Metschnikowia pulcherrima Selected Strain for Ethanol Reduction in Wine: Influence of Cell Immobilization and Aeration Condition

One of the most important problems in the winemaking field is the increase of ethanol content in wine. Wines with high ethanol level negatively affect wine flavor and human health. In this study, we evaluated the use of a selected strain of Metschnikowia pulcherrima in immobilized form and under different aeration conditions, to reduce the ethanol content evaluating the volatile profile of the resulting wines. In a preliminary screening the best conditions regarding free/immobilized cells, static/aerated fermentation and inoculation level were identified. Bench-Top fermentation trials with different aeration conditions showed that the use of M. pulcherrima selected strain with aeration flow of 20 mL/L/min during the first 72 h of fermentation, led an ethanol reduction of 1.38% (v/v) in comparison with Saccharomyces cerevisiae control strain. The analytical profile of the resulting wines did not show any negative feature. Indeed, the concentration of ethyl acetate, that above its sensory threshold impacts negatively the wine sensory profile, was found at an acceptable level. On the other hand, an increase in the concentration of significant fruity and flower compounds was found.


Introduction
Today there is an increasing interest toward the reduction of ethanol content in wines. Indeed, the impact of climate change upon the global production of grapes and the current demand of well-structured and high phenolic content wines determine a generalized increase of ethanol concentration of wines [1][2][3]. Among the various methodologies proposed for the reduction of alcohol content in wine, the microbiological approach seems quite promising to avoid negative variations to the final wine composition [4]. In this regard, Saccharomyces cerevisiae, the main yeast species responsible for alcoholic fermentation during winemaking, is not the best yeast species for reducing alcohol levels in wine [5]. Indeed, in general, S. cerevisiae commercial starter strains showed similar high ethanol yields, so the researchers focused their attention on the development of new S. cerevisiae strains with modified fermentative pathway to obtain wines with reduced alcohol content [6][7][8][9]. On the other hand, non-Saccharomyces yeast species have shown great potential to produce less ethanol content in wine [10][11][12]. Indeed, non-Saccharomyces yeast can divert carbon away from ethanol production affecting ethanol yield, fermentation efficiency, biomass production and final by-products. In addition, respiro-fermentative regulatory mechanisms exhibited by some non-Saccharomyces yeasts (Crabtree negative) are different from that exhibited by S. cerevisiae (Crabtree positive). This metabolic behaviour can be exploited to reduce ethanol production through partial and controlled aeration of the grape juice [2,11,13,14]. A previous work [12] showed that the use of immobilized selected strains of Starmerella bombicola, Metschnikowia pulcherrima, Hanseniaspora osmophila and Hanseniaspora uvarum in sequential fermentation with S. cerevisiae, could be a suitable strategy to reduce the ethanol content in wine. Sequential fermentation allows to exploit the metabolism of non-Saccharomyces yeast while the immobilization procedure allows for high density of cells. In particular, immobilized cells of S. bombicola and M. pulcherrima under anaerobic conditions led an ethanol reduction of 1.6% and 1.4% respectively, exhibiting an increase of some key aroma compounds. Immobilized cells allow a large amount of cells in confined condition determining high reaction rate, specific physiological conditions and possible reuse. Recently, the use of non-Saccharomyces belonging to M. pulcherrima, Torulaspora delbrueckii and Zygosaccharomyces bailii species was investigated, at different aeration conditions to obtain wines with reduced ethanol content but with valuable volatile profile [15].
In the present work we evaluated the use of M. pulcherrima in sequential fermentations (M. pulcherrima/S. cerevisiae) at different aeration conditions to reduce the alcohol content and maintaining, at the same time, a good aromatic profile of wine. After a preliminary screening to optimize the modalities of inoculum and the effect of aeration, bench-top fermentation trials were carried out. The ethanol reduction and analytical composition of the final wines were evaluated.

Yeast Strains
M. pulcherrima DiSVA 269 used in this study was obtained from the Yeast Collection of Department of Life and Environmental Sciences (DiSVA) of Polytechnic University of Marche (Italy) and it was previously selected and evaluated in sequential fermentation trials in immobilized form and free cells under different aeration conditions [12,15]. This yeast was used in sequential fermentation trials with S. cerevisiae commercial strain Lalvin EC1118 (Lallemand Inc., Toulouse, France). S. cerevisiae strain was also used in pure culture as control. These strains were maintained on Yeast extract-Peptone-Dextrose (YPD) agar medium (Oxoid, Basingstoke, UK) at 25 • C for 48-72 h, and then stored at 4 • C.

Biomass Production and Immobilization Procedures
The biomass for free and immobilized trials was obtained using Modified YPD medium (0.5% yeast extract, 0.1% peptone, 2% dextrose, all w/v). M. pulcherrima cells were incubated at 25 • C for 72 h in a rotary shaker (150 rpm). The biomass for cell immobilization (5% wet w/v) was mixed with 2.5% Na-alginate (Carlo Erba, Milan, Italy) following the procedures described in Canonico et al. [12].

Screening Optimization Cell Concentration on Synthetic Grape Juice (SGJ)
A screening to optimize the immobilized cell concentration of M. pulcherrima was carried out in SGJ prepared following the procedures of Ciani and Ferraro [16] and Canonico et al. [12]. Briefly, SGJ was prepared using three different solutions: solution A (sugars); solution B, (tartaric acid, malic acid and citric acid); solution C (vitamins and survival factors). The three solutions were sterilized separately and then combined aseptically. The fermentation trials were set up in 100 mL flask that contained 70 mL SGJ under static and a rotary shaker (150 rpm) at 22 • C in triplicate. The inoculum for the immobilized cells of M. pulcherrima was 10%, 5%, 1% of beads (wet w/v) which corresponds to an inoculum of 1 × 10 8 cells/mL, 5 × 10 7 cells/mL and 1 × 10 7 cells/mL respectively. The same inoculum was carried out using free cells of M. pulcherrima. After 72 h of fermentation, free S. cerevisiae cells (1 × 10 6 cell/mL) were inoculated into the partially fermented grape juice without the removal of free or immobilized cells. At the end of fermentation, fermented products were subjected to chemical and microbiological analysis. The cell concentration trial that exhibited the best ethanol reduction was chosen to carry out fermentation in Natural Grape Juice (NGJ).

Fermentation Trials on Natural Grape Juice (NGJ)
For the fermentation trials in NGJ it was used sterile-filtered (membrane Ø = 0.45 µm) Verdicchio grape juice, an autochthonous white grape variety of Marche region (central Italy). Grape juice was obtained during 2015 vintage and maintained frozen until the use. The main characteristics of the grape juice were: pH 3.39; total acidity 4.42 g/L; total SO 2 , 34 mg/L; malic acid, 2.7 g/L; initial sugar content, 204 g/L; yeast assimilable nitrogen, 90 mg N/L. Immobilized M. pulcherrima cells were evaluated in sequential fermentation trials on Verdicchio grape juice (initial inoculation level 5 × 10 7 cell/mL selected by the previous screening). After 72 h of fermentation, free S. cerevisiae cells (1 × 10 6 cell/mL) were inoculated into the partially fermented grape juice without removal of free or immobilized cells of M. pulcherrima.
The fermentations were carried out in 2 L bench-top bioreactor (Biostat ® B; B. Braun Biotech Int., Goettingen, Germany) with 1.5 L of Verdicchio grape juice in gentle agitation (60 rpm/min) at 22 • C with the following aeration conditions during the initial 72 h: (i) no aeration; (ii) aeration flow of 1 mL/L/min; (iii) 20 mL/L/min. The inoculation level was 5 × 10 7 cell/mL of immobilized cells (corresponding to 5% wet w/v of beads). Control trials were carried out using free S. cerevisiae (10 6 cell/mL) under no aeration condition. The fermentation kinetics was determined by the evolution of sugar consumption using a specific enzymatic kit (Megazyme International Wicklow Ireland).
The cell release was analyzed by colony forming unit (CFU) counts on Lysine Agar (Oxoid, Hampshire, UK), a selective medium that does not support the growth of S. cerevisiae [17]. The fermentations were carried out in triplicate.

Analytical Procedures
The Official European Union Methods [18] were used to determine the ethanol content and volatile acidity g acetic acid/L). Acetaldehyde, ethyl acetate, n-propanol, isobutanol, amyl and isoamyl alcohols were quantified by direct injection into a gas-liquid chromatography system (GC-2014; Shimadzu, Kjoto, Japan). Each sample was prepared and analysed following the procedures of Canonico et al. [19]. Briefly, the main volatile compounds were determined by Solid-phase microextraction (HS-SPME) method. Five mL of each sample was placed in vial containing 1 g NaCl closed with a septum-type cap. HS-SPME was carried out under magnetic stirring for 10 min at 25 • C. After this period, an amount of 3-octanol as the internal standard (1.6 mg/L) was added and the solution was heated to 40 • C and extracted with a fiber Divinylbenzene/Carboxen/Polydimethylsiloxane (DVB/CAR/PDMS) fibre (Sigma-Aldrich, St. Louis, Missouri, USA ) for 30 min by insertion into the vial headspace. The compounds were desorbed by inserting the fibre into a Shimadzu gas chromatograph GC injector for 5 min. A glass capillary column was used: 0.25 µm Supelcowax 10 (length, 60 m; internal diameter, 0.32 mm). The fibre was inserted in split-splitless mode. The compounds were identified and quantified by comparisons with external calibration curves for each compound. The glucose and fructose (K-FRUGL), glycerol (K-GCROL), and succinic acid (K-SUCC) concentrations were determined using specific enzyme kits (Megazyme International, Wicklow Ireland).

Statistical Analysis
Analysis of variance (ANOVA) was applied to the experimental data for the analytical characters of wine. The means were analyzed using the STATISTICA 7 software (Stat Soft, Inc, Tulsa, OK, USA). The significant differences were determined using Duncan tests, and the data were considered significant if the associated p-values were <0.05.

Screening to Assess Aeration, Different Modalities of Immobilization and Inoculation Levels of M. pulcherrima Sequential Fermentation in Synthetic Grape Juice (SGJ)
The preliminary screening was carried out under static and agitation conditions (150 rpm) in SGJ at 22 • C ± 1 • C and inoculated with M. pulcherrima (free and immobilized cells), at concentration 10 8 cell/mL (10% wet w/v of beads), 5 × 10 7 cell/mL (5% wet w/v of beads) and 10 7 cell/mL (1% wet w/v  Table 1. M. pulcherrima/S. cerevisiae sequential fermentation reduced the ethanol content in almost all conditions tested. M. pulcherrima in immobilized form improved the ethanol reduction in comparison with free cells in both static and agitated fermentation conditions. Agitation condition determined a further reduction in ethanol content. A significant ethanol reduction was obtained with 10% and 5% of M. pulcherrima immobilized cells both in static and agitation condition while 1 × 10 7 (1% immobilized cells) did not show significant reduction in comparison with free cells and S. cerevisiae control strain. Regarding the volatile acidity, only M. pulcherrima immobilized cells (10%) in agitation condition exhibited a significant increase in comparison with the other fermentation trials. Considering the results obtained in this preliminary screening, M. pulcherrima at inoculum level 5% corresponding to 5 × 10 7 cell/mL (the lowest active concentration) was selected for bench-top fermentation trials in NGJ using the following different fermentation conditions: semianaerobic condition (gently agitation 60 rpm), aeration flow of 1 mL/L/min and 20 mL/L/min.  The preliminary screening was carried out under static and agitation conditions (150 rpm) in SGJ at 22 °C ± 1°C and inoculated with M. pulcherrima (free and immobilized cells), at concentration 10 8 cell/mL (10% wet w/v of beads), 5 × 10 7 cell/mL (5% wet w/v of beads) and 10 7 cell/mL (1% wet w/v of beads). After 72 h of fermentation, S. cerevisiae EC 1118 was inoculated (10 6 cells/mL). The results of the main oenological parameter are shown in Table 1.

Sugar Consumption in Natural Grape Juice (NGJ) and the Main Analytical Characters
M. pulcherrima/S. cerevisiae sequential fermentation reduced the ethanol content in almost all conditions tested. M. pulcherrima in immobilized form improved the ethanol reduction in comparison with free cells in both static and agitated fermentation conditions. Agitation condition determined a further reduction in ethanol content. A significant ethanol reduction was obtained with 10% and 5% of M. pulcherrima immobilized cells both in static and agitation condition while 1 × 10 7 (1% immobilized cells) did not show significant reduction in comparison with free cells and S. cerevisiae control strain. Regarding the volatile acidity, only M. pulcherrima immobilized cells (10%) in agitation condition exhibited a significant increase in comparison with the other fermentation trials. Considering the results obtained in this preliminary screening, M. pulcherrima at inoculum level 5% corresponding to 5 × 10 7 cell/mL (the lowest active concentration) was selected for bench-top fermentation trials in NGJ using the following different fermentation conditions: semianaerobic condition (gently agitation 60 rpm), aeration flow of 1 mL/L/min and 20 mL/L/min.  At 72 h of fermentation, before the S. cerevisiae inoculum, M. pulcherrima with an aeration of 20 mL/L/min, exhibited a higher sugar consumption in comparison with the other sequential fermentation trials. On the other hand, at this time S. cerevisiae pure culture exhibited the faster sugar consumption in comparison with all the other trials. During the whole fermentation process, the sequential fermentation trials with 20 mL/L/min and 1 mL/L/min exhibited an intermediate consumption between S. cerevisiae pure culture and M. pulcherrima/S. cerevisiae in semianaerobic sequential fermentation that was the slowest fermentation. However, all sequential fermentations completely consumed the initial sugar content in grape juice and M. pulcherrima cell release from beads was lower than 10 3 cell/mL in all trials. The preliminary screening was carried out under static and agitation conditions (150 rpm) in SGJ at 22 °C ± 1°C and inoculated with M. pulcherrima (free and immobilized cells), at concentration 10 8 cell/mL (10% wet w/v of beads), 5 × 10 7 cell/mL (5% wet w/v of beads) and 10 7 cell/mL (1% wet w/v of beads). After 72 h of fermentation, S. cerevisiae EC 1118 was inoculated (10 6 cells/mL). The results of the main oenological parameter are shown in Table 1.

Sugar Consumption in Natural Grape Juice (NGJ) and the Main Analytical Characters
M. pulcherrima/S. cerevisiae sequential fermentation reduced the ethanol content in almost all conditions tested. M. pulcherrima in immobilized form improved the ethanol reduction in comparison with free cells in both static and agitated fermentation conditions. Agitation condition determined a further reduction in ethanol content. A significant ethanol reduction was obtained with 10% and 5% of M. pulcherrima immobilized cells both in static and agitation condition while 1 × 10 7 (1% immobilized cells) did not show significant reduction in comparison with free cells and S. cerevisiae control strain. Regarding the volatile acidity, only M. pulcherrima immobilized cells (10%) in agitation condition exhibited a significant increase in comparison with the other fermentation trials. Considering the results obtained in this preliminary screening, M. pulcherrima at inoculum level 5% corresponding to 5 × 10 7 cell/mL (the lowest active concentration) was selected for bench-top fermentation trials in NGJ using the following different fermentation conditions: semianaerobic condition (gently agitation 60 rpm), aeration flow of 1 mL/L/min and 20 mL/L/min.  At 72 h of fermentation, before the S. cerevisiae inoculum, M. pulcherrima with an aeration of 20 mL/L/min, exhibited a higher sugar consumption in comparison with the other sequential fermentation trials. On the other hand, at this time S. cerevisiae pure culture exhibited the faster sugar consumption in comparison with all the other trials. During the whole fermentation process, the sequential fermentation trials with 20 mL/L/min and 1 mL/L/min exhibited an intermediate consumption between S. cerevisiae pure culture and M. pulcherrima/S. cerevisiae in semianaerobic sequential fermentation that was the slowest fermentation. However, all sequential fermentations completely consumed the initial sugar content in grape juice and M. pulcherrima cell release from beads was lower than 10 3 cell/mL in all trials. The preliminary screening was carried out under static and agitation conditions (150 rpm) in SGJ at 22 °C ± 1°C and inoculated with M. pulcherrima (free and immobilized cells), at concentration 10 8 cell/mL (10% wet w/v of beads), 5 × 10 7 cell/mL (5% wet w/v of beads) and 10 7 cell/mL (1% wet w/v of beads). After 72 h of fermentation, S. cerevisiae EC 1118 was inoculated (10 6 cells/mL). The results of the main oenological parameter are shown in Table 1.

Bench-Top Fermentation Trials
M. pulcherrima/S. cerevisiae sequential fermentation reduced the ethanol content in almost all conditions tested. M. pulcherrima in immobilized form improved the ethanol reduction in comparison with free cells in both static and agitated fermentation conditions. Agitation condition determined a further reduction in ethanol content. A significant ethanol reduction was obtained with 10% and 5% of M. pulcherrima immobilized cells both in static and agitation condition while 1 × 10 7 (1% immobilized cells) did not show significant reduction in comparison with free cells and S. cerevisiae control strain. Regarding the volatile acidity, only M. pulcherrima immobilized cells (10%) in agitation condition exhibited a significant increase in comparison with the other fermentation trials. Considering the results obtained in this preliminary screening, M. pulcherrima at inoculum level 5% corresponding to 5 × 10 7 cell/mL (the lowest active concentration) was selected for bench-top fermentation trials in NGJ using the following different fermentation conditions: semianaerobic condition (gently agitation 60 rpm), aeration flow of 1 mL/L/min and 20 mL/L/min.  At 72 h of fermentation, before the S. cerevisiae inoculum, M. pulcherrima with an aeration of 20 mL/L/min, exhibited a higher sugar consumption in comparison with the other sequential fermentation trials. On the other hand, at this time S. cerevisiae pure culture exhibited the faster sugar consumption in comparison with all the other trials. During the whole fermentation process, the sequential fermentation trials with 20 mL/L/min and 1 mL/L/min exhibited an intermediate consumption between S. cerevisiae pure culture and M. pulcherrima/S. cerevisiae in semianaerobic sequential fermentation that was the slowest fermentation. However, all sequential fermentations completely consumed the initial sugar content in grape juice and M. pulcherrima cell release from beads was lower than 10 3 cell/mL in all trials. The preliminary screening was carried out under static and agitation conditions (150 rpm) in SGJ at 22 °C ± 1°C and inoculated with M. pulcherrima (free and immobilized cells), at concentration 10 8 cell/mL (10% wet w/v of beads), 5 × 10 7 cell/mL (5% wet w/v of beads) and 10 7 cell/mL (1% wet w/v of beads). After 72 h of fermentation, S. cerevisiae EC 1118 was inoculated (10 6 cells/mL). The results of the main oenological parameter are shown in Table 1.

Sugar Consumption in Natural Grape Juice (NGJ) and the Main Analytical Characters
M. pulcherrima/S. cerevisiae sequential fermentation reduced the ethanol content in almost all conditions tested. M. pulcherrima in immobilized form improved the ethanol reduction in comparison with free cells in both static and agitated fermentation conditions. Agitation condition determined a further reduction in ethanol content. A significant ethanol reduction was obtained with 10% and 5% of M. pulcherrima immobilized cells both in static and agitation condition while 1 × 10 7 (1% immobilized cells) did not show significant reduction in comparison with free cells and S. cerevisiae control strain. Regarding the volatile acidity, only M. pulcherrima immobilized cells (10%) in agitation condition exhibited a significant increase in comparison with the other fermentation trials. Considering the results obtained in this preliminary screening, M. pulcherrima at inoculum level 5% corresponding to 5 × 10 7 cell/mL (the lowest active concentration) was selected for bench-top fermentation trials in NGJ using the following different fermentation conditions: semianaerobic condition (gently agitation 60 rpm), aeration flow of 1 mL/L/min and 20 mL/L/min.  At 72 h of fermentation, before the S. cerevisiae inoculum, M. pulcherrima with an aeration of 20 mL/L/min, exhibited a higher sugar consumption in comparison with the other sequential fermentation trials. On the other hand, at this time S. cerevisiae pure culture exhibited the faster sugar consumption in comparison with all the other trials. During the whole fermentation process, the sequential fermentation trials with 20 mL/L/min and 1 mL/L/min exhibited an intermediate consumption between S. cerevisiae pure culture and M. pulcherrima/S. cerevisiae in semianaerobic sequential fermentation that was the slowest fermentation. However, all sequential fermentations completely consumed the initial sugar content in grape juice and M. pulcherrima cell release from beads was lower than 10 3 cell/mL in all trials. The preliminary screening was carried out under static and agitation conditions (150 rpm) in SGJ at 22 °C ± 1°C and inoculated with M. pulcherrima (free and immobilized cells), at concentration 10 8 cell/mL (10% wet w/v of beads), 5 × 10 7 cell/mL (5% wet w/v of beads) and 10 7 cell/mL (1% wet w/v of beads). After 72 h of fermentation, S. cerevisiae EC 1118 was inoculated (10 6 cells/mL). The results of the main oenological parameter are shown in Table 1.

Sugar Consumption in Natural Grape Juice (NGJ) and the Main Analytical Characters
M. pulcherrima/S. cerevisiae sequential fermentation reduced the ethanol content in almost all conditions tested. M. pulcherrima in immobilized form improved the ethanol reduction in comparison with free cells in both static and agitated fermentation conditions. Agitation condition determined a further reduction in ethanol content. A significant ethanol reduction was obtained with 10% and 5% of M. pulcherrima immobilized cells both in static and agitation condition while 1 × 10 7 (1% immobilized cells) did not show significant reduction in comparison with free cells and S. cerevisiae control strain. Regarding the volatile acidity, only M. pulcherrima immobilized cells (10%) in agitation condition exhibited a significant increase in comparison with the other fermentation trials. Considering the results obtained in this preliminary screening, M. pulcherrima at inoculum level 5% corresponding to 5 × 10 7 cell/mL (the lowest active concentration) was selected for bench-top fermentation trials in NGJ using the following different fermentation conditions: semianaerobic condition (gently agitation 60 rpm), aeration flow of 1 mL/L/min and 20 mL/L/min.  At 72 h of fermentation, before the S. cerevisiae inoculum, M. pulcherrima with an aeration of 20 mL/L/min, exhibited a higher sugar consumption in comparison with the other sequential fermentation trials. On the other hand, at this time S. cerevisiae pure culture exhibited the faster sugar consumption in comparison with all the other trials. During the whole fermentation process, the sequential fermentation trials with 20 mL/L/min and 1 mL/L/min exhibited an intermediate consumption between S. cerevisiae pure culture and M. pulcherrima/S. cerevisiae in semianaerobic sequential fermentation that was the slowest fermentation. However, all sequential fermentations completely consumed the initial sugar content in grape juice and M. pulcherrima cell release from beads was lower than 10 3 cell/mL in all trials. ).

Sugar Consumption in Natural Grape Juice (NGJ) and the Main Analytical Characters
At 72 h of fermentation, before the S. cerevisiae inoculum, M. pulcherrima with an aeration of 20 mL/L/min, exhibited a higher sugar consumption in comparison with the other sequential fermentation trials. On the other hand, at this time S. cerevisiae pure culture exhibited the faster sugar consumption in comparison with all the other trials. During the whole fermentation process, the sequential fermentation trials with 20 mL/L/min and 1 mL/L/min exhibited an intermediate consumption between S. cerevisiae pure culture and M. pulcherrima/S. cerevisiae in semianaerobic sequential fermentation that was the slowest fermentation. However, all sequential fermentations completely consumed the initial sugar content in grape juice and M. pulcherrima cell release from beads was lower than 10 3 cell/mL in all trials.
The main analytical characters of M. pulcherrima sequential fermentations under different aeration conditions at 72 h (before inoculum of S. cerevisiae) and at the end of fermentation are reported in Table 2. Results of volatile acidity and glycerol content at 72 h did not show significant differences unlike ethanol content that showed significant differences among the trials. S. cerevisiae pure cultures displayed the highest production while the ethanol formation by M. pulcherrima was related to aeration indicating that the oxygen positively affects the fermentation rate of M. pulcherrima.
M. pulcherrima sequential fermentation with aeration flow of 1 mL/L/min and 20 mL/L/min increased significantly the glycerol content in the final wine in comparison with semianaerobic fermentation trials (M. pulcherrima/S. cerevisiae and S. cerevisiae pure culture). Similarly, to 72 h of fermentation, acetic acid content of final wines showed a comparable value among the trials.
The results of the final ethanol content highlighted a relevant differentiation among the fermentation trials. In the condition tested in semianaerobic condition M. pulcherrima/S. cerevisiae, there was an ethanol reduction of 0.51% (v/v) in comparison with S. cerevisiae control trial. This value was comparable with the ethanol reduction obtained with M. pulcherrima sequential fermentation with an aeration flow of 1mL/L/min. A significant increase of ethanol reduction of ca. 1.30% (v/v) was obtained with M. pulcherrima sequential fermentation with an aeration flow 20 mL/L/min. The data of ethanol yield reflected followed this trend.

The Main By-Products and Volatile Compounds of Sequential Fermentation with M. pulcherrima in Different Aeration Conditions in NGJ
To evaluate the influence of different aeration conditions on the aromatic profile of wines, the main volatile compounds produced in NGJ were assayed and the results were reported in Table 3.
A significant increase in ethyl acetate content in fermentation carried out with M. pulcherrima with aeration flow of 20 mL/L/min was found. However, differently from the results shown in similar conditions (sequential fermentation and aeration flow during the first 72 h) using free cells [15], the increase of ethyl acetate with immobilized cells was restrained and under the negative sensory threshold for this compound (175 mg/L) [20]. Moreover, this condition (immobilized cells and aeration flow of 20 mL/L/min) determined a significant increase of desirable volatile compounds such as ethyl butyrate and isoamyl acetate. In semianerobic condition, M. pulcherrima sequential fermentation exhibited a significantly higher amount of phenyl ethyl acetate in comparison with the other trials. Regarding the main higher alcohols M. pulcherrima sequential fermentation with aeration flow 1 mL/L/min exhibited a significant increase in amylic and isoamylic alchols, while with aeration flow 20 mL/L/min increased significantly the production of β-phenyl ethanol and isobutanol respectively. A significant increase of acetaldehyde content showed sequential fermentations under semianaerobic condition and 20 mL/L/min of air flow. However, acetaldehyde at these concentrations did not negatively influence the aromatic profile of the final wines. Regarding the carboxylic acids, a significant increase was found only for the diethyl succinate content in the trial with 20 mL/L/min of air flow, while no significant differences were exhibited in butyric acid content.   .49 ± 0.08 b 44.5 ± 0.04 b S. cerevisiae semianaerobic 0.47 ± 0.01 a 0.33 ± 0.00 a 8.65 ± 0.06 a 38.6 ± 0.13 a S. cerevisiae semianaerobic 5.15 ± 0.06 b 0.38 ± 0.02 b 12.00 ± 0.01 a 46.5 ± 0.01 a Data are means ± standard deviations from three independent experiments. Data with different superscript letters ( a,b,c ) within each column are different homogeneous groups according to Duncan tests (0.05%).