Lactobacillus acidophilus Induces Cytokine and Chemokine Production via NF-κB and p38 Mitogen-Activated Protein Kinase Signaling Pathways in Intestinal Epithelial Cells

Abstract

Intestinal epithelial cells can respond to certain bacteria by producing an array of cytokines and chemokines which are associated with host immune responses. Lactobacillus acidophilus NCFM is a characterized probiotic, originally isolated from human feces. This study aimed to test the ability of L. acidophilus NCFM to stimulate cytokine and chemokine production in intestinal epithelial cells and to elucidate the mechanisms involved in their upregulation. In experiments using intestinal epithelial cell lines and mouse models, we observed that L. acidophilus NCFM could rapidly but transiently upregulate a number of effector genes encoding cytokines and chemokines such as interleukin 1α (IL-1α), IL-1β, CCL2, and CCL20 and that cytokines showed lower expression levels with L. acidophilus NCFM treatment than chemokines. Moreover, L. acidophilus NCFM could activate a pathogen-associated molecular pattern receptor, Toll-like receptor 2 (TLR2), in intestinal epithelial cell lines. The phosphorylation of NF-κB p65 and p38 mitogen-activated protein kinase (MAPK) in intestinal epithelial cell lines was also enhanced by L. acidophilus NCFM. Furthermore, inhibitors of NF-κB (pyrrolidine dithiocarbamate [PDTC]) and p38 MAPK (SB203580) significantly reduced cytokine and chemokine production in the intestinal epithelial cell lines stimulated by L. acidophilus NCFM, suggesting that both NF-κB and p38 MAPK signaling pathways were important for the production of cytokines and chemokines induced by L. acidophilus NCFM.

INTRODUCTION

The human gastrointestinal (GI) tract, which is populated by a complex mixture of more than 10¹⁴ microorganisms, is lined by a single monolayer of intestinal epithelial cells (IEC) (7). IEC are recognized as immunological sentinels of the GI tract and play a key regulatory role in maintaining host innate and adaptive mucosal immunity (16, 40). IEC act as the first line of host defense against a pathogenic bacterial invasion or inflammatory stimuli by secreting an array of cytokines and chemokines, which affect the immune cells scattered in the GI tract and recruit immune cells to the GI tract, respectively (13, 19, 24, 36). Because IEC are continually exposed to the GI tract microbiota, it is clear that commensal bacteria should not elicit as intense an inflammatory response as pathogenic bacteria (31). In addition, some investigators showed that IEC remain hyporesponsive to nonpathogenic commensal bacteria (23, 29). However, it has also been reported that IEC, exposed to some commensal bacteria, such as Bacillus subtilis, Bacteroides ovatus, Escherichia coli, Lactobacillus rhamnosus, Bifidobacterium lactis, Lactobacillus casei, or Lactobacillus acidophilus, could produce inflammatory cytokines (e.g., interleukin 1 [IL-1], IL-8, and tumor necrosis factor alpha [TNF-α]) or chemokines (e.g., CCL2 and CCL20) (6, 12, 21, 33, 40).

Probiotics exert beneficial effects on the health of the host through establishing mutualistic relationships with the IEC (22). Some strains have been shown to enhance the host immune responses by regulating cytokine and chemokine production (12, 18, 21, 33, 39, 40). Of these, the strain Lactobacillus acidophilus NCFM is a well-characterized probiotic bacterium, with several reports showing beneficial effects on the host (1, 20, 32, 35). These studies have shown that L. acidophilus NCFM is able to modulate the production of inflammatory mediators, such as TNF-α, IL-1β, CCL2, and IL-6, in dendritic cells (DC) and IEC (32, 40). However, little is known about the basic molecular mechanism of L. acidophilus NCFM regulation of the host immune responses.

IEC sense bacteria through expression of conserved pattern recognition receptors (PRRs), such as the Toll-like receptors (TLRs) (21, 26). Some studies have shown that TLR2 and TLR4 were constitutively expressed both in IEC lines and in primary IEC isolated from intestinal tissue (3, 21). These receptors activated nuclear factor kappa B (NF-κB) and mitogen-activated protein kinase (MAPK), the immune-related transcriptional factors that induced the synthesis of cytokines and chemokines (25). It has been reported that B. lactis, the dominant microbial population group in the human GI tract, induced inflammatory cytokine IL-6 production through NF-κB and p38 MAPK signaling pathways in IEC (33). L. casei could activate these signaling pathways in production of innate cytokines, such as TNF-α and IL-12, in spleen cells (18). Miettinen et al. also showed that L. rhamnosus GG can initiate NF-κB, STAT1, and STAT3 DNA-binding activity in human macrophages (27). Therefore, it is likely that the activation of these transcriptional factors of host cells by L. acidophilus plays important roles in the generation of immune-related cytokines and chemokines that function to benefit the host.

In this study, we examined the ability of L. acidophilus NCFM to stimulate cytokine and chemokine production in native IEC and IEC lines, and elucidated the mechanisms involved. We found that L. acidophilus NCFM could rapidly but transiently induce cytokine and chemokine production, and cytokines showed lower expression levels than chemokines. Furthermore, our research suggested that the activation of TLR2-mediated NF-κB and p38 MAPK signaling pathways played a key role in the production of cytokines and chemokines in IEC.

MATERIALS AND METHODS

Bacterial strain and culture conditions.

L. acidophilus NCFM was obtained from American Type Culture Collection (ATCC) (Rockville, MD). For stimulation experiments, the bacteria were anaerobically grown at 37°C in de Man, Rogosa, and Sharp broth (MRS broth) (Difco, Detroit, MI) overnight prior to use. The bacterial cells were harvested by centrifugation (4,000 × g, 10 min) at stationary phase, washed twice with sterile phosphate-buffered saline (PBS), and then diluted with Dulbecco's modified Eagle's minimal essential medium (DMEM) (GIBCO-BRL, Grand Island, NY) and sterile 10% skimmed milk for in vitro and in vivo experiments, respectively. The number of bacterial cells was determined by the plate counting agar method.

Cell culture.

Human colorectal adenocarcinoma cell line Caco-2 cells were purchased from ATCC and maintained in an incubator at 37°C, 5% CO₂, in DMEM supplemented with 10% heat-inactivated fetal bovine serum (FBS) (NQBB, Australia), 1% nonessential amino acids, 10 U/ml penicillin, and 10 μg/ml streptomycin. The Caco-2 cells (3 × 10⁶ cells/well), which were used for stimulation experiments, were allowed to attach and grow in plastic six-well culture plates (Costar; Corning). Cell culture medium was changed every second day for approximately 17 days until the cells reached full differentiation and polarization (38). Subsequently, the Caco-2 cells were used in experimental investigations as specified below.

Stimulation experiment.

Before stimulation, polarized epithelial cell monolayers were washed twice with prewarmed PBS, and then the cells were incubated with the bacteria suspensions at a multiplicity of infection (MOI) (ratio of bacteria number to epithelial cell number) of 10, which did not affect the composition of the culture medium and IEC viability (21), for various times at 37°C and 5% CO₂. Culture medium was used as a negative control. Where indicated, the experiments were terminated by thoroughly washing the cells with cold PBS.

Animal studies.

BALB/c mice, 10 to 12 weeks old, weighing from 20 g to 24 g, were used to study the in vivo kinetics of how L. acidophilus NCFM induced cytokine and chemokine expression. The mice were housed in plastic cages kept at a constant room temperature of 22 ± 2°C and relative humidity of 55% ± 5% and exposed to a 12-h light/dark cycle. They had free access to a conventional balanced diet and distilled water. The experimental group was administered intragastrically with L. acidophilus NCFM diluted in 10% skimmed milk at a clinically relevant concentration of 10⁹ CFU/ml for a week (32), and the daily suspension intake of bacteria was 1.0 ± 0.1 ml/mouse. The mice that were intragastrically administered sterile 10% skimmed milk alone were used as negative controls. Mice in each group were sacrificed 1, 3, 5, and 7 days after the initial intragastric administration. The cecum and colon were removed, washed in cold PBS, and then placed in liquid nitrogen (LIN) immediately after the mice were sacrificed.

Inhibitor treatment.

Prior to stimulation with L. acidophilus NCFM, the polarized Caco-2 cells were incubated with the NF-κB inhibitor (pyrrolidine dithiocarbamate [PDTC], 40 μM; Sigma) and p38 MAPK inhibitor (SB203580, 20 μM; Sigma) for 30 min. Afterwards, the cells were washed twice with prewarmed PBS and then exposed to L. acidophilus NCFM for 2 h at an MOI of 10. The experiments were terminated by thoroughly washing the cells with cold PBS, and then total RNA was prepared for real-time reverse transcription-PCR (RT-PCR).

RNA isolation and real-time RT-PCR.

RNA from cell lines or cecum and colon was extracted using TRIzol (Invitrogen, Carlsbad) by repetitive pipetting (17). The purity and integrity of RNA were evaluated by spectrophotometry and electrophoresis on 1% agarose gels. cDNA was synthesized using the cDNA RT reagent kit (Takara, Dalian, China) according to the manufacturer's protocol. Real-time RT-PCRs were carried out using the ABI Prism 7500 system using SYBR green buffer according to the manufacturer's instructions (Applied Biosystems), subjected to 30-s denaturation at 95°C, followed by 40 cycles of 5 s at 95°C and 34 s at 60°C. The sequences of specific primers used in the PCR are shown in Table 1. The data were analyzed by using ABI Prism 7500 system sequence detection software. All gene quantifications were performed with glyceraldehyde-3-phosphate dehydrogenase (GAPDH) as an internal standard, and the relative quantification of gene expression was analyzed by using the standard formula 2^{−[(Et − Rt) − (Ec − Rc)]}. C_T is the cycle number where the amplified target reaches the defined threshold; Et is the C_T of the experimental gene in treated samples, Rt is the C_T of GAPDH in treated samples, Ec is the C_T of the experimental gene in control samples, and Rc is the C_T of GAPDH in control samples (28). Application plot and dissociation curves were used for the examination of the amplified products.

Table 1.

Primer sequences for cytokines and chemokines for real-time RT-PCR

Gene product	Primer sequence (5′ → 3′)a	Fragment size (bp)
Human
CCL2 (MCP-1)	F, CTCAGCCAGATGCAATCAATG	129
	R, AGATCACAGCTTCTTTGGGACAC
CCL20 (MCP-3α)	F, TTGACTGCTGTCTTGGATAC	150
	R, TCTGTTTGGATTTGCG
IL-1β	F, GTGGCAATGAGGATGACTTGTTC	130
	R, TTGCTGTAGTGGTCGGAG
IL-1α	F, AGAAGACAGTTCCTCCATTG	136
	R, CTTGGATGTTTAGAGGTTTC
GAPDH	F, AACGGATTTGGTCGTATTG	214
	R, GCTCCTGGAAGATGGTGAT
Mouse
CCL2	F, ACGTGTTGGCTCAGCCAGA	136
	R, ACTACAGCTTCCTTTGGGACACC
CCL20	F, TACTGCTGGCTCACCTC	112
	R, ATCTGTCTTGTGAAACCC
IL-1β	F, AAGTTGACGGACCCCA	126
	R, GTGATACTGCCTGCCTGA
IL-1α	F, TCTGCCATTGACCATCTC	183
	R, AATCTTCCCGTTGCTTG
GAPDH	F, GCCTGGAGAAACCTGCC3′	200
	R, ATACCAGGAAATGAGCTTGACA

^aF, forward; R, reverse.

Western blot analysis.

Caco-2 cells, which were treated with L. acidophilus NCFM or DMEM, were lysed in lysis buffer (50 mM Tris-HCl, 100 mM NaCl, 1 mM EDTA, 1 mM EGTA, 10% NP-40 and sodium deoxycholate, 10 mg/ml aprotinin and leupeptin, 100 mg/ml phenylmethylsulfonyl fluoride [PMSF], 400 μM Na₃VO₄, and 5 Mm NaF), incubated at 4°C for 30 min, and centrifuged at 13,000 × g for 10 min at 4°C. The supernatants were transferred to fresh tubes and stored at −70°C until required. The protein concentration in the supernatants was determined by Bradford's method. Approximately 20 μg protein per lane was loaded on a sodium dodecyl sulfate–12% polyacrylamide gel and then transferred to polyvinylidene fluoride membranes (Millipore; Bedford) in 25 mM Tris base, 190 mM glycine, and 20% methanol using a wet blotter. Subsequently, the membranes were blocked with 5% bovine serum albumin (BSA) in Tris-buffered saline (TBS) supplemented with 0.1% Tween 20 for 1 h and washed with TBS supplemented with 0.1% Tween 20 for 5 min three times. Afterwards, the membranes were incubated at 4°C overnight with rabbit anti-Ser(p)-NF-κB p65 (phosphor-specific Ser536), anti-Th(p)-p38 MAPK (phosphor-specific Thr180/Tyr182) (Cell Signaling Technology, Inc., Beverly, MA), anti-TLR2, and anti-GAPDH (Santa Cruz Biotechnology, Santa Cruz, CA). After incubation with horseradish peroxidase (HRP)-conjugated anti-rabbit antibody, the membranes were incubated with ECL chemiluminescence reagent (TransGen Biotech, Beijing, China), and the film was then exposed to the membranes.

Statistical analysis.

Data were expressed as means ± standard deviations (SD) of triplicates. The statistical significance of the difference between the two means was evaluated by using Student's t test. P values of <0.05 were considered significant.

RESULTS

Kinetics of cytokine and chemokine expression in Caco-2 cells stimulated with L. acidophilus NCFM.

In order to assess the effect of L. acidophilus NCFM on cytokine and chemokine production in IEC, Caco-2 cells were incubated with bacteria at an MOI of 10 for 0, 2, 4, 8, and 12 h, and cytokines and chemokines, including IL-1α, IL-1β, CCL2, and CCL20, associated with the host immunity, were measured. As shown in Fig. 1, L. acidophilus NCFM induced cytokine and chemokine expression with the same kinetics, and the expression of these genes was significantly upregulated (P < 0.05) at 2 h after bacterial stimulation, except for that encoding IL-1β, which was significantly upregulated (P < 0.05) at 4 h. All gene expression peaked at 4 h after stimulation and then gradually declined. The chemokine mRNA expression represents a higher fold change than the cytokines.

Fig 1.

Kinetics of cytokine and chemokine expression in Caco-2 cells stimulated with L. acidophilus NCFM. Caco-2 cells were stimulated with L. acidophilus NCFM at an MOI of 10 for 0, 2, 4, 8, and 12 h. Caco-2 cells treated with DMEM were used as a control. Total cellular RNA was extracted at different time points and analyzed by real-time RT-PCR. Each bar represents the combined mean value (± SD) for three experiments. ∗, P < 0.05; ∗∗, P < 0.01. CTR, control.

Kinetics of cytokine and chemokine expression in mice administered L. acidophilus NCFM intragastrically.

In order to further investigate whether L. acidophilus NCFM can induce cytokine and chemokine expression in vivo, BALB/c mice were administered bacteria intragastrically for 0, 1, 3, 5, and 7 days. Figure 2 shows that L. acidophilus NCFM could induce cytokine and chemokine production, and the trends of gene expression levels were comparative in vivo and in vitro. Expression of both cytokines and chemokines was highest on day 5 after the initial bacterial association in mice. However, the expression of these genes was significantly upregulated (P < 0.05) on day 5 except for that encoding IL-1α, the expression of which was not significant (P > 0.05) compared to that for the control group during the bacterial association. Similar to the in vitro data, the expression level of cytokines was lower than that observed for the chemokines. The above results indicated that L. acidophilus NCFM had the ability to regulate transient cytokine and chemokine expression both in vitro and in vivo.

Fig 2.

Kinetics of cytokine and chemokine expression in mice intragastrically administered L. acidophilus NCFM. The BALB/c mice, which were 10 to 12 weeks old, were administered L. acidophilus NCFM intragastrically for 0, 1, 3, 5, and 7 days. Mice intragastrically administered sterile skimmed milk were used as controls. The mice (n = 3) were killed, and the IEC (the cecum and colon) were isolated. Total RNA was extracted and analyzed by real-time RT-PCR. Each bar represents the combined mean value (± SD) for three experiments. ∗, P < 0.05. CTR, control.

Induction of TLR2 in Caco-2 cells by L. acidophilus NCFM.

Expression of the pattern recognition receptors, TLRs, plays an essential role in activation of the host immune responses, and TLR2 has been shown to be activated by Gram-positive bacteria (6, 18). Therefore, we investigated whether L. acidophilus NCFM could induce TLR2 expression in IEC. Caco-2 cells were treated with bacteria at an MOI of 10 for 0, 0.5, 1, 2, and 4 h. As shown in Fig. 3, TLR2 was induced after stimulation with L. acidophilus NCFM, and the activation was started as early as 0.5 h after treatment. The data suggested that probiotic L. acidophilus NCFM could upregulate the expression of the pattern recognition receptor molecule TLR2 in Caco-2 cells.

Fig 3.

Changes in expression levels of TLR2 in Caco-2 cells after stimulation with L. acidophilus NCFM. Caco-2 cells were stimulated with L. acidophilus NCFM at an MOI of 10 for 0, 0.5, 1, 2, and 4 h. Total protein was extracted as described in Materials and Methods. Levels of TLR2 and the internal standard protein, GAPDH, were measured by Western blotting with antibodies against TLR2 and GAPDH. Data show one representative experiment of three independent experiments.

Activation of NF-κB and p38 MAPK signaling pathways in Caco-2 cells stimulated with L. acidophilus NCFM.

TLRs have been shown to lead to the activation of the NF-κB and p38 MAPK signaling pathways, which were important in the production of many immune-related factors, including cytokines and chemokines (25). Therefore, in order to examine whether the NF-κB and p38 MAPK signaling pathways have been activated, the activation state of these two signaling pathways was studied when Caco-2 cells were stimulated with L. acidophilus NCFM at an MOI of 10 for 0 to 4 h. As shown in Fig. 4A, L. acidophilus NCFM could activate the p38 MAPK signaling pathway in IEC. The levels of p38 MAPK phosphorylation increased until 2 h and then slowly decreased, despite persistent bacterial stimulation. To verify the activation of the NF-κB signaling pathway, cell lysates were analyzed for levels of phosphorylated NF-κB p65, since phosphorylation of the NF-κB p65 subunit was associated with the activation of the NF-κB signaling pathway (11). L. acidophilus NCFM could rapidly activate the NF-κB signaling pathway with a kinetics similar to that for the p38 MAPK signaling pathway (Fig. 4B). These results demonstrated that the NF-κB and p38 MAPK signaling pathways were transiently activated when L. acidophilus NCFM stimulated Caco-2 cells and that this activation occurred before a significant increase in cytokine and chemokine expression (Fig. 1).

Fig 4.

Phosphorylation of NF-κB p65 and p38 MAPK in Caco-2 cells stimulated with L. acidophilus NCFM. Caco-2 cells were stimulated with L. acidophilus NCFM at an MOI of 10 for 0, 0.5, 1, 2, and 4 h. Total protein was extracted as described in Materials and Methods. (A) Levels of the phosphorylated forms of p38 MAPK and the internal standard protein, GAPDH, were measured by Western blotting with antibodies against Th(p)-p38 MAPK and GAPDH. (B) Levels of the phosphorylated forms of NF-κB p65 and the internal standard protein, GAPDH, were measured by Western blotting with antibodies against Ser(p)-NF-κB p65 and GAPDH. Data show one representative experiment of three independent experiments.

To further test whether the NF-κB and p38 MAPK signaling pathways are necessary for cytokine and chemokine production, the Caco-2 cells were stimulated with or without the existence of PDTC, a specific inhibitor for NF-κB, or SB203580, a specific inhibitor for p38 MAPK. The Caco-2 cells were preincubated with PDTC (40 μM) or SB203580 (20 μM) for 30 min and then treated with L. acidophilus NCFM for 2 h. Inhibition of the NF-κB or p38 MAPK signaling pathway resulted in a partial yet significant decline (P < 0.05) in cytokine and chemokine expression compared to results for the uninhibited groups treated with bacteria only (Fig. 5). The above results suggested that L. acidophilus NCFM could rapidly induce IL-1α, IL-1β, CCL2, and CCL20 production through NF-κB and p38 MAPK signaling pathways in Caco-2 cells.

Fig 5.

Suppression of L. acidophilus NCFM-stimulated cytokine and chemokine production by the NF-κB or p38 MAPK inhibitor. Caco-2 cells were preincubated with SB203580 (20 μM) or PDTC (40 μM) for 30 min and then stimulated with L. acidophilus NCFM at an MOI of 10 for 2 h. Total RNA was extracted and analyzed by real-time RT-PCR. Each bar represents the combined mean value (± SD) for three experiments. *, P < 0.05 compared to the uninhibited groups treated with L. acidophilus NCFM only. CTR, control.

DISCUSSION

It is well known that cytokines and chemokines, which affect the immune cells scattered in the GI tract and recruit immune cells to the GI tract, respectively, play a major role in mediating immune and intestinal inflammatory responses (13, 19, 24, 36). Recently, it was been reported that commensal bacteria, such as L. rhamnosus, L. acidophilus, and E. coli, could upregulate the production of many members of the cytokine and chemokine family, such as IL-1, CCL2, and CCL20 (3, 11, 21), although some studies have shown that the intestine appeared to be tolerant toward commensal bacteria (23, 29).

In line with these studies, our data also showed that L. acidophilus NCFM induced the production of some cytokines (IL-1α and IL-1β) and chemokines (CCL2 and CCL20) which were of crucial importance in the control of normal homeostasis and host gut immunity. IEC showed a rapid but transient upregulation of cytokines and chemokines (Fig. 3) despite the persistence of bacterial stimulation. Cytokines and chemokines are a “double-edged sword,” and they play an important role in enhancing host immunity, but uncontrolled overexpression has been implicated in epithelial tissue damage and many intestinal pathologies, including chronic intestinal inflammation, especially in the genetically susceptible (33). The nonoverexpression of cytokines and chemokines as a result of L. acidophilus NCFM treatment suggested that the normal IEC had developed feedback mechanisms to control mucosal immune responses to constant challenge by commensal bacteria (33). It has been demonstrated that IEC, which remain hyporesponsive to commensal bacteria, can respond to nonpathogenic bacteria in the presence of human peripheral blood mononuclear cells (PBMC), suggesting that bacterial signaling in the intestinal tract requires a network of cellular interactions (10). However, we found that L. acidophilus NCFM exerted an inflammatory activation pattern in vivo similar to that in vitro with respect to cytokine (IL-1α and IL-1β) and chemokine (CCL2 and CCL20) expression. However, the fold change of immune-related gene expression in vivo was relatively lower than that in vitro, which may be explained by the fact that bacteria populations existing in the epithelial surfaces are complex and interactions might occur between different bacteria in vivo (21). Moreover, we found that L. acidophilus NCFM-induced cytokine and chemokine upregulation appeared to be strain specific. L. acidophilus JCM 1132^T did not stimulate cytokine expression in IEC (12), and L. acidophilus X37 also did not induce cytokine production, except for that of IL-8 but with a low expression level (40). However, the factors by which L. acidophilus species mediate different interactions with IEC remain unclear, and further studies are necessary to analyze the different cytokine and chemokine secretion of IEC stimulated with various L. acidophilus strains.

IEC sense commensal bacteria through expression of pattern recognition molecules, such as TLRs, which are thought to recognize the signature molecules of microorganisms during the early period of innate immune responses (21, 26). It has been reported that IEC could induce TLR2 and TLR4 when responding to commensal bacteria, but TLR2 was mainly involved in response to Gram-positive bacteria (6, 18, 21, 26). Commensal bacteria, such as L. casei, L. rhamnosus, L. plantarum and B. lactis, all activated TLR2 in many cells, including IEC and macrophages (6, 18, 33). In this study, we found that the expression of TLR2 was upregulated in a rapid manner in IEC after treatment with L. acidophilus NCFM compared to results for the unstimulated controls (Fig. 3), which is in line with a study that showed that L. acidophilus NCFM could activate TLR2 in HEK293 cells (20), while others indicated that the mouse fetal epithelial cells were nonresponsive to the expression of TLR2 after L. acidophilus NCFM stimulation (40). It is likely that different cells respond differently even to the same bacteria (5).

The consequences of signaling through TLRs have been reported to trigger both NF-κB and p38 MAPK activation. These play important roles in the production of cytokines and chemokines involved in regulating immune responses (25). Y. G. Kim and colleagues have shown that the p38 MAPK signaling pathway was important for the production of cytokines in L. casei-treated mouse spleen cells, whereas NF-κB P65 also contributed but to a lesser extent (18). B. lactis has also been shown to induce cytokine IL-6 gene expression in IEC through the NF-κB and p38 MAPK signaling pathways (33). In this study, phosphorylation of NF-κB p65 and p38 MAPK in Caco-2 cells was shown to be rapidly but transiently enhanced in the L. acidophilus NCFM-treated groups (Fig. 4), indicating that both signaling pathways could be activated by L. acidophilus NCFM. Consistent with our findings, previous studies also showed that the direct contact of L. acidophilus NCFM with IEC was able to activate the NF-κB pathway (32). Inhibition of the NF-κB or p38 MAPK signaling pathway, using the specific inhibitor PDTC or SB203580, respectively, significantly reduced cytokine (IL-1α and IL-1β) and chemokine (CCL2 and CCL20) production in Caco-2 cells after stimulation by L. acidophilus NCFM (Fig. 5). These results suggested that activation of both NF-κB and p38 MAPK could play an important role in augmenting the production of cytokines and chemokines by L. acidophilus NCFM. It has been reported that p38 MAPK had numerous direct and indirect interactions with NF-κB (4, 34), so it is necessary to further examine the role of the interactions of the p38 MAPK and NF-κB signaling pathways in cytokine and chemokine production.

Interestingly, both in vivo and in vitro data demonstrated that the cytokines (IL-1α and IL-1β) showed a lower expression level with L. acidophilus NCFM treatment than chemokines (CCL2 and CCL20) (Fig. 1 and 2). Some studies also showed that expression of proinflammatory cytokines secreted by IEC stimulated with an agonist or bacteria was generally much lower than that observed for the chemokines (8, 15). IL-1α and IL-1β are proinflammatory mediators which have been shown to induce chemokine responses. IL-1α can upregulate CCL20 mRNA expression and protein production in IEC lines, including Caco-2 cells and HT-29 cells (14). IL-1β has been shown to significantly induce CCL2 and CCL20 expression in Caco-2 cells or macrophages (10, 39). Fichorova et al. and Perkins also demonstrated that IL-1 would induce the secretion of chemokines such as CCL2 via the NF-κB signaling pathway (9, 30). In line with these reports, the NF-κB signaling pathway was reported to be important for IL-1β-stimulated CCL2 production in rat astrocytes, and the MAPK signaling pathway also contributed (37). In addition, IL-1β was able to induce phosphorylation of p38 MAPK in IEC-6 cells (2). Taken together, the synergism between cytokines, chemokines, and L. acidophilus NCFM may be explained as follows. L. acidophilus NCFM induced an early-phase response with subsequent cytokine (IL-1α and IL-1β) and chemokine (CCL2 and CCL20) production through the TLR2-mediated NF-κB and p38 MAPK signaling pathways in Caco-2 cells. Then, the secreted cytokines (IL-1α and IL-1β) might have further stimulated the cells through the NF-κB and p38 MAPK signaling pathways, which initiated a late-phase response to express the chemokines (CCL2 and CCL20). However, further studies would be required to determine whether IL-1α and IL-1β have important roles as chemokine-inducing factors in L. acidophilus NCFM-stimulated Caco-2 cells.

In this study, our data demonstrated that the commensal bacterium L. acidophilus NCFM can induce cytokine and chemokine production in IEC, with the cytokines showing a lower expression level with bacterial treatment than chemokines. This may help to provide important insights to elaborate the host immune responses triggered by probiotic bacteria. Moreover, L. acidophilus NCFM could induce TLR2 signaling to trigger cytokine and chemokine expression in IEC through the NF-κB and p38 MAPK signaling pathways, and the activation in IEC after L. acidophilus NCFM stimulation is rapid but transient. Although we examined the signaling pathways involved, the study does not fully reveal the mechanisms, and further research is needed. Together, this study allows for a better understanding of how L. acidophilus NCFM contributes to the immune responses of the host, and it will be important in establishing the basis for further studies on the molecular mechanisms of interactions between commensal bacteria and the host.