AJP - GI AJP: Gastrointestinal and Liver Physiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Am J Physiol Gastrointest Liver Physiol 292: G767-G778, 2007. First published November 30, 2006; doi:10.1152/ajpgi.00447.2006
0193-1857/07 $8.00
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
292/3/G767    most recent
00447.2006v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Web of Science (5)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Sun, J.
Right arrow Articles by Hobert, M. E.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Sun, J.
Right arrow Articles by Hobert, M. E.

INFLAMMATION/IMMUNITY/MEDIATORS

Flagellin-induced tolerance of the Toll-like receptor 5 signaling pathway in polarized intestinal epithelial cells

Jun Sun, Pamela E. Fegan, Anjali S. Desai, James L. Madara, and Michael E. Hobert

Department of Pathology, The University of Chicago, Chicago, Illinois

Submitted 27 September 2006 ; accepted in final form 26 November 2006


   ABSTRACT
TOP
 ABSTRACT
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 
Salmonella typhimurium is a gram-negative enteric pathogen that invades the mucosal epithelium and is associated with diarrheal illness in humans. Flagellin from S. typhimurium and other gram-negative bacteria has been shown to be the predominant proinflammatory mediator through activation of the basolateral Toll-like receptor 5 (TLR5). Recent evidence has shown that prior exposure can render immune cells tolerant to subsequent challenges by TLR ligands. Accordingly, we examined whether prior exposure to purified flagellin would render human intestinal epithelial cells insensitive to future contact. We found that flagellin-induced tolerance is common to polarized epithelial cells and prevents further activation of proinflammatory signaling cascades by both purified flagellin and Salmonella bacteria but does not affect TNF-{alpha} stimulation of the same pathways. Flagellin tolerance is a rapid process that does not require protein synthesis, and that occurs within 1 to 2 h of flagellin exposure. Prolonged flagellin exposure blocks activation of the NF-{kappa}B, MAPK, and phosphoinositol 3-kinase signaling pathways and results in the internalization of a fraction of the basolateral TLR5 without affecting the polarity or total expression of TLR5. After removal of flagellin, cells require more than 24 h to fully recover their ability to mount a normal proinflammatory response. We have found that activation of phosphoinositol 3-kinase and Akt by flagellin has a small damping effect in the early stages of flagellin signaling but is not responsible for tolerance. Our study indicates that inhibition of TLR5-associated IL-1 receptor-associated kinase-4 activity occurs during the development of flagellin tolerance and is likely to be the cause of tolerance.

inflammation

THE EPITHELIAL CELLS LINING the human gastrointestinal tract act as a physical barrier between the luminal contents and the rest of the body (9, 16), and thus they provide the first line of defense against invading pathogens. Acting as guardians, enterocytes are able to recognize invading pathogens and initiate rapid secretion of proinflammatory cytokines that orchestrate the recruitment of immune cells to the site of infection. Pathogen detection is achieved through the family of Toll-like receptors (TLRs), which is comprised of at least 11 members that recognize a variety of pathogen-associated molecular patterns, including lipopolysaccharide (LPS) and flagellin (reviewed in Ref. 2). The extracellular domains of these receptors impart ligand-binding specificity, whereas the intracellular domains have a higher degree of homology and interact with cytosolic signaling proteins including MyD88, IL-1 receptor-associated kinase (IRAK), and tumor necrosis factor receptor-associated factor 6 (48), ultimately resulting in activation of nuclear factor kappa-B (NF-{kappa}B), mitogen-activated protein kinase (MAPK) pathways, and proinflammatory cytokine gene expression.

One enteric pathogen that initiates this innate immune response in the intestinal epithelium is the gram-negative pathogen Salmonella enterica serovar Typhimurium (S. typhimurium). S. typhimurium is the leading cause of gastroenteritis in humans, and, in vivo, likely breaches the epithelial barrier by first invading enterocytes and M cells of the ileum and colon (22). Flagellin from S. typhimurium is known to be a ligand for the Toll-like receptor 5 (TLR5), whose ligation induces proinflammatory gene expression in epithelial cells (11, 14, 42, 49). In polarized epithelial cells, TLR5 is localized to the basolateral membrane and thus is segregated from the luminal contents that contain flagellin (14, 15, 42). Under normal circumstances, this allows a rapid proinflammatory response to occur only when flagellin from an enteric pathogen like S. typhimurium has breached the epithelial barrier. The secreted flagellin can activate proinflammatory signaling through the basolateral TLR5, and type III effectors may provide the stimulus by which flagellin is transcytosed across the epithelium (42). However, in Crohn's disease, flagellin from commensal gut flora is believed to be the instigator of an aberrant mucosal immune response and chronic inflammation (29). This is likely to be a result of these genetically susceptible individuals having a leaky bowel mucosa (40, 45, 46), allowing flagellin from commensal gut bacteria to initiate an inappropriate immune response.

It is clear that the host innate immune response to bacterial components must be regulated to prevent an unrestrained or prolonged inflammatory immune response. Systemic LPS is known to be the principal initiator of septic shock; however, mechanisms seem to be in place that can limit the pathological effects of LPS and induce a state of hyporesponsiveness or tolerance (51). Recent in vitro studies have demonstrated that prior exposure to LPS can induce a state of tolerance in neutrophils and macrophages (31, 35). Other studies using cultured monocytes and T cells have demonstrated that prior exposure to flagellin also induces a state of tolerance to subsequent stimulation by flagellin (36). Both LPS- and flagellin-induced tolerance occurs within hours after the initial exposure (31, 35, 36); however, the precise mechanisms that regulate these induced states of tolerance have not been fully defined. Recently, the mechanisms that negatively regulate the innate immune response have begun to be investigated, and many seem to be both cell type and TLR specific. Possible mechanisms that have been identified thus far include a reduction in the surface expression of TLR2, TLR4, and the TLR4 coreceptor MD-2 (1, 41), the expression of soluble TLR2, TLR4, and MyD88 (6, 21, 24), and the expression of numerous intracellular inhibitors and regulators like the suppressor of cytokine signaling 1 and Toll-interacting protein (reviewed in Ref. 28).

The importance of flagellin-induced mucosal inflammation in both self-limiting gastroenteritis and chronic inflammation has been clearly demonstrated; however, the mechanisms by which it may be controlled in the polarized epithelium remain unknown. Recently, much attention has been focused on the development of tolerance to TLR ligands in immune cells. The mechanisms adopted by these nonpolarized cells to limit the immune response may, in fact, be quite different from those employed by the highly polarized intestinal epithelium. Indeed, the signaling pathways initiated by transmembrane receptors appear to be dictated by the membrane in which the receptor is localized (3, 23). Segregation of TLR5 receptors at the basolateral surface, and the compartmentalization of various cytoplasmic and membrane-associated signaling molecules, may provide for quite different mechanisms of tolerance in the polarized epithelium (18). In this study, we have used polarized epithelial cell model systems to explore the possibility that prior exposure to flagellin would result in functional downregulation of TLR5 responses and induce a state of hyporesponsiveness or tolerance to subsequent challenges with flagellin in polarized intestinal epithelial cells. We show here for the first time that flagellin-induced self-tolerance appears to be common to polarized epithelial cells and induces the internalization of TLR5 from the basolateral cell surface. We demonstrate that flagellin tolerance is a rapid, dose-dependent, protein synthesis-independent phenomenon that prevents flagellin signaling through the NF-{kappa}B, MAPK, and phosphoinositol 3-kinase (PI3K) pathways. Activation of PI3K by flagellin seems to damp the proinflammatory response but is not required for flagellin tolerance to develop. Furthermore, recovery from the state of flagellin tolerance requires ~24 h of incubation in flagellin-free medium before cells will again respond like nontolerant cells. The inhibition of IRAK-4 activity is associated with the flagellin-induced tolerance. It seems likely that the rapid and prolonged tolerance to flagellin will help to protect the host epithelium from an uncontrolled immune response and may well aid in the healing process after the initial inflammatory insult.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 
Cell culture and cell lines. T84 cells were grown in a 1:1 mixture of Dulbecco-Vogt modified Eagle's medium and Ham's F-12 medium supplemented with 15 mM HEPES buffer (pH 7.5) (Cambrex Bio Science, Walkersville, MD), 14 mM NaHCO3, 100 U/ml penicillin, 100 µg/ml streptomycin, and 5% newborn calf serum. HT-29, Caco2-BBE, and MDCK II cells were grown in DMEM supplemented with 10% FBS, 2 mM glutamine, 0.1 mM nonessential amino acids, 100 U/ml penicillin, 100 µg/ml streptomycin (Invitrogen, Carlsbad, CA), at 37°C in a humidified atmosphere containing 5% CO2. Monolayers of T84, HT-29, Caco2-BBE, and MDCK II cells were grown on permeable supports (0.33 or 4.67 cm2, 0.4 µm pore) (Costar, Cambridge, MA) and utilized 6–14 days (T84), 4 days (MDCK II, HT-29), or 14 days (Caco2-BBE) after being plated, as described previously (30). For clarity, the permeable supports with the attached monolayer of cells will be referred to as "monolayers." MDCK II cell lines expressing HA epitope tagged human TLR5 were generated by transfecting cells with the pUNO-HA-TLR5 plasmid (Invivogen) using the Lipofectamine 2000 reagent (Invitrogen) according to the manufacturer's recommendations. Transfected cells were selected by using media containing 10 µg/ml blasticidin (Invivogen). Individual clones were isolated by limiting dilution and screened for the expression of TLR5-HA using both HA and TLR5 antibodies.

Bacterial strains, growth conditions, and flagellin purification. Salmonella typhimurium 14028s or {chi}3306 [wild type (WT)] and Escherichia coli F-18 (commensal gut strain) were grown as previously described (32, 34). E. coli F-18 (33) was provided by Beth McCormick (Harvard University Medical School, Boston, MA). Bacterial cultures were prepared for experimental use under nonagitated, microaerophilic conditions by culturing of 0.04 ml of a stationary-phase culture in 40 ml of Luria broth overnight (15–18 h) at 37°C as previously described (32). Bacterial overnight cultures were concentrated 33-fold in Hanks’ balanced salt solution (HBSS) supplemented with 10 mM HEPES, pH 7.4. Flagellin was purified from WT S. typhimurium and E. coli F18 as previously described (14). Briefly, S. typhimurium or E. coli F-18 supernatants were filtered through 0.2-µm filters, concentrated 100-fold with Amicon 30-kDa filters (Millipore, Billerica, MA), and boiled for 20 min. The pH of the flagellin concentrate was adjusted to pH 6.0 and incubated with 0.5 ml High S cation exchange support (pH 6.0) (Bio-Rad Laboratories, Hercules, CA) and then polymyxin-B agarose (Sigma-Aldrich, St. Louis, MO) for 1 h at room temperature with mixing. Resins were removed by centrifugation. Flagellin concentrate was diluted fivefold in 20 mM Tris, pH 8.0, and applied to a 2-ml High Q anion exchange column (Bio-Rad). The column was washed with 10 ml of 20 mM Tris, pH 8.0, at a flow rate of 1.5 ml/min. A 0.0–1.0 M NaCl linear gradient was applied to the column at a flow rate of 1.5 ml/min. Fractions (1.5 ml) were collected and analyzed for flagellin by Western blot. Fractions containing flagellin were pooled and assayed for bioactivity and concentration.

Measurement of IL-8 secretion. Confluent monolayers of T84, HT-29, Caco2BBE, and MDCK cells, grown on 6.5-mm-diameter (0.33 cm2) collagen-coated cell culture inserts were washed three times with HBSS, placed into 300 µl of HBSS, and incubated for 30 min at 37°C. Monolayers were placed in dry wells, and 25 µl of Salmonella-containing HBSS (1.6 x 1010 bacteria/ml) were placed on the apical surface of each monolayer. This inoculum has been previously shown to correspond to 30 bacteria per cell (32). One hour later, the monolayers were returned to the wells containing 300 µl HBSS. Five hours after the bacteria were added, basolateral cell supernatants were removed and assayed for IL-8. Human IL-8 from T84, HT-29, and Caco2BBE cells was measured by ELISA as previously described (32). Canine IL-8 from MDCK cells was measured by ELISA in 96-well plates (Linbro/Titertek: ICN Biochemicals, Costa Mesa, CA) coated overnight with 1 µg/ml anti-rabbit IL-8 monoclonal antibody and detected with rabbit anti-canine IL-8 polyclonal antibody from Dr. Randal Mrsny (Genentech, South San Francisco, CA).

Antibodies and reagents. The following reagents were used: recombinant human IL-8 and TNF-{alpha}, human IL-8 and phospho-Akt antibodies (R&D Systems, Minneapolis, MN); human IL-8 antibody (Endogen, Woburn, MA); rabbit IL-8 and canine IL-8 antibodies (Genentech, San Francisco, CA); peroxidase-conjugated anti-rabbit antibodies (Kirkegaard & Perry Laboratories, Gaithersburg, MD); peroxidase-conjugated anti-mouse and peroxidase-conjugated anti-rabbit antibodies (Amersham Life Science, Arlington Heights, IL); pan Akt, phospho-p44/42, phospho-p38, phospho-SAPK/JNK, phospho-p65 antibodies (Cell Signaling Technology, Beverly, MA); I{kappa}B{alpha} antibody, TLR4 antibody (Santa Cruz Biotechnology,); TLR5, IRAK-4 antibodies (Axxora, San Diego, CA), epidermal growth factor receptor (EGFR), and HA antibodies (Upstate, Lake Placid, NY); actin antibody, cycloheximide, wortmannin, E. coli LPS, mycrosistin-LR, myelin basic protein (Sigma-Aldrich), LY294002 (Promega, Madison, WI).

FACS assay for TLR-4 and TLR-5 expression. Epithelial cells were trypsinized and incubated with FITC-conjugated TLR-5 or TLR-4 antibodies for 30 min in the dark according to the manufacturer's protocol (Abcam, Cambridge, UK). Cells were analyzed by flow cytometry.

Cell surface biotinylation. Cell culture inserts were rinsed twice in ice-cold PBS supplemented with 0.1 mM CaCl2 and 1 mM MgCl2 (PBS-CM) and incubated for 30 min on ice with 1 mg/ml sulfo-NHS-biotin (Pierce Chemical, Rockford, IL) in PBS-CM added to either the apical or basolateral compartments. Biotinylation was quenched by incubating monolayers in 50 mM NH4Cl. Cells were washed three times in PBS and lysed for 10 min on ice in a solution containing 1% (wt/vol) Triton X-100, 20 mM Tris, pH 8.0, 50 mM NaCl, 5 mM EDTA, 0.2% (wt/vol) BSA supplemented with protease inhibitors. Biotinylated proteins were isolated by incubation with streptavidin-agarose beads (Pierce Chemical) for 16 h at 4°C. Proteins bound to agarose beads were solubilized by boiling for 5 min in sample buffer (50 mM Tris, pH 6.8, 100 mM dithiothreitol, 2% SDS, 0.1% bromophenol blue, 10% glycerol), separated by SDS-PAGE, and transferred to nitrocellulose membrane for immunoblotting.

Cell extraction and immunoblotting. Cells were rinsed twice in ice- cold HBSS, lysed in protein loading buffer (50 mM Tris, pH 6.8, 100 mM dithiothreitol, 2% SDS, 0.1% bromophenol blue, 10% glycerol), and sonicated with a Branson Sonifier 450 (Branson, Danbury, CT). Equal amounts of total cell protein were separated by SDS-PAGE, transferred to nitrocellulose according to standard procedures, and processed for immunoblotting with specific antibodies. Immune complexes were visualized with the appropriate peroxidase-conjugated antibodies and developed by use of an ECL kit (Amersham Pharmacia Biotech, Piscataway, NJ). Chemiluminescent signals were collected and scanned from BioMax film (Kodak, Rochester, NY) with a Scanjet 7400c backlit flatbed scanner (Hewlett-Packard, Palo Alto, CA). For figures, the contrast of images was adjusted, arranged, and labeled in Adobe Photoshop and Adobe Illustrator (Adobe Systems, San Jose, CA). Bands were quantified using NIH Image software. The digital images are representative of the original data.

Quantitative real-time PCR analysis. Total RNA was extracted from epithelial cell monolayers using TRIzol reagent (Invitrogen, Carlsbad, CA) and reverse transcribed using the iScript cDNA synthesis kit (Bio-Rad, Hercules, CA) according to the manufacturer's directions. The RT cDNA reaction products were subjected to quantitative real-time PCR using the MyiQ single-color real-time PCR detection system (Bio-Rad) and iQ SYBR green supermix (Bio-Rad) according to the manufacturer's directions. IL-8 cDNA was amplified by using primers to the human IL-8 gene that are complementary to regions in exon 1 (5'-TGCATAAAGACATACTCCAAACCT) and overlapping the splice site between exons 3 and 4 (5'-AATTCTCAGCCCTCTTCAAAAA). All expression levels were normalized to the GAPDH levels of the same sample, using forward (5-CTTCACCACCATGGAGAAGGC) and reverse (5'-GGCATGGACTGTGGTCATGAG) primers for GAPDH. Percent expression was calculated as the ratio of the normalized value of each sample to that of the corresponding untreated control cells. All real-time PCR reactions were performed in triplicate. All PCR primers were designed using Lasergene software (DNAStar, Madison, WI).

Coimmunoprecipitation and IRAK kinase assay. IRAK kinase activity was measured as previously described (27, 36), with the following modifications. Cell culture inserts were incubated with or without flagellin for various times, rinsed twice in ice-cold HBSS, and lysed in ice-cold coimmunoprecipitation buffer [0.4% IGEPAL, 50 mM Tris·HCl, pH 7.5, 150 mM NaCl, 1 mM EDTA, 1 mM sodium fluoride, 1 mM sodium orthovanadate, and a protease inhibitor cocktail (Roche Diagnostics, Indianapolis, IN)]. Particulate matter was removed by centrifugation, and lysates were incubated with anti-HA antibody and protein A-agarose (Upstate) for 2 h at 4°C with gentle agitation. The immunoprecipitates were washed three times with coimmunoprecipitation buffer and each sample was divided into two parts. One half was solubilized in protein loading buffer (50 mM Tris, pH 6.8, 100 mM dithiothreitol, 2% SDS, 0.1% bromophenol blue, 10% glycerol) for Western blot analysis, and the other half was washed three times with IRAK reaction buffer (20 mM Tris·HCl, pH 7.5, 20 mM Mg2Cl, 20 mM beta-glycerol phosphate, 1 mM sodium orthovanadate, 1 µM microcystin-LR). ATP (5 µM), myelin basic protein (2 µg), and 10 µCi of [32P]-ATP were added to each reaction and incubated at 37°C for 30 min. Protein loading buffer 5x was added to stop the reaction and equal volumes of the IRAK kinase reaction products were boiled, separated by SDS-PAGE, and the band was visualized by autoradiography.

Statistics. Results are expressed as means ± SD. Student's t-test was used to compare results, with statistical significance assumed at P < 0.05. Individual experiments were performed independently three or more times.


   RESULTS
TOP
 ABSTRACT
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 GRANTS
 REFERENCES
 
Prolonged flagellin exposure induces a state of flagellin tolerance, in a variety of polarized epithelial cell lines, that is dose dependent. Previous studies from our laboratory and others have demonstrated that flagellin is required for Salmonella-induced proinflammatory responses in epithelial cells (11, 14, 42). Both intact Salmonella bacteria and purified Salmonella flagellin activate host cell proinflammatory signaling cascades that result in the degradation of the inhibitor of kappa-B alpha (I{kappa}B{alpha}), the translocation of NF-{kappa}B to the nucleus, and the expression of a variety of proinflammatory cytokines. We began our study by asking whether prior exposure to purified S. typhimurium flagellin would prevent polarized epithelial cell monolayers from responding to subsequent challenges with WT Salmonella bacteria, purified Salmonella flagellin, or TNF-{alpha}. For these studies, we chose to use four different polarized epithelial cell lines to determine whether any observed flagellin effects were restricted to one cell line or could be seen in a variety of epithelial cells. The human intestinal epithelial cell lines T84, Caco2BBE, and HT-29, and the Madin-Darby canine kidney cell line (MDCKII) were grown on permeable membrane supports (cell culture inserts) to produce confluent monolayers of polarized cells. This cell culture method is thought to more closely approximate the epithelial monolayers found in the human intestine. The epithelial monolayers were then incubated with or without purified flagellin (100 ng/ml) added to the basolateral medium for 24 h.

When untreated monolayers were challenged for 5 h with TNF-{alpha} (100 ng/ml), WT Salmonella, or purified flagellin (100 ng/ml) they responded by secreting interleukin-8 (IL-8) into the basolateral medium (Fig. 1). Monolayers that had been previously incubated with basolateral flagellin for 24 h did not secrete IL-8 in response to the secondary stimulation with WT Salmonella or flagellin but remained responsive to challenge with TNF-{alpha} (Fig. 1). This indicates that flagellin tolerance is specific for flagellin/TLR5-mediated signaling and does not induce cross-tolerance for TNF-{alpha} signaling. Prolonged exposure to TNF-{alpha} induced self-tolerance but did not affect the proinflammatory response to flagellin or Salmonella (unpublished results).


Figure 1
View larger version (27K):
[in this window]
[in a new window]

  		Fig. 1. Flagellin-induced self-tolerance in epithelial cells. Monolayers of T84 (A), MDCK II (B), HT-29 (C), and Caco2 BBE (D) cells grown on cell culture inserts were incubated with or without flagellin (100 ng/ml) added to the basolateral compartment for 24 h. Cells were then washed and incubated with TNF-{alpha}, Salmonella typhimurium [wild-type (WT)], or flagellin (100 ng/ml) for 5 h. Basolateral medium was collected and analyzed by ELISA for the presence of IL-8. Data presented are means ± SD from a single experiment assayed in triplicate and are representative of results obtained in 3 separate experiments.

 
Taken together, we found that prior exposure to flagellin produced a functional downregulation of TLR5 signaling in all four epithelial cell lines tested, suggesting that flagellin-induced tolerance may be a common regulatory mechanism employed by all epithelial cells. Additionally, we found that the response was not unique to Salmonella flagellin, since monolayers responded in a similar manner to flagellin isolated from the gut commensal E. coli strain F-18 (not shown). The finding that flagellins from both pathogenic and commensal bacteria have similar effects is consistent with other recent reports of the proinflammatory nature of flagellin from commensal E. coli (5).

We next sought to determine whether flagellin tolerance was dependent on the dose of flagellin applied. One assay commonly used to demonstrate Salmonella-induced activation of the proinflammatory signaling pathway is the degradation of the NF-{kappa}B inhibitory molecule I{kappa}B{alpha}. Thus we used the degradation of I{kappa}B{alpha} to determine the concentration of flagellin necessary to induce tolerance in epithelial cells. Epithelial monolayers were incubated for 24 h with various concentrations of basolateral flagellin and then challenged a second time with basolateral flagellin (100 ng/ml). Complete inhibition of I{kappa}B{alpha} degradation required primary flagellin concentrations of 25 ng/ml or higher (Fig. 2A), whereas concentrations from 5 to 10 ng/ml were only partially inhibitory. In addition to I{kappa}B{alpha} degradation, the activation of members of the MAPK family of signaling molecules, specifically p38 MAPK (52), are thought to play an important role in the proinflammatory signaling of TLR5. We next examined the activity of p38 using antibodies that recognize phosphorylated (active) p38. As expected the phosphorylation of p38 paralleled I{kappa}B{alpha} degradation over the range of flagellin concentrations used. Flagellin concentrations of 25 ng/ml and higher completely prevented p38 phosphorylation, and lower concentrations partially inhibited p38 activity (Fig. 2A). This can be clearly seen when the relative I{kappa}B{alpha} and phospho-p38 band intensities after secondary flagellin treatment were determined and compared with the corresponding untreated controls (Fig. 2B).


Figure 2
View larger version (29K):
[in this window]
[in a new window]

  		Fig. 2. Dose dependence of flagellin tolerance in epithelial cells. A: monolayers of epithelial cells grown on cell culture inserts were incubated with flagellin at the concentrations indicated added to the basolateral compartment for 24 h. Cells were then washed and incubated with or without flagellin (100 ng/ml) for 30 min, lysed, and equal amounts of total protein were separated by SDS-PAGE, transferred to nitrocellulose, and probed for I{kappa}B{alpha} and phospho-p38. B: band intensity was quantified using a back-lit scanner and NIH Image software and is presented as percent of untreated control. These data are from a single experiment that is representative of results obtained in 3 separate experiments.

 
Taken together, these results suggest that, for flagellin tolerance to occur, a certain threshold of flagellin-TLR5 binding must be achieved. Additionally, these results demonstrate that flagellin tolerance occurs early in the proinflammatory signaling cascade, blocking both the NF-{kappa}B and MAPK signaling pathways.

Flagellin tolerance affects the NF-{kappa}B and MAPK pathways in a time-dependent manner. Thus far we had examined flagellin-induced tolerance after 24 h of exposure to flagellin. Our next objective was to determine the length of time cells must be exposed to flagellin before a state of tolerance ensues. We incubated epithelial monolayers with basolateral flagellin (100 ng/ml) for times ranging from 24 h to 5 min before a second 30-min incubation with flagellin (100 ng/ml). After the second flagellin incubation, the TLR5-signaling capacity was assessed by the degradation of I{kappa}B{alpha} and the phosphorylation (activation) of p38 MAPK, JNK, and p44/42 MAPK (ERK1/2). As seen in Fig. 3A, our data show that flagellin tolerance was achieved with as little as 1 to 2 h of prior exposure to flagellin and that there was no significant difference in the amount of inhibition with longer flagellin-exposure periods. Similar results were obtained when the secondary stimulus was WT S. typhimurium instead of purified flagellin (not shown).


Figure 3
View larger version (52K):
[in this window]
[in a new window]

  		Fig. 3. Time course of flagellin-induced tolerance in epithelial cells. A: monolayers of epithelial cells grown on cell culture inserts were incubated with or without flagellin (100 ng/ml) added to the basolateral compartment for the times indicated (1° flagellin). Cells were washed and incubated with basolateral flagellin (100 ng/ml) for 30 min (2° flagellin). Control monolayers were incubated with buffer alone. Equal amounts of total cell lysates were separated by SDS-PAGE, transferred to nitrocellulose, and probed with the antibodies indicated. Bands were visualized using ECL (Amersham). B: monolayers of epithelial cells grown on cell culture inserts were incubated with or without flagellin (100 ng/ml) added to the basolateral compartment for the times indicated. Cells were washed and incubated with either basolateral TNF-{alpha} (15 min), apical WT Salmonella (1 h), or basolateral flagellin (100 ng/ml) (30 min). Control monolayers were incubated with buffer alone. Equal amounts of total cell lysates were separated by SDS-PAGE, transferred to nitrocellulose, and probed with antibodies indicated, and bands were visualized using ECL. Results are representative of at least 3 separate experiments.

 
We then directly compared the effects of 24 h and 1 h of flagellin exposure. In both the 24-h and 1-h flagellin-treated cells, flagellin tolerance prevented both I{kappa}B{alpha} degradation as well as the activation of p38, JNK, and ERK1/2 by both WT Salmonella and purified flagellin (Fig. 3B). As seen before, flagellin tolerance had no effect on TNF-{alpha} stimulation of these pathways (Fig. 3B). A separate dose-response analysis of TNF-{alpha} stimulation indicated that even at low TNF-{alpha} concentrations (1–10 ng/ml), flagellin-induced tolerance did not alter TNF-{alpha} activation of the proinflammatory pathways (unpublished data).

Flagellin tolerance reduces the basolateral surface expression of TLR5, but does not alter total TLR5 expression or require protein synthesis. One way in which cells can limit their response to an external stimulus is to remove the receptor from the cell surface. This is a well-understood mechanism of desensitization for the EGFR (50) and has recently been reported as a potential mechanism of tolerance for the Toll-like receptor 4 (TLR4) (39, 41). We next sought to determine whether polarized epithelial cells used a mechanism of TLR5 internalization to limit their response to bacterial flagellin. Using domain-specific cell-surface biotinylation to isolate receptors from the apical or basolateral surfaces of polarized epithelial cell monolayers, we examined the effect of prior exposure to flagellin on the cell-surface expression of TLR5. In untreated epithelial monolayers, TLR5 was localized predominantly to the basolateral membrane in each of the cell lines tested (Fig. 4, A and B). This is consistent with previous observations of TLR5 surface distribution in polarized epithelial monolayers (14, 42). When epithelial monolayers were incubated for 24 h with basolateral flagellin (100 ng/ml), the highly polarized surface distribution of TLR5 was unaffected, and total TLR5 expression levels remained constant (Fig. 4, A and B). Interestingly, the basolateral expression level was reduced by as much as 10–50% compared with the corresponding untreated controls (Fig. 4, A and B). Since the total TLR5 levels remain unchanged, it is likely that the internalized receptors are not degraded, but they may be sequestered in internal compartments. A basolateral marker EGFR was used as control for the specific effect that flagellin has on TLR5. Flagellin treatment did not change the distribution of EGFR (Fig. 4A).


Figure 4
View larger version (40K):
[in this window]
[in a new window]

  		Fig. 4. Toll-like receptor 5 (TLR5) expression in epithelial cells. A: monolayers of MDCK II epithelial cells grown on permeable cell culture inserts were incubated with or without flagellin (100 ng/ml) added to the basolateral media for 24 h, biotinylated at either the apical (Ap) or basolateral (BL) surfaces, and lysed, and biotinylated proteins isolated by affinity chromatography. Biotinylated proteins were separated by SDS-PAGE, transferred to nitrocellulose, probed with anti-TLR5 antibody, and visualized by ECL. A basolateral marker epidermal growth factor receptor (EGFR) was used as control. TLR5 band intensity was determined and is presented as percent of the untreated basolateral controls. Aliquots of the biotinylated cell lysates were run on a separate gel and probed for total TLR5. B: monolayers of T84 epithelial cells grown on permeable cell culture inserts were incubated with or without flagellin (100 ng/ml) added to the basolateral media for 24 h, biotinylated at either the apical or basolateral surfaces, and lysed, and biotinylated proteins isolated by affinity chromatography. Biotinylated proteins were separated and probed with anti-TLR5 antibody. TLR5 band intensity was determined and is presented as percent of the untreated basolateral controls. Aliquots of the biotinylated cell lysates were run on a separate gel and probed for total TLR5. C: monolayers of epithelial cells grown on permeable cell culture inserts were incubated with or without cycloheximide (CHX; 10 µg/ml) before and during incubation with flagellin. Cells were then incubated with or without flagellin for 30 min and lysed, and immunoblots were probed for I{kappa}B{alpha}.

 
We next determined whether protein synthesis was required for flagellin-induced tolerance to occur. Several time points were examined for the following reasons: the 1-h time point was the earliest at which we had previously observed flagellin tolerance, and the 4- and 6-h time points would help to determine whether protein synthesis was required for prolonged flagellin tolerance. Monolayers were incubated without or with 10 µg/ml cycloheximide for 15 min before and during the 1-, 4-, and 6-h incubations with flagellin. The cells were washed and incubated again with or without flagellin for 30 min before being assayed for I{kappa}B{alpha} degradation. For each of the flagellin pretreatment time points, 1, 4, and 6 h, cycloheximide did not prevent flagellin tolerance (Fig. 4C). In the cycloheximide-treated cells a noticeable decrease in I{kappa}B{alpha} expression was seen in both the control and flagellin-treated cells by 6 h (Fig. 4C). Long-term cycloheximide treatment reduced the viability of the monolayers and decreased I{kappa}B{alpha} expression to undetectable levels, making it difficult to determine whether protein synthesis is necessary at later time points, between 12 and 24 h. Our data demonstrate that inhibition of protein synthesis by cycloheximide treatment before and during incubation with flagellin was not able to block the induction of flagellin tolerance, thus indicating that protein synthesis is not required for the rapid onset or prolongation of flagellin tolerance.

LPS does not cross-regulate flagellin signaling in polarized epithelial cells. Recent reports indicate that LPS can cross-regulate the response to other TLR ligands in macrophages (10, 26, 36, 43, 44). To determine whether this might also occur in polarized epithelial cells, we asked whether prior exposure to purified LPS would be able to cross-regulate flagellin signaling. The epithelial cell lines, T84, Caco2-BBE, HT-29, and MDCK, showed expression of TLR4 by Western blot (Fig. 5A). Additionally, we also confirmed the expression of the TLR4 by FACS in the human epithelial cells (data not shown). As a representative cell line, T84 monolayers were incubated at 37°C with LPS (2 µg/ml) added to both the apical and basolateral compartments for 0- to 24-h time periods and then assayed for I{kappa}B{alpha} degradation and phosphorylation of NF-{kappa}B (p65). Whereas the response to purified flagellin was rapid, LPS stimulation had very little effect on I{kappa}B{alpha} degradation and the phosphorylation of p65 was delayed until between 4 and 8 h after the addition of LPS (Fig. 5B). Interestingly, there appeared to be a transient reduction in TLR4 expression, followed by an upregulation between 2 and 8 h, followed by a reduction at 24 h of LPS stimulation.


Figure 5
View larger version (40K):
[in this window]
[in a new window]

  		Fig. 5. Lack of lipopolysaccharide (LPS) cross-regulation of flagellin signaling. A: TLR4 expression in T84, Caco2-BBE, HT29, and MDCK cell lines. The epithelial cell lines showed expression of TLR4 by Western blot with anti-TLR 4 antibody. B: T84 monolayers were incubated at 37°C with LPS (2 µg/ml) added to both the apical and basolateral compartments for 0 to 24 h. Equal amounts of total cell lysates were separated by SDS-PAGE, transferred to nitrocellulose, and probed with anti-I{kappa}B{alpha} anti-phosphorylation of NF-{kappa}B (p65) or anti-TLR4 antibodies, followed by a secondary horseradish peroxidase (HRP)-conjugated antibody. Bands were visualized using ECL and are representative of results obtained in 4 separate experiments. C: T84 monolayers were incubated at 37°C with flagellin (100 ng/ml) or LPS (2 µg/ml) added to both the apical and basolateral compartments for 24-h time periods. Cells were then washed and incubated with or without TNF-{alpha} (100 ng/ml), flagellin (100 ng/ml), or LPS (2 µg/ml) for 0.5 h. The total cell lysates were assayed for I{kappa}B{alpha} degradation and phosphorylation of NF-{kappa}B (p65).

 
To determine whether LPS treatment could alter the response to flagellin, monolayers were incubated for 24 h with either flagellin or LPS, then washed and incubated with TNF-{alpha} (100 ng/ml) or flagellin (100 ng/ml) and analyzed by immunoblot for I{kappa}B{alpha} levels. Exposure to LPS for 24 h did not affect the ability of TNF-{alpha} or flagellin (Fig. 5C, lanes 5 and 6) to induce I{kappa}B{alpha} degradation, compared with monolayers that received flagellin pretreatment and obtained the tolerance to flagellin stimulation (Fig. 5C, lane 3). Interestingly, the epithelial monolayers were only mildly responsive to LPS treatment.

Recovery from flagellin-induced tolerance. We have shown that flagellin tolerance is very rapid. Our next objective was to determine how long the state of tolerance would persist after removal of flagellin. As before, monolayers were incubated with basolateral flagellin for 24 h and then washed extensively to remove any residual flagellin. The washed monolayers were then incubated in fresh medium for times ranging from 30 min to 24 h followed by a second 30-min challenge with flagellin. Total protein and RNA were extracted from duplicate monolayers and used to determine TLR5 signaling activity by I{kappa}B{alpha} immunoblot and IL-8 message levels using quantitative real-time PCR. Immunoblot of total cell protein for I{kappa}B{alpha} degradation showed that monolayers gradually recovered their ability to respond to flagellin, beginning 1 h after flagellin washout and proceeding through 24 h of recovery (Fig. 6A). Quantification of the relative I{kappa}B{alpha} band intensities revealed that flagellin-tolerant monolayers are likely to require more than 24 h to recover before being able to mount a full response to flagellin compared with untreated controls (Fig. 6B). However, monolayers showed a normal phospho-p38 response by 12 h (Fig. 6B), indicating that the I{kappa}B{alpha} and MAPK pathways may have differing sensitivities to TLR5 activation upon recovery. Additionally, we used quantitative real-time PCR to determine when monolayers regained their ability to synthesize IL-8 in response to flagellin. As seen in Fig. 6C, cells required between 16 and 24 h of recovery after flagellin washout before being able to produce IL-8 message levels that were comparable to those of control nontolerant cells incubated with flagellin. Taken together, these data indicate that flagellin-induced tolerance has a long-lasting effect on the ability of cells to respond to subsequent challenges with flagellin.


Figure 6
View larger version (30K):
[in this window]
[in a new window]

  		Fig. 6. Recovery after flagellin-induced tolerance. Monolayers of epithelial cells grown on cell culture inserts were incubated with or without flagellin (100 ng/ml) added to the basolateral compartment for 24 h. Cells were then washed and incubated in normal media for the times indicated before being incubated with or without basolateral flagellin (100 ng/ml) for 30 min. A: equal amounts of total cell lysates were separated by SDS-PAGE, transferred to nitrocellulose, and probed with anti- I{kappa}B{alpha}, phospho-p38, and TLR5 antibodies followed by a secondary HRP-conjugated antibody. Bands were visualized using ECL. B: band intensity was determined at each time point and normalized to the corresponding untreated control for I{kappa}B{alpha} and to the flagellin treated control for phospho-p38. The upper and lower dashed lines represent the maximum and minimum I{kappa}B{alpha} and phospho-p38 levels in control cells. C: monolayers were treated as above and incubated for an additional 2 h, and equal amounts of total RNA were extracted. Relative IL-8 expression was determined by real-time PCR. The dashed line represents the maximal IL-8 expression level in nontolerant flagellin-stimulated control cells (100%). Results are representative of at least 3 separate experiments.

 
PI3K is activated by flagellin but does not contribute to flagellin tolerance. Recent studies have identified PI3K as both a positive and negative regulator of TLR2, TLR4, and TLR9 signaling in monocytes and dendritic cells (reviewed in Ref. 12). As such, we sought to determine whether PI3K might play a role in flagellin tolerance of the TLR5 signaling pathway in polarized epithelial cells. Epithelial cell monolayers were incubated for various times with basolaterally applied flagellin, and PI3K activity was assessed by the phosphorylation of Akt/protein kinase B. Akt phosphorylation began to appear 30 min after flagellin application, peaked at ~60 min, and was declining by 2 h (Fig. 7A). This activation of the PI3K/Akt pathway is much more rapid than the 2-h incubation necessary to elicit a response in epithelial cells exposed to WT Salmonella (20). This is probably due to the fact that Salmonella must translocate flagellin across the monolayer to activate basolateral TLR5 before Akt phosphorylation can occur.


Figure 7
View larger version (36K):
[in this window]
[in a new window]

  		Fig. 7. PI3K and Akt activity during flagellin-induced tolerance. A: monolayers were incubated with or without basolateral flagellin for the times indicated, and total cell protein was separated by SDS-PAGE, transferred to nitrocellulose, and immunoblotted for phospho-Akt and total Akt. B: monolayers were incubated with or without wortmannin (250 nM) 15 min before and during incubation with or without basolateral flagellin for 1 h. Cells were washed and incubated with or without wortmannin and flagellin for an additional 30 min. Equal amounts of total cell protein were immunoblotted for I{kappa}B{alpha}, phospho-p38, phospho-Akt, and total Akt. C: monolayers were incubated with or without flagellin for times indicated, washed, and incubated with or without flagellin for 60 min. Equal amounts of total cell protein were immunoblotted for phospho-Akt and total Akt. D: monolayers were incubated with or without flagellin or LY294002 (200 µM) for 2 h, washed, and incubated with or without EGF for 60 min. Equal amounts of total cell protein were immunoblotted for phospho-Akt and total Akt. E: monolayers were incubated with or without flagellin or LY294002 for the times indicated, washed, and incubated with or without flagellin for 60 min. Total RNA was extracted and used for quantitative real-time PCR of IL-8, *P < 0.05. All blots were visualized using ECL (Amersham). Results are representative of at least 3 separate experiments.

 
Having determined that peak Akt phosphorylation can be detected in epithelial monolayers after 60 min of flagellin incubation, we took advantage of the PI3K inhibitors wortmannin and LY294002 to ascertain whether blocking of PI3K activity would affect flagellin tolerance. As seen in Fig. 7B, primary incubation with flagellin prevented the activation of both the NF-{kappa}B and MAPK signaling pathways as evidenced by the presence of I{kappa}B{alpha} and phospho-p38 after a second 30-min dose of flagellin. The addition of the PI3K inhibitor LY294002 or wortmannin before and during the primary incubation with flagellin prevented Akt phosphorylation but was unable to inhibit flagellin tolerance (Fig. 7B). Interestingly, we also observed that flagellin tolerance prevented the flagellin stimulation of Akt phosphorylation (Fig. 7, B and C), demonstrating that flagellin tolerance can also impede signaling through the PI3K/Akt pathway.

We next asked whether the inhibition of PI3K signaling by flagellin was restricted to TLR5 activation of the pathway or was instead a general block of PI3K stimulation. Monolayers were incubated with flagellin for 4 h, followed by incubation with epidermal growth factor (EGF) for an additional hour. As seen in Fig. 7D, prior incubation with flagellin did not prevent the subsequent EGF activation of PI3K whereas LY294002 treatment completely blocked EGF induced Akt phosphorylation. This indicates that the loss of PI3K signaling after flagellin tolerance seen in Fig. 7, B and C only applies to flagellin signaling through TLR5 and is not a result of a general inhibition of PI3K activity. To further examine the role of PI3K in flagellin tolerance we measured the induction of IL-8 message by real-time PCR. As expected, when monolayers were incubated with flagellin alone, a significant amount of IL-8 mRNA was detected, whereas monolayers pretreated with flagellin for 4 or 6 h had levels that were only slightly above background after secondary flagellin stimulation (Fig. 7E). However, monolayers that had been pretreated with LY294002 had significantly higher IL-8 mRNA levels (P < 0.05) in each of the treatment categories when compared with their vehicle (DMSO)-treated counterparts (Fig. 7E). Thus it appears that the rapid activation of PI3K/Akt signaling by flagellin stimulation may help to modify or damp the acute flagellin-induced proinflammatory signals but does not seem to contribute to the prolonged flagellin tolerance seen in epithelial monolayers.

TLR5 interacts with IRAK-4 during flagellin tolerance. Flagellin has been shown to activate IRAK-1 kinase activity in mouse macrophage-like and human THP-1 cells (36, 37). The phosphorylation and subsequent degradation of some of the downstream signaling molecules like IRAK-1 and IRAK-4 have been implicated in other systems as a means by which to induce tolerance to other TLR ligands (45). However, in our polarized epithelial cell system we observed transient IRAK-1 phosphorylation (Fig. 8A) but were unable to detect any change in the expression of IRAK-1. Recently, others have reported that TLR signaling can be inhibited by a failure to recruit the key mediator, IRAK-4 (6). To establish whether TLR5 associates with IRAK-4 during the development of flagellin tolerance, a series of coimmunoprecipitation experiments were conducted using MDCK cell lines that express HA-tagged TLR5. Monolayers were incubated for various times in the presence of basolateral flagellin. TLR5-HA was immunoprecipitated from cell lysates using an anti-HA antibody, and probed by Western blot for IRAK-4. IRAK-4 coimmunoprecipitated with TLR5 at all time points; however, the higher molecular weight phosphorylated form was only seen between 30 min and 3 h (Fig. 8B). It indicated that flagellin treatment changed the phosphorylation of the TLR5-associated IRAK-4. However, we did not see a reduction in the total IRAK-4 after flagellin treatment in cell lysates by Western blot (Fig. 8B).


Figure 8
View larger version (51K):
[in this window]
[in a new window]

  		Fig. 8. Coimmunoprecipitation of IRAK-4 and IRAK kinase activity during flagellin-induced tolerance. A: monolayers of epithelial cells grown on cell culture inserts were incubated with or without flagellin (100 ng/ml) added to the basolateral compartment for 24 h. Equal amounts of total cell lysates were separated by SDS-PAGE, transferred to nitrocellulose, and probed with anti-phospho-IRAK-1 or IRAK antibodies. B: monolayers of MDCK cells expressing TLR5-HA were incubated with or without basolateral flagellin (100 ng/ml) for the times indicated. Cells were lysed and incubated with anti-HA antibody and protein A beads to immunoprecipitate TLR5-HA and any associated proteins. Total cell lysates for immunoprecipitation were also analyzed by Western blot for IRAK-4. C: IRAK4 kinase activity in coimmunoprecipitates was determined by in vitro IRAK kinase assay using myelin basic protein (MBP) as the substrate. Monolayers of epithelial cells grown on cell culture inserts were incubated with or without flagellin (100 ng/ml) added to the basolateral compartment for 24 h. Cells were then washed and incubated in normal media for the times indicated before being incubated with or without basolateral flagellin (100 ng/ml) for 30 min. As described in the MATERIALS AND METHODS, the immunoprecipitates pulled down with anti-HA antibodies were divided in half. One half was solubilized in protein loading buffer for Western blot analysis, and the other half was used for IRAK reaction. The IRAK kinase reaction products were separated by SDS-PAGE, and the band was visualized by autoradiography. Enhanced phosphorylated forms of the MBP indicated the increased activity of IRAK-4. D: IRAK-4 kinase activity in coimmunoprecipitates was determined by in vitro IRAK kinase assay with flagellin (100 ng/ml) added to the basolateral compartment for 30–90 min. The phosphorylated forms of MBP indicated the activity of IRAK-4.

 
To determine whether the IRAK-4 associated with TLR5 was kinase active, an in vitro IRAK kinase assay was performed on the coimmunoprecipitated proteins using the substrate myelin basic protein. A 30-min incubation with basolateral flagellin was sufficient to induce IRAK kinase activity and IRAK-4 phosphorylation (Fig. 8C). However, after a 24-h incubation with basolateral flagellin, the cells lacked IRAK kinase activity and were no longer able to be activated by secondary flagellin stimulation (Fig. 8C). Interestingly, earlier experiments in macrophage cell lines showed a rapid activation of IRAK kinase that returned to basal levels by 60 min (37); however, in epithelial cells, we saw a more prolonged IRAK activation lasting at least 90 min (Fig. 8D).


   DISCUSSION
TOP
 ABSTRACT
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
GRANTS
 REFERENCES
 
In this study, we have characterized the response of polarized intestinal epithelial cells to prolonged exposure to flagellin. Whereas much attention has been focused on understanding the proinflammatory pathways activated by flagellin in epithelial cells, very little is known about how these pathways may respond to continuous exposure to flagellin. We and other investigators have previously demonstrated that TLR5 is exclusively localized to the basolateral membrane of polarized epithelial cells (14, 42), physically segregating it from the luminal contents. This spatial separation of TLR5 and flagellin is thought to allow epithelial cells to mount an innate immune response only when flagellin has breached the epithelium, be it the result of an invading pathogen like Salmonella or a leaky barrier, as may be the case in Crohn's disease.

We have shown that the initial exposure of the basolateral membrane to flagellin causes the rapid activation of multiple signaling cascades including the NF-{kappa}B, MAPK, and PI3K pathways and that it ultimately results in the secretion of the proinflammatory cytokine IL-8. We have demonstrated that prolonged incubation with basolateral flagellin produces a state of tolerance in the epithelial monolayer such that it becomes insensitive to further stimulation by flagellin. However, these same cells remain responsive to TNF-{alpha} stimulation, suggesting that flagellin tolerance occurs at an early step in the TLR5 signaling pathway. The ability to become tolerant to flagellin also seems to be common to epithelial cells in general since we were able to induce flagellin tolerance in all of the polarized epithelial cell lines tested. Interestingly, although each epithelial cell line was able to respond to flagellin and developed flagellin tolerance in a similar manner, differences were observed in the magnitude of IL-8 secretion in response to flagellin (Fig. 1). This might be explained in part by differences in TLR5 expression in the cell lines. However, other studies have showed that TLR5 expression levels do not necessarily correlate with the robustness of the flagellin response (49), and we were unable to detect any significant differences in total TLR5 expression among the cell lines tested (unpublished data). This suggests that differences in an as-yet-unidentified TLR5 adaptor molecule or one or more of the downstream signaling components may be responsible for the observed differences in flagellin-induced IL-8 secretion in the epithelial cell lines.

Flagellin-induced tolerance is a rapid process that does not require protein synthesis and is dose dependent. This tells us several things about the development of flagellin tolerance in epithelial cells. The speed at which flagellin tolerance is achieved and the lack of a requirement for protein synthesis suggest that the inhibitory molecule(s) or mechanism(s) are already in place when cells first come into contact with flagellin. The synthesis of an inhibitory molecule does not appear to be necessary, at least in the early stage of flagellin tolerance. We have observed a prolonged upregulation of I{kappa}B{alpha} expression during flagellin tolerance by Western blot (Fig. 3, A and B) and real-time PCR (unpublished data), but this is unlikely to be the basis of flagellin tolerance since it is unable to inhibit TNF-{alpha}-induced IL-8 secretion (Figs. 1 and 3). The phosphorylation and subsequent degradation of some of the downstream signaling molecules such as IRAK-1 and IRAK-4 have been implicated in other systems as a means by which to induce tolerance to other TLR ligands (17). In our polarized epithelial cell system, we observed transient IRAK-4 phosphorylation (Fig. 8A). We did observe a lack of IRAK kinase activity associated with TLR5 after prolonged flagellin stimulation. This leads us to speculate that the lack of IRAK-4 kinase activity is vital to the development of flagellin tolerance. In addition, our finding that recovery from flagellin tolerance is protracted also points toward the synthesis of a cofactor, and preliminary results indicate that protein synthesis is required for recovery. Alternatively, TLR5 itself may be modified by the initial interaction with flagellin, and the receptors may need to be regenerated during the recovery process. Studies are currently under way to explore these possible mechanisms.

Cross-regulation between TLR-signaling pathways has also become an area of great interest. Recent reports have demonstrated cross-regulation between the LPS and a variety of TLR ligands in macrophages/monocytes and T cells (10, 26, 36, 43, 44). Initially, we had sought to determine whether LPS might also cross-regulate flagellin signaling and vice versa; however, in our hands the intestinal epithelial cells were only minimally responsive to LPS treatment. We found that LPS pretreatment of monolayers had no effect on the signaling capacity of flagellin (Fig. 5). This finding is consistent with reports that intestinal epithelial cells downregulate TLR4 and MD-2 expression and are unresponsive to LPS treatment (1, 38). This seems to be a result of differentiation in polarized epithelial cells (25). However, other groups have noted responses to LPS in intestinal epithelial cells (8, 19). These results indicate that the LPS/flagellin cross-regulated tolerance observed in macrophages does not necessarily apply to polarized epithelial cells. At the same time it is in contrast to other observations of the functional expression of TLR4 at the plasma membrane (7, 8), and the intracellular localization of TLR4 in the Golgi apparatus (19) of intestinal epithelial cells. Taken together with the previous observation that LPS is released into intracellular vesicles during S. typhimurium invasion of epithelial cells (13), this may indicate an intracellular signaling role for TLR4. If these two separate findings are correct, one may then be able to speculate that flagellin tolerance does cross-regulate LPS/TLR4 signaling since flagellin pretreatment blocked S. typhimurium proinflammatory signals (Figs. 1 and 3).

We have shown here for the first time that prolonged exposure to flagellin results in TLR5 internalization from the cell surface without reducing total TLR5 expression. After incubation with flagellin for 24 h, polarized epithelial cell monolayers showed a 10–50% reduction in the cell surface expression of TLR5 depending on the cell line being tested, yet each cell line was completely tolerant to further flagellin treatment (Fig. 4). This suggests that complete removal of TLR5 from the cell surface is not necessary for flagellin tolerance; however, internalization of receptors may be needed for activation of tolerance-inducing signaling cascades. This is consistent with other reports of reduced TLR4 cell surface expression during the development of LPS tolerance in macrophages (39). However, it is in contrast to other reports using Jurkat T-cells that did not have a reduction in surface TLR5 expression after incubation with flagellin (36). The differences seen in basolateral TLR5 expression hint at the possibility that only a subset of TLR5 receptors can be internalized; however, the precise differences in the putative receptor subsets remain elusive. It seems likely that the internalization of TLR5 after prolonged ligand exposure may be cell type dependent, but it is still unclear whether receptor internalization is necessary for TLR5 activation or required for the development of flagellin tolerance. Taken together with our dose response data this would suggest that a certain threshold of surface receptors is required to activate proinflammatory pathways. Internalization of some, but not all, of the surface receptors may prevent the threshold of activation from being reached, or at least it may damp the proinflammatory response.

The results presented in this report demonstrate that the prolonged exposure of polarized human intestinal epithelial cells to flagellin results in a state of tolerance to subsequent contact with flagellin that is dependent on the interaction of TLR5 with and activity of IRAK-4. The mechanism and biological relevance of flagellin tolerance in intestinal disease is a topic of great interest since flagellin from commensal enteric microflora has been identified as the predominant antigen in Crohn's disease (47). Recent evidence from the Dextran sulfate sodium-colitis mouse model indicates that the lack of MyD88-dependent TLR signaling results in the development of more severe colitis, suggesting that TLR signaling may have a protective role against the development of colitis (4). The precise role of flagellin tolerance during chronic inflammation and self-limiting enteritis should be explored.


   GRANTS
TOP
 ABSTRACT
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
REFERENCES
 
This work was supported by National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) Grants DK-063288 (to M. E. Hobert), and DK-35932 and DK-47662 (to J. L. Madara). Core facilities at the University of Chicago Digestive Disease Research Core Center were supported by a grant from the NIDDK, P30 DK-42086.


   ACKNOWLEDGMENTS
 
We thank our colleagues for thoughtful comments and critical reading of this manuscript and Sumalatha Kuppireddi for technical assistance.


   FOOTNOTES
 

Address for reprint requests and other correspondence: J. Sun, Dept. of Pathology, The Univ. of Chicago, 5841 S. Maryland Ave., Chicago, IL 60637 (e-mail: jsun{at}bsd.uchicago.edu)

The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.


   REFERENCES
TOP
 ABSTRACT
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 

  1. Abreu MT, Vora P, Faure E, Thomas LS, Arnold ET, Arditi M. Decreased expression of Toll-like receptor-4 and MD-2 correlates with intestinal epithelial cell protection against dysregulated proinflammatory gene expression in response to bacterial lipopolysaccharide. J Immunol 167: 1609–1616, 2001.[Abstract/Free Full Text]
  2. Akira S. Mammalian Toll-like receptors. Curr Opin Immunol 15: 5–11, 2003.[CrossRef][Web of Science][Medline]
  3. Amsler K, Kuwada SK. Membrane receptor location defines receptor interaction with signaling proteins in a polarized epithelium. Am J Physiol Cell Physiol 276: C91–C101, 1999.[Abstract/Free Full Text]
  4. Araki A, Kanai T, Ishikura T, Makita S, Uraushihara K, Iiyama R, Totsuka T, Takeda K, Akira S, Watanabe M. MyD88-deficient mice develop severe intestinal inflammation in dextran sodium sulfate colitis. J Gastroenterol 40: 16–23, 2005.[CrossRef][Web of Science][Medline]
  5. Bambou JC, Giraud A, Menard S, Begue B, Rakotobe S, Heyman M, Taddei F, Cerf-Bensussan N, Gaboriau-Routhiau V. In vitro and ex vivo activation of the TLR5 signaling pathway in intestinal epithelial cells by a commensal Escherichia coli strain. J Biol Chem 279: 42984–42992, 2004.[Abstract/Free Full Text]
  6. Burns K, Janssens S, Brissoni B, Olivos N, Beyaert R, Tschopp J. Inhibition of interleukin 1 receptor/Toll-like receptor signaling through the alternatively spliced, short form of MyD88 is due to its failure to recruit IRAK-4. J Exp Med 197: 263–268, 2003.[Abstract/Free Full Text]
  7. Cario E, Brown D, McKee M, Lynch-Devaney K, Gerken G, Podolsky DK. Commensal-associated molecular patterns induce selective toll-like receptor-trafficking from apical membrane to cytoplasmic compartments in polarized intestinal epithelium. Am J Pathol 160: 165–173, 2002.[Abstract/Free Full Text]
  8. Cario E, Rosenberg IM, Brandwein SL, Beck PL, Reinecker HC, Podolsky DK. Lipopolysaccharide activates distinct signaling pathways in intestinal epithelial cell lines expressing Toll-like receptors. J Immunol 164: 966–972, 2000.[Abstract/Free Full Text]
  9. Diamond JM. Twenty-first Bowditch lecture. The epithelial junction: bridge, gate, and fence. Physiologist 20: 10–18, 1977.[Medline]
  10. Dobrovolskaia MA, Medvedev AE, Thomas KE, Cuesta N, Toshchakov V, Ren T, Cody MJ, Michalek SM, Rice NR, Vogel SN. Induction of in vitro reprogramming by Toll-like receptor (TLR)2 and TLR4 agonists in murine macrophages: effects of TLR "homotolerance" vs. "heterotolerance" on NF-kappa B signaling pathway components. J Immunol 170: 508–519, 2003.[Abstract/Free Full Text]
  11. Eaves-Pyles T, Murthy K, Liaudet L, Virâag L, Ross G, Soriano FG, Szabâo C, Salzman AL. Flagellin, a novel mediator of Salmonella-induced epithelial activation and systemic inflammation: I kappa B alpha degradation, induction of nitric oxide synthase, induction of proinflammatory mediators, and cardiovascular dysfunction. J Immunol 166: 1248–1260, 2001.[Abstract/Free Full Text]
  12. Fukao T, Koyasu S. PI3K and negative regulation of TLR signaling. Trends Immunol 24: 358–363, 2003.[CrossRef][Web of Science][Medline]
  13. Garcia-del Portillo F, Stein MA, Finlay BB. Release of lipopolysaccharide from intracellular compartments containing Salmonella typhimurium to vesicles of the host epithelial cell. Infect Immun 65: 24–34, 1997.[Abstract]
  14. Gewirtz AT, Navas TA, Lyons S, Godowski PJ, Madara JL. Cutting edge: bacterial flagellin activates basolaterally expressed TLR5 to induce epithelial proinflammatory gene expression. J Immunol 167: 1882–1885, 2001.[Abstract/Free Full Text]
  15. Gewirtz AT, Simon PO Jr, Schmitt CK, Taylor LJ, Hagedorn CH, O’Brien AD, Neish AS, Madara JL. Salmonella typhimurium translocates flagellin across intestinal epithelia, inducing a proinflammatory response. J Clin Invest 107: 99–109, 2001.[Web of Science][Medline]
  16. Gumbiner B. Structure, biochemistry, and assembly of epithelial tight junctions. Am J Physiol Cell Physiol 253: C749–C758, 1987.[Abstract/Free Full Text]
  17. Hatao F, Muroi M, Hiki N, Ogawa T, Mimura Y, Kaminishi M, Tanamoto K. Prolonged Toll-like receptor stimulation leads to down-regulation of IRAK-4 protein. J Leukoc Biol 76: 904–908, 2004.[Abstract/Free Full Text]
  18. Hershberg RM. The epithelial cell cytoskeleton and intracellular trafficking. V. Polarized compartmentalization of antigen processing and Toll-like receptor signaling in intestinal epithelial cells. Am J Physiol Gastrointest Liver Physiol 283: G833–G839, 2002.[Abstract/Free Full Text]
  19. Hornef MW, Frisan T, Vandewalle A, Normark S, Richter-Dahlfors A. Toll-like receptor 4 resides in the Golgi apparatus and colocalizes with internalized lipopolysaccharide in intestinal epithelial cells. J Exp Med 195: 559–570, 2002.[Abstract/Free Full Text]
  20. Huang FC, Li Q, Cherayil BJ. A phosphatidyl-inositol-3-kinase-dependent anti-inflammatory pathway activated by Salmonella in epithelial cells. FEMS Microbiol Lett 243: 265–270, 2005.[CrossRef][Web of Science][Medline]
  21. Iwami KI, Matsuguchi T, Masuda A, Kikuchi T, Musikacharoen T, Yoshikai Y. Cutting edge: naturally occurring soluble form of mouse Toll-like receptor 4 inhibits lipopolysaccharide signaling. J Immunol 165: 6682–6686, 2000.[Abstract/Free Full Text]
  22. Jones BD, Ghori N, Falkow S. Salmonella typhimurium initiates murine infection by penetrating and destroying the specialized epithelial M cells of the Peyer's patches. J Exp Med 180: 15–23, 1994.[Abstract/Free Full Text]
  23. Kuwada SK, Lund KA, Li XF, Cliften P, Amsler K, Opresko LK, Wiley HS. Differential signaling and regulation of apical vs. basolateral EGFR in polarized epithelial cells. Am J Physiol Cell Physiol 275: C1419–C1428, 1998.[Abstract/Free Full Text]
  24. LeBouder E, Rey-Nores JE, Rushmere NK, Grigorov M, Lawn SD, Affolter M, Griffin GE, Ferrara P, Schiffrin EJ, Morgan BP, Labéta MO. Soluble forms of Toll-like receptor (TLR)2 capable of modulating TLR2 signaling are present in human plasma and breast milk. J Immunol 171: 6680–6689, 2003.[Abstract/Free Full Text]
  25. Lee SK, Il Kim T, Kim YK, Choi CH, Yang KM, Chae B, Kim WH. Cellular differentiation-induced attenuation of LPS response in HT-29 cells is related to the down-regulation of TLR4 expression. Biochem Biophys Res Commun 337: 457–463, 2005.[CrossRef][Web of Science][Medline]
  26. Lehner MD, Morath S, Michelsen KS, Schumann RR, Hartung T. Induction of cross-tolerance by lipopolysaccharide and highly purified lipoteichoic acid via different Toll-like receptors independent of paracrine mediators. J Immunol 166: 5161–5167, 2001.[Abstract/Free Full Text]
  27. Li L, Cousart S, Hu J, McCall CE. Characterization of interleukin-1 receptor-associated kinase in normal and endotoxin-tolerant cells. J Biol Chem 275: 23340–23345, 2000.[Abstract/Free Full Text]
  28. Liew FY, Xu D, Brint EK, O’Neill LA. Negative regulation of toll-like receptor-mediated immune responses. Nat Rev Immunol 5: 446–458, 2005.[CrossRef][Web of Science][Medline]
  29. Lodes MJ, Cong Y, Elson CO, Mohamath R, Landers CJ, Targan SR, Fort M, Hershberg RM. Bacterial flagellin is a dominant antigen in Crohn disease. J Clin Invest 113: 1296–1306, 2004.[CrossRef][Web of Science][Medline]
  30. Madara JL, Parkos C, Colgan S, MacLeod RJ, Nash S, Matthews J, Delp C, Lencer W. Cl- secretion in a model intestinal epithelium induced by a neutrophil-derived secretagogue. J Clin Invest 89: 1938–1944, 1992.[Web of Science][Medline]
  31. McCall CE, Grosso-Wilmoth LM, LaRue K, Guzman RN, Cousart SL. Tolerance to endotoxin-induced expression of the interleukin-1 beta gene in blood neutrophils of humans with the sepsis syndrome. J Clin Invest 91: 853–861, 1993.[Web of Science][Medline]
  32. McCormick BA, Colgan SP, Delp-Archer C, Miller SI, Madara JL. Salmonella typhimurium attachment to human intestinal epithelial monolayers: transcellular signalling to subepithelial neutrophils. J Cell Biol 123: 895–907, 1993.[Abstract/Free Full Text]
  33. McCormick BA, Franklin DP, Laux DC, Cohen PS. Type 1 pili are not necessary for colonization of the streptomycin-treated mouse large intestine by type 1-piliated Escherichia coli F-18 and E. coli K-12. Infect Immun 57: 3022–3029, 1989.[Abstract/Free Full Text]
  34. McCormick BA, Miller SI, Carnes D, Madara JL. Transepithelial signaling to neutrophils by salmonellae: a novel virulence mechanism for gastroenteritis. Infect Immun 63: 2302–2309, 1995.[Abstract]
  35. Medvedev AE, Kopydlowski KM, Vogel SN. Inhibition of lipopolysaccharide-induced signal transduction in endotoxin-tolerized mouse macrophages: dysregulation of cytokine, chemokine, and toll-like receptor 2 and 4 gene expression. J Immunol 164: 5564–5574, 2000.[Abstract/Free Full Text]
  36. Mizel SB, Snipes JA. Gram-negative flagellin-induced self-tolerance is associated with a block in interleukin-1 receptor-associated kinase release from toll-like receptor 5. J Biol Chem 277: 22414–22420, 2002.[Abstract/Free Full Text]
  37. Moors MA, Li L, Mizel SB. Activation of interleukin-1 receptor-associated kinase by gram-negative flagellin. Infect Immun 69: 4424–4429, 2001.[Abstract/Free Full Text]
  38. Naik S, Kelly EJ, Meijer L, Pettersson S, Sanderson IR. Absence of Toll-like receptor 4 explains endotoxin hyporesponsiveness in human intestinal epithelium. J Pediatr Gastroenterol Nutr 32: 449–453, 2001.[CrossRef][Web of Science][Medline]
  39. Nomura F, Akashi S, Sakao Y, Sato S, Kawai T, Matsumoto M, Nakanishi K, Kimoto M, Miyake K, Takeda K, Akira S. Cutting edge: endotoxin tolerance in mouse peritoneal macrophages correlates with down-regulation of surface toll-like receptor 4 expression. J Immunol 164: 3476–3479, 2000.[Abstract/Free Full Text]
  40. Oshitani N, Watanabe K, Nakamura S, Fujiwara Y, Higuchi K, Arakawa T. Dislocation of tight junction proteins without F-actin disruption in inactive Crohn's disease. Int J Mol Med 15: 407–410, 2005.[Web of Science][Medline]
  41. Otte JM, Cario E, Podolsky DK. Mechanisms of cross hyporesponsiveness to Toll-like receptor bacterial ligands in intestinal epithelial cells. Gastroenterology 126: 1054–1070, 2004.[CrossRef][Web of Science][Medline]
  42. Reed KA, Hobert ME, Kolenda CE, Sands KA, Rathman M, O’Connor M, Lyons S, Gewirtz AT, Sansonetti PJ, Madara JL. The Salmonella typhimurium flagellar basal body protein FliE is required for flagellin production and to induce a proinflammatory response in epithelial cells. J Biol Chem 277: 13346–13353, 2002.[Abstract/Free Full Text]
  43. Sato S, Nomura F, Kawai T, Takeuchi O, Mèuhlradt PF, Takeda K, Akira S. Synergy and cross-tolerance between toll-like receptor (TLR) 2- and TLR4-mediated signaling pathways. J Immunol 165: 7096–7101, 2000.[Abstract/Free Full Text]
  44. Sato S, Takeuchi O, Fujita T, Tomizawa H, Takeda K, Akira S. A variety of microbial components induce tolerance to lipopolysaccharide by differentially affecting MyD88-dependent and -independent pathways. Int Immunol 14: 783–791, 2002.[Abstract/Free Full Text]
  45. Söderholm JD, Olaison G, Lindberg E, Hannestad U, Vindels A, Tysk C, Järnerot G, Sjödahl R. Different intestinal permeability patterns in relatives and spouses of patients with Crohn's disease: an inherited defect in mucosal defence? Gut 44: 96–100, 1999.[Abstract/Free Full Text]
  46. Söderholm JD, Olaison G, Peterson KH, Franzén LE, Lindmark T, Wirén M, Tagesson C, Sjödahl R. Augmented increase in tight junction permeability by luminal stimuli in the non-inflamed ileum of Crohn's disease. Gut 50: 307–313, 2002.[Abstract/Free Full Text]
  47. Sitaraman SV, Klapproth JM, Moore DA 3rd, Landers C, Targan S, Williams IR, Gewirtz AT. Elevated flagellin-specific immunoglobulins in Crohn's disease. Am J Physiol Gastrointest Liver Physiol 288: G403–G406, 2005.[Abstract/Free Full Text]
  48. Takeda K, Akira S. TLR signaling pathways. Semin Immunol 16: 3–9, 2004.[CrossRef][Web of Science][Medline]
  49. Tallant T, Deb A, Kar N, Lupica J, de Veer MJ, DiDonato JA. Flagellin acting via TLR5 is the major activator of key signaling pathways leading to NF-kappa B and proinflammatory gene program activation in intestinal epithelial cells. BMC Microbiol 4: 33, 2004.[CrossRef][Medline]
  50. Wiley HS, Herbst JJ, Walsh BJ, Lauffenburger DA, Rosenfeld MG, Gill GN. The role of tyrosine kinase activity in endocytosis, compartmentation, and down-regulation of the epidermal growth factor receptor. J Biol Chem 266: 11083–11094, 1991.[Abstract/Free Full Text]
  51. Yoza B, LaRue K, McCall C. Molecular mechanisms responsible for endotoxin tolerance. Prog Clin Biol Res 397: 209–215, 1998.[Web of Science][Medline]
  52. Yu Y, Zeng H, Lyons S, Carlson A, Merlin D, Neish AS, Gewirtz AT. TLR5-mediated activation of p38 MAPK regulates epithelial IL-8 expression via posttranscriptional mechanism. Am J Physiol Gastrointest Liver Physiol 285: G282–G290, 2003.[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 Am. J. Physiol. Lung Cell. Mol. Physiol.Home page
K. Kato, W. Lu, H. Kai, and K. C. Kim
Phosphoinositide 3-kinase is activated by MUC1 but not responsible for MUC1-induced suppression of Toll-like receptor 5 signaling
Am J Physiol Lung Cell Mol Physiol, September 1, 2007; 293(3): L686 - L692.
[Abstract] [Full Text] [PDF]

Home page
 Am. J. Physiol. Gastrointest. Liver Physiol.Home page
O. Shibolet and D. K. Podolsky
TLRs in the Gut.IV. Negative regulation of Toll-like receptors and intestinal homeostasis: addition by subtraction
Am J Physiol Gastrointest Liver Physiol, June 1, 2007; 292(6): G1469 - G1473.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
292/3/G767    most recent
00447.2006v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Web of Science (5)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Sun, J.
Right arrow Articles by Hobert, M. E.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Sun, J.
Right arrow Articles by Hobert, M. E.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Visit Other APS Journals Online
Copyright © 2007 by the American Physiological Society.