Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Group IIA, secretory phospholipase A₂ (sPLA₂), is a member of a family of enzymes that hydrolyze the sn-2 ester bond from cellular phospholipids, liberating fatty acids, and lysophospholipids. When the released fatty acid is arachidonic acid, several proinflammatory lipid mediators (e.g., prostaglandins and leukotrienes) are formed. The lysophospholipid can be cytotoxic and under certain conditions serves as the precursor to platelet-activating factor. Thus, sPLA₂ has the capacity to release several highly reactive pharmacological agents and thereby could be a key player in promoting or amplifying inflammatory disease processes.

The physiological role of sPLA₂ is unknown. However, an antibacterial role against both Gram-negative and Gram-positive organisms has been suggested for sPLA₂ (Wright et al., 1990; Weiss et al., 1994; Weinrauch et al., 1996). An antibacterial action of sPLA₂ against Gram-negative bacteria required the presence of other antibacterial agents such as bactericidal/permeability-increasing protein (Wright et al., 1990; Weiss et al., 1994), whereas against Gram-positive bacteria, this requirement was not necessary (Weinrauch et al., 1996). Thus, sPLA₂ may serve as one of the body's natural host-defense mechanisms. There also is evidence to implicate sPLA₂ in various inflammatory disease states associated with the proinflammatory mediators produced as a consequence of the catalytic activity of this protein. Elevated levels of sPLA₂ have been reported in various body fluids from humans with several inflammatory conditions, including systemic inflammatory response syndrome encompassing sepsis and multiple organ trauma (Vadas and Pruzanski, 1993), acute pancreatitis (Nevalainen et al., 1993), arthritis (Pruzanski et al., 1991; Kortekangas et al., 1994), and inflammatory bowel disease (Minami et al., 1993, 1994). Levels of sPLA₂ have been shown to correlate with the severity and prognosis of some of these disorders (Vadas and Pruzanski, 1986, 1993; Nevalainen et al., 1993). The proposed pathophysiological effects of sPLA₂ would result from an excessive and uncontrolled host response, leading to the overproduction of sPLA₂ and subsequent release of proinflammatory mediators.

The role of sPLA₂ in various inflammatory conditions will be determined only when potent, specific inhibitors of sPLA₂ are developed and evaluated in the clinic. As a step toward achieving this goal, we herein describe some pharmacological studies with LY315920 (Fig. 1), a potent and selective inhibitor of the catalytic activity of human, group IIA, nonpancreatic sPLA₂ enzyme (Draheim et al., 1996).

View larger version (12K): [in this window] [in a new window]   Fig. 1.   Structural formula of LY315920, [[3-(aminooxoacetyl)-2-ethyl-1-(phenylmethyl)-1H-indol-4-yl]oxy]acetate.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Chromogenic Assay

The effects of LY315920 on the catalytic activities of human, group IIA, nonpancreatic sPLA₂ and human, group IB, pancreatic sPLA₂ were measured using a modified chromogenic mixed micelle assay (Reynolds et al., 1992). Briefly, LY315920 was evaluated in the assay, which contained final concentrations of 0.96 mM racemic 1,2-bis(thioheptanoyl)-1,2-dideoxyphosphatidylcholine, 0.27 mM Triton X-100, and 0.12 mM 5,5'-dithiobis[2-nitrobenzoic acid] (DTNB). Concentration-response curves were generated with 16 nM recombinant human sPLA₂ for 30 min at 40°C in a microtiter plate format. IC₅₀ values were determined as well as mole fraction for 50% inhibition (Xi₅₀). Mole fraction represented a dimensionless number derived by dividing the IC₅₀ value by the concentration of lipid in the assay mixture (Jain et al., 1989). IC₅₀ values were determined in triplicate.

Deoxycholate/Phosphatidylcholine Assay

The hydrolytic activities of animal sera/plasma samples were measured using a modified ¹⁴C-labeled deoxycholate/phosphatidylcholine (DOC/PC) mixed micelle assay (Schadlich et al., 1987; Draheim et al., 1996). This assay uses 3 mM sodium deoxycholate, 1 mM 1-palmitoyl-2-¹⁴C-oleoyl-sn-glycero-3-phosphocholine, and 3 nM recombinant group IIA human sPLA₂ or recombinant group IB, human pancreatic sPLA₂. When evaluating sPLA₂ activity in serum, 5 to 20 µl of a test serum sample was assayed in 140 µl of total volume. In conditions where LY315920 was tested in vitro, 10 or 20 µl of active serum plus 10 µl of various test concentrations in dimethyl sulfoxide (DMSO) solution were used.

Cytosolic PLA₂ Activity Assay

Recombinant human cytosolic PLA₂ (cPLA₂) was produced in a baculovirus expression system and purified using sequential ion exchange and hydrophobic interaction chromatographies (Becker et al., 1994). cPLA₂ activity was assessed using a slight modification of the chromogenic assay described by Reynolds et al. (1994). The final assay contained 2 mM 1-hexadecyl-2-arachidonoylthio-2-deoxy-sn-glycero-3-phosphocholine, 1 mM Triton X-100, and 30% glycerol in 80 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid, pH 7.4, 150 mM NaCl, 10 mM CaCl₂, and 1 mg/ml BSA. The selected concentrations of inhibitor or vehicle (DMSO) was added in 5-µl volume to the substrate that had been aliquoted into a 96-well format. The reaction was initiated by adding 750 ng of cPLA₂, in a 20-µl volume of 1× assay buffer, to the wells; shaking the plate for 30 s on high to mix; and then incubating for 60 min at 37°C. To quench the reaction, 20 µl of a 12.5 mM DTNB/475 nM EGTA mixture was added to the wells. The plate was shaken for 30 s on high and allowed to incubate for an additional 3 min at room temperature to give the DTNB chromophore time to fully develop. Absorbance was measured at 405 nm.

Cyclooxygenase-1 and Cyclooxygenase-2 Activity Assays

The effects of LY315920 and selected inhibitors were evaluated on cyclooxygenase-1 (COX-1) and cyclooxygenase-2 (COX-2) activity using a slight modification of the human whole blood assay described by Brideau et al. (1996). Briefly, COX-1 activity was evaluated by monitoring the release of thromboxane A₂ (TXA₂) from human platelet-rich plasma (PRP) stimulated with human thrombin (3 U/ml) over a 10-min period. The drug or vehicle was added to the PRP 5 min before the thrombin challenge. The release of TXA₂, measured as its stable metabolite, thromboxane B₂ (TXB₂), was quantified with a competitive enzyme immunoassay (EIA) kit according to manufacturer's instructions. COX-2 activity was assessed by incubating human whole blood with 1 µg/ml lipopolysaccharide (LPS) for 20 h and measuring the release of prostaglandin E₂ (PGE₂) using an EIA kit according to manufacturer's instructions. Selected concentrations of drugs or vehicle were added to the blood before the start of the 20-h incubation.

Release of TXA₂ from Guinea Pig Bronchoalveolar Lavage Cells

In Vitro Experiments. The methodology for these experiments has been described in detail previously (Fleisch et al., 1996). Briefly, male Hartley guinea pigs (Charles River Laboratories, Portage, MI) weighing 300 to 500 g were anesthetized with 40 mg/kg sodium pentobarbital. The abdominal aorta was severed through a ventral midline abdominal incision, and the trachea was cannulated. Bronchoalveolar lavage (BAL) was performed with a 5-ml glass syringe connected to a cannula. The syringe contained 4 ml of oxygenated Krebs' bicarbonate solution. The procedure was done three times in a 1.5-min period with the same solution to obtain a high yield of cells. Three to 3.5 ml of BAL fluid was recovered and placed in a polypropylene test tube. Aliquots of BAL fluid (0.5 ml) were transferred to 12 × 75-mm polypropylene test tubes and preincubated at 37°C for 5 min in a shaking water bath. LY315920 or vehicle control, active recombinant human sPLA₂, and 1.2 mM final concentration of Ca⁺⁺ were added to the cell suspension. The mixture was gently vortexed for 15 s, and the incubation continued. In some experiments, N-formyl-methionyl-leucyl-phenylalanine (fMLP) or arachidonic acid was substituted for human sPLA₂. After 60 min, the reaction was stopped, and the supernatant fluids were decanted into polypropylene microtubes and assayed for TXA₂ in terms of its stable metabolite TXB₂ using an EIA as described by Fleisch et al. (1996).

Ex Vivo Experiments. Male Hartley guinea pigs weighing 300 to 500 g were anesthetized with 40 mg/kg sodium pentobarbital. The lateral saphenous vein was cannulated, through which vehicle or increasing amounts of LY315920 were administered. The catheter was then removed, and the vein was clamped shut. A ventral midline abdominal incision was made to expose and sever the abdominal aorta 5 min after vehicle or drug administration. The remainder of the procedure followed that enumerated above for the in vitro test system.

Guinea Pig Lung Pleural Strips

The methodology for these experiments has been described in detail previously (Snyder et al., 1993). Briefly, male Hartley strain guinea pigs (500-650 g) were sacrificed by cervical dislocation followed by decapitation. Dorsal pleural strips (4 × 1 × 25 mm) were dissected from intact parenchymal segments (8 × 4 × 25 mm) cut parallel to the outer edge of the lower lung lobes. Two adjacent pleural strips, obtained from a single lobe and representing a single tissue sample, were tied independently to a metal support rod and attached by cotton thread to a Grass (Quincy, MA) force displacement transducer (FT 03C). Changes in isometric tension were displayed on a Modular Instruments (Malvern, PA) monitor and thermal recorder. All tissues were suspended in 10-ml, jacketed tissue baths that contained a modified Krebs' buffer solution of the following composition: 118.2 mM NaCl, 4.6 mM KCl, 2.5 mM CaCl₂·2H₂O, 1.2 mM MgSO₄·7H₂O, 24.8 mM NaHCO₃, 1.0 mM KH₂PO₄, and 10.0 mM dextrose. The tissue bath buffer was aerated (95% O₂/5% CO₂) and maintained at 37°C. Pleural strips from opposite lobes of the same lung were used for paired experiments. Initially, the tissues were challenged 3 times with KCl (40 mM) to test for tissue viability and to obtain a consistent response.

Cumulative concentration-response curves were obtained from pleural strips by increasing the agonist concentration in the tissue bath by half-log₁₀ increments while the previous concentration remained in contact with the tissues (van Rossum, 1963). Agonist concentration was increased after reaching the plateau of the contraction elicited by the preceding concentration. One concentration-response curve was obtained from each tissue. To minimize variability between tissues obtained from different animals, contractile responses were expressed as a percentage of the maximal response obtained with the final KCl challenge for each tissue. When studying the effects of LY315920 on the agonist-induced contractile effects, the drug and its vehicle (DMSO) were added to the tissues 30 min before starting the sPLA₂ or arachidonic acid concentration-response curves.

The data analysis has been previously described in detail (Dillard et al., 1996). Briefly, data from different experiments were pooled and presented as a percentage of the maximal KCl responses (mean ± S.E.M.). To estimate the drug-induced rightward shifts in the concentration-response curves, the curves were analyzed simultaneously using statistical nonlinear modeling methods similar to those described by Waud (1976). The model includes four parameters: the maximum tissue response that is assumed to be the same for each curve, the ED₅₀ for the control curve, the steepness of the curves, and the pA₂ or apparent K_B (the concentration of antagonist that requires a 2-fold increase in agonist to achieve an equivalent response). The Schild slope was determined to be 1, which is consistent with the assumptions of a competitive antagonist. The potency of LY315920 was defined in this assay in terms of apparent K_B, which could be interpreted as the dissociation constant of the inhibitor.

Agonist-Induced Mediator Release and Contraction of Lung Pleural Strips

In these experiments, guinea pig lung pleural strips were challenged with a bolus concentration of human sPLA₂ or arachidonic acid in the absence and presence of drug. LY315920 at the indicated concentrations or vehicle was freshly prepared and mixed in Krebs' solution before experimentation. Spontaneous and agonist-induced release of mediators were measured in the presence of drug or vehicle over a 30-min time period. For the spontaneous release, the bath fluid was changed, drug or vehicle was added to the baths, and samples of bath fluid (300 µl/bath) were collected and immediately replaced with fresh buffer. This sample served as the zero time collection point. A second sample, with buffer replacement, was collected from each bath 30 min later, from which the spontaneous release of TXA₂ over 30 min was determined and served as the baseline level for agonist-induced mediator release. The tissues were then challenged with agonist, and a third sample of bath fluid was taken 30 min later. The collected bath fluid was stored in microfuge tubes at 20°C until subsequent measurements of the stable metabolite TXB₂ were made with an EIA. Contractile responses of tissues resulting from the application of the agonists were recorded and expressed as milligrams of tension and as a percentage of their maximal responses to the final KCl (40 mM) challenge (see above). Tissues were finally removed from the baths, detached from support rods, lightly blotted onto absorbent paper, and weighed. The concentration of TXB₂ was expressed as picograms per gram of tissue.

Effects of LY315920 in Transgenic Mice Expressing Human sPLA₂ Protein

The development and phenotype of these transgenic animals have been previously described (Fox et al., 1996). These mice were used to evaluate the activity of LY315920 in vivo. Retro-orbital blood samples were obtained with the animals under metaphane anesthesia before and at selected times after drug administration. LY315920 (0.3-3 mg/kg) or vehicle (5% DMSO, 5% ethanol, and 30% polyethylene glycol 300 in water) was administered i.v. in a volume of 0.15 ml or by gavage at 1 ml/100 g (v/w of mouse). Serum sPLA₂ activity (using the DOC/PC assay described above) was determined for each blood sample. The effect of LY315920 was measured as a percent change in serum sPLA₂ activity from the time zero (predrug bleed) for each mouse. Slight activity was detected in serum of nontransgenic littermates, and this activity was averaged (n = 3-5 determinations/assay) and subtracted from the activity recorded from the transgenic animals, which rendered percent change values of >100% for maximal inhibition.

Drugs and Chemicals

Indomethacin, sodium azide, EDTA, EGTA, deoxycholate, Triton X-100, HEPES, DMSO, DTNB, BSA (fatty acid free), Tween 20, and fMLP were obtained from Sigma Chemical Co. (St. Louis, MO). Arachidonic acid was purchased from Nu Chek Prep, Inc. (Elysian, MN). LPS (Escherichia coli O55:B5) was purchased from Difco (Detroit, MI). Human -thrombin was obtained from Enzyme Research Lab (West Bank, IN). TXB₂ and PGE₂ EIA kits and standards were purchased from Cayman Chemical Co. (Ann Arbor, MI). NS-398 was purchased from BIOMOL (Plymouth Meeting, PA). 1-Palmitoyl-2-¹⁴C-oleoyl-sn-glycero-3-phosphocholine was obtained from Dupont NEN (Wilmington, DE). Racemic 1,2-bis(thioheptanoyl)-1,2-dideoxyphosphatidylcholine, recombinant human group IB, pancreatic sPLA₂, and 1-hexadecyl-2-arachidonoylthio-2-deoxy-sn-glycero-3-phosphocholine were obtained from Shionogi & Co. LY315920 was synthesized by Eli Lilly and Company (Indianapolis, IN). Recombinant human nonpancreatic sPLA₂ was cloned, expressed, and purified from Syrian hamster AV12 cells (Eli Lilly and Company).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Effect of LY315920 on Recombinant Human sPLA₂. The intrinsic activity of LY315920 as an inhibitor of human, group IIA sPLA₂ was initially defined using the chromogenic mixed micelle assay (Reynolds et al., 1992). LY315920 was able to inhibit human sPLA₂ activity with an IC₅₀ value of 9 ± 1 nM (mean ± S.E.M., n = 4) or a mole fraction, which is defined as the IC₅₀/total concentration of lipid in the assay mixture, of 7.3 × 10⁶. A radiometric DOC/PC assay (Schadlich et al., 1987) was used to gain a more accurate measure of the potency of LY315920 against human sPLA₂. Under these conditions, LY315920 produced an IC₅₀ of 5.9 ± 0.4 nM (mean ± S.E.M., n = 9) or a mole fraction of 1.5 × 10⁶. In contrast, when tested against human recombinant, group IB, pancreatic sPLA₂ in the chromogenic assay, LY315920 produced an IC₅₀ of 367 ± 86 nM (mean ± S.E.M., n = 3) or a mole fraction of 2.9 × 10⁴. When human pancreatic sPLA₂ was used as the enzyme in the DOC/PC assay, LY315920 inhibited its activity with an IC₅₀ of 217 ± 3 nM (mean ± S.E.M., n = 3) or a mole fraction of 5.4 × 10⁵ demonstrating a 40-fold selectivity as an inhibitor of human group IIA, SPLA₂. LY315920 was also evaluated as an inhibitor of human cPLA₂. At concentrations up to 200 µM (n = 3), LY315920 failed to inhibit the catalytic activity of cPLA₂.

Lack of Effect of LY315920 on COX-1 and COX-2. Thrombin-stimulated release of TXA₂ from human PRP was used to evaluate the actions of LY315920 (Fig. 2A). LY315920 had no effect on the TXB₂ levels except at extremely high concentrations (100 µM), where TXB₂ levels were slightly reduced. In contrast, indomethacin and NS-398, a selective COX-2 inhibitor (Masferrer et al., 1994), reduced TXB₂ levels in a concentration-dependent manner with IC₅₀ values of 0.92 ± 0.35 and 21 ± 10 µM, respectively. Likewise, LY315920 failed to alter LPS-induced release of PGE₂ from human whole blood (Fig. 2B). Indomethacin and NS-398 reduced PGE₂ production in a concentration-related manner. The potency of NS-398 in this assay (IC₅₀ = 0.19 ± 0.15 µM) was nearly 10-fold greater than that of indomethacin (IC₅₀ = 1.11 ± 0.82 µM).

View larger version (24K): [in this window] [in a new window]   Fig. 2.   Effect of LY315920 on blood assay that assesses cyclooxygenase activities. A, thrombin-stimulated TXA2 production from PRP was used as an indicator of COX-1 activity. After thrombin stimulation of PRP, TXB2 levels of 272, 538, and 539 pg/ml were measured in control samples. B, LPS-stimulated production of PGE2 from human whole blood reflected COX-2 activity. Control levels of PGE2 released from three separate blood samples stimulated with LPS (1 µg/ml) were 382, 300, and 324 pg/ml. Symbols represent the mean ± S.E.M. of three experiments.

Effect of LY315920 on sPLA₂ Activity in Serum of Various Species. sPLA₂ can be released as a result of the clotting process. Thus, sPLA₂ activity could be measured from the serum of rat, rabbit, and guinea pig using the DOC/PC assay. Human serum was also assayed, but no measurable activity could be detected with this assay. Human recombinant sPLA₂ was added to human serum to assess the in vitro potency of LY315920. Table 1 summarizes the inhibitory potency (IC₅₀ values) of LY315920 on sPLA₂ activity in sera from selected species and sPLA₂-spiked human serum.

Effect of LY315920 on Human sPLA₂-Mediated Release of TXA₂ from Guinea Pig BAL Cells In Vitro. This technique was used as a functional assay to evaluate the potency of LY315920 in vitro and ex vivo. In BAL cells challenged with human sPLA₂, LY315920 (0.1-3 µM) reduced the formation of thromboxane in a concentration-related manner with an IC₅₀ of approximately 8 × 10⁷ M (Fig. 3). In control experiments (n = 7), LY315920 (3 µM) had no effect when fMLP (2 × 10⁷ M), a chemotactic peptide, was used to stimulate a comparable release of thromboxane. Similarly, generation of TXA₂ by arachidonic acid was not influenced by prior exposure to LY315920 but could be inhibited by indomethacin (Fig. 4).

View larger version (15K): [in this window] [in a new window]   Fig. 3.   Inhibition of human sPLA2-induced release of thromboxane from guinea pig BAL cells by LY315920 (0.1-3 µM) in vitro. Symbols represent the mean ± S.E.M. of the number of experiments (N).

View larger version (24K): [in this window] [in a new window]   Fig. 4.   Lack of inhibitory effect of LY315920 (3 µM) on arachidonic acid-induced thromboxane generation from guinea pig BAL cells. Indomethacin (10 µM) was used as a positive control. Symbols represent the mean ± S.E.M. of the number of experiments (N).

Effect of LY315920 on Generation of TXA₂ from Guinea Pig BAL Cells Ex Vivo. Consistent inhibition of sPLA₂ activity in BAL fluid was observed 5 min after the i.v. administration of LY315920. The human sPLA₂-induced generation of TXA₂ on BAL cells from guinea pigs pretreated with LY315920 (3, 10, and 30 mg/kg i.v.) was reduced compared with BAL cells from vehicle-treated controls. The calculated ED₅₀ for LY315920 was 16.1 mg/kg (Fig. 5).

View larger version (15K): [in this window] [in a new window]   Fig. 5.   LY315920 administered i.v. inhibited human sPLA2-induced release of thromboxane from guinea pig BAL cells. Symbols represent the mean ± S.E.M. of the number of experiments (N).

Effect of LY315920 on Human sPLA₂-Induced Contractions of Guinea Pig Lung Pleural Strips. The effects of human sPLA₂ on guinea pig lung pleural strips have been well characterized (Snyder et al., 1993). Challenging the lung pleural strips with human sPLA₂ (0.1-10 µg/ml) produced concentration-dependent contractile responses (Fig. 6). The ED₅₀ value for the control curves was 1.22 ± 0.12 µg/ml or 87.6 ± 5.06 nM (mean ± S.E.M., n = 32). In the presence of LY315920 (0.03-3 µM), these contractile responses were suppressed in a concentration-dependent manner (Fig. 6). An apparent K_B, the concentration of antagonist that requires a 2-fold increase in agonist to achieve an equivalent response, was calculated to be 83 ± 14 nM, which approached the stiochiometric limit of this assay. In contrast, when arachidonic acid was used as the agonist, the ensuing concentration-response curves were not altered by LY315920 (Fig. 7). The concentration of LY315920 (10 µM) used in this experiment was 3-fold greater than that which nearly abolished the human sPLA₂ contractile responses.

View larger version (29K): [in this window] [in a new window]   Fig. 6.   Effect of LY315920 (0.03-3 µM) on contractile responses induced by human sPLA2 on guinea pig lung pleural strips. Symbols represent mean ± S.E.M. of the number of tissues in parentheses and are expressed as percentage of maximal KCl (40 mM) responses. The final KCl response averaged 444 ± 20 mg (mean ± S.E.M.) for 32 control tissues.

View larger version (20K): [in this window] [in a new window]   Fig. 7.   Effect of LY315920 (10 µM) on contractile responses induced by arachidonic acid on guinea pig lung pleural strips. Symbols represent mean ± S.E.M. of the number of tissues in parentheses and are expressed as percentage of maximal KCl (40 mM) responses. The final KCl response averaged 451 ± 26 mg (mean ± S.E.M.) for eight control tissues.

Effect of LY315920 on TXA₂ Release from Guinea Pig Lung Pleural Strips. The lung pleural strip assay was used to monitor the release of thromboxane in the tissue bath fluid after agonist challenge. The spontaneous release of TXA₂ from resting guinea pig lung pleural strips over a 30-min period was measured as 336 ± 63 pg/g tissue (mean ± S.E.M., n = 4) from the bath fluid. Challenging the tissues with human sPLA₂ (10 µg/ml bolus) produced a contractile response associated with marked elevations in TXA₂ levels in the bath fluid of the vehicle-treated tissues (Table 2). The amount of TXA₂ released into the bath fluid by human sPLA₂ is nearly 20-fold greater than that seen in resting tissues during a similar time period. LY315920 (3 µM) had no effect on the spontaneous release of TXA₂ (229 ± 16 pg/g tissue, mean ± S.E.M., n = 4) but virtually abolished the contractile responses and TXA₂ release by human sPLA₂ (Table 2). Challenge of the tissues with arachidonic acid produced contractile responses and TXA₂ that could be measured in the bath fluid. A concentration of arachidonic acid was chosen (0.6 µg/ml) that released amounts of TXA₂ similar to those seen with human sPLA₂. When the tissues were challenged with this concentration of arachidonic acid, LY315920 (3 µM) failed to alter either the contractile responses or the level of TXA₂ released into the bath (Table 2). The degree of inhibition of the contractile responses and TXA₂ release was >90% versus human sPLA₂ and <10% versus arachidonic acid.

Effects of LY315920 in Transgenic Mice Expressing Human sPLA₂. The phenotype of the transgenic mouse expressing human sPLA₂ has been previously characterized (Fox et al., 1996). LY315920 (0.3-3 mg/kg), administered i.v., abolished serum sPLA₂ activity at all three doses tested (Fig. 8A). As time progressed after dosing, the inhibitory effect gradually diminished over the 4-h observation period. This was especially evident with the lowest dose. No inhibitory effect was observed in vehicle-treated transgenic mice. Oral administration of LY315920 (0.3-3 mg/kg) to transgenic mice resulted in similar inhibition of serum sPLA₂ activity (Fig. 8B). Within 30 min of dosing, serum sPLA₂ activity was abolished at the highest doses, whereas only 75% inhibition was observed at the 0.3 mg/kg dose. This inhibitory effect of LY315920 was well maintained by all three doses over the 4-h observation period.

View larger version (24K): [in this window] [in a new window]   Fig. 8.   Effect of i.v. (A) and oral (B) administration of LY315920 (0.3-3 mg/kg) or vehicle on sPLA2 enzyme activity in serum of transgenic animals expressing the human sPLA2 protein. Symbols represent the mean ± S.E.M. of four mice per group. Drug was administered at time 0. The dpm released at time 0 ranged from 1133 to 4357 and from 1094 to 4695 in the i.v. and oral studies, respectively.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

LY315920 was shown to be a potent, selective inhibitor of the group IIA, sPLA₂. In the isolated chromogenic enzyme assay, inhibitory activity (IC₅₀ = 9 ± 1 nM) approached the stoichiometric limit of the assay because 16 nM was the total concentration of human sPLA₂ enzyme in the assay. We turned to the DOC/PC assay for a more accurate measure of its potency. Under these conditions, the IC₅₀ was reduced to 5.9 ± 0.4 nM, which is equivalent to a mole fraction of 1.5 × 10⁶. LY315920 was able to inhibit sPLA₂ activity in sera from various species (3-8 nM), demonstrating its potential use in different animal models of inflammation. LY315920 acts at the active site of the human enzyme (R. W. Schevitz, unpublished observation) in a manner similar to findings with structural analogs that were previously cocrystallized in the active site (Schevitz et al., 1995). Selectivity of LY315920 was apparent by the 40-fold weaker activity against group IB, human pancreatic sPLA₂. Furthermore, LY315920 failed to inhibit the catalytic activity of cPLA₂. Lack of activity against cPLA₂ is not surprising based on marked differences between cPLA₂ and sPLA₂ that include but are not limited to size, distribution, lack of primary structure homology, mechanistic differences like the role of Ca⁺⁺ in enzyme function, and fatty acid requirements in the sn-2 position of glycerophospholipid.

Selectivity also was observed in cellular assays that demonstrated thromboxane synthase and the constitutive and inducible forms of cyclooxygenase were not inhibited by LY315920. Thrombin-stimulated human PRP was used to evaluate the effect of LY315920 on COX-1. LY315920 was inactive except at the highest concentration (100 µM), where a slight reduction in TXB₂ levels was observed. Thrombin-induced release of TXA₂ from platelets has been well characterized and is mediated via the activation of cPLA₂ (Kramer et al., 1993). This would be further evidence that LY315920 does not inhibit cPLA₂ or thromboxane synthase activity. Similarly, LPS-induced COX-2 activity in human whole blood (Brideau et al., 1996) was not altered by LY315920. Indomethacin and NS-398 were used as positive controls in these experiments. Indomethacin was a more potent COX-1 inhibitor and NS-398 was more potent inhibiting COX-2 activity that confirms previous reports (Prasit et al., 1995; Brideau et al., 1996).

We further characterized LY315920 in vitro using two functional assays that were developed specifically to evaluate inhibitors of human sPLA₂. Guinea pig lung pleural strips and BAL cells isolated from guinea pig lung have been well characterized (Snyder et al., 1993; Fleisch et al., 1996). Both of these assays were considered functional assays because human sPLA₂ hydrolyzed membrane phospholipids in tissues and cells rather than artificial, unnatural substrates, which were used in the isolated enzyme assays. Contractile responses, in the lung pleural strip assay, were mediated by TXA₂ that was formed after the catalytic release of arachidonic acid. LY315920 suppressed the contractile responses in a concentration-related manner with an apparent K_B of 83 ± 14 nM, which approaches the stiochiometric limit of this assay, similar to the chromogenic assay. When the PLA₂ step was circumvented by challenging tissues with arachidonic acid, the contractile responses were not significantly altered by LY315920. These observations indicated that LY315920 was not acting as a cyclooxygenase inhibitor or a thromboxane receptor antagonist and that it did not possess any nonspecific tissue-depressant effects.

In this assay, TXB₂ levels were measured in the bath fluid after the bolus challenge with human sPLA₂ or arachidonic acid. LY315920 abolished the release of TXA₂ and the contractile responses mediated by human sPLA₂ without altering those induced by arachidonic acid. The degree of inhibition of the contractile responses and release of TXA₂ was similar, demonstrating that the contractile responses were mediated by the formation of TXA₂. Interestingly, LY315920 did not inhibit the spontaneous release of TXA₂, suggesting this process was mediated by an endogenous PLA₂ other than sPLA₂. Cytosolic PLA₂ may account for the spontaneous release of TXA₂ because LY315920 has no inhibitory effect on this enzyme.

BAL cells isolated from guinea pig lung served as the substrate for human sPLA₂, which resulted in the formation of TXA₂. LY315920 was able to inhibit the human sPLA₂-mediated release of TXA₂ with an IC₅₀ of 790 nM. The 10-fold difference in potency relative to the lung pleural strip assay is difficult to explain but may be due to nonspecific binding of the inhibitor to BAL proteins. Challenge of BAL cells with either arachidonic acid or fMLP resulted in the generation of TXA₂, which was not altered by LY315920. Therefore, both functional assays demonstrated the specificity of LY315920 as an inhibitor of sPLA₂.

In vivo pharmacological activity of LY315920 was assessed using two different animal systems. First, transgenic mice have been generated that express human sPLA₂ protein in their serum (Fox et al., 1996). Maintenance of these mice with the use of ZnSO₄ in their drinking water induced the metallothionein gene promotor, which increased the expression of the sPLA₂ protein to high levels (300 ng/ml). These are in the range observed in patients undergoing a systemic inflammatory response to sepsis (Nevalainen et al., 1992; Rintala and Nevalainen, 1993). They do not, however, cause pathology in the mice, presumably due to the absence of coexisting pathology, which may predispose them to the actions of sPLA₂ (see Fox et al., 1996). Thus, these transgenics were used to study the pharmacodynamic effects of LY315920. After the i.v. or oral administration of LY315920, serum sPLA₂ activity was inhibited in a dose-dependent manner over the 4-h time course. This suggests that LY315920 might have use for acute or chronic indications. Activities of sPLA₂ measured in these mice were not confounded by an endogenous sPLA₂ because this strain of mouse (C57BL/6) is genetically sPLA₂ deficient (Kennedy et al., 1995; MacPhee et al., 1995). In support, we found serum PLA₂ activity in nontransgenic littermates to be near the detection limit of the assay.

Second, in vivo characteristics of LY315920 were assessed in a series of ex vivo experiments using guinea pig BAL cells. LY315920 was administered i.v., and 5 min later the BAL procedure was performed. Under these conditions, BAL cells demonstrated a dose-related reduction in the ability to generate TXA₂ when challenged with human sPLA₂. Therefore, LY315920 was able to distribute into the lung fluid in a pharmacologically active concentration (ED₅₀ = 16 mg/kg) after i.v. administration. Recently, in a guinea pig model of LPS-induced acute lung injury, sPLA₂ was shown to hydrolyze pulmonary surfactant (Arbibe et al., 1998), a process that is blocked by LY311727 (Mihelich et al., 1997), an analog of LY315920. These observations indicate for the first time that mammalian, group IIA, sPLA₂ is pathologic in an animal model of human disease. The ability of LY315920 to distribute to the lung could have potential therapeutic value because the lung is a major target organ in human sepsis and sPLA₂ levels are markedly elevated (Vadas and Pruzanski, 1993).

Several promising inhibitors of sPLA₂ have been identified that block the release of arachidonic acid and have the potential to limit the formation of a plethora of proinflammatory lipid mediators (Tramposch et al., 1992; Glaser et al., 1993; Beaton et al., 1994; Marshall et al., 1995; Tibes and Friebe, 1997). The potencies of these inhibitors have been expressed in molar terms and ranged from 4 µM for WAY 121520 (Glaser et al., 1993) to 21 nM for FPL 67047 (Beaton et al., 1994). However, because of the complex nature of the catalytic process, which requires an aggregated substrate, and the fact that assay conditions vary between laboratories, true potency comparisons can be made only by evaluating these inhibitors in terms of mole fraction (Xi₅₀), a dimensionless term (Jain et al., 1989). The typical potency, in Xi₅₀ terms, of these inhibitors is approximately 0.001, which is 1000-fold weaker than that of LY315920. Besides the weaker activity, other problems have been associated with these inhibitors; for example, BMS 181162 (Xi₅₀ = 0.013, Tramposch et al., 1992) showed limited bioavailability, which necessitated its use as a topical agent. We calculated an Xi₅₀ of 0.01 for SB 203347 (Marshall et al., 1995). This compound showed in vivo efficacy in an LPS-induced mouse shock model. However, the strain of mouse used in that study was genetically deficient in group IIA sPLA₂ (Kennedy et al., 1995; MacPhee et al., 1995), which indicates that the beneficial effects of SB 203347 on these mice were not the result of its ability to inhibit group IIA sPLA₂. Thus, LY315920 appears to be significantly more potent as an inhibitor of sPLA₂ than any compound described to date.

In summary, we have demonstrated that LY315920 is a selective, stoichiometric inhibitor of sPLA₂. With LY315920, the stage is set to address the pathophysiology of sPLA₂ in humans. Given the potency, selectivity, and potential therapeutic use of this novel molecule, we propose that LY315920 represents the first in a new class of anti-inflammatory agents carrying the designation of SPI, sPLA₂ inhibitor.