compared to the free acetaldehyde treatment, there is significant formation of polymeric pigments.

The polymeric pigments formed were further examined using the Harbertson-Adams assay based on protein precipitation.156,255 This allowed differentiation of the polymeric pigments

108 109 as large or small polymeric pigments based on their interaction with bovine serum albumin

(BSA). Small polymeric pigments (SPP) are defined as those that do not precipitate with BSA and large polymeric pigments (LPP) as those that do precipitate.255 SPP and LPP have been shown to increase during winemaking compared to the starting fruit.255

Acetaldehyde treatment significantly increased the amount of LPP and SPP compared to both control groups (Table 5.2); however, there was no effect of sulfonate treatment on these parameters compared to control. There was a significant difference between the bisulfite treated group and the sulfonate samples. The effect of acetaldehyde on LPP and SPP shows that its reactions are contributing to both of these groups of pigments. The fraction of pigments precipitated as LPP are likely reaction products of acetaldehyde and anthocyanins with wine tannins. SPP may be vitisins and smaller reaction products from anthocyanins and small flavanols.66,255

Table 5.2. Color parameters from the Harbertson-Adams assay for control, bisulfite treated, acetaldehyde treated, and sulfonate treated wines after 8 weeks.a

Control Bisulfite Acetaldehyde Sulfonate

Anthocyanin (mg/L M3G) 255.6 ± 3.4 b 338.7 ± 4.9 c 210.3 ± 6.2 a 257.4 ± 1.7 b

LPP (au) 0.31 ± 0.06 a 0.29 ± 0.07 a 0.48 ± 0.08 b 0.33 ± 0.03 a

SPP (au) 1.78 ± 0.05 b 1.37 ± 0.04 a 2.32 ± 0.06 c 1.90 ± 0.08 b a Averaged results are shown ± one standard deviation of the mean, and results in the same row with different letters (a, b) are significantly different (p < 0.05).

The results of this experiment demonstrate the significant impact of low concentrations of acetaldehyde as well as the possible influence of bound acetaldehyde on red wine color post- fermentation. Acetaldehyde treatment significantly improved every measure of color stability

109 110 from the Somers assay and Harbertson-Adams assay. Sulfonate treatment improved some color parameters but had significantly less effect than treatment with acetaldehyde alone, suggesting that bound acetaldehyde may contribute to color stability but to a lesser extent. Based on our results obtained under model conditions, we would expect the impact of bound acetaldehyde to become more evident with time.

5.4.3 Effect of Exogenous Acetaldehyde and Oxygenation on Red Wine Color

Acetaldehyde is typically added indirectly using oxygenation techniques like micro- oxygenation. Oxygen ingress allows the formation of acetaldehyde by metal-catalyzed oxidation of ethanol (Figure 5.2). Assuming there are no side reactions involving reaction intermediates,

31 one mole of acetaldehyde would be formed by each mole of O2 consumed. Based on this ideal ratio, equimolar amounts of O2 and acetaldehyde were compared in this experiment.

Oxygenation was achieved by periodic aeration. Laurie et al. demonstrated that periodic aeration achieves similar effects as micro-oxygenation of the same total dose of DO.30 By comparing the effect of oxidation by DO and endogenous acetaldehyde to the assumed concentration as exogenous acetaldehyde we are addressing two main issues - the success of each treatment in improving wine quality and the efficiency with which oxygen is converted to acetaldehyde.

Wines treated with endogenous acetaldehyde (i.e., resulting from the hydroxyl radical- mediated oxidation of ethanol) and exogenous acetaldehyde were compared to a control using a modified Somers assay for characterizing wine pigments. The results of this assay can be found in

Table 5.3. Overall, these parameters showed significant effects of both oxidation treatments.

Both treatment groups had significantly different results for both chemical ages and total anthocyanins. Acetaldehyde treatment also had significant differences from control but not from 110 111 oxygen treatment for degree of ionization of anthocyanins, total phenolics, and SO2-corrected color density. Acetaldehyde treatment was significantly different from both the control and the oxygen treated wine for hue, color density, and SO2-resistant pigments.

Table 5.3. Color parameters from the modified Somers assay for control, acetaldehyde treated, and oxygenated (periodic aeration) wines after 8 weeks.a

Control Acetaldehyde Oxygen

Chemical age 1 0.409 ± 0.009 a 0.458 ± 0.009 b 0.455 ± 0.009 b

Chemical age 2 0.135 ± 0.003 a 0.178 ± 0.002 b 0.170 ± 0.010 b

Degree of ionization of anthocyanins (%) 25.6 ± 0.5 a 30.5 ± 0.7 b 27.7 ± 1.7 ab

Total anthocyanins (mg/L) 170.3 ± 2.9 b 143.9 ± 2.2 a 140.7 ± 6.6 a

Color density (au) 6.07 ± 0.05 a 6.49 ± 0.06 b 6.00 ± 0.12 a

a b ab Color density, SO2-corrected (au) 6.03 ± 0.08 6.45 ± 0.10 6.09 ± 0.11

Hue 0.654 ± 0.002 b 0.618 ± 0.002 a 0.662 ± 0.012 b

a b a SO2-resistant pigments (au) 3.63 ± 0.05 3.97 ± 0.06 3.65 ± 0.09

Total phenolics (au) 19.6 ± 0.2 b 19.0 ± 0.3 a 19.2 ± 0.4 ab a Averaged results are shown ± one standard deviation of the mean, and results in the same row with different letters (a, b) are significantly different (p < 0.05).

The results of the Somers assay illustrate the shifts in wine pigments one would expect given these oxidation treatments. Both treatments resulted in a shift away from monomeric anthocyanins demonstrated by the effects on chemical age and total anthocyanins. However, only acetaldehyde treatment showed evidence of the formation of the more beneficial and stable polymeric pigments. The change in hue is likely due to the shifts in color seen from ethylidene bridging and pyranoanthocyanin formation. Direct condensation products of anthocyanins typically do not have a change in observed color.253

111 112

The significant impact of acetaldehyde on SO2-resistant pigments represents its role in polymeric pigment formation. Most likely, these stable pigments are pyranoanthocyanins and pyranoanthocyanin-flavanols, which are formed as a result of ethylidene bridging of flavonoids.

Other oxidation products of anthocyanins, which may be formed as a result of DO, are not necessarily stable to SO2 bleaching.

The wines were then analyzed using the Harbertson-Adams assay to differentiate large and small polymeric pigments by BSA precipitation. As seen in Table 5.4, both the acetaldehyde and oxygen treatments did have significant effects on the pigments characterized by this assay.

The concentration of anthocyanins was shown to decrease by both treatments as was seen in the

Somers assay. The concentration of LPP and SPP did also increase with each treatment. The concentration of LPP in the oxygen treated wine was higher but not significantly different while the SPP was significantly different from the control. LPP and SPP were significantly higher in the acetaldehyde treated wine compared to the control. There were no significant differences between the oxidation treatments. These results show that acetaldehyde, endogenous or exogenous, is contributing to the formation of large and small polymeric pigments.

Table 5.4. Color parameters from the Harbertson-Adams assay for control, acetaldehyde treated, and oxygenated (periodic aeration) wines after 8 weeks.a

Control Acetaldehyde Oxygen

Anthocyanin (mg/L M3G) 249.0 ± 3.0 b 214.4 ± 5.1 a 211.3 ± 7.6 a

LPP (au) 0.43 ± 0.05 a 0.62 ± 0.07 b 0.56 ± 0.08 ab

SPP (au) 2.05 ± 0.07 a 2.47 ± 0.10 b 2.34 ± 0.15 b a Averaged results are shown ± one standard deviation of the mean, and results in the same row with different letters (a, b) are significantly different (p < 0.05).

112 113

While both oxidation treatments affected important pigment characteristics of the wine, exogenous acetaldehyde was more effective at improving overall color stability. The most important indicator of this improvement is the significant increase in SO2-resistant pigments by acetaldehyde treatment. Although oxygen treatment was seen to decrease monomeric anthocyanin concentrations by the same amount, only the acetaldehyde treatment converted anthocyanins to

SO2-resistant pigments.

These results suggest that the conversion of oxygen to acetaldehyde is not an efficient process. Instead of the ideal conversion of one mole of O2 to one mole of acetaldehyde, there appears to be less than equimolar acetaldehyde formed based on the results shown here. Other side reactions are likely occurring from intermediates of oxidation and could be deleterious to wine components. Specifically, reactions of hydroxyl radicals and quinones can contribute to undesirable side reactions.8,10,11 Anthocyanins may also be consumed by direct condensation reactions that may not contribute to color stability.66,253 Oxygen exposure is therefore an indirect and non-selective method of generating endogenous acetaldehyde.

Acetaldehyde, delivered either exogenously or generated endogenously, plays a significant role in the reactions of flavonoids in red wine. As was shown under model wine conditions, acetaldehyde is reactive when added as its bound form, 1-hydroxyethanesulfonate, at low pH. In Cabernet Franc wine, bound acetaldehyde was able to improve several important pigment parameters. By comparing exogenous acetaldehyde to oxygenation, we showed that the formation of endogenous acetaldehyde is indirect and does not produce the benefits of the equivalent amount of acetaldehyde. In both wine experiments, we demonstrated that a small amount of exogenous acetaldehyde added post-fermentation was able to significantly improve all color parameters examined including SO2-resistant pigments. These results demonstrate the

113 114 importance of the reactions of acetaldehyde in red wine and the potential for exogenous acetaldehyde to be used to gain these benefits.

5.5 Acknowledgements

The authors thank Denise Gardner for her assistance with wine production as well as for her insightful discussions throughout the study.

114

Chapter 6

Conclusions and Recommendations for Future Work

Results from my work have demonstrated several significant and novel facets of acetaldehyde reactivity in red wine. In chapter 2, I showed that exogenous acetaldehyde can be used to significantly improve color stability during the fermentation of a red wine. In chapter 3, I characterized the reactions of the specific flavonoids involved and established the activity of acetaldehyde in the presence of equimolar bisulfite. In chapter 4, I confirmed the relationship between dissociation and reactivity of several aldehyde-bisulfite adducts. And in chapter 5, I demonstrated the efficacy of low concentrations of exogenous acetaldehyde in red wine, the reactivity acetaldehyde from its bound form, and the benefits of exogenous acetaldehyde over oxygenation. This work has advanced a basic understanding of the chemistry that underlies aldehyde-phenolic interactions, but also provides practical information for winemakers and a possible novel technique for red wine production.

6.1 Reaction of Flavonoids with Acetaldehyde

Mechanistic details for the reaction of acetaldehyde were elucidated using model wine studies. Using catechin as a model flavan-3-ol, it was established that pH accelerates the rate of reaction with acetaldehyde and that dissolved oxygen concentration has no effect on this rate.

More significantly, it was also shown that acetaldehyde reacts with flavonoids in the presence of equimolar bisulfite when the aldehyde is assumed to be bound and inert. This is the first evidence for the activity of acetaldehyde at these relative concentrations of bisulfite. This reactivity was

then shown in systems containing malvidin-3-glucoside as a model anthocyanin where the effects on color stability parameters (e.g., SO2-resistant pigments) were demonstrated. Finally, the products of these reactions were confirmed using a novel technique, MALDI-TOF MS.

In order to further understand the nature of the reactions of flavonoids with acetaldehyde, there are several possible routes for future research. First, the reaction products should be further characterized by isolating and purifying specific compounds of interest, namely larger ethyl- bridged products and pyranoanthocyanins. Subsequent structural analysis by MS and NMR would allow confirmation of the proposed structures. Monitoring specific products over time would also be instrumental in understanding the evolution of ethyl-bridged compounds through acid- catalyzed cleavage to pyranoanthocyanin formation. Finally, the reactions of oligomeric tannins could be studied using purified tannins as starting materials. This would allow the isolation of the impact of acid-catalyzed cleavage on tannin size and the reaction of oligomeric tannins with acetaldehyde and anthocyanins, characteristics that are difficult to understand in the heterogeneous mixture of wine tannins. Methods including small-angle x-ray scattering (SAXS) and gel permeation chromatography (GPC) could also be used to characterize changes in the size and shape of the tannin after reaction with acetaldehyde.

6.2 Reactivity of Aldehydes from α-Hydroxyalkylsulfonates

Since acetaldehyde was shown to be reactive in the presence of an equimolar concentration of bisulfite, I then pursued studies on other aldehydes when added to systems as their bound bisulfite adducts. I synthesized these α-hydroxyalkylsulfonate adducts as their sodium salts and added them to model wine systems. This study was able to demonstrate the differences in reactivity of a variety of aldehydes based on their structures. It also showed the direct effect of the dissociation constant of the bisulfite adduct on the activity of the aldehyde. 116

The aldehyde with the highest dissociation constant, benzaldehyde, had no significant difference between adding the aldehyde and the aldehyde-bisulfite adduct. The aldehyde with the lowest dissociation constant, formaldehyde, had a significant effect of the aldehyde-bisulfite adduct wherein there was no reactivity seen.

Based on the large differences seen between aldehydes and the impact of Kd on the reactivity of their bisulfite adducts, this is an area that deserves further research. There are many aldehydes present in wine including glyoxylic acid, glyceraldehyde, and pyruvic acid. These aldehydes should be characterized in similar studies and attempts should be made to synthesize their bisulfite adducts. Their oligomeric catechin products should also be characterized and compared. The free bisulfite concentrations from each of these adducts could be characterized as the aldehydes react and are consumed. Dissociation of aldehydes from their sulfonates could also be investigated by analyzing the exchange of free and bound aldehydes in mixtures. As many aldehydes, including those studied in this work, have distinctive aromas, the impact of bisulfite concentration on odor should be quantified using GC and sensory analysis.

6.3 Exogenous Acetaldehyde as Wine Treatment

In several experiments, it was established that exogenous acetaldehyde additions can be used to improve the color stability of red wine. Exogenous acetaldehyde was first used during the fermentation of a red wine. When wines were analyzed after the completion of primary fermentation, the high acetaldehyde treatment (1000 mg/L) significantly improved color stability and lowered protein precipitation. This confirmed the potential for treatment of red wines early in their production with exogenous acetaldehyde. These results were also the impetus for the experiments on exogenous acetaldehyde treatment that followed.

117

In subsequent studies, treatment of red wine after primary fermentation with lower concentrations of acetaldehyde (500 µM, 22 mg/L) proved to be very effective. This treatment significantly improved all measures of color stability including SO2-resistant pigments.

Monomeric anthocyanins were successfully converted to stable, polymeric pigments by reaction with acetaldehyde and flavanols. These results demonstrate the viability of exogenous acetaldehyde treatment as a method for improving wine quality.

In the final study that I conducted in red wine, exogenous acetaldehyde was compared to a typical oxygenation regime used for wine production. The concentration of acetaldehyde added was based on the reduction of oxygen in ideal conditions, where one mole of oxygen would form one mole of endogenous acetaldehyde. Both treatments improved the color of the wines after 8 weeks; however, the results of this experiment clearly showed that oxygenation is an indirect method of improving color. Though the same amount of monomeric anthocyanins was lost with each treatment, only exogenous acetaldehyde treatment converted these anthocyanins to SO2- resistant pigments. Exogenous acetaldehyde is therefore a more direct and selective method of improving wine color than oxygenation. These results demonstrate that winemakers could use acetaldehyde as a tool similar to those used for oxygenation (e.g., micro-oxygenation).

Exogenous acetaldehyde treatment of red wines merits more research based on its success in these studies. Tannin additions should be combined with acetaldehyde treatment to see if this can significantly improve color stability, especially in low tannin wines. Wines with variable ratios of tannins/anthocyanins should also be treated to identify optimal ratios for these reactions.

Long-term effects on color and mouthfeel should be determined using chemical and sensory analyses. In comparison to oxygenation treatment, aroma should also be considered, as the loss of desirable aromas has been associated with oxygenation. Aroma compounds of acetaldehyde and oxygen treated wines could be compared using sensory and GC techniques. The effects of

118

oxygenation compared to acetaldehyde should be determined at a larger scale so micro- oxygenation equipment could be used for a more relevant comparison for winemakers.

6.4 Concluding Remarks

Acetaldehyde has been assumed to be inert in the presence of sulfur dioxide as it forms an adduct with bisulfite. We have shown that acetaldehyde is able to react in the presence of equimolar bisulfite. We also demonstrated that when added as its bound form, 1- hydroxyethanesulfonate, acetaldehyde is able to react. These results indicate that acetaldehyde in wine, even when assumed bound, may be contributing to the benefits associated with acetaldehyde reactions.

Although the importance of acetaldehyde in beneficial reactions is known, there has been little research based on the utilization of exogenous acetaldehyde as a treatment for red wine.

Academic research and industry practices have focused on the acetaldehyde formed from passive and deliberate exposure to oxygen. This has included methods optimized to form acetaldehyde without accumulating oxygen like micro-oxygenation. We have shown that the endogenous acetaldehyde formed during oxygenation does not produce the same advantageous results seen with an equimolar treatment of exogenous acetaldehyde.

Exogenous acetaldehyde treatment at reasonable, low concentrations (~20 mg/L) after primary fermentation can significantly improve color stability of red wine. This technique could be used by winemakers to the improve color, mouthfeel, and quality of the finished wines. The results of this and future work will enable winemakers to take full advantage of acetaldehyde and its role in color stability.

119

References

1. Ribéreau-Gayon, P., Dubourdieu, D. & Donèche, B. Handbook of enology. (John Wiley,

2006).

2. Tao, Y., García, J. F. & Sun, D.-W. Advances in wine aging technologies for enhancing

wine quality and accelerating wine aging process. Crit. Rev. Food Sci. Nutr. 54, 817–835

(2014).

3. U.S. Environmental Protection Agency. Acetaldehyde. (Environmental Criteria and

Assessment Office, Office of Health and Environmental Assessment, Office of Research

and Development, 1987).

4. CFR - Code of Federal Regulations Title 21. Available at:

http://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfcfr/CFRSearch.cfm?fr=182.60.

(Accessed: 19th January 2016)

5. Amoore, J. E. & Hautala, E. Odor as an ald to chemical safety: Odor thresholds compared

with threshold limit values and volatilities for 214 industrial chemicals in air and water

dilution. J. Appl. Toxicol. 3, 272–290 (1983).

6. Liu, S.-Q. & Pilone, G. J. An overview of formation and roles of acetaldehyde in

winemaking with emphasis on microbiological implications. Int J Food Sci Tech 35, 49–61

(2000).

7. Lachenmeier, D. W. & Sohnius, E.-M. The role of acetaldehyde outside ethanol metabolism

in the carcinogenicity of alcoholic beverages: Evidence from a large chemical survey. Food

Chem. Toxicol. 46, 2903–2911 (2008).

121

8. Waterhouse, A. & Laurie, V. Oxidation of wine phenolics: A critical evaluation and

hypotheses. Am. J. Enol. Vitic. 3, 306–313 (2006).

9. Singleton, V. L. Oxygen with Phenols and Related Reactions in Musts, Wines, and Model

Systems: Observations and Practical Implications. Am. J. Enol. Vitic. 38, 69–77 (1987).

10. Danilewicz, J. C. Review of Reaction Mechanisms of Oxygen and Proposed Intermediate

Reduction Products in Wine : Central Role of Iron and Copper. Am. J. Enol. Vitic. 54, 73–

85 (2003).

11. Cheynier, V. V., Atanasova, V., Fulcrand, H., Mazauric, J.-P. & Moutounet, M. Oxygen in

wine and its role in phenolic reactions during ageing. in Proceedings of the ASVO seminar

uses of gases in winemaking 23–27 (2002).

12. Ribereau-Gayon, J. Contribution a l’etude des oxydations et reductions dans les vins.

(Delmas, 1933).

13. Vidal, J. C., Dufourcq, T., Boulet, J. C. & Moutounet, M. Les apports d’oxygene au cours

des traitements des vins. Bilan des observations sur site, 1ere partie. Rev. FRANCAISE

OENOLOGIE 24–31 (2001).

14. Castellari, M., Simonato, B., G. B. Tornielli, Spinelli, P. & R. Ferrarini. Effects of different

enological treatments on dissolved oxygen in wines. Ital. J. Food Sci. 16, 387–396 (2004).

15. Oxidation in foods and beverages and antioxidant applications. (Woodhead Pub, 2010).

16. Ugliano, M. Oxygen contribution to wine aroma evolution during bottle aging. J. Agric.

Food Chem. 61, 6125–6136 (2013).

17. Smith, P. A., McRae, J. M. & Bindon, K. A. Impact of winemaking practices on the

concentration and composition of tannins in red wine: Impact of winemaking practices on

tannins. Aust. J. Grape Wine Res. n/a–n/a (2015). doi:10.1111/ajgw.12188

121 122

18. Gómez-Plaza, E. & Cano-López, M. A review on micro-oxygenation of red wines: Claims,

benefits and the underlying chemistry. Food Chem. 125, 1131–1140 (2011).

19. Paul, R. Micro-oxygenation – where now? in Proceedings of the ASVO seminar uses of

gases in winemaking 18–22 (2002).

20. Laurie, V. F., Law, R., Joslin, W. S. & Waterhouse, A. L. In situ measurements of dissolved

oxygen during low-level oxygenation in red wines. Am. J. Enol. Vitic. 59, 215–219 (2008).

21. Cano-López, M. et al. Effect of micro-oxygenation on color and anthocyanin-related

compounds of wines with different phenolic contents. J. Agric. Food Chem. 56, 5932–41

(2008).

22. Osborne, J. P., Mira de Orduña, R., Pilone, G. J. & Liu, S. Q. Acetaldehyde metabolism by

wine lactic acid bacteria. FEMS Microbiol. Lett. 191, 51–5 (2000).

23. Burns, T. R. & Osborne, J. P. Impact of Malolactic Fermentation on the Color and Color

Stability of Pinot noir and Merlot Wine. Am. J. Enol. Vitic. 64, 370–377 (2013).

24. Burns, T. R. & Osborne, J. P. Loss of Pinot noir Wine Color and Polymeric Pigment after

Malolactic Fermentation and Potential Causes. Am. J. Enol. Vitic. 66, 130–137 (2015).

25. Anli, R. E. & Cavuldak, Ö. A. A review of microoxygenation application in wine. J. Inst.

Brew. 118, 368–385 (2012).

26. Schmidtke, L. M., Clark, A. C. & Scollary, G. R. Micro-Oxygenation of Red Wine:

Techniques, Applications, and Outcomes. Crit. Rev. Food Sci. Nutr. 51, 115–131 (2011).

27. du Toit, W. J., Marais, J., Pretorius, I. S. & du Toit, M. Oxygen in Must and Wine: A

review. Afr J Enol Vitic 27, 76–94 (2006).

28. Philips, C. Micro-Oxygenation Controllers. Wine Business Monthly 20–23 (2009).

29. Flecknoe-Brown, A. ‘Diffusive’ methods of oxygenating wine: simpler, better, lower-cost

barrel replacement. Aust. N. Z. Grapegrow. Winemak. (2006).

122 123

30. Laurie, V. F., Salazar, S., Campos, M. I., Caceres-Mella, a. & Pena-Neira, a. Periodic

Aeration of Red Wine Compared to Microoxygenation at Production Scale. Am. J. Enol.

Vitic. 65, 254–260 (2014).

31. Danilewicz, J. C. Reactions Involving Iron in Mediating Catechol Oxidation in Model

Wine. Am. J. Enol. Vitic. 64, 316–324 (2013).

32. Wildenradt, H. L. & Singleton, V. L. The production of aldehydes as a result of oxidation of

polyphenolic compounds and its relation to wine agin. 25, (1974).

33. Margalith, P. Z. Flavor Microbiology. (Charles G. Thomas, 1981).

34. Carlton, W. K., Gump, B., Fugelsang, K. & Hasson, A. S. Monitoring Acetaldehyde

Concentrations during Micro-oxygenation of Red Wine by Headspace Solid-Phase

Microextraction with On-Fiber Derivatization. J. Agric. Food Chem. 55, 5620–5625 (2007).

35. Kilmartin, P. A. in Advances in Food and Nutrition Research (ed. Taylor, S. L.) 61, 149–

186 (Academic Press, 2010).

36. González-del Pozo, A., Arozarena, Í., Noriega, M.-J., Navarro, M. & Casp, A. Short- and

long-term effects of micro-oxygenation treatments on the colour and phenolic composition

of a Cabernet Sauvignon wine aged in barrels and/or bottles. Eur. Food Res. Technol. 231,

589–601 (2010).

37. McRae, J. M., Kassara, S., Kennedy, J. a, Waters, E. J. & Smith, P. a. Effect of wine pH and

bottle closure on tannins. J. Agric. Food Chem. 61, 11618–27 (2013).

38. Miller, D. M., Buettner, G. R. & Aust, S. D. Transition metals as catalysts of ‘autoxidation’

reactions. Free Radic. Biol. Med. 8, 95–108 (1990).

39. Welch, K. D., Davis, T. Z., Van Eden, M. E. & Aust, S. D. Deleterious iron-mediated

oxidation of biomolecules1 6. Free Radic. Biol. Med. 32, 577–583 (2002).

123 124

40. Guyot, S., Vercauteren, J. & Cheynier, V. Structural determination of colourless and yellow

dimers resulting from (+)-catechin coupling catalysed by grape polyphenoloxidase.

Phytochemistry 42, 1279–1288 (1996).

41. Li, H., Guo, A. & Wang, H. Mechanisms of oxidative browning of wine. Food Chem. 108,

1–13 (2008).

42. Danilewicz, J. C., Seccombe, J. T. & Whelan, J. Mechanism of interaction of polyphenols,

oxygen, and sulfur dioxide in model wine and wine. Am. J. Enol. Vitic. 59, 128–136 (2008).

43. Barril, C., Clark, A. C. & Scollary, G. R. Chemistry of ascorbic acid and sulfur dioxide as

an antioxidant system relevant to white wine. Anal. Chim. Acta 732, 186–193 (2012).

44. Culleré, Laura, Cacho, J. & Ferreira, V. An Assessment of the Role Played by Some

Oxidation-Related Aldehydes in Wine Aroma. J. Agric. Food Chem. 55, 876–881 (2007).

45. Blanchard, L., Darriet, P. & Dubourdieu, D. Reactivity of 3-Mercaptohexanol in Red Wine:

Impact of Oxygen, Phenolic Fractions, and Sulfur Dioxide. Am. J. Enol. Vitic. 55, 115–120

(2004).

46. Nikolantonaki, M., Chichuc, I., Teissedre, P.-L. & Darriet, P. Reactivity of volatile thiols

with polyphenols in a wine-model medium: Impact of oxygen, iron, and sulfur dioxide.

Anal. Chim. Acta 660, 102–109 (2010).

47. Laurie, V. F. et al. Reactivity of 3-sulfanyl-1-hexanol and catechol-containing phenolics in

vitro. Food Chem. 131, 1510–1516 (2012).

48. Nikolantonaki, M. et al. Identification of Adducts between an Odoriferous Volatile Thiol

and Oxidized Grape Phenolic Compounds: Kinetic Study of Adduct Formation under

Chemical and Enzymatic Oxidation Conditions. J. Agric. Food Chem. 60, 2647–2656

(2012).

124 125

49. Ugliano, M. et al. Evolution of 3-Mercaptohexanol, Hydrogen Sulfide, and Methyl

Mercaptan during Bottle Storage of Sauvignon blanc Wines. Effect of Glutathione, Copper,

Oxygen Exposure, and Closure-Derived Oxygen. J. Agric. Food Chem. 59, 2564–2572

(2011).

50. Nikolantonaki, M. & Waterhouse, A. L. A Method To Quantify Quinone Reaction Rates

with Wine Relevant Nucleophiles: A Key to the Understanding of Oxidative Loss of

Varietal Thiols. J. Agric. Food Chem. 60, 8484–8491 (2012).

51. Kreitman, G. Y. Investigating the Formation and Fate of Ethyl Radicals in Wine. (2013).

52. Elias, R. J., Andersen, M. L., Skibsted, L. H. & Waterhouse, A. L. Key Factors Affecting

Radical Formation in Wine Studied by Spin Trapping and EPR Spectroscopy. Am. J. Enol.

Vitic. 60, 471–476 (2009).

53. Laurie, V. F. & Waterhouse, A. L. Oxidation of glycerol in the presence of hydrogen

peroxide and iron in model solutions and wine. Potential effects on wine color. J. Agric.

Food Chem. 54, 4668–4673 (2006).

54. Es-Safı, N.-E., Le Guernevé, Christine, Fulcrand, H., Cheynier, V. & Moutounet, M. New

Polyphenolic Compounds with Xanthylium Skeletons Formed through Reaction between

(+)-Catechin and Glyoxylic Acid. J. Agric. Food Chem. 47, 5211–5217 (1999).

55. Elias, R. J. & Waterhouse, A. L. Controlling the Fenton Reaction in Wine. J. Agric. Food

Chem. 58, 1699–1707 (2010).

56. Gislason, N. E., Currie, B. L. & Waterhouse, A. L. Novel Antioxidant Reactions of

Cinnamates in Wine. J. Agric. Food Chem. 59, 6221–6226 (2011).

57. Kreitman, G. Y., Laurie, V. F. & Elias, R. J. Investigation of Ethyl Radical Quenching by

Phenolics and Thiols in Model Wine. J. Agric. Food Chem. 61, 685–692 (2013).

125 126

58. Danilewicz, J. C. Reaction of Oxygen and Sulfite in Wine. Am. J. Enol. Vitic.

ajev.2015.15069 (2015). doi:10.5344/ajev.2015.15069

59. Hoffmann, M. R. & Edwards, J. O. Kinetics of the oxidation of sulfite by hydrogen

peroxide in acidic solution. J. Phys. Chem. 79, 2096–2098 (1975).

60. Boulton, R. B., Singleton, V. L., Bisson, L. F. & Kunkee, R. E. in Principles and Practices

of Winemaking 448–473 (Springer US, 1999).

61. De Azevedo, L. C. et al. Evaluation of the formation and stability of hydroxyalkylsulfonic

acids in wines. J. Agric. Food Chem. 55, 8670–8680 (2007).

62. Andersen, M. L., Outtrup, H. & Skibsted, L. H. Potential Antioxidants in Beer Assessed by

ESR Spin Trapping. J Agric Food Chem 48, 3106–3111 (2000).

63. Vinson, J. a, Su, X., Zubik, L. & Bose, P. Phenol antioxidant quantity and quality in foods:

fruits. J. Agric. Food Chem. 49, 5315–21 (2001).

64. Simonetti, P., Pietta, P. & Testolin, G. Polyphenol Content and Total Antioxidant Potential

of Selected Italian Wines. J. Agric. Food Chem. 45, 1152–1155 (1997).

65. Flavonoids: Chemistry, Biochemistry and Applications. (Taylor & Francis, 2006).

66. Fulcrand, H., Dueñas, M., Salas, E. & Cheynier, V. Phenolic Reactions during Winemaking

and Aging. Am. J. Enol. Vitic. 57, 289–297 (2006).

67. Mazza, G. Anthocyanins in grapes and grape products. Crit. Rev. Food Sci. Nutr. 35, 341–

371 (1995).

68. Heredia, F. J., Francia-aricha, E. M., Rivas-gonzalo, J. C., Vicario, I. M. & Santos-buelga,

C. Chromatic characterization of anthocyanins from red grapes - I . pH effect. Food Chem.

63, 491–498 (1998).

69. Bridle, P. & Timberlake, C. F. Anthocyanins as natural food colours—selected aspects.

Food Chem. 58, 103–109 (1997).

126 127

70. Pascual-Teresa, S. de & Sanchez-Ballesta, M. T. Anthocyanins: from plant to health.

Phytochem. Rev. 7, 281–299 (2007).

71. He, F. et al. Anthocyanins and Their Variation in Red Wines I. Monomeric Anthocyanins

and Their Color Expression. Molecules 17, 1571–1601 (2012).

72. He, F., He, J.-J., Pan, Q.-H. & Duan, C.-Q. Mass-spectrometry evidence confirming the

presence of pelargonidin-3-O-glucoside in the berry skins of Cabernet Sauvignon and Pinot

Noir (Vitis vinifera L.). Aust. J. Grape Wine Res. 16, 464–468 (2010).

73. Alcalde-Eon, C., Escribano-Bailón, M. T., Santos-Buelga, C. & Rivas-Gonzalo, J. C.

Changes in the detailed pigment composition of red wine during maturity and ageing. Anal.

Chim. Acta 563, 238–254 (2006).

74. He, F. et al. Anthocyanins and their variation in red wines II. Anthocyanin derived pigments

and their color evolution. Molecules 17, 1483–1519 (2012).

75. Sacchi, K. L., Bisson, L. F. & Adams, D. O. A Review of the Effect of Winemaking

Techniques on Phenolic Extraction in Red Wines. Am. J. Enol. Vitic. 56, 197–206 (2005).

76. Busse-Valverde, N., Gómez-Plaza, E., López-Roca, J. M., Gil-Muñoz, R. & Bautista-Ortín,

A. B. The Extraction of Anthocyanins and Proanthocyanidins from Grapes to Wine during

Fermentative Maceration Is Affected by the Enological Technique. J. Agric. Food Chem.

59, 5450–5455 (2011).

77. Cheynier, V., Prieur, C., Guyot, S., Rigaud, J. & Moutounet, M. in Wine (ed. Watkins, T.

R.) 661, 81–93 (American Chemical Society, 1997).

78. Somers, T. C. The polymeric nature of wine pigments. Phytochemistry 10, 2175–2186

(1971).

79. Ribereau-Gayon, P., Sudraud, P., Milhe, J. C. & Canbas, A. Recherches technologiques sur

les composés phénoliques des vin rouges. Conn Vigne Vin 133–144 (1970).

127 128

80. Nagel, C. W. & Wulf, L. W. Changes in the Anthocyanins, Flavonoids and

Hydroxycinnamic Acid Esters during Fermentation and Aging of Merlot and Cabernet

Sauvignon. Am. J. Enol. Vitic. 30, 111–116 (1979).

81. Salas, E., Fulcrand, H., Meudec, E. & Cheynier, V. Reactions of Anthocyanins and Tannins

in Model Solutions. J. Agric. Food Chem. 51, 7951–7961 (2003).

82. Vidal, S. et al. Taste and mouth-feel properties of different types of tannin-like

polyphenolic compounds and anthocyanins in wine. Anal. Chim. Acta 513, 57–65 (2004).

83. Iacobucci, G. A. & Sweeny, J. G. The chemistry of anthocyanins, anthocyanidins and

related flavylium salts. Tetrahedron 39, 3005–3038 (1983).

84. Timberlake, C. F. & Bridle, P. Flavylium salts, anthocyanidins and anthocyanins. I.—

Structural transformations in acid solutions. J. Sci. Food Agric. 18, 473–478 (1967).

85. Brouillard, R. & Dubois, J.-E. Mechanism of the structural transformations of anthocyanins

in acidic media. J. Am. Chem. Soc. 99, 1359–1364 (1977).

86. Cheynier, V. & Dueñas-Paton, M. Structure and properties of wine pigments and tannins.

Am. J. Enol. Vitic. 57, 298–305 (2006).

87. Jurd, L. Anthocyanins and related compounds. I. Structural transformations of flavylium

salts in acidic solutions. J. Org. Chem. 28, 987–991 (1963).

88. Boulton, R. The Copigmentation of Anthocyanins and Its Role in the Color of Red Wine: A

Critical Review. Am. J. Enol. Vitic. 52, 67–87 (2001).

89. Boulton, R. B., Singleton, V. L., Bisson, L. F. & Kunkee, R. E. Principles and Practices of

Winemaking. (Springer Science & Business Media, 2013).

90. Castañeda-Ovando, A., Pacheco-Hernández, M. de L., Páez-Hernández, M. E., Rodríguez,

J. A. & Galán-Vidal, C. A. Chemical studies of anthocyanins: A review. Food Chem. 113,

859–871 (2009).

128 129

91. Timberlake, C. F. & Bridle, P. The effect of processing and other factors on the colour

characteristics of some red wines. Vitis 15, 37–49 (1976).

92. Timberlake, C. F. & Bridle, P. Interactions between anthocyanins, phenolic compounds, and

acetaldehyde and their significance in red wines. Am. J. Enol. Vitic. 27, 97–105 (1976).

93. Fulcrand, H., Doco, T., Es-Safi, N.-E., Cheynier, V. & Moutounet, M. Study of the

acetaldehyde induced polymerisation of flavan-3-ols by liquid chromatography-ion spray

mass spectrometry. J. Chromatogr. A 752, 85–91 (1996).

94. Salas, E. et al. Characterization of pigments from different high speed countercurrent

chromatography wine fractions. J. Agric. Food Chem. 53, 4536–46 (2005).

95. Atanasova, V., Fulcrand, H., Le Guernevé, C., Cheynier, V. & Moutounet, M. Structure of a

new dimeric acetaldehyde malvidin 3-glucoside condensation product. Tetrahedron Lett.

43, 6151–6153 (2002).

96. Francia-Aricha, E., Guerra, M., Rivas-Gonzalo, J. C. & Santos-Buelga, C. New anthocyanin

pigments formed after condensation with flavanols. J. Agric. Food Chem. 45, 2262–2266

(1997).

97. Monagas, M., Núñez, V., Bartolomé, B. & Gómez-Cordovés, C. Anthocyanin-derived

Pigments in Graciano, Tempranillo, and Cabernet Sauvignon Wines Produced in Spain. Am.

J. Enol. Vitic. 54, 163–169 (2003).

98. Monagas, M., Bartolomé, B. & Gómez-Cordovés, C. Updated knowledge about the

presence of phenolic compounds in wine. Crit. Rev. Food Sci. Nutr. 45, 85–118 (2005).

99. Escribano-Bailón, T., Alvarez-García, M., Rivas-Gonzalo, J. C., Heredia, F. J. & Santos-

Buelga, C. Color and stability of pigments derived from the acetaldehyde-mediated

condensation between malvidin 3-O-glucoside and (+)-catechin. J. Agric. Food Chem. 49,

1213–7 (2001).

129 130

100. Cheynier, V., Es-Safi, N. E. & Fulcrand, H. Structure and colour properties of anthcoyanins

and related pigments. in 23–25 (M.I.M Mosquera et al. (Eds.), 1999).

101. Bakker, J. & Timberlake, C. F. Isolation, Identification, and Characterization of New Color-

Stable Anthocyanins Occurring in Some Red Wines. J. Agric. Food Chem. 45, 35–43

(1997).

102. Es-Safi, N. E., Cheynier, V. & Moutounet, M. Role of aldehydic derivatives in the

condensation of phenolic compounds with emphasis on the sensorial properties of fruit-

derived foods. J. Agric. Food Chem. 50, 5571–5585 (2002).

103. Baranowski, E. S. & Nagel, C. W. Kinetics of malvidin-3-glucoside condensation in wine

model systems. J Food Sci 419–429 (1983).

104. Roggero, J. P., Coen, S., Archier, P. & Rocheville-Divorne, C. Etude par C.L.H.P. de la

reaction glucoside de malvidine-acetaldehyde-compose phenolique. Connaiss Vigne Vin 21,

163–168 (1987).

105. Bakker, J., Picinelli, A. & Bridle, P. Model wine solutions: colour and composition changes

during ageing. Vitis 111–118 (1993).

106. Rivas-Gonzalo, J. C., Bravo-Haro, S. & Santos-Buelga, C. Detection of compounds formed

through the reactin of malvidin-3-monoglucoside and catechin in the presence of

acetaldehyde. J Agric Food Chem 1444–1449 (1995).

107. Dallas, C., Ricardo-Da-Silva, J. M. & Laureano, O. Products formed in model wine

solutions involving anthocyanins, procyanidin B2, and acetaldehyde. J. Agric. Food Chem.

44, 2402–2407 (1996).

108. Dallas, C., Ricardo-da-Silva, J. M. & Laureano, O. Interactions of Oligomeric Procyanidins

in Model Wine Solutions Containing Malvidin-3-Glucoside and Acetaldehyde. J. Sci. Food

Agric. 70, 493–500 (1996).

130 131

109. Es-Safi, N. E. et al. Kinetic studies of acetaldehyde-induced condensation of flavan-3-ols

and malvidin-3-glucoside model solutions systems. in Polyphenols Communications 96

(Vercauteren, J., Cheze, C., Dumon, M.C., Weber, J.F., (Eds.), 1996).

110. Escribano-Bailón, T., Dangles, O. & Brouillard, R. Coupling reactions between flavylium

ions and catechin. Phytochemistry 1583–1592 (1996).

111. Benabdeljalil, C. et al. Evidence of new pigments resulting from reaction between

anthocyanins and yeast metabolites [hemisynthesis, fermentation products]. Sci. Aliments

Fr. (2000).

112. Sarni-Manchado, P., Fulcrand, H., Souquet, J.-M., Cheynier, V. & Moutounet, M. Stability

and Color of Unreported Wine Anthocyanin-derived Pigments. J. Food Sci. 61, 938–941

(1996).

113. Mateus, N., Oliveira, J., Santos-Buelga, C., Silva, A. M. S. & de Freitas, V. NMR structure

characterization of a new vinylpyranoanthocyanin–catechin pigment (a portisin).

Tetrahedron Lett. 45, 3455–3457 (2004).

114. Mateus, N. et al. A new vinylpyranoanthocyanin pigment occurring in aged red wine. Food

Chem. 97, 689–695 (2006).

115. Mateus, N. et al. Isolation and Structural Characterization of New Acylated

Anthocyanin−Vinyl−Flavanol Pigments Occurring in Aging Red Wines. J. Agric. Food

Chem. 51, 277–282 (2003).

116. Mateus, N., Silva, A. M. S., Rivas-Gonzalo, J. C., Santos-Buelga, C. & de Freitas, V. A

New Class of Blue Anthocyanin-Derived Pigments Isolated from Red Wines. J. Agric.

Food Chem. 51, 1919–1923 (2003).

131 132

117. Remy, S., Fulcrand, H., Labarbe, B., Cheynier, V. & Moutounet, M. First confirmation in

red wine of products resulting from direct anthocyanin – tannin reactions. J. Sci. Food

Agric. 80, 745–751 (2000).

118. Brouillard, R., Iacobucci, G. A. & Sweeny, J. G. Chemistry of anthocyanin pigments. 9.

UV-visible spectrophotometric determination of the acidity constants of apigeninidin and

three related 3-deoxyflavylium salts. J. Am. Chem. Soc. 104, 7585–7590 (1982).

119. Atanasova, V., Fulcrand, H., Cheynier, V. & Moutounet, M. Effect of oxygenation on

polyphenol changes occurring in the course of wine-making. Anal. Chim. Acta 458, 15–27

(2002).

120. Santos-buelga, C. & Scalbert, A. Proanthocyanidins and tannin-like compounds – nature ,

occurrence , dietary intake and effects on nutrition and health. J. Sci. Food Agric. 1094–

1117 (2000).

121. Ozdal, T., Capanoglu, E. & Altay, F. A review on protein–phenolic interactions and

associated changes. Food Res. Int. 51, 954–970 (2013).

122. Cos, P. et al. Proanthocyanidins in health care: current and new trends. Curr. Med. Chem.

11, 1345–59 (2004).

123. Quideau, S., Deffieux, D., Douat-Casassus, C. & Pouységu, L. Plant polyphenols: chemical

properties, biological activities, and synthesis. Angew. Chem. Int. Ed Engl. 50, 586–621

(2011).

124. Prieur, C., Rigaud, J., Cheynier, V. & Moutounet, M. Oligomeric and Polymeric

Procyanidins from Grape Seeds. Phytochemistry 34, 781–784 (1994).

125. Souquet, J.-M., Cheynier, V., Brossaud, F. & Moutounet, M. Polymeric proanthocyanidins

from grape skins. Phytochemistry 43, 509–512 (1996).

132 133

126. Vivas, N. & Glories, Y. Role of Oak Wood Ellagitannins in the Oxidation Process of Red

Wines During Aging. Am. J. Enol. Vitic. 47, 103–107 (1996).

127. Harbertson, J. F. et al. Variability of Tannin Concentration in Red Wines. Am. J. Enol.

Vitic. 59, 210–214 (2008).

128. Mercurio, M. D. & Smith, P. A. Tannin quantification in red grapes and wine: comparison

of polysaccharide- and protein-based tannin precipitation techniques and their ability to

model wine astringency. J. Agric. Food Chem. 56, 5528–37 (2008).

129. Bennick, a. Interaction of Plant Polyphenols With Salivary Proteins. Crit. Rev. Oral Biol.

Med. 13, 184–196 (2002).

130. Gawel, R. Red wine astringency: a review. Aust. J. Grape Wine Res. 4, 74–95 (1998).

131. Varela, P. & Gámbaro, A. Sensory descriptive analysis of Uruguayan Tannat wine:

Correlation to quality assessment. J. Sens. Stud. 21, 203–217 (2005).

132. Bajec, M. R. & Pickering, G. J. Astringency: mechanisms and perception. Crit. Rev. Food

Sci. Nutr. 48, 858–75 (2008).

133. Jöbstl, E., O’Connell, J., Fairclough, J. P. a & Williamson, M. P. Molecular model for

astringency produced by polyphenol/protein interactions. Biomacromolecules 5, 942–9

(2004).

134. Mcrae, J. M. & Kennedy, J. A. Wine and Grape Tannin Interactions with Salivary Proteins

and Their Impact on Astringency: A Review of Current Research. Molecules 16, 2348–

2364 (2011).

135. Hagerman, A. E. & Butler, L. G. Condensed tannin purification and characterization of

tannin-associated proteins. J. Agric. Food Chem. 28, 947–952 (1980).

133 134

136. Kauffman, D. L., Keller, P. J., Bennick, A. & Blum, M. Alignment of Amino Acid and

DNA Sequences of Human Proline-rich Proteins. Crit. Rev. Oral Biol. Med. 4, 287–292

(1993).

137. Boze, H. et al. Proline-rich salivary proteins have extended conformations. Biophys. J. 99,

656–65 (2010).

138. Cala, O. et al. The colloidal state of tannins impacts the nature of their interaction with

proteins: the case of salivary proline-rich protein/procyanidins binding. Langmuir ACS J.

Surf. Colloids 28, 17410–8 (2012).

139. Noble, A. Why do wines taste bitter and feel astringent? in Wine Flavor 156–165 (A.L.

Waterhouse and S.E. Ebeler (Eds.), 1998).

140. Noble, A. C. Bitterness in wine. Physiol. Behav. 56, 1251–1255 (1994).

141. Harbertson, J. F., Kilmister, R. L., Kelm, M. a & Downey, M. O. Impact of condensed

tannin size as individual and mixed polymers on bovine serum albumin precipitation. Food

Chem. 160, 16–21 (2014).

142. Zeller, W. E. et al. Protein Precipitation Behavior of Condensed Tannins from Lotus

pedunculatus and Trifolium repens with Different Mean Degrees of Polymerization. J.

Agric. Food Chem. 63, 1160–1168 (2015).

143. Sarni-Manchado, P., Cheynier, V. & Moutounet, M. Interactions of grape seed tannins with

salivary proteins. J. Agric. Food Chem. 47, 42–7 (1999).

144. Sarni-Manchado, P. & Cheynier, V. Study of non-covalent complexation between catechin

derivatives and peptides by electrospray ionization mass spectrometry. J. Mass Spectrom.

JMS 37, 609–16 (2002).

145. Vidal, S. et al. The mouth-feel properties of grape and apple proanthocyanidins in a wine-

like medium. J. Sci. Food Agric. 83, 564–573 (2003).

134 135

146. Cala, O. et al. NMR and molecular modeling of wine tannins binding to saliva proteins:

revisiting astringency from molecular and colloidal prospects. FASEB J. Off. Publ. Fed.

Am. Soc. Exp. Biol. 24, 4281–90 (2010).

147. Poncet-Legrand, C. et al. Poly(l-proline) interactions with flavan-3-ols units: Influence of

the molecular structure and the polyphenol/protein ratio. Food Hydrocoll. 20, 687–697

(2006).

148. Cala, O. et al. Towards a Molecular Interpretation of Astringency: Synthesis, 3D Structure,

Colloidal State, and Human Saliva Protein Recognition of Procyanidins. Planta Med. 77,

1116–1122 (2011).

149. Vidal, S. et al. The mouth-feel properties of polysaccharides and anthocyanins in a wine

like medium. Food Chem. 85, 519–525 (2004).

150. Riou, V., Vernhet, A., Doco, T. & Moutounet, M. Aggregation of grape seed tannins in

model wine—effect of wine polysaccharides. Food Hydrocoll. 16, 17–23 (2002).

151. Fontoin, H., Saucier, C., Teissedre, P.-L. & Glories, Y. Effect of pH, ethanol and acidity on

astringency and bitterness of grape seed tannin oligomers in model wine solution. Food

Qual. Prefer. 19, 286–291 (2008).

152. Gawel, R., Iland, P. G. & Francis, I. L. Characterizing the astringency of red wine: a case

study. Food Qual. Prefer. 12, 83–94 (2001).

153. Gambuti, A., Rinaldi, A., Ugliano, M. & Moio, L. Evolution of phenolic compounds and

astringency during aging of red wine: effect of oxygen exposure before and after bottling. J.

Agric. Food Chem. 61, 1618–27 (2013).

154. Rinaldi, A., Gambuti, A. & Moio, L. Application of the SPI (Saliva Precipitation Index) to

the evaluation of red wine astringency. Food Chem. 135, 2498–504 (2012).

135 136

155. Harbertson, J. F., Kennedy, J. A. & Adams, D. O. Tannin in Skins and Seeds of Cabernet

Sauvignon , Syrah , and Pinot noir Berries during Ripening. Am. J. Enol. Vitic. 1, 54–59

(2002).

156. Heredia, T. M., Adams, D. O., Fields, K. C., Held, P. G. & Harbertson, J. F. Evaluation of a

Comprehensive Red Wine Phenolics Assay Using a Microplate Reader. Am. J. Enol. Vitic.

4, 497–502 (2006).

157. Tanaka, T., Takahashi, R., Kouno, I. & Nonaka, G. Chemical evidence for the de-

astringency (insolubilization of tannins) of persimmon fruit. J. Chem. Soc. [Perkin 1] 3013–

3022 (1994).

158. Saucier, C., Little, D. & Glories, Y. First evidence of acetaldehyde-flavanol condensation

products in red wine. Am. J. Enol. Vitic. 48, 370–373 (1997).

159. Saucier, C., Guerra, C., Pianet, I., Laguerre, M. & Glories, Y. (+)-Catechin - Acetaldehyde

Condensation Products in Relation to Wine-Ageing. Phytochemistry 46, 229–234 (1997).

160. Saucier, C., Bourgeois, G. & Vitry, C. Characterization of (+)-Catechin - Acetaldehyde

Polymers: A Model for Colloidal State of Wine Polyphenols. J. Agric. … 1045–1049

(1997).

161. Es-Safi, N.-E. E., Fulcrand, H., Cheynier, V. & Moutounet, M. Competition between (+)-

catechin and (-)-epicatechin in acetaldehyde-induced polymerization of flavanols. J. Agric.

Food Chem. 47, 2088–95 (1999).

162. Drinkine, J., Lopes, P., Kennedy, J. A., Teissedre, P.-L. & Saucier, C. Analysis of

Ethylidene-Bridged Flavan-3-ols in Wine. J. Agric. Food Chem. 55, 1109–1116 (2007).

163. Drinkine, J., Lopes, P., Kennedy, J. a, Teissedre, P.-L. & Saucier, C. Ethylidene-bridged

Flavan-3-ols in red wine and correlation with wine age. J. Agric. Food Chem. 55, 6292–9

(2007).

136 137

164. Vidal, S., Cartalade, D., Souquet, J., Fulcrand, H. & Cheynier, V. Changes in

proanthocyanidin chain length in winelike model solutions. J. Agric. Food Chem. 50, 2261–

2266 (2002).

165. Haslam, E. In vino veritas: oligomeric procyanidins ageing of red wines. Phytochemistry

19, 2577–2582 (1980).

166. Kontoudakis, N. et al. Influence of wine pH on changes in color and polyphenol

composition induced by micro-oxygenation. J. Agric. Food Chem. 59, 1974–84 (2011).

167. Tao, J., Dykes, S. I. & Kilmartin, P. A. Effect of SO2 concentration on polyphenol

development during red wine micro-oxygenation. J. Agric. Food Chem. 55, 6104–6109

(2007).

168. Ribereau-Gayon, P., Pontallier, P. & Glories, Y. Some Interpretations of Colour Changes in

Young Red Wines During Their Conservation. J Sci Food Agric 34, 505–516 (1983).

169. Monagas, M., Bartolome, B. & Gomez-Cordoves, C. Evolution of polyphenols in red wines

from Vitis vinifera L. during aging in the bottle. Eur. Food Res. Technol. 220, 331–340

(2005).

170. Cejudo-Bastante, M. J., Hermosín-Gutiérrez, I. & Pérez-Coello, M. S. Micro-oxygenation

and oak chip treatments of red wines: Effects on colour-related phenolics, volatile

composition and sensory characteristics. Part II: Merlot wines. Food Chem. 124, 738–748

(2011).

171. Sánchez-Iglesias, M., González-Sanjosé, M. L., Pérez-Magariño, S., Ortega-Heras, M. &

González-Huerta, C. Effect of Micro-Oxygenation and Wood Type on the Phenolic

Composition and Color of an Aged Red Wine. J. Agric. Food Chem. 57, 11498–11509

(2009).

137 138

172. Nguyen, D.-D., Nicolau, L., Dykes, S. I. & Kilmartin, P. A. Influence of Microoxygenation

on Reductive Sulfur Off-Odors and Color Development in a Cabernet Sauvignon Wine. Am.

J. Enol. Vitic. 61, 457–464 (2010).

173. Pérez-Magariño, S., Sánchez-Iglesias, M., Ortega-Heras, M., González-Huerta, C. &

González-Sanjosé, M. L. Colour stabilization of red wines by microoxygenation treatment

before malolactic fermentation. Food Chem. 101, 881–893 (2007).

174. Cejudo-Bastante, M. J., Pérez-Coello, M. S. & Hermosín-Gutiérrez, I. Effect of wine micro-

oxygenation treatment and storage period on colour-related phenolics, volatile composition

and sensory characteristics. LWT - Food Sci. Technol. 44, 866–874 (2011).

175. del Carmen Llaudy, M. et al. Influence of Micro-Oxygenation Treatment before Oak Aging

on Phenolic Compounds Composition, Astringency, and Color of Red Wine. J. Agric. Food

Chem. 54, 4246–4252 (2006).

176. Kennedy, J. A. & Jones, G. P. Analysis of proanthocyanidin cleavage products following

acid-catalysis in the presence of excess phloroglucinol. J. Agric. Food Chem. 49, 1740–

1746 (2001).

177. Kennedy, J. a & Taylor, A. W. Analysis of proanthocyanidins by high-performance gel

permeation chromatography. J. Chromatogr. A 995, 99–107 (2003).

178. McRae, J. M., Schulkin, A., Kassara, S., Holt, H. E. & Smith, P. A. Sensory Properties of

Wine Tannin Fractions: Implications for In-Mouth Sensory Properties. J. Agric. Food

Chem. 61, 719–727 (2013).

179. Aleixandre-Tudo, J. l. et al. Effect of acetaldehyde addition on the phenolic substances and

volatile compounds of red Tempranillo wines. Aust. J. Grape Wine Res. n/a–n/a (2016).

doi:10.1111/ajgw.12203

138 139

180. Aleixandre-Tudó, J. L. et al. Impact of Caffeic Acid Addition on Phenolic Composition of

Tempranillo Wines from Different Winemaking Techniques. J. Agric. Food Chem. 61,

11900–11912 (2013).

181. Research and Markets. Wine: Global Industry Almanac. 1–32 (2015).

182. Bruwer, J., Saliba, A. & Miller, B. Consumer behaviour and sensory preference differences:

implications for wine product marketing. J. Consum. Mark. 28, 5–18 (2011).

183. Gawel, R., Iland, P. G. & Francis, I. L. Characterizing the astringency of red wine: a case

study. Food Qual. Prefer. 12, 83–94 (2001).

184. Varela, P. & Gámbaro, A. Sensory descriptive analysis of Uruguayan Tannat wine:

Correlation to quality assessment. J. Sens. Stud. 21, 203–217 (2005).

185. Gawel, R. Red wine astringency: a review. Aust. J. Grape Wine Res. 4, 74–95 (1998).

186. Cheynier, V. & Dueñas-Paton, M. Structure and properties of wine pigments and tannins.

Am. J. Enol. Vitic. 57, 298–305 (2006).

187. Vidal, S. et al. Taste and mouth-feel properties of different types of tannin-like

polyphenolic compounds and anthocyanins in wine. Anal. Chim. Acta 513, 57–65 (2004).

188. Monagas, M., Bartolome, B. & Gomez-Cordoves, C. Evolution of polyphenols in red wines

from Vitis vinifera L. during aging in the bottle. Eur. Food Res. Technol. 220, 331–340

(2005).

189. Bajec, M. R. & Pickering, G. J. Astringency: mechanisms and perception. Crit. Rev. Food

Sci. Nutr. 48, 858–75 (2008).

190. Jöbstl, E., O’Connell, J., Fairclough, J. P. a & Williamson, M. P. Molecular model for

astringency produced by polyphenol/protein interactions. Biomacromolecules 5, 942–9

(2004).

139 140

191. Mcrae, J. M. & Kennedy, J. A. Wine and Grape Tannin Interactions with Salivary Proteins

and Their Impact on Astringency: A Review of Current Research. Molecules 16, 2348–

2364 (2011).

192. Atanasova, V., Fulcrand, H., Cheynier, V. & Moutounet, M. Effect of oxygenation on

polyphenol changes occurring in the course of wine-making. Anal. Chim. Acta 458, 15–27

(2002).

193. Es-Safi, N. E., Fulcrand, H., Cheynier, V. & Moutounet, M. Studies on the acetaldehyde-

induced condensation of (-)-epicatechin and malvidin 3-O-glucoside in a model solution

system. J. Agric. Food Chem. 47, 2096–102 (1999).

194. Saucier, C., Guerra, C., Pianet, I., Laguerre, M. & Glories, Y. (+)-Catechin - Acetaldehyde

Condensation Products in Relation to Wine-Ageing. Phytochemistry 46, 229–234 (1997).

195. Alcalde-Eon, C., Escribano-Bailón, M. T., Santos-Buelga, C. & Rivas-Gonzalo, J. C.

Changes in the detailed pigment composition of red wine during maturity and ageing. Anal.

Chim. Acta 563, 238–254 (2006).

196. Saucier, C., Little, D. & Glories, Y. First evidence of acetaldehyde-flavanol condensation

products in red wine. Am. J. Enol. Vitic. 48, 370–373 (1997).

197. Danilewicz, J. C. Review of Reaction Mechanisms of Oxygen and Proposed Intermediate

Reduction Products in Wine : Central Role of Iron and Copper. Am. J. Enol. Vitic. 54, 73–

85 (2003).

198. Waterhouse, A. & Laurie, V. Oxidation of wine phenolics: A critical evaluation and

hypotheses. Am. J. Enol. Vitic. 3, 306–313 (2006).

199. Carlton, W. K., Gump, B., Fugelsang, K. & Hasson, A. S. Monitoring Acetaldehyde

Concentrations during Micro-oxygenation of Red Wine by Headspace Solid-Phase

Microextraction with On-Fiber Derivatization. J. Agric. Food Chem. 55, 5620–5625 (2007).

140 141

200. Somers, T. C. & Evans, M. Spectral Evaluation of Young Red Wines: Anthocyanin

Equilibria, Total Phenolics, Free and Molecular SO, ‘ Chemical Age’. J. Sci. Food Agric.

28, 279–287 (1977).

201. Drinkine, J., Lopes, P., Kennedy, J. a, Teissedre, P.-L. & Saucier, C. Ethylidene-bridged

Flavan-3-ols in red wine and correlation with wine age. J. Agric. Food Chem. 55, 6292–9

(2007).

202. Somers, T. C. & Evans, M. E. Wine quality: Correlations with colour density and

anthocyanin equilibria in a group of young red wines. J. Sci. Food Agric. 25, 1369–1379

(1974).

203. Kontoudakis, N. et al. Influence of wine pH on changes in color and polyphenol

composition induced by micro-oxygenation. J. Agric. Food Chem. 59, 1974–84 (2011).

204. González-del Pozo, A., Arozarena, Í., Noriega, M.-J., Navarro, M. & Casp, A. Short- and

long-term effects of micro-oxygenation treatments on the colour and phenolic composition

of a Cabernet Sauvignon wine aged in barrels and/or bottles. Eur. Food Res. Technol. 231,

589–601 (2010).

205. Gambuti, A., Rinaldi, A., Ugliano, M. & Moio, L. Evolution of phenolic compounds and

astringency during aging of red wine: effect of oxygen exposure before and after bottling. J.

Agric. Food Chem. 61, 1618–27 (2013).

206. Gómez-Plaza, E. & Cano-López, M. A review on micro-oxygenation of red wines: Claims,

benefits and the underlying chemistry. Food Chem. 125, 1131–1140 (2011).

207. Sampaio, T. L., Kennedy, J. A. & Vasconcelos, M. C. Use of Microscale Fermentations in

Grape and Wine Research. Am. J. Enol. Vitic. 58, 534–539 (2007).

208. Mercurio, M. D., Dambergs, R. G., Herderich, M. J. & Smith, P. A. High throughput

analysis of red wine and grape phenolics-adaptation and validation of methyl cellulose

141 142

precipitable tannin assay and modified Somers color assay to a rapid 96 well plate format. J.

Agric. Food Chem. 55, 4651–7 (2007).

209. Waterhouse, A. L. in Current protocols in food analytical chemistry (ed. Wrolstad, R.)

I1.1–I1.8 (Wiley, 2001).

210. Heredia, T. M., Adams, D. O., Fields, K. C., Held, P. G. & Harbertson, J. F. Evaluation of a

Comprehensive Red Wine Phenolics Assay Using a Microplate Reader. Am. J. Enol. Vitic.

4, 497–502 (2006).

211. Rinaldi, A., Gambuti, A. & Moio, L. Application of the SPI (Saliva Precipitation Index) to

the evaluation of red wine astringency. Food Chem. 135, 2498–504 (2012).

212. Etievant, P. X. in Volatile Compounds in Food and Beverages (ed. Maarse, H.) 483–546

(1991).

213. Mercurio, M. D. & Smith, P. A. Tannin quantification in red grapes and wine: comparison

of polysaccharide- and protein-based tannin precipitation techniques and their ability to

model wine astringency. J. Agric. Food Chem. 56, 5528–37 (2008).

214. Sheridan, M. K. & Elias, R. J. Exogenous acetaldehyde as a tool for modulating wine color

and astringency during fermentation. Food Chem. 177, 17–22 (2015).

215. Drinkine, J., Glories, Y. & Saucier, C. (+)-Catechin-aldehyde condensations: Competition

between acetaldehyde and glyoxylic acid. J. Agric. Food Chem. 53, 7552–7558 (2005).

216. Sonni, F. et al. Impact of glutathione on the formation of methylmethine-and

carboxymethine-bridged (+)-catechin dimers in a model wine system. J. Agric. Food Chem.

59, 7410–7418 (2011).

217. Hornsey, I. S. The Chemistry and Biology of Winemaking. (Royal Society of Chemistry,

2007).

142 143

218. Mercurio, M. D., Dambergs, R. G., Herderich, M. J. & Smith, P. A. High throughput

analysis of red wine and grape phenolics-adaptation and validation of methyl cellulose

precipitable tannin assay and modified Somers color assay to a rapid 96 well plate format. J.

Agric. Food Chem. 55, 4651–7 (2007).

219. Morata, A., Gómez-Cordovés, M. C., Calderón, F. & Suárez, J. A. Effects of pH,

temperature and SO2 on the formation of pyranoanthocyanins during red wine fermentation

with two species of Saccharomyces. Int. J. Food Microbiol. 106, 123–129 (2006).

220. Garde-Cerdán, T. & Ancín-Azpilicueta, C. Effect of SO2 on the formation and evolution of

volatile compounds in wines. Food Control 18, 1501–1506 (2007).

221. Timberlake, C. F. & Bridle, P. Anthocyanins: colour augmentation with catechin and

acetaldehyde. J. Sci. Food Agric. 28, 539–544 (1977).

222. Vivas, N. et al. Differentiation of proanthocyanidin tannins from seeds, skins and stems of

grapes (Vitis vinifera) and heartwood of Quebracho (Schinopsis balansae) by matrix-

assisted laser desorption/ionization time-of-flight mass spectrometry and

thioacidolysis/liquid c. Anal. Chim. Acta 513, 247–256 (2004).

223. Mane, C. et al. Assessment of the Molecular Weight Distribution of Tannin Fractions

through MALDI-TOF MS Analysis of Protein − Tannin Complexes. Anal Chem 79, 2239–

2248 (2007).

224. Oliveira, J., Alhinho da Silva, M., Teixeira, N., De Freitas, V. & Salas, E. Screening of

Anthocyanins and Anthocyanin-Derived Pigments in Red Wine Grape Pomace Using LC-

DAD/MS and MALDI-TOF Techniques. J. Agric. Food Chem. 150513092027007–

150513092027007 (2015). doi:10.1021/acs.jafc.5b00256

225. Takano, T., Murakami, T., Kamitakahara, H. & Nakatsubo, F. Mechanism of formaldehyde

adsorption of (+)-catechin. J. Wood Sci. 54, 329–331 (2008).

143 144

226. Schwarz, M., Wabnitz, T. C. & Winterhalter, P. Pathway Leading to the Formation of

Anthocyanin−Vinylphenol Adducts and Related Pigments in Red Wines. J. Agric. Food

Chem. 51, 3682–3687 (2003).

227. Cruz, L. et al. Role of vinylcatechin in the formation of pyranomalvidin-3-glucoside-(+)-

catechin. J. Agric. Food Chem. 56, 10980–10987 (2008).

228. Pechamat, L., Zeng, L., Jourdes, M., Ghidossi, R. & Teissedre, P.-L. Occurrence and

formation kinetics of pyranomalvidin-procyanidin dimer pigment in merlot red wine:

impact of acidity and oxygen concentrations. J. Agric. Food Chem. 62, 1701–5 (2014).

229. He, J., Carvalho, A. R. F., Mateus, N. & De Freitas, V. Spectral features and stability of

oligomeric pyranoanthocyanin-flavanol pigments isolated from red wines. J. Agric. Food

Chem. 58, 9249–58 (2010).

230. Pissarra, J. et al. Formation of new anthocyanin-alkyl/aryl-flavanol pigments in model

solutions. Anal. Chim. Acta 513, 215–221 (2004).

231. Nykänen, L. Formation and Occurrence of Flavor Compounds in Wine and Distilled

Alcoholic Beverages. Am. J. Enol. Vitic. 37, 84–96 (1986).

232. Frivik, S. K. & Ebeler, S. E. Influence of sulfur dioxide on the formation of aldehydes in

white wine. Am. J. Enol. Vitic. 54, 31–38 (2003).

233. Vanderlinde, R. Etude de certains caracteres des eaux-de-vie: Role des composes

carbonyles. (Univ. de Bordeaux II, 1995).

234. Rogerson, F. s. s. & De Freitas, V. a. p. Fortification Spirit, a Contributor to the Aroma

Complexity of Port. J. Food Sci. 67, 1564–1569 (2002).

235. Elias, R. J., Laurie, V. F., Ebeler, S. E., Wong, J. W. & Waterhouse, A. L. Analysis of

selected carbonyl oxidation products in wine by liquid chromatography with diode array

detection. Anal. Chim. Acta 626, 104–110 (2008).

144 145

236. Nagata, Y. Measurement of odor threshold by triangle odor bag method. Odor Meas. Rev.

Off. Odor Noise Vib. Environ. Manag. Bur. Minist. Environ. Gov. Jpn. Tokyo Jpn. 118–127

(2003).

237. Dravnieks, A., McDaniel, H. C. & Powers, J. J. Comparison by twelve laboratories of the

odor qualities of nine chemicals sniffed from the bottle and as gas-liquid chromatography

effluents. J. Agric. Food Chem. 27, 336–346 (1979).

238. Godish, T., Davis, W. T. & Fu, J. S. Air Quality, Fifth Edition. (CRC Press, 2014).

239. Buttery, R. G., Guadagni, D. G. & Ling, L. C. Flavor compounds. Volatilities in vegetable

oil and oil-water mixtures. Estimation of odor thresholds. J. Agric. Food Chem. 21, 198–

201 (1973).

240. Hui, Y. H., Chen, F. & Nollet, L. M. L. Handbook of Fruit and Vegetable Flavors. (John

Wiley and Sons, 2010).

241. Laurie, V. F. & Waterhouse, A. L. Glyceraldehyde bridging between flavanols and

malvidin-3-glucoside in model solutions. J. Agric. Food Chem. 54, 9105–11 (2006).

242. Es-Safi, N. E., Le Guernevé, C., Cheynier, V. & Moutounet, M. New phenolic compounds

obtained by evolution of (+)-catechin and glyoxylic acid in hydroalcoholic medium.

Tetrahedron Lett. 41, 1917–1921 (2000).

243. Hélène Fulcrand, C. B. A new class of wine pigments generated by reaction between

pyruvic acid and grape anthocyanins. Phytochemistry, 47, 1401-1407. Phytochemistry 47,

1401–7 (1998).

244. Es-Safi, N. E., Cheynier, V. & Moutounet, M. Study of the reactions between (+)-catechin

and furfural derivatives in the presence or absence of anthocyanins and their implication in

food color change. J. Agric. Food Chem. 48, 5946–5954 (2000).

145 146

245. Pissarra, J., Mateus, N., Rivas-Gonzalo, J., Santos Buelga, C. & Freitas, V. Reaction

Between Malvidin 3-Glucoside and (+)-Catechin in Model Solutions Containing Different

Aldehydes. J. Food Sci. 68, 476–481 (2003).

246. Pissarra, J. et al. Structural Characterization of New Malvidin 3-Glucoside − Catechin Aryl/

Alkyl-Linked Pigments. J Agric Food Chem 52, 5519–5526 (2004).

247. Pissarra, J. et al. Isolation and structural characterization of new anthocyanin-alkyl-catechin

pigments. Food Chem. 90, 81–87 (2005).

248. Chapman, N. Correlation Analysis in Chemistry: Recent Advances. (Springer Science &

Business Media, 1978).

249. Mulliez, M. & Naudy, C. Synthèse d’α-hydroxysulfinates. Tetrahedron 49, 2469–2476

(1993).

250. Green, L. R. & Hine, J. The pH independent equilibrium constants and rate constants for

formation of the bisulfite addition compound of isobutyraldehyde in water. J. Org. Chem.

39, 3896–3901 (1974).

251. Blackadder, D. A. & Hinshelwood, C. The Kinetics of the Decomposition of the Addition

Compounds formed by Sodium Bisulphite and a Series of Aldehydes and Ketones. Part II.

General Discussion of Energy-Entropy Relations. 2720–2727 (1958).

252. Olson, T. M., Boyce, S. D. & Hoffmann, M. R. Kinetics, thermodynamics, and mechanism

of the formation of benzaldehyde-sulfur(IV) adducts. J. Phys. Chem. 90, 2482–2488 (1986).

253. Salas, E. et al. Demonstration of the occurrence of flavanol–anthocyanin adducts in wine

and in model solutions. Anal. Chim. Acta 513, 325–332 (2004).

254. Burroughs, L. F. & Sparks, a H. Sulphite-binding Power of Wines and Ciders. I.

Equilibrium Constants for the Dissociation of Carbonyl Bisulphite Compounds. J. Sci. Food

Agric. 24, 187–198 (1973).

146 147

255. Harbertson, J. F., Picciotto, E. A. & Adams, D. O. Measurement of Polymeric Pigments in

Grape Berry Extracts and Wines Using a Protein Precipitation Assay Combined with

Bisulfite Bleaching. Am. J. Enol. Vitic. 4, 301–306 (2003).

147

Vita

Marlena K. Sheridan

Education Ph.D. Food Science, The Pennsylvania State University, University Park, PA, 2016 B.A. Chemistry, Barnard College, Columbia University, New York, NY, 2011

Publications Sheridan, M.K. & Elias, R.J. Exogenous acetaldehyde as a tool for modulating wine color and astringency during fermentation. Food Chem. 2015, 177, 17–22.

Presentations Sheridan, M.K. & Elias, R.J. Where has all the color gone? Oxidation techniques for improving color stability of PA red wines. PA Wine Marketing and Research Board Symposium, State College, PA, March 2016, Oral Presentation. Sheridan, M.K. & Elias, R.J. Effect of Sulfite on the Reactivity of Exogenous Acetaldehyde with Wine Flavonoids. 251st National Meeting of the American Chemical Society, San Diego, CA, March 2016, Oral Presentation. Sheridan, M.K. & Elias, R.J. Investigating the Role of Acetaldehyde in Tannin Modification and Color Stability of Red Wines. 66th Annual American Society for Enology and Viticulture National Conference, Portland, OR, June 2015, Poster Presentation. Sheridan, M.K. The Effect of Acetaldehyde on Red Wine Color Stability and Astringency. PA Wine Marketing and Research Board Symposium, State College, PA, April 2015, Oral Presentation. Sheridan, M.K. & Elias, R.J. Effect of Exogenous Acetaldehyde during Fermentation on Red Wine Tannins and Astringency. 65th Annual American Society for Enology and Viticulture National Conference, Austin, TX, June 2014, Poster Presentation. Sheridan, M.K.; Orman, M.; Tolentino, J.; Weena, U.; Stein, A.; Merrer, D.C. Intermolecular Chemistry of Dichlorocarbene Additions to Strained C-C π Bonds – Finding the Dynamics Threshold. 241st National Meeting of the American Chemical Society, Anaheim, CA, March 2011, Poster Presentation.

Awards PA Wine Marketing and Research Program Grant Winner (2015) Title: Oxidation Techniques for Improving Color Stability of Pennsylvania Red Wines 1st Place in Physical Sciences and Engineering, Gamma Sigma Delta Research Expo (2015) American Society for Enology and Viticulture – Eastern Section Scholarship (2015) American Society for Enology and Viticulture Scholarship (2015) Penn State College of Agricultural Sciences Competitive Grants Winner (2014) Title: Investigating Acetaldehyde-Induced Condensation of Wine Tannins and Anthocyanins