Role of calcium-independent phospholipase A 2 in complement-mediated glomerular epithelial cell injury

Transcription

1 Page 1 of 52 Articles in PresS. Am J Physiol Renal Physiol (January 2, 2008). doi: /ajprenal Role of calcium-independent phospholipase A 2 in complement-mediated glomerular epithelial cell injury Daniel Cohen 2, Joan Papillon 1, Lamine Aoudjit 1, Hongping Li 1, Andrey V. Cybulsky 1,2,3 and Tomoko Takano 1,2,3 Department of Medicine 1 and Physiology 2, McGill University, Montreal, Quebec, Canada. 3 authors contributed equally Running head: ipla 2 in complement-mediated glomerular epithelial injury Correspondence to; Tomoko Takano, MD, PhD Division of Nephrology, McGill University 3775 University Street, Rm 236 Montreal, Quebec H3A2B4 Phone: Fax: tomoko.takano@mcgill.ca 1 Copyright 2008 by the American Physiological Society.

2 Page 2 of 52 2 ABSTRACT In experimental membranous nephropathy, complement C5b-9-induced glomerular epithelial cell (GEC) injury leads to morphological changes in GEC and proteinuria, in association with phospholipase A 2 (PLA 2 ) activation. The present study addresses the role of calciumindependent PLA 2 (ipla 2 ) in GEC injury. ipla 2 -short and ipla 2 were expressed in cultured rat GEC and normal rat glomeruli. To determine if ipla 2 is involved in complement-mediated arachidonic acid (AA) release, GEC were stably transfected with ipla 2 or ipla 2 cdnas (GEC-iPLA 2 ; GEC-iPLA 2 ). Compared with control cells (GEC-Neo), the GEC-iPLA 2 and GEC-iPLA 2 demonstrated greater expression of ipla 2 proteins and activities. Complementmediated release of [ 3 H]AA was augmented significantly in GEC-iPLA 2, as compared with GEC-Neo, and the augmented [ 3 H]AA release was inhibited by the ipla 2 -directed inhibitor bromoenol lactone (BEL). For comparison, overexpression of ipla 2 also amplified [ 3 H]AA release after incubation of GEC with H 2 O 2, or chemical anoxia followed by re-exposure to glucose (in vitro ischemia-reperfusion injury). In parallel with release of [ 3 H]AA, complementmediated production of prostaglandin E 2 was amplified in GEC-iPLA 2. Complement-mediated cytotoxicity was attenuated significantly in GEC-iPLA 2, as compared with GEC-Neo, and the cytoprotective effect of ipla 2 was reversed by BEL, and in part by indomethacin. Overexpression of ipla 2 did not amplify complement-dependent [ 3 H]AA release, but nonetheless attenuated complement-mediated cytotoxicity. Thus, ipla 2 may be involved in complement-mediated release of AA. Expression of ipla 2 or ipla 2 induces cytoprotection against complement-dependent GEC injury. Modulation of ipla 2 activity may prove to be a novel approach to reducing GEC injury. Key words: ipla 2 -, ipla 2 -, cytoprotection 2

3 Page 3 of 52 3 INTRODUCTION Members of the phospholipase A 2 (PLA 2 ) family of enzymes hydrolyze fatty acids from the sn-2 position of phospholipids, generating free fatty acids and lysophospholipids (34, 43). Both products of this reaction have signaling properties. Moreover, arachidonic acid (AA) may be further metabolized by cyclooxygenase (COX) and lipoxygenase enzymes to bioactive eicosanoids. The PLA 2 s may be subdivided into three categories, including secretory PLA 2 s (spla 2, Groups I, II, III, V, VII-XIV), calcium-dependent PLA 2 s, e.g. cytosolic PLA 2 (cpla 2, Group IV), and the most recently discovered calcium-independent PLA 2 s (ipla 2 s Group VI) (34, 43). cpla 2 plays an important role in complement-mediated glomerular epithelial cell (GEC) injury in the passive Heymann nephritis (PHN) model of membranous nephropathy (19). GEC, otherwise known as podocytes, are an important component of the glomerular permselective barrier (33, 38, 42). In PHN and human membranous nephropathy, deposition of antibodies in the subepithelial region of the glomerulus leads to activation of complement and assembly of the C5b-9 membrane attack complex in GEC plasma membranes, leading to GEC injury and proteinuria. Studies in cultured GEC and in PHN indicate that C5b-9 induces sublethal GEC injury, which is associated with activation of protein kinases and phospholipases (including cpla 2 ), upregulation of COX-2, production of reactive oxygen species, induction of endoplasmic reticulum stress response, and other effects. These pathways ultimately contribute to changes in GEC lipid composition and function, reorganization of the actin cytoskeleton, and displacement of filtration slit diaphragm proteins (19). Previous studies have demonstrated that assembly of C5b-9 in cultured GEC is linked to the activation of cpla 2, and that glomeruli of rats with PHN demonstrate increased cpla 2 activity 3

4 Page 4 of 52 4 compared with control glomeruli (23). Activation of cpla 2 leads to various cellular effects. First, in GEC, activation of cpla 2 perturbs the membrane of the endoplasmic reticulum and contributes to induction of endoplasmic reticulum stress. Preconditioning of rats with compounds that induce endoplasmic reticulum stress can limit proteinuria in PHN (22). Second, activation of cpla 2 may contribute directly to the exacerbation of C5b-9-induced GEC injury (18, 23, 37). Finally, the AA mobilized by cpla 2 is metabolized by COX enzymes (44) to bioactive eicosanoids. Glomeruli from rats with PHN show increased COX-1 and COX-2 protein expression, and enhanced production of prostaglandin (PG) E 2 and thromboxane A 2. Blockade of COX enzymes reduces urinary protein excretion in PHN and in human membranous nephropathy (45, 46). The purpose of the present study was to determine if complement could stimulate the activity of ipla 2. Among the various members of the ipla 2 family, we focused on two isoenzyme subfamilies, so-called ipla 2 -VIA-1 and 2 (ipla 2 -short and ipla 2 -long), and ipla 2 -VIB (ipla 2 ), each with unique structural features, especially in the N terminus (29, 31, 43). Both ipla 2 -short and ipla 2 -long are functional enzymes, which differ by a 54 amino acid prolinerich insertion sequence (13). Knockout mice lacking ipla 2 show only impaired sperm motility (7), while transgenic mice overexpressing the same isoform are prone to ventricular tachyarrhythmias (30). ipla 2 is a membrane bound enzyme, with competing peroxisomal (C terminal) and mitochondrial (N terminal) localization signals (31). ipla2 also localizes in the endoplasmic reticulum (17). ipla 2 gene transcription and translation are complex (31). ipla 2 has multiple promoter sites, as well as a transcriptional repressor site in the 5' untranslated region. The full length 2.4 kb transcript contains 13 exons and includes four AUG translation initiation codons generating as many as 10 splice variants, which appear as 88, 77, 74, and 63 4

5 Page 5 of 52 5 kda immunoreactive proteins on SDS-PAGE and immunoblotting (47). ipla 2 enzymes are active in basal phospholipid turnover, phospholipid accumulation during the G1/M phase of mitosis, and are involved at various levels of apoptotic cascades (2, 5). ipla2 was shown to be cytoprotective against oxidant insults in renal cells, presumably by removing oxidized phospholipids (17). In contrast, inhibition of rabbit renal proximal tubule cell microsomal ipla 2 during cisplatin induced apoptosis reduced annexin V staining, chromatin condensation, and caspase-3 activation indicating that ipla 2 inhibition was cytoprotective (16). Mitochondrial ipla 2 has also been shown to actively participate in the permeability pore transition of the mitochondria and the release of the proapoptotic cytochrome c (25). A recent, study showed that ipla 2 knockout mice display multiple bioenergetic dysfunctional phenotypes (32). In addition, several potential pathophysiological roles have been described for the ipla 2 s, including AA release for eicosanoid production, tumorigenesis, cell injury, apoptosis, chemotaxis, and others (2, 5, 35). In the present study, we have used cultured GEC to demonstrate that C5b-9 can stimulate an ipla 2 to increase free [ 3 H]AA. Overexpression of ipla 2 in GEC augmented complementdependent [ 3 H]AA release, and production of PGE 2. Furthermore, overexpression of ipla 2, as well as ipla 2 limited complement-induced GEC injury via eicosanoids, suggesting that activation of ipla 2 is cytoprotective. 5

6 Page 6 of 52 6 MATERIALS AND METHODS Materials Tissue culture media, zeomycin, hygromycin, and Lipofectamine 2000 were from Invitrogen- Life Technologies (Burlington, ON). Electrophoresis reagents were from Bio-Rad Laboratories (Mississauga, ON). Mouse anti-ipla 2, anti-cox-1 and anti-cox-2 antibodies were from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). Enhanced chemiluminescence (ECL) detection reagents were from Amersham Bioscience (Baie d Urfé, QC). Bromoenol lactone (BEL) and PGE 2 enzyme immunoassay kits were purchased from Cayman Chemical (Ann Arbor, MI). [ 3 H]AA (100mCi/mmol), [ 14 C]arachidonyl-phosphatidylethanolamine (PE; 48 mci/mmol), and [ 14 C]arachidonyl-phosphatidylcholine (PC; 53 mci/mmol), were purchased from Perkin-Elmer Life & Products (Boston, MA). Mouse ipla 2 long and ipla 2 short cdnas subcloned into pcdna 3.1 were kindly provided by Dr. Suzanne Jackowski (University of Tennessee Health Centre, Memphis, TN), while human ipla 2 cdna in pcdna 1.1 was provided by Drs. Richard Gross and David Mancuso (Washington University School of Medicine, St. Louis, MO). ipla 2 antibody was kindly provided by Dr. Makoto Murakami (Showa University, Tokyo, Japan). Unless otherwise specified, all other chemicals and biochemical reagents were purchased from Sigma-Aldrich Canada (Mississauga, ON). Cell culture, radiolabeling, and transfection Rat GEC culture and characterization have been described previously (14, 40). According to established criteria, the cells demonstrated cytotoxic susceptibility to low doses of aminonucleoside of puromycin, presence of junctional complexes by electron microscopy, and expression of a variety of GEC antigens. GEC were maintained in K1 medium. COS cells and 6

7 Page 7 of 52 7 Chinese hamster ovary (CHO) cells were cultured in DMEM-10% fetal calf serum. Phospholipids were radiolabeled to isotopic equilibrium by incubating cells with [ 3 H]AA (0.125 µci/ml K1 medium) for 72 h (20). At the end of incubations, supernatants and cells were collected, lipids were extracted with chloroform:methanol (1:1.2) and 0.2% formic acid, followed by chloroform, and were then dried under nitrogen. After reconstitution in chloroform:methanol (2:1) and addition of cold AA and 1,2-diacylglycerol (DAG) standards, the free AA and DAG were separated from other lipids by thin layer chromatography in hexane:diethyl ether:acetic acid (80:20:2). Relevant bands were scraped and radioactivity was quantified in a liquid scintillation counter (18, 37). It should be noted that when using this method, we cannot exclude some oxidation of [ 3 H]AA. For expression of ipla 2 enzymes, plasmids containing ipla 2 short or ipla 2 cdnas were transfected in GEC using the Lipofectamine 2000 reagent. Following selection by Neomycin and expansion, all subclones were tested for overexpression of the enzyme by immunoblotting. Confirmation of enhanced enzyme activity was determined using an in vitro PLA 2 assay (described below). Characterization of GEC that overexpress cpla 2 was published previously (18, 37). Incubation of GEC with complement GEC were incubated with rabbit anti-gec antiserum (5% v/v) in modified Krebs-Henseleit buffer containing 145 mmol/l NaCl, 5 mmol/l KCl, 0.5 mmol/l MgSO 4, 1 mmol/l Na 2 HPO 4, 0.5 mmol/l CaCl 2, 5 mmol/l glucose, 20 mmol/l Hepes, ph 7.4 for 30 min at 22 C. The cells were then incubated at 37 C with normal human serum (NS, with full complement activity) or heat-inactivated serum (HIS, incubated at 56 C for 30 min in order to inactivate complement) as 7

8 Page 8 of 52 8 indicated. In experiments designed to measure Ca 2+ -independent [ 3 H]AA release, antibodysensitized GEC were incubated with C8-deficient human serum (C8DS) in buffer containing 0.5 mm Ca 2+ for 40 min. Then, GEC were washed, and incubated with or without purified C8 (2.0 µg/ml) and C9 (1.5 µg/ml) for 20 min in buffer containing 1 mm EGTA (18, 37). Preparation of cell extracts and in vitro assay of PLA 2 activity Cells were collected in homogenization buffer containing 50 mm Hepes, 0.25 M sucrose, 1mM EDTA, 1mM EGTA, 20 µm leupeptin, 20 µm pepstatin, 0.1 mm PMSF, ph 7.4 (18, 37). Following centrifugation for 3 min at 1,500 g, the pellet was homogenized with 25 strokes in a Wheaton homogenizer, centrifuged for 5 min at 2,500 g, and the supernatant was collected. The remaining pellet was again homogenized and centrifuged for 5 min with the supernatant being pooled with that collected earlier. The supernatant was centrifuged at 15,000 g for 20 minutes to remove any remaining insoluble membrane fractions. To assay PLA 2 activity (18, 37), exogenous 2-[ 14 C]arachidonyl-PE or PC (substrate; 3.5 µm) was dried under a stream of nitrogen, and was reconstituted in 2 µl of DMSO. Cell extracts diluted in homogenization buffer (48 µl) were incubated with substrate for 45 min at 37 C. In some experiments, PLA 2 activity was measured after adding CaCl 2 to the homogenization buffer to a final Ca 2+ concentration of 2 mm. The reaction was stopped by adding 50 µl of ethanol containing cold AA and acetic acid. [ 14 C]AA was separated from other lipids by thin layer chromatography as described above. Immunoblotting Immunoblotting was described previously (45). Glomeruli were isolated from rat kidney 8

9 Page 9 of 52 9 cortices by differential sieving (purity of glomeruli was >95%) (22). Briefly, cells or isolated glomeruli were lysed in buffer containing 10 mm sodium pyrophosphate, 25 mm NaF, 2 mm Na 3 VO 4, 2% Triton X-100, 250 mm NaCl, 20 mm Tris, 2 mm EDTA, 2 mm EGTA, and protease inhibitor cocktail (Roche Diagnostics), ph 7.4. After centrifugation at 14,000g, soluble components were collected and the protein concentration was quantified using the Bio-Rad assay. Equal quantities of proteins were separated by SDS-PAGE under reducing conditions. Proteins were then transferred to a nitrocellulose membrane and the membrane was blocked at room temperature for 60 min with 5% dry milk in TBS-T wash buffer containing 10 mm Tris, ph 7.5, 50 mm NaCl, 2.5 mm EDTA, 0.05% Tween 20. The nitrocellulose membrane was subsequently incubated with primary antibody for 60 min at 22 C, washed three times in TBS-T wash buffer, and incubated with the appropriate secondary antibody conjugated to horseradish peroxidase for 60 min at 22 C. Immunoreactive proteins were then visualized using enhanced chemiluminescence (ECL). Reverse transcriptase polymerase chain reaction (RT-PCR) RNA samples were isolated from cells or normal rat glomeruli using TRIzol reagent (Invitrogen) according to the manufacturer s instructions (3). RT-PCR was performed with the Invitrogen Superscript II One-Step PCR Kit. Primers designed for ipla 2 flanked the insertion site of exon 9 (forward: TTGGGGCAGAAGTAGACACC, reverse: GTCAGCATCACCTTGGGTTT) to distinguish short and long isoforms. Primers for ipla 2 were forward: ATTGATGGTGGAGGAACAAGA and reverse: ATGGCCTGCCACATTTTATAC. PCR for ipla 2 and ipla 2 were performed with 35 cycles of 30 s at 94 o C, 1 min at 55 o C, and 1 min at 72 o C. Experimental conditions for EP4 and -actin 9

10 10 Page 10 of 52 PCR were described previously (3). PCR products were separated by agarose gel electrophoresis and visualized after staining with ethidium bromide. PGE 2 Assay At the end of incubations with complement, supernatants were collected for measurement of PGE 2. The samples were further diluted fold in assay buffer, and PGE 2 was quantified using an enzyme immunoassay kit from Cayman Chemical, according to the manufacturer s instructions (45). Measurement of intracellular Ca 2+ concentration The method has been described in detail (20). Briefly, GEC were loaded in suspension with 5 µm fura-2-acetoxymethyl ester in buffer containing 0.5 mm CaCl 2 for 30 min at 37 o C. After washing, aliquots of cells were placed into a spectrofluorometer in Ca 2+ -free Krebs-Henseleit buffer, containing 2 mm EGTA. Fluorescence of fura-2 was monitored continuously (emission wavelength of 510 nm, excitation wavelengths alternating between 340 and 380 nm) at 37 o C. Release of Ca 2+ from intracellular storage sites was initiated by the addition of 5 µm ionomycin. Lactate dehydrogenase (LDH) assay GEC were incubated with antibody and complement as described above. At the end of incubations, supernatants were collected and the remaining cell contents were solubilized with Triton-X 100 (1%). The supernatant and Triton fractions were assayed for LDH by incubating 100 µl of the sample in 1.5 ml glycine (58 mm)/lactate buffer (0.2 M), containing NAD (4.8 mg/ml) at 37 C, and monitoring absorbance at 340 nm in a spectrophotometer. Specific LDH 10

11 Page 11 of release was calculated using the formula (NS-HIS)/(100-HIS) x 100 (40). Statistics Data are presented as mean ± SEM. The t statistic was used to determine significant differences between two groups. For more than two groups, one-way analysis of variance (ANOVA) was used to determine significant differences among groups. Where significant differences were found, individual comparisons were made between groups using the t statistic, and adjusting the critical value according to the Bonferroni method (Fig. 5-8). Two-way analysis of variance was used to determine significant differences in multiple measurements among groups (Fig. 3B and 4B). 11

12 12 Page 12 of 52 RESULTS C5b-9 induces Ca 2+ -independent release of [ 3 H]AA in GEC Previous studies demonstrated that in GEC, assembly of C5b-9 leads to an increase in cytosolic Ca 2+ concentration, which is primarily due to Ca 2+ influx (20). Furthermore, Ca 2+ influx activates the Ca 2+ -dependent cpla 2, leading to release of [ 3 H]AA (37). In the first set of experiments, we assessed if C5b-9 could also induce [ 3 H]AA release in the absence of Ca 2+ influx. Since antibody-dependent activation of complement (i.e. the classical pathway) is Ca 2+ - dependent, we incubated antibody-sensitized GEC with C8DS in the presence of Ca 2+. As a second step, GEC were incubated with or without purified C8 and C9 without Ca 2+ (i.e. in the presence of EGTA). The C8DS leads to assembly of the nascent C5b-7 complex, while addition of C8 and C9 allows assembly of the C5b-9 transmembrane channel (in the absence of Ca 2+ ). Formation of C5b-9 resulted in a significant release of [ 3 H]AA, suggesting activation of an ipla 2 (Fig. 1A). C5b-9 also increased [ 3 H]DAG (Fig. 1A), which reflects activation of phospholipase C. It should be noted that the incubation of GEC with C8 and C9 in buffer containing EGTA not only results in chelation of extracellular Ca 2+, but would most likely also result in depletion of intracellular Ca 2+ stores. To verify this effect of EGTA, we measured changes in intracellular Ca 2+ concentration by monitoring fluorescence of the Ca 2+ indicator, fura-2. Addition of ionomycin to GEC in buffer containing 2 mm EGTA resulted in a Ca 2+ spike, reflecting release of Ca 2+ from intracellular storage sites (Fig. 1B, top panel). In contrast, if GEC were preincubated with 2 mm EGTA for 20 min prior to the addition of ionomycin, the Ca 2+ spike was almost completely abolished (Fig. 1B, bottom panel), indicating that intracellular Ca 2+ stores were depleted. In preliminary studies, we determined that a 40 min incubation after the addition 12

13 Page 13 of of C8 and C9 to the GEC with assembled C5b-7 complexes was required to increase free [ 3 H]AA significantly. Thus, the effect of C5b-9 on [ 3 H]AA occurred while Ca 2+ stores were largely depleted. Expression of ipla 2 isoforms Several isoforms of the Group VI ipla 2 enzyme have been described, including ipla 2 long, ipla 2 short and ipla 2. To confirm endogenous expression of ipla 2 in cultured rat GEC, as well as in rat glomeruli, RT-PCR was performed using primers to regions flanking exon 9 of the ipla 2 isoform, and the coding sequence of ipla 2 (Fig. 2). PCR products consistent with ipla 2 short (Fig. 2A), as well as ipla 2 (Fig. 2B) were identified in GEC and glomeruli, and identities were confirmed by purification of bands and DNA sequencing. It should be noted that ipla 2 RT-PCR in GEC and glomeruli showed additional faint bands, with a band that appeared to correspond to the predicted PCR product for ipla 2 long. However, we were not able to confirm the identity of this band by purification and DNA sequencing, so it is not possible to make a firm conclusion regarding expression of ipla 2 long. ipla 2 protein expression was studied by immunoblotting. GEC and glomeruli express ipla 2, and on SDS-PAGE, migration of this protein was slightly retarded, as compared with COS cells transfected with ipla 2 short (Fig. 2C). Nevertheless, the protein migrated faster than the ipla 2 long isoform. ipla 2 has four potential translational initiation sites yielding proteins of 88, 77, 74, and 63 kda respectively (47). GEC and glomeruli predominantly express the 88 kda isoform of ipla 2 (Fig. 2D). Overexpression of ipla 2 enzymes 13

14 14 Page 14 of 52 To strengthen the link between complement and ipla 2 activation, we transfected GEC with ipla 2 and ipla 2 short cdnas. Clones were screened for stable expression by immunoblotting. Several clones of GEC transfected with ipla 2 (GEC-iPLA 2 ) expressed the 88, 77, and 74 kda isoforms of the enzyme (Fig. 3A). Clones that express ipla 2 short (GECiPLA 2 ) were also produced, and a representative clone is presented (Fig. 4A). It should be noted that the transfected enzyme migrates slightly faster than the endogenous isoform. To ensure that the expressed ipla 2 proteins were enzymatically active, we further tested the extracts of stable subclones for ipla 2 activity, as monitored by release of [ 14 C]AA from exogenous [ 14 C]arachidonyl-PE. GEC transfected with ipla 2 (Fig. 3B) or ipla 2 short (Fig. 4B) showed augmented [ 14 C]AA release in a concentration dependent manner, as compared with GEC-Neo. Interestingly, release of [ 14 C]AA from exogenous [ 14 C]PC was relatively weak (data not shown), suggesting that at least in our assay system, [ 14 C]PE is a preferred substrate for ipla 2. This observation is in keeping with an earlier study (35). Stimulus-coupled [ 3 H]AA release in GEC stably transfected with ipla 2 The effect of complement on [ 3 H]AA release was studied in GEC-iPLA 2, and was compared with GEC-Neo. GEC-iPLA 2 tended to have slightly increased basal levels of free [ 3 H]AA, in keeping with previous studies (35). Complement stimulated [ 3 H]AA release in GEC-Neo, whereas overexpression of ipla 2 enhanced complement-mediated [ 3 H]AA release significantly (Fig. 5A). As expected, complement-mediated [ 3 H]AA release in GEC-iPLA 2 was inhibited significantly by the ipla 2 -directed inhibitor, BEL (1)(Fig. 5A). This result confirms that in these cells, a significant portion of complement-dependent [ 3 H]AA release is due to ipla 2 activity. BEL did not have a significant effect on [ 3 H]AA release in control cells, suggesting that 14

15 Page 15 of this [ 3 H]AA release may have been mediated by another PLA 2, e.g. cpla 2 (18, 37). In previous studies, we demonstrated that in GEC, C5b-9 can induce [ 3 H]AA release by activating cpla 2 (37). BEL was shown to have an inhibitory concentration (IC 50 ) 250-fold greater for ipla 2 over cpla 2 (1), nevertheless, we wanted to rule out the possibility that BEL had inhibited cpla 2. To determine if BEL could affect cpla 2 activity, we monitored PLA 2 activity in COS cells that had been transiently transfected with cpla 2. In the presence of 1 mm EGTA (Ca 2+ -free buffer), PLA 2 activity in COS cell extracts was low (Table 1). In the presence of 2 mm free Ca 2+, PLA 2 activity increased more than 6-fold, reflecting Ca 2+ -dependent activation of cpla 2. PLA 2 activity was not affected by BEL in Ca 2+ -replete buffer, indicating that BEL is not an effective inhibitor of cpla 2 activity (Table 1). For comparison, we also studied how other stimuli that may injure GEC may affect free [ 3 H]AA. Overexpression of ipla 2 enhanced release of [ 3 H]AA by H 2 O 2 (Fig. 5B), and by chemical anoxia/re-exposure to glucose (recovery), an in vitro model of ischemia-reperfusion injury (24) (Fig. 5C). These effects of H 2 O 2 (Fig. 5B) and anoxia/recovery (Fig. 5C) in GECiPLA 2 were reduced by BEL, although the effect did not reach statistical significance in the case of H 2 O 2. Thus, ipla 2 is involved in stimulus-coupled [ 3 H]AA release, as the liberation of [ 3 H]AA in response to complement, H 2 O 2 and chemical anoxia/recovery is, at least in part, mediated by ipla 2. In previous studies, we demonstrated that overexpression of cpla 2 in GEC (~5-10-fold above endogenous levels) amplified the C5b-9-dependent release of [ 3 H]AA (18, 37). In contrast, the increases in free [ 3 H]AA induced by H 2 O 2 or anoxia/recovery (Fig. 5B and C) were not amplified in GEC that overexpress cpla 2 (Table 2), suggesting lack of involvement of cpla 2. Furthermore, these increases in [ 3 H]AA were attenuated in the presence of BEL (Table 15

16 16 Page 16 of 52 2). Together, the results suggest that ipla 2 (and possibly other ipla 2 s), but not cpla 2, is activated by H 2 O 2 and anoxia/recovery, whereas both ipla 2 and cpla 2 are activated by complement. In some cells, AA may be derived from metabolism of DAG. Thus, in the above experiments, we also monitored changes in [ 3 H]DAG. Complement (Fig. 1), H 2 O 2 and anoxia/recovery (data not shown) increased [ 3 H]DAG production, but the increases were not affected by ipla 2 overexpression (data not shown). Furthermore, BEL did not reduce [ 3 H]DAG production (data not shown), even though BEL is reported to inhibit phosphatidic acid phosphohydrolase, which mediates generation of DAG from phosphatidic acid. Based on these results, it is unlikely that inhibition of DAG production was the mechanism by which BEL attenuated AA release in GEC. ipla 2 -dependent AA release is coupled to production of PGE 2 The above experiments demonstrated that complement can induce release of [ 3 H]AA via activation of ipla 2. In the next set of experiments, we investigated if free AA could be metabolized via COX to eicosanoids. We monitored production of PGE 2, which was shown to be a major metabolite of COX in GEC (45). GEC-Neo produced a small amount of PGE 2 in response to complement, which appeared to be inhibited by indomethacin but not by BEL (HIS: 16±9, HIS/BEL: 7±4, NS: 46±15, NS/BEL: 50±14, NS/Indo: 18±2 pg/ml, Fig. 6). However, when analyzed together with GEC-iPLA 2, these changes were not statistically significant. Overexpression of ipla 2 did not affect basal PGE 2 production significantly during short term incubation (Fig. 6, 60 min incubation), while there was a small increase in PGE 2 in culture media of GEC-iPLA 2, as compared with GEC-Neo, after 2 days of cell culture (GEC-Neo: 284 ± 47 16

17 Page 17 of pg/ml, GEC-iPLA 2 : 519 ± 8 pg/ml, p<0.05, 3 experiments). However, after incubation with complement, PGE 2 production was ~20-fold greater in GEC-iPLA 2 relative to GEC-Neo (Fig. 6). The effect of complement was inhibited significantly by BEL, and by the non-selective COX inhibitor, indomethacin (Fig. 6). Thus, complement-induced production of PGE 2 parallels the release of [ 3 H]AA (Fig. 5A). Overexpression of ipla 2 limits complement-mediated GEC injury We reported previously that overexpression of cpla 2 in GEC increased the susceptibility to complement-mediated cytotoxicity (18). We next tested if overexpression of ipla 2 has a similar effect. Complement mediated cytotoxicity was monitored as release of LDH. Antibodysensitized GEC-Neo or GEC-iPLA 2 were incubated with normal serum at serially-increasing concentrations (to assemble C5b-9) for 100 min. The protocol allows for complement to activate ipla 2, but with increasing incubation time and complement dose, a portion of the cells will undergo lysis. LDH release was attenuated significantly in GEC-iPLA 2, as compared with GEC-Neo (Fig. 7A). Further studies in GEC-iPLA 2 showed that the cytoprotective effect of ipla 2 was reversed by the addition of BEL, and to a lesser extent by the addition of indomethacin (Fig. 8A). Indomethacin did not reverse the cytoprotective effect of ipla 2 completely, despite complete inhibition of PGE 2 production (Fig. 6). Therefore, part of the cytoprotective effect of ipla 2 may have been independent of eicosanoid production. In keeping with previous results, GEC-Neo tended to be protected by the addition indomethacin, since the activation of other PLA 2 s by complement was shown to involve mobilization of AA, and downstream metabolism to bioactive eicosanoids with a cytotoxic effect (46). To further elucidate which COX enzyme was responsible for the cytoprotective effect of 17

18 18 Page 18 of 52 ipla 2, we included isoform-specific COX inhibitors, NS 398 (COX-2) and SC560 (COX-1) in incubations with complement (Fig. 8B). Both inhibitors partially reversed the cytoprotective effect of ipla 2, suggesting that both COX isoforms were involved. Previously, we showed that C5b-9 upregulates the expression of COX-2 in GEC, although expression of COX-1 was not affected (45). Interestingly, COX-2 was upregulated in resting GEC-iPLA 2, i.e. independently of a complement effect (Fig. 9A). In contrast, expression of COX-1 was not significantly different between GEC-iPLA 2 and GEC-Neo (Fig. 9B). These results provide further support for the view that following activation of ipla 2, both COX enzymes are involved in metabolism of AA. In GEC, the principal receptor for PGE 2 is EP 4 (3), and earlier studies have demonstrated that signaling via EP 4 is cytoprotective (3). Overexpression of ipla 2 in GEC did not, however, affect expression of EP 4 receptor mrna (Fig. 10). Thus, the cytoprotective effect of ipla 2 and eicosanoids most likely occurs independently of changes in EP 4 receptor expression. Role of ipla 2 in complement-dependent GEC injury By analogy to the studies that addressed ipla 2 (described above), we employed GECiPLA 2 to study complement-mediated effects on [ 3 H]AA release and cell injury. In contrast to GEC-iPLA 2, there was no amplification of complement-mediated [ 3 H]AA release in GECiPLA 2. Levels of free [ 3 H]AA in GEC-Neo and GEC-iPLA 2 after exposure to HIS were 1.5 ± 0.2 and 2.5 ± 0.3 % of total radioactivity, respectively, while after exposure to NS, [ 3 H]AA increased to 4.1 ± 0.7 (p<0.0001) and 3.3 ± 0.2 % of total radioactivity, respectively (the increase in GEC-iPLA 2 was not statistically significant; 3 experiments). After incubation of antibodysensitized GEC with serially-increasing concentrations of complement, LDH release was 18

19 Page 19 of attenuated significantly in GEC-iPLA 2, as compared with GEC-Neo (Fig. 7B). However, in contrast to GEC-iPLA2, neither BEL nor indomethacin reversed the protective effect of ipla 2 (Fig. 8C). Thus, while overexpression of ipla 2 in GEC was not involved in complementinduced [ 3 H]AA release, it was nonetheless cytoprotective. 19

20 20 Page 20 of 52 DISCUSSION The present study demonstrates that in GEC, assembly of C5b-9 resulted in a significant release of [ 3 H]AA in Ca 2+ -free medium, suggesting activation of an ipla 2 (Fig. 1). GEC endogenously express ipla 2 short, although on SDS-PAGE, the protein appeared to be slightly greater in molecular mass, as compared with transfected ipla 2 short in COS cells (Fig. 2). Perhaps the endogenous ipla 2 short undergoes a post-translational modification in GEC, which accounts for the slight increase in size. GEC also express the 88 kda isoform of ipla 2 (Fig. 2). Both ipla 2 short and ipla 2 were also expressed in isolated normal glomeruli (Fig. 2). Overexpression of ipla 2 in GEC augmented complement-dependent [ 3 H]AA release, as compared with GEC-Neo, and furthermore, the [ 3 H]AA release in GEC-iPLA 2 was attenuated by BEL (Fig. 5). These results indicate that activation of complement is coupled with the activation of ipla 2. It should also be noted that the AA mobilized by ipla 2 was accessible to COX, and could be metabolized to prostanoids (Fig. 6). An earlier study also showed coupling of ipla 2 with COX-1 in HEK293 cells (35). In contrast to ipla 2, overexpression of ipla 2 in GEC did not amplify complement-dependent [ 3 H]AA release. By analogy to ipla 2, overexpression of cpla 2 in GEC was previously shown to augment complement-dependent [ 3 H]AA release and prostanoid production, as compared with GEC-Neo. However, overexpression of group IIA spla 2 in GEC was similar to ipla 2, i.e. neither enzyme amplified complement-dependent [ 3 H]AA release (37). Previous studies also demonstrated that complement induces serine 505 phosphorylation of cpla 2 in GEC, in association with an increase in enzymatic activity (18), and that glomerular cpla 2 is phosphorylated in PHN (21). This posttranslational modification of cpla 2, which is generally associated with activation, has been described with multiple agonists (27). ipla 2 activity has been reported to increase after 20

21 Page 21 of stimulation with PMA or ATP, but the mechanism is unknown at present, and will require further study (16, 28). An alternate mechanism of ipla 2 activation may involve modification of phospholipids (e.g. by C5b-9-induced deformability of the membrane lipid bilayer, or by lipid peroxidation) (9, 36), such that there is increased substrate availability to ipla 2. The overexpression models of the various PLA 2 s are invaluable in defining links of C5b-9 assembly with PLA 2 isoform activation. We have expressed these PLA 2 s at levels modestly above endogenous, but the results need to be interpreted with some caution. Furthermore, overexpression does not allow for a direct comparison of the amounts of [ 3 H]AA release by cpla 2 vs ipla 2. Nevertheless, release of [ 3 H]AA by complement in the presence of Ca 2+ was substantially greater than that in the absence of Ca 2+ (37), and overexpression of cpla 2 generally resulted in a greater amplification of complement-dependent [ 3 H] release, as compared with ipla 2. Thus, it is likely that cpla 2 is a quantitatively more important enzyme than ipla 2. To further address the potential functional importance of ipla 2 in GEC, we examined the effects of H 2 O 2 and anoxia/recovery on [ 3 H]AA release. Both of these stimuli model effects of ischemia-reperfusion injury in cultured cells (24). By analogy to complement, overexpression of ipla 2 enhanced release of [ 3 H]AA by H 2 O 2 and by anoxia/recovery (Fig. 5), providing further support for the view that ipla 2 activity can be enhanced by specific stimuli, including ischemia-reperfusion. In contrast to complement, the increases in free [ 3 H]AA induced by H 2 O 2 or anoxia/recovery (Fig. 5B and C) were not amplified in GEC that overexpress cpla 2 (Table 2). Other studies have demonstrated that release of AA in response to H 2 O 2 was mediated by cpla 2 in renal tubular epithelial and mesangial cells (26, 41). Thus, H 2 O 2 -coupled AA mobilization may be cell-specific (10). 21

22 22 Page 22 of 52 Various PLA 2 enzymes have been shown to regulate pathways of cell injury in several experimental disease models (5, 11, 12, 15, 39). In vivo, assembly of C5b-9 in GEC is associated with sublethal GEC injury and proteinuria; however, it is not practical to assay proteinuria in cell culture models. Instead, in cultured GEC, we have assayed C5b-9-mediated injury as cytolysis (LDH release), and it has been shown that cytolysis in culture correlates with proteinuria in vivo, in the context of injury associated with generation of prostanoids (46), p38 activation (3) and ER stress (22). The present study demonstrates that overexpression of ipla 2 attenuated complement-induced GEC injury, and the effect was reversed by BEL (Fig. 8). In addition, the attenuation of complement-induced GEC injury was partially reversed in the presence of a nonselective COX inhibitor, as well as with selective inhibitors of the COX-1 and -2 isozymes (Fig. 8). Thus, cytoprotection was, at least in part, mediated via generation of prostanoids, most likely through both COX-1 and COX-2. In support of this view are our earlier results, which show that GEC express COX-1 constitutively, and that C5b-9 induces COX-2 expression (45). Interestingly, while GEC-iPLA 2 also express COX-1 constitutively, expression of COX-2 was markedly upregulated in these cells under basal conditions (Fig. 9). The mechanism by which ipla 2 may be enhancing basal COX-2 expression will require elucidation, but it is noteworthy that similar COX-2 upregulation was observed in HEK293 cells overexpressing ipla 2 and stimulated with interleukin-1/fetal calf serum (35). However, despite the increase in basal COX-2 expression in GEC-iPLA 2, basal PGE 2 production was only modestly increased in these cells only after long term culture (Fig. 6), indicating that AA substrate was rate limiting. COX inhibitors did not enhance complement-induced injury in GEC-Neo, and actually tended to reduce cytotoxicity (Fig. 8). This result is in keeping with earlier studies in GEC-Neo, which showed that COX inhibitors protected the cells from complement-induced cell death. Finally, it 22

23 Page 23 of should be noted that inhibition of COX in PHN in vivo with nonselective or with COX-2- selective inhibitors reduced proteinuria (45, 46). There would seem to be an apparent contradiction in that production of prostanoids was cytoprotective in GEC-iPLA 2, but exacerbated cytotoxicity in other GEC lines (46). Possibly, individual prostanoids downstream of COX (and/or intracellular localization of their production) are responsible for inducing a cytoprotective vs cytotoxic effect. For example, in earlier studies in GEC-Neo, where COX inhibition protected the cells from complement-induced injury, the protective effect of the COX inhibitor could be reversed by the addition of a thromboxane A 2 analog, but not PGE 2 (46). Among the PGE 2 receptors, GEC predominantly express EP 4, as well as some EP 1 receptor (3). Previously, it was demonstrated that exogenous PGE 2 was protective against apoptosis induced by serum deprivation, and that these effects were reversed by an EP 4 receptor antagonist. Moreover, exogenous PGE 2 could also attenuate puromycin aminonucleoside-induced GEC injury, indicating that PGE 2 is cytoprotective (3). In the present study, we have shown that GEC-iPLA 2 express EP 4 receptor mrna, at levels similar to GEC- Neo (Fig. 10). Thus, a PGE 2 -EP 4 receptor pathway could mediate the cytoprotective effect ipla 2 in complement-dependent injury. Further studies will be required to confirm this hypothesis more directly. In GEC-iPLA 2, BEL reversed the cytoprotective effect of ipla 2 to a greater extent than indomethacin (Fig. 8), suggesting that mechanisms distinct from AA metabolism, but nonetheless related to ipla 2 catalytic activity also contributed to the cytoprotective effects of ipla 2. One possible mechanism is that the ipla 2 enzyme facilitates activation of the ER stress response, a cytoprotective response shown to be activated in GEC exposed to complement C5b-9 (22). At least some ipla 2 isoforms may be localized at the membrane of the ER (17); 23

24 24 Page 24 of 52 consequently, ipla 2 could perturb the ER membrane sufficiently to initiate the ER stress response, which may limit complement-induced injury. Additional studies will be required to determine the subcellular localization of ipla 2 in GEC, in particular, whether ipla 2 is found at the ER. Of interest, ipla 2 protected mitochondria in renal proximal tubular cells from oxidant-induced lipid peroxidation and dysfunction (28), and ipla 2 knockout mice show defects in mitochondrial functions and mitochondrial lipid composition (32). Thus, mitochondria are another potential target of ipla 2 in GEC. The present study demonstrates that overexpression of ipla 2 in GEC also attenuated complement-induced GEC injury (Fig. 7), and this effect occurred in the absence of stimulated [ 3 H]AA release, and was not blocked by acute incubation with BEL and indomethacin (Fig. 8). Another study failed to demonstrate stimulus-coupled AA release with ipla 2, but there are also reports of protein kinase C and calmodulin involvement in ipla 2 activation (2, 5). In the absence of complement-induced [ 3 H]AA release or acute inhibition, our results suggest that overexpression of ipla 2 may have induced preconditioning of GEC, such that the cells were able to resist complement-mediated injury. There were no significant differences in basal free [ 3 H]AA levels between GEC-Neo and GEC-iPLA 2, but it is possible that there was enhanced turnover of [ 3 H]AA in the latter. This view is consistent with earlier studies, which have determined that ipla 2 may be involved in both deacylation of membrane phospholipids and production of lysophospholipid acceptors (4, 6, 8). The mechanisms of phospholipid metabolism and cytoprotection in GEC-iPLA 2 will require further study. While the role of complement in the PHN model of membranous nephropathy is appreciated, the ensuing biological changes in GEC are complex, and include both pro-injurious and protective stimuli (19). The role of PLA 2 s in complement-mediated GEC injury has been 24

25 Page 25 of addressed in a number of studies (19). The present study suggests that in addition to cpla 2, ipla 2 may be involved in the pathogenesis of PHN, as complement-mediated GEC injury was reduced in the context of ipla 2 overexpression. Based on previous studies, there are cytoprotective and injurious aspects related to the complement-mediated activation of cpla 2 and production of prostanoids, both in cell culture and in vivo models. Further studies will be required to delineate these effects more precisely, and define the interplay between the various PLA 2 isoforms. Current treatment approaches to idiopathic membranous nephropathy are nonspecific and not very effective. Modulation of specific PLA 2 family enzymes may represent a mechanism of limiting GEC injury and maintaining the integrity of the permselective barrier. GRANTS This study was supported by grants from the Canadian Institute of Health Research (AVC, TT) and Kidney Foundation of Canada (AVC, TT), Scholarships from the Fonds de la Recherche en Santé du Québec (AVC, TT), and the Catherine McLaughlin Hakim Chair (AVC). 25

26 26 Page 26 of 52 FIGURE LEGENDS Figure 1. A. C5b-9 induces Ca 2+ -independent release of [ 3 H]AA in GEC. GEC, labeled with [ 3 H]AA, were incubated with antibody and 10% C8-deficient serum (C8DS) in buffer containing 0.5 mm Ca 2+ for 40 min. Then, GEC were washed, and incubated with or without purified C8 (2.0 µg/ml) and C9 (1.5 µg/ml; to assemble C5b-9) in buffer containing 1 mm EGTA. [ 3 H]AA and [ 3 H]DAG were measured after 40 min. *p<0.025 C8DS+C8+C9 vs C8DS alone, 5 experiments. B. GEC, loaded with fura-2, were placed into buffer containing 2 mm EGTA. Ionomycin (5 µm) was added after 3 min (top panel) or 23 min (bottom panel), and fura-2 fluorescence was monitored for at least another 7 min. Results are presented as the ratio (R) of fluorescence at 340/380 nm excitation, which is proportional to intracellular Ca 2+ concentration. Tracings are representative of 3 experiments. Figure 2. Expression of ipla 2. Endogenous expression of ipla 2 (A) and ipla 2 mrnas (B) in GEC and rat glomeruli was assessed using RT-PCR. Glomeruli were isolated from adult rat kidneys by differential sieving. Control lanes (-) were devoid of reverse transcriptase. A. GEC express ipla 2 short mrna and possibly some ipla 2 long. ipla 2 RT-PCR in Chinese hamster ovary cells (CHO), which express only the short isoform (577 bp), is shown for comparison. B. GEC express ipla 2 mrna. Expression of ipla 2 mrnas in glomeruli is similar to expression in GEC (A and B). Expression of ipla 2 (C) and ipla 2 proteins (D) in GEC and glomeruli was assessed by immunoblotting. COS cells transfected with the ipla 2 short and long isoforms (C), and COS cells transfected with ipla 2 (D) are shown for comparison. C. GEC express an immunoreactive ipla 2 protein whose molecular mass appears in between the long and short isoforms expressed by COS cells. D. GEC express ipla 2. The 26

27 Page 27 of lower molecular mass bands in glomeruli and GEC (C), and COS-iPLA2 (D) are probably nonspecific. Figure 3. Overexpression of ipla 2 in GEC. A. Lysates of control GEC and clones of GEC stably transfected with ipla 2 (GEC-iPLA 2 ) were immunoblotted with an antibody to ipla 2. GEC-iPLA 2 express 88, 77, and 74 kda splice variants of the enzyme. Expression of endogenous ipla 2 in parental GEC is not visible at this exposure. B. Serially-increasing amounts of cell extracts from GEC-iPLA 2 and GEC-Neo were incubated with [ 14 C]PE for 45 min at 37 C in the absence of Ca 2+. The data were analyzed using two-way ANOVA. The increase in PLA 2 activity was significant. p< ipla 2 vs Neo, 6 measurements. Figure 4. Overexpression of ipla 2 in GEC. A. Lysates of control GEC and a clone of GEC stably transfected with ipla 2 (GEC-iPLA 2 ) were immunoblotted with an antibody to ipla 2. COS cells are shown for comparison. In GEC-iPLA 2, the transfected enzyme migrates slightly faster than the endogenous, and it comigrates with ipla 2 short in transfected COS cells. B. Serially-increasing amounts of cell extracts from GEC-iPLA 2 and GEC-Neo were incubated with [ 14 C]PE for 45 min at 37 C in the absence of Ca 2+. The data were analyzed using two-way ANOVA. p<0.05, ipla 2 vs Neo, 6 measurements. Figure 5. Effects of complement, H 2 O 2 and anoxia/recovery on [ 3 H]AA release in GEC-iPLA 2. A. Complement-induced release of [ 3 H]AA is enhanced in GEC that overexpress GEC-iPLA 2. GEC-iPLA 2 and GEC-Neo were labeled with [ 3 H]AA, and incubated with anti-gec antiserum for 30 min at 22 C in the presence or absence of the ipla 2 -directed inhibitor, BEL (25 µm). The 27

28 28 Page 28 of 52 cells were then incubated at 37 C with 3% NS (to form C5b-9) or HIS for 60 minutes. *p< NS vs HIS (GEC-Neo), **p< NS vs HIS (GEC-iPLA 2 ) and p<0.003 GEC-iPLA 2 vs GEC-Neo (NS), + p<0.001 NS vs NS/BEL (GEC-iPLA 2 ), 4 experiments. BEL tended to increase [ 3 H]AA release in control cells modestly, but the change was not significant. B. H 2 O 2 -induced release of [ 3 H]AA is enhanced in GEC that overexpress GEC-iPLA 2. [ 3 H]AA-labeled GECiPLA 2 and GEC-Neo were preincubated with or without BEL for 30 minutes at 37 C and were then incubated with 1 mm H 2 O 2 for 30 min. *p<0.02 H 2 O 2 vs Control (GEC-iPLA 2 ) and p<0.01 GEC-iPLA 2 vs GEC-Neo (H 2 O 2 ), 3 experiments. C. Anoxia/recovery (Anoxia)-induced release of [ 3 H]AA is enhanced in GEC that overexpress GEC-iPLA 2. Anoxia was induced by incubating cells with 10 mm 2-deoxyglucose and 10 µm antimycin A in glucose-free buffer at 37 C for 90 min. This step was followed by recovery in glucose-replete medium for 30 min at 37 C. *p<0.05 Anoxia vs Control (GEC-iPLA 2 ) and p<0.05 GEC-iPLA 2 vs GEC-Neo (Anoxia), **p<0.05 Anoxia vs Anoxia/BEL (GEC-iPLA 2 ), 3 experiments. Figure 6. Complement-induced production of PGE 2 is enhanced in GEC that overexpress GECiPLA 2. Incubation with complement is described in Fig. 5. In some experiments, BEL (25 µm) or the COX inhibitor, indomethacin (10 µm) were included in the incubations. GEC-iPLA 2 produced significantly more PGE 2 than GEC-Neo in response to complement, and this increase was attenuated by the addition of BEL or indomethacin. *p< GEC-iPLA 2 vs GEC-Neo (NS), **p< NS/BEL vs NS (GEC-iPLA 2 ), + p< NS/indomethacin vs. NS (GECiPLA 2 ), 5 experiments. 28

29 Page 29 of Figure 7. Overexpression of ipla 2 (A), or ipla 2 (B) limits complement-mediated GEC injury. GEC-iPLA 2, GEC-iPLA 2 and GEC-Neo were incubated with anti-gec antiserum and then with 2.5, 5.0, or 7.5% NS for 100 min (HIS in control incubations). Cell injury was monitored as release of LDH. A. *p< GEC-iPLA 2 vs GEC-Neo (5.0% NS), **p<0.025 GEC-iPLA 2 vs GEC-Neo (7.5% NS), 3 experiments. Similar results were obtained when these experiments were performed using another GEC-Neo clone. B. *p<0.005 GEC-iPLA 2 vs GEC- Neo (2.5% NS), **p< GEC-iPLA 2 vs GEC-Neo (5.0% NS), + p< GEC-iPLA 2 vs GEC-Neo (7.5% NS), 4 experiments. Figure 8. The protective effect of ipla 2, but not ipla 2 on complement-mediated injury is prostaglandin-dependent. A. GEC-iPLA 2 and GEC-Neo were preincubated with or without BEL (25 µm) or indomethacin (10 µm) for 30 min. Cells were then incubated with anti-gec antiserum and 5% NS (HIS in control) for 100 min. *p< GEC-iPLA 2 vs GEC-Neo (NS), **p< NS/BEL vs NS (GEC-iPLA2), + p< NS/indomethacin vs NS (GEC-iPLA 2 ), 3 experiments. B. GEC-iPLA 2 and GEC-Neo were preincubated with or without NS 398 (10 µm) or SC560 (10 µm) for 30 min. Cells were then incubated as described above. *p< GEC-iPLA 2 vs GEC-Neo (NS), **p<0.04 NS/NS398 vs NS (GEC-iPLA2), + p<0.002 NS/SC560 vs NS (GEC-iPLA 2 ), 4 experiments. C. GEC-iPLA 2 and GEC-Neo were treated as in Fig. 8A. *p<0.01 GEC-iPLA 2 vs GEC-Neo (NS), 4 experiments. Figure 9. Expression of COX-2, but not COX-1, is upregulated in GEC-iPLA 2. Representative immunoblots demonstrate expression of COX-2 (A) and COX-1 proteins (B) in GEC-Neo, and GEC-iPLA 2. Expression in transiently transfected COS cells is shown for comparison. 29

30 30 Page 30 of 52 Figure 10. EP 4 receptor mrna is expressed in GEC-Neo and GEC-iPLA 2. Semi-quantitative RT-PCR demonstrates that there are no significant differences in the amount of EP 4 receptor mrna in the two cell lines. -actin was used as a loading control. 30

31 Page 31 of Table 1. Effect of BEL on cpla 2 activity Buffer % [ 14 C]AA Release 1 mm EGTA 8.3 ± mm EGTA+BEL 6.8 ± mm Calcium 48.8 ± 0.5* 2 mm Calcium+BEL 48.7 ± 2.0* PLA 2 activity was measured in extracts of COS cells that had been transiently transfected with cpla 2 by monitoring release of [ 14 C]AA from exogenous [ 14 C]PC. PLA 2 activity in Ca 2+ -replete buffer was not affected by BEL. *p< vs 1mM EGTA, 6 measurements. 31

32 32 Page 32 of 52 Table 2. Effects of H 2 O 2 or Anoxia/Recovery on Free [ 3 H]AA A GEC-Neo GEC-cPLA 2 Untreated BEL 139 ± ± 15 H 2 O ± 64* 460 ± 28* H 2 O 2 +BEL 196 ± 17** 135 ± 11** B GEC-Neo GEC-cPLA 2 Untreated BEL 153 ± ± 23 A/R 256 ± ± 19 + A/R+BEL 176 ± ± 21 Overexpression of cpla 2 in GEC (GEC-cPLA 2 ) does not enhance [ 3 H]AA release after incubation with H 2 O 2 or Anoxia/Recovery (A/R). Values represent [ 3 H]AA release in treated/untreated cells (%). A: *p< H 2 O 2 vs untreated (GEC-Neo and GEC-cPLA 2 ), **p< H 2 O 2 +BEL vs H 2 O 2 (GEC-Neo and GEC-cPLA 2 ), 3 experiments. B: + p< A/R vs untreated (GEC-Neo and GEC-cPLA 2 ), ++ p<0.01 A/R+BEL vs A/R (GEC-Neo), 4 experiments. There are no significant differences between GEC-Neo and GEC-cPLA 2 after treatment with H 2 O 2 or A/R. 32

33 Page 33 of REFERENCES 1. Ackermann EJ, Conde-Frieboes K, and Dennis EA. Inhibition of macrophage Ca(2+)- independent phospholipase A 2 by bromoenol lactone and trifluoromethyl ketones. J Biol Chem 270: , Akiba S, and Sato T. Cellular function of calcium-independent phospholipase A 2. Biol Pharm Bull 27: , Aoudjit L, Potapov A, and Takano T. Prostaglandin E 2 promotes cell survival of glomerular epithelial cells via the EP4 receptor. Am J Physiol Renal Physiol 290: F , Baburina I, and Jackowski S. Cellular responses to excess phospholipid. J Biol Chem 274: , Balsinde J, and Balboa MA. Cellular regulation and proposed biological functions of group VIA calcium-independent phospholipase A 2 in activated cells. Cell Signal 17: , Balsinde J, Balboa MA, and Dennis EA. Antisense inhibition of group VI Ca2+independent phospholipase A 2 blocks phospholipid fatty acid remodeling in murine P388D1 macrophages. J Biol Chem 272: , Bao S, Miller DJ, Ma Z, Wohltmann M, Eng G, Ramanadham S, Moley K, and Turk J. Male mice that do not express group VIA phospholipase A 2 produce spermatozoa with impaired motility and have greatly reduced fertility. J Biol Chem 279: , Barbour SE, Kapur A, and Deal CL. Regulation of phosphatidylcholine homeostasis by calcium-independent phospholipase A 2. Biochim Biophys Acta 1439: 77-88,

34 34 Page 34 of Bhakdi S, and Tranum-Jensen J. Membrane damage by complement. Biochim Biophys Acta 737: , Birbes H, Gothie E, Pageaux JF, Lagarde M, and Laugier C. Hydrogen peroxide activation of Ca(2+)-independent phospholipase A(2) in uterine stromal cells. Biochem Biophys Res Commun 276: , Bonventre JV. The 85-kD cytosolic phospholipase A 2 knockout mouse: a new tool for physiology and cell biology. J Am Soc Nephrol 10: , Chilton F. Would the real role(s) for secretory PLA 2 s please stand up. J Clin Invest 97: , Chiu CH, and Jackowski S. Role of calcium-independent phospholipases (ipla(2)) in phosphatidylcholine metabolism. Biochem Biophys Res Commun 287: , Coers W, Reivinen J, Miettinen A, Huitema S, Vos JT, Salant DJ, and Weening JJ. Characterization of a rat glomerular visceral epithelial cell line. Exp Nephrol 4: , Cummings BS, McHowat J, and Schnellmann RG. Phospholipase A(2)s in cell injury and death. J Pharmacol Exp Ther 294: , Cummings BS, McHowat J, and Schnellmann RG. Role of an endoplasmic reticulum Ca2+-independent phospholipase A 2 in cisplatin-induced renal cell apoptosis. J Pharmacol Exp Ther 308: , Cummings BS, McHowat J, and Schnellmann RG. Role of an endoplasmic reticulum Ca(2+)-independent phospholipase A(2) in oxidant-induced renal cell death. Am J Physiol Renal Physiol 283: F ,

35 Page 35 of Cybulsky AV, Monge JC, Papillon J, and McTavish AJ. Complement C5b-9 activates cytosolic phospholipase A 2 in glomerular epithelial cells. Am J Physiol 269: F , Cybulsky AV, Quigg RJ, and Salant DJ. Experimental membranous nephropathy redux. Am J Physiol Renal Physiol 289: F , Cybulsky AV, Salant DJ, Quigg RJ, Badalamenti J, and Bonventre JV. Complement C5b-9 complex activates phospholipases in glomerular epithelial cells. Am J Physiol 257: F , Cybulsky AV, Takano T, Papillon J, Bijian K, and Guillemette J. Activation of the extracellular signal-regulated kinase by complement C5b-9. Am J Physiol Renal Physiol 289: F , Cybulsky AV, Takano T, Papillon J, Khadir A, Liu J, and Peng H. Complement C5b-9 membrane attack complex increases expression of endoplasmic reticulum stress proteins in glomerular epithelial cells. J Biol Chem 277: , Cybulsky AV, Takano T, Papillon J, and McTavish AJ. Complement-induced phospholipase A 2 activation in experimental membranous nephropathy. Kidney Int 57: , Dagher PC. Modeling ischemia in vitro: selective depletion of adenine and guanine nucleotide pools. Am J Physiol Cell Physiol 279: C , Gadd ME, Broekemeier KM, Crouser ED, Kumar J, Graff G, and Pfeiffer DR. Mitochondrial ipla 2 activity modulates the release of cytochrome c from mitochondria and influences the permeability transition. J Biol Chem 281: , Han WK, Sapirstein A, Hung CC, Alessandrini A, and Bonventre JV. Cross-talk between cytosolic phospholipase A 2 alpha (cpla 2 alpha) and secretory phospholipase A 2 35

36 36 Page 36 of 52 (spla 2 ) in hydrogen peroxide-induced arachidonic acid release in murine mesangial cells: spla2 regulates cpla 2 alpha activity that is responsible for arachidonic acid release. J Biol Chem 278: , Hirabayashi T, and Shimizu T. Localization and regulation of cytosolic phospholipase A(2). Biochim Biophys Acta 1488: , Kinsey GR, McHowat J, Beckett CS, and Schnellmann RG. Identification of calciumindependent phospholipase A 2 gamma in mitochondria and its role in mitochondrial oxidative stress. Am J Physiol Renal Physiol 292: F , Larsson PK, Claesson HE, and Kennedy BP. Multiple splice variants of the human calcium-independent phospholipase A 2 and their effect on enzyme activity. J Biol Chem 273: , Mancuso DJ, Abendschein DR, Jenkins CM, Han X, Saffitz JE, Schuessler RB, and Gross RW. Cardiac ischemia activates calcium-independent phospholipase A 2 beta, precipitating ventricular tachyarrhythmias in transgenic mice: rescue of the lethal electrophysiologic phenotype by mechanism-based inhibition. J Biol Chem 278: , Mancuso DJ, Jenkins CM, Sims HF, Cohen JM, Yang J, and Gross RW. Complex transcriptional and translational regulation of iplagamma resulting in multiple gene products containing dual competing sites for mitochondrial or peroxisomal localization. Eur J Biochem 271: , Mancuso DJ, Sims HF, Han X, Jenkins CM, Guan SP, Yang K, Moon SH, Pietka T, Abumrad NA, Schlesinger PH, and Gross RW. Genetic Ablation of Calcium-independent Phospholipase A 2 {gamma} Leads to Alterations in Mitochondrial Lipid Metabolism and 36

37 Page 37 of Function Resulting in a Deficient Mitochondrial Bioenergetic Phenotype. J Biol Chem 282: , Mundel P, and Shankland SJ. Podocyte biology and response to injury. J Am Soc Nephrol 13: , Murakami M, and Kudo I. Phospholipase A 2. J Biochem (Tokyo) 131: , Murakami M, Masuda S, Ueda-Semmyo K, Yoda E, Kuwata H, Takanezawa Y, Aoki J, Arai H, Sumimoto H, Ishikawa Y, Ishii T, Nakatani Y, and Kudo I. Group VIB Ca2+independent phospholipase A 2 gamma promotes cellular membrane hydrolysis and prostaglandin production in a manner distinct from other intracellular phospholipases A 2. J Biol Chem 280: , Neale TJ, Ojha PP, Exner M, Poczewski H, Ruger B, Witztum JL, Davis P, and Kerjaschki D. Proteinuria in passive Heymann nephritis is associated with lipid peroxidation and formation of adducts on type IV collagen. J Clin Invest 94: , Panesar M, Papillon J, McTavish AJ, and Cybulsky AV. Activation of phospholipase A 2 by complement C5b-9 in glomerular epithelial cells. J Immunol 159: , Pavenstadt H, Kriz W, and Kretzler M. Cell biology of the glomerular podocyte. Physiol Rev 83: , Portilla D. Role of fatty acid beta-oxidation and calcium-independent phospholipase A 2 in ischemic acute renal failure. Curr Opin Nephrol Hypertens 8: , Quigg RJ, Cybulsky AV, Jacobs JB, and Salant DJ. Anti-Fx1A produces complementdependent cytotoxicity of glomerular epithelial cells. Kidney Int 34: 43-52,

38 38 Page 38 of Sapirstein A, Spech RA, Witzgall R, and Bonventre JV. Cytosolic phospholipase A 2 (PLA 2 ), but not secretory PLA 2, potentiates hydrogen peroxide cytotoxicity in kidney epithelial cells. J Biol Chem 271: , Shankland SJ. The podocyte's response to injury: role in proteinuria and glomerulosclerosis. Kidney Int 69: , Six DA, and Dennis EA. The expanding superfamily of phospholipase A(2) enzymes: classification and characterization. Biochim Biophys Acta 1488: 1-19, Smith WL, Garavito RM, and DeWitt DL. Prostaglandin endoperoxide H synthases (cyclooxygenases)-1 and -2. J Biol Chem 271: , Takano T, and Cybulsky AV. Complement C5b-9-mediated arachidonic acid metabolism in glomerular epithelial cells : role of cyclooxygenase-1 and -2. Am J Pathol 156: , Takano T, Cybulsky AV, Cupples WA, Ajikobi DO, Papillon J, and Aoudjit L. Inhibition of cyclooxygenases reduces complement-induced glomerular epithelial cell injury and proteinuria in passive Heymann nephritis. J Pharmacol Exp Ther 305: , Yang J, Han X, and Gross RW. Identification of hepatic peroxisomal phospholipase A(2) and characterization of arachidonic acid-containing choline glycerophospholipids in hepatic peroxisomes. FEBS Lett 546: ,

39 Page 39 of 52 Figure 1. A. C5b-9 induces Ca2+-independent release of [3H]AA in GEC. GEC, labeled with [3H]AA, were incubated with antibody and 10% C8-deficient serum (C8DS) in buffer containing 0.5 mm Ca2+ for 40 min. Then, GEC were washed, and incubated with or without purified C8 (2.0 g/ml) and C9 (1.5 g/ml; to assemble C5b-9) in buffer containing 1 mm EGTA. [3H]AA and [3H]DAG were measured after 40 min. *p<0.025 C8DS+C8+C9 vs C8DS alone, 5 experiments. B. GEC, loaded with fura-2, were placed into buffer containing 2 mm EGTA. Ionomycin (5 M) was added after 3 min (top panel) or 23 min (bottom panel), and fura-2 fluorescence was monitored for at least another 7 min. Results are presented as the ratio (R) of fluorescence at 340/380 nm excitation, which is proportional to intracellular Ca2+ concentration. Tracings are representative of 3 experiments. 190x254mm (300 x 300 DPI)

40 Figure 2. Expression of ipla2. Endogenous expression of ipla2= (A) and ipla2> mrnas (B) in GEC and rat glomeruli was assessed using RT-PCR. Glomeruli were isolated from adult rat kidneys by differential sieving. Control lanes (-) were devoid of reverse transcriptase. A. GEC express ipla2= short mrna and possibly some ipla2= long. ipla2= RT-PCR in Chinese hamster ovary cells (CHO), which express only the short isoform (577 bp), is shown for comparison. B. GEC express ipla2> mrna. Expression of ipla2 mrnas in glomeruli is similar to expression in GEC (A and B). Expression of ipla2= (C) and ipla2> proteins (D) in GEC and glomeruli was assessed by immunoblotting. COS cells transfected with the ipla2= short and long isoforms (C), and COS cells transfected with ipla2> (D) are shown for comparison. C. GEC express an immunoreactive ipla2= protein whose molecular mass appears in between the long and short isoforms expressed by COS cells. D. GEC express ipla2>. The lower molecular mass bands in glomeruli and GEC (C), Page 40 of 52

41 Page 41 of 52 and COS-iPLA2g (D) are probably nonspecific. 190x254mm (300 x 300 DPI)

42 Figure 3. Overexpression of ipla2> in GEC. A. Lysates of control GEC and clones of GEC stably transfected with ipla2> (GEC-iPLA2>) were immunoblotted with an antibody to ipla2>. GEC-iPLA2> express 88, 77, and 74 kda splice variants of the > enzyme. Expression of endogenous ipla2> in parental GEC is not visible at this exposure. B. Serially-increasing amounts of cell extracts from GEC-iPLA2> and GEC-Neo were incubated with [14C]PE for 45 min at 37 C in the absence of Ca2+. The data were analyzed using two-way ANOVA. The increase in PLA2 activity was significant. p< ipla2> vs Neo, 6 measurements. 190x254mm (300 x 300 DPI) Page 42 of 52

43 Page 43 of 52 Figure 4. Overexpression of ipla2= in GEC. A. Lysates of control GEC and a clone of GEC stably transfected with ipla2= (GEC-iPLA2=) were immunoblotted with an antibody to ipla2=. COS cells are shown for comparison. In GEC-iPLA2=, the transfected enzyme migrates slightly faster than the endogenous, and it comigrates with ipla2= short in transfected COS cells. B. Serially-increasing amounts of cell extracts from GEC-iPLA2= and GEC-Neo were incubated with [14C]PE for 45 min at 37 C in the absence of Ca2+. The data were analyzed using two-way ANOVA. p<0.05, ipla2= vs Neo, 6 measurements. 190x254mm (300 x 300 DPI)

44 Figure 5. Effects of complement, H2O2 and anoxia/recovery on [3H]AA release in GECiPLA2>. A. Complement-induced release of [3H]AA is enhanced in GEC that overexpress GEC-iPLA2>. GEC-iPLA2> and GEC-Neo were labeled with [3H]AA, and incubated with anti-gec antiserum for 30 min at 22 C in the presence or absence of the ipla2-directed inhibitor, BEL (25 µm). The cells were then incubated at 37 C with 3% NS (to form C5b-9) or HIS for 60 minutes. *p< NS vs HIS (GEC-Neo), **p< NS vs HIS (GECiPLA2>) and p<0.003 GEC-iPLA2> vs GEC-Neo (NS), +p<0.001 NS vs NS/BEL (GECiPLA2>), 4 experiments. BEL tended to increase [3H]AA release in control cells modestly, but the change was not significant. B. H2O2-induced release of [3H]AA is enhanced in GEC that overexpress GEC-iPLA2>. [3H]AA-labeled GEC-iPLA2> and GEC-Neo were preincubated with or without BEL for 30 minutes at 37 C and were then incubated with 1 mm H2O2 for 30 min. *p<0.02 H2O2 vs Control (GEC-iPLA2>) and p<0.01 GEC-iPLA2> vs Page 44 of 52

45 Page 45 of 52 GEC-Neo (H2O2), 3 experiments. C. Anoxia/recovery (Anoxia)-induced release of [3H]AA is enhanced in GEC that overexpress GEC-iPLA2>. Anoxia was induced by incubating cells with 10 mm 2-deoxyglucose and 10 µm antimycin A in glucose-free buffer at 37 C for 90 min. This step was followed by recovery in glucose-replete medium for 30 min at 37 C. *p<0.05 Anoxia vs Control (GEC-iPLA2>) and p<0.05 GEC-iPLA2> vs GEC-Neo (Anoxia), **p<0.05 Anoxia vs Anoxia/BEL (GEC-iPLA2>), 3 experiments. 190x254mm (300 x 300 DPI)

46 Figure 6. Complement-induced production of PGE2 is enhanced in GEC that overexpress GEC-iPLA2>. Incubation with complement is described in Fig. 5. In some experiments, BEL (25 µm) or the COX inhibitor, indomethacin (10 µm) were included in the incubations. GEC-iPLA2> produced significantly more PGE2 than GEC-Neo in response to complement, and this increase was attenuated by the addition of BEL or indomethacin. *p< GECiPLA2> vs GEC-Neo (NS), **p< NS/BEL vs NS (GEC-iPLA2>), +p< NS/indomethacin vs. NS (GEC-iPLA2>), 5 experiments. 190x254mm (300 x 300 DPI) Page 46 of 52

47 Page 47 of 52 Figure 7. Overexpression of ipla2> (A), or ipla2= (B) limits complement-mediated GEC injury. GEC-iPLA2>, GEC-iPLA2= and GEC-Neo were incubated with anti-gec antiserum and then with 2.5, 5.0, or 7.5% NS for 100 min (HIS in control incubations). Cell injury was monitored as release of LDH. A. *p< GEC-iPLA2> vs GEC-Neo (5.0% NS), **p<0.025 GEC-iPLA2> vs GEC-Neo (7.5% NS), 3 experiments. Similar results were obtained when these experiments were performed using another GEC-Neo clone. B. *p<0.005 GEC-iPLA2= vs GEC-Neo (2.5% NS), **p< GEC-iPLA2= vs GEC-Neo (5.0% NS), +p< GEC-iPLA2= vs GEC-Neo (7.5% NS), 4 experiments. 190x254mm (300 x 300 DPI)

48 190x254mm (300 x 300 DPI) Page 48 of 52

49 Page 49 of 52 Figure 8. The protective effect of ipla2>, but not ipla2 on complement-mediated injury is prostaglandin-dependent. A. GEC-iPLA2> and GEC-Neo were preincubated with or without BEL (25 µm) or indomethacin (10 µm) for 30 min. Cells were then incubated with anti-gec antiserum and 5% NS (HIS in control) for 100 min. *p< GEC-iPLA2> vs GEC-Neo (NS), **p< NS/BEL vs NS (GEC-iPLA2>), +p< NS/indomethacin vs NS (GEC-iPLA2>), 3 experiments. B. GEC-iPLA2> and GEC-Neo were preincubated with or without NS 398 (10 µm) or SC560 (10 µm) for 30 min. Cells were then incubated as described above. *p< GEC-iPLA2> vs GEC-Neo (NS), **p<0.04 NS/NS398 vs NS (GEC-iPLA2>), +p<0.002 NS/SC560 vs NS (GEC-iPLA2>), 4 experiments. C. GEC-iPLA2 and GEC-Neo were treated as in Fig. 8A. *p<0.01 GEC-iPLA2 vs GEC-Neo (NS), 4 experiments.

50 190x254mm (300 x 300 DPI) Page 50 of 52

51 Page 51 of 52 Figure 9. Expression of COX-2, but not COX-1, is upregulated in GEC-iPLA2>. Representative immunoblots demonstrate expression of COX-2 (A) and COX-1 proteins (B) in GEC-Neo, and GEC-iPLA2>. Expression in transiently transfected COS cells is shown for comparison. 190x254mm (300 x 300 DPI)

52 Figure 10. EP4 receptor mrna is expressed in GEC-Neo and GEC-iPLA2>. Semiquantitative RT-PCR demonstrates that there are no significant differences in the amount of EP4 receptor mrna in the two cell lines. =-actin was used as a loading control. 190x254mm (300 x 300 DPI) Page 52 of 52