3.1 Preliminary screening of yeast strains
Post-morphological examination (both micro and macro), we identified 16 isolates that were growing on YEPD agar as yeast strains. The cluster analysis of the rep-PCR of these 16, based on a coefficient of similarity of 76%, resulted in eight clusters (Fig. 3). We identified the isolates in each cluster by sequencing the D1/D2 region of the 26S rRNA gene, followed by a BLAST search at GenBank (Fig. 3). Approximately 31.25% of the yeast isolates were found in cluster Ⅶ, which were identified as Wickerhamomyces anomalus (HN007, HN010), Wickerhamomyces mori (HN012, HN015), and Dabaryomyces hansenii (HN014). Approximately 18.75% of the yeast isolates were found in cluster V and shared 100% homology with GenBank sequences of Saccharomyces cerevisiae (HN006, HN009) and Saccharomyces kudriavzevii (HN016). The cluster Ⅰ and Ⅷ isolates were unambiguously identified as Hanseniaspora vineae (HN001, HN005) and Pichia kluyveri (HN002, HN004), respectively. The remaining four clusters (Ⅱ, Ⅲ, Ⅳ, and Ⅵ) were identified as Trichosporon asahii (HN003), Kluyveromyces lactis (HN008), Galactomyces geotrichum (HN013), and Yarrowia lipolytica (HN011), respectively, each showing 100% homology to GenBank sequences.
We performed the olfactory “sniff” test on the YEPD medium plates to preliminarily determine the aroma producing ability of the obtained 16 yeast strains. Table 1 presents the results of this analysis. We found that different genera of the isolates presented various aromas, differentiated by different intensities of sourness, fruity, floral, cheese-like, and cream-like smells. Saccharomyces cerevisiae (HN006, HN009) and Saccharomyces kudriavzevii (HN016) produced different intensities of alcoholic aroma in the YEPD agar plates. Both Dabaryomyces hansenii (HN014), and Yarrowia lipolytica (HN011) produced characteristic cheese-like smells with intermediate strength. Each of the following strains: Trichosporon asahii (HN003), Wickerhamomyces anomalus (HN007, HN010), Wickerhamomyces mori (HN012, HN015), Pichia kluyveri (HN002, HN004), and Kluyveromyces lactis (HN008) produced pleasant fragrance of pineapple fruit, apple and rose fruit, apple fruit, banana fruit, and strawberry fruit, respectively, and among which strain Wickerhamomyces anomalus (HN010) possessed the highest intensity. While the strain Hanseniaspora vineae (HN001, HN005) generated an unpleasant sour flavor, the strain Galactomyces geotrichum (HN013) produced a cream-like aroma, which was reported to have wide application in the fermentation of dairy products (Jacques et al., 2017).
Based on these results, nine strains (HN002, HN003, HN006, HN008, HN009, HN010, HN011, HN013, and HN014), which produced a pleasant aroma with moderate and strong intensity were selected for the follow-up experimental studies.
These nine strains were further examined for Chinese chestnuts wine fermentation, and each aroma profile of the obtained wine was measured by the SPME-GC-MS technique to further evaluate their performance in the accumulation of aroma compounds for the production of the Chinese chestnuts wine. Table 2 presents the list of all identified volatile compounds, which were mainly characterized by the following functional groups: 10 volatile acids (6.744–14.721 µg/mL), 14 esters (3.502–16.456 µg/mL), 11 alcohols (1.890–8.689 µg/mL), 6 aldehydes (0.309–3.974 µg/mL), 3 ketones (0.108–0.604 µg/mL), 3 Alkanes (0.373–1.188 µg/mL), and, at a lower content, by 2 volatile phenols (0–0.409 µg/mL).
Apart from presenting the overall aromatic profile of each tested strain, Table 2 also highlighted the differential behavior exhibited by strains belonging to the same species. We observed a variation in the total concentration of each type of volatile compounds based on different strains. The samples produced by non-Saccharomyces yeasts (HN002, HN003, HN008, HN010, HN011, HN013, and HN014) possessed higher amounts of volatile acids compared with those of Saccharomyces yeasts (HN006, HN009), especially the wine fermentation using the strain HN010 showed the highest concentration (15.721 µg/mL) of various acids, with the exception of butanoic acid, hexanoic acid, 2-methyl-butanoic acid, and nonanoic acid. Butanoic acid and hexanoic acid are generally present in the Chinese strong flavor liquor, and their high content can impart an unpleasant flavor to the wine (Aslankoohi et al., 2016). The presence of 2-methyl-butanoic acid in the strains HN011 and HN014 imparts cheese-like and sweet characteristics (Marycarmen et al., 2020), which were consistent with the results of our preliminary olfactory “sniff” test. Nonanoic acid, detected only in strain HN009, had green and fat characteristics (Aslankoohi et al., 2016). No significant difference was observed in the type and concentration of volatile acids between HN006 and HN009.
We observed a significant difference in the concentrations of esters in the wines produced using different cultures (Table 2). We found a significantly higher concentration of esters in the non-Saccharomyces culture wines compared with the Saccharomyces culture wines, which was probably attributed to its low lipase activity (Table 2). The highest concentration of esters (16.456 µg/mL) was found in the wine prepared using HN010, which was 3.22 and 3.70 times higher than that in the Saccharomyces yeast strain of HN006, HN009, respectively. Amongst all types of esters, the concentration of ethyl acetate was found to be the highest for each strain, ranging from 2.187 µg/mL (HN006) to 16.456 µg/mL (HN010). Ethyl esters, typically described to possess a “fruity and flower” aroma (Laura et al., 2019), were the largest ester-based compounds in the Chinese chestnut wine. We found extremely high amounts of ethyl acetate (sweet, fruity), phenylethyl acetate (rosy), and ethyl caproate (apple peel, fruity) in the Chinese chestnut wine produced by HN010 (Marycarmen et al., 2020), along with high-molecular-weight ethyl esters, such as ethyl laurate, ethyl palmitate, ethyl oleate, and ethyl linoleate, which were reported to promote the accumulation of the after-taste in the Chinese strong flavor liquor (Yan et al., 2019).
Higher alcohols, produced by yeast, originate either from the degradation of branched-chain amino acids or directly from sugar fermentation. We found that the higher alcohols constituted the major quantitative component in Saccharomyces (HN006, HN009)-fermented chestnut wine samples, and were significantly higher than those obtained by non-Saccharomyces-fermented wines (Table 2). The wine samples brewed using the strain HN006 exhibited the highest concentration of total higher alcohols, which included isoamyl alcohol (3.987 µg/mL), 1-nonanol (0.254 µg/mL), 1-hexanol (0.323 µg/mL), 2,3-butanediol (0.914 µg/mL), isooctanol (0.165 µg/mL), 2-methyl-1-propanol (1.851 µg/mL), enanthol (0.294 µg/mL), 3-ethoxy-1-propanol (0.754 µg/mL), 1-octen-3-ol (0.653 µg/mL), octanol (0.572 µg/mL), and phenylethyl alcohol (0.354 µg/mL). Of these, isoamyl alcohol, 2,3-butanediol, and phenylethyl alcohol have previously been reported to be present in fermented alcoholic beverages as characteristic flavor compounds, such as Chinese strong flavor liquor (Yan et al., 2019).
Aldehydes are produced from amino acids either by Strecker degradation or by transamination, followed by decarboxylation, and they can be easily reduced to alcohols. We observed that the Saccharomyces-fermented wines possessed a significantly higher concentration of aldehydes compared with the non-Saccharomyces-fermented wines. This trend was consistent with the trends of higher alcohols (Table 2).
Additionally, we detected three aldehydes in the Chinese chestnut wine samples, and they were found to be most abundant in the strain HN006. Among these, 2-octanone, which imparts a cream-like flavor to the wine body (Yan and Dong, 2019), was found in the Chinese chestnut wine samples. Also, 2-nonanone, which has a fruity and flowery note (Yan and Dong, 2019), was found in each chestnut wine sample fermented using strains HN005, HN006, HN007, and HN008.
We found the highest levels of alkanes (1.188 µg/mL) in the samples fermented by strain HN006, and they constituted tetramethylethylene (0.387 µg/mL), 2-methyl-2-butene (0.687 µg/mL), and phenylethylene (0.114 µg/mL). Additionally, two volatile phenols, namely 4-vinylphenol, and 4-vinylguaiacol, were also identified to be present in the highest amount in the samples with strain HN006, and are known to possess a characteristic spice-like aroma (Yan et al., 2019).
We created a heatmap representation to allow a rapid visual assessment of the similarities and differences in the volatile profile of the strains between different samples. Additionally, the acids, alcohols, and esters represented the major functional groups for all the yeast strains (Fig. 4). Figure 4 shows the differences in the relative abundance of compounds among different strains, revealing the presence of different flavor profiles in different strains. For instance, the Saccharomyces strains (HN006, HN009), produced high levels of ethyl esters, isoamyl alcohol, and 2-methyl-1-propanol, along with the non-Saccharomyces strains (HN004 and HN010). Notably, the strain HN010 showed relatively larger amounts of volatile compounds compared with other strains, which indicated that it might produce wine with a more harmonious wine aroma. These results also suggested that different yeast strains, especially non-Saccharomyces yeasts, imparted a specific influence on wine flavor. However, their fermentation characteristics are not yet fully understood, and further research is required to identify the wine fermentation capacity of each strain.
3.1.1 Testing for fermentation ability
We tested the fermentation performance of the obtained nine yeast strains to identify the optimal yeast strain for preparing the Chinese chestnut wine. Figure 5(A-B) show the effects of yeast strains on the concentrations of ethanol and reducing sugar during the fermentation of Chinese chestnut wine. We observed a slight increase in the concentration of ethanol for Saccharomyces strains (HN006, HN009) during the initial 24 h of fermentation, which increased significantly and reached the maximum value after 120 h, whereas the ethanol content produced by non-Saccharomyces strains (HN002, HN003, HN008, HN010, HN011, HN013, and HN014) got slowly accumulated throughout the fermentation, and their corresponding concentrations were lower than those produced by the Saccharomyces strains. Table 3 also summarizes the kinetic parameters of the ethanol fermentation with different strains. The strain HN006 generated the maximum concentration of ethanol (90.220 ± 0.874 g/L), had the highest ethanol productivity (0.752 ± 0.026 g/L/h), and sugar utilization (99.5%) after 120 h (Table 3). Also, this strain achieved the highest final ethanol yield of 0.467 ± 0.043 g/g. Thus, we observed that the strain HN006 showed the high fermentation performance of Chinese chestnut wine.
3.1.2 Sensory evaluation
A sensory evaluation was performed to identify the yeast strain that had the potential to produce a pleasant tasting wine. Sensory characteristics are vital markers to evaluate the wine quality, amongst which sensory appraisal is a commonly accepted method. A significant difference was observed in different quality attributes, such as aroma, after taste, wine taste, mouthfeel, and overall quality (Fig. 6). Among these wine samples, the wine prepared using strain HN010 was superior to those with other strains based on its highest scores for taste (7.37), aroma (7.76), mouth feel (7.64), after taste (6.90), and overall acceptability (7.12), which agreed with the result of aromatic profile of the volatile compounds determined by SPME-GC-MS. The Saccharomyces strains (HN006, HN009) received the lowest scores in the sensory analysis. Thus, this test identified and verified that the taste-profile of the Chinese chestnut wine made using the yeast strain HN010 was appreciated by the testers and had an affirmative effect on the overall sensory features.
Thus, the non-Saccharomyces strain, HN010, showed the highest amounts of volatile compounds and best sensory characteristics but with poor ethanol fermentation capacity, and the Saccharomyces strain, HN006, showed the highest efficiency in ethanol production and consumption of reducing sugar. Therefore, both these yeast strains were selected to be used as mixed inoculum for producing an improved version of the Chinese chestnut wine.