Isolation, identification and potential probiotic characterization of lactic acid bacteria from Thai traditional fermented food

The probiotic potential of lactic acid bacteria (LAB) isolated from Thai traditional fermented food was investigated. Forty-two samples were collected from four markets in Phra Nakhon Si Ayutthaya Province. Out of 50 isolated LAB, 6 (a3, f4, f8, K1, K4 and K9) obtained from pla-ra and bamboo shoot pickle samples showed high tolerance to gastrointestinal tract conditions. These isolates were selected to identify and characterize their probiotic properties. Isolate a3 was identified as Weissella thailandensis, isolates f4 and f8 were identified as belonging to Enterococcus thailandicus and isolates K1, K4 and K9 were determined as Limosilactobacillus fermentum. All six LAB exhibited high autoaggregation ability (93.40–95.01%), while W. thailandensis isolate a3 showed potential for coaggregation in almost all the pathogenic bacteria tested. Cell-free supernatant (CFS) obtained from all isolates did not inhibit Staphylococcus aureus. CFS derived from L. fermentum isolate K4 showed the most efficient antimicrobial activity, in particular against Gram-negative bacteria, while L. fermentum isolate K4 presented high surface hydrophobicity in the presence of xylene and n-hexane. All LAB isolates were found to be resistant to clindamycin and nalidixic acid, whereas E. thailandicus isolate f8 exhibited resistance to most of the antibiotics tested. L. fermentum isolate K4 showed promise as a suitable probiotic candidate for future applications in the food industry due to tolerance to gastrointestinal tract conditions with high surface hydrophobicity and inhibited most of the pathogens tested.


Sample collection
Forty-two samples of Thai traditional fermented food such as bamboo shoots and ginger pickles, phak-gard-dorng, pla-ra and Thai fermented pork sausage were collected from four markets in Phra Nakhon Si Ayutthaya Province as Grand, Chao Phrom, DD and 4 Right Development. All samples were stored at 30 ℃ until required for use.

Isolation of lactic acid bacteria (LAB)
LAB were isolated on MRS agar (HiMedia, India) supplemented with bromocresol purple as an indicator and incubated at 37 ℃ for 24-48 h. Single bacterial colonies that produced a yellow zone were selected according to morphological differences. Gram-staining and catalase tests were carried out for 135 colonies. Only Gram-positive and catalase negative colonies were sub-cultured in MRS broth and kept in glycerol before experimental use.

Tolerance to simulated gastric and intestinal fluid
Tolerance to simulated gastric and intestinal fluid was evaluated according to the method of Arboleya et al. [15]. Briefly, cells of isolated LAB from 16 h MRS culture were harvested by centrifugation (7,500 × g at 4 ℃ for 10 min), washed twice with phosphate buffered saline (PBS) pH 7.4 and resuspended in 500 µL of the same buffer. A 100 µL aliquot of cell suspension was added to 900 µL of simulated gastric juice (NaCl 125 mM, KCl 7 mM, NaHCO 3 45 mM and pepsin 3 g/L), with final pH adjusted to 2.5 or bile juice (45 mM NaCl, 1 g/L pancreatin and 3 g/L bile salt, adjusted to pH 8.0). Suspensions were then incubated at 37 ℃ for 2 h under anaerobic condition. The samples were aliquoted at time 0 and after 2 h. These samples were serially diluted in 0.85% NaCl and viable cells were counted by the spread plate method using MRS agar, and incubated at 37 ℃ for 48 h. The percentage of LAB survival was determined as follows: Survival (%) = log CFU of viable cell after 2 h of incubation × 100 Survival at more than 75% was considered tolerant to gastrointestinal conditions.

Genotypic identification using 16S ribosomal DNA
Bacteria showing high tolerance to gastrointestinal conditions were molecularly identified using polymerase chain reaction (PCR), and pure cultures were sequenced by Macrogen Inc (Seoul, Republic of South Korea). Primers 785F 5' (GGATTAGATACCCTGGTA) 3' and 907R 5' (CCGTCAATTCMTTTRAGTTT) 3' were used to amplify the 16S rDNA gene. The PCR reaction was carried out with 20 ng of genomic DNA as the template in a 30 µL reaction mixture using an EF-Tag (SolGent, Korea) as follows: activation of Taq polymerase at 95 ℃ for 2 min, followed by 35 cycles of 95 ℃ for 1 min, 55 ℃, and 72 ℃ for 1 min each, with a final 10-min step at 72 ℃. The amplification products were purified with a multiscreen filter plate (Millipore Corp., Bedford, MA, USA). Sequencing reaction was performed using a PRISM BigDye Terminator v3.1 Cycle Sequencing Kit. DNA samples containing the extension products were added to Hi-Di formamide (Applied log CFU of initial viable cell Biosystems, Foster City, CA, USA). The mixture was incubated at 95 ℃ for 5 min, followed by 5 min on ice, and then analyzed by an ABI Prism 3730XL DNA analyzer (Applied Biosystems, Foster City, CA, USA). A homology search was performed using BLAST software, and 16S rDNA sequences of the isolated LAB were deposited in GenBank.

Autoaggregation and coaggregation assays
Autoaggregation and coaggregation assays were determined according to Re et al. [16]. Isolated LAB were cultured at 37 ℃ for 16 h in MRS broth. The cells were harvested by centrifugation (7,500 × g at 4 ℃ for 10 min), washed twice, resuspended in PBS buffer pH 7.4 and adjusted for optical density at 600 nm of 1.0. The suspensions were then incubated at 37 ℃ for 4 h and 0.1 mL of the upper suspension was aliquoted and added to 1.9 mL PBS. The suspensions were measured at OD600 nm, with percentage autoaggregation evaluated as follows: Autoaggregation (%) = 1 -(OD 600 of upper suspension) × 100 Coaggregation of isolated LAB and 8 pathogenic bacteria were investigated. Cell suspensions were prepared similarly to the autoaggregation evaluation. A 1 mL aliquot of isolated LAB was mixed with 1 mL of pathogenic strain and incubated at 37 ℃ for 4 h. The mixtures were measured at OD 600 at time 0 and 4 h. Coaggregation percentage was calculated as follows: Coaggregation (%) = OD 600 of initial cell -OD 600 of suspension after 4 h incubation × 100 2.3.4. Surface hydrophobicity Surface hydrophobicity was measured according to Guan et al. [17]. The isolated LAB was grown in MRS broth at 37 ℃ for 24 h and the cell pellet was harvested by centrifugation. The cells were resuspended and washed twice in PBS buffer. The OD 600 of suspension was adjusted to 0.5 ± 0.05, then 4 mL of cell suspension were mixed with 0.4 mL xylene or hexane, with PBS buffer used as the control. The aqueous phase was determined at OD 600 after 1 h of incubation at room temperature. Surface hydrophobicity was calculated as follows: Surface hydrophobicity (%) = OD 600 of control -OD 600 of suspension after 1 h incubation ×100

Antimicrobial activity
Antibacterial activity of cell-free supernatant (CFS) of isolated LAB was investigated against eight pathogenic bacteria by the agar well diffusion method [18]. Overnight cultures of isolated LAB were collected by centrifugation (7,500 × g at 4 ℃ for 10 min). The CFS was divided into two groups. Group 1 was unadjusted for pH, while group 2 was adjusted to pH 6.5 using 1 N NaOH to eliminate inhibition from pH reduction. The CFS samples were filter-sterilized through 0.45 µm pore size filters. h. The OD 600 was adjusted to 1 and spread on nutrient agar. Then, a hole with a diameter of 5 mm is punched aseptically with a sterile cork borer and 30 µL of CFS were added to each hole and incubated at 37 ℃ for 24 h. The diameter of the inhibition zone was measured, with Streptomycin (10 µg/mL) used as the control.

Antibiotic susceptibility
The antibiotic susceptibility of isolated LAB was assessed using the agar well diffusion method according to Tagg and McGiven [18]. The isolated LAB was spread on MRS agar and antibiotic solution containing penicillin (10 µg), ampicillin (10 µg), amoxicillin (30 µg), clindamycin (2 µg) and nalidixic acid (30 µg) was added to each well. The plates were incubated at 37 ℃ for 24 h and the inhibition zones were measured. Results were interpreted according to the cut-off levels proposed by Charteris et al. [19].

Statistical analyses
All experiments were carried out in triplicate. Statistical analyses were performed using SPSS Version 16.0 (SPSS Inc., USA). Duncan's multiple range test was used to determine the level of significant differences (P < 0.05).

Isolation of LAB
Fifty colonies were isolated from 42 samples on MRS agar. Among these Gram-positive and catalase-negative strains, 26 strains were isolated from pla-ra obtained from all four markets, 19 strains were obtained from bamboo shoot pickles obtained from Chao Phrom and four Right Development markets, 3 strains from phak-gard-dorng obtained from Chao Phrom market and 2 strains from Thai fermented pork sausage obtained from four Right Development market. LAB were not observed from ginger pickle (Table 1). Results showed that pla-ra and bamboo shoot pickle were the main sources of LAB, concurring with Miyashita et al. [20] [21] also reported that Tetragenococcus, Halanaerobium and Lactobacillus were the dominant bacteria in pla-ra from Northeastern Thailand, while Thakur et al. [22], Lindayani et al. [23] and Mir et al. [24] isolated Lactobacillus spp. and Streptococcus spp. from bamboo shoot pickle and L. plantarum from fermented vegetables. All isolates were used to further investigate intolerance to simulated gastrointestinal conditions.

Tolerance to simulated gastric juice and bile salt
High tolerance capabilities against simulated gastric juice and bile salt were associated with the bilayer membrane structure, which enables easy tolerance of inverse condition [25]. A 75% survival rate of LAB after 2 h of incubation in simulated gastric juice and bile salt was considered the cut-off level for tolerance. Fifty isolates from Thai traditional fermented food showed varied tolerance to simulated gastric juice and bile salt. Most isolates exhibited low survival in simulated gastric juice, and these isolates were not further tested. The remaining three isolates (a3, f4 and f8) from pla-ra and three isolates (K1, K4 and K9) from bamboo shoot pickle showed high tolerance to acidic condition with survival rate of 78.47-93.01%. Isolate f4 showed the strongest tolerance to simulated gastric juice in comparison with the other isolates ( Table 2). Resistant to bile salt is a significant property of a potential probiotic because the small intestine and colon have high concentrations of bile salt that is toxic to cells. Most LAB isolates displayed tolerance to bile salt at higher than 75% survival, except isolates H4, H6, K5, K6, K8, S2, S4, t4, t5 and q2. Out of the 50 isolates, 6 were selected for identification with their higher tolerances to gastrointestinal tract conditions.

Genotypic identification using 16S ribosomal DNA.
Isolate a3 was identified as Weissella thailandensis, whereas isolates f4 and f8 were identified as belonging to Enterococcus thailandicus (Table 3). W. thailandensis and E. thailandicus were previously isolated from pla-chom and fermented sausage (mum) in Thailand, respectively [26,27]. To the best of my knowledge, this is the first report of isolation of W. thailandensis and E. thailandicus from pla-ra. Isolates K1, K4 and K9 were determined as belonging to Limosilactobacillus fermentum, consistent with Chen et al. [28] and Behera and Balaji [29] who reported that L. fermentum was isolated from jiang-sun (fermented bamboo shoot) in Taiwan and dry fermented bamboo shoot in India, respectively. The high potential tolerance in simulated gastric juice of E. thailandicus isolates f4, f8 and L. fermentum isolates K1, K4 and K9 concurred with Hanghshenas et al. [30] who reported that Enterococcus and Lactobacillus strains showed high (51-82%) and moderate survival rates (71-76%) at low pH values, respectively. This finding suggested that these 6 isolates have the potential to survive in the human gastrointestinal tract and can likely survive passage through the stomach and small and large intestines.

Autoaggregation and coaggregation properties.
The autoaggregation property of probiotics plays a major role in the colonization of host epithelial cells, resulting in improvement of host defense mechanisms against the gut [31]. Autoaggregation at higher than 40% is required for a potential probiotic strain [32]. Six LAB exhibited autoaggregation ranging from 93.40-95.01%, demonstrating good adhesion ability (Table 4). There were no significant differences in autoaggregation among the six isolates. Autoaggregation of E. thailandicus isolates f4 and f8 and L. fermentum isolates K1, K4 and K9 were 93.40-94.83% and 94.57-95.01%, respectively and higher than those of Enterococcus spp. from fermented sausage (29.0-67.0%) [33] and Lactobacilli from cocoa fermentation (31.18%) [34]. Results implied that these isolates can protect themselves from environmental stresses, and have the capacity to survive in gastrointestinal tract conditions [35]. Coaggregation is related to the ability to interact with other bacteria to form a barrier that prevents colonization by pathogenic bacteria. W. thailandensis isolate a3 showed higher coaggregation with other pathogenic bacterial LAB isolates, except with E. faecalis and K. pneumoniae, while E. thailandicus isolates f4 and L. fermentum isolate K4 and K9 had low coaggregation ability against pathogenic bacteria. All LAB isolates showed the highest coaggregation with S. aureus (10.15-38.50%), while L. fermentum K1, K4 and K9 aggregated with E. coli (5.56-8.62%) at a lower level than L. plantarum aggregation with E. coli O157:H7 (21.49-27.25%) [36]. Results indicated that isolate a3 prevented colonization of most pathogenic bacteria on the intestine. Lowercase letters indicate significant differences in percentage coaggregation among the 6 isolates against the same pathogenic bacteria (P < 0.05).
Capital letters indicate significant differences in percentage coaggregation of each isolate against 8 pathogenic bacteria (P < 0.05). * indicates no significant difference in percentage autoaggregation or coaggregation (P > 0.05).

Surface hydrophobicity
Significant differences in surface hydrophobicity values were found between LAB, even within the same species. L. fermentum isolate K4 showed high surface hydrophobic values as 39.15% and 78.09% with xylene and n-hexane, respectively (Figure 1). These values were lower and higher than Kocabay et al. [37], who reported surface hydrophobicity in the presence of xylene at 63.94% and nhexane 57.59%, respectively. Surface hydrophobic values of L. fermentum isolate K4 also differed from Tamang and Tamang [38] who reported that L. fermentum isolated from fermented bamboo products of Arunachal Pradesh in India did not have a hydrophobic nature. Results indicated that L. fermentum isolate K4 may favor strong colonization through interaction with the organic mucin layer of the gut and play an important role in the adhesion of bacteria to epithelial cells [39]. E. thailandicus isolate f4 had surface hydrophobicity value of 42.83% in the presence of n-hexane, higher than reported by Santos et al. [40] who noted that E. faecium EM485 and E. faecium EM 925 had surface hydrophobicity values of 8.18% and 11.33% in the presence of n-hexadecane, respectively. Although, both xylene and n-hexane were non-polar solvents, most LAB isolates had higher surface hydrophobicity values with n-hexane than xylene, except W. thailandensis isolate a3 and L. fermentum isolate K9 that had surface hydrophobicity of 0% in the presence of both organic solvents. This finding differed from Xiong et al. [41] who reported that W. confusa possessed high surface hydrophobicity. n-hexane. Lowercase letters indicate significant differences in surface hydrophobicity percentages among 6 isolates in the presence of xylene or n-hexane (P < 0.05). Capital letters indicate significant differences in surface hydrophobicity percentages of each isolate in the presence of xylene and n-hexane (P < 0.05). * indicates no significant difference in surface hydrophobicity percentage (P > 0.05).
Capital letters indicate significant differences in diameter of the inhibition zone of CFS obtained from each isolate against the 8 pathogenic bacteria (P < 0.05). * indicates no significant difference in the diameter of the inhibition zone (P > 0.05).

Antibiotic susceptibility.
All LAB isolates were resistant against clindamycin and nalidixic acid, as antibiotics that inhibited protein synthesis and nucleic acid synthesis, respectively. They were also susceptible or moderately susceptible to ampicillin (Table 6). W. thailandensis isolate a3 and E. thailandicus isolates f4 and f8 exhibited resistance to penicillin. W. thailandensis isolate a3 had a different antibiotic susceptibility profile than the profile reported by Kamboj et al. [44]. They found that Weissella spp. exhibited susceptibility to penicillin and clindamycin, while E. thailandicus isolates f4 and f8 had different results from those reported by Wu et al. [45] who noted that E. thailandicus TC1 was found to be susceptible to penicillin. E. thailandicus isolate f8 showed high potential for resistance to 4 out of 5 antibiotics, whereas L. fermentum isolates K1, K4 and K9 only exhibited resistance to clindamycin and nalidixic acid. This result differed from EgervÄrn et al. [46] who indicated that L. plantarum exhibited low minimum inhibitory concentration (MICs) for ampicillin and clindamycin.

Conclusions
Pla-ra and bamboo shoot pickle from markets in Phra Nakhon Si Ayutthaya were identified as major sources for isolation of LAB. Six isolating of LAB showing high tolerance to gastrointestinal tract conditions were identified as W. thailandensis, E. thailandicus and L. fermentum. L. fermentum isolate K4 showed outstanding probiotic properties such as tolerance to simulated gastric juice and bile salt with survival of 90.75 and 94.64%, high surface hydrophobicity, inhibition of various pathogenic bacteria and resistance to clindamycin and nalidixic acids. Isolate K4 showed potential as a candidate for further in vivo investigations to evaluate potential health benefits and safety assessment of probiotics for applications in the food industry.