Probiotic Properties of Lactic Acid Bacteria with High Conjugated Linoleic Acid Converting Activity Isolated from Jeot-Gal, High-Salt Fermented Seafood

Conjugated linoleic acid (CLA) isomers are potent health-promoting fatty acids. This study evaluated the probiotic properties of 10 strains of high CLA-producing lactic acid bacteria (LAB) isolated from Jeot-gal, a high-salt, fermented seafood. Two isolates, Lactiplantibacillus plantarum JBCC105683 and Lactiplantibacillus pentosus JBCC105676, produced the largest amounts of CLA (748.8 and 726.9 μg/mL, respectively). Five isolates, L. plantarum JBCC105675, L. pentosus JBCC105676, L. pentosus JBCC105674, L. plantarum JBCC105683, and Lactiplantibacillus paraplantarum JBCC105655 synthesized more cis-9, trans-11-CLA than trans-10, cis-12-CLA (approximately 80:20 ratio). All the strains survived severe artificial acidic environments and showed antimicrobial activity and strong adhesion capability to Caco-2 cells as compared to the commercial strain Lactocaseibacillus rhamnosus GG. Among them, Pediococcus acidilactici JBCC105117, L. paraplantarum JBCC105655, and L. plantarum JBCC105683 strongly stimulated the immunological regulatory gene PMK-1 and the host defense antimicrobial peptide gene clec-60 in Caenorhabditis elegans. Moreover, three strains showed a significant induction of tumor necrosis factor-α, interleukin (IL)-1β, IL-6, IL-12, and IL-10 production in RAW 264.7 macrophages, indicating that they were promising candidates for probiotics with high CLA-converting activity. Our results indicate that the newly isolated CLA-producing LAB might be useful as a functional probiotic with beneficial health effects that modulate the immune system.


Introduction
Jeot-gals are traditional Korean fermented, high-salt seafoods prepared from various fishes, shrimps, oysters, fish eggs, and fish intestines with final salt concentrations of up to 30% (w/w) [1,2]. During fermentation for several months, or several years for some Jeot-gals, the unique microbial consortium results in lots of aerobic and anaerobic microorganism developments, including lactic acid bacteria (LAB) during the maturing period [3]. LAB account for 70~90% and consist mainly of the genus Lactobacillus and the genus Weissella [4]. Many studies have shown that functional strains, such as halophillic and halotolerant bacteria [5][6][7][8] and γ-aminobutyric acid or bacteriocin isolated from Jeotgals, are indications of useful microbial sources of LAB [9,10]. Hence, the LAB from Jeot-gals were expected to possess some novel and useful properties as marine environments are different from terrestrial environments.
CLA is a mixture of the positional and geometric isomers of linoleic acid (LA, cis-9, cis-12-octadecadienoic acid) synthesized naturally in the human colon. It usually exists in two biologically active isomers, cis-9, trans-11-CLA and trans-10, cis-12-CLA, which evaluated the probiotic activities of these strains, including their ability to survive under extreme conditions, their antimicrobial activity, their adhesion to human intestinal epithelial cells, and their immune modulation in vivo by C. elegans as well as in vitro by inflammatory cytokine assay.

Chemicals and Media
All the standard samples of LA, CLA (cis-9, trans-11-CLA and trans-10, cis-12-CLA), heptadecanoic acid as an internal standard, and the antibiotics were purchased from Sigma-Aldrich (St. Louis, MO, USA). The de Man, Rogosa, and Sharpe (MRS) medium for LAB cultivation and nutrient broth, brain heart infusion broth, and the PP medium for pathogenic microorganisms were obtained from BD Difco Laboratories (Detroit, MI, USA). All the other chemicals and reagents used were of analytical grade. Lactocaseibacillus rhamnosus GG (ATCC53103) was used as a control for the probiotic characterization.

Screening and Identification of LAB with LA to CLA Conversion Activity
Rapid bacterial screening for the LAB converting of LA to CLA was conducted in two steps: (1) LA resistance [34], with some modifications, and (2) UV spectral scan analysis. First, approximately 10 g of Jeot-gal samples were homogenized at 500 rpm in a rotor-stator homogenizer (U-TRON LAB/P, ESYNDMT, Sasang-gu, Busan Korea) and serially diluted 10-fold with saline (0.85%). The diluted homogenized samples were plated on MRS agar plates containing 0.2% (v/v) LA with 1% (v/v) Tween 80. After the inoculated plates were incubated at 30 • C for 48 h, only strains showing large and wellformed colonies were selected for the next steps. The selected strains were cultured in 5 mL MRS broth containing LA (5 mg/mL) with 1% (v/v) Tween 80 and scanned using a UV/Visible spectrophotometer (Beckman Coulter, Brea, CA, USA). Second, fatty acids in the culture supernatant were extracted using isopropanol (99.9%)/hexane extraction to confirm the conversion of CLA from LA by UV spectral scan analysis, as described by Liu et al. [11]. A 1 mL sample and 2 mL of isopropanol were mixed vigorously for 1 min. Next, 1.5 mL of hexane was added and vortexed for 2 min at room temperature. The upper layer obtained after centrifugation (4000× g for 5 min) was used for downstream analysis. The fatty acid extraction samples obtained above were scanned from 200 to 300 nm using a Beckman DU 800 UV/Visible spectrophotometer (Beckman, Fullerton, CA, USA) according to Liu et al. [11]. A characteristic absorption peak at 228-235 nm was expected, indicating the possible presence of CLA with conjugated double bonds in the extraction samples. Absorbance was measured in 1 cm quartz cuvettes at room temperature, and spectral graph data and absorbance values of the samples were obtained. For identification of the obtained strains, 16s rRNA sequencing was used as described by Kim and Baik [35].
The CLA isomers were methylated for GC analysis and quantified using methyl esters prepared by acid-catalyzed methylation methods by gas chromatographic analysis. The hexane solvent in the extracted samples was dried with nitrogen gas, and then the residue was added to 1 mL of the internal standard solution prepared by dissolving heptadecanoic acid in isooctane (2,2,4-trimethylpentane) to produce a final concentration of 1 mg/mL. After 4 mL of 0.5 N sodium methoxide (CH 3 ONa, 2 g NaOH in 100 mL methanol, w/v) was added, the sample was heated at 80 • C for 5 min and cooled to room temperature. Next, 5 mL of 14% boron trifluoride-methanol (BF 3 /CH 3 OH; w/v) was added, and the sample was heated at 80 • C for 10 min. After cooling to room temperature, the methylated sample was extracted with 3 mL of isooctane by vortexing for 30 s. Subsequently, 5 mL of saturated sodium chloride was added and vortexed for 30 s. The upper layer was dried over anhydrous sodium sulfate and analyzed by GC (Hewlett-Packard 5890, Agilent Technologies Inc., Santa Clara, CA, USA) equipped with a flame ionization detector and a Supelco SP-2560 fused silica capillary column (100 m × 0.25 mm, i.d., 0.2 µm film thickness; Supelco Inc., Bellefonte, PA, USA). Fatty acid methyl esters were separated Microorganisms 2021, 9, 2247 4 of 16 as described by Beppu et al. [36]. Heptadecanoic acid was used as an internal standard, and the CLA peak was identified by comparing the retention time to the standards.

Acid and Bile Tolerance of Isolated Strains
Each LAB (8-9 log colony-forming units [CFU]/mL) was harvested by centrifugation (4000× g at 4 • C for 10 min) and suspended in an equal volume of the MRS broth, which was adjusted to pH 2.0 with 5.0 M HCl to investigate acid tolerance. After incubation at 37 • C for up to 2 h, the survival rate was evaluated by determining the viable cell counts of the samples after serial dilution in sterilized water and incubation at 37 • C for 48 h. For bile tolerance, 1% of the cell suspension was inoculated onto plates supplemented with 0, 0.3%, 1%, 3%, and 5% (w/v) oxgall (Sigma-Aldrich, St. Louis, MO, USA), and the survival rate was evaluated by determining the viable cell counts as described above.

Antimicrobial Activity
The antimicrobial activities against various bacterial pathogens were evaluated against pathogenic microorganisms, such as

Cell Adherence Assay
The adherence assay of LAB to Caco-2 cells in vitro was carried out as previously described [35], with minor modifications. The Caco-2 intestinal epithelial cells (5 × 10 4 cells/well) were cultured in MEM (Lonza, Basel, Switzerland) containing 10% heat-inactivated fetal bovine serum (Gibco, Grand Island, NY, USA) in a 5% CO 2 humidified incubator (NU-5800; NuAire, Caerphilly, UK). The Caco-2 cell monolayers were incubated with LAB (8-9 log CFU/mL) for 2 h at 37 • C and washed with PBS to remove unadhered bacteria and treated with 0.1 mL of 0.05% (v/v) Triton X-100 to extract the adhered bacteria. The cells were then serially diluted, plated on MRS agar, and incubated for 48-72 h at 37 • C for bacterial adhesion calculation based on the number of viable bacteria in the original suspension and the cell lysates and expressed as log CFU/mL.

In Vivo Caenorhabditis Elegans Assay for Immune Response
For the in vivo assay of the LAB, the C. elegans CF512 fer-15(b26)II;fem-1(hc17)IV (fer-15;fem-1 worms) strain was routinely maintained on nematode growth medium (NGM) plates seeded with Escherichia coli OP50. The LAB were sub-cultured (9 log CFU/mL) three times before use and exposed to the C. elegans on NGM plates containing nystatin for 5 days. Ten worms were randomly picked, washed twice with M9 buffer, and placed on brain heart infusion plates containing kanamycin and streptomycin. After exposure to gentamycin (5 µL of a 25 µg/mL solution) for 5 min, the worms were washed three times with M9 buffer and then pulverized using a pestle (Kontes Glass Inc., Vineland, NJ, USA) in a 1.5 mL Eppendorf tube containing M9 buffer supplemented with 1% Triton X-100. After serial dilution in the M9 buffer, the worm lysate was plated on the MRS agar (pH 5.0) plates, incubated for 48 h at 37 • C, and then counted for live bacterial cells. The results were compared with those obtained using L. rhamnosus GG as a positive control. To elucidate the transcriptional host responses [pmk-1::GFP in wild-type N2], the nematodes were induced by exposure to selected LAB strains for 24 h. The animals were mounted on glass slides with 2% agarose pads, anesthetized with 10 mM NaN 3 , and quickly visualized using an Ax-ioImager Z1 fluorescence microscope (Zeiss, Oberkochen, Germany). Quantitative reverse transcription polymerase chain reaction (qRT-PCR) analysis was performed to evaluate the impact of conditioning with the LAB strains on clec-60, a gene associated with nematode immune responses. Transcript levels were measured in young adult C. elegans N2 Bristol Microorganisms 2021, 9, 2247 5 of 16 wild-type, conditioned with the LAB strains for 24 h. The total RNA from the worms was quickly isolated following the TRIzol (Invitrogen, Waltham, MA, USA) method and purified using the RNeasy Mini Kit (QIAGEN, Germantown, MD, USA), including an on-column DNase digestion with RNase-free DNase (QIAGEN, Germantown, MD, USA). After RNA isolation, the total RNA (50 ng) was used for quantitative real-time PCR (qRT-PCR) using the SuperScript III Platinum SYBR green one-step qRT-PCR kit (Invitrogen, Waltham, MA, USA). The qRT-PCR was performed using the StepOne™ Real-Time PCR System (Applied Biosystems, Foster City, CA, USA). Primers were designed using Primer3Input software (v0.4.0) and are listed as follows: clec-60 (5 -ACGGGCAAGTTATTGGAGAG-3 and 5 -ACACGGTATTGAATCCACGA-3 ) and snb-1 (5-CCGGATAAGACCATCTTGACG-3 and 5-GACGACTTCATCAACCTGAGC-3. The control gene snb-1 was used to normalize the gene-expression data.

qRT-PCR for Inflammatory Cytokine Assay
RAW 264.7 cells (5 × 10 4 cells/well) were seeded into 96-well plates and heat-killed LAB (100 • C, 30 min) suspended in PBS (pH 7.4) were added to the wells at a final concentration of 1 × 10 8 cells/mL. The effects of the strains on immune cell proliferation were determined using the MTT assay. RAW 264.7 cells were stimulated with 1 mg/mL lipopolysaccharide (LPS) and the LAB. Total cellular RNA was extracted from RAW 264.7 cells using a NucleoSpin RNA kit (Macherey-Nagel, Düren, Germany). The total RNA (500 ng) was reverse-transcribed to cDNA using a Quick reverse transcription system (ReverTraAce qPCR RT Master Mix; TOYOBO, Osaka, Japan). The cDNAs were amplified by PCR using SYBR Green Mastermix (TOYOBO, Osaka, Japan) and specific primers (Supplementary Table S1). The qRT-PCR was performed as follows: initial denaturation at 95 • C for 2 min, 39 cycles at 95 • C for 15 s, 57 • C for 20 s, and 72 • C for 30 s (CFX96 Touch TM Real-Time PCR Detection System, Bio-Rad Laboratories, Hercules, CA, USA). Target-gene expression was analyzed as the relative quantity of the target gene using the 2 −∆∆CT threshold cycle method compared to the glyceraldehyde-3-phosphate dehydrogenase as the reference gene used to normalize the gene expression data.

Statistical Analysis
All samples were evaluated in triplicate and the data are expressed as the mean ± standard deviation. One-way analysis of variance was performed using SPSS software (version 20.0; SPSS, Inc., Chicago, IL, USA). Post hoc tests were performed using Duncan's test for multiple comparisons at a significance level of 0.05. Principal component analysis (PCA) was conducted and visualized using MetaboAnalyst version 5.0 software (20 May 2021, https://www.metaboanalyst.ca).

Screening of CLA-Producing LAB
An LA concentration of 0.5% at 37 • C inhibited the growth of the LA-tolerant LAB. However, when the LA concentration in the MRS agar medium decreased to 0.2%, the LAB formed colonies of various sizes surrounded by a halo on the medium (Supplementary Figure S1). The isolates were classified according to halo size: 63 were large halo producers, 38 were medium producers, and 89 were low producers and subjected to additional screening by UV spectral scan analysis. Among the 63 selected isolates that formed large halos with large colonies, only 24 isolates exhibited characteristic absorption peaks at 228-235 nm, indicating the presence of conjugated double bonds (Supplementary Figure S2). When the CLA productivity was measured at 233 nm, the 24 strains which exhibited a characteristic peak could produce above 50.0 µg/mL of CLA and the 10 LAB strains of JBCC105611, JBCC105117, JBCC105675, JBCC105634, JBCC105676, JBCC105674, JBCC105683, JBCC105655, JBCC105645, and JBCC105686 that produced more than 80.0 µg/mL CLA (from 86.0 to 125.5 µg/mL) were ultimately selected for further experiments (Table 1).

Isomeric Composition Analysis of CLA by GC
To analyze the geometric isomeric ratio of cis-9, trans-11-CLA, and trans-10, cis-12-CLA, which show significantly different biological activities [37], GC analysis was performed, which revealed that all the selected strains showed different geometric isomeric ratios, as shown in Table 1. The Lactobacillus plantarum JBCC105683 and L. pentosus JBCC105676 strains produced the largest amounts of CLA (748.8 and 726.9 µg/mL, respectively) in the presence of 600 µg/mL of substrate, showing high conversion rates of 124.8% and 121.2%, respectively. They also showed high conversion rates, of approximately 40%, in the presence of 1000 µg/mL of LA. The lower conversion capacity to CLA at higher LA concentrations can be ascribed to the antimicrobial ability of LA [38]. In addition, different CLA isomer ratios were observed for each strain. In particular, L. paraplantarum JBCC105611, P. acidilactici JBCC105117, L. paraplantarum JBCC105634, L. plantarum JBCC105645, and L. mesenteroides JBCC105686 produced similar proportions of the two isomers (an approximately 50:50 ratio), whereas L. plantarum JBCC105675, L. pentosus JBCC105676, L. pentosus JBCC105674, L. plantarum JBCC105683, and L. paraplantarum JBCC105655 synthesized more cis-9, trans-11-CLA than trans-10, cis-12-CLA (approximately 80:20 ratio).

pH and Bile Salt Tolerance of CLA-Converting LAB
As shown in Table 2, the 10 selected LAB strains showed relatively high survival rates at pH 2.0 for 2 h, even though the survival rate declined after 2 h. L. paraplantarum JBCC105634, L. paraplantarum JBCC105655, and L. pentosus JBCC105674 showed the highest acid tolerance. The bile tolerance of the selected CLA-producing LAB is shown in Table 3. Although the cell viability of the CLA-producing LAB slightly declined with the increasing bile concentration, nearly all the selected strains grew in 5% bile salts.

Antimicrobial Activity
Previous studies have suggested that LAB in the gut microflora play an important role as a barrier against pathogenic strains in the GI tract and inhibit attachment to cultured uroepithelial cells, intestinal cells, and mucus [39]. As shown in Table 4, all the selected LAB strains inhibited the growth of all the selected pathogenic strains (Supplementary Figure S4). In particular, strong antimicrobial activities against S. aureus were observed for all the selected LAB, showing nearly identical results to the control strain, L. rhamnosus GG. However, the capacity to inhibit pathogenic bacteria, such as E. coli and B. cereus, by all the selected LAB, was higher than that of the control strain, indicating their suitability as probiotics. Other pathogenic bacteria, including S. epidermidis, P. aeruginosa, and P. putida also showed similar inhibition activities compared to the control strains. The antimicrobial activity of the CLA-producing LAB isolated in this study inhibited both gram-positive and gram-negative pathogens more effectively than the control strain.

Intestinal Adhesion Ability of LAB Strains to Caco-2 Cells
To estimate the adhesion capacity of the high CLA-producing LAB, we examined their adhesion to the human epithelial cell line Caco-2. As shown in Figure 1, the 10 selected bacterial cells showed variable adhesion capacity depending on the strain. The isolates L. plantarum JBNU105683, L. paraplantarum JBCC105655, L. plantarum JBNU105645, and L. paraplantarum JBCC105634 showed slightly higher adhesion capacities to human epithelial cells (8.20-8.46 log CFU/mL) (p > 0.05) at levels comparable to those of the commercial probiotic strain, L. rhamnosus GG (8.17 log CFU/mL). L. plantarum JBCC105675, L. mesenteroides JBCC105686, L. pentosus JBCC105676, L. pentosus JBCC105674, and L. paraplantarum JBCC105611 exhibited intermediate (7.38-7.81 log CFU/mL) capacities. The P. acidilactici JBCC105117 strains exhibited low adhesion capacity (6.89 log CFU/mL). Thus, the adherence of the isolates in the GI tract may extend the residence time in the host. epithelial cells (8.20-8.46 log CFU/mL) (p > 0.05) at levels comparable to those of the commercial probiotic strain, L. rhamnosus GG (8.17 log CFU/mL). L. plantarum JBCC105675, L. mesenteroides JBCC105686, L. pentosus JBCC105676, L. pentosus JBCC105674, and L. paraplantarum JBCC105611 exhibited intermediate (7.38-7.81 log CFU/mL) capacities. The P. acidilactici JBCC105117 strains exhibited low adhesion capacity (6.89 log CFU/mL). Thus, the adherence of the isolates in the GI tract may extend the residence time in the host.

Effects of LAB Strains on Growth and Cytokine Secretion in RAW 264.7 Cells
To determine the toxicity of the isolates, RAW 264.7 cells were treated with the LAB for 72 h and the cell viability was evaluated by the MTT assay. The treatment with L. paraplantarum JBCC105611, L. paraplantarum JBCC105634, and L. mesenteroides JBCC105686 resulted in similar cell viabilities compared to L. rhamnosus GG and other isolates, with no effects on cell toxicity (Supplementary Figure S5). This indicates that none of the selected strains in this study were cytotoxic to RAW 264.7 cells. We demonstrated the production of different pro-and anti-inflammatory responses in murine macrophage RAW 264.7 to examine the immune-modulating functions of the heat-killed LAB. As shown in Figure 2, some strains showed similar cytokine production patterns, except for transforming growth factor-β. The three selected strains, L. paraplantarum JBCC105655, L. plantarum JBCC105683, and L. pentosus JBCC105676, induced a significantly higher secretion of proinflammatory cytokines TNF-α, IL-1β, IL-12, and IL-6 in RAW 264.7 cells compared to the positive control L. rhamnosus GG. IL-10, an anti-inflammatory cytokine, was also stimulated by isolated LAB compared to L. rhamnosus GG, except for the strains L. paraplantarum JBCC105634 and L. plantarum JBCC105645. In particular, L. plantarum JBCC105655, L. plantarum JBCC105683, and P. acidilactici JBCC105117 highly stimulated

Effects of LAB Strains on Growth and Cytokine Secretion in RAW 264.7 Cells
To determine the toxicity of the isolates, RAW 264.7 cells were treated with the LAB for 72 h and the cell viability was evaluated by the MTT assay. The treatment with L. paraplantarum JBCC105611, L. paraplantarum JBCC105634, and L. mesenteroides JBCC105686 resulted in similar cell viabilities compared to L. rhamnosus GG and other isolates, with no effects on cell toxicity (Supplementary Figure S5). This indicates that none of the selected strains in this study were cytotoxic to RAW 264.7 cells. We demonstrated the production of different pro-and anti-inflammatory responses in murine macrophage RAW 264.7 to examine the immune-modulating functions of the heat-killed LAB. As shown in Figure 2, some strains showed similar cytokine production patterns, except for transforming growth factor-β. The three selected strains, L. paraplantarum JBCC105655, L. plantarum JBCC105683, and L. pentosus JBCC105676, induced a significantly higher secretion of pro-inflammatory cytokines TNF-α, IL-1β, IL-12, and IL-6 in RAW 264.7 cells compared to the positive control L. rhamnosus GG. IL-10, an anti-inflammatory cytokine, was also stimulated by isolated LAB compared to L. rhamnosus GG, except for the strains L. paraplantarum JBCC105634 and L. plantarum JBCC105645. In particular, L. plantarum JBCC105655, L. plantarum JBCC105683, and P. acidilactici JBCC105117 highly stimulated macrophage cells to simultaneously produce IL-10 and pro-inflammatory IL-12, exhibiting higher levels than those obtained with LPS.

Effects of LAB Strains on In Vivo Caenorhabditis Elegans Assay for Immune Response
Four strains, including P. acidilactici JBCC105117, L. paraplantarum JBCC105611, JBCC105655, and L. plantarum JBCC105683 strongly stimulated PMK-1 expression ( Figure  3). When the transcript levels were measured in young adult fer-15;fem-1 worms conditioned with four LAB strains for 24 h to quantitatively verify their immune activity, P. acidilactici JBCC105117, L. paraplantarum JBCC105655, and L. plantarum JBCC105683 strongly stimulated the expression of clec-60, which is consistent with the pmk-1::GFP assay ( Figure 4). Unexpectedly, L. paraplantarum JBCC105611 showed very low expression of clec-60, despite the increased induction activity of pmk-1::GFP. The reason for this is still unclear, but one likely explanation is the presence of diverse factors related to the immune pathway differences, such as TNF-α, IL-1β, IL-12, and IL-6. In order to confirm our immune results from the nematode innate immune system, we examined the immune-modulating functions of heat-killed lactic acid bacteria in the murine macrophage cell line RAW 264.7. Figure 2. Pro-inflammatory (A-D) and anti-inflammatory cytokine (E,F) production by RAW 264.7 cells exposed to various heat-killed L. rhamnosus GG and selected lactic acid bacteria strains. RAW 264.7 cells were cultured in the medium (control), LPS (1 mg/mL), or heat-killed bacteria (8 log CFU/mL) for 48 h. Lactocaseibacillus rhamnosus GG, LGG; Pa, Pediococcus acidilactici; Lpa, Lactobacillus paraplantarum; Lp, Lactobacillus plantarum; Lpen, Lactobacillus pentosus; Leu, Leuconostoc mesenteroids. * p < 0.05 compared to the negative control; # p < 0.05 compared to LGG.

Effects of LAB Strains on In Vivo Caenorhabditis Elegans Assay for Immune Response
Four strains, including P. acidilactici JBCC105117, L. paraplantarum JBCC105611, JBCC105655, and L. plantarum JBCC105683 strongly stimulated PMK-1 expression (Figure 3). When the transcript levels were measured in young adult fer-15;fem-1 worms conditioned with four LAB strains for 24 h to quantitatively verify their immune activity, P. acidilactici JBCC105117, L. paraplantarum JBCC105655, and L. plantarum JBCC105683 strongly stimulated the expression of clec-60, which is consistent with the pmk-1::GFP assay (Figure 4). Unexpectedly, L. paraplantarum JBCC105611 showed very low expression of clec-60, despite the increased induction activity of pmk-1::GFP. The reason for this is still unclear, but one likely explanation is the presence of diverse factors related to the immune pathway differences, such as TNF-α, IL-1β, IL-12, and IL-6. In order to confirm our immune results from the nematode innate immune system, we examined the immune-modulating functions of heat-killed lactic acid bacteria in the murine macrophage cell line RAW 264.7.

PCA
To comprehensively evaluate the selected LAB strains, we assessed their relationships with the probiotic phenotypes (acid tolerance, bile tolerance, cell adhesion, antibiotic resistance, and antimicrobial activity), immune response factors (IL-1β and IL-10), and CLA productivity subjected to PCA. As illustrated in Figure 5, PC1 (variance 36.4%) was mainly related to CLA productivity, acid/bile tolerance, and cell adhesion ability. PC2 (variance 26.9%) expressed cytokine expression and antimicrobial activity. In particular, three strains, L. paraplantarum JBCC105655, L. plantarum JBCC105683, and L. pentosus JBCC105674, located in the right part of the PC1 plane, were characterized by high CLA productivity and high probiotic properties. The other groups, including L. paraplantarum JBCC105655, L. plantarum JBCC105683, and P. acidilactici JBCC105117, displayed high levels of IL-1β or IL-10 expression and antimicrobial activity. In particular,

PCA
To comprehensively evaluate the selected LAB strains, we assessed their relationships with the probiotic phenotypes (acid tolerance, bile tolerance, cell adhesion, antibiotic resistance, and antimicrobial activity), immune response factors (IL-1β and IL-10), and CLA productivity subjected to PCA. As illustrated in Figure 5, PC1 (variance 36.4%) was mainly related to CLA productivity, acid/bile tolerance, and cell adhesion ability. PC2 (variance 26.9%) expressed cytokine expression and antimicrobial activity. In particular, three strains, L. paraplantarum JBCC105655, L. plantarum JBCC105683, and L. pentosus JBCC105674, located in the right part of the PC1 plane, were characterized by high CLA productivity and high probiotic properties. The other groups, including L. paraplantarum JBCC105655, L. plantarum JBCC105683, and P. acidilactici JBCC105117, displayed high levels of IL-1β or IL-10 expression and antimicrobial activity. In particular, L. paraplantarum JBCC105655 and L. plantarum JBCC105683 significantly (p < 0.05) stimulated the expression of both pro-and anti-inflammatory cytokines.

Discussion
LAB are important producers of valuable CLAs that exert health-promoting effects. Recent studies have focused on producing functionally enhanced fermented foods and using CLA-producing strains as probiotic strains. In this study, we aimed to find CLAconverting LAB from Jeot-gal, a traditional Korean fermented seafood, due to its severe production environments with high salt content. We successfully isolated CLAconverting LAB by using systematic three-step screening approaches for LA tolerance, UV spectrum, and GC analysis and found 10 strains, including L. paraplantarum (3), L. (20%), which is identical to L. plantarum lp15 from naturally fermented Chinese pickles (75:25) [38]. CLA isomer production by L. plantarum involves four enzymes that catalyze hydration/dehydration (CLA-HY), oxidation of the hydroxyl groups and reduction of the oxo groups (CLA-DH), migration of carbon-carbon double bonds (CLA-DC), and saturation of carbon-carbon double bonds [45,46]. A recent study showed that the cis9, trans11-CLA isomer accumulation by L. plantarum is significantly related to α-enolase, a multifunctionalanchorless-surface protein that plays a role in cell detoxification from polyunsaturated fatty acids, such as linoleic acid, along with the linoleate isomerase complex [47]. Nevertheless, the reason why bacteria exhibit different isomer metabolism remains unclear. L. plantarum NCUL005, isolated from natural sauerkraut, produced CLA isomers in a 30:70 ratio [29]. The cis9, trans11-CLA isomer showed a more potent antiproliferative effect on the viability of two cancer cell lines (SW480 and HT-29) than the c9, t11 isomer [48].
Resistance to low pH and bile salts in the acidic environment of the stomach and upper part of the intestine has been regarded as an important property in the assessment of probiotic strains. The low pH of 1.0 to 2.5 and bile concentrations of 0.3-0.4% are strong barriers against the entry of bacteria into the intestinal tract [49]. Tolerance under these conditions is a stronger discriminative parameter for selecting acid-tolerant bacteria for probiotics [50]. It seems that the high CLA-converting strains exhibited higher survival than other strains of L. plantarum, L. acidophilus, and L. paracasei, which showed an average survival rate of less than 40% after exposure to pH 2.0 [51,52]. In particular, L. paraplantarum JBCC105634, L. paraplantarum JBCC105655, and L. pentosus JBCC105674 exhibited survival rates above 50% after exposure to pH 2.0. Once the LAB reach the small intestinal tract, their bile resistance is crucial because the bacterial cell membrane, which consists of lipids and fatty acids, is very susceptible to the detergent-like characteristics of the bile salts [50]. Although the cell viability of the CLA-producing LAB slightly declined with an increase in bile concentration (p < 0.05), almost all the selected strains were able to tolerate the bile salts even at high concentrations. P. acidilactici JBCC105117 showed significantly high survival rates of 99.4% and 96.1% at 3% and 5% bile salts, respectively (p < 0.05).
In our study, we evaluated the selected LAB using a Caenorhabditis elegans surrogate in vivo model to determine their potential for immunity. Our results clearly showed that two strains, L. paraplantarum JBCC105655 and L. plantarum JBCC105683, induced a significantly higher secretion of TNF-α, IL-1β, IL-12, and IL-10 in RAW 264.7 cells, compared to L. rhamnosus GG as a positive control. In addition, P. acidilactici JBCC105117 stimulated macrophage cells to simultaneously produce IL-10 and pro-inflammatory IL-12. However, we could not observe any distinct expression of immune-modulating functions with L. paraplantarum JBCC105611, which clearly matched the results described above, except for the PMK-1 signaling test. Taken together, these three LAB strains showed distinct pro-and anti-inflammatory cytokine expression levels, indicating that they may play an important role in balancing the immune response. As we used not only heat-killed bacterial preparations in order to remove possible bacterial metabolite interruptions but also the nematode innate immune system of C. elegans in vivo, the observed immune functions of the three strains might be related to both the surface properties and the inner antimicrobial peptides. It is well known that the cell-wall structure of non-pathogenic grampositive bacteria acts as an excellent inducer of immune responses, such as that against pathogenic gram-negative bacteria [53]. It was also shown that different LAB stimulate different levels of various cytokines, such as TNF-α, IFN-γ, and IL-12. Although we did not measure IFN-γ production by the three LAB strains directly in our studies, the selected strains must be effective for immune function as macrophage-derived IL-12 stimulates IFN-γ production in T cells and NK cells, which can activate the specific immune responses of the Th-2 pathway, as well as IgE secretion in mice [54]. In addition, the secretion of Th-2 cytokines plays a major role in the perpetuation of immunological responses in allergic diseases [55]. Interestingly, the cytokine IL-12 induction in the three strains was higher than in the controls, including LPS. Excessive production of IL-12 can impair the organ-specific auto-immunity balance. However, IL-10 suppresses IL-12 production and supports the function of the regulatory T cells. Their mutually antagonistic functions have been reported by several authors [56,57]. Thus, the balanced production of IL-10 and IL-12 by the three strains inhibits polarization of the immune response, which is related to the Th1-type immune response, whereas the Th2-type is important for host immunity. L. plantarum isolated from Kimchi has been reported to induce a macrophage-derived Th1 response, which may be helpful for anti-allergic effects in vitro [58]. The innate immune response serves not only as the first line of defense, but also plays a crucial role in the development of subsequent adaptive immune responses. Therefore, the release of pro-and anti-inflammatory cytokines from immune cells stimulated by selected lactic acid bacteria, L. plantarum JBCC105655, L. plantarum JBCC105683, and P. acidilactici JBCC105117, may modulate the innate and adaptive immune system and inflammatory response. In addition, the conjugated linoleic acid (CLA), which exhibits anti-inflammatory and anti-carcinogenic properties, produced by the bacterial strains in this study may have a positive effect on the intestinal environment. Furthermore, kinetic studies of each enzyme are needed in future investigations.

Conclusions
In this study, LAB from Jeot-gals were identified and found to have diverse CLA production yields and isomer ratios. Moreover, these LAB showed potential probiotic properties, suggesting their potential as beneficial probiotics that produce CLA as a functional ingredient. P. acidilactici JBCC105117, L. paraplantarum JBCC105655, and L. plantarum JBCC105683 strongly stimulated the immunological regulatory gene PMK-1 and a host defense antimicrobial peptide gene, clec-60, in C. elegans and produced the significant induction of tumor necrosis factor-α, interleukin (IL)-1β, IL-6, IL-12, and IL-10 in RAW 264.7 macrophages, indicating that they are good candidates for probiotics with high CLA-converting activity. The LAB from Jeot-gals are good candidates for manufacturing functionally enhanced fermented foods as functional starters with the beneficial effects of CLA.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/ 10.3390/microorganisms9112247/s1, Figure S1: MRS plate agar medium containing 0.2% linoleic acid, Figure S2: UV spectral scan of hexane extracts, Figure S3: Phylogenetic tree based on 16S rRNA analysis of LAB isolated from Jeot-gals, Figure S4: Antibacterial activity of some CLA-producing LAB against pathogenic bacteria, Figure S5: The effects of selected lactobacillus strains on RAW 264.7 cell proliferation by using the MTT colorimetric assay, Table S1: Primer sequences of cytokines used in qRT-PCR analysis.