Influence of Isolation Source on the Probiotic Properties and Health Benefits of Yeasts: Insights from Metabarcoding and Cultivation Approaches

Abstract

The study aimed to identify potential sources of novel probiotic yeasts exhibiting health-promoting properties. A combination of metabarcoding analysis and cultural methods was employed to investigate and isolate yeasts from various sources, including rice wine, palm wine, fermented shrimp paste at different stages of natural fermentation, and lychee peels. The two analytical methods revealed distinct yeast profiles, and each source exhibited a unique composition of yeast species. Through metabarcoding and cultural methods, it was demonstrated that lychee peels harbored a greater diversity of genera compared to other sources. The evaluation of the probiotic properties of yeasts revealed that lychee peel yielded the highest proportion of isolates with potential probiotic activity (53.33%), followed by palm wine (25%), fermented shrimp paste (10%), and rice wine (9.09%). Moreover, yeast isolates with health-promoting properties as evaluated in this study, including Starmerella meliponinorum L12 and Pichia terricola L9 from lychee peels, demonstrated notable antioxidant activity and cholesterol-reducing properties, respectively. These findings represent the first report providing initial insights into the influence of yeast sources and serve as a guideline for the targeted selection of yeasts with specific probiotic and health-promoting attributes.

1. Introduction

Probiotics have become increasingly popular as more information comes out about health promotion. The World Health Organization defines probiotics as “live microorganisms which when administered in adequate amounts confer a health benefit on the host” [1]. The properties of probiotics must exhibit in vitro characteristics to survive during the passage through the human gastrointestinal tract, tolerance to low pH and bile salt, capacity to control pathogenic microorganisms, and efficient adherence to intestinal epithelial cells [2]. Yeast holds several advantages in comparison to bacteria, as they are normally resistant to antibiotics, which are necessary for their resistance during antibiotic treatment, and there is no transfer of genes related to antibiotic resistance [3]. However, a key property of probiotic yeasts is that they should remain susceptible to certain antifungal drugs [1]. In 1920, Henri Boulard isolated a yeast species from the peel of tropical fruits (lychee and mangosteen) named Saccharomyces cerevisiae var. boulardii [4] and, subsequently, studied its probiotic properties. There are relatively few yeast strains currently available on the market as probiotics, with the primary ones being Saccharomyces cerevisiae var. boulardii, Saccharomyces cerevisiae, and Kluyveromyces fragilis (B0399) [5,6]. These reasons led to an interest in the study of other yeast species with probiotic properties. Fermented foods, consumed since 8000 BC and representing nearly one-third of global diets [7], are a rich source of diverse, generally recognized as safe (GRAS) microorganisms [8], including resilient yeasts from high-stress environments that exhibit probiotic potential by withstanding gastric and intestinal conditions, making them ideal for functional food development [9,10]. Probiotics offer health benefits such as balancing gut microbiota, reducing inflammatory bowel disease, boosting immunity, and lowering the risk of cancer, obesity, and NCDs [11]. They may also help prevent coronary heart disease by reducing cholesterol and controlling blood pressure [12]. Some probiotics have antioxidant activities, including ROS scavenging, metal ion chelation, and enzyme inhibition [13]. Yeasts are gaining interest as probiotics due to their larger cell size and the presence of cell wall components such as mannose and glucan. Importantly, probiotic yeasts exhibit natural resistance to antibiotic treatments, allowing them to maintain their beneficial effects even during antibiotic therapy. This inherent antibiotic resistance makes them a particularly promising alternative to bacterial probiotics, whose viability and efficacy are often compromised by concurrent antibiotic use. Furthermore, yeasts produce beneficial metabolites such as fatty acids, amino acids, and B vitamins, adding to their functional potential [14]. Several reports have highlighted the ability of specific yeast strains to produce high levels of conjugated linoleic acid (CLA), particularly the CLA 10-12 isomer, and propionic acid. These compounds are associated with numerous health benefits, including anticancer and anti-inflammatory properties, antiplatelet aggregation, hypocholesterolemic effects, and the prevention of fatty liver [15,16]. Furthermore, certain yeast strains exhibit anti-inflammatory properties, as evidenced by their ability to significantly reduce the production of the pro-inflammatory cytokine IL-1β [16]. Yeast cell wall components, such as β-glucan, or the yeast biomass itself, also play a vital role in preventing Candida virulence and stimulating the immune system. Similarly, yeasts such as Saccharomyces cerevisiae and various non-Saccharomyces species produce alcoholic signaling molecules, including phenylethanol and tryptophol. Numerous animal studies, along with a limited number of clinical trials, have demonstrated the role of probiotics and their metabolites, such as short-chain fatty acids (SCFAs), tryptophol, and phenylethanol, in promoting human health and mitigating disease [17]. SCFAs, including acetic, propionic, and butyric acids, are particularly notable for their immunomodulatory, antimicrobial, antioxidant, and cholesterol-lowering properties [7]. Overall, probiotic yeasts represent a promising alternative, with antioxidant and cholesterol-lowering properties being crucial for reducing the risk of NCD-related diseases. These traits could serve as key criteria for selecting yeast probiotics, offering significant market potential. Further studies are needed to support and expand their applications. Metabarcoding analysis offers a comprehensive approach to profiling microbial communities in complex environmental samples. It is a molecular technique that combines DNA barcoding with high-throughput sequencing to identify and characterize multiple taxa [18]. This method involves the amplification of specific marker genes (e.g., 16S rRNA for bacteria, ITS for fungi) using universal primers, allowing the detection of both culturable and non-culturable organisms. This comprehensive insight can guide the identification of potential sources for target microbes, such as probiotics, by highlighting promising candidates and their ecological roles. Such information enables the strategic design of isolation approaches to focus on microbial strains with desired functional properties.

This study aimed to evaluate potential sources using metabarcoding analysis and culture-dependent techniques for the investigation and isolation of yeasts having probiotic properties for further selection as probiotic candidates with potential antioxidant and cholesterol-lowering properties. The insights obtained can be used as a guideline for a more targeted selection of yeast sources with the specific desired traits.

2. Materials and Methods

2.1. Sample Collection and Preparation

The samples (rice wine, palm wine, and fermented shrimp paste during the early to final stages of natural fermentations) were collected from the traditional fermented food manufactured in Thailand. The palm wine sample after 12 h was collected directly from a hanging pot and kept for natural fermentation in the fermentation pot for 24 h. Lychee was obtained from an organic orchard located in the northern province of Thailand. All collected samples were immediately placed in sterile self-sealing bags and transported back to the laboratory within 24 h in a low-temperature sampling box. Each sample was divided into two portions: the first was stored at −20 °C for DNA extraction and analysis of the fungal community structure via metabarcoding, while the second was kept at 4 °C for yeast isolation. The commercial yeast, Saccharomyces cerevisiae var. boulardii CNCM I-745 (SB) (Bioflor, BIOCODEX, Gentilly, France) was used as the reference strain.

Ten grams of each sample (rice wine, palm wine, and fermented shrimp paste) were mixed with 90 mL of sterilized peptone water (0.1%, w/v), and the mixture was homogenized in a Stomacher for 2 min at the maximum speed. For the lychee peel sample, the goal was to isolate surface-associated yeasts. Lychee (25 g) was rinsed in 225 mL of 0.1% peptone water with 0.01% Tween 80 by orbital shaking in a flask at 150 rpm for 30 min to facilitate the detachment of surface microorganisms [19]. Homogenized samples were used for metabarcoding DNA extraction and yeast isolation.

2.2. Metabarcoding Analysis

The extraction of microbial genomic DNA from the samples was performed using the bead-beating method using the PowerSoil Pro Kit (Qiagen, Berlin, Germany) according to the manufacturer’s instructions. The DNA concentrations and the quality assessment of the extracted DNA templates were obtained from the NanoDrop Spectrophotometer (Thermo Scientific, Waltham, MA, USA). The ITS gene was amplified using ITS1F (5′-CTTGGTCATTTAGAGGAAGTAA-3′) and ITS2R (5′-GCTGCGTTCTTCATCGATGC-3′) primers, targeting ITS1-2 regions, and 2X sparQ HiFi PCR master Mix (QuantaBio, Beverly, MA, USA). The PCR reaction was run with an initial step at 98 °C for 2 min and 30 cycles of denaturation at 98 °C for 20 s, annealing at 60 °C for 30 s, and extension at 72 °C for 30 s, followed by a single-step final extension step at 72 °C for 1 min. Cluster generation and 250 bp paired-end read sequencing were performed on an Illumina MiSeq at Omics Sciences and Bioinformatics Center (Chulalongkorn University, Bangkok, Thailand). Bioinformatic analysis was performed using the QIIME 2 data analysis package (version 2023.9). Raw sequencing data were demultiplexed and quality-filtered using the q2-demux plugin, followed by denoising with DADA2. Amplicon Sequence Variants (ASVs) with total read counts below 2 were removed to eliminate low-frequency noise. The representative ASVs were aligned using the BLAST algorithm for taxonomic classification [20,21].

2.3. Isolation and Characterization of Yeasts

Yeast was isolated from each sample using the spread plate technique. The mixed samples were serially diluted in 0.1% peptone water, from which 0.1 mL of an aqueous dilution was spread on the yeast malt agar (YMA) medium plates. The plates were incubated at 30 °C for 48–96 h. The colonies from extensive growth plates were picked following Harrison’s disc method [22]. Then, 20% of the total yeast isolates of each such morphologically classified group were selected as representative of that group and purified by restreaking onto YMA and then maintained and subcultured until identified. All isolates of purified yeast cultures were preserved in YM broth supplemented with 20% (v/v) glycerol and stored at −80 °C.

2.4. Identification and Phylogenetic Analysis of the Potential Probiotic Yeasts

The yeast strains that demonstrated potential probiotic characteristics were selected for molecular identification. DNA was extracted from pellets of yeast cells by boiling according to procedures described by Silva et al. [23]. The D1/D2 domain of the 26S rDNA was amplified by PCR with the forward primer NL1 (5′-GCATATCAATAAGCGGAG GAAAAG-3′) and reverse primer LS2 (5′-ATTCCCAAACAACTCGACTC-3′). PCR reactions were performed with 2X PCR SuperMix (BIO-HELIX, Taipei, Taiwan). Amplification was performed in the reaction mixture using 2X PCR SuperMix (BIO-HELIX, Taipei, Taiwan). The PCR reaction was run with an initial step at 95 °C for 5 min and 30 cycles of denaturation at 95 °C for 1 min, annealing at 52 °C for 2 min, and extension at 72 °C for 2 min, with a final extension at 72 °C for 7 min. The PCR products were sent to a commercial sequencing facility (Macrogen, Seoul, Republic of Korea) for direct sequencing. DNA base sequences were analyzed by homology and sequence similarity comparison with those in the GenBank database (National Center for Biotechnology Information, NCBI) using the Blast program with default settings. Molecular Evolutionary Genetic Analysis (MEGA version 11) software was used for phylogenetic analyses. The closest homologous to the sequences were selected, and multiple sequence alignments were carried out using the Clustal W program in the MEGA 11 software [24].

2.5. Evaluation of Probiotic Properties of the Yeast Isolates

2.5.1. Acid and Bile Salt Tolerance

Yeast isolate fresh cultures were prepared by inoculating a single colony of isolate in 10 mL of YM broth and incubating at 30 °C for 24 h to obtain vegetative cells (7–8 log CFU/mL). A total of 1.0 mL of fresh culture was inoculated with 9.0 mL of artificial gastric juice containing 0.3% (w/v) pepsin in phosphate buffer (pH 2.0, adjusted with 0.1 M HCl), following 3 h at 37 °C. On the other hand, tolerance to bile salt was determined by inoculating 1.0 mL of fresh culture in 9.0 mL of phosphate buffer containing 0.3% (w/v) ox bile (pH 7.0, adjusted with 0.1 M NaOH) and incubating for 4 h at 37 °C. Then viable cells were enumerated by the plate count technique on YMA and incubated for 48 h at 30 °C. Survival rate was expressed by comparing the population at zero time and after incubation in artificial gastric juice or bile salt fluid. The percentage of survival was calculated according to Equation (1) [25]:

Survival (%) = (CFU viable cells survived/CFU initial cells inoculated) × 100

(1)

2.5.2. Cell Surface Hydrophobicity

The hydrophobicity properties of the yeast cells were determined according to Amorim et al. [26]. The yeast cells were cultured in YM broth at 30 °C for 24 h and centrifuged at 8000× g for 10 min. The cell pellet was washed twice with sterile PBS and resuspended in the same buffer to achieve an OD 600 nm of approximately 0.4–0.6 (OD initial). The adjusted yeast cell suspension (4.0 mL) was mixed thoroughly with xylene (1.0 mL) using a vortex mixer for 30 s. The mixture was then kept at room temperature for 2 h to separate the phases. The aqueous phase was measured at OD 600 nm. The absorbance of the aqueous phase was measured at 600 nm. The percentage of hydrophobicity was calculated according to Equation (2), where OD_initial and OD_final are the absorbances (at 600 nm) before and after extraction with solvent.

Hydrophobicity (%) = [1 − ((OD_initial − OD_final)/OD_initial)] × 100

(2)

2.6. Evaluation of Health Benefit Properties of the Potential Probiotic Yeasts

2.6.1. Test of Antioxidant Activity

Yeast isolates were tested for their ability to scavenge the DPPH (1,1-Dipheny-l-2-Picrylhydrazyl) radical following the method described by Tang et al. [27]. Yeast cultures (1 mL) in YM broth (7–8 log CFU/mL) were centrifuged at 10,000 rpm for 5 min, followed by two washes with a sterile 0.85% NaCl solution. The resulting pellet was resuspended in 1.0 mL of the same solution. Next, 1.0 mL of the cell suspension was transferred into a fresh tube, and then 1.0 mL of a DPPH radical solution (0.2 mM in ethanol) was added. The reaction solution was thoroughly mixed and then incubated at room temperature in the dark for 30 min. The control group contained an equal volume of deionized water instead of the sample. The blank group included an equal volume of ethanol instead of the DPPH radical solution. The absorbance of the solution was measured at 517 nm after centrifugation at 6000× g for 10 min. The percentage of scavenging activity was calculated according to Equation (3):

Scavenging activity (%) = [1 − (A_sample − A_blank)/A_control] × 100

(3)

2.6.2. Test of Cholesterol Reduction

Cholesterol reduction was evaluated following the method described by Ngongang et al. [28,29]. Yeast isolate fresh cultures were prepared by culturing in YM broth and incubated at 30 °C for 24 h to obtain vegetative cells (7–8 log CFU/mL). Half of the tubes were supplemented with 0.4% bile salt, followed by the addition of 1% cholesterol in all the tubes. They were inoculated with 1% into the cultures, while, under the same conditions, four other tubes were prepared free of isolates and used as controls. They were incubated at 37 °C for 24 h. After incubation, cultures were centrifuged, and unutilized cholesterol was estimated in the supernatant by measuring the optical density at 540 nm and compared to the control. The percentage of cholesterol reduction was calculated according to Equation (4), where OD_initial and OD_final are the absorbances (at 540 nm) before and after incubation for 24 h.

Cholesterol reduction (%) = (OD_initial − OD_final)/OD_initial × 100

(4)

2.7. Statistical Analysis

For the analysis of alpha and beta diversities of microbial communities, different indices of alpha diversity (including Shannon, Observed, and Faith’s phylogenetic diversity (PD) (Faith’s PD) index) were analyzed using the Kruskal–Wallis test. The beta diversity measures (Bray–Curtis and Jaccard index) were analyzed using Permutational Multivariate Analysis of Variance (PERMANOVA) by the q2-diversity alpha plugin in QIIME2. Experimental data were presented as mean ± standard deviation (SD) from experiments conducted in triplicate. Analysis of variance (ANOVA) and multiple comparisons by Duncan’s test were assessed using IBM-SPSS statistics package version 22 (SPSS Inc., Chicago, IL, USA). A statistically significant difference was considered at a significance level of p < 0.05.

3. Results

3.1. Yeast Composition in Each Source of Isolate Through Metabarcoding and Cultural Analyses

To explore the influence of environmental conditions on yeast communities, a metabarcoding approach targeting ITS genes was employed, offering comprehensive insights into naturally occurring yeast populations. Sequencing reads were classified at the family level using QIIME 2 (version 2023.9), and representative Amplicon Sequence Variants (ASVs) were analyzed taxonomically via Blast alignment. The alpha diversity was evaluated using the Shannon, Observed, and Faith’s PD index, which reflects the richness and evenness of microbial communities within each group. The statistical significance was assessed using the Kruskal–Wallis test. Significant differences in alpha diversity were observed at p < 0.05, with the lychee peel group exhibiting the highest diversity. The beta diversity was performed by Bray–Curtis and Jaccard indexes to assess compositional differences between groups using PERMANOVA at p < 0.05. The results indicated significant clustering of microbial communities across sample groups, as shown in Table S1. Yeast communities were expressed as relative abundances, representing the percentage of a yeast genus relative to total organisms in the environment. Results summarized in Figure 1 revealed distinct yeast profiles for each source type: fermented foods exhibited lower diversity compared to lychee peels, which showed a more complex profile. Dominant fungal phyla included Ascomycota in palm wine (99.89%) and fermented shrimp paste (88.29%), Zygomycota in rice wine (89.1%), and Basidiomycota in lychee peels (74.28%).

Figure 1a highlights that lychee peels and fermented shrimp paste exhibited the highest diversity of genera. In lychee peels, Inocybe (50.41%) was the most abundant genus, followed by Symmetrospora (20.59%), Starmerella (16.61%), Wickerhamiella (7.79%), Macalpinomyces (2.26%), Candida (1.02%), Humaria (0.94%), and others (0.38%). Fermented shrimp paste was dominated by Hanseniaspora (86.43%), with lesser contributions from Lachancea (11.25%), Candida (1.51%), Rhizopus (0.46%), Discosia (0.23%), and Tortispora (0.16%). Rice wine primarily contained the filamentous fungus Rhizopus (89.10%), along with Candida (6.72%), Lachancea (2.90%), and Aspergillus (1.28%), while palm wine was dominated by Starmerella (90.47%), followed by Wickerhamiella (9.42%) and Schwanniomyces (0.11%). The cultural plating results, illustrated in Figure 1b and Table S2, revealed distinct morphological colony types, with 44 yeast strains isolated across sources. Lychee peel yielded the most isolates with diverse morphologies, aligning with the metabarcoding findings.

Both metabarcoding analysis and the culturing plate method were employed to study the community yeast sources, with results indicating that lychee peels harbored a greater variety of genera compared to other sources, as shown in Figure 1a,b. However, inconsistencies were observed between the microbial profiles obtained by the two methods, as illustrated in Figure 1c. For rice wine, the culturing plate method identified Saccharomyces cerevisiae as the most abundant yeast species, whereas metabarcoding analysis indicated a higher abundance of Rhizopus, followed by Candida and Lachancea. In palm wine, Hanseniaspora sp. were most abundant based on the culturing plate method, while metabarcoding analysis highlighted Starmerella and Wickerhamiella as dominant. Fermented shrimp paste, analyzed through the culturing plate method, showed the prevalence of Schwanniomyces etchellsii and an unidentified species, whereas metabarcoding analysis revealed higher abundances of Hanseniaspora and Lachancea. For lychee peels, the culturing plate method detected Pichia terricola, Kurtzmaniella quercitrusa, Starmerella meliponinorum, and Meyerozyma guilliermondii, while metabarcoding analysis indicated dominance by Inocybe, Symmetrospora, and Starmerella. These differences demonstrate how the two approaches complement each other in identifying the diversity of yeast communities.

3.2. Identification and Phylogenetic Analysis

The yeast isolates were identified as species using 26S rRNA gene sequences and compared with related yeast species in the NCBI database. The isolates belonged to Saccharomyces, Hanseniaspora, Schwanniomyces, Zygoascus, Kurtzmaniella, Pichia, Meyerozyma, and Starmerella genera, as shown in Table S3. Although 26S rDNA sequencing provides reliable genus-level identification, integrating ITS region sequencing would further enhance species-level resolution and accuracy [30]. In future studies, additional DNA regions, as well as whole-genome sequencing, could be employed to achieve more comprehensive taxonomic and functional characterization of selected isolates. Probiotic candidates require to be properly identified at the genus, species, and strain levels. Selected strain candidates should be characterized in accordance with several safety traits before a possible application. Figure 2 exhibited a close relationship with isolates using the neighbor-joining algorithm. The phylogenetic tree of potential probiotic yeasts indicates that the yeast candidate strains showed a different clade from Candida alblicans, which is an opportunistic pathogenic yeast. Therefore, they should be suitable for further study to explore their general probiotic properties and potential health benefits.

3.3. Probiotic Properties

3.3.1. Acid and Bile Salt Tolerance

Tolerance to gastric acidity and bile salt resistance are considered key property requirements for probiotics, which enable them to survive during passage through the gastrointestinal tract. All 44 yeast isolates obtained from the four sources were evaluated and compared to SB, which was used as the control. The yeast isolates with tolerance to gastrointestinal fluids were mainly the strains isolated from lychee peels, followed by rice wine and palm wine, as shown in Figure 3a. The number of isolates that showed acid tolerance with more than 50% survival, similar to SB, included 11 isolates from lychee peels and 6 isolates each from palm wine and rice wine. In contrast, there were only two isolates from fermented shrimp paste. In addition, the yeast isolates that showed bile salt tolerance with more than 50% survival, like SB, were mainly isolated from lychee peels, fermented shrimp paste, and palm wine, with the number of isolates being 13, 7, and 6, respectively. There were only two isolates from rice wine that showed bile salt tolerance.

3.3.2. Cell Surface Hydrophobicity

Cell surface hydrophobicity was assessed using the microbial adhesion to hydrocarbons (MATH) method, with xylene serving as the non-polar solvent. Yeast affinity for xylene reflects surface hydrophobicity, which is relevant to adhesion to intestinal epithelial cells that possess apolar structures such as lipid bilayers [31]. All 23 yeast isolates that passed the acid and bile salt tolerance tests were evaluated. Of these, only 12 exhibited hydrophobicity values greater than 50%, similar to or superior to those of SB, as shown in Figure 3b. The yeast isolates with the highest hydrophobicity values were primarily obtained from lychee peels, while those from other sources showed significantly lower values.

3.4. Potential Health Benefit Properties

3.4.1. Antioxidant Activity

The antioxidant capacity, evaluated by measuring the scavenging effects on DPPH radicals, is shown in Figure 4a. All the strains tested exhibited antioxidant activity at different levels. The highest scavenging activity was exhibited by Starmerella meliponinorum L12 (65.10%), Saccharomyces cerevisiae R2 (58.25%), and Hanseniaspora opuntiae P7 (51.78%), which were isolated from lychee peels, rice wine, and palm wine, respectively, with SB, a commercial strain, showing 54.72%. The other strains exhibited scavenging activity lower than 35%.

3.4.2. Cholesterol-Lowering Capacity

High serum cholesterol levels are linked to an increased risk of coronary heart disease, and the use of probiotics to reduce serum cholesterol levels has attracted much interest. The cholesterol-lowering ability of the isolated yeasts is shown in Figure 4b. Pichia terricola L9, isolated from lychee peels, demonstrated the highest cholesterol reduction ability (65%), while the other yeast strains, including SB, exhibited cholesterol reduction abilities of less than 40%.

4. Discussion

As a result of metabarcoding analysis on yeast communities, the abundance of yeast genera in various sources was as follows: lychee peels (Symmetrospora, Starmerella, Wickerhamiella, Macalpinomyces, Candida, and others); fermented shrimp paste (Hanseniaspora, Lachancea, Candida, and Tortispora); rice wine (Candida and Lachancea); and palm wine (Starmerella, Wickerhamiella, and Schwanniomyces). Notably, genera such as Candida, Hanseniaspora, Starmerella, and Lachancea, reported for their probiotic potential [32,33], were identified across multiple sources, with lychee peel emerging as a promising source for probiotic yeasts due to its high diversity and abundance. Yeast species on fruit surfaces must adapt to fluctuating conditions, possibly through nutrient sensing and transporter gene expression, which require further study. Osmotic diffusion may facilitate substrate translocation from inner tissues to the surface, supporting yeast growth [34]. Furthermore, under stress conditions, yeasts may adapt by efficiently utilizing unfamiliar molecules as nutrient sources. The ability to hydrolyze cutin, waxes, and hemicellulosic components may explain the frequent isolation and adaptive properties of certain species, potentially linked to probiotic properties, highlighting the need for further investigation.

While metabarcoding analysis provides broad-spectrum insights into microbial communities, it has limitations that may lead to misidentification, including primer bias, limited marker gene resolution, and the inability to accurately detect low-abundance species, especially when DNA from non-viable cells is present. Dominant species contribute more template DNA, leading to overrepresentation, whereas rare taxa may be underrepresented or missed entirely [19]. Despite these limitations, metabarcoding serves as a valuable complement to traditional culture-based methods by enabling the identification of viable but non-culturable species. While culture-based techniques remain essential for isolating viable strains for further development and commercial applications, metabarcoding data can serve as a guide for targeting specific yeast genera with desirable traits.

Yeasts in natural ecosystems, particularly those associated with food fermentation, have been extensively studied, often using the culturing plate method. Saccharomyces is typically the most abundant yeast genus in alcoholic beverages and plays a central role in alcohol production. However, non-Saccharomyces yeasts are crucial in enhancing sensory properties, such as flavor and aroma. Several studies suggest that non-Saccharomyces yeasts dominate during the early stages of fermentation [35]. Notably, when Saccharomyces yeasts are not dominant, they are less frequently observed using both culture-dependent and non-culture-dependent methods [19].

In rice wine, traditional culturing methods identified Saccharomyces cerevisiae as the most abundant yeast species, while metabarcoding analysis showed a higher presence of Rhizopus, a filamentous fungus. This higher Rhizopus abundance may result from detecting DNA from non-viable cells. Both Rhizopus and Saccharomyces play key roles in rice wine fermentation. Rhizopus produces enzymes like amylase and glucoamylase during the initial stages by breaking down starch into glucose [36,37], while Saccharomyces metabolizes glucose and amino acids in later stages to produce ethanol and flavor compounds [38]. Similar dominant groups, including Saccharomyces, Rhizopus, and lactic acid bacteria (LAB), have been reported in Japanese sake and Wuyi Hong Qu glutinous rice wine [20,39]. Factors influencing fungal diversity in rice wine include fermentation processes, local raw materials, environmental conditions, and starter cultures. Yeasts with alcohol-tolerance mechanisms such as specialized cell walls and transmembrane proteins support better nutrient uptake and metabolite excretion, positioning them as potential probiotics. Yeast isolates from harsh fermentation environments have shown promising probiotic properties [10].

In palm wine, the culturing plate method identified Hanseniaspora sp. as the most abundant yeast species, aligning with findings from toddy fermentation in India, where metabarcoding analysis highlighted Hanseniaspora as the dominant genus, followed by Saccharomyces, Lachancea, Starmerella, and Torulaspora [40]. However, in this study, metabarcoding analysis revealed Starmerella and Wickerhamiella as the dominant genera instead, illustrating differences in the microbial profiles. Non-Saccharomyces yeasts were dominant in the palm wine ecosystem, with Starmerella, Wickerhamiella, and Schwanniomyces as key species, consistent with previous observations. Reports indicate non-Saccharomyces yeasts such as Metschnikowia pulcherrima, Lachancea thermotolerans, Torulaspora delbrueckii, and Zygosaccharomyces spp. are effective in producing wines with lower ethanol content under oxygenated conditions [41]. The dominance of non-Saccharomyces yeasts, particularly Hanseniaspora, suggests a lower alcohol production in palm wine compared to rice wine. This may influence the probiotic potential and cell properties of yeasts, particularly in low-alcohol versus high-alcohol fermentations [42]. These differences remain an area of active research, particularly in terms of their probiotic effects.

The microbial profile of fermented shrimp paste, based on the culturing plate method, showed an abundance of Schwanniomyces etchellsii and an unidentified species, while metabarcoding analysis revealed a higher presence of Hanseniaspora and Lachancea. Previous studies on fermented fishery products have primarily focused on LAB rather than yeast. Yeasts such as Metschnikowia, Wickerhamomyces, and Candida are typically dominant in these products [33]. Additionally, other yeast species like Zygosaccharomyces rouxii, Debaryomyces sp., and Hansenula sp. have been identified in the context of traditionally fermented fish in Thailand and Ghana [43]. In this study, the most abundant yeast genus in fermented shrimp paste during the early stages of fermentation was Hanseniaspora, followed by Lachancea, Candida, Rhizopus, and other genera. This pattern differs from previous findings, potentially influenced by manufacturing processes, raw materials, and variations in salt content. Salt type and concentration are key factors influencing microbial communities in fermentation, as they select microorganisms of interest for the development of novel probiotic strains for osmotolerant and lipid-utilizing yeasts [44,45]. Further investigation is required to explore this potential.

For lychee peels, the culturing plate method detected Pichia terricola, Kurtzmaniella quercitrusa, Starmerella meliponinorum, and Meyerozyma guilliermondii, while metabarcoding analysis revealed the dominance of Symmetrospora, Starmerella, and Wickerhamiella. The surface of fruits is a natural environment for various yeasts, which are among the first to appear. Common yeasts found on fruit surfaces include Rhodotorula, Sporobolomyces, Cryptococcus, Torulopsis, Candida, Pichia, Hansenula, Kloeckera, Hanseniaspora, and, less frequently, Saccharomyces and Schizosaccharomyces [46]. Fernández Pacheco et al. [47] also identified Candida, Diutina, Hanseniaspora, Meyerozyma, Pichia, Rhodotorula, and Torulaspora on the surface of Brazilian fruits like acerola, pitanga, orange, and cajamanga. This study further identified diverse yeast genera on lychee peel, such as Inocybe, Symmetrospora, Starmerella, Wickerhamiella, Macalpinomyces, Candida, and Humaria. Lychee peel holds great potential as a source of novel yeasts, including those with probiotic properties like SB, which is derived from similar environments [4]. The abundant yeasts identified through both cultural and non-cultural methods highlight the need to further explore their probiotic potential and beneficial properties.

In summary, distinct yeast profiles were observed across different sources, even under similar environmental conditions. Cultural and non-cultural methods revealed different findings, emphasizing the importance of combining both approaches for a comprehensive understanding. The predominant yeast genera across these sources were as follows: Candida and Lachancea in rice wine; Starmerella and Wickerhamiella in palm wine; Hanseniaspora and Lachancea in fermented shrimp paste; and Symmetrospora, Starmerella, and Wickerhamiella on lychee peels. Many of these genera are already reported as potential probiotics, as previously mentioned. Minor isolates are also being tested for their probiotic properties, with further investigation into the influence of their environmental origins.

The preliminary screening of probiotic microorganisms requires their survival under gastrointestinal tract conditions. As a result, 23 isolates showed tolerance to acid and bile salt with more than 50% survival as SB. They were selected to continue the study on cell adhesion properties. Interestingly, the number of yeast isolates with tolerance to gastric acid and bile salt is mostly isolated from lychee peels. These may demonstrate that yeast in natural habitats with high stress, such as low nutrients, low water activity (a_w), exposure to sunlight and UV, as on fruit peels, could be a key factor associated with yeast adaptation leading to a tolerance strain with potential probiotic activity. The stresses could affect the change in the physiological and compositional properties of the yeast cell wall [10]. The cell wall is a highly dynamic structure whose biochemical and biophysical properties can be finely tuned as yeast cells encounter different stresses [48]. The cell wall maintains cell shape and integrity, and, together with the plasma membrane, forms the cell envelope, serving as the first line of defense against multiple adverse environmental conditions [48]. Heat stress also induces changes in the morphology of the yeast cell surface with the formation of circular structures in the surface of heat-stressed cells [49]. Moreover, exposure to sunlight and UV, like on fruit peels, could be associated with oxidative stress to yeast adaptation, leading to a tolerant strain. Pandey et al. [50] reported that oxidative stress caused by exposure to a harmful environment induces distinct responses in Saccharomyces cerevisiae. The thickness and composition of the cell wall and plasma membrane contribute to decreased cell permeability, which enhances resistance to pro-oxidant compounds by restricting their diffusion into the cell [51].

Hydrophobicity evaluation can serve as a preliminary test for the ability of probiotics to adhere to epithelial cells. This reflects the nature of the source, as fruit surfaces tend to be hydrophobic to reduce water absorption, maintain peel integrity, and protect against pests and microbial infections [52]. Therefore, microbes surviving in such environments are likely to exhibit higher hydrophobicity and better adaptation to nutrient consumption. According to De Souza et al. [53], microorganisms with higher hydrophobicity are more likely to adhere to epithelial cells. Additionally, hydrophobicity rates higher than 40% indicate that yeast isolates are hydrophobic [54]. A high capacity for cell surface hydrophobicity is crucial, as it allows yeast strains to effectively attach to hydrophobic surfaces, such as the human gastric mucosa and colon, thereby enabling them to confer health-promoting benefits to the host [55].

In the final part of this study, twelve strains were screened for health-promoting properties beyond probiotic traits, focusing on antioxidant capacity and cholesterol reduction. Yeasts isolated from lychee peels, Starmerella meliponinorum L12 and Pichia terricola L9, demonstrated notable antioxidant activity and cholesterol-reducing properties, respectively, superior to those from other sources. Previous reports have described various mechanisms that may be involved in the antioxidant activity of yeast cells, such as the composition of the cell wall and other substances, including antioxidant enzymes like superoxide dismutase, glutathione peroxidase, and catalase. In addition, the composition of the yeast cell wall correlated with the content of (1→3)-β-D-glucan and other β-glucans, which are also involved in antioxidant activity [56]. Particularly, β-glucans are recognized as compounds with various biological activities, including anti-inflammatory, antioxidative (in terms of free radical scavenging), and immunomodulatory effects [56]. Several studies have reported that Saccharomyces cerevisiae produces various antioxidant metabolites, including beta-glucan, GABA, mannans, and mannoproteins, as well as acids and alcohols, which contribute to its ability to combat oxidative stress and offer potential health benefits [57,58]. However, other genera of yeast have rarely been reported to exhibit antioxidant abilities through the production of antioxidant metabolites. This observation highlights the potential for further in-depth studies, such as whole-genome analysis, to investigate and confirm these properties. Interestingly, the source of yeast isolates significantly influences their ability to produce metabolites, as environmental factors, ecological niches, and stress conditions can affect their metabolic pathways. Stressful environments enable yeasts to develop resilience mechanisms, resulting in key traits such as the production of substances that protect against oxidative damage, resistance to dehydration or oxygen depletion, enhanced cell wall formation, and internal system adaptations [9,10]. Additionally, Siesto et al. [59] reported that the source (or strain) of yeast is an important factor affecting β-glucan content of the cell wall.

Several studies also reported that probiotics can be an effective tool for lowering cholesterol levels, particularly probiotic bacteria such as Lactobacillus reuteri NCIMB 701089, Lactobacillus fermentum NCIMB 2797, and Lactobacillus casei strains AP, which have demonstrated cholesterol reduction percentages ranging from 40% to 70% [60,61]. The mechanisms related to cholesterol reduction by probiotics have been proposed, including (i) the inhibition of cholesterol synthesis by hypocholesterolemia factors such as uric acid, lactose, orotic acid, and whey protein [62], as well as (ii) the deconjugation of bile acids via bile salt hydrolase (BSH) [63]. This study examines the cholesterol reduction potential of yeast, which may produce BSH that deconjugates bile salts, making them less soluble and reducing their reabsorption in the intestines. As a result, the host utilizes more cholesterol to synthesize new bile acids, thereby lowering blood cholesterol levels.

The Pichia genus (formerly Hansenula) has been extensively utilized in biotechnology and molecular biology, as reported in previous studies [64]. Additionally, Pichia terricola and other non-Saccharomyces yeasts play a vital role in enhancing wine quality by improving its aroma and flavor profile [65]. Interestingly, Ivashov et al. [66] reported that some Pichia, such as P. pastoris, possess the potential protein Are2p, associated with steryl ester synthases for lipid esterification, which exhibits superior cholesterol-lowering capabilities compared to Saccharomyces cerevisiae. As mentioned earlier, in addition to simple components like sugars, organic acids, and amino acids, the fruit surface also contains outer structural components of the exocarp. The affinity of yeasts toward fresh fruits may determine the specificity of their association with fruit surfaces. Establishing these associations is important to understand how yeasts contribute to the degradation of the cuticular layer, primarily through the production of extracellular enzymes like lipases, cutinases, and other enzymes (collectively referred to as hydrolases) [67]. These enzymes enable yeasts to break down the cuticular waxes, which are primarily composed of long-chain fatty acids, alcohols, esters, and hydrocarbons [6], into smaller carbon sources that the yeasts can utilize for growth and metabolism. While waxes are esters of fatty acids and alcohols, cholesterol is a hydrophobic sterol. Yeasts that produce hydrolases (e.g., lipases and esterases) for degrading waxes, such as those isolated from lychee peels, may also be capable of hydrolyzing cholesterol esters into free cholesterol and free fatty acids. Despite their structural differences, both cholesterol and waxes are hydrophobic, making them suitable substrates for broad-specificity enzymes. This suggests that yeasts could possess overlapping enzymatic capabilities for metabolizing both waxes and cholesterol. In addition to general hydrolases, some microorganisms produce enzymes specific to cholesterol degradation, including cholesterol oxidases, sterol side-chain oxidases, and cytochrome P450 monooxygenases [68]. These enzymes enable microbes to effectively metabolize cholesterol and its derivatives. Based on the properties of the Pichia yeast isolate, further research is required to develop yeast strains optimized for cholesterol reduction. Key areas of focus include identifying and characterizing the enzymes involved and understanding the genetic and metabolic pathways that enable these processes. Environmental conditions likely influence these adaptive mechanisms, as yeasts must convert available substrates into nutrients for survival.

This insight suggests that alternative yeasts may offer promising potential for biotechnological applications, particularly in the probiotic field, with the ability to promote specific health benefits, such as cholesterol reduction, that may help mitigate chronic diseases like cardiovascular disease, as well as antioxidant properties that contribute to life extension.

5. Conclusions

Exploring and investigating potential probiotic yeast candidates from various sources using cultural and non-cultural methods revealed distinct findings. Each source exhibited unique yeast profiles influenced by environmental conditions affecting yeast communities, underscoring the importance of combining both approaches for a comprehensive understanding. Remarkably, among the sources tested in this study, lychee peels mainly isolated yeast with potential probiotic strains. Furthermore, potential probiotic yeasts isolated from lychee peels, Starmerella meliponinorum L12 and Pichia terricola L9, demonstrated antioxidant activity and cholesterol-reducing properties, respectively, superior to those of other sources. This highlights the influence of natural habitats with high-stress conditions, such as low nutrient levels, low a_w, and exposure to sunlight and UV radiation, as seen on fruit peels, as key factors driving yeast adaptation and the development of tolerant strains with potential probiotic properties. These findings highlight the potential of natural sources such as food surfaces to harbor novel probiotic strains for highly adaptive strains for food and health applications. Although this study presents promising in vitro screening results, further evaluation is essential to confirm the functional efficacy and safety of the selected strains. Such evaluations should include whole-genome sequencing analysis (in silico), comprehensive in vitro safety assessments following established guidelines, and in vivo validation using appropriate animal models.