Characterization and comparison of yeasts from different sources for some probiotic properties and exopolysaccharide production

A total of 134 strains of yeasts isolated from fruits and vegetables (certain berry fruits, apples, pomegranates, carrots and grapes), free-range chicken feces and dairy products, and from 10 commercial yeast preparations were identified and subjected to analyses to determine their in vitro probiotic properties. Based on 26S rRNA sequence analysis all ten isolates from commercial products were identified as Saccharomyces cerevisiae , and the natural isolates as Candida corpophila , Candida diddensiae , Clavispora lusitaniae , Hanseniaspora opuntiae , Hanseniaspora uvarum , Kazachstania bovina , Kluyveromyces marxianus , Metschnikowia. pulcherrim a, Metschnikowia sp . , Meyerozyma carribbica , Pichia kluyveri and Wickerhamomyces anomalus . All isolates were found to be resistant to simulated gastric juice at pH 2.5 for 2 h and were able to grow at both 30 and 37 °C. The exopolysaccharide (EPS) production of isolates from commercial preparations and from natural sources varied between 249-275.22 and 27.95-272.22 mg/L, respectively. Two of the natural isolates had levels of EPS production comparable to the natural strains ( S. cerevisiae T8-3C and S. cerevisiae P25-1) with 264.63 and 272.53 mg/L, respectively. Osolates were also investigated for autoaggregation and coaggregation abilities. The highest coaggregation ability was determined for the Saccharomyces cerevisiae P25-1 strain against Staphylococcus aureus (ATCC 25923).


Introduction
As the food industry continues developing new products and processes, consumers are focused on the food safety, diet and health aspects of their food. Yeasts are important microorganisms in the fermentation of foods and beverages, and some yeasts have been used as biocontrol agents and novel probiotics (Fleet, 2007). During the last decades, since they have shown numerous beneficial effects on human health, their usage as probiotics has been increasing (Gil-Rodríguez et al., 2015). Probiotics are desirable and natural tools to providing balance to the intestinal microflora. They are consumed either as food or non-food preparations. New species of probiotics are constantly being identified (Gotcheva et al., 2002). Although most studies and applications of microbial benefits have focused on lactic acid bacteria (LAB), considerable efforts have also beeen directed towards researching yeasts for their beneficial effects (Muccilli & Restuccia, 2015;Suvarna et al., 2018).
Probiotic yeasts are non-pathogenic strains generally belonging to species of Saccharomyces cerevisiae (Rajkowska & Kunicka-Styczynska, 2010;Fakruddin et al., 2017). For yeasts intended to be used as probiotics, each strain must be characterized in vitro and in vivo (Moslehi-Jenabian et al., 2010;Suarez & Guevara, 2018). Among the yeast used as probiotics strains must have the ability to show acid tolerance, bile resistance and inhibition of the pathogen adhesion to mucus and epithelial cells (Food and Agriculture Drganization of the United Nations, 2002). Exopolysaccharides (EPSs) are produced by some yeast species, and some of these polysaccharides are useful in the food, cosmetic and pharmaceutical industries due to their specific physico-chemical and rheological properties (Pavlova et al., 2004;Kuncheva et al., 2007).
Some different yeast species such as Debaryomyces hansenii, Torulaspora delbrueckii, Kluyveromyces lactis, Kluyveromyces marxianus and Kluyveromyces lodderae have shown the ability to survive in the gastrointestinal tract and to inhibit enteropathogens (Moslehi-Jenabian et al., 2010). S. cerevisiae var. boulardii is a species most often used as a probiotic (Gil-Rodríguez et al., 2015), and is reported to be efficacious in the prevention or the recurrence of antibiotic associated diarrhea and colitis in humans ( Van der Aa Kühle et al., 2005). Numerous in vitro and in vivo studies indicated that S. boulardii can prevent severe diarrhea and enterocolitis induced by a range of enteric bacteria (Hu et al., 2018). These pathogens are Clostridium difficile, Vibrio cholerae, Salmonella enterica subsp. enterica serovar Typhi (Salmonella Typhimurium), Shigella flexneri, enterohemorrhagic Escherichia coli and enteropathogenic E. coli (Chen et al., 2013;Syal & Vohra, 2013;Xu et al., 2018).
On this study, isolation, characterization, and identification of yeast strains from natural sources and retail cultures were investigated. Some probiotic properties of the isolates were determined. The tested strains were compared with retail yeast strains with in vitro setting for preselection for future researches.

Isolation
Sixteen samples were prepared from natural sources as fruits and vegetables (certain berry fruits, apples, pomegranates, carrots and grapes), fourteen from free-range chicken feces, four from dairy products and ten commercial yeast preparations were used for isolation. Each sample was first homogenized in sterile phosphate saline (PBS); pH 7.0. Samples were spread on malt-extract glucose yeast-extract-peptone (MGYP) agar and plates were incubated at 30 °C for 48 h. After incubation, isolated colonies were propagated for simple staining microscopic observation, and the cultures were maintained in MGYP broth with 15% glycerol at -20 °C (Syal & Vohra, 2013).

Growth at 30 and 37 °C and determination of EPS production
Yeast isolates were screened for growth at two different temperatures and for qualitative analysis of EPS production. Cultures activated by two transfers in MGYP broth were inoculated at 1% and then were incubated at 30 and 37 °C for 2-5 days. Growth was evaluated by visual observation (Psomas et al., 2001). EPS production was conducted according to Sourabh et al. (2011) and Syal & Vohra (2013). Dvernight cultures were streaked on the surface of ruthenium red milk agar and Petri plates were incubated at 37 °C for 24 h. Non-ropy strains formed red colonies as a result of the staining of the microbial cell wall, while ropy isolates producing EPSs formed white colonies.

Survival in gastric juice
Activated cells were collected by centrifugation (7,000 g, 15 min.) and inoculated at the level of approx. 10 6 CFU/mL in a simulated gastric juice prepared according to Corcoran et al. (2005) and Cassanego et al. (2017). After incubation of yeast cultures for 18 h in 25 mL MGYP broth, the cultures were centrifuged at 7,000 g at 4 °C for 15 min, washed once in an equal volume of cold 1/4 Ringer's solution and centrifuged under the same conditions. Pellets were then resuspended in a volume of simulated gastric juice equal to the original culture and incubated at 37 °C for 2.5 h with constant stirring. Then samples were serially diluted in maximum-recovery diluent, plated on MGYP agar, and incubated at 37 °C for 48 h. The survival rate was calculated as the percentage of colonies observed on MYPG agar after exposure to pH 2.5 for 2 h compared to the initial cell concentration (Corcoran et al., 2005).

Autoaggregation assay
The cells were harvested by centrifugation at 5,000 g for 15 min, washed twice and resuspended in PBS to obtain viable counts of approximately 10 6 CFU/mL. Cell suspensions (4 mL) were mixed by vortexing for 10 s and autoaggregation was determined during 5 h of subsequent incubation at room temperature. Every hour 0.1 mL of the upper suspension was transferred to another tube with 3.9 mL of PBS and the absorbance (A) was measured at 600 nm. The autoaggregation percentage is expressed as Equation 1I: whereI: At represents the absorbance at time 1, 2, 3, 4 or 5 h and A 0 the absorbance at t 0 (Kos et al., 2003). Percentage values of autoaggregation <30% were considered low, between 30% and 60% intermediate, and >60% high at room temperature for 2 h. This assay was performed in duplicate (Binetti et al., 2013).

Coaggregation assay
The method for preparing the cell suspensions for coaggregation was the same as that for the autoaggregation assay. For determining coaggregation properties Staphylococcus aureus (ATCC 25923), Listeria monocytogenes (RSKK 472), E. coli D157I:H7 (ATCC 35150), and S. Typhimurium (ATCC 700408) were used. Equal volumes (2 mL) of bacteria and yeast cell suspensions were mixed together in pairs by vortexing for 10 s. Control tubes were set up at the same time, each containing 4 mL of yeast or bacteria suspension alone. The absorbance (A) at 600 nm of the suspensions was measured after mixing and after 5 h of incubation at room temperature. Samples were taken the same way as in the autoaggregation assay. The percentage of coaggregation was calculated using the Equation 2I: whereI: x and y represent each of the two strains in the control tubes, and (x + y) the mixture (Kos et al., 2003). Percentage values of coaggregation <30% were considered low, between 30 and 60% intermediate, and >60% high, at room temperature for 3 h. This assay was performed in duplicate (Binetti et al., 2013).

Quantitative analysis of EPS production
Dvernight cultures were heated at 100 °C for 5 min and then centrifuged at 5,000 g for 10 min at 20 °C. The supernatant was removed, and the pellet was suspended in 1 mL of 85% (w/v) TCA and centrifuged again. Aqueous phases were decanted, and the pellet suspended in 1 mL ethanol, washed twice more with ethanol, and the final pellet was dissolved in 1 mL of distilled water (Marshall & Rawson, 1999). The phenol sulfuric acid colorimetric test for polysaccharides was used and the amount of EPSs determined in terms of glucose according to the glucose standard curves (Dubois et al., 1956).

Isolation, preselection and identification
Natural and commercial yeast samples were plated on MGYP agar plates as described above. On all, 144 yeasts were selected from agar plates prepared from 44 different materials according to colony and cell morphologies. The ability of each of the isolates to grow at 37 °C (human body temperature) and 30 °C was investigated. Fifty-seven isolates grew at 37 °C as well as 30 °C. The remaining eighty-seven isolates showed weak or no growth at 37 °C. The strains able to grow at both temperatures were selected for further investigation. Dne of the important criteria for the selection of probiotic yeasts is the ability to grow at 37 °C (Van der Aa Kühle et al., 2005). Some researchers investigated growth at 37 °C as the first step of probiotic selection in yeast isolates from traditional Ondian foods (Syal & Vohra, 2013), infant feces and feta cheese (Psomas et al., 2001).
Recently, there is a growing interest in the isolation of microbial EPSs. This compound has been found to have many different industrial applications because of its wide diversity instructural and chemical properties. When added to food, EPSs can positively improve the rheological properties and sensory qualities of the final product (Ramirez, 2016). Moreover, the roles of EPSs in probiotic activities have been determined to be important; some of these includes prebiotic potential, ability to adhese to the intestinal epithelium (Sarıkaya et al., 2016), and contributions to human health (Ramirez, 2016), as well as antioxidant, antitumor, antiaging and immunostimulant activities (Cheng et al., 2012;Kšonžeková et al., 2016). Consequently, EPS producing probiotic cultures are reported to positively contribute to human health. Syal & Vohra (2013) mentioned that S. cerevisiae strains and some other yeast strains isolated from traditional fermented food products have the EPS producing capacity. For these reasons, ability to grow at 37 °C and EPS producing capacity were conducted for preselection of the isolates. On this study, thirty-five of the natural isolates and ten isolates from commercial products survived at 37 °C and also were shown to be EPS producers by the qualitative method described above. These isolates were selected for use in the investigation of further probiotic properties. This initial screening decreased the number of yeasts to be identified and investigated for other probiotic properties.
A total of forty five isolates were identified by sequencing the D1/D2 domains of the large subunit of the 26S rRNA gene. All isolates (SB codes) from commercial products were identified as S. cerevisiae. The isolates T8-3C from chicken feces and P25-1 from cheese samples were also identified as S. cerevisiae. Sequence analyses of the yeast isolates extracted from different materials revealed that six of the strains belonged to Kazachstania bovina, five to K. marxianus, four to Metschnikowia pulcherrima, three to Candida albicans and Pichia kluyveri each, two to Clavispora lusitaniae, Hanseniaspora opuntiae, Hanseniaspora uvarum and Metschnikowia sp. each, and one to Candida corpophila, Candida diddensiae, Meyerozyma carribbica, Wickerhamomyces anomalus each.

Survival rate of isolates in simulated gastric juice
All of the forty-five yeasts could survive in simulated gastric juice at pH 2.5 for 2.5 h. Survival rates were consistently higher than 90%. Osolation sources and identification results and log reduction at pH 2.5 for 2.5 h of the strains are given in Table 1.
The reduction was determined less than 1 log for all yeasts. Survival of both yeasts was similar, showing a reduction of less than 1 log after treatment. After 2.5 h incubation at pH 2.5, fifteen of thirty five yeasts from natural samples showed good viability; 8 and 4 of the yeasts showed 0.20-0.50 and 0.5-0.7 log 10 CFU/mL reduction, respectively and the decrease of the 8 yeasts count was just 0.1-0.2 log 10 CFU/mL. While seven of the yeasts from commercial preparates survived at the same number, three of the strains showed 0.1-0.2 log 10 CFU/mL decrease.
According to Syal & Vohra (2013), similar results have been reported by several in vitro studies, especially in the yeasts belonging to the species of Saccharomyces, Debaryomyces and Kluyveromyces. Similar results have been reached by using yeasts isolated from broiler excreta; the strains were also capable of surviving and growing under stress conditions such as 2.0, 2.5 and 3.0 pH and 0.3-0.6% bile salts concentrations with survival rates higher than 98% (García-Hernández et al., 2012). Van der Aa Kühle et al. (2005) also demonstrated that all yeasts survived after 4 h of incubation at pH 2.5. All tested yeasts isolated from chicken feces and kefirs showed high survival for 8 h of incubation at pH 2.5 (Rajkowska & Kunicka-Styczynska, 2010).

EPS production
EPS production of the forty-five isolates varied almost tenfold, between 27.95 and 275.22 mg/L. According to the findings, EPS production of the strains isolated from commercial yeasts was between 249-275 mg/L. S. cerevisia T8-3C and P25-1, K. bovina T1-3H produced EPS more than 250 mg/L, whereas eight strains produced EPS between 200 and 250 mg/L. The EPS production of S. cerevisiae T8-3C and S. cerevisiae P25-1 showed almost the same level of EPS production with commercial S. cerevisiae Many bacteria and yeasts are able to produce EPS, which may show variations in monomer composition, types of branching, and molecular weight. Therefore, the functionality of EPSs produced from microorganisms is immensely varied (Schmid et al., 2016). Gientka et al. (2016) investigated the influence of carbon sources on EPS biosynthesis of the strains Candida famata and Candida guilliermondii isolated from kefirs. The strains were determined to be good EPS producers in their previous study (Gientka & Madejska, 2013). EPS production resulting from growth on different carbon sources ranged from 4.13 to 7.15 g/L. The highest biomass yield was reported for C. guilliermondii after cultivation on maltose and the maximum EPS production was determined as 0.505 and 0.321 mg/L for C. guilliermondii and C. famata, respectively. On another research, Ghada et al. (2012) investigated Rhodotorula glutinins isolated from soil, and they determined the optimum culture conditions of EPS synthesis. Maximum EPS production was 2.6 g/L and the specific production was 0.34 g/g based on dry weight cell with 0.1 of consumed glucose. Review of the literature indicate that EPS production efficiency varies heavily depending on the ambient conditions to which the isolates are subjected (Schmid et al., 2016). Further optimization studies are underway for EPS production of yeasts tested in the present study.

Autoaggregation/coaggregation assays
The capability of forming aggregates is one of the most desirable characteristics of a potential probiotic microorganism, because aggregation of microorganisms affects microbial adherence to the intestines, thus providing potential competitive advantage in the colonization of the GO tract (García-Cayuela et al., 2014). Binetti et al. (2013) described the values of autoaggregation in percentages; <30% being low, between 30 and 60% being intermediate, and >60%. Prior research studies have used various time intervals and total incubation periods (Syal & Vohra, 2013;Dgunremi et al., 2015). On our study, autoaggregation (%) abilities of the strains were measured every 1 h over a period of 5 h.
Most of the strains showed similar mean values of autoaggregation around 90% after 5 h incubation. Dnly four isolates (M. pulcherrima MS1-1B and MS1-3C, K. bovina T1-3G, P. kluyveri U1-1A) exhibited autoaggregation abilities below 90% after 5 h. The autoaggregation of commercial isolates and the other isolates from natural samples were determined as 56.64-93.97 and 41.85-97.59%, respectively after 2 h incubation at room temperature ( Figure 2). The results obtained by Gil-Rodríguez et al. (2015) showed autoaggregation for yeasts isolated from foods at 1.1 to 85.8% at 2 h, and from 83.3 to 99.8% at 24 h, the results suggesting that autoaggregation percentages are strongly strain dependent.

Conclusion
Probiotics play a critical and vital role in human nutrition. On recent years, new strain isolation, characterization and verification of potential health benefits particularly related with probiotic traits have been a very attractive area for researchers. Probiotic properties are strain specific, therefore new strains must be well characterized. On this study, good survival in gastric juice, high percentages of autoaggregation/coaggregation and production of EPS were determined for some of the natural isolates investigated. On this study, natural S. cerevisiae strains showed high as EPS production as much as commercial isolates and had good features in terms of autoaggregation and coaggregation. Pichia kluyveri YB1-1A and Pichia kluyveri T8-1C were aggregated in the first hour of incubation period. Even though the strains were not able to produce EPS as much as S. cerevisiae, they produced EPS between 100 to 150 mg/L. The results indicated that, among the isolated strains, S. cerevisiae T8-3C and P25-1, Pichia kluyveri YB1-1A and T8-1C showed notable potential probiotic properties. These yeasts were selected to be used in the present study since they have been verified to be reliable for food industry and biotechnological applications. This research represents a study of probiotic yeast selection; to declare them  as effective probiotics, these strains would undoubtedly require further studies and animal trials. Food and pharmacy industries may benefit from these strains as new food supplements or pharmaceutical preparations.