Major Volatile Aroma Composition, Organic Acids, and Color Characteristics: A Comparative Study of Berries and Wines from Vitis amurensis and Its Interspecific Hybrids

TL;DR

This study compares the aroma, acid, and color profiles of wines from Vitis amurensis and its hybrids, identifying key compounds that contribute to wine quality and color stability.

Contribution

The study identifies specific volatile aroma compounds and organic acids unique to Vitis amurensis that enhance wine quality and color stability.

Findings

Hexanal is the primary aroma compound in Vitis amurensis berries.

1-Hexanol and isoamyl acetate distinguish V. amurensis wines from hybrids.

V. amurensis wines retain more pigments and anthocyanins, resulting in intense red-purple color.

Abstract

Understanding the chemical basis for the quality differentiation of wines is essential for breeding and quality control. This study performed a comparative analysis of volatile aroma compounds (VACs), organic acids, and color characteristics in berries and wines from Vitis amurensis (V. amurensis), its interspecific hybrids, and Vitis labrusca (V. labrusca). Hexanal was identified as the primary contributor to grape aroma in V. amurensis berries. 1-Hexanol and isoamyl acetate were the key aroma marker distinguishing wines produced from V. amurensis from its interspecific hybrids. V. amurensis exhibited enhanced metabolism of C6 alcohols, aldehydes, and esters during fermentation. Its wines showed the highest pigment retention and significantly greater anthocyanin content, resulting in the most intense and stable red-purple color compared to other genotypes. Correlation analyses indicate high anthocyanin retention and high color intensity in wines produced from V. amurensis were associated with its high organic acid levels, particularly influenced by tartaric acid, lactic acid, and citric acid. This work clarifies the chemical evolution underlying V. amurensis wine’s sensory superiority and supports breeding, quality control, and product development.

Figures

Full text

1. Introduction

Vitis amurensis Rupr., a native Chinese grapevine species used for wine production, exhibits exceptional cold hardiness [1], high levels of bioactive compounds [2,3], and a unique aroma profile [4]. As a cold-climate wine grape, it holds significant economic and geographical importance in China. Since the 1950s, the Institute of Special Animal and Plant of CAAS has bred several V. amurensis cultivars, including ‘Beibinghong’, ‘Beiguohong’, ‘Zijing’ series, etc.

Grape cultivar is a key determinant of wine sensory quality, with aroma, acidity, and color serving as primary indicators [5]. Different cultivars impart distinct flavor profiles; for example, Cabernet Sauvignon is known for green and spicy notes attributed to 4-ethylguaiacol, cis-2-exen-1-ol, 1-octen-3-ol, 1-hetanol, and (Z)-aok-lactone [6], while Beibinghong ice wine exhibits honey, floral, and peach aromas driven mainly by β-damascenone, linalool, and lactones [7]. Although over 1000 volatile compounds have been identified in wine, only a limited number significantly influence sensory perception due to differences in odor thresholds and synergistic effects [8,9].

Organic acid composition also varies considerably among cultivars. While V. vinifera grapes such as Cabernet Sauvignon typically contain 6–8 g/L of total acids [10,11], V. amurensis berries exhibit notably higher acidity (11.85–22.95 g/L) [4]. Furthermore, V. amurensis wines contain 5 to 10 times more total anthocyanins than V. vinifera wines, resulting in intense color saturation [12]. Beyond varietal traits, fermentation plays a critical role in shaping wine composition through microbial activity that modifies aroma, organic acids, and color stability [13,14]. Thus, both grape cultivar and fermentation processes are essential in defining wine flavor [15,16,17]. Though current research remains limited to several cultivars [18,19,20,21], comparative studies on the flavor characteristics of wines produced from V. amurensis and its interspecific hybrids remain limited.

HS-GC-IMS is a widely adopted technique for the comprehensive profiling of volatile compounds in complex matrices such as wine [12,22]. Multivariate statistical analyses, including principal component analysis (PCA) and orthogonal partial least squares discriminant analysis (OPLS-DA), are employed to visualize sample clustering, identify differential metabolites, and elucidate the key volatile markers responsible for sensory distinctions among different cultivars or treatments [7].

Wines produced from V. amurensis represents a distinctive category within the Chinese wine sector. While its overall production remains limited, this variety has established a stable industry presence in core growing regions—such as the Tonghua region of Jilin Province—largely attributable to the unique characteristics of this grape cultivar [23]. The quality of wines produced from V. amurensis is formally recognized through certifications such as the Geographical Indication (GI) designation, which reinforces the strong linkage between its origin and sensory profile [24]. Consumer demand for wines produced from V. amurensis has been growing steadily, fueled by increasing interest in regional specialties, perceived health benefits, and flavor diversity [25]. The typicity of wines produced from V. amurensis—marked by a characteristic aroma, intense color, and high acidity—is intrinsically associated with the terroir of its cold-climate origins, which serves as a key geographical determinant of its style [4]. Building on this context, the present study hypothesizes that V. amurensis and its interspecific hybrids undergo distinct metabolic shifts during fermentation, and that their enhanced sensory properties originate from a defined chemical basis. This hypothesis was examined through systematic compositional and statistical analyses to better understand the role of fermentation in shaping the sensory attributes of the resulting wine. Therefore, this study aims to investigate the aroma, organic acid, and color profiles of berries and wines from V. amurensis and its interspecific hybrids using HS-GC-IMS, HPLC, and CIELab analyses, combined with multivariate statistics. The study elucidates the chemical evolution underlying V. amurensis wine’s sensory superiority, providing a theoretical basis for breeding, quality control, and product development.

2. Materials and Methods

2.1. Vineyard Site and Characteristics

Grape berries of Vitis amurensis and its interspecific hybrids were harvested at the commercial maturity stage on 22 September 2020, from the National Vitis amurensis Germplasm Repository in Zuojia, Jilin Province, China (44°04′ N, 126°05′ E). The vineyard is characterized by a gently south-facing slope with dark brown forest soil. The regional climate is a temperate continental monsoon climate. The meteorological data for the year 2020 were as follows: an average annual temperature of 3.6 °C, with maximum and minimum temperatures of 34 °C and −28 °C, respectively; an annual precipitation of 679 mm; and an average relative humidity of 67%. The vines were trained to a unilateral cordon system with a planting density of 1.0 m (within rows) × 2.5 m (between rows). Conventional vineyard management practices for fertilization, irrigation, and pest control were applied uniformly across the experimental block.

2.2. Chemicals

Yeast CEC01 (food grade) was purchased from Angel Yeast Co., Ltd. (Yichang, China). Potassium metabisulfite (food grade) was obtained from SAS SOFRALAB OENOFRANCE (Magenta, France). Sodium hydroxide and phosphoric acid (analytical grade) were supplied by Beijing Chemical Industry Group Co., Ltd. (Beijing, China). Methanol (chromatographic grade) was acquired from Tedia Company, Inc. (Fairfield, OH, USA). 4-Methyl-2-pentanol (chromatographic grade) was purchased from Shanghai Lianshuo Biotechnology Co., Ltd. (Shanghai, China). Organic acids (chromatographic grade), including malic acid, succinic acid, citric acid, glacial acetic acid, lactic acid, and tartaric acid, were obtained from Shanghai Yuanye Bio-Technology Co., Ltd. (Shanghai, China).

2.3. Raw Materials

A total of 16 grape cultivars were utilized in this study, comprising 7 V. amurensis (Beiguolan (BGL), Beiguohong (BGH), Zuoshanyi (ZSY), Zuoshaner (ZSE), Shuanghong (SH), Shuangyou (SY), Shuangfeng (SF)), 4 V. amurensis–V. vinifera hybrid (Beibinghong (BBH), Xuelanhong (XLH), Zuohongyi (ZHY), Zuoyouhong (ZYH)), 4 V. vinifera–V. amurensis hybrid (Beimei(BM), Beihong (BH), Beichun (BC), Gongniang No. 1 (GNYH)), and 1 V. labrusca (Beida (BD)) cultivar. The full list of cultivars, including their abbreviations, is provided in Table S1. For sampling, nine vines per cultivar with moderate growth vigor were selected and tagged. Three vines constituted one biological replicate, resulting in three replicates per cultivar. From each vine, thirty berries were sampled from the upper, middle, and lower canopy regions. The collected berry samples were immediately transported back to the laboratory, rapidly frozen in liquid nitrogen, and stored at −80 °C for subsequent biochemical analysis.

2.4. Detection of Basic Agronomic Traits and Physicochemical Properties of the Grape Berry

For each cultivar, five clusters were randomly selected. The length and width of these clusters were measured. From each cluster, 100 berries (20 berries per cluster) were randomly sampled for berry trait analysis. Individual berry quality was determined by weighing. Berry diameter and length were measured using a digital vernier caliper. The soluble solids content (SSC) of the fresh grape berries and wines was measured using a handheld refractometer. The total acidity was determined titrimetrically according to the method of Cao [12]. Data for the basic agronomic traits are presented in Table S2.

2.5. Preparation of Wine Samples

The winemaking process was carried out following the method of Cao [12] with modifications. For each cultivar, 8 kg of processed must was fermented in 12 L stainless steel tanks, with three biological replicates per cultivar. The grapes clusters were sorted, washed, crushed, and destemmed. The resulting must was transferred to fermentation tanks, and sulfur dioxide was added at a concentration of 50 mg/L. The must underwent a 24 h cold maceration at 16 °C. Subsequently, the temperature was raised to 26 ± 1 °C for alcoholic fermentation, which was initiated by inoculating with 200 mg/L of activated CEC01 yeast. Alcoholic fermentation was considered complete when the soluble solids content stabilized at approximately 8° Brix. Upon completion of alcoholic fermentation, the temperature was lowered to 23 °C for a one-month period of post-fermentation. Subsequently, the wine was bottled and stored at 16 °C. Finally, samples were aliquoted into 15 mL centrifuge tubes and stored at −80 °C for subsequent analysis.

2.6. Volatile Aroma Analysis

Volatile aroma compounds in fresh grape berries and wine samples were determined by headspace gas chromatography–ion mobility spectrometry (HS-GC-IMS) (G.A.S., Dortmund, Germany), following the method of Cao [12] with modifications. For berry samples, the fresh fruits were homogenized, and 1 mL of the mash was accurately weighed into a 20 mL headspace vial. For wine samples, 1 mL of the supernatant was directly transferred to the vial. Subsequently, 10 μL of 4-methyl-2-pentanol (20 ppm) was added to each vial as an internal standard. The vials were incubated at 60 °C with agitation at 500 rpm for 10 min for headspace generation. A 100 μL aliquot of the headspace gas was automatically injected into the HS-GC-IMS system via a heated syringe (65 °C). Separation was performed on a WAX capillary column (15 m × 0.53 mm, 1 μm film thickness) maintained at 60 °C. High-purity nitrogen was used as the carrier gas, with the flow rate programmed as follows: 2 mL/min for the initial 2 min, increased to 10 mL/min from 2 to 10 min, then raised to 100 mL/min from 10 to 20 min, and finally maintained at 100 mL/min from 20 to 30 min. The total runtime was 30 min. The temperature of the ion mobility spectrometer was set at 45 °C. Qualitative identification was primarily performed by matching against the instrument’s built-in NIST and IMS databases. In GC-IMS, a single volatile compound often appears as two adjacent spots of monomers (M) and dimers (D), which were considered as different states of the same compound. There relative concentration were summed for calculating the relative content. Quantitative analysis was conducted by calculating the ratio of each component’s peak area to that of the internal standard, thereby determining their relative concentrations.

2.7. Relative Odor Activity Value (rOAV) Assessment

rOAV is a relative index used to evaluate the contribution of volatile compounds to the overall flavor in food [26]. The calculation formula is as follows:

$$[eqn]$$

C_i_ represents the relative concentration of compound i (µg/L); T_i_ represents the odor threshold of compound i (µg/L); C_max_ represents the relative concentration of the compound with the highest OAV value (µg/L); T_max_ represents the odor threshold of the compound with the highest OAV value (µg/L). Compounds with rOAV ≥ 1.0 are considered key aroma compounds, while those with 0.1 ≤ rOAV < 1.0 are regarded as compounds with modifying effects on the aroma.

2.8. Electronic Nose Analysis

Volatile aroma compounds were determined using a PEN3 electronic nose (Airsense Analytics Co., Ltd., Schwerin, Germany) equipped with an array of ten metal oxide sensors with varying specificities. The sensor characteristics are as follows: W1C (aromatic compounds), W5S (nitrogen oxides), W3C (aromatic amines), W6S (hydrogen), W5C (alkanes, aromatic compounds), W1S (short-chain alkanes), W1W (terpene compounds), W2S (alcohols, aldehydes, ketones), W2W (organic sulfides, terpenes), and W3S (substances > 100 mg/kg). For measurement, 1.5 mL of each sample was transferred into a 10 mL headspace vial and equilibrated at room temperature for 30 min. The headspace gas was then automatically sampled, and the sensor responses were recorded for 120 s. After each analysis, the sensor chamber was purged with clean air until the signals returned to baseline prior to the introduction of the next sample. The entire procedure was conducted following the method described by Liu [22].

2.9. Organic Acids Analysis

Organic acids were determined using a High-Performance Liquid Chromatography (HPLC) system (Agilent Technologies, Inc., Beijing, China) equipped with a UV-DAD fluorescence detector. The method was adapted from Cao [12] with minor modifications. Separation was achieved on a Eternal XT-C18 column (4.6 mm × 250 mm, 5 μm) maintained at 25 °C. The mobile phase consisted of methanol and an aqueous solution (pH 2.3) in a ratio of 3:97 (v/v), delivered under isocratic conditions at a flow rate of 0.4 mL/min. The injection volume was 10 μL, and detection was performed at a wavelength of 210 nm. Qualitative analysis is performed by comparing the retention times and characteristic spectra of target compounds in the sample with organic acid reference standards; quantitative analysis of the target compounds is conducted using the external standard method with a calibration curve. This method effectively avoids co-elution and exhibits high stability and accuracy.

Individual stock solutions of tartaric acid (10.0090 g/L), succinic acid (2.0040 g/L), citric acid (1.0100 g/L), glacial acetic acid (10.4475 g/L), lactic acid (10.6128 g/L), and L-malic acid (6.0600 g/L) were prepared in methanol and diluted to volume in 10 mL volumetric flasks. These stock solutions were subsequently diluted by factors of 2, 4, 8, 10, and 16 to construct a series of calibration standards (Table 1). All standard solutions were filtered through 0.22 μm membranes prior to HPLC analysis. Grape berry (after deseeding and homogenization) and wine samples were diluted twofold with the aqueous mobile phase, filtered through a 0.22 μm syringe filter, and then subjected to HPLC analysis.

2.10. Color Analysis and Characterization

2.10.1. Determination of Total Anthocyanins

Measurement of total anthocyanin content was carried out by the pH differential method [27]. Sample pretreatment: For grape samples, the entire berry was ground into a homogenate. For wine samples, the liquid was first filtered through a 0.22 μm aqueous phase membrane and then set aside for subsequent analysis. A volume of 1 mL of sample was transferred to a 10 mL centrifuge tube and mixed with 9 mL of pH 3.0 sodium citrate–citric acid buffer. The mixture was vortexed thoroughly and then equilibrated in the dark for 60 min. Following this, two separate dilutions were prepared: 1 mL of the above diluted sample was added to 9 mL of KCl-HCl buffer (pH 1.0), and another 1 mL was added to 9 mL of NaOAc-HAc buffer (pH 4.5). Both solutions were mixed well and allowed to react in the dark for 90 min. The absorbance of each solution was then measured at 518 nm and 700 nm using a LAMBDA 365 UV/Vis spectrophotometer (PerkinElmer Corporate Management (Shanghai) Co., Ltd., Shanghai, China).

The total anthocyanin concentration was calculated using the following formula:

$$[eqn]$$ $$[eqn]$$

C represents the total anthocyanins concentration (unit: mg/L), MW represents the molecular weight of malvidin-3-glucoside (493.2), DF represents the dilution factor of the sample solution, ε represents the molar extinction coefficient of malvidin-3-glucoside (28,000 L·mol^−1^·cm^−1^), l represents the diameter of the optical path of the cuvette (1 cm), and A represents the absorbance value of the sample.

2.10.2. Analysis of Wine Color

The color of the wine samples was characterized using the CIE (1976) Lab color space, following a procedure adapted from the method of Zhang [14] with minor modifications. Briefly, each wine sample was initially diluted sixfold with distilled water. The diluted sample was then filtered through a 0.22 μm aqueous phase membrane and subjected to full-wavelength scanning from 400 nm to 700 nm using a spectrophotometer. A 1 cm path length quartz cuvette was employed for the measurement, with distilled water serving as the blank control. Absorbance values were specifically recorded at 450 nm, 520 nm, 570 nm, and 630 nm for the subsequent calculation of the CIELab parameters: L* (lightness), a* (red-green component), b* (yellow-blue component), and C*ab (chroma).

For digital color visualization, a separate aliquot of each wine was diluted threefold with distilled water. The corresponding CIELab parameters (L, a*, b*) obtained from the spectrophotometric analysis were input into Color Express version 1.0.0.1700 (TEMBO EXPRESS (HONG KONG) LIMITED, Hong Kong, China) to generate representative color cards.

2.11. Statistical Analysis

All experimental data are presented as the mean ± standard deviation from three independent replicates. Normality was assessed using the Shapiro-Wilk test. Data conforming to a normal distribution were analyzed with parametric tests (independent samples t-test or one-way ANOVA) to compare group differences, while non-normal data were analyzed using non-parametric alternatives (Mann-Whitney U or Kruskal-Wallis H tests). All analyses were performed in SAS 9.4 (SAS Institute Inc., Cary, NC, USA), with p < 0.05 considered statistically significant. A radar chart was constructed based on the statistical results using Microsoft Excel 2010. For multivariate analysis, Orthogonal Projections to Latent Structures Discriminant Analysis (OPLS-DA) was conducted using SIMCA 14.1 to identify key variables based on Variable Importance in Projection (VIP) values. Additionally, Principal Component Analysis (PCA), clustering heatmaps, and differential bar charts were generated using the Metware Cloud platform https://cloud.metware.cn (accessed on 29 July 2025), a freely accessible online tool for data analysis.

3. Results and Discussion

3.1. Aroma Analysis

HS-GC-IMS was employed to analyze the aroma compounds in berries and wines of 16 grape cultivars, including V. amurensis, its interspecific hybrids, and V. labrusca. A total of 72 volatile compounds were detected, among which 40 were unambiguously identified. Specifically, 36 aroma compounds were identified in berry samples, while 23 were identified in the corresponding wines (Figure 1, Table S3 [4,28,29,30,31,32,33,34,35]).

3.1.1. Aroma Characteristics of 16 Cultivars of Berries

A comprehensive analysis of the 36 volatile compounds identified in the berry samples revealed profound differences in volatile aroma compounds among the 16 cultivars. The composition of aroma classes differed significantly (p < 0.05), with the average relative content across cultivars ranking as follows: aldehydes (9.98–81.69%) > esters (1.17–71.64%) > alcohols (13.83–36.65%) > ketones (0.82–7.31%) > acids (0.94–2.28%) (Figure 1a, Table S4). Notably, esters and aldehydes showed the highest coefficients of variation (CV), highlighting their significant contribution to aroma diversity across cultivars—a pattern attributable to genetic differences (Table 2). Based on the volatile aroma compound profiles, 16 cultivars were classified into three distinct groups by cluster heat map analysis (Figure 2): Group 1 (ZYH, ZHY), Group 2 (BD, SY, SF, SH, BH, BC), and Group 3 (ZSE, BGH, BBH, BGL, BM, ZSY, GNYH, XLH). The grouping strongly reflected genetic relationships. In Group 1, ZYH and ZHY share a close pedigree. All cultivars in Group 2 contain V. amurensis genes. In Group 3, BGH shared a consistent aroma profile with its parental line, ZSE. Significant intra-specific variation was observed. For example, among the V. amurensis-V. vinifera hybrids, XLH and BBH exhibited a low-ester trait, whereas ZHY and ZYH displayed a high-ester trait. Furthermore, although BM, BH, and BC (all V.v_V.a) share an identical parentage (Muscat Hamburg × V. amurensis), BM exhibited significantly higher concentrations of 3-methylbutanal and 2-methylbutanal (p < 0.05), highlighting how minor genetic or epigenetic differences can lead to distinct aroma phenotypes. Notably, V. amurensis and some of its interspecific hybrids exhibited high levels of C6 aldehydes, including (E)-2-hexenal and hexanal, which are synthesized via the lipoxygenase (LOX) pathway as regulated by LOX genes [36,37,38]. This enrichment was particularly prominent in V. amurensis, which aligns with previous findings by Li [4]. This consistent pattern not only defines the core herbaceous aroma of this species but also supports the hypothesis that these volatiles are integral to its constitutive chemical defense system, aligning with its pest and disease resistance [39]. Trace ethanol detected in berry samples likely originated from both enhanced alcohol dehydrogenase (ADH) activity during post-harvest cold storage and natural fermentation occurring in fruit tissues prior to incubation [40].

Volatile compounds are classified based on their relative odor activity value (rOAV): compounds with rOAV ≥ 1.0 are considered key aroma compounds with dominant sensory impact, whereas those with 0.1 ≤ rOAV < 1.0 are regarded as modifying compounds that contribute to aroma complexity and nuance [26]. A total of five key aroma compounds (hexanal, 3-methylbutanal, 2-methylpropanal, 2-methylbutanal, and (E)-2-hexenal) and one modifying compound (methyl 2-methylbutanoate) were identified in the berry samples (Table S6). Hexanal exhibited the highest odor activity, particularly in V. amurensis, indicating its major contribution to the primary contributor to herbaceous notes. (E)-2-hexenal, also a major herbaceous odorant, similarly showed elevated rOAVs in V. amurensis. Although the relative content of (E)-2-hexenal was significantly higher than that of hexanal in berries across different cultivars, the lower odor threshold of hexanal resulted in its greater contribution to the herbaceous aroma at the sensory level. The compounds 3-methylbutanal, 2-methylbutanal, and 2-methylpropanal were identified as important contributors to smoky aroma nuances. Notably, the rOAV ranges of these five key aroma compounds were generally higher in V. amurensis and its interspecific hybrids than in V. labrusca, indicating that V. amurensis and its interspecific hybrids possess more pronounced herbaceous and smoky aroma characteristics compared to V. labrusca. Meanwhile, methyl 2-methylbutanoate displayed consistent rOAVs across 16 cultivars, playing a modest modifying role with fruity undertones. The above results indicate that the aroma characteristics are significantly influenced by the genetic background [41,42].

3.1.2. Aroma Characteristics of Wines

Volatile aroma profiling of the 16 wines (Table S5) showed that alcohols and esters were the predominant compounds, accounting for 64.38–71.24% and 24.51–29.41% of the total volatiles, respectively, followed by aldehydes (2.15–3.19%), ketones (0.51–2.32%), and acids (0.33–1.03%).

A total of 14 key odor-active compounds (rOAV > 1) were identified (Table S7), primarily ketones, alcohols, and esters, contributing to smoky, herbaceous, alcoholic, and fruity aroma notes [32]. Cluster heatmap analysis of the 14 key odor-active compounds revealed significant differences (p < 0.05) in the enrichment patterns of these odorants among wines from different species (Figure 3a). The clustering effectively discriminated the four wine types: wines from V. amurensis and its interspecific hybrids showed significantly lower levels of ethyl acetate, ethyl propanoate, and ethyl butanoate, but higher levels of other esters (isoamyl acetate, ethyl isobutyrate, butyl acetate, ethyl 3-methylbutanoate) and alcohols (ethanol, 3-methyl-1-butanol) compared to wines produced from V. labrusca (p < 0.05). These differences primarily corresponded to variations in fruity and alcoholic attributes (Figure 3b–e). Notably, wines produced from V. amurensis exhibited the highest enrichment of 1-hexanol, 3-pentanone, isoamyl acetate, ethyl isobutyrate, and butyl acetate, highlighting their prominent herbaceous and fruity character. Wines produced from V. amurensis interspecific hybrids showed the highest similarity in aroma composition.

To investigate the overall aromatic characteristics of wines from different groups, the electronic nose was employed for detection. Radar plots of the electronic nose data indicated broadly similar aromatic profiles across the samples, albeit with varying response intensities (Figure 4a–d). Sensors W1S, W5S, W2W, and W2S exhibited strong responses, suggesting these as key features potentially reflecting the high content of alcohols, aldehydes, ketones, and esters-consistent with the odor-active compounds identified. Electronic nose data reveal that there are differences of orders of magnitude in the overall odor signal strength among different cultivars. A considerable number of volatile compounds have been scientifically confirmed to possess specific sensory properties. There exists a potential association between the sensory attributes of these volatile compounds and evaluators’ cognitive perceptions, including affective and memory-related factors. Sensory evaluation methods based on descriptive analysis may introduce substantial variation in experimental results due to individual differences among assessors [43,44]. The electronic nose, valued for its high sensitivity, speed, objectivity, and non-destructive nature, is suitable for distinguishing overall aromatic profiles. However, it is susceptible to environmental interference and struggles to accurately identify individual components within complex mixtures.

The OPLS-DA performed on the relative content of volatile aroma compounds revealed that wines produced from V. amurensis were distinct from the wines of V. amurensis interspecific hybrids, and V. labrusca, reflecting significant differences in their aroma profiles (Figure 5a–c). The model’s validity was confirmed through 200 permutation tests.

Key discriminant compounds were identified through sequential screening based on VIP > 1, p < 0.05, and rOAV > 0.1, revealing distinct differential markers across comparisons: 1-Hexanol, isoamyl acetate, ethyl acetate, and 3-pentanone for V. amurensis versus V. amurensis-V. vinifera hybrids (Figure 5d); isoamyl acetate and 1-hexanol for V. amurensis versus V. viniferas-V. amurensis hybrids (Figure 5e); and isoamyl acetate, 1-Hexanol, ethyl acetate, and 3-methyl-1-butanol for V. amurensis versus V. labrusca (Figure 5f). Studies have shown that isoamyl acetate is also a key differential marker between wines produced from V. amurensis and V. vinifera (Cabernet Sauvignon and Syrah) [12]. Notably, isoamyl acetate and 1-Hexanol emerged as common differential markers, with 1-hexanol consistently exhibiting the highest relative concentration and rOAV values in V. amurensis across all comparisons, highlighting its chemotaxonomic significance for this genotype. While their overall aromatic architecture resembles that of typical V. vinifera wines (e.g., Cabernet Sauvignon) [16], the composition of key aroma markers is distinctly different, thereby endowing wines produced from V. amurensis with a recognizable and independent flavor identity.

3.1.3. Analysis of the Aroma Differences Between Wines and Berries

Integrated analysis of volatile compounds during berry-to-wine fermentation reveals both conserved and group-specific metabolic patterns across four fermentation groups. A core set of compounds consistently demonstrated differential expression, including universally down-regulated compounds ((E)-2-hexenal, hexanal, pentanal, 2-methylpropanal, propanal, acetoin, hexyl acetate) and universally up-regulated fermentation-derived compounds (1-Propanol, ethanol, 2-Methyl-1-propanol, 3-Methyl-1-butanol, ethyl hexanoate, ethyl isobutyrate, isobutyl acetate, butyl hexanoate. Among them, ethyl isobutyrate and butyl hexanoate were newly synthesized during fermentation.). Several compounds displayed group-specific presence as differential metabolites. Ethyl 3-methylbutanoate, butyl acetate, and 2-butanone were exclusively up-regulated in B1 vs. A1, B3 vs. A3, and B4 vs. A4, respectively. Methyl 2-methylbutanoate was exclusively down-regulated in B3 vs. A3. 2-propanol, 2,3-pentanedione, propyl acetate, ethyl butanoate, ethyl 3-methylbutanoate, and ethyl 2-methylbutanoate were exclusively down-regulated in B4 vs. A4 (Figure 6).

The fermentation trials indicated that the overall metabolic pattern of aroma compounds in wines produced from V. amurensis aligns with previous studies [45]. However, we observed an intriguing metabolic trend: V. amurensis and its interspecific hybrids exhibited highly similar differential aroma profiles during fermentation. This similarity was primarily characterized by the up-regulation of esters with rOAV > 1 in the wines and the down-regulation of aldehydes with rOAV > 1 in the berries, underscoring fermentation as a key determinant of the wine’s overall aromatic profile. Notably, fermentation did not substantially alter the relative concentrations of acetaldehyde and 1-hexanol, yet these compounds emerged as primary contributors to the smoky and herbaceous notes in the wine [35], respectively. This observation suggests their origin in the grape varieties themselves. Their prominence as key aroma contributors was likely accentuated by the marked decrease in the relative concentration of hexanal—initially the compound with the highest rOAV—during fermentation. This dynamic highlights the presence of interactions among volatile compounds in shaping the final wine aroma profile. Conversely, compounds such as 1-propanol, 2-methyl-1-propanol, isobutyl acetate, and butyl hexanoate—despite being up-regulated during fermentation—did not contribute substantially to the wine’s aroma due to their high sensory thresholds. This reinforces that a compound’s role as a key aroma compound is jointly determined by its relative concentration, odor threshold, and interactive effects.

We also focused on the transformation of C6 aldehydes, alcohols, and esters. V. amurensis berries contained substantially higher levels of key C6 aldehydes ((E)-2-hexenal and hexanal, which impart herbaceous notes) than the other genotypes. Specifically, compared to its interspecific hybrids and V. labrusca, (E)-2-hexenal was 1.5-, 2-, and 8-fold higher, and hexanal was 1.4-, 1.7-, and 9-fold higher in V. amurensis, respectively, while 1-hexanol showed similar or moderately elevated levels. Following fermentation, the herbaceous-associated C6 aldehydes were nearly depleted. Concurrently, 1-hexanol increased variably across wines, with V. amurensis showing a 1.3-fold rise. More strikingly, fruity C6 esters, predominantly ethyl hexanoate, underwent marked accumulation—most notably in V. amurensis (20-fold) wines, which exhibited the highest concentrations. In the resulting wines, 1-hexanol remained highest in V. amurensis, exceeding levels in its hybrid and V. labrusca. The enhanced metabolism of C6 compounds in wines produced from V. amurensis may be linked to the indigenous microbiota [46,47].

Terpenes and phenols were not detected in the samples, likely due to their low concentrations in the matrix and the limited sensitivity of the analytical method, which may have prevented their detection or resulted in only minimal signals [48]. Future studies employing more sensitive detection techniques are warranted to fully characterize the aroma metabolome of V. amurensis wines. Our findings provide a theoretical foundation for investigating how indigenous fermentation microbiota shape the aroma formation in wines produced from V. amurensis.

3.2. Analysis of Differences in Organic Acids

The transformation from grape berries to wines significantly altered the organic acid composition across all 16 cultivars (Figure 7, Tables S8 and S9). Total organic acid content decreased markedly by 26% after fermentation (p < 0.05). Notably, wines produced from V. amurensis maintained significantly higher total acidity compared to other genetic backgrounds (p < 0.05). The fermentation process induced intense acid transformations. Following fermentation, V.a exhibited the greatest reduction in total organic acid content (28%), accompanied by a substantial alteration in organic acid composition (p < 0.05). Notably, lactic acid in wines produced from V. amurensis reached the highest concentration-—approximately 9243 times that in the berries—-while malic acid showed the most pronounced decline (by 31-fold). The inverse relationship between malic acid depletion and lactic acid accumulation provides clear evidence of spontaneous malolactic fermentation, likely mediated by endogenous microbial communities [14].

Genetic background significantly influenced acid profiles. Wines produced from V. amurensis exhibited distinct chemotaxonomic characteristics: lower tartaric acid (shared with V. amurensis-V. vinifera hybrids and V. labrusca) and malic acid content, but elevated lactic and succinic acid levels compared to other groups (p < 0.05). This specific acid configuration, reduced sharp-tasting acids coupled with enhanced flavor-enhancing acids, suggests wines produced from V. amurensis potentially develop a smoother, more complex sensory profile.

3.3. Analysis of Total Anthocyanins and Color

Anthocyanin profiling revealed substantial compositional changes during the berry-to-wine transformation in all cultivars. Wines from V. amurensis exhibited significantly higher pigment retention than those from other genetic backgrounds (Figure 8). Total anthocyanin content varied considerably both among wines from different species and within the same species. After fermentation, the anthocyanin retention rates were 63.3% for V. amurensis, 24.4% for V. amurensis-V. vinifera hybrids, 28.8% for V. vinifera-V. amurensis hybrids, and 8.5% for V. labrusca. The total anthocyanin content in V. amurensis berries was 1.7, 2.7, and 13.3 times higher than in its interspecific hybrids and V. labrusca berries, respectively. This initial advantage was further amplified in the corresponding wines, where the anthocyanin content of wines produced from V. amurensis exceeded that of its interspecific hybrids and V. labrusca by factors of 4.3, 6.0, and 98.7, respectively. With its high anthocyanin content and strong fermentation stability, V. amurensis was well suited for the production of high-quality, deeply colored wines. This enhanced stability is likely due to the characteristically high abundance of glycosylated anthocyanins in V. amurensis, which are known to be more resistant to degradation [49]. The influence of interspecific variation in anthocyanin profiles and their chemical stability among diverse V. amurensis cultivars on the chromatic and sensory properties of the wine remains a key knowledge gap, necessitating further systematic investigation.

Colorimetric analysis using CIELab parameters revealed distinct visual characteristics across the wine groups (Table S10, Figure 9b). Wines produced from V. amurensis exhibited the most favorable color profile, characterized by significantly lower L values, indicating darker appearance, and higher a*, b*, and Cab values (p < 0.05). The higher hue angle (h°ab ranged from 34.02° to 69.70°) and the highest total color difference (ΔEab ranged from 1886.43 to 3793.22) of wines produced from V. amurensis resulted in their most vibrant and intense red-purple coloration (Table S10, Figure 9a). Wines produced from V. amurensis have exceptional color intensity and stability that significantly exceed those of other varieties. The preserved anthocyanin content in wines produced from V. amurensis demonstrated their potential for producing wines with intense coloration.

A practical limitation of the CIELab method is that the intense color of the wine necessitates dilution for accurate measurement, as does the RGB method. Thus, while these values do not represent the wine’s absolute color, they offer an objective and standardized basis for comparison.

Correlation analysis elucidated the fundamental relationships between chemical composition and color characteristics (Figure 10). Total anthocyanin content showed strong positive correlations with desirable color parameters including a* and Cab, while exhibiting a significant negative correlation with L. Anthocyanins provide the fundamental chemical basis for the enhanced color properties. Among organic acids, tartaric acid mirrored this correlation pattern, suggesting its supportive role in color, potentially through co-pigmentation mechanisms [50]. Conversely, citric acid demonstrated inverse relationships with favorable color parameters. Total anthocyanin content showed a highly significant positive correlation with tartaric acid, lactic acid, and total organic acid content, indicating that the high anthocyanin retention and high color intensity in wines produced from V. amurensis were associated with its high organic acid levels, particularly influenced by tartaric acid, lactic acid, and citric acid. These findings align with established research on the pH-dependent behavior of anthocyanin stability [51,52], confirming that both anthocyanin concentration and the specific acid composition collectively govern the final wine color.

In conclusion, this study establishes that the exceptional color characteristics of wines produced from V. amurensis—characterized by deep red hues with high saturation and brightness—result directly from their high anthocyanin content interacting favorably with specific organic acids, particularly tartaric acid.

4. Conclusions

This study demonstrates that V. amurensis possesses comprehensive advantages in winemaking characteristics compared to its interspecific hybrids and V. labrusca, particularly in terms of aroma composition, organic acid metabolism, and color stability. Regarding aroma, V. amurensis exhibits prominent metabolism of C6 compounds during fermentation: the berries are rich in C6 aldehydes that impart herbaceous notes, which are subsequently transformed into significantly higher levels of C6 alcohols (1-Hexanol) and C6 esters (ethyl hexanoate) post- fermentation. Notably, 1-Hexanol serves as a key aroma marker distinguishing wines produced from V. amurensis. Furthermore, this research reveals that the contribution of a compound to key aromas is determined not solely by its concentration, but by the combined effects of its relative concentration, odor threshold, and interactions with other volatile constituents. In terms of organic acid, wines produced from V. amurensis showed the most significant reduction in total organic acid content (28%) after fermentation, accompanied by a distinct metabolic transformation—achieving the highest concentration of lactic acid and the most pronounced decrease in malic acid. This shift has important implications for both taste profile and microbial stability. For color attributes, wines produced from V. amurensis displayed the highest anthocyanin retention rate (61.2%) and an exceptional red-purple hue. Color intensity strongly correlated with total anthocyanin content, with V. amurensis consistently outperforming other genotypes in this metric. In conclusion, the metabolic properties exhibited by V. amurensis during fermentation—particularly its advantages in C6 aroma compound conversion, organic acid restructuring, and anthocyanin preservation—collectively contribute to the superior quality of its wines in aroma, acidity, and color. These findings provide a scientific basis for the enological application of V. amurensis.