Volatilomic and Sensorial Profiles of Cabernet Sauvignon Wines Fermented with Different Commercial Yeasts

Abstract

Volatilomic and sensory analyses of wine are excellent tools for enologists and winemakers when selecting commercial yeast based on the production of metabolites related to desirable wine characteristics. Integrating this holistic approach could lead to the terroir description, characterization, and quality control improvement of the vinification process. Volatilomic and sensory profiles of Cabernet Sauvignon Mexican wines fermented with three commercial yeasts (WLP740, ICVD254, and ICVD80) were obtained using HS-SPME-GC-qTOF/MS and CATA evaluation. A total of 100 volatile compounds were identified, with unique entities per strain. WLP740 wines were rated as high quality, presenting fruity and minty aromas with fewer off-aromas, while ICVD254 wines showed higher levels of compounds associated with off-notes and were rated as low quality. ICVD80 wines were of medium quality, with fruity esters and higher alcohols descriptors. Volatilomic profiles highlighted the role of specific compounds in differentiating strains and sensory attributes, while yeast selection significantly impacts wine aroma and quality. The authors acknowledge the need for further analyses, including an increased sample size, yeast species, diverse vineyards, and vinification processes, which will result in a solid and robust methodology.

1. Introduction

Cabernet Sauvignon, a globally highly esteemed red wine grape variety, is celebrated for its complex aromatic profile, which significantly influences consumer perception and preference. Such complexity arises from a multitude of volatile compounds, including esters, alcohols, acids, terpenes, and norisoprenoids, synthesized through grape metabolism (primary aromas), fermentation processes (secondary aromas), and chemical reactions while aging (tertiary aromas) [1,2,3,4]. Thus, understanding the volatilome, the complete set of volatile compounds, of Cabernet Sauvignon (CS) is essential for enhancing wine quality and tailoring sensory attributes to meet market demands [5,6,7,8,9].

Primary aromas are described as floral and fruity, while secondary aromas arise from metabolism by yeast, and the aromas formed as byproducts of lactic acid bacteria in malolactic fermentation (MLF) have buttery, creamy, and lactic descriptors; tertiary aromas are described as complex aromas, comprising wood, spices, caramel, and smoke [1,2,3]. Saccharomyces cerevisiae, the principal yeast species used in winemaking, converts sugars into ethanol and contributes to the formation of a diverse array of volatile compounds, mainly esters and higher alcohols, that define a wine’s aroma [3,4,10]. The selection of specific yeast strains can influence the concentration and composition of these volatiles, allowing winemakers to modulate sensory characteristics [10]. For instance, different commercial yeast strains have been shown to produce varying levels of ethyl esters and higher alcohols, leading to distinct fruity and floral notes in the final product [7,11,12]. Also, winemakers are focusing on mixed cultures of Saccharomyces and non-Saccharomyces to enhance floral and fruity aromas, reduce off-aromatic compounds, improve the nitrogen content in must, and take advantage of their β-glucosidase and proteolytic activities [13,14,15]. Furthermore, the role of yeast in shaping wine’s pyrazine content, a compound associated with green and herbaceous aromas, has been highlighted in recent studies, illustrating how strain choice can mitigate or enhance these sensory attributes [16].

Terroir encompasses the unique combination of soil, climate, topography, and vineyard microbial biodiversity, profoundly affecting the grape composition and, consequently, the wine’s volatilome [17,18,19]. In recent years, warmer temperatures and lower water availability have impacted the grapevine vegetative cycle and, consequently, grape maturation, which increased the sugar content, decreased the flavonoid content and titratable acidity, and therefore led to a higher pH. A fermentative technique to reduce alcohol content in wine due to a high sugar content consists of using Saccharomyces and non-Saccharomyces yeasts that are able to metabolize sugars through other pathways producing organic acids and aromatic compounds [13,20,21]. The microbial terroir, mainly native yeast populations on grape surfaces and in the winery environment, contributes to the distinctive aromatic qualities of wines from different regions [22]. Studies have demonstrated that autochthonous S. cerevisiae strains can enhance regional wine characteristics by producing unique volatile profiles compared to commercial strains [23,24]. Additionally, the interaction between terroir and winemaking techniques, such as fermentation temperature and yeast inoculation, plays a significant role in determining the final wine aromatic expression. Terroir influence extends to non-volatile precursors transformed during fermentation, further enhancing wine’s complexity [19,22,25].

Quality control during winemaking is crucial to ensure consistency and quality in the final product. Volatilomic analysis through advanced analytical techniques, such as gas chromatography coupled with quadrupole time-of-flight mass spectrometry (GC-qTOF-MS), enables the precise identification and quantification of volatile compounds. This analytical approach facilitates the assessment of fermentation dynamics, the impact of yeast strain selection, and the influence of terroir on wine aroma. By implementing volatilomic analyses, winemakers can make informed decisions to optimize fermentation conditions, select appropriate yeast strains, and implement quality control measures that enhance the desired aromatic properties of Cabernet Sauvignon wines [5,23,26,27]. Complements of GC-qTOF-MS with other advanced methods, such as liquid chromatography coupled with quadrupole time-of-flight mass spectrometry (LC-qTOF-MS), can help to identify non-volatile precursors of aromatic compounds to further deepen our wine chemistry understanding [28].

In addition to yeast and terroir, post-fermentation processes, including malolactic fermentation and aging, contribute significantly to the volatilome. Malolactic fermentation, often carried out by lactic acid bacteria, reduces wine acidity and adds buttery and nutty aromas by producing diacetyl and other volatiles [29]. Aging in oak barrels or stainless steel further influences the wine’s aromatic profile by introducing compounds such as vanillin, lactones, and tannins, which enhance its complexity and structure [30]. The integration of these processes underlines the importance of a holistic approach to winemaking, where each step plays a role in shaping the wine’s final aromatic profile.

This study analyzed the volatilome and sensorial profiles of Cabernet Sauvignon wines from the same wine region, fermented with three different commercial S. cerevisiae yeast strains. Statistical analysis using principal component analysis (PCA) was successfully applied to identify entities that influence the strain wine composition. This study aimed to elucidate how yeast strain selection influences the composition of volatile compounds, focusing on aromatic properties and the expression of terroir in the wines.

2. Materials and Methods

2.1. Cabernet Sauvignon Young Wines

Commercial vinification of young dry Cabernet vintage 2021 wines from San Vicente Valley (Ensenada, Mexico) was achieved by two wineries. Three commercial yeasts, WLP740 (White Labs, Inc., San Diego, CA, USA), ICVD80, and ICVD254 (LALLEMAND, Inc., Montreal, QC, Canada), were inoculated (as suggested by the manufacturer) for alcoholic fermentation (AF) in stainless steel tanks, with a controlled temperature, and pumping over two times per day (

Table 1. Vinification conditions.

Tank	Harvest TSS	Yeast	AF (Days)	AF T (°C)	PFM (Days)
1 (69 hL)	22.5 °Brix	WLP740	12	20	3
2 (107 hL)	22.2 °Brix	ICVD80	12	21	6
3 (25 hL)	23.0 °Brix	ICVD254	12	22	6

TSS: Total soluble solids; AF: alcoholic fermentation; AF T: alcoholic fermentation temperature; PFM: post-fermentation maceration.

). After AF, the must was pressed and transferred to another steel tank for malolactic fermentation (MF), then it was stabilized, bottled, and sealed with a natural cork. Wine samples were collected (two 750 mL bottles per tank) and stored under wine cellar conditions (16–18 °C and relative humidity of 80%) until sensory analysis. Upon bottle opening, volatilome samples were collected in 50 mL tubes and stored at −50 °C until analyzed.

2.2. Volatilome Analysis

2.2.1. Data Acquisition (HS-SPME-GC-qTOF/MS)

Wine samples were analyzed in triplicate using a 7890B GC System (Agilent Technologies, Santa Clara, CA, USA), coupled to an autosampler PAL3 System with a head-space solid-phase micro-extraction (HS-SPME) module (CTC Analytics AG, Zwingen, Switzerland), attached to a 7200 quadrupole-time-to-flight mass spectrometer (qTOF/MS) (Agilent Technologies, Santa Clara, CA, USA). A 50/30 µm DVB/CAR/PDMS Stable Flex fiber (Supelco, Bellefonte, PA, USA) was used for volatile extraction. The chromatographic separation of the volatile wines was performed on a DB-WAX (30 m/250 µm/0.25 µm) GC column (Agilent Technologies, Santa Clara, CA, USA). Detailed parameters are specified in our previous work [5]. Briefly, 10 mL of sample and 3 g of NaCl were added to a 20 mL amber SPME glass vial and sealed with an aluminum cap and an 18 mm blue PTFE/silicone septum (Agilent Technologies, Santa Clara, CA, USA). Sample extraction was performed at 40 °C/30 min after conditioning at 40 °C/5 min; desorption was completed at 240 °C/10 min in splitless mode; for separation, the flow rate was set at 0.8 mL/min using He as carrier gas (RT locked with 2-Undecanone at 35.461 min). The oven’s initial temperature was at 40 °C for 5 min, increasing 3 °C/min to 180 °C, then 30 °C/min to 220 °C for 10 min; detection was set in a mass range of 30 to 400 amu at an acquisition rate of 2.5 spectra/s, 70 eV with an electron ionization (EI) source at 230 °C in profile mode.

2.2.2. Quality Control

Quality control (QC) was assessed by analyzing pooled wine (PW) and pooled wine spiked (PWS) with 2.5 ng/L of α-pinene, p-cymene (Honeywell Fluka™, Morristown, NJ, USA), β-pinene, and 2-undecanone (Sigma-Aldrich, St. Louis, MO, USA) standards. PW and PWS were analyzed at the beginning, middle, and end of each batch sequence. Mass calibration was also performed initially and after running five samples to ensure mass accuracy. Samples were analyzed in triplicate and randomized.

2.2.3. Data Processing/Mining and Identification

A recursive analysis was performed (explained in detail in [5]) using Agilent Mass Hunter WorkStation Unknowns Analysis software version 10.0 to perform the extraction, alignment, and identification of components present in all wine samples. The Accurate Mass Flavours Database [31] and NIST 17 were used to identify components and to create an internal database with the Retention Time (RT), Retention Index (RI), compound name, and mass spectrum with the exact mass to ease component extraction, alignment, and identification (level 2 and 4, as proposed by Sumner et al., 2007 [32]). Level 2 identification was assigned when the match factor was >70, exact-mass accuracy < 20 ppm of the mass spectrum, and ΔRI < 30; if one of these criteria was not met, the component was identified as Unknown + RI, which is classified as a level 4 identification.

2.2.4. Data Analysis and Interpretation

Extracted, aligned, and identified components were imported to Agilent Mass-Hunter WorkStation Mass Profiler Professional version 15.0 for data interpretation and analysis. A completely randomized design was considered; samples were grouped by yeast (WLP740, ICVD80, and ICVD254), normalized with Median Fold Change (PQN) using PW, and the baseline was transformed with Pareto scaling. A principal component analysis (PCA) was performed on all entities present in at least 60% of all samples in at least one group, using analysis of variance, covariance, and standard deviations. Also, a Venn diagram was made to obtain all the entities shared by the yeasts, and a one-way ANOVA with Tukey HSD test (p < 0.05) was performed; relative abundance data were represented as a stacked column chart made in Microsoft Excel version 2025.

2.3. Sensory Analysis

A descriptive double-blind sensorial analysis was performed by a panel of six winemakers (6 to 45 years of experience) in a tasting room at the Autonomous University of Baja California, Ensenada Campus. The tasting room was fitted with black curtains, and panelists’ seats were distributed in such a way as to avoid communication between them. To reduce wine color bias on odor and aroma descriptors, red lights were placed above each seat. Wine samples (50 mL) were served in standard wine glasses in a randomized order. Global quality (GQ) parameters were determined in three different categories: (a) nose parameters (odor intensity and odor complexity); (b) mouthfeel parameters (aroma intensity, aroma complexity, equilibrium and body, aromatic persistence); and (c) visual parameters (hue and color intensity). Quantitative parameters were evaluated on a 7-level scale, as described elsewhere [33], where 1 is “null quality” and 7 is “maximum quality”; meanwhile, qualitative parameters (descriptors) were modified to fit Cabernet Sauvignon wines in a Check All That Apply (CATA) method (Table S1). Verbally informed consent was obtained from the panelists for their participation and their consent to publish this paper. The sensory analysis procedure was established following the recommendations of [33]. Panelists’ training on the procedure was accomplished with aged Cabernet Sauvignon wines from San Vicente Valley as a preliminary study. Descriptive analysis and a one-way ANOVA were performed to evaluate the effect of yeast type on the global quality score, and the comparison of means was performed by Tukey’s test at a significance of p < 0.05, using NCSS 2007 Software (Kaysville, UT, USA).

2.4. Tasting Panel

The sensory panel consisted of six wine experts (five males and one female), including enologists and winemakers, with 6 to 45 years of professional experience and ages ranging from 40 to 66 years. To ensure quality control in the sensory evaluation, the first sample was presented again at the end of the session. Panelist repeatability was required to meet a threshold of less than 15% relative standard deviation (RSD) and a standard deviation below 1 in GQ scores. Based on these criteria, one panelist (Panelist 4) was disqualified from the analysis (

Table 2. Quality control for sensory analysis.

Panelist	Expertise (Years)	Age (Years)	GQ1	GQ2	GQ $\overline{\mathbf{X}}$ (SD)	RSD
1	6	40	4.52	4.00	4.26 (0.37)	8.63
2	44	68	3.35	4.13	3.74 (0.55)	14.75
3	45	62	5.80	5.35	5.57 (0.32)	5.71
4 *	30	65	5.10	3.55	4.32 (1.10)	25.34
5	6	47	4.76	4.66	4.71 (0.07)	1.50
6	25	66	2.76	2.30	2.53 (0.33)	12.86

GQ1: global quality of first wine; GQ2: global quality of repetition; $\overline{\mathrm{X}}$: mean; SD: standard deviation; RSD: relative standard deviation. * Panelist disqualified.

).

3. Results

3.1. Volatilome

Volatilome analysis of young Cabernet Sauvignon wines from San Vicente Valley fermented with commercial yeast strains revealed a total of 100 entities (

Table 3. Metabolites detected in wines fermented with ICVD254, ICVD80, and WLP740.

			ICVD254		ICVD80		WLP740
#	RI	Compound	Ab	%Ab	Ab	%Ab	Ab	%Ab	Aromatic Descriptors
1	873	Ethyl formate	4.5 × 106 c	0.081	3.7 × 106 b	0.100	2.9 × 106 a	0.070	Like grapes, cognac, and melon
2	907	Ethyl acetate	4.9 × 108	8.844	7.2 a × 108	19.466	8.7 × 108	20.996	Anise, balsam, ethereal, fruity
3	913	Methyl alcohol	1.2 × 107	0.213	1.3 × 107	0.350	1.5 × 107	0.366	Alcoholic, sweet, pungent
4	965	Ethyl propanoate	5.6 × 106	0.101	4.6 × 106	0.124	3.4 × 106	0.084	Sweet, fruity, rum, juicy, grape
5	972	Ethyl isobutyrate	1.6 × 107	0.286	1.6 × 107	0.441	ND	ND	Fruity, sweet, ethereal
6	1021	Isobutyl acetate	8.1 × 106	0.147	7.4 × 106	0.202	7.7 × 106	0.188	Apple, banana, bitter
7	1043	Ethyl butyrate	4.4 × 107 a	0.802	6.2 × 107 b	1.678	7.3 × 107 b	1.764	Apple, banana, cognac
8	1047	1-Propanol	5.4 × 107	0.983	4.4 × 107	1.196	4.7 × 107	1.143	Alcoholic, fermented, musty
9	1051	Methyl thiolacetate	ND	ND	7.3 × 105	0.020	6.5 × 105	0.016	Sulfurous, eggy, cheese, dairy
10	1061	Ethyl 2-methylbutyrate	1.8 × 106	0.033	4.1 × 106	0.110	2.5 × 106	0.061	Sharp, sweet, green, apple
11	1062	2,3-Pentanedione (3)	ND	ND	ND	ND	1.0 × 106	0.025	Butter, caramel, cheese, cream
12	1067	Ethyl 3-methylbutyrate	2.6 × 106 a	0.048	7.2 × 106 b	0.197	3.7 × 106 a	0.089	Fruity, sweet, apple, pineapple
13	1109	Isobutyl alcohol	4.4 × 108 a	8.037	1.3 × 108 b	3.610	9.1 × 107 b	2.205	Bitter, ether, solvent, wine
14	1128	Isoamyl acetate	1.6 × 108 a	2.954	4.3 × 108 b	11.799	4.4 × 108 b	10.756	Banana, bitter, fruity, solvent
15	1161	1-Butanol	9.1 × 106	0.166	8.9 × 106	0.241	1.03 × 107	0.250	Fruity, fusel, medicine, oil
16	1181	Isoamyl propionate (2)	ND	ND	2.2 × 106	0.061	ND	ND	Sweet, banana, pineapple, ripe
17	1184	Methyl hexanoate (3)	ND	ND	ND	ND	9.7 × 105	0.023	Ethereal, pineapple, apricot
18	1188	D-Limonene (1)	4.0 × 105	0.007	ND	ND	ND	ND	Lemon, orange
19	1212	Unknown 1212 (1)	1.7 × 109	30.594	ND	ND	ND	ND	N/A
20	1216	2-Methyl-1-butanol	8.7 × 105	0.016	4.0 × 105	0.011	4.2 × 105	0.010	Malt
21	1236	Ethyl hexanoate	5.7 × 108	10.291	2.8 × 108	7.645	8.3 × 108	20.133	Apple peel, banana, green
22	1263	Styrene (1)	1.1 × 106	0.020	ND	ND	ND	ND	Balsamic, floral, gasoline, plastic
23	1267	1-Pentanol (1)	4.0 × 106	0.072	ND	ND	ND	ND	Balsam, fusel, oil, sweet
24	1273	Isoamyl butyrate (3)	ND	ND	ND	ND	1.1 × 106	0.027	Fruity, green, apricot, pear
25	1282	Hexyl acetate(	9.7 × 106	0.175	1.0 × 107	0.282	1.60 × 107	0.387	Apple, banana, bitter, green
26	1310	Acetoin	1.6 × 106 a	0.030	2.6 × 106 a,b	0.072	5.0 × 106 b	0.121	Sweet, buttery, creamy, dairy
27	1303	Isopentyl 3-methylbutyrate (3)	ND	ND	ND	ND	1.3 × 106	0.032	Apple, green, mango, ripe, sweet
28	1307	Ethyl 3-hexanoate (1)	1.0 × 106	0.018	ND	ND	ND	ND	Fresh, fruity, green, metallic
29	1335	Hexyl formate	2.7 × 106	0.049	5.3 × 106	0.145	3.2 × 106	0.079	Nutty
30	1348	4-Methyl-1-pentanol	5.0 × 106 a	0.091	1.5 × 107 b	0.412	9.4 × 106 b	0.229	Apple, plum, banana, sweet
31	1349	Ethyl heptanoate	7.4 × 106 b	0.134	2.2 × 106 a	0.059	1.3 × 106 a	0.031	Balsam, creamy, fruity, milky
32	1358	Ethyl 2-hexanoate (1)	3.7 × 106	0.068	ND	ND	ND	ND	Fruity, green, juicy, rum, sweet
33	1359	Ethyl lactate	1.1 × 108	2.065	1.2 × 108	3.333	8.2 × 107	1.985	Butter, butterscotch, fruit, sharp
34	1373	1-Hexanol	3.2 × 108 c	5.861	2.0 × 108 b	5.442	1.3 × 108 a	3.076	Alcoholic, ethereal, floral, fruity
35	1384	trans-3-Hexen-1-ol	2.9 × 106 b	0.052	1.7 × 106 a,b	0.046	1.1 × 106 a	0.027	Green, cortex, leafy, floral, petal
36	1394	3-Ethoxy-1-propanol (1)	1.1 × 106	0.020	ND	ND	ND	ND	Fruit
37	1403	cis-3-Hexen-1-ol (1)	2.3 × 106	0.042	ND	ND	ND	ND	Green, leafy
38	1404	Methyl octanoate	2.0 × 106	0.036	2.2 × 106	0.060	1.7 × 106	0.042	Aldehydic, green, herbal, orange
39	1410	3-Octanol	1.5 × 106	0.027	ND	ND	1.2 × 106	0.028	Earthy, mushroom, dairy, musty
40	1423	Unknown 1423 (1)	3.8 × 105	0.007	ND	ND	ND	ND	N/A
41	1434	Unknown 1434 (1)	5.6 × 105	0.010	ND	ND	ND	ND	N/A
42	1436	Unknown 1436	8.5 × 105	0.015	8.7 × 105	0.024	ND	ND	N/A
43	1444	Unknown 1444	8.7 × 105	0.016	7.4 × 105	0.020	1.4 × 106	0.033	N/A
44	1453	Ethyl octanoate	5.9 × 108 b	10.745	5.1 × 106 a	0.140	5.1 × 108 b	12.319	Apricot, banana, brandy, fat, fruity
45	1460	1,3-Dichlorobenzene	1.1 × 106	0.019	9.3 × 105	0.025	9.2 × 105	0.022	N/A
46	1466	1-Octen-3-ol	4.1 × 106 b	0.073	4.0 × 106 b	0.107	2.8 × 106 a	0.069	Earthy, fishy, fungal, green
47	1474	1-Heptanol (3)	ND	ND	ND	ND	5.5 × 106	0.133	Chemical, coconut, green, herbal
48	1475	Isoamyl hexanoate (2)	ND	ND	2.0 × 106	0.053	ND	ND	Apple, banana, green, pineapple
49	1478	Acetic acid	1.1 × 108	2.074	8.4 × 107	2.290	5.7 × 107	1.390	Pungent, sharp, sour, vinegar
50	1482	Unknown 1482	1.02 × 106	0.019	4.8 × 105	0.013	5.5 × 105	0.013	N/A
51	1505	2-Ethyl-1-hexanol	5.2 × 105 a	0.009	3.8 × 106 b	0.102	3.7 × 106 b	0.089	Citrus, floral, fresh, green, oily
52	1521	Unknown 1521 (1)	6.1 × 105	0.011	ND	ND	ND	ND	N/A
53	1527	Unknown 1527	2.7 × 106 c	0.048	1.5 × 106 b	0.040	1.0 × 106 a	0.025	N/A
54	1535	2-Nonanol	2.2 × 106	0.041	2.8 × 106	0.078	ND	ND	Cheese, citrus, creamy, cucumber
55	1547	Unknown 1547 (1)	1.10 × 106	0.020	ND	ND	ND	ND	N/A
56	1549	Unknown 1549	ND	ND	1.3 × 106	0.037	6.8 × 105	0.017	N/A
57	1549	Ethyl nonanoate	1.5 × 106	0.026	2.0 × 106	0.053	ND	ND	Fruity, natural, rose, rum, tropical
58	1557	2,3-Butanediol	5.3 × 107	0.955	4.7 × 107	1.275	3.9 × 107	0.948	Buttery, creamy, fruit, onion
59	1559	Ethyl dl-2-hydroxycaproate	1.6 × 106 a	0.029	2.7 × 106 b	0.073	1.2 × 106 a	0.028	N/A
60	1562	beta-Linalool	1.3 × 106	0.023	1.5 × 106	0.042	1.1 × 106	0.026	Blueberry, citrus, floral, flower
61	1574	1-Octanol	6.3 × 106 a	0.113	8.0 × 106 b	0.219	5.1 × 106 a	0.124	Aldehyde, burnt, chemical, green
62	1585	Isoamyl lactate	1.9 × 106 a	0.035	3.5 × 106 b	0.094	1.3 × 106 a	0.032	Fruity, creamy, nutty
63	1593	Propylene Glycol	1.8 × 107	0.323	1.7 × 107	0.453	1.1 × 107	0.261	Alcoholic, odorless
64	1611	Unknown 1611	1.8 × 106	0.033	2.6 × 106	0.072	ND	ND	N/A
65	1610	Methyl decanoate (3)	ND	ND	ND	ND	3.6 × 105	0.009	Floral, fruity, oily, wine
66	1633	Unknown 1633	6.3 × 105	0.011	6.2 × 105	0.017	ND	ND	N/A
67	1637	Ethyl 2-furoate	1.5 × 105	0.003	1.6 × 105	0.004	1.5 × 105	0.004	Fruity, floral
68	1646	Oxolan-2-one	1.6 × 106	0.029	1.6 × 106	0.044	1.1 × 106	0.026	Creamy, oily, fatty, caramel
69	1649	Ethyl methylbutanedionate (1)	2.90 × 105	0.005	ND	ND	ND	ND	No flavor nor odor reported
70	1657	Ethyl decanoate	2.2 × 107 a	0.394	2.0 × 108 b	5.560	1.2 × 108 a	2.814	Apple, brandy, fruity, grape, oily
71	1676	1-nonanol	4.5 × 106 a,b	0.081	6.4 × 106 b	0.174	3.5 × 106 a	0.086	Bitter, clean, dusty, fatty, floral
72	1676	Ethyl benzoate	4.5 × 105	0.008	3.3 × 105	0.009	ND	ND	Fruity, dry, musty, sweet
73	1691	Diethyl succinate	2.1 × 108 b	3.761	3.1 × 108 c	8.321	1.5 × 108 a	3.522	Apple, apricot, chocolate, cooked
74	1702	3-Nonen-1-ol (1)	2.8 × 105	0.005	ND	ND	ND	ND	Fresh, waxy, green, melon rind
75	1707	Ethyl 9-decenoate	1.9 × 106 a	0.035	7.8 × 106 b	0.212	8.0 × 106 b	0.194	Fruity, fatty (1)
76	1715	alpha-Terpineol	2.7 × 105	0.005	3.2 × 105	0.009	1.8 × 105	0.004	Anise, citrus, floral, minty, pine
77	1735	3-(methylthio)-1-propanol	1.3 × 106 a	0.024	4.0 × 106 b	0.108	1.8 × 106 a	0.044	Onion, potato, soup, sulfurous
78	1738	Unknown 1738	2.6 × 105	0.005	2.3 × 105	0.006	2.3 × 105	0.005	N/A
79	1758	Unknown 1758	2.7 × 105	0.005	ND	ND	1.8 × 105	0.004	N/A
80	1768	Unknown 1768	3.9 × 106	0.071	9.6 × 106	0.260	4.4 × 106	0.107	N/A
81	1777	1-Decanol	1.1 × 106	0.020	1.4 × 106	0.037	9.1 × 105	0.022	Clean, fat, fatty, floral, orange
82	1781	Citronellol	4.2 × 105	0.008	5.6 × 105	0.015	3.0 × 105	0.007	Citrus, floral, geranium, leather
83	1795	Methyl salicylate	2.0 × 105 b	0.004	2.4 × 105 b	0.006	1.5 × 105 a	0.004	Wintergreen, minty
84	1806	Ethyl phenylacetate	4.5 × 105 b	0.008	4.4 × 105 b	0.012	2.4 × 105 a	0.006	Sweet, floral, honey, rose, cocoa
85	1807	Unknown 1807	5.4 × 105 b	0.010	6.9 × 105 b	0.019	2.5 × 105 a	0.006	N/A
86	1816	Unknown 1816	2.0 × 106	0.037	3.3 × 106	0.089	2.3 × 106	0.055	N/A
87	1825	Phenethyl acetate	1.7 × 106 a	0.031	5.3 × 106 b	0.144	4.9 × 106 a,b	0.120	Floral, fruity, honey, rose, sweet
88	1830	beta-Damascenone	3.7 × 105	0.007	7.0 × 105	0.019	7.6 × 105	0.019	Apple, honey, rose, smoky, sweet
89	1841	Ethyl dodecanoate	4.9 × 105 a	0.009	4.1 × 106 b	0.112	3.1 × 106 a,b	0.075	Clean, floral, leafy, soapy, sweet
90	1854	3-Methylbutyl decanoate (2)	ND	ND	2.0 × 105	0.005	ND	ND	Waxy, banana, fruity, sweet
91	1864	Benzyl alcohol	3.8 × 106 b	0.070	5.5 × 106 c	0.149	2.6 × 106 a	0.062	Balsamic, cherry, floral, flower
92	1879	Ethyl isopentyl succinate (1)	2.0 × 106 a	0.036	6.4 × 106 b	0.173	2.2 × 106 a	0.053	Found in wine and beer
93	1887	Unknown 1887	2.5 × 105	0.005	ND	ND	ND	ND	N/A
94	1887	2-Phenylethanol	4.5 × 108 a	8.103	7.9 × 108 b	21.388	5.2 × 108 a	12.571	Rose, honey, sweet berry, hyacinth
95	2006	4-Ethylguaiacol (1)	4.3 × 105	0.008	ND	ND	ND	ND	Bacon, clove, phenolic, smoky
96	2123	Unknown 2123	7.5 × 106	0.135	ND	ND	1.6 × 107	0.401	N/A
97	2093	Octanoic acid (2)	ND	ND	2.8 × 107	0.762	ND	ND	Cheese, cheesy, fatty, oily, rancid
98	2197	4-Ethylphenol	3.2 × 106 c	0.057	4.8 × 105 b	0.013	3.1 × 105 a	0.007	Phenolic, castoreum, smoky
99	2313	Unknown 2313 (2)	ND	ND	3.2 × 105	0.009	ND	ND	N/A
100	2323	2,4-Di-tert-butylphenol	1.7 × 105	0.003	2.7 × 105	0.007	2.0 × 105	0.005	Phenolic

# of compounds elucidated. Unique metabolites of ⁽¹⁾ ICVD254, ⁽²⁾ ICVD80, and ⁽³⁾ WLP740; RI: Retention Index; Ab= Relative abundance; %Ab= relative abundance percentage. Means with different letters within a row indicate statistical differences (p < 0.05) in compounds present in all wines.

). Wines fermented with ICVD254 showed 87 entities, compared to 75 in ICVD80 and 72 in WLP740. Interestingly, each wine exhibited unique compounds; 16 were exclusive to ICVD254 (including five Unknown components), 5 were unique to ICVD80, and 6 were unique to WLP740. A shared core of 61 compounds was found across all wine samples, although relative abundance varied (Table 3).

Principal component analysis (PCA) grouped wines according to the yeast strain, explaining 66.23% of the total variance (PC1= 36.21% and PC2 = 30.02%), underscoring the impact of yeast on volatile profiles (Figure 1a). The distribution of compound loadings (Figure 1b) and their sensorial attributes allowed for the interpretation of PCA quadrants. PC1 score was related to less favorable aromas for young wines, meanwhile PC2 was related to floral and fruity characteristics. The PC1-/PC2- quadrant comprises fruity, floral, butter, and sharp aromas; PC1+/PC2- comprises balsamic, fresh, lemon, and orange; PC1-/PC2+ has sweeter characteristics, such as apple and banana; lastly, PC1+/PC2+ yielded fresh, citrus, and balsamic characteristics (Supplementary Materials, Table S2). This classification indicates that wines fermented with WLP740 exhibited more floral, fruity, and buttery notes, while ICVD254 wines were richer in balsamic and fresh characteristics. ICVD80 wines displayed a profile similar to ICVD254, but with a higher abundance of fruity and floral metabolites (Figure 1a). The most abundant compounds per yeast were ethyl acetate (#2) for WLP740, 2-phenylethanol (#94) for ICVD80, and Unknown 1212 (#19) for ICVD254. Notably, Unknown 1212 was detected only in ICVD254 wines and may be tentatively identified as 3-methyl-1-butanol (level 4 identification by Sumner et al., 2007 [32]) when comparing its mass fragmentation and Retention Index with the NIST Chemistry WebBook [34].

Several unique compounds in ICVD254 (level 2 identification, Sumner et al., 2007 [32]) such as D-limonene (#18), ethyl 3-hexanoate (#28), and 3-ethoxy-1-propanol (#32), which present fruity, citrus, green, leafy, and rum-like aromas, appeared during alcoholic fermentation and persisted through vinification. Similarly, these compounds, including Unknown 1212, were absent in the musts and wines of ICVD80 and WLP740. Conversely, compounds such as 3-methylbutyl decanoate (#90; waxy, banana, fruity, cheesy) and octanoic acid (#97; fatty, oily) were exclusive to ICVD80, while WLP740 wines uniquely featured methyl hexanoate (#17; fruity, pineapple, strawberry) and isopentyl 3-methylbutyrate (#27; apple, mango), along with 2,3-pentanedione (#11; butter, caramel), 1-heptanol (#47; coconut, herbal), and methyl decanoate (#65; floral, fruity).

The abundance of volatile compounds, their interaction with the wine matrix, and their odor thresholds are critical to the wine’s aromatic profile [2]. While unique compounds highlight the impact of specific yeast strains, common metabolites likely reflect the influence of terroir, encompassing grape origin, fermentation, and vinification techniques [17,19,22]. The 61 compounds common to all wines are therefore attributed to Cabernet Sauvignon wines from San Vicente. Although different commercial yeasts were used, yeast-derived esters often have a greater impact on varietal differentiation than fruit-derived esters [35]. Also, some varietal aromas only emerge during fermentation due to the hydrolysis of carbohydrate-bound precursors [36]. ICVD80 wines showed significantly higher concentrations (p < 0.05) of compounds such as ethyl 3-methyl butyrate (#12), ethyl dl-2-hydroxy caproate (#59), ethyl decanoate (#70), 1-nonanol (#71), diethyl succinate (#73), 3-(methylthio)-1-propanol (#77), benzyl alcohol (#91), ethyl isopropyl succinate (#92), and 2-phenylethanol (#94). In contrast, ethyl formate (#1), isobutyl alcohol (#13), 1-hexanol (#34), Unknown 1527 (#53), and 4-ethylphenol (#98) were significantly more abundant in ICVD254 wines (Figure 2).

Overall, 32 compounds were discriminated between wines fermented with ICVD254, ICVD80, and WLP740 (p < 0.05, Figure 3). Notably, 1-hexanol (#34) and 4-ethylphenol (#98) were most abundant in ICVD254 wines (p < 0.001). These compounds are associated with balsamic, phenolic, smoky, and creamy notes, with minor fruity or floral nuances at lower concentrations. Compounds such as ethyl formate (#1), ethyl heptanoate (#31), and trans-3-hexen-1-ol (#35) followed a similar trend, with the highest abundance in ICVD254, moderate abundance in ICVD80, and the lowest abundance in WLP740.

3.2. Sensory Analysis

The sensory evaluation assessed the global quality (GQ) of wines based on nasal (odor) and retronasal (aroma) intensity and complexity, as well as balance, body, aromatic persistence, and color (

Table 4. Relative quality of wines.

Parameters (Values)	WLP740 (RelAb)	ICVD80 (RelAb)	ICVD254 (RelAb)
Nose Parameters (30%)
RQ odor intensity (12%)	0.50	0.45	0.41
RQ odor complexity (18%)	0.81	0.63	0.68
Mouthfeel parameters (60%)
RQ aromatic intensity (10%)	0.47	0.45	0.38
RQ aromatic complexity (15%)	0.71	0.56	0.48
RQ balance and body (25%)	1.12	0.94	0.75
RQ aromatic persistency (10%)	0.52	0.37	0.32
Appearance parameters (10%)
RQ color (6%)	0.36	0.34	0.31
RQ color intensity (4%)	0.24	0.21	0.18
Global Quality	4.75 a High	3.96 a,b Medium	3.52 b Low

RQ: Relative quality; RelAb: Relative abundance. Different letters in global quality scores indicate significant differences (p < 0.05).

). The global quality of wines can be assessed as maximum quality (6.17–7.00), very high quality (5.31–6.15), high quality (4.46–5.30), medium quality (3.61–4.45), low quality (2.76–3.60), very low quality (1.91–2.75), and null quality (1.00–1.90) based on their quantitative scores.

Descriptive analysis focused on aromatic markers of each wine. All samples prominently exhibited berry notes in both odor and aroma, with overall similarities in olfactory perception but greater differences in retronasal complexity. Wine balance—defined as the harmonious relationship among its sensory components—is essential for consumer acceptance. While typically linked to sweetness, acidity, bitterness, and saltiness, wine balance also includes astringency and alcoholic sensations [37] (Figure 4). In this context, wines fermented with WLP740 had greater balance compared to those fermented with ICVD80 and ICVD254.

4. Discussion

Volatilomics is a relatively new term used in metabolomics to elucidate a wide range of volatile metabolites [8]. Its application in wine has grown significantly, as its relation to sensory profiling is widely reported in [7,38]. Young Cabernet Sauvignon (CS) wines are typically described as tannic, complex, and structured with good balance and acidity. Their aromatic profiles, referred to as varietal characteristics, commonly include black currant, green bell pepper, mint, cassis, and sometimes eucalyptus notes [39]. These sensory profiles are influenced by the region of cultivation, vinification techniques [26,27,40], and microbial activity, mainly fermented with Saccharomyces cerevisiae species. However, the volatilomic impact of both commercial and non-commercial strains remains insufficiently explored [23,25,41,42]. In this study, CS wines fermented with WLP740 presented a strong smell of berries, followed by mint, vanilla, and eucalyptus. During feedback sessions, panelists identified these as essential characteristics of CS from Ensenada, Baja California. Additionally, panelists classified WLP740 wines as high quality, citing their balanced profile, and higher aromatic intensity, complexity, and persistency (Table 4 and Figure 4). Although eucalyptus was a perceived descriptor across all wines, neither eucalyptol nor menthol, two compounds typically responsible for eucalyptus notes, were detected. This observation aligns with the understanding that aroma perception depends not only on individual compounds but also on their thresholds, odor activity values, and metabolite interactions that could mask another aroma [1,2,43,44].

Eucalyptus aromas can also be described as fresh, minty, and citrusy, as the compounds present in the PC1+/PC2+ quadrant (Figure 1), where ICVD254 is distributed. Cassis aroma was the second most frequently mentioned descriptor and is linked to fruity, green, herbaceous, and balsamic aromas, corresponding to the compound distribution seen in WLP740 and ICVD80 wines. The high-quality classification of WLP740 wines was associated with their balanced aromatic profiles, including berries, mint, vanilla, and eucalyptus, and their lower abundance of off-aroma compounds compared to ICVD80 and ICV254 (Table 4 and Figure 3). Furthermore, ICVD80 wines were rated as medium quality due to their greater aromatic intensity, complexity, and persistence compared to ICVD254 low-quality wines, even though there were no statistical differences (p > 0.05, Table 4). ICVD80 wines exhibited more esters linked to fruity and floral notes, but also a higher abundance of higher alcohols associated with less desirable descriptors such as balsamic, leafy, and green, which are typically less favorable in young dry red wines (Table 3).

Industrial S. cerevisiae strains, believed to descend from a few ancestral strains, were genetically and phenological differentiated through food fermentations and evolved into separate lineages, each used for specific applications, such as wine-making [45]. ICVD254 and ICVD80 were first isolated in the 1990s from different regions in France, south and north from Rhône, respectively (https://www.lallemand.com/). In contrast, limited information is published regarding the regional origin or metabolic characteristics of WLP740, a liquid-form yeast commercialized by White Labs (https://www.whitelabs.com/). While commercial yeasts are selected mainly for their sugar fermentation capabilities, glucosidase activity, and killer factor presence, understanding their influence on local musts is essential for winemaking decisions.

Recent studies have examined the impact of commercial S. cerevisiae strains on varietal wines by evaluating their chemical composition and sensory attributes. For instance, Shiraz wines from South Australia fermented with ten different commercial S. cerevisiae strains presented differences in tannin, polyphenol, and polysaccharide contents. Even though targeted volatile analysis could not discriminate between all strains, sensory analysis linked astringency, color, and opacity with tannin concentration. Notably, yeast contributions to aromatic profiles have been studied to enhance wine quality [3,10]; however, studies on their impact across grape varietals and regions, especially in Mexican wines, remain scarce.

In China, research on CS wines has explored how the terroir influences their chemical and sensory profiles. Targeted volatile analysis did not discriminate between five subregions in Ningxia, despite distinct soil characteristics [41]. However, indicators such as chromaticity, total phenols, ethyl isobutyrate, n-decanoic acid, and (-)-epigallocatechin were linked to regional identity, influenced largely by winemaking techniques [46]. To characterize terroir effectively, the role of vinification, particularly microbial contributions to secondary metabolite formation, must be considered [25]. Correlation between aroma compounds and different yeasts, Saccharomyces spp. and non-Saccharomyces, has been identified and used to discriminate between winemaking regions [22,23,25,42]. Aroma-related metabolites such as 2-methyl-1-butanol, 3-methyl-1-butanol, phenethyl ethanol, isoamyl acetate, ethyl caproate, and ethyl caprylate were used to discriminate between CS produced in different regions inoculated with native yeasts. Furthermore, ethyl esters like ethyl acetate, ethyl hexanoate, ethyl caprylate, and ethyl heptanoate were negatively correlated with non-Saccharomyces yeasts; meanwhile, β-damascenone, α-ionone, 4-terpineol, and 1-nonanal had a positive correlation [25].

β-damascenone, a yeast-derived compound linked in CS wines to wine region production [5,23,27,29,47], was detected in all samples, showing no difference between ICVD254, ICVD80, and WLP740 wines (p > 0.05), reinforcing that all wines were from the same region (San Vicente Valley, Ensenada, Mexico). In Chinese CS wines, higher alcohols were more abundant than ethyl esters, although ethyl lactate was the most prevalent compound [29]. Similar compounds, including isobutyl alcohol, 2-phenylethanol, 1-propanol, 1-pentanol, ethyl acetate, ethyl lactate, isopentyl acetate, phenethyl acetate, ethyl hexanoate, ethyl octanoate, ethyl decanoate, isoamyl acetate, isoamyl hexanoate, and isoamyl octanoate, have been documented across CS producing regions [5,27,29] and were also found in this study. These compounds, mainly higher alcohols and ethyl esters, are bacterial and yeast metabolic byproducts [48,49,50,51]. Therefore, their profiles can offer insights into the microorganisms involved in vinification.

Wines fermented with ICVD254 were perceived as low quality. Notably, 1-hexanol (#34) and 4-ethylphenol (#98), both associated with off-aromas in wines depending on their concentration [3,52,53], were more abundant in ICVD254 wines compared to those fermented with ICVD80 and WLP740 (Figure 3). 1-hexanol is synthesized by yeasts via the Erlich pathway, the abundance of which can vary depending on the yeast strain, must, amino acid composition, and bacterial communities during the vinification process [29,53,54]. 4-ethylphenol can be produced by S. cerevisiae in low amounts; however, it is primarily associated with Brettanomyces/Dekkera, a spoilage yeast introduced through overripe or contaminated grapes [52,55,56]. After spontaneous MLF, a common process in red wine vinification, 1-hexanol and 4-ethylphenol abundance increased by 5.2% and 57.2%, respectively [54]. Descriptors for these aromas include balsamic, creamy, alcoholic, ethereal, green, leafy, phenolic, and smoky notes, with minor floral and fruity notes at low abundances. Furthermore, ethyl formate (#1), ethyl heptanoate (#31), and trans-3-hexen-1-ol (#35) share 1-hexanol and 4-ethylguaiacol tendencies (ICVD254>ICVD80>WLP740), contributing to the wines’ lower sensory quality.

Isobutyl alcohol (#13), which is associated with bitter, ether, alcohol, and solvent aroma descriptors, negatively correlates with strawberry, lactic, red fruit, coconut, wood, and vanilla aromas [43,44]. Defining wine aroma is a complex process as it is affected by numerous volatile compound interactions, whose abundances and presence are influenced by many factors. For instance, isobutyl alcohol and 3-methyl-1-butanol have been related to a negative role in wine aroma quality, as their interactions suppress fruity and woody aromas, while leaving leather or inky characteristics unaffected [44]. Based on these findings, it is plausible that compound #19 (Unknowns 1212, tentatively identified as 3-methyl-1-butanol) interacted with isobutyl alcohol in ICVD254 wines, suppressing fruity aromas, masking favorable fruity notes, and contributing to their low quality. It is noteworthy that commercial yeast producers recommend blending wines fermented with ICVD254 and ICVD80 to enhance wine complexity (https://www.lallemandwine.com/). Also, this pair of higher alcohols reduces the perceptible impact of 2-phenylethanol (#94) [44], a compound with rose, honey, sweet berry, and hyacinth descriptors [49,56,57], further diminishing the wine’s aromatic appeal.

Compounds associated with apple, banana, sweet, buttery, apricot, plum, and brandy aromatic descriptors, such as diethyl succinate (#73), ethyl decanoate (#70), ethyl butyrate (#7), acetoin (#26), ethyl octanoate (#44), and isoamyl acetate (#14), were more abundant in WLP740 than in ICVD80, which, in turn, showed higher abundance than ICVD254. Also, WLP740 had the lowest abundance of 1-hexanol (#34) and isobutyl alcohol, both of which are commonly linked to off-aromas, potentially contributing to its perceived high quality. Although ICVD80 wines had similar fruity and floral aromatic descriptors to those of WLP740, they also showed a higher abundance of higher alcohol-related off-aromas, which may explain their medium quality classification.

The WLP740 yeast strain yielded Cabernet Sauvignon wines from San Vicente, Ensenada, México, with more intense fruity and floral attributes and fewer off-aroma descriptors compared to those fermented with ICVD254. Fruity and floral characteristics are primarily associated with acetate and ethyl esters, which are synthesized via amino acid and fatty acid metabolism, respectively [1,3,56]. The formation of these esters is influenced by multiple factors, making their prediction during fermentation challenging. Acetate ester production is closely linked to higher alcohols, whose synthesis increases at elevated fermentation temperatures; however, excessively high temperatures can reduce the overall abundances of both acetate and ethyl esters [56,57]. Moreover, low oxygen conditions can enhance higher alcohol formation [1,58]. Therefore, optimized vinification practices are essential to preserve ester-derived aromas. In this study, fermentation with ICVD254 occurred at higher temperatures and for a longer duration than with WLP740, likely contributing to a lower ester concentration in the resulting wines and, consequently, a less favorable aromatic profile.

5. Conclusions

Volatilomic and sensory analyses of wine are excellent tools for enologists and winemakers in selecting commercial yeasts based on the production of metabolites related to desirable wine characteristics. Applying this holistic approach could lead to terroir description and characterization, which may include evaluating and selecting wild yeast strains that are capable of enriching regional wine terroir. These methods are sensible enough for other uses, such as terroir descriptors for a grape variety grown in different locations/regions. The authors acknowledge that this study represents an initial step in understanding the impact of commercial Saccharomyces cerevisiae strains on the volatile and sensory profiles of regional wines. While these findings provide valuable insights into the influence of these commercial yeasts on Cabernet Sauvignon wines from the San Vicente Valley, additional replicates across multiple vintages, vineyards, and environmental conditions are necessary to strengthen the conclusions and improve reproducibility. Nonetheless, this work lays a solid foundation for future research and contributes meaningfully to the ongoing characterization of how commercial yeast strains interact with regional terroir, offering winemakers data-driven tools to guide strain selection in quality-focused production.