The 85-kD Cytosolic Phospholipase A2 Knockout Mouse
1999; American Society of Nephrology; Volume: 10; Issue: 2 Linguagem: Inglês
10.1681/asn.v102404
ISSN1533-3450
Autores Tópico(s)Retinoids in leukemia and cellular processes
ResumoPhospholipases are enzymes that hydrolyze phospholipids (Figure 1). These enzymes may be placed into two broad classes: (1) acyl hydrolases (phospholipase A1 [PLA1], which hydrolyzes the lipid at the sn-1-acyl ester bond; phospholipase A2 [PLA2], which acts at the sn-2 position; and phospholipase B [PLB], which acts at both the sn-1 and sn-2 positions); and (2) phosphodiesterases (phospholipase C [PLC], which cleaves the glycerol-phosphate bond; and phospholipase D [PLD], which removes the base from the phospholipid). PLA2s generate a free fatty acid and lysophospholipid. They are involved in signal transduction, dietary lipid degradation, microbial defense, and represent an important component of snake venoms and pancreatic secretions. Because the sn-2 position of phospholipids of mammalian cells is enriched with arachidonic acid, the substrate for cyclooxygenases, lipoxygenases, and P450 monooxygenases, the regulation of PLA2 activity has important implications for the control of prostaglandin, leukotriene, and other eicosanoid production. Arachidonic acid can also be nonenzymatically oxidized to isoprostanes. Although arachidonic acid can be released from phospholipids by mechanisms not involving PLA2 (1), PLA2s determine most arachidonic acid release in response to physiologic and pathophysiologic stimuli. The degree to which regulation of eicosanoid synthesis also depends on changes in cyclooxygenase or lipoxygenase activities is not always easy to determine because the latter enzymes are also regulated and there are multiple PLA2, cyclooxygenase, and lipoxygenase enzymes with different temporal regulatory features and localization within the cell. Furthermore, there may be regulatory interactions between PLA2s and expression of genes encoding cyclooxygenases and lipoxygenases.Figure 1.: Sites of action of phospholipases on phospholipids.Arachidonic acid and its metabolites, the eicosanoids, have been implicated in the control of a number of physiologic processes in the kidney (Table 1), including regulation and distribution of renal blood flow, sodium and water reabsorption by the nephron, urinary concentrating ability, GFR, control of renin release, and activation of potassium channels, to name a few. In other tissues, these enzymes have been implicated in control of voltage-dependent (2,3) and ligand-gated channels (4,5), modulation of neurotransmitter uptake (6,7) and release (8,9), blood vessel tone (10), and enhancement of the phosphorylation of cellular proteins (11).Table 1: Roles proposed for eicosanoids in the kidneyIn addition to its role in physiologic processes, PLA2 activity is important for pathophysiologic states. PLA2 activation is associated with membrane degradation, changes in plasma and mitochondrial membrane bioenergetics and permeability (12, 13, 14), and altered behavior of transporters. Enzymes that metabolize arachidonic acid produce free radicals with destructive potential (15). Metabolites of arachidonic acid can cause vasoconstriction and hence compromise blood flow and oxygen delivery to tissue. Metabolites can serve as, or modulate the production of, inflammatory mediators (16,17) and hence are important modulators of the inflammatory response including that associated with glomerulonephritis and tubular interstitial diseases of the kidney (18). Leukotrienes potentiate neutrophil adhesion to the endothelium (19). If the PLA2 substrate lipid is a 1-O-alkyl, 2 acyl phospholipid, the 1-O-alkyl-lysophospho-lipid (lyso-PAF) produced can then be acetylated by a CoA-independent transacylase at the sn-2 position to form platelet-activating factor (PAF) (20). PAF is chemotactic for neutrophils. Characteristics of PLA2 The number of identified members of the PLA2 superfamily has grown dramatically over the past few years. Different PLA2 have different substrate specificities, Ca2+ sensitivities, cellular localizations, and inhibitor susceptibilities (21). To understand the role of PLA2 in any biologic process, therefore, it is important to understand which enzymes participate in the process being studied, and how each one is regulated. The PLA2 superfamily can be divided into at least 10 groups which can be further subdivided based on their amino acid structure and other characteristics (21). There are both secretory and intracellular forms of PLA2. The secretory forms of the enzyme are prominent components of pancreatic exocrine secretions, snake and insect venoms, and inflammatory joint fluid. These secreted enzymes are generally small proteins (molecular weight varying from 12,000 to 18,000 daltons), which have very high specific activities and generally require a calcium concentration greater than 0.1 mM to be maximally active (22). Many of these small molecular mass enzymes have been purified, and their amino acid sequences have been determined (23). Until recently, much less was known about intracellular nonsecreted forms of PLA2. It was not known which forms were involved in the hormonal regulation of arachidonic acid release. Both Ca2+ -dependent and -independent forms of intracellular PLA2 have now been identified, and some intracellular forms have specific activities approaching that of the secretory forms of PLA2 (24). Intracellular forms of the PLA2 may be cytosolic, associated with membrane compartments, or integral membrane proteins. Our laboratory has focused on the study of the group IV cytosolic 85-kD PLA2 (cPLA2) (25, 26, 27, 28, 29, 30, 31). cPLA2 is widely distributed in expression in human and rodent tissue. cPLA2 is cytosolic in localization but becomes associated with membranes in response to increases in intracellular free calcium concentrations in the physiologic (nanomolar to low micromolar) range. cPLA2 has a preference for arachidonic acid at the sn-2 position of its substrate phospholipid. We initially identified this large molecular mass form of PLA2 in mesangial cell extracts and found that the form was regulated in a stable manner by calcium and protein kinase C in response to vasopressin (32). We also found it to be increased in activity in cytosolic extracts derived from mesangial cells exposed to epidermal growth factor (33). We purified cPLA2 from the cytosolic extracts of rat and bovine kidneys (29, 34). Mouse and human cDNA for cPLA2 were cloned (35, 36). A predicted protein size of 85.2 kD was determined. cPLA2 is not dependent on calcium for its catalytic activity as are many other forms of PLA2; calcium is required for cPLA2 to associate with its substrate (32, 37). cPLA2 translocates to the nuclear membrane in response to increases in cytosolic free calcium concentration (38, 39). cPLA2 has been implicated in many physiologic and pathophysiologic processes in the kidney (Table 2), as well as other organs and cell culture systems (Table 3). Most of these conclusions regarding the role of cPLA2 derive from studies performed with inhibitors or, in a small number of cases, with antisense constructs. The specificity of inhibitors, however, is usually inversely proportional to the length of time that the agent is available for general use. When initially introduced, there are generally strong claims regarding specificity that become more difficult to support as the agent is recognized to have effects other than the ones for which the inhibitor was originally designed. In addition, the large number of PLA2 enzymes with cross-talk among different forms (40) makes it particularly difficult to assign a particular pathophysiologic or physiologic response to a specific form of the enzyme, using traditional pharmacologic approaches.Table 2: Conditions where kidney cell cPLA2 has been implicatedaTable 3: Conditions where cPLA2 has been implicated (excluding kidney)aTable 3: ContinuedThe cPLA2 Knockout Mouse To address the physiologic and pathophysiologic roles of cPLA2, we have created a homozygous null (cPLA2-/-) mouse (28). We used gene targeting in mouse embryonic stem cells to disrupt an exon containing Ser228, which is critical for the catalytic activity of the cPLA2 gene (41), generating a null allele. The absence of cPLA2 in homozygous cPLA2-/- mice was confirmed by Western blotting (Figure 2), biochemical activity measurements of cell extracts, and column chromatography of kidney and brain. Heterozygous (cPLA2+/-) mice appeared normal and, when interbred, yielded litters of normal size with a predicted Mendelian genetic distribution. Homozygote null mice develop normally, gain weight at a rate equal to that of wild-type animals, and have life spans of > 1.5 yr. There were no significant differences in mean arterial pressure between wild-type cPLA2+/+ and homozygous cPLA2-/- siblings. Likewise, there were no differences in baseline arterial pH, Pco2, or rectal temperature between the two groups. The presence of normal-appearing viable cPLA2-/- progeny indicate that cPLA2 is not necessary for normal embryonic development. This does not mean that the protein is without a role in normal development, but if it does play a role, there are other compensatory influences in the cPLA2-/- animal that maintain a normal developmental phenotype in the absence of cPLA2.Figure 2.: Western blot analysis of organs from cytosolic phospholipase A2 -/- (cPLA2 -/-) and cPLA2 +/+ littermates. Immunodetection of cPLA2 was performed on cytosolic fractions, using a polyclonal antibody raised against the first 129 amino acid residues of cPLA2. Note the absence of cPLA2 in the organs from the null mice. LLC-PLA2 cells are LLC-PK1 kidney epithelial cells that have been stably transfected with the cPLA2 cDNA. Reprinted with permission from reference (28).An abnormality that did become apparent when we tried to breed the mice was a dramatic effect on reproduction in the cPLA2-/- females (28). They produce small litters and the pups usually do not survive. The dead offspring were larger than normal but had no organ abnormalities on post mortem examination. Removal of offspring at day 18 of pregnancy from a cPLA2-/- mother yielded four normal viable pups. These results are reminiscent of the findings when mice null for the Ptgs 1 gene (which encodes cyclooxygenase-1) are interbred. Almost all of the pups born when both parents are homozygous null are dead, although the litters are normal in size. In contrast to our findings, however, the Ptgs 1 mutant females produce normal viable litters if the males are heterozygotes or Ptgs 1+/+ (42). Mating to cPLA2+/+ or cPLA2+/- males did not result in normal reproductive capacity in the cPLA2-/- mice. Male cPLA2-/- mice had normal fertility when crossed with wild-type or heterozygous females. There are defects in various aspects of reproduction in cPLA2-/- females: ovulation, oocyte transport or implantation, and parturition. cPLA2 is present in ovarian tissue (43). Uterine PLA2 activity is increased at the same time that PGE2 and PGF2α levels are increased during ova implantation (44). In cPLA2-/- females that failed to deliver after matings, we observed that initial plug formation was followed several days later by formation of another vaginal plug. Pregnancy in the cPLA2-/- females often fails after implantation (approximately day 4.5) during the period when decidualization takes place in normal murine pregnancy. Prostaglandins are involved in the increased vascular permeability of the endometrium during implantation and in the induction of decidualization (45). Absence of a fetal factor that is elaborated through a cPLA2-dependent pathway may also contribute to the reproductive abnormalities. Eicosanoids have been implicated in early fetalmaternal interactions (46). PGE2 and PGF2α, produced by the maternal-fetal-placental compartments, which include decidua, chorion, amnion, trophoblasts, and myometrial smooth muscle cells, are important for uterine contractions and cervical dilation during labor (47, 48). Although other forms of PLA2 have been identified along with cPLA2 in the uterus, including the cervix (48, 49), the dead offspring produced by the cPLA2-/- female mice indicate that cPLA2 is a critical determinant of normal parturition. Thus, despite the potential for redundancy of multiple PLA2 enzymes, our data show that these other enzymes cannot rescue the fertility defects associated with cPLA2 deficiency. We have begun to explore pathophysiologic roles of cPLA2 in whole animal models, as well as cells isolated from the animals. The cPLA2-/- mouse affords the opportunity to study the effects of this enzyme on ischemic injury to the kidney, brain, and other organs. In the brain, the effects of deletion of cPLA2 on eicosanoid, lysophospholipid, PAF, hydroxyl ion production, glutamate, and activation of other forms of PLA2 in response to ischemia can be examined. PLA2s and arachidonic acid metabolites have been implicated in postischemic neuronal injury (26, 27, 31, 50, 51). The pathophysiology of brain ischemia and a number of other neuropathologic conditions involve excessive glutamatergic action. Both N-methyl-D-aspartate (NMDA) and non-NMDA glutamate receptor agonists cause PLA2 activation and arachidonic acid release (52, 53, 54). We have found that glutaminergic stimulation results in a robust stable activation of cPLA2 (30). PLA2 activity increases glutamate binding to non-NMDA receptors (4, 55) and potentiates the action of glutamate at NMDA receptors (5). Another mechanism by which PLA2 activation potentiates glutamate actions is through the action of arachidonic acid to both prolong the action of glutamate by blocking glutamate reuptake (7, 56, 57) and increase presynaptic release of glutamate (58). Furthermore, eicosanoids can be chemotactic and enhance the inflammatory response seen after ischemia with resulting microvascular occlusion and enhanced secondary ischemia. cPLA2 activity may also adversely affect cell viability in the postischemic brain by direct actions on plasma and mitochondrial membranes (12, 51, 59, 60). Lysophospholipids and reactive oxygen species generated during reperfusion partly as a result of metabolism of arachidonic acid by cyclooxygenases and lipoxygenases (61) may contribute to irreversible neuronal injury. cPLA2 acts synergistically with reactive oxygen species generated during reperfusion (62) to cause cellular injury (60), likely due to enhanced susceptibility of peroxidized membranes to the action of PLA2 (12). Furthermore, oxygenated metabolites of arachidonic acid may themselves be toxic (63). Upon middle cerebral artery occlusion with an intraluminal filament, there was a reduction in cerebral blood flow to approximately 20% of baseline in both cPLA2+/+ and cPLA2-/- animals (28). Cerebral blood flow returned to the baseline within 30 min of removal of the intraluminal filament. No differences were found in mean arterial BP, and arterial Po2, Pco2, and pH between wild-type and knockout mice during ischemia and reperfusion. There was 34% less brain infarct volume as well as less brain edema and fewer functional neurologic deficits after middle cerebral artery occlusion for 2 h in cPLA2-/- mice. Neuroprotection was seen in brains examined up to at least 3 d after the transient ischemic episode. There were no systematic differences in vascular anatomy of the brain between wild-type and cPLA2-/- animals. Thus, our results clearly demonstrate that cPLA2 activity is a critical determinant of functional injury after an ischemic insult to the brain. Another example of the way the cPLA2-/- mouse can provide insight into tissue injury mechanisms is the response to the neurotoxin, 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) (64). We found that cPLA2-/- mice had significantly less depletion of brain striatal dopamine after administration of MPTP when compared with either littermate cPLA2+/+ or cPLA2+/- mice. The MPTP metabolite 1-methyl-4-phenylpyridinium (MPP+) is a selective inhibitor of complex I of the electron transport chain (65). Resistance to MPTP-induced neurotoxicity could be due to reduced release of fatty acids and lysophospholipids, and reduced reactive oxygen metabolite (66) production, leading to less excitotoxicity and mitochondrial damage. MPTP neurotoxicity may also be reduced in the cPLA2-/- animals due to decreased excitotoxic mechanisms and decreased production of PAF. Although the mechanism remains unclear, the present results further implicate activation of cPLA2 in neuronal injury. Because MPTP is used to model Parkinson's disease, the results also suggest that the development of selective cPLA2 inhibitors might be a useful therapeutic strategy for the treatment of Parkinson's disease. This is another example of the utility of the cPLA2-/- mouse as a model. Oozumi et al. have also produced a cPLA2-/- mouse (76). They found that ovalbumin-induced anaphylactic responses were significantly reduced, as was their bronchial reactivity to methacholine, demonstrating the importance of cPLA2 for lung inflammation. Clearly, it must be recognized that not all effects of cPLA2 are detrimental. There have been described beneficial effects of a number of products of arachidonic acid. PGE2 and prostacyclins are vasodilators and therefore may be protective in pathophysiologic states associated with high circulating levels of vasoconstrictors, such as angiotensin II and norepinephrine. Prostaglandins have been reported to protect the kidney against ischemic and toxic injury (67). As another example of beneficial actions, PGE2 elevates intracellular cAMP, which inhibits monocyte function and selectively blocks secretion of cytokines by Th1 cells (68). Thus, the cPLA2-/- animal may demonstrate detrimental as well as beneficial effects due to the absence of cPLA2. The beneficial effects we find when studying ischemic or MPTP injury in the brain must therefore be considered a "net" effect. It must also be recognized that the cPLA2-/- animal may upregulate compensatory mechanisms. Inhibiting the enzyme in the normal animal or cell may not be totally equivalent to working with cPLA2-/- animals or cells. Inhibitors, as described previously, however, are limited by specificity, access to cPLA2, and unknown actions. The cPLA2-/- animal provides a specific way to eliminate cPLA2 action and, when coupled with in vitro studies in which antisense and dominant negative approaches are done side by side with studies comparing cPLA2+/+ and cPLA2-/- cells, provides a very good way of defining the role of cPLA2. Future Research Opportunities There are many ways in which this model can be used in renal physiology and pathophysiology in whole organ, isolated tubule, isolated cell, and cell culture models. Questions may be asked such as: How is urinary prostaglandin excretion altered in the cPLA2 mouse? Is cPLA2 involved in the selective production of certain eicosanoids? For example, is thromboxane production affected more than prostaglandin production? What is the effect of absence of cPLA2 on urinary concentrating capacity? How is sodium reabsorption affected by the absence of cPLA2? Are the cPLA2-/- mice less susceptible to glomerulonephritis or diabetic nephropathy? How is sodium and water reabsorption in different nephron segments affected? How does the susceptibility to renal ischemic injury compare in cPLA2-/- and cPLA2+/+ littermate controls? How are the changes in blood flow and inflammatory responses of the kidney changed in response to obstruction of the ureter in cPLA2-/- animals? In preliminary experiments, we have found evidence for a urinary concentration defect in cPLA2-/- mice that is associated with abnormal proximal aquaporin 1 (69). When one addresses these questions, however, a number of challenges become immediately apparent. The animals, mice, are small and therefore measurements are more difficult. Even the simple collection of urine, especially in a water-deprived animal, requires some level of innovation. Many of the physiologic and pathophysiologic concepts on which the questions are based have been formulated from experiments on rats or dogs, and do not necessarily apply in the same way in the mouse. For example, a particular eicosanoid that may be critical for sodium transport in the rat may not be produced by the mouse kidney. The effects of aging must be taken into account. It is not sufficient to study young animals since interesting and very relevant processes may only be uncovered when the animals are challenged at an older age. Finally, there may be compensatory mechanisms in the cPLA2-/- animals that may affect the phenotype. The knockout model also affords the opportunity to collect cells, such as leukocytes, platelets, or macrophages, and study these in vitro as well as establish cPLA2-/- cells in culture. Stimulated peritoneal macrophages from the cPLA2 null mouse do not produce prostaglandin E2 (PGE2), leukotriene B4, or leukotriene C4 (28). Whereas macrophages from wild-type littermate mice respond to lipopolysaccharide with a marked PGE2 release, there is no detectable PGE2 production produced by macrophages from cPLA2-/- mice. Although the peritoneal macrophage is known to have other forms of PLA2 (21, 70), these data indicate that the primary form of PLA2 responsible for phorbol myristate acetate and Ca2+ ionophore-induced arachidonate release and lipopolysaccharide-induced PGE2 production is cPLA2. It would be very difficult to be as confident of such a conclusion using inhibitors. We have generated mesangial cells and embryonic fibroblasts from cPLA2-/- and littermate cPLA2+/+ mice. Once one has these cells, a large number of hypotheses may be tested. One can evaluate, for example, the role of cPLA2 in cell proliferation, cell volume regulation, hyperglycemia-induced matrix production. In preliminary studies, cPLA2-/- mesangial cells have different growth characteristics in the presence of serum when compared with cells derived from cPLA2+/+ littermates (71). The role of cPLA2 in excitotoxicity, injury and repair, and glutamate release in vitro can be evaluated using mesangial cells, fibroblasts, smooth muscle, leukocytes, neurons, or astrocytes, isolated from cPLA2-/- and cPLA2+/+ mice. The role of cPLA2 relative to other forms of PLA2s in signal transduction, in activation of various genes, and in coupling with cyclooxygenases may be evaluated in ways hitherto not feasible. The role of cPLA2 in cell death may be explored independent of the toxicities of inhibitors or antisense constructs. From the list of processes that cPLA2 has been implicated in (Tables 1 through 3), it is clear that the number of processes that one can study with this new experimental tool is quite large. The cell lines derived from cPLA2-/- and cPLA2+/+ animals also afford the opportunity to evaluate the role of eicosanoids on gene transcription. For example, the eicosanoid 15-deoxy-Δ12,14-prostaglandin J2 is a ligand for the peroxisome proliferator-activated receptor (PPARγ) (72), a member of the nuclear receptor superfamily. PPARγ contains a DNA-binding domain that binds to hormone response elements in promoters of target genes. Cells that do not contain cPLA2 may not make the eicosanoids necessary for interactions with PPAR and therefore may have significant alterations in lipid signaling. 8 (S)-hydroxyeicosatetraenoic acid (8-S-HETE) and leukotriene B4 activate PPARα (73). PPARα has been localized to the proximal tubule and medullary thick ascending limb, whereas PPARγ has been found in the medullary collecting duct and epithelium lining the papilla (74). It is important to confirm, however, that the differences seen between cPLA2-/- and cPLA2+/+ cell lines are in fact due to the absence of cPLA2 in the cPLA2-/- line. For example, it is possible that other molecules are upregulated or altered in expression and the differences seen are due to those rather than the absence of cPLA2. In cyclooxygenase 1 and 2 null cells, there is a compensatory PGE2 biosynthesis that may be related in part to an increase in cPLA2 activity (75). We have chosen to introduce cPLA2 back into the null cell lines using adenoviral approaches since epithelial cells are infected with adenovirus with high efficiency. Conclusion Our findings show that although there are many forms of PLA2, and in many cells more than one form is present, the functions of these forms with regard to generation of arachidonic acid for eicosanoid synthesis are not redundant. There are specific roles played by cPLA2. The use of "knockout" technology to generate a cPLA2-/- mouse provides a hitherto unavailable opportunity to dissect the role of cPLA2 both in the whole animal as well as at a cellular and molecular level. Important issues in organ and vascular physiology can be approached in vivo. Important questions relating to the role of cPLA2 in inflammatory and toxic responses can be studied in vivo. In vitro, one can ask how this enzyme is involved in signal transduction, gene regulation, and cell homeostasis. Breeding cPLA2-/- animals with animals that are null for a different gene or have been altered to overexpress a particular gene allows for the study of interactions among proteins at a whole animal as well as cellular level. This is an exciting new tool that allows the investigator to explore the role of proteins by taking a genetic approach rather than a pharmacologic one. The cPLA2-/- mice will yield a great deal of insight into the role of arachidonic acid and eicosanoids and the specific coupling of this enzyme to eicosanoid generation in many physiologic and pathophysiologic states. Acknowledgments Dr. Bonventre is supported by National Institutes of Health Grants DK 39773 (MERIT Award), DK 38452, and NS 10828. The author acknowledges the many individuals who have contributed to the work described, in particular Drs. Adam Sapirstein, Michael Moskowitz, and Zhihong Huang, Mr. M. Reza Taheri, and Ms. Eileen O'Leary.
Referência(s)