MD-2 controls bacterial lipopolysaccharide hyporesponsiveness in human intestinal epithelial cells

Available online at www.sciencedirect.com

Life Sciences 82 (2008) 519 – 528 www.elsevier.com/locate/lifescie

MD-2 controls bacterial lipopolysaccharide hyporesponsiveness in human intestinal epithelial cells Christelle Lenoir, Catherine Sapin, Alexis H. Broquet, Anne-Marie Jouniaux, Sabine Bardin, Isabelle Gasnereau, Ginette Thomas, Philippe Seksik, Germain Trugnan, Joëlle Masliah, Maria Bachelet ⁎ INSERM U538, Centre Hospitalo-Universitaire Saint Antoine, 27 rue de Chaligny, 75012 Paris, France Received 13 September 2007; accepted 9 December 2007

Abstract Intestinal epithelial cells (IEC) have adapted to the presence of commensal bacteria through a state of tolerance that involves a limited response to lipopolysaccharide (LPS). Low or absent expression of two LPS receptor molecules, the myeloid differentiation (MD)-2 receptor, and toll-like receptor (TLR)4 was suggested to underlie LPS tolerance in IEC. In the present study we performed transfections of TLR4 and MD-2 alone or combined in different IEC lines derived from intestinal cancer (Caco-2, HT-29, and SW837). We found that LPS responsiveness increased more than 100-fold when IEC were transfected with MD-2 alone, but not TLR4. The release of interleukin (IL)-8, but also the expression of cyclooxygenase (Cox-)2 and the related secretion of prostaglandin (PG)E2 were coordinately stimulated by LPS in IEC transfected with MD-2 alone. Supernatants collected from MD-2-transfected IEC supported LPS activation of naïve HT-29, providing additional support to the concept that MD-2 alone endows IEC with LPS responsiveness. LPS responsiveness detected at concentrations as low as 110 pg/ml, and maximal values obtained by 10 ng/ml were clearly beyond those evoked by classical stimuli as IL-1β. In polarized cells, apical LPS stimulation was markedly more efficient than basolateral. Our data contradict previous opinion that both TLR4 and MD-2 limit IEC response to LPS, and emphasize the prominent role of MD-2 in intestinal immune responses to Gram-negative bacteria. © 2008 Elsevier Inc. All rights reserved. Keywords: HT29; Caco-2; SW837; TLR4; Cox-2; PGE2

Introduction In mammalian cells, the recognition of pathogen-associated invariant molecular patterns such as the bacterial endotoxin lipopolysaccharide (LPS) relies on the presence of a receptor complex which includes two indispensable molecules: the transmembrane toll like receptor (TLR)4 together with myeloid differentiation (MD)-2, a secreted glycoprotein that binds LPS (Medzhitov et al., 1997; Shimazu et al., 1999; Ulevitch and Tobias, 1999; Da Silva Correia et al., 2001; Visintin et al., 2001). In responsive cells such as monocytes, LPS-mediated activation leads to nuclear factor (NF)κ-B transcriptional

⁎ Corresponding author. Tel.: +33 1 40 01 13 34; fax: +33 1 40 01 13 90. E-mail address: [email protected] (M. Bachelet). 0024-3205/$ - see front matter © 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.lfs.2007.12.007

activity that switches on the transcription of a large number of genes including those encoding inflammatory cytokines and chemokines (Medzhitov et al., 1997; Tak and Firestein, 2001). Stimulation by LPS also leads to the activation of a set of mitogen activated protein kinases (MAPKs): extracellular signaling regulated kinases (ERKs), c-Jun N-terminal kinases (JNKs), and p38 MAPKs (Sweet and Hume, 1996; Yang et al., 2000). The distinct classes of MAPKs play important roles in various cellular events including the innate immune response, cell proliferation, differentiation, and apoptosis (Kyriakis and Avruch, 2001). Especially, the protein kinases p38 and JNK play an essential role in LPS responses by phosphorylating transcription factors including components of the AP-1 dimer, c-jun, and ATF-2 (Swantek et al., 1997; Smith et al., 2000). These factors are involved in the transcriptional activation of inducible genes including cyclooxygenase (Cox)-2, one rate

520

C. Lenoir et al. / Life Sciences 82 (2008) 519–528

Abcam (Cambridge, UK). Mouse monoclonal anti-human Cox-2, and goat polyclonal anti-human actin antibodies were from Santa Cruz Biotechnology (Santa Cruz, CA). Horseradish-peroxidaseconjugated or fluorescently labelled secondary antibodies were from Rockland (Gilbertsville, PA) or Jackson Immunoresearch (West Grove, PA). Blasticidin, the HA-tagged TLR4, and MD-2 expression vectors (punoha-htlr4a, puno-hmd-2), and the empty vector were from Cayla Invivogen (Toulouse, France). Acrylamide-bis-acrylamide was from Q-Biogene (Montreal, Canada). Dulbecco's modified Eagle's medium (DMEM), RPMI culture medium, fetal calf serum, penicillin-streptomycin, non essential amino acids, phosphate buffered saline (PBS), Trizol reagent, MMLV Reverse Transcriptase, Platinum® Taq DNA Polymerase were from Invitrogen (Paisley, UK). All other reagents not mentioned were from Sigma-Aldrich.

limiting enzyme in the conversion of arachidonic acid to prostaglandin (PG)E2 and other prostanoids (Simmons et al., 2004). Under physiological conditions, intestinal epithelium down regulates responses to bacteria to allow tolerance to beneficial commensal microflora (Duchmann et al., 1995; Abreu et al., 2001; Hornef et al., 2002). Low or absent expression of molecules involved in the recognition of bacterial patterns such as LPS has been reported to explain local bacterial hyporesponsiveness in the gut (Abreu et al., 2001; Dziarski et al., 2001). In agreement, co-transfection of TLR4 and MD-2 restored the ability of IEC to respond to LPS as attested by NFκB activation and IL-8 secretion (Abreu et al., 2001). It was suggested that IEC only lack those two recognition molecules, while the remaining intracellular LPS-dependent signaling pathway is unaltered. Further support to this proposal relies on the notion that LPS-mediated activation in many cells shares the intracellular signaling pathway utilized by IL-1β which is functional in IEC (O'Neill and Dinarello, 2000). However, later reports suggest a more prominent role for MD-2 in the physiologic mechanism that maintains LPS hyporesponsivenes in the intestine (Cario et al., 2006). Therefore, the current study was designed to provide further insight on the relative contribution of TLR4 and MD-2 in LPS responsiveness observed in IEC. We chose Caco-2, HT29, and SW837 IEC lines which are derived from human intestinal cancer since they are well characterized for NF-κB activation, and exhibit different levels of Cox-2 expression (Komatsu et al., 2005; Tsujii et al., 1997; Kakiuchi et al., 2002). Our results demonstrated that transfection with MD-2 alone was sufficient to confer LPS sensitivity as attested by a great increase in IL-8 and PGE2 secretion that correlated with an increased expression of Cox-2. We also observed an MD-2 like activity in culture supernatants from MD-2 transfected cells suggesting that MD-2 may be secreted into the medium. Our data provide novel basis to confine the functional LPS hyporesponsiveness of human IEC to defective expression and/or function of MD-2.

Electroporation was performed according to the manufacturer's instructions (Amaxa GmbH, Köln, Germany). Caco-2, HT29, and SW837 cells (5 × 105) were subjected to centrifugation and the pellet was resuspended in 100 µl of the Amaxa nucleofactor solution T or R, depending on the manufacturer's recommendations. TLR4-HA or/and MD-2 expression plasmids (5 µg) were mixed with 100 µl of cell suspension, transferred into Amaxa certified cuvettes (2 mm width) and electroporated with an Amaxa nucleofector apparatus using the O17 and W17 programmes according to the protocol supplied by the manufacturer. After electroporation, cells were resuspended in complete fresh medium and seeded in 24-well plates for 96 h until experiments. To establish stable clones, blasticidin (30 µg/ml) was added to the medium 2 days after transfection for 10 days and the medium was changed every 2 days. Resistant clones were isolated using cloning glass cylinders and tested by IL-8 secretion. Selected clones were further cultured under 10 µg/ml blasticidin.

Materials and methods

Cell culture and cell treatments

Reagents, antibodies, and expression vectors

HT29, SW837, and Caco-2 cells were obtained from the European Collection of Cell Cultures (ECACC, Wiltshire, UK). HT29, and SW837 cells were plated in 12–24 wells culture plates (Beckton Dickinson, Franklin Lakes, NJ), cultured in DMEM supplemented with 10% heat inactivated fetal calf serum, and 1% antibiotics at 37 °C in 5% CO2/air atmosphere until confluence. The human colon cancer cell line Caco-2 was maintained for 21 days in DMEM supplemented with 20% heat inactivated fetal calf serum, 1% non essential amino acids, and 1% penicillin-streptomycin at 37 °C in 10% CO2/air atmosphere. To allow separate access to the apical or basolateral medium of polarized Caco-2, cells were plated on tissueculture-treated polycarbonate Transwell filters (12 mm diameter, 0.4 µm pore size, Costar, Corning, NY), at a density of 100 × 103 cells per filter and grown for 21 days. Stimulation was performed either on the apical (upper chamber) or the basolateral (lower chamber) side of the cell layer. Cell polarization was confirmed by measuring transepithelial electrical

LPS (Escherichia coli 055:B5), phenol extracted and further purified by ion-exchange chromatography (N 1% protein, N 1% RNA), peptidoglycan (PG) (from Staphylococcus aureus), human recombinant IL-1β, and Bay 11-7082 were purchased from Sigma-Aldrich (Saint Louis, MO). SB203580 and SP600125 were from Biomol (Plymouth, PA). The protease inhibitor cocktail and the lactate dehydrogenase viability determination kit were from Roche Diagnostics GmbH, (Mannheim, Germany). Ficoll-Paque™ Plus, and the immunoblot detection system (ECL Plus), were from Amersham Biosciences (Buckinghamshire, UK). Bicinchoninic acid (BCA) protein assay reagent was from Pierce (Rockford, IL). Goat polyclonal antihuman TLR4 that recognizes the extracellular domain of the protein, IL-8 detection ELISA (DuoSet), and PGE2 immunoassay (EIA) kits were from R&D systems (Minneapolis, MN). Rabbit polyclonal anti-human MD-2 antibody was purchased from

Transfection of human IEC and generation of stably transfected cell lines

C. Lenoir et al. / Life Sciences 82 (2008) 519–528

resistance by the Millicell electrical resistance system (Millipore, Saint Quentin en Yvelines, France), and was N450 Ω/cm2. Peripheral blood human mononuclear cells were isolated from buffy coats of healthy donors through Ficoll Paque™ Plus density gradient centrifugation and further purified by 45 min adherence onto 12 wells culture plates. Cells were cultured in RPMI supplemented with 10% heat inactivated fetal calf serum and antibiotics at 37 °C in 5% CO2/air atmosphere. For cell treatments, different doses of LPS (1 pg/ml to 10 µg/ml), 100 ng/ml PG or 25 ng/ml IL-1β were dissolved in sterile PBS and added to the culture medium. In some experiments, cell culture supernatants from HT-29 cells (control or MD-2transfected) were collected following 24-h culture and concentrated 2-fold before addition to control HT-29 with or without LPS. At the end of the stimulation, supernatants were collected and stored at − 80 °C. Cells were washed twice in PBS and lysed in PBS containing 1% Triton X-100 and the protease inhibitor cocktail. In experiments designed to elucidate the role of NF-κB, p38, and JNK in LPS-mediated activation, cells were pre-incubated for 1 h with inhibitors (Bay 11-7082, SB203580, and SP600125) and further exposed to LPS for 18 h. Extraction of RNA and reverse transcription-polymerase chain reaction (RT–PCR) analysis for TLR4 and MD-2 Total RNA was isolated from Caco-2, HT-29, SW837, and peripheral monocytes, using Trizol reagent according to the manufacturer's instructions. First strand cDNA was synthesized from 5 µg of total RNA in a final volume of 20 µl containing 50 mM Tris-HCl (pH 8.4), 75 mM KCl, 3 mM MgCl2, 10 mM DTT, 0.5 mM each dNTP, 500 ng oligod(T)12–18 primer and 200 U of M-MLV reverse transcriptase for 1 h at 37 °C followed by 5 min at 75 °C. Subsequently, 2 µl of cDNA was specifically amplified by PCR with 1 U Platinum® Taq DNA polymerase in a solution containing 20 mM Tris-HCl (pH 8.4), 50 mM KCl, 1.5 mM MgCl2, 0.2 mM each dNTP and 25 pmoles of each primer. β-actin primers (forward: ATCATGTTTGAGA and reverse: TTGCGCTCAGGAGGAAGCAAT) were used as internal control. The TLR4 and MD-2 oligonucleotide primers were previously described (Medvedev et al., 2002). PCR amplifications were performed for 30 or 35 cycles at 94 °C for 15 s, 57 °C for 45 s, 72 °C for 45 s and a final primer extension at 72 °C for 5 min. The PCR products were run on 2% agarose gels, stained with ethidium bromide, and visualized with a transilluminator. The identity of all fragments was confirmed by sequencing. Flow cytometry Cells were harvested by adding a solution of 1 mM EDTA/ 1 mM EGTA in PBS (calcium and magnesium free) accompanied by gentle rocking to remove the cells from the plate. This procedure may take longer than normal trypsinization but avoids stripping of surface antigens. Cells were centrifuged at 1000 ×g for 5 min at 4 °C and the remainder of the protocol was performed on ice. Cells were fixed in 100 µl paraformaldehyde solution (3%) for 10 min, and exposed to primary antibodies

521

(diluted to 1/50 in PBS containing 1% BSA) for 1 h. Staining was performed by resuspending cells in species-specific FITCconjugated secondary antibodies diluted to 1/100 in PBS-BSA. In negative controls for background staining, paraformaldehyde-fixed cells were incubated with the secondary antibody alone. These values were not different from control values obtained with an isotype-specific FITC-conjugate secondary antibody. Fluorescence and light scatters were analysed in a BD Biosciences fluorescence activated cell sorter for green fluorescence detected through a 525-nm filter. Additional methods Protein concentrations were determined in cell lysates using BCA protein assay reagents and bovine serum albumin as standard according to the manufacturer's instructions. Quantitative levels of IL-8, and PGE2 were determined in cell supernatants by commercially available ELISA and EIA kits, respectively, according to the instructions of the manufacturer. Cell viability was monitored by measuring the release of lactate dehydrogenase using a commercial kit, as specified by the manufacturer. Statistics Data are presented as mean± SEM and comparisons were performed using the Wilcoxon's signed rank test. A p value≤ 0.05 was considered statistically different. Results Overexpression of MD-2 alone was sufficient to confer LPS responsiveness to hyporesponsive IEC Secretion of interleukin-8 The role of MD-2 and TLR4 in LPS responsiveness was initially studied in standard non transfected IEC or after transient transfection with MD-2 alone, TLR4 alone, or cotransfection with both MD-2 and TLR4 human cDNAs. Cells were stimulated overnight with LPS (100 ng/ml), IL-1β (25 ng/ ml) used as positive control to stimulate IL-8 secretion, or PG (100 ng/ml), a TLR2 ligand (Dziarski et al., 2001). Non transfected HT29, and SW837 cells exhibited a slight to moderate response to LPS while Caco-2 appeared unresponsive (Fig. 1A, B and C). However, as could be expected, IL-1β significantly increased IL-8 secretion by non transfected HT29, SW837 and Caco-2 cells (Fig. 1A, B and C). Values obtained in standard non transfected cells were indistinguishable from those observed in cells transfected with either the empty vector or TLR4 alone (Fig. 1A, B and C). However, HT29 became highly responsive to LPS when transfected with MD-2 alone: IL-8 concentrations were increased by 102 fold, and 121 fold above basal levels in HT29 cells transfected with MD-2 alone or cotransfected with TLR4/MD-2, respectively (Fig. 1A). Thus, cotransfection with TLR4 only slightly increased the response observed following transfection with MD-2 alone in HT29 IEC. Of note, LPS responses observed in MD-2 transfected HT29

522

C. Lenoir et al. / Life Sciences 82 (2008) 519–528

Fig. 1. MD-2 confers LPS responsiveness to human IEC as attested by IL-8 secretion. HT29 (A), SW837 (B), and Caco-2 (C) cells were transiently transfected with the empty vector (control), TLR4 alone, MD-2 alone or co-transfected with TLR4/MD-2. Cells were left untreated (open bars) or stimulated overnight with 100 ng/ml LPS (filled bars), 25 ng/ml IL-1β (hatched bars) or 100 ng/ml PG (dotted bars). Cell stimulations started at 72 h after transient tranfections and supernatants collected were stored at − 80 °C until measuring IL-8 by ELISA. Values represent means ± SEM from 4 independent experiments. (D) IL-8 secretion was measured in supernatants from control HT29 (open bars) or stable MD-2 transfects (filled bars) stimulated overnight with increasing concentrations of LPS (1 pg/ml to 10 µg/ml). Data represent means ± SEM from 4 independent experiments. (E) For comparison, IL-8 secretion was measured in supernatants from human monocytes stimulated overnight with LPS at the same concentrations (means ± SEM, n = 4).

were significantly higher than those induced by IL-1β, considered as a potent stimulus for IL-8 secretion in IEC, which shares most of the LPS intracellular signaling machinery. As observed in HT29, transient transfection of SW837 cells with MD-2 alone induced a 113 fold increase in LPS responsiveness that was only slightly increased in cells cotransfected with TLR4 (Fig. 1B). In Caco-2 cells, transfection with MD-2 alone induced a less intense but significantly increased response to LPS, and this effect was further increased when TLR4 was also transfected (Fig. 1C). We next performed stable transfections of MD-2 in HT29, and as expected, these

cells exhibited a strong response to LPS in a dose dependent fashion. LPS stimulated IL-8 secretion at concentrations as low as 10 pg/ml, to reach a maximal effect by 10 ng/ml that decreased at higher LPS concentrations, as illustrated in Fig. 1D. This diminished response to high doses of LPS could not be accounted for by LPS toxicity since no alterations in cell viability were noticed as attested for by the release of lactate dehydrogenase. This effect might rather reflect the establishment of a desensitization state after long exposure to high doses of LPS, as described for repeated stimulations with LPS (Otte et al., 2004). Both the concentration of LPS required

C. Lenoir et al. / Life Sciences 82 (2008) 519–528

and the kinetics of IL-8 secretion were found similar in MD-2 transfected IEC as compared to monocytes. Maximal responses to LPS observed in IEC were, however, always far below normal monocytic cells (Fig. 1E). MD-2 effect was specific for LPS since no significant response was observed in the presence of the Gram-positive bacterial peptidoglycan (PG, 100 ng/ml), known to mediate signaling through TLR2 (Fig. 1A, B and C). Basal levels of IL-8 secretion were also significantly increased in non stimulated MD-2 and MD-2/TLR4 transfected cells, as

523

compared to non transfected cells or cells transfected with the empty vector, particularly in HT29 cells (175 pg/mg proteins in non transfected HT29, and 515 pg/mg proteins in MD-2 transfects, p b 0.01, n = 4) (Fig. 1A). These basal activated levels were also observed when cells were incubated with medium containing low LPS fetal calf serum (b10 EU). Thus, transfection with MD-2 activated IEC in the absence of LPS suggesting that MD-2 not only binds LPS but might also affect the mechanism that triggers cellular activation.

Fig. 2. MD-2 confers LPS responsiveness to human IEC as attested by PGE2 secretion. HT29 (A), SW837 (B), and Caco-2 cells (C) were transiently transfected with the empty vector (control), TLR4 alone, MD-2 alone or co-transfected with TLR4/MD-2. Cells were left either untreated (open bars) or stimulated overnight with 100 ng/ml LPS (filled bars), 25 ng/ml IL-1β (hatched bars) or 100 ng/ml PG (dotted bars). PGE2 was measured by EIA in the same supernatants used to evaluate IL-8. Data represent means ± SEM from 4 independent experiments. (D) PGE2 secretion was measured in supernatants from control HT29 (open bars) or MD-2 stable transfects (filled bars) stimulated overnight with increasing concentrations of LPS (1 pg/ml to 10 µg/ml). Data presented are means ± SEM from 4 independent experiments. (E) PGE2 secretion was measured in supernatants from human monocytes (means ± SEM, n = 4).

524

C. Lenoir et al. / Life Sciences 82 (2008) 519–528

Fig. 3. Dose-dependent stimulation of Cox-2 expression by LPS in HT29 IEC stably transfected with MD-2. Exposure of HT-29 stable transfects expressing MD-2 to increasing doses of LPS (1 pg/ml to 10 µg/ml, overnight) resulted in a dose-dependant increase of Cox-2 expression detected with a specific monoclonal antibody approximately at 70–72 kDa by immunoblot analysis as detailed in Materials and methods. No detectable response was observed in control HT29. Data are representative from at least 3 experiments with similar results.

Secretion of prostaglandin E2 To evaluate whether MD-2 overexpression could affect the LPS-mediated stimulation of the eicosanoid cascade in IEC we measured the secretion of PGE2, the major arachidonic acid metabolite produced by IEC. An initial set of experiments was performed in transiently transfected IEC. Subsequent experiments were performed with HT29 cells stably transfected with MD-2 alone. As observed for IL-8, transfection with TLR4 alone had no significant effect on LPS-induced PGE2 secretion in IEC. However, as compared to control levels, secreted PGE2 increased approximately by 21 fold following exposure to LPS in HT29 cells tranfected with MD-2 alone, and this level was only weakly increased when TLR4 was co-tranfected (25 fold increase) (Fig. 2A). Likewise, exposure to LPS of SW837 cells transfected with MD-2 induced a 20 fold increased PGE2 secretion as compared with a 21 fold increase in the same cells co-transfected with TLR4, and MD-2 (Fig. 2B). In Caco-2 cells, maximal responses were observed in cells co-transfected with MD-2 and TLR4, though transfection with MD-2 alone induced a significant increase in PGE2 secretion (Fig. 2C). In HT29 cells stably transfected with MD-2, LPS induced a dose-dependent increase of PGE2 secretion, as shown in Fig. 2D. Again, the effect of LPS was higher in HT29 and SW837 cells as compared to Caco-2 but did not reach the levels of monocytic cells (Fig. 2E). In agreement with IL-8 results, transfected IEC remained unresponsive to PG. Expression of cyclooxygenase-2 Cox-2 is one of the major targets for LPS stimulation in monocytic cells, and plays a crucial role in the biosynthesis of PGE2. The effect of LPS was therefore investigated also at the level of Cox-2 expression by immunoblot. Low constitutive levels of Cox-2 were observed in the three cell lines studied though HT29 and SW837 cells expressed higher levels as compared to Caco-2. The highest expression as well as the

highest response to LPS was observed in SW837 cells (data not shown). Increased expression of Cox-2 was always associated with increased cyclooxygenase activity, as determined by the production of PGE2. In experiments shown in Fig. 3, control HT29 or cells stably transfected with MD-2 were left untreated or exposed to increasing concentrations of LPS (1 pg/ml to 10 µg/ml, overnight). Fig. 3A shows that untreated control HT29 cells exhibit low but detectable basal levels of Cox-2 that were poorly stimulated by LPS. However, LPS promoted an intense increase of Cox-2 expression in MD-2 transfected HT29 IEC. This effect was dose dependent though very high doses appeared less efficient, and maximal stimulation was observed between 10-100 ng/ml LPS (Fig. 3B). As could be expected, the occurrence of Cox-2 appeared proportional to the secretion of PGE2. Culture supernatants from MD-2 transfected HT-29 confer LPS responsiveness to control HT29 In these experiments we investigated the capacity of culture supernatants collected from MD-2 transfected HT29 IEC to support LPS activation. As illustrated in Fig. 4, levels of LPS activation attested by IL-8 secretion were greatly enhanced in control HT-29 activated in the presence of culture supernatants from HT-29-MD-2 as compared with cells activated in the presence of culture supernatants from control HT-29. Supernatants from MD-2 transfected cells had no enhancing effect when cells were stimulated with gram-positive PG (not shown), indicating that this effect was specific for the interaction of IEC with LPS. The capacity to confer LPS responsiveness of supernatants from MD-2 transfected cells was, however, less efficient that transfection. LPS-mediated stimulation is a polarized event in responsive polarized IEC The apical versus basolateral LPS stimulating activity was tested in stably transfected polarized Caco-2 cells grown for 3 weeks on Transwell filters. As shown in Fig. 5, when added

Fig. 4. HT-29 IEC stably transfected with MD-2 secrete MD-2-like activity. Control HT-29 IEC were incubated overnight with supernatants collected from control or MD-2 transfected IEC, in the absence or in the presence of LPS (10 ng/ml). Results are the mean ± SEM from two separate experiments measured in triplicate.

C. Lenoir et al. / Life Sciences 82 (2008) 519–528

525

LPS-mediated effects in transfected IEC involve mainly NF-kB for IL-8 secretion, and MAPKs for PGE2 secretion

Fig. 5. LPS stimulates stably transfected polarized Caco-2 cells preferentially via the apical pole. Stably transfected Caco-2 cells were grown on Transwell filter until spontaneous polarization (3 weeks) and then stimulated overnight with LPS (10 ng/ml) on the apical (upper chamber) or the basolateral (lower chamber) pole. IL-8 was measured in the apical (white bars) or the basolateral (black bars) supernatants. Results are the mean ± SEM from four separate experiments ⁎p b 0.01 versus basolateral secretion.

onto the apical pole of Caco-2 cells LPS induced a strong secretion of IL-8 that was significantly higher on the basolateral side as compared to the apical side. Basolateral stimulation of LPS was clearly less efficient though also resulted in a higher basolateral secretion of IL-8.

NF-κB and MAPK activation have been involved in LPSmediated activation of responsive cells (Swantek et al., 1997; Abreu et al., 2001; Wadleigh et al., 2000). The effect of 1 h pretreatment with various pharmacological inhibitors of these signaling cascades was evaluated and compared regarding IL-8 and PGE2 secretion. Pre-treatment of stable MD-2 transfected HT29 cells with the NF-κB inhibitor Bay 11-7082 strongly reduced (84%) LPS-mediated IL-8 secretion. The effect observed with MAPK inhibitors was clearly less efficient, as compared to NF-κB inhibition (59% for SB203580 and 49% for SP600125) (Fig. 6A). Comparatively, the relative efficiency of these MAPKs inhibitors was significantly higher when considering PGE2 secretion: 93% for SB203580, and 89% for SP600125, while the effect of NF-κB inhibition was less efficient and only attained 47% inhibition (Fig. 6B). Thus, NF-κB is mainly responsible for elevated IL-8 secretion while MAPKs may be mainly required for induction of Cox-2 expression and PGE2 release following stimulation of IEC by LPS. Standard human IEC express detectable levels of endogenous MD-2, and TLR4 mRNA It is clear from RT–PCR data that at the level of mRNA IEC express endogenous TLR4 and MD-2. Though endogenous levels of mRNA for TLR4 were lower in Caco-2 cells, similar

Fig. 6. Effect of NF-κB, p-38 and JNK MAPKs inhibitors on LPS-dependent IL8 and PGE2 secretion by stable IEC transfects. (A) IL-8 secretion was measured by ELISA in culture supernatants from HT29 cells stably transfected with MD2. Cells were either left untreated (open bars) or stimulated overnight with 10 ng/ ml LPS (filled bars). The p-38 inhibitor SB203580 (20 µM, sloped hatched bars), the JNK inhibitor SP600125 (20 µM, dotted bars), and the NF-κB inhibitor Bay 11-7082 (25 µM, horizontally hatched bars) were added to the culture medium 1 h before stimulation with LPS. Data represent means ± SEM from 4 independent experiments. ⁎p b 0.05 vs LPS alone. (B) Secretion of PGE2 was measured by EIA in the same supernatants used for IL-8 determination. ⁎p b 0.05 vs LPS alone.

Fig. 7. Endogenous expression of TLR4 and MD-2 mRNA in human IEC and monocytes. Cultures of standard HT29, SW837, Caco-2, and monocytes were examined for expression of TLR4, and MD-2 mRNA by RT–PCR, using specific probes as described in Materials and methods. Actin was used as control for RNA integrity and C-RTase as control for reverse transcriptase. PCR products were electrophoresed on agarose gels and photographed.

526

C. Lenoir et al. / Life Sciences 82 (2008) 519–528

levels were observed in HT29, SW837, and human monocytes, the highly LPS-responsive cell used as positive control (Fig. 7). Endogenous MD-2 mRNA levels were maximal in human monocytes, appeared low in Caco-2, and slightly higher in HT29 and SW837 (Fig. 7). Thus, standard IEC most strikingly differ from monocytes regarding their lower mRNA content of MD-2. Unlike MD-2, comparable levels of TLR4 mRNA were observed in high responsive monocytes and low responsive IEC.

stable transfects expressing MD-2, moderate levels of cell surface expression were observed for this protein (43% positive cells). However, these levels were higher than those observed in control HT29 (12%, Fig. 8). The level of surface TLR4 appeared unaffected by MD-2 transfection in HT29 IEC (62% in control cells and 57% in MD-2 stable transfects). Cell membrane expression of MD-2 and TLR4 in human monocytes is also illustrated in Fig. 8 for comparison. Discussion

Increased cell surface expression of MD-2 in HT29 IEC stably transfected with MD-2 Cell surface expression levels of TLR4 and MD-2 were further analyzed by flow cytometry in IEC. Analysis of control HT-29 cells revealed low levels of cell surface MD-2, as compared with moderate levels of TLR4 (Fig. 8). In HT29

We report here that IEC hyporesponsiveness to Gramnegative bacteria can be mainly attributed to defective expression of one molecule involved in the recognition of LPS: the small glycosylated protein, MD-2. Our data suggested that in normal IEC TLR4 was sufficiently expressed while MD-2 was below levels required to support an efficient LPS signaling.

Fig. 8. Expression of cell surface TLR4 and MD-2 in HT29 IEC: flow cytometric study. MD-2 and TLR4 surface expression were analysed by flow cytometry in non permeabilized HT29 IEC (control or stably transfected with MD-2). For comparison, cell surface of MD-2 and TLR4 was measured in the same conditions in human monocytes. The left panels represent MD-2 surface expression while the right panels represent TLR4 expression. Shaded areas represent background staining with the FITC-labelled secondary antibody alone (or the isotype specific FITC-conjugated secondary antibody) and the open areas show the fluorescence staining with MD-2 and TLR4 specific antibodies (numbers in brackets represent percent of positive cells). Flow cytometric profiles shown in each panel are acquired in the same experiment, and are representative of at least 3 unrelated experiments.

C. Lenoir et al. / Life Sciences 82 (2008) 519–528

Accordingly, transfection of IEC with MD-2 alone but not TLR4 conferred LPS responsiveness that led to a coordinate upregulation of Cox-2 expression, PGE2, and IL-8 secretion. LPS stimulated polarized IEC preferentially on the apical side, thus where Gram-negative bacteria should occur. Culture supernatants from MD-2-transfected HT29 greatly enhanced the capacity of LPS to stimulate control HT29 implying that soluble secreted MD-2 might confer LPS responsiveness to hyporesponsive IEC. Prior exposure of stable MD-2 transfects to pharmacological inhibitors suggested that activation of p38 and JNK MAPKs was mainly involved in LPS-mediated upregulation of Cox-2 while IL-8 secretion primarily depended on NF-κB activation. Thus, the LPS-induced signaling pathway observed in IEC leads to distinct activation of NF-κB and MAPK activities. This is not an unexpected result, since both pathways are known to be involved in the adverse effects of LPS in cells from the monocytic lineage (Sweet and Hume, 1996; Yang et al., 2000). However, their involvement and their relative contribution to the LPS-mediated response had not been so far examined in IEC. Since we found that human intestinal IEC constitutively express MD-2, we suggest that the protein needs to be in excess, as compared to TLR4, to have significant biological activity. In agreement, a recent report (Mullen et al., 2003) demonstrated that optimal LPS receptor function requires multiple molecules of MD-2 bound to TLR4, and small changes in MD-2 concentrations induce large changes in LPS recognition. Significant differences in MD-2 expression between control hyporesponsive and transfected responsive IEC were mainly observed at the surface level. One can expect, therefore, that cell surface availability of MD-2 mainly regulates TLR4 function and thereby LPS responsiveness in the intestinal epithelium. We found, however, that maximal responses to LPS might differ strikingly, while MD-2 expression only exhibited moderate differences. At identical stimulation conditions Caco-2 cells released approximately 0.5 ng/mg proteins of IL-8, as compared to 3 ng/mg proteins in HT-29 and more than 3000 ng/mg proteins in monocytes. Higher responses observed in monocytic cells might be explained by the presence in these cells of higher cell surface expression of MD-2 or other constituents of the LPS receptor complex as LBP and CD14 that greatly enhance the final response (Ulevitch and Tobias, 1999). Further explanations for these differences observed between transfected IEC, and normal monocytes might involve other differences at the level of specific signaling pathways. The existence of an additional, so far unidentified component of the LPS receptor might also be considered. So far, one study has been performed in transfected IEC indicating that both TLR4 and MD-2, but not MD-2 alone, were required to allow LPS responsiveness in these cells (Abreu et al., 2001). Our hypothesis to explain this discrepancy relies on the efficiency of transient transfection that appeared limited in the latter study. Because low transfection efficiency may lead to insufficient expression or function of MD-2, in the present study we chose a highly efficient nucleofector technology that allowed us to optimize the process, and to perform stable transfections. Otherwise, our results support previous findings that human IEC are defective in the surface recognition for LPS,

527

while the remaining intracellular signaling pathway leading to LPS-mediated activation is able to function. We provide here, however, a more thorough exploration, and add novel evidence that extends previous findings, and point out the crucial need for membrane MD-2 in the LPS response leading to increased Cox2 expression in human IEC. This is a subject of importance since the metastatic potential of human IEC was shown to be correlated to Cox-2 expression (Tsujii et al., 1997; Kakiuchi et al., 2002), and the mechanisms leading to Cox-2 overexpression in human cancer remain poorly understood. Therefore, dysregulated responses to bacteria should be related not only to inflammatory bowel diseases but also to aberrant overexpression of Cox-2 observed in tumor cells. This is relevant to what we know about the markedly elevated risk of cancer in a variety of chronic inflammatory conditions (Potter and Ulrich, 2006). Other studies performed both in vivo and in vitro in cells from knockout mice support the concept that MD-2 confers responsiveness to cells expressing TLR4 alone (Shimazu et al., 1999; Nagai et al., 2002). Additional in vitro data on LPS receptor function were mainly obtained using transient transfection and NF-κB reporter gene assay systems in human embryonic epithelial kidney cells which are easier to transfect than IEC. However, the specificity of the endogenous pathways may be cell-dependent, and data obtained with different cells should not be extrapolated to human IEC that, unlike other cell types, functionally develop an LPS tolerant phenotype. IEC are the first barrier between luminal bacteria and the mucosal innate immune system and the potential mechanisms by which IEC limit LPS-dependent activation that control bacterial tolerance might be crucial to understand so far unresolved events occurring during dysregulated host-microbial interactions in these cells. Though the understanding of the role of MD-2 in the recognition and signaling of LPS has intensively grown from the beginning of the 21st century, few studies have focused on the incidence of MD-2 on human diseases. Circulating MD-2 activity was recently demonstrated in patients with septic shock (Pugin et al., 2004) and might represent an important target for anti-inflammatory strategies in different chronic inflammatory (or infectious) diseases. Acknowledgements The authors thank Michel Kornprobst for assistance in flow cytometry, and Lucette Groisard for technical assistance. Alexis H. Broquet was supported by a grant from the Research Ministry of France (PRFMMIP). References Abreu, M., Vora, P., Faure, E., Thomas, L.S., Arnold, E.T., Arditi, M., 2001. 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. Journal of Immunology 167, 1609–1617. Cario, E., Golenbock, D.T., Visintin, A., Runzi, M., Gerken, G., Podolsky, D.K., 2006. Trypsin-sensitive modulation of intestinal epithelial MD-2 as mechanism of lipopolysaccharide tolerance. Journal of Immunology 176, 4258–4266.

528

C. Lenoir et al. / Life Sciences 82 (2008) 519–528

Da Silva Correia, J., Soldau, K., Christen, U., Tobias, P.S., Ulevitch, R.J., 2001. Lipopolysaccharide is in close proximity to each of the proteins in its membrane receptor complex. Transfer from CD14 to TLR4 and MD-2. Journal of Biological Chemistry 276, 21129–21135. Duchmann, R., Kaiser, I., Hermann, E., Mayet, W., Ewe, K., Meyer zum Buschenfelde, K.H., 1995. Tolerance exists towards resident intestinal flora but is broken in active inflammatory bowel disease. Clinical and Experimental Immunology 102, 445–447. Dziarski, R., Wang, Q., Miyake, K., Kirschning, C.J., Gupta, D., 2001. MD-2 enables toll-like receptor 2 (TLR2)-mediated responses to lipopolysaccharide and enhances TLR2-mediated responses to gram-negative bacteria and their cell wall components. Journal of Immunology 166, 1938–1944. Hornef, M.W., Frisan, T., Vandewalle, A., Normark, S., Richter-Dahlfors, A., 2002. Toll-like receptor 4 resides in the Golgi apparatus and colocalizes with internalized lipopolysaccharide in intestinal epithelial cells. Journal of Experimental Medicine 195, 559–570. Kakiuchi, Y., Tsuji, S., Tsujii, M., Murata, H., Kawai, N., Yasumaru, M., Kimura, A., Komori, M., Irie, T., Miyoshi, E., Sasaki, Y., Hayashi, N., Kawano, S., Hori, M., 2002. Cyclooxygenase-2 activity altered the cellsurface carbohydrate antigens on colon cancer cells and enhanced liver metastasis. Cancer Research 62, 1567–1572. Komatsu, K., Buchanan, F.G., Katkury, S., Morrow, J.D., Inoue, H., Otaka, M., Watanabe, S., DuBois, R.N., 2005. Oncogenic potential of MEK1 in rat intestinal epithelial cells is mediated via cyclooxygenase-2. Gastroenterology 129, 577–590. Kyriakis, J.M., Avruch, J., 2001. Mammalian mitogen-activated protein kinase signal transduction pathways activated by stress and inflammation. Physiological Reviews 81, 807–869. Medvedev, A.E., Lentschat, A., Wahl, L.M., Golenbock, D.T., Vogel, S.N., 2002. Dysregulation of LPS-induced toll-like receptor 4-MyD88 complex formation and IL-1 receptor-associated kinase 1 activation in endotoxin tolerant cells. Journal of Immunology 169, 5209–5216. Medzhitov, R., Preston-Hurlburt, P., Janeway Jr., C.A., 1997. A human homologue of the Drosophilla Toll protein signals activation of adaptive immunity. Nature 388, 394–397. Mullen, G.E.D., Kennedy, M.N., Visintin, A., Mazzoni, A., Leifer, C.A., Davis, D.R., Segal, D.M., 2003. The role of disulfide bonds in the assembly and function of MD-2. Proceedings of the National Academy of Sciences USA 100, 3919–3924. Nagai, Y., Akashi, S., Nagafuku, M., Ogata, M., Iwakura, Y., Akira, S., Kitamura, T., Kosugi, A., Kimoto, M., Miyake, K., 2002. Essential role of MD-2 in LPS responsiveness and TLR4 distribution. Nature Immunology 3, 667–672. O'Neill, L.A.J., Dinarello, C.A., 2000. The IL-1 receptor/toll-like receptor superfamily:crutial receptors for inflammation and host defense. Immunology Today 21, 206–209.

Otte, J.M., Cario, E., Podolsky, D.K., 2004. Mechanisms of cross hyporesponsiveness to toll-like receptor bacterial ligands in intestinal epithelial cells. Gastroenterology 128, 1054–1070. Potter, J.D., Ulrich, C.M., 2006. Cox-2 and gastric cancer: more on inflammation and neoplasia. Gastroenterology 130, 2198–2200. Pugin, J., Stem-Voeffray, S., Daubeuf, B., Matthay, M.A., Elson, G., 2004. Soluble MD-2 activity in plasma from patients with severe sepsis and septic shock. Blood 104, 4071–4079. Shimazu, R., Akashi, S., Ogata, H., Nagai, Y., Fukudome, K., Miyake, K., Kimoto, M., 1999. MD-2, a molecule that confers lipopolysaccharide responsiveness on tolllike receptor 4. Journal of Experimental Medicine 189, 1777–1782. Simmons, D.L., Botting, R.M., Hla, T., 2004. Cyclooxygenase isozymes: the biology of prostaglandin synthesis and inhibition. Pharmacological Reviews 56, 387–437. Smith, J.L., Schaffner, A.E., Hofmeister, J.K., Hartman, M., We, G., Forsthoefel, D., Hume, D.A., Ostrowski, M.C., 2000. ets-2 is a target for an akt (protein kinase B)/jun N-terminal kinase signaling pathway in macrophages of motheatenviable mutant mice. Molecular and Cellular Biology 20, 8020–8034. Swantek, J.L., Cobb, M.H., Geppert, T.D., 1997. Jun N-terminal kinase/stressactivated protein kinase (JNK/SAPK) is required for lipopolysaccharide stimulation of tumor necrosis factor alpha (TNF-alpha) translation: glucocorticoids inhibit TNF-alpha translation by blocking JNK/SAPK. Molecular and Cellular Biology 17, 6274–6282. Sweet, M.J., Hume, D.A., 1996. Endotoxin signal transduction in macrophages. Journal of Leukocyte Biology 60, 8–26. Tak, P.T., Firestein, G.S., 2001. NF-kB: a key role in inflammatory diseases. Journal of Clinical Investigation 107, 7–11. Tsujii, M., Kawano, S., DuBois, R.N., 1997. Cyclooxygenase-2 expression in human colon cancer cells increases metastatic potential. Proceedings of the National Academy of Sciences USA 94, 3339–3340. Ulevitch, R.J., Tobias, P.S., 1999. Recognition of Gram-negative bacteria and endotoxin by the innate immune system. Current Opinion in Immunology 11, 19–22. Visintin, A., Mazzoni, A., Spitzer, J.A., Segal, D.A., 2001. Secreted MD-2 is a large polymeric protein that efficiently confers lipopolysaccharide sensitivity to toll-like receptor 4. Proceedings of the National Academy of Sciences USA 98, 12156–12161. Wadleigh, D.J., Reddy, S.T., Kopp, E., Ghosh, S., Herschman, H.R., 2000. Transcriptional activation of the cyclooxygenase-2 gene in endotoxin-treated RAW 264.7 macrophages. Journal of Biological Chemistry 275, 6259–6266. Yang, H., Young, D.W., Gusovsky, F., Chow, J.C., 2000. Cellular events mediated by lipopolysaccharide-stimulated toll-like receptor 4. Journal of Biological Chemistry 275, 20861–20866.