Microbial Biocapsules as Generally Recognized-As-Safe Fungal-Based Immobilized Cell Technology for Precision Sequential Fermentations of Grape Must

Abstract

This work focuses on the production of a white wine with a specific organoleptic profile by means of sequential fermentation using immobilized yeast in a system known as “microbial biocapsules”. Three fermentation conditions were created: sequential fermentation with immobilized yeast (SqFMB) employing a matrix composed by Aspergillus oryzae (pellet-forming fungus recognized as GRAS), sequential fermentations with non-immobilized yeast cells (SqF), and a control of spontaneous fermentation (SpF). To carry out these fermentations, Pedro Ximénez grape must was used and two non-Saccharomyces yeast strains, Debaryomyces hansenii LR1 and Metschnikowia pulcherimma Primaflora, and the Saccharomyces cerevisiae X5 strain were used. The wines produced were subjected to microbiological and chemical analyses in which metabolites that positively influence the wine profile, such as 1,1-diethoxyethane and decanal, are only produced in the SqFMB condition, and others, like nonanal, were detected in higher concentrations than in SqF and SpF. Microbiological analyses show that less non-Saccharomyces yeasts were isolated in the SqFMB condition than in SqF, which indicates an efficiency in the inoculation and removal method proposed. These results conclude that microbial biocapsules seem to be a good yeast carrier for wine elaboration; however, modulation of some variables like yeast concentration inocula, the employment of preadaptation methods or the use of yeast species with higher fermentative power need to be tested to improve the novel methodology.

1. Introduction

Sequential fermentation is a method employed in winemaking, utilizing various yeast strains in a specific sequence to tailor the properties of the wine. This process involves the use of non-Saccharomyces yeasts to ferment independently for a specified duration, followed by the inoculation of Saccharomyces cerevisiae yeasts to end the fermentation [1,2,3]. This technique can be applied to lower the ethanol content in wine, elevate the aroma, and enhance its overall quality [2,4]. It enables the creation of distinctive flavors and aromas that are unattainable with a single yeast strain [2,5]. Although sequential fermentation serves as a valuable tool for winemakers aiming to craft high-quality wine with distinctive flavors and aromas, precise process control is essential to prevent competition with wild yeast and S. cerevisiae and to ensure that non-Saccharomyces yeast strains can produce desired flavors and aromas before S. cerevisiae takes over fermentation [4,6].

Immobilized Cell Technologies (ICTs) enable more precise control in sequential fermentations. ICT aims to enclose cells within a matrix, ensuring they fix yet allowing substrate and product diffusion [7]. Unlike suspended cells, it ensures precise control over yeast inoculation, duration, and removal, thus boosting the competitiveness and metabolic activity of strains in sequential fermentations [4]. In studies by Ciani et al. [8], the authors experimented with sequential fermentations using immobilized Candida stellata in 2.5% Na-alginate beads and S. cerevisiae, resulting in improved fermentation rates, increased wine ethanol levels, and reduced unwanted byproducts (i.e., acetaldehyde and acetoin). By maintaining immobilized C. stellata cells 3 days in the grape must before adding S. cerevisiae, the same authors achieved wines with enhanced profiles, including reduced sugars, controlled acetaldehyde and acetoin, and higher concentrations of fruity and floral aroma compounds [9]. Later, Genisheva et al. [10] conducted sequential alcoholic and malolactic fermentations using immobilized S. cerevisiae on grape stems and skins and bacteria on grape skins, yielding 13 vol% ethanol wines with good quality in terms of acidity, alcohol content, and sugar content. A notable reduction of 67% in malic acid concentration was observed, demonstrating greater efficiency compared to traditional methods. Lastly, Canonico et al. [4] employed immobilized non-Saccharomyces yeasts (Starmerella bombicola, Metschnikowia pulcherrima, Hanseniaspora osmophila, and Hanseniaspora uvarum) to start fermentation followed by free S. cerevisiae cells. This innovative approach significantly reduced ethanol levels by up to 1.6% v/v (with S. bombicola) while generating beneficial fermentation byproducts (glycerol and succinic acid for S. bombicola, geraniol for M. pulcherrima, and isoamyl acetate and isoamyl alcohol for H. osmophila) and avoiding unfavorable ones (e.g., ethyl acetate).

In this work, we advocate employing a natural cell carrier made up of inactive Generally Recognized-As-Safe (GRAS) fungal hollow spherical pellets assembling ICT referred to as “microbial biocapsules” to produce a distinctive white wine with specific sensorial properties—earthy, flower aromas, low-bodied and a clean taste void of off-flavors—from unfiltered grape must (imitating industrial settings) through sequential fermentations. Both this ICT system and a prior iteration using a different assembly technique called “yeast biocapsules” have undergone trials for the production of alcoholic beverages [11,12].

2. Materials and Methods

2.1. Microorganisms, Media and Growth Conditions

For the sequential fermentations, three distinct yeast strains—Debaryomyces hansenii LR1, M. pulcherrima Primaflora, and S. cerevisiae X5 (CECT13015)—were selected due to their respective metabolic characteristics and their fermentative power outlined in

Table 1. Yeast selected, time interval in days at the sequential fermentation, yeast metabolic features, and wine sensory properties expected.

Yeast Selected	Time Interval in Days at the Sequential Fermentation	Metabolic Feature	Wine Sensory Properties Expected	References
Debaryomyces hansenii LR1	1–3	Production of phenethyl alcohol and α-terpineol. Low fermentative power	Floral aroma	[4,13]
Metschnikowia pulcherrima Primaflora	3–6	Low glycerol production, high ester production and production of pulcherrimine. High fermentative power	Low body, enhanced varietal aroma and wine microbial stability	[4,13,14]
Saccharomyces cerevisiae X5	6–14	Low impact on grape compounds and highly fermentative	Varietal sensory profile, no residual sugars and high ethanol concentration	[4,15]

.

To ensure the yeasts in Table 1 create desired sensory qualities without introducing wine defects, fermentations were conducted using each strain independently in Pedro Ximénez must (details in the Section 2.3).

For the production of fungal pellets as cell-immobilizing supports, we used the filamentous fungus Aspergillus oryzae 76-2 (FST 76-2), a pelletizable fungus widely used in the food sector and GRAS [16], obtained from the Agricultural Chemistry, Edaphology and Microbiology Department of the University of Córdoba (Spain) and Phaff Culture Collection at the University of California, Davis (USA). This fungus was sporulated on solid medium containing 17 g/L cornmeal agar, 1 g/L yeast extract, 2 g/L glucose, and 20 g/L agar.

Yeast cells were seeded on YPD agar (10 g/L yeast extract, 20 g/L peptone, 20 g/L dextrose and 20 g/L agar) at 28 °C for 1–3 days and subsequently pre-cultured on liquid YPD medium for one day at 175 rpm and at 28 °C before being inoculated into the grape must or immobilized on fungal pellets.

2.2. Yeast Immobilization Procedure

To produce the immobilizing carriers, A. oryzae 76-2 fungal spores were inoculated on Fungal Pelletization Medium, or FPM, containing 60 g/L glucose, 0.5 g/L KCl, 3 g/L yeast extract, 3 g/L NaNO₃, 1 g/L K₂HPO₄, 0.5 g/L MgSO₄ and 0.01 g/L FeSO₄ adjusted to 5.5 with HCl, to reach a final concentration of 0.1 × 10⁶/mL. The spore suspension was grown on a rotary shaker at 250 rpm and 28 °C for 72 h; 4–6 mm Ø fungal pellets were obtained and were further inactivated by autoclaving (121 °C, 20 min) to avoid competition with the immobilized yeasts and to avoid influencing the sensory chemical properties of the wine (Figure 1). The inactivated pellets were kept in sterile distilled water at 4 °C until use.

To immobilize the yeast cells in the fungal pellets, the pre-cultured cells of the above-mentioned yeast strains were pelleted by centrifugation (5000 rpm, 15 min; Hettich Zentrifugen Rotina 38R, ø 15 cm) and mixed with the produced pellets in a 1:1 wet mass ratio and sterile distilled water (in volume 5 times higher than the wet weight of the yeasts) in a Falcon. This suspension was vortexed and kept under vacuum (<0.3 atm pressure) for 1 min using a Bonsenkitchen system (Oakwood, GA, USA), thus forcing the yeast cells into the pellet matrix. To confirm infusion, the optical density at 580 nm was measured spectrophotometrically on the cell suspension before and after vacuum (a 20% reduction was confirmed). Pellets with yeast cells were immersed in YPD culture media (20 mL per gram of wet mass of fungal pellet–yeast cells) and cultured 24 h, at 175 rpm and 28 °C, to promote yeast growth from within the fungal pellet.

Finally, the already-formed microbial biocapsules (MBs) were rinsed with sterile deionized water to remove the cells located on the surface, thus avoiding their release and proliferation when being used in the subsequent fermentations. The MBs were kept at 4 °C in Falcons until use. Prior to inoculation into the grape must, the biocapsules were reactivated by culturing them in YPD medium at 28 °C for 72 h and finally rinsed with sterile distilled water. The methodology of biocapsule assembly is subjected to a provisional patent (Application number: 63/411,843) co-owned by the University of Cordoba and the University of California, Davis.

2.3. Fermentation Conditions

The grape must in this work came from Pedro Ximénez, a prominent grape variety within the Montilla-Moriles Protected Designation of Origin (PDO) from Pérez Barquero SA winery. The must had a pH value of 3.9, 12.9° Baume (equivalent to 234 g of sugar per liter and 23.4° Brix), 13.8% (v/v) probable ethanol, SO₂ concentration of 98 mg/L, and 4.8 g H₂T/L total acidity after adjusting the pH to 3.4 by adding tartaric acid (C₄H₆O₆) to avoid bacteria propagation [17].

Three different fermentation conditions were designed to evaluate how sequential inoculations of immobilized and non-immobilized yeast cells affected grape must fermentation. These included sequential fermentations with yeasts immobilized in MBs (SqFMB), sequential fermentations with non-immobilized yeast cells (SqF), and a control using spontaneous fermentation (SpF). Each condition comprised three biological replicates, with each replicate containing 500 mL of grape must in a sterile 1 L Erlenmeyer flask set in a thermostatic chamber at 15 °C, under static conditions until completion (weight loss due to CO₂ release less than 1 g/day). Under SqFMB and SqF, the grape musts were inoculated with a yeast population (free or immobilized) sufficient to reach a final concentration of 1 × 10⁶ cells/mL. In SqFMB, MBs were removed before inoculating the next MB type.

In the SqFMB condition, to determine the concentration of immobilized yeast cells, and to calculate the number of grams of biocapsules needed to reach the desired final concentration per replicate, five randomly selected MBs were first drained for 5 s and weighed on a precision balance. They were then placed in a tissue grinder (Kisker Biotech, Steinfurt, Germany) to which 100 mM NaCl was added, ensuring that the MBs were totally submerged in the solution and subsequently crushed. After crushing the MBs and obtaining a homogeneous suspension, this was sonicated for 20 min to separate yeast cells from fungal hyphae. Finally, the yeast cells were counted by the capsule with the conventional microscope and Neubauer chamber to calculate the number of cells per MB unit or gram of MBs. The calculated amount of MBs was placed in food-grade mesh bags with plastic cords on one side to ease inoculation and removal (Figure 2). Mesh bags prevent MB dispersion, and the plastic cords allow us to manipulate the mesh bags, minimizing the contamination risk.

During the fermentation process, the three must wines were sampled 4 times for subsequent microbiological analysis. However, in SqFMB and SqF must wine conditions, samples were taken right before the inoculation of D. hansenii LR1 (T0), before M. pulcherrima Primaflora inoculation (T1), before inoculation of S. cerevisiae X5 (T2) and at the end of the fermentation (Tf) (Figure 3).

2.4. Microbiological Analysis

Samples were cultured in YPD agar medium as a general culture medium for fungi, lysine agar medium (44.5 g/L glucose, 17.8 g/L agar, 1 g/L lysine and trace elements in concentrations lower than 1 g/L; OXOID CM 0191B) as selective culture medium for non-Saccharomyces, and WL agar medium (50 g/L dextrose, 4 g/L yeast extract, 5 g/L tryptone and 0.022 g/L bromocresol green; OXOID CM 0501) as a differential culture medium. The bromocresol green gives the blue-green appearance to the WL medium and acts as a pH indicator, turning to yellow upon acid production by the microorganisms. The plates of the three culture media were incubated in an oven at 28 °C for 72 h to allow the growth of the microorganisms at a suitable temperature.

A visual characterization of the isolated colonies in the different media was performed, and subsequently, morphological groups were established. The different morphological groups were seeded on lysine agar plates in order to determine if these visually differentiated groups belonged to non-Saccharomyces or if, on the contrary, they were potential Saccharomyces yeasts.

2.5. Measurement of Enological Parameters

pH, ethanol, titratable acidity, and volatile acidity were determined following OIV protocols (International Organisation of Vine and Wine, 2024) [18]. The pH was measured in a Crison GLP 21 + pH meter [17]. An Alcolyzer 3001 alcohometer (Anton Paar; Graz, Austria) was used to analyze the ethanol content while the acetic acid and glycerol content (g/L) was determined using the Y15/C chemical analyzer using an absorbance of 500 nm (Biosystems; Barcelona, Spain).

2.6. Quantification of Major Aroma Compounds and Polyols

The major volatiles or compounds found  >10 mg/L that are evaporative under room temperature and influence the wine organoleptic characteristics were analyzed by gas chromatography on the Agilent 6890 GC (Palo Alto, Santa Clara, CA, USA). This chromatograph is equipped with a flame ionization detector (FID) and with a “CP-Wax 57 CB” column prepared for analysis with dimensions of 60 m × 0.25 μm × 0.2 μm, injecting 0.7 μL of sample per replicate; the frame time for the total elution of the major volatiles and polyols quantified was 80 min [19]. Prior to sample injection into the chromatograph, the wine sample was treated by adding 1 mL of 1.018 g L⁻¹ 4-methyl-2-pentanol (CAS 108-11-2) as internal standard in a 14% (v/v) ethanol solution and 0.2 g of solid calcium carbonate to a total volume of 10 mL. This mixture was then shaken for 30 s in an ultrasonic bath and finally subjected to centrifugation at 5000 rpm for 10 min at 2 °C temperature to remove tartaric acid from the wine. The resulting supernatant was injected for analysis [17].

2.7. Quantification of Minor Aroma Compounds

Minor volatile compounds were analyzed using the platform SBSE-TD-GC-MS (Stir Bar Sorptive Extraction–Thermal Desorption–Gas Chromatography–Mass Spectrometry) composed of an Agilent-7890A GC coupled to an MSD 5975C (Wilmington, DE, USA) and a Multi-Purpose Sampler (MPS) from Gerstel (GmbH & Co. KG–Mülheim an der Rhur, Deutschland). The Chemstation software v. 02. 02. 1431 from Agilent and Maestro from Gerstel were used for platform control and chromatographic data processing. The minor volatile compounds were extracted by the SBSE technique, using a twister (10 mm long, 0.5 mm thick film) coated with polydimethylsiloxane (PDMS). Briefly, the procedure was as follows: 1 mL of wine sample, 0.1 mL of internal standard solution (0.4116 gL⁻¹ hexyl butyrate (CAS 2639-63-6) in absolute ethanol), and 8.9 mL of 12% (v/v) ethanol solution containing 2.6 gL⁻¹ tartaric acid and 2.2 gL⁻¹ potassium bitartrate (pH 3.5) were added to a 10 mL vial. Then, the twister was placed in the vial and stirred at 1200 rpm and 20 °C for 120 min to favor the adsorption of compounds in a Variomag Multipoint 15 magnetic stirrer (Thermo Fisher Scientific, Waltham, MA, USA). The twister was removed, rinsed with water, and dried and then placed in a desorption tube to be transferred by MPS to the Thermal Desorption Unit (TDU) from Gerstel where the volatiles were desorbed and transferred to the GC system. An HP-5MS-fused silica capillary column (60 m × 0.25 mm i.d., 0.25 μm film) from Agilent Technologies was used at an initial oven temperature of 50 °C (2 min), increased at 4 °C/min to 190 °C for 10 min. The MSD operated at 70 eV in the electron impact mode (EI), with a mass range of 35–550 Da at a temperature of 150 °C. All samples were analyzed in triplicate. The time frame for the total elution of the minor volatiles quantified was 60 min. The quantification of minor volatile compounds was carried out using a calibration table built with standard solutions, containing a known concentration of each compound.

2.8. Sensory Analyses

The wines were evaluated by a panel of eight expert judges from the Department of Agricultural Chemistry, Edaphology and Microbiology of the University of Córdoba (Spain). To carry out the evaluation, the panel used the OIV (2021) tasting sheet (International Organisation of Vine and Wine, 2021). This tasting sheet allowed the different attributes to be evaluated in terms of sight (cleanliness, aspects other than cleanliness), smell (authenticity, positive intensity, quality), and taste (authenticity, positive intensity, harmonious persistence, quality). To ensure proper conditions, all wine samples were stored for 24 h at 4 °C prior to analysis.

Each fermentation condition was evaluated in a random order and the wine samples (30 mL) were presented to the tasters tempered at 12–15 °C using standardized wine glasses (NF V09-110 AFNOR, 1995), according to the requirements of the ISO 3591 standards [18] listed but without indicating the condition (blind tasting).

2.9. Statistical Analysis

For the statistical analysis of all data matrices, the statistical software package Statgraphics (Centurion v. 16.1.11) was used. Multiple-Variable Analysis (MVA) was performed in order to identify significant differences between the three types of wine obtained. In addition, a Principal Component Analysis (PCA) was performed to obtain a global interpretation of the most relevant quantitative information contained in the volatile compound data matrix.

3. Results and Discussion

The role of yeasts in the aroma composition of wines has been universally recognized [20]. The use of microorganisms with specific metabolic traits in sequential fermentations to obtain wines with lower ethanol content, more complexity or less sulfite levels is a trending topic [15,21,22]. In this study, we inoculated specific yeast strains on designated days for specified durations, immobilizing them within inactivated GRAS fungal pellets to create microbial biocapsules. This approach aimed to produce intricate wines from Pedro Ximénez must within a controlled environment.

3.1. Alcoholic Fermentation Rates

The mass loss evolution due to CO₂ release during the fermentation process reported the highest fermentation rates on day 6: SqFMB (14.18 ± 0.23 g CO₂/day), SqF (14.2 ± 0.2 g CO₂/day) and SpF (14.24 ± 0.1 g CO₂/day). The alcoholic fermentation rate by SqFMB, SqF and SpF follows the kinetic models described by Bely et al. [23]. The represented data generate a sigmoid curve, characteristic of the Gompertz fermentation equation and the Lineweaver–Burk graph, as they reflect the concentration of CO₂ produced over time [24]. CO₂ release remained constant from day 11 until day 15, when all fermentations were stopped. Similar patterns between conditions may be since the yeast present in the must dominated over the inoculated yeast at the beginning of sequential fermentations, specifically D. hansenii LR1, a non-fermentative yeast sensitive to high ethanol concentrations [25]. Additionally, this is an aerobic yeast that releases less CO₂ than fermentative yeasts [26]. The must microbiota may have greater fermentative power and resistance to osmotic stress and ethanol than D. hansenii LR1, which may explain the similar results in all conditions.

3.2. Microbiological Analysis

Table 2. Non-Saccharomyces (NS) and Saccharomyces (S) percentages (%) in each sampling time on the day of the yeast inoculation (T0), day 3 after the inoculation (T3), day 6 after the inoculation (T6) and at the end of the fermentation or day 15 after inoculation (Tfinal) of sequential fermentations with yeasts immobilized in MBs (SqFMB), sequential fermentations with non-immobilized yeast cells (SqF), and a control using spontaneous fermentation (SpF).

	T0	T3	T6	Tfinal
SqFMB	70 NS–30 S	22 NS–78 S	13 NS–87 S	38 NS–62 S
SqF	70 NS–30 S	24 NS–76 S	31 NS–69 S	62 NS–38 S
SpF	70 NS–30 S	29 NS–71 S	19 NS–81 S	27 NS–73 S

shows the percentage of non-Saccharomyces (NS) and Saccharomyces (S) yeast UFCs obtained in lysine agar in each condition. While a similar ratio of NS/S was observed in the first three days for all conditions studied, SqF shows the highest value from day 6 onwards. This could be due to an increased dominance of M. pulcherrima (inoculated at day 3) metabolism and physiology which allow the yeast to stay active and grow until the end of the fermentation [27]. However, this does not occur in the SqFMB condition, which shows an efficiency in the inoculation and removal method proposed in this work.

3.3. General Enological Parameters

A slight reduction in pH on the SpF wines was observed between conditions, while no significant differences in other general oenological parameters were observed (

Table 3. Concentration of metabolites detected in the fermented musts. CAS: identification number assigned by the Chemical Abstracts Service. OT: odor threshold. OAV: Odor Activity Value. SqFMB: Wine obtained from sequential fermentation with microbial biocapsules. SqF: Wine obtained from sequential fermentation with free yeasts. SpF: Wine obtained with spontaneous fermentation with indigenous microbiota. GC–FID: gas chromatography–flame ionization detector; HG: homogenous group; N.f.: not found in the bibliography; OIV: International Organisation of Vine and Wine; SBSE–TD–GC–MS: Stir Bar Sorptive Extraction–Thermal Desorption–Gas Chromatography–Mass Spectrometry; ±: Standard deviation; abc: homogeneous group among groups of sampling., ND: not detected. The different letters indicate homogeneous groups which significantly differ statistically in the parameters between wines (p < 0.05, F-test). Concentrations overpassing an OAV of 1 are bolded.

Library/ID	Method of Detection	CAS	OT (mg/L)	Odor/Flavor Description	SqFMB	OAV	SqF	OAV	SpF	OAV
Ethanol (g/L)	Parameters measured according OIV	64-17-5	10	Alcoholic	130.70 ± 0.05 a	13,070	130.67 ± 0.49 a	13,060	131.27 ± 0.25 a	13,120
Volatile acidity (acetic acid) (mg/L)		64-19-7	200	Vinegar	200 ± 10 a	1	200 ± 10 a	1	210 ± 10 a	1.05
pH					3.13 ± 0.005 b		3.13 ± 0.00 b		3.11 ± 0.005 a
Fixed acidity (equiv/L)					6.1 ± 0.026 a		6.09 ± 0.035 a		6.08 ± 0.005 a
Glucose (g/L)		50-99-7			0.45 ± 0.06 a		0.56 ± 0.07 a		0.53 ± 0.05 a
Acetaldehyde (mg/L)	GC–FID	75-07-0	10	Over-ripe apple	107.59 ± 10.23 a	10.76	99.90 ± 4.78 a	9.99	106.18 ± 2.93 a	10.61
Ethyl acetate (mg/L)		141-78-6	7.5	Pineapple, varnish, balsamic	31.72 ± 0.84 a	4.23	34.15 ± 1.27 a	4.55	33.44 ± 1.84 a	4.45
1,1-Diethoxyethane (mg/L)		105-57-7	1	Green fruit, liquorice	10.56 ± 0.65 b	10.56	0.00 ± 0 a	0	0 ± 0.00 a	0
Methanol (mg/L)		67-56-1	668	Chemical, medicinal	61.68 ± 4.64 a	0.092	53.91 ± 4.22 a	0.081	51.41 ± 6.85 a	0.077
1-Propanol (mg/L)		71-23-8	830	Ripe fruit, alcohol	45.49 ± 0.92 a	0.055	48.07 ± 1.78 a	0.058	46.17 ± 1.11 a	0.056
Isobutanol (mg/L)		78-83-1	40	Alcohol, wine-like, nail polish	35.23 ± 0.45 a	0.88	35.28 ± 1.28 a	0.88	35.44 ± 1.82 a	0.890
2-Methylbutanol (mg/L)		137-32-6	N.f	Cooked roasted aroma with fruity or alcoholic undertones	49.66 ± 1.97 a	-	51.12 ± 1.45 a	-	50.34 ± 0.99 a	-
Isoamyl alcohol (mg/L)		123-51-3	30	Alcohol, nail polish	243.05 ± 5.55 a	8.10	257.55 ± 7.8 a	8.59	248.78 ± 8.49 a	8.29
Acetoin (mg/L)		513-86-0	30	Buttery, cream	69.42 ± 8.61 b	2.31	79.08 ± 4.51 b	2.64	30.92 ± 15.40 a	1.03
Ethyl lactate (mg/L)		97-64-3	100	Strawberry, raspberry, buttery	15.88 ± 0.53 a	0.16	16.08 ± 0.32 a	0.16	15.8 ± 0.10 a	0.16
Diethyl succinate (mg/L)		123-25-1	100	Over-ripe, lavender	10.61 ± 1.76 a	0.11	9.68 ± 2.63 a	0.10	7.95 ± 1.14 a	0.08
2-Phenylethanol (mg/L)		60-12-8	10	Rose, honey	56.98 ± 12.11 a	5.70	49.19 ± 3.82 a	4.92	43.3 ± 1.13 a	4.33
2,3-Butanediol l (mg/L)		24347-58-8	668	Buttery, creamy	550.1 ± 218.87 a	0.82	475.64 ± 147.38 a	0.71	350.3 ± 81.47 a	0.52
2,3-Butanediol m (mg/L)		5341-95-7	668	Buttery, creamy	168.04 ± 83.47 a	0.25	116.45 ± 61.84 a	0.17	79.42 ± 32.82 a	0.12
Glycerol (g/L)		56-81-5		Confers body and smoothness and a sweet taste	6.53 ± 0.11 a	-	6.69 ± 0.1 a	-	6.54 ± 0.06 a	-
Ethyl isobutyrate (mg/L)	SBSE-GCMS	97-62-1	0.015	Sweet, rubber	0.013 ± 0.002 a	0.84	0.014 ± 0.006 a	0.97	0.015 ± 0.006 a	1
Ethyl butyrate (mg/L)		105-54-4	0.02	Fruity, sweet, apple	0.10 ± 0.005 a	5.24	0.11 ± 0.003 a	5.55	0.11 ± 0.009 a	5.50
Hexanal (mg/L)		66-25-1	25	Green	0.006 ± 0.0008 a	0.00024	0.0056 ± 0.0003 a	0.00022	0.005 ± 0.0008 a	0.00021
Butyl acetate (mg/L)		123-86-4	4.6	Banana, ripe pear, glue	0.0021 ± 0.0002 a	0.00046	0.0023 ± 0.0006 a	0.00049	0.0024 ± 0.0006 a	0.00053
Furfural (mg/L)		98-01-1	0.77	Burnt almond, incense, floral	0.51 ± 0.2 a	0.66	0.30 ± 0.19 a	0.39	0.26 ± 0.23 a	0.35
Ethyl 2-methylbutanoate (mg/L)		7452-79-1	0.002	Fruity, strawberry, apple, blackberry	ND	-	ND	0	0.00088 ± 0.00028 a	0.44
Ethyl 3-methylbutyrate (mg/L)		108-64-5	0.0007	Fruity, strawberry, apple	0.0022 ± 0.00018 a	3.14	0.0025 ± 0.00024 a	3.51	0.0023 ± 0.00031 a	3.29
2-Furanmethanol (mg/L)		98-00-0	N.f.	Burnt sugar	0.0035 ± 0.0017 a	-	ND	0	0.0026 ± 0.00040 a	-
Isoamyl acetate (mg/L)		123-92-2	0.03	Banana	2.20 ± 0.13 a	73.42	2.36 ± 0.080 a	78.78	2.31 ± 0.25 a	77.08
3-Heptanone (mg/L)		106-35-4	N.f.	Fruity, green, fatty, sweet	0.00019 ± 0.00005 a	-	0.00068 ± 0.00028 b	-	0.00032 ± 0.000035 a	-
2-Acetylfuran (mg/L)		1192-62-7	N.f.	Sweet, almondy, nutty, brown and toasted	0.017 ± 0.0072 a	-	0.014 ± 0.007 a	-	0.0084 ±0.0047 a	-
γ-Butyrolactone (mg/L)		96-48-0	100	Coconut, caramel	20.69 ± 3.57 b	0.21	12.66 ± 3.81 a	0.13	7.82± 1.04 a	0.078
5-Methylfurfural (mg/L)		620-02-0	1.1	Toasted, bitter almond, spicy	0.11 ± 0.060 b	0.098	ND	0	0.051 ± 0.056 ab	0.046
Ethyl hexanoate (mg/L)		123-66-00	0.005	Banana, green apple	0.45 ± 0.0096 a	89.48	0.48 ± 0.0085 a	96.39	0.48 ± 0.031 a	96.00
Hexyl acetate (mg/L)		142-92-7	0.67	Apple, cherry, pear, floral	0.10 ± 0.0048 a	0.15	0.11 ± 0.0022 a	0.17	0.10 ± 0.010 a	0.15
2-Ethylhexanol (mg/L)		104-76-7	8	Fruity, floral	0.052 ± 0.0056 a	0.0065	0.052 ± 0.0088 a	0.0065	0.039 ± 0.0047 a	0.0049
2-Phenylacetaldehyde (mg/L)		122-78-1	0.001	Flowery, rose	0.013 ± 0.0018 a	13.07	0.013 ± 0.0014 a	12.98	0.014 ± 0.0017 a	14.00
Octanol (mg/L)		111-87-5	0.12	Intense citrus, roses	0.057 ± 0.016 a	0.48	0.045 ± 0.011 a	0.38	0.034 ± 0.011 a	0.28
Acetophenone (mg/L)		98-86-2	0.065	Mildly sweet	0.0028 ± 0.0025 a	0.044	0.0019 ± 0.0021 a	0.030	0.0025 ±0.0011 a	0.038
Ethyl heptanoate (mg/L)		106-30-9	0.0022	Fruity, pineapple, sweet	0.0006 ± 0.00003 a	0.26	0.0006 ± 0.00003 a	0.26	0.0005 ± 0.00007 a	0.23
Nonanal (mg/L)		124-19-6	0.0011	Floral, orange–rose odor or waxy and green	0.0037 ± 0.00095 b	3.42	0.0014 ± 0.0013 a	1.32	ND	0
Ethyl octanoate (mg/L)		106-32-1	0.005	Pineapple, pear, soapy	0.31 ± 0.034 a	62.33	0.37 ± 0.029 b	74.45	0.36 ± 0.016 ab	72.00
Decanal (mg/L)		112-31-2	0.001	Grassy, orange skin-like	0.013 ± 0.00031 a	12.86	ND	0	ND	0
Octyl acetate (mg/L)		112-14-1	50	Apple, mushroom, herbal	0.0012 ± 0.0014 a	0.000024	0.00035 ± 0.00015 a	7.00 × 10−6	ND	0
B-Citronellol (mg/L)		7540-51-4	0.1	Floral	0.039 ± 0.017 a	0.39	0.040 ± 0.0039 a	0.4	0.033 ± 0.0038 a	0.33
2-Phenethyl acetate (mg/L)		103-45-7	0.25	Fruity, rose, sweet, honey	2.30 ± 0.15 a	9.20	2.03 ± 0.16 a	8.12	2.21 ± 0.086 a	8.84
1-Decanol (mg/L)		112-30-1	0.4	Fatty	0.013 ± 0.00031 a	0.032	0.01015 ± 0.0057 a	0.025	0.0099 ± 0.0030 a	0.025
E-Citral (mg/L)		5392-40-5	0.03	Citrus, Lemon	ND	0	ND	0	0.0030 ± 0.0012 a	0.1
2,4-Decadienal (mg/L)		25152-84-5	0.00007–0.01	Fatty, Citrus	0.00030 ± 0.00006 a	-	0.00032 ± 0.00003 a	-	0.00026 ± 0.000035 a	-
2-Methoxy-4-vinylphenol (mg/L)		7786-61-0	0.004	Apple, Spicy	0.24 ± 0.011 a	60	0.24 ± 0.023 a	60	0.22 ± 0.031 a	55
G-Nonalactone (mg/L)		104-61-0	0.03	Coconut, sweet, and stone fruit	0.015 ± 0.0017 a	0.51	0.016 ± 0.0017 a	0.53	0.014 ± 0.0028 a	0.46
Ethyl decanoate (mg/L)		110-38-3	1.1	Sweet, fruity, nuts and dried fruit	0.44 ± 0.11 a	0.40	0.51 ± 0.10 a	0.46	0.54 ± 0.08 a	0.49
E-Geranyl acetone (mg/L)		3796-70-1	0.06	Floral	ND	-	0.00082 ± 0.00007 a	0.014	ND	0
Z-Geranyl acetone (mg/L)		3796-70-1	0.06	Floral	0.0026 ± 0.00013 a	0.044	0.0026 ± 0.00006 a	0.044	0.0025 ± 0.00012 a	0.042
Dodecanol (mg/L)		112-53-8	0.007	Fatty	0.012 ± 0.0016 ab	1.67	0.013 ± 0.0031 c	1.86	0.0054 ± 0.0048 a	0.77
Ethyl undecanoate (mg/L)		627-90-7	N.f	N.f.	0.00061 ± 0.00004 a	-	0.00064 ± 0.00003 a	-	0.00063 ± 0.000021 a	-
Ethyl dodecanoate (mg/L)		106-33-2	3.5	Floral, fruity	0.42 ± 0.049 a	0.12	0.49 ± 0.074 a	0.14	0.53 ± 0.063 a	0.15
2- Phenethyl hexanoate (mg/L)		6290-37-5	N.f	Fruity	0.0013 ± 0.00014 a	-	0.0015 ± 0.00019 a	-	0.0014 ± 0.00022 a	-
Trans-Methyl Dihydrojasmonate		24851-98-7	N.f	N.f.	0.011 ± 0.0012 a	-	0.013 ± 0.0052 a	-	0.0084 ± 0.00059 a	-
Ethyl myristate (mg/L)		124-06-1	494	Mild waxy, soapy	0.039 ± 0.0029 a	0.000078	0.040 ± 0.0038 a	0.00008	0.040 ± 0.0020 a	0
Phenethyl benzoate (mg/L)		94-47-3	N.f	N.f	0.0015 ± 0.00011 a	-	0.0018 ± 0.0004 a	-	0.0015 ± 0.00018 a	-
Ethyl hexadecanoate (mg/L)		628-97-7	1.5	Fatty, rancid, fruity, sweet	0.13 ± 0.005 a	0.086	0.13 ± 0.019 a	0.086	0.12 ± 0.015 a	0.080

). Alcoholic fermentations were concluded in all studied conditions as reported by the high ethanol content and low glucose/fructose concentrations (below 4 g/L). The highest volatile acidity values were observed in the SpF condition, although they were all below 1.2 g/L (Table 3) in all conditions, which indicates that the wines were not acidified by opportunistic acetic acid bacteria or yeasts, which can impact negatively on the organoleptic properties. On the other hand, the pH is slightly higher in the SqFMB and SqF conditions, probably due to lower volatile acidity and, consequently, a lower concentration of acetic acid.

Non-Saccharomyces fermentative yeasts such as M. pulcherrima and Lachancea thermotolerans are employed as starter cultures in must fermentations due to their ability to produce wine with a lower ethanol content, lactic acid by L. thermotolerans, or pulcherrimin by M. pulcherrima, respectively [28,29,30]. Our results indicate that the yeast sequential inoculations employed had little impact on the general oenological parameters, which could be due to a high competition with indigenous microbiota from the must [30].

3.4. Volatilome Analysis

Fifteen different compounds were measured by a gas chromatography-flame ionization detector (GC–FID) with 1,1-diethoxyethane and acetoin accounting for significant differences (p value ≤ 0.05) and quantified over their odor threshold (OT). The presence of forty-two compounds was reported by using SBSE coupled with gas chromatography/mass spectrometry (GC/MS). Twelve of them presented significant differences (p value ≤ 0.05) from which four (nonanal, ethyl octanoate, decanal and dodecanol) surpassed the OT (Table 3).

Assessment of the potential impact of volatile compounds on aroma can be determined by the Odor Activity Value (OAV). The OAV quantifies the significance of a specific compound to the overall odor of a sample. It is calculated as the ratio of the compound to its OT [31]. From the total volatile compounds (major and minor) analyzed, eighteen, sixteen and fifteen overpassed an OAV of 1 in SqFMB, SqF and Spf respectively (Table 3), meaning that these compounds can be perceived. The presence of more volatile compounds surpassing an OAV of 1 in SqFMB and SqF reinforces the hypothesis that sequential fermentation of musts increases the complexity of wines [32,33]. Furthermore, the odor description (Table 3) of compounds that exceeded the odor threshold in SqFMB contributes to a floral and fruity profile, which aligns with one of the study’s objectives.

Each of these metabolites is associated with an aroma descriptor providing a unique character to the wine. 1,1-Diethoxyethane and decanal are associated with green fruit and orange-skin like aroma, respectively, and were only detected in the SqFMB condition with an OAV of 10.56 and 12.86, respectively. Nonanal, associated with floral and orange-rose odor, was found in SqF (OAV of 1.32) and SqFMB (OAV of 3.42), being more concentrated in the second condition. Acetoin, ethyl octanoate and dodecanol were detected in all conditions, from which acetoin and dodecanol were quantified in higher concentrations in both sequential fermentations and are related to buttery and fatty notes, respectively, while ethyl octanoate contributes with fruity aromas like pineapple and pear. These results could be related to the influence of the inoculated yeasts and format. Indeed, mixed cultures of M. pulcherrima and S. cerevisiae produce wines with more ethyl octanoate, nonanal and dodecanol than those fermented with monocultures of both yeast species during alcoholic fermentation of Leon Millot grape juice [34]. Moreover, M. pulcherrima is described as a higher acetoin producer than S. cerevisiae during alcoholic fermentation of grape must, which explains the higher concentration of these compounds in the SqF and SqFMB conditions [29,34]. 1,1-Diethoxyethane is a chemical combination of acetaldehyde with ethanol. This compound is characteristic of velum wines such as Sherry, and in this work, it is only detected at 10.56 ± 0.65 mg/L in the SqFMB condition, which indicates that its production is induced by the cell–hyphae attachment. However, previous use of an S. cerevisiae strain with the MB immobilization technique in an alcoholic fermentation did not produce this compound [11]. These authors used a single yeast species for the fermentation process and the must was previously filtered. For this reason, the results cannot be properly compared so further experiments will be required to confirm if 1,1-diethoxyethane is produced due to this immobilization process.

PCA (Figure 4) was performed with the data matrix built of 62 compounds quantified in three biological replicates and analyzed by triplicate for each of the three variables (SqFMB, SqF and SpF). The performed PCA resulted in three PCs which account for 59.97% of the total variance (28.77% for PC1; 18.38% for PC2; 12.82% for PC3). Supplementary Table S1 contains the contribution (loadings) of each compound to every PC. PC1 is mainly affected by four compounds with loads higher than 0.200 (diethyl succinate, 2-phenylethanol and the levo and meso forms of 2,3-butanediol), while five affect PC2 positively (3-heptanone, hexyl acetate, G-nonalactone, E-geranyl acetone and dodecanol) and one negatively (2-furanmethanol). PC3 is mostly positively influenced by four compounds (ethanol, volatile acidity, ethyl isobutyrate and 2-phenylacetaldehyde) and negatively by two (fixed acidity and z-geranyl acetone). The three-dimensional plot (Figure 4) explains the behavior of each fermentation strategy. Each group is well defined and separated from the rest. SqFMB exhibited positive values in PC1 and PC2 and negative in PC3; SqF presented positive values from each PC, and SpF was located in an area with positive values from PC3 and negatives from PC1 and PC2.

The odorant descriptor and odor threshold were taken from [31,35,36,37,38,39,40,41,42,43,44,45,46].

The use of MB as ICT to carry a sequential fermentation of Pedro Ximénez must result in a dry wine which does not differ significantly in its general oenological parameters from the one obtained with a sequential fermentation of free yeasts or a spontaneous fermentation. This could be related with the selection of a low-fermentative yeast such as D. hansenii in the beginning (T0) of SqMB and SqF conditions [47]. As previously discussed, yeast immobilization provides several advantages over free yeasts, such as cell reutilization and the reproducibility of the process [48]. Our results indicate that MB provides not only a good carrier technology for fermentation but the possibility to obtain a more complex profile of wine, as has occurred in our case, with the detection of 1,1-Diethoxyethane and the higher dodecanol production in the SqFMB condition.

3.5. Sensory Analyses

Table 4. Wine scores from judges (n = 8). SqFMB: Wine obtained from sequential fermentation with microbial biocapsules. SqF: Wine obtained from sequential fermentation with free yeasts. SpF: Wine obtained with spontaneous fermentation with indigenous microbiota. The different letters indicate homogeneous groups which significantly differ statistically in the parameters between wines (p < 0.05, F-test).

Attributes	SqFMB	Sqf	Spf
Visual	6.10 ± 0.68 a	6.11 ± 0.53 a	6.10 ± 0.68 a
Smell	8.77 ± 0.87 a	8.55 ± 2.02 a	8.48 ± 0.35 a
Taste	8.90 ± 0.51 a	8.49 ± 1.09 a	9.16 ± 0.77 a
Overall quality	9.55 ± 0.39 a	9.33 ± 0.58 a	9.89 ± 0.51 a
Total points	83.85 ± 4.85 a	81.2 ± 10.68 a	84.19 ± 4.44 a

shows the means, standard deviations and HG calculated for the sensory attributes (visual, smell, taste and overall quality) evaluated. These data indicate that there are no significant sensorial differences between the wines obtained through different fermentation strategies. The wines obtained present very similar organoleptic profiles, and all of them achieved a score between 80 and 84 points (out of 100) (the wine advocate, 2021) and were rated as barely above average to very good. No one of the judgers detected any deficiencies in any of the wines obtained.

4. Conclusions

The study of different yeast inoculation methods for sequential must fermentations shows that microbial biocapsules could be used to produce more complex wines throughout yeast selection. The microbial biocapsule matrix composed of A. oryzae (pellet-forming fungus recognized as GRAS) seems to be a good yeast carrier for this purpose, as some metabolites, such as 1,1-diethoxyethane and decanal, are only produced with this inoculation method, and others, like nonanal, were detected in higher concentrations. However, if modulation of the main oenological parameters and major volatile compounds is desired, variables like yeast concentration at each timepoint of the sequential fermentation, yeast preadaptation methods, and the use of yeast species with higher fermentative power need to be improved to overcome the indigenous microbiota competition in unfiltered musts.

In conclusion, while our study primarily addressed the qualitative aspects of winemaking by focusing on organoleptic evaluations and identifying methodological improvements, we acknowledge the importance of meeting both qualitative and quantitative criteria for successful winemaking. Once these methodological refinements are fully implemented, we anticipate that the optimized process will not only enhance the sensory quality of the wine but also improve its potential for price-competitive production. Future work will focus on validating the process against both criteria to ensure the production of high-quality, cost-effective wine.

5. Patents

The microbial biocapsule assembling methodology is subjected to a provisional patent (Application number: 63/411,843).