Department of Microbiology and Infectious Disease, Health Sciences Centre, University of Calgary, 3330 Hospital Drive NW, Calgary, Alberta, Canada T2N 4N1. E-mail: email@example.com
Prostaglandin E2 (PGE2) is one of the most important biologically active prostanoids found throughout the gastrointestinal tract. Despite the fact that PGE2 regulates many physiological functions of the gut including mucosal protection, gastrointestinal secretion and motility, it is implicated in the pathophysiology of inflammatory bowel diseases (IBD) and colorectal neoplasia. The varied biological functions exerted by PGE2 are through the pharmacologically distinct, G-protein coupled plasma membrane receptors termed EP receptors. Disruptions of various prostanoid receptor genes have helped in unravelling the physiological functions of these receptors. To date, all four subtypes of EP receptors have been individually knocked out in mice and various phenotypes have been reported for each subtype. Similarly, in vitro and in vivo studies using EP receptor agonists and antagonists have helped in uncoupling the diverse functions of PGE2 signalling involving distinct EP receptors in the gut. In this review, we will summarize and conceptualize the salient features of EP receptor subtypes, their regional functions in the gut and how expressions of EP receptors are altered during disease states.
The interests in understanding prostaglandin (PG)-mediated functions of the gastrointestinal (GI) tract have grown steadily since the discovery of PGs in 1957. In recent years, technological advances in the field of molecular biology have helped researchers make substantial progress towards unraveling the complex functions of PGs in the gut. It is only now that the functional roles of the various PGs in the GI tract are established. Apparently, each of the bioactive PGs exhibits versatility and diversity in functions in various segments of the gut. However, the mechanisms by which they exert their biological functions are not clearly known. To discern the variable roles of PGs, it is imperative to understand thoroughly the mode of action of these mediators, including the pharmacology, expression and distribution of their respective receptors in the gut. Nevertheless, insights into PG receptor biology in the GI tract will be of fundamental importance in developing strategies for pharmacological intervention in various GI ailments.
PGs are 20-carbon fatty acid derivatives that are found ubiquitously in all tissues and organs and mediate variety of physiological and pathological functions. They are synthesized in the cell from different essential fatty acid precursors, including arachidonic acid (AA). PGs derived from AA are termed series-2 PGs, which consist of prostaglandin E2 (PGE2), prostaglandin D2 (PGD2), prostaglandin I2 (PGI2), prostaglandin F2α (PGF2α) and thromboxane A2 (TXA2) (Calder, 2001). All of these PGs share a common initial biosynthetic pathway that begins with the hydrolysis of cell-membrane phospholipids, mediated by the enzyme phospholipase A2 (PLA2), which is found mostly in cellular membranes including the plasma membrane (Murakami et al., 1997). Diverse physiological and pathological stimuli can result in the activation of PLA2 to liberate AA from membrane phospholipids into the cytoplasm (Murakami and Kudo, 2002). Upon release, AA is converted into unstable endoperoxide intermediates, PGG2 and PGH2 (Hamberg et al., 1974) by the action of cyclooxygenase (COX) in a rate limiting enzymatic reaction. Three isoforms of COX have been identified to date. The constitutively expressed COX-1 and the inducible COX-2 are the most important isoforms (Smith et al., 1994, 1996). COX-3 is a splice variant of COX-1 and is mostly expressed in the brain and the heart (Chandrasekharan et al., 2002). The oxygenated intermediate PGH2 is in turn metabolized by cell-specific synthases and isomerases into PGD2, PGE2, PGF2α, PGI2, and TXA2 (Vane et al., 1998; Figure 1). Prostanoids are released outside the cell immediately after their synthesis where they exert their biological functions through their interaction with the cell surface prostanoid receptors in an autocrine or paracrine fashion (Narumiya, 1994). Alternatively, the action of prostanoids are terminated when they are transported across the cell membrane into the cytoplasmic compartment with the help of PG transporters (Kanai et al., 1995) where they are acted upon by oxidizing and reducing enzymes, 15 hydroxy PG dehydrogenases and delta 13–15-ketoprostaglandin reductase, respectively (Tai et al., 2002).
Prostanoid receptors belong to the family of Rhodopsin-type receptors characterized by their seven transmembrane domains that are intracellularly coupled to different subunits of G proteins (Breyer et al., 2001). Five major types of prostanoid receptors, namely; D prostanoid (DP), E prostanoid (EP), F prostanoid (FP), I prostanoid (IP) and T prostanoid (TP), which include six subtypes namely; DP1, DP2, EP1, EP2, EP3 and EP4, have been described for the prostanoids PGD2, PGE2, PGF2α, PGI2 and TXA2. The structures, properties and functions of most of these receptors have previously been reviewed (Narumiya et al., 1999). Investigation of the functional role of these receptors in gut physiology as well as pathophysiology is presently an important pursuit. For example, the role of DP, IP, EP and TP receptors in the propulsive peristaltic movement of the guinea-pig small intestine has been studied and the receptors responsible for differential peristaltic motor effect identified (Shahbazian et al., 2002). Similarly, the role of DP, FP, IP, TP and the EP receptor subtypes in the development of carcinogen-induced aberrant crypt foci in colon has been studied extensively (Watanabe et al., 1999; Mutoh et al., 2002).
It is clear that the different prostanoids synthesized and their interactions with respective cellular receptors are responsible for the varied biological actions of the host. As the cellular components of the GI tract are known to possess the machinery for the biosynthesis of PG, it is not surprising that PGs are produced throughout the gut. Among the different PGs that are produced, PGE2 is considered to be the most important for normal physiological functions of the GI tract including gastric mucosal protection and motility and is also implicated in the pathology of various disease conditions such as inflammatory bowel disease (IBD) (Ahrenstedt et al., 1994), entero-invasive bacterial diseases (Reseta and Barrett, 2002) and colorectal cancers (Eberhart et al., 1994). Diversity in the cellular functions of PGE2 is attributed to its binding to four different subtypes of EP receptors that in turn propagates signals through alteration in the intracellular calcium (Ca2+) or cyclic adenosine monophosphate (cAMP) levels. This results in the activation of an array of kinases modulating diverse cellular functions (Figure 2). Although, signalling through different EP receptor determines the various effects of PGE2, regional and differential expression of EP receptors in the GI tract is critical for determining its biological functions. Based on this premise, we will summarize in this review the salient features of EP receptor subtypes, their regional functions in the gut and their differential expression during normal and disease states.
EP1 is a 42-kDa protein that has been cloned and expressed from humans and rodents (Funk et al., 1993; Watabe et al., 1993; Okuda et al., 1996). Two splice variants of EP1 receptor have been reported (Okuda et al., 1996). Among the four subtypes of EP receptors, EP1 has the least affinity for PGE2 (dissociation constant (KD) of 16–25 nM, (Table 2)). It binds PGs in the rank order PGE2>sulprostone, iloprost>PGE1>misoprostol, M&B-28767>PGF2α>PGD2 (Abramovitz et al., 2000; Table 3). A few selective agonists for EP1 like ONO-DI-004, 17-phenyl trinor PGE2 are available, 17-phenyl trinor PGE2 being the most well known of them. Other agonists including sulprostone, carbacyclin and enprostil exhibit the highest affinity for EP1, but are also very potent EP3 agonists. ONO-AE-829, ONO-8711, ONO-8713, SC-19220 and AH6809 are used as antagonists for the EP1 receptor but AH6809 is a weak antagonist (Woodward et al., 1995, 2005).
Table 2. Dissociation constants (Kd) of PGE2 for EP receptor subtypes in various species
Kd (nM) (human)
Kd (nM) (rat)
Kd (nM) (mouse)
Abbreviations: EP, E prostanoid; PGE2, prostaglandin E2.
Before the discovery of the EP2 receptor in 1994, the cloned EP4 receptor was referred to as EP2 (Bastien et al., 1994; Regan et al., 1994a). The EP2 receptor is a 53 kDa protein that has been cloned and expressed from humans, bovine, rabbits, rodents and several other species (An et al., 1993; Katsuyama et al., 1995; Guan et al., 1996; Nemoto et al., 1997; Arosh et al., 2003). The affinity of PGE2 for EP2 receptor differs greatly across species. The rat EP2 receptor has a significantly higher affinity for PGE2 than human or mouse. The mouse EP2 receptor shows the least affinity (Table 2). The relative affinity of binding of various EP ligands for the mouse EP2 receptor was PGE1, PGE2>16, 16-dimethyl-PGE2>11-deoxy-PGE1>butaprost>AH13205, misoprostol> AH-6809. Similarly, the human EP2 receptor binds PGE2 and PGE analogues with a rank order of PGE2>PGE1>16, 16-dimethyl-PGE2>11-deoxy-PGE1>butaprost, 1-OH-PGE1, M&B-28767> sulprostone (Table 3). Butaprost and ONO-AE1-259-01 are highly selective EP2 receptor agonists, whereas AH-6809 is a well-known EP2 antagonist.
The EP3 receptor of humans and rodents has been successfully cloned and expressed (Sugimoto et al., 1992; Yang et al., 1994). The EP3 is unique among the prostanoid family of receptors in the sense that multiple alternatively spliced variants of EP3 exist that can activate contrasting second messenger signalling (Pierce and Regan, 1998). To date, eight different isoforms for human EP3 have been reported. In general, the EP3 receptor has relatively higher affinity for PGE2 reflected by a very low KD value of 0.33–2.9 (Table 2). The rank order of affinity of PGE2 to the mouse EP3 receptor is sulprostone, M&B-28767, PGE2, PGE1, 11-deoxy-PGE1, 16,16-dimethyl-PGE2, misoprostol>1-OH-PGE1. Different PGE analogs, SC-46275, ONO-AE-248, sulprostone, GR 63799X and 11-deoxy-PGE1 are used as agonists for EP3 with SC-46275 and ONO-AE-248 being highly selective. L826266 is an EP3 antagonist.
The EP4 receptor cDNA encodes a 487–513 amino-acid polypeptide that has been cloned and expressed in vitro from human, rodents, rabbit, bovine and several other species (An et al., 1993; Honda et al., 1993; Breyer et al., 1996; Boie et al., 1997). They have very high affinity for PGE2 with a KD value of 0.59–1.27 (Table 2). The rank order of affinity of PGE ligands for the mouse EP4 receptor was PGE2, PGE1>11-deoxy-PGE1, 16, 16-dimethyl-PGE2, misoprostol>1-OH-PGE1 and M&B-28767 (Table 3). At present, there are few selective agonists or antagonists available for EP4. ONO-AE1-329 is a selective EP4 agonist whereas L161982 and ONO-AE3-208 are selective antagonists.
Diversity in EP receptor signalling and functions
Diverse factors govern the outcome of EP receptor signalling. Basically, structural, pharmacological and functional differences that exist between the subtypes of EP receptors determine the biological effects of PGE2. Structurally, differences observed in the length and composition of amino acids of the N and the C-termini, the second and the third extracellular as well as the intracellular domains are associated with variations in receptor signalling. In fact, differences in the length of C-termini of EP3 isoforms are attributed for their differential activation of varied second messenger pathways (Pierce and Regan, 1998). Likewise, the extracellular sequence of the EP2 receptor is a critical determinant of its structure and function (Stillman et al., 1999). Similarly, a cluster of hydrophobic aromatic amino acids in the second intracellular loop of EP2 but not EP3 isoform β is absolutely essential for activation of Gαs subunit (Sugimoto et al., 2003, 2004). Pharmacologically, EP receptors exhibit differences in binding to various ligands (Table 1) and surprisingly, the affinity of these receptors to its principal ligand, PGE2, greatly varies between receptor subtypes (Tables 2 and 3). In fact, affinity of PGE2 to its receptor may depend on the state of coupling of G protein subunits. A G-protein coupled receptor displays more affinity for PGE2 than a free uncoupled receptor. Apart from differences in the ligand affinity to EP receptors, the phenomena of receptor desensitization and internalization may regulate preferential signalling through certain EP receptor subtypes. Agonist-induced desensitization is a common phenomenon in G-protein coupled receptor that is characterized by the loss of receptor signalling. The EP receptors have been shown to undergo PGE2-induced desensitization. Interestingly, a differential desensitization of EP receptor subtypes has been observed wherein, the EP4 receptor is highly desensitized in contrast to the EP2 receptor which is not (Nishigaki et al., 1996). Similarly, differences in ligand-induced internalization of EP receptors have been reported. EP4 and EP3 isoform-I are readily internalized when activated by PGE2 whereas EP2 as well as EP3 isoform-III and IV are not. Interestingly, the C-terminus of EP4 has been implicated in its ligand-induced sequestration (Desai et al., 2000; Bilson et al., 2004). Functionally, differences among EP receptors directly correlate with the type of signal that it transduces. For example, receptor activation leading to intracellular Ca2+ mobilization is associated with the contraction of smooth muscle cells, whereas increase in cytoplasmic cAMP levels is associated with its relaxation. Evidently, signalling through EP1 receptors increases intracellular Ca2+ levels whereas, EP2 and EP4 have been shown to increase cytoplasmic cAMP. Signalling via EP3 receptors is unique wherein cAMP levels are decreased (Narumiya et al., 1999). Clearly, diverse factors govern the fate of signalling through EP receptors and all these may play a critical role in determining the differential biological effect of PGE2 in the GI tract.
Table 1. Ki values of PGE2 for heterologously expressed human and mouse EP receptors
Abbreviations: EP, E prostanoid; PGE2, prostaglandin.
Inhibitory constant (Ki) values are expressed in nM.
Downstream signalling pathways affected by EP receptors
Coupling of PGE2 or the specific receptor agonists to EP receptors results in their activation and induces signalling cascade inside the cell (Figure 2). Interestingly, signalling through different subtypes of EP receptors seems to alternate and sometimes overlap; yet they are unique in terms of their signalling outcomes. EP1 receptors mediate signalling events by activation of phospholipase C and elevation of cytoplasmic signalling intermediates including inositol triphosphate, diacylglycerol and Ca2+. Coupling of PGE2 to EP1 activates protein kinase Cα and c-Src. In addition to directly activating downstream kinases, signalling through EP1 can also transactivate HER's-2/Neu tyrosine kinase receptor, which is mediated by c-Src resulting in upregulation of vegetative endothelial growth factor-C (Su et al., 2004). Recently, it was reported (Han and Wu, 2005) that c-Src mediated the transactivation of epidermal growth factor receptor by the EP1 receptors through activation of protein kinase B (Akt) that promotes cell proliferation and invasion. These data corroborate with those obtained from studies in EP1 receptor knockout mice, thus implicating a role of EP1 receptors in colon carcinogenesis (Watanabe et al., 1999).
The EP2 and EP4 receptors are linked to the stimulation of cAMP/protein kinase A (PKA) signalling through the sequential activation of Gαs and adenylate cyclase. Contradictory reports associate this signalling pathway to growth and proliferation. Increased cAMP production, upon activation of EP2 and EP4 has been shown to have an antiproliferative effect in human gastric carcinoma cell lines (Okuyama et al., 2002). In contrast, activation of PKA is increasingly being linked to proliferation of various epithelial cell types. Phosphorylation of PKA is coupled to the regulation of glycogen synthase kinase-3 (GSK-3) and Akt (Filippa et al. 1999; Li et al., 2000). PKA phosphorylates and activates Akt kinase, which indirectly inhibits GSK-3. Inhibition of GSK-3 decreases the inhibitory phosphorylation of cytosolic β-catenin that promotes the translocation of β-catenin to the nucleus resulting in cellular proliferation (Cadigan and Nusse, 1997). A recent study has directly linked EP2 receptor activation to cellular proliferation. Upon activation the Gαs subunit of EP2 receptors can directly associate with the regulator of G protein signalling domain of Axin, which inactivates and releases GSK-3β from the Axin complex causing β-catenin activation and nuclear translocation (Castellone et al., 2005). Until recently, it was believed that the phosphoinositide-3-kinase (PI3K) signaling pathway was activated only by the EP4 receptor. However, recent studies have shown that PGE2 signalling through EP2 receptors can activate both PI3K and Akt by using the free βγ subunit of G-protein (Castellone et al., 2005).
Although EP2 and EP4 receptors are capable of stimulating both PKA and PI3K signalling pathway, a receptor-specific signalling outcome has always been observed. For example, a study showed that PGE2 stimulation of EP4, but not EP2 receptor that are stably expressed in HEK cells lead to the phosphorylation of the extracellular signal-regulated kinases (ERKs) by a PI3K-dependent mechanism and induced the expression of early growth response factor-1 (Fujino et al., 2003). The differences in the signalling potential of both EP2 and EP4 receptors are corroborated by the fact that both receptors mediate the phosphorylation of cAMP responsive element binding protein (CREB) by different signalling pathways (Fujino et al., 2005). Signalling through EP2 receptors can decrease the inhibitory tyrosine phosphorylation of PTEN, a phosphatase that can act as a PI3K pathway inhibitor (White et al., 2005). A novel EP2 receptor-signalling pathway, which can transactivate the EGF receptor leading to increased migration and invasion of colon cancer cells, has also been reported (Pai et al., 2002; Buchanan et al., 2003).
The EP3 receptors are unique in their ability to couple to multiple G proteins. They couple and activate the Gi subunits, which results in the inhibition of adenylyl cyclase. Apart from activation of Gi subunits, signalling through EP3 receptors can also activate Gs resulting in cAMP production. Evidence also indicates that the EP3 receptor can activate the small G protein Rho and its target p160 Rho-A binding kinase ROK α (Katoh et al., 1998; Tamma et al., 2003). EP3 receptors have also been shown to activate the Ras signalling pathway leading to cancer (Yano et al., 2002).
Distribution of EP receptors in the gut
Although PGE2 is produced by a variety of cells in the GI mucosa and is found throughout the gut, its cellular targets in the mucosa and the resulting physiological changes in the GI tract are predominantly determined by the presence and the distribution of EP receptors. Pharmacological as well as cellular localization studies have been aimed towards identifying the type and pattern of EP receptor distribution and their density in major cell types in the GI mucosa. However, very little is known of the regional distribution or the temporal changes in the pattern of EP receptor expression and function under normal and disease states in the gut.
EP1 receptor distribution
EP1 receptor expression has been reported in the GI tract of various species. In rat, EP1 mRNA expression is detected at the gastric, small intestinal and colonic muscle layers (Ding et al., 1997). A non-radioactive in situ hybridization technique used to study the pattern of localization of EP receptors in the rat GI tract revealed the presence of EP1 receptors in gastric chief cells that secrete pepsinogen, parietal cells that secrete hydrochloric acid and in mucus secreting gastric epithelial cells. In the small intestine, EP1 receptor expression was noticed only in goblet cells whereas in the large intestine, EP1 receptor expression was seen in goblet cells and in other epithelial cell types. Expression of EP1 receptors has been reported in enteric glial cells (Northey et al., 2000). In mice, EP1 mRNA expression was noticed in the muscularis mucosa (Morimoto et al., 1997). In rabbits, EP1 receptors are highly expressed on the intestinal brush border membranes of intestinal villi. Goblet cells of the jejunum and ileum also show considerable expression. In contrast, goblet cells of the duodenum show a lack of expression of EP1 receptors. Interestingly, EP1 receptors are also highly expressed in the neurons of the myenteric and submucosal ganglia throughout the rabbit intestinal tract (Grasa et al., 2006).
EP2 receptor distribution
EP2 receptors are normally expressed in rat gastric mucous cells, in the goblet cells of the small intestine and also in various epithelial cells of the large intestine (Northey et al., 2000). In mice, an abundant expression of EP2 receptors in the stomach and ileum has been detected by Northern blot analysis (Katsuyama et al., 1995). A recent report suggested a relatively uniform expression of EP2 receptors throughout the mouse GI tract except for the ileum and caecum, where it is weakly expressed. Interestingly, expression of EP2 receptors depends on the state of differentiation of the epithelial cells. Undifferentiated crypt epithelial cells predominantly express EP2 receptors on their nuclear membranes whereas the highly differentiated epithelial cells at the apex of the villi express these receptors on their plasma membrane (Houchen et al., 2003). In rabbits, EP2 receptors are localized on the intestinal brush border membrane of the villi and are also strongly expressed in goblet cells throughout the small intestine. EP2 receptors are weakly expressed in neurons but are completely absent in the smooth muscle cells of the small intestine (Grasa et al., 2006). In humans, EP2 receptor expression is restricted to the luminal surface of the gastric epithelium (Takafuji et al., 2002) and at the apex of the colonic mucosa (Takafuji et al., 2000). In guinea-pigs, the presence of EP2 receptors in the neurons of the enteric nerve plexus in the ileum was detected through pharmacological studies (Lawrence et al., 1992).
EP3 receptor distribution
EP3 receptor gene expression appears to be localized to the muscular parts of the rat intestine whereas in the stomach EP3 mRNA expression is seen in the mucosal layer especially in parietal cells (Ding et al., 1997). In most rodents, EP3 receptor expression is predominantly found in gastric parietal and small intestinal goblet cells. They are also highly expressed in enteric glial and myenteric neuronal cells (Morimoto et al., 1997; Takahashi et al., 1999; Northey et al., 2000). EP3 receptors are highly expressed in rabbit intestinal circular and longitudinal smooth muscles of the duodenum and in the circular muscles of the ileum (Grasa et al., 2006). In humans, EP3 receptor expressions are observed throughout the gastric epithelium (Takafuji et al., 2002) and in the apex of colonic mucosa (Takafuji et al., 2000).
EP4 receptor distribution
EP4 receptor mRNA are mainly expressed in the rodent intestinal mucosal layers and in the parietal as well as the mucosal epithelial cells of the gastric mucosa (Ding et al., 1997; Northey et al., 2000). In the ileum, EP4 receptor expression was demonstrated on mature enterocytes of the villi (Morimoto et al., 1997). Normal EP4 receptor mRNA expression was observed in rabbit gastric epithelial cells (Takahashi et al., 1999). However, EP4 receptor proteins were not detected in any of the segments (Grasa et al., 2006). In humans, EP4 expression was modest in the gastric epithelium but intense expression of EP4 was detected in lamina propria mononuclear cells (Takafuji et al., 2002). In the colon EP4 receptors are strongly expressed in the lateral crypt epithelia (Takafuji et al., 2000) and lamina propria mononuclear cells of CD+ T lymphocytes (Cosme et al., 2000).
Physiological role of EP receptors in the GI tract
The role of PGE2 in GI physiology has been distinctly proved in various animal models. Presently, agonist/antagonist-based approaches along with EP receptor knockout strategies are used to determine the role that each of the EP receptors plays in GI physiology. Using these approaches, the role that EP receptor subtype(s) play in gastric acid secretion, GI motility and mucosal barrier functions including bicarbonate secretion, mucus secretion and epithelial cytoprotection have been elucidated.
Gastric acid secretion
Gastric acid is the main secretion of the stomach. A dual action of PGE2 on gastric acid secretion was observed in rats, wherein, lower concentrations inhibited and higher concentration stimulated secretion (Ding et al., 1997). The inhibitory action of PGE2 on acid secretion in rats was mediated by the EP3 receptor whereas the stimulatory effect was due to the EP4 receptor. It is interesting to note that both EP3 and EP4 receptors are present in parietal cells and acid-producing chief cells in the gastric mucosa. Parietal cells can respond to histamine in releasing acid into the gastric lumen. An EP4 agonist, ONO-AE1-329, was shown to stimulate gastric acid secretion through histamine released from entero chromaffin cells (Kato et al., 2005).
The peristaltic movement of the GI tract is maintained through coordinated contraction and relaxation of GI smooth muscle and PGE2 has been reported to play a major role in GI motility. In rabbits, a complex mechanism involving activation of both EP3 and EP1 receptors was shown to be responsible for the contraction of the small intestine. A direct effect of EP3-mediated activation of intestinal smooth muscles and an indirect effect of EP1-mediated stimulation of myenteric neurons acts together during PGE2-induced intestinal contractions (Grasa et al., 2006). In rats, inhibition of small intestinal hypermotility was mediated through EP4 receptors (Kunikata et al., 2002). Studies conducted in EP receptor knockout mice clearly underscore the importance of EP1 and EP3 receptors in the contraction of longitudinal smooth muscles of gastric fundus and ileum and EP4 receptors in the relaxation (Okada et al., 2000). In guinea-pig ileum, pharmacological investigation revealed that EP3 receptor stimulation mediates contraction of circular smooth muscles and increases peristalisis, whereas EP2 mediates relaxation and decreases peristalisis (Botella et al., 1993; Shahbazian et al., 2002).
GI bicarbonate secretion is one of the first lines of defense of the host epithelial cells against the harsh acidic environment of the stomach and the duodenum. Bicarbonates help in maintaining a narrow zone of neutral pH just above the mucosal lining that offers a protective barrier against diffusing acid (Takeuchi et al., 1997). PGE2 has a stimulatory effect on the GI bicarbonate secretion. In fact, low physiological concentrations of PGE2 in the duodenum would induce an acid-provoked bicarbonate secretion (Hirokawa et al., 2004). An agonist–antagonist-based approach in rats, as well as studies involving EP receptor knockout mice have proved the involvement of different EP receptor subtypes in bicarbonate secretion in different segments of the gut. In the stomach of rats and mice, EP1 receptors are responsible for bicarbonate secretion whereas, the same effect was mediated through EP3 receptors in the duodenum (Takeuchi et al., 1999a). In fact, the presence of EP3 receptors in the duodenum is essential for duodenal bicarbonate secretion to counter luminal acid-induced mucosal damage (Takeuchi et al., 1999b). In addition to EP3 receptors, EP4 receptors have also been proved to be involved in duodenal bicarbonate secretion (Aoi et al., 2004). In humans, EP4 receptors are solely responsible for duodenal bicarbonate secretion, which is in contrast to that seen in rats and mice, where EP3 receptors mediate duodenal bicarbonate secretion (Larsen et al., 2005). This study suggests that species differences occur with regard to EP receptor function(s) and therefore caution should be taken in ascribing a particular function to a particular EP receptor.
Mucins are polymers made of glycoproteins that are secreted by the mucous cells of the stomach and the goblet cells of the intestine that form a protective covering over the lining of the mucosa. They form the major component of the innate immune response of the GI mucosa. PGE2 is reported to be a potent mucin secretagogue. It strongly stimulates mucin secretion from rat gastric epithelial cells (Tani et al., 1997). More recently, the molecular mechanism of PGE2-induced-mucin gene transcription via the ERK MAPK/RSK1/CREB pathway has been elucidated (Cho et al., 2005). PGE2-induced mucus secretion from the gastric epithelial cells of rabbit are mainly mediated through EP4 receptors in a cAMP/PKA-dependent manner (Takahashi et al., 1999). Mucin exocytosis from the antral mucous cells of guinea-pigs is also mediated through EP4 but activated EP1 receptors are shown to potentiate the effect of mucus secretion mediated by EP4 (Ohnishi et al., 2001). An agonist-based approach in identifying the EP receptors responsible for mucin exocytosis in the rat colon and in a human colonic epithelial cell line LS174T revealed that coupling of PGE2 with EP4 receptors was essential for mucin secretion (Belley and Chadee, 1999).
PGE2 plays an important role in protecting the GI mucosa from injuries caused by the harsh environment of the lumen. The mechanism by which PGE2 exerts its protective effects on GI mucosa from noxious agents is termed cytoprotection. For example, PGE2 exerts a potent cytoprotective effect on gastric glandular cells against indomethacin-induced injury, which is independent of neural, vascular and hormonal factors (Brzozowski et al., 2005). Current studies are underway to identify EP receptor subtype that mediates this protective action. It has been reported that, in an acute rat esophagitis model, low doses of PGE2 confer a protective effect that is mediated through EP1 receptors (Yamato et al., 2005). Similarly, in the stomach of rat, EP1 receptors are essential for offering protection against indomethacin or ethanol-induced injury (Araki et al., 2000; Suzuki et al., 2001). In rodents, EP1 receptors also mediate adaptive gastric cytoprotection, under conditions in which initial exposure to mild irritants prevents subsequent damage by more severe irritants (Takeuchi et al., 2001). Activation of the EP1 receptors helps in maintaining gastric mucosal integrity (Takeuchi et al., 2002). In the small intestine of rats, EP3 and EP4 receptors offer cytoprotection against indomethacin-induced injury possibly through increased mucus secretion, enteropooling and through inhibition of hypermotility mediated by EP4 alone (Kunikata et al., 2002). Activation of both EP2 and EP4 receptors in the guinea-pig gastric mucosal cells offers cytoprotection against ethanol-induced apoptosis (Hoshino et al., 2003). In mice, EP2 receptors are essential for preventing radiation-induced apoptosis of crypt epithelial cells at the jejunum (Houchen et al., 2003).
Differential expression of EP receptors during pathological conditions in the gut
This area is perhaps the least studied aspect of EP receptor biology in the gut. To complicate the issue, the limited data that are available does not always discriminate whether EP receptor expressions are altered in individual cells, mucosa or submucosa or the full thickness of the gut. Nevertheless, these data on differential expression of EP receptors during various pathological conditions such as radiation injury, tumorigenesis and inflammation may explain the modus operandi of PGE2 in exerting its pathological functions as against its well-known physiological roles.
A regional and temporal difference in the expression of EP receptors following radiation-induced injury in mice has been observed. Significant changes in EP2 and EP4 mRNA and protein levels were noticed in the jejunum and colon after radiation-induced injury. EP2 receptor expression increased steeply whereas a decrease in EP4 expression was observed. Interestingly, EP2 receptor expression corresponded with epithelial restitution and early crypt morphogenesis of the injured tissues (Houchen et al., 2003).
A differential expression of EP receptors in the colon during early stages of tumorigenesis is observed in rodents and humans. The most notable changes reflect an increase in the expression of EP1 and EP2 receptors and a marked decrease in EP3 receptors. However, EP4 receptor levels remained constant (Shoji et al., 2004). Upregulation of EP1 receptors is clearly observed during early tumorigenesis that helps in cellular proliferation and in the antiapoptotic effects observed in cancer cells, indicating a major role played by EP1 receptors in cancer (Kawamori et al., 2005). A few studies involving RT–PCR analysis have revealed that as well as EP1, EP2 and EP4 receptor mRNAs expression are increased in azoxymethane-induced colorectal cancer tissues (Mutoh et al., 2002; Kawamori et al., 2003). Similarly, EP2 and EP4 mRNA levels are also increased in ApcΔ716 mouse small intestinal and colonic polyps (Sonoshita et al., 2001). These studies implicate EP2 and EP4 receptors in tumorigenesis. The signalling mediated by EP1/2/4 receptors that lead to cellular proliferation and tumorigenesis was briefly discussed under the previous section ‘Downstream signalling pathways affected by EP receptors.’’ Apart from the upregulation and activation of certain EP receptors during cancer, it is interesting to note that EP3 receptor is downregulated. Downregulation of the EP3 receptor is linked to the hypermethylation of the EP3 receptor gene, thus inhibiting the suppressive role of EP3 in colon cancer development. Increased expression of EP4 receptors on CD3+ lymphocytes adjacent to the tumour was also evident in human gastric mucosal cancers but the role of T lymphocyte EP4 signalling in cancer is not clear (Takafuji et al., 2002). An increased expression of EP2, EP3 and EP4 receptor mRNA levels was observed in rat oesophageal squamous cell dysplasia and Barrett's metaplasia induced by duodenal contents reflux; however, no changes in the mRNA levels of the EP1 receptor was observed (Jang et al., 2004). This study suggests that regional differences may exist for the role of EP receptor subtypes in GI tumorigenesis.
A differential expression of EP receptors on the GI mucosa may occur during inflammation. One interesting study showed defined patterns and identified striking differences of mucosal expression of the EP4 receptors between cells in normal human colonic lamina propria and those from the inflamed human colon (Cosme et al., 2000). In particular, mucosal EP4 receptor expression was upregulated in T lymphocytes. In ulcerative colitis, apart from an increase in EP4 receptor expression in lamina propria T lymphocytes, an increase in the levels of EP2 and EP3 receptors was apparent in epithelial cells (Takafuji et al., 2000). At present, there are no reports on EP receptor expression levels in the gut during microbial or parasitic infections.
Role of EP4 receptors in colitis
Even though PGE2 levels are significantly increased during IBD, the functional role PGE2 and individual EP receptors play in the pathogenesis of IBD remains undefined. PGE2–EP receptor coupling may have a critical role in homeostasis or in the onset of GI inflammation and/or tissue repair. The limited data to date suggest that signalling via EP receptors can predetermine whether PGE2 exerts a proinflammatory or anti-inflammatory effect in the gut. For example, studies in vitro, to address early responses of PGE2 in a variety of colonic epithelial cell lines clearly demonstrate that PGE2 couples via high-affinity EP4 receptors to upregulate IL-8 mRNA expression and protein secretion confirming a proinflammatory role for PGE2 (Yu and Chadee, 1999). IL-8 is a potent chemokine that can attract and activate neutrophils to cause nonspecific tissue damage important in the onset of colonic inflammation. In contrast, in EP4 receptor knockout mice, a 7-day regime of dextran sodium sulphate-induced colitis was reported to be more severe than that in wild-type controls. This suggests that signalling via EP4 receptors may play a critical role in maintaining normal mucosal integrity and/or to promote healing (Kabashima et al., 2002). Unfortunately, the role of EP4 signaling in either epithelial cells or the submucosal cells that induce an anti-inflammatory effect has not been elucidated. Moreover, it is not clear whether epithelial barrier functions or cytokine productions are altered in EP4 knockout animals. Similarly, studies in rats using an EP4 receptor agonist reported the suppression of colitis caused by extended DSS treatment through upregulation of an anti-inflammatory cytokine, IL-10 (Nitta et al., 2002). Suppression of Th1-type response, coupled with evidence of increased expression of EP4 receptors on the mucosal T lymphocytes in inflamed colonic mucosa (Cosme et al., 2000), suggests a fundamental role for PGE2-EP4 signalling in T lymphocytes in repair and restitution of damaged tissues. It appears that EP4 signalling outcomes differ between the early onset (pro-inflammatory on mucosal epithelial cells) and the late progressive stages of colitis (anti-inflammatory on immune cells in the lamina propria). Collectively, these studies emphasize the importance of PGE2-EP4 receptor signalling events in the cell types present in colonic mucosa as a fundamental determinant of colitis. More studies are needed to define the diverse functions of PGE2-EP receptor coupling on various cell types before we can model their role in normal and disease states.
Recent evidence clearly suggests that PGE2-EP receptor signalling plays an important role in the GI tract. Importantly, PGE2-EP receptor coupling on different cell types may exert either a pro-inflammatory or anti-inflammatory response. In general, it appears that EP1, EP2 and EP4 receptors are major determinants in the early stages of intestinal tumorigenesis and inflammation, whereas EP4 receptors on immune cells may promote the restitution of colitis/inflammation. There maybe redundancy of EP receptor signalling and EP receptor functions may vary from onset to different stages of disease (initiation versus progression). However, it appears that all receptors are needed for various normal functions of the GI tract. As regional, temporal and species differences exist in EP receptor expression and distribution in the GI tract, future research should aim at identifying EP receptor expression in various cellular components of the mucosa across the length and the breath of the gut. As regional and species differences are apparent with regards to EP receptor functions, it is dangerous to ascribe specific functions for EP receptors in the gut. A priority is to define the distinct EP receptor signalling events in various cell types in the gut. Knowledge of receptor expression coupled with insights into specific signalling events in specific cell types may help the cause of understanding distinct roles that EP receptors play in the GI tract. More research is also needed to define the activity and the toxicity of single/combination EP receptor functions under normal and inflamed/diseased conditions. The long-term goal of future research in EP receptor biology should aim at identifying the Achilles’ heel at the receptor level as well as at the signalling that it propagates. This would help in effective pharmacological intervention of various GI ailments whose pathologies are mediated by PGE2.
Research in Dr Chadee's laboratory is supported by grants from the Crohn's and Colitis Foundation of Canada, the Canadian Institute for Health Research, the Canadian Foundation for Innovation, the Natural Sciences and Engineering Research Council of Canada and the Canadian Association of Gastroenterology-Industry-CIHR Research and Fellowship Awards. Dr Chadee holds a Canada Research Chair in Gastrointestinal Inflammation.