Abstract
The intense passion fruit-type aroma of New Zealand Sauvignon blanc wines, attributed to high concentrations of the varietal thiols 3-mercaptohexan-1-ol (3-MH) and 3-mercaptohexan-1-ol acetate (3-MHA), is known to diminish rapidly over just one year in the bottle. It is thus important to understand the processes by which the aroma compounds are lost during storage to improve the shelf life of these wines. The loss of varietal thiols has been linked to polyphenol oxidation, catalyzed by metals, a process that can be inhibited by antioxidants such as sulfur dioxide (SO2), ascorbic acid, and glutathione. The evolution of 3-MH and 3-MHA was monitored in commercially bottled New Zealand Sauvignon blanc wines, stored in the dark at 15°C for 7 months, together with changes in polyphenol content and small molecule antioxidants. The acetate ester 3-MHA was the least stable of the varietal thiols and declined steadily in concentration following first-order kinetic decay, whereas 3-MH barely decreased over the first 3 months of bottle storage, and was instead followed by an increase in concentration after a further 4 months. These results pointed to hydrolysis of 3-MHA to 3-MH as the major loss mechanism in the commercial wines during the initial months of storage. Concomitantly, the flavan-3-ols exhibited a greater susceptibility to oxidation and polymerization reactions compared with the more stable hydroxycinnamic acids. Glutathione concentrations also declined steadily over the first year and would thus only provide protection against oxidative thiol losses up to that point. Free SO2 remained relatively stable in the commercially bottled Sauvignon blanc wines after an initial drop in concentration associated with oxygen entry at bottling.
Two varietal thiols of interest for passion fruit and grapefruit aromas in Sauvignon blanc wines are 3-mercaptohexan-1-ol (3MH) and 3-mercaptohexan-1-ol acetate (3MHA) (Tominaga et al. 1998a). Concentrations of 3-MHA reported in Sauvignon blanc wines range from not detected to several hundred ng/L (Tominaga et al. 1998b, Mateo-Vivaracho et al. 2010). However, Sauvignon blanc wines from Marlborough have contained levels as high as 2500 ng/L (Lund et al. 2009b). Given a perception threshold of 4 ng/L in model wine (Tominaga et al. 1996), the sensory impact of 3-MHA may be considerable in young wines. At concentrations greater than ~50 ng/L, 3-MHA can suppress the impact of the more intense green capsicum notes produced by methoxypyrazines (Campo et al. 2005). 3-MH is the most abundant varietal thiol, with levels frequently above the 60 ng/L perception threshold.
One issue surrounding Sauvignon blanc wines high in tropical fruit characters is the rapid loss of these aromas in the first one to two years in the bottle. Previous studies have shown that 3-MH is susceptible to polyphenol-mediated oxidation processes, which are accelerated by the presence of catalytic iron and lessened by the provision of free sulfur dioxide (Blanchard et al. 2004, Danilewicz et al. 2008, Nikolantonaki et al. 2010). In discussions on the effect of wine aging and changes in wine aroma profiles, 3-MHA and 3-MH have typically been grouped together as susceptible to oxidative degradation (Ugliano et al. 2009) because they possess a sulfhydryl group (-SH).
We have previously compared the loss characteristics of 3-MHA and 3-MH in research-scale wines that were manually bottled (Herbst et al. 2008). While 3-MHA was confirmed as the less stable of the two varietal thiols, a high ingress of oxygen at bottling adversely affected various wine oxidation parameters. The aim of the present study has been to examine the stability of the varietal thiols 3-MH and 3-MHA over a 7-month period following commercial bottling, monitor associated changes in polyphenol and antioxidant content, and thus determine the major loss mechanisms of 3-MH and 3-MHA in Sauvignon blanc wines.
Materials and Methods
Wine.
Sauvignon blanc wines, predominantly made from grapes from the Marlborough region, were obtained from Pernod Ricard (n = 4) and Villa Maria (n = 3) directly after commercial bottling under screwcap closure in 2007, using bottling procedures such as vacuum bottle fillers to minimize oxygen exposure during bottling. The bottles were stored upright in the dark in a temperature-controlled room at 15°C.
Reagents and solutions.
Hydrochloric acid (37%, reagent grade) and sodium hydroxide (pellets, ≥99%, reagent grade), di-sodium hydrogen phosphate dihydrate (≥99.5%), sodium acetate trihydrate (99.5–100.5%), and sodium sulfate anhydrous (powder, extra pure, 98.5–100.5%) were obtained from Scharlau (Barcelona, Spain). 4-Hydroxymercuribenzoic acid (p-HMB) sodium salt (≥95.0% Hg), l-cysteine hydrochloride hydrate (99%), and butylated hydroxyanisole (BHA) were from Sigma-Aldrich (Castle Hill, NSW, Australia). Tris (ultrapure, ≥99.9%) was provided by Applichem (Darmstadt, Germany) and Dowex (1 × 2, Cl− form, strongly basic, 50–100 mesh) was a product of Sigma-Aldrich. 5,5′-Dithio-bis-(2-nitrobenzoic acid) (99%) was from Acros Organics (Geel, Belgium). Ethyl acetate (≥99.7%, LC-MS Chromasolv, Fluka, Castle Hill, Australia) and dichloromethane (for gas chromatography; SupraSolv, Merck, Darmstadt, Germany) were used as solvents. 3-Mercaptohexan-1-ol (≥98%) was from Interchim (Montluçon, France), and 3-mercaptohexan-1-ol acetate (≥98%) was from Oxford Chemicals (Hartlepool, UK). As internal standards, 3-mercapto[1-2H2]hexan-1-ol ([1-2H2]3-MH) and 3-mercapto[1-2H2]hexan-1-ol acetate ([1-2H2]3-MHA) were synthesized by a previous method (Hebditch et al. 2007). Nitrogen (food grade) and helium (instrument grade) were sourced from BOC Gases (Auckland, New Zealand).
Catechin hydrate (≥98%), epicatechin (≥90%), caffeic acid (≥98%), trans-p-coumaric acid (≥98%), trans-ferulic aid, rutin hydrate (≥94%), l-ascorbic acid (≥99%), and l-glutathione reduced (≥98%) were purchased from Sigma-Aldrich. Acetic acid glacial (HPLC-grade, ≥99.8%) was obtained from Scharlau (Barcelona, Spain) and acetonitrile (HPLC grade) was from Merck. All reagents were used without further purification.
Solutions were prepared using grade 1 water (Barnstead Nanopure Diamond Water Purification System, Thermo Scientific, Waltham, MA) with a resistivity of 18.2 MΩ/cm at 25°C or absolute ethanol (≥99.5%, Univar, Ajax Finechem, Auckland, New Zealand). Model white wine base was prepared by dissolving 5 g/L l-(+)-tartaric acid (≥99%, Scharlau) in an ultrapure water/ethanol (88:12 v/v) solution. After adjustment of the pH to 3.20, using NaOH (10 N, 1 N) and HCL (1 N), the medium was purged with nitrogen for 15 min to lower the dissolved oxygen content.
Varietal thiols analysis.
The 3-MH and 3-MHA concentrations were assayed in duplicate according to a previous method (Tominaga et al. 1998b) with some modification. 5 mL of p-HMB (1 mM in a 0.1 M Tris solution) was added to 50 mL wine, followed by further additions: 0.5 mL of a 2 mM BHA solution and 50 μL of a isotopically labeled 22 μM 3-MH ([1-2H2]3-MH) and 2.8 μM 3-MHA ([1-2H2]3-MHA) mix, used as internal standards for 3-MH and 3-MHA, respectively. After pH adjustment to 7.00 ± 0.05 (10 N, 1 N NaOH; 1 N HCl), the sample was loaded onto a strongly basic anion exchange column (Dowex), which had been previously activated using 0.1 M HCl, and then rinsed with ultrapure water. After percolation of the sample, the column was washed with 50 mL 0.1 M sodium acetate buffer (pH 6.00). The varietal thiols were released from the thiol-p-HMB-complex fixed onto the column by percolating 50 mL of 50 mM l-cysteine solution (400 mg in 0.1 M sodium acetate buffer), adjusted to pH 6.00. The eluate was extracted twice with dichloromethane (4 and 2 mL) after addition of 0.5 mL ethyl acetate. The collected organic phase was dried over sodium sulfate anhydrous, filtered through silanized glass wool (Supelco, Bellefonte, PA), then concentrated under nitrogen flow to ~25 μL. The gas chromatographic analysis of varietal thiols was carried out using an 6890N gas chromatograph (Agilent, Santa Clara, CA) equipped with a 7683B automatic liquid sampler, a G2614A autosampler, and a 5973 mass selective detector. The inlet temperature was held at 240°C. 5 μL of the sample was injected in split mode with a split ratio of 5:1 and split flow of 5 mL/min and delivered onto a Agilent HP-INNOWax capillary column (60 m × 0.252 mm i.d., 0.25 μm film) using helium as carrier gas (112 kPa) at an initial flow rate of 1 mL/min (for 38.30 min), raised to 2.4 mL/min after separation of the compounds of interest. The initial oven temperature (50°C for 5 min) was ramped to 115°C at a rate of 3°C/min, then raised to 150°C at 40°C/min and held for 3 min, followed by a further increase to 173°C at 3°C/min, and finally to 250°C at 70°C/min (held for 17 min) before dropping to 50°C at 40°C/min (held for 3 min). The temperature of the interface line was 230°C. The ion source, operating in electron impact mode at 70 eV, was held at 230°C. The quadrupole temperature was set at 150°C. The varietal thiols and internal standards were detected in SIM mode selecting the following ions (m/z) for identification (in combination with the retention time given by an injection of individual standard); the quantifier ion is listed first: 118 and 103 for [1-2H2]3-MHA, 116 and 101 for 3-MHA, 136 and 102 for [1-2H2]3-MH, and 134 and 100 for 3-MH. Standard curves were obtained with nine calibration points by adding increasing quantities of the reference standards, prepared according to a previous method (Ellman 1959), to 50 mL Sauvignon blanc wine (70 to 2600 ng/L for 3-MHA; 500 to 18,000 ng/L for 3-MH). The linear regression and recovery was very good for both thiols. All of the samples were analyzed in duplicate.
pH and sulfur dioxide analyses.
All pH measurements were performed on a Thermo Orion 420A plus benchtop pH- meter (Thermo Scientific), calibrated at pH 4.00 and 7.00. Flow injection analysis, using the FIAstar 5000 Analyzer in 2-channel configuration with Sampler 5027 (FOSS, Hillerød, Denmark) controlled by SoFIA software, allowed simultaneous, single analysis of free and total sulfur dioxide (SO2) in duplicate Sauvignon blanc wines.
Polyphenols, ascorbic acid, and glutathione analysis.
A reversed-phase high-performance liquid chromatography (RP-HPLC) method (Kilmartin et al. 2002) was adapted for white wines and modified (with inclusion of an electrochemical detector) to separate and detect ascorbic acid and glutathione simultaneously with polyphenols in a single run. About 2 mL wine was filtered through a 0.45 μm membrane filter (Minisart RC 15; Sartorius, Göttingen, Germany), of which 20 μL was directly injected into an Agilent 1100 series instrument, equipped with a G1315B diode array detector (DAD) and coupled to a Coulochem III electrochemical detector (ECD) (model 5340; ESA Laboratories, Chelmsford, MA) operating in DC mode, consisting of a guard cell (model 5020) and a dual-electrode analytical cell (model 5010). The guard cell was located before the injector, set at a potential of +850 mV to eliminate oxidizable impurities in the mobile phase that may interfere with the baseline stability. The DAD was set at 260 nm for ascorbic acid (Lopes et al. 2006b), 280 nm for flavan-3-ols, 320 nm for hydroxycinnamates (including grape reaction product; GRP), and 365 nm for flavonols. The ECD (for glutathione) was kept at a potential of +750 mV and +800 mV for electrode 1 and 2, respectively, after reviewing the hydrodynamic voltammetry plot (graph of response versus potential) using a previous approach (Smith et al. 1995). The mobile phase consisted of a ternary solvent mixture: (A) ultrapure water, (B) ultrapure water/acetic acid (CH3COOH) solution (95:5 v/v), and (C) acetonitrile (CH3CN), delivered at a constant flow rate of 0.8 mL/min onto a Luna C18 column (250 × 4.6 mm i.d., 5 μm particle size) (Phenomenex, Torrance, CA) operated at 25°C. The solvent gradient of the RP-HPLC analysis can be found in Table 1. Polyphenols and ascorbic acid were identified using retention times and absorption wavelength maxima given by the respective reference standard or published in the literature if no standard was commercially available, while glutathione was identified using the retention time only. Quantitative determination was carried out by external standard calibration curves obtained by serial dilution of the reference compounds with model white wine and expressed as mg/L, encompassing a typical concentration range for the compound of interest. Since no commercial standards were available for the tartrate esters, they were quantified and expressed as mg/L of their corresponding free hydroxycinnamic acids. The 5-point standard calibration curves were linear within the calibration range. Recoveries close to 100% were obtained for all standards. Wine samples were analyzed in duplicate (single injection from the duplicate bottles).
Color analysis.
A single measurement of absorbance at 420 nm (A420) was taken against a blank (model white wine) of filtered wine (0.45 μm membrane filter; Minisart RC 15; Sartorius), based on a previous procedure (Iland 2004). The 420 nm absorbance was used as additional marker of oxidation progress, and was taken with a Spectronic Genesys 10 UV spectrophotometer (Thermo Scientific).
Statistical analysis.
One-way analysis of variance (ANOVA) was performed for all analytical parameters using SPSS Statistics 17.0 (SPSS Inc., Chicago, IL). When significant differences (p < 0.05) were indicated, the Tukey honestly significant difference (HSD) test (p < 0.05) was used to evaluate differences between means over time for each Sauvignon blanc wine. Linear relationships (*, **, and *** indicating significance at p < 0.05, 0.01, and 0.001, respectively) between compositional variables were measured using the Pearson’s correlation coefficient (r).
Results and Discussion
Varietal thiols.
The initial concentration of 3-MHA in the seven New Zealand Sauvignon blanc wines from the 2007 vintage ranged from 359 to 2358 ng/L (mean = 906 ± 701 ng/L). This varietal thiol has a very low perception threshold of just 4 ng/L (Tominaga et al. 1996) and has been given the sensory descriptor of sweet-sweaty passion fruit, an attribute that has been rated high in intensity for wines from the Marlborough region (Lund et al. 2009b). The levels of 3-MH, which highly correlated with those for 3-MHA (r = 0.948*** [n = 14]), varied between 2450 and 10,993 ng/L (mean = 5049 ± 3410 ng/L). The sensory descriptor passion fruit skin/stalk has been applied to this thioalcohol (Lund et al. 2009b), which exhibits a higher sensory threshold of 60 ng/L (Tominaga et al. 1996). These thiol concentrations (Table 2) fall within the range observed for 15-month-old New Zealand Sauvignon blanc wines (Lund et al. 2009b), but are higher than values reported for 14 Marlborough Sauvignon blanc wines (mean values of 136 ng/L for 3-MHA and 3164 ng/L for 3-MH), but with variable wine age and unknown transportation and storage histories (Mateo-Vivaracho et al. 2010). The concentrations obtained here are also higher for both varietal thiols than those obtained for a series of 16 Marlborough Sauvignon blanc research-scale wines in a previous study (Herbst et al. 2008), where the maximum concentrations were 440 ng/L for 3-MHA and 1950 ng/L for 3-MH, attributed to extensive oxidation during wine handling together with the contribution of evaporation losses due to prolonged exposure to air.
The quantitative changes in varietal thiols up to 7 months postbottling are shown for a representative high-thiol Sauvignon blanc wine (Figure 1). After 3 months of storage at 15°C in the dark, between 29 and 46% (40 ± 6% on average) of 3-MHA was lost across the seven wines, and only 31 ± 7% of the initial concentration was present 7 months after bottling. A 60% loss of 3-MHA was observed in French rosé wines within 3 months of storage (Murat 2005); after one year in the bottle this varietal thiol had disappeared completely.
Given that 3-MHA is an acetate ester, hydrolysis as a loss mechanism (Figure 2) should be considered together with oxidation of the thiol moiety, as observed with 3-MH (Blanchard et al. 2004, Nikolantonaki et al. 2010). The acid-catalyzed hydrolysis of acetate esters has been analyzed as pseudo-first-order kinetic reactions (Ramey and Ough 1980). The rate of this reaction then depends on the concentration of only one reactant [A], the acetate ester, and is proportional to its initial concentration [A0]. Thus, the reaction rate follows first-order kinetic decay and is expressed as:
The first-order rate constant k [time−1] for the 3-MHA hydrolysis was calculated using the integrated first-order rate equation:
The pseudo-first-order rate constants for the seven Sauvignon blanc wines (Table 3) were similar for the two time periods tested (0 to 3 months, averaging at k0–3 = 0.17 month−1; 3 to 7 months, averaging at k3–7 = 0.17 month−1). This value is equivalent to a half-life of 4 months, meaning that after 12 months of storage at 15°C (three half-lives), only 12.5% of the initial 3-MHA starting concentration would be predicted. Should oxidative degradation have been important with the commercial wines, a greater loss would have been expected during the first 3 months in the bottle, when the oxygen present at bottling is consumed and affects wine components such as an initial loss of SO2 (Brajkovich et al. 2005, Lopes et al. 2009). As demonstrated for a representative example (Figure 3), the rate constants for manually bottled Sauvignon blanc wines from a previous study, where higher levels of oxygen were present at bottling (Herbst et al. 2008), were higher for the first half of a year (k0–3 = 0.37 month−1 in the example here, compared to an average of 0.22 ± 0.07 month−1 across 16 wines). With the earlier research wines, some oxidative 3-MHA losses are indicated in addition to the hydrolysis loss mechanism, possibly due to oxidation of 3-MHA directly or to oxidation of 3-MH with the effect of accelerating the hydrolysis of 3-MHA through removal of 3-MH as a hydrolysis product. A significant decline in acetate esters (isoamyl acetate, hexyl acetate, phenylethyl acetate) was noted after 13 weeks of oxidative storage (Ferreira et al. 1997). We can also note that the wines sampled at 0 months went through the same bottling operation as the wines stored for 3 to 7 months and were thus exposed to the same conditions and potential evaporative losses. Under the conditions of the commercial bottling operations used for the seven wines of this study, the oxygen exposure at bottling was minimal, and the rate constants for the 0-to-3-month and the 3-to-7-month periods were nearly identical. Therefore, the 3-MHA loss appears to be almost entirely due to hydrolysis. In previous studies on the hydrolysis of acetate esters, the effect of pH on hydrolysis rate has been demonstrated, where a 0.63 pH unit change in one model solution study (from 3.58 to 2.95) increased the rate of loss for the acetate esters by 2.5- to 4-fold (Ramey and Ough 1980). With one exception (at pH 2.99), the pH values for the Sauvignon blanc wines in this study covered a narrow range from 3.16 to 3.30 (Table 3) and no correlation was observed between first-order rate constants and wine pH. The influence of further wine components, of variable concentrations, needs to be considered here.
By contrast, 3-MH levels barely declined after the first 3 months (5 ± 8% on average), but increased in all cases after a further 4 months of storage (Figure 1). This increase was likely due to the hydrolysis of 3-MHA to 3-MH, but could include 3-MH derived from further thiol precursors present in the wines (Capone et al. 2010) or by breakdown of any 3-MH disulfide present, as recently identified in botrytized Sauternes wines (Sarrazin et al. 2010). In the previous study using manually bottled Sauvignon blanc wines, 3-MH levels increased (between 1 and 63%) during the first 3 months (Herbst et al. 2008). Over the following 9 months, 3-MH concentrations declined steadily (39 ± 9% on average) in the manually bottled wines, and remained rather stable and well above the perception threshold in all of the wines up to 24 months postbottling. In French rosé wines, an average 3-MH loss of 35% was observed after 3 months of storage. In the French rosé wines, only half of the initial 3-MH content was left after 12 months of storage in the bottle (Murat et al. 2003, Murat 2005), while concentrations in red Bordeaux wines have been shown to diminish gradually during barrel aging, with only 25% on average of initial levels remaining after about 12 months (Blanchard et al. 2004). The sulfhydryl function of 3-MH is thus susceptible to oxidation. The extent of 3-MH oxidative loss over the 7 months in the wines of this study was not great, consistent with the minimal oxygen exposure in the commercial bottling operations. The total varietal thiol content, calculated in nmol/L, dropped by 9 ± 8% on average after 3 months in the bottle (presumably due to the presence of oxygen introduced at bottling) and remained relatively stable over the following 4 months, during which time the loss of 3-MHA was largely matched by an increase in 3-MH content.
Antioxidants.
The evolution of free SO2 and glutathione in the commercial wines is shown (Table 2). Ascorbic acid was not detected in any of these wines. As seen elsewhere (Brajkovich et al. 2005, Lopes et al. 2009), free SO2 content dropped over the first 3 months by 32 ± 9% on average (ranging from 18 to 44%), but remained relatively stable in the 4 months thereafter. This observation is in agreement with research showing that oxygen ingress into wine bottles is very important at bottling and affects SO2 losses in the first month of storage more than in the time thereafter (Lopes et al. 2006b). This initial decline in SO2 content reflects the oxygen input at bottling, mainly in the headspace above the wine (Kontoudakis et al. 2008). On the other hand, storage time had little effect (p = 0.097 [n = 14]) on the total SO2 concentration in the commercially bottled Sauvignon blanc wines up to the 7-month time point. These results are in contrast to those obtained for the earlier manually bottled Sauvignon blanc wines, where a high oxygen exposure at bottling led to a continuous decline in free and total SO2 throughout the first year in the bottle (Herbst et al. 2008). Glutathione, which diminishes relatively rapidly during storage (Lavigne et al. 2007), followed a similar trend to 3-MHA throughout the 7-month trial period (r = 0.655** [n = 21]), with levels declining between 49 and 77% (64 ± 9% on average) within 3 months of bottling. After 7 months, only 19 ± 4% on average of the initial concentration (averaging at 10.2 ± 3.2 mg/L) was found in the bottle. The rate of glutathione degradation in the commercially bottled wines was lower than in the previous study of manually bottled wines (Herbst et al. 2008) due to lower oxygen exposure at bottling.
Color.
The absorbance at 420 nm in the commercially bottled wines differed significantly (p < 0.05 [n = 14]) across the trial period. The A420 values decreased 3 months after bottling from 0.052 ± 0.007 to 0.049 ± 0.006, an average 7% drop, followed by an absorbance increase of 13% over the next 4 months of storage (Table 2). This trend has also been seen for accelerated white wine storage tests (Pérez-Zúñiga et al. 2000, Sioumis et al. 2005), revealing that browning exhibited a biphasic course, in that a decline in A420 can at times be observed prior to longer-term increases. This observation is in contrast to the manually bottled Sauvignon blanc wines, which exhibited a steady increase in A420 values throughout the trial period because of higher oxygen exposure at bottling (Herbst et al. 2008). It is possible that the initial decrease in A420 was due to some particulars of the polyphenol oxidation process, in that the quinones originally formed by polyphenol oxidation could take part in further reactions. Coupled oxidation with phenolic and nonphenolic species (e.g., antioxidants such as SO2 and glutathione) could reduce it back to less colored hydroquinone forms, resulting in a temporary drop in browning. Polymerization reactions involving polyphenol oxidation produced could then lead to an increase in the concentration of more intensely colored products, responsible for the browning observed thereafter.
Polyphenols.
Eight polyphenol compounds were identified in the New Zealand Sauvignon blanc wines: the flavan-3-ols (+)-catechin and (−)-epicatechin, the hydroxycinnamates trans-caftaric acid, grape reaction product (GRP), and cis- and trans-p-coutaric acid, and their respective hydrolyzed forms, trans-caffeic acid and trans-p-coumaric acid. No flavonols were detected in these wines, even though they have been detected in Sauvignon blanc juices, particularly from heavier press fractions (Patel et al. 2010). The total flavan-3-ol content at bottling was not particularly high, with (+)-catechin predominant (mean = 3.6 ± 0.8 mg/L) over (−)-epicatechin (1.7 ± 0.3 mg/L) (Table 4). The average total hydroxycinnamic acid concentration was 27.1 ± 5.8 mg/L (22.0 to 35.8 mg/L); GRP was the most abundant at 12.7 ± 2.1 mg/L (8.8 to 15.7 mg/L), followed by trans-caftaric acid at 8.5 ± 4.2 mg/L (4.8 to 16.0 mg/L) and smaller quantities of cis- and trans-p-coutaric acid, trans-caffeic acid, and trans-p-coumaric acid.
The quantitative changes in polyphenols up to 7 months of storage are shown (Table 4). A storage time of 7 months already had a significant effect on both (+)-catechin (p < 0.05) and (−)-epicatechin (p < 0.001), with greater losses occurring over the first 3 months in the bottle, leading to a total decline of 25 ± 7% in (+)-catechin and 37 ± 11% in (−)-epicatechin. These results are in accordance with other reports (Recamales et al. 2006, Herbst et al. 2008, Hernanz et al. 2009, Kallithraka et al. 2009) and demonstrate the greater susceptibility of flavan-3-ols to oxidation and polymerization reactions than the hydroxycinnamic acids (Simpson 1982, Gonzales Cartagena et al. 1994).
By contrast, trans-caftaric acid and trans-p-coutaric acid levels were not significantly affected by storage time (p > 0.05 [n = 42]), although average total losses of 10 ± 7% of trans-caftaric acid were observed. Similar observations were made for another white wine variety after almost one year of storage (up to 5% losses) (Vrhovšek and Wendelin 1998), whereas other authors found a more pronounced decline in trans-caftaric acid (Recamales et al. 2006, Hernanz et al. 2009, Kallithraka et al. 2009) or increase in trans-p-coutaric acid (Recamales et al. 2006, Hernanz et al. 2009) in other white wines during a similar storage period. The cis configuration of coutaric acid appeared less stable, dropping significantly (p < 0.05) after 7 months in the bottle by 13 ± 5% on average. Even higher losses (45 ± 9%) could be seen for the GRP (p < 0.001), in accordance with Vrhovšek and Wendelin 1998. Conversely, trans-p-coumaric acid increased continuously by 35 ± 20% (p < 0.05) during storage.
Polyphenols have been observed to lower headspace volatility of certain flavor compounds (Dufour and Bayonove 1999, Aronson and Ebeler 2004). In one study, the addition of (+)-catechin (10 mg/L) lowered the perception of 3-MHA, but increased the perception of 3-MH in diluted base wine (Lund et al. 2009a). Consequently, the observed decline of this flavan-3-ol would work in the direction of lessening the intensity of the sweet-sweaty passion fruit aroma and could increase the passion fruit skin/stalk character. Furthermore, the “green” characters will become more prevalent with time, because methoxypyrazines are relatively stable during bottle aging (Maga 1990, Marais 1998) and both (+)-catechin and trans-caffeic acid have been demonstrated to suppress the perception of 2-methoxy-3-isobutylpyrazine (Lund et al. 2009a), which imparts capsicum and green gooseberry aroma notes. Methoxypyrazines are likely to influence the perception of the passion fruit-type characters in young Sauvignon blanc wines already, since they have been shown to exhibit a negative effect on the tropical fruit notes (related positively to 3-MHA) in white wine matrices, even if they are not being perceived directly themselves (Campo et al. 2005). The loss of flavan-3-ols and 3-MHA would reinforce each other in allowing methoxypyrazines to exert a greater impact on Sauvignon blanc wine aroma with storage time, with some influence also expected from changes in the hydroxycinnamic acids.
Conclusions
The hydrolysis of 3-MHA was shown to be the predominant cause for the gradual decline of this acetate ester in New Zealand Sauvignon blanc wines. 3-MHA hydrolysis was also responsible for the initial increase in 3-MH content in these same wines, which as a thioalcohol will be lost via oxidation during bottle aging. The contribution of oxidative 3-MHA degradation in screwcapped Sauvignon blanc wine is negligible when bottled commercially. The rapid decline of 3-MHA, combined with the greater stability of 3-MH, may well explain some of the changes in aroma profiles of New Zealand Sauvignon blanc wines during bottle storage. The removal of 3-MHA at the expense of 3-MH (due to hydrolysis) is expected to lower the intensity of the sweet-sweaty passion fruit aroma, while leaving a more passion fruit skin/stalk character. Any masking or synergistic effects that 3-MHA undergoes in conjunction with other compounds (e.g., methoxypyrazines, esters, and polyphenols) would also modify the aroma profile of the wines during bottle storage. The antioxidant glutathione would only provide protection from oxidative aroma losses during the first year or so of bottle aging, since it was found to disappear rapidly with time. On the other hand, free SO2 remained relatively stable in commercially bottled New Zealand Sauvignon blanc wines after an initial drop in concentration due to oxygen introduced at bottling.
Acknowledgments
Acknowledgments: This research was supported by the New Zealand Foundation for Research, Science & Technology (grant UOAX0404) and New Zealand Winegrowers.
The authors acknowledge Pernod Ricard New Zealand and Villa Maria Estate for their assistance with the supply of Sauvignon blanc wines and also Pernod Ricard New Zealand for the use of a FIAstar 5000 Analyzer for sulfur dioxide analyses at their winery in Auckland.q
- Received February 1, 2011.
- Revision received June 1, 2011.
- Accepted June 1, 2011.
- Published online December 1969
- © 2011 by the American Society for Enology and Viticulture