Prostaglandins and Inflammation – PMC – PubMed Central (PMC)

Arterioscler Thromb Vasc Biol. Author manuscript; available in PMC 2012 May 1.

Published in final edited form as:

PMCID: PMC3081099

NIHMSID: NIHMS271399

Institute for Translational Medicine and Therapeutics, University of Pennsylvania, Philadelphia, Pa

Prostaglandins are lipid autacoids derived from arachidonic acid. They both sustain homeostatic functions and mediate pathogenic mechanisms, including the inflammatory response. They are generated from arachidonate by the action of cyclooxygenase (COX) isoenzymes and their biosynthesis is blocked by nonsteroidal anti-inflammatory drugs (NSAIDs), including those selective for inhibition of COX-2. Despite the clinical efficacy of NSAIDs, prostaglandins may function in both the promotion and resolution of inflammation.

This review summarizes insights into the mechanisms of prostaglandin generation and the roles of individual mediators and their receptors in modulating the inflammatory response. Prostaglandin biology has potential clinical relevance for atherosclerosis, the response to vascular injury and aortic aneurysm.

Inflammation is the immune systems response to infection and injury and has been implicated in the pathogeneses of arthritis, cancer and stroke, as well as in neurodegenerative and cardiovascular disease. Inflammation is an intrinsically beneficial event that leads to removal of offending factors and restoration of tissue structure and physiological function. The acute phase of inflammation is characterized by the rapid influx of blood granulocytes, typically neutrophils, followed swiftly by monocytes that mature into inflammatory macrophages that subsequently proliferate and thereby affect the functions of resident tissue macrophages. This process causes the cardinal signs of acute inflammation: rubor (redness), calor (heat), tumor (swelling) and dolor (pain). Once the initiating noxious stimulus is removed via phagocytosis, the inflammatory reaction can decrease and resolve. During the resolution of inflammation, granulocytes are eliminated and macrophages and lymphocytes return to normal pre-inflammatory numbers and phenotypes. The usual outcome of the acute inflammatory program is successful resolution and repair of tissue damage, rather than persistence and dysfunction of the inflammatory response, which can lead to scarring and loss of organ function. It may be anticipated, therefore, that failure of acute inflammation to resolve may predispose to auto-immunity, chronic dysplastic inflammation and excessive tissue damage (1).

Prostaglandins play a key role in the generation of the inflammatory response. Their biosynthesis is significantly increased in inflamed tissue and they contribute to the development of the cardinal signs of acute inflammation. While the pro-inflammatory properties of individual prostaglandins during the acute inflammatory response are well established, their role in the resolution of inflammation is more controversial.

In this review, we will discuss the biosynthesis of and response to prostaglandins and the pharmacology of their blockade in orchestrating the inflammatory response, with particular regard to cardiovascular disease.

Prostaglandins and thromboxane A2 (TXA2), collectively termed prostanoids, are formed when arachidonic acid (AA), a 20-carbon unsaturated fatty acid, is released from the plasma membrane by phospholipases (PLAs) and metabolized by the sequential actions of prostaglandin G/H synthase, or cyclooxygenase (COX), and respective synthases.

There are four principal bioactive prostaglandins generated in vivo: prostaglandin (PG) E2 (PGE2), prostacyclin (PGI2), prostaglandin D2 (PGD2) and prostaglandin F2 (PGF2).They are ubiquitously produced usually each cell type generates one or two dominant products - and act as autacrine and paracrine lipid mediators to maintain local homeostasis in the body. During an inflammatory response, both the level and the profile of prostaglandin production changes dramatically. Prostaglandin production is generally very low in uninflamed tissues, but increases immediately in acute inflammation prior to the recruitment of leukocytes and the infiltration of immune cells.

Prostaglandin production () depends on the activity of prostaglandin G/H synthases, colloquially known as COXs, bifunctional enzymes that contain both cyclooxygenase and peroxidase activity and which exist as distinct isoforms referred to as COX-1 and COX-2 (2). COX-1, expressed constitutively in most cells, is the dominant source of prostanoids that subserve housekeeping functions, such as gastric epithelial cytoprotection and homeostasis (3). COX-2, induced by inflammatory stimuli, hormones and growth factors, is the more important source of prostanoid formation in inflammation and in proliferative diseases, such as cancer (3). However, both enzymes contribute to the generation of autoregulatory and homeostatic prostanoids, and both can contribute to prostanoid release during inflammation.

Biosynthetic pathway of prostanoids.

PGH2 is produced by both COX isoforms and it is the common substrate for a series of specific isomerase and synthase enzymes that produce PGE2, PGI2, PGD2, PGF2 and TXA2. COX-1 couples preferentially, but not exclusively, with thromboxane synthase (TXS), prostaglandin F synthase, and the cytosolic (c) prostaglandin E synthase (PGES) isozymes (4). COX-2 prefers prostaglandin I synthase (PGIS) and the microsomal (m) PGES isozymes, both of which are often coinduced along with COX-2 by cytokines and tumor promoters (4).

The profile of prostanoid production is determined by the differential expression of these enzymes within cells present at sites of inflammation. For example, mast cells predominantly generate PGD2 while macrophages produce PGE2 and TXA2 (5). In addition, alterations in the profile of prostanoid synthesis can occur upon cellular activation. While resting macrophages produce TXA2 in excess of PGE2, this ratio changes to favor PGE2 production after bacterial lipopolysaccharide (LPS) activation (5).

Prostaglandins exert their effects by activating rhodopsin-like seven transmembrane spanning G protein-coupled receptors (GPCRs; ). The prostanoid receptor subfamily is comprised of eight members: EP1 (E prostanoid receptor 1), EP2, EP3, and EP4 subtypes of the PGE receptor, PGD receptor (DP1), PGF receptor (FP), PGI receptor (IP), and TX receptor (TP) (6). Two additional isoforms of the human TP, (TP, TP) and FP (FPA FPB) and eight EP3 variants are generated through alternative splicing, which differ only in their C-terminal tails (7). In addition, there is another GPCR termed chemoattractant receptor-homologous molecule expressed on T helper 2 (TH2) cells (CRTH2 or DP2) that responds to PGD2 but belongs to the family of chemokine receptors (8). CRTH2 is a member of the fMLP chemoattractant receptor superfamily ().

Figure modified with permission from Shimizu (181).

Signal transduction of Prostanoid Receptors

Prostanoid receptors couple to a range of intracellular signaling pathways that mediate the effects of receptor activation on cell function. EP2, EP4, IP, and DP1 receptors activate adenylyl cyclase via Gs, increasing intracellular cyclic adenosine monophosphate (cAMP). EP1 and FP activate phosphatidylinositol metabolism via Gq, leading to the formation of inositol trisphosphate with mobilization of intracellular free calcium. In addition to signaling through Gq, the FP receptor couples to the small G-protein Rho via a Gq-independent mechanism (9). TP couples mainly to two types of G-proteins, the Gq (Gq, G11 G15, G16) and the G13 (G12, G13) families, resulting in the activation of phospholipase C and guanine nucleotide exchange factor of the small G protein Rho (RhoGEF), respectively. Additionally, TP can also be coupled via Gh to phospholipase C as well as via Gi and Gs to adenylate cyclase. Both TP isoforms are coupled to phospholipase C activation, but TP stimulates adenylyl cyclase, whereas TP inhibits it. EP3 isoforms can couple via Gi or G12 to elevation of intracellular Ca2+, inhibition of cAMP generation, and activation of the small G protein Rho (10). The DP2/CRTH2 couples to a Gi-type G protein to inhibit cAMP synthesis and elevate intracellular Ca2+. However, the effects of prostanoids on these G proteincoupled signaling pathways may change as a function of ligand concentration or structure (6).

Some, but not all prostanoid receptors exhibit the capacity to dimerize which may alter ligand affinity and/or preference for downstream signaling pathways. Thus, while DP1/DP2 dimers appear not to form, under similar conditions, IP and TP receptors can associate to form homo- and heterodimers. TP-TP heterodimers enhance the response to activation by free radicalcatalyzed isoprostanes (11). Furthermore, dimerization of IP and TP enables cAMP formation through TP receptor activation, a cellular outcome typically observed with IP activity (12). Moreover EP1 receptor activation has been shown to modulate 2- adrenoreceptor function in bronchial airways via formation of a heterodimeric complex (13).

The two cyclooxygenase isoforms, COX-1 and COX-2, are targets of nonsteroidal anti-inflammatory drugs (NSAIDs). These drugs are competitive active site inhibitors of both COXs. Although both COXs exist as homodimers, only one partner is used at a time for substrate binding (14). COX-1/COX-2 heterodimers may also exist, but their role in biology remains to be established (15). NSAIDs bind to and inactivate the COX site at only one of the monomers of the COX dimer and this is sufficient to shut down prostanoid formation (14). The other monomer appears to play an allosteric function. The peroxidase capacity of both proteins is unaltered by NSAIDs.

The clinical efficacy of structurally distinct NSAIDs, all of which share this capacity for prostanoid inhibition, points to the importance of these mediators in the promotion of pain, fever and inflammation (16). The dramatic increase of COX-2 expression upon provocation of inflammatory cells, its expression in inflamed tissues and the assumption that inhibition of COX-1 derived prostanoids in platelets and gastric epithelium explains NSAID evoked gastrointestinal adverse effects and provided a rationale for development of NSAIDs designed to be selective for inhibition of COX-2 for treating arthritis and other chronic inflammatory diseases (17).

Although COX-2 appears to be the dominant source of prostaglandin formation in inflammation, there is some suggestion that both isoforms of the human enzyme may contribute to the acute inflammatory response. COX-1 is constitutively expressed in resident inflammatory cells and there is evidence for induction of COX-1 during LPS-mediated inflammatory response and cellular differentiation (18). Both COX isoforms are co-expressed in circulating inflammatory cells ex vivo and in both inflamed rheumatoid arthritis (RA) synovium and in atherosclerotic plaques obtained from patients (19, 20). Human data are compatible with COX-1-derived products driving the initial phase of an acute inflammation, with COX-2 upregulation occurring within several hours (4). However, controlled clinical trials to test the comparative efficacy of NSAIDs that inhibited both COXs versus COX-2 alone were never performed at scale. Such trials were designed to seek divergence in the incidence of gastrointestinal adverse effects rather than to assess comparative clinical efficacy.

Studies in both COX-1- and COX-2- knockout (KO) mice reveal impaired inflammatory responses, although the impacts of gene deletion diverge in time course and intensity. Mice deficient in COX-1, but not COX-2, exhibit a reduction in AA-induced ear edema although AA induces an equivalent inflammatory response in wild-type (WT) and COX-2-deficient mice (2123). By contrast, the level of edema induced by the tumor promoter tetradecanoyl phorbol acetate (TPA) was not significantly different between WT, COX-1-deficient and COX-2-deficient mice (2123). The ear inflammation studies indicate that COX-1, as well as COX-2, may contribute to inflammatory responses and the isoform responsible for the inflammation may depend on the type of inflammatory stimulus and/or the relative levels of each isoform in the target tissue.

Similarly, arthritis models exhibit a significant reliance on either the COX-1 or COX-2 isoforms for the development of clinical synovitis. This varies, depending on the experimental model used. Thus, in the K/BxN serumtransfer model of arthritis, COX-1 derived prostaglandins, in particular PGI2, make a striking contribution to the initiation and perpetuation of arthritis (24). In a collagen- induced arthritis (CIA) model, COX-2- deletion suppresses considerably synovial inflammation and joint destruction, whereas arthritis in COX- 1deficient mice is indistinguishable from controls (25, 26).

COX-2 deletion also suppresses acute inflammation in the air pouch model. Here, the COX-2 inhibitor, NS-398, administered six hours after carageenan treatment, reduced PG production in wild-type mice to levels comparable to those seen in COX-2-KO mice, and was also effective during the early stages of inflammation (27). When compared to wild-type mice, the deficiency of COX-2 reduces the level of PGE2 production by approximately 75%, while the deficiency of COX-1 reduces the PGE2 level by 25% during this early stage. By day 7 following carrageenan treatment, higher numbers of inflammatory cells were present in the pouch fluid of COX-2-KO mice, and little resolution of inflammation was apparent compared to wild-type or COX-1-KO mice (27). These findings indicate that both COX isoforms contribute to PG production during inflammation and also that COX-2-derived PGs appear to be important in both the acute inflammatory process and in the resolution phase. A contribution of COX-2 to both phases of inflammation was also reported in other models. Gilroy et al. reported that COX-2 expression and PGE2 levels increased transiently early in the course of carrageenan-induced pleurisy in rats (28). Later in the response, COX-2 was induced again to even greater levels and generated anti-inflammatory prostaglandins, such as PGD2 and 15-deoxy-1214-PGJ2 (15d-PGJ2) but only low levels of the proinflammatory PGE2. Further support for an anti-inflammatory role of COX-2 in this model was the finding that late administration of the COX-2selective inhibitor, NS-398, exacerbates the inflammatory response. Furthermore, Wallace et al. observed that in the paw carrageenan model, the resultant inflammation resolves within 7 days in wild-type mice but is unaltered over this period in COX-2deficient mice (29). An anti-inflammatory role of COX-2 derived prostanoids has also been reported in models of inflammatory colitis and allergic airway disease (30, 31). Thus, COX-2 appears to have a dual role in the inflammatory process, initially contributing to the onset of inflammation and later helping to resolve the process. While COX-2 does play a role in supporting resolution of this process in some models of inflammation, it is unclear which products of the enzyme might contribute in which settings to this effect. This is exemplified by the case of 15d-PGJ2. Long touted as an endogenous ligand to peroxisome proliferator-activated receptor gamma (PPAR), it remains to be established that endogenous concentration sufficient to subserve this function are formed in any model of resolving inflammation. Erroneously elevated levels have been reported using a variety of immunoassays, but the concentrations of bound and free compound documented to be formed by quantitative mass spectrometry fall far short of the EC50 for PPAR activation (32). It is one thing to show that putative pro-resolution products can be formed in vitro and that the synthetic compounds do exert pro- resolution actions when administered in vivo and another to document that the concentrations formed in vivo in the setting of inflammation are sufficient and necessary to mediate resolution.

In the case of atherosclerosis, deletion or inhibition of COX-2 has been shown variously to retard, accelerate or leave unaltered atherogenesis in mouse models (5). This may reflect an impact postnatally of disruption of the many roles of the enzyme in development in COX-2 KOs; differences in timing of interventions with COX-2 inhibitors and a failure in most cases to define biochemically the selectivity for inhibition of COX-2 of the drug regimen deployed. In contrast, COX-1 deletion markedly attenuated lesion development in the Apolipoprotein E (ApoE) KO mouse, as does inhibition of COX-1 and COX-2 together in low-density lipoprotein receptor (LDLR)- KO model (33, 34). Thus, products of COX-1 like TxA2 promote atherogenesis while there is more ambiguity around the role of COX-2. Nevertheless, deletion of the IP, the receptor for the major COX-2 product, PGI2, fosters the initiation and early development of atheroscelrosis in hyperlipidemic mice (35).

PGE2 is one of the most abundant PGs produced in the body, is most widely characterized in animal species, and exhibits versatile biological activities. Under physiological conditions, PGE2 is an important mediator of many biological functions, such as regulation of immune responses, blood pressure, gastrointestinal integrity, and fertility. Dysregulated PGE2 synthesis or degradation has been associated with a wide range of pathological conditions (36). In inflammation, PGE2 is of particular interest because it is involved in all processes leading to the classic signs of inflammation: redness, swelling and pain (37). Redness and edema result from increased blood flow into the inflamed tissue through PGE2-mediated augmentation of arterial dilatation and increased microvascular permeability (37). Pain results from the action of PGE2 on peripheral sensory neurons and on central sites within the spinal cord and the brain (37). PGE2 is synthesized from PGH2 by cPGES or mPGES-1 and mPGES-2 (38). cPGES is constitutively and abundantly expressed in the cytosol of various tissues and cells and it requires glutathione (GSH) as a cofactor (39). The role of cPGES and even its ability to form prostaglandin E2 is controversial. cPGES seems capable of converting COX-1-, but not COX-2-derived PGH2 to PGE2 in cells, particularly during the immediate PGE2-biosynthetic response elicited by Ca2+ evoked stimuli. Localization of cPGES in the cytosol may allow coupling with proximal COX-1 in the endoplasmic reticulum (ER) in preference to distal COX-2 in the perinuclear envelope (39). Functional coupling of cPGES with COX-1 suggests that the functions of cPGES in vivo overlap significantly, if not entirely, with COX-1. CPGES-deficient mice were developed, but have not been particularly informative in addressing the importance of cPGES-derived PGE2, because deletion of this enzyme results in perinatal lethality (40).

mPGES-1 is a member of the MAPEG (membrane-associated proteins involved in eicosanoid and glutathione metabolism) superfamily and like cPGES, it requires GSH as cofactor (41). mPGES-1 is a perinuclear protein that is markedly induced by cytokines and growth factors and downregulated by anti-inflammatory glucocorticoids as in the case of COX-2 (42, 43). It is functionally coupled with COX-2 in marked preference to COX-1 (44). Constitutive expression of mPGES-1 in certain tissues and cell types was also reported.

The generation of mPGES-1-deficient mice has revealed the dominant role of this enzyme in PGE2 generation relevant to promotion of inflammation. In CIA, a disease model of human RA, mPGES-1 null mice exhibited a reduced incidence and severity of disease compared with wild-type controls (45). This difference was not associated with alterations in interleukin (IL)-6 production by peritoneal macrophages or significant differences in circulating IgG2a anticollagen antibodies. Likewise, in collagen-antibody-induced arthritis (CAIA), another model of RA that does not involve the activation of the immune system, mPGES-1 null mice had a similar incidence, but a lesser severity of arthritis than wild-type mice, as well as a 50% reduction in paw levels of PGE2 (46). In the same study, it was also observed that the migration of macrophages following peritoneal injection of thioglycollate was strikingly reduced in mPGES-1-null mice relative to wild-type mice (46).

The formation of inflammatory granulation tissue and attendant angiogenesis in the dorsum induced by subcutaneous implantation on the paw of a cotton thread was significantly reduced in mPGES-1 KO mice as compared with WT mice (46). In this model, mPGES-1 deficiency was also associated with reduced induction of vascular endothelial cell growth factor (VEGF) in the granulation tissue. These results indicate that mPGES-1-derived PGE2, in cooperation with VEGF, may play a critical role in the development of inflammatory granulation and angiogenesis, thus eventually contributing to tissue remodeling.

Together, these findings illustrate that deletion or inhibition of mPGES- 1 markedly reduces inflammatory response, in several mouse models. The pro-inflammatory effect of mPGES-1-derived PGE2 was also observed in atherosclerosis. Deletion of mPGES-1 retarded atherogenesis in fat-fed hyperlipidaemic mice, in both sexes (47). Interestingly, in addition to the expected depression of PGE2 production, deletion of mPGES-1 permits rediversion of the PGH2 substrate to other PG synthases, with, for example, augmented formation of PGI2 and PGD2 (47). This complicates the selection of mPGES-1 as a drug target. Thus, elevated PGI2 may contribute to the more benign cardiovascular profile of mPGES-1 deletion compared to COX-2 deletion or inhibition- less predisposition to hypertension and thrombogenesis. However, this same effect may attenuate relief of pain, in the case of augmented PGI2, or may mediate adverse effects on bronchial tone, allergic inflammatory disease or the sleepwakefulness cycle, in case of elevated PGD2. This may account for the less impressive effectiveness of disrupting mPGES-1 vs COX-2 in some models of pain (48). Recently Geisslingers group has provided evidence that mPGES-1 derived PGE2 may contribute both to promotion and resolution of neuroinflammation in mice (49). Finally, the major substrate products of rediversion will be influenced by the dominant cell type in a particular setting. For example, augmented PGI2 may contribute to the restraint of atherogenesis in hyperlipidemic mice consequent to mPGES-1 deletion (47). However, it remains to be seen whether mPGES-1 inhibition in the setting of established atherosclerosis causes regression or possibly accelerates further progression of disease, due to endoperoxide rediversion to TxA2 in macrophage rich plaques.

mPGES-2 is synthesized as a Golgi membrane-associated protein, and the proteolytic removal of the N-terminal hydrophobic domain leads to the formation of a mature cytosolic enzyme. This enzyme is constitutively expressed in various cells and tissues and is functionally coupled with both COX-1 and COX-2 (37). mPGES-2-deficient mice showed no specific phenotype and no alteration in PGE2 levels in several tissues or in LPS-stimulated macrophages (50). Studies with PGES null mice have revealed that cross regulation between the different PGES isoforms may function, on occasion, as a compensatory mechanism. For example, mPGES-1-null mice exhibit a delayed increase in urinary PGE2 excretion in response to acute water loading, coincident with enhanced renal medullary expression of cPGES but not of mPGES-2 (51). Similar evidence suggestive of cross regulation was observed with the COXs. Thus, deletion of COX-2 in macrophages is associated with upregulated expression of COX-2 in vascular smooth muscle cells in atherosclerotic plaque (52).

After PGE2 is formed, it is actively transported through the membrane by the ATP-dependent multidrug resistance protein-4 (MRP4) or diffuses across the plasma-membrane to act at or nearby its site of secretion (53).

PGE2 then acts locally through binding of one or more of its four cognate receptors (), termed EP1EP4 (45). Among the four EPs, EP3 and EP4 receptors are the most widely distributed with their mRNAs being expressed in almost all mouse tissues and have the highest affinity for binding PGE2. In contrast, the distribution of EP1 mRNA is restricted to several organs, such as the kidney, lung, and stomach, and EP2 is the least abundant of the EP receptors. Both EP1 and EP2 bind PGE2 with lower affinity (54). Each EP subtype shows a distinct cellular localization within tissues (54).

One of the lessons learned from the knock-out mouse studies is that PGE2 can exert both pro-inflammatory and anti-inflammatory responses, and these actions are often produced through regulation of receptor gene expression in relevant tissues. For example, hyperalgesia, a classic sign of inflammation, is mediated mainly by PGE2 through EP1 receptor signaling that acts on peripheral sensory neurons at the site of inflammation, as well as on central neuronal sites (55). Other studies have also implicated EP3 receptor in the inflammatory pain response mediated by low doses of PGE2 (56).

EP2 and EP4 redundantly mediate development of paw swelling associated with collagen-induced arthritis (57). Likewise, studies of carrageenin induced paw edema and carrageenin-induced pleurisy both revealed participation of EP2 and EP3 in inflammatory exudation (58). The EP4 receptor also appears to play a pro-inflammatory role in the pathogenesis of rheumatoid arthritis. PGE2 produced by rheumatoid synovium has been implicated in IL-6 production and joint destruction (59, 60). Mice deficient in the EP4 but not in the EP1, EP2, or EP3 receptors exhibit an attenuated response in the collagen antibody induced arthritis model, with significantly lower levels of the inflammatory cytokines IL-6 and IL-1 and a dramatic reduction in the clinical signs of disease (61). Anti-inflammatory actions of prostaglandins are seen typically in allergic or immune inflammation and are usually balanced by pro-inflammatory actions of other prostaglandins. Such contrasting biology is evident between the PGI2-IP and TxA2-TP pathways in cardiovascular disease and between the PGD2-DP and the PGE2-EP3 pathways in elicitation of allergic asthma (62, 63).

PGE2, binding to different EP receptors, can regulate the function of many cell types including macrophages, dendritic cells and T and B lymphocytes leading to both pro- and anti-inflammatory effects. As pro-inflammatory mediator, PGE2 contributes to the regulation of the cytokine expression profile of dendritic cells (DC)s and has been reported to bias T cell differentiation towards a T helper (Th) 1 or Th2 response (35). A recent study showed that PGE2-EP4 signaling in DCs and T cells facilitates Th1 and IL-23-dependent Th17 differentiation (64). Additionally, PGE2 is fundamental to induction of a migratory DC phenotype permitting their homing to draining lymph nodes (65, 66). Simultaneously, PGE2 stimulation early during maturation induces the expression of costimulatory molecules of the TNF superfamily on DCs resulting in an enhanced T cell activation (67). In contrast, PGE2 has also been demonstrated to suppress Th1 differentiation, B cell functions and allergic reactions (68). Moreover, PGE2 can exert anti-inflammatory actions on innate immune cells like neutrophils, monocytes and NK cells (68).

PGE2 can thus modulate various steps of inflammation in a context-dependent manner and coordinate the whole process in both pro-inflammatory and anti-inflammatory directions. This dual role of PGE2 and its receptors in modulating the inflammatory response has been observed in several disorders. In atherosclerosis, EP4 deficiency promotes macrophage apoptosis and suppresses early atherosclerosis in LDLR/ mice chimeric for EP4/ in hematopoietic cells after 8 weeks on a Western diet (69). EP2 deficiency in hematopoietic cells revealed a trend for similar, but modest, effects on atherosclerosis (69). In the same study macrophage EP4 appeared to play a proinflammatory role in the early stages of atherosclerosis by regulating production of inflammatory cytokines such as IL-1, IL-6 and monocyte chemotactic protein-1 (MCP-1) (69). In contrast, EP4 deletion in bone marrow-derived cells enhanced local inflammation (increased expression of chemotactic proteins, including MCP-1 and IL-10; and increased inflammatory cells, such as macrophages and T cells) and altered lesion composition (increased smooth muscle cells within plaque) but did not alter plaque size or morphology in established atherosclerosis (after 10 weeks of high-fat diet) (70).

PGE2 also plays contrasting roles during neuroinflammation. LPS-induced PGE2 synthesis causes deleterious effects in neurons resulting in lesions or enhanced pain transmission (7173). However, PGE2 also has anti-inflammatory properties. It mediates bradykinin-induced neuroprotection and blocks LPS and ATP-induced cytokine synthesis in cultured microglia and in neuron-glia cocultures (74, 75). The anti-inflammatory and neuroprotective effects of PGE2 are mediated via microglial EP2- and EP4- receptors. Recently, it has been reported that PGE2 limits cytokine and prostaglandin synthesis mainly through EP2 activation in a model of LPS-induced neuroinflammation and that mPGES-1 is a critical enzyme in this negative feedback regulation (49).

PGI2 is one of the most important prostanoids that regulates cardiovascular homeostasis. Vascular cells, including endothelial cells, vascular smooth muscle cells (VSMCs) and endothelial progenitor cells (EPCs), are the major source of PGI2 (76).

It is generated by the sequential action of COX and PGIS, a member of the cytochrome P450 superfamily that specifically converts PGH2 to PGI2. PGIS co-localizes with COX in the ER, plasma membrane and nuclear membrane (77). PGIS is constitutively expressed in endothelial cells where it couples with COX-1 (78), although COX-2-dependent PGI2 production by endothelial cells has been reported to be modulated in vitro by thrombin, shear stress, oxidized LDL, hypoxia and inflammatory cytokines and it is synchronized by up-regulation of COX-2 (79, 80). In vivo studies in mice and humans showed that COX-2 was the dominant source of PGI2 (81).

Once generated, PGI2 is released to act upon neighboring VSMCs as well as circulating platelets. Indeed, PGI2 exerts its effects locally, is not stored, and is rapidly converted by non-enzymatic processes to an inactive hydrolysis product, 6-keto-PGF1 (82).

PGI2 is a potent vasodilator, and an inhibitor of platelet aggregation, leukocyte adhesion, and VSMC proliferation (75). PGI2 is also anti-mitogenic and inhibits DNA synthesis in VSMC (83). These actions of PGI2 are mediated through specific IP receptors (). This receptor is expressed in kidney, liver, lung, platelets, heart, and aorta (84). There is inconclusive evidence that some effects of PGI2 on the vasculature might be mediated by the PPAR pathway, in addition to the classical IP cAMP signaling pathway (85). PGI2 can indeed activate PPAR , however, just like 15- dPGD2 and PPAR , it is unclear that it represents a biological target at concentrations of the ligand attained in vivo (86).

Although IP-deficient mice mature normally without suffering from spontaneous thrombosis, both the response to thrombogenic stimuli and VSMC proliferation in response to vascular injury are enhanced compared with control littermates (87). Mice lacking IP are also sensitive to dietary salt induced hypertension (88) and exhibit accelerated atherogenesis with enhanced platelet activation and increased adhesion of leukocytes on the vessel walls in both the LDL-receptor and ApoE KO models (35, 89).

In addition to its cardiovascular effects, PGI2 is an important mediator of the edema and pain that accompany acute inflammation. PGI2 is rapidly produced following tissue injury or inflammation and it is present at high concentrations in inflammatory milieus (90). PGI2 is the most abundant prostanoid in synovial fluid in human arthritic knee joints as well as in peritoneal cavity fluid from mice injected with irritants (91, 92). In IP receptor deficient mice, potentiation of bradykinin-induced microvascular permeability by PGI2 is abolished; in addition, these mice have substantially reduced carrageenan- induced paw edema (93). The level of paw edema observed in IP-deficient mice was equivalent to that of indomethacin treated controls and indomethacin treatment of IP-deficient animals did not induce a further decrease in swelling, indicating that PGI2 -IP receptor signaling is the major prostanoid pathway that mediates the acute inflammatory response in this model. It has also been suggested that bradykinin induces PGI2 formation leading to enhancement of microvascular permeability and edema (94). The IP receptor has been shown to mediate nociceptive pain during acute inflammation. IP receptor mRNA is present in dorsal root ganglion neurons including those that express substance P, a marker for nociceptive sensory neurons (95). IP receptor-deficient mice have an attenuated writhing response following intraperitoneal injection of either acetic acid or PGI2, indicating that the IP receptor plays a role in mediating peripheral nociceptive sensitization to inflammatory stimuli (93). The IP receptor is also expressed in the spinal cord and has been implicated in spinal pain transmission in response to peripheral inflammation (96). IP antagonists were shown to reduce pain responses in several models, including acetic acid-induced abdominal constriction, mechanical hyperalgesia produced by carrageenan and pain associated with models of osteoarthritis and inflammatory arthritis (97, 98). In contrast to the pro-inflammatory effects of IP receptor activation in nonallergic acute inflammation, some studies have suggested that IP receptor signaling suppresses Th2- mediated allergic inflammatory responses (99). IP receptor mRNA is up-regulated in CD4+ Th2 cells and inhibition of PGI2 formation by the COX-2 inhibitor, NS-398, during antigen-induced airway inflammation results in greater lung Th2-mediated lung inflammation (99). PGI2 has been suggested to exert this effect in part by enhancing Th2 cell production of the anti- inflammatory cytokine IL-10. This immunosuppressive role for the IP receptor in Th2-mediated inflammation is supported by the observation that in ovalbumin (OVA)-induced asthma, IP deletion results in increased antigen-induced leukocyte accumulation in bronchoalveolar lavage fluid and peribronchiolar and perivascular inflammatory infiltration (100). Thus, PGI2 may shift the balance within the immune system away from a Th2 dominant response and inhibit allergic inflammation.

PGD2 is a major eicosanoid that is synthesized in both the central nervous system (CNS) and peripheral tissues and appears to function in both an inflammatory and homeostatic capacity (101). In the brain, PGD2 is involved in the regulation of sleep and other CNS activities, which includes pain perception (102, 103). In peripheral tissues, PGD2 is produced mainly by mast cells, but also by other leukocytes, such as DCs and Th2 cells (104106). Two genetically distinct PGD2-synthesizing enzymes have been identified, including hematopoietic- and lipocalin type PGD synthases (H-PGDS and L-PGDS, respectively). H-PGDS is generally localized to the cytosolic of immune and inflammatory cells, whereas L-PGDS is more resigned to tissue-based expression (107).

PGD2 can be further metabolized to PGF2, 9,11-PGF2 (the stereoisomer of PGF2) and the J series of cyclopentanone PGs, including PGJ2, 12-PGJ2, and 15d-PGJ2 (108). Synthesis of J series PGs involves PGD2 undergoing an initial dehydration reaction to produce PGJ2 and 15d-PGJ2, after which PGJ2 is converted to 15d-PGJ2 and 12-PGJ2 via albumin-dependent and albumin-independent reactions, respectively (109).

PGD2 activity is principally mediated through DP or DP1 and CRTH2 or DP2, as described previously (). Also 15d-PGJ2, binds with low affinity the nuclear PPAR (110). PGD2 has long been associated with inflammatory and atopic conditions, although it might exert an array of immunologically relevant anti-inflammatory functions as well.

PGD2 is the predominant prostanoid produced by activated mast cells, which initiate IgE-mediated Type I acute allergic responses (104, 111). It is well established that the presence of an allergen triggers the production of PGD2 in sensitized individuals. In asthmatics, PGD2 which can be detected in the bronchoalveolar lavage fluid within minutes, reaches biologically active levels at least 150-fold higher than pre-allergen levels (112). PGD2 is produced also by other immune cells such as antigen-presenting dendritic cells and Th2 cells, suggesting a modulatory role for PGD2 in the development of antigen-specific immune system responses (104, 105). PGD2 challenge elicits several hallmarks of allergic asthma such as bronchoconstriction and airway eosinophil infiltration (113, 114), and mice that overexpress lipocalin-PGD synthase have elevated PGD2 levels and an increased allergic response in the OVA-induced model of airway hyperreactivity (115).

The pro-inflammatory effects of PGD2 appear to be mediated by both DP1 and DP2/CRTH2 receptors. Because both receptors bind PGD2 with similar high affinity, PGD2 produced by activated mast cells or T cells would be capable of activating multiple signaling pathways leading to different effects, depending on whether the DP1 or DP2/CRTH2 receptors or both are locally expressed.

The DP1 receptor is expressed on bronchial epithelium and has been proposed to mediate production of chemokines and cytokines that recruit inflammatory lymphocytes and eosinophils, leading to airway inflammation and hyperreactivity seen in asthma (116). Mice deficient in DP1 exhibit reduced airway hypersensitivity and Th2-mediated lung inflammation in the OVA-induced asthma model, suggesting that the DP1 plays a key role in mediating the effects of PGD2 released by mast cells during an asthmatic response (62). Furthermore, PGD2 may inhibit eosinophil apoptosis via the DP1 receptor (117).

DP1 antagonists exert anti-inflammatory properties in several experimental models including inhibition of antigen-induced conjunctival microvascular permeability in guinea pigs and OVA-induced airway hyperreactivity in mice (118, 119).

DP2/CRTH2 receptors contribute largely to pathogenic responses by mediating inflammatory cell trafficking and by modulating effector functions. PGD2 released from mast cells may mediate recruitment of Th2 lymphocytes and eosinophils directly via the DP2/CRTH2 receptor. In humans, the DP2/CRTH2 receptor is expressed on Th2 lymphocytes, eosinophils, and basophils (8, 120, 121), and an increase in DP2/CRTH2+ T-cells has been positively associated with certain forms of atopic dermatitis (122). The DP2/CRTH2 receptor has been demonstrated to mediate PGD2-stimulated chemotaxis of these cells in vitro and leukocyte mobilization in vivo (123).

In contrast to the pro-inflammatory role of PGD2 in allergic inflammation, PGD2 may act to inhibit inflammation in other contexts. The DP1 receptor is expressed on dendritic cells that play a key role in initiating an adaptive immune response to foreign antigens. PGD2 activation of the DP1 receptor inhibits dendritic cell migration from lung to lymph nodes following OVA challenge, leading to reduced proliferation and cytokine production by antigen specific T cells (124). DP1 activation also reduces eosinophilia in allergic inflammation in mice and mediates inhibition of antigen-presenting Langerhans cell function by PGD2 (125, 126). As mentioned, PGD2 and its degradation product, 15d-PGJ2 have been suggested as the COX-2 products involved in the resolution of inflammation (28, 127). Administration of a COX-2 inhibitor during the resolution phase exacerbated inflammation in a carageenan-induced pleurisy model (28). In a zymosan-induced peritonitis model, deletion of H-PGDS induced a more aggressive inflammatory response and compromised resolution, findings that were moderated by addition of a DP1 agonist or 15d-PGJ2 (123). While, these data appear to implicate PGD2 and 15d-PGJ2 in resolution, there is a large disparity between the nanomolar to picomolar amounts of 15d-PGJ2 detected by physicochemical methodology in in vivo settings and the amount needed to have a biologic effect in vitro on PPAR or NFB (micromolar amounts) (32, 128, 129). This discrepancy is supported by recent data reported in zymosan-induced peritonitis, where we observed evoked biosynthesis of PGD2 only during the proinflammatory phase, but not during resolution. Consistent with this observation, DP2/CRTH2 deletion reduced the severity of acute inflammation, but neither DP1 or DP2/CRTH2 deletion impacted resolution. Although 15d-PGJ2 is readily detected when synthetic PGD2 was infused into rodents (130), endogenous biosynthesis of 15d-PGJ2 was not detected during promotion or resolution of inflammation (Mehta J et al., 2010; unpublished data).

PGD2 may play a role in the evolution of atherosclerosis. In the context of inflamed intima, PGD2 in part derives from H-PGDS-producing inflammatory cells that are chemotactically compelled to permeate the vasculature (131, 132). Additionally, L-PGDS expression is induced by laminar sheer stress in vascular endothelial cells and is actively expressed in synthetic smooth muscle cells of atherosclerotic intima and coronary plaques of arteries with severe stenosis (133135). PGD2 has been shown to inhibit expression of pro-inflammatory genes, such as iNOS and plasminogen activator inhibitor (136, 137). Indeed, L-PGDS deficiency accelerates atherogenesis (138).

In summary, studies with COX-2 inhibitors suggest that products of this enzyme may play a role in resolution in several models of inflammation. However, it the identity of such products, whether formed directly by COX-2 or due to substrate diversion consequent to COX-2 inhibition, remains, in many cases, to be established.

PGF2 is synthesized from PGH2 via PGF synthase, and it acts via the FP, which couples with Gq protein to elevate the intracellular free calcium concentration (). Two differentially spliced variants of the sheep FP receptor ortholog have been reported: FPA and FPB, which differ from each other in the length of their C-terminal tails (139). The FP receptor is the least selective of the prostanoid receptors in binding the principal endogenous prostaglandins; both PGD2 and PGE2 ligate the FP with EC50s in the nanomolar range (140). PGF ring compounds can be formed as minor products from other prostaglandins. For example, enzymatic reduction of 9-keto group of PGE compounds by 9-ketoreductases results in either 9a-hydroxyl, yielding PGF compounds or more rarely, a 9b-hydroxyl, yields PGF compounds (141). PGF ring metabolites may also be formed from PGD ring compounds by 11-keto reductases (142). 15-Keto-dihydro-PGF2, a major stable metabolite of PGF2 that reflects in vivo PGF2 biosynthesis, is found in larger quantities first in the peripheral plasma and later on in the urine both in basal physiological conditions and in certain physiological and pathophysiological situations like acute and chronic inflammation (143).

PGF2, derived mainly from COX-1 in the female reproductive system, plays an important role in ovulation, luteolysis, contraction of uterine smooth muscle and initiation of parturition (144, 145). Recent studies have shown that PGF2 also plays a significant role in renal function (146), contraction of arteries (147), myocardial dysfunction (148, 149), brain injury (150) and pain (151). Analogs of PGF2 have previously been developed for estrus-synchronization and abortion in domestic animals (152, 153) and to influence human reproductive function (154). FP agonists are being widely used worldwide to reduce intraocular pressure in the treatment of glaucoma (155).

Administration of PGF2 leads to acute inflammation, and NSAIDs inhibit PGF2 biosynthesis both in vitro and in vivo (144). In models of acute inflammation evoked biosynthesis of PGF2 may coincide with free radical catalyzed generation of F ring isoprostanes, indices of lipid peroxidation (156, 157). The tachycardia induced in wild-type mice by injection of LPS is greatly attenuated in FP-deficient or TP-deficient mice and is completely absent in mice lacking both of these receptors (148). A recent study reported that deletion of FP selectively attenuates pulmonary fibrosis without a change in alveolar inflammation after microbial invasion (158). Elevated biosynthesis of PGF2 has been reported in patients suffering from rheumatoid arthritis, psoriatic arthritis, reactive arthritis, and osteoarthritis (159).

Cardiovascular risk factors such as diabetes, obesity, smoking, and thickening of intima-media ratio in the carotid artery have been variably associated with elevations in PGF2 metabolites, together with IL-6 and acute phase proteins in body fluids (160, 161). Deletion of the FP reduces blood pressure and retards the attendant atherogenesis in hyperlipidemic mice despite the absence of detectable FP in large blood vessels and their atherosclerotic plaques (162). PGF2 is the most abundant prostanoid formed by human umbilical cord endothelial cells in response to laminar shear stress that upregulates expression of COX-2 (163).

The emerging role of PGF2 in acute and chronic inflammation opens opportunities for the design of new anti-inflammatory drugs.

TXA2, an unstable AA metabolite with a half-life of about 30 seconds, is synthesized from PGH2 via TXS and it is non-enzymatically degraded into biologically inactive TXB2. TXA2 is predominantly derived from platelet COX-1, but can also be produced by other cell types, including by macrophage COX-2(164, 165).

TXA2 activity is principally mediated through the TP, which couples with Gq, G12/13 and multiple small G proteins which in turn regulate several effectors, including phospholipase C, RhoGEF and adenylyl cyclase (166, ). TP and TP, two spliced isoforms of TP in humans, communicate with different G proteins and undergo heterodimerization, resulting in changes in intracellular traffic and receptor protein conformations. Only the TP protein is expressed in mice.

TP activation mediates several physiological and pathophysiological responses, including platelet adhesion and aggregation, smooth muscle contraction and proliferation, and activation of endothelial inflammatory responses (167). TP function is regulated by several factors, such as oligomerization, desensitization and internalization, glycosylation, and cross-talk with receptor tyrosine kinases (167).

Although TXA2 is the preferential physiological ligand of the TP receptor, PGH2 particularly, also can activate this receptor (168). Additionally, isoprostanes (nonenzymatic free radical-catalyzed peroxidative products of polyunsaturated fatty acids) and hydroxyeicosatetraenoic acids (HETEs, generated by lipoxygenases and cytochrome P450 monooxygenases or formed by nonenzymatic lipid peroxidation) are also potent agonists at TP receptors (169, 170).

Epoxyeicosatrienoic acid (cytochrome P450 metabolites of AA) dihydro derivatives, in contrast, are selective endogenous antagonists of TP (171). However, whether PGH2, isoprostanes or HETEs significantly contribute to the responses attributed to TP activation in vivo is still to be investigated. For example, TP activation by isoprostanes may play an important role in clinical settings of oxidative stress, such during reperfusion after organ transplant.

TP deficient mice are normotensive but have blunted vascular responses to TP agonists and show a tendency to bleeding (172).

The deletion of TP decreases vascular proliferation and platelet activation in response to vascular injury, delays atherogenesis and prevents angiotensin II- and L-NAME-induced hypertension and the associated cardiac hypertrophy (87, 89, 173, 174).

In septic shock models, TP deletion or TP antagonism protected against various LPS-induced responses, such as the increase in inducible nitric oxide synthase expression, acute renal failure, and inflammatory tachycardia (175, 176), suggesting a potential role of TXA2 as pro-inflammatory mediator.

The phenotype of TXS deficient mice is much less pronounced, perhaps because TXA2 is only one of the endogenous ligands of TP and more likely because the deletion of this enzyme may redirect the PGH2 toward other countervailing synthases (177).

This review has described a stunning complexity of evidence about the role of prostanoids in inflammation. Different products have conflicting effects on both the promotion and resolution of inflammation. The same product formed by different enzymes COX-1 or COX-2 may either promote or resolve inflammation. Products of the same enzyme may promote or resolve inflammation in different models. Different cell types that predominate at varying stages of disease evolution generate prostanoids that have contrasting effects on inflammation. Individual prostanoids overlap considerably in their biological effects with other mediators.

These observations prompt several questions.

Given this complex array of biological effects mediated by prostanoids, how does their general inhibition, high up in the cascade, result in drugs that are reasonably well tolerated and reasonably effective? Aspirin and the myriad NSAIDs, including acetaminophen and those developed to inhibit selectively COX-2, are amongst the most commonly consumed drugs on the planet. Hard evidence with which to address this question is in short supply, but lets speculate. Aspirin at doses less than 100 mg per day or greater than 1 gm per day have equivalent effects on platelet COX-1 derived TxA2. They both suppress it completely. However, increasing daily doses of aspirin increasingly inhibit PGI2 coincidentally with this effect. We dont have direct randomized comparisons across doses, but indirect comparisons suggest that the cardioprotective efficacy of aspirin may be progressively attenuated as the dose is increased. Similarly, locally formed, COX-1 derived PGE2 and PGI2 are protective of gastroduodenal epithelial integrity and platelet COX-1 derived TxA2 contributes to hemostatic competence. Disruption of these pathways by NSAIDs that inhibit COX-1 is thought to account for NSAID induced ulcers. However, NSAIDs selective for inhibition of COX-2 only halve the comparative incidence of serious gastrointestinal events. This may in part reflect their impact on gastroduodenal epithelial COX-2 dependent prostanoids that accelerate ulcer healing. In these examples, inhibitors high up in the pathway confer benefit, but it is a net benefit and of course the margin of that benefit may vary substantially between individuals. We poorly understand inter-individual differences in anti-inflammatory efficacy amongst the reversibly acting NSAIDs.

Prostanoids tend to be relatively weak agonists in systems where their blockade has resulted in clinical efficacy. For example, TxA2 is a relatively weak platelet agonist compared to thrombin and there is a considerable amount of redundancy in the system.

Why does blockade of just one of the many pathways of platelet activation result in an effect so great that its impact can be detected by as crude and instrument as a randomized clinical trial? Similarly why does blockade of sulfidopeptide leukotrienes alone amongst many bronchoconstrictors result in clinical efficacy in asthma or why do other mediators, such as NO, not substitute for the cardioprotective effects of PGI2, suppressed by NSAIDs selective for inhibition of COX-2? Perhaps drug efficacy is because eicosanoids often tend to function as amplifying signals for other, more potent agonists. Activate platelets with thrombin, ADP or collagen and release of TxA2 amplifies and sustains the aggregation response. Perhaps this is why aspirin is so effective in the secondary prevention of myocardial infarction or stroke. As for why other mediators do not step in to substitute for suppressed prostanoids, this may speak to their singular importance in circumstances of phenotypic perturbations, as discussed next.

Given the myriad biological effects of these compounds, how are drugs that shut down their synthesis even tolerated? NSAIDs may indeed result in life threatening gastrointestinal or cardiovascular adverse events, but only in a small minority (maybe 12%) of patients exposed. This may reflect the fact that prostanoid formation is a homeostatic response system. Under physiological conditions, trivial amounts of these compounds are formed and their biological importance is, in many cases marginal. However, when a system is stressed they may become pivotal. Examples include their essential role in the maintenance of renal blood flow under renoprival conditions, their antihypertensive effects in patients infused with vasopressors or their antithrombotic effects in patients at increased risk of thrombogenesis. For example, deletion of the IP does not result in spontaneous thrombosis but rather accentuates the response to thrombogeneic stimuli. Similarly, although pre-existing cardiovascular disease was often used to dilute the legal liability of the sponsor in cases where patients suffered myocardial infarctions (MIs) while taking coxibs, the relative risk of MIs on celecoxib relates to the underlying burden of cardiovascular disease, an expected consequence of suppression of COX-2 derived PGI2 (164).

Given the contrasting effects of prostanoids, should we move down the pathway to get a more targeted and safer response? Presently, we have almost no data that address this question. In the case of downstream PG synthase versus COX-2 inhibition, the experience with mPGES-1 deletion highlights the complexity of the comparison. Here, substrate rediversion to PGIS, may attenuate the cardiovascular risk of COX-2 inhibition, but dilute analgesic efficacy. Another example is the comparison between low dose aspirin for cardioprotection and TP antagonism. The latter strategy avoids PGI2 suppression, but this is very modest, on average ~15%, with low dose aspirin. One would need an enormous clinical trial to detect that theoretical benefit. Alternatively, the antagonist, unlike aspirin, might block TP activation by unconventional ligands, such as isoprostanes and HETEs. However, while these compounds can activate the TP, their relevance in vivo, even in settings of tissue reperfusion where they are formed in excess, remains speculative. One might of course model this comparison and use COX-1 knock down mice which mimic the asymmetric impact on platelet COX-1 of low dose aspirin, to address the question. Finally, as the overlap in the biological consequences of activating several prostanoid receptors the IP, EP2, EP4, DP1 or the FP, TP, EP1 groups for example, hint at the possibility that efficacy might be diluted as one moves downstream to target just one prostanoid in the pathway. Our poor understanding of such potential functional redundancy at the receptor level, never mind insight into the implications of receptor dimerization leave these as open questions for drug development.

Surely, this complexity suggests that experiments in animals are going to be of virtually no value when it comes to predicting drug effects in vivo? Certainly the limitations of standard experimental paradigms apply in this pathway as in others. We study mouse models of atherosclerosis that fail to undergo spontaneous plaque fissure and thrombotic occlusion of vital arteries to reach conclusions about prevention or provocation of MI. We extrapolate from drug action on murine tolerance of a hotplate to patients undergoing molar extraction in an effort to divine therapies for elderly ladies with osteoarthritis of the knees. These and many such examples inspire caution in the field of translational therapeutics. However, consistency of evidence from different model systems across multiple species can, especially when integrated with independent lines of evidence, predict with some confidence the outcome in randomized clinical trials of drug action in this pathway. Such was the case in the prediction and mechanistic elucidation of the cardiovascular hazard from NSAIDs (178).

Finally, is it likely that this pathway will yield more therapeutic opportunities? Besides drugs already approved for various indications - aspirin and the myriad NSAIDs, analogues of PGE2, PGF2 and PGI2, TP and leukotriene antagonists, currently inhibitors of mPGES-1, 5 lipoxygenase and its activating factor (FLAP) and antagonists of DP1, DP2 and EP4 are all undergoing clinical evaluation. Many other targets in the eicosanoid pathway are undergoing preclinical evaluation, just as others, such as the multiple secretory phospholipases and the soluble epoxide hydrolase, emerge. This coincides with new information about old targets.

Why do we need to inhibit both COXs, not just COX-1 to see gastrointestinal injury in model systems (29)? Why is a form of aspirin confined to platelet inhibition in the presystemic circulation associated with a reduction in the incidence of colon cancer (179, 180)? Much remains to be discovered about the biology of the eicosanoids and from this is likely to come new therapeutic opportunity.

Prostanoids can promote or restrain acute inflammation. Products of COX-2 in particular may also contribute to resolution of inflammation in certain settings. Presently, we have little information on which products of COX-2 might subserve this role or indeed if the dominant factors reflect rediversion of the arachidonic acid substrate to other metabolic pathways consequent to deletion or inhibition of COX-2. As with cyclopentanone prostanoids, many arachidonate derivatives, including transcellular products, when synthesized and administered as exogenous compounds, can promote resolution in models of inflammation. However, rigorous physico-chemical evidence for the formation of the endogenous species in relevant quantities to subserve this role in vivo is limited. Elucidation of whether and how prostanoids might restrain inflammation and how substrate modification, such as with fish oils, might exploit this understanding is currently a focus of much research from which novel therapeutic strategies are likely to emerge.

The work that formed the basis for opinions expressed in this review was supported by National Institutes of Health grants HL062250, HL083799 and HL054500 (G.A.F). Dr FitzGerald is the McNeil Professor of Translational Medicine and Therapeutics. He has consulted in the past year for Astra Zeneca, Daiichi Sankyo, Logical Therapeutics, Lilly, and Nicox on NSAIDs and related compounds. The other authors report no conflicts.

The rest is here:
Prostaglandins and Inflammation - PMC - PubMed Central (PMC)