Mechanisms in bradykinin stimulated arachidonate release and synthesis of prostaglandin and platelet activating factor

Regulatory mechanisms in bradykinin (BK) activated release of arachidonate (ARA) and synthesis of prostaglandin (PG) and platelet activating factor (PAF) were studied in bovine pulmonary artery endothelial cells (BPAEC). A role for GTP binding protein (G-protein) in the binding of BK to the cells was determined. Guanosine 5-O- (thiotriphosphate), (GTPτS), lowered the binding affinity for BK and increased the Kd for the binding from 0.45 to 1.99 nM. The Bmax remained unaltered at 2.25 × 10-11 mole. Exposure of the cells to aluminium fluoride also reduced the affinity for BK. Bradykinin-induced release of ARA proved pertussis toxin (PTX) sensitive, with a maximum sensitivity at 10 ug/ml PTX. GTPτS at 100 μM increased the release of arachidonate. The effect of GTPτS and BK was additive at suboptimal doses of BK up to 0.5 nM but never exceeded the levels of maximal BK stimulation at 50 nM. PTX also inhibited the release of ARA induced by the calcium ionophore, A23187. Phorbol 12-myristate 13-acetate or more commonly known as tetradecanoyl phorbol acetate (TPA) itself had little effect on release by the intact cells. However, at 100 nM it augmented the BK activated release. This was downregulated by overnight exposure to TPA and correlated with down-regulation of protein kinase C (PKC) activity. The down-regulation only affected the augmentation of ARA release by TPA but not the original BK activated release. TPA displayed a similar, but more potent amplification of PAF synthesis in response to both BK or the calcium ionophore A23187. These results taken together point to the participation of G-protein in the binding of BK to BPAEC and its activation of ARA release. Possibly two types of G-protein are involved, one associated with the receptor, the other activated by Ca2+ and perhaps associated with phospholipase A2 (PLA2). Our results further suggest that a separate route of activation, probably also PLA2 related, takes place through a PKC catalysed phosphorylation.


Introduction
The exposure of endothelial cells to bradykinin (BK), a nonapeptide with inflammatory as well as vasoactive functions, results in the activation of a number of early metabolic events as determined in a number of cell types. These events include a biphasic increase in cytosolic Ca2+,1-4 a transient membrane hyperpolarization, the activation of K + channels, 6 formation of diacylglycerol (DAG) and phosphatidic acid, 7 release of arachidonate (ARA) from phospholipid stores and synthesis of prostaglandin (PG) and platelet activating factor (PAF). Many of these events involve the activation of phospholipid related hydrolases including phospholipase A 2 (PLA2) and phospholipase C (PLC).
Whereas the generation of inositol triphosphate (IP3) and DAG are due to the activation of PLC, cumulative evidence suggests that the release of free arachidonic acid from phospholipid stores, at least in endothelial cells, involves PLA2 activation by BK. 1 Although this release of fatty acid has been well described, the sequence of events which follows the binding of BK to the cell and results in the activation of PLA2 remains to be fully elucidated. The resulting increase in cytosolic Ca 2 + is certainly a necessary step. However regulatory systems such as GTP-binding proteins (G-protein) and DAG activated protein kinase C (PKC) have (C) 1992 Rapid Communications of Oxford Ltd. been implicated in the process. 11 '12 In this series of experiments we attempt to link some of these second messengers to elucidate the process of activation of PLA2 by BK in the bovine pulmonary artery endothelial cell (BPAEC).

Materials and Methods
Cell culture: Endothelial cells were isolated without the use of proteolytic enzymes. A freshly obtained calf pulmonary artery was cut open and lightly scraped with a scalpel. The resulting clumps of endothelial cells were placed into 13 The homogeneity of endothelial cell cultures was determined morphologically and histochemically. At confluence the cells displayed a typical cobblestone appearance and stained positive for factor VIII antigen according to the method of Weinberg et al.14 Endothelial cells were maintained in 25cm 2 flasks containing 6 ml McCoy's 5A medium, 20% FBS, 50 #g/ml streptomycin and 50units/ml penicillin. The culture medium was replaced every 3 days and the cells passaged biweekly. Cells to be used for experiments were passaged from 25 cm 2 flasks into 24-well plates, 20 cm 2 dishes or 60 cm 2 dishes at a density of 20 000 cells/cm 2 with trypsin (0.05%). The culture medium was replaced after 6 days. Two to 3 days after the last feeding, these cultures were used for the various experiments.
Release of arachidonate: Confluent cells in 24-well plates were prelabelled with 3H-arachidonate [79.9 Ci/mmol, New England Nuclear, MA] (0.2/tCi per well) for 18 h. They were then washed once with McCoy's medium and preincubated for 2 h in McCoy's medium containing 1% FBS (0.5 ml per well). This process was found previously to increase the potential of the cell to synthesize prostaglandin. 13 In experiments with phorbol ester, 50/al of a 10 x concentration of TPA was added to the wells during the last 10 min of the preincubation. Cells were then incubated in McCoy's medium containing 2 mg/ml bovine serum albumin (BSA) (Sigma, St Louis, MO, essentially fatty acid free from essentially globulin free) and the indicated additions. After 10 min the medium was removed and centrifuged (800 g,). Radioactivity was determined in a 200 #1 aliquot of the supernatant. Radioimmunoassay (RIA ) for prostaglandins Antibodies to 6-keto PGFI (6-K-PGFI) were prepared in our laboratory. PGI2 concentrations were determined as its stable degradation product 6-K-PGFI. Crossreactivity of the antisera against nontargeted PGs was less than 4%. 11 The radioimmunoassay was performed as described previously. 13 Platelet activating factor: The synthesis of PAF was determined as described previously, is Generally the procedure of McIntyre et al. was followed. 9 Confluent cells (20 cm 2 Petri dishes) were incubated with 3H-acetate [ Binding of BK to cells: The binding of BK to cells was carried out as described previously. 16 Binding was done in 24-well plates at 4C for 2 h. The cultures were washed three times with 1 ml of phosphatebuffered saline (72 mM NaC1, 1.6 mM KC1, 5 mM NaiHPO4, 0.9mMKH2PO4) at 4C. This was followed by a 15 min equilibration with 0.5 ml of modified Hank's Balanced Salt Solution (HBSS), pH 7.3, containing 0.05% bovine serum albumin (BSA), 2 mM bacitracin, 10 mM HEPES, 120 mM N-methyl-D-glucamine (replacing NaC1), 0.65 mM CaCl2, 0.25 mM MgC12 and 0.25 mM MgSO4. This binding medium was removed and replaced with fresh, chilled binding medium containing the 3H-bradykinin [90 Ci/mmol, Amersham, IL]. Nonspecific binding was determined in the presence of 3/,M unlabelled bradykinin. At the end of the incubation, the medium was aspirated and the cells were washed five times with 1 ml of the Modified Balanced Salt Solution with 0.2% BSA. This was followed by two washes with 1 ml of phosphate-buffered saline. Bound radioactivity was determined by solubilizing the cells with 0.5 ml of 0.2% sodium dodecyl sulphate. Radioactivity was quantitated in New England Nuclear [Boston, MA] 963 using an LKB Rackbeta Counter.
Protein kinase C: To determine PKC activity, cells were grown to confluence in 60 cm 2 dishes. They were treated with various effectors and then washed with cold PBS before harvesting by scraping into sample buffer (20mM Tris buffer, pH 7.5 containing 0.33 M sucrose, 2 mM EDTA, 0.5 mM EGTA, and 100/g/ml leupeptin). The scraped cells were centrifuged at 2 000 x g, and the PKC was solubilized in 1% Triton x 100 in sample buffer for 30 min on ice. The insoluble material was removed by centrifugation and the supernatant containing the PKC was absorbed to DE52 resin. The resin was then washed with 10 ml sample buyer. PKC activity was eluted from the resin with sample buffer containing 100 mM NaC1. Activity was determined as described by Navarro et al. 17 Aliquots of the enzyme were mixed with 10 mM MgC12, 100/M .32p ATP (1 000 dpm/pmol), 50/g of histone III-S with or without 1 mM CaC12, 5 #g of phosphatidylserine, and 20 ng of phorbol dibutyrate (PDBu) in a final volume of 50/1. Samples were incubated for 10 min at 30C. The reaction was stopped by spotting 25/1 of the reaction mixture onto Whatman 3MM paper and then washing the filter papers in 10% trichloroacetic acid containing 10 mM sodium pyrophosphate. Protein kinase C activity was determined by subtracting the amount of 32p incorporation into histone in the absence of added Ca2+, phosphatidylserine, and PDBu.
Assay for sn-1,2-diacylgcerol: Diacylglycerol was determined according to the method of Preiss et al.18 Cells were grown in 20 cm 2 plates and treated as indicated in the figure legends. Lipids were extracted and the dried lipids solubilized in 20/1 of an octyl-B-D-glycoside/cardiolipin solution (7.5% octyl-B-D-glycoside, 5 mM cardiolipin in 1 mM diethylenetriaminepentaacetic acid (DETAPAC) by sonication in a bath sonicator). Fifty microlitres reaction buffer (100 mM imidazole HC1, pH 6.6, 100 mM NaC1, 25 mM MgCI2 and 2 mM EGTA), 2/1 100 mM dithiothreitol, 10/1 diluted membranes containing DAG kinase (5/.zg protein) and 8 #1 water were then added. The reaction was started with the addition of 10/1 10mM (.2p) ATP prepared in 100 mM imidazole, 1 mM DETAPAC, pH 6.6. After 30min at 25C the reaction was stopped with chloroform/methanol and the lipids extracted. The chloroform layer was washed twice with 2 ml 1% HC10 4. The volume of the chloroform layer was measured and an aliquot removed and dried under nitrogen for TLC. Silica Gel 60 thin layer chromatography plates were activated by running in acetone and air dried immediately before spotting samples. Plates were developed with chloroform: methanol:acetic acid (65:15:5), air dried and spots located by autoradiography. Radioactivity was quantitated by counting the scraped silica in a scintillation counter. The amount of sn.-1,2-DAG present in the original sample was calculated from the sample volumes and the specific activity of the ATP.

Results
The effect of GTPS on binding of BK to BPAEC: To determine interaction between bradykinin, G-protein and phorbol 12-myristate 13-acetate (TPA) we first looked at the effect of GTPS on the binding of BK to endothelial ceils. This was done in intact cells which were permeabilized with saponin. A typical binding of BK to endothelial cells is illustrated as a binding curve (inset) and its Scatchard plot in Fig. 1. The permeabilization itself had no effect on the binding. The untreated cells bound with a Kd of 0.45 nM. Adding GTP:S (100 #M) to the incubation solution had a marked negative effect on the binding, increasing the Kd to 1.99nM. Bmax stayed constant as 2.2 x 10 -11 mole. GDPflS, structurally similar to GTPS but not an activator of G-protein, had no effect on binding (data not shown). In a separate experiment, aluminium fluoride, which dissociates the Gprotein, reduced the binding of BK significantly as illustrated in Fig. 2. ATP (10/M) had no effect.
The eect of GTPS on release of A RA: As illustrated in Fig. 3, the release of arachidonate from endothelial cells preincubated with H-ARA is related to the GTP:S concentration up to approximately 100#M. This release is time dependent as illustrated in Fig. 4. In this figure the release of arachidonate by GTP'S (100/M) is compared to that by BK (50 nM). At 2 min after addition, BK more than doubles the release of label while GTP:S has little effect. Some effect by GTP,S is seen at 5 and 10 min, but this is small in comparison to BK. At 20 min after stimulation the effect by GTP:S is 70% that of BK.
The effect of pertussis toxin (PTX) was tested with regard to the release of arachidonate and PG synthesis. As illustrated in Fig. 5, PTX had a negligible inhibition of release by cells not treated with BK. The release caused by BK (50 nM) was approximately 12-fold the basal value. PTX reduced the BK activated increase of arachidonate release to approximately twice that of the basal value. Interestingly, PTX also blocked the release caused by A23187 to a similar degree. The effect was related to the PTX concentration as illustrated in Fig. 6. Some effect is seen at 1 ng/ml; by 10 ng/ml maximum effect is seen. No further inhibition is detected at 100 ng/ml. The interaction between BK and GTP:S on the release of arachidonate is illustrated in Fig. 7.
GTPq:S alone at 100 #M increased release. BK increased release from 0.5 nM up to 5 nM. At suboptimal concentrations of BK, the addition of GTPq:S increased release above that of BK up to Phorbol 12-myristate 13-acetate: Exposure of BPAEC to TPA (100 nM) for 10 min in absence of BK had no effect on the release of arachidonate (Fig. 8).
However, as illustrated in the same figure, TPA increased BK stimulated release. When the cells were treated overnight with TPA (500 nM) to down-regulate PKC activity and then were pretreated with fresh TPA and exposed to BK the  response to BK remained unaltered, but no augmentation by TPA was observed (Fig. 8).
Determination of PKC activity, as illustrated in Fig.  8 (insert) shows that PKC was indeed downregulated by exposure of the cells to 500 nM TPA overnight. In Fig. 9  Cultures not treated with TPA 11 Cultures treated with 100 nM TPA production was determined at various concentrations of BK from 0.01 to 50 nM. In this case a maximum was reached between 0.5 and 2 nM BK.
The interaction of TPA with BK and GTP:S is illustrated in Fig. 11. The release of arachidonate was determined with either BK or GTPq:S in the presence or absence of TPA. In this case TPA basically doubled the eEect of BK. It proved to have no effect on GTP:S stimulated release.

Discussion
In the experiments described we investigated the high affinity B 2 binding site of BK in intact, viable  [] Cultures treated with either (nothing, GTP:S or BK). Culture treated as above -t-100 nM TPA. endothelial cells. We found evidence in these cells of only one high affinity B 2 site. 16  The generation of IP in BPAEC in response to BK has been shown to be G-protein linked but not sensitive to PTX. 25 Results here show a clear inhibition of ARA release by PTX in the same cells. Our observations further illustrate that PTX inhibits not only BK-activated release but also A21387-activated release. This suggests that the PTX sensitive G-protein may be functioning beyond the Ca 2+ mobilization step. Nakashima et al. reported previously that N-formyl-methionylleucyl-phenylalanine (fMLP) activated neutrophil release of arachidonate and PG synthesis. 26 They also reported that G-protein is involved in this interaction. They proposed that in the neutrophil G-protein lowers the requirement of PLA2 for Ca 2 +.
One explanation for these observations is that the PTX insensitive G-protein is associated directly with the BK receptor while the PTX sensitive G-protein is perhaps associated directly with PLA2.
The G-protein associated with the receptor may be involved in the BK regulation of cytosolic Ca 2+ concentrations. G-protein was recently shown to participate in the regulation of receptor operated Ca 2+ channels in platelets. 27 Also, a recent report illustrated that the complex A1Factivated calcium influx in endothelial cells. 7 The BK receptor may form a part of a Ca 2 + transporter. 28 The PTX sensitive G-protein in BPAEC may be one of the subtypes of G i. Clark et al. 29 have identified a 41 kDa protein as ADP ribosylated by PTX, consistent with the molecular weight of Gi. Lee et al. have demonstrated that endothelial cells express mRNA for all three subtypes of Gi. 3 Our results illustrate that TPA is also involved in the release of arachidonate in BPAEC. Generally phorbol ester has been shown to inhibit the action of PLC in the generation of IP3 .31 This inhibition is related to PKC. With regard to the activation of PLA2 by phorbol ester, the literature is more tenuous. For example, in the experiments using MDCK cells, 31 TPA itself stimulated the release of arachidonic acid which was inhibited by 1-(5isoquinolinylsulphonyl)-2-methyl piperazine (H7), an inhibitor of PKC. In another report using MDCK cells, the BK response was only slightly sensitive to PKC inhibitors (sphingosine, H7, staurosporine). 32 The BK-stimulated release of arachidonic acid could not be enhanced by TPA in PKC down-regulated BPAEC. However, in these cells TPA itself stimulated the release of arachidonate which was attenuated (50%) by PKC inhibitors. 32 Phorbol ester was reported to augment BK-stimulated PG synthesis in other cell types such as Swiss 3T3. 33 Other effectors such as thrombin, which activates PG and PAF synthesis in human endothelial cells, are augmented by TPA. 34 In the described experiments, TPA alone did little in the absence of BK. However, in the presence of BK the release of arachidonate was synergistic. This synergism by phorbol was abolished when PKC was down-regulated by incubating the cells overnight with TPA. However, the BK-stimulated release was totally unaffected by the downregulation of PKC. This action by TPA appears related to PLA2. This is suggested in Fig. 9 by the TPA-augmented, BK-induced synthesis of PAF. In these experiments the synthesis of PAF was determined through the remodelling pathway which utilizes PLA2 to remove fatty acid from the sn-2 position of 1-O-alkyl-2-acyl-sn-glyceryl-3-phosphocholine and incorporates acetate into the lyso-PAF via lyso-PAF acetyl-CoA acetyltransferase (LPAT) to form the 1-O-alkyl-2-acetyl-snglyceryl-3-phosphocholine (PAF). It is possible that TPA in conjunction with BK also activates LPAT. This is suggested by the considerably larger augmentation by TPA of BK-stimulated PAF synthesis than of ARA release. In fact Heller et al. 35 reported recently that thrombin-activated PAF synthesis in human umbilical vein endothelial cells involved the activation of LPAT. TPA alone did not activate LPAT. The effect of thrombin plus TPA on LPAT activity was not determined.
TPA also did not affect GTPS activation of release (Fig. 11), suggesting that the two events represent separate paths to the activation of PEA 2. As we illustrated, BK does generate DAG (Fig.  10), and thus may, under certain conditions, additionally activate the release of arachidonate through the action of protein kinase C.