Yeast alter micro-oxygenation of wine: oxygen consumption and aldehyde production

Transcription

1 Research Article Received: 17 July 2016 Revised: 27 January 2017 Accepted article published: 9 February 2017 Published online in Wiley Online Library: (wileyonlinelibrary.com) DOI /jsfa.8252 Yeast alter micro-oxygenation of wine: oxygen consumption and aldehyde production Guomin Han, a,c* Michael R Webb, b Chandra Richter, d Jessica Parsons d and Andrew L Waterhouse b* Abstract BACKGROUND: Micro-oxygenation (MOx) is a common winemaking treatment used to improve red wine color development and diminish vegetal aroma, amongst other effects. It is commonly applied to wine immediately after yeast fermentation (phase 1) or later, during aging (phase 2). Although most winemakers avoid MOx during malolactic (ML) fermentation, it is often not possible to avoid because ML bacteria are often present during phase 1 MOx treatment. We investigated the effect of common yeast and bacteria on the outcome of micro-oxygenation. RESULTS: Compared to sterile filtered wine, Saccharomyces cerevisiae inoculation significantly increased oxygen consumption, keeping dissolved oxygen in wine below 30 μgl 1 during micro-oxygenation, whereas Oenococcus oeni inoculation was not associated with a significant impact on the concentration of dissolved oxygen. The unfiltered baseline wine also had both present, although with much higher populations of bacteria and consumed oxygen. The yeast-treated wine yielded much higher levels of acetaldehyde, rising from 4.3 to 29 mg L 1 during micro-oxygenation, whereas no significant difference was found between the bacteria-treated wine and the filtered control. The unfiltered wine exhibited rapid oxygen consumption but no additional acetaldehyde, as well as reduced pyruvate. Analysis of the acetaldehyde-glycerol acetal levels showed a good correlation with acetaldehyde concentrations. CONCLUSION: The production of acetaldehyde is a key outcome of MOx and it is dramatically increased in the presence of yeast, although it is possibly counteracted by the metabolism of O. oeni bacteria. Additional controlled experiments are necessary to clarify the interaction of yeast and bacteria during MOx treatments. Analysis of the glycerol acetals may be useful as a proxy for acetaldehyde levels Society of Chemical Industry Keywords: microbial metabolism; oxygen; acetaldehyde; pyruvic acid; glycerol acetal INTRODUCTION Oxidation is an essential process for the development of red wine color, particularly in tannic reds. Although some white wines can age well with the almost hermetic screwcap bottle seals available today, oxygen has been considered as a factor essential to wine aging; as noted by Pasteur, c est l oxygene qui fait le vin. 1 For many red wines, the addition of low levels of oxygen, micro-oxygenation (MOx), or other practices that expose wine to scant oxygen, such as barrel aging, are carried out aiming to improve color stability and reduce the perceptions of tannin and vegetal aroma. 2 The basic chemistry of wine oxidation was initially described as a reaction between oxygen and a catechol phenolic to yield peroxide and a quinone. The peroxide reacts further with alcohols (and other wine constituents) to produce aldehydes and ketones. 3 There have been many recent investigations of oxidation chemistry that have clarified many important details in the reaction pathways and control factors in the reactions, and a few reviews canleadtomuchoftheliterature, 4 7 although this is an active field with frequent new reports. Aldehydes and ketones are integral fermentation metabolites, as well as common byproducts of chemical oxidation of major wine components. 8,9 In red wines, aldehydes are key to the color stabilization observed with wine oxidation. 2 These products lead to the formation of ethyl-bridged tannins and tannin-anthocyanins products, 10 and also give rise to many other reactions, including those with malvidin-3-glucoside to produce a pyranoanthocyan, vitisin B. 11 The pyranoanthocyanins are resistant to SO 2 bleaching and ph-induced color changes 12, which is the characteristic property of stable wine pigments. In addition, acetaldehyde in wine reacts with the glycerol present, resulting in the formation of four heterocyclic acetal alcohols that have been associated with aging and the aroma of aged madeiras and ports. 13 Correspondence to: AL Waterhouse, Department of Viticulture and Enology, One Shields Avenue, University of California, Davis, CA 95616, USA. alwaterhouse@ucdavis.edu; or G Han, School of Enology, Binzhou Medical University, Yantai, Shandong , PR China. hanguomin@bzmc.edu.cn. a School of Enology, Binzhou Medical University, Yantai, Shandong, PR China b Department of Viticulture and Enology, University of California, Davis, CA, USA c College of Enology, Northwest A&F University, Yangling, Shaanxi, PR China d Viticulture, Chemistry and Enology, E&J Gallo Winery, Modesto, CA, USA J Sci Food Agric (2017) Society of Chemical Industry

2 G Han et al. Oxygen is also important for ensuring the viability of yeast during fermentation, and adding oxygen during fermentation can promote yeast growth, as well as confer yeast with a higher resistance to ethanol and a higher fermentation activity, and also decrease the production of sulfur compounds. 14 However, the relationship between oxygen and yeast during MOx, after the completion of alcohol fermentation, is poorly documented. On the other hand, fino style sherry is made by yeast acting on dry wine in the presence of oxygen to produce acetaldehyde from ethanol. 15 It has also been observed that MOx in the presence of lees can improve the color stability of red wine pigments Oxygen is reported to delay the onset of malolactic fermentation (MLF) 19 and, because of this effect, it is typically recommended that early MOx treatments should be completed prior to inoculation with malolactic bacteria (MLB) or after ML fermentation is complete. Additionally, MLB consume acetaldehyde, which is an essential intermediate in color stabilization, 20 and so conducting MOx when bacteria are present in high populations may decrease the desired color stabilizing effect of acetaldehyde. Although MOx is widely used, its interaction with the yeast or bacteria in wine is poorly understood. In the present study, the interaction between MOx and the most common wine microbes is addressed directly using yeast or bacteria inoculations, with the results being compared against a sterile filtered control and an unfiltered reference wine, typical of wine production. A number of oxidation and color parameters were measured. MATERIALS AND METHODS Wine samples During the 2012 vintage, a California north coast Cabernet Sauvignon wine that had unexpectedly started MLF was obtained from E&J Gallo Winery (Modesto, CA, USA). The base parameters (mean ± SD) of wine at the start of the experiment were: ethanol content, ± 0.30% v/v; ph, 3.93 ± 0.01; titratable acidity, 6.70 ± 0.09 g L 1 expressed as tartaric acid; volatile acidity, 0.43 ± 0.05 g L 1 expressed as acetic acid; malic acid, 1360 mg L 1 ; free SO 2,12mgL 1 ;andtotalso 2,21mgL 1. The wine selected had just completed yeast fermentation and had not been treated with micro-oxygenation. By the time that it was transferred to the research facility, it had spontaneously started malolactic fermentation. Although it is not conventional to treat wines with MOx during MLF, winemakers report that these fermentations often start during post-fermentation (phase 1) MOx treatment and so this situation was not exceptional. In this instance, the unplanned MLF in the unfiltered wine provided some possible insight into the proper management of MOx. For the experiment, 120 L of wine was left unfiltered and 360 L of wine was sterile filtered, which was membrane filtered using an AcroPak 1000 filter (1.0, 0.5 and 0.45 μm filters) (Pall, New York, NY, USA) in a single pass directly into tanks under nitrogen-protection conditions. For the filtered wine, the first 120 L was treated with Saccharomyces cerevisiae strain EC1118 (Lallemand Inc., Montreal, Canada.) (cultured in 50 ml of grape juice for 1 day initially at room temperature, and then gradually mixed with 500 ml of Cabernet Sauvignon wine for adaptation to alcohol, after which the yeast was collected by centrifugal precipitation and then added to the experimental wine 3 days later) at 0.1 g L 1 ;the second 120 L was treated with g of Viniflora Oenos (Hansen, Hoersholm, Denmark) rehydrated in 10 ml of distilled water; and the third portion was not treated to yield four different treatments: filtered (F), filtered with yeast (F + Y), filtered with MLB (F + B) and Figure 1. Schematic of MOx hardware. unfiltered (unf). Each wine was then treated with Mox (+M) or without. The treatments were carried out in duplicate at 18 Cin 23-L stainless steel tanks. All tanks were filled by weight with 22 kg of wine per tank. Micro-oxygenation trials MOx was carried out using a technique in which 23-L cylindrical stainless steel tanks were fitted with a 2.5-inch Tri Clover (Alfa Laval, Guildford, UK) compatible ferrule. This was sealed with a Tri Clover cap using a clamp. The cap was modified by drilling holes in square pattern and welding 1 4 inch stainless steel tubing to the cap as shown in Fig. 1. One tube ended in a Swagelok connector (Swagelok, Solon, OH, USA) that was sealed to fluorinated ethylene-propylene tubing (FEP-188x250; Ozone Solutions, Inc., Hull, IA, USA), which was pressurized with oxygen (65 psi, 450 kpa) to allow for a controlled diffusion of oxygen into the tanks. 21 Samples were removed though a second opening sealed with valve and terminating in rubber septum ensuring air exclusion from the tank when sampling. Oxygen was monitored by attaching a 5-mm nuclear magnetic resonance (NMR) tube to the third opening and placing a luminescent Nomasense dot (Nomacorc, Zebulon, NC, USA) toward the bottom (see below). The tanks also included an overpressure valve to release pressure, if needed. Each tank was stirred using magnetic stir bars to ensure that the oxygen was thoroughly mixed into the wine as diffusion occurred. The successful use of this type of equipment for small scale MOx treatment has been described previously. 6,22 The key issue is how much oxygen is transferred to the liquid over time and a diffusion system accomplishes this very well on a small scale. The control tanks were sealed with plain Tri Clover caps. Oxygen treatment was calibrated in the assembled tanks by pressurizing the treatment tube with oxygen and then monitoring the appearance of oxygen in a deoxygenated 12% ethanol solution. The dissolved oxygen concentration was detected non-invasively via a flat-bottomed NMR tube fitted on the end with a Presens Pst-3 luminescent oxygen sensing dot (PreSens Precision Sensing GmbH, Regensburg, Germany). Oxygen was then measured using the fiber-optic probe and a NomaSense O2 P6000 (Nomacorc). Testing showed that the amount of oxygen entering the tanks was consistent day to day with mixing. Oxygen was added at a rate of 30 mg L 1 month 1 over the treatment wileyonlinelibrary.com/jsfa 2017 Society of Chemical Industry J Sci Food Agric (2017)

3 Yeast alter micro-oxygenation of wine period of 18 days, typical of phase 1 MOx in production (18 mg L 1 added in total). During MOx treatment, oxygen concentrations in wine were monitored approximately every 24 h. 23 Wine sampling Microbial samples for plating and carbonyl compounds were analyzed at five stages: initial sampling (1 day for 4 h after inoculation), as well as 4, 9, 14 and 18 days after the initiation of MOx treatment. Acetaldehyde acetals were analyzed at three stages: initial sampling (1 day), 9 and 18 days after initiation of MOx treatment. Duplicate samples were obtained from the duplicate treatments, and each wine sample was analyzed in duplicate. Some samples were frozen at 80 C to create a library of samples for later testing. Microbial plating Each sample was plated on differential media: YPD (10% yeast extract, 20% peptone, 20% dextrose and 2% agar), ML [MRS media (Difco, Franklin Lakes, NJ, USA): 55% MRS broth powder, 10% fructose, 0.05% Tween 80 and 20% agar plus 0.05 g L 1 Delvocid (DSM Food Specialties, Delft, The Netherlands) and 25 mg L 1 kanamycin] and NL (MRS media: 55% Difco MRS broth powder, 10% fructose, 0.05% Tween 80 and 20% agar plus 0.05 g L 1 Delvocid and 0.05 g L 1 nisin). Plates were incubated at 28 Candread after 5 days (YPD) or 7 days (ML and NL). Carbonyls analysis For analyzing carbonyls, a recently improved liquid chromatography method based on the reaction with 2,4-dinitrophenyl hydrazine (DNPH) was used. 24 Briefly, derivatizations were conducted manually in 2.0-mL glass vials (15 85 mm; Fisher Scientific Co., Pittsburgh, PA, USA) with Teflon (Dupont, Wilmington, DE, USA) lined caps. Sample aliquots (100 μl) were dispensed into a vial, followed by 20 μl of freshly prepared 1120 mg L 1 sulfur dioxide solution, and then 20 μl of 25% sulfuric acid (v/v) was added, followed by 140 μl of8gl 1 DNPH reagent. The solution was allowed to react for 15 min at 65 C. Derivatized samples were analyzed by HPLC-DAD-MS within 10 h, storing samples at room temperature. The chromatographic conditions used were: sample injection volume 15 μl; flow rate 0.75 ml min 1 ; column temperature 35 C; mobile-phase solvents (A) 0.5% (v/v) formic acid in water and (B) acetonitrile; gradient elution protocol (v/v), 35 60% B (8 min), 60 90% B (13 min), 90 95% B (15 min, 2-min hold), 95 35%B (16 min, 4-min hold), with a total run time of 20 min. The identification of the observed carbonyls was based on their retention time compared with those of the carbonyl standards tested at 365 nm, as well as their mass spectral characteristics. Acetal analysis Acetal analysis was performed as described previously. 25 Wine samples were stirred with anhydrous sodium sulfate (0.1 g ml 1 ) until the salt was completely dissolved. Wine aliquots (2 ml) were dispensed into clear glass vials (19 65 mm). Synthesized 2 H 3 -deuterium-labelled acetals (25 μl of 800 ppm in water) were spiked into wine sample aliquots or synthetic wine. Wine samples were extracted one time with 2 ml of ethyl acetate. Extracts were dried over Na 2 SO 4 (0.5 1 g) and placed in amber vials (2 ml) for analysis by gas chromatography-mass spectroscopy on a DB-Wax column (Agilent Technologies Inc., Santa Clara, CA, USA). One microliter of the sample extract was injected at 240 C. The molecular (m/z 117), qualitative (m/z 103) and quantitative (m/z 87/88) ions for the four glycerol acetal isomers were measured in selective ion monitoring mode (SIM). Quantitation of the deuterated internal standards was performed by measuring the m/z 91 and 92 ions in SIM. Statistical analysis Analysis of variance was used for data analysis and Tukey s honestly significant difference test procedure was used to make multiple comparisons. Correlation and regression analysis between acetaldehyde and acetaldehyde acetals were analyzed using a standard correlation and regression analysis. These analyses were performed using SPSS, version 19.0 (IBM Corp., Armonk, NY, USA), which was also used to evaluate the linear relationship between acetaldehyde and acetaldehyde acetals. All data are the means of four values (two experimental replicates; the averages of two analytical replicates). RESULTS AND DISCUSSION Oxygen levels during treatment The concentration of dissolved oxygen in the wine of each treatment tank is shown in Fig. 2. In the Mox-treated wines (Fig. 2A, B), the oxygen levels in the filtered (F + M) and the filtered wine with bacteria (F + B + M) showed very similar behavior. After the second day, oxygen concentrations dropped slowly, until day 6, and then started to rise and continued to do so until the end of the experiment. The almost identical change in oxygen concentration in the filtered and bacteria-inoculated samples indicates that the bacteria had a negligible interaction with the oxygen. Because oxygen was being added at 1 mg L 1 day 1, the almost slight drop up to day 6 actually reflects that approximately one oxygen saturation from air (8 mg L 1 ) was actually absorbed by the wine with a drop of approximately 0.2 mg L 1. In addition, the rise of these two treatments to approximately 2.2 mg L 1 on day 18 reflects the fact that the wine has consumed approximately 16 mg L 1 of oxygen during that time, or 0.5 mmol L 1. On the other hand, the two MOx treatments with yeast present, the yeast inoculated sample (F + Y + M) and the unfiltered wine (unf + M),bothshowedarapiddropinoxygenconcentration,with the yeast inoculated wine immediately consuming the oxygen by day 2 and the unfiltered wine reaching zero at day 4. In both cases, the levels never rose above zero for the rest of the experiment. This behavior suggests an alternative mechanism for oxygen consumption, which is different from the chemical pathways available to the filtered wine, and the behavior of the F + Y + M treatment points to yeast as the responsible microbe. This would not be too surprising because fino Sherry is made by yeast oxidizing ethanol to acetaldehyde. 15 In the case of the unfiltered wine, the presence of lees may be a complicating factor because lees are known to consume oxygen, and the oxygen consumption could be attributable to the presence of nonviable yeast in lees However, Table 1 showsthat S. cerevisiae survived in unf + M wine during MOx, whereas unf wine without MOx lost all viable yeast. Thus, it could be concluded that the surviving, viable yeast in unf + M wine did consume oxygen. Future experiments should clarify the role of lees versus viable yeast in consuming oxygen during MOx treatments. In addition, the yeast in the unf treatment were in a different metabolic state than those added in the F + Y + Mtreatment at the beginning of the treatment, and future studies should consider comparing fresh versus residual yeast after fermentation. It is possible that, during the treatment, the yeast in the two treatments reached similar states. The yeast activity contrasts with the J Sci Food Agric (2017) 2017 Society of Chemical Industry wileyonlinelibrary.com/jsfa

4 G Han et al. Figure 2. The evolution of dissolved oxygen in wine with and without micro-oxygenation. (A, B) The evolution of dissolved oxygen in wine with micro-oxygenation; (C, D) The evolution of dissolved oxygen in wine without micro-oxygenation; F, wine with sterile filtration; F + M, F wine with micro-oxygenation treatment; F + Y,FwinewithS. cerevisiae inoculation; F + Y + M, F + Y wine with micro-oxygenation treatment; F + B, F wine with O. oeni inoculation; F + B + M, F + B with micro-oxygenation treatment; unf, wine without sterile filtration; unf + M, unf wine with micro-oxygenation treatment. F + B + M treatment, which appears to show, as noted above, that the Oenococcus bacteria do not consume any oxygen under these conditions. In the wine without any oxygen treatment (Fig. 2C, D), the residual oxygen from the transfer process was quickly depleted, reaching 5 15 μgl 1 by day 4 in all tanks and remaining in that range for the balance of the experiment. Effect of micro-oxygenation on microbial populations and malic acid Microbial population changes occurred over the course of the MOx treatment and were quantified (Table 2). There were no culturable yeast or bacteria in the wine after membrane filtration andnomorethan20cfuml 1 Oenococcus oeni growth was found in the Mox-treated filtered wine (F + M) after 18 days. It is also worth noting that no acetic acid bacteria were detected in any of the wines over the study period, which can consume oxygen and release acetic acid. These data demonstrate that this sterile filtration was successful, with only minimal microbial contamination. The F + BandF+ B + M wines were inoculated with a population suitable to stimulate malolactic fermentation, 10 6 CFU ml As noted in Table 1, the populations were sustained but not stable and varied widely during the experiment, increasing up to day 14 and then dropping, with no notable differences between the controlandoxygen-treatedwines.thebacteriawerealsofairly active in terms of a malolactic fermentation, consuming approximately 400 mg L 1 malic acid in both wines (Fig. 3), although this was hardly as active as the unf samples where the malic acid almost depleted, introducing a variable that should be better controlled in the future. However, it appears that, under these conditions, the bacteria appeared to not interact with the oxygen in any manner because the oxygen had no effect on the metabolism of malic acid or the bacterial populations, and the presence of the bacteria did not substantially alter the oxygen concentration. Other studies have shown that, in the presence of oxygen, Oenococcus do produce more diacetyl than without, 30 and so there may be some interaction involving oxygen. The bacterial population in the unfiltered wines, which had spontaneously started malolactic fermentation prior to the experiment, was very high and remained high throughout the 18 days. This high population was also very effective in metabolizing the residual malic acid, consuming over 1000 mg L 1, and dropping to wileyonlinelibrary.com/jsfa 2017 Society of Chemical Industry J Sci Food Agric (2017)

5 Yeast alter micro-oxygenation of wine Table 1. Microbial populations during each treatment showing concentration of cells present (CFU ml 1 ) Wines Microorganism Initial wines Day 4 Day 9 Day 14 Day 18 F ND ND ND ND ND ND F + M O. oeni ND F + Y S. cerevisiae E F + Y + M S. cerevisiae E 10 3 F + B O. oeni > > F + B + M O. oeni > > unf S. cerevisiae ND ND O. oeni unf + M S. cerevisiae O. oeni ND, not detected. F, wine with sterile filtration; F + M, F wine with micro-oxygenation treatment; F + Y, F wine with S. cerevisiae inoculation; F + Y + M, F + Y wine with micro-oxygenation treatment; F + B, F wine with O. oeni inoculation; F + B + M, F + B with micro-oxygenation treatment; unf, wine without sterile filtration; unf + M, unf wine with micro-oxygenation treatment. Table 2. Concentration of DO (μgl 1 ), selected carbonyls (mg L 1 ) and acetals after 18 days of treatment Wine DO Pyruvic acid Acetaldehyde cdl tdl cd td unf + M ± 5.74 a ± 0.73 a 3.01 ± 0.34 a 1.35 ± a 1.14 ± a ± a ± a F + M ± b ± 4.65 b 6.60 ± 0.96 a 1.41 ± ac 1.16 ± a ± a ± a F + Y + M ± 2.22 a ± 0.37 bc ± 4.69 b 2.11 ± b 1.52 ± b ± b ± b F + B + M ± b ± 2.64 c 6.79 ± 0.84 a 1.40 ± a 1.14 ± a ± a ± a unf ± 2.63 ac ± 1.35 d 2.49 ± 0.34 a 1.36 ± a 1.12 ± a ± a ± a F 6.04± 0.38 c 48.1 ± 3.74 b 4.89 ± 0.37 a 1.38 ± a 1.14 ± a ± a ± a F + Y ± 1.81 d ± 5.63 b 8.21 ± 1.44 γc 1.55± c 1.24 ± a ± a ± a F + B 9.06± 0.33 βd 52.1 ± 3.83 bc 3.43 ± 0.60 αa 1.37± a 1.12 ± a ± a ± a F, wine with sterile filtration; F + M, F wine with micro-oxygenation treatment; F + Y,, F wine with S. cerevisiae inoculation; F + Y + M, F + Ywinewith micro-oxygenation treatment; F + B, F wine with O. oeni inoculation; F + B + M, F + B with micro-oxygenation treatment; unf, wine without sterile filtration; unf + M, unf wine with micro-oxygenation treatment; DO, dissolved oxygen; cdl, cis-dioxolane; tdl, trans-dioxolane; cd, cis-dioxane; td, trans-dioxane. Different letters indicate statistical differences (P < 0.05). Latin lowercase letters (a, b, c) are used to compare MOx wines and nomox wines with different treatments. approximately 100 mg L 1 by day 18. However, as noted above, the bacteria in these samples were also not affected by the presence or lack of oxygen, at least in terms of the culturable populations. Thus, it would appear that with the higher bacterial population in the unfiltered wine samples (unf and unf + M), malic acid was rapidly consumed, whereas, in the inoculated samples, only a modest fraction of the malic acid was metabolized. Most winemakers would not start MOx treatment during malolactic fermentation, although the oft-stated reason is that this could stimulate acetic acid formation by the bacteria. In the present study, we did not detect the formation of new acetic acid in any sample. However, winemakers also report that ML fermentations often start without prompting during Phase 1 MOx treatment, although they do not generally see any problems arising as a result. As described below, there may be a more important reason to avoid MOx treatment when bacteria are present and active. Although the presence of oxygen appeared to have little effect on the bacteria, there was a different response by yeast. In the inoculated samples, both tanks supported yeast until the end of the experiment, although the MOx wines had approximately double the population. In the unfiltered wine, yeast were no longer culturable at the end of the experiment, whereas there was a viable population in the MOx-treated wines. Comparing the effect of MOx in these two treatments strongly suggests that the yeast were consuming oxygen to enhance survival. It should be noted that, although there were viable populations in these samples, the cell density was between 1.5 and CFU ml 1,which is a very small fraction of the density observed during a sugar fermentation. This observation is difficult to compare to sherry production where the yeast are present as a biofilm on the wine surface and populations in the liquid are generally not reported. 31 Changes in acetaldehyde and other carbonyls With the micro-oxygenation treatment, there was some production of acetaldehyde in the filtered wine via oxidation (F + M), although the amount was small, rising from 4.3 to 6.6 mg L 1 for an increase of 2.3 mg L 1 or mmol L 1 (Fig. 4A). This was a surprisingly small amount considering how much oxygen was consumed (16 mg L 1 or 0.5 mmol L 1 ) clearly indicating that other oxidation reactions occur. A very similar pattern was observed for the bacteria-inoculated MOx wine (F + B + M), where there was a small increase in acetaldehyde during the 18 days, rising from 4.3 To 6.8 mg L 1. This again supports the observations noted above, that bacteria (O. oeni) had little interaction with oxygen. With yeast present, however, the outcome was very different. Although there was some increased acetaldehyde with yeast and no oxygen (Fig. 4B), the amount produced increased substantially J Sci Food Agric (2017) 2017 Society of Chemical Industry wileyonlinelibrary.com/jsfa

6 G Han et al. Figure 3. Malic acid concentration in the wines after 18 days of treatment. F, wine with sterile filtration; F + M, F wine with micro-oxygenation treatment; F + Y, F wine with S. cerevisiae inoculation; F + Y + M, F + Ywine with micro-oxygenation treatment; F + B, F wine with O. oeni inoculation; F + B + M, F + B with micro-oxygenation treatment; unf, wine without sterile filtration; unf + M, unf wine with micro-oxygenation treatment. two cases is likely caused by the consumption of both free and SO 2 -bound acetaldehyde by bacteria, as reported previously, 32,33 and, in the filtered wine, the small drop is likely from reactions with flavonoids or other nucleophiles. Thus, the presence of bacteria had a small but significant (P < 0.05) effect on the concentration of acetaldehyde during the 18 days of treatment (Table 2). The yeast treatment with no oxygen did show a substantially increased level of acetaldehyde, from 4.3 to 8.2 mg L 1 (90%). This production may have been a result of the condition of the yeast prior to inoculation, where the yeast are grown under aerobic conditions and had probably accumulated cofactors to facilitate ethanol oxidation. 34 In the case of the unfiltered wine, there was a large drop in pyruvate with oxygen (Fig. 4C) and a larger drop in the anaerobic treatment (Fig. 4D). It would appear that the high density of bacteria in these treatments metabolized the pyruvic acid; 35 however, pyruvic acid levels were not affected so much in F + BwineandF+ B + M wine and, in contrast, there was a small increase in the bacterial MOx treatment (Fig. 4C). This would be an important finding to follow-up because pyruvate reacts with anthocyanins to create vitisin A and related compounds, which are important red wine pigments. with oxygen present, rising from 4.3 to 29 mg L 1, and was significantly higher (P < 0.05) than other MOx treatment after 18 days (Table 2). The increase of 25 mg L 1 is equal to 0.59 mmol L 1,and so the yeast are very efficient in terms of utilizing the two equivalents of oxidizing capacity of 0.5 mmol L 1 oxygen, despite the modest density of approximately 5000 CFU ml 1. The most surprising result, however, was that, with the unfiltered wine (unf + M), which had both yeast and bacteria present (Table 1), the oxygen was rapidly consumed, although there was no increase in acetaldehyde. One reason may be that the oxygen was utilized by the dead yeast cells, although, as shown in Table 1, there were considerable active yeast in unf + Mwine.IntheunF wine without oxygen, yeast viability declined, strongly suggesting that the presence of oxygen was affecting yeast metabolic activity in the unf + M treatment. If that were the case, then the reduced acetaldehyde is best explained as a result of the bacteria consuming the yeast-generated acetaldehyde. These bacteria have previously been observed to consume acetaldehyde, leading to reduced color stabilization in ML-treated wines. 32,33 This same high population of O. oeni rapidly consumed all the malic acid (Fig. 2) and so it may be reasonable to expect rapid metabolism of other compounds as well. This finding is most revealing because it suggests that, when undertaking a phase I MOx treatment, it may be important to take both yeast and bacterial populations into account. In particular, the presence of a high bacterial population may well negate the desired effect of the oxidation, removing the acetaldehyde and other carbonyls that would react with anthocyanins to produce stabilized pigments. However, this initial observation must be corroborated with confirmatory controlled experiments. It should also be noted that a post ML MOx treatment would ordinarily have involved an SO 2 addition, and that addition would likely eliminate bacteria, or nearly so, greatly diminishing their ability to consume any acetaldehyde produced during MOx treatment. Acetaldehyde levels did not change as much as the F + Y + M treatment in any of the control treatments with no oxygen (Fig. 4B). The filtered control wine had a very small drop of 7%, whereas the bacterial and unfiltered treatments showed modest (21% and 24%, respectively) decreases in levels (Fig. 4B). The drop in these latter The relationship between acetaldehyde acetals and acetaldehyde Heterocyclic acetal alcohols have been identified and reported as potential age markers of Madeira wine. 36 However, there have been few reports about the influence of MOx treatment on these compounds. 21 Table 2 shows that the levels of these acetals varied between 0.3 and 2.5 mg L 1, and these levels appeared to increase along with acetaldehyde levels, especially F + Y + M wines, which were significantly higher (P < 0.05) than other MOx wines, and so the acetal and acetaldehyde levels were compared (Fig. 5). Pearson correlation analysis between single acetals and acetaldehyde (Fig. 5) showed that cis-dioxolane, trans-dioxolane, cis-dioxane and trans-dioxane were all positively correlated with acetaldehyde at a very significant level (P < 0.001). Regression analysis also demonstrated a strong relationship (P < 0.001) between single acetal and acetaldehyde, and all models produced a coefficient of determination (r 2 ) equal to or greater than This showed that these four selected heterocyclic acetal alcohols were all well correlated with acetaldehyde. This close correlation also shows that it may be possible to relate the dioxane or dioxolane levels to acetaldehyde levels in wine, as reported recently. 37 Acetaldehyde was the main oxygenation product during Mox; therefore, these oxidation markers observed in Madeira wine may also be useful indicators of the cumulative extent of oxidation, which can reflect the oxidation process of wine timely and accurately, such as MOx, in other wines. From the data provided in Table 2, most acetaldehyde acetals were in the dioxolane form, accounting for 78% of the total, which also had higher r 2 in the form of model tested (Fig. 5). These results revealed that dioxolane may better reflect the concentration of acetaldehyde, the main product of wine oxidation. Thus, cis-dioxolane and trans-dioxolane were better indicators than cis-dioxane and trans-dioxane during wine micro-oxygenation. CONCLUSIONS It is clear that wine micro-oxygenation is not solely a chemical process. Yeast and bacteria also can have a major impact on the treatment. Thus, yeast can be considered as a catalyst to accelerate wileyonlinelibrary.com/jsfa 2017 Society of Chemical Industry J Sci Food Agric (2017)

7 Yeast alter micro-oxygenation of wine Figure 4. The evolution of carbonyls in wine during micro-oxygenation. (A, B) acetaldehyde; (C, D) pyruvic acid; F, wine with sterile filtration; F + M, F wine with micro-oxygenation treatment; F + Y, F wine with S. cerevisiae inoculation; F + Y + M, F + Y wine with micro-oxygenation treatment; F + B, F wine with O. oeni inoculation; F + B + M, F + B with micro-oxygenation treatment; unf, wine without sterile filtration; unf + M, unf wine with micro-oxygenation treatment. microbes present, and future testing should be able to provide the detailed information necessary for creating improved winemaking management tools. Finally, analysis of the dioxolanes may be a facile method of analyzing acetaldehyde in wine. ACKNOWLEDGEMENTS G. Han was supported by a grant from the China Scholarship Council (grant number: ). We also thank E&J Gallo Winery for their support of this project. Figure 5. The correlation of acetaldehyde and acetals in wine during micro-oxygenation across all treatments. Means of duplicate analyses at days 1, 9 and 18. the formation of acetaldehyde, whereas bacteria may prove to negate the effect. This strongly suggests that proper management of the micro-oxygenation process requires an assessment of the REFERENCES 1 Pasteur ML, Etudes sur le vin. Librairie F. Savy, Paris (1875). 2 Gómez-Plaza E and Cano-López M, A review on micro-oxygenation of red wines: Claims, benefits and the underlying chemistry. Food Chemistry 125: (2011). 3 Wildenradt HL and Singleton VL, The production of aldehydes as a result of oxidation of polyphenolic compounds and its relation to wine aging. Am J Enol Vitic 25: (1974). 4 Danilewicz JC, Review of oxidative processes in wine and value of reduction potentials in enology. Am J Enol Vitic 63:1 10 (2012). 5 Waterhouse AL and Laurie VF, Oxidation of wine phenolics: a critical evaluation and hypotheses. Am J Enol Vitic 57: (2006). J Sci Food Agric (2017) 2017 Society of Chemical Industry wileyonlinelibrary.com/jsfa

8 G Han et al. 6 Gomez-Plaza E and Cano-Lopez M, A review on micro-oxygenation of red wines Claims, benefits and the underlying chemistry. Food Chem 125: (2011). 7 Ugliano M, Oxygen contribution to wine aroma evolution during bottle aging. J Agric Food Chem 61: (2013). 8 Prokai L, Szarka S, Wang X and Prokai-Tatrai K, Capture of the volatile carbonyl metabolite of flecainide on 2,4-dinitrophenylhydrazine cartridge for quantitation by stable-isotope dilution mass spectrometry coupled with chromatography. J Chromatogr A 1232: (2012). 9 Nikolantonaki M, Magiatis P and Waterhouse AL, Direct analysis of free and sulfite-bound carbonyl compounds in wine by two-dimensional quantitative proton and carbon nuclear magnetic resonance spectroscopy. Anal Chem 87: (2015). 10 Atanasova V, Fulcrand H, Cheynier V and Moutounet M, Effect of oxygenation on polyphenol changes occurring in the course of wine-making. Anal Chim Acta 458:15 27 (2002). 11 Bakker J and Timberlake CF, Isolation, identification, and characterization of new color-stable anthocyanins occurring in some red wines. J Agric Food Chem 45:35 43 (1997). 12 Fulcrand H, Benabdeljalil C, Rigaud J, Cheynier V and Moutounet M, A new class of wine pigments generated by reaction between pyruvic acid and grape anthocyanins. Phytochemistry 47: (1998). 13 Perestrelo R, Barros AS, Câmara JS and Rocha SM, In-depth search focused on furans, lactones, volatile phenols, and acetals as potential age markers of Madeira wines by comprehensive two-dimensional gas chromatography with time-of-flight mass spectrometry combined with solid phase microextraction. JAgric Food Chem 59: (2011). 14 Du Toit W, Marais J, Pretorius IS and Du Toit M, Oxygen in must and wine: a review. South Afr J Enol Vitic 27:76 94 (2006). 15 Pozo-Bayon MA and Moreno-Arribas MV, Sherry wines, in Advances in Food and Nutrition Research, Vol. 63: Speciality Wines, ed. by Jackson RS. Elsevier Academic Press Inc., San Diego, CA, pp (2011). 16 Del Barrio-Galán R, Pérez-Magariño S and Ortega-Heras M, Techniques for improving or replacing ageing on lees of oak aged red wines: the effects on polysaccharides and the phenolic composition. Food Chem 127: (2011). 17 Sartini E, Arfelli G, Fabiani A and Piva A, Influence of chips, lees and micro-oxygenation during aging on the phenolic composition of a redsangiovesewine. Food Chem 104: (2007). 18 Baiano A, Scrocco C, Sepielli G and Del Nobile MA, Wine processing: a critical review on physical, chemical, and sensory implications of innovative vinification procedures. Crit Rev Food Sci Nutr 56: (2016). 19 Kunkee RE, Some roles of malic acid in the malolactic fermentation in wine making. FEMS Microbiol Lett 88:55 72 (1991). 20 Moreno-Arribas MV and Polo MC, Wine Chemistry and Biochemistry. Springer, New York (2009). 21 Gambuti A, Han G, Peterson AL and Waterhouse AL, Sulfur Dioxide and Glutathione Alter the Outcome of Microoxygenation. American Journal of Enology and Viticulture (2015) in press. 22 Dykes S, The Effect of Oxygen Dosage Rate on the Chemical and Sensory Changes Occurring During Micro-Oxygenation of New Zealand Red Wine. University of Auckland, Auckland (2007). 23 Diéval JB, Vidal S and Aagaard O, Measurement of the oxygen transmission rate of co-extruded wine bottle closures using a luminescence-based technique. Packag Technol Sci 24: (2011). 24 Han G, Wang H, Webb MR and Waterhouse AL, A rapid, one step preparation for measuring selected free plus SO 2 -bound wine carbonyls by HPLC-DAD/MS. Talanta 134: (2015). 25 Peterson AL, Gambuti A and Waterhouse AL, Rapid analysis of heterocyclic acetals in wine by stable isotope dilution gas chromatography mass spectrometry. Tetrahedron 71: (2015). 26 Salmon JM, Fornairon-Bonnefond C and Mazauric JP, Interactions between wine lees and polyphenols: influence on oxygen consumption capacity during simulation of wine aging. J Food Sci 67: (2002). 27 Salmon J-M, Fornairon-Bonnefond C, Mazauric J-P and Moutounet M, Oxygen consumption by wine lees: impact on lees integrity during wine ageing. Food Chem 71: (2000). 28 Fornairon-Bonnefond C and Salmon J-M, Impact of oxygen consumption by yeast lees on the autolysis phenomenon during simulation of wine aging on lees. J Agric Food Chem 51: (2003). 29 Costello P, The chemistry of malolactic fermentation, in Malolactic Fermentation in Wine: Understanding the Science and the Practice,ed. by Specht KS. Lallemand Inc., Montréal, pp (2005). 30 Nielsen JC and Richelieu M, Control of flavor development in wine during and after malolactic fermentation by Oenococcus oeni. Appl Environ Microbiol 65: (1999). 31 Alexandre H, Flor yeasts of Saccharomyces cerevisiae their ecology, geneticsandmetabolism. Int J Food Microbiol 167: (2013). 32 Osborne J, Mira de Orduna R, Pilone G and Liu SQ, Acetaldehyde metabolism by wine lactic acid bacteria. FEMS Microbiol Lett 191:51 55 (2000). 33 Osborne J, Dubé Morneau A and Mira de Orduna R, Degradation of free and sulfur-dioxide-bound acetaldehyde by malolactic lactic acid bacteria in white wine. J Appl Microbiol 101: (2006). 34 Mauricio JC, Moreno JJ and Ortega JM, In vitro specific activities of alcohol and aldehyde dehydrogenases from two flor yeasts during controlled wine aging. J Agric Food Chem 45: (1997). 35 Wells A and Osborne JP, Impact of acetaldehyde- and pyruvic acid-bound sulphur dioxide on wine lactic acid bacteria. Lett Appl Microbiol 54: (2012). 36 Pereira AC, Reis MS, Saraiva PM and Marques JC, Analysis and assessment of Madeira wine ageing over an extended time period through GC MS and chemometric analysis. Anal Chim Acta 660:8 21 (2010). 37 Peterson AL and Waterhouse AL, 1 H-NMR: a novel approach to determining the thermodynamic properties of acetaldehyde condensation reactions with glycerol, (+)-catechin, and glutathione in model wine. J Agric Food Chem 64: (2016). wileyonlinelibrary.com/jsfa 2017 Society of Chemical Industry J Sci Food Agric (2017)