Sulfur dioxide and glutathione promote acetaldehyde accumulation under aerobic conditions in a model wine system: Impacts on hydroxyl radical formation

Abstract

The addition of antioxidants is essential in red wine aging, as it influences chemical oxidation processes. This study investigated the effects of free sulfur dioxide (SO₂) and glutathione (GSH), both individually and combined, on oxidative changes in a model wine system. Dissolved oxygen (DO), acetaldehyde, hydroxyl radicals (HO·), ferrous ions (Fe²⁺), SO₂, and GSH concentrations were monitored. Results showed that 30 mg/L SO₂ treatments effectively inhibited the Fenton reaction and accumulation of HO· and acetaldehyde, while higher SO₂ treatments (≥60 mg/L SO₂) promoted the production of acetaldehyde. All the GSH treatments increased the accumulation of acetaldehyde and HO· (fluorescence intensity = 686). Addition of SO₂: GSH treatment 30:20 and 30:80 further promoted HO· accumulation (fluorescence intensity = 2785). This study demonstrates the roles of SO₂ in radical scavenging and autoxidation, and acetaldehyde accumulation during wine oxidation can be effectively controlled and regulated by adjusting the GSH-to-SO₂ ratio.

Graphical abstract

Highlights

• •Assessing the Impact of SO₂ Dosages on Acetaldehyde Accumulation.
• •Evaluating the Role of Glutathione Dosages in Modulating Acetaldehyde Levels.
• •Investigating the Effect of SO₂ & Glutathione on Hydroxyl Radical Formation.
• •Integrating SO₂ and Glutathione Applications: Insights for Wine Production.
• •Provide new information on acetaldehyde accumulation in wine oxidation.

1. Introduction

It is well accepted that oxygen plays an important role in red winemaking, which influences the composition and quality of wine. In red winemaking, a small amount of oxygen is often introduced through microoxygenation to soften tannins (Gambuti et al., 2016), stabilize wine colour (McRae et al., 2015), and reduce vegetal aromas (Ugliano, 2013). However, an excessive amount of oxygen exposure can result in wine faults and colour loss (Gambuti et al., 2015). Sulfur dioxide (SO₂) is traditionally used in winemaking as an effective antioxidant agent, though some individuals may be sensitive to SO₂ and experience allergic reactions (Newair et al., 2023). FAO/WHO and OIV set limits on SO₂ intake (0.7 mg/kg body weight) and concentration in wine (e.g. 150 mg/L for dry red wines) (Newair et al., 2023; OIV, 2024). In recent years, alternative antioxidants for winemaking have attracted great attention of researchers. Several previous studies have demonstrated that glutathione (GSH) can act as an effective antioxidant to prevent wine oxidation (Gambuti et al., 2017; Nikolantonaki et al., 2018; Webber et al., 2017).

1.1. Mechanism of chemical oxidation in wine

The general mechanism of chemical oxidation in wine is a widely accepted theory (Fig. 1). Normally, oxygen in its triplet state does not directly react with phenolic compounds at lower pH (Danilewicz, 2011). Instead, Reactive Oxygen Species (ROS) such as superoxide anion (O₂·^-), hydroperoxyl (HOO·), hydroxyl (HO·), peroxyl (ROO·), and hydrogen peroxide (H₂O₂) generated via Fenton reactions involving dissolved oxygen in the presence of transitional metals [(Fe²⁺) and (Cu²⁺)], are responsible for the oxidation reaction of various compounds in wine (Oliveira et al., 2011). These radicals oxidize phenolic compounds such as catechin and gallic acid, which leads to the formation of quinones and hydrogen peroxide. Hydrogen peroxide is further involved in oxidation to produce hydroxyl radicals, which oxidize ethanol to acetaldehyde, a key marker of wine oxidation. An excess amount of acetaldehyde (typically >25–40 mg/L) has negative impacts on wine quality and aroma.

Fig. 1.

Chemical oxidation in wine: pathways of acetaldehyde, hydroxyl radicals, and SO₂ autoxidation.

Notes: a. The hydroperoxyl radical (HOO·) generated in the Fenton reaction oxidizes phenolic compounds and produces semiquinone and quinone, with itself being reduced to hydrogen peroxide (Waterhouse and Laurie, 2006);

b. H₂O₂ further reacts with a ferrous (Fe²⁺) or cuprous (Cu⁺) ions in a Fenton reaction to generate hydroxyl radicals (HO·) in wine (Elias and Waterhouse, 2010);

c. Ethanol reacts with HO· to produce acetaldehyde and water (Gambuti et al., 2015; Elias and Waterhouse, 2010);

d. Bisulfite (HSO₃⁻) in wine reacts with H₂O₂ to form sulfate (SO₄²⁻) and water (Waterhouse et al., 2016);

e. HSO₃⁻ reduces quinones formed during oxidation back to phenols (Waterhouse et al., 2016);

f. HSO₃⁻ reacts with quinones to form sulfonic acid or bisulfite adducts (Maria and Waterhouse, 2012);

g. The oxidation of HSO₃⁻ is catalyzed by Fe³⁺ to produce sulfite radicals (SO₃·^-), during which Fe³⁺ is reduced back to its ferrous state (Fe²⁺) (Danilewicz, 2007);

h. SO₃·^- reacts rapidly with dissolved oxygen to produce peroxomonosulfate radicals (SO₅·^-), which can restore the catalytic function by oxidizing Fe²⁺ to Fe³⁺ (Danilewicz, 2007);

i. SO₅·^- oxidizes HSO₃⁻ to produce SO₄²⁻ and sulfate radicals (SO₄·^-);

j. SO₄·^- oxidizes HSO₃⁻ to form SO₃·^- and SO₄²⁻ (Danilewicz, 2007);

k. Catechols from phenolic compounds can scavenge SO₅·^-, thereby terminating the chain reaction and hindering SO₂ autoxidation (Danilewicz, 2007);

l. During the radical chain propagation, SO₄·^- oxidizes ethanol to acetaldehyde in wine (Danilewicz, 2007).

1.2. Role of SO₂ in wine oxidation

In red wine making, 30 mg/L SO₂ is normally added at the crushing stage as an antioxidant, and 60–120 mg/L SO₂ in white wine making. At wine pH, up to 94 % of SO₂ exists in its bisulfite (HSO₃⁻) form, with only a small portion of it existing as molecular SO₂ (Danilewicz, 2007). It can bind a variety of molecules, including acetaldehyde, pyruvic acid, glucose, and phenolic compounds (e.g. anthocyanin, caffeic acid, and coumaric acid) (Oliveira et al., 2011).

As an antioxidant, HSO₃⁻ reacts with H₂O₂ to form sulfate (SO₄²⁻) and water (Fig. 1, d), competing with the Fenton reaction and preventing the formation of HO·. This stops further oxidation (Waterhouse et al., 2016). HSO₃⁻ also reduces quinone back to phenol (Fig. 1, e) (Waterhouse et al., 2016). In addition, the formation of sulfonic acid or bisulfite adducts can also prevent quinones from further oxidation (Fig. 1, f) (Maria and Waterhouse, 2012).

In the presence of transition metals like Fe and Cu, SO₂ undergoes a complex autoxidation, which significantly accelerates the oxidation reaction rate (Gambuti et al., 2015). HSO₃⁻ is catalyzed by Fe³⁺ to produce sulfite radicals (SO₃·^-), which further react with the dissolved oxygen rapidly to produce peroxomonosulfate radicals (SO₅·^-) (Fig. 1g and h). These radicals are strong oxidizing agents that regenerate metal catalysts Fe and Cu. The resulting SO₃·^- then participates in the above reactions, consuming additional oxygen molecules and propagating the chain reaction (Danilewicz, 2007). It is important to note that catechols from phenolic compounds can scavenge SO₅·^-, preventing the chain reaction and hindering SO₂ autoxidation (Fig. 1, k).

1.3. Role of glutathione in wine oxidation

Glutathione (GSH) is a tripeptide composed of glutamic acid, cysteine, and glycine (Dienes-Nagy et al., 2022), known for its antioxidant and detoxifying properties (Pastore et al., 2003). It is synthesized in both the cytosol and chloroplasts in plant cells (Jez et al., 2011). In grape juice, GSH content ranges from 10 to 100 mg/L. For example, Semillon contains 10–20 mg/L, and Sauvignon Blanc contains 30–40 mg/L, with lower levels in white wines compared to grape juice (Fracassetti et al., 2011). In 2015, OIV approved the addition of GSH (up to 20 mg/L) in winemaking (OIV, 2015). GSH can act as a protective agent to preserve the varietal aroma and prevent wine browning in white wine (Kritzinger et al., 2012). This can be attributed to the reduction of quinone or the formation of glutathione conjugates (Newair et al., 2023). However, its role in acetaldehyde and hydroxyl radical formation remains unclear.

Most existing studies are focused on the changes in phenolic composition and acetaldehyde accumulation in the presence of SO₂. However, the antioxidant effects of SO₂ and GSH and their combination on HO· and acetaldehyde formation remain poorly investigated. The hypothesis was that the application of SO₂ and GSH, as antioxidants, would suppress the accumulation of hydroxyl radicals and acetaldehyde by interfering with free radical generation in chemical oxidation and stabilizing reactive intermediates. In this study, the effects of varying dosages of SO₂ and GSH, both individually and in combination were evaluated regarding the accumulation of HO·, acetaldehyde, and Fe²⁺ concentration during oxidation in a model wine system. Specifically, the research investigated how SO₂ dosage shifts its primary role from radical scavenging to autoxidation, thereby altering the oxidative profile of the system.

2. Materials and methods

2.1. Chemicals

Acetonitrile and methanol were obtained from Fisher Scientific (Fair Lawn, NJ, USA). Acetaldehyde, glutathione, and DNPH-acetaldehyde hydrazine standard, and DNPH (30 % water, m/m) were obtained from Sigma-Aldrich (St. Louis, MO, USA). All the above solvents were HPLC grade. Ortho-phosphoric acid, boric acid, and tartaric acid were obtained from Tianjin Kemio Chemical Reagent Co., Ltd. (Tianjin, China). Potassium metabisulfite (PMS), sodium hydroxide, hydrogen peroxide (30 % v/v), ammonium acetate, anhydrous ethanol, acetic acid, ascorbic acid, hydrochloric acid, and sulfuric acid were obtained from Sinopharm Chemical Reagent Co., Ltd. (Shanghai China). Ethylene diamine tetraacetic acid (EDTA) was purchased from Aladdin (Shanghai, China). Dimethyl Sulfoxide (DMSO) was obtained from Solarbio (Beijing, China). Catechin, ferrozine, dithiothreitol (DTT), naphthalene-2,3-dicarboxaldehyde (NDA), and 1,4-cyclohexanedione (CHD) were obtained from Macklin (Shanghai, China). Potassium dihydrogen phosphate and ferrous sulfate heptahydrate (FeSO₄·7H₂O) were obtained from Tianjin Damao Chemical Reagent Partnership Enterprise (Tianjin, China).

2.2. Experimental design

The experimental design consisted of three trials (Table S1). Trial 1 and Trial 2 investigated the effect of individual SO₂ (30, 60, and 120 mg/L) and GSH (20, 40, and 80 mg/L) on the accumulation of HO·, acetaldehyde, and Fe²⁺ concentration during oxidation in a model wine system, respectively. Trial 3 evaluated the synergistic/combined effects of both antioxidants at SO₂:GSH ratios of 30:20, 30:80, 120:20 and 120:80.

2.3. Preparation of model wines

The model wines were prepared as described by Danilewicz and Wallbridge (2010) with modifications. 12 % v/v aqueous ethanol was used to make model wine. The model wine contained tartaric acid (8 g/L), FeSO₄·7H₂O (248.9 mg/100 mL), and catechin (500 mg/L). The pH of the model wine was adjusted to 3.6 using 10 M NaOH.

Air-tight anaerobic glass bottles (Capacity = 0.63 L, Inner diameter = 8.5 mm, Butyl rubber stopper length = 0.9 mm) with pre-attached dissolved oxygen sensors were used. The model wine of 610 mL was pre-poured in the bottle (with left of 20 mL of headspace). Potassium metabisulfide (PMS) and glutathione (GSH) were added to model wines according to Table S1. All bottles were placed in the incubator (model SPL-450, Tianjin Laboratory Instrument Equipment Co., Ltd., China) in the dark at 30 °C. The experiment was carried out in triplicate (n = 3). Oxygen was added to model wines as described by Ferreira et al. (2015). The experiment was initiated 12 h after the completion of oxygen exposure. Dissolved oxygen (DO) levels were monitored and regular samples were taken until the DO reached a plateau. Samples were taken according to specific concentration changes: initially, samples were collected for every 2 mg/L decrease in DO; as the rate of DO depletion slowed toward the end of the aging process, the sampling interval was narrowed to 1 mg/L. Samples were taken using a syringe by puncturing the butyl rubber stopper, to ensure that no exogenous oxygen entered the bottle during the aging process. The concentrations of SO₂ and GSH were selected based on a designed concentration gradient: one level within the legal limit, one level exceeding the legal limit, and one substantially excessive level. This gradient was established to systematically evaluate the concentration-dependent effects. The free SO₂ concentration was measured immediately after each sampling, and the rest of the samples were stored at −80 °C for further chemical analyses.

2.4. Chemical analyses

2.4.1. Determination of dissolved oxygen level (DO)

Dissolved oxygen level in model wines was measured as described by Han et al. (2019). The DO levels were measured in triplicate.

2.4.2. Determination of free SO₂ concentration

The free SO₂ concentration was measured as described by Iland et al. (2004). The free SO₂ concentrations were measured in triplicate.

2.4.3. Determination of glutathione concentration (GSH) by HPLC-FLD

The GSH concentration in model wines was measured as described by Webber et al. (2014). Shimadzu LC-20AT equipped with InertSustain C18 column (5 μm, 250 × 4.6 mm) and Fluorescence Detector (FLD) was used for quantification. GSH was eluted under the following conditions: oven temperature, 40 °C; the mobile phase, methanol-phosphate buffer (15:85 v/v) with pH adjusted to 7.5; the flow rate of mobile phase, 1 mL/min; The excitation and emission wavelengths were 467 nm and 525 nm, respectively. GSH standard was used to create the calibration curve (0–60 mg/L). The GSH concentrations in samples were measured in triplicate.

2.4.4. Determination of Hydroxyl Radicals (HO·)

The fluorescence intensity (FI) of hydroxyl radicals (HO·) was measured as described by Tai et al. (2002). The fluorescence intensity of hydroxyl radical is proportional to its concentration; therefore, of FI was used to represent the concentration of hydroxyl radical. The hydroxyl radical was measured in triplicate.

2.4.5. Determination of acetaldehyde by HPLC

The concentration of acetaldehyde was measured as described by Han et al. (2015). Acetaldehyde was quantified using a Shimadzu LC-20AT HPLC instrument equipped with a UV detector (Shimadzu, China) and a C18 column (5 μm, 250 × 4 mm) (Agilent). The wavelength was 365 nm with an injection sample volume of 20 μL; the column temperature was set at 35 °C; and the mobile phase flow rate was set at 0.75 mL/min. Acetonitrile (mobile phase A) and water (mobile phase B) were used for elution. The following elution gradients were used: 0.0 min, 35 % B; 25 min, 60 % B; 30 min, 90 % B; 32.5 min, 95 % B; 35 min, 35 % B; and 40.0 min, 35 % B. DNPH-acetaldehyde hydrazine standard was used to create the calibration curve (0–50 mg/L). The acetaldehyde concentrations were measured in triplicate.

2.4.6. Determination of ferrous ions (Fe²⁺)

The concentration of ferrous ions (Fe²⁺) was measured using a UV spectrophotometer as described by Nguyen and Waterhouse (2018). FeSO₄ was used to create the calibration curve (0–20 mg/L). The Fe²⁺ concentrations in samples were measured in triplicate.

2.5. Statistical analysis

Analysis of Variance (ANOVA), Duncan's multiple range test, Pearson's correlation coefficients, least significant difference (LSD), and principal component analysis (PCA) were analyzed using IBM SPSS Statistics 25 (Model 25.0.0, IBM Corporation, USA).

3. Results and discussion

3.1. Free SO₂ and oxidative changes during aging

Free SO₂ concentration had a significant impact on the consumption of DO in the model wine (p < 0.05) (Fig. 2A). For example, in the control, no oxygen was consumed over first 36 h while with the highest level of SO₂, the lowest DO (2 mg/L) was observed. This is similar to a previous study in wine oxidation which had shown that on Day 15, 25 mg/L SO₂ treatment consumed less DO (approximately 0.3 mg/L) than 65 mg/L SO₂ (Gambuti et al., 2015).

Fig. 2.

Change of dissolved oxygen (DO) during bottle aging (A), Initial Oxygen Consumption Rate (IOCR) with varying SO₂ levels (B), Change of free SO₂ concentration during bottle aging with varying SO₂ levels (C). The error bar in the above figures represents the standard deviation of the repeated trials (n = 3). Value not sharing the same letter differ significantly (p < 0.05). The vertical bar in bold represents the least significant difference (LSD).

SO₂ concentration also had an impact on the rate of DO consumption in the early stage of aging (Fig. 2A). The most rapid decrease in DO was observed in the treatment with 120 mg/L of SO₂, followed by the treatment with 60 mg/L SO₂, the treatment with 30 mg/L SO₂, and finally the control; such trend was also observed in the Initial Oxygen Consumption Rate (IOCR) (Fig. 2B). The highest IOCR (4.6 mg/L/12h) was found in the model wine treated with 120 mg/L SO₂, and the lowest IOCR (0.2 mg/L/12h) was found in the control. The Average Oxygen Consumption Rate (AOCR) (Table S2) is similar to Initial Oxygen Consumption Rate (IOCR). The highest AOCR (0.53 mg/L/12h) was found in the model wine treated with 120 mg/L SO₂, and the lowest AOCR (0.17 mg/L/12h) was found in the control. Similar findings were demonstrated by Motta et al. (2022), showed that the wines treated with 40 mg/L of SO₂ exhibited an oxygen consumption rate of 0.053 mg/L/12h, which was 2.4 times greater than the 20 mg/L SO₂ treatment (0.022 mg/L/12h).

This behavior might be expected since SO₂ is the ultimate reducing agent in most wines. However, the presence of additional SO₂ would not necessarily have this effect unless it was involved in a slow step in the oxidation cascade (Fig. 1, d, e, and f).

As SO₂ plays an antioxidant role in wine, free SO₂ is gradually depleted during aging. Fig. 2C demonstrates a decreasing trend in the remaining SO₂ concentration across all treatments. After 204 h, free SO₂ was completely depleted in 30 and 60 mg/L treatments, and approximately 45 mg/L free SO₂ still remained in 120 mg/L treatment.

The findings in the current study align with the established antioxidant mechanism of SO₂ in wine. During aging, there are two main steps of chemical oxidation. The first step involves the formation of ROS from dissolved oxygen through the catalysis of metal ions; the second step is the Fenton reaction between H₂O₂ and Fe²⁺ that generates HO· which reacts with all substances present in the solution, nearly in proportion to their concentration (Danilewicz, 2007; Waterhouse and Laurie, 2006). SO₂ can react with H₂O₂ to form sulfate (SO₄²⁻) and water, thereby terminating the Fenton reaction and preventing further oxidation (Danilewicz, 2007; Waterhouse et al., 2016). These reactions collectively contribute to the depletion of SO₂. The reducing action of SO₂ drives the oxidation reactions toward the formation of oxidative products, thereby enhancing the consumption of dissolved oxygen.

Fig. 3 demonstrates the relationship between the concentration of acetaldehyde and different SO₂ treatments. At the end of aging, the lower acetaldehyde concentrations were found in the control (3.9 mg/L) and treatment 30 mg/L SO₂ (3.1 mg/L). The highest acetaldehyde concentration was observed in the 120 mg/L SO₂ treatment. This finding is different from what is typically expected, as the addition of SO₂ should suppress oxidation products. This might be attributed to the varying chemical reaction pathways of SO₂ at different concentrations.

Fig. 3.

The relationship between the concentration of acetaldehyde and different SO₂ treatments. The error bar in the above figures represents the standard deviation of the repeated trials (n = 3). Values not sharing the same letter differ significantly (p < 0.05).

Fig. 4 shows the correlation between acetaldehyde, HO·, and Fe²⁺ concentration. At the end of aging, the lowest FI of HO· was found in 120 mg/L SO₂ treatment with (426), followed by 60 mg/L (864), then 30 mg/L SO₂ treatments (757), and finally, the control (964). This can be attributed to the addition of SO₂ inhibiting the accumulation of HO·, because the HSO₃⁻ formed can react with H₂O₂ to form sulfate (SO₄²⁻) and water, thus terminating the Fenton reaction (Danilewicz, 2007).

Fig. 4.

The relationship between acetaldehyde concentration, fluorescence intensity (FI) of hydroxyl radical (HO·), and ferrous ion concentration in model wines with varying SO₂ levels.

HO· is involved in the formation of acetaldehyde (Danilewicz, 2007). With increasing SO₂ dosage, stronger correlation between acetaldehyde and HO· were observed. For example, linear relationships were observed between acetaldehyde and HO· concentrations in different SO₂ treatments, the R-squares (R²) for control, 30, and 60 mg/L (R² = 0.951) SO₂ treatments are 0.725, 0.869, and 0.951 (Fig. 4). However, the R² value (0.800) for the 120 mg/L SO₂ treatment was lower than that of other SO₂ treatments. The trend indicates a quadratic relationship, and the result of 120 mg/L SO₂ treatment is different from the others, which may be attributed to the autoxidation of SO₂.

The production of both SO₄·^- and HO· led to acetaldehyde accumulation. When SO₂ concentration was 30–60 mg/L or lower, the reaction between SO₂ and H₂O₂ is the predominant pathway, inhibiting the formation of HO· and the oxidation of ethanol. Consequently, a lower acetaldehyde concentration was observed. Elias and Waterhouse (2010) studied the formation of 1-hydroxylethyl radicals in ethanol oxidation and demonstrated that the formation of these radicals was completely inhibited in the presence of 64 mg/L of SO₂.

When SO₂ concentration was 120 mg/L, the autoxidation of SO₂ became the predominant reaction. In the presence of transitional metals (Fe³⁺), SO₃·^- radicals were generated, leading to an increase in Fe²⁺ concentration. During the radical chain propagation, SO₄·^- was formed, which oxidized ethanol to acetaldehyde in wine (Fig. 1, l) (Danilewicz, 2007). In 120 mg/L SO₂ treatment, the generation of HO· was initially inhibited during the early stage of aging, while the concentration of acetaldehyde increased rapidly. Additionally, a higher concentration of Fe²⁺ was observed, which showed a significant (p < 0.05) correlation with the concentration of acetaldehyde (Table S3). As free SO₂ was gradually consumed, the predominant reaction shifted back, and the hydroxyl radical concentration increased, slowing the production of acetaldehyde.

3.2. Glutathione and oxidative changes during aging

Glutathione (GSH) had a significant impact on the consumption of DO in the model wine (p < 0.05) (Fig. 5A). For example, the highest DO (9 mg/L) was found in control at the end of model wine aging, while the lowest DO (5.8 mg/L) was found in the model wine treated with 80 mg/L of GSH. That can be attributed to the strong reducing properties of the sulfhydryl structure (-SH) contained in GSH, which can react with quinones and reduce them to phenols (Gambuti et al., 2017; Newair et al., 2023; Waterhouse and Nikolantonaki, 2015). This accelerated DO consumption indirectly.

Fig. 5.

Change of dissolved oxygen (DO) during bottle aging (A), Initial Oxygen Consumption Rate (IOCR) with varying GSH level (B), Change of GSH concentration during bottle aging with varying GSH level (C). The error bar in the above figures represents the standard deviation of the repeated trials (n = 3). Values not sharing the same letter differ significantly (p < 0.05). The vertical bar in bold represents the least significant difference (LSD).

GSH concentration also had an impact on the rate of DO consumption in the early stage of aging (Fig. 5A). The most rapid decrease in DO was observed in the treatment with 80 mg/L of GSH, followed by the treatment with 20 and 40 mg/L GSH, and finally the control. A similar trend was observed in IOCR (Fig. 5B) and AOCR (Table S2). The highest IOCR (1.4 mg/L/12h) was found in the model wine treated with 80 mg/L GSH, and the lowest IOCR (0.4 mg/L/12h) was found in the control. There was no significant difference between 20 and 40 mg/L GSH treatment. The highest AOCR (0.98 mg/L/12h) was found in the model wine treated with 80 mg/L GSH, and the lowest AOCR (0.50 mg/L/12h) was found in the control. Like SO₂, GSH plays an antioxidant role in wine; it is also gradually depleted during aging. However, the difference between these two antioxidants is that GSH was consumed much quickly than SO₂. Fig. 5C demonstrated a decreasing trend in remaining GSH concentration across all treatments. At 36 h, the amount of remaining GSH in all treatments was below 5 mg/L.

Fig. 6 demonstrates the correlation between acetaldehyde, HO·, and Fe²⁺ concentration. At the end of aging, the lowest FI of HO· was found in control (338), followed by 20 mg/L GSH (491), then 40 mg/L GSH (662), and finally 80 mg/L GSH (686). A strong correlation was found between the concentration of HO· and acetaldehyde accumulation (p < 0.05) (Table S4). The production of HO· led to acetaldehyde accumulation. Linear relationships were observed between acetaldehyde and HO· concentrations in different GSH treatments: control (R² = 0.822), 20 (R² = 0.971), 40 (R² = 0.952), and 80 mg/L (R² = 0.991) GSH treatments (Fig. 6). The highest concentrations of acetaldehyde were observed in 40 and 80 mg/L GSH treatment. In addition, a lower concentration of Fe²⁺ was observed in all GSH dosages at the end of the aging experiment. Such findings indicated that GSH did not inhibit the Fenton reaction, but instead, GSH facilitated the reduction reaction of quinones to phenols (Sonni et al., 2011) and simultaneously promoted the transformation of metal ions, thereby driving the Fenton reaction forward, which led to the production of HO· and accumulation of acetaldehyde. This is in contrast to the results of the experiments of Gambuti et al. (2015), in which GSH inhibited the accumulation of acetaldehyde during micro-oxidation in their experiments. The differences of acetaldehyde accumulation observed could be attributed to variations in system composition. The previous study used red wine, which contains a complex mixture of phenolic compounds with diverse structures and radical scavenging capabilities. In the current study, the model wine only contained catechol, representing a simpler system with lower phenolic concentrations. Furthermore, the different oxygen exposure method was used compared to the current study.

Fig. 6.

The relationship between acetaldehyde concentration, fluorescence intensity (FI) of hydroxyl radical (HO·), and ferrous ion concentration in model wines with varying GSH levels.

In summary, the distinct antioxidant functions of GSH and SO₂ were demonstrated through the detection of hydroxyl radicals FI and acetaldehyde concentration. In model wine, GSH primarily acts as an antioxidant by reducing quinones to scavenge radicals during the initial stage of oxidation. However, as quinones are reduced back to phenols, these phenols can promote the forward progression of the oxidation reaction by converting Fe²⁺ to Fe³⁺. This resulted in the accumulation of HO· and acetaldehyde. The rapid consumption of GSH by quinones and its potential role in ion recycling appear to be the dominant processes, leading to the observed net increase. In contrast, SO₂ was effective in both the reduction reaction of quinones and H₂O₂, thereby interrupting the oxidation reaction. In addition, SO₂ inhibits the Fenton reaction and HO· concentration, while its impact on acetaldehyde concentration depends on its dosage. Conversely, the addition of GSH promoted the accumulation of both HO· and acetaldehyde.

3.3. SO₂ - GSH interaction and oxidative development

When SO₂ and GSH were applied together, the DO level was primarily influenced by higher SO₂ dosage (Fig. 7). At the end of the aging experiment, the highest DO was found in control (6.5 mg/L), followed by the SO₂: GSH treatment 30:20 (5.2 mg/L) and 30:80 (4.0 mg/L), and with the lowest level in 120:20 (2.3 mg/L) and 120:80 (2.3 mg/L). This trend was aligned with findings from Fracassetti et al. (2013), the DO in the treatment with high SO₂ dosage reached plateau at the end of the aging experiment (SO₂ dosage = 50 mg/L); and the DO in the combined treatment (SO₂ = 20 mg/L, GSH = 67.5 mg/L) was decreased to approximately 2.0 mg/L. In addition, the remaining DO in all the combined treatments was lower than the DO of SO₂: GSH treatment 20:20 (7 mg/L) in a previous study by Giménez et al. (2023). Treatments with higher final dissolved oxygen concentrations exhibited lower average oxygen consumption rates. The highest AOCR (0.40 mg/L/12h) was found in treatments SO₂: GSH 120:20 and 120:80, and the lowest AOCR (0.12 mg/L/12h) was found in the control (see Table S2).

Fig. 7.

Change of dissolved oxygen (DO) during bottle aging. The error bar in the above figure represents the standard deviation of the repeated trials (n = 3). The vertical bar in bold represents the least significant difference (LSD).

Table 1 illustrates the remaining SO₂ and GSH concentrations across different treatments. Overall, both SO₂ and GSH were consumed during the aging experiment. The dosage of SO₂ had a significant impact on the remaining GSH concentration (p < 0.05). At the same GSH dosage (20 mg/L or 80 mg/L), there were significant differences (p < 0.05) between the remaining GSH concentration in treatment 30:20 and 120:20, and 30:80 and 120:80. At the same SO₂ dosage (30 mg/L or 120 mg/L), there were no significant differences in the remaining SO₂ concentration when treated with different concentrations of GSH. These findings align with the previous observations, indicating that the dosage of GSH did not influence the consumption of SO₂ (Díaz et al., 2021).

Table 1.

Remaining SO₂ and GSH concentrations across different treatments during aging.

Time	GSH (20 mg/L)		GSH (80 mg/L)		SO2 (30 mg/L)		SO2 (120 mg/L)
	30:20	120:20	30:80	120:80	30:20	30:80	120:20	120:80
0 h	20.00 ± 0.0a	20.00 ± 0.0a	80.00 ± 0.00a	80.00 ± 0.00a	30.00 ± 0.00a	30.00 ± 0.00a	120.00 ± 0.00a	120.00 ± 0.00a
21 h	6.28 ± 0.52b	5.29 ± 0.75c	19.42 ± 0.57b	18.07 ± 0.76c	23.52 ± 3.00b	19.68 ± 2.20c	109.44 ± 9.44b	111.84 ± 3.00 ab
60 h	3.14 ± 0.22e	3.95 ± 0.38d	9.95 ± 0.92de	10.69 ± 1.01d	13.44 ± 1.66d	13.44 ± 1.66d	82.08 ± 9.44c	84.48 ± 3.33c
156 h	2.47 ± 0.15f	3.23 ± 0.16e	9.22 ± 1.00e	7.33 ± 0.81f	0.00 ± 0.00e	0.00 ± 0.00e	42.24 ± 4.40d	46.08 ± 2.88d
300 h	0.00 ± 0.00g	0.00 ± 0.00g	3.42 ± 0.42g	6.34 ± 0.44f	0.00 ± 0.00e	0.00 ± 0.00e	1.92 ± 1.66f	12.00 ± 2.20e

Note: Different letters (a-g) indicate means that are significantly different at p < 0.05.

PCA was performed to investigate the overall effects of the combined application of SO₂ and GSH on acetaldehyde and hydroxyl radical accumulation during wine aging (Fig. 8). The first two principal components explained 49 % and 44 % of the variance in the dataset, respectively. Broadly speaking, the distribution of wines on biplot gave rise to three main groupings: those mainly clustered in the two lower quadrants, those in the upper left quadrant, and those in the upper right quadrant.

Fig. 8.

Principal component analysis (PCA) biplot of the effect of the combined application of SO₂ and GSH on acetaldehyde and hydroxyl radical accumulation.

Time was a major factor influencing the accumulation of acetaldehyde and HO·. The treatments that had less than 60 h of aging were in the two lower quadrants; this indicated that the concentration of acetaldehyde and HO· was below the averages. In addition, the higher acetaldehyde and HO· concentrations were mainly found in the treatments after 60 h of aging.

The treatments with the SO₂ dosage of 120 mg/L (with ≥60 h of aging) were all in the upper right quadrant. The vectors associated with this quadrant were acetaldehyde and Fe²⁺. The higher level of acetaldehyde and Fe²⁺ can be attributed to the autoxidation of SO₂, which was the predominant reaction. In the presence of transitional metals (Fe³⁺), SO₃·^- radicals were generated, leading to an increase in Fe²⁺ concentration. Additionally, SO₂ can react with H₂O₂ to form sulfate (SO₄²⁻) and water, thereby terminating the Fenton reaction and removing HO· (Danilewicz, 2007; Waterhouse et al., 2016).

The treatments with the SO₂ dosage of 30 mg/L and the control (with ≥156 h of aging) were all in the upper left quadrant. The vector associated with this quadrant was HO·. In the treatment with a lower SO₂ dosage, GSH had improved the accumulation of HO· because GSH facilitated the reduction reaction of quinones to phenols (Sonni et al., 2011) and simultaneously promoted the transformation of metal ions, thereby driving the Fenton reaction forward, which led to the production of HO·.

Free SO₂ of 30 mg/L is often added to wine. The current study showed that GSH in wine can potentially improve the accumulation of HO·. Therefore, the extra addition of GSH in winemaking should be avoided. In addition, further investigation on alternative antioxidants is required to remove HO·.

Based on the accumulation analysis of acetaldehyde and HO· in this experiment, the treatment with 30 mg/L SO₂ addition is the optimal solution for antioxidant effect, which has the lowest accumulation of acetaldehyde and HO·. An additional addition of GSH is not recommended, because the presence of GSH promotes the Fenton reaction, leading to the accumulation of HO·. Further research is required to investigate alternative antioxidants capable of mitigating and regulating acetaldehyde accumulation within a real wine system.

4. Conclusion

This study further investigated the antioxidant roles of SO₂ and GSH, both individually and combined, during wine oxidation. Two mechanisms of SO₂ in wine oxidation were demonstrated: (i) reacting with H₂O₂ to inhibit the Fenton reaction, thereby preventing the accumulation of HO· and acetaldehyde, and (ii) forming SO₄·^-, which promotes ethanol oxidation and acetaldehyde formation. In addition, SO₂ plays dual roles in radical scavenging and autoxidation, which also influence the accumulation of hydroxyl radicals and acetaldehyde. At a concentration of 30 mg/L SO₂, inhibition of the accumulation of acetaldehyde was the dominant effect; at 60 and 120 mg/L SO₂, the accumulation of acetaldehyde became more pronounced. GSH was found to facilitate the Fenton reaction during chemical oxidation. When both antioxidants were present simultaneously, SO₂ played the primary antioxidant role. In the SO₂: GSH treatment 120:20 and 120:80, acetaldehyde accumulation was influenced by both HO· and sulfate radicals generated in SO₂ autoxidation, with GSH playing a minor role. In contrast, at the SO₂: GSH treatment 30:20 and 30:80, the addition of GSH further promoted HO· accumulation. Further research is needed to explore alternative antioxidants capable of removing and regulating acetaldehyde accumulation by adjusting the GSH-to-SO₂ ratio, as well as the impact of these adjustments on wine colour.

Author statement

We declare that this manuscript is original, has not been published before and is not currently being considered for publication elsewhere.

We confirm that the manuscript has been read and approved by all named authors and that there are no other persons who satisfied the criteria for authorship but are not listed. We further confirm that the order of authors listed in the manuscript has been approved by all of us.

We understand that the Corresponding Author is the sole contact for the Editorial process. He is responsible for communicating with the other authors about progress, submissions of revisions and final approval of proofs.

All authors as follows:

●Jiaqi Wang (First Author): Investigation, Formal analysis, Writing - original draft, Writing - Review& Editing;

●Heqiang Chang: Investigation, Writing - original draft;

●Junzhe Wang: Writing - original draft;

●Yuhang Sun: Investigation, Formal analysis, Writing - original draft;

●Qinglong Wang: Writing - original draft;

●Lingmin Dai: Conceptualization, Funding acquisition, Methodology, Writing - original draft;

●Zhijing Ye: Writing - Review& Editing, Supervision;

●Guomin Han (Corresponding Author): Conceptualization, Funding acquisition, Methodology, Writing - Review& Editing, Supervision.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Handling Editor: Professor Aiqian Ye

Appendix A. Supplementary data

The following is the Supplementary data to this article: