Isolation, characterization of Weissella confusa and Lactococcus lactis from different milk sources and determination of probiotic features

Abstract  

This study aimed to investigate the probiotic properties of Lactic Acid Bacteria (LAB) isolates derived from various milk sources. These isolates identified based on their morphological characteristics and 16S rRNA gene sequencing. Four strains of Lactococcus lactis and two strains of Weissella confusa were identified with over 96% 16S rRNA gene similarity according to the NCBI-BLAST results. The survival of the isolates was determined in low pH, pepsin, bile salts, and pancreatin, and their adhesion ability was assessed by in vitro cell adhesion assay, hydrophobicity, auto- and co-aggregation, and safety criteria were determined by hemolytic, gelatinase activities, and DNAse production ability tests. The results showed that the LAB isolates had different levels of resistance to various stress factors. L. lactis subsp. cremoris MH31 showed the highest resistance to bile salt, while the highest pH resistance was observed in L. lactis MH31 at pH 3.0. All the isolates survived in pepsin exposure at pH 3.0 for 3 h. The auto-aggregation test results showed that all strains exhibited auto-aggregation ranging from 84.9 to 91.4%. Co-aggregation percentage ranged from 19 – 54% and 17 – 57% against Escherichia coli ATCC 25922 and Staphylococcus aureus ATCC 29213, respectively. The hydrophobicity capacity of the LAB isolated ranged from 35–61%. These isolates showed different adhesion abilities to Caco-2 cells (81.5% to 92.6%). None of the isolates exhibited DNase, gelatinase and hemolytic activity (γ-hemolysis). All results indicate that these LAB strains have the potential to be used as probiotics.

Introduction

Recent scholarly attention has focused on the microbiota present in human and other mammalian milk, recognizing their dual role as both a source of nutrition for neonates and a dynamic habitat for a plethora of microorganisms. This milieu provides an ideal environment that supports the proliferation of a diverse array of microorganisms encompassing both symbiotic and potentially pathogenic entities. These microbial populations are pivotal in establishing the neonatal intestinal microbiota and fortifying the host defense mechanism [1, 2]. The transmission of microorganisms, including lactic acid bacteria (LAB), occurs through dietary means, perpetuating their survival across generations. LAB, which includes species from genera such as Lactococcus, Pediococcus, Streptococcus, Enterococcus, Lactobacillus, Bifidobacterium, and Weissella, play a crucial role in the formation of the intestinal flora of mammalian neonates, a process that is initiated through breast milk [2–5].

Probiotics are live microorganisms that confer health benefits to the host when consumed in adequate amounts. These organisms play a crucial role in regulating the intestinal flora and can prevent and control diseases. LAB strains, a group of important probiotic microorganisms, are capable of resisting the harsh gastrointestinal environment. LAB can also stimulate the immune system, prevent and treat diarrhea, and improve lactose digestion [6]. Additionally, there is evidence to suggest that these microorganisms may possess anticancer properties [7, 8], antimicrobial activity against pathogens [9], and a positive effect on the treatment of Alzheimer's disease when used in conjunction with specific compounds [10].

Several LAB genera have been isolated from various sources and exhibit diverse biological activities [11–13]. The species isolated from human milk predominantly belong to the genera Lactococcus, Pediococcus, Weissella, Lactobacillus, and Leuconostoc [14–16]. Reports indicate that Lactobacillus and Bifidobacterium were isolated from Buffalo milk [17, 18], while Lactobacillus, Lactococcus, Limosilactobacillus, Weissella, and Lactiplantibacillus [19–21] were isolated from goat and sheep milk. Additionally, Leuconostoc, Streptococcus, Bacillus, Lactococcus, Lactobacillus, and Enterococcus were isolated from donkey milk [22–24].

It has been noted that the effects of probiotics and prebiotics on gut health vary depending on the species and strain of microorganisms [25, 26]. The diversity of plant species in natural grazing areas, the use of antibiotics, and the natural diet of animals can impact the microbiota of milk [27–29]. This highlights the significance of natural milk sources as an essential and irreplaceable resource for the isolation of LAB. The characterization of the microbiota in these natural milk sources is crucial for the collection of new probiotic bacteria for industrial applications [30]. Despite the isolation of probiotic microorganisms from various sources, the isolation of effective strains is still required [31, 32].

Weissella confusa has garnered the attention of researchers due to its potential properties. Its metabolic capabilities beyond lactic acid fermentation, such as exopolysaccharide production and antifungal activity, suggest that it may be a promising candidate for various biotechnological applications. This microorganism is also a member of the human microflora and has been identified as a potential probiotic [33–37]. In addition to its presence in milk, it has been isolated from fermented dough, soil, plants, freshwater lakes, spontaneously fermented products, and animal foods.

From an economic standpoint, there is considerable consumer interest in functional foods, and milk sources possess unique biological properties that make them highly attractive. These properties include high digestibility, distinct alkalinity, high buffering capacity, and therapeutic values in medicine, as well as their accessibility and affordability [38]. The present study focused on identifying potential probiotic bacteria in various milk sources, including breast, donkey, and buffalo milks. The isolates were characterized based on their survival in low pH, gastric juice, bile salts, and pancreatin, and in vivo tests were conducted to evaluate their colonization properties, hemolytic activity, and antagonistic activity against test pathogens.

Methods

Isolation of lactic acid bacteria from different milk sources

Milk samples were collected under aseptic conditions from 3 voluntary humans, 3 donkeys, and 3 buffaloes in sterile bottles and transported to the laboratory while maintaining at the cold chain. The samples were processed two hours after arrival. LAB were isolated from different milk source by using 10^–3 to 10^–5 dilutions with phosphate buffered saline (PBS g / L: 80.1 NaCl; 1.44 Na₂HPO₄; 0.2 KCl; 0.24 KH₂PO₄ pH 7.4), plated on to de Man, Rogosa and Sharpe (MRS; Merck, Germany) agar and incubated at 37  C for 24 h. Four colonies from each sample were picked up and transferred to MRS broth. Cultures incubated overnight at 37° C were washed twice with PBS and spread on MRS agar to check the purity of the isolates. Then they were stored in a 30% glycerol solution at -20  C. Preliminary identification was based on Gram staining and catalase activity.

DNA extraction

Genomic DNA isolation was done according to the protocol Osmanagaoglu and Kıran [39]. Briefly, 1.5 mL of the overnight cultures centrifuged at 3000 × g for 10 min. The supernatant was removed, and the pellet was resuspended in a 200 µL spheroplast buffer (10% sucrose, 25 mM Tris pH 8.4, 25 mM EDTA pH 8.0, 2 mg /mL lysozyme and 0.4 mg mL⁻¹ RNase A). For cell lysis to occur, the tubes were incubated at 37° C for 10 min. after vortexing. Then 50 µL of lysis buffer 1 (5% SDS) and 50 µL of lysis buffer 2 (5 M NaCl) were added to the tubes, shaken gently, and then incubated at 65° C for 5 min. At the end of the time, 100 µL of neutralization buffer (60 mL of 5 M potassium acetate, 11.5 mL of glacial acetic acid, and 28.5 mL of dH₂O) was added to the tubes and placed on ice for 5 min. Samples were centrifuged at 18000 × g and 4° C for 15 min. 400 µL of the supernatant was transferred to new tubes and the same volume of isopropanol was added. Samples kept at room temperature (RT: 23 ± 2° C) for 5 min. DNA was precipitated by centrifuging the samples at 18000 × g for 15 min at RT. Supernatant was removed and the pellet was washed with 70% ethanol and centrifugation at 18,000 × g at RT for 5 min. Pellets were air dried and re-suspended in 50 μL 1 × Tris EDTA (TE) buffer pH 8 and stored at -20 °C.

Determination of RAPD-PCR profiles

The intraspecific biodiversity of isolates from LAB was investigated by Randomly Amplified Polymorphic DNA-Polymerase Chain Reaction (RAPD-PCR) analysis using the following universal primers: OPA -7 (5'GAAACGGGTG 3') and OPA14 (5'TGCTGCAGGT 3'). PCR was performed by using a NyxTechnik thermal cycler (NyxTechnik, USA) and Taq DNA polymerase (Promega, USA). PCR conditions were designed 95° C for 10 min and 30 cycles at 95° C 45 s, 36° C 45 s and 72° C 45 s followed by a final extension at 72° C for 10 min.

Genotypic identification using 16S ribosomal RNA and evolutionary analyzes of isolates

Isolates were molecularly identified by amplifying 16S rRNA gene region polymerase chain reaction (16S rRNA-PCR), and pure cultures were sequenced by BM Software Consulting and Laboratory Systems Limited Company (Ankara, Türkiye). Primers 27-F (5’-AGA GTT TGA TCC TGG CTC AG-3’) and 1492-R (5’-CTA CGG CTA CCT TGT TAC GA-3’) were used to amplify the 16S rRNA gene. PCR was performed: at 95° C for 5 min and 35 cycles at 95° C 1 min, 58° C 1 min and 72° C 1 min, followed by final extension at 72° C for 10 min [40]. PCR sequence analysis was performed by BM Laboratory Systems (Ankara, Türkiye). Sequence analysis results were opened in Finch TV 1.4.0 program and converted to FASTA format. The sequences in FASTA format were compared with all sequences identified in the NCBI database using the program BLAST. The 16S rRNA sequences of the isolated LAB were deposited in GenBank and the accession number was assigned. Evolutionary analyses were performed in MEGA7 [41] using reference sequences (L. lactis strain NCDO 604 NR_040955.1; W. confusa strain JCM 1093 NR_040816.1; L. lactis subsp. cremoris HM058872.1; W. confusa KU324936.1; Lactiplantibacillus plantarum MH899299.1; L. cremoris NR_040954.1) from the NCBI database. Evolutionary analysis was carried out using the Neighbor-Joining method.

Production of bacteriocin

The antimicrobial potential of the LAB isolates was evaluated using the agar well diffusion method according to the recommendations of Gupta et al. [8]. Isolates were grown in MRS broth at 37° C for 24 h and then centrifuged (6000 × g, 4° C, 20 min) to collect the supernatant (cell-free supernatant, CFS). The CFS was filtered through a 0.22 μm syringe filter. CFS was adjusted to pH 6.5 with 0.1N NaOH. Bacterial indicator (Listeria monocytogenes; NCTC 5348, Salmonella enterica Typhimurium; ATCC 14028, Escherichia coli O157:H7; ATCC 35150) were grown overnight in Tryptic Soy Broth (TSB; Biolife, Italiana) broth, Pediococcus acidilactici, Lactococcus lactis, Lactobacillus sakei were grown in MRS broth 18 h. Overnight cultures were diluted to 10⁷ CFU / mL and plated on MRS agar with TSA Soft Agar (0.8%) and allowed to grow for 6 h at 37° C. Wells with a diameter of 6 mm were made on the surface plate using by sterile glass tips, and 50 μL CFS was added to each well. Plates were first pre-incubated for 2–3 h at RT to allow diffusion of the test material into the agar, and then the plates were incubated for an additional 18 h at 37° C. The antibacterial spectrum of the isolates was determined by measuring the zones.

Antibiotic sensitivity test and MIC test

The LAB isolates were tested for antibiotic susceptibility to erythromycin (15 μg), vancomycin (30 μg), tetracycline (30 μg), chloramphenicol (30 μg), and kanamycin (30 μg) antibiotic discs (Himedia Ltd., India) the method proposed by Bauer et all [42]. Freshly cultured bacterial suspension (after dilution to approximately 10⁵–10⁶ CFU / mL) was evenly distributed on the surface of MRS agar plate. Antibiotic platelets were added to the plates and then incubated at 37° C for 24–48 h. After incubation, the diameters of the zones of inhibition were measured, and the degree of inhibition was classified as resistant (R, zone diameter ≤ 14 mm), intermediate (I, zone diameter 15–19 mm), or susceptible (S, zone diameter > 20 mm).

Antibiotic susceptibility of the isolated LAB was evaluated using the MIC test according to the internationally accepted standards of the Clinical and Laboratory Standard Institute [43]. Vancomycin, ampicillin, penicillin-G and streptomycin were used 0.5, 1, 2, 4, 8, 16, 32, 64, 128, 256, 512, 1024, 2048 μg / mL concentrations. Results were interpreted according to the cut-off levels proposed by European Food Safety Authority (EFSA) [40].

Tolerance to low pH

Method of Maragkoudakis et. al. [44] was followed to determine tolerance to low pH. The isolates were grown at 37° C for 18 h and the isolates were taken into 1 mL Eppendorf tubes and centrifuged at 12000 × g for 5 min at + 4° C, and the supernatant was removed. The pellet was washed three times with PBS (pH 7.4) buffer. The PBS buffer to be used for determining the pH resistance was adjusted to be pH 2.0 and 3.0 with 5 M HCl. The pellets of each strain were added to the tubes containing the pH adjusted buffers and the tubes were vortexed. As a control, PBS with a pH 7.4 was used. Tubes were incubated at 37° C. The samples taken from these tubes at 0, 1^st and 3^rd h were diluted and poured to MRS medium. After incubation at 37° C for 48 h, the colonies in the petri dishes were counted. Values are calculated as CFU / mL.

Tolerance to bile salt

The isolates were evaluated for effect of bile salt on the viability of the strains using Maragkoudakis et al. method [44]. MRS broth containing 0.3%, 0.5% and 0.1% bile salts were used for experiment, and MRS broth without bile salts were used as control. The overnight culture was centrifuged at 12000 × g for 5 min at + 4° C and supernatant was removed. The pellet was washed twice with PBS (pH 7.4) buffer, and it was dissolved in MRS broth containing 0.3%, 0.5% and 0.1% bile salt and incubated at 37° C for 4 h. Samples were taken at 0,1^st and 3^rd and 4^th h subjected to serial dilutions, and smears were planted on MRS solid medium in 3 parallels. After 24 h of incubation at 37° C, the colonies were counted. Values were calculated as log CFU / mL.

Bacterial resistance to pepsin, and pancreatin

Bacterial cultures growth in media for18 h and then centrifuged at 10000 × g for 5 min. Supernatant was removed, and pellet washed twice with PBS buffer. Then the pellet was suspended in PBS buffer (pH 8.0) including 1 mg / mL pancreatin (Sigma-Aldrich, USA). The same procedure was performed in PBS buffer (pH 2.0 and 3.0) supplemented with 3 mg /mL pepsin (Sigma-Aldrich, UK). PBS was used as control (pH7). Samples were incubated at 37° C (pepsin) for 3 h and at 37° C (pancreatin) for 4 h. The samples were inoculated on MRS agar at different time intervals, then incubated at 37° C for 24 h. Results were evaluated by counting the CFU / mL [44].

Auto-aggregation Assay

The auto-aggregation test was performed according to Collado et al. [45] with minor modifications. LAB were grown at 37° C for 20 h in MRS broth. Cells were collected by centrifuge (10.000 × g for 10 min at 4° C) and washed twice with PBS (pH 7.2) and then re-suspended in PBS. Absorbance at a wavelength of OD₆₀₀ nm (A_600nm) was adjusted to 0.25 ± 0.05 to standardize the number of bacteria (10⁷ –10⁸ CFU / mL). Then, the bacterial suspensions were incubated in 1 mL aliquots at 37° C, which were monitored at different times at OD₆₀₀ (0 or 5 h). Auto-aggregation percentage was expressed as [1 − A_t/A₀] × 100, where A_t represents the absorbance at time t = 5 h, and A₀ the absorbance at t = 0.

Co-aggregation

Aggregation properties of the selected LAB were performed according to the procedure described by Huligere et al., [31]. Bacterial suspension for co-aggregation was prepared as described above Auto-aggregation Assay. The LAB suspension was mixed with in the ratio of 2:1, each of the 3 pathogenic strains (E. coli ATCC 25922, S. aureus ATCC 29213 and S. Typhimurium ATCC 14028) and incubated at 37 °C for 4 h. The suspensions were monitored OD₆₀₀ nm at different times (0, 2^nd, and 4^th h). The co-aggregation percentage was expressed as:

$$\mathrm{Co}-\mathrm{aggregation}\%=({\mathrm{A}}_{\mathrm{L}}+{\mathrm{A}}_{\mathrm{P}})-{\mathrm{A}}_{\mathrm{mix}}/({\mathrm{A}}_{\mathrm{L}}+{\mathrm{A}}_{\mathrm{P}})\times 100$$

where A_mix signifies [the absorbance of the LAB mixture + pathogen at the different time], and A_L + A_P denotes [the absorbance of the LAB mixture + pathogen at 0 h].

Hydrophobicity

The degree of cell surface hydrophobicity of the LAB isolates was assessed via measuring microbial adhesion to hydrocarbons (MATH) in xylene (Sigma-Aldrich, Germany) [46, 47]. Briefly, overnight cultures of isolates were centrifuged at 5,000 × g for 3 min at 4° C. The pellets were washed twice with PBS and re-suspended in PBS, then the OD₆₀₀ of the isolates was adjusted by spectrophotometry (Biochrom genequant, UK) in the range of 1.0 ± 0.1. Each bacterial suspension (3 mL) was mixed with 1 mL of xylene (Macklin, Shanghai, China), swirled for 2 min, then incubated at 37° C for 1 h. The water layer was carefully aspirated and measured at OD600 (Biochrom genequant, UK). Cell surface hydrophobicity (%) was calculated the equation as follows:

$$\mathrm{Hydrophobicity}\%=[-(1-{\mathrm{A}}_{\mathrm{t}})/{\mathrm{A}}_0\times 100]$$

Where A₀ denotes the optical density at 0 h and A_t stands for the optical density at 1 h.

Hemolytic activity

Bacterial cultures were grown overnight on Columbia agar (Oxoid, Ireland) containing 5% (v/v) sheep blood (Oxoid) and incubated for 24 h at 37 °C. Colonies were classified as hemolytic positive if they had a halo (green area for α-hemolysis, clear area for β-hemolysis, and no halo for y-hemolysis) [47]. S. aureus ATCC 25923, E. coli ATCC 25922 used as positive control.

DNase activity

To evaluate capacity to produce deoxyribonuclease (DNase) enzymes of LAB isolates were streaked onto DNase agar plates (BioLife Italiana, Milano, Italia) and incubated 48 h at 37° C. After incubation time 1N HCl was added agar plates and DNase activity was evaluated by the formation of the clear zone around the strains. A distinct zone surrounding the colonies found evidence of positive DNase activity [29]. S. aureus ATCC 29213 used as a positive control.

Gelatinase activity

The gelatinase producing activity of LAB was performed according to Zhang et al. [46]. Briefly, 1 mL of 24 h incubated LAB was spotted on MRS agar with 5% (w/v) gelatin (Solarbio, China), and the plates were incubated anaerobically at 37° C for 72 h, then cooled at 4° C for 4 h. The opaque halo around the colony is considered to be a positive result of gelatinase production.

In vitro cell adhesion assay

Adhesive capacity was determined based on the previous method with minor adjustments [39, 47]. In brief, Caco-2 cells were seeded in 24-well plates (5 × 10⁴ / well) and cultured in Dulbecco’s Modified Eagle’s Medium (DMEM) (Sigma-Aldrich, Germany) with 10% Stain Buffer (FBS) to obtain a confluent monolayer cell. The isolates were cultured for 24 h at 37° C in MRS broth and harvested by centrifugation (5000 × g, 3 min, 4° C). The pellets were washed twice in sterile PBS pH 7.4. Then, the bacterial solution was re-suspended in PBS and finally the cell density was adjusted to 10⁸ CFU / mL with DMEM medium. The cell monolayers were washed three times with PBS, then 1 mL of serum-free DMEM medium and 50 µL of the bacterial suspension were added to each well in three replicates, and the 24-well plates were incubated for 90 min at 37° C and 5% CO₂. First, the culture medium was aspirated, and each well was carefully washed three times with sterile PBS to remove non-adherent bacteria. Then, 100 μl of 0.1% Tween 80 (Isochem, USA) was added to each well and the plates were incubated at 37° C for 30 min to completely digest the cells. At the end of the time, the wells were again washed three times with PBS and the cells were transferred to new, fresh Eppendorf tubes. An amount of 100 μl of this suspension was added to a tube containing 900 μl of sterile PBS pH 7.4 (1:10 dilution), and serial decimal dilutions were prepared. From each dilution, 100 μl was added to MRS agar plates, which were then incubated at 37 °C for 18 h. The number of bacteria adhering to the Caco-2 cells was counted on the plates and expressed as CFU.

Statistical analysis of experimental data

For statistical analysis, t-test and analyses of variance (ANOVA) were used in SPSS 24 program. One-Sample T test and One-Way Anova were used for determination of significance and TUKEY test were used for post hoc analysis. A confidence interval of 95% (P < 0.05) was considered as significant. All assays were done in triplicate.

Results

Identification of LAB

A total of 20 bacterial colonies were detected in the tested milks, and 9 of them were identified as LAB through Gram staining and catalase tests. These LAB isolates were Gram-positive and catalase negative. The universal OPA7 primer and the RAPD-PCR method were employed for intraspecific identification of the LAB isolates. Six isolates with different band profiles were selected from the nine isolates (buffalo milk: MH10, MH13; human milk: MH24, MH26; donkey milk MH29, MH31). For preliminary evaluation, six isolates that differed according to their RAPD profiles were selected (Fig. 1). These six isolates were then subjected to 16S rRNA sequence analyses, and BLAST results showed very high homology with other reference strains deposited in the NCBI GenBank. Based on 16S rRNA sequencing results, the four strains were identified as Lactococcus lactis and two strains as Weissella confusa. These strains were designated as L. lactis MH10, L. lactis subsp. cremoris MH13, W. confusa MH24, W. confusa MH26, L. lactis MH29, and L. lactis MH31 (Table 1). Finally, a dendrogram was constructed using MEGA 7 to show the phylogenetic relationship between the six strains and reference strains in the NCBI database (Fig. 2) [48–50]. The phylogenetic analysis revealed that the strains obtained from the same milk source were grouped within the same clade. Notably, W. confusa MH24 and MH26, isolated from different human milk samples, were found to be part of the same clade and demonstrated 100% similarity with the reference strain. Similarly, L. lactis MH29 and MH31, isolated from donkey milk and the reference strain, showed 100% similarity. However, L. lactis MH10, isolated from buffalo milk, was grouped in a different clade but still shared 93% similarity with the reference strain. Lastly, L. lactis subsp. cremoris MH13, isolated from buffalo milk, was found within the same clade as other L. lactis strains and showed 100% similarity with the reference strain.

Fig. 1.

Dendrogram shows RAPD_PCR analysis of LAB isolates using by OPA7 primer (isolate 29: Lactococcus lactis MH29, isolate 30: L. lactis MH100, isolate 31: L. lactis MH10), breast milk (isolate 24: Weissella confusa MH24, isolate 26: W. confusa MH26) and buffalo milk (isolate 10: L. lactis MH10, isolate 11: L. lactis MH11, isolate 12: L. lactis MH12, isolate 13: L. lactis subsp. cremoris MH13, M: Marker)

Table 1.

Accession number of 16S rRNA gene sequence of LAB isolate. Lactococcus lactis MH10, L. lactis subsp. cremoris MH13, Weissella confusa MH24, W. confusa MH26, L. lactis MH29, L. lactis MH31

Isolate	Similarity based on 16S rRNA	Strain code	Isolation source	Sequence length (bp)	Similarity (%)	NCBI Accession number
10	Lactococcus lactis	MH10	Buffalo	1151	97.56	MW633195
13	L. lactis subsp. cremoris	MH13	Buffalo	110	96.59	MW633196
24	Weissella confusa	MH24	Human	1166	98.11	MW633198
26	W. confusa	MH26	Human	1151	100	MW633199
29	L. lactis	MH29	Donkey	1152	97.33	MW633200
31	L. lactis	MH31	Donkey	1101	97.28	MW633201

Fig. 2.

Phylogenetic relationships according to the nucleotide sequences of the 16S rRNA fragments of the strains identified with their GenBank accession numbers. NR_040816.1: W. confusa strain JCM 1093, HM058872.1: L. lactis subsp. cremoris, KU324936.1: W. confusa, MH899229.1: Lactococcus lactis, NR_040954.1: L. cremoris, 10: L. lactis MH10, 13: L. lactis subsp. Cremoris MH13, 24: W. confusa MH24, 26: W. confusa MH26, 29: L. lactis MH29, 30: L. lactis MH30 31: L. lactis MH31

Antibiotic sensitivity test and MIC test

According to the European Food Safety Authority (EFSA), it is recommended to test the use of antibiotics that inhibit cell wall synthesis and protein synthesis for selected probiotics [40]. For this reason, antibiotics that inhibit cell wall synthesis and protein synthesis were selected, and the disc diffusion method was performed. The results were interpreted according to the break-point recommended by the Clinical And Laboratory Standards Institute (CLSI) [43] guidelines: isolates with an inhibition zone less than or equal to 14 mm were considered resistant (R), those with a diameter greater than 20 mm were considered susceptible (S), and those with a zone diameter between 15 and 20 mm were considered intermediate (I) [51–53]. The percent-wise LAB isolates showed resistance to kanamycin (100%) and gentamicin (83%) in the disc diffusion test (Table 2). Additionally, LAB isolates showed resistance to streptomycin (100%), vancomycin (83%), penicillin-G (50%), and ampicillin (17%) according to the MIC test (Table 3).

Table 2.

Antibiotic resistant of isolated LAB disc diffusion test result

	*Antibiotics
Isolate	Erythromycin (15 µg)	Tetracycline (30 µg)	Kanamycin (30 µg)	Vancomycin (30 µg)	Gentamicin (10 µg)	Chloramphenicol (30 µg)
L. lactis MH10	20.0 ± 1.0 (S)	25.5 ± 1.0 (S)	14.0 ± 1.0 (R)	18.0 ± 1.0 (I)	9.0 ± 1.0 (R)	16.0 ± 1.0 (I)
L. lactis subsp. cremoris MH13	20.5 ± 1.0(S)	26 ± 1.0 (S)	8.0 ± 1.0 (R)	18.0 ± 1.0 (I)	7.0 ± 1.0 (R)	19.5 ± 1.0 (I)
W. confusa MH24	26.0 ± 1.0 (S)	20.5 ± 1.0 (S)	7.5 ± 1.0 (R)	7.0 ± 1.0 (R)	11.0 ± 1.0 (R)	20.0 ± 1,0 (S)
W. confusa MH26	19.5 ± 1.0 (I)	22.0 ± 1.0 (S)	4.0 ± 1.0 (R)	7.0 ± 1.0 (R)	15.0 ± 1.0 (I)	22.0 ± 1.0 (S)
L. lactis MH29	19.0 ± 1.0 (I)	25.0 ± 1.0 (S)	7.5 ± 1.0 (R)	5.0 ± 1.0 (R)	5.0 ± 1.0 (R)	15.5 ± 1.0 (I)
L. lactis MH31	19.0 ± 1.0 (I)	9.0 ± 1.0 (R)	3.5 ± 1.0 (R)	17.0 ± 1.0 (I)	7.5 ± 1.0 (R)	20.0 ± 1.0 (S)

"-" No zone. S: > 20 (mm), I: 15–19 (mm), R: ≤ 14 (mm)

Table 3.

MIC distributions of four antibiotics for isolated LAB

	*Antibiotics (μg mL−1)
Isolate	Vancomycin	Ampicillin	Penisilin-G	Streptomisin
L. lactis MH10	< 0.5 (S)	< 0.5 (S)	8 (R)	128 (R)
L. lactis subsp. cremoris MH13	512 (R)	< 0.5 (S)	< 0.5 (I)	64 (R)
W. confusa MH24	512 (R)	1 (S)	< 0.5(I)	1024 (R)
W. confusa MH26	512 (R)	4 (S)	2 (R)	2048 (R)
L. lactis MH29	512 (R)	16 (R)	< 0.5 (I)	> 2048 (R)
L. lactis MH31	512 (R)	4 (S)	4 (R)	> 2048 (R)

^*S: Susceptible, R: Resistant, I: Intermediate

Tolerance to low pH

The results of the viability assay for the 6 LAB under low pH conditions revealed a reduction in bacterial viability. After just 3 h of exposure to low pH, bacterial viability was noted to be only MH31 (1.67 log / CFU), MH29 (1.33 log / CFU), MH26 (1.33 log / CFU), and MH13 (1.43 log / CFU) at pH2. At an extreme pH of 2.0, no surviving viable colonies were detected after 3 h. However, at pH 3.0, viability decreased at a rate of 0.7–4.7 log. The highest resistance was observed in L. lactis MH31 (6.75 log). The results revealed that there wasn’t statistically significant difference (P > 0.05) in the number of viable cells in the control, pH 2, and pH 3 groups at 0, 1, and 3 time zones.

Tolerance pepsin and pancreatin

To evaluate the probiotic potential of six selected LAB, their resistance to pepsin (at pH 2.0 and 3.0) and pancreatin was tested. No viable strains were detected in the presence of pepsin (pH 2.0), although they survived pepsin exposure at pH 3.0, with isolates W. confusa MH24 and L. lactis MH29 exhibiting survival rates of 6.52 log (P < 0.05) (Fig. 4a, b). The LAB displayed good tolerance to pancreatin exposure, with viability rates varying between 6–7.34 log. Among the six LAB tested, W. confusa MH26 was found to be the most sensitive to pancreatin, with a survival rate of 6.57 log, while L. lactis MH13 demonstrated the highest survival rate of 7.34 log and didn’t show any statistically significant difference compared to the control group (P > 0.05) (Fig. 5).

Fig. 4.

Effect of pepsin on the viability of lactic acid bacteria. a Pepsin pH 2 (F (2,55) = 46.828 (P < 0.05)), (b) Pepsin pH 3 (F (2,55) = 14.126; (P < 0.05)). 10: L. lactis MH10, 13: L. lactis subsp. cremoris MH13, 24: W. confusa MH24, 26: W. confusa MH26, 29: L. lactis MH29, 31: L. lactis MH31. The error bar indicated that standard error. The error bar indicated that standard error. Data are presented as mean ± standard deviation. Tukey tests revealed that the means of survival rate with different pH interval and superscripts (a-c) substantially different from one another

Fig. 5.

Effect of pancreatin on the viability of lactic acid bacteria. 10: L. lactis MH10, 13: L. lactis subsp. cremoris MH13, 24: W. confusa MH24, 26: W. confusa MH26, 29: L. lactis MH29, 31: L. lactis MH31. The error bar indicated that standard error. Data are presented as mean ± standard deviation. Tukey tests revealed there isn't significant differences (F (1,35) = 0.083; 0.251; 0.025; 0.08; (P > 0.05))

The viability of LAB cultures exposed to different concentrations of bile salts (0.3%, 0.5%, and 1%) was moderately affected. L. lactis subsp. cremoris MH13 demonstrated the highest resistance to bile salts (Fig. 6a, b, c). Viable cell numbers were measured in the first, third, and fourth time zones, revealing no statistically significant difference between the control, bile 0.3%, bile 0.5%, and bile 1% groups (P < 0.05).

Fig. 6.

Effect of 0,3% (a), 0,5% (b), 1% (c) bile salt on the viability of lactic acid bacteria. 10: L. lactis MH10, 13: L. lactis subsp. cremoris MH13, 24: W. confusa MH24, 26: W. confusa MH26, 29: L. lactis MH29, 31: L. lactis MH31. The error bar indicated that standard error. The error bar indicated that standard error. Data are presented as mean ± standard deviation. Tukey tests revealed that the means of survival rate with different pH interval and superscript (a) substantially different from one another (F (3,72) = 17.776 (P < 0.05))

Production of bacteriocin

The bacteriocins possess the potential to exert either bactericidal or bacteriostatic impact on vulnerable cells, and they are ribosomally synthesized antimicrobial peptides that can target bacteria of the same species (narrow spectrum) or across genera (broad spectrum) [49, 50]. However, following the bacteriocin test, no bacteriocin production could be established against the examined indicator microorganisms by the strains.

Auto-aggregation assay

The auto-aggregation of cells associated with the interaction of various surface components of cells has been found to have beneficial effects, and this property is closely related to the probiotics [33]. The auto-aggregation of individual species was observed in Lactococcus lactis MH10 (86.9%), L. lactis subsp. cremoris MH13 (85%), W. confusa MH24 (87.1%), W. confusa MH26 (91.1%), L. lactis MH29 (91.4%), and L. lactis MH31 (84.9%). A statistically significant difference was found between the samples collected at different times (P < 0.05) (Fig. 7a).

Fig. 7.

The auto-aggregation and coaggregation of lactic acid bacteria. (a): auto-aggregation (P < 0.05) (b): coaggregation E. coli ATCC 25922 (P < 0.05) (c): coaggregation S. aureus ATCC 25923 (P < 0.05) (d): coaggregation S. Typhmurium (P < 0.05). 10: L. lactis MH10, 13: L. lactis subsp. cremoris MH13, 24: W. confusa MH24, 26: W. confusa MH26, 29: L. lactis MH29, 31: L. lactis MH31. The error bar indicated that standard error. Data are presented as mean ± standard deviation. One-sample T tests revealed there are significant differences between the different timelines

Co-aggregation

The results of the co-aggregation test, conducted over a period of 4 h, indicated that all six isolates exhibited a range of co-aggregation percentages with E. coli ATCC 25922, varying from 19–41%. Similarly, the co-aggregation percentages with S. aureus ATCC 29213 and S. Typhimurium ranged from 32–54% and 17–57%, respectively. A statistically significant difference was observed between the isolates at the 2^nd and 4^th hours against different pathogens (P < 0.05) (Fig. 7b, c, d).

Hydrophobicity

The hydrophobicity capacity of the LAB isolates was observed to range from 35–61% upon using xylene. MH24 exhibited the highest hydrophobicity, while MH29 displayed the minimum. According to statistical test results, significant differences (P < 0.05) were observed (Fig. 8a).

Fig. 8.

The hydrophobicity and Caco-2 cell adhesion ability of LAB isolates. (a) Hydrophobicity percentages of LAB isolates to xylene (P < 0.05) (b) The Caco-2 cell adhesion ability of LAB isolates (P < 0.05). The error bar indicated that standard error. Data are presented as mean ± standard deviation. One-sample T tests revealed there are significant differences

Hemolytic activity, DNase activity and gelatinase activity

The safety of the probiotic isolates was determined through the performance of three tests: Hemolytic activity, DNase activity, and gelatinase activity. Results of the hemolytic activity test demonstrated that no strain exhibited any hemolytic activity (γ-hemolysis) (Fig. 9). Additionally, the isolates were devoid of DNase and gelatinase activity, as evident from the results (Fig. 10). Considering the safety tests, it can be inferred that the isolates are safe for use.

Fig. 9.

The Hemolytic activity (a) W. confusa MH24, (b) L. lactis MH31, (c) S. aureus ATCC 25923, (d) E. coli ATCC 25922

Fig. 10.

DNAse activity W. confusa MH24, 29: L. lactis MH29, 31: L. lactis MH31, S. aureus ATCC 25923

In vitro cell adhesion assay

Figure 8b displays the adhesion abilities of the isolates to Caco-2 cells, with varied levels of adhesion ranging from 81.5% to 92.6%. MH24 exhibited the highest adhesion ability at 92.6%. The in vitro cell adhesion assay One-Sample Statistics test result revealed significant differences (P < 0.05).

Discussion

Milk sources are considered to be the best food for living organisms as they provide a combination of essential nutrients such as proteins, carbohydrates, lipids, minerals, and vitamins, further augmented by the presence of probiotic bacteria, particularly lactic acid bacteria (LAB) [54, 55]. Utilization of LAB as probiotics has gained widespread acceptance in the treatment of intestinal diseases in both humans and animals, as they promote the overall health of the host by maintaining a balance of the intestinal microbiota [39, 53, 56]. Studies have shown that the beneficial effects of probiotics on intestinal health differ based on the type and strain of microorganisms [24, 57]. Therefore, although probiotic microorganisms have been isolated from various sources, there is still a need to identify more effective strains. The objective of our research was to determine the similarity of LAB content in milk from different organisms that reside in the same region and to evaluate the potential probiotic properties and safety of LAB isolated from these sources. The LAB strains were investigated for their potential probiotic properties, including pepsin, pancreatin, bile salt, and acid tolerance, auto-aggregation, co-aggregation, antibiotic sensitivity, hydrophobicity, hemolytic activity, gelatinase activity, DNAse activity, in vitro cell adhesion, and production of antimicrobial compounds.

Nine of the twenty isolates were identified as LAB based on preliminary tests. The selected strains were subjected to RAPD marker-based genetic characterization [58], and six of them were further analyzed (Fig. 1). The selected six strains were identified using 16S rRNA sequencing, and based on the phylogenetic analysis, they were grouped into two distinct clusters. Cluster 1 corresponded to Lactococcus lactis strains, while cluster two corresponded to Weissella confusa strains. The isolates were designated in Table 1, and their 16S rRNA sequencing results were compared with findings from similar studies on milks from various regions of the world. The majority of LAB isolated from various milk and milk products contained Lactobacillus, Lactococcus, Streptococcus, Leuconostoc, Bifidobacteria, Pediococcus [59, 60] and Weissella confusa [25, 30]. When compared with other studies, it was determined that LAB strains obtained from donkey milk from different regions could be similar (Lactobacillus paracasei, Lactococcus lactis, and Carnobacterium maltaromaticum) [61] or completely different (four Enterococcus, five Streptococcus, and one Pediococcus) [62] from the current research. It has been determined that the content of buffalo milk may vary (L. fermentum, L. casei, L. acidophilus and Bifidobacterium longum) [17] or remain similar (L. lactis subsp. lactis Lac3) [63] in comparison to the present study. Two strains of W. confusa were isolated from human milk sources. The presence of these strains is believed to be a result of early maternal breastfeeding. Cabrera-Rubio et al. [64] reported that this species is considered one of the most prevalent in the colostrum of Finnish women. It is acknowledged that bacteria isolated from human milk may show variability, with Kang [15] and Osmanagaoğlu [39] isolating Lactobacillus rhamnosus and Pediococcus pentasaceus respectively, while Weissella species were not found in their studies. This can be attributed to differences in diet, antibiotic consumption, and geographical location.

According to EFSA, the use of antibiotics that inhibit cell wall synthesis and protein synthesis is generally recommended for the selection of probiotic strains [40, 48]. The disk diffusion test results were evaluated based on the breakpoint recommended by the Clinical and Laboratory Standards Institute (CLSI) [43, 65] as follows: isolates with an inhibition zone less than or equal to 14 mm were considered resistant (R), those with a diameter greater than 20 mm were considered susceptible (S), and those with a zone diameter between 15 and 19 mm were considered intermediate (I) [51, 66]. Our study agrees with previous research, which has shown that Lactobacilli are resistant to antibiotics that inhibit protein synthesis and aminoglycosides, and glycopeptide [67, 68]. Intrinsic resistance is chromosomally encoded and does not pose a risk of horizontal metastasis in non-pathogenic bacteria. However, acquired resistance, which is caused by the transfer of resistance genes from the environment or other bacteria, can spread horizontally among bacteria and is a concern [64].

All probiotic strains must be able to survive the adverse conditions present in the gastrointestinal tract, such as gastric acid and bile salts, which range in pH from 1.5 to 4.5 with a mean value of 3.0, and 0.03–0.3% (w/v) concentration, respectively [64, 69, 70]. To assess the tolerance of isolates, we tested them at pH 2–3 and bile salt concentrations of 0.1%, 0.5%, and 1% [11, 71–73]. We found that the viability of LAB was highest in low pH3 conditions, with L. lactis MH31 showing the highest resistance (Fig. 3). At pH 2, only L. lactis MH31, L. lactis MH29, W. confusa MH26, and L. lactis subsp. cremoris MH13 survived, with bacterial viability ranging from 1.67 log / CFU to 1.43 log / CFU. However, no surviving viable colonies were detected after 3 h at an extreme pH 2.0. Similar results were reported by Enrique et al. [26], who showed that survival was better at pH 3.0. Simone et al. [74, 75] also reported that resistance of the isolate was observed after exposure to acidified media, except for pH 2. For the other treatments, pH 3 and pH 4, no significant differences were observed when comparing the samples with the control (pH 7) during the incubation period, which ranged from 0 to 4^th h. Our study aimed to evaluate the tolerance of potential probiotic strains during their passage through the Gastrointestinal Tract (GIT) Conditions system. To determine the probiotic power of 6 selected LAB, we conducted experiments to test their resistance to pepsin (at pH 2.0 and 3.0) and pancreatin. No viable strains were detected in the presence of pepsin (pH 2.0), while they survived pepsin exposure at pH 3.0, with survival of isolates W. confusa MH24 and L. lactis MH29 at 6.52 log (Fig. 4a, b). The LAB survived well to pancreatin exposure, and the viability rates varied between 6–7.34 log. Among the 6 LAB tested, isolate W. confusa MH26 (6.57 log) was most sensitive to pancreatin, while isolate L. lactis subsp. cremoris MH13 (7.34 log) had the highest survival rate according to the pancreatin (Fig. 5). In previous studies, cell viability to pepsin was 94.32—91.22% and pancreatic resistance was 97.32—93.11% [76]. According to the report by Cervantes-Elizarrarás [25] LAB isolates had a viability of between 63.2–96.3% under simulated GIT. This finding is in good agreement with those observed by previous studies [75, 77, 78]. It is suggested that LAB plays an active role in lowering cholesterol levels because it contains enzymes that can deconjugate bile salts to amino acids and cholesterol [79]. The exposure of LAB cultures to different concentrations of bile salts (0.3%, 0.5% and 1%) moderately affected the viability of the strains. L. lactis subsp. cremoris MH13 showed the highest resistance to bile salt (Fig. 6a, b, c). These results are consistent with the activity (or partial activity) of bile salt hydrolase found in some LAB [78, 80, 81]. In contrast, Bautista-Gallego et al. [82] reported the absence of this activity in LAB.

Fig. 3.

Effect of low pH on the viability of lactic acid bacteria. 10: L. lactis MH10, 13: L. lactis subsp. cremoris MH13, 24: W. confusa MH24, 26: W. confusa MH26, 29: L. lactis MH29, 31: L. lactis MH31. The error bar indicated that standard error. Data are presented as mean ± standard deviation. Tukey tests revealed there isn't significant differences (F (2,17) = 0.57; (P > 0.05))

The physiological properties of LAB, specifically the production of antimicrobial compounds and lactic acid in combination with bile salts, can significantly reduce or inhibit the growth of pathogens in the gut [79]. The results of the bacteriocin test were negative, indicating that the strains tested did not produce bacteriocins against the investigated indicator microorganisms. Bacteriocins are ribosomally synthesized antimicrobial peptides that can have a bactericidal or bacteriostatic effect on sensitive cells, and their activity is specific to the strain producing them [51, 52]. The production of bacteriocins is influenced by various factors, including the physiological state of the indicator cells and the presence of temperature, pH, and other antimicrobial compounds [83]. It is important to conduct wide-ranging tests on different indicator strains to fully assess the effectiveness of bacteriocins.

Auto-aggregation, co-aggregation, and hydrophobicity are important properties that affect the colonization of bacteria in the gut [82, 84]. Auto-aggregation is the ability of similar cell types to attach to each other, while co-aggregation is the combination of organisms from different species [82, 85]. High auto-aggregation capacity enables microorganisms to achieve higher population density and stability in the host gut, reducing exposure to adverse conditions. High co-aggregation capacity can prevent pathogens from colonizing the host gastrointestinal tract [30]. The mean percentage of auto-aggregation in individual species was determined in Lactococcus lactis MH3 (86.9%), L. lactis subsp. cremoris MH4 (85%), W. confusa MH6 (87.1%), W. confusa MH7 (91.1%), L. lactis MH8 (91.4%), L. lactis MH9 (84.9%) (Fig. 7a). Previous investigations have also reported a wide range of auto-aggregation in Lactobacillus, ranging from 5–68% [86, 87], and up to 28.8–87.7% [75]. The highest hydrophobicity was observed in W. confusa MH24 (61%), while the lowest was in L. lactis MH29 (35%) (Fig. 8a). The coaggregation percentages were observed at the end of 4 h, ranging from Escherichia coli ATCC 25922 (19–41%), Staphylococcus aureus ATCC 29213 (32–54%), and Salmonella Typhimurium ATCC 14028 (17–57%) (MH24-MH13, MH26-MH31, W. confusa MH24—L. lactis subsp. cremoris MH13), which is consistent with previous studies suggesting that strong adhesion is correlated with high hydrophobicity and high aggregation ability [88, 89].

Probiotic microorganisms should maintain a safe profile and should not induce hemolysis or gelatin liquefaction within the host's body, nor should they possess DNAse activity. Hemolysis and gelatin hydrolysis are two of the primary virulence factors associated with pathogenic bacteria. Hemolysin / cytolysin, a bacterial toxin, is secreted by certain enterococcal strains and is one such virulence factor. Microorganisms that failed to produce visible changes on blood agar were classified as non-hemolytic cells [87, 89, 90]. The positive control strains, Staphylococcus aureus ATCC 25923 and Escherichia coli ATCC 25922, exhibited distinct clear and clean zones around them due to the complete lysis of red blood cells on the media. However, the LAB isolates did not form clear or transparent zones on the blood agar and were not γ-hemolytic (Fig. 9). All strains tested negative for gelatinase and DNAse activity (Fig. 10).

The in vitro adhesion assay utilizing Caco-2 cells has been a commonly employed method for identifying probiotics. The results of the adhesion test reveal that six isolates demonstrated adhesion rates of 81.8% and 92.6%. A correlation between hydrophobicity and cell adhesion was observed, with the highest levels of both parameters noted in strain W. confusa MH24 (61%-92.6%). These findings are consistent with previous studies (Fig. 8b) [91–93].

Conclusion

In this study, the authors focused on LAB isolation from different milk sources and their potential probiotic features. All isolated strains showed noteworthy results in terms of low pH3, auto- and coaggregation, hydrophobicity, cell adhesion capability, which are essential parameters for probiotics. Furthermore, all the strains were tested safety profiles, which were antibiotic susceptibility, hemolytic, gelatinase and DNAse activity. And the results show that all strains have safety characteristics. Especially W. confusa MH24 showed high hydrophobicity and Caco-2 cell adhesion. All results indicate that these LAB strains, especially W. confusa MH24 could be potential probiotics. Additional in vivo research is required for further assessment.

Author contributions

All authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by Melda ONUR, and Harun ÖNLÜ. The first draft of the manuscript was written by Harun ÖNLÜ, and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Data availability

The authors declare that the data supporting the findings of this study are available within the paper and its Supplementary Information files. Should any raw data files be needed in another format they are available from the corresponding author upon reasonable request.

Declarations

Conflict of interest

The authors declare no competing interests.