Probiotic Yeasts and How to Find Them—Polish Wines of Spontaneous Fermentation as Source for Potentially Probiotic Yeasts

One approach towards maintaining healthy microbiota in the human gastrointestinal tract is through the consumption of probiotics. Until now, the majority of probiotic research has focused on probiotic bacteria, but over the last few years more and more studies have demonstrated the probiotic properties of yeast, and also of species besides the well-studied Saccharomyces cerevisiae var. boulardii. Probiotic strains have to present the ability to survive in harsh conditions of the host body, like the digestive tract. Must fermentation might be an example of a similar harsh environment. In the presented study, we examined the probiotic potential of 44 yeast strains isolated from Polish wines. The tested isolates belonged to six species: Hanseniaspora uvarum, Pichia kluyveri, Metschnikowia pulcherrima, Metschnikowia ziziphicola, Saccharomyces cerevisiae and Starmerella bacillaris. The tested strains were subjected to an assessment of probiotic properties, their safety and their other properties, such as enzymatic activity or antioxidant properties, in order to assess their potential usefulness as probiotic yeast candidates. Within the most promising strains were representatives of three species: H. uvarum, M. pulcherrima and S. cerevisiae. H. uvarum strains 15 and 16, as well as S. cerevisiae strain 37, showed, among other features, survivability in gastrointestinal tract conditions exceeding 100%, high hydrophobicity and autoaggregation, had no hemolytic activity and did not produce biogenic amines. The obtained results show that Polish wines might be a source of potential probiotic yeast candidates with perspectives for further research.


Introduction
The human microbiota significantly influence the well-being of the human host, and may participate in the development of a wide variety of diseases.One of the approaches towards maintaining healthy microbiota is through probiotic consumption [1,2].The World Health Organization defines probiotics as live microorganisms that, when administered in sufficient amounts, confer a health benefit.While the majority of probiotic research has focused on bacterial strains (mainly Lactobacillus (according to previous nomenclature) and Bifidobacterium), the exploration of probiotic yeast is just gaining momentum due to its potential in terms of unique features and potential therapeutic applications [3][4][5].Most probiotic yeast research focuses on Saccharomyces cerevisiae var.boulardii, the only species of yeast that has been tested well and is widely used as a probiotic [6][7][8][9].However, more and more research is focused on the search for new strains with probiotic properties, not only Saccharomyces cerevisiae var.boulardii [4,[10][11][12].Examples of non-Saccharomyces species with probiotic properties may be Debaryomyces hansenii and Kluyveromyces marxianus [13][14][15].
Fermented foods and beverages have been a source of probiotic microorganisms in the human diet since the dawn of time.The coexistence of yeasts and lactic acid bacteria often occurs in a synergistic manner, where they mutually enhance their growth and survival [16][17][18].Examples of fermented products that can be sources for the isolation of probiotic strains are cheese, dough, dairy products, fermented fruits and vegetables, and traditional food and beverages [19][20][21].Among such products, wines deserve attention as a source of probiotic microorganisms due to the richness of their microbiota [22][23][24][25][26][27][28].Examples of such potentially probiotic isolates include bacteria like Lactobacillus fermentum, Lactobacillus rhamnosus, Pediococcus pentosaceus [23,29] and both Saccharomyces and non-Saccharomyces yeasts like Candida spp., Hanseniaspora spp., Pichia spp.and Torulaspora delbrueckii [27,28,30].Over recent years, with the warming of Poland's climate, there has been an increased interest in viticulture, resulting in the growth of vineyards by over 550 in the last decade [25,31].At the time of publication preparation, only two publications characterizing native yeast strains of Polish wines and one metagenomic study were available, but none of them were dedicated to exploring the probiotic potential of the tested strains [22,25,32].To the best of our knowledge, this is the first study focusing on the aspect of probiotic potential of yeast isolates derived from Polish wines and may provide some new perspectives on wine-derived yeasts besides the typical look inside their influence on wine's enological properties, such as aroma, acidity, texture, etc.
The main aim of the study was the isolation of probiotic yeast candidate strains from spontaneously fermented Polish wines and the in vitro testing of their potentially probiotic and safety properties.

Yeast Strains
Forty-four yeast strains were obtained from the culture collection of the Department of Biotechnology, Microbiology and Human Nutrition, Faculty of Food Science and Biotechnology, University of Life Sciences in Lublin.The strains originated from spontaneously fermented wines produced from grapes of the Regent variety sourced from three Polish vineyards, "Dom Bliskowice" (DB), "Małe Dobre" (MD) and "Winnica Janowiec" (WJ), located in the Małopolska Vistula Gorge region in 2019.Fresh cultures for the following experiments were grown for 48 h at 28 • C in YPD (yeast extract peptone dextrose) broth (BTL, Łódź, Poland).A commercial probiotic yeast, Saccharomyces cerevisiae var.boulardii CNCM I-745 (Enterol, Biocodex, France), was used as a positive control.To ensure consistency in the measurements, the initial optical density at 600 nm (OD 600 ) of each culture was determined.Subsequently, the results were standardized to an OD 600 value of 1.0 for each experiment, unless stated otherwise.This normalization allowed for accurate and comparable data analysis across different cultures.The strains were preserved in freezing conditions (−20 • C) with glycerol (20%).The code names for individual strains with vineyard origin are listed in Table 1.

Yeast Identification
To verify the identification of the strains obtained from the collection, they were subjected to species identification based on ITS1-5.8SrDNA-ITS2 regions.Strains' identification was obtained via 5.8S-ITS gene sequencing.DNA isolation was performed according to the procedure of Genomic Mini AX Yeast (A&A Biotechnology, Gda ńsk, Poland).Amplification of the region was performed with the primers ITS1 (5 TCCGTAGGTGAACCTGCGG-3 ) and ITS4 (5 -TCCTCCGCTTATTGATATGC-3 ) using PCR Mix Plus Green (A&A Biotechnology, Gda ńsk, Poland) with both primers concentration 0.5 µM and 1 µL of fungal DNA template [33].Amplification conditions: cycle of initial denaturation at 95 • C for 5 min, followed by 34 cycles: denaturation at 95 • C for 1 min, annealing at 56 • C for 1 min, extension at 72 • C for 2 min with final extension step at 72 • C for 10 min.Sequencing of the amplicons was performed using Genomed (Warsaw, Poland).Contigs were assembled with DNA Sequence Assembler v5 (Heracle BioSoft, www.DnaBaser.com)(accessed on 10 January 2023).Obtained contigs were compared to the known sequences of the ITS region in the NCBI GenBank database via alignment with the BLAST algorithm and evolutionary analyses were conducted in MEGA11.The phylogenetic tree was created based on the ITS region sequences using the maximum likelihood method with the best-fit Kimura 2-parameter model with a discrete gamma distribution (+G), with 1000 bootstrap replication [34].For checking the ability to grow at human body temperature, 10 µL of fresh yeast culture was inoculated into 1 mL of YPD (Yeast Peptone Dextrose) broth and was incubated at 37 • C for 48 h.Survival under these conditions was assessed via growth or no growth after 48 h.If the strain did not show the ability to grow in the above-mentioned conditions, it was removed from further analysis.

Survival and Growth under Gastrointestinal Tract Conditions
All strains undergo in vitro digestion according to Fernandez-Pacheco et al., 2018, with modification [12].One milliliter of fresh culture was centrifuged (5000 rpm, 10 min), biomass was washed with YPD broth in PBS (phosphate-buffered saline) at pH = 2 supple-mented with 3 mg/mL pepsin, centrifuged under the same conditions and after washing, inoculated into 1 mL of gastric solution containing YPD supplemented with 3 mg/mL pepsin in phosphate-buffered saline at pH = 2 to simulate gastric conditions.The mixture was then incubated at 37 • C for 3 h.Following this, the gastric solution was centrifuged (5000 rpm, 10 min) and the biomass was washed with an intestinal solution composed of YPD broth containing 1 mg/mL pancreatin and 0.5% bile salts, adjusted to pH = 8 and was transferred to the same intestinal solution and incubated at 37 • C for 22 h.To assess survival and growth of cells under gastrointestinal conditions, the spread plate technique on YPD agar was used after serial dilution preparations [35,36].
Survival rate and growth was calculated as % by the formula: survival rate [%] = concentration of living cells after in vitro digestion/ concentration of living cells before in vitro digestion × 100%

Hydrophobicity of Cell Surface
Cell surface hydrophobicity was tested according to Amorim et al., 2018 [37].One milliliter of fresh culture was centrifuged (5000 rpm, 10 min), biomass was washed twice and was resuspended in 5 mL of PBS (phosphate-buffered saline) at pH = 7.Three milliliters were blended with 1 mL of xylene.The mixture was vigorously shaken for 2 min and then allowed to settle undisturbed at 37 • C for 30 min, facilitating the complete separation of the phases.Subsequently, the aqueous phase was carefully removed, and the absorbance at 600 nm was measured spectrophotometrically.The decrease in the absorbance was taken as the measure of cell surface hydrophobicity calculated using the formula bellow: where OD initial and OD final are the absorbance (at 600 nm) before and after extraction with xylene [37][38][39].
The yeasts were grown in YPD broth at 37 • C for 24 h.After that, they were centrifuged (5000 rpm, 10 min) and washed twice with PBS at pH = 7.The biomass was then resuspended in PBS.To assess autoaggregation, 3 mL of the cell suspension was vortexed for 10 s.Autoaggregation was determined spectrophotometrically after 2, 4 and 24 h of incubation at 37 • C via absorbance (A) (OD 600 ) and it was expressed as: where A t is the absorbance at 2, 4 or 24 h and A 0 is at zero time [37].

Antioxidant Activity
Antioxidant activity assay was carried out according to Gil-Rodriguez et al., 2015 [40].One milliliter of yeast culture in YPD broth was centrifuged (5000 rpm, 10 min), washed twice with a sterile solution 0.9% NaCl and the pellet was resuspended in 1 mL of 0.9% NaCl.Next, 800 µL of the cell suspension was transferred to a new tube, to which 1 mL of a DPPH (2,2-diphenyl-1-picrylhydrazyl) solution (0.2 mM in methanol) was added.The mixture was vortexed and incubated in darkness at room temperature for 30 min.Following incubation, the reaction tubes were centrifuged (12,000 rpm, 5 min) and 300 µL of the resulting supernatant was transferred to 96-well plates for measurement of the absorbance at 517 nm (A 517 ).The percentage of reduction in DPPH was then calculated using the following formula [40]:

Percentage of reduction of DPPH [%] = [(A 517 control − A 517 sample )/A 517 control ] × 100%
where A 517 control represents the absorbance of the control (DPPH solution without yeast) and A 517 sample represents the absorbance of the sample (yeast culture treated with DPPH solution).

Antimicrobial Activity
The well-diffusion method was employed to evaluate the antimicrobial activity of the yeast strains against various bacterial species [41,42], including Escherichia coli, Salmonella enterica, Staphylococcus aureus, Bacillus cereus, Listeria monocytogenes and Enterococcus faecalis.The bacterial strains were cultivated in BHI (brain heart infusion) broth at 37 • C overnight.Then, the bacterial cultures were individually inoculated with 1mL of bacterial suspension onto BHI agar plates, spread and left to allow for the liquid absorption.Afterward, three wells (5 mm diameter each) were cut in the center of every plate at equal intervals between each other and inoculated with fresh yeast culture.The plates were incubated for 24 h at 37 • C, to evaluate bacterial growth inhibition.

Hemolytic Activity
To test hemolytic activity, yeasts were streaked onto blood agar plates (Columbia agar supplemented with 5% defibrinized sheep's blood) (Biomaxima, Lublin, Poland).The plates were then incubated at 37 • C for 72 h.Following incubation, the plates were observed for any visible signs of hemolysis.As a positive control, Staphylococcus aureus, known to exhibit hemolysis, was included in the experiment.

Biogenic Amine Production
To evaluate strains' ability to produce biogenic amines, the method suggested by Aslankoohi et al., 2016 was used [43].Yeasts were streaked onto YPD agar plates supplemented with 0.006% bromocresol purple and an amino acid mix (tyrosine, histidine, phenylalanine, leucine, tryptophan, arginine and lysine in equal ratios) with a total mass concentration of 1%.For the detection of biogenic amines (BAs), the plates were incubated at 30 • C for 7 days.Throughout the incubation period, the growth of strains and any changes in the color of the medium were monitored on a daily basis.This monitoring aimed to identify the presence of biogenic amines, which could be indicated by specific changes in the appearance of the medium.In yeast strains that produce biogenic amines, the process of amino acid decarboxylation resulted in the immediate appearance of a purple halo surrounding the growth area.On the other hand, in the strains that did not produce biogenic amines, the growth area exhibited a yellow halo surrounding it, which was attributed to glucose fermentation.During the growth of the strain, pH reduction occurred, leading to the medium gradually turning purple [43].

Enzymatic Activity
We determined the enzymatic activities of the strains using the API ZYM system (bio-Merieux, Craponne, France) according to the manufacturer's recommendations.Yeast cell suspensions were transferred into the wells of the API ZYM strips and incubated at 37 • C for 4 h.After the incubation period, one drop each of the reagents ZYM A and ZYM B was added to each well.Color changes observed in the wells indicating positive enzymatic reactions were noted and used for evaluation of the results on the basis of the API ZYM color chart [44].

Statistical Analysis
The results were analyzed with Statistica version 13.3 (2017) for Windows (StatSoft Inc., Tusla, OK, USA).The one-way ANOVA followed by Tukey HSD post hoc tests (p < 0.05) was used to compare the results for: survivability, hydrophobicity and antioxidant activity for each strain, and mixed two-way ANOVA for level of autoaggregation for testing strains after 2, 4 and 24 h.

Yeast Identification
All strains were identified to the species level (Table 1).All forty-four identified strains belong to five yeast genera: Saccharomyces (29.54% of strains), Starmerella (25% of strains), Hanseniaspora (22.72% of strains), Metschnikowia (18.2% of strains) and Pichia (4.5% of strains).The results of the phylogenetic analysis are presented in Figure 1.Out of forty-four tested strains only twenty-one isolates showed an ability to survive and grow at 37 • C. Within tested strains, the highest percentage of survived strains was shown by Pichia kluyveri (100% of strains survived), followed by Saccharomyces cerevisiae (76.9% of strains survived), Metschnikowia spp.(62.5% of strains survived) and Hanseniaspora uvarum (40% of strains survived).None of the tested Starmerella bacillaris isolates survived.Detailed data are presented in Figure 2. All strains that did not demonstrate the ability to grow at 37 • C were excluded from further studies.

Survival and Growth under Gastrointestinal Tract Conditions
As the next step, the remaining strains were subjected to in vitro digestion.Out of twenty-one strains subjected to the procedure, eleven strains showed the ability to survive simulated gastrointestinal conditions.Eight of the eleven strains showed better survivability than the positive control (Saccharomyces cerevisiae var.boulardii CNCM I-745 with survivability at 48.7%).Five strains demonstrated survivability above 100%: 16_Hans_uvarum (147.8%),15_Hans_uvarum (106.6%),37_Sacch_cerevisiae (104.6%),34_Metsch_pulcherrima (100.4%) and 36_Metsch_pulcherrima (100.1%).This may suggest that the conditions in the gastrointestinal tract for the five mentioned strains could be prevailing for their growth compared to standard culture conditions.Detailed data are presented in Figure 3.All strains that did not demonstrate the ability to survive gastrointestinal tract conditions were excluded from further studies and are not included in Figure 3.

Hydrophobicity of Cell Surface
The ability to adhere to intestinal epithelial cells is a crucial requirement for the colonization of potentially probiotic strains in the gastrointestinal tract, as it helps them avoid immediate elimination through peristalsis and gives a competitive advantage within the gastrointestinal tract.Many authors suggest that strains with high hydrophobicity exhibit high adherence to intestinal cell lines [45,46].All, except three strains: 36_Metsch_pulcherrima (54.63%), 32_Pich_kluyveri (59.42%) and 13_Metsch_ziziphicola (76.53%), showed hydrophobicity above 80% which may suggest their ability to quickly adhere to the mucosa.Detailed data are presented in Figure 4.

Survival and Growth under Gastrointestinal Tract Conditions
As the next step, the remaining strains were subjected to in vitro digestion.Out of twenty-one strains subjected to the procedure, eleven strains showed the ability to survive simulated gastrointestinal conditions.Eight of the eleven strains showed better survivability than the positive control (Saccharomyces cerevisiae var.boulardii CNCM I-745 with survivability at 48.7%).Five strains demonstrated survivability above 100%: 16_Hans_uvarum (147.8%),15_Hans_uvarum (106.6%),37_Sacch_cerevisiae (104.6%),34_Metsch_pulcherrima (100.4%) and 36_Metsch_pulcherrima (100.1%).This may suggest that the conditions in the gastrointestinal tract for the five mentioned strains could be prevailing for their growth compared to standard culture conditions.Detailed data are presented in Figure 3.All strains that did not demonstrate the ability to survive gastrointestinal tract conditions were excluded from further studies and are not included in Figure 3.

Hydrophobicity of Cell Surface
The ability to adhere to intestinal epithelial cells is a crucial requirement for the colonization of potentially probiotic strains in the gastrointestinal tract, as it helps them avoid immediate elimination through peristalsis and gives a competitive advantage within the gastrointestinal tract.Many authors suggest that strains with high hydrophobicity exhibit high adherence to intestinal cell lines [45,46].All, except three strains: 36_Metsch_pulcherrima (54.63%), 32_Pich_kluyveri (59.42%) and 13_Metsch_ziziphicola (76.53%), showed hydrophobicity above 80% which may suggest their ability to quickly adhere to the mucosa.Detailed data are presented in Figure 4.

Antimicrobial Activity
Only strain 15_Hans_uvarum showed weak antimicrobial activity against Staphylococcus aureus; other strains did not inhibit the growth of pathogenic bacteria.

Hemolytic Activity
None of strains tested in the experiment presented hemolytic activity.

Biogenic Amine Production
From all tested strains, three strains: 27_Sacch_cerevisiae, 32_Pich_kluyveri and 34_Metsch_pulcherrima produced a purple halo around the growth area, which indicate the presence of biogenic amines [43].

Enzymatic Activity
The API ZYM assay showed various activity profiles within the tested strains.All strains presented activity of alkaline phosphatase, esterase (C4), esterase lipase (C8) and acid phosphatase, but none of the strains presented activity of trypsin, α-galactosidase, β-

Hemolytic Activity
None of strains tested in the experiment presented hemolytic activity.

Biogenic Amine Production
From all tested strains, three strains: 27_Sacch_cerevisiae, 32_Pich_kluyveri and 34_Metsch_pulcherrima produced a purple halo around the growth area, which indicate the presence of biogenic amines [43].

Enzymatic Activity
The API ZYM assay showed various activity profiles within the tested strains.All strains presented activity of alkaline phosphatase, esterase (C4), esterase lipase (C8) and acid phosphatase, but none of the strains presented activity of trypsin, α-galactosidase, β-glucuronidase nor α-fucosidase.The differences in enzymatic activity between tested strains are presented in Table 2.  1 Strain number (see Table 1).

Discussion
At the time of preparation for publication, to the best of our knowledge, nobody has published studies focused on the aspect of probiotic potential of yeasts isolated from Polish wines.
The forty-four strains examined belonged to six species: Hanseniaspora uvarum, Pichia kluyveri, Metschnikowia pulcherrima, Metschnikowia ziziphicola, Saccharomyces cerevisiae, and Starmerella bacillaris.Despite limited research on yeast biodiversity in Polish wines, our findings mostly align with the lists of species described by Drozdz et al. and Cioch-Skoneczny et al. [22,32], but include one species not mentioned by the authors-Starmerella bacillaris.The species, S. bacillaris, known also as Candida zemplinina, is frequently isolated from grapes and wines, and may affect the chemical composition of the musts and wines by its ability to produce various metabolites which influence their enological properties [48][49][50].All species identified in the study commonly occur and are isolated from grape, must and wine environments worldwide [51][52][53].
Despite challenging environmental conditions for microorganisms during the wine fermentation process, only 11 out of the 44 strains subjected to the initial experiments demonstrated potential for further research.Two of the most discriminative tests for the experiments were the ability to survive at 37 • C and the ability to survive and grow under gastrointestinal tract (GIT) conditions, eliminating 52.3% and 47.6% of the strains at each of those steps.Among the isolates able to survive the two initial steps of the experiment were the strains belonging to H. uvarum, P. kluyveri, M. pulcherrima, M. ziziphicola and S. cerevisiae.Besides their enological potential, none of S. bacillaris strains tested in the study were able to survive and grow at 37 • C, which stands in opposition to the results obtained by Shen et al., who described S. bacillaris CC-PT4 as showing properties as a probiotic candidate [54].Current data about potentially probiotic strains belonging to S. bacillaris are limited to the data of Shen et al. and do not include exact information about the source of isolation of each strain and the information is mainly limited to the list of fruits bought from greengrocers in China.Possibly the reason for differences in the tolerance for the host temperature and the ability to survive under GIT conditions between S. bacillaris strains from this study and Shen et al.'s study might be differences in climate conditions at the place of isolate's origin, but such a hypothesis cannot be verified without more detailed data.
Gastrointestinal tract conditions play an important role in human health protection.serving as a barrier between the external environment and the body and helping to prevent the entry of harmful substances and pathogens.Such conditions also play important role in survival of potential probiotics [40].In our study, eight of the eleven strains presented better survival rate than Saccharomyces cerevisiae var.boulardii CNCM I-745 (positive control), with survivability at 48.7%.Two strains for both Hanseniaspora uvarum and Metschnikowia pulcherrima, and one strain of Saccharomyces cerevisiae exceeded 100% survival rate, which may suggest that GIT conditions are more favorable for them than typical culture conditions.Similar results for wine-derived strains were published by Vergara Alvarez et al. [27].Such properties may result from the adaptation of the strains to the conditions of wine fermentation, where low pH is common [55,56].
Pizzolitto et al. suggest that an autoaggregation rate at 80% or higher should be considered as high [47].For all examined strains, autoaggregation rates increased with time exceeding above 80% after 24 h and were higher than values obtained for strains derived from various food environments tested by Fernandez-Pacheco [12].Similar results for wine-derived strains were presented by Vergara Alvarez et al. [27].
According to Gil-Rodriguez et al., 2015, the antioxidant activity of yeasts can be a result of their cell walls and other cellular compounds and sometimes even higher than in lactic acid bacteria [40].In the study, antioxidant activity levels of tested isolates were close to those of the control, S. cerevisiae var.boulardii.
Despite some yeast strains exhibiting antimicrobial properties (i.e., M. pulcherrima known for its ability to produce antimicrobial pulcherrimin), none of the tested strains showed such properties against tested pathogenic bacteria, which was consistent with the results of other researchers [30,57,58].
As a safety aspect, no hemolytic activity was present in our wine-derived strains.That is a result comparable to those obtained by Corbu et al., 2023, andFernández-Pacheco et al., 2021 [59,60].However, the strains 27_Sacch_cerevisiae, 32_Pich_kluyveri and 34_Metsch_pulcherrima showed the ability to produce biogenic amines (BAs) that excludes them as candidates for potentially probiotic strains [43], despite their promising properties in previous assays, due to potential risks for consumers [61,62].Both Caruso et al., 2001, and Delgado-Ospina et al., 2021, showed that strains belonging to M. pulcherrima, S. cerevisiae and P. kluyveri might produce BAs [63,64].Comparing the number of the strains producing BAs to the total number of strains in their and our study (where only three of all strains produced BAs) the result can be surprising, especially in the context of places of isolation-wine, in which the presence of biogenic amines formed during the fermentation process is not unusual and influences the sensory properties of wine.Our findings are closer to results obtained by Landete et al., 2007, where none of the examined wine-derived yeast strains produced BAs [65].
All tested strains presented a wide range of enzymatic activity including enzymes that break down proteins and sugars like esterases, lipase, arylamidases and α-glucosidase, which may confer a potential improvement on food digestion.Moreover, none of the strains presented any activity of trypsin (the enzyme that can be related with pathogenicity of some microorganisms in human) or β-Glucuronidase related to transformation of pre-carcinogens into carcinogens and stimulation of colon cancer [66,67].Activity of enzyme aminopeptidases like arylamidases and esterases may influence product properties because of their biotransformation ability [21,[67][68][69].Four strains also presented activity of N-acetyl-β-glucosaminidase, an enzyme with the primary functions of targeting and hydrolyzing oligosaccharides containing chitin [70] that may possibly arise in the biocontrol and biotransformation of some fungal-and invertebrate-based foods.

Conclusions
Our data show that Polish wines may be a source of yeast isolates with probiotic potential.Environmental conditions that occur during the must fermentation can help to preselect yeast for some desirable probiotic traits like low pH tolerance.Although isolates obtained during the study were sensitive for high temperature, 52.3% of strains were not able to grow and survive at 37 • C. One of the potential hypotheses of such a result may be due to the climate in Poland, which, despite global warming, does not force the local yeast populations to evolve towards higher temperature tolerance, but further studies will be needed.
The strains with the highest resistance to the conditions in the gastrointestinal tract belong to the species: Hanseniaspora uvarum (15_Hans_uvarum and 16_Hans_uvarum), Metschnikowia pulcherrima (34_Metsch_pulcherrima and 36_Metsch_pulcherrima) and Saccharomyces cerevisiae (37_Sacch_cerevisiae).These strains exceeded the 100% survival rate at GIT conditions and at least doubled the survival rate showed by commercial probiotic yeast Saccharomyces cerevisiae var.boulardii CNCM I-745, which suggests their high adaptation to such harsh conditions and may allow a reduction in dosage in the case of probiotic administration.
Besides their ability to survive in GIT conditions, four of the five mentioned isolates showed high hydrophobicity and autoaggregation, which are highly desirable traits in probiotic strains, with the exception of 36_Metsch_pulcherrima with the lowest hydrophobicity rate among all tested strains.Antioxidant activities of tested isolates were close to the control S. cerevisiae var.boulardii.Unfortunately, the strains did not show the ability to inhibit the growth of common foodborne pathogens.
The selected strains showed valuable enzymatic activities necessary in food digestion and did not have activities of undesirable enzymes such as trypsin and β-glucuronidase.
As a safety aspect, none of the strains presented hemolytic activity, but 34_Metsch_ pulcherrima produced biogenic amines; thus, regardless of its other promising characteristics, it cannot be considered as a probiotic yeast candidate.
Here, we can suggest some improvements in the experimental design in future-in case of screening high number of isolates, we suggest screening the tolerance for both temperature and GIT condition at first and moving forward to testing the traits related to the virulence and safety aspects, which will allow a reduction in costs and workload, in the case of experiments where high number of isolates may present such undesirable traits.
In our opinion, Polish wines may be a source of potentially probiotic yeasts and based on our results we suggest 15_Hans_uvarum, 16_Hans_uvarum and 37_Sacch_cerevisiae as encouraging probiotic yeast candidates with perspectives for further research in aspects including food science, development of probiotic formulations, etc.

Foods 2023 , 17 Figure 1 .
Figure 1.Maximum likelihood tree based on ITS sequences of the studied isolates.

3. 2 .Figure 1 .
Figure 1.Maximum likelihood tree based on ITS sequences of the studied isolates.

Table 1 .
Code names for tested strains.

Table 2 .
Results of API-ZYM assay for selected yeast strains.