Probiotic effect on Helicobacter pylori attachment and inhibition of inflammation in human gastric epithelial cells

Bacterial strains and culture conditions

The probiotic strains used were as follows: L. acidophilus, L. salivarius, C. butyricum, B. licheniformis, B. infantis, B. longum and L. bulgaricus, purchased from Siliankang, Jinshuangqi, Changlekang and Taiwan Yaxin (Table I). The probiotics were cultured on De Man, Rogosa and Sharpe agar plates (MRS; Oxoid; Thermo Fisher Scientific, Inc.) at 37°C in an anaerobic humidified environment for 24 h. Live bacteria cells were obtained by centrifuging the cultures and washing with sterile PBS three times. The precipitate were resuspended with RPMI 1640 medium (Gibco; Thermo Fisher Scientific, Inc.) and adjusted to 3×10⁸ colony-forming units (CFU)/ml.

H. pylori ATCC43504 strain was purchased from the American Type Culture Collection and the multidrug resistant H. pylori strain (HP021) was isolated from the gastric biopsy of a patient with chronic atrophic gastritis and confirmed to have resistance to clarithromycin and levofloxacin. H. pylori ATCC43504 strain expresses CagA and VacA proteins, which induces nuclear factor (NF)-κB and interleukin (IL)-8 expression (21). Both H. pylori strains were grown on Columbia blood agar plates supplemented with 7% defibrinated horse blood (Beijing Biotek Medical Device, Ltd.) for at least 3 days at 37°C under microaerophilic conditions (5% O₂, 10% CO₂, and 85% N₂). Normal human gastric epithelial GES-1 cells (American Type Culture Collection) were cultured at 37°C in RPMI 1640 medium supplemented with 10% fetal bovine serum (Biological Industries) and 100 µg/ml penicillin/streptomycin in a humidified atmosphere with 5% CO_2. The medium was changed every other day. Before experiments were initiated, the cells were plated in 96-well plates at 5×10³ cells/well or 2.5×10⁵ cells/well in 6-well plates for 24 h in serum-free RPMI-1640 medium.

Cell-free supernatant (CFS) preparation and H. pylori ATCC43504–lipopolysaccharide (LPS) culture

Cultures of the various probiotic strains were grown in MRS broth (Oxoid; Thermo Fisher Scientific, Inc.) in an anaerobic humidified environment at 37°C for 96 h. CFS from the probiotics was prepared by centrifuging (12,000 × g; 4°C; 10 min) the respective MRS broths. CFS were filtered through a 0.2 µM filter and stored at −20°C.

Culture broth from H. pylori ATCC43504 was centrifuged (8,000 × g; 4°C; 10 min) and washed three times then resuspended in PBS. H. pylori lipopolysaccharide (HP-LPS) was obtained using an LPS extraction kit (Intron Biotechnology, Inc.) according to the manufacturer's instructions. HP-LPS concentrations were determined with a kinetic Limulus Amebocyte Lysate Assay kit (Xiamen Limulus Reagent Biotechnology Co., Ltd.).

GES-1 cell viability

Probiotics and CFS are toxic to GES-1 cells at high concentrations and during long incubations (17). To determine the optimal concentration and incubation time for each probiotic and CFS, GES-1 cells were infected with CFS and probiotics at multiplicities of infection (MOI) of 100 and 1,000 in antibiotic-free RPMI 1640 medium at 37°C, 5% CO₂ for up to 8 h. Viable GES-1 cell numbers were determined by trypan blue staining following incubation for 2, 4, 6 and 8 h at 37°C. Non-infected cell cultures served as controls. Cells that were not stained with trypan blue were counted as viable cells.

GES-1 proliferation

GES-1 cells were seeded into 96-well plates at 1.5×10⁵ cells/ml in 100 µl culture volume for 24 h. CFS (100 µl) and probiotics were then added to the GES-1 cells at 10², 10³, 10⁴, 10⁵, 10⁶, 10⁷, 10⁸ and 10⁹ CFU/ml. Each condition was performed in triplicate and cultured for 24 h. Cells were washed three times with sterile PBS solution then 20 µl of 4 mg/ml MTT reagents was added. The cells were cultured for another 4 h and then centrifuged for 5 min at 800 × g. The supernatant was discarded and 150 µl dimethyl sulfoxide was added. Absorbance was measured at 490 nm.

Assessment of adhesion ability

Adhesion ability of probiotics on GES-1 cells was determined as previously described by de Klerk et al (22). In brief, probiotics were suspended in RPMI-1640 medium to a concentration of 3×10⁸ CFU/ml. GES-1 cells cultured on slides in 6-well plates were infected with probiotics at a MOI of 100. Following 4 h of incubation, each well was washed three times with sterile PBS to remove any unbound bacteria. Cells were fixed with 4% paraformaldehyde at room temperature for 30 min and then the bacteria were Gram stained at room temperature for 5min. The cell monolayer was washed with tap water and the cells were observed at a magnification of ×1,000 under light microscope oil immersion. The adherence of probiotics was calculated as previously described (23,24) and was as follows: Cell adhesion index=number of adhering probiotics/total number of cells in the field of view ×100. Cell adhesion rate=number of cells adhered by probiotics/total number of cells in the field of view ×100. Cell adhesion index indicates the number of probiotics adhering to each cell, and cell adhesion rate represents the proportion of cells adhered by probiotics in total cells.

Inhibition of H. pylori growth

H. pylori ATCC43504 and HP021 were evenly seeded on Columbia agar plates without antibiotics. Holes were introduced to the agar plates with a sterile oxford cup then the bottom of holes were sealed with 0.8% agar liquid. Live probiotics (120 µl at 3×10⁹ CFU/ml) were suspended in MRS broth and CFS (120 µl) was added to the holes in the plates. The plates were incubated under microaerophilic conditions for 72 h at 37°C, and then the diameters of the inhibition zones were measured. PBS and MRS medium were used as negative controls.

Adhesion of H. pylori to GES-1 cells

For the infection studies, GES-1 cells were grown on microtiter plates to form a confluent monolayer. The concentration of each probiotic was adjusted to 1.5×10⁷ CFU/ml. H. pylori ATCC43504 and HP021 concentrations were then adjusted to 1.5×10⁷ CFU/ml. GES-1 cells were pre-treated with 50 µl of live probiotics for 2 h before infection (pre-treated group) or following infection (post-treated group). GES-1 cells were infected for 2 h with 50 µl of live H. pylori ATCC43504 or HP021. Subsequently, each well was washed three times to remove any non-adherent H. pylori. Urease activity was determined using a modified phenol red method (18,25,26). In brief, 200 µl of urease test solution (20% [w/v] urea and 0.012% phenol red in phosphate buffer; pH 7.0) was added to each well of a microtiter plate. The plate was then incubated at 37°C for 1 h. The absorbance at 550 nm was measured with a microtiter plate spectrophotometer (BioTek Instruments, Inc.). The adherence of H. pylori was calculated as described by Chen et al (18) and was as follows: Adherence=([Optical density experimental-optical density negative]/[optical density positive-optical density negative] ×100.

The negative control contained only GES-1 cells and the positive control contained both GES-1 cells and H. pylori, which were used to establish 100% adherence.

ELISA for interleukin (IL)-8 detection

LPS is a pathogenic factor of H. pylori which can induce GES-1 cells to produce IL-8 (27). Probiotics with proven adhesive ability were added to GES-1 cells (MOI of 100) and incubated for 2 h. Following washing with PBS to remove bacilli, HP-LPS (final concentration of 7,000 endotoxin units [EU]/ml) was added and the cells were incubated for another 6 h. The final culture supernatants were centrifuged for 10 min at 12,000 × g and 4°C to remove bacteria and cell debris. Supernatants were then aliquoted and stored at −80°C. IL-8 concentration in the supernatant was determined using a commercially available ELISA kit (cat. no. VAL103; R&D Systems, Inc.). Absorbance values were measured at 450 nm using a microplate reader. Each sample was measured three times.

Preparation of cellular lysates and western blot analysis

GES-1 cells were pre-treated for 2 h with probiotics, followed by HP-LPS stimulation for 60 or 120 min at a final concentration of 7,000 EU/ml. Cytoplasmic and nuclear extracts were then isolated using a Nuclear Extract kit (Active Motif, Inc.). Protein concentrations were determined with an enhanced bicinchoninic acid protein assay kit.

Each extract was mixed with ×5 loading buffer and boiled for 5 min. The extracted proteins were then aliquoted and stored at −80°C. A total of 30 µg of protein was loaded into each lane, separated via 8% SDS-PAGE and then electrotransferred to a polyvinylidene difluoride membrane. Membranes were blocked at room temperature with 5% fat-free dried milk in Tris buffered saline containing 0.1% Tween-20 (TBST) for 2 h. Following this, membranes were incubated overnight at 4°C with primary antibodies anti-toll-like receptor 4 (TLR4; cat. no. AF1478-SP; 1:1,000; R&D Systems, Inc.), anti-NF-κB p65 (cat. no. 8242P; 1:1,000; Cell Signaling Technology, Inc.), and anti-NFKB inhibitor-α (IκBα; cat. no. 4814P; 1:1,000; Cell Signaling Technology, Inc.). Anti-lamin B1 (cat. no. 66095-1-Ig; 1:5,000; ProteinTech Group, Inc.) and anti-GAPDH (cat. no. 60004-1-Ig; 1:2,000; ProteinTech Group, Inc.) were used to verify equal protein loading. Following washing three times in TBST, the membranes were incubated with horseradish peroxidase conjugated goat anti-rabbit (cat. no. ZB-5301; 1:10,000; OriGene Technologies, Inc.) and goat anti-mouse (cat. no. ZB-2305; 1:5,000; OriGene Technologies, Inc.) secondary antibodies at room temperature for 1 h.

The proteins were visualized using an enhanced chemiluminescence reagent (Beyotime Institute of Biotechnology). The integrated optical density (IOD) of each band was analyzed using Image-Pro Plus software v6.0 (National Institutes of Health.). Each TLR4 and IκBα band value was normalized as the ratio of the IOD to the GAPDH band. Each NF-κB p65 band value was normalized as the ratio of the IOD to the lamin B1 band.

Statistical analysis

Statistical analyses were performed by SPSS v17.0 software (SPSS, Inc.). Graphs were generated using GraphPad Prism 5.2 software (GraphPad Software, Inc.). All data were presented as the mean value ± standard deviation. Multiple comparisons were evaluated by one-way analysis of variance followed by the Student-Newman-Keuls and least significant difference post hoc tests. The Dunnett method was used for comparisons with the control group. P<0.05 was considered to indicate statistical significance.

GES-1 cell viability and growth rate following incubation with probiotics and CFS

The morphological characteristics of the probiotics and H. pylori were first evaluated following Gram staining (Fig. 1). The probiotics in this study were all Gram-positive bacteria. L. acidophilus was chain rod shaped, L. bulgaricus, B. longum and B. licheniformis were long rod shaped, B. infantis and L. salivarius were short rod shaped. C. butyricum contained giant spore and was enlarged at one end in a drum-hammer shape. H. pylori was a gram-negative bacterium, which is S-shaped or spiral. Furthermore, the viability of GES-1 cells incubated with CFS and different titers of probiotics (MOI of 100 and 1,000) was determined by assessing the percentage of cells not stained with trypan blue following 2, 4, 6 and 8 h of incubation. At a MOI of 100, the cytotoxicity results indicated that the CFS of L. acidophilus, L. bulgaricus, L. salivarius, C. butyricum, B. infantis and B. longum were toxic to GES-1 cells. The CFS of B. licheniformis did not show any toxicity towards GES-1 cells with an average survival rate at 8 h of 96%. L. acidophilus, L. bulgaricus, L. salivarius, Clostridium butyricum, Bifidobacterium infantis, Bifidobacterium longum and B. licheniformis did not show any toxicity towards GES-1 cells after 4 h of incubation, with an average survival rate of 97% and no significant differences between these groups and the control (Fig. 2A). However, GES-1 cell viability decreased gradually after 4 h. At 6 h, the most significant decrease in viability was observed in B. longum-treated cells, which had a 92% survival rate. Notably, the viability of these cells was significantly lower compared with the cells treated with the other six bacteria (P<0.05; Fig. 2A). GES-1 cell viability following B. licheniformis treatment decreased at the slowest rate, with an average survival rate of 99.7%. This was significantly higher than the survival rate of the cells treated with other bacteria (P<0.05; Fig. 2A). There was no significant differences in the average GES-1 survival rates following treatment with L. acidophilus, L. bulgaricus, L. salivarius, C. butyricum, B. longum and B. infantis at 6 h. At 8 h, the survival rate of the GES-1 cells decreased further with cells treated with L. bulgaricus, demonstrating significantly lower survival rate (82.3%) compared with the cells treated with the other six probiotics (P<0.05; Fig. 2A). Furthermore, the survival rate of cells treated with B. licheniformis was significantly increased (96.7%) at 8 h compared with cells treated with the other six bacteria (P<0.05; Fig. 2A).

At a MOI of 1,000, all seven probiotic strains exhibited no toxicity towards GES-1 cells at 2 h (Fig. 2B). However, GES-1 cell viability decreased following 4 h of incubation, with cells treated with L. bulgaricus (10.4% at 4 h) exhibiting significantly reduced viability compared with cells treated with the other six bacterial strains (P<0.05; Fig. 2B). GES-1 cell viability following B. licheniformis treatment demonstrated the slowest decrease, with the average survival rate of cells 100% at 4 h, which was significantly higher when compared with cells treated with the other bacterial strains (P<0.05; Fig. 2B). The viability of B. licheniformis-treated GES-1 cells decreased significantly after 6 h compared with the control group (P<0.05; Fig. 2B). Furthermore, the viability of these cells was 42.7% at 8 h, whereas GES-1 cells treated with the other six bacteria did not survive past 8 h. There were no significant differences between L. acidophilus-, L. salivarius-, B. longum-, B. infantum- and C. butyricum-treated cells (Fig. 2B). The probiotic half maximal inhibitory concentration (IC₅₀) for each group is listed in Table II, with B. longum demonstrating the lowest IC₅₀ and B. licheniformis demonstrating the highest IC₅₀.

Adhesion of probiotics to GES-1 cells

Probiotics adhered to GES-1 cells with different cell adhesion indexes and cell adhesion rates (Table III; Fig. 3). Staining of GES-1 cells for each probiotic group varied because metabolites produced by different probiotics were able to change the pH values of the culture medium, thereby affecting viability and staining (28). The pH values of the culture medium were pH 4 (L. acidophilus), pH 5 (L. bulgaricus), pH 3.8 (B. infantis), pH 4.2 (B. longum), pH 4.8 (L. salivarius), pH 5.5 (C. butyricum) and pH 6.8 (B. licheniformis).

As indicated in Fig. 4, the cell adhesion index was consistent with the cell adhesion rate, with the highest to lowest results as follows: L. acidophilus>L. bulgaricus> B. infantis>B. longum>L. salivarius> B. licheniformis> C. butyricum. There was no significant difference in the cell adhesion index between C. butyricum and B. licheniformis (Fig. 4A); however, there were significant differences between the other five bacterial groups when compared with each other (P<0.05; Fig. 4A). There was no significant difference in the cell adhesion rate between L. bulgaricus and L. acidophilus. In addition, no significant differences in the cell adhesion rate were observed between B. longum, C. butyricum, B. licheniformis and L. salivarius (Fig. 4B).

Live probiotics and CFS inhibit H. pylori growth

CFS from probiotic culture and live probiotics were screened for inhibitory effects against HP021 by measuring the diameters of inhibitory growth zones on agar plates (Table IV). Fresh MRS medium, used as a negative control, also exhibited significant activity against HP021 and formed inhibition zones of ~27 mm, which was larger than all the tested probiotics (P<0.05; Table IV). This suggested that there were likely unknown anti-HP substances in MRS medium. Probiotics consumed some of the anti-HP substances of MRS, and likely secreted some anti-HP substances. Live L. bulgaricus demonstrated significantly greater anti-H. pylori activity compared with the other live probiotics, with inhibition zones of ~25 mm (P<0.05; Table IV). Live B. licheniformis did not display any anti-H. pylori activity. By contrast, there were no significant differences in inhibition of H. pylori growth by live L. acidophilus, B. infantis, B. longum, L. salivarius and C. butyricum. The CFS from all the probiotics also inhibited H. pylori activities; however, there were no significant differences between the probiotic groups (Table IV). Similar results were observed when H. pylori ATCC43504 was used (data not shown).

Probiotics and CFS suppress adhesion of H. pylori to GES-1 cells

The urease activity of HP021 and ATCC43504 were examined to evaluate the effect of CFS and probiotics on HP021 and H. pylori ATCC43504 adherence to GES-1 cells. The adherence rate of H. pylori was 100% without probiotic or CFS treatment. Results indicated that pre- or post-treatment with B. licheniformis CFS did not reduce the adhesion of H. pylori to GES-1 cells compared with HP group (Fig. 5A). However, urease activity was significantly reduced following probiotic treatment, with the exception of C. butyricum and B. licheniformis treatment compared with the HP group (Fig. 5A). No significant differences between the two H. pylori strains were found (the data not shown). The adherence rate of H. pylori dropped to ~50% with L. acidophilus treatment. Compared with non-treated cells, the adherence rate of HP021 did not decrease with C. butyricum or B. licheniformis treatment (Fig. 5B). These results demonstrated that pre-treatment with L. acidophilus, L. bulgaricus, L. salivarius, B. infantis and B. longum inhibited the adherence of H. pylori to GES-1 cells.

Certain probiotics significantly affect IL-8 production in GES-1 cells

Under normal physiological conditions, GES-1 cells secrete a small amount of IL-8 (29). The addition of HP-LPS to GES-1 cells for 6 h significantly increased the IL-8 levels compared with non-treated cells (P<0.05; Fig. 6). No decrease in IL-8 secretion following treatment with the CFS of B. licheniformis was observed compared with HP-LPS group. However, following pre-treatment with L. bulgaricus and L. salivarius, there was a significant decrease in IL-8 secretion induced by HP-LPS compared with HP-LPS group (P<0.05; Fig. 6). Notably, IL-8 levels following HP-LPS plus L. acidophilus, B. infantis or B. longum treatment were higher compared with HP-LPS only treatment.

Probiotics attenuate the TLR4-NF-κB p65 signaling pathway following HP-LPS activation

To explore the molecular mechanism of H. pylori on GES-1 growth, expression levels of markers associated with inflammation, TLR-4, IκBα and NF-κB, p65 were detected. H. pylori ATCC43504 contains CagA protein that induces activation of NF-κB and IL-8 whereas HP021 does not contain CagA protein. Therefore, H. pylori ATCC43504 was selected for the NF-κB activation and IL-8 production experiment. L. bulgaricus reduced IL-8 production most significantly. Therefore, L. bulgaricus was selected for the NF-κB activation experiment. Treatment of GES-1 cells with 7,000 EU/ml of HP-LPS upregulated the expression of TLR-4, increased the degradation of IκBα and induced the translocation of NF-κB p65 into nucleus in a time-dependent manner (Fig. 7A and B). Pre-treatment for 2 h with viable L. bulgaricus (MOI of 100) significantly inhibited the effects of HP-LPS on the TLR4/NFκB pathway by increasing cytoplasmic IκBa and decreasing nuclear NF-κB p65 levels (Fig. 7C-E). The levels of NFκB and TLR-4 in L. bulgaricus pre-treatment 60 and 120 min groups were significantly lower than those of HP-LPS 60 and 120 mins groups, respectively (P<0.05; Fig. 7C and E). The level of IκBα in L. bulgaricus pre-treatment 120 min group was significantly higher than HP-LPS 120 min group (P<0.05; Fig. 7D). These results indicated that the TLR4/NF-κB signaling pathway may be a critical target for probiotics to alter GES-1 cell biological functions.