Anti-inflammatory effects of Bifidobacterium longum subsp infantis secretions on fetal human enterocytes are mediated by TLR-4 receptors

The pathogenesis of necrotizing enterocolitis (NEC) involves the use of the Toll-like receptor-4 (TLR-4). Probiotics or their secretions protect premature infants against NEC through anti-inflammatory mechanisms. Using secretions from Bifidobacterium longum supp infantis, we show that fetal enterocyte reduction of IL-6 activation after IL-1β stimulation is mediated through TLR-4 and specifically affects IL-1 receptor-associated kinase 2 (IRAK-2) mRNA and c-Jun and c-Fos phosphorylation of the activator-protein 1 (AP-1) transcription factor. These observations may help in future prevention of NEC.

Abstract

The therapeutic and preventive application of probiotics for necrotizing enterocolitis (NEC) has been supported by more and more experimental and clinical evidence in which Toll-like receptor 4 (TLR-4) exerts a significant role. In immune cells, probiotics not only regulate the expression of TLR-4 but also use the TLR-4 to modulate the immune response. Probiotics may also use the TLR-4 in immature enterocytes for anti-inflammation. Here we demonstrate that probiotic conditioned media (PCM) from Bifidobacterium longum supp infantis but not isolated organisms attenuates interleukin-6 (IL-6) induction in response to IL-1β by using TLR-4 in a human fetal small intestinal epithelial cell line (H4 cells), human fetal small intestinal xenografts, mouse fetal small intestinal organ culture tissues, and primary NEC enterocytes. Furthermore, we show that PCM, using TLR-4, downregulates the mRNA expression of interleukin-1 receptor-associated kinase 2 (IRAK-2), a common adapter protein shared by IL-1β and TLR-4 signaling. PCM also reduces the phosphorylation of the activator-protein 1 (AP-1) transcription factors c-Jun and c-Fos in response to IL-1β stimulation in a TLR-4-dependent manner. This study suggests that PCM may use TLR-4 through IRAK-2 and via AP-1 to prevent IL-1β-induced IL-6 induction in immature enterocytes. Based on these observations, the combined use of probiotics and anti-TLR-4 therapy to prevent NEC may not be a good strategy.

NEW & NOTEWORTHY

The pathogenesis of necrotizing enterocolitis (NEC) involves the use of the Toll-like receptor-4 (TLR-4). Probiotics or their secretions protect premature infants against NEC through anti-inflammatory mechanisms. Using secretions from Bifidobacterium longum supp infantis, we show that fetal enterocyte reduction of IL-6 activation after IL-1β stimulation is mediated through TLR-4 and specifically affects IL-1 receptor-associated kinase 2 (IRAK-2) mRNA and c-Jun and c-Fos phosphorylation of the activator-protein 1 (AP-1) transcription factor. These observations may help in future prevention of NEC.

human intestinal epithelial cells (IEC) are the first line of defense by the innate immune system because of their critical role in host interactions with both invading pathogens and innocuous commensal microbiota (23). IEC recognize bacterial components via pattern-recognition receptors (PRRs) to maintain intestinal homeostasis (44). Failure of homeostasis may lead to the development of intestinal inflammatory conditions such as necrotizing enterocolitis (NEC), which occurs principally in premature infants. NEC is the leading cause of morbidity and mortality from gastrointestinal disease in the preterm infant population (14, 39). Intestinal inflammation in prematures is likely caused by a combination of IEC immaturity and microbiota imbalance. Toll-like receptor 4 (TLR-4), a major PRR present in IEC, exists on the enterocyte surface in increased amounts in the premature intestine (18, 33). TLR-4 is involved in both induction of tolerance as well as inflammation suggesting conflicting functions (6, 19). IL-1β, secreted by both immune cells and IECs (4, 51), is one of the major inflammatory cytokines contributing to the development of NEC. IL-1β activates a signaling pathway via the IL-1β receptor (IL-1βR). which shares an intracellular Toll/interleukin-1 receptor (TIR) homology domain with TLR-4. These two receptor pathways also share adapter proteins involved in signal transduction such as interleukin-1 receptor-activated protein kinase 2 (IRAK-2) (21, 41). These complexities in cell signaling must be appropriately regulated to maintain intestinal homeostasis. However, how IL-1β and TLR-4 signals cross talk in immature enteroctye inflammation is not completely understood.

A growing body of experimental and clinical evidence supports the concept that probiotics may have a preventative or therapeutic application in several gastrointestinal inflammatory disorders, including NEC. It also appears that TLR-4 may play a significant role (1, 25, 28). In immune cells, such as the macrophage, probiotics not only regulate the expression of TLRs but also use these receptors to stimulate or modulate an innate immune response (29). Bifidobacterium longum subsp infantis is a common commensal found in the newborn intestine (46). This organism has been shown to have many useful functions in the management of experimental (45) and clinical NEC (1, 2). We have reported that a secreted factor(s) from this probiotic can prevent IL-6-induction in response to IL-1β stimulation in immature enterocytes (13). This secreted factor has been partially characterized as a 5- to 10-kDa glycan or glycolipid (8, 13). In this study, we have begun to determine the mechanism by which this factor inhibits inflammation in human premature enterocytes and in enterocytes from NEC patients. The process appears to be mediated via the TLR-4 receptor. Furthermore, these secretions affect signal transduction down-regulation of AP-1 transcription factors c-Jun and c-Fos phosphorylation as well as the IRAK-2 gene.

MATERIALS AND METHODS

Bacterial cultures and isolation of probiotic-conditioned media.

Bifidobacterium longum subsp infantis (B. infantis) (ATCC No. 15697) was obtained from American Type Culture Collection (ATCC, Manassas, VA), cultured as recommended by ATCC, and stored individually in Mann-Rogosa-Sharpe (MRS) broth (DIFCO; BD Bioscience, Franklin Lakes, NJ) containing 20% glycerol at −80°C. A 50-ml Falcon tube containing MRS broth supplemented with 0.5 g/l of cysteine was inoculated with a single colony of B. infantis. The inoculum was cultured at 37°C in an anaerobic chamber (Thermo Forma 1029; Thermo Fisher Scientific) until it reached a stationary phase of growth (OD600 >1.0) as described previously (8, 13). Probiotic conditioned media (PCM) at the stationary growth phase was prepared by centrifugation of probiotic cultures at 1,600 g for 10 min and then by use of 0.22-μm filtration to eliminate residual bacteria The efficacy of bacterial depletion from the conditioned media was determined by plating and dilutions. The tested filtrate was then used in PCM experiments. Free bacteria in the absence of PCM was developed as follows: the pellet of the filtered bacteria was washed with PBS once to remove additional PCM and resupended in antibiotic-free H4 media as isolated bacteria. Isolated, PBS washed B. infantis was also heated to 100°C for 15 min to prevent further secretion of PCM. Isolated free and heated free bacteria were then exposed to H4 cells at a dose of 1 × 10⁷ organisms for 1 and 24 h before exposure to 1 ng/ml of recombinant human IL-1β for 24 h. Cell supernatants were then tested by ELISA for IL-6 production.

H4 cell line and NEC enterocytes subjected to TLR-4 siRNA transfection.

H4 cells, a human fetal nontransformed primary small intestinal cell line characterized by our laboratory(42), were cultured in Dulbecco′ modified Eagle's medium (DMEM; GIBCO Thermo Fisher Scientific, Woburn, MA) with 10% fetal bovine serum (FBS; Mediatech, Manassa, VA), 0.5 U/ml insulin (Eli Lilly, Indianapolis, IN), 2 mM l-glutamine, 0.1 mM MEM nonessential amino acids, 10 mM HEPES buffer, 100 unit/ml penicillin, and 100 μg/ml streptomycin (all purchased from GIBCO Thermo Fisher Scientific). H4 cells were transfected with stealth human TLR-4 siRNA or control siRNA following the manufacturer's instructions (Invitrogen Thermo Fisher Scientific, Grand Island, NY). Briefly, for each well of sixwell plates to be transfected, RNAi and Lipofectamine RNAiMAX complexes were prepared as follows: 25 pmol of diluted RNAi in 400 μl Opti-MEM I medium without serum; and 4 μl of Lipofectamine RNAiMAX were added to each well containing the diluted RNAi molecules and incubated for 20 min at room temperature. H4 (3 × 10⁵) cells were added in 600 μl of antibiotic free H4 growth media and then added to each well resulting in 25 nM RNAi in the culture. After 24 h of incubation at 37°C, an additional 1 ml of H4 growth media without antibiotics was added to each well resulting in a 12.5 nM final concentration. All transfection reagents were purchased from Invitrogen Thermo Fisher Scientific. Efficiency was analyzed by real-time quantitative reverse transcription PCR (qRT-PCR) and red oligo staining 72 h after the transfection (the knockdown rate was 71.6%, and transfection rate was 94.8%). The cells were incubated with PCM or MRS control media for 24 h beginning at 48 h after transfection and then exposed or not to 1ng/ml of recombinant human IL-1β (R&D Systems, Minneapolis, MN) for 6 and 24 h. The total RNA at 6 h or the supernatants at 24 h were collected and stored at −80° or −20°C, respectively, for a later real-time qRT-PCR or enzyme-linked immunosorbent assay (ELISA) analysis.

NEC-IEC were isolated and cultured from the viable margins of resected ileal NEC tissues from a NEC neonate at 25-wk gestation (13, 36) . NEC-IEC were cultured in Opti-Minimal Essential Medium (Opti-MEM; GIBCO Thermo Fisher Scientific) supplemented with 10% heat-inactivated FBS (Mediatech), 0.2 U/ml insulin (Eli Lilly), 20 ng/ml epidermal growth factor (EGF; EMD Millipore, Billerica, MA), and a 1% antibiotic-antimycotic cocktail (GIBCO Thermo Fisher Scientific) in a sterile cell culture humidifier at 32°C and 5% CO2. TLR-4 transfection was applied as described with the H4 cell line. Efficiency was analyzed by real time qRT-PCR (96 h after the transfection) The knockdown rate reached 71%. The cells were incubated with PCM or control MRS media for 24 h beginning at 48 h after transfection and then exposed or not to 1 ng/ml of recombinant human IL-1β (R&D Systems, Minneapolis, MN) for 24 h. The supernatants were then collected and stored at −20°C for a later ELISA analysis.

Human fetal small intestinal xenografts and TLR-4 siRNA transfection.

Human fetal small intestinal xenografts, as described previously (37, 38), were used to confirm the H4 cell line effect. Briefly, 2 cm of fetal small intestine were implanted subcutaneously into severe combined immunodeficiency (SCID) mice. Successful xenografts, as determined by viability testing of representative sections (37, 38), were harvested at 30 wk after transplantation. Based on previous studies (37, 38), other publications (24, 30), and the manufacturer's instructions for use of Lipofectamine RNAiMAX (Invitrogen Thermo Fisher Scientific), xenografts were cut into 3-mm pieces and immediately incubated in a premixed media (mixed 30 min before the tissue was added) containing 100 μl of Opti-MEM I medium (GIBCO Thermo Fisher Scientific), 7.5 pmol siRNA, and 1 μl of Lipofectamine RNAiMAX per tissue in a 1.5 ml-eppendorf tube at room temperature for 30 min. Tissues and media (one tissue per well) were then transferred to 48-well BD-Falcon Tissue Culture Plates (Becton Dickinson, Franklin Lakes, NJ) and 150 μl of antibiotic-free organ culture media made by Opti-MEM I medium supplemented with 10% FBS, 2 mM l-glutamine, 10 mM HEPES buffer, 0.5 U/ml insulin, and 200 ng/ml EGF (EMD Millipore), 5 mg/ml transferrin, 0.06 mM sodium selenite, 200 nM hydrocortisone (Sigma, St. Louis, MO) resulting in 30 nM RNAi per culture. This mixture was incubated at 37°C for 24 h, and then 250 μl of additional antibiotic free organ culture media were added to each well resulting in a final concentration of 15 nM RNAi. The transfection efficiency was analyzed by real-time PCR 72 h after the transfection. The knockdown rate was 91%. The tissues were pretreated with PCM or MRS for 24 h starting at 48 h after transfection and then exposed or not exposed to 1ng/ml recombinant human IL-1β (R&D Systems) for 24 h. Supernatants were collected and stored at −20°C for ELISA analysis. All experiments utilizing human tissues were previously approved by the Partners Human Research Committee (Protocol No. 1999-P-003833). All animal experiments were previously approved by the Subcommittee on Research Animal Care, Massachusetts General Hospital (Protocol No. 2005N000040).

Fetal mouse model.

C57BL/6J wild-type (WT) mice and TLR-4^−/− and TLR2^−/− mice, both on a C57BL/6J background (Jackson Laboratory), were used. All mice were bred and housed in a specific pathogen free facility. Animals were given water and standard laboratory chow ad libitum. Timed pregnant mice were established by pairing 10–12-wk old female mice with proven breeder males just before the end of the daily light cycle. The following morning, each female was examined for the presence of an ejaculatory plug in the vagina. When noted, the female was placed in a dated cage and considered pregnant, e.g., embryonic day (E) 0.5. Pups were delivered by caesarean section at E 18.5 (33). Ileal tissues were then collected for experiments. Animal procedures had been previously approved by the Massachusetts General Hospital Subcommittee on Research Animal Care and Use committee (A3596-01).

Fetal mouse ileal organ culture.

WT, TLR-4^−/−, and TLR-2^−/− mouse fetal ileal tissues harvested at E 18.5 were cut into 3-mm pieces and maintained in organ culture media as described in the xenograft organ culture section but supplemented with an antibiotic-antimycotic cocktail (100 unit/ml penicillin, 100 μg/ml streptomycin, and 0.25 μg/ml fungizone antimycotic) (GIBCO Thermo Fisher Scientific). After 2 h at 37°C, tissues were pretreated with PCM from B. infantis or MRS media alone as a control for 24 h and then stimulated with 1 ng/ml of recombinant mouse IL-1β (R&D Systems) for 24 h. Supernatants were collected and stored at −20°C for ELISA analysis.

ELISA.

Levels of IL-6 were measured in culture supernatants using ELISA kits (R&D Systems) according to the manufacturer's instructions. IL-6 was quantified in each supernatant in triplicate. Colorimetric results were read at a wavelength of 450 nm. Values were normalized to total protein in cells or organ cultures. Protein was determined by a bicinchoninic acid protein assay (Pierce, Rockford, IL) modified for 96-well microtiter plates according to the manufacturer's protocol.

Immunofluorescent staining.

H4 cells cultured on Falcon Culture Slides (Thermo Fisher Scientific) were transfected with human stealth TLR-4 siRNA or control siRNA and then pretreated with PCM for 24 h as described above and stimulated with IL-1β for 2 h. The cells were washed with phosphate-buffered saline (PBS; GIBCO Thermo Fisher Scientific) once and fixed in 4% paraformaldehyde in PBS for 15 min at room temperature (RT). Cells were washed three times with PBS to remove paraformaldehyde and then blocked in blocking buffer (1× PBS/3% bovine serum albumin/0.3% Triton X-100) for 60 min at RT. The cells were then incubated with either anti-phospho-c-Jun (P-c-Jun; 1:80) or anti-phospho-c-Fos (P-c-Fos; 1:800) in antibody dilution buffer (1× PBS/1% BSA/0.3% Triton X-100) overnight at 4°C. After being washed three times with PBS, stained with goat anti-rabbit Cy3 (1:500) at RT for 60 min, and washed again with PBS for three times, the cell membrane was stained by anti-phospholipid antibody conjugated with fluorescein isothiocyanate (FITC; 1:500) for 1 h at RT. The cells were then counterstained with 1, 5-bis {[α-(di-methylamino) ethyl] amino}-4,8-dihydroxyanthracene-9,10-dione (DRAQ5) (1:1, 000) (Life Technologies, Grand Island, NY) for 20 min at RT for nuclear visualization. Specimens were then washed three times with PBS, mounted with ProLong Antifade Reagent (Life Technologies) and analyzed with a fluorescent Leica confocal microscope. The fluorescent intensity was analyzed with ImageJ (fiji-win64) software. Corrected total cell fluorescence (CTCF) = integrated density − (area of selected cell × the mean fluorescence of background readings) (31) was then calculated.

Real-time quantitative reverse transcription PCR.

The RNA RNeasy Mini kit (Qiagen, Valencia, CA) was used for extraction of total RNA from H4 cells. RNA was reverse transcribed with random hexamers using an Advantage RT-for -PCR kit (Clontech, Mountain View, CA). The cDNA was amplified using iQ SYBR Green Supermix (Bio-Rad, Philadelphia, PA) and 500 nM of each primer specified. β-Actin primers were amplified in all samples. Triplicate cDNA samples were amplified with the following primers: mean threshold cycle (CT) values of each transcript were normalized by subtracting the mean CT value for the β-actin transcript of that sample. The change in the normalized transcript level was expressed relative to the control sample with a change of n in CT representing a 2n or greater-fold difference, as described previously (32). Primer sequences used in this study were as follows: human-IRAK-2, forward, 5′-CCAGGCAACCGATGACTTCAA-3′, and reverse, 5′-TGGGGTGGCAGCATCTAAGA-3′; human-β-actin, forward, 5′-CATGTACGTTGCTATCCAGGC-3′, and reverse, 5′-CTCCTTAATGTCACGCACGAT-3′.

Statistical analysis of data.

Results were expressed as the mean ± SE. The unpaired Student's t-test was used to compare the mean of two groups. One-way ANOVA was used to compare the mean of multiple groups. Differences of P < 0.05 were considered significant (*P < 0.05, **P < 0.01, †P < 0.001, ‡P < 0.0001) (GraphPad Prism 6).

RESULTS

TLR-4 was necessary for the prevention of IL-6 induction by IL-1β after PCM exposure in H4 cells.

To determine the role of the TLR-4 receptor in PCM mediated anti-inflammation after IL-1β stimulation on H4 cells, TLR-4 knockdown cells were pretreated with or without PCM before exposure to IL-1β and IL-6 levels measured. The results are shown in Fig. 1. As expected, IL-1β-induced H4 cell IL-6 secretion was abolished by PCM pretreatment; PCM alone had no effect on it (Fig. 1B). However, TLR-4 knockdown reversed the PCM effect on IL-6 induction by IL-1β. In fact, PCM exposure to TLR-4 knockdown H4 cells alone increased the IL-6 response (Fig. 1C), which was additive with PCM and IL-1β.

Fig. 1.

Toll-like receptor-4 (TLR-4) was required for probiotic conditioned media (PCM) prevention of IL-1β-induced IL-6 in H4 cells. A: H4 control (Con) and TLR-4 knockdown cells were pretreated with or without PCM and then exposed to IL-1β for 24 h. B and C: cells were treated with PCM alone. The supernatants were collected to measure IL-6 levels. Inhibition of IL-6 production after PCM occurred only in control and not TLR-4 knockdown cells. PCM in TLR-4 knockdown calls enhanced the IL-6 response in a manner additive to IL-1β. The means ± SE were from triplicate wells and are representative of 3 separate experiments with similar results. *P < 0.05, **P < 0.01, †P < 0.001 (A: one-way ANOVA and post hoc tests; B and C: Student's unpaired t-test).

TLR-4 was necessary for PCM prevention of IL-1β-induced IL-6 in human fetal small intestinal xenografts.

To provide additional evidence that TLR-4 was critical for PCM prevention of IL-1β- induced IL-6 in immature human enterocytes, human fetal small intestinal xenografts were used. Organ cultures were pretreated with or without PCM before exposure to IL-1β. The supernatants of the organ cultures were collected to determine IL-6 levels. The results (Fig. 2, A and B) showed that TLR-4 gene knockdown reversed the effects of PCM in preventing IL-1β-induced IL-6 induction in human fetal small intestinal xenografts. PCM alone in TLR-4 gene knockdown organ cultures increased the IL-6 response (Fig. 2C), which was additive with PCM and IL-1β.

Fig. 2.

TLR-4 was required for PCM prevention of IL-1β induced IL-6 in human fetal small intestinal xenografts. Control (con) and TLR-4 knockdown fetal intestinal xenografts were studied. A: the xenografts were pretreated with or without PCM and then stimulated with IL-1β. B and C: the xenografts were also treated with PCM alone. PCM inhibited IL-6 secretion in control but not TLR-4 knockdown xenografts. PCM alone in TLR-4 knockdown xenografts actually increased IL-6 secretion which was additive with IL-1β stimulation. The means ± SE were from triplicate wells and are representative of 3 separate experiments with similar results, *P < 0.05, **P < 0.01, †P < 0.001, ‡P < 0.0001 (A: one-way ANOVA and post hoc tests; B and C: Student's unpaired t-test).

TLR-4 was necessary for PCM prevention of IL-1β-induced IL-6 in mouse fetal ileum.

To investigate if TLR-4 was necessary in PCM prevention of IL-1β-induced IL-6 in immature mouse small intestine, an E 18.5 mouse fetal ileal organ culture model (33) was used in WT, TLR-4^−/−, and TLR2^−/− mice. Mouse fetal ileum were pretreated with or without PCM before exposure to IL-1β. The supernatants were collected to detect IL-6 levels (Fig. 3).The results showed that PCM attenuated IL-1β-induced Il-6 in WT (Fig. 3A) and TLR2^−/− (Fig. 3C) mice but not in TLR-4^−/− (Fig. 3B) mice. As with H4 cells and fetal xenografts, PCM alone increased IL-6 in TLR-4^−/− fetal mice (Fig. 3E), which was additive with PCM and IL-1β.

Fig. 3.

TLR-4 was required for PCM reduction of IL-1β-induced IL-6 in fetal mouse small intestinal organ cultures. Embryonic 18.5 day (E 18.5) mouse fetal small intestinal organ culture were pretreated with and without PCM in wild-type (WT) mice (A and D), TLR-4^−/− mice (B and E), and TLR2^−/− mice (C and F). and then stimulated with or without IL-1β. IL-6 levels in the supernatants were assayed. PCM inhibition of IL-6 secretion after IL-1β stimulation occurred in WT and TLR2^−/− but not TLR-4 organ cultures. PCM alone increased IL-6 secretion and was additive with IL-1β. The means ± SE are from triplicate wells and are representative of 3 separate experiments with similar results, **P < 0.01, †P < 0.001 (A-C: one-way ANOVA and post hoc tests; D–F: Student's unpaired t-test).

B. infantis organisms alone failed to reduce IL-1β induced IL-6 in H4 cells.

To determine if B. infantis organisms alone in the absence of PCM could prevent IL-1β induced IL-6, H4 cells were pretreated with isolated and with isolated, heat-inactivated B. infantis for 1 and 24 h before stimulation with IL-1β. The results showed that isolated B. infantis at 1 h but not at 24 h (at a time with PCM had reaccumulated) failed to prevent IL-6 production. In contrast, isolated, heat-inactivated B. infantis failed to prevent IL-6 production both at both 1 and 24 h presumably because the inactivated cells did not produce PCM (Fig. 4).

Fig. 4.

Isolated Bifidobacterium longum subsp infantis organisms in the absence of PCM failed to prevent IL-6 induction by IL-1β in H4 cells H4 cells were incubated with isolated (A and B) and isolated, heat-inactivated B. infantis (C and D) for 1 and 24 h before exposure to IL-1β, Isolated B. infantis at 1 h but not at 24 h failed to inhibit IL-1 production. Isolated, heat-treated B. infantis organisms at both 1 and 24 h failed to inhibit IL-6 production. The means ± SE are from triplicate wells and are representative of 3 separate experiments with similar results. †P < 0.001 (one-way ANOVA and post hoc tests).

TLR-4 was required for PCM prevention of IL-1β-induced IL-6 in NEC-IEC.

To investigate the role of the TLR-4 receptor in PCM prevention of IL-1β induced IL-6, NEC-IEC and TLR-4 knockdown NEC-IEC cells were exposed to IL-1β with and without PCM pretreatment. The cell supernatants were collected for IL-6 levels. The results are shown in Fig. 5. In a manner similar to H4 cells, IL-1β-induced NEC-IEC IL-6 secretion and PCM pretreatment abolished this response (Fig. 5A), PCM alone had no effect on it (Fig. 5B). However, TLR-4 knockdown reversed the role of PCM prevention of IL-6 induction by IL-1β. In fact, PCM exposure alone to TLR-4 knockdown NEC-IEC increased the IL-6 response (Fig. 5C), which was additive with PCM and IL-1β.

Fig. 5.

TLR-4 was required for PCM prevention of IL-1β-induced IL-6 in necrotizing enterocolitis intestinal epithelial cells (NEC-IEC). Control (con) and TLR-4 knockdown NEC-IEC cells were pretreated with or without PCM and then exposed to IL-1β (A) or treated with PCM alone (B and C). IL-6 levels from control but not TLR-4 knockdown NEC-IEC cells was inhibited. PCM alone in TLR-4 NEC-IEC knockdown cells stimulated IL-6 in an additive manner with IL-1β. The means ± SE were from triplicate wells and are representative of 3 separate experiments with similar results. *P < 0.05, **P < 0.01, †P < 0.001, ns no significant difference (A: one-way ANOVA and post hoc tests; B and C: Student's unpaired t-test).

B. infantis secretion inhibition of IRAK-2 mRNA was TLR-4 dependent.

We have previously reported by microarray analysis of fetal enterocyte (H4) RNA after exposure to B. infantis secretions followed by an IL-1β stimulus that the IRAK-2 gene was inhibited (17). To investigate if TLR-4 was involved in IRAK-2 signaling regulation after B. infantis secretion exposure, control and TLR-4 knockdown H4 cells were tested (Fig. 6). The results (Fig. 6A) showed that PCM pretreatment in controls but not TLR-4 knockdown H4 cells reduced IRAK-2 mRNA expression. However, PCM in TLR-4 knockdown cells actively increased IRAK-2 mRNA expression (Fig. 6, B and C) rather than attenuating it and was additive with IL-1β. To provide evidence that IRAK-2 plays a key role in anti-inflammation in immature enterocytes, IRAK-2 was knocked down in H4 cells. The results (Fig. 6D) showed that IRAK-2 gene knockdown significantly reduced the IL-6 response to IL-1β.

Fig. 6.

PCM attenuation of interleukin-1 receptor-associated kinase 2 (IRAK-2) mRNA was TLR-4 dependent. A: control (con) and TLR-4 knockdown H4 cells were pretreated with or without PCM, exposed to IL-1β and then IRAK-2 mRNA measured. TLR-4 knockdown H-4 cells failed to attenuate the IRAK-2 increase. PCM alone had no effect on IRAK-2 mRNA fold levels in controls (B) but increased IRAK-2 mRNA fold levels in TLR-4 knockout H4 cells (C). D: when control and IRAK-2 knockdown H4 cells were exposed to IL-1β, IL-6 levels were reduced only in the IRAK-2 knockdown cells. The means ± SE were from triplicate wells and are representative of 3 separate experiments with similar results, *P < 0.05, **P < 0.01, †P < 0.001 (one-way ANOVA and post hoc tests).

PCM from B. infantis secretions reduced IL-1β-stimulated P-c-jun and P-c-fos expression in control but not TLR-4 knockdown H4 cells.

To investigate if AP-1 transcription factor activation by IL-1β was involved in PCM anti-inflammation and if this process was TLR-4-dependent (Fig. 7), P-c-Jun and P-c-Fos was determined in control (Fig. 7, A1 and A2), and TLR-4 knockdown H4 cells (Fig. 7, B1 and B2) after pretreatment with PCM and before exposure to IL-1β. P-c-Jun staining was used. The results showed that P-c-Jun was increased significantly after IL-1β stimulation and PCM stimulation reduced the response in control but not in TLR-4 knockdown cells. The P-c-Jun increase by IL-1β was additive by pretreatment with PCM in TLR-4 knockdown cells In like manner, P-c-Fos was increased by IL-1β and PCM reduced the effect in control but not in TLR-4 knockdown cells (data not shown).

Fig. 7.

PCM from B. infantis secretions prevented IL-1β-induced P-c-Jun expression and was TLR-4 dependent. Control and TLR-4 knockdown H4 cells were pretreated with or without PCM and then exposed or not exposed to IL-1β before P-c-Jun was detected by immunofluorescent staining (red). The nuclei were counterstained by DRAQ5 (blue), and the cell membrane was stained green. P-c-Jun levels were increased by IL-1β and inhibited by PCM in control (A1) but not TLR-4 knockdown H4 cells (B1). PCM alone in TLR-4 knockdown cells increased P-c-Jun and was additive with IL-1β. The images are representative of 3 separate experiments with similar results. The analysis (A2 and B2) applied by corrected total cell fluorescence (CTCF) was then determined. *P < 0.05, † P < 0.001, ns, no significant difference, one-way ANOVA and post hoc tests. Amplified ×600; scale bar = 20 μm; n = 30–40 cells/treatment.

DISCUSSION

This laboratory has had a long-standing interest in the interaction between colonizing bacteria and the developing human intestine. We have shown that colonizing bacteria interacting with the premature human intestine preferentially stimulates inflammation over immune homeostasis (32, 38, 36). We have also reported that the stimulation by certain bacteria can mediate excessive inflammation through activation of immature innate immune inflammatory response genes in fetal enterocytes (13, 35). Yet, there is also clinical and experimental evidence to suggest that probiotics and their secretions may also have a protective effect against this enhanced inflammatory response (1, 45) suggesting a paradoxical situation. This observation may be important clinically. A major intestinal inflammatory condition, necrotizing enterocolitis (NEC), occurs commonly in premature infants <1,500 g (14, 39).

NEC is associated with multiple risk factors after premature birth including hypoxic-ischemic events, the use of formula feeding, and a dysbiotic bacterial colonization resulting in an increased likelihood of severe intestinal inflammation, coagulation necrosis and gastrointestinal bleeding (14, 39). We believe that a major pathogenic mechanism of NEC is the accentuated response of the immature intestine to colonizing bacteria. Yet, paradoxically a major probiotic that has been shown clinically to protect against the expression of NEC is Bifidobacterium longum subsp infantis (3, 20, 45). Accordingly, we have carefully studied the mechanism by which this probiotic and its secretions alter enterocyte inflammation in the immature human intestine.

In this study we focused on the effect of B. infantis secretions on the IL-1β-induced inflammatory response in premature human enterocytes. IL-1β is an important inflammatory cytokine associated with the pathogenesis of NEC and functions by mediating inflammatory cell recruitment and amplifying the innate immune response via enterocyte signaling (47, 49). The gut is colonized with trillions of beneficial commensal microorganisms that normally maintain a mutually beneficial symbiotic relationship with the host. To maintain intestinal homeostasis, both IEC and gut-associated immune cells recognize bacterial components via PRRs to conserve a balance between tolerance to the large communities of resident luminal bacteria while at the same time being able to mount an inflammatory response against pathogens (34). TLRs are a major class of PRRs that are present on IECs and mucosal immune cells which are involved in the induction of both tolerance and inflammation (15).

TLR-4 is one of 10 or more members of the TLR family. TLR-4 is expressed at low levels and internally in adult enterocytes but its expression is much higher on the surface in human fetal enterocytes (18, 33) suggesting its possible importance in immature intestinal bacterial-enterocyte interaction. The network of TLR-4 and IL-1β signaling is interwoven as they share several adaptor proteins such as IRAK-2. Accordingly, a study of the interaction between TLR-4 and IL-1β signaling in immature enterocyte inflammation may provide a better understanding of the pathogenesis and prevention of NEC.

Although the pathogenesis of NEC is incompletely understood, experimental and clinical evidence support the preventive application of probiotics in its intestinal inflammation. Furthermore, TLR-4 is implicated in its expression (25, 27, 48). We have found that B. infantis secretions prevent IL-1β-induced IL-6 stimulation and modulate TLR-4 expression in immature enterocytes (13, 35). However, the underlying molecular mechanism in the enterocyte is not completely understood. TLR-4 appears to have a dual role in this process with controversial functions noted in breast cancer (19), asthma and chronic obstructive pulmonary disease (COPD) (5). TLR-4 also mediates intestinal inflammation in NEC (15, 18, 25) and these investigators suggest that the use of a TLR-4 antagonist may be one approach to the treatment of NEC. However, we now provide evidence to suggest that TLR-4 may also participate in anti-inflammation under specific conditions such as with B. infantis secretions of anti-IL-1β-induced inflammation. This suggests that there may be crosstalk between TLR-4 and IL-1β signaling in immature intestinal inflammation. This study was designed to address this issue.

Accordingly, we used a human fetal small intestinal epithelial cell line (H4 cells) and knocked down its TLR-4 gene. The cells were then pretreated with or without probiotic secretions and stimulated with IL-1β before measuring IL-6. The results showed that B. infantis secretions prevented IL-6 induction in response to IL-1β in control but not in TLR-4 gene knockdown cells. These data suggest that TLR-4 is a participant in the B. infantis secretion effect in prevention of the IL-6 inflammatory response in immature H4 cells. We also showed that PCM alone in TLR-4 H4 knockdown cells has an additional inflammatory effect that was additive when given with IL-1β suggesting a different mechanism for inflammation. Unfortunately, at this time we have no studies to determine this mechanism. To confirm this observation, we used organ cultures from human fetal small intestinal xenografts after TLR-4 knockdown and determined that indeed B. infantis secretions have a role in protection from inflammation via the TLR-4 receptor. In addition, we extended these observations in an in vivo fetal mouse model by applying similar conditions to embryo 18.5 day (E 18.5) fetal small intestinal organ cultures (33) from control, TLR-4^−/−, and TLR2^−/− mouse strains. The B. infantis secretions attenuated IL-6 induction in response to IL-1β in both control and TLR2^−/− mice but not in TLR-4^−/− mice. These results suggest that B. infantis secretions require the TLR-4 but not the TLR2 receptor to prevent or attenuate IL-1β - induced IL-6 induction. However, the underlying signaling mechanism was still not completely understood. When these experiments were repeated with H4 cells exposed to isolated B. infantis washed free of PCM for 1 h or allowed to incubate with H4 cells for 24 h, a period of time presumably allowing for the reaccumulation of PCM, inhibition was only apparent with 24 h incubation. When isolated B. infantis washed free of PCM were inactivated by heat treatment, presumably preventing the resecretion of PCM, for 1 and 24 h, no inhibition of IL-1 was noted suggesting that B. infantis H4 interaction alone was not involved in the anti-inflammatory effect against IL-1β stimulation. Furthermore, using primary intestinal enterocytes isolated from a resected NEC intestine, we confirmed these observations under conditions of NEC.

IL-1βR through conserved cytosolic regions called Toll- and IL-1R-like (TLR) domains (41) rapidly assembles two intracellular signaling steps, myeloid differentiation primary response gene 88 (MYD88) and IRAK-4, which form a stable IL-1β-inducible initial signaling module (7, 26). This is paralleled by the autophosphorylation of IRAK-4, which subsequently phosphorylates IRAK-1 and IRAK-2 and is followed by the recruitment and oligomerization of tumor necrosis factor-associated factor (TRAF) 6 (10, 11, 21). IRAK-1 and 2 function as both adaptor and protein kinases to transmit downstream signals. Complexes of IRAK-1, IRAK-2, and TRAF6 dissociate from the initial receptor complex. Cells lacking these proteins have impaired activation of the transcription factors nuclear factor-κB (NF-κB) and activator protein 1 (AP-1) (11, 16, 22, 50). Accordingly, disruption of these signaling pathways by PCM represents a mechanistic step in anti-inflammation and the role of TLR-4 in this process suggests its importance.

In previous reported studies from this laboratory (17), we exposed H4 cells to B. infantis secretions before an inflammatory stimulus of IL-1β. RNA from these cells was subjected to transcription profiling. The results showed that IRAK-2 mRNA was up-regulated by IL-1β and pretreatment with B. infantis secretions prevented this up-regulation. To further substantiate this observation and to determine the role of TLR-4 in this process, we knocked down the TLR-4 gene in H4 cells and pretreated the cells with or without B. infantis secretions before the IL-1β stimulation. IRAK-2 mRNA expression was then measured. The results showed that in both control and TLR-4 knockdown cells, IRAK-2 mRNA expression increased significantly. B. infantis secretion pretreatment prevented IRAK-2 mRNA induction only in control cells. These results suggest that IL-1β acts through the IRAK-2 adaptor protein in fetal human enterocytes to induce IL-6 by a TLR-4 independent pathway and B. infantis secretions may prevent this process by inhibiting IRAK-2 mRNA induction. Furthermore TLR-4 gene knockdown actually increased IRAK-2 mRNA expression in PCM pretreated cells compared with the absence of pretreatment after IL-β stimulation suggesting an enhancing affect of IL-6 in the absent of the TLR-4. These data suggest that B. infantis secretion prevention of IL-1β induced IL-6 occurs through TLR-4 to regulate IRAK-2 mRNA expression in H4 cells.

We have previously reported that AP-1 is a specific nuclear transcriptional factor for IL-1β-induced IL-6 secretion but only in immature enterocytes (9). AP-1 is composed of c-Jun and c-Fos which are phosphorylated in their active form. To investigate the role of AP-1 in probiotic secretion prevention of IL-1β-induced IL-6 in H4 cells, we used TLR-4 gene knockdown cells and pretreated these cells with or without secretions before stimulation with IL-1β. Immunofluorescent staining was used to test phospho-c-Jun (P-c-Jun) and phospho-c-Fos (P-c-Fos). B. infantis secretion exposure prevented both the P-c-Jun and P-c-Fos increase in control but not TLR-4 knockdown cells. These observations suggested that in immature enterocytes B. infantis secretion prevention of IL-1β signaling was TLR-4 dependent and mediated through IRAK-2, a common adaptor protein of IL-1β, as well as AP-1 genes. It should be noted that these studies of human fetal intestine are done with ex vivo techniques in which the stimulus (IL-1β and PCM) do not necessarily interact with the luminal surface of the intestine alone. Accordingly, to be more physiologically relevant to the human premature, in vivo studies need to be done. However, access to these tissues is difficult and ex vivo followed by clinical studies appear to be necessary.

We have previously shown that in fetal human enterocytes that TLR-4 is increased and expressed on the fetal enterocyte surface rather than within the cell (33). This expression in the human fetal intestine could serve another purpose. Here we speculate that TLR-4 expression on fetal human enterocytes may have an added function of decreasing inflammation developmentally just as toll in Drosophila fetuses (43) has been shown to be a developmental regulator and not an innate inflammatory receptor (5, 12, 40, 43) as they function in the adult. Further studies are required to support this speculation.

It is important to recognize the potential conflict between using TLR-4 antagonists and B. infantis secretions to treat premature infants to prevent NEC based on the observations in this study. These observations should be extended to clinical trials to determine if the use of B. infants or its secretions in prematures at risk for NEC may be preventative.

GRANTS

This work was supported by National Institutes of Health Grants R01-HD-12437, R0-1-HD-059126, P01-DK-033506, and P30-DK-040561 (to W. A. Walker).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

D.M. and W.A.W. conception and design of research; D.M., W.Z., and K.G. performed experiments; D.M. analyzed data; D.M., H.N.S., and W.A.W. interpreted results of experiments; D.M. prepared figures; D.M. drafted manuscript; H.N.S. and W.A.W. approved final version of manuscript; W.A.W. edited and revised manuscript.