Article

Top Abstract Article References

Phospholipase A₂s (PLA₂s) represent a superfamily of esterases that hydrolyze the sn-2 ester bond in phospholipids releasing free fatty acids and lysophospholipids. The ubiquitous nature of PLA₂s highlights the important role they play in many biological processes, including the generation of proinflammatory lipid mediators such as prostaglandins and leukotrienes, and the regulation of lipid metabolism (Glaser, 1995). Since 1997, PLA₂s have been classified according to their nucleotide sequence (Balsinde et al., 1999). At this time, at least 10 groups have been described (I-X) with each group having at least one member and a majority containing at least two members. Individual members of each group are designated by capital letters. There is significant confusion in the field of PLA₂ because many of the identified PLA₂s are not associated with specific cellular activities and functions, and cellular activities and functions of PLA₂s are not associated with identified PLA₂s. A previous classification system is based on whether the PLA₂ is secreted from the cell (sPLA₂), Ca²⁺-dependent and cytosolic (cPLA₂) or Ca²⁺-independent (iPLA₂). This older classification system still remains and retains some value at this time (Table 1). sPLA₂ isoforms require millimolar amounts of Ca²⁺ for activity, have low molecular masses (14-18 kDa), and demonstrate no selectivity for arachidonylated phospholipids (Types I-III, V, IX, and X). cPLA₂ isoforms are found in the cytosol, have a higher molecular mass (~85 kDa), require micromolar amounts of Ca²⁺ for translocation to membrane phospholipids, and are selective for arachidonylated phospholipids (Types IVA and B). The iPLA₂s are located in both the cytosol (Balsinde and Dennis, 1996a) and membrane fractions (McHowat and Creer, 1998) (Types IVC, VI, VIIB, and VIII). They do not require Ca²⁺ for activity and have molecular masses ranging from 29 to 85 kDa. Within certain PLA₂ groups, such as iPLA₂s, there exist multiple splice variants of the same gene resulting in the expression of two "catalytically distinct" iPLA₂ isoforms (Larsson et al., 1998; Ma et al., 1999).

The involvement of PLA₂s in inflammation is the result of their ability to mobilize arachidonic acid from phospholipids. Arachidonic acid serves as a substrate for prostaglandin H synthase 1 and 2 (COX-1 and -2, respectively), resulting in the production of prostaglandins. Prostaglandins activate cellular receptors resulting in the subsequent initiation of signal cascades involving G-proteins and cyclic AMP (Cirino, 1998). Thus, PLA₂s have an important role in cellular injury via their ability to mediate inflammatory responses.

Measurement and Chemical Inhibition of PLA₂ Activity

Chemical inhibitors of PLA₂s play an important role in elucidating the actions of specific PLA₂s. The study of PLA₂ inhibitors is a critical area of investigation due to the potential pharmacological benefit of these compounds in the treatment of inflammation and cell injury, and as a tool to investigate the role of PLA₂s in physiological functions and in cellular injury and death. As the number and roles of PLA₂s have increased in recent years the need for isoform-selective inhibitors has become critical. Many of the early inhibitors of PLA₂ (e.g., dibucaine, mepacrine) were neither isoform-specific nor potent. More recently, several new PLA₂ inhibitors have been developed. Table 2 lists various PLA₂ inhibitors including the isoforms they inhibit and the type of inhibition. A review by Glaser (1995) discusses the kinetics of PLA₂ inhibition and lists several criteria for potential PLA₂ inhibitors.

Many PLA₂ inhibitors originally thought to be selective for a specific PLA₂ are now known to inhibit other isoforms. For example, methyl arachidonyl fluorophosphonate, an inhibitor of cPLA₂ with an IC₅₀ of ~0.5 µM for purified cPLA₂, also inhibits iPLA₂ purified from murine macrophage-like P388D₁ cells, exhibiting half-maximal inhibition at 0.5 µM (Lio et al., 1996). In fact, most cPLA₂ inhibitors also inhibit iPLA₂ (Balsinde et al., 1999). One protocol that has been used to overcome this problem is to use a combination of inhibitors to differentiate between PLA₂ isoforms. For example, if a process is blocked by methyl arachidonyl fluorophosphonate and arachidonyl trifluoromethyl ketone (AACOCF₃) but not bromoenol lactone (BEL, a specific inhibitor of iPLA₂) then it is likely that cPLA₂ but not iPLA₂ isoforms are involved in the process (Balsinde et al., 1999). The reason most inhibitors of cPLA₂ can inhibit iPLA₂ is both isoforms have a serine in their active site and these inhibitors contain a serine-reactive group. In contrast, BEL does not react with this serine group but interacts with other amino acids in the active site of iPLA₂. Unfortunately, many studies have used these inhibitors to examine the role of PLA₂ isoforms without verifying that selective inhibition of PLA₂ activity occurs or have just used one inhibitor. Thus, the use of chemical inhibitors of PLA₂ requires careful characterization to ensure that selective inhibition occurs in a given model.

Numerous studies have used arachidonic acid release as a marker for PLA₂ activity. Typically this method involves prelabeling cells with [³H]arachidonic acid followed by the measurement of [³H] release from cells upon exposure to an agonist or toxicant. There are several problems with this method. First, arachidonic acid release is an indirect measure of PLA₂ activity. Second, multiple pathways may cause free arachidonate production, resulting in an overestimation of PLA₂ activity. Arachidonic acid release may be due to activation of intracellular phospholipid metabolism independent of PLA_2, such as arachidonoyl-CoA synthetase, CoA-dependent acyltransferase, and CoA-independent transacylase, and contribute to [³H] release (Lio et al., 1996). Apoptotic bodies that are released as a result of apoptosis also contain arachidonic acid and can contribute to [³H] release. Another potential problem is the phospholipid pools that arachidonic acid is incorporated into are not usually determined (e.g., plasmenylcholine, phosphatidylcholine, and alkylacyl glycerophosphorylcholine). This should be determined under control conditions to determine whether arachidonic acid incorporation has reached equilibrium and to determine whether the labeled arachidonic acid is preferentially incorporated into one specific phospholipid pool. Finally, care should be taken to ensure that arachidonic acid release occurs before increases in markers of cell death.

Direct measurement of PLA₂ activity using synthetic phospholipid substrates that are also the endogenous phospholipids for PLA₂-catalyzed hydrolysis can alleviate some of the above problems. Although some of these substrates are available commercially, many are not. However, they can be synthesized, and PLA₂ activity measurements obtained using these substrates can be determined in subcellular fractions. Using plasmenylcholine, phosphatidylcholine, and alkylacyl glycerophosphorylcholine substrates, it is possible to determine whether PLA₂ activity is influenced by the covalent linkage of the sn-1 fatty acid. The selectivity of PLA₂ for arachidonylated substrates can be determined using substrates with oleic acid or arachidonic acid at the sn-2 position. Proper use of these substrates may indicate if a PLA₂ isoform has a preference for a specific phospholipid (i.e., those with covalent linkages at the sn-1 or sn-2 position). When studying the effect of PLA₂ inhibition on cellular injury and death, careful selection of multiple inhibitors is key with special attention paid to experiments verifying that PLA₂ activity is being inhibited. Activity should be measured using a method that relies on the use of endogenous substrates of PLA₂ (i.e., plasmenylcholine, phosphatidylcholine, alkylacyl glycerophosphorylcholine).

Molecular Modulation of PLA₂ Activity

A number of studies have used advances in molecular biology to overcome some of the problems involved with chemical inhibitors of PLA₂. For example, cell lines that overexpress certain types of PLA₂ isoforms, antisense oligonucleotides that decrease specific PLA₂ isoforms, and transgenic mice that are deficient or "overexpress" PLA₂ isoforms have been developed. Overexpression of PLA₂ isoforms allows one to study the effect of increased activity of a specific PLA₂ isoform. Sapirstein et al. (1996) overexpressed human cPLA₂ and sPLA₂ in LLC-PK₁ cells and used these cells to study the role of these PLA₂ isoforms in oxidant-induced cell injury (see the section on oncosis).

Antisense oligonucleotide technology has been used to decrease specific PLA₂ isoforms. Locati et al., (1996) used an antisense oligonucleotide directed against codons 4 through 9 of human cPLA₂ to produce a 57% decrease in cPLA₂ protein levels in cultures of human monocytes. When these cells were exposed to monocyte chemotactic protein, arachidonate release was 19% of cells treated with the same oligonucleotide with four mismatched bases or an unrelated antisense oligonucleotide. Woo et al. (2000) also showed that the same antisense oligonucleotide to cPLA₂ inhibited Rac-mediated c-Jun N-terminal kinase activation in Rat-2 fibroblast cells if it is cotransfected with the Rac plasmid. Interestingly, the effect of the antisense oligonucleotide was similar to cells treated with AACOCF₃. Thus, antisense oligonucleotide technology provides an additional mechanism by which levels of a specific PLA₂ isoform can be decreased to explore its role in cell injury and death.

"Knockout" mice that lack specific PLA₂ isoforms are another model that can be used to study specific PLA₂ functions. A homozygous null (cPLA₂^/) mouse has been produced that develops normally and has weight gain and life span equal to that of wild type mice (cPLA₂^+/+) (Bonventre, 1999). The cPLA₂^/ mice did display abnormal reproduction, resulting in small litters and a high death rate. Removal of offspring after 18 days of pregnancy resulted in normal mice, indicating that cPLA₂ plays a critical role in parturition. This model has been used to study the role of cPLA₂ in ischemic injury to the kidney, brain, and other organs (Bonventre, 1999). In general, studies demonstrate that deletion of cPLA₂ results in decreased postischemic injury. Deletion of cPLA₂ also protects against 1-methyl-4-phenyl-1,2,3,6-tetrahyropyridine (MPTP)-induced injury in the brain as cPLA₂^/ mice were able to resist the dopamine-depleting effects of MPTP compared to wild type mice (Klivenyi et al., 1998). The decrease in injury seen in these studies is hypothesized to result from decreased production of lipid mediators of injury such as eicosanoids, lysophospholipids, and oxidative species that are derived from metabolism of arachidonic acid by cyclooxygenases and lipooxygenases.

In contrast to knockout mice, investigators have over expressed specific PLA₂ isoforms in mice to study its role in mediating injury. Laine et al. (2000) over-expressed group II PLA₂ (sPLA₂) in sPLA₂ deficient mice and reported that mice expressing sPLA₂ were more resistant to Staphylococcus aureus and Escherichia coli infection than sPLA₂-deficient mice. Specifically, mice over-expressing group II PLA₂ had lower rates of mortality and less bacterial growth in body fluids and organs when compared with mice lacking group II PLA₂.

The use of over-expression, antisense oligonucleotides, and knockout models should increase our knowledge of the roles and mechanisms of specific PLA₂ in cell injury and death. Future efforts will undoubtedly focus on applying the technologies perfected with cPLA₂ and sPLA₂ to other isoforms to elucidate the overlapping roles of these enzymes in mediating both oncosis and apoptosis.

The Role of PLA₂ Isoforms in Oncosis

Although the role of PLA₂ in oncosis (cell death characterized by cell and organelle swelling, ATP depletion, increased plasma membrane permeability, release of macromolecules, and inflammation) has been studied over the past 20 years, much remains unknown. It was originally postulated that during oncosis, PLA₂ activity increased, accelerated membrane phospholipid hydrolysis, and, in turn, increased plasma membrane permeability and cell lysis (Sevanian, 1988). Typically, experiments demonstrated that the PLA₂ inhibitors, mepacrine or dibucaine, decreased cell lysis following an injurious insult. Unfortunately, in many cases the investigators did not document increases in PLA₂ activity or verify that PLA₂ inhibitors were indeed inhibiting PLA₂. Furthermore, because the number of PLA₂ isoforms known and their characteristics were limited, the experiments were crude in nature. Still some useful information can be gained from these studies. For example, dibucaine and mepacrine decreased the toxicity of tert-butyl hydroperoxide (a model oxidant), and the reduced toxicity correlated with their inhibition of arachidonic acid release (Schnellmann et al., 1994). The PLA₂ inhibitors did not decrease the ability of antimycin A (a mitochondrial inhibitor) nor carbonyl cyanide p-trifluoromethoxyphenylhydrazone (an uncoupler of oxidative phosphorylation) to cause toxicity. Thus, even though specific PLA₂ inhibitors were not used in this study it was determined that the role of PLA₂ in renal cell oncosis depends on the stimulus.

The development of more selective inhibitors has increased our knowledge of the role of PLA₂s in oncosis. For example, Kohjimoto et al. (1999) showed that preincubation of Madin-Darby canine kidney cells with AACOCF₃ (5-10 µM) significantly reduced the toxicity of oxalate. Arachidonic acid release occurred before cell death and was inhibited by the cPLA₂ inhibitor AACOCF₃, but not inhibitors of sPLA₂ (oleyloxyethyl phosphorylcholine; 20 µM) and iPLA₂ (BEL; 10 µM). Thus, this study used multiple inhibitors of PLA₂ and the measurement of PLA₂ activity (albeit, an indirect measure) to suggest that cPLA₂ is involved in renal cell oncosis.

Sapirstein et al. (1996) showed that cPLA₂ was involved in oxidant-induced oncosis by over-expressing cPLA₂ or sPLA₂ in LLC-PK₁ cells and demonstrating that cells over-expressing cPLA₂ were more susceptible to H₂O₂ toxicity, whereas over-expression of sPLA₂ did not increase H₂O₂ toxicity. Briefly, cPLA₂ and sPLA₂ were over-expressed in LLC-PK₁ cells, and activity was determined by measurement of both 1-steroyl-2-[1-¹⁴C]arachidonyl phosphatidylcholine as a substrate in vitro and release of [³H]arachidonic acid from cells. It was shown that over-expression of cPLA₂, but not sPLA_2, increased H₂O₂ toxicity. The increase in H₂O₂ toxicity was not due to decreases in the activity of the antioxidant defense enzymes, superoxide dismutase, catalase, or glutathione peroxidase. Interestingly, chelation of cytosolic-free Ca²⁺ protected cells from H₂O₂ toxicity, suggesting a key role for Ca²⁺ in the mediation of cPLA₂-mediated oncosis in renal cells.

Data from many studies implicate cPLA₂ as an important mediator of oxidant damage in cells; however, the exact mechanism of cPLA₂-mediated cellular injury has yet to be determined. Sapirstein et al. (1996) hypothesized that oxidant-induced damage may direct cPLA₂ activity to a specific subcellular location where it produces injury. An oxidant-induced rise in cytosolic free Ca²⁺ and, in turn, Ca²⁺ binding to the Ca²⁺-lipid-binding domain of cPLA₂ would initiate translocation to intracellular membranes. Experiments to determine alterations in cPLA₂ cellular localization before and during toxic injury would greatly aid in elucidating the site of action of cPLA₂.

Sevanian (1988) proposed that cellular insults result in prolonged activation of PLA₂ isoforms (Fig. 1). Consequently, many of the products formed by hydrolysis of phospholipids (free fatty acids, lysophospholipids) may decrease membrane integrity by acting as detergents and altering membrane fluidity. In addition, the release of membrane phospholipids as a result of oxidant-induced lipid peroxidation and/or PLA₂ metabolism may decrease membrane integrity independent of free fatty acids or lysophospholipids. The free fatty acid and lysophospholipids may serve as precursors for biologically active metabolites and promote inflammation and may further increase the activity of PLA₂s by themselves (see below) (McHowat et al., 1993). Alternatively, or perhaps in tandem with released membrane phospholipids, increased cytosolic-free Ca²⁺ can increase cPLA₂ activity. If the accumulation of free fatty acids and lysophospholipids or the loss of membrane phospholipids contributes directly to cellular injury and death, then inhibition of PLA₂s would be protective.

View larger version (100K): [in this window] [in a new window]   Fig. 1.   Potential roles of PLA2s in cell injury. 1, an injury stimulus may increase cytosolic free Ca2+ ([Ca2+]i), activate PLA2 or disrupt membrane phospholipids directly; 2, increases in [Ca2+]i facilitate the translocation of cPLA2 to the cellular membrane where it, as well as membrane bound and other PLA2 isoforms, metabolizes phospholipids to arachidonic acid and a lysophospholipid. All of these processes can lead to decreased membrane phospholipids. 3, the injury stimulus may also cause membrane lipid peroxidation leading to the formation of oxidized arachidonylated and other phospholipids; 4, PLA2 may function to remove peroxidized phospholipids and minimize membrane dysfunction; 5, the loss of membrane phospholipids, increases in lysophosphatidyl lipids acting as detergents, and lipid peroxidation may all decrease membrane integrity and/or function.

In contrast to the above scenario it has been proposed that PLA₂ isoforms may serve to decrease free radical-induced membrane phospholipid damage by hydrolyzing oxidized phospholipids from the membrane. The hydrolysis of oxidized phospholipids by PLA₂s facilitates the removal of these damaged lipids from the cells and decreases toxicity (Fig. 1) (Salgo et al., 1993). Furthermore, subsequent reacylation (by enzymes such as CoA-dependent acyltransferase, and CoA-independent transacylase) of the phospholipids results in the return of normal functions. This cycle is analogous to DNA repair of damaged bases, but for membrane phospholipids. If PLA₂s that hydrolyze damaged phospholipids from membranes are inhibited then cellular injury would increase. In support of this hypothesis, oxidized phospholipids are substrates for cPLA₂ and can decrease the Ca²⁺ requirement for purified cPLA₂, thereby enhancing its activity (Rashba-Step et al., 1997). Key determinants of whether PLA₂s are increasing or decreasing toxicity is the location of the membrane phospholipids being released, the type of phospholipid (phosphatidylcholine, phosphatidyethanolamine, phosphatidylserine, etc.), the PLA₂ isoforms responsible, and the stimulus of injury. For oxidative injury, incurred by H₂O₂ or other toxicants, cPLA₂ activity appears to increase toxicity and most studies report a central role for Ca²⁺ in mediating oncosis. Very little work has examined the role of iPLA₂ in oncosis.

Many studies have demonstrated that sPLA₂s in snake or bee venom are responsible for cellular injury. There is a large amount of evidence to suggest a role for inflammation, but studies have shown that the toxicity of sPLA₂s may be independent of its ability to produce arachidonic acid. Furthermore, investigators have shown that sPLA₂ requires specific membrane phospholipids to mediate cellular injury. For example, recombinant human and venom-derived sPLA₂ are indirectly cytolytic to human erythrocytes, erythroleukemia, and U937 cells in a manner dependent on the presence of liposomal phospholipids (phosphatidylcholine and phosphatidylethanolamine) (Vadas, 1997). Interestingly, human sPLA₂ was cytolytic only in the presence of phosphatidylethanolamine. Thus, phospholipid metabolites of PLA₂ other than arachidonic acid can be mediators of cellular injury (i.e., lysophospholipids). However, sPLA₂ is believed to be responsible for the bulk of arachidonic acid released into the extracellular milieu as a result of its extracellular location/action (Balsinde and Dennis, 1996b). This is also believed to occur because of the increase in oxidized phospholipids being translocated to the extracellular surface of the membrane, secondary to cellular injury, which make them more accessible to sPLA₂ (Balsinde and Dennis, 1996b).

Role of PLA₂ Isoforms in Apoptosis

In contrast to oncosis, apoptosis is characterized by cell shrinkage, chromatin condensation, plasma membrane budding, caspase activation, and is ATP-dependent. Similar to the role of PLA₂s in oncosis, the role of PLA₂s in apoptosis appears to be dependent on the stimulus of apoptosis and the cell type being targeted. For example, Atsumi et al. (1998) suggested that cPLA₂ does not have a role in Fas-induced apoptosis as caspase-3 cleaved and inactivated cPLA₂ in human leukemic U937 cells exposed to Fas (Fig. 2). Enari et al. (1996) supported this hypothesis by demonstrating that cPLA₂ was not needed for Fas-induced apoptosis in mouse L929 cells expressing human Fas. Finally, cPLA₂ has been shown to be a substrate for human caspase-1 and -8, as both caspases degraded and inactivated cPLA₂ in vitro (Adam-Klages et al., 1998; Luschen et al., 1998). Thus, cPLA₂ does not appear to play a significant role in Fas-mediated apoptosis. Despite these studies several questions remain. For example, is cleavage and inactivation of cPLA₂ by caspases a required event for Fas-induced apoptosis or is cPLA₂ inactivated to decrease the formation of proinflammatory prostanoids during apoptosis?

View larger version (15K): [in this window] [in a new window]   Fig. 2.   Differences in the roles of PLA2s in TNF and Fas-induced apoptosis. It has been proposed that activation of caspases (e.g., caspase-3) results in the activation of cPLA2 in TNF-induced apoptosis. cPLA2 may be needed for the activation of downstream caspases, arachidonic acid production, and the progression of TNF-induced apoptosis. It is not known if cPLA2-mediated activation of caspases is a direct result of arachidonic acid production or lysophospholipids and other metabolites. The role of iPLA2 during TNF-induced apoptosis is not known. Fas-induced apoptosis does not require cPLA2 activity and cPLA2 is degraded by caspase-3. Caspase-1 and -8 can also degrade and inactivate cPLA2 in vitro. Inhibition of iPLA2 during Fas-induced apoptosis decreased arachidonic acid release and reduced cell death, but the mechanism of such events and the role of caspases in these processes are not known.

Although cPLA₂ is not needed for Fas-induced apoptosis, several studies report that iPLA₂ mediates several signal transduction processes associated with apoptosis such as Fas-induced arachidonic acid release and membrane remodeling (Fig. 2) (Balsinde and Dennis, 1996; Atsumi et al., 1998). One study showed that Fas-induced arachidonic acid release in U937 cells undergoing apoptosis is mediated by iPLA₂ and inhibition of iPLA₂ decreased Fas-induced cell death (Atsumi et al., 1998). Furthermore, iPLA₂ levels did not decrease while cPLA₂ inactivation occurred via caspase-mediated cleavage. Other functions for iPLA₂ in apoptosis may include the generation of lipid signaling molecules that regulate ion channel activity (Ma et al., 1997). Although many of these processes have been proposed as key events in apoptosis, very little work has been done correlating these events to the genesis of apoptosis and the activity of iPLA₂.

In contrast to the hypothesis that cPLA₂ does not play a role in apoptosis, several studies have reported that cPLA₂ is needed for apoptosis and that inhibition of cPLA₂ decreased apoptosis. cPLA₂ is required and appears to be the rate-limiting step in tumor necrosis factor  (TNF)-induced apoptosis in several cell types (Enari et al., 1996; Ilic et al., 1998). In this case, cPLA₂ is activated by caspase-3 and is needed to activate downstream caspases (Wissing et al., 1997), but the mechanism of action of PLA₂ on caspases downstream of caspase-3 is not known (Fig. 2). These studies, as well as those listed above, demonstrate that Fas and TNF cause apoptosis by different pathways and that cPLA₂ has distinctly different roles in each pathway (Fig. 2). In a system where apoptosis was caused by removal of extracellular matrix survival factors (fibronectin) or focal adhesion kinases, inhibition of cPLA₂ with AACOCF₃ significantly improved the survival of cell lines undergoing apoptosis (Ilic et al., 1998). Interestingly, the activation of caspase and protein kinase C in this model were thought to occur downstream of cPLA₂ activation. Although cPLA₂ inhibition has been shown to decrease apoptosis, several studies demonstrate that inhibition of PLA₂ increases apoptosis. For example, Miao et al. (1997) reported that apoptosis in human umbilical vein endothelial cells induced by deprivation of fibroblast growth factor and serum was increased by preincubation of the cultures with PLA₂ inhibitors (manoalide, 3-(4-octadecyl)benzoylacrylic acid, and oleyloxyethylphosphorylcholine).

Whether cPLA or iPLA₂ has no role in apoptosis, is required for apoptosis, or inhibit apoptosis depends on the stimuli of injury and the model system studied. To increase our knowledge of the role of these PLA₂s in apoptosis, a precise examination of their activity using endogenous substrates and correlation to the activation of the apoptotic cascade (caspase-9 activation, cytochrome c release, etc.) is needed. Balsinde and Dennis (1996b) have hypothesized that iPLA₂ and cPLA₂ have distinct roles, with iPLA₂ responsible for membrane remodeling (maintenance of membrane integrity and phospholipid content) and cPLA₂ responsible for arachidonic acid production. Similarly, PLA₂ localization also may be a key determinant in the role of PLA₂ isoforms in cellular injury. For example, membrane-bound PLA₂ isoforms may serve to regulate membrane fluidity/integrity during apoptosis, although cPLA₂ isoforms may respond to fluxes in Ca²⁺ and the release of membrane phospholipids.

Activation of PLA₂

A key determinant of the role of PLA₂s in oncosis and apoptosis is the mechanism of PLA₂ regulation/activation during these processes. It is likely that events that cause oncosis elicit a set of signals that activate PLA₂s differentially than the signals elicited when apoptosis is induced. Increased PLA₂ activity can be caused by agents that produce alterations in membrane phospholipids that result in the exposure of preferential lipid substrates (Sevanian, 1988; Salgo et al., 1993; Sapirstein et al., 1996). For example, oxidative stress may lead to the rearrangement of membrane phospholipids such that the sn-2 fatty acids are more accessible to PLA₂ (Balsinde and Dennis, 1996b). Furthermore, excessive toxicant and/or oxidant injury may result in the release of intact membrane phospholipids themselves exposing the sn-2 ester bond. As the result of either of the above processes, PLA₂ activity increases and any agent that inhibits the access of PLA₂ to either exposed or released membrane phospholipids would decrease PLA₂ activity. In support of this hypothesis, agents that bind to phospholipids such as lipocortins and annexins can inhibit the ability of PLA₂ to hydrolyze phospholipids (Buckland and Wilton, 1998).

Alterations in cellular membrane phospholipid integrity may be one process that results in modifications in PLA₂ activity but several studies have reported that increased PLA₂ activity occurs independently of significant phospholipid alterations. One hypothesis is that cellular injury may cause a rise in intracellular Ca²⁺, activation of protein kinase C (PKC) and PKC-mediated activation of PLA₂. In support of this hypothesis, Chen et al. (1999) demonstrated a link between increases in Ca²⁺, PKC- activation, and the activation of cPLA₂, and Akiba et al. (1999) reported that zymosan stimulated iPLA₂-mediated arachidonic acid release via a PKC-dependent mechanism. In vitro, both PKC and protein kinase A (PKA) can phosphorylate cPLA₂ but the phosphorylation did not lead to a corresponding increase in activation (Leslie, 1997). In vivo it is not known whether PKC and PKA regulate cPLA₂ by direct phosphorylation.

The signaling cascade involved in activation of cPLA₂ by mitogen-activated protein kinase (MAPK) has been studied also. Nemenoff et al. (1993) demonstrated that p42 MAPK phosphorylated cPLA₂ and increased its activity in vitro, and later studies demonstrated this event in cell lines (Leslie, 1997). The phosphorylation of cPLA₂, at serine 505, occurs before the increases in intracellular Ca²⁺ that facilitate the binding of the lipid-binding domain of cPLA₂ to phospholipids, promoting its translocation to cellular membranes and arachidonic acid release. Recently, a negative feedback loop for cPLA₂ activation by MAPK has been proposed (Xing et al., 1999). In this model, purinergic receptor activation results in MAPK activation followed by activation of cPLA₂ and an increase in arachidonic acid. The increase in cellular arachidonate levels is followed by an increase in prostaglandin E₂, which in turns activates adenyl cyclase and PKA. Activation of these enzymes decreases MAPK and cPLA₂ activity. This feedback loop only was seen with an agonist of purinergic receptors and was not seen in adrenergic receptor activation of MAPK, suggesting that cPLA₂ regulation can occur by multiple processes that appear to depend on the stimuli. cPLA₂ can be activated by pathways independent of MAPK also as studies have shown that okadaic acid can increase cPLA₂ activity in a Ca²⁺-independent manner and induce phosphorylation of cPLA₂ at serine 727 rather than serine 505 (de Carvalho et al., 1996). Serine 727 is not within a consensus site of MAPK but appears to be a site for a basotrophic kinase (Leslie, 1997).

If PLA₂ activation in a given model depends on PKC, PKA, cAMP, or MAPK activation then inhibition of these compounds may inhibit PLA₂ isoforms during cellular injury. Understanding of the signaling pathways involved in the activation/deactivation of PLA₂ during cellular injury will point to key events that can be used to prevent the cellular injury. Furthermore, to date, there is limited information regarding the regulation of iPLA₂ or sPLA₂ by these pathways.

Future Directions

The role of PLA₂ in cell injury and the potential benefits of pharmacological inhibition have been studied for over 20 years. Within the last 5 years additional PLA₂ isoforms have been identified and characterized, including the discovery of catalytically different splice variants of iPLA₂, but the role of these new isoforms in cell injury needs to be explored. The use of over-expression and knockout mice models and antisense technology has increased our knowledge of these enzymes but these advances need to be expanded to encompass more PLA₂ isoforms and cell injury studies. This information along with transgenic mice models can be used to design therapeutic treatments for organ (e.g., brain, kidney) injury, develop more potent anti-inflammatory inhibitors, and study the mechanisms of apoptosis and oncosis. Careful analysis of PLA₂ isoforms in general and in specific models must be considered at every step.