Comprehensive in vitro probiotic screening and ITS-based molecular identification and characterization of yeast strains isolated from agro-industrial and rumen sources Authors: Fereshteh Falah, Negin Ghazanfari, Atieh Rahimi, Damoon Ghofrani Tabari, Morteza Kashaninejad, Yasamin Yousefi, Behrooz Alizadeh Behbahani, Alireza Vasiee Published: 2026 ABSTRACT BackgroundProbiotic yeasts, particularly Saccharomyces cerevisiae, are receiving increasing attention for their health-promoting effects. However, identifying safe and efficient strains from diverse and underexplored ecological niches remains challenging. This study addressed this gap by isolating and characterizing yeast strains from agro-industrial byproducts and rumen-derived sources, followed by a comprehensive evaluation of their probiotic attributes and safety profiles.MethodsFifty-four yeast isolates were subjected to a multi-step in vitro screening workflow, including morphological assessment, safety evaluation (hemolytic, gelatinase, and DNase activities), gastrointestinal tolerance tests, cell surface characterization (hydrophobicity, auto-aggregation, co-aggregation), antimicrobial activity, antioxidant capacity, and antibiotic susceptibility. Molecular identification of selected strains was performed using ITS sequencing and phylogenetic analysis.ResultsA total of ten strains passed the primary safety and functional criteria. Among them, four top-performing isolates showed exceptional probiotic potential, exhibiting gastrointestinal survival rates > 80%, cell surface hydrophobicity > 80%, and co-aggregation with Escherichia coli ATCC 25922 > 70%. These isolates also produced antimicrobial inhibition zones > 12 mm and demonstrated DPPH radical scavenging activity > 80%. After excluding three antibiotic-resistant strains, nine final candidates were confirmed by PCR, comprising four S. cerevisiae and five Pichia fermentans strains.ConclusionsYeasts isolated from non-conventional agro-industrial and rumen-derived sources demonstrated strong in vitro probiotic properties and fulfilled essential safety requirements. These findings highlight their potential application as functional ingredients in food and feed formulations. Nevertheless, in vivo studies are required to validate their efficacy and safety in host systems. Background Probiotics are defined as live microorganisms that, when administered in adequate amounts, confer health benefits to the host, primarily through modulation of the gastrointestinal tract microbiota [1]. The majority of research concerning probiotic applications in food systems has traditionally focused on the most extensively characterized probiotic bacteria, belonging to the genera Bifidobacterium, Lacticaseibacillus, and Lactiplantibacillus [2, 3]. In recent decades, however, the potential of yeast strains as probiotics has received increasing attention, particularly due to their prospective roles in promoting gut health. Despite the continued predominance of bacterial probiotics, yeasts have emerged as promising candidates owing to their eukaryotic nature and considerably larger cell size—approximately an order of magnitude greater than that of bacteria. This substantial size advantage enables yeasts to function as a steric barrier against pathogenic microorganisms, thereby contributing to their probiotic efficacy [4–6]. Probiotic yeasts have recently emerged as a promising avenue in the development of next-generation probiotics. Several yeast strains, including Kluyveromyces marxianus, Kluyveromyces lactis, Pichia kluyveri, Pichia pastoris, Candida famata, Candida tropicalis, and Saccharomyces boulardii, have been proposed as potential probiotic candidates in multiple studies [4, 7, 8]. Compared with lactic acid bacteria (LAB), probiotic yeasts present several notable advantages: (1) enhanced resistance to harsh gastrointestinal conditions, facilitating improved survival through the stomach and efficient colonization of the intestinal tract, and (2) a broader functional repertoire, including the production of bioactive compounds that promote the growth of beneficial microbial populations [5]. Probiotics are widely recognized for their ability to withstand acidic environments, inhibit intestinal bacterial pathogens, exert anti-inflammatory effects, modulate and stimulate the immune system, maintain and restore intestinal epithelial integrity, and enhance nutrient bioavailability [9, 10]. Among all yeast genera, only a limited number of S. cerevisiae strains are commercially available as probiotics for human consumption [11]. Probiotic yeasts confer substantial health benefits, including the prevention of allergic reactions and the enhancement of immune system function. They interact directly with immune cells, such as macrophages and dendritic cells, thereby stimulating the production of cytokines, including interleukin-10 and interferon-gamma [12–14]. In addition, S. cerevisiae strains can modulate the gut microbiota by inhibiting pathogenic organisms, supporting the growth of beneficial microbes, and alleviating gastrointestinal disorders such as diarrhea, constipation, and inflammation [15–17]. These yeasts further contribute to improved digestion and enhanced nutrient absorption [18, 19]. In livestock production, probiotic yeasts play a pivotal role in enhancing animal health and nutrition. Historically employed as feed additives, yeasts provide fiber, proteins, B vitamins, minerals, and organic acids [20]. Both viable and non-viable yeast preparations have been associated with improved feed efficiency, weight gain, and overall animal performance [21]. Reports indicate that S. cerevisiae has significantly influenced poultry health, leading to reduced plasma cholesterol levels, improved feed intake, and increased egg production [22, 23]. Moreover, it has been incorporated into ruminant diets to mitigate lactate accumulation and prevent rumen acidosis [24]. The inclusion of S. cerevisiae cell walls in animal feed has also been shown to reduce aflatoxin B1 contamination and protect against infectious diseases [25, 26]. Collectively, these effects underscore the critical benefits of probiotic yeasts in promoting livestock growth and health, preventing and managing diseases, and reducing the reliance on antibiotics in animal production systems. Although S. cerevisiae is a well-established probiotic, most research has concentrated on strains from conventional commercial sources. Consequently, a significant knowledge gap exists in the bioprospecting of probiotic yeasts from unique and underexplored ecological niches. In particular, rumen-derived and agro-industrial environments in Iran represent a largely untapped reservoir of potentially robust and novel strains that are uniquely adapted to local substrates and conditions. Therefore, the objectives of this study were: (1) to isolate and screen yeast strains from these specific Iranian sources, (2) to systematically evaluate their probiotic properties and safety profiles in vitro, and (3) to molecularly identify the most promising candidates. We hypothesized that these non-conventional, locally derived sources would yield novel strains exhibiting superior probiotic characteristics. Materials and methods Sample collection A diverse array of food and animal samples was collected from Mashhad, Iran, to isolate potential probiotic yeasts. The samples were classified into several categories: (1) fermented fruit byproducts, including apple, carrot, and mulberry pomace; (2) grain-based sources, such as traditional bread and fermented grains; (3) dairy-based fermentations, including kefir and kombucha mixtures; and (4) ruminant-associated sources, such as corn silage and rumen fluids. A detailed description of each sample is provided in Table 1. Two rumen-fistulated Holstein dairy cows from the Ferdowsi University of Mashhad dairy farm (Mashhad, Iran), with an average milk yield of 32 ± 2.5 kg/day and approximately 150 days in milk, were housed in individual pens and fed a total mixed ration formulated according to NRC (2001) recommendations. The diet had a forage-to-concentrate ratio of 40:60 and was designed to meet the nutrient requirements of mid-lactation, potentially pregnant cows producing approximately 32 kg of milk per day. The ration contained roughly 17.8% crude protein (CP) and 30.5% neutral detergent fiber (NDF) on a dry matter basis. Cows were fed twice daily at 07:00 a.m. and 04:00 p.m., with ad libitum access to clean water and mineral blocks. Following a 7-day adaptation period, rumen fluid samples were individually collected from each cow via the rumen cannula, 4 h after the morning feeding. Each sample was filtered through four layers of cheesecloth, transferred into pre-warmed bottles, immediately placed in an ice bucket at 4 °C, and transported to the laboratory within 15 min for subsequent isolation and analysis of S. cerevisiae. Table 1Description of sources, origins, assigned codes, and colony morphology of the 54 yeast strains isolated for probiotic screeningSourcesOriginCollected fromIsolated StrainsColony morphologyFermented apple pomace (AP)FruitFermentation, Fruit and vegetable market, Mashhad, Iran2Medium size, smooth and cream-coloredFermented carrot pomace (CP)4Small to Medium size, round and butteryFermented mulberry pomace (BP)2Large, Flat and cream- opaque coloredTraditional bread (YB)GrainFermentation, Traditional bakery, Mashhad, Iran4Small to Medium size, round and butteryFermented grains (SC)8Small, smooth and bread-like odorLocal bread yeast (YM)2Medium, Flat and cream- coloredkombucha mix (KN)DairyFermentation, Traditional dairy products, Mashhad, Iran4Small to Medium size, round and butteryCrystal kefir mix (CK)2Medium size, smooth and cream-coloredKefir and milk mix (MK)6Large, Irregular and cream-coloredRumen fluid 1 (R1)Feed StockCorn silage, Rumen fluids6Medium, regular edge and cream-coloredRumen fluid 2 (R2)8Small, smooth and creamyCorn silage (SR)6Medium, regular edge and cream-colored Table 1Description of sources, origins, assigned codes, and colony morphology of the 54 yeast strains isolated for probiotic screeningSourcesOriginCollected fromIsolated StrainsColony morphologyFermented apple pomace (AP)FruitFermentation, Fruit and vegetable market, Mashhad, Iran2Medium size, smooth and cream-coloredFermented carrot pomace (CP)4Small to Medium size, round and butteryFermented mulberry pomace (BP)2Large, Flat and cream- opaque coloredTraditional bread (YB)GrainFermentation, Traditional bakery, Mashhad, Iran4Small to Medium size, round and butteryFermented grains (SC)8Small, smooth and bread-like odorLocal bread yeast (YM)2Medium, Flat and cream- coloredkombucha mix (KN)DairyFermentation, Traditional dairy products, Mashhad, Iran4Small to Medium size, round and butteryCrystal kefir mix (CK)2Medium size, smooth and cream-coloredKefir and milk mix (MK)6Large, Irregular and cream-coloredRumen fluid 1 (R1)Feed StockCorn silage, Rumen fluids6Medium, regular edge and cream-coloredRumen fluid 2 (R2)8Small, smooth and creamyCorn silage (SR)6Medium, regular edge and cream-colored Description of sources, origins, assigned codes, and colony morphology of the 54 yeast strains isolated for probiotic screening Isolation of strains After fully homogenizing 1 g of each sample with 0.85% sodium chloride, a serial dilution was performed. Following this, 100 µL of the diluted samples were inoculated onto the Yeast Glucose Chloramphenicol (YGC) agar culture medium, and the plates were incubated at 29 °C for 24–48 h. To purify the yeast-like isolates, linear streaking was performed across the surface of the YGC agar after the isolates had been seeded into the YGC broth culture medium. Ultimately, 54 isolated strains were obtained for further research and stored as stock at -80 °C for future usage [27]. Evaluation of isolated yeast strains’ safety Hemolytic activity Hemolytic activity was assessed by inoculating 10 µL of yeast cell culture onto the surface of Tryptone Soy Agar (TSA) 10% enriched with 5% (v/v) sheep blood fibrin. A positive reaction is indicated by the presence of lytic or clear zones surrounding the colonies after 48 h of incubation at 37 °C. Hemolytic strains were discarded, as they are not considered potential probiotic yeasts [28, 29]. Gelatinase activity The synthesis of gelatinase by isolated yeast strains was examined using Tryptone-Neopeptone-Dextrose (TND) agar media supplemented with 0.4% gelatin. To perform the assessment, 10 µL of the yeast culture was inoculated onto the TND agar media, and the plates were incubated at 37 °C for 48 h. After incubation, a saturated ammonium sulfate solution was added to the plates. A positive response is indicated by the formation of clear zones surrounding the colonies [30]. Deoxyribonuclease (DNase) activity The surface of DNase Test Agar medium is inoculated with a microbial suspension, and the plates are incubated at 37 °C for 48 h. After incubation, a 1 N hydrochloric acid (HCl) solution is applied to the plates. This results in the formation of clear zones surrounding colonies that produce DNase [31]. Evaluation of probiotic potential of yeast isolates The control strain, S. cerevisiae PTCC 5052, obtained from the Industrial Microorganism Collection Center of Iran was used as the positive control in all probiotic tests described below. Measuring the survivability of isolated yeasts in Gastrointestinal simulations A buffer solution consisting of 2.5 g/L NaCl, 0.6 g/L KH2PO4, 0.11 g/L CaCl2, and 0.37 g/L KCl was prepared to simulate gastric juice. This solution was autoclaved at 121 °C for 15 min after adjusting the pH to 2 with a 1 M HCl solution. The resulting solution was then mixed with 0.01 g/L lysozyme and 0.133 g/L pepsin. The next step involved centrifuging the activated yeast cells from the YGC broth culture medium at 10,000 × g for 10 min at 4 °C. The resulting pellet was washed using the previously prepared buffer (pH 2), but without the enzymes. Subsequently, the cells were resuspended in the gastric juice solution until the optical density (OD) at 600 nm reached 0.5. The sample was then incubated at 37 °C with shaking at 200 rpm for 3 h to simulate peristaltic movements. For pancreatic digestion, pancreatic juice was prepared using 0.3 g/L bile salts, 0.1 g/L pancreatin, 26.9 g/L Na2HPO4, and 8.5 g/L NaCl, with the pH adjusted to 8.5. After two washes with sterile Phosphate Buffered Saline (PBS) and resuspension in the simulated pancreatic solution, the cells extracted from the gastric digestion phase were centrifuged at 5,000 × g for 10 min at 4 °C. Cell viability was assessed at the beginning and end of each gastrointestinal simulation by culturing on YGC agar. The viability was determined using the following equation after shaking the suspension at 200 rpm for 3.5 h at 37 °C. The commercial probiotic yeast S. cerevisiae PTCC 5052 was used as the positive control, while sterile PBS (without cells) served as the negative control [3]: 1\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\mathrm{Survival}\;\mathrm{Rate}\;(\%)=(\mathrm{Log}10\;\mathrm{Nt}\;/\;\mathrm{Log}10\;\mathrm N0)\;\times100$$\end{document} where N0 is the number of viable yeast cells in the original culture suspension, and Nt is the number of viable yeast cells during the gastric digestion simulation. Enzymatic potential of isolated strains To assess amylolytic activity, 2 g of agar and 1 g of starch were added to a nutrient broth medium. After inoculating the isolates onto the plates, they were incubated for 48 h. Following incubation, iodine crystals were applied to the plates. A positive result is indicated by the presence of light zones surrounding the inoculation point, while a negative result is characterized by the absence of a halo. For the evaluation of cellulolytic activity, a culture medium was prepared by dissolving 0.9 g of nutrient broth, 2 g of agar, and 1 g of carboxymethyl cellulose (CMC) in 100 mL of distilled water. After inoculation, the plates were incubated for 48 h and then stained first with Congo red dye and subsequently with NaCl. The presence of a clear zone surrounding the inoculated colony indicates a positive result, whereas its absence signifies a negative outcome. Proteolytic activity was measured using a 1% casein agar culture medium. Plates were spot-inoculated and incubated for 48 h. After incubation, the plates were immersed in frozen 1% acetic acid. A favorable result is indicated by the formation of a clear zone, while a negative result is indicated by its absence [32]. For urease activity analysis, the strains were cultivated in a medium adjusted to pH 7, containing 0.01 g of phenol, 20 g/L urea, 9.5 g/L Na2HPO4, and 9.1 g/L KH2PO4. A positive urease reaction converts urea into ammonia and carbon dioxide, resulting in the alkalization of the medium and a pink hue development. All yeast strains were incubated in the urea-containing medium for 24 h at 37 °C [33]. Hydrophobicity After being cultivated in YGC broth for 24 h at 29 °C, the yeasts were centrifuged at 4,000 × g for 15 min at 4 °C in order to assess the isolates’ affinity for hydrocarbons. A spectrophotometer (OD1) was used to measure the absorbance at a wavelength of 600 nm after the produced pellet had been cleaned with sterile PBS solution and its OD had been adjusted to 0.6. Subsequently, 1 mL of chloroform was combined with 3 mL of the cell suspension, and the mixture was incubated for 1 h at 37 °C. After carefully removing the aqueous phase, the absorbance at 600 nm was measured (OD2). Using the following formula, the percentage of surface hydrophobicity was determined. Cell surface hydrophobicity of the isolates was compared with the probiotic control strain S. cerevisiae PTCC 5052, and PBS served as the negative control [31, 34]: 2\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\mathrm{Hydrophobicity}\;\%= (A0 - A1 /A0) \times 100$$\end{document} Auto-aggregation The fresh yeast culture was first centrifuged for 15 min at 4 °C at 3,000 × g. A spectrophotometer (A0) was used to measure the absorbance at a wavelength of 600 nm after the produced pellet had been cleaned twice with sterile PBS solution and its OD had been adjusted to 0.6. The following step involved vortexed 4 mL of the produced suspension well and incubating it for 4 h at 37 °C. Following incubation, the suspension’s absorbance at 600 nm was determined (A1). The following formula will be used to calculate the yeast cells’ rate of auto-aggregation. Auto-aggregation ability was benchmarked against the positive control strain S. cerevisiae PTCC 5052. PBS without cells served as the negative control [35, 36]: 3\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\mathrm{Auto}-\mathrm{aggregation}\;\%= 1 - [(A0 /A1)] \times 100$$\end{document} Co-aggregation It was investigated whether the yeast isolates could co-aggregate with the pathogen E. coli ATCC 25922. The fresh yeast culture was first centrifuged at 3,000 × g for 15 min at 4 °C to perform this test. After adjusting the OD of the produced pellet to 0.6 and washing it twice with sterile PBS solution, a spectrophotometer was used to measure the absorbance at 600 nm. Following this, the E. coli ATCC 25922, cultivated in Mueller Hinton Broth (MHB) for 24 h at 37 °C, were suspended in PBS at an OD of 600 nm = 0.7 and vortexed for 10 s. They were then rinsed twice with PBS. Equal amounts (2 mL) of the bacterial and yeast cultures were mixed and vortexed for 10 s. The absorbance of the samples was measured at a wavelength of 600 nm. For control groups, 2 mL of the strain under study were added. The microbial suspensions were cultured for 4 h at 37 °C, and a spectrophotometer was used to measure the absorbance at 600 nm once more. Ultimately, the following formula was employed to determine the concentration of yeast cell co-aggregation. The probiotic reference strain S. cerevisiae PTCC 5052 was used as the positive control, while pathogen–PBS mixtures without yeast cells served as the negative control [31]: 4\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\mathrm{Co}-\mathrm{aggregation}\;(\%)=\lbrack(\mathrm{Ax}+\mathrm{Ay})/2-\mathrm A\;(\mathrm x+\mathrm y)\rbrack/\lbrack(\mathrm{Ax}+\mathrm{Ay})/2\rbrack\times100$$\end{document} where x and y stand for bacterial and yeast cultures, x + y for the mixed culture, and A for the absorption rate. Isolates containing pathogenic bacteria that had a biomass concentration below 5% were thrown out. Measurement of the antimicrobial activity of isolated strains The antibacterial activity of the isolated strains against E. coli ATCC 25922 was determined using the well diffusion assay (WDA). For this purpose, yeast cells were cultivated in YGC broth. Following cultivation, the yeast culture was centrifuged at 7,000 × g for 5 min at 4 °C to separate the cells from the supernatant. A volume of 50 µL of the supernatant was then placed into wells that were created on the surface of the Brain Heart Infusion (BHI) agar growth medium. Prior to adding the supernatant, the agar had been inoculated with 0.1 mL of the pathogenic bacterium E. coli ATCC 25922 to establish a uniform bacterial lawn. After the supernatant was added to the wells, the plates were incubated for 24 h at 37 °C. The antimicrobial activity of the isolated yeast strains was assessed based on the diameter of the inhibitory zone that developed around each well. The results were measured in millimeters, and a larger zone of inhibition indicated a stronger antibacterial effect of the yeast supernatant against E. coli ATCC 25922. The explicit diameters of the zones were recorded and analyzed to compare the antimicrobial efficacy of different yeast isolates. For benchmarking antimicrobial activity, S. cerevisiae PTCC 5052 was used as the positive control, while uninoculated YGC broth served as the negative control [28]. Effect of Yeast Strain R1b on E. coli ATCC 25922 Morphology Using Scanning Electron Microscopy (SEM) To investigate the effect of yeast strain R1b on the morphology of E. coli ATCC 25922, co-culture experiments were performed. E. coli ATCC 25922 was grown in sterile nutrient broth at 37 °C until reaching the logarithmic growth phase. The yeast strain R1b was then added at a defined ratio (1:1 based on OD600 or Colony Forming Unit (CFU)/mL), and the culture was incubated for an additional 24 h at 37 °C under aerobic conditions. After incubation, samples were harvested and fixed with 2.5% (v/v) glutaraldehyde in PBS (pH 7.2) for 2 h at 4 °C. The fixed samples were washed three times with PBS and post-fixed with 1% osmium tetroxide (OsO₄) for 1 h. Dehydration was carried out through a graded ethanol series (30%, 50%, 70%, 90%, and 100%) for 10 min at each step. The dehydrated samples were subjected to critical-point drying to preserve structural integrity and then sputter-coated with a thin layer of gold–palladium to enhance conductivity. Finally, the prepared specimens were examined using a scanning electron microscope (Leo, Germany) at different magnifications to evaluate morphological alterations in E. coli ATCC 25922 cells induced by the yeast strain R1b [37]. Investigation of antibiotic resistance in isolated strains To determine the antibiotic resistance of the isolated yeast strains, the disk diffusion method was employed. This method involves inoculating the surface of YGC agar with a 1% active yeast suspension to create a uniform yeast lawn. After the agar had been inoculated, antibiotic disks were carefully placed on the surface. The antibiotics used in this investigation included: Erythromycin (15 µg/disc), Ampicillin (10 µg/disc), Kanamycin (30 µg/disc), Gentamicin (10 µg/disc), Tetracycline (30 µg/disc). The inoculated plates were incubated for 48 h at 37°C. After incubation, the inhibition zones surrounding each antibiotic disk were measured in millimeters. Due to the absence of standardized clinical breakpoints (e.g., CLSI) for probiotic yeasts from non-clinical sources, the interpretation of results was performed as follows: a strain was categorized as ‘resistant’ to an antibiotic if it exhibited no zone of inhibition (i.e., a diameter of 6 mm, equal to the disk size), and was categorized as ‘susceptible’ if any clear zone of inhibition (> 6 mm) was observed. This approach allowed for the screening of acquired resistance profiles among the isolates. To validate the antibiotic susceptibility assay, a known antibiotic-resistant reference strain was included as the positive control, while a sensitive strain and medium without cells served as negative controls. Zones of inhibition produced by the isolates were compared with those of the controls to ensure assay accuracy [38]. Measurement of the antioxidant activity of isolated strains The antioxidant activity of the yeast isolates will be assessed using the 1,1-diphenyl-2-picrylhydrazyl (DPPH) assay. To prepare for the assay, the yeast cultures were centrifuged for 10 min at 4 °C at 3,000 × g. The resulting supernatants were then transferred to fresh tubes and sterilized using a 0.22 μm filter. For the assay, 1 mL of a 0.2 mM DPPH solution in methanol was mixed with 800 µL of each yeast supernatant. After thoroughly vortexed the mixtures for 2 min, they were allowed to sit at room temperature in the dark for 30 min to enable the reaction to occur. Following the incubation period, the absorbance of the mixtures was measured at a wavelength of 517 nm using a spectrophotometer. The radical-scavenging capacity was determined using the following formula (DPPH solution mixed with sterile broth served as the negative control) [38]: 5\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\mathrm{Radical}\;\mathrm{scavenging}\;(\%)\;=1-\lbrack{\mathrm A}_{(\mathrm{Sample})}/{\mathrm A}_{(\mathrm{Blank})}\rbrack\;\times\;100$$\end{document} Identification of isolated strains To identify the isolated yeast strains, Polymerase Chain Reaction (PCR) amplification of the ITS region was performed. DNA was extracted using the Dena Zist Asia extraction kit (Mashhad, Iran). The universal fungal primers ITS1 (TCCGTAGGTGAACCTGCGG) and ITS4 (TCCTCCGCTTATTGATATGC) were used to amplify the ITS1–5.8 S–ITS2 region. Each 25 µL PCR reaction mixture contained 2 µL of extracted DNA, 2 µL of dNTPs, 1 µL of each primer, 16.3 µL of sterile distilled water, 0.2 µL of Taq DNA polymerase, and 2.5 µL of reaction buffer. PCR cycling conditions were as follows: initial denaturation at 95 °C for 5 min; 35 cycles of denaturation at 95 °C for 30 s, annealing at 60 °C for 30 s, and extension at 72 °C for 30 s; followed by a final extension at 72 °C for 5 min. PCR products were analyzed by agarose gel electrophoresis using a 0.9% agarose gel prepared in 0.5× TBE buffer, supplemented with 1 µL SYBR Safe dye. Wells were loaded with 5 µL of PCR product mixed with loading dye, alongside a Kb1 DNA ladder (Sinaclon, Iran). A negative control (no-template reaction) was included. Electrophoresis was performed at 95 V for 30 min, and bands were visualized using a gel documentation system. PCR products displaying the expected ~ 550–800 bp band were purified and submitted for Sanger sequencing (Kian Danesh Varzan Novin, Tehran, Iran). The obtained ITS sequences were quality-checked and analyzed using the NCBI BLASTn algorithm against the nucleotide database. Species identification was assigned based on ≥ 98% sequence identity and ≥ 97% query coverage relative to reference sequences, which are commonly accepted thresholds for reliable yeast species discrimination [39]. To confirm species-level identification and to visualize the genetic relatedness among the selected yeast isolates, a phylogenetic analysis was performed based on the ITS1–5.8S–ITS2 region. ITS sequences were aligned using the ClustalW algorithm, and a Neighbor-Joining phylogenetic tree was constructed using standard distance-based methods. The phylogenetic analysis was performed as a supportive and illustrative approach to demonstrate species-level clustering of the isolates rather than to infer detailed evolutionary relationships. Due to the conserved nature of the ITS region at the species level and the limited number of closely related taxa included in this study, bootstrap analysis was not emphasized [27]. Statistical analysis Statistical analysis was conducted using SPSS software version 20 to calculate the mean and standard deviation of the data. Graphical representations were created using Excel and Prism software. To assess the presence or absence of statistically significant differences, a one-way ANOVA (with Duncan’s post-hoc test) was performed. A p-value of less than 0.05 was considered statistically significant. All experiments were performed in triplicate. Results and discussion Screening of isolated strains based on morphology and staining To identify the morphological features of the 54 isolates obtained from various sources, an initial screening test was conducted using culture methods and morphological analysis (Table 2). The colonies were subsequently stained with lactophenol cotton blue. The results indicated that 14 (MK2, MK22, SC1, SC11, SR2, SR22, R2cc, R2aa, R2a, R2c, R1aa, R1cc, R1c and R1a) of the 54 isolates were disqualified from further testing as they did not meet the necessary staining and morphological criteria. Their staining properties (cream to white) and colony appearance (smooth to velvety and spherical) were not significantly different from those of S. cerevisiae colonies (p < 0.05). The decision to eliminate these isolates was primarily based on characteristics such as colony size, shape, color, and microscopic appearance. The primary aim of this preliminary screening was to minimize errors and ensure the accurate differentiation of S. cerevisiae strains for further investigation. Consequently, only the strains that exhibited the required morphological traits and staining characteristics were advanced to the subsequent phases of the study for a more comprehensive evaluation of their probiotic qualities. Table 2Identification codes of the 54 initial microbial isolates obtained from various agro-industrial and animal-derived sources prior to morphological and probiotic screeningAPYB1YM1MK1CP22R1cBPR1bbYB11SC3CP2KN1MK11KN11KN2R2bMK2SC44CP1R1bKN22R2aMK22AP1CP11SC4R1aaR1aSC11CK1SC34R2dR1ccSR22SR33SC22SR11R2ddSC2R2cYMMK3R2ccSC1SR2SR1MK33YB22SR3BP1YB2CKR2bbSC33 Table 2Identification codes of the 54 initial microbial isolates obtained from various agro-industrial and animal-derived sources prior to morphological and probiotic screeningAPYB1YM1MK1CP22R1cBPR1bbYB11SC3CP2KN1MK11KN11KN2R2bMK2SC44CP1R1bKN22R2aMK22AP1CP11SC4R1aaR1aSC11CK1SC34R2dR1ccSR22SR33SC22SR11R2ddSC2R2cYMMK3R2ccSC1SR2SR1MK33YB22SR3BP1YB2CKR2bbSC33 Identification codes of the 54 initial microbial isolates obtained from various agro-industrial and animal-derived sources prior to morphological and probiotic screening Results of the safety evaluation for the isolated strains A series of safety assessments were conducted to evaluate the profiles of the 40 remaining strains. These included assays for hemolytic, gelatinase, and DNase activities. Hemolytic activity is a primary safety concern and a key screening criterion for the selection of newly proposed probiotic strains. Only non-hemolytic strains (γ-hemolysis) are considered suitable for further studies, as they are deemed safer for human and animal consumption. In contrast, β-hemolysis is closely associated with pathogenicity and the destruction of red blood cells, which can lead to significant adverse health effects [40]. Gelatinase is a type of protease that may enhance specific fermentation processes within the food industry and support bacterial survival and adaptation by facilitating the metabolism of gelatin and other proteins. However, high gelatinase activity in probiotic strains can be problematic, as it may lead to degradation of the host’s extracellular matrix and mucosal barriers. Such degradation can result in tissue damage and increased susceptibility to infection [41]. DNase activity, mediated by the enzyme deoxyribonuclease, also plays a crucial role in assessing strain safety. There are two main types of DNase: DNase I, which functions optimally at neutral pH (6.5–8.0) and requires divalent ions like calcium (Ca²⁺) and magnesium (Mg²⁺); and DNase II, which is active under acidic conditions and does not depend on divalent ions. Because of its role in cellular degradation, excessive DNase activity can contribute to pathological conditions, including autoimmune diseases such as systemic lupus erythematosus. The regulation of DNase activity—often through natural, conventional, or chemical inhibitors—is therefore essential in mitigating such risks [42]. Together, hemolytic, gelatinase, and DNase activities serve as widely accepted markers of potential tissue and cellular damage. Strains exhibiting any of these traits may be classified as potentially pathogenic or hazardous [43]. Following this safety screening, hemolytic, gelatinase, and DNase activities were quantified by measuring the diameter of the lytic or hydrolysis zones. Hemolytic reactions on blood agar ranged from 0 mm (γ-hemolysis) to 4–6 mm clear zones (β-hemolysis). Gelatinase-positive strains exhibited hydrolysis zones of 3–5 mm, while DNase-producing strains showed clear zones of 2–4 mm. Based on the safety evaluation (hemolysis, gelatinase, and DNase assays), 16 strains were excluded due to measurable lytic or hydrolytic activities. These strains exhibited clear zones within the predefined exclusion thresholds and were therefore removed from subsequent probiotic testing. Only strains that showed complete absence of all three enzymatic activities progressed to the next phase. Supporting these findings, Fernández-Pacheco et al. (2021) evaluated 20 yeast strains with probiotic potential and reported that none exhibited DNase, hemolytic, or gelatinase activity, underscoring their suitability for use in the food industry [44]. Similarly, Dessalegn and Andualem (2023) investigated lactic acid bacteria and yeasts isolated from Ersho samples (a traditional Ethiopian fermented beverage), finding that strains with hemolytic and gelatinase activity, as well as those producing biogenic amines, had reduced probiotic suitability due to their adverse biological effects [45]. Results of the evaluation of probiotic potential of isolated strains Simulation of gastric and intestinal conditions (Gastrointestinal System) A simulation test designed to replicate gastric and intestinal conditions was conducted as part of the probiotic evaluation process. This test aimed to assess the resilience of the surviving yeast strains to environmental factors mirroring those encountered in the gastrointestinal system, such as gastric acidity and varying intestinal conditions. The purpose of this test was to evaluate the yeasts’ ability to endure and function within the harsh environments of the stomach and intestines. Since gastric acid and bile salts in the gut represent the primary biological barriers that a probiotic strain must navigate after ingestion to reach its target site, assessing resistance to low pH and bile exposure is a crucial criterion in strain selection [46]. The survival percentages of the isolated strains under simulated gastrointestinal conditions varied significantly (p < 0.05). The data revealed that survival rates ranged from 20% (the lowest) to over 90% (the highest) among the tested strains (Fig. 1). Out of the 24 strains evaluated, 4 strains (SR11, YB22, SC22 and R2bb) exhibited survival rates below 40% and were unable to withstand the simulated gastric and intestinal environments. Due to a marked decrease in survival in both the alkaline intestinal and acidic gastric conditions, these strains were excluded from further studies. These results indicate that these strains lacked the capacity to endure the challenging conditions of the gastrointestinal tract and sustain probiotic functionality during transit from the stomach to the intestines. The ability of a probiotic to survive both the acidic environment of the stomach and the alkaline conditions of the intestines is essential, making this screening step vital. Such attributes facilitate the successful colonization of yeasts within the intestines and their effective passage through the gastrointestinal tract. Additionally, this assay aided in identifying strains with limited potential for probiotic activity in actual gastrointestinal tract. Therefore, it was crucial to eliminate these four strains to ensure the selection of the most appropriate candidates for subsequent phases of research and a more precise evaluation of their probiotic characteristics. The significance of gastrointestinal simulation experiments for the preliminary screening and selection of promising probiotic strains is clearly illustrated in this phase. Fig. 1Survival rates of 24 isolated yeast strains and a control strain under Gastrointestinal system conditions. The chart displays the mean percentage of survival following sequential exposure to simulated gastric fluid (green bars) and simulated intestinal fluid (blue bars). Values represent mean ± SD (n = 3) Survival rates of 24 isolated yeast strains and a control strain under Gastrointestinal system conditions. The chart displays the mean percentage of survival following sequential exposure to simulated gastric fluid (green bars) and simulated intestinal fluid (blue bars). Values represent mean ± SD (n = 3) The observed variability in survival rates among the isolates, particularly the robust performance of rumen-derived strains, can be largely attributed to strain-specific differences in cell wall architecture. The yeast cell wall, primarily composed of β-glucans, mannoproteins, and chitin, acts as a physical barrier against environmental stresses [47]. It is hypothesized that strains isolated from harsh environments like the rumen have evolved thicker or more compact cell walls with higher glucan-to-mannan ratios. This structural adaptation likely prevents the permeation of protons and the emulsifying action of bile salts, thereby maintaining cellular homeostasis and ensuring viability during gastrointestinal transit [5]. Furthermore, the survival rates observed in our yeast isolates are comparable to, and in some cases exceed, those reported for probiotic lactic acid bacteria isolated from Iranian traditional sourdough and dairy products [48]. This suggests that indigenous yeast strains from rumen and agro-industrial sources possess robustness traits similar to established bacterial probiotics from regional fermented foods. In a 2019 study by Khisti et al., the probiotic properties of yeast cells isolated from larval feces were examined. The results indicated that all isolated cultures displayed robust tolerance to bile at a concentration of 1.2%, and these strains demonstrated the capability to grow across a pH range of 1.5 to 10 [49]. Twelve isolates with high survival rates under in vitro digestion conditions were selected from a study by Alkalbani et al. (2022), which investigated the survival of 105 yeast isolates from fermented dairy and non-dairy products in a laboratory setting. The findings indicated that the growth of all yeast isolates declined under simulated digestion conditions [50]. Cell surface Hydrophobicity, aggregation and enzymatic activity Figure 2 presents the results of three key assays designed to evaluate the probiotic potential of the remaining yeast strains: hydrophobicity, auto-aggregation, and co-aggregation. These tests are essential for screening strains that possess the ability to colonize the gastrointestinal tract and interact with host and microbial components. The cell surface hydrophobicity assay assessed each strain’s ability to adhere to intestinal epithelial cells and interact with hydrophobic environments, a crucial trait for colonization and microbial competition in the gut. Hydrophobicity is strongly associated with a strain’s capacity to persist in the intestinal environment and form stable associations with host tissues and other members of the microbiota [34, 51]. Given the much larger size of yeast cells—approximately ten times larger than bacterial cells—they require more surface area for adhesion, which may influence their overall adhesive behavior [50]. Fig. 2Probiotic-related cell surface characteristics of the selected yeast isolates. The bar chart illustrates the percentage of auto-aggregation (blue bars), cell surface hydrophobicity (orange bars), and co-aggregation with the pathogenic indicator strain E. coli ATCC 25922 (grey bars). Values represent mean ± SD (n = 3) Probiotic-related cell surface characteristics of the selected yeast isolates. The bar chart illustrates the percentage of auto-aggregation (blue bars), cell surface hydrophobicity (orange bars), and co-aggregation with the pathogenic indicator strain E. coli ATCC 25922 (grey bars). Values represent mean ± SD (n = 3) Enzymatic profiling included the assessment of proteolytic, cellulolytic, amylolytic, and urease activities. These enzymes play a pivotal role in nutrient degradation and absorption within the host gut. Proteases contribute to the breakdown of proteins into peptides and amino acids, facilitating microbial growth and metabolism [52]. Amylases and cellulases enable the conversion of complex carbohydrates like starch and cellulose into fermentable sugars, serving as energy sources during microbial colonization [53, 54]. Although urease activity was tested in all strains, none displayed positive activity. The universal absence of urease activity across all isolated strains is a significant safety attribute. Urease catalyzes the hydrolysis of urea into ammonia, leading to intestinal alkalization. An elevated gut pH and excessive ammonia accumulation can be toxic to the intestinal epithelium and may compromise the mucosal barrier. Therefore, the urease-negative phenotype of these isolates ensures that they will not induce ammonia toxicity or disrupt the protective acidic environment of the gut, reinforcing their suitability for application in food and feed [55]. Auto-aggregation and co-aggregation assays evaluated the strains’ abilities to aggregate with themselves and with pathogenic microbes, respectively. High auto-aggregation supports prolonged colonization of the gut by promoting microbe-host and microbe-microbe interactions [56]. Co-aggregation, on the other hand, reflects the ability of probiotic strains to competitively inhibit pathogen adherence and create protective microbial barriers [49]. Notably, due to their size and density, yeast cells tend to sediment faster and form aggregates more readily than bacteria [57]. Three of the remaining 20 strains were excluded (CK, YM and KN22) from further study after the initial tests revealed their failure to meet the required experimental standards. The findings indicated that strain R2b displayed the highest hydrophobicity percentage at 90%, while strains CK1, KN2, and KN22 exhibited significantly lower hydrophobicity percentages, each below 30% (p < 0.05). These strains were unable to demonstrate adequate hydrophobicity characteristics, which adversely affected their ability to adhere to intestinal cells and the gut microbiota. Notably, the cell surface hydrophobicity recorded for our top strains (> 80%) aligns with recent findings by Nami et al. [48], who reported hydrophobicity values up to 76.24% for Lactobacillus strains. This high hydrophobicity in our yeast isolates reinforces their potential for strong adhesion to the intestinal epithelium, a critical prerequisite for colonization, although strictly verified here under in vitro conditions. In tests evaluating amylolytic, cellulolytic, and proteolytic activities, these isolates performed poorly. Additionally, the urease test yielded negative results across all strains, indicating their inability to absorb complex substances in the gastrointestinal tract, a crucial requirement for probiotic activity. The CP1, R2b, and AP strains exhibited auto-aggregation levels exceeding 80%, while strains CK1 and KN22 showed the lowest auto-aggregation percentages, roughly around 20% and 30%, respectively. In the co-aggregation assessment, CP1 and R2b strains recorded the highest co-aggregation percentage at 75%, whereas CK1 had the lowest at less than 20%, with KN22 and SR33 showing approximately 20% (p < 0.05). These results suggest that these strains displayed limited potential for interacting with gut microbiota, as well as a diminished capacity to contend with pathogens, evidenced by their inability to effectively coalesce with other microbes or even their own strains. Only those strains that passed these rigorous tests progressed to the subsequent phases of the study. The experiments particularly evaluated the capacity of probiotics to compete, support, and interact with the gut flora. To optimize the research process and select strains with greater potential based on their probiotic qualities, these three strains were removed from consideration. This screening underlines the necessity of thorough and multidimensional testing to ensure the safety and effectiveness of probiotics. Such rigorous evaluation is essential for confirming the ability of the remaining strains to support the gut health and enhance microbial balance in the intestinal environment. In 2022, Alkalbani et al. investigated the auto-aggregation percentage of 12 yeast strains isolated from food sources after 24 h of incubation at 37 °C. They also examined the aggregation percentages in the presence of pathogens such as E. coli O157:H7, Salmonella Typhimurium, Listeria monocytogenes, and Staphylococcus aureus at 3, 6, and 24 h of incubation, as well as the hydrophobicity of these strains against hexadecane, xylene, and octane. During the 3, 6, and 24 h incubation periods, the yeast isolates exhibited significant auto-aggregation with percentages ranging from 37.6% to 66%, 44.5% to 84.0%, and 50.7% to 85.8%, respectively. Notably, the aggregation capacity for S. Typhimurium increased throughout the incubation period for 3 to 24 h at 37 °C. In contrast, the yeast strains demonstrated the highest aggregation rates with L. monocytogenes compared to the other pathogens during the incubation period. Additionally, among the 12 isolates, hexadecane and octane exhibited higher hydrophobicity than xylene [50]. In 2023, Kil et al. isolated 100 yeast strains from rice wine and investigated their adhesion, auto-aggregation, and co-aggregation characteristics. The adhesion test to Caco-2 cells revealed that all strains exhibited adhesion capacities ranging from 4% to 15%. Some strains performed better or comparably to the S. boulardii CNCM I-745 control sample in the co-aggregation test. Most of the isolated yeasts demonstrated strong auto-aggregation traits. Notably, colony formation, which influences host colonization, accounted for 80% of the auto-aggregation observed after a 24 h incubation period. The pathogens S. aureus ATCC 25922, Enterococcus faecalis ATCC 29212, and E. coli K88 were able to co-aggregate with the auto-aggregating strains [58]. The probiotic potential and enzyme profile of the bacterium Pichia kudriavzevii Y33, isolated from traditional mango pickle, were studied by Lata et al. in 2022. The β-galactosidase activity of P. kudriavzevii Y33 was found to be 0.054 U/mg. Additionally, the specific activity of P. kudriavzevii Y33 was 3.64 U/mg, indicating significant amylase activity. The phytase enzyme enhances the nutritional content of food and feed by releasing inorganic phosphorus, with the specific activity of phytase in P. kudriavzevii Y33 recorded at 4.36 U/mg. Furthermore, the lipase specific activity of P. kudriavzevii Y33 was determined to be 0.564 U/mg [54]. Antimicrobial activity of the isolated strains In this phase of the probiotic evaluation process, the antimicrobial properties of the remaining strains were assessed using the WDA to determine their ability to inhibit the growth of E. coli ATCC 25922 (Fig. 3). Probiotic yeasts exert their antimicrobial effects through several key mechanisms, including adhesion to pathogens, modification of the structure and function of intestinal epithelial cells, and the production of specific inhibitory metabolites. Various metabolites such as phenolic compounds, ethanol, organic acids, carbon dioxide, hydrogen peroxide, and mucins are synthesized by different yeast strains [59]. When probiotic microorganisms adhere to pathogens, they can obstruct the pathogens’ access to nutrients, prevent their adhesion to intestinal epithelial cells, and facilitate their expulsion from the gastrointestinal tract through peristalsis [60]. Fig. 3Antimicrobial activity of the selected yeast isolates against the pathogenic indicator E. coli ATCC 25922, as determined by the well diffusion assay. The bars represent the mean diameter of the inhibition zone (mm). Values are mean ± SD (n = 3) Antimicrobial activity of the selected yeast isolates against the pathogenic indicator E. coli ATCC 25922, as determined by the well diffusion assay. The bars represent the mean diameter of the inhibition zone (mm). Values are mean ± SD (n = 3) Antimicrobial activity is a crucial attribute for any probiotic microorganism aimed at preventing and treating enteropathogenic infections within the host [61, 62]. Prior to this study, 17 strains remained from earlier experimental phases. In the current investigation, each strain demonstrated inhibitory activity against foodborne pathogens. According to the results from the WDA, four of the remaining 17 strains (R1bb, SC3, CP11, and SR33) were excluded from further research due to the small size of their inhibitory zones, measuring less than 9 mm, which suggested limited antimicrobial effectiveness. The inability of these four strains to produce potent antimicrobial compounds was evident in their insufficient antibacterial activity against E. coli ATCC 25922. Conversely, strains CP1, SC2, and AP1 exhibited the largest inhibition zones, surpassing 12 mm in diameter. Comparatively, the commercial probiotic control strain, S. cerevisiae PTCC 5052, exhibited a significantly smaller inhibition zone of 10.83 mm under identical conditions (p < 0.05). This quantitative difference demonstrates that the top-performing isolates (e.g., CP1 with 13.11 mm and SC2 with 12.67 mm) possess superior antimicrobial potency compared to the standard reference strain. Following this assessment, only 13 strains remained that demonstrated the capacity to generate strong antimicrobial agents and effectively inhibit E. coli ATCC 25922 growths. Four strains were eliminated during this screening phase due to their inadequate production of antimicrobial compounds, preventing them from advancing in the study. This stringent screening process significantly contributed to our ability to identify strains with considerable probiotic potential and robust antimicrobial properties. Consequently, the WDA underscores the strains’ roles in maintaining gut health and their ability to combat infections effectively. In 2021, Chelliah et al. studied the antibacterial activity and probiotic potential of six food-isolated strains of S. boulardii. They tested strains against various pathogens, including E. coli, S. aureus, E. faecalis, Micrococcus luteus, Klebsiella pneumoniae, Salmonella typhi, Salmonella paratyphi A, Salmonella paratyphi B, Proteus mirabilis, Vibrio cholerae, Shigella flexneri, Shigella dysenteriae, and Pseudomonas aeruginosa. Although all strains exhibited considerable sensitivity to various pathogens, they showed the least sensitivity against S. paratyphi A and P. aeruginosa. This reinforces the importance of identifying effective probiotic strains that can withstand pathogenic challenges and support gastrointestinal health [63]. Results of SEM SEM revealed clear morphological alterations in E. coli ATCC 25922 cells after co-cultivation with yeast strain R1b. In contrast to the smooth and intact rod-shaped morphology, treated bacteria exhibited irregular surfaces, deformation, and partial collapse of cell walls (Fig. 4). Quantitative analysis of randomly selected fields of view indicated that approximately 70% of the E. coli ATCC 25922 population exhibited significant morphological deformations, such as cell collapse and wrinkling, compared to the intact cells observed in the control group. Many cells showed wrinkling and elongation, while others appeared aggregated and entangled, indicating structural stress. The observed changes suggest that the interaction with yeast R1b disrupts the integrity of the bacterial envelope, possibly through secreted metabolites, cell surface interactions, or competitive mechanisms. Such damage may impair bacterial viability and growth, supporting the inhibitory role of R1b against E. coli ATCC 25922. Similar morphological disruptions have been reported in previous studies where probiotics or yeast-derived factors exert antimicrobial effects. These findings highlight the potential of strain R1b as a biocontrol agent by compromising the structural stability of pathogenic bacteria. Fig. 4ASEMs showing the morphological effects on E. coli ATCC 25922 cells after co-cultivation with the yeast strain R1b. The image reveals significant damage to the bacterial cells, including irregular surfaces, cellular deformation, and wrinkling of the cell envelope, indicating a loss of structural integrity. Scale bar = 1 μm ASEMs showing the morphological effects on E. coli ATCC 25922 cells after co-cultivation with the yeast strain R1b. The image reveals significant damage to the bacterial cells, including irregular surfaces, cellular deformation, and wrinkling of the cell envelope, indicating a loss of structural integrity. Scale bar = 1 μm Antibiotic resistance evaluation of the isolated strains Antibiotic resistance in yeasts arises through two main mechanisms: intrinsic and acquired resistance. Intrinsic resistance refers to the natural ability of yeast species to withstand certain antibiotics due to structural and physiological traits, such as robust cell walls and the presence of efflux pumps that expel antibiotics. Acquired resistance in yeasts is less common but may occur through chromosomal mutations or gene amplification. While horizontal gene transfer is a well-documented mechanism in bacteria, it is rare in yeasts. In laboratory settings, yeasts can acquire antibiotic resistance genes via engineered plasmids, but natural plasmid-mediated horizontal transfer between yeasts is not a primary mechanism of resistance. Understanding these mechanisms is essential for managing antibiotic resistance in clinical, industrial, and probiotic applications of yeast [64, 65]. Table 3 illustrates the findings of the antibiotic resistance evaluation. Assessing the isolated strains’ resistance to erythromycin, ampicillin, tetracycline, gentamicin, and kanamycin was a critical component in determining their probiotic potential. Probiotic strains should ideally exhibit susceptibility to antibiotics to avoid the development of antibiotic resistance. The presence of antibiotic-resistant probiotics poses a significant risk, as it can facilitate the transmission of resistant traits to pathogenic bacteria. The horizontal transfer of antibiotic resistance genes to pathogens raises safety concerns regarding probiotic use. Thus, antibiotic-sensitive probiotics are generally preferred [50]. Given the global health crisis of antibiotic resistance, verifying that probiotic candidates do not serve as reservoirs for transferable resistance genes is a paramount safety requirement [66]. Therefore, the primary objective of this evaluation was to identify whether any of the strains had unintentionally developed resistance to specific antibiotics. Antibiotic resistance is a key consideration in selecting probiotic strains, as these microorganisms must effectively combat infections without acquiring resistance to medications that could compromise user health. Prior to conducting this assessment, 13 isolated strains were selected based on the outcomes of earlier phases of the study. The test results indicated that three out of these 13 strains—designated (YM1, SC33, and AP1) exhibited resistance to all tested antibiotics, leading to their exclusion from further research. This removal was essential to ensure that only strains devoid of undesirable antibiotic resistance were included in the study, which also mitigates the risk of propagating antibiotic resistance from the gut microbiota to other areas of the body. Regarding the retained strains presented in Table 3, the observed resistance to specific antibacterial agents (e.g., Ampicillin) is classified as intrinsic resistance. Unlike bacteria, yeasts naturally lack the specific cellular targets (such as peptidoglycan cell walls) that these antibiotics attack [66]. Consequently, this form of resistance is inherent to the yeast physiology and is not carried on mobile genetic elements like plasmids. This distinction is critical for safety, as it implies there is minimal risk of horizontal gene transfer of resistance determinants to the gut microbiota, a major concern often associated with bacterial probiotics [67]. Consequently, the results of this evaluation underscore the importance of carefully considering antibiotic resistance when selecting probiotic strains, in order to safeguard public health and maintain the efficacy of antibiotics. In 2021, Fernández-Pacheco et al. investigated the antibiotic resistance of 20 yeast strains using both antifungal drugs (including nystatin, ciclopirox olamine, clotrimazole, and fluconazole) and antibiotics (such as erythromycin, gentamicin, chloramphenicol, tetracycline, and ampicillin). The analysis revealed that all yeast strains displayed some level of resistance to the antibiotics tested. Conversely, with the exception of Hanseniaspora osmophila, which did not produce an inhibitory zone in response to ciclopirox olamine, all twenty yeast strains were susceptible to the antifungal agents examined in their study. These findings further highlight the necessity of evaluating antibiotic resistance in probiotic strains to ensure their safety and effectiveness, ultimately supporting their role in promoting gut health without contributing to antibiotic resistance issues [68]. Table 3Antibiotic susceptibility profiles of the 13 selected yeast isolates against 5 common antibiotics. The results were determined using the disk diffusion methodAntibioticsAPControlCP1SR3AP1YM1YB2R2bSC33SR1CKKN2R1bErythromycin‒+‒+‒‒+‒‒‒+‒‒Ampicillin‒‒+‒‒‒‒‒‒‒‒‒+Kanamycin‒‒‒‒‒‒‒+‒+‒‒‒Gentamycin++++‒‒++‒++++Tetracycline+‒++‒‒‒‒‒‒‒+‒ Table 3Antibiotic susceptibility profiles of the 13 selected yeast isolates against 5 common antibiotics. The results were determined using the disk diffusion methodAntibioticsAPControlCP1SR3AP1YM1YB2R2bSC33SR1CKKN2R1bErythromycin‒+‒+‒‒+‒‒‒+‒‒Ampicillin‒‒+‒‒‒‒‒‒‒‒‒+Kanamycin‒‒‒‒‒‒‒+‒+‒‒‒Gentamycin++++‒‒++‒++++Tetracycline+‒++‒‒‒‒‒‒‒+‒ Antibiotic susceptibility profiles of the 13 selected yeast isolates against 5 common antibiotics. The results were determined using the disk diffusion method Antioxidant activity evaluation of the isolated strains The antioxidant activity of yeasts is supported by several key mechanisms. A major contributor is the production of antioxidant enzymes, such as catalase and superoxide dismutase (SOD), which help neutralize reactive oxygen species (ROS) and reduce oxidative stress. Another important molecule is glutathione, a tripeptide produced by many yeast species that acts as a powerful intracellular antioxidant. Structural components of the yeast cell wall, particularly β-glucans and mannans, also contribute to antioxidant activity by scavenging free radicals and forming a barrier against oxidative damage. While yeasts do not typically produce significant amounts of polyphenols, they may interact with or bind these compounds when present in their environment. Some strains may also synthesize limited amounts of antioxidant vitamins, depending on the species and growth conditions. Together, these factors highlight the antioxidant potential of yeasts and support their use in health-promoting and probiotic applications [69]. In this phase of evaluating probiotic potential, the antioxidant activity of the isolated strains was assessed using the DPPH free radical scavenging method. This test aimed to examine the ability of the strains to neutralize free radicals and combat oxidative stress. The findings are illustrated in Fig. 5. The DPPH test is one of the most widely employed methods for evaluating antioxidant activity, utilizing spectrophotometry to measure free radical inhibition. The color change of DPPH serves as an indicator of the scavenging ability of the samples [70]. Before conducting this assay, 10 isolated strains were selected from previous experiments. All yeast strains exhibited antioxidant activity by absorbing environmental free radicals; however, the CP1 strain was excluded due to its low antioxidant activity and limited ability to inhibit free radicals. The DPPH test results indicated that CP1 could only inhibit 31.24% of the DPPH free radicals. In contrast, strains R2b, SR1, and YB2 demonstrated the highest antioxidant potentials, with inhibition rates of 82.83%, 80.25%, and 79.34%, respectively (p < 0.05). The remarkable scavenging capacity observed in these top-performing strains is likely attributable to their superior production of intracellular antioxidant metabolites, particularly glutathione, as well as cell wall polysaccharides like β-glucans. These compounds function as electron donors, effectively neutralizing DPPH radicals and protecting cells against oxidative damage [65]. This test is crucial, as probiotic activity can help mitigate oxidative and inflammatory damage in the gut and other body systems [71]. The DPPH test results indicated that the eliminated strain had a lower capacity to reduce free radicals than the remaining strains, which effectively suppressed a higher percentage of DPPH free radicals. This characteristic suggests that the strain’s probiotic qualities may be negatively impacted by its inadequate ability to produce antioxidant compounds. Thus, it was necessary to exclude this strain from further study to ensure the selection of strains with robust antioxidant capabilities that can protect cells from oxidative and inflammatory damage. Through our screening process, we successfully identified strains with potent antioxidant activity that are beneficial for gastrointestinal and overall health. In a related study conducted in 2021, Fernández-Pacheco et al. evaluated the probiotic potential and antioxidant activity of 10 Saccharomyces yeast strains and 10 non-Saccharomyces yeast strains using the DPPH assay. The results demonstrated that all examined yeasts exhibited antioxidant potential by scavenging environmental free radicals. Most strains—including several non-Saccharomyces strains—exhibited favorable traits. Specifically, two strains of Saccharomyces and one strain of Pichia kudriavzevii showed the highest percentage of DPPH reduction, with respective values of 33.71%, 32.66%, and 33.42% [44]. Fig. 5DPPH radical scavenging activity of the 10 selected yeast isolates. The chart displays the mean percentage of radical scavenging activity for each strain. Radical scavenging activity (%) reported as mean ± SD (n = 3) DPPH radical scavenging activity of the 10 selected yeast isolates. The chart displays the mean percentage of radical scavenging activity for each strain. Radical scavenging activity (%) reported as mean ± SD (n = 3) Molecular identification of isolated strains Table 4 presents the findings of the molecular identification process. The PCR method was employed to accurately identify the species and genus of the yeast strains that successfully passed all probiotic and safety assessments. This technique is known for its reliability and precision in molecular identification of microorganisms by targeting specific genetic sequences, enabling accurate identification at both the genus and species levels. The identified strains belong to the probiotic yeast genera Pichia and Saccharomyces. Molecular identification plays a critical role in probiotic research, especially in selecting the appropriate probiotic strains. This phase of the study not only assessed the probiotic properties but also provided a clear understanding of the strains’ genus and species, which is essential for conducting further research and for the effective utilization of specific strains in probiotic products. Thanks to the high accuracy of PCR-based strain identification, the selected probiotic strains are confidently identified [72]. Moreover, molecular identification is vital for eliminating any potential identification errors and ensuring the accuracy of the study findings. Table 4Final species identification of the selected probiotic yeast strains based on PCR analysisCodeNameAccession No.R1bSaccharomyces cerevisiae strain JBM-IAUF-5PX111597YB2Pichia fermentans strain PM79OP696694SR1Pichia fermentans strain P9AOP696690Control Saccharomyces cerevisiae PTCC 5052APPichia fermentans strain HR-1 DS1308KM029994SC2Saccharomyces cerevisiae strain N85PX498691KN2Pichia fermentans strain PM19OP696693SR3Saccharomyces cerevisiae strain Yea12PX111598CKPichia fermentans strain XZY223-2MW710386R2bSaccharomyces cerevisiae strain TU21KU535593 Table 4Final species identification of the selected probiotic yeast strains based on PCR analysisCodeNameAccession No.R1bSaccharomyces cerevisiae strain JBM-IAUF-5PX111597YB2Pichia fermentans strain PM79OP696694SR1Pichia fermentans strain P9AOP696690Control Saccharomyces cerevisiae PTCC 5052APPichia fermentans strain HR-1 DS1308KM029994SC2Saccharomyces cerevisiae strain N85PX498691KN2Pichia fermentans strain PM19OP696693SR3Saccharomyces cerevisiae strain Yea12PX111598CKPichia fermentans strain XZY223-2MW710386R2bSaccharomyces cerevisiae strain TU21KU535593 Final species identification of the selected probiotic yeast strains based on PCR analysis Species-level identification of strain SC2 was confirmed based on its highest BLASTn similarity to an unsubmitted ITS sequence of S. cerevisiae strain N85. As no publicly accessible accession number was available for this reference sequence, identification relied on the closest matching unsubmitted ITS entry in the NCBI database. While our in vitro results indicate promising probiotic potential for the selected S. cerevisiae and P. fermentans strains, several limitations should be acknowledged. First, all evaluations were conducted in vitro, and additional in vivo studies are required to confirm their efficacy and safety in a living host. Second, antimicrobial activity was assessed exclusively against E. coli ATCC 25922, and future studies should evaluate these strains against a broader spectrum of relevant pathogens. Third, the application of stringent selection criteria resulted in a limited number of final candidate strains, potentially constraining the exploration of greater genetic and functional diversity. Finally, although in vitro safety assessments were performed, in vivo validation is necessary to comprehensively establish the safety of these strains prior to their widespread use in food and feed applications. The ITS-based phylogenetic analysis revealed two distinct clades, clearly separating the isolates into S. cerevisiae and P. fermentans. All four S. cerevisiae isolates (PX111597, PX111598, KU535593, and PX498691) clustered tightly, indicating strong intra-species similarity. The remaining isolates formed a well-supported P. fermentans clade (Fig. 6). This clustering pattern is fully consistent with the BLASTn identification results, confirming accurate species-level assignment for all isolates. Phylogenetic analysis based on the ITS1–5.8 S–ITS2 region provided robust support for the molecular identification of the isolates. The clear separation between the S. cerevisiae and P. fermentans clades reflects the expected evolutionary divergence between these genera. The tight clustering of the S. cerevisiae isolates, including PX498691, indicates high genetic relatedness within this group. Similarly, the consistent grouping of the P. fermentans isolates confirms their correct taxonomic assignment. Collectively, these results highlight the reliability of ITS sequencing for yeast species identification and validate the accuracy of the taxonomic classifications employed in this study [73]. Fig. 6Neighbor-joining phylogenetic tree constructed based on ITS1–5.8S–ITS2 sequences of the selected yeast isolates. The tree illustrates species-level clustering of the isolates into Saccharomyces cerevisiae and Pichia fermentans. The tree is presented as a simplified representation intended to support molecular identification rather than detailed evolutionary inference Neighbor-joining phylogenetic tree constructed based on ITS1–5.8S–ITS2 sequences of the selected yeast isolates. The tree illustrates species-level clustering of the isolates into Saccharomyces cerevisiae and Pichia fermentans. The tree is presented as a simplified representation intended to support molecular identification rather than detailed evolutionary inference Conclusion This study highlights agro-industrial byproducts and rumen-derived environments as promising reservoirs for discovering resilient probiotic yeasts. The nine isolates identified showed a favorable combination of functional traits, including strong tolerance to simulated gastrointestinal conditions, high hydrophobicity and aggregation capacity, and notable antimicrobial and antioxidant activities. These characteristics indicate an inherent adaptability to stressful environments and suggest a meaningful potential for modulating gut microbial balance. Overall, the functional profile of these strains supports their applicability as novel probiotic candidates for food and feed formulations. However, because all evaluations were performed in vitro, further in vivo investigations are necessary to confirm their safety, persistence within the host, and biological efficacy prior to any commercial implementation. CRediT authorship contribution statement Fereshteh Falah: Supervision, Investigation, Methodology, Writing – review & editing. Negin Ghazanfari & Atieh Rahimi: Writing – original draft, Investigation, Methodology. Damoon Ghofrani Tabari, Morteza Kashaninejad, Yasamin Yousefi and Behrooz Alizadeh Behbahani: Methodology, Writing – review & editing. Alireza Vasiee: Supervision, Data Curation, Writing – review & editing.