Cytosolic phospholipase A2: physiological function and role in disease

The group IV phospholipase A2 (PLA2) family is comprised of six intracellular enzymes (GIVA, -B, -C, -D, -E, and -F) commonly referred to as cytosolic PLA2 (cPLA2)α, -β, -γ, -δ, -ε, and -ζ. They contain a Ser-Asp catalytic dyad and all except cPLA2γ have a C2 domain, but differences in their catalytic activities and subcellular localization suggest unique regulation and function. With the exception of cPLA2α, the focus of this review, little is known about the in vivo function of group IV enzymes. cPLA2α catalyzes the hydrolysis of phospholipids to arachidonic acid and lysophospholipids that are precursors of numerous bioactive lipids. The regulation of cPLA2α is complex, involving transcriptional and posttranslational processes, particularly increases in calcium and phosphorylation. cPLA2α is a highly conserved widely expressed enzyme that promotes lipid mediator production in human and rodent cells from a variety of tissues. The diverse bioactive lipids produced as a result of cPLA2α activation regulate normal physiological processes and disease pathogenesis in many organ systems, as shown using cPLA2α KO mice. However, humans recently identified with cPLA2α deficiency exhibit more pronounced effects on health than observed in mice lacking cPLA2α, indicating that much remains to be learned about this interesting enzyme.

transport ( 14,19,20 ). This is supported by a recent study showing that cPLA 2 regulates formation of tubules involved in clathrin-independent endocytic traffi cking ( 21 ). Therefore in studies designed to understand the role of cPLA 2 ␣ in regulating cellular processes, it is important to consider that it has a functional role independent of eicosanoid production.
GROUP IVA cPLA 2 (cPLA 2 ␣ ) It has been over 25 years since a soluble calcium-regulated high-molecular-weight PLA 2 that exhibited arachidonic acid selectivity was identifi ed in cells and purifi ed (22)(23)(24)(25)(26)(27)(28)(29). Cloning of PLA 2 ␣ revealed the C2 domain for calcium regulation ( 30,31 ). The identifi cation of the active site residues, characterization of stimulus-dependent phosphorylation by MAPKs, and structural elucidation then followed (32)(33)(34)(35)(36). Over the years, we have learned a great deal about the regulation of cPLA 2 ␣ and its role in initiating the production of bioactive lipid mediators . cPLA 2 ␣ activation results in the production of diverse lipid mediators derived from its products, arachidonic acid and lysophospholipids. Leukotrienes are potent pro-infl ammatory mediators produced by cells involved in the infl ammatory response, particularly neutrophils, macrophages, and mast cells ( 37 ). Prostaglandins (PGs) are made by many cell types and have very diverse functions. In addition to their role in regulating homeostatic processes, they can promote all the cardinal signs of acute infl ammation and participate in the maintenance of chronic infl ammation ( 38,39 ). However they facilitate the resolution of infl ammation by altering the balance of pro-infl ammatory and antiinflammatory cytokines, and by enhancing clearance of apoptotic neutrophils ( 38,(40)(41)(42)(43). cPLA 2 ␣ is also implicated in the production of the pro-resolving lipid mediators (lipoxins) and bioactive lysophospholipids (or derived from lysophospholipids) such as platelet-activating factor and lysophosphatidylinositol (44)(45)(46). The generation of mice defi cient in cPLA 2 ␣ has revealed its role in regulating normal physiological processes and disease pathogenesis in rodents ( 47,48 ). More recently, the identifi cation of humans with cPLA 2 ␣ defi ciency has emphasized its important role in human health (49)(50)(51). This review will provide a brief overview of the regulation of cPLA 2 ␣ and primarily focus on reviewing its function in specifi c organs of mice and humans.
REGULATION OF cPLA 2 ␣ cPLA 2 ␣ is widely expressed in cells throughout all tissues in mice and humans. It is a highly conserved enzyme with mouse and human sharing 95% amino acid identity, suggesting similarities in regulation and function ( 31 ). In addition to PLA 2 activity, cPLA 2 ␣ catalyzes other enzymatic reactions through the Ser-Asp catalytic dyad, including PLA 1 , lysophospholipase, and transacylase activity, although the physiological relevance of these activities in vivo is exhibits preference for hydrolysis of arachidonic acid from phospholipid substrates that occurs in cells stimulated with diverse agonists ( 4 ). Most cells contain several types of PLA 2 enzymes and studies have suggested that they have distinct functions. For example, a comparison of the function of cPLA 2 ␣ and iPLA 2 ␤ in mast cells and macrophages found that cPLA 2 ␣ , but not iPLA 2 ␤ , selectively hydrolyzes arachidonic acid-containing phospholipids in response to cell stimulation ( 12,13 ).

GROUP IV PLA 2 FAMILY
The six members of the cPLA 2 family (cPLA 2 ␣ , -␤ , -␥ , -␦ , -, and -) share only about 30% homology, and have differences in enzymatic properties, tissue expression, and subcellular localization, suggesting that they are not redundant ( 4,14 ). Analysis of the interfacial kinetic and binding properties of the cPLA 2 family showed that they exhibit very different relative lysophospholipase, PLA 2 , and PLA 1 activities, inhibitor sensitivities, calcium dependence, and activation by anionic phospholipids, highlighting potential differences in regulation and function ( 10 ). With the exception of the established role of cPLA 2 ␣ in initiating production of lipid mediators, little is known about the in vivo function of other cPLA 2 isoforms. However, there is evidence emerging from studies using cultured cells that members of the cPLA 2 family, including cPLA 2 ␣ , play a role in regulating membrane traffi cking that does not involve production of oxygenated metabolites of arachidonic acid. PLA 2 (and PLA 1 ) enzymes are implicated in regulating intracellular membrane traffi cking by participating in the formation of transport carriers from donor membrane ( 15 ). Membrane budding requires alterations in membrane curvature that can occur by formation of lysophospholipids. Four cytoplasmic PLAs, including cPLA 2 ␣ , PAF-AH Ib, iPLA 2 ␤ , and iPLA 1 ␥ , are implicated in the formation of tubules from the endoplasmic reticulum (ER)-Golgi intermediate compartment and the Golgi apparatus, maintenance of Golgi structure, and transport through the Golgi. The details of the role of these enzymes have recently been reviewed ( 15,16 ). cPLA 2 ␣ regulates intra-Golgi transport by mediating the formation of tubules that connect the Golgi stacks ( 17 ). The localization of cPLA 2 ␣ at the Golgi in endothelial cells has also been shown to regulate the transport of junction proteins (vascular endothelial-cadherin, occludin, claudin-5) from the Golgi to cell-cell junctions ( 18 ). In these studies it was shown that the enzymatic activity of cPLA 2 ␣ is required for regulating transport, but oxygenated metabolites of arachidonic acid are not involved. Considering that these are fundamental processes important for cell function, it is surprising that the cPLA 2 ␣ KO mouse does not exhibit any major phenotype. Golgi transport was found to be normal in cells defi cient in cPLA 2 ␣ , but results of a siRNA screen suggested that GVIII PLA 2 (PAF-AH) compensates for the loss of cPLA 2 ␣ in regulating Golgi transport ( 17 ). Other group IV PLA 2 s have been shown to localize to endocytic vesicles, suggesting a possible role in regulating analysis of the human cPLA 2 ␣ gene ( 73 ). This IL-1 ␤responsive distal regulatory element contains an AP-1 site, which regulates both basal and IL-1 ␤ -induced expression of cPLA 2 ␣ . Binding of c-Jun to this site acts as a repressor in the basal state and an activator in response to IL-1 ␤ , and an important role for C/EBP ␤ as a transcriptional activator of cPLA 2 ␣ was demonstrated. cPLA 2 ␣ is rapidly activated in cells by posttranslational processes, including increases in intracellular calcium and phosphorylation by MAPK ( Fig. 1 ). These signaling pathways are activated in cells through engagement of many types of receptors, indicating that cPLA 2 ␣ activation and arachidonic acid release occur commonly in response to cell stimulation. There have been a number of reviews that detail the regulation of cPLA 2 ␣ by posttranslational processes (2)(3)(4)74 ). In brief, increases in intracellular calcium promote the translocation of cPLA 2 ␣ from the cytosol to intracellular membrane (75)(76)(77)(78). Calcium binds to the Nterminal C2 domain of cPLA 2 ␣ that increases the hydrophobicity of the calcium binding loops, which penetrate the membrane bilayer ( 79-83 ) ( Fig. 2 ). Differences in phospholipid binding specifi city of the C2 domains of protein kinase C (anionic phospholipids) and cPLA 2 ␣ (phosphatidylcholine) play an important role in determining unknown ( 2,4,32,35 ). The intense focus on cPLA 2 ␣ stems from its preferential hydrolysis of sn -2 arachidonic acid, and its well-established role in initiating the release of arachidonic acid for the production of lipid mediators ( 52 ). cPLA 2 ␣ KO mice have provided a source for cells, including (but not limited to) mast cells, neutrophils, macrophages, platelets, endothelial cells, and lung fi broblasts, to establish a role for cPLA 2 ␣ in mediating arachidonic acid release and production of lipid mediators ( 47,(53)(54)(55)(56)(57)(58)(59). cPLA 2 ␣ is regulated at the transcriptional level and by posttranslational mechanisms. It is basally expressed in many cells, due in part to transcription factor II D binding to the TATA-less promoter (60)(61)(62). cPLA 2 ␣ expression is also induced transcriptionally through a number of signaling pathways involving Ras and MAPKs, and transcriptional activators NF-B, Krüppel-like factor, hypoxiainducible factor, Sp1, and c-Jun have been described (63)(64)(65)(66)(67)(68)(69)(70). Pro-infl ammatory cytokines induce expression of human cPLA 2 ␣ mRNA that is blocked by glucocorticoids ( 71 ). However, glucocorticoids are not universally suppressive because they paradoxically enhance cPLA 2 ␣ expression in human amnion fi broblasts (see Reproduction section below) ( 72 ). Recently the fi rst cytokine-dependent enhancer element has been identifi ed by DNase I hypersensitive site Fig. 1. Regulation of cPLA 2 ␣ . cPLA 2 ␣ -mediated arachidonic acid (AA) release occurs following cell stimulation that results in increases in intracellular Ca 2+ and kinase activation. In the cytosol, cPLA 2 ␣ is bound to p11/annexin A2 and is released from this complex when phosphorylated on S727. Calcium binds to the calcium binding loops (CBL, green) at the membrane binding face of the C2 domain. This increases the hydrophobicity of the C2 domain and promotes preferential binding of cPLA 2 ␣ to the Golgi and, at higher levels of intracellular calcium, to the ER and nuclear envelope. Basic residues (orange patch) in the C2 domain are implicated in binding to ceramide-1-P in the Golgi. Association of the catalytic domain with membrane is mediated in part by a tryptophan residue in the catalytic domain (W464, red patch) that stabilizes cPLA 2 ␣ on the membrane. Phosphorylation of S505 by MAPKs in the catalytic domain enhances cPLA 2 ␣ activity. Interaction of basic residues in the catalytic domain (blue patch) with anionic components in the membrane, such as polyphosphoinositides, optimizes catalytic activity. AA released by cPLA 2 ␣ is metabolized by COX-1 and -2, which are primarily localized to the ER, but also occur on the Golgi. Leukotriene production occurs at the nuclear envelope, the site of 5-LO translocation in response to increases in calcium. 5-LO binds to the scaffold protein FLAP, an AA binding protein, which also binds LTC 4 synthase (syn) for production of LTC 4 . results show that the ability of cPLA 2 ␣ to release arachidonic acid in cells involves calcium-dependent membrane binding and optimization of catalytic activity by phosphorylation and interaction of basic residues in the catalytic domain with anionic components, perhaps polyphosphoinositides, in the membrane ( 92 ).
The translocation of cPLA 2 ␣ to intracellular membranes raises the interesting question of how cPLA 2 ␣ -derived arachidonic acid couples to the downstream enzymes, 5-lipoxygenase (5-LO) and cyclooxygenase (COX)-1 and -2 ( Fig. 1 ). There are general similarities in the regulation of cPLA 2 ␣ and 5-LO involving both calcium and phosphorylation ( 109 ). 5-LO, a soluble enzyme (cytoplasmic or intranuclear) in resting cells, translocates to the nuclear envelope in response to calcium, which binds to an N-terminal C2like domain (110)(111)(112)(113)(114)(115). Leukotriene production requires interaction of 5-LO with 5-LO-activating protein (FLAP), an integral nuclear envelope protein that functions as an arachidonic acid binding protein ( 116 ). It has recently been shown that leukotriene synthesis involves activationdependent formation of supramolecular complexes on the nuclear envelope composed of FLAP with 5-LO and FLAP with leukotriene C 4 synthase ( 117,118 ). In addition, cPLA 2 ␣ -derived arachidonic acid was shown to affect the conformation of FLAP and enhance its ability to recruit 5-LO for complex formation ( 119 ). It is not known whether cPLA 2 ␣ physically associates with these complexes or whether arachidonic acid released in their vicinity diffuses through the membrane to initiate leukotriene production. their distinct subcellular targeting to the plasma membrane and intracellular membranes, respectively (84)(85)(86). The catalytic domain is then positioned on the membrane by calcium-independent mechanisms, in part through a tryptophan residue (W464) that stabilizes membrane binding (87)(88)(89)(90). cPLA 2 ␣ preferentially translocates to the Golgi apparatus and, at higher intracellular calcium concentrations, to the ER and nuclear envelope ( 76-78, 91, 92 ) ( Fig. 1 ). Binding of cPLA 2 ␣ to the Golgi may, in part, involve interaction of the C2 domain with sphingolipids, including ceramide-1-P and lactosylceramide (93)(94)(95)(96). Translocation of cPLA 2 ␣ to membrane is necessary but not suffi cient for cPLA 2 ␣ to release arachidonic acid ( 92 ). The catalytic activity of cPLA 2 ␣ is enhanced by phosphorylation of S505 by the MAPKs, extracellular regulated protein kinases, and p38 ( 33,34,97,98 ). More recent studies suggest that cPLA 2 ␣ is also activated by c-jun N-terminal kinases ( 99,100 ). Activation of cPLA 2 ␣ in vascular smooth muscle cells by calcium/calmodulin-dependent kinase II that phosphorylates S515 has been described ( 101 ). Phosphorylation of cPLA 2 ␣ on S727 by MAPK-interacting kinase does not enhance catalytic activity per se, but blocks its binding to an inhibitory complex in the cytosol composed of p11(S100A10)/annexin A2 ( 74, 102-105 ) ( Fig. 1 ). In addition to phosphorylation, the activity of cPLA 2 ␣ is regulated by basic residues in the catalytic domain that are the site for activation by polyphosphoinositides ( 89,(106)(107)(108). The basic residues are required for the ability of cPLA 2 ␣ to release arachidonic acid in cells but are not required for translocation of cPLA 2 ␣ to the Golgi ( 92 ). Collectively, the In all of these patients, a common clinical fi nding is abnormalities in the gastrointestinal (GI) tract. However, there are differences in the clinical presentations and severity of disease. In one subject, two rare heterozygous single bp mutations in PLA2G4A were found; one mutation (S111P) occurred in the C2 domain of cPLA 2 ␣ and the other allele had a mutation in the catalytic domain (R485H) ( Fig. 2 ) ( 49 ). The expression of cPLA 2 ␣ protein and PLA 2 activity in platelet lysates was reduced compared with normals. The patient exhibited a global reduction in eicosanoid production with reduced urinary metabolites of thromboxane (TX)A 2 , PGI 2 , PGD 2 , and PGE 2 , and undetectable leukotriene E 4 in urine. The production of leukotriene B 4 (LTB 4 ) from leukocytes in A23187-stimulated blood was reduced by 97%. The patient had occult GI blood loss and anemia in childhood with more acute GI bleeding in his fourth decade with recurrent ulcerations in the ileum and jejunum of the small intestine. The benefi cial effect of misoprostol suggests that cPLA 2 ␣ -mediated PGE 2 production is important for function of the small intestine. Unlike the effect of nonsteroid anti-infl ammatory drug toxicity, the patient did not exhibit gastroduodenal ulcerations, and the colon was normal. The results support a role for cPLA 2 ␣ in providing arachidonic acid for a large proportion of eicosanoid production in humans. Other PLA 2 s may contribute to the residual eicosanoids produced, or the patient may retain a low level of cPLA 2 ␣ activity. To investigate the effect of the mutations on cPLA 2 ␣ function in more detail, we compared the catalytic activity of WT and mutant forms of purifi ed cPLA 2 ␣ in vitro, and compared their cellular function by expression in cells lacking endogenous cPLA 2 ␣ ( 124 ). cPLA 2 ␣ with the S111P mutation showed slightly reduced catalytic activity and interfacial binding in vitro at saturating calcium. When expressed in cells, there was less translocation of cPLA 2 ␣ with the S111P mutation to the Golgi in response to serum stimulation and reduced arachidonic acid release. In contrast, cPLA 2 ␣ with the R485H mutation translocated to the Golgi in serum-stimulated cells, but failed to release arachidonic acid and was inactive in vitro. Consequently, the mutations in the C2 and catalytic domains of cPLA 2 ␣ compromise function, but do not result in a complete loss of protein or activity.
Another patient, and his twin sister, with inherited cPLA 2 ␣ deficiency also exhibited GI abnormalities from early childhood characterized by duodenal ulcers that were corrected by misoprostol treatment ( 51 ). The incidence of small intestinal ulcers was not evaluated. Homozygous bp mutations (D575H) in the cPLA 2 ␣ catalytic domain were identifi ed ( Fig. 2 ). cPLA 2 ␣ protein was not detected in lysates from patients with the homozygous mutations. Unlike the patients described above, the mutations resulted in global diathesis (see Hemostasis section below).
The most severe clinical manifestation of inherited cPLA 2 ␣ defi ciency occurred in siblings diagnosed with cryptogenic multifocal ulcerating stenosing enteritis due to homozygous deletion mutations in PLA2G4A ( 50 ). The affected male had severe peptic ulcers and bleeding beginning at an early age (4 years) then over the next several years COX-1 and -2 are also found in internal membranes, but unlike cPLA 2 ␣ and 5-LO, they are integral membrane proteins primarily localized in the ER ( 120 ). However, it was recently shown that COX-2 localizes to the Golgi in cancer cell lines detected by using a COX-2-specifi c fl uorescent probe ( 121 ). Another study investigating the posttranslational processing and degradation of COX-2 found catalytically active COX-2 (but not COX-1) localized to the Golgi as a result of anterograde traffi cking from the ER, a step required for posttranslational modifi cation prior to COX-2 degradation through the ER-associated degradation pathway ( 122 ). It is interesting that microsomal PGE synthase-1 (mPGES-1) was also found at the Golgi, suggesting that this site has a system for PGE 2 production involving cPLA 2 ␣ /COX-2/mPGES-1 ( 122 ). Another report showed that COX-1, but not COX-2, is constitutively found at the Golgi in airway epithelial cells, where it colocalizes with cPLA 2 ␣ ( 123 ). The localization of cPLA 2 ␣ , COX, mPGES-1, and leukotriene synthetic enzymes at intracellular membranes, including the ER, nuclear envelope, and Golgi, may effi ciently coordinate the production of certain eicosanoids. However, several downstream enzymes involved in PG and leukotriene production are cytosolic, and not apparently subject to stimulus-dependent translocation, although the sub-cellular organization of enzymes involved in eicosanoid production is only beginning to be elucidated.

ROLE OF cPLA 2 ␣ IN NORMAL PHYSIOLOGICAL PROCESSES AND DISEASE
The fi rst characterization of mice with cPLA 2 ␣ gene disruption published in 1997 was followed by a large body of work that provides a comprehensive view of cPLA 2 ␣ function in normal processes and disease pathogenesis in mice ( 47,48 ). The cPLA 2 ␣ KO mouse develops normally and lives a normal life span with no overt adverse health effects. Considering the evolutionary conservation of cPLA 2 ␣ and its role in eicosanoid production, similarities in the regulation and function of cPLA 2 ␣ in rodents and humans is expected . However, as discussed in this review, there are deleterious effects on the health of humans with cPLA 2 ␣ defi ciency, suggesting differences in the role of cPLA 2 ␣ and eicosanoids in mouse models and humans. This could be due to a number of factors, including differences in expression of other types of PLA 2 enzymes that could compensate and differences in the tissue-specifi c expression of receptors for lipid mediators that could infl uence their function. In many cases, the mechanisms responsible for observed phenotypes in human and mouse cPLA 2 ␣ defi ciencies are not well understood. This is understandable considering that cPLA 2 ␣ is the fi rst regulatory enzyme that releases the lipid mediators, arachidonic acid and lysophospholipids, which are precursors for an enormous number of diverse bioactive lipid metabolites that can have opposing effects.

Gastrointestinal tract
The consequences of inherited cPLA 2 ␣ defi ciency in several human patients have now been described (49)(50)(51). against colon cancer in mice induced by the colon carcinogen, azoxymethane, which induces increased tumor multiplicity in heterozygous and KO mice compared with WT mice ( 131 ). The increased colon tumorigenesis in cPLA 2 ␣ mutant mice occurs despite a decrease in PGE 2 production. In contrast, genetic ablation of cPLA 2 ␣ protects against cancer development in mouse models of adenomatous polyposis (Apc ⌬ 716 KO mice and Apc Min mice) ( 125,132 ). Apc ⌬ 716 KO mice develop small intestinal polyps that express high levels of COX-2 and cPLA 2 ␣ mRNA compared with normal small intestine ( 125,133 ). Genetic deletion of cPLA 2 ␣ in Apc ⌬ 716 KO mice leads to a reduction in size of the polyps but has no effect on the number of polyps unlike COX-2 KO mice ( 125,133 ). In contrast, knocking out cPLA 2 ␣ in Apc Min mice results in a reduction in size and a dramatic decrease in the number of small intestinal polyps, although there is no effect on tumor number in the large intestine ( 132 ). The results highlight differences in the role of cPLA 2 ␣ in promoting tumorigenesis in the small and large intestine. Several lines of evidence suggest that tumorigenesis in the small intestine involves the cPLA 2 ␣ /COX-2/PGE 2 pathway rather than a role for arachidonic acid itself in promoting apoptosis ( 132,134 ). Understanding how cPLA 2 ␣ regulates tumorigenesis is complex because it is upstream of many lipid mediators including arachidonic acid, diverse PGs, leukotrienes, and lysophospholipids. In addition, the large repertoire of receptors for lipid mediators is differentially expressed in tissues and promotes signaling pathways with distinct effects on cell function. This adds another level of complexity in understanding the role of cPLA 2 ␣ -derived bioactive lipids in organ-specifi c cancer development.

Liver
Studies have shown that PGs promote increased triglyceride storage in adipocytes and hepatocytes, prompting investigations on the role of cPLA 2 ␣ in regulating lipid storage and liver damage (135)(136)(137)(138)(139)(140). It is of interest that even under normal dietary conditions, cPLA 2 ␣ regulates lipid storage in adipose tissue and liver ( 138 ). cPLA 2 ␣ KO mice fed a normal diet have reduced adipose tissue mass compared with WT mice, and lower triglyceride deposition in the liver, that correlates with lower serum levels of PGE 2 . An extension of this work showed that cPLA 2 ␣ KO mice are protected from fatty liver damage induced by a high-fat diet ( 139 ). Mice with cPLA 2 ␣ defi ciency have lower serum PGE 2 levels, which may afford protection because the administration of PGE 2 enhances hepatic triacylglycerol content, whereas inhibition of COX-2 prevents fat deposition in rodents ( 137,141 ). Obesity is associated with development of nonalcoholic fatty liver disease with pathological manifestations including nonalcoholic steatohepatitis, fi brosis, or cirrhosis. Recently, it was shown that cPLA 2 ␣ KO mice fed a high-fat diet are protected from hepatic liver deposition and development of hepatic fi brosis ( 140 ). Similar results are observed in mice subjected to repeated administration of carbon tetrachloride that induces fi brosis without lipid deposition. The mechanism is developed more severe disease involving the ileum, stomach, esophagus, duodenum, biliary system, and liver, although the colon remained normal. He developed type 2 diabetes, peripheral neuropathy, and osteoporosis. His sister exhibited similar severe GI disease beginning at age 2 years that worsened over time. She experienced infections with Campylobacter enteritis and Salmonella enteriditis , and developed Candida septicemia, staphylococcus lung infection, and acute respiratory distress syndrome. Other extra-intestinal disease included acute renal failure, endometriosis, left ventricular concentric hypertrophy, fi brotic bladder with stones, and infertility. The siblings contained homozygous 4 bp deletions in the PLA2G4A gene predicted to result in a frameshift and premature stop codon with the loss of 43 amino acids at the C terminus of cPLA 2 ␣ (from V707) ( Fig. 2 ). cPLA 2 ␣ was undetectable by immunohistochemical analysis of small bowel biopsy and could not be detected in peripheral blood mononuclear cells by Western blot. Many factors could contribute to the individual phenotypic differences in disease severity associated with inherited cPLA 2 ␣ defi ciency, but the studies show a critical role for cPLA 2 ␣ and eicosanoids in maintaining the integrity of the gut, particularly the small intestine.
The small intestine of the cPLA 2 ␣ KO mouse has numerous small ulcers, but there appears to be little effect of the lesions on the health of mutant mice ( 125 ). However, there is a role for cPLA 2 ␣ in protecting the GI tract of mice from chemical toxicity induced by COX inhibitors and dextran sodium sulfate (DSS) ( 126,127 ). Injury induced by COX inhibitors in cPLA 2 ␣ KO mice did not occur in the colon, but was extensive in the small intestine, the site of greatest drug absorption ( 126 ). Differences in toxicity of cPLA 2 ␣ WT and KO mice did not correlate with PGE 2 levels, but may have been due to greater mitochondrial toxicity in KO mice. In the DSS model, there was extensive damage in the colons of cPLA 2 ␣ KO mice that correlated with lower levels of PGs in colon tissue compared with DSS-treated WT mice ( 127 ). Mice defi cient in mPGES-1 also exhibited greater DSS toxicity in the colon, suggesting an important protective role for PGE 2 . In contrast to cPLA 2 ␣ KO mice, which recovered from DSS toxicity to a similar extent as WT mice, the recovery of mPGES-1 KO mice was compromised. The studies in mice and humans show an important role for cPLA 2 ␣ in protecting the GI tract. However, levels of eicosanoids in the GI tract of cPLA 2 ␣ KO mice are only partially reduced, suggesting a role for other PLA 2 s. This may contribute to the less severe health effects of cPLA 2 ␣ defi ciency in mice compared with humans. There has not been a systematic comparison of the relative levels of the various types of PLA 2 enzymes expressed in different tissues of mice and humans to speculate on the PLA 2 s that may account for eicosanoid production in the GI tract of mice.
Although results support a role for PGs in protecting the integrity of the intestine, COX-2 expression and PGE 2mediated infl ammation are associated with development of human colorectal cancer, consistent with epidemiological evidence that nonsteroid anti-infl ammatory drug use is protective (128)(129)(130). However, cPLA 2 ␣ protects exhibit decreased fertility, small litter size, and delayed onset of labor ( 47,48,(148)(149)(150). In cPLA 2 ␣ KO mice, blastocyst attachment is delayed due to a loss of maternal cPLA 2 ␣ , but decidualization is unaffected ( 151 ). Implantation beyond the normal "window" results in defective embryo development, abnormal uterine spacing of the embryos and resorption contributing to the small litter size, and defects in parturition ( 151 ). A later study emphasized the importance of cPLA 2 ␣ function early in pregnancy because the defect in parturition in cPLA 2 ␣ KO mice could be attributed to the deferred embryo implantation ( 152 ). cPLA 2 ␣ KO mice showed less pronounced defects in ovulation and fertilization compared with COX-2 KO mice ( 149,151,152 ). A more detailed study showed that cPLA 2 ␣ modestly regulates ovulation and fertilization by promoting the induction of COX-2 for PG production ( 153 ). A role for cPLA 2 ␣ in regulating male sexual maturation has also been described ( 154 ). Male cPLA 2 ␣ KO mice are fertile but exhibit abnormal histology of the testes showing a reduced number of interstitial cells. Compared with WT mice, cPLA 2 ␣ KO males have delayed spermatogenesis and reduced testosterone production, although these defects normalized with age.
It is diffi cult to extrapolate from studies in mouse models in understanding the specifi c role of cPLA 2 ␣ in human reproduction, which has many unique regulatory features not shared by other mammalian species ( 155 ). The female patient with homozygous deletion mutations in cPLA 2 ␣ associated with cryptogenic multifocal ulcerating stenosing enteritis is infertile, although the basis for this is unknown, and may be due to the severe consequences of the disease ( 50 ). PGs regulate labor and birth in humans, and an important source of PGs is amnion fi broblasts in fetal membranes ( 156,157 ). There is an increase in cortisol secretion and PG production during parturition, a paradoxical event because cortisol suppresses PG production in most tissues ( 158 ). Glucocorticoids have been shown to induce mRNA and protein expression of cPLA 2 ␣ and COX-2 in human amnion fi broblasts ( 72,159 ). The induction of cPLA 2 ␣ by glucocorticoids involves activation of G ␣ s leading to cyclic AMP response element-binding protein-1 phosphorylation , which interacts with the glucocorticoid receptor on the glucocorticoid response element in the cPLA 2 ␣ promoter ( 160 ).

Kidney
cPLA 2 ␣ KO female mice exhibit defects in renal function, possibly due to a defi ciency in PGE 2 production, which regulates various aspects of kidney physiology ( 161 ). Mice lacking cPLA 2 ␣ have a urine-concentrating defect that develops with age. Based on abnormal aequorin staining in proximal tubules, the defect in cPLA 2 ␣ KO female mice may be due to diminished water reabsorption. There is also evidence that cPLA 2 ␣ regulates Na + -HCO 3 -transport by angiotensin II in kidney proximal tubules ( 162 ). The role of cPLA 2 ␣ in kidney function of humans is poorly understood. The female sibling with the cPLA 2 ␣ deletion developed acute renal failure requiring hemodialysis, although whether this is a direct effect of cPLA 2 ␣ deletion is not clear ( 50 ). It has recently been thought to involve a role for cPLA 2 ␣ in production of monocyte chemoattractant protein-1 and infi ltration of macrophages in these models. Previous results have shown that leukotrienes and PGs regulate induction of monocyte chemoattractant protein-1, a chemotactic factor implicated in development of hepatic fi brosis (142)(143)(144)(145). Eicosanoids secreted by cells infl uence gene expression through paracrine or autocrine interaction with their specifi c G proteincoupled receptors ( 146,147 ). For example in an infectious disease model using macrophages from WT and cPLA 2 ␣ KO mice, we found that cPLA 2 ␣ activation promotes an autocrine loop involving production of PGs that engage prostanoid receptors, increase cAMP, and globally affect expression of genes involved in host defense and infl ammation ( 43 ).

Hemostasis
Considering the importance of eicosanoids in regulating hemostasis, the role of cPLA 2 ␣ in TXA 2 production, platelet activation, and hemostasis has been investigated. Platelets from cPLA 2 ␣ KO mice stimulated with collagen produce much less TXA 2 , and bleeding time is increased compared with WT mice ( 59 ). However, the low level of TXA 2 produced by other phospholipases in collagen-stimulated cPLA 2 ␣ KO platelets is suffi cient for platelet aggregation. In vivo, cPLA 2 ␣ KO mice have increased bleeding time and are protected from thromboembolism that correlates with reduced TXA 2 levels and vasoconstriction. cPLA 2 ␣ appears to play a discrete role in the complex regulation of platelet activation in mice.
Studies of humans with inherited cPLA 2 ␣ defi ciency have shown an important role of cPLA 2 ␣ in regulating platelet function and hemostasis. In the patient with two heterologous bp mutations in the coding regions of PL-A2G4A , the PLA 2 activity in platelet lysates was reduced and levels of TXB 2 and 12-hydroxyeicosatetraenoic acid in serum during blood clotting (platelet-derived) were much lower than normal levels ( 49 ). Platelet aggregation and ATP release were compromised indicating impaired function. In patients with homozygous deletion mutants of cPLA 2 ␣ associated with cryptogenic multifocal ulcerating stenosing enteritis ( 50 ), similar abnormalities of platelet function were observed, including defects in collageninduced aggregation and reduced levels of TXB 2 production. The results indicate that cPLA 2 ␣ plays a dominant role in providing arachidonic acid for the COX-1 and 12-lipoxygenase products in platelets, although the patients described above did not exhibit generalized clotting deficiency. In contrast, the siblings with inherited cPLA 2 ␣ defi ciency due to homozygous bp mutations had lifelong bleeding diathesis characterized by spontaneous mucocutaneous bleeding in addition to GI bleeding. The bleeding defect correlated with defective platelet function and extremely low levels of serum TXB 2 ( 51 ).

Reproduction
PGs produced through the COX-1 and COX-2 pathways play an important role in regulating female reproduction, therefore it is not surprising that cPLA 2 ␣ KO female mice genetic ablation (cPLA 2 ␣ KO C57Bl/6 mice) results in improved motor defi ciencies and less tissue damage from SCI. Another study reported a complex role for different PLA 2 enzymes (cPLA 2 ␣ , sPLA 2 GIIA, and iPLA 2 GVIA) in protecting or exacerbating SCI ( 177 ). All of the PLA 2 s were upregulated after SCI. In contrast to Liu et al. ( 176 ), cPLA 2 ␣ was found to protect from SCI, which was determined using specifi c inhibitors and cPLA 2 ␣ KO mice (Balb/c), but sPLA 2 GIIA and iPLA 2 GVIA were detrimental. Results suggested that the benefi cial effect of cPLA 2 ␣ in SCI involves PGE 2 signaling through the EP 1 receptor. A confounding factor in evaluating the role of cPLA 2 ␣ in these mouse models is the infl uence of genetic differences in mouse strains. For example, the commonly used mouse strains, BALB/c and C57BL/6, generally exhibit distinct polarization of helper T cell responses, with BALB/c predisposed to Th2 and C57BL/6 to Th1 responses ( 178 ). In addition, C57BL/6 mice lack sPLA 2 GIIA due to a natural deletion in the gene that does not occur in BALB/c mice ( 179 ).
A role for cPLA 2 ␣ in mouse models of neurodegenerative disease such as Alzheimer's disease has been investigated ( 180,181 ). The A ␤ peptides that are implicated in Alzheimer's neurodegeneration exert toxic effects in neurons, which are in part mediated by cPLA 2 ␣ ( 182, 183 ). Genetic ablation of cPLA 2 ␣ improves cognitive function in a mouse model of familial Alzheimer's disease, and protects from the toxic effects of intra-cerebroventricular injection of A ␤ oligomers ( 184,185 ). A recent study found that A ␤ peptides activate cPLA 2 ␣ in rat primary cortical neurons, resulting in increased expression of ␤ -amyloid precursor protein by an autocrine pathway involving PGE 2 production and increases in cAMP ( 186 ). Additional studies also suggest that cPLA 2 ␣ activation is involved in promoting neurodegeneration that occurs in prion diseases. In primary cortical neurons, the prion-derived peptide (PrP82-146) induces cPLA 2 ␣ activation and arachidonic acid release that correlates with synapse damage and cell death ( 187 ). Treatment of neurons with PLA 2 inhibitors (AACOCF3, MAFP) and a PGE 2 receptor antagonist (AH13205), protects from synapse degeneration ( 188 ). In a model of Parkinson's disease, cPLA 2 ␣ contributes to neurotoxicity induced by 1-methyl-4-phenyl-1,2,3,6tetrahydropyridine ( 189 ). The reduction in dopamine induced by 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine is signifi cantly less in KO than in WT mice. It is not known how cPLA 2 ␣ contributes to neurodegeneration in these models, but mechanisms involving free radical production, effects on intracellular membrane traffi cking, and arachidonic acid metabolites have been proposed.
It is important to consider underlying differences that have been found in the architecture and lipid composition of the brain in cPLA 2 ␣ WT and KO mice. A recent study showed by microscopy that cortical neurons in cPLA 2 ␣ KO mice exhibit abnormal architecture including larger nuclei, increased number of nucleoli, and aggregated intra-nuclear ribosomes ( 190 ). Abnormalities in the nuclear envelope, nuclear pore, and synapses were observed. The authors speculated that this could be due to the role of shown that cPLA 2 ␣ plays a role in mediating the onset of nonobstructive neonatal hydronephrosis, induced in mice by treatment with 2,3,7,8-tetrachlorodibenzo-p-dioxin, which increases expression of cPLA 2 ␣ in the kidneys ( 163 ). Results showed that cPLA 2 ␣ KO neonates have reduced incidence and severity of hydronephrosis that correlates with reduced expression of COX-2, mPGES-1, and lower urinary PGE 2 compared with 2,3,7,8-tetrachlorodibenzo-p-dioxintreated WT pups. The COX-2/PGE 2 pathway is also implicated in the pathogenesis of dioxin-induced hydronephrosis ( 164 ). The toxic effects of dioxins are mediated by the arylhydrocarbon receptor, which binds to the second intron of Pla2g4a and mediates its induction ( 165,166 ). The arylhydrocarbon receptor pathway also promotes signaling cascades for posttranslational activation of cPLA 2 ␣ by increasing calcium and phosphorylation ( 167,168 ).

Nervous system
Many studies have implicated cPLA 2 ␣ in mediating several disease processes in the nervous system including postischemic brain injury, spinal cord injury (SCI), and neurodegenerative diseases, as will be discussed below. However, there is a role for cPLA 2 ␣ in regulating neuronal homeostasis, particularly long-term depression and longterm potentiation involved in synaptic plasticity. Longterm depression is blocked in brain slices from cPLA 2 ␣ KO mice by treatment with cPLA 2 ␣ (pyrrolidine-1) and COX-2 inhibitors ( 169 ). The defect in long-term depression by cPLA 2 ␣ defi ciency or COX-2 inhibition attenuates optokinetic eye movement adaptation, suggesting that cPLA 2 ␣regulated long-term depression is important for motor learning. Long-term potentiation is also important for motor learning and a recent report found that long-term potentiation in cerebellar parallel fi ber-Purkinje cell synapses is abolished in cPLA 2 ␣ KO mice ( 170,171 ).
The earliest work investigating the role of cPLA 2 ␣ in brain injury found that cPLA 2 ␣ KO mice are protected from postischemic brain injury induced by middle cerebral artery occlusion ( 47 ). Infarct volumes are reduced in KO mice after reperfusion; this is particularly evident in posterior brain slices . The KO mice also exhibit fewer neurological defi cits compared with WT mice. No differences in the vascular anatomy of the brain in KO and WT mice were noted. An extension of this work found that cPLA 2 ␣ contributes to the early events leading to injury during cerebral ischemia ( 172 ). cPLA 2 ␣ WT mice exhibit greater COX-2 expression, PGE 2 production, and reactive oxygen species after ischemia, and more disruption of neuron morphology than KO mice. Other studies have reported a role for cPLA 2 ␣ in regulating the expression of COX-2 in the brain ( 173 ). After systemic LPS, there is less COX-2 expression and PGE 2 production in cPLA 2 ␣ KO compared with WT mice ( 174 ). COX-2 is induced in neuronal and nonneuronal cells of the brain during infl ammation and promotes injury ( 175 ).
A role for cPLA 2 ␣ in regulating SCI has also been investigated. One report found that cPLA 2 ␣ expression is increased in SCI, particularly in neurons and oligodendrocytes ( 176 ). It was shown that inhibiting cPLA 2 ␣ (AACOCF 3 ) or inhibitors, PF-5212372, blocks eicosanoid production by stimulated human lung mast cells and lung homogenates. However, in human lung homogenates, which contain a complex cell mix, the inhibitor was less effective at blocking PGE 2 production than other eicosanoids, suggesting that other PLA 2 s, particularly sPLA 2 s, may contribute to arachidonic acid release for eicosanoid production ( 1,196 ). The inhibitor also blocked contraction of isolated human bronchial rings induced by AMP. The results provide convincing evidence for a role of cPLA 2 ␣ in mediating responses in allergic asthma. Analysis of the promoter region of PLA2G4A has shown that microsatellite fragments are shorter in patients with severe asthma than healthy controls ( 197 ). A comparison of peripheral blood monocytes from severe asthmatics with short and long microsatellite sequences in the PLA2G4A promoter found a correlation between the shorter sequences with increased expression of cPLA 2 ␣ mRNA and protein, and greater eicosanoid production ( 197,198 ).
Results of studies comparing cPLA 2 ␣ WT and KO mice have also supported a role for cPLA 2 ␣ in mediating acute lung injury ( 199 ). One model induced lung injury by intravenous administration of endotoxin and fungal cell walls, zymosan; the other model involved intratracheal administration of hydrochloric acid. These models simulate endotoxemia and acid aspiration, which are causes of acute respiratory distress syndrome. In both models of lung injury, the cPLA 2 ␣ KO mice were protected, exhibiting less protein leakage and neutrophil recruitment into the lung than WT mice, that correlated with reduced levels of TXB 2 , LTB 4 , and cysteinyl leukotrienes. However, in a cecal ligation and puncture model of experimental sepsis, there was no difference in survival of cPLA 2 ␣ WT and KO mice, although there was lower production of eicosanoids (with the exception of 12-hydroxy-eicosatetraenoic acid) and cytokines (IL-6 and CCL2) in the mutant mice ( 200 ). It is possible that lipid mediators do not play a dominant role in regulating this sepsis model, or that just targeting one PLA 2 isoform is not suffi cient ( 201 ). A study using an antisense approach to inhibit both sPLA 2 -IIa and cPLA 2 ␣ showed improved survival of rats in a cecal ligation and puncture model ( 202 ). Another consideration is that cPLA 2 ␣ mediates the production of diverse lipid mediators that may have opposing effects, some are pro-infl ammatory and others anti-infl ammatory ( 201 ).
The opposing role of cPLA 2 ␣ -derived lipid mediators is illustrated in the regulation of pulmonary fibrosis, a chronic disease characterized by the accumulation of myofi broblasts and deposition of extracellular matrix ( 203 ). Chronic infl ammation often leads to fi brosis due to aberrant wound healing resulting in permanent scarring and organ failure ( 204 ). A study of pulmonary fi brosis induced by bleomycin found that fi brosis, collagen synthesis, and infl ammation were reduced in cPLA 2 ␣ KO mice compared with WT mice ( 205 ). In WT mice, the effects of bleomycin were accompanied by increases in eicosanoids in bronchoalveolar lavage fl uid that were markedly lower in bleomycin-treated cPLA 2 ␣ KO mice. Although disease parameters were not completely attenuated in KO mice, cPLA 2 ␣ in regulating membrane structure and curvature. In addition, a study comparing the brain lipid content of cPLA 2 ␣ WT and KO mice found altered brain phospholipid composition in mutant mice but no differences in the concentrations of nonesterifi ed fatty acids in the plasma or brain ( 191 ). There were decreased levels of esterifi ed arachidonic acid in phosphatidylinositol and decreases of several fatty acids including linoleic, arachidonic, and docosahexaenoic acids in phosphatidylcholine in cPLA 2 ␣ KO mice compared with WT mice. However, an interesting fi nding was the increased incorporation and turnover of AA in phosphatidylinositol and phosphatidylcholine in cPLA 2 ␣ KO mice, perhaps due to compensation from other PLA 2 s. Consequently the differences observed in phospholipid brain composition in cPLA 2 ␣ WT and KO mice may refl ect both a lack of cPLA 2 ␣ and compensation from other enzymes with differences in lipid specifi city as noted by the authors ( 191 ). The lipid alterations in KO mice may infl uence membrane properties contributing to the structural differences noted in cortical neurons, and could also infl uence the response of mutant mice to brain injury. This again raises the question of whether responses of cPLA 2 ␣ KO mice are due to a primary effect of the loss of cPLA 2 ␣ or secondary compensatory changes that occur.

Lung
There is extensive information investigating the role of eicosanoids in regulating lung diseases, including fi brosis, acute lung injury, and cancer. Comparisons of cPLA 2 ␣ KO and WT mice have provided interesting insight into how cPLA 2 ␣ regulates these processes. When the cPLA 2 ␣ KO mouse was initially developed, the role of cPLA 2 ␣ in regulating type I allergic response (anaphylaxis) in the lung was investigated ( 48 ). To induce systemic anaphylaxis KO and WT mice were immunized by intraperitoneal injection of ovalbumin (OVA) followed by intravascular OVA injection to elicit anaphylaxis. In WT mice, there was narrowing of the airway and alveolar thickening that was not evident in KO mice. Lung resistance was lower and recovery was faster in KO mice than WT mice after OVA challenge. Bronchial reactivity to methacholine challenge was dramatically reduced in KO mice. Airway hyperreactivity is also reduced in 5-LO KO mice, suggesting that cPLA 2 ␣mediated production of leukotrienes mediates allergic responses ( 192 ). In contrast, allergic lung responses are exacerbated in COX-1 and COX-2 KO mice compared with WT mice, suggesting PGs play a protective role ( 193 ).
Over the past few years, Wyeth has developed indole compounds that selectively inhibit human cPLA 2 ␣ and have been tested in a sheep model of allergic asthma ( 194 ). Because diverse lipid mediators are implicated in mediating asthmatic responses in this model, they reasoned that broad inhibition with a cPLA 2 ␣ inhibitor would be benefi cial. The indole cPLA 2 ␣ inhibitors reduced late phase bronchoconstriction and airway hyperreactivity in allergen-challenged sheep. The indole inhibitor also blocked allergen-dependent induction of novel genes expressed in cells from asthmatics ( 195 ). One of the most potent indole from collagen-induced arthritis ( 228 ). Based on the collective involvement of eicosanoids in this autoimmune disease, it has been suggested that a cPLA 2 ␣ inhibitor would have therapeutic value by simultaneously blocking eicosanoids from the COX and 5-LO pathways ( 228 ). In rheumatoid arthritis, bone homeostasis is disrupted due, in part, to chronic infl ammation resulting in joint destruction and bone loss. The role of cPLA 2 ␣ in regulating infl ammatory bone resorption induced in mice with endotoxin has been investigated ( 229 ). In response to endotoxin (ip), cPLA 2 ␣ KO mice exhibit less bone loss and reduced PGE 2 production. The results indicate that cPLA 2 ␣ mediates osteoclastic bone resorption due to endotoxininduced PGE 2 production by osteoblasts. The role of cPLA 2 ␣ in regulating autoimmune disease in nonobese diabetic (NOD) mice has also been investigated ( 230 ). NOD mice spontaneously develop autoimmune diabetes that has similarities to human type 1 diabetes. In contrast to collageninduced arthritis, cPLA 2 ␣ plays a protective role in the development of diabetes. NOD mice with genetic ablation of cPLA 2 ␣ have a higher incidence of diabetes than cPLA 2 ␣ WT NOD mice. Results suggest that cPLA 2 ␣ plays a protective role during the diabetic stage, possibly due to the ability of PGE 2 produced by macrophages to suppress the production of the pro-infl ammatory cytokine, TNF ␣ .
Experimental autoimmune encephalomyelitis (EAE) is an infl ammatory disease in mice that has features of the human autoimmune disease multiple sclerosis ( 231 ). There are conflicting data on the role of specific eicosanoids in regulating EAE based on studies with 5-LO KO mice, in which EAE is exacerbated, or using an LTB 4 receptor antagonist, which protects ( 232 ). The role of PGs is not clear because COX-2 KO mice are not protected but a COX-2 inhibitor is benefi cial in EAE ( 233,234 ). A study of EAE induced in cPLA 2 ␣ WT and KO mice on the susceptible C57Bl/6 background found that KO mice are resistant to disease ( 235 ). Results suggested that cPLA 2 ␣ contributes in the induction and effector phase by contributing to the development of Th1 responses. Because the development of the immune system may be altered in KO mice, a follow up study tested the effect of cPLA 2 ␣ , 5-LO, and COX inhibitors on the development of EAE in WT mice ( 236 ). The cPLA 2 ␣ inhibitor, WAY-196025, used at the time of immunization to promote EAE blocked the development of EAE, and when used at the start of disease reduced the duration of EAE relapses. Inhibitors of COX-1/2 and 5-LO also reduced disease severity. Results using the selective cPLA 2 ␣ inhibitor are consistent with the studies using the cPLA 2 ␣ KO mice. A recent study also showed that another selective cPLA 2 ␣ inhibitor (2-oxoamide, AX059) protected rats from EAE that correlated with increased production of Tregs ( 237 ). However, the confl icting studies on the role of specifi c eicosanoids in EAE highlight potential diffi culties in interpreting results from KO mice, which can exhibit compensatory effects or changes that occur during embryogenesis. However, there are also complications in the use of inhibitors involving dosing and time of administration, as well as unknown off-target effects. cPLA 2 ␣ may contribute to the pathogenesis of fi brosis due to production of pro-infl ammatory lipid mediators. This is supported by studies using mice with targeted disruption of 5-LO and the cysteinyl leukotriene receptors 1 and 2, which show reduced bleomycin-induced pulmonary fi brosis (206)(207)(208). In contrast, COX-2 defi ciency promotes pulmonary fi brosis, and an important role for PGE 2 and PGI 2 in protecting the lung has emerged (209)(210)(211)(212). Consequently cPLA 2 ␣ is upstream of pathways that produce mediators with positive and negative effects on lung disease. There has been considerable interest in the role of lysophosphatidic acid (LPA) in promoting fi brosis in many organ systems, including the lung ( 213, 214 ). LPA levels increase during fi brosis and it exerts a profi brotic effect primarily through the LPA 1 receptor. Several pathways for the synthesis of LPA have been described ( 215,216 ). Although there is correlative data that cPLA 2 ␣ may play a role in the initial production of lysophospholipids that are metabolized to LPA by autotaxin, there is as yet little defi nite data to support a role for cPLA 2 ␣ in LPA production ( 217,218 ).
Mouse models have been used to investigate the role of cPLA 2 ␣ in regulating lung cancer. Using the chemical carcinogen urethane to initiate lung cancer, cPLA 2 ␣ KO mice developed fewer lung tumors than WT mice, and much lower levels of tumor-associated PGs ( 219 ). Another study investigated whether the presence of cPLA 2 ␣ in the tumor microenvironment infl uenced lung cancer progression and metastasis ( 220 ). The injection of mouse lung tumor cells into mouse lungs induced primary tumors of similar size in cPLA 2 ␣ WT and KO mice. However, KO mice had fewer secondary tumor metastases, less lymph node spread and increased survival. cPLA 2 ␣ is thought to contribute to tumorigenesis, in part by enhancing the viability and proliferation of endothelial cells leading to formation of vascular networks in the tumor microenvironment ( 221,222 ). This process diminishes the effectiveness of radiotherapy. It has recently been shown that a cPLA 2 ␣ inhibitor (PLA-695) blocks pro-survival signaling in endothelial cells and renders cancer cells radiosensitive in a simulated tumor microenvironment ( 223 ). Combined therapy (PLA-695 and irradiation) also reduces lung tumor growth in mice. Results suggest that cPLA 2 ␣ -mediated production of lysophospholipid metabolites promotes vascular endothelial cell proliferation, migration, and tumor vascularization ( 224 ). Therefore a number of lipid mediators, including arachidonic acid, eicosanoids, and lysophospholipids, may contribute to cancer development.

Autoimmunity
Chronic infl ammation is a characteristic of certain autoimmune diseases, such as rheumatoid arthritis. Collageninduced arthritis in mice has pathological features resembling rheumatoid arthritis and the role of eicosanoids has been studied extensively. Interestingly, there is evidence that both leukotrienes and PGs contribute to disease pathogenesis based on studies using FLAP inhibitors, LTB 4 receptor antagonist, and COX-2 KO mice (225)(226)(227). Therefore, it is not surprising that cPLA 2 ␣ KO mice are also protected Therefore multiple approaches are required to evaluate the role of these pathways in disease pathogenesis. As an example, a recent study used bone marrow transplantation to specifi cally block PGE 2 synthesis, or EP receptor expression, in peripheral immune cells and showed that PGE 2 /EP4 signaling promotes development of EAE in mice ( 238 ). This approach demonstrates the importance of investigating the role of specifi c cell types and receptor pathways for understanding how lipid mediators regulate disease processes in vivo.

CONCLUDING REMARKS
Due to the key role of cPLA 2 ␣ in mediating lipid mediator production and its widespread tissue expression, it has been implicated in regulating homeostatic processes and disease pathogenesis throughout all organ systems. The function of cPLA 2 ␣ in regulating certain physiological systems, such as female reproduction and hemostasis, is more straightforward and can be attributed to specifi c types of lipid mediators. However, elucidating the specifi c mechanism for cPLA 2 ␣ action in disease processes in many cases is not known and understandably diffi cult to evaluate. The cPLA 2 ␣ KO mouse has provided a wealth of information; however, multiple approaches are needed to determine cPLA 2 ␣ function. Genetic differences in mouse strains lead to diffi culties in interpretation of results, and compensatory changes in the conventional KO mouse are a complicating issue. The use of cPLA 2 ␣ conditional KO mice with ablation in specifi c tissues would be useful to bypass changes occurring during embryogenesis and development. The availability of specifi c inhibitors is also important to understand cPLA 2 ␣ function in animal models in vivo, and in human cells ( 2 ). The primary function of cPLA 2 ␣ is due to its role in mediating production of eicosanoids, however, there is evidence that cPLA 2 ␣ regulates processes such as membrane traffi cking that are not linked to production of eicosanoids, adding an additional complexity to understanding its function ( 16,17,74 ). Many studies, primarily from comparisons of cPLA 2 ␣ WT and KO mice, have suggested that cPLA 2 ␣ is a useful therapeutic target for treating a plethora of diseases. There are a number of issues in considering cPLA 2 ␣ as a therapeutic target, especially because most of the studies are based on disease models comparing cPLA 2 ␣ WT and KO mice, which in some cases may not accurately refl ect processes contributing to human disease. This is apparent from the differences observed between mice and humans as a consequence of cPLA 2 ␣ defi ciency. Another complication is that cPLA 2 ␣ is the fi rst regulatory enzyme in the pathway for production of numerous lipid mediators that often have opposing effects, although targeting cPLA 2 ␣ may be benefi cial in some diseases where both 5-LO and COX metabolites contribute to disease, as suggested in studies of asthma and CIA ( 194,228 ). However, the essential role of cPLA 2 ␣ in human health, particularly for maintenance of the small intestine and female reproduction, is also a consideration for targeting cPLA 2 ␣ in disease.