Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2021 Sep 1.
Published in final edited form as: Shock. 2020 Sep;54(3):394–401. doi: 10.1097/SHK.0000000000001444

EP2 receptor blockade attenuates COX-2 upregulation during intestinal inflammation

Jamie Golden 1,2, Laura Illingworth 1, Patil Kavarian 1, Oswaldo Escobar 1, Patrick Delaplain 1,2, Mubina Isani 1,2, Jin Wang 1, Joanna Lim 1,2, Jordan Bowling 1,2, Brandon Bell 1, Christopher P Gayer 1,2, Anatoly Grishin 1,2, Henri R Ford 1,2,3
PMCID: PMC7051888  NIHMSID: NIHMS1538385  PMID: 31490357

Abstract

High levels of PGE2 have been implicated in the pathogenesis of intestinal inflammatory disorders such as necrotizing enterocolitis (NEC) and peritonitis. However, PGE2 has a paradoxical effect: its low levels promote intestinal homeostasis, whereas high levels may contribute to pathology. These concentration-dependent effects are mediated by four receptors, EP1-EP4. In this study, we evaluate the effect of blockade of the low affinity pro-inflammatory receptors EP1 and EP2 on expression of COX-2, the rate limiting enzyme in PGE2 biosynthesis, and on gut barrier permeability using cultured enterocytes and three different models of intestinal injury. PGE2 upregulated COX-2 in IEC-6 enterocytes, and this response was blocked by the EP2 antagonist PF-04418948, but not by the EP1 antagonist ONO-8711 or EP4 antagonist E7046. In the neonatal rat model of NEC, EP2 antagonist and low dose of COX-2 inhibitor Celecoxib, but not EP1 antagonist, reduced NEC pathology as well as COX-2 mRNA and protein expression. In the adult mouse endotoxemia and cecal ligation/puncture models, EP2, but not EP1 genetic deficiency decreased COX-2 expression in the intestine. Our results indicate that the EP2 receptor plays a critical role in the positive feedback regulation of intestinal COX-2 by its end-product PGE2 during inflammation and may be a novel therapeutic target in the treatment of NEC.

Keywords: EP2 receptor, COX-2, intestinal inflammation, rats, mice

Introduction

The intestine contains high loads of bacteria that are normally symbiotic with the host (1). Under homeostatic conditions, the intestinal epithelium serves as a selective barrier that prevents bacteria and toxins from entering the systemic circulation. However, in the settings of critical illnesses, gut barrier dysfunction allows for the translocation of bacteria and toxins, activation of the mucosal immune system, and production of high levels of inflammatory factors that further exacerbate epithelial injury (2). Gut-origin sepsis can result from many disorders, including inflammatory bowel disease, peritonitis, and necrotizing enterocolitis (NEC). Thus, alleviating intestinal barrier breakdown and the subsequent dysregulated host response to infection are key therapeutic strategies to decrease morbidity and mortality secondary to inflammatory disorders.

Elucidating the inflammatory mechanism that leads to intestinal barrier dysfunction is critical to our understanding of gut-origin sepsis and to the development of novel therapies. High levels of cyclooxygenase-2 (COX-2) and its product, prostaglandin E2 (PGE2), have been implicated in the inflammatory cascade leading to intestinal barrier failure in NEC (3), inflammatory bowel disease (4), peritonitis (5), systemic sepsis and shock (6). PGE2 is the major prostanoid in the intestine, and COX-2 is the rate-limiting enzyme in its production. COX-2 and PGE2 have paradoxical effects in the intestine. Low levels of PGE2 play a key role in intestinal homeostasis, while high levels of PGE2 due to sustained upregulation of COX-2 during inflammation, lead to intestinal barrier breakdown. COX-2 may be further upregulated by its products, resulting in high levels of inflammatory prostanoids and exacerbation of inflammatory response. This positive feedback regulation has been described in macrophages (7) , colon cancer cells (8), fibroblasts (9), and prostate cancer cells (10).

PGE2 acts through four G protein-coupled 7-transmembrane receptors, EP1-EP4. EP1 and EP2 are low affinity receptors, whereas EP3 and EP4 are high-affinity receptors (11). The variety of EP receptors may explain the concentration-dependent effects of PGE2 in the intestine. Activation of the high affinity receptors strengthens the intestinal barrier, prevents bacterial invasion, and inhibits upregulation of inflammatory mediators (12). These protective effects are functionally related to the stimulation of mucus and fluid secretion and the inhibition of intestinal hypermotility (13). Therefore, it is not surprising that global inhibition of COX-2 by non-steroid anti-inflammatory drugs or gene knockouts has been found to exacerbate intestinal barrier dysfunction and increase mortality during experimental sepsis (14, 15).

The low-affinity receptors EP1 and EP2 are activated by high levels of PGE2 such as are seen during inflammation. EP1 plays a role in intestinal smooth muscle contraction, mucus secretion, gastric bicarbonate secretion, and tumorigenesis (11). EP2 has been identified as a key pro-inflammatory prostanoid receptor (16-18). EP2 signaling facilitates cytokine expression and promotes T cell differentiation (19-21). EP2-targeted therapy has been shown to be beneficial in inflammatory disorders such as neurodegeneration, endometriosis, and cancer (18, 22). Thus, targeting the EP2 receptor may allow therapeutic reduction of intestinal inflammation while preserving prostanoid-mediated gut protection. However, before this approach can be implemented, we need better understanding of the roles of EP2 in the regulation of intestinal prostanoids and barrier permeability. The aim of this study was to define the role of EP2 receptor inhibition in the feedback regulation of COX-2 in the intestine by its end-product, PGE2 under diverse inflammatory scenarios including necrotizing enterocolitis, endotoxemia, and sepsis. We hypothesized that EP2 mediates upregulation of COX-2 by PGE2 in the intestinal epithelium. We demonstrate that EP2 blockade reduces induction of COX-2 by PGE2 in cultured enterocytes in vitro, and attenuates intestinal expression of COX-2 in experimental NEC and experimental peritonitis.

Materials and Methods

Chemicals and antibodies.

Key reagents were purchased from the following suppliers: PGE2, PF-04418948, ONO-8711, and E7046, Cayman Chemical (Ann Arbor, MI); β-actin antibody, LPS from E. coli O127:B8, FITC-dextran 4000, forskolin, U73122, and H89, Sigma-Aldrich (St. Louis, MO); MLC and phospho-MLC antibodies, Abcam (Cambridge, MA), COX-2 antibody (Western blot), Cell Signaling Technology (Danvers, MA); COX-2 antibody M-19 (immunostaining) and secondary antibodies for Western blots, Santa Cruz Biotechnology (Santa Cruz, CA). Secondary antibodies for immunostaining and ProLong Diamond Antifade Mountant with DAPI were from ThermoFisher Scientific (Canoga Park, CA).

Cell Culture.

Rat small intestinal IEC-6 cells (American Type Culture Collection, Manassas, VA, passages 17-25) were grown at 37°C, 5% CO2 and 100% humidity in Dulbecco-modified Eagle medium with 5% heat-inactivated fetal bovine serum, and 0.1 U/ml insulin-transferrin-selenium. Cell cultures of 70-90% confluence were used in experiments.

Immunoblotting.

Cell monolayers or mucosal scrapings were rinsed with cold phosphate-buffered saline (PBS) and lysed on ice with RIPA buffer (150 mM NaCl, 50 mM Tris pH 7.5, 1 mM EDTA, 1 mM EGTA, 1% Triton X-100, 0.1% SDS, 0.5% sodium deoxycholate, 1 mM PMSF) for 10 min. After clearing by centrifugation at 10,000 rpm for 10 min at 4°C, the lysates were mixed with Laemmli buffer and boiled for 1 min. Proteins (20 μg/lane) were separated by electrophoresis through denaturing 10% SDS-polyacrylamide gel. Gels were electroblotted onto nitrocellulose membranes. The membranes were blocked for 1 h at room temperature in blocking buffer (PBS with 0.1% Tween-20 and 3% bovine serum albumin) and incubated with primary antibody as recommended by the manufacturer, followed by incubation with an appropriate secondary antibody for 1 h at room temperature. After extensive washing in PBS, membranes were soaked in luminol-peroxide reagent and exposed to x-ray film. Protein bands were quantified using GelDoc system (Bio-Rad, Hercules, CA).

Immunofluorescence microscopy.

4 μm paraffin sections of terminal ileum were deparaffinized, rehydrated and blocked in 10% normal donkey serum in PBS. The sections were incubated with primary antibody, washed, and incubated with secondary FITC-conjugated antibodies. Slides were mounted with ProLong Diamond Antifade Mountant with DAPI. Images were acquired using DMI 6000B microscope (Leica, Buffalo Grove, IL). FIJI software was used for fluorescence quantification. For comparisons, sections were processed on the same slide and photographed at the same exposure.

Real time RT-PCR.

Total RNA was isolated from ileal mucosal scrapings using Trizol (ThermoFisher Scientific). The first strand cDNA was synthesized using the iScript cDNA Synthesis Kit (Bio-Rad). Real time PCR was performed on LightCycler 480 using either SYBR Green 1 Master Mix (Roche Diagnostics, Indianapolis, IN). Rat COX-2 (ATG TGC ACT ACG GTT ACA AAA GT and TGA ACT CTC TCC TCA GAA GAA CC) and HPRT (ACA GGC CAG ACT TTG TTG GAT and GGC CAC AGG ACT AGA ACG TC) primers were custom-synthesized by Integrated DNA Technology (Coralville, IA).

Animals.

All animal experiments have been approved by the Institutional Animal Care and Use Committee at Children’s Hospital Los Angeles. Timed pregnant rats were purchased from Envigo (Placentia, CA). C57Bl/6J mice were bred in-house or purchased from Jackson Laboratories (Sacramento, CA). EP2 knockout mice were purchased from Jackson Laboratories and bred homozygous. EP1 knockout mice (a generous gift from Dr. M. D. Breyer, Vanderbilt University) were extensively backcrossed to C57Bl/6J.

Rat NEC.

A well-established rat model of NEC was used (23). Newborn rats were separated from their mothers at birth and kept in an incubator at 30°C and 90% humidity. They were subjected to a regimen of feeding with formula (200 μL of 60% Similac, Abbott Nutritional, Columbus, OH and 40% Esbilac canine milk replacement, Pet AG, Hampshire, IL) and hypoxia (95% N2, 5% O2 for 10 min) every 8 h. Cronobacter muytjensii 51329 (ATCC) was added to formula at 107 cfu/ml . On day of life 4, rats were euthanized and conventional histologic NEC scores from 0 to 4 (0, normal architecture; 1, epithelial sloughing and/or submucosal edema; 2, destruction of villus tips; 3, damage extending to the base of the villi; 4, complete obliteration of epithelium) were assigned by pathologist blinded to treatment groups upon microscopic examination of H&E-stained sections of terminal ileum, based on the extent of epithelial damage (23).

Mouse experimental peritonitis.

Experimental peritonitis was induced in 8-12 week old mice by i.p. injection with 30 mg/kg LPS or by cecal ligation and puncture (CLP). In CLP, animals were anesthetized with isoflurane, the cecum was externalized through a left lower quadrant incision, ligated at its base with a silk suture, and punctured twice using a 21 gauge needle prior to internalization in the peritoneal cavity. The incision was closed with a 5-0 silk suture and staples. Sham operation was performed without ligation and puncture. Mice were sacrificed 12-16 h following induction of peritonitis.

Assessment of barrier function.

Mice were orally gavaged with 200 μL of 22 mg/mL FITC-dextran in 0.9% NaCl 4 h prior to sacrifice for CLP and 6 h prior to sacrifice for LPS injection. Blood was collected by cardiac puncture under isoflurane anesthesia and preserved by adding EDTA to 10 mM. Serum was prepared by centrifugation at 1,000 × g for 5 min. Concentration of FITC-dextran in serum was determined by fluorometry.

Statistics.

Data was analyzed using Student’s t-test with Bonferroni correction or chi-square test where appropriate, using GraphPad Prism 6.0. Data are presented as mean±SEM unless otherwise noted. A p-value of ≤ 0.05 was considered significant.

Results

PGE2 induces COX-2 via EP2 receptor-dependent mechanisms in IEC-6 cells

To determine if COX-2 is upregulated by PGE2 in these cells, IEC-6 intestinal epithelial cells of rat origin were treated with 0-125 μM PGE2 for 12 h. Expression of Cox-2 protein was examined by Western blotting. PGE2 increased Cox-2 expression in a dose-dependent manner with a 2.5 fold peak at 100 μM (Fig. 1A), demonstrating that high levels of PGE2 indeed induce COX-2 expression in enterocytes.

Fig. 1.

Fig. 1.

Open in a new tab

PGE2 induces COX-2 in IEC-6 cells in an EP2-dependent manner. A, dose-response of Cox-2 protein induction by PGE2. B, effects of EP antagonists on induction of Cox-2 by 100 μM PGE2. ONO-8711 (ONO, a specific antagonist of EP1), PF-04418948 (PF, a specific antagonist of EP2), or E7046 (E, a specific antagonist of EP4) were added at 10 μM, 50 μM, and 50 μM, respectively, 10 min prior to treatment with PGE2. Cell lysates were prepared 12 h post treatment. C, effects of signaling mediator inhibitors on induction of Cox-2 by PGE2. U73122, forskolin, and H89, specific inhibitors of phospholipase C, adenylate cyclase, and protein kinase A, respectively, were added at 10 μM prior to treatment with PGE2. D, effects of EP antagonists on PGE2-induced phosphorylation of myosin light chain (MLC). IEC-6 cells were pre-treated with indicated antagonists at 50 μM for 10 min, followed by 100 μM PGE2 for 5 min. Representative blots are shown. β-actin or MLC re-probes are shown to demonstrate lane load. Bar graphs show average ratios of Cox-2/β-actin or p-MLC/MLC band densities, relative to control ratios. *, significant differences from untreated or no inhibitor controls (p<0.05, n>3, Student’s t test).

To elucidate the roles of EP receptors in positive feedback regulation of COX-2, we used receptor-specific antagonists and pathway-specific inhibitors. IEC-6 cells were pre-treated with EP1 antagonist ONO-8711, EP2 antagonist PF-04418948, or EP4 antagonist E-7046 before treatment with PGE2, and expression of COX-2 was examined. Whereas EP1 and EP4 antagonists did not significantly inhibit COX-2 induction, EP2 antagonist blocked it (Fig. 1B). Cox-2 induction was also blocked by inhibitors of EP2 downstream signaling mediators adenylate cyclase (forskolin) or protein kinase A (H89), but not by the inhibitor of EP1 downstream signaling mediator phospholipase C (U73122). To ascertain that EP1 antagonist was active under conditions of our experiment, we examined effects of antagonists on PGE2-induced phosphorylation of myosin light chain kinase (MLC), a response mediated by the EP1-PLC-IP3-Ca2+-MLCK cascade. As expected, EP1 antagonist, but not EP2 or EP4 antagonists, inhibited this response (Fig. 1D). Unfortunately, there is no known test for EP4 antagonist in IEC-6 cells. Our results suggest that EP2, but not EP1 or EP4 signaling is critical to the induction of COX-2 by PGE2. Previously we have shown that Cox-2 induction in IEC-6 cells is accompanied by accumulation of stable PGE2 catabolites, indicating increased COX-2 enzyme activity (5).

Inhibition of EP2 decreases intestinal injury and COX-2 expression in the rat model of NEC

To evaluate the roles of COX-2 and EP receptors in intestinal inflammation in vivo, we first utilized a well-established rat model of NEC. High levels of COX-2 and PGE2 have been previously implicated in the pathogenesis of NEC (3), making this model a useful tool to investigate the contribution of EP receptors. Neonatal rats were subjected to the NEC-inducing regimen of formula feeding and hypoxia, with an additional C. muytjensii 51329 challenge, with or without EP1 antagonist ONO-8711, or EP2 antagonist PF-04418948, or low dose of the specific Cox-2 inhibitor Celecoxib. The latter treatment attenuates Cox-2 activity without completely blocking it (5). Animals were euthanized on day of life 4 and NEC pathology was scored on a scale of 0-4 by examination of hematoxylin-eosin (H&E)-stained ileal sections. Treatment with the EP1 antagonist did not change intestinal injury, whereas treatment with the EP2 antagonist significantly decreased NEC scores (Fig. 2A, B). As expected, low dose of Cox-2 inhibitor also decreased NEC scores (Fig. 2A, B). These results are consistent with roles of EP2 and elevated Cox-2 activity in the intestinal pathology during NEC.

Fig. 2.

Fig. 2.

Open in a new tab

Effects of EP receptor antagonists and Cox-2 inhibitor on rat NEC pathology. Newborn rats (n=19 in each group) were orally inoculated with 107 cfu C. muytjensii during first feeding. Animals were injected i.p. daily with 200 μl normal saline (Ctrl), or 10 mg/kg ONO-8711 (ONO), or 20 mg/kg PF-04418948 (PF), or 1 mg/kg Celecoxib, as indicated. A, NEC scores. *** and **, significant differences from control group, p=0.0004 and 0.0082, respectively, χ2 test. B, representative images of NEC pathology - H&E-stained ileal sections. Bar=100 μm. Note destruction of villus architecture in Ctrl and ONO and normal villus architecture in PF and Celecoxib.

To evaluate effects of EP signaling inhibition on intestinal COX-2 expression during NEC, we compared Cox-2 immunofluorescence in the intestines of animals treated with either EP1 or EP2 specific antagonist and subjected to the NEC-inducing regimen of formula feeding and hypoxia. There was no change in COX-2 immunofluorescence in animals treated with the EP1 antagonist compared to control. However, animals treated with the EP2 antagonist had significantly lower levels of epithelial COX-2 immunofluorescence (Fig. 3A). To corroborate immunofluorescence analysis, intestinal COX-2 mRNA expression was evaluated using real time PCR. EP2 antagonist, but not EP1 antagonist significantly reduced COX-2 mRNA expression (Fig. 3B). These data indicate that EP2, but not EP1, positively regulates intestinal COX-2 expression during NEC.

Fig. 3.

Fig. 3.

Open in a new tab

EP2 antagonist reduces Cox-2 expression in experimental NEC. A, representative images of ileal sections from 4 day old rats subjected to the NEC-inducing regimen + C. muytjensii, with or without treatment with ONO-8711 (ONO) or PF-04418948 (PF), as indicated. Green, Cox-2 immunofluorescence, blue, DAPI-stained nuclei. Bar=100 μm. Graph shows relative green pixel densities in high power frame (HPF) randomly placed over epithelium. Data are average ± SEM of 3 independent experiments, 25 HPFs per image. *, significant difference from control, p=0.026, Student’s t test. B, levels of COX-2 mRNA in the intestinal mucosa of 4 day old rats subjected to the NEC-inducing regimen and treated with or without EP antagonists as indicated. COX-2 mRNA levels were normalized to HPRT transcript levels in the same sample. Results are average of at least 5 independent experiments in each group. **, significant difference from control, p=0.002, Student’s t test).

EP2 receptor deficiency decreases COX-2 upregulation during experimental peritonitis

We next evaluated the role of the EP2 receptor in inflammatory gut barrier failure in mice, which allows genetic ablation of EP receptors. ep1−/− and ep2−/− mice are healthy under normal conditions and do not have gastrointestinal defects (24). We employed endotoxemia and CLP models of experimental peritonitis, which is known to be associated with high levels of COX-2 (5). The endotoxemia and CLP models each have their advantages and disadvantages. LPS injection, which causes a rapid transient induction of high levels of inflammatory cytokines , is more controlled than CLP but is less clinically relevant, whereas polymicrobial sepsis secondary to CLP, which closely resembles the progression, characteristics, and cytokine response of clinical sepsis, is less controlled (25).

In the endotoxemia model, WT and congenic EP1 or EP2 receptor knockout mice of both sexes were injected with saline (control) or LPS. In the polymicrobial sepsis model, mice were subjected to sham operation or CLP. Mice were orally gavaged with FITC-dextran at the time of treatment. Intestinal barrier breakdown was evaluated by measuring FITC-dextran fluorescence in the serum and by examining epithelial morphology. COX-2 protein expression in terminal ileum samples was determined using Western blotting and immunofluorescence. FITC-dextran translocation to serum dramatically increased following LPS injection or CLP in all mice regardless of their EP receptor genotype (Fig. 4A). Experimental peritonitis in both models was associated with massive epithelial sloughing, but there were no apparent signs of destruction of the epithelial architecture. WT and EP receptor knockout mice had no apparent differences in epithelial morphology or extent of epithelial sloughing associated with experimental peritonitis. Fig. 4B shows changes in ep2−/− mice; similar changes were observed in WT and ep1−/− mice (data not shown). EP2 gene knockout, but not EP1 gene knockout, significantly attenuated Cox-2 protein expression following either LPS injection or CLP (Fig. 5). Cox-2 immunofluorescence followed the same pattern (Fig. 6, results for LPS injection not shown). According to these results, neither EP1 deficiency, nor EP2 deficiency significantly protect the gut barrier during experimental peritonitis. However, EP2 deficiency blunts peritonitis-associated COX-2 induction in the mucosa.

Fig. 4.

Fig. 4.

Open in a new tab

Effects of peritonitis on barrier function and epithelial morphology. A, relative serum FITC-dextran levels following saline/sham, LPS injection, and CLP. Open boxes, WT; hatched and solid boxes, ep1−/− and ep2−/−, respectively (homozygous ep1 and ep2 knockouts) ***, significant differences from controls, p<0.001, n≥12 in each group, Student’s t test. B, Representative H&E-stained sections of small intestine of ep2−/− mice subjected to sham operation, LPS injection, or CLP. Bar=1 mm. Arrowheads indicate masses of sloughed epithelium. Framed areas are shown at higher magnification on the right to demonstrate degraded sloughed epithelium.

Fig. 5.

Fig. 5.

Open in a new tab

Genetic ablation of EP2 reduces levels of mucosal Cox-2 protein in experimental peritonitis. Mice of indicated genotypes were subjected to sham operation, i.p. injection of 30 mg/kg LPS, or CLP. Cox-2 protein levels in the small intestine were determined by Western blotting. β-actin re-probe is shown to demonstrate lane load. The image is representative of 5 independent experiments. Cox-2 band densities, which were normalized to β-actin density in the same sample, are shown relative to WT sham on the graph below. *, significant difference from respective WT, p<0.05, Student’s t test.

Fig. 6.

Fig. 6.

Open in a new tab

EP2 deficiency reduces Cox-2 immunofluorescence following CLP. Representative images of small intestinal epithelium from WT, ep1−/−, or ep2−/− mice subjected to sham operation or CLP. Green, Cox-2 immunofluorescence, blue, DAPI-stained nuclei. Bar=200 μm. Graph shows relative green pixel densities in high power frame (HPF) randomly placed over epithelium. Open bars, WT; hatched bars, ep1−/−, solid bars, ep2−/−. Data are average ± SEM of 3 independent experiments, 25 HPFs per image. *, significant difference from control, p=0.044, Student’s t test.

Discussion

In this report, we present evidence for the role of the PGE2 receptor EP2 in the positive feedback regulation of COX-2 by its end-product, PGE2, during intestinal inflammation. COX-2 is known to be upregulated during intestinal inflammation, and EP receptors are expressed in the intestinal epithelium (3, 5). We show that COX-2 is induced by PGE2 in cultured enterocytes and that this induction is attenuated when cells are pre-treated with an EP2 antagonist, but not with an EP1 or EP4 antagonist, thus implicating EP2 as a mediator of this response. Induction of COX-2 in IEC-6 cells requires relatively high concentration of PGE2, which is consistent with the positive feedback regulation occurring specifically under inflammatory conditions characterized by high levels of prostanoid production. Although we do not have direct evidence for the autocrine loop positively regulating COX-2 expression by PGE2, the fact that EP2 blockade reduces COX-2 expression indicates that high levels of Cox-2 seen during intestinal inflammation depend on PGE2-EP receptor signaling, thus indirectly pointing to the autocrine regulation. The autocrine regulation of COX-2 has been previously described (10).

One might argue that 100 μm concentration of PGE2 is not physiologic. However, PGE2 is rapidly degraded in the organismal milieu, therefore its local concentrations at production sites are impossible to measure and are thus unknown. Because COX-2 and EPs are co-localized at apical membranes, concentrations of PGE2 in the vicinity of EP2 receptors may be very high. High levels of PGE2 are required for COX-2 induction in the epithelioid cell lines that we used. In the intestinal mucosa lower concentrations may be sufficient (8).

Since our in vitro results pointed to potential positive feedback regulation of PGE2 production in the intestinal epithelium, we set to determine whether this mechanism plays a role in vivo. To this end, we evaluated COX-2 expression and contribution of low affinity EP receptors (EP1 and EP2) in rodent models of intestinal inflammation. In the neonatal rat model of NEC, EP2 antagonist, but not EP1 antagonist, decreased COX-2 upregulation. In the adult mouse models of endotoxemia and polymicrobial sepsis, COX-2 upregulation was attenuated by EP2 genetic deficiency, but not by EP1 deficiency. Taken together, these results indicate the role of EP2 in the upregulation of the intestinal COX-2 under a variety of inflammatory scenarios.

Since many different cell types may produce prostanoids, it was not clear which particular cell types are responsible for high local levels of inflammatory PGE2 in the intestine. Our immunofluorescence microscopy experiments (Fig. 3A and 6) clearly point to the epithelium as the tissue within the intestine whereby the majority of COX-2 induction occurs during inflammation. Although COX-2 expression in the intestinal epithelium may be regulated systemically, induction of COX-2 by PGE2 in cultured enterocytes is consistent with direct epithelial response to locally produced inflammatory prostanoids.

The role of COX-2 and PGE2 in intestinal inflammation and barrier breakdown is paradoxical. Low levels of COX-2 and PGE2 are necessary to maintain intestinal homeostasis while high levels are seen during inflammation and are detrimental to intestinal barrier integrity. We have shown that COX-2 is induced by its end-product PGE2 in enterocytes, which, at least in vitro, requires certain threshold concentration of PGE2. Our data indicating the critical role of EP2 in this process supports the known pro-inflammatory effect of this PGE2 receptor.

EP2 is known to be pro-inflammatory. We have shown that it is the receptor involved in positive feedback regulation of COX-2 in enterocytes. Due to the existence of multiple concentration-dependent physiologic responses to prostanoids, global inhibition of COX-2 may interfere with housekeeping functions of prostanoids and cause a variety of adverse effects such as hypertension, cardiac failure, coronary atherosclerosis, renal failure, stroke, and myocardial infarction (26, 27). By contrast, targeted inhibition of the individual EP receptors may blunt inflammation without inhibiting the homeostatic effects of COX-2 and PGE2.

We report the effects of EP2 receptor modulation in three physiologically distinct models of in vivo intestinal inflammation. High levels of COX-2 and PGE2 have been shown to play a role in the pathogenesis of NEC, LPS induced peritonitis, and CLP induced sepsis (5, 6, 28, 29). We used a well-established NEC model that allows for the evaluation of neonatal inflammation in an immature intestinal epithelium. In this model, an EP2 antagonist was found to decrease both NEC intestinal injury score and COX-2 expression, while an EP1 antagonist had no effect. These findings in combination with our in vitro data suggest that EP2 is critical for COX-2 positive feedback induction and subsequent epithelial damage during NEC.

EP2 genetic deficiency attenuated COX-2 expression following LPS injection or CLP in adult mice. However, this was not associated with protection against barrier breakdown. This may be due to the multifactorial character of intestinal barrier breakdown during adult peritonitis involving inflammatory cytokines and systemic stresses in addition to prostanoids (25, 30), and points to differences between NEC and peritonitis in the pathogenesis of gut barrier breakdown.

Our findings show that EP2 receptor modulation may have potential therapeutic benefits in the treatment of barrier failure during NEC.

Acknowledgments

We thank Dr. Larry Wang for help with NEC pathology and Dr. Mike Breyer for the gift of ep1 knockout mice. This study was supported by NIH Grant AI 014032 to H.R.F.

Footnotes

Authors declare no conflict of interested in association with this study.

References

  • 1.Defazio J, Fleming ID, Shakhsheer B, Zaborina O, Alverdy JC: The opposing forces of the intestinal microbiome and the emerging pathobiome. Surg Clin North Am 94(6):1151–1161, 2014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Klingensmith NJ, Coopersmith CM: The Gut as the Motor of Multiple Organ Dysfunction in Critical Illness. Crit Care Clin 32(2):203–212, 2016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Lugo B, Ford HR, Grishin A: Molecular signaling in necrotizing enterocolitis: regulation of intestinal COX-2 expression. J Pediatr Surg 42(7):1165–1171, 2007. [DOI] [PubMed] [Google Scholar]
  • 4.MacDermott RP: Alterations in the mucosal immune system in ulcerative colitis and Crohn’s disease. Med Clin North Am 78(6):1207–1231, 1994. [DOI] [PubMed] [Google Scholar]
  • 5.Short SS, Wang J, Castle SL, Fernandez GE, Smiley N, Zobel M, Pontarelli EM, Papillon SC, Grishin AV, Ford HR: Low doses of celecoxib attenuate gut barrier failure during experimental peritonitis. Lab Invest 93(12):1265–1275, 2013. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Li W, Li H, Mu Q, Zhang H, Yao H, Li J, Niu X: Protective effect of sanguinarine on LPS-induced endotoxic shock in mice and its effect on LPS-induced COX-2 expression and COX-2 associated PGE2 release from peritoneal macrophages. Int Immunopharmacol 22(2):311–317, 2014. [DOI] [PubMed] [Google Scholar]
  • 7.Hinz B, Brune K, Pahl A: Prostaglandin E(2) upregulates cyclooxygenase-2 expression in lipopolysaccharide-stimulated RAW 264.7 macrophages. Biochem Biophys Res Commun 272(3):744–748, 2000. [DOI] [PubMed] [Google Scholar]
  • 8.Yoshida K, Fujino H, Otake S, Seira N, Regan JW, Murayama T: Induction of cyclooxygenase-2 expression by prostaglandin E2 stimulation of the prostanoid EP4 receptor via coupling to Galphai and transactivation of the epidermal growth factor receptor in HCA-7 human colon cancer cells. Eur J Pharmacol 718(1-3):408–417, 2013. [DOI] [PubMed] [Google Scholar]
  • 9.Faour WH, He Y, He QW, de Ladurantaye M, Quintero M, Mancini A, Di Battista A: Prostaglandin E(2) regulates the level and stability of cyclooxygenase-2 mRNA through activation of p38 mitogen-activated protein kinase in interleukin-1 beta-treated human synovial fibroblasts. J Biol Chem 276(34):31720–31731, 2001. [DOI] [PubMed] [Google Scholar]
  • 10.Tjandrawinata RR, Hughes-Fulford M: Up-regulation of cyclooxygenase-2 by product-prostaglandin E2. Adv Exp Med Biol 407:163–170, 1997. [DOI] [PubMed] [Google Scholar]
  • 11.Dey I, Lejeune M, Chadee K: Prostaglandin E2 receptor distribution and function in the gastrointestinal tract. Br J Pharmacol 149(6):611–623, 2006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Moreno JJ: Eicosanoid receptors: Targets for the treatment of disrupted intestinal epithelial homeostasis. Eur J Pharmacol 796:7–19, 2017. [DOI] [PubMed] [Google Scholar]
  • 13.Kunikata T, Tanaka A, Miyazawa T, Kato S, Takeuchi K: 16,16-Dimethyl prostaglandin E2 inhibits indomethacin-induced small intestinal lesions through EP3 and EP4 receptors. Dig Dis Sci 47(4):894–904, 2002. [DOI] [PubMed] [Google Scholar]
  • 14.Fredenburgh LE, Velandia MM, Ma J, Olszak T, Cernadas M, Englert JA, Chung SW, Liu X, Begay C, Padera RF, et al. : Cyclooxygenase-2 deficiency leads to intestinal barrier dysfunction and increased mortality during polymicrobial sepsis. J Immunol 187(10):5255–5267, 2011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Loftin CD, Tiano HF, Langenbach R: Phenotypes of the COX-deficient mice indicate physiological and pathophysiological roles for COX-1 and COX-2. Prostaglandins Other Lipid Mediat 68-69:177–185, 2002. [DOI] [PubMed] [Google Scholar]
  • 16.Ansari KM, Sung YM, He G, Fischer SM: Prostaglandin receptor EP2 is responsible for cyclooxygenase-2 induction by prostaglandin E2 in mouse skin. Carcinogenesis 28(10):2063–2068, 2007. [DOI] [PubMed] [Google Scholar]
  • 17.Li S, Xu X, Jiang M, Bi Y, Xu J, Han M: Lipopolysaccharide induces inflammation and facilitates lung metastasis in a breast cancer model via the prostaglandin E2-EP2 pathway. Mol Med Rep 11(6):4454–4462, 2015. [DOI] [PubMed] [Google Scholar]
  • 18.Ma X, Aoki T, Tsuruyama T, Narumiya S: Definition of Prostaglandin E2-EP2 Signals in the Colon Tumor Microenvironment That Amplify Inflammation and Tumor Growth. Cancer Res 75(14):2822–2832, 2015. [DOI] [PubMed] [Google Scholar]
  • 19.Boniface K, Bak-Jensen KS, Li Y, Blumenschein WM, McGeachy MJ, McClanahan TK, McKenzie BS, Kastelein RA, Cua DJ, de Waal Malefyt R: Prostaglandin E2 regulates Th17 cell differentiation and function through cyclic AMP and EP2/EP4 receptor signaling. J Exp Med 206(3):535–548, 2009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Ganesh T: Prostanoid receptor EP2 as a therapeutic target. J Med Chem 57(11):4454–4465, 2014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Sheibanie AF, Khayrullina T, Safadi FF, Ganea D: Prostaglandin E2 exacerbates collagen-induced arthritis in mice through the inflammatory interleukin-23/interleukin-17 axis. Arthritis Rheum 56(8):2608–2619, 2007. [DOI] [PubMed] [Google Scholar]
  • 22.Jiang J, Dingledine R: Prostaglandin receptor EP2 in the crosshairs of anti-inflammation, anti-cancer, and neuroprotection. Trends Pharmacol Sci 34(7):413–423, 2013. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Nadler EP, Dickinson E, Knisely A, Zhang XR, Boyle P, Beer-Stolz D, Watkins SC, Ford HR: Expression of inducible nitric oxide synthase and interleukin-12 in experimental necrotizing enterocolitis. J Surg Res 92(1):71–77, 2000. [DOI] [PubMed] [Google Scholar]
  • 24.Kobayashi T, Narumiya S: Function of prostanoid receptors: studies on knockout mice. Prostaglandins Other Lipid Mediat 68-69:557–573, 2002. [DOI] [PubMed] [Google Scholar]
  • 25.Remick DG, Newcomb DE, Bolgos GL, Call DR: Comparison of the mortality and inflammatory response of two models of sepsis: lipopolysaccharide vs. cecal ligation and puncture. Shock 13(2):110–116, 2000. [DOI] [PubMed] [Google Scholar]
  • 26.Komers R, Anderson S, Epstein M: Renal and cardiovascular effects of selective cyclooxygenase-2 inhibitors. Am J Kidney Dis 38(6):1145–1157, 2001. [DOI] [PubMed] [Google Scholar]
  • 27.Konstam MA, Weir MR: Current perspective on the cardiovascular effects of coxibs. Cleve Clin J Med 69 Suppl 1:SI47–52, 2002. [DOI] [PubMed] [Google Scholar]
  • 28.Bergholz R, Zschiegner M, Eschenburg G, Wenke K, Tiemann B, Roth B, Appl B, Reinshagen K, Sommerfeldt D, Ridderbusch I: Mucosal loss with increased expression of IL-6, IL-8, and COX-2 in a formula-feeding only neonatal rat model of necrotizing enterocolitis. J Pediatr Surg 48(11):2301–2307, 2013. [DOI] [PubMed] [Google Scholar]
  • 29.Strong VE, Mackrell PJ, Concannon EM, Naama HA, Schaefer PA, Shaftan GW, Stapleton PP, Daly JM: Blocking prostaglandin E2 after trauma attenuates pro-inflammatory cytokines and improves survival. Shock 14(3):374–379, 2000. [DOI] [PubMed] [Google Scholar]
  • 30.McCarron EP, Williams DP, Antoine DJ, Kipar A, Lemm J, Stehr S, and Welters ID: Exploring the translational disconnect between the murine and human inflammatory response: analysis of LPS dose-response relationship in murine versus human cell lines and implications for translation into murine models of sepsis. J Inflamm Res 8:201–209, 2015. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES