Lipopolysaccharide Induces Cyclooxygenase-2 in Intestinal Epithelium via a Noncanonical p38 MAPK Pathway This information is current as of May 29, 2013.
Anatoly V. Grishin, Jin Wang, Douglas A. Potoka, David J. Hackam, Jeffrey S. Upperman, Patricia Boyle, Ruben Zamora and Henri R. Ford J Immunol 2006; 176:580-588; ; http://www.jimmunol.org/content/176/1/580
Subscriptions Permissions Email Alerts
This article cites 53 articles, 26 of which you can access for free at: http://www.jimmunol.org/content/176/1/580.full#ref-list-1 Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscriptions Submit copyright permission requests at: http://www.aai.org/ji/copyright.html Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/cgi/alerts/etoc
The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 9650 Rockville Pike, Bethesda, MD 20814-3994. Copyright © 2006 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606.
Downloaded from http://www.jimmunol.org/ by guest on May 29, 2013
References
The Journal of Immunology
Lipopolysaccharide Induces Cyclooxygenase-2 in Intestinal Epithelium via a Noncanonical p38 MAPK Pathway1 Anatoly V. Grishin,2* Jin Wang,* Douglas A. Potoka,† David J. Hackam,† Jeffrey S. Upperman,† Patricia Boyle,* Ruben Zamora,† and Henri R. Ford*
B
acterial colonization of the gut, together with formula feeding and hypoxia, are well-established risk factors in necrotizing enterocolitis (NEC),3 a severe intestinal inflammation in preterm infants (1–3). According to the existing model of NEC, bacteria breaching the gut barrier cause inflammation by stimulating production of inflammatory factors in the specialized cells of the innate immune system, such as macrophages or monocytes (1, 2). However, enterocytes have been reported to produce a number of proinflammatory molecules in response to bacterial stimuli (4 – 6), suggesting an intriguing possibility that the intestinal epithelium may be a front-line sensor of bacterial presence in NEC. Cyclooxygenases (COXs), key enzymes in the biosynthesis of prostanoids, play important roles in the gut. COXs are critical for the maintenance of the intestinal epithelium (7). COX inhibitors damage the gastrointestinal tract, which can be alleviated by synthetic prostanoids (8). Inflammatory bowel disease is associated
*Division of Pediatric Surgery, Childrens Hospital Los Angeles, Los Angeles, CA 90027; and †Division of Pediatric Surgery, Children’s Hospital of Pittsburgh, Pittsburgh, PA 15213 Received for publication January 6, 2005. Accepted for publication September 28, 2005. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1 This work was supported by a Start-Up Grant from the Research Advisory Committee, Children’s Hospital of Pittsburgh (to A.V.G.), and National Institutes of Health Grants R01-AI-494-73 and R01-Al-1403 (to H.R.F.). 2 Address correspondence and reprint requests to Dr. Anatoly V. Grishin, Division of Pediatric Surgery, Childrens Hospital Los Angeles, 4661 Sunset Boulevard, Los Angeles, CA 90027. E-mail address:
[email protected] 3 Abbreviations used in this paper: NEC, necrotizing enterocolitis; COX, cyclooxygenase; MKK, MAPK kinase; FAK, focal adhesion kinase; ATF2, activating transcription factor 2.
Copyright © 2005 by The American Association of Immunologists, Inc.
with high levels of COX (9 –11). COX knockout mice are prone to intestinal disorders (12, 13). High levels of intestinal COX have been found in an animal model of NEC (14). On one hand, inhibitors of COX, including glucocorticoids and nonsteroidal antiinflammatory drugs, have been implicated as a risk factor of NEC (3). On the other hand, COX-2 is proinflammatory, and abnormally high levels of this enzyme may be pathogenic during the intestinal inflammation (11, 15). These data indicate that COX may play both protective and deleterious roles in NEC. COX-2 is induced by cytokines, mitogens, and stresses at either transcriptional or posttranscriptional levels (reviewed in Ref. 16). Except for involvement of transcription factor AP-1 (17) and p38 MAPK (18, 19), the mechanisms of COX-2 regulation in enterocytes remain unknown. In other cell types, expression of COX-2 may be regulated by transcription factors NF-B (20 –23) and AP-l (24, 25); MAPK of the ERK, JNK, and p38 families (20 –23, 26 – 35); surface receptors and their adapter proteins (20, 21, 23, 36); protein kinases of PKC (23, 30, 32, 35) and Src (28) families; and PI3K (20, 35). The roles of individual MAPK (29, 37) and NF-B (38) remain controversial. Because different cells may use different signaling pathways to regulate COX-2, data obtained in other cell types cannot be extrapolated to enterocytes without proof. LPS, a component of Gram-negative bacteria, triggers inflammation by binding its cognate receptors on the surface of innate immune system sensory cells (macrophages) and stimulating production of proinflammatory molecules (reviewed in Ref. 39). Enterocytes, like several other cell types, also may be responsive to LPS (40 – 43). The molecular mechanisms of LPS responses in enterocytes remain largely unknown. However, knowledge of these mechanisms is critically important for the design of antiinflammatory therapies for NEC. Previously, we have shown that LPS stimulates the production of COX-2 in the IEC-6 enterocyte cell line and that this response is mediated by p38 MAPK (44). In 0022-1767/05/$02.00
Downloaded from http://www.jimmunol.org/ by guest on May 29, 2013
Necrotizing enterocolitis (NEC), a severe intestinal inflammation in neonates, occurs following bacterial colonization of the gut. LPS-induced production of inflammatory factors in immature enterocytes may be a factor in NEC. Previously, we described LPS-induced p38 MAPK-dependent expression of cyclooxygenase-2 (COX-2) in rat IEC-6 cells. In this study, we examine COX-2 expression in newborn rat intestinal epithelium and further characterize the mechanisms of COX-2 regulation in enterocytes. Induction of NEC by formula feeding/hypoxia increased phospho-p38 and COX-2 levels in the intestinal mucosa. Celecoxib, a selective COX-2 inhibitor, exacerbated the disease, suggesting a protective role for COX-2. COX-2 was induced in the intestinal epithelium by LPS in vivo and ex vivo. The latter response was attenuated by the p38 inhibitor SB202190, but not by inhibitors of ERK, JNK, or NF-B. In IEC-6 enterocytes, COX-2 was induced by the expression of MAPK kinase 3 EE (MKK3EE), a constitutive activator of p38, but not of activators of ERK or JNK pathways. However, neither MKK3/6 nor MKK4, the known p38 upstream kinases, were activated by LPS. Dominant-negative MKK3 or MKK4 or SB202190 failed to prevent LPS-induced, p38-activating phosphorylation, ruling out important roles of these kinases or p38 autophosphorylation. LPS increased COX-2 and activating phosphorylation of p38 with similar dose-response. Blockade of LPS-induced expression of COX-2-luciferase reporter and destabilization of COX-2 message by SB202190 indicate that p38 regulates COX-2 at transcription and mRNA stability levels. Our data indicate that p38-mediated expression of COX-2 proceeds through a novel upstream pathway and support the role of the neonate’s enterocytes as bacterial sensors. The Journal of Immunology, 2006, 176: 580 –588.
The Journal of Immunology this study, we provide evidence for the involvement of COX-2 in a rat model of NEC and further examine the mechanisms of LPSinduced expression of COX-2 in the intestinal epithelium and cultured enterocytes. We demonstrate that NF-B does not mediate LPS-induced expression of COX-2 in enterocytes and that activation of p38 alone is sufficient for this response. Surprisingly, we found that the response to LPS in enterocytes is independent of MAPK kinase (MKK)3/MKK6 or MKK4, the classical activators of p38, and of p38 autophosphorylation, suggesting the existence of a novel inflammatory response pathway upstream of p38. We show that p38 regulates expression of COX-2 in enterocytes at the levels of transcription and mRNA stability.
Materials and Methods Reagents
Animal experiments All experiments were approved by the Animal Research and Care Committee at Children’s Hospital of Pittsburgh. Induction of experimental NEC in newborn rats was described previously (45). Briefly, newborn rats were gavage-fed with rodent formula and subjected to hypoxia three times a day for 10 min at 5% O2, 95% N2. Control animals stayed with their mothers. Celecoxib (10 mg/kg) was administered i.p. once daily. At day 4, mortality was scored, and surviving animals from both treatment and control groups were sacrificed. NEC in the surviving group was graded on 0 –3 scale (0, no inflammation; 2, moderate; and 3, severe inflammation) by a pathologist blinded to the groups. In ex vivo experiments, mucosal scrapings aseptically prepared from the terminal ileum of newborn rats were transiently cultured in DMEM (Invitrogen Life Technologies) containing 5% FCS, 10 mM L-glutamine, and 0.5 U/ml insulin at 37°C, 10% CO2, and 100% humidity.
Cell culture, Western and Northern blots, immunofluorescent microscopy, RT-PCR Cells of the rat intestinal cell line IEC-6 (American Type Culture Collection), passages 16 –30, were grown to 70 –90% confluence. Aliquots of whole-cell lysates containing 50 g of protein were analyzed by Western blotting as suggested by Ab manufacturers. For Northern analysis, poly(A) RNA was extracted using the Dynabeads kit (Dynal Biotech). Northern blots were performed using the NorthernMax kit (Ambion); the probe was 32 P-labeled fragment of rat COX-2 open reading frame (nt 116 – 872) generated by PCR. Paraffin sections of the gut were decorated with primary Ab followed by Texas Red-conjugated secondary Ab and photographed using a DMRA microscope (Leica Microsystems). For RT-PCR analysis, total RNA prepared using TRIzol (Invitrogen Life Technologies) was reversetranscribed with Moloney RNA-dependent DNA polymerase followed by amplification with COX-2 primers (TATAAATGTGACTGTACCCGGAC and GAATATCACACACTCTGTTGTGC) and RPS17 primers (AAAA CATCGGTCTAGGCTTCAG and AGAACTTCTGGAACTGCTTCTTG). The number of amplification cycles was adjusted such that input RNA was limiting for PCR product yield. PCR products were separated on ethidium bromide-agarose gels, and intensities of 756-bp COX-2 bands were compared. For COX-2 mRNA stability measurements, IEC-6 cells pretreated with LPS for 4 h were incubated with or without 5 g/ml ␣-amanitin. Levels of COX-2 transcript at various time points of ␣-amanitin chase were determined by RT -PCR.
DNA constructs and transfections pGL3-COX-2 reporter construct containing the full COX-2 mouse promoter was provided by Dr. D. Dixon (Vanderbilt University, Nashville, TN). pGL3 basic and pGL3 control plasmids were purchased from Promega. pcDNA3-MKK4 AA and pRc/RSV-MKK3 EE were provided by Dr. R. Davis (University of Massachusetts, Worcester, MA). pRev-TREJNK-MKK7 was a gift from Dr. U. Rapp (University of Wuerzburg, Germany). pOPI3-p38 AF was provided by Dr. P. Scherer (Albert Einstein College, New York, NY). pOPRSVI/MCS-RafCAAX was constructed by inserting the Raf XhoI-XbaI fragment from pCMV-RafCAAX (Clontech Laboratories) into XhoI-Xbal-cut pOPRSVI/MCS (Stratagene). pCMVIB␣ AA and a tetracycline-regulated retroviral expression kit were purchased from Clontech Laboratories. pcDNA3Zeo-JNK was a gift from Dr. D. Altschuler (University of Pittsburgh, Pittsburgh, PA); the dominantnegative AF mutant derivative was constructed by site-directed mutagenesis. pOPRSVI/MCS-MKK3 EE was constructed by inserting the HindIIISpeI MKK3 fragment of pRc/RSV-MKK3 EE into XhoI-SpeI-cut pOPRSVI/MCS with XhoI-HindIII adapter. Dominant-negative pRc/RSVMKK3 AA plasmid was constructed by site-directed mutagenesis. Entire coding regions and junctions of each new construct were sequenced to verify the absence of spurious mutations. IEC-6 cells were transfected using Lipofectamine 2000 (Invitrogen Life Technologies) as directed by the manufacturer or coinfected with pRevTRE-MKK7 and pRev-TetON viruses (Clontech Laboratories). Stable transfectant clones were selected with G418, hygromycin B, or zeocin, depending on the plasmid. Luciferase reporter plasmids were cotransfected with pcDNA3, and stable transfectants were selected on G418. Expression of transgenes in stable transfectants was verified by Western blots or luciferase assays.
Luciferase assays Luciferase activity in cell lysates was measured using the luciferase activity kit (Sigma-Aldrich) as directed by the manufacturer.
Results COX-2 is activated in experimental NEC In our rat model of NEC, gavage feeding of newborn rats with formula, combined with hypoxia, results in intestinal inflammation that resembles human NEC. Morphological findings include macroscopic ischemic changes in the terminal ileum and microscopic damage to the intestinal epithelium characterized by neutrophil influx, edema, and epithelial destruction (45). We refer to this condition as NEC hereon. To test whether there is a correlation between NEC and COX-2 expression, we examined COX-2 protein levels in ileal mucosal scrapings from newborn rats in the hypoxia/ formula and control groups. At day 4 after birth, the ileal mucosa of the formula/hypoxia group had significantly higher levels of COX-2 protein, compared with the control group (Fig. 1A). Elevated expression of COX-2 in ileal mucosa was also observed 6 h after gavaging newborn animals with LPS solution (Fig. 1B). Thus, elevated ileal COX-2 was observed both in experimental NEC and after exposure to LPS, a condition that mimics the onset of bacterial colonization of the gut. To understand the role of COX-2 expression in NEC, we examined the effects of the COX-2 inhibitor celecoxib on the incidence, mortality, and severity of this disease. Table I shows that, at a dose of 10 mg/kg applied daily, celecoxib by itself induced mild gut inflammation and caused death in 25 and 11% of the animals, respectively. Combined with the NEC-inducing regimen, celecoxib exacerbated the disease, increasing all pathologic manifestations ⬃2-fold. These data may point to the protective role of COX-2 in NEC. LPS induces expression of COX-2 and activates p38 in ileal mucosa ex vivo Based on our findings in vivo, we hypothesized that enterocytes are capable of producing COX-2 in response to treatment with LPS. In our previous report, we described LPS-induced expression of COX-2 in the IEC-6 cell line (44). We next asked whether this
Downloaded from http://www.jimmunol.org/ by guest on May 29, 2013
The following reagents were used: M2 anti-FLAG Ab, actinomycin D, ␣-amanitin, PD98059, SP600125, SB202190, and Bay11-7082 (Sigma-Aldrich); anti-focal adhesion kinase (FAK), anti-actin, anti-Raf, anti-p65 NF-B Ab, and NF-B binding-site oligos (Santa Cruz Biotechnology); anti-p38, anti-Myc, anti-phospho-p38, anti-IB, anti-activating transcription factor 2 (ATF2), anti-phospho-ATF2, anti-phospho-c-Jun, antiphospho-MKK3/6, and anti-phospho-MKK4 Ab (Cell Signaling Technology); anti-COX-2 Ab (Cayman Chemical); and peroxynitrite and SN50 (Alexis). LPS from Escherichia coli 0127:B8 (Sigma-Aldrich) was either used without additional purification or purified by adsorption on a Dowex AG1 ion exchanger (Bio-Rad) and DNase I digestion followed by proteinase K digestion and phenol extraction. Both preparations were indistinguishable in their ability to activate p38 or induce COX-2 in our experiments.
581
582
LPS-INDUCED COX-2 EXPRESSION IN ENTEROCYTES Table I. Celecoxib increases mortality, incidence, and severity of NECa Groups
% Mortality
% NEC
Mean NEC Grade
BF (n ⫽ 8) BF ⫹ celecoxib (n ⫽ 9) F/H (n ⫽ 22) F/H ⫹ celecoxib (n ⫽ 22)
0 11.0 22.7 40.9b
0 25.0 35.3 84.6c
N/A 0.5 ⫾ 0.1 1.0 ⫾ 0.25 1.75 ⫾ 0.36
a Animals in the breast-fed (BF) and formula/hypoxia (F/H) groups were injected with either celecoxib or 0.9% NaCl daily, and surviving animals were sacrificed on day 4 after birth. NEC was diagnosed by macroscopic observation and histology. b p ⫽ 0.08 by 2. c p ⬍ 0.02 against all groups by Fisher’s exact test.
response could be demonstrated in the intestinal epithelium. Indeed, LPS strongly induced expression of COX-2 protein in ilea1 mucosal scrapings of newborn rats ex vivo following 6-h exposure (Fig. 2A). Although samples from different animals differed somewhat in basal and induced levels of COX-2, all samples demonstrated considerable induction. Similarly, LPS activated p38 following 15-min exposure of mucosal scrapings ex vivo, as judged by increase in activating phosphorylation (Fig. 2B). Activation of p38 was apparent at a concentration of LPS as low as 5 ng/ml and increased in a dose-dependent manner up to 500 ng/ml (Fig. 2C), reaching a plateau at 0.5–10 g/ml (data not shown). The doseresponse of COX-2 protein expression closely followed that of p38 activation (Fig. 2C), suggesting the possibility of a causal relationship between these two processes. Activation of p38 in mucosal scrapings ex vivo peaked at 15 min and returned to baseline in 1 h (see Fig. 5A), which is similar to the time course of activation in IEC-6 cells (44). COX-2 protein levels started increasing after l h and reached a plateau after 4 – 8 h of LPS exposure (Fig. 2D). To examine the requirement for p38 in LPS-induced expression of COX-2 in mucosal scrapings, we used pharmacologic inhibitors of MAPK. In agreement with data in IEC-6 cells (44), SB202190, the specific inhibitor of p38, but not SP600125 or PD98059, inhibitors of JNK and ERK pathways, dramatically attenuated COX-2 expression in mucosal scrapings (Fig. 2E), indicating the requirement of p38 but not other MAPK in enterocytes from newborn rats. Because mucosal scrapings may contain cell types other than enterocytes (e.g., leukocytes, fibroblasts, and smooth muscle cells), it was possible that nonenterocyte cells are responsible for the observed increases in p38 activation and COX-2 expression in mucosal scrapings. However, immunofluorescent microscopy of gut segments treated with LPS ex vivo revealed that p38 activation and COX-2 expression are largely confined to the villus epithelium (Fig. 3A). The same was true for p38 activation and COX-2 expression in the rat model of NEC (Fig. 3B).
NF-B is not required for LPS-induced expression of COX-2 In our previous experiments, LPS-induced expression of COX-2 was not attenuated by the proteasome inhibitor MG132 in IEC-6 cells (44). Because one of the effects of MG132 is inhibition of the transcription factor NF-B, this result tentatively indicated the lack of a role for NF-B. Other investigators have found that NF-B mediates COX-2 induction in several cell types, including macrophages (21), renal tubular epithelium (22), and HeLa cells (23). To clarify this issue, we performed additional experiments in mucosal scrapings and IEC-6 cells. First, we used MG132 as well as Bay11-7082 and SN50, two additional pharmacologic inhibitors of NF-B with different mechanisms of action, to probe LPS-induced expression of COX-2 in mucosal scrapings. Unlike MG132 that blocks proteasome-dependent degradation of the inhibitory IB subunits, Bay11-7082 inhibits the IB kinase IKK, and SN50 peptide prevents nuclear import of NF-B by binding its nuclear localization signal. None of these inhibitors attenuated LPS-induced COX-2 expression (Fig. 4A). MG132 itself strongly induced COX-2. We believe that this induction is the consequence of strong p38 activation by MG132 that is independent of NF-B inhibition (data not shown). To demonstrate that these inhibitors are active in enterocytes, we examined their effect on the activation of NF-B in IEC-6 cells using EMSA with NF-B binding-site DNA probe. This assay demonstrated a dramatic increase in NF-B binding following stimulation with LPS (Fig. 4B, lanes 1 and 4), consistent with previously reported activation of NF-B by LPS in the IEC-6 line (41). Binding was specific, because it was completely abolished by a point mutation in NF-B consensus binding sequence (lane 2) and modified by the addition of antiNF-B Ab (lane 3). MG132 significantly attenuated LPS-induced NF-B activation (lane 5), and Bay11-7082 completely abrogated it, demonstrating that these two inhibitors are effective in IEC-6 cells. To corroborate data obtained with pharmacologic inhibitors, we examined the effect of inhibition of NF-B by a dominantnegative IB mutant. In stable IEC-6 transfectants expressing the IB AA phosphorylation-defective mutant, LPS-induced expression of COX-2 was indistinguishable from vector-transfected cells (Fig. 4C), confirming that NF-B is not required for this response. Activation of p38 is sufficient for the induction of COX-2 To demonstrate that p38 pathway is not only necessary but also sufficient for COX-2 up-regulation in enterocytes, we next assessed levels of COX-2 in IEC-6 stable transfectants conditionally expressing constitutively activated signal transducers that belong to different signaling pathways. RafCAAX, the constitutive activator of the ERK pathway, and MKK3 EE, the constitutive activator of the p38 pathway, were expressed in an isopropyl thiogalactoside-inducible manner using the LacSwitch regulated expression system. JNK-MKK7 fusion, the constitutive activator of the JNK pathway, was expressed in a doxycycline-inducible
Downloaded from http://www.jimmunol.org/ by guest on May 29, 2013
FIGURE 1. A, Expression of intestinal COX-2 in newborn rats following hypoxia/formula feeding. Five newborn rats were either subjected to formula feeding/hypoxia regimen for 4 days (three animals) or left with their mothers for 4 days (two animals). Expression of COX-2 in mucosal scrapings from terminal ileum was analyzed by Western blotting. Blot for actin is shown to demonstrate equal lane load. B, Expression of intestinal COX-2 following introduction of LPS into gut lumen. Ten newborn rats were treated by gavaging of either 200 l of 0.5 mg/ml LPS in distilled water (five animals) or 200 l of distilled water (five animals). Following treatments, animals were kept in a humidity- and temperature-controlled incubator for 6 h. Mucosal scrapings from terminal ileum were analyzed for expression of COX-2 and actin.
The Journal of Immunology
583
FIGURE 3. Activation of p38 and expression of COX-2 in the intestinal epithelium. A, Segments of distal ileum from newborn rats were incubated in DMEM or DMEM plus 0.5 g/ml LPS for the time indicated. B, Newborn rats were subjected to the formula feeding/hypoxia regimen (FFH) or were breast-fed by their mothers (BF) for a number of days indicated. Activation of p38 and expression of COX-2 in paraffin sections of ileum was examined by decorating with anti-phospho-p38 or anti-COX-2 Ab and Texas Red-conjugated secondary Ab. Diamidino phenylindolestained nuclei appear in blue. Day 4 samples are shown at both lower and higher magnification. Note differences in phospho-p38 and COX-2 levels and localization patterns in the intestinal epithelium of BF and FFH animals.
Downloaded from http://www.jimmunol.org/ by guest on May 29, 2013
FIGURE 2. LPS-induced expression of COX-2 and activation of p38 in ileal mucosal scrapings. A, Ileal mucosal scrapings from four newborn rats were incubated in medium with or without 0.5 g/ml LPS for 6 h, and expression of COX-2 was examined by Western blotting. B, Ileal mucosal scrapings from five newborn rats were incubated in medium with or without 0.5 g/ml LPS for 15 min. Expression and phosphorylation of p38 was examined by Western blotting. C, Mucosal scrapings from a newborn rat were divided into eight equal aliquots and incubated for 15 min with indicated concentrations of purified LPS in medium (two top rows). Aliquots of mucosal scrapings from a different newborn rat were incubated at the same concentrations of purified LPS for 6 h (two bottom rows). Expression of p38 and COX-2, as well as phosphorylation of p38, were examined. D, Aliquots of mucosal scrapings from a newborn rat were treated with 0.5 g/ml purified LPS for indicated times, and expression of COX-2 was examined. E, Aliquots of mucosal scrapings from a newborn rat were pretreated with or without 20 M PD98059, 5 M SP600125, or 10 M SB202190 for 20 min and then treated with 0.5 g/ml purified LPS for 6 h. Expression of COX-2 was examined by Western blotting. C–E, Representative blots from two independent experiments. Anti-FAK blots are shown to demonstrate equal lane load.
584
LPS-INDUCED COX-2 EXPRESSION IN ENTEROCYTES
manner using the Rev-TRE retroviral regulated expression system. Regulated (as opposed to constitutive) expression was chosen to avoid generation of suppressor mutations in stable transfectants expressing potentially harmful constitutive activators of mitogenic or stress responses. As shown in Fig. 5, induced expression of MKK3 EE, but not RafCAAX or JNK-MKK7, dramatically increased p38 phosphorylation and COX-2 protein level. Induced expression of JNK-MKK7 resulted in increased phosphorylation of the JNK substrate c-Jun (Fig. 5B), as expected for constitutively active JNK. These data indicate that activation of p38 alone is sufficient for COX-2 up-regulation in IEC-6 enterocytes and that p38 is not redundant with ERK or JNK in this role. LPS-induced activation of p38 does not require MKK3/6, MKK4, or p38 autophosphorylation The fact that constitutive MKK3 activates p38 and induces COX-2 may suggest that MKK3 and its closely related homolog, MKK6, are the physiologic activators of p38 in response to LPS. To test the physiologic role of MKK3/6 as well as of JNK/p38-specific kinase MKK4 as mediators of LPS-induced p38 phosphorylation, we examined activating phosphorylation of these kinases in mucosal scrapings following exposure to LPS. Neither MKK3/6 nor MKK4 were detectably phosphorylated following exposure to 0.5 g/ml LPS for 5 min to 2 h; in the same cells, p38 was strongly phosphorylated at 15 and 30 min (Fig. 6A, left). The lack of LPSinduced MKK3/6 or MKK4 phosphorylation was not due to low sensitivity of their phosphospecific Abs, because UV-induced phosphorylation of all three kinases was quite apparent (Fig. 6A, right). One possible explanation could be that activating phosphorylation of MKK3/6 or MKK4 beyond detection limit may still be sufficient to cause robust activation of p38. To test this possibility, we examined LPS-induced phosphorylation of p38 in stable IEC-6 transfectants expressing dominant-negative mutants of MKK3 or MKK4. None of these transgenes caused detectable attenuation of
FIGURE 5. Constitutive activation of p38 pathway, but not ERK or JNK pathways, stimulates COX-2 expression. A, IEC-6 cells were stably cotransfected with pCMVLacI and one of the following: pOPRSVI/MCS vector, pOPRSVI/MCS-RafCAAX, or pOPRSVI/MCS-MKK3 EE. Transfectants were incubated for 12 h in DMEM or DMEM plus 2 mM isopropyl thiogalactoside. Expression of the transgenes, COX-2, phospho-p38, p38, and FAK was examined by Western blotting. B, IEC-6 cells were coinfected with pRevTetON and either pREV-TRE vector or pRev-TRE-JNK-MKK7 retroviruses. Cells were incubated with or without 10 g/ml doxycycline for 12 h. Expression of COX-2, Myc-tagged JNK-MKK7, as well as activating phosphorylation of c-Jun and p38, were examined by Western blotting. FAK blot shows equal lane load. Data are representative of at least two independent experiments.
Downloaded from http://www.jimmunol.org/ by guest on May 29, 2013
FIGURE 4. NF-B does not mediate LPS-induced expression of COX-2 in enterocytes. A, Aliquots of mucosal scrapings from a newborn rat were pretreated with DMSO (control), 20 M Bay 11-7082, 5 M MG132, or 100 g/ml SN50 for 20 min, and then treated with 0.5 g/ml LPS for 6 h as indicated, and expression of COX-2 was examined by Western blot. B, Nuclear extracts from IEC-6 cells that were pretreated with or without 5 M MG132 or 20 M Bay11-7082 for 20 min and then treated with 0.5 g/ml LPS for 10 min were incubated with 32P-labeled wild-type or mutant NF-B binding-site oligonucleotides and anti-p65 NF-B Ab, as indicated. Binding reaction products were analyzed by nondenaturing electrophoresis and autoradiography. Positions of NF-B-DNA complexes are indicated. C, IEC-6 cells stably transfected with pcDNA3 vector or pCMV-IB␣ AA were treated with 0.5 g/ml LPS for 6 h. Expression of COX-2 and IB was examined by Western blotting. FAK probing in A and B demonstrates equal lane load. Data are representative of two independent experiments.
The Journal of Immunology
585
LPS-induced p38 phosphorylation (Fig. 6B), confirming the lack of role for MKK3 or MKK4 in this response. An MKK-independent mechanism of p38 activation was recently described that involves autophosphorylation of p38 following association with the adapter protein TAB1 (46). Unlike MKKdependent phosphorylation, this mechanism depends on the kinase activity of p38. To investigate the possible role of p38 autophosphorylation in LPS-induced activation of p38, we examined the effect of p38 pharmacological inhibition. SB202190 inhibits the activity of p38 and thus its autophosphorylation, but it does not affect the phosphorylation of p38 by other kinases. Pretreatment of newborn rat mucosal scrapings with 20 M SB202190 for 30 min did not cause significant attenuation of LPS-induced p38 phosphorylation. SB202190 also had no effect or had only a weak effect on peroxynitrite- and TNF-␣-induced p38 phosphorylation (Fig. 6C), indicating that p38 autophosphorylation does not play a significant role in these responses in the intestinal epithelium. The unique character of the pathway of LPS-induced p38 activation is demonstrated by its specific inhibition with a novel antiinflammatory drug CNI-1493. This drug completely abrogates LPS-induced activation of p38 while having no effect on activation by peroxynitrite and TNF-␣ (Fig. 6D), as well as other cytokines and stresses (data not shown). p38 mediates LPS-induced ATF2 activation p38 is known to activate multiple downstream pathways by phosphorylating transcription factors, cytosolic kinases, and other effector proteins. One of the known targets is the transcription factor ATF2, a subunit of the AP-1 complex that is activated upon p38dependent phosphorylation at the Thr69/71 site (47). AP-1 has been reported to activate transcription from the COX-2 promoter (17). To test whether ATF2 is involved in response to LPS, we examined LPS-induced activating phosphorylation of this protein. In LPS-treated IEC-6 cells, ATF2 phosphorylation closely followed that of p38 activation (Fig. 7A). Because ATF2 can be phosphorylated by both p38 and JNK, and JNK is responsive to LPS in IEC-6 cells (44), it was possible that JNK rather than p38 phos-
FIGURE 7. LPS-induced activation of ATF2 is mediated by p38. A, Time course of ATF2 activation following treatment of IEC-6 cells with 0.5 g/ml LPS. B, Aliquots of mucosal scrapings from a newborn rat were pretreated with or without 10 M SB202190 or 5 M SP600125 for 20 min and then treated with 0.5 g/ml purified LPS for 15 min. Phosphorylation of ATF2 was examined by Western blotting. Actin blot shows equal lane load. C, IEC-6 cells were stably transfected with pcDNA3, pOPRSVIp38 AF, or pcDNA3Zeo-JNK AF. Transfectants were treated with or without 0.5 g/ml LPS for 15 min. Expression of HA-tagged JNK, FLAGtagged p38, and phosphorylation of ATF2 and c-Jun were examined by Western blotting. FAK blot shows equal lane load.
Downloaded from http://www.jimmunol.org/ by guest on May 29, 2013
FIGURE 6. LPS-induced activation of p38 does not involve MKK3/6, MKK4, or p38 autophosphorylation. A, Aliquots of mucosa1 scrapings from a newborn rat were incubated with 0.5 g/ml purified LPS for indicated times. Aliquots of mucosal scrapings from another animal were exposed to 50 mJ/cm2 UV, returned to growth medium, and incubated for indicated times. Activating phosphorylation of MKK3/6, MKK4, and p38 was examined by Western blotting. p38 blot shows equal lane load. B, IEC-6 cells stably transfected with pcDNA3 vector, pRC-RSV-MKK3 AA, or pcDNA3-MKK4 AA were treated with 0.5 g/ml LPS for 15 min. Phosphorylation of p38 and expression of FLAG-tagged MKK3 and MKK4 transgenes were examined by Western blotting. p38 blot shows equal lane load. C, Aliquots of mucosal scrapings from a newborn rat were pretreated for 30 min with or without 20 M SB202190 and then treated with 0.5 g/ml LPS, 50 M peroxynitrite, or 5 ng/ml TNF-␣ for 15 min. D, IEC-6 cells were pretreated with vehicle alone or 10 M CNI-1493 for 20 min and then treated with 0.5 g/ml LPS, 50 M peroxynitrite, or 5 ng/ml TNF-␣. Expression and phosphorylation of p38 in C and D were examined by Western blotting. All data are representative of at least two independent experiments.
586
LPS-INDUCED COX-2 EXPRESSION IN ENTEROCYTES
phorylates ATF2. To examine the roles of p38 and JNK in ATF2 phosphorylation, we used pharmacologic and genetic approaches. In newborn rat mucosal scrapings, LPS-induced ATF2 phosphorylation was significantly attenuated by the p38 inhibitor, but not by the JNK inhibitor (Fig. 7B), indicating a major role of p38. In IEC-6 cells, stable expression of dominant-negative p38, but not dominant-negative JNK, abrogated the LPS-induced ATF2 phosphorylation (Fig. 7C). These data support the major role for p38, but not JNK, in LPS-induced activation of ATF2 in enterocytes. p38 activates COX-2 promoter and increases COX-2 mRNA stability We next investigated whether p38 regulates transcription of COX-2 mRNA in enterocytes. In IEC-6 cells, transcription inhibitor actinomycin D significantly attenuated induction of COX-2 protein by LPS, hydrogen peroxide, and osmotic stress (Fig. 8A), indicating that these responses depend, at least partially, on new RNA synthesis. In IEC-6 cells, levels of COX-2 transcript increased following treatment with LPS, hydrogen peroxide, or 1 M Table II. LPS activates COX-2-luciferase transcriptional reportera
Transgene
Treatment
Luciferase Activity (cpm/mg protein)
None pGL3 basic pGL3 basic pGL3-COX-2 pGL3-COX-2 pGL3-COX-2 pGL3 control pGL3 control
None None LPS None LPS LPS ⫹ SB202190 None LPS
26 ⫾ 10 85 ⫾ 16 90 ⫾ 16 640 ⫾ 60 3,300 ⫾ 400 380 ⫾ 50 350,000 ⫾ 20,000 330,000 ⫾ 30,000
a IEC-6 cells were stably transfected with the indicated luciferase reporter constructs. Transfectants were treated with 0.5 g/ml LPS and 20 M SB202190 for 6 h as indicated. Luciferase activity (counts per milligram of protein) was measured by scintillation counting.
glycerol, and transcript accumulation was attenuated by pretreatment with p38 inhibitor (Fig. 8B, left). The increase in COX-2 mRNA was also observed in newborn rat mucosal scrapings treated with LPS ex vivo (Fig. 8B, right). To investigate activation of the COX-2 promoter by LPS directly, we used a COX-2 transcriptional reporter, the full-size COX-2 promoter fused to the firefly luciferase cDNA (COX-2-luc). Treatment with LPS increased luciferase activity in a COX-2-luc transfectant ⬃5-fold (Table II), demonstrating transcriptional induction of COX-2 by LPS. SB202190 abrogated this response, indicating its dependence on p38. LPS did not appreciably increase luciferase activity in positive and negative control transfectants, indicating specific COX-2 promoter activation by this agent. To examine the effect of p38 on COX-2 mRNA stability, we measured the COX-2 mRNA decay rate in the presence or absence of SB202190. This inhibitor dramatically destabilized COX-2 transcripts, reducing their half-life to ⬃1 h from that of 8 h in the untreated control (Fig. 8C), indicating that p38 activity is required to stabilize the COX-2 message. These findings show that both mechanisms of p38-dependent COX-2 mRNA regulation, transcription initiation and mRNA degradation, are active in enterocytes.
Discussion In this report, we describe elevated expression of COX-2 in the intestine of newborn rats that develop experimental NEC. Moreover, we show that COX-2 is up-regulated in the intestinal epithelium following introduction of LPS into the gut lumen of newborn rats, in the ileal mucosa following treatment with LPS ex vivo, and in LPS-treated IEC-6 enterocytes. These findings support the previously proposed idea that expression of COX-2 in the neonatal intestine may play a role in the development of NEC (14). The fact that celecoxib exacerbated the pathology of NEC in our model suggests a protective role for COX-2. However, it would be premature to draw definitive conclusions based on experiments with a single dose of one COX-2 inhibitor. Other doses and/or other
Downloaded from http://www.jimmunol.org/ by guest on May 29, 2013
FIGURE 8. p38 regulates levels of COX-2 transcript and COX-2 mRNA stability in enterocytes. A, IEC-6 cells were pretreated with 5 M actinomycin D in the medium for 30 min and then treated with or without 0.5 g/ml LPS, 200 M H2O2, or 1 M glycerol for 1 h. Cells treated with glycerol were transferred to regular medium with actinomycin D followed by incubation of all samples for additional 12 h. Expression of COX-2 was examined by Western blotting. FAK blot shows equal lane load. Data are representative of three independent experiments. B, Left, IEC-6 cells were pretreated with or without 10 M SB202190 for 20 min and then treated with 0.5 g/ml LPS, 200 M H2O2, or 1 M glycerol for 4 h. Levels of COX-2 mRNA were examined by Northern blotting. B, Right, Aliquots of mucosal scrapings from a newborn rat were treated with or without 0.5 g/ml LPS for 4 h, and expression of COX-2 mRNA was examined by Northern blotting. GAPDH blot demonstrates equal lane load. C, IEC-6 cells were pretreated with 0.5 g/ml LPS for 4 h followed by the addition of 5 g/ml ␣-amanitin with or without 10 M SB202190. Levels of COX-2 and ribosomal protein S17 control mRNAs at indicated time points after the addition of ␣-amanitin were examined by RT-PCR.
The Journal of Immunology
This study identifies p38 as a signaling mediator of COX-2 induction by LPS in enterocytes. Other pathogen-activated signaling molecules, including ERK, JNK, and NF-B, may regulate the production of other pro- and anti-inflammatory factors in enterocytes, contributing to the overall inflammatory response. Systematic dissection of these pathways will lead to better understanding of molecular mechanisms of NEC and identification of molecular targets for therapeutic intervention.
Acknowledgments We thank Drs. Phil Scherer, Roger Davis, Ulf Rapp, Daniel Altschuler, and Dan Dixon for gifts of plasmids and Carmel Portugal for assistance with Western blotting.
Disclosures The authors have no financial conflict of interest.
References 1. Noerr, B. 2003. Current controversies in the understanding of necrotizing enterocolitis. Adv. Neonatal Care 3: 107–120. 2. Ford, H. R., D. L. Sorrells, and A. S. Knisely. 1996. Inflammatory cytokines, nitric oxide, and necrotizing enterocolitis. Semin. Pediatr. Surg. 5: 155–159. 3. Guthrie, S. O., P. V. Gordon, V. Thomas, J. A. Thorp, J. Peabody, and R. H. Clark. 2003. Necrotizing enterocolitis among neonates in the United States. J. Perinatol. 23: 278 –285. 4. Ohno, Y., J. Lee, R. D. Fusunyan, R. P. MacDermott, and I. R. Sanderson. 1997. Macrophage inflammatory protein-2: chromosomal regulation in rat small intestinal epithelial cells. Proc. Natl. Acad. Sci. USA 94: 10279 –10284. 5. Longo, W. E., L. J. Damore, J. E. Mazuski, G. S. Smith, N. Panesar, and D. L. Kaminski. 1998. The role of cyclooxygenase-l and cyclooxygenase-2 in lipopolysaccharide and interleukin-l-stimulated enterocyte prostanoid formation. Mediat. Inflamm. 7: 85–91. 6. Cario, E., G. Gerken, and D. K. Podolsky. 2002. “For whom the bell tolls!”: innate defense mechanisms and survival strategies of the intestinal epithelium against lumenal pathogens. Z. Gastroenterol. 40: 983–990. 7. Longo, W. E., N. Panesar, J. Mazuski, and D. L. Kaminski. 1998. Contribution of cyclooxygenase-l and cyclooxygenase-2 to prostanoid formation by human enterocytes stimulated by calcium ionophore and inflammatory agents. Prostaglandins Other Lipid Mediat. 56: 325–339. 8. Whittle, B. J. 2003. Gastrointestinal effects of nonsteroidal anti-inflammatory drugs. Fundam. Clin. Pharmacol. 17: 301–313. 9. Wallace, J. L. 2001. Prostaglandin biology in inflammatory bowel disease. Gastroenterol. Clin. North Am. 30: 971–980. 10. Subbaramaiah, K., K. Yoshimatsu, E. Scherl, K. M. Das, K. D. Glazier, D. Golijanin, R. A. Soslow, T. Tanabe, H. Naraba, and A. J. Dannenberg. 2004. Microsomal prostaglandin E synthase-l is overexpressed in inflammatory bowel disease: evidence for involvement of the transcription factor Egr-l. J. Biol. Chem. 279: 12647–12658. 11. Kankuri, E., K. Vaali, R. Korpela, I. Paakkari, H. Vapaatalo, and E. Moilanen. 2001. Effects of a cyclooxygenase-2 preferential agent nimesulide on TNBSinduced acute inflammation in the gut. Inflammation 25: 301–310. 12. Loftin, C. D., H. F. Tiano, and R. Langenbach. 2002. Phenotypes of the cyclooxygenase-deficient mice indicate physiological and pathophysiological roles for cyclooxygenase-l and cyclooxygenase-2. Prostaglandins Other Lipid Mediat. 68 – 69: 177–185. 13. Sigthorsson, G., R. J. Simpson, M. Walley, A. Anthony, R. Foster, C. Hotz-Behoftsitz, A. Palizban, J. Pombo, J. Watts, S. G. Morham, and I. Bjarnason. 2002. cyclooxygenase-1 and 2, intestinal integrity, and pathogenesis of nonsteroidal anti-inflammatory drug enteropathy in mice. Gastroenterology 122: 1913–1923. 14. Chung, D. H., R. T. Ethridge, S. Kim, S. Owens-Stowall, A. Hernandez, D. R. Kelly, and B. M. Evers. 2001. Molecular mechanisms contributing to necrotizing enterocolitis. Ann. Surg. 233: 835– 842. 15. Tanabe, T., and N. Tohnai. 2002. Cox isozymes and their gene structures and expression. Prostaglandins Other Lipid Mediat. 68 – 69: 95–114. 16. Ramsay, R. G., D. Ciznadija, M. Vanevski, and T. Mantamadiotis. 2003. Transcriptional regulation of cyclooxygenase expression: three pillars of control. Int. J. Immunopathol. Pharmacol. 16: 59 – 67. 17. Guo, Y. S., M. R. Hellmich, X. D. Wen, and C. M. Townsend, Jr. 2001. Activator protein-1 transcription factor mediates bombesin-stimulated cyclooxygenase-2 expression in intestinal epithelial cells. J. Biol. Chem. 276: 22941–22947. 18. Tessner, T. G., F. Muhale, S. Schloemann, S. M. Cohn, A. Morrison, and W. F. Stenson. 2002. Basic fibroblast growth factor up-regulates cyclooxygenase-2 in 1407 cells through p38 MAP kinase. Am. J. Physiol. 284:G269 –G279. 19. Arbabi, S., M. R. Rosengart, I. Garcia, S. Jelacic, and R. V. Maier. 2001. Epithelial cyclooxygenase-2 expression: a model for pathogenesis of colon cancer. J. Surg. Res. 97: 60 – 64. 20. Lee, J. Y., J. Ye, Z. Gao, H. S. Youn, W. H. Lee, L. Zhao, N. Sizemore, and D. H. Hwang. 2003. Reciprocal modulation of Toll-like receptor-4 signaling pathways involving MyD88 and phosphatydylinositol 3-kinase/AKT by saturated and polyunsaturated fatty acids. J. Biol. Chem. 278: 37041–37051.
Downloaded from http://www.jimmunol.org/ by guest on May 29, 2013
COX-2 inhibitors may prove beneficial. Because COX-2 has roles in intestinal homeostasis and inflammation, its effects may be double-edged. Abnormally low and abnormally high levels of this enzyme may both contribute to the pathogenesis of NEC. Elevated intestinal COX-2 in NEC is more likely a result of systemic inflammation than a direct effect of formula feeding/hypoxia. Indeed, levels of COX-2 do not increase after 1 day of formula feeding/hypoxia, and no inflammation is observed at this time (R. Zamora, unpublished data). Induction of COX-2 by gavaging LPS in vivo, as well as by LPS treatment of mucosal scrapings ex vivo and IEC-6 cells in vitro, argue, however, that enterocytes are capable of LPS-induced up-regulation of COX-2 in the absence of systemic inflammation. The concept of enterocytes possessing the attributes of the sensory cells of the innate immune system, such as LPS responsiveness, has been proposed recently (4 –7, 42, 43, 48 –50); our results lend it further support. Sensing bacterial components by enterocytes is required for the maintenance of the adult intestine (51) and may be critically important in the newborn during the establishment of the intestinal microflora. One of the goals of this study was to identify signaling pathways that mediate up-regulation of COX-2 in the neonatal intestine. The following observations point to a key role of the p38 MAPK. First, the dose-response of LPS-induced COX-2 expression is similar to that of p38 activation. Second, COX-2 induction is abrogated by the specific inhibitor of p38 but not by inhibitors of ERK, JNK, or NF-B pathways. Third, constitutively active MKK3, the selective activator of p38, but not constitutive Raf or JNK, causes COX-2 up-regulation in the absence of external stimuli. Our results agree with the known role of p38 in COX-2 expression in enterocyte cell lines (18, 19), macrophages (21, 24), monocytes (46), and other cell types (18, 23, 32, 38, 52). However, unlike several other studies, this work does not support a positive role for NF-B or MAPK other than p38 in COX-2 expression in enterocytes. We demonstrate, for the first time, that p38 activation is not only necessary but also sufficient for COX-2 expression. Another novel finding is that LPS may activate p38 via a noncanonical upstream pathway that does not involve MKK3/6, MKK4, or p38 autophosphorylation. In addition to identifying a key role for p38, our study addresses the mechanisms of p38-mediated regulation of COX-2 in enterocytes. Because treatment with LPS increases COX-2 transcript levels, and the p38 specific inhibitor SB202190 blocks this increase, the regulation occurs, at least partially, at the level of mRNA. According to the published data, COX-2 can be regulated in various cell types at the level of transcription (17, 21, 23–25, 30), transcript stability (18, 29, 31–33, 37, 52), or both (22, 34, 53). We have found that inhibition of transcription dramatically decreases LPS-induced COX-2 protein accumulation, indicating that induced expression of COX-2 depends at least partially on new RNA synthesis. Furthermore, experiments with the COX-2-luciferase transcriptional reporter confirmed that up-regulation of COX-2 by LPS involves the activation of transcription from the COX-2 promoter. Taken together, these observations prove that p38 positively regulates transcription of the COX-2 gene in enterocytes, presumably by phosphorylating and activating transcription factors that bind to the COX-2 promoter. One such transcription factor could be ATF2 whose LPS-induced activating phosphorylation closely followed activation of p38 in our experiments. Specific p38 inhibitor dramatically reduced the half-life of COX-2 mRNA, indicating that p38 also regulates expression of COX-2 in enterocytes by regulating stability of COX-2 transcripts. Thus, our findings support p38-dependent regulation of COX-2 mRNA in IEC-6 cells at both the transcriptional and mRNA stability levels.
587
588
37. Guan, Z., S. Y. Buckman, A. P. Pentland, D. J. Templeton, and A. R. Morrison. 1998. Induction of cyclooxygenase-2 by the activated MEKK13 SEK1/ MKK43p38 mitogen-activated protein kinase pathway. J. Biol. Chem. 273: 12901–12908. 38. Hendrickx, N., C. Volanti, U. Moens, O. M. Seternes, P. de Witte, J. R. Vandenheede, J. Piette, and P. Agostinis. 2003. Up-regulation of cyclooxygenase-2 and apoptosis resistance by p38 MAPK in hypericin-mediated photodynamic therapy of human cancer cells. J. Biol. Chem. 278: 52231–52239. 39. Beutler, B., K. Hoebe, X. Du, and R. J. Ulevitch. 2003. How we detect microbes and respond to them: the Toll-like receptors and their transducers. J. Leukocyte Biol. 74: 479 – 485. 40. Fusunyan, R. D., N. N. Nanthakumar, M. E. Baldeon, and W. A. Walker. 2001. Evidence for an innate immune response in the immature human intestine: Tolllike receptors on fetal enterocytes. Pediatr. Res. 49: 589 –593. 41. De Plaen, I. G., X. W. Qu, H. Wang, X. D. Tan, L. Wang, X. B. Han, R. A. Rozenfeld, and W. Hsueh. 2002. Endotoxin, but not platelet-activating factor, activates nuclear factor-B and increases IB␣ and IB turnover in enterocytes. Immunology 106: 577–583. 42. Ruemelle, F. M., J. F. Beaulieu, S. Dionne, E. Levy, E. G. Seidman, N. Cerf-Bensussan, and M. J. Lentze. 2002. Lipopolysaccharide modulation of normal enterocyte turnover by Toll-like receptors is mediated by endogenously produced tumor necrosis factor ␣. Gut 51: 842– 848. 43. Nanthakumar, N. N., R. D. Fusunyan, I. Sanderson, and W. A. Walker. 2000. Inflammation in the developing human intestine: a possible pathophysiologic contribution to necrotizing enterocolitis. Proc. Natl. Acad. Sci. USA 97: 6043– 6048. 44. Grishin, A., J. Wang, D. Hackam, F. Qureshi, J. Upperman, R. Zamora, and H. R. Ford. 2004. p38 MAP kinase mediates endotoxin-induced expression of cyclooxygenase-2 in enterocytes. Surgery 136: 329 –335. 45. Nadler, E. P., E. Dickinson, A. Knisely, X. R. Zhang, P. Boyle, D. Beer-Stolz, S. C. Watkins, and H. R. Ford. 2000. Expression of inducible nitric oxide synthase and interleukin-12 in experimental necrotizing enterocolitis. J. Surg. Res. 92: 71–77. 46. Ge, B., H. Gram, F. Di Padova, B. Huang, R. J. Ulevitch, Y. Luo, and J. Han. 2002. MAPKK-independent activation of p38␣ mediated by TAB1-dependent autophosphorylation of p38␣. Science 295: 1291–1294. 47. Livingstone, C., G. Patel, and N. Jones. 1995. ATF-2 contains a phosphorylationdependent transcriptional activation domain. EMBO J. 14: 1785–1797. 48. Cetin, S., H. R. Ford, L. R. Sysko, C. Agrawal, J. Wang, M. D. Neal, C. Baty, G. Apodaca, and D. J. Hackam. 2004. Endotoxin inhibits intestinal epithelial restitution through activation of Rho-GTPase and increased focal adhesions. J. Biol. Chem. 279: 24592–24600. 49. Otte, J. M., E. Cario, and D. K. Podolsky. 2004. Mechanisms of cross hyporesponsiveness to Toll-like receptor bacterial ligands in intestinal epithelial cells. Gastroenterology 126: 1054 –1070. 50. Meyer, T. A., Y. Noguchi, C. Ogle, G. Tiao, J. J. Wang, J. E. Fischer, and P. O. Hasselgren. 1994. Endotoxin stimulates interleukin-6 production in intestinal epithelial cells: a synergistic effect with prostaglandin E2. Arch. Surg. 129: 1290 –1294. 51. Rakoff-Nahum, S., J. Paglino, F. Eslami-Varzaneh, S. Edberg, and R. Medzhitov. 2004. Recognition of commensal microflora by Toll-like receptors is required for intestinal homeostasis. Cell 118: 229 –241. 52. Faour, W. H., A. Mancini, Q. W. He, and J. A. Di Battista. 2003. T cell-derived interleukin-17 regulates the level and stability of cyclooxygenase-2 mRNA through restricted activation of the p38 mitogen-activated protein kinase cascade: role of distal sequences in the 3⬘-untranslated region of cyclooxygenase-2 mRNA. J. Biol. Chem. 278: 26897–26907. 53. Dean, J. L., M. Brook, A. R. Clark, and J. Saklatvala. 1999. p38 mitogen-activated protein kinase regulates cyclooxygenase-2 mRNA stability and transcription in lipopolysaccharide-treated human monocytes. J. Biol. Chem. 274: 264 –269.
Downloaded from http://www.jimmunol.org/ by guest on May 29, 2013
21. Yeo, S. J., D. Gravis, J. G. Yoon, and A. K. Yi. 2003. Myeloid differentiation factor 88-dependent transcriptional regulation of cyclooxygenase-2 expression by CpG DNA: role of NF-B and p38. J. Biol. Chem. 278: 22563–22573. 22. Cheng, H. F., and R. C. Harris. 2002. cyclooxygenase-2 expression in cultured cortical thick ascending limb of Henle increases in response to decreased extracellular ionic content by both transcriptional and posttranscriptional mechanisms: role of p38-mediated pathways. J. Biol. Chem. 277: 45638 – 45643. 23. Deva, R., P. Shankaranarayanan, R. Ciccoli, and S. Nigam. 2003. Candida albicans induces selectively transcriptional activation of cyclooxygenase-2 in HeLa cells: pivotal roles of Toll-like receptors, p38 mitogen-activated protein kinase, and NF-B. J. Immunol. 171: 3047–3055. 24. Caivano, M., B. Gorgoni, P. Cohen, and V. Poli. 2001. The induction of cyclooxygenase-2 mRNA in macrophages is biphasic and requires both CCAAT enhancer-binding protein  (C/EBP-) and C/EBP-␦ transcription factors. J. Biol. Chem. 276: 48693– 48701. 25. Deng, W. G., Y. Zhu, and K. K. Wu. 2003. Role of p300 and PCAF in regulating cyclooxygenase-2 promoter activation by inflammatory mediators. Blood 103: 2135–2142. 26. Huh, Y. H., S. H. Kim, S. J. Kim, and J. S. Chun. 2003. Differentiation statusdependent regulation of cyclooxygenase-2 expression and prostaglandin E2 production by epidermal growth factor via mitogen-activated protein kinase in articular chondrocytes. J. Biol. Chem. 278: 9691–9697. 27. Yang, T., J. M. Park, L. Arend, Y. Huang, R. Topaloglu, A. Pasumarthy, H. Praetorius, K. Spring, J. P. Briggs, and J. Schnermann. 2000. Low chloride stimulation of prostaglandin E2 release and cyclooxygenase-2 expression in a mouse macula densa cell line. J. Biol. Chem. 275: 37922–37929. 28. Yang, T., Y. Huang, L. E. Heasley, T. Berl, J. B. Schnermann, and J. P. Briggs. 2000. MAPK mediation of hypertonicity-stimulated cyclooxygenase-2 expression in renal medullary collecting duct cells. J. Biol. Chem. 275: 23281–23286. 29. Xu, K., A. M. Robida, and T. J. Murphy. 2000. Immediate-early MEK-1-dependent stabilization of rat smooth muscle cell cyclooxygenase-2 mRNA by G␣(q)coupled receptor signaling. J. Biol. Chem. 275: 23012–23019. 30. Subbaramaiah, K., J. C. Hart, L. Noron, and A. J. Dannenberg. 2000. Microtubule-interfering agents stimulate the transcription of cyclooxygenase-2: evidence for involvement of ERK1/ERK2 and p38 mitogen-activated protein kinase pathways. J. Biol. Chem. 275: 14838 –14845. 31. Lasa, M., S. M. Abraham, C. Boucheron, J. Saklatvala, and A. R. Clark. 2002. Dexamethasone causes sustained expression of MAPK phosphatase 1 and phosphatase-mediated inhibition of MAPK p38. Mol. Cell. Biol. 22: 7802–7811. 32. Kim, S. J., S. G. Hwang, I. C. Kim, and J. S. Chun. 2003. Actin cytoskeletal architecture regulates nitric oxide-induced apoptosis, dedifferentiation, and cyclooxygenase-2 expression in articular chondrocytes via mitogen-activated protein kinase and protein kinase C pathways. J. Biol. Chem. 278: 42448 – 42456. 33. Jang, B. C., T. Sanchez, H. J. Schaefers, O. C. Trifan, C. H. Liu, C. Cremion, C. K. Huang, and T. Hla. 2000. Serum withdrawal-induced posttranscriptional stabilization of cyclooxygenase-2 mRNA in MDA-MB-231 mammary carcinoma cells requires the activity of the p38 stress-activated protein kinase. J. Biol. Chem. 275: 39507–39515. 34. Matsuura, H., M. Sakaue, K. Subbaramaiah, H. Kamitani, T. E. Eling, A. J. Dannenberg, T. Tanabe, H. Inoue, J. Arata, and A. M. Jetten. 1999. Regulation of cyclooxygenase-2 by interferon ␥ and transforming growth factor ␣ in normal human epidermal keratinocytes and squamous carcinoma cells: role of mitogen-activated protein kinases. J. Biol. Chem. 247: 29138 –29148. 35. Barry, O. P., M. G. Kazanietz, D. Pratico, and G. A. FitzGerald. 1999. Arachidonic acid in platelet microparticles up-regulates cyclooxygenase-2-dependent prostaglandin formation via a protein kinase C/mitogen-activated protein kinasedependent pathway. J. Biol. Chem. 274: 7545–7556. 36. Brewer, J. A., B. Khor, S. K. Vogt, L. M. Muglia, H. Fujiwara, K. E. Haegele, B. P. Sleckman, and L. J. Muglia. 2003. T cell glucocorticoid receptor is required to suppress cyclooxygenase-2-mediated lethal immune activation. Nat. Med. 9: 1318 –1322.
LPS-INDUCED COX-2 EXPRESSION IN ENTEROCYTES
